Aistė Aleksandravičienė

Aurelijus Morkevičius, PH.D.

MagicGrid® BOOK OF KNOWLEDGE

A Practical Guide to Systems Modeling using
MagicGrid from Dassault Systèmes
2nd edition
Aistė Aleksandravičienė
Aurelijus Morkevičius, PH.D.

MagicGrid® BOOK OF KNOWLEDGE

A Practical Guide to Systems Modeling using
MagicGrid from Dassault Systèmes
2nd edition

Kaunas, 2021
ISBN 978-609-454-554-2

© 2018-2021, No Magic, Inc. All Rights reserved.

While No Magic, Inc. has taken precautions to ensure accuracy and completeness, the information contained in
this book is provided for information purposes only and No Magic, Inc. does not make any representations as to
its accuracy or completeness or any commitment as to future availability of functionality or products. No Magic,
Inc. disclaims any liability, loss, or risk incurred as a result of the use of any information or advice contained in
this book, either directly or indirectly. Individual results may vary. No part of this publication may be reproduced
in any form or by any means without the prior written permission of No Magic, Inc. All other trademarks are the
property of their respective owners.
TABLE OF CONTENTS

FOREWORD BY OLIVIER SAPPIN 7

FOREWORD BY AURELIJUS MORKEVIČIUS 8
CONTRIBUTORS 9
REVIEWERS 10
PREFACE 11
PROBLEM DOMAIN 15
Black-box perspective 15
Stakeholder Needs 16
Step 1. Organizing the model for stakeholder needs 17
Step 2. Creating a table for stakeholder needs 18
Step 3. Capturing stakeholder needs 19
Capturing information directly in the model 20
Copying information from a Microsoft Excel spreadsheet 20
Synchronizing information from a Microsoft Excel spreadsheet 22
Importing information from a ReqIF file 24
Synchronizing information from IBM Rational DOORS 26
Step 4. Grouping stakeholder needs 29
Step 5. Numbering stakeholder needs 31
Step 6. Categorizing stakeholder needs 33
Stakeholder Needs done. What's next? 34
System Context 35
Step 1. Organizing the model for the system context 36
Step 2. Capturing system contexts 37
Step 3. Creating an ibd for a system context 38
Step 4. Capturing participants of the system context 38
Step 5. Specifying interactions between the participants of system context 41
Step 6. Adding item flows to the system context 42
System Context done. What's next? 44
Use Cases 45
Step 1. Organizing the model for use cases 47
Step 2. Creating a diagram for use cases 47
Step 3. Capturing use cases 50
Step 4. Creating a diagram for specifying the use case scenario 50
Step 5. Creating swimlanes and setting allocation mode 51
Step 6. Specifying steps of the use case scenario 53
Step 7. Adding item flows to the use case scenario 58
Step 8. Supplementing the use case scenario with a parallel flow 61
Use Cases done. What's next? 62
Measures of Effectiveness 63
Step 1. Organizing the model for MoEs 64
Step 2. Creating a block for capturing MoEs 64
Step 3. Capturing MoEs for the SoI 65
MoEs done. What's next? 69

MagicGrid® Book of Knowledge | 3

 White-box perspective 70
Functional Analysis 71
Step 1. Organizing the model for functional analysis 73
Step 2. Creating an activity diagram to decompose a function 73
Step 3. Specifying the white-box scenario 76
Step 4. Creating an activity decomposition map 78
Functional Analysis done. What's next? 79
Conceptual Subsystems 80
Step 1. Organizing the model for conceptual subsystems communication 83
Step 2. Creating a bdd for specifying conceptual interfaces 83
Step 3. Capturing conceptual interfaces 85
Step 4. Creating an ibd for capturing conceptual subsystems 87
Step 5. Capturing conceptual subsystems 89
Step 6. Specifying interactions between conceptual subsystems and the outside of the SoI 90
Step 7. Specifying interactions among conceptual subsystems 94
Step 8. Allocating functions by conceptual subsystems and synchronizing item flows 99
Conceptual Subsystems done. What's next? 105
MoEs for Subsystems 106
Step 1. Organizing the model for MoEs for subsystems 106
Step 2. Capturing MoEs for subsystems 107
MoEs for Subsystems done. What's next? 110
Traceability to Stakeholder Needs 111
Step 1. Organizing the model for traceability 113
Step 2. Creating a matrix for capturing refine relationships 113
Step 3. Capturing refine relationships 115
Problem Domain done. What's next? 117

SOLUTION DOMAIN 118

Building the logical architecture of the SoI 121
System Requirements 122
Step 1. Creating and organizing a model for system requirements 124
Step 2. Creating a diagram for system requirements 126
Step 3. Specifying system requirements 126
Step 4. Establishing traceability to stakeholder needs 128
Step 5. Establishing traceability to the rest of the problem domain model 131
System Requirements done. What's next? 133
System Structure 134
Step 1. Organizing the model for the LSA 136
Step 2. Creating a bdd for capturing the LSA 137
Step 3. Capturing logical subsystems 137
Step 4. Establishing traceability to the problem domain model 140
Step 5. Capturing logical interfaces 142
System Structure done. What's next? 149
System Behavior 150
System Parameters 152
Step 1. Organizing the model for system parameters 154
Step 2. Specifying the MoE for capturing the total energy consumption 155
Step 3. Defining a method for calculating the total energy consumption 158
Step 4. Specifying system parameters for calculating the total energy consumption 160
Step 5. Binding constraint parameters to corresponding value properties 162
Step 6. Performing early system requirements verification 166
Step 7. Storing values in the model 172
System Parameters done. What's next? 176

4 | MagicGrid® Book of Knowledge

 Traceability to System Requirements 177
Step 1. Creating a matrix for capturing satisfy relationships 178
Step 2. Capturing satisfy relationships 179
LSA done. What's next? 180
Building the logical architecture of subsystems 181
Subsystem Requirements 182
Step 1. Creating and organizing a model for subsystem requirements 184
Step 2. Creating a diagram for subsystem requirements 186
Step 3. Specifying subsystem requirements 186
Step 4. Specifying traceability relationships 188
Subsystem Requirements done. What's next? 193
Subsystem Structure 194
Step 1. Organizing the model for subsystem structure 195
Step 2. Getting ready to capture the subsystem structure 196
Step 3. Capturing components of the Cooling System 200
Step 4. Creating an ibd for specifying interactions 202
Step 5. Specifying interactions with the outside 204
Step 6. Specifying interactions within the Cooling System 209
Subsystem Structure done. What's next? 214
Subsystem Behavior 215
Step 1. Organizing the model for subsystem behavior 217
Step 2. Creating a diagram for capturing Cooling System states 217
Step 3. Capturing states of the Cooling System 218
Step 4. Specifying event occurrences on transitions 220
Step 5. Specifying internal behaviors of the state 225
Step 6. Synchronizing item flows from the structure model 240
Subsystem behavior done. What's next? 242
Subsystem Parameters 243
Step 1. Organizing the model for subsystem parameters 245
Step 2. Specifying the MoE for capturing the total energy consumption for cooling 246
Step 3. Defining a method to calculate the total energy consumption for cooling 248
Step 4. Specifying subsystem parameters for calculating the total energy consumption for cooling 249
Step 5. Binding constraint parameters to appropriate value properties 249
Step 6. Performing early subsystem requirements verification 251
Step 7. Recalculating subsystem parameters in different states 252
Subsystem parameters done. What's next? 255
Traceability to Subsystem Requirements 256
Step 1. Creating a dependency matrix for capturing satisfy relationships 257
Step 2. Capturing satisfy relationships to subsystem requirements 258
Step 3. Traceability to and revision of system requirements 260
LSSA done. What's next? 263
Building system configuration model 265
System Configuration Structure 266
Step 1. Creating and organizing a model for system configuration structure 268
Step 2. Getting ready to capture the structure of the system configuration 269
Step 3. Capturing the integrated structure of the system configuration 273
Step 4. Verifying interface compatibility 277
System Configuration Structure done. What's next? 282
System Configuration Behavior 283
Step 1. Starting simulation of the system configuration block 284
Step 2. Analyzing the integrated behavior of the system configuration 285
System Configuration Behavior done. What's next? 289

MagicGrid® Book of Knowledge | 5

 System Configuration Parameters 290
Step 1. Redefining the MoE 291
Step 2. Creating a parametric diagram 291
Step 3. Updating bindings 293
System Configuration Parameters done. What's next? 295
Traceability from System Configuration to System Requirements 296
Solution Domain done. What's next? 297

IMPLEMENTATION DOMAIN 298

Implementation Requirements 298

ANNEX A: SAFETY & RELIABILITY ANALYSIS 301

Stakeholder needs: safety and reliability requirements 302
Conceptual FMEA in Black Box view 302
Functional FMEA in Black Box view 304
Functional FMEA coverage analysis 306
Addressing safety and reliability requirements in Problem domain 307
FMEA at the White Box view 311
Addressing safety and reliability concerns in the LSA 312
Addressing safety and reliability concerns at the subsystem level 315
Cooling Subsystem level 315
Filtering Subsystem level 317
Logical subsystems FMEA at the subsystem level 328
Addressing safety and reliability concerns at the configuration level 329

ANNEX B: FROM SYSTEM OF SYSTEMS TO SYSTEM ARCHITECTURE 330

System of Systems Engineering 330
Transitioning from SoS to SA 331

AFTERWORD 335
GLOSSARY 336
BIBLIOGRAPHY 344
ABOUT THE AUTHORS 345

6 | MagicGrid® Book of Knowledge

FOREWORD BY OLIVIER SAPPIN
Model-based systems engineering (MBSE) is a relatively young practice for developing complex systems.
The promise of MBSE is to be able to progressively model and simulate systems behavior in accordance
with the needs & scenarios of usage, at any time. This promise is about avoiding the discipline-specific
tunnel effect and giving project stakeholders the ability to continuously monitor the evolution of system
development. One of the key success factors of MBSE is to apply best practices based on a sound systems
engineering methodology.

The MagicGrid Book of Knowledge (BoK) focuses on how practitioners implement MBSE. MagicGrid has
been developed to provide a framework that can be easily adopted, implemented, and modified to grow
with your modeling and systems engineering practice. This book teaches you how to use CATIA® Magic
software to design and optimize the systems and products of today and tomorrow. It is the culmination of
many years of research, and I am proud to put the Dassault Systèmes name behind this effort.

Thank you to all the magicians who made this happen, especially our industry process consultants team
who have engaged with customers over thousands of hours to learn, digest, develop, and teach this
framework. I would also like to thank our partners and customers who contributed to peer review of this
work prior to its release.

— Olivier Sappin, CATIA CEO

Dassault Systèmes

MagicGrid® Book of Knowledge | 7

FOREWORD BY AURELIJUS
MORKEVIČIUS
The MagicGrid BoK is a unique, model-based systems engineering experience. This book describes one of a
million possible straightforward ways to develop a system model from A to Z. It is an answer to all the
questions you may have, starting from “why” and ending with “how”, which few of today’s available
sources of information can answer.

When I started MBSE training and consulting 10 years ago, the first thing that came to my mind was that
there was a lack of an unambiguous approach to developing system models using SysML. Every customer
was developing models differently. Moreover, every consultant used different approaches to train
customers in this discipline. It should come as no surprise that SysML was (and still is) criticized for being
method-agnostic. There was no other way forward than to initiate internal discussions within the team,
and to start collecting and summarizing our experience in the form of a framework for architecting
systems using a sort of “vanilla” SysML. The framework comes first, then adjusting to the language
second. I am certain that this is the only approach that is 100 percent aligned with SysML. Extensions and
customizations are possible; however, they are not necessary.

While collecting our knowledge, I did some scientific studies with a local university to prove my ideas were
correct. Together with my team and colleagues, I wrote a couple of papers and delivered multiple
presentations and tutorials at various events worldwide, e.g., the INCOSE International Symposium,
INCOSE local chapter events, and the MBSE Cyber Experience Symposium, to name just a few. Publicity,
and a lot of positive feedback, gave me the trust and self-confidence to continue working on MagicGrid.

After working with multiple customers from various industries, I do agree that there is no single unified
approach to this issue, and there cannot be one. Every industry and every customer has their own
specifics, and we have taken this into account. MagicGrid has evolved as a foundation and collection of
best practices that can be modified or extended to support specific customer needs. Know-how captured in
this approach is highly influenced by the INCOSE handbook, ISO/IEC/IEEE 15288, the Unified Architecture
Framework (previously known as UPDM) and development efforts. Companies which influenced our
efforts included Bombardier Transportation, Kongsberg Defense and Airspace, Ford Motor Company, John
Deere, BAE Systems, and many others which I personally, along with fellow team members and
colleagues, have been working with in recent years.

I’m proud that besides the MagicGrid framework, we have managed to put together a detailed step-by-
step tutorial and publish it both as a book and an e-book. Please note that this book can only get better
with your help. Please do not hesitate to contact us and provide your feedback to make the second edition
of the book even better.

My hope is that this book will be a lighthouse for all MBSE practitioners sailing in the dark, and a light
breeze for those sailing in the sunlight.

— Aurelijus Morkevičius, PhD

Dassault Systèmes

8 | MagicGrid® Book of Knowledge

CONTRIBUTORS
Andrius Armonas, PhD, [email protected]

Barry Papke, [email protected]

Daniel Brookshier, [email protected]

Edita Milevičienė, [email protected]

Elona Paulikaitytė, [email protected]

Gintarė Kriščiūnienė, [email protected]

Osvaldas Jankauskas, [email protected]

Rokas Bartkevičius, [email protected]

Ronald Kratzke, [email protected]

Saulius Pavalkis, PhD, [email protected]

Solange Muhayimana, [email protected]

Thomas Marchand, [email protected]

Tomas Juknevičius, [email protected]

Žilvinas Strolia, [email protected]

MagicGrid® Book of Knowledge | 9

REVIEWERS
Andrew Terbrock
Antonio Tulelli
Arnaud Durantin
Atif Mahboob
Bernhard Meyer
Bryan K. Pflug
Carmelo Tommasi
Kris Gough
Christian Guist
Christian Konrad
Damian Delic
Darius Šilingas
David D. Walden
David Hughes
Jeff Tipps
Jon Symons
Gauthier Fanmuy
Himanshu Upadhyay
Holger Gamradt
James Towers
Jonathan Jansen
Macaulay Osaisai
Michael Schneider
Michael Elliot
Oliver Nagel
Olivier Casse
Paolo Neri
Pierfelice Ciancia
Piotr Malecki
Prem Sharma
Raymond Kempkens
Robert Marino
Sagar Nirgudkar
Shelley Higgins
Sven Degenkolbe
Thomas Eickhoff
Timo Wekerle
Toshihiro Obata
Ulf Koenemann
Zackary Bogner

10 | MagicGrid® Book of Knowledge

PREFACE
Why this book?

Any theory needs practice. An approach becomes useful when it is provided together with comprehensive
instructions on how to use it in practice. The MagicGrid framework for Model-Based Systems Engineering
(MBSE) by CATIA® No Magic is no exception to this rule. This was the main reason we have decided to
write this book.

So, if you’re new at MBSE or look forward to learn about the MagicGrid framework from A to Z, this book
is for you. This book is for you, too, if you wish to learn how to use CATIA Magic software for MBSE in
combination with OMG® Systems Modeling Language (SysML)™ and MagicGrid. This book is for you as
well, if you go for comprehensive instructions, illustrated by the case study of modeling an easy-to-
understand real-world system.

The book is easier to read if you are familiar with UML/SysML and basic concepts of CATIA Magic software
for MBSE; that is, Magic Cyber-Systems Engineer / Cameo Systems Modeler or Magic Systems of Systems
Architect / Cameo Enterprise Architecture along with the Cameo DataHub plugin installed on the top, or
Magic Software Architect / MagicDraw with the SysML and Cameo DataHub plugins, as of version 19.0,
generally referred to as the modeling tool.

Why MagicGrid?

MagicGrid is the bridge between the theory and practice, bringing three major MBSE components –
method, language, and modeling tool – together.

The idea of developing a new framework for MBSE emerged while working with organizations from
various industry sectors, such as defense, automotive, aerospace, and healthcare, to name a few. They
needed an unambiguous approach for developing system models using SysML, the critical enabler for
MBSE, as defined by the International Council on Systems Engineering (INCOSE). However, quite a
few methods for MBSE were available in the market at that time, and the existing ones were too abstract
for solving a real-world problem. The practice also revealed that every industry and every customer has
their own specifics, and cannot use the standard method without applying any modifications or
extensions to it.

The MagicGrid framework has evolved by summarizing the experience of numerous MBSE adoption
projects, as a foundation and collection of best practices that can be modified or extended to support
specific customer needs.

The framework is completely applicable in practice for the following reasons:

• It is fully compatible with “vanilla” SysML, which means, no extensions for the standard SysML are
required.
• It clearly defines the modeling process, which is based on the best practices of the systems
engineering process.
• It is tool-independent, as long as that tool supports SysML.

MagicGrid® Book of Knowledge | 11

MagicGrid 101

As the title reveals, the MagicGrid framework can be represented as a Zachman style matrix (see the
following figure), and is designed to guide the engineers through the modeling process and answer their
questions, like “how to organize the model?”, “what is the modeling workflow?”, “what model artifacts
should be produced in each step of that workflow?”, “how these artifacts are linked together?”, and so on.

As you can see, the framework defines the Problem, Solution, and Implementation domains for
developing the system of interest (SoI). Each domain is represented as a separate row of the MagicGrid
framework. The row that represents the Problem domain splits into two, to convey that the Problem
domain should be defined by looking at the SoI from the black-box and then from the white-box
perspective. The row that represents the solution domain splits into numerous inner rows to convey that
the solution architecture of the system can be specified in multiple levels of detail. The row that represents
the Implementation domain doesn’t split and is not fully highlighted like the upper rows, to convey
that everything except the implementation requirements specification in this domain is not a part of MBSE
and therefore appears outside the scope of the matrix.

Each domain definition includes different aspects of the SoI. These aspects match the four pillars of the
SysML, that is, Requirements, Structure, Behavior, and Parameters (also known as Parametrics); plus, the
Safety & Reliability aspect added in the 2nd edition of this book. They are represented as columns of the
matrix.

A cell at the intersection of some row and column represents a view of the system model, which can
consist of one or more presentation artifacts. The presentation artifact can be a diagram, including the
elements that can be used for modeling there, matrix, map, or table.

Though it is not conveyed in the framework, more than one solution architecture can be provided for a
single problem. In such case, a trade-off analysis is performed to choose an optimal solution for
SoI implementation.

12 | MagicGrid® Book of Knowledge

Alignment to ISO/IEC/IEEE 15288

Before we explore MagicGrid in detail, it is important to disclose how it fits into the overall systems
engineering life cycle defined by ISO/IEC/IEEE 15288.

MagicGrid actually addresses the technical ISO/IEC/IEEE 15288 processes only.

To start with the Business or Mission Analysis process, it is considered to be the input to the Problem
domain of the MagicGrid and is usually described by using other technologies, like OMG® Unified
Architecture Framework (UAF)®, which can be utilized in the same modeling tool, prior to applying the
MagicGrid framework for developing the identified systems.

The Problem domain of this framework covers the entire Stakeholder Needs and Requirements Definition
process. The subprocesses are mapped as follows:

• Stakeholder Needs Definition subprocess to the Stakeholder Needs cell of MagicGrid.

• Operational Concept and Other Life Cycle Concepts Development subprocess to the System Context,
Use Cases, and Measures of Effectiveness (MoEs) cells, all together.
• Transformation of Stakeholder Needs into Stakeholder Requirements subprocess to the Conceptual
Subsystems, Functional Analysis, and MoEs for Subsystems cells, all together.

The System Requirements Definition process is carried out in the System Requirements cell, although it can
be extended in the Subsystem Requirements and Component Requirements cells, too.

Right after system requirements are completed, the Architecture Definition process takes place. In
MagicGrid, it covers all system-level and subsystem-level cells of the Solution domain, except the System
Requirements and Subsystem Requirements cells. The Architecture Definition process also covers the
component-level cells, but only partially. These cells are covered by the Design Definition process, too. This
is where the architecture relates to design, or in other words, MBSE meets Model-Based Design (MBD),
and activities of the Assess architecture candidates subprocess are performed. Once they are completed, the
Design Definition process is carried out.

Finally, the Implementation process transforms requirements, architecture, and design into actions that
create the system element according to the practices of the selected implementation technology, using
appropriate technical specialties or disciplines. It is covered in the Implementation domain of MagicGrid.

MagicGrid® Book of Knowledge | 13

Case study

In this book, the modeling framework applies to the specification and architecture of the Vehicle Climate
Control System (VCCS). Technical accuracy and the feasibility of the actual solution proposed are not high
priorities.

How to read this book

It’s important to understand that this book doesn’t replace the modeling tool documentation. It is
intended to supplement it.

The structure of the book corresponds the layout of the MagicGrid framework. It is divided into three main
chapters, each describing the modeling workflow within a single domain. These are the Problem
domain (including the Black-box perspective and White-box perspective sub-chapters), the Solution
domain (including sub-chapters that correspond different phases of creating the solution domain model),
and the Implementation domain.

While chapters consist of sub-chapters, the latter consist of multiple sections, each named after relevant
cell of the MagicGrid framework. The order of these cells defines the modeling workflow. Every section
includes the brief overview of the cell to explain the purpose of modeling in this cell, who is responsible for
the results, and how to model. It also gives the list of basic steps for the modeler to go through to achieve
the expected result. Each step is described in the related sub-section, which provides answers to the
questions, like “what to do?”, “why to do it?”, and “how to do it?”. Follow these instructions from the very
beginning until the end of the book to create a sample model of the VCCS by yourself. If the version of
your modeling tool is 2021x or later, you are welcome to study the sample model there; it is saved as VCCS
(MG BoK v2 Sample).mdzip.

There are information boxes marked with the symbol, in the text. These boxes are designated to
provide you with additional information, tips and tricks for smarter modeling, or references to more
detailed information sources.

Moreover, the second edition of the book includes:

• Annex A, which provides comprehensive instructions on how to enrich the system model with
safety and reliability information.
• Annex B, which provides a detailed description of transitioning from System of Systems to a single
system architecture in the model-based environment.

All of the keywords and acronyms are defined in the Glossary.

How to provide feedback

We appreciate any feedback from our readers for the next version of this book. Send us an email
at [email protected].

14 | MagicGrid® Book of Knowledge

PROBLEM DOMAIN
The purpose of the problem domain definition is to analyze stakeholder needs and refine them with
SysML model elements to get the clear and coherent description of what functions are expected from the
SoI in the defined environment.

As shown in the following figure, the problem domain analysis is performed in two phases. In the first
phase, the SoI is considered a black box. The main focus is on how it interacts with the defined
environment without getting any knowledge about its internal structure and behavior (i.e., the operational
analysis of the SoI is performed). In the second phase, the black box is opened and the SoI is analyzed
from the white-box perspective, which helps to understand the expected behavior and conceptual
structure of the SoI (i.e., the functional analysis of the SoI is performed). It is also important to note that
the problem definition is not about the logical or the physical architecture of the SoI.

As you can see in the following figure, both phases of the problem domain definition consist of specifying
the requirements, structure, behavior, and parameters of the SoI. The only difference is in perspective.

Black-box perspective
Problem domain analysis and definition from the black-box perspective is the first step towards specifying
what the SoI shall do fulfill the stakeholders' expectations. The goal of this phase is to understand how
the SoI interacts with its environment to solve the given challenges. As the SoI is considered a black box,
the analysis is not concerned with its internal structure and behavior. The analysis is performed to define
the main inputs and outputs, black-box functions, and quantifiable characteristics of the SoI, while
operating in a variety of system contexts.

MagicGrid® Book of Knowledge | 15

The following pages describe how to capture the problem domain definition from the black-box
perspective in the SysML model.

Stakeholder Needs
What is it?

The cell represents information gathered from various stakeholders of the SoI. This information includes
primary user needs, system-related government regulations, industry standards, policies, procedures, and
internal guidelines, etc.

Stakeholder needs can be captured by interviewing stakeholders, giving them questionnaires, discussing
needs in focus groups, or studying documents written in diverse formats. Though this information is raw,
it does not need to be specially rewritten. However, it needs to be analyzed and refined in other cells of the
problem domain model. Once this is done, traceability relationships are created to convey which artifacts
from the problem domain model refine every stakeholder need (see Chapter Traceability to Stakeholder
Needs). These refinements will serve as a basis for specifying system requirements (see Chapter System
Requirements).

Who is responsible?

The elicitation and analysis of stakeholder needs can be performed by the Requirements Engineer or
Business and Mission Analyst.

16 | MagicGrid® Book of Knowledge

How to model?

Stakeholder needs can be captured directly in the modeling tool, or imported from other tools and
formats, including:

• Requirements management tools (i.e., Requirements Engineer (TRM) (as of 19.0 SP4), IBM®
Rational® DOORS®)
• Spreadsheets (i.e., Microsoft Excel file)
• ReqIF file (which contains data exported from another tool, such as, IBM® Rational® DOORS®, PTC
Integrity Modeler, Polarion® REQUIREMENTS™, or Siemens Teamcenter)

In the modeling tool, a single stakeholder need can be captured as a SysML requirement and have a
unique identification number, name, and textual specification. The entire list of stakeholder needs can be
displayed in a SysML requirement table (see the following figure) or diagram. Stakeholder needs can be
hierarchically related to each other (grouped by source, for example). It can be two categories of them:
functional and non-functional. This facilitates further analysis of stakeholder needs.

Tutorial
• Step 1. Organizing the model for stakeholder needs
• Step 2. Creating a table for stakeholder needs
• Step 3. Capturing stakeholder needs
• Step 4. Grouping stakeholder needs
• Step 5. Numbering stakeholder needs
• Step 6. Categorizing stakeholder needs

Step 1. Organizing the model for stakeholder needs

A well-organized model is easier to read, understand, and maintain; thus, we recommend organizing your
model into packages. Packages in a SysML model are very similar to folders in a file system: while folders
are used to organize files (including other folders), packages are used to organize diagrams and elements
(including other packages).

According to the design of the MagicGrid framework, model artifacts that capture stakeholder needs
should be stored under the structure of packages, as displayed in the following figure. As you can see, the

MagicGrid® Book of Knowledge | 17

top-level package represents the domain, the medium-level package represents the perspective, and the
bottom-level package represents the cell.

 You can skip this step (as well as other steps of establishing the model structure), if you use the
MagicGrid v2 QuickStart or MagicGrid v2 Blank template. They both come with the 2021x version of
the modeling tool, and include a predefined structure based on the MagicGrid framework.

 Using numbers in package names enables you to preserve the package order in the model tree.

To organize the model for stakeholder needs

1. Right-click the Model package (this is the default name of the root package) and select Create
Element.
2. In the search box, type pa, the first two letters of the element type Package, and press Enter.
3. Type 1 Problem Domain to specify the name of the new package and press Enter.
4. Right-click the 1 Problem Domain package and select Create Element.
5. Repeat steps 2 and 3 to create the package named 1 Black Box.
6. Right-click the 1 Black Box package and select Create Element.
7. Repeat steps 2 and 3 to create the package named 1 Stakeholder Needs.

Step 2. Creating a table for stakeholder needs

There are many ways to capture stakeholder needs in the SysML model. You can do this directly in the
Model Browser or over a presentation view (i.e., a requirement diagram or table), or another diagram,
which can display requirements (for example, block definition diagram (bdd)). In this tutorial, we utilize
the requirement table, as it is the best way to get stakeholder needs into the model from an external
source, such as Microsoft Excel spreadsheet or IBM® Rational® DOORS®.

It is important to notice that the MagicGrid BoK always describes one possible way to complete the task
(for example, we choose to create requirement table to capture and display stakeholder needs for the
reason explained above), but this does not mean that you cannot do this in another way. For a
comprehensive description of all the alternatives, refer to the latest documentation of the modeling tool.

18 | MagicGrid® Book of Knowledge

To create the SysML requirement table for capturing stakeholder needs

1. Right-click the 1 Stakeholder Needs package and select Create Diagram.

2. In the search box, type rt, where r stands for requirement and t for table, and then double-press Enter.
The table is created.

 Note that the table is named after the package where is it stored. This name suits for the table,
too. You need only remove the sequence number from its name.

3. Select the 1 Stakeholder Needs package and drag it to the Scope box in the Criteria area of the table.
From now on, the table is updated every time you append the contents of the related package.

Step 3. Capturing stakeholder needs

Suppose you need to capture the following stakeholder needs into your model:

# Name Text
1 Ambient Temperature I want to see the ambient temperature on the screen
or some other output device.
2 Comfortable Temperature I'd like to feel comfortable temperature while being in
the cabin.
3 Desired Temperature It should be a possibility to easily specify the desired
temperature.
4 Heating & Cooling The unit must be able to heat and cool.
5 Energy Consumption I'd like the unit to consume as little energy as
possible.
6 Manual Control I should be able to start and stop climate control by
myself.
7 Sound Level Climate control unit in max mode shall not be louder
than engine.
8 Total Mass Mass of the unit shall not exceed 2 percent of the
total car mass.

A single stakeholder need can be stored as a SysML requirement, which has a unique identification, name,
and text specification.

MagicGrid® Book of Knowledge | 19

You can get the stakeholder needs into your model by one of the following means:
• Capturing information directly in the model
• Copying information from a Microsoft Excel spreadsheet
• Synchronizing information from a Microsoft Excel spreadsheet
• Importing information from a ReqIF file
• Synchronizing information from IBM Rational DOORS

Capturing information directly in the model

To capture stakeholder needs directly in the model

1. Open the table created in step 2 of this cell tutorial, if not already opened.
2. On the toolbar of the table, click Add New and select Requirement.

3. In the newly created row, type the name and text of the first stakeholder need displayed in the table,
in step 3 of this cell tutorial.
4. Repeat steps 2 and 3 until you have all items in the table, as displayed in the following figure.

 As you can see, items created in the SysML requirements table also appear in the Model
Browser.

Copying information from a Microsoft Excel spreadsheet

If you are provided with a list of stakeholder needs in a Microsoft (MS) Excel spreadsheet, you can easily
transfer this information to the SysML model by utilizing the copy-paste capability of the modeling tool.

20 | MagicGrid® Book of Knowledge

To copy stakeholder needs from the MS Excel spreadsheet

1. Open the spreadsheet and select the contents you want to copy to your model.

2. Copy the contents:

• Press Ctrl+C on Windows or Linux.

• Press Command+C on macOS.

3. Switch to the modeling tool and open the SysML requirement table created in step 2 of this cell
tutorial, if not already opened.
4. Paste the contents to the table:

• Press Ctrl+V on Windows or Linux.

• Press Command+V on macOS.

5. Select Requirement as the type for the pasted data.

MagicGrid® Book of Knowledge | 21

 6. Wait while stakeholder needs are pasted into the table.

 As you can see, items copied to the SysML requirements table also appear in the Model
Browser.

Synchronizing information from a Microsoft Excel spreadsheet

Another option to bring stakeholder needs from an MS Excel spreadsheet into a SysML model is to
synchronize the model with the spreadsheet.

To synchronize stakeholder needs from the MS Excel spreadsheet

1. Open the table created in step 2 of this cell tutorial, if not already opened.
2. On the toolbar of the table, click and then choose Excel/CSV File > Select File.
3. In the Excel/CSV Sync Options dialog, select the MS Excel file from your file system.

4. Click Details to expand the dialog.

22 | MagicGrid® Book of Knowledge

5. Map columns of the SysML requirement table to columns of the MS Excel spreadsheet. For this, do
the following:

a. Select a column on the left and drag it to the relevant column on the right.
b. Release the mouse.

 The modeling tool might require you to refresh the dialog before setting the mappings.
Therefore, if you see this notification at the bottom of the dialog, click Refresh.

6. Click OK.
7. When the notification appears on the bottom right corner of the modeling tool window (see the
following figure), click Read from File.

MagicGrid® Book of Knowledge | 23

 8. Wait while stakeholder needs are synchronized with the table. The relevant notification displays in the
bottom right corner, when synchronization is completed.

 Items synchronized to the SysML requirements table also appear in the Model Browser.

As you can see, all the items in the table are highlighted in green. The legend above reveals that all new
items are highlighted like this. You can change the legend colors or hide the legend from the table. This,
along with more information about using legends, can be found in the documentation of the modeling
tool.

Importing information from a ReqIF file

At times, stakeholder needs are gathered by using one of these tools:

• IBM® Rational® DOORS®
• PTC Integrity Modeler
• Polarion® REQUIREMENTS™
• Siemens Teamcenter
• Requirements Engineer (TRM)

In such a case, information between a requirements management tool and the modeling tool can be
transferred via a Requirements Interchange Format (ReqIF) file. Thus, getting stakeholder needs into your
model includes:

1. Exporting requirements from the requirements management tool to a ReqIF file.

2. Importing the ReqIF file to the SysML model in the modeling tool.

We skip the first step, assuming that you’ve already been provided with the ReqIF file.

24 | MagicGrid® Book of Knowledge

To import stakeholder needs from the ReqIF file

1. From the main menu, select File > Import From > Requirements Interchange Format (ReqIF) File.
2. Select the ReqIF file form your file system and click Open.
3. In the Select Package for New Data dialog, select the package where you want to store the imported
stakeholder needs (i.e., the package 1 Stakeholder Needs).

4. Click OK.
5. Wait while stakeholder needs are imported into the 1 Stakeholder Needs package. Since that package is
the scope of the SysML requirement table (see step 2 of this cell tutorial), all imported items also
appear in the table.

MagicGrid® Book of Knowledge | 25

Synchronizing information from IBM Rational DOORS

If stakeholder needs are gathered in an IBM® Rational® DOORS® project (see the following figure), you
can get them into the SysML model by synchronizing data between DOORS and the modeling tool.

 The modeling tool can be synchronized with DOORS only if it has the Cameo DataHub Plugin
installed.

In this tutorial, we assume that DOORS are used only for capturing stakeholder needs, and further analysis
is performed in the SysML model. For this reason, we need to establish one-way synchronization between
DOORS and the modeling tool. In one-way synchronization, items are copied from DOORS to
the modeling tool, but no items are ever copied back from the modeling tool to DOORS.

 Although the following procedure defines the steps for synchronization with IBM® Rational®
DOORS® Next Generation 6.x, it is also possible to synchronize requirements from IBM® Rational®
DOORS® 8.0-8.3 and 9.0-9.6 as well as from IBM® Rational® DOORS® NG 4.x and 5.x.

To synchronize stakeholder needs from DOORS NG 6.x to the modeling tool

1. Be sure you have Cameo DataHub Plugin 19.0 or later installed on top of the modeling tool.
2. Connect to the DOORS NG 6.x repository:

a. From the main menu of the modeling tool, select Tool > DataHub > DataHub Explorer. The
Cameo DataHub Explorer panel opens to the right of the modeling tool window.
b. On the toolbar of the open panel, click the Add Data Source button.

26 | MagicGrid® Book of Knowledge

 c. Provide the required information and click Create, as shown in the following figure.

d. Wait while the modeling tool connects to the DOORS NG 6.x repository. After the operation is
completed, the tree of the Cameo DataHub Explorer panel displays projects from the DOORS
repository (as packages). One of the projects is VCCU Project (Requirements).

3. Synchronize data:

a. In the tree of the Cameo DataHub Explorer panel, select the Stakeholder Needs package and drag
it onto the 1 Black Box package in the Model Browser on the left of the modeling tool window.

MagicGrid® Book of Knowledge | 27

  In this case, you should ignore the inner package, 1 Stakeholder Needs, created in step 1 of
this cell tutorial. If you decide to ignore this recommendation and drag the items of
stakeholder needs to the package 1 Stakeholder Needs one by one (ignoring the package
CCU Stakeholder Needs), keep in mind that you risk missing the items that will be created in
the DOORS project afterwards. In that case, only individual items are synchronized, not
the whole project.

b. In the open dialog, do the following:

i. Click One-way Synch to Cameo Systems Modeler to select the synchronization direction.
ii. Specify the mappings:

1. Folder to Package
2. Requirement to Requirement

iii. Click the OK button.

c. Wait for the message announcing that synchronization is completed. Synchronized items in both
trees become marked with S (see the following figure).

4. Expand the contents of the 1 Black Box package and see the Stakeholder Needs package with all its
data imported into your model.

 “S” indicates synchronized data, and “1” the synchronization source. As you have selected
One-way Synch to Cameo Systems Modeler, the DOORS NG 6.x becomes the single source for
synchronization.

28 | MagicGrid® Book of Knowledge

Step 4. Grouping stakeholder needs

 This does not apply if stakeholder needs are synchronized from external software for requirements
management (for example, IBM® Rational® DOORS®), because all the changes you make in the
SysML model will be overwritten after routine synchronization. Grouping, numbering, and
categorization of stakeholder needs in that case should be performed in the requirements
management tool.

After you have stakeholder needs in your model, you can arrange them into groups according to their
nature. These groups can be system-related government regulations, user needs, industry standards, or
business requirements, among others.

For grouping stakeholder needs, you need to create extra elements of the SysML requirement type, each
representing a single group of stakeholder needs and generally referred to as a grouping requirement. A
grouping requirement might not have any text itself, only a unique identification number and name.

Now let’s analyze the stakeholder needs of the Vehicle Climate Control Unit in your model. Actually, the
majority of them are user needs, and therefore can be grouped under the User Needs requirement. The
stakeholder need captured as the Total Mass requirement sounds like a design constraint. For this reason, it
should appear under the Design Constraints requirement. The top-level grouping requirement simply
named Stakeholder Needs can also be created, to complete the requirements hierarchy.

To group stakeholder needs

1. Create the User Needs requirement:

a. In the Model Browser, right-click the 1 Stakeholder Needs package and select Create Element.
b. In the search box, type re – the first letters of the requirement type, and press Enter. A new
element of the requirement type is created.
c. Type User Needs to specify the name of the requirement and press Enter.

2. In the Model Browser, select all the items except the Total Mass requirement (the first item in the list)
and the User Needs requirement (the last item in the list).

MagicGrid® Book of Knowledge | 29

  To select the set of adjacent items in the tree, click the first one, press Shift, and while holding
it down, select the last one.

3. Drag the selection onto the User Needs requirement (see the figure on the left). It becomes a grouping
requirement (see the figure on the right).

4. Repeat the step 1 to create the Design Constraints requirement.

5. Select the Total Mass requirement and drag it onto the Design Constraints requirement. The latter
becomes a grouping requirement.
6. Repeat step 1 to create the Stakeholder Needs requirement.
7. Select both grouping requirements: User Needs and Design Constraints.
8. Drag the selection onto the Stakeholder Needs requirement. It becomes the top-level grouping
requirement. See appropriate changes in the SysML requirement table.

 Be sure the table is displayed in the Compact tree mode. For this, click the Options button
on the table toolbar and then select Display Mode > Compact Tree.

30 | MagicGrid® Book of Knowledge

Step 5. Numbering stakeholder needs

 This does not apply if stakeholder needs are synchronized from external software for requirements
management (for example, IBM® Rational® DOORS®), because all the changes you make in the
SysML model will be overwritten after the routine synchronization. Grouping, numbering, and
categorization of stakeholder needs in that case should be performed in the requirements
management tool.

SysML requirements, which we utilize for capturing and storing stakeholder needs, may have hierarchical
numbers. The hierarchical number should not be mixed with the unique identifier, which can never be
changed.

Stakeholder needs are by default numbered using natural numbers, from 1 to infinity. In order to
distinguish stakeholder needs from system or implementation requirements, we recommend numbering
them with special prefixes (i.e., SN-, where S stands for stakeholder and N for needs).

After organizing stakeholder needs into groups, their numbering became inaccurate, as it now begins from
11. Let's fix this, too.

To number stakeholder needs with the prefix SN-

1. In the Model Browser, right-click Stakeholder Needs (the top-level grouping requirement), and
select Element Numbering. The Element Numbering dialog opens.
2. In the dialog, click the empty value cell of the Prefix column to initiate the editing, type SN- (see the
following figure), and press Enter. The hierarchical number of the Stakeholder Needs requirement is
updated to SN-11.

MagicGrid® Book of Knowledge | 31

 3. Click Details to see more options, if they are not seen yet.
4. Click to select the Renumber From Initial Value check box.
5. Click the Renumber Recursively button. Every item of stakeholder needs becomes numbered with the
prefix SN- starting from number 1.

 If you want to make sure that all the items are renumbered, expand the model tree on the left
side of the dialog to see the entire hierarchy of the stakeholder needs for the VCCU.

6. Click OK to close the dialog.

32 | MagicGrid® Book of Knowledge

Step 6. Categorizing stakeholder needs

 This does not apply if stakeholder needs are synchronized from external software for requirements
management (for example, IBM® Rational® DOORS®); all the changes you make in the SysML
model will be overwritten after the routine synchronization. Grouping, numbering, and
categorization of stakeholder needs should be performed in the requirements management tool.

Stakeholder needs can be either functional or non-functional. Functional stakeholder needs specify what
the SoI is expected to do. They are further refined by use cases and use case scenarios (see Chapter Use
Cases). Non-functional stakeholder needs that identify quantifiable characteristics of the SoI are refined
by measures of effectiveness (see Measures of Effectiveness).

Categorization facilitates further analysis of stakeholder needs, especially when dealing with a large scope
of information. This step may be skipped if the case does not require it.

While non-functional stakeholder needs can be captured as SysML requirements, functional stakeholder
needs should be converted to SysML functional requirements. In technical terms, functional requirement
(together with interface, performance, design, and other more specific requirements, which are not
needed thus far) is a sub-type of the requirement and inherits all its features.

In the given sample, the majority of stakeholder needs (except Sound Level, Energy Consumption, and Total
Mass) are functional. Therefore, they should be refactored to SysML functional requirements.

To convert SysML requirements to functional requirements

1. In the Model Browser or on the requirements table, select all requirements except Sound Level, Energy
Consumption, and Total Weight.

 To select a set of non-adjacent items in the tree, click them one by one while holding down the
Ctrl key.

2. Right-click the selection and then choose Refactor > Convert To > More Specific > Functional
Requirement.

MagicGrid® Book of Knowledge | 33

As a result, all the requirements are converted to functional, and F instead of R is displayed on each
element icon.

Stakeholder Needs done. What's next?

• Stakeholder needs are analyzed to identify system contexts (see Chapter System Context).
• Functional stakeholder needs are further analyzed and refined by use cases and use case scenarios
(see Chapter Use Cases).
• Non-functional quantifiable stakeholder needs are refined by measures of effectiveness (see
Chapter Measures of Effectiveness).

34 | MagicGrid® Book of Knowledge

System Context
What is it?

System context (or operating environment) determines an external view of the system. The system context
introduces all external entities that do not belong to the system but do interact with it. More than one
system context can be defined for a single SoI. Besides the system itself (currently considered to be a black
box), the collection of elements in the particular system context can include external systems (natural or
artificial) and users (humans, organizations) that interact with the SoI to exchange data, matter, energy, or
even human resources.

In overall, this cell produces:

• Definitions of system contexts
• Participants of each system context: SoI, supposed users of the system, other systems, etc.
• Interactions between participants of each system context
• Items that are exchanged over these interactions

Who is responsible?

System contexts can be identified and specified by the Systems Analyst or Systems Engineer.

How to model?

System context diagram pictures the SoI at the center, with no details of its interior structure, surrounded
by all its interacting entities, such as supposed users and external systems to name a few. In the SysML
model, system contexts can be captured as elements of the SysML block type. All the entities can be
captured as SysML blocks, too. As defined by SysML, once the entity block is used in the system context
block, a part property typed by the former, is created for the latter.

 We recommend using blocks rather than actors (stick man notation) to capture even human-nature
entities (for example, users of the SoI) because the actor cannot be a part of the system context in
the SysML model.

A SysML internal block diagram (ibd) can serve to represent the entities of the particular system context,
and interactions between them. Each entity is represented as a part property of the related system context
block.

Interactions within the system context can be captured as connectors with one or more item flows
between those part properties. Data, matter, or energy are items that can flow. They can be stored in the
model as signals, blocks, or value types, depending on their nature, and in general are refereed as
exchange items.

MagicGrid® Book of Knowledge | 35

As you can see in the following figure, system context diagrams can be supplemented with various
images, so that the result could be presented to stakeholders, including even those who are incapable
of reading models.

Tutorial
• Step 1. Organizing the model for the system context
• Step 2. Capturing system contexts
• Step 3. Creating an ibd for a system context
• Step 4. Capturing participants of the system context
• Step 5. Specifying interactions between the participants of system context
• Step 6. Adding item flows to the system context

Step 1. Organizing the model for the system context

Following the structure of the MagicGrid framework, system contexts and owned-by-them SysML internal
block diagrams, together with the elements they represent, should be stored in a separate package inside
the 1 Black Box package. We recommend naming the package after the cell, that is, 2 System Context.

To organize the model for the system context

1. Right-click the 1 Black Box package and select Create Element.

2. In the search box, type pa (the first two letters of the element type Package) and press Enter.
3. Type 2 System Context to specify the name of the new package and press Enter.

36 | MagicGrid® Book of Knowledge

Step 2. Capturing system contexts
System contexts can be identified by analyzing stakeholder needs (see Chapter Stakeholder Needs).
Stakeholder needs of the Vehicle Climate Control Unit enable you to identify a single context: the Vehicle In
Use. It can be captured in your model as an element of the SysML block type.

To create a block that captures the Vehicle In Use system context

1. Right-click the 2 System Context package and select Create Element.

2. In the search box, type bl (the first letters of the required element type) and press Enter.
3. Type Vehicle In Use to specify the name of the new block and press Enter. The block is created and
represented in the Model Browser.

It is important to understand that the SoI may operate in more than one system context. To illustrate this,
we need to capture one more system context. In this example, we will use the Vehicle in Maintenance (see
the following figure). You can create the appropriate block on your own by following the procedure above.

The recently captured system context is not included in further analysis. It is not mentioned in the
stakeholder needs and has been created for demonstration purposes only.

MagicGrid® Book of Knowledge | 37

Step 3. Creating an ibd for a system context
To specify participants of the Vehicle In Use system context, you first need to create a SysML internal block
diagram for the block that captures that system context.

To create an IBD for the Vehicle In Use block

1. Right-click the Vehicle In Use block and select Create Diagram.

2. In the search box, type ibd, the acronym of the SysML internal block diagram, and then double-press
Enter. The diagram is created and represented in the Model Browser under the Vehicle In Use block.

 The diagram is named after the block that owns it. As this name perfectly suits the diagram,
there is no need to change it.

Step 4. Capturing participants of the system context

Entities that participate in the system context can be captured in the model as the part properties of that
system context, where each part property is typed by the block, that captures the appropriate entity (for
example, the SoI, its supposed users, or some external system).

 Here you should remember the SysML Specification, which tells you that blocks should be used for
defining concepts and parts are used for specifying the usage of these concepts. Any concept, once
defined in your model, can be used in many system contexts. For example, the Vehicle Climate
Control Unit SoI defined as a block can type a part property of the Vehicle in Use system context and
part property of the Vehicle in Maintenance system context, meaning that it participates in both.

Stakeholder needs indicate that the Vehicle In Use system context includes the following participants:
• Vehicle Climate Control Unit, the SoI
• Vehicle Occupant (driver or passenger), the supposed user of the SoI
• Vehicle Cabin, the closed environment that must be provided with a comfortable air temperature
• Temperature Sensor, the external system to provide an ambient temperature to the SoI

To capture participants of the Vehicle In Use system context

1. Open the ibd created in step 3 of this cell tutorial, if not yet opened.

38 | MagicGrid® Book of Knowledge

2. Make sure the Type Selection Mode is on in the diagram. Otherwise, parts created in this diagram will
not be typed by blocks.

 In this mode, when you create a part property, the tool offers you a list of blocks, from which
you can choose the part property type. If the necessary block does not exist yet, you can
simply create it by typing its name directly on the part property shape.

3. Click the Part Property button on the diagram palette and then click an empty place on the diagram
pane. An unnamed part property is created, and the list of existing blocks to type it is offered.
4. Type VCCU next to the colon (":") directly on the part property shape and press Enter. The VCCU block
to type the just-created part property emerges in the Model Browser.

 When the Type Selection Mode is enabled, typing on the part property shape always defines
the name of the new block. This new block types that part property, but not the name of the
part property. If you want to specify the name as well, type it before the colon (":") on the
shape. However, names are not necessary, since part properties created in this diagram can be
easily identified by their types. Names are useful when you want to specify more than one
usage of the same block (for example, the front-left, front-right, rear-left, and rear-right wheels
of the vehicle).

MagicGrid® Book of Knowledge | 39

 5. Repeat steps 3 and 4 to create the rest of part properties: one typed by the Vehicle Occupant block,
one by the Cabin block, and one by the Temperature Sensor block.

 Since you need to create more than one part property in succession, use the functionality of
the Sticky button (see the following figure), so it is not necessary to click the Part Property
button on the palette each time you want to create a new part property. When the Sticky
mode is enabled, you can create as many elements of the selected type as you need. Press Esc
when you are finished.

When you are finished, you have four part properties and the same number of blocks that type them in
your model.

To keep your model well-organized, we recommend storing the blocks that capture users and other
systems in a separate package within the 2 System Context package. For this, you need to create an internal
package and drag the blocks to the new location in the Model Browser. Hold down the Shift key to select
more than one block.

40 | MagicGrid® Book of Knowledge

Step 5. Specifying interactions between the participants of system context
The stakeholder needs indicate that the Vehicle Climate Control Unit interacts with all participants of the
Vehicle In Use system context. We need to capture this information in the model.

Remember that interactions can be specified as connectors with one or more item flows between
appropriate part properties. We will start with drawing the connectors.

To draw a connector between the :VCCU and :Vehicle Occupant part properties

1. Select the :VCCU part property and click the Connector button on its smart manipulator toolbar.
2. Select the :Vehicle Occupant part property. The connector is created.

MagicGrid® Book of Knowledge | 41

You can create other connectors using this same method. Once you have finished, your ibd should look
very similar to the one below.

Step 6. Adding item flows to the system context

To begin, we first need the items that can flow from one part property to another across the connector and
back. We recommend storing the elements that capture these items in a separate package (see the
following figure). The contents of this package will be used in the white-box analysis model, too.
Therefore, instead of creating the new package directly under the 2 System Context package (which is a
part of the black-box analysis model) create it under the 1 Problem Domain package.

If we assume that you have created the 3 Exchange Items package (the package number is 3 because
number 2 will be used for the White Box package) on your own, we can move directly to creating items that
can flow and assigning item flows to the connectors. One or more item flows can be assigned on a single
connector.

Stakeholder needs indicate that:

• Vehicle Occupant shall be able to turn the VCCU on and off, and specify the desired temperature (see
SN-1.1.2, SN-1.1.6)
• VCCU shall provide a comfortable air temperature for the Vehicle Occupant sitting in the Cabin (see
SN-1.1.7)
• VCCU shall obtain the ambient temperature from the Temperature Sensor of the vehicle (see
SN-1.1.5)
• VCCU shall display status and temperature information (see SN-1.1.5)

42 | MagicGrid® Book of Knowledge

These statements enable you to identify the following exchange items:
• User Command
• Desired Temperature
• Air
• Comfortable Air
• Ambient Temperature
• Information

Air and Comfortable Air are matter; thus, they can be captured in the model as blocks. The rest of the
items can be captured as signals, because they mean data.

To capture the User Command, Desired Temperature, and Information in the model and specify them as
items conveyed by the item flows between the :VCCU and :Vehicle Occupant part properties

1. Right-click the 3 Exchange Items package and select Create Element.

2. In the search box, type si (the first two letters of the element type Signal) and press Enter.
3. Type User Command to specify the name of the new signal and press Enter.
4. Drag the signal to the connector between :VCCU and :Vehicle Occupant part properties in the Vehicle In
Use ibd.
5. In the open dialog, select From Vehicle Occupant To VCCU as the direction of the flow and click Finish.
A new item flow conveying the User Command signal is created and displayed on the connector.
6. Repeat steps 1 to 3 to create the Desired Temperature signal.
7. Drag the signal to the same connector.
8. In the open dialog, select From Vehicle Occupant To VCCU as the direction of the flow and click Finish.
A new item flow conveying the Desired Temperature signal is created and displayed on the connector.
9. Repeat steps 1 to 3 to create the Information signal.
10. Drag the signal to the same connector.
11. In the open dialog, click Finish. A new item flow conveying the Information signal is created and
displayed on the connector.

MagicGrid® Book of Knowledge | 43

In the same manner, you can capture the remaining exchange items and then specify item flows on other
connectors in the Vehicle In Use ibd. Once finished, your ibd should look very similar to the one below.

As you can see in the figure below, system context diagrams can be supplemented with various images so
the result could be presented to stakeholders, including even those who are incapable of reading models.
For more information, see the latest documentation of the modeling tool.

System Context done. What's next?

• System contexts are used to analyze the expected behavior of the SoI (see Chapter Use Cases).
Behavior analysis begins with identifying the use cases of the SoI, and each use case belongs to one
or more system contexts of that SoI.
• Participants of each system context captured as part properties of that system context are also used
in the Use Cases cell. They can be represented as swimlanes in the SysML activity diagram that
describes a use case scenario.
• Exchange items identified in the System Context cell are used in the use case scenario as object flow
items and pin types.
• After the block that captures the system of interest is created, quantitative characteristics of the
system describing non-functional stakeholder needs can be captured. Thus, you can switch
to Measures of Effectiveness.

44 | MagicGrid® Book of Knowledge

Use Cases
.What is it?

In this cell, functional stakeholder needs are refined with use cases and use case scenarios. In comparison
to stakeholder needs, use cases are more precise in telling what people expect from the system and what
they want to achieve by using it. Each use case must belong to one or more system contexts defined in
the System Context cell (see Chapter System Context).

Overall, this cell produces:

• Functional use cases that provide meaningful value to the user.
• Use case scenarios on how the system is expected to interact with the user and/or other systems.

Who is responsible?

Use cases and their scenarios can be specified by the Systems Analyst or Systems Engineer.

How to model?

In the modeling tool, use cases of the system can be captured by utilizing the infrastructure of the SysML
use case diagram. Every use case diagram should be created considering one of the system contexts
defined in the System Context cell (see Chapter System Context). Use cases can be captured as elements
of the use case type. A single use case can be performed in different contexts.

Entities that perform one or more use cases of the system context should be represented as blocks in the
use case diagram. Here you should remember that these blocks type part properties of the system context.
As you can see in the following figure, even humans, such as the vehicle occupant, can be represented as
blocks (with or without images applied).

MagicGrid® Book of Knowledge | 45

  We recommend using blocks rather than actors (stick man notation) to capture even human-nature
entities (for example, users of the SoI) because the actor cannot be a part of the system context in
the SysML model.

Each use case must have a name and primary scenario. Alternative scenarios are optional. A use case
scenario can be captured in a form of SysML activity or sequence diagram. In the activity diagram, SoI,
supposed users of the system, and/or other systems can be represented as swimlanes (activity partitions).
In the sequence diagram, they are represented as lifelines. The SoI is meanwhile considered a black box.

46 | MagicGrid® Book of Knowledge

Tutorial
• Step 1. Organizing the model for use cases
• Step 2. Creating a diagram for use cases
• Step 3. Capturing use cases
• Step 4. Creating a diagram for specifying the use case scenario
• Step 5. Creating swimlanes and setting allocation mode
• Step 6. Specifying steps of the use case scenario
• Step 7. Adding item flows to the use case scenario
• Step 8. Supplementing the use case scenario with a parallel flow

Step 1. Organizing the model for use cases

Following the structure of the MagicGrid framework, use cases and their scenarios in the form of SysML
activity or sequence diagrams, with elements they represent, should be stored in a separate package inside
the 1 Black Box package. We recommend naming the package after the cell, that is, 3 Use Cases.

To organize the model for the use cases

1. Right-click the 1 Black Box package and select Create Element.

2. In the search box, type pa (the first two letters of the element type Package) and press Enter.
3. Type 3 Use Cases to specify the name of the new package and press Enter.

Step 2. Creating a diagram for use cases

You must create a SysML use case diagram to begin capturing use cases of the Vehicle In Use system
context (see Chapter System Context). You need to display the shape of the block that captures the system
context on the diagram pane as well. The Vehicle Occupant block, which represents the user of the
SoI, should also be displayed on the diagram pane.

MagicGrid® Book of Knowledge | 47

To create a diagram for capturing use cases performed in the Vehicle In Use system context

1. Right-click the 3 Use Cases package and select Create Diagram.

2. In the search box, type uc, where u stands for use and c for case, and then press Enter. The diagram is
created.

 If you have more than one system context, we recommend storing use cases of each system
context in a separate package. In this case, you do not need to.

3. Type Use Cases of Vehicle In Use SC to specify the name of the new diagram and press Enter again.

4. In the Model Browser, select the Vehicle In Use block. For this, use the quick find capability:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type Vehicle In U.
c. When you see the Vehicle In Use block selected in the search results list, press Enter, as shown in
the following figure. The Vehicle In Use block is selected in the Model Browser.

5. Drag the Vehicle In Use block to the newly created use case diagram pane. The shape of the Vehicle In
Use block appears on the diagram.

48 | MagicGrid® Book of Knowledge

 6. Repeat the sub-steps of Step 4 to find the Vehicle Occupant block and drag it to that diagram pane as
well. The shape of the Vehicle Occupant block appears on the diagram, too.
7. If the relationship is displayed on the diagram too, select it and press Delete to remove its symbol
from the diagram. When the dialog asks if you want to remove the relationship from the model, click
No.
8. Select the Vehicle In Use block and drag the bottom right corner of the symbol to enlarge it. It must be
larger to nest the use cases inside.

As you can see in the figure below, the Vehicle Occupant block can be displayed as an image you applied on
that block in the ibd of the Vehicle In Use system context (see Chapter System Context). For this, select the
block and click the Show As Image button on the diagram toolbar.

MagicGrid® Book of Knowledge | 49

Step 3. Capturing use cases
By analyzing stakeholder needs, we can see that the main goal of the VCCU is to provide a comfortable
temperature for all vehicle occupants while they are in the cabin. This goal can be captured in the model as
the Provide Comfortable Temperature use case of the Vehicle In Use system context. Let’s capture it in the
Use Cases of Vehicle In Use SC diagram.

To capture the Provide Comfortable Temperature use case

1. Open the use case diagram created in step 2 of this cell tutorial, if not yet opened.
2. Click the Use Case button on the diagram palette and move the mouse over the shape of the Vehicle In
Use block.
3. When you see the blue border around the shape of the Vehicle In Use block, click it. An unnamed use
case is created within the shape of the block.
4. Type Provide Comfortable Temperature to specify the name of this use case and press Enter.
5. Double-click the use case to open its Specification and make sure that the Vehicle In Use block is set as
its subject.

6. Select the Vehicle Occupant block (with or without image applied) that represents the user of the VCCU,
click the Association button on its smart manipulator toolbar, and then select the Provide
Comfortable Temperature use case to relate these two elements.

Step 4. Creating a diagram for specifying the use case scenario

A use case scenario can be captured in a SysML activity or sequence diagram. The SysML activity is more
suitable for specifying abstract behavior definitions, and the sequence diagram should be used when you
need to distinguish between synchronous and asynchronous interactions.

In this example, we chose the SysML activity diagram.

50 | MagicGrid® Book of Knowledge

To create a SysML activity diagram for specifying the Provide Comfortable Temperature use case scenario

1. In the Use Cases of Vehicle In Use SC diagram, right-click the shape of the Provide Comfortable
Temperature use case and select Create Diagram.
2. In the search box, type act, the acronym of the SysML activity diagram, and press Enter. The diagram
is created and opened. It is owned by the SysML activity with the same name.

 If you refer to the use case diagram, which displays the Provide Comfortable Temperature use
case, you can see that the shape of that use case is decorated with the rake ( ) icon. This
decoration means that the use case contains an internal diagram, the Provide Comfortable
Temperature activity diagram. Double-clicking the shape of the use case opens that internal
diagram.

Step 5. Creating swimlanes and setting allocation mode

Swimlanes in the activity diagram are used to represent the participants of the related system context (see
Chapter System Context). They allow you to establish allocation relationships between the structural and
behavioral parts of the model. Use these relationships in the black-box analysis model to specify which
participant of the system context is responsible for performing which step of the related use case scenario.

Remember that participants of the system context are captured as its part properties. When you create
swimlanes in the activity diagram of the nested use case, the modeling tool automatically recognizes and
suggests the part properties of the nesting system context for you to select for representation.

Additionally, you need to enable the allocation to usage mode in your activity diagram. This mode allows
you to convey that allocations are established considering the system context. Otherwise, allocations are
generic, unrelated to any system context.

Now let's represent the participants of the Vehicle In Use system context in the Provide Comfortable
Temperature use case scenario.

MagicGrid® Book of Knowledge | 51

To create swimlanes in the Provide Comfortable Temperature activity diagram

1. Open the activity diagram created in step 4 of this cell tutorial, if not yet opened.
2. Click the Vertical Swimlanes button on the activity diagram palette and then click an empty space on
that diagram pane.
3. In the open dialog, select Usage (Action allocated to Part) to enable the allocation to usage mode and
then click OK.

 By default, this mode is enabled in both MagicGrid templates.

4. In the Represent Properties dialog, click OK to represent all the participants of the related system
context.

52 | MagicGrid® Book of Knowledge

  As in this case, the dialog always displays the list of part properties that belong to the related
system context. Technically speaking, the related system context is the subject of the use case,
which encloses the activity diagram (see step 3 of this cell tutorial).

5. Click the OK button on the dialog. The swimlanes are displayed on the diagram.

 Swimlanes are always arranged alphabetically from A to Z, but the buttons and (which
appear when you select the header of the swimlane) enable you to rearrange them according
to your preference in the activity diagram.

When you are finished, you're ready to specify the Provide Comfortable Temperature use case scenario.

Step 6. Specifying steps of the use case scenario

The steps of the use case scenario can be captured in the activity diagram as call behavior actions. As
defined by SysML, a call behavior action may have an assigned behavior, captured in the model as an
activity, state machine, or interaction. As an activity, the behavior can be created along with the
referencing call behavior action, if the Behavior Creation Mode is enabled in your model. We recommend
modeling in this mode for the following reasons:
• It is easier to specify recurrent steps. Any step created in the model once, as an activity assigned to
some call behavior action, can then be referred by other call behavior actions in one or more activity
diagrams.
• When it is time to decompose a step, there is no need to create the referred activity. It is already
created and assigned as a behavior to the call behavior action which captures that step.
• The Activity Decomposition Map does not include call behavior actions. To ensure you have the
entire decomposition of the use case steps, you need to have them captured as activities.

To capture the steps of the main flow of the Provide Comfortable Temperature use case scenario

1. Be sure the Behavior Creation Mode is enabled in the diagram (see the following figure). Otherwise,
activities will not be created and assigned as behaviors to call behavior actions.

MagicGrid® Book of Knowledge | 53

 2. Click the Initial Node button on the diagram palette and then click an empty place on the diagram
pane, within the :Vehicle Occupant swimlane. The initial node is created and displayed on the diagram.
3. Click the Control Flow button on the smart manipulator toolbar of the initial node and then click
an empty place on the diagram pane in the same swimlane. An unnamed action with a new activity
assigned as its behavior is created.
4. Type Turn on Climate Control to name the activity and press Enter.

 Be sure to type the name after the colon (":"). Otherwise, the name is given to the action, but
not to the referred activity.

5. Click the Control Flow button on the smart manipulator toolbar of the newly created action and
then click an empty place on the diagram pane, within the :VCCU swimlane. An unnamed action with
a new activity assigned as its behavior is created.
6. Type Check System to name the activity. Remember to type after the colon. Press Enter after you are
finished.
7. Repeat steps 5 and 6 as many times as you need to create the actions you see in the following figure.
8. Select the shape of the Stop Climate Control action and click the Control Flow button on its smart
manipulator toolbar.

54 | MagicGrid® Book of Knowledge

 9. Right-click an empty place on the diagram pane and select Activity Final. The final node is created
and displayed on the diagram.

Now let's split the flow into two: the main flow, when the system can be started after checking, and the
alternative flow, when it cannot. To do this, we need to insert a decision node between the Check System
and Start Climate Control actions, and a merge node before the activity final node.

To update the use case scenario with an alternative flow

1. Click the Decision button on the diagram palette and then click the control flow between the Check
System and Start Climate Control actions. The decision node is inserted.

 Before inserting the new shape, you can make more space between the existing shapes on the
diagram pane. To do this, click the Pusher button on the diagram palette.

2. Select the control flow between the decision node and the Start Climate Control action, type [OK, and
press Enter (the closing bracket is added automatically). As a result, the OK guard is specified for this
control flow and displayed between square brackets on the diagram.

MagicGrid® Book of Knowledge | 55

  If you open the Specification of the control flow, you can see that the text between the squire
brackets has been set as the Guard property value.

3. Click the Merge button on the diagram palette and then click the control flow before the activity final
node.
4. Select the decision node, click the Control Flow button on its smart manipulator toolbar, then click
the merge node. The control flow is created.
5. Select that control flow, type [Not OK, and press Enter. The Not OK guard is specified for this control
flow and displayed between square brackets on the diagram.

After you finish, your Provide Comfortable Temperature diagram should look very similar to the one in the
following figure.

There is one more alternative flow that must be specified in the scenario, which says the vehicle occupant
can either set the desired temperature or enjoy the ambient temperature after they evaluate the comfort in
the cabin. For this, you need to insert another pair of decision and merge nodes in the diagram.

56 | MagicGrid® Book of Knowledge

To update the use case scenario with another alternative flow

1. Click the Decision button on the diagram palette and then click the control flow between the Evaluate
Comfort and Set Temperature actions. The decision node is inserted.
2. Select the control flow between the decision node and the Set Temperature action, type [Not
Comfortable, and press Enter. The Not Comfortable guard is specified for this control flow and displayed
between square brackets on the diagram.
3. Click the Control Flow button on the smart manipulator toolbar of the newly created decision node
and then select the Enjoy Temperature action. A new control flow is created.
4. Immediately type [Comfortable. The Comfortable guard is specified for this control flow and displayed
between square brackets on the diagram.
5. Click the Merge button on the diagram palette and then click the control flow between the Start
Climate Control and Measure Temperature actions.
6. Select the control flow between the Get Comfortable Air and Enjoy Temperature actions, click its arrow
end, and then drag it to the newly created merge node. The target node of the control flow changes.

After you finish, your Provide Comfortable Temperature diagram should look very similar to the one in the
following figure.

The simulation engine of the modeling tool, based on the OMG® Executable UML (fUML)™
standard, enables you to validate the specified behavior and see if your model is correct. To start the
simulation session, click the Run button on the Provide Comfortable Temperature activity diagram
toolbar. Then in the Simulation panel, click the Start button to start the model execution.
So far, we have used control flows to connect actions. In the next step we will specify what data, matter, or
energy flows between actions.

MagicGrid® Book of Knowledge | 57

Step 7. Adding item flows to the use case scenario
Now we will update the Provide Comfortable Temperature use case scenario with item flows, by
synchronizing them from the Vehicle In Use ibd created during the system context analysis (see Chapter
System Context). This helps to identify and eliminate logical inconsistencies across the structural and
behavioral parts in the problem domain model, such as the existence of a questionable item flow that does
not appear in any use case scenario, or identifying the need for a new item flow.

Before adding item flows to the activity diagram which captures the Provide Comfortable Temperature use
case scenario, we need to look at the Vehicle In Use ibd first. Let's start from the interaction between the
:Vehicle Occupant and the :VCCU part properties. We see that the vehicle occupant provides control
commands to and sets the desired temperature for the VCCU (see the highlighted fragment of the diagram
below).

The item flows conveying these signals must be realized between these part properties in the use case
scenario, as well. Remember that part properties specified in the Vehicle In Use ibd are represented by
swimlanes in the Provide Comfortable Temperature activity diagram. Item flows should be conveyed by
object flows, not control flows. Therefore, we need to convert relevant control flows to object flows.

To realize item flows conveying the User Command signal between the actions allocated to the :Vehicle
Occupant and the :VCCU part properties

1. Select the control flow between the Turn On Climate Control and Check System actions.
2. Right-click the selection and choose Refactor > Convert To > Object Flow. The control flow turns to
the object flow, and pins are created for both actions.
3. In the Model Browser, under the 3 Exchange Items package, select the User Command signal.
4. Drag the signal to the output pin of the Turn On Climate Control action. The User Command type is set
as the type for the output pin.
5. Drag the signal to the input pin of the Check System actions. The User Command type is set as the type
for the input pin.
6. Select the object flow and click Item Flow Manager button from its smart manipulator toolbar.
7. In the open dialog, select the item flow that conveys the User Command signal (see the following
figure). The item flow is realized between the relevant actions.

58 | MagicGrid® Book of Knowledge

8. Click Close to close the dialog.
9. Repeat the steps 1 to 8 for the Turn Off Climate Control and Stop Climate Control actions.

When you finish, your diagram should look the same as in the figure below.

 You may notice that pins display their types only. For this, do the following:
1. Press down the Alt key and select any input pin. All input pins are selected on the diagram.
2. Right-click the selection, click Show Name to clear the selection, and then click to select the
Show Type check box.

3. Repeat the steps above to hide the output pin types.

MagicGrid® Book of Knowledge | 59

To realize item flows conveying the Desired Temperature signal between the actions allocated to the :Vehicle
Occupant and the :VCCU part properties

1. Select the Set Temperature action and click the Object Flow button on its smart manipulator
toolbar.
2. Click the Reach Desired Temperature action. The object flow with pins for both actions is created.
3. In the Model Browser, under the 3 Exchange Items package, select the Desired Temperature signal.
4. Drag the signal to the output pin of the Set Temperature action. The Desired Temperature type is set as
the type for the output pin.
5. Drag the signal to the input pin of the Reach Desired Temperature actions. The Desired Temperature type
is set as the type for the input pin.
6. Select the object flow and click Item Flow Manager button on its smart manipulator toolbar.
7. In the open dialog, select the item flow that conveys the Desired Temperature signal. The signal
becomes the conveyed item of the object flow.
8. Click Close to close the dialog.
When you finish, your diagram should look the same as in the figure below.

 You may notice that pins don't display their types, and only item flows are shown on the diagram.
This makes the diagram less crowded and easier to read. If you want to get rid of pin types, do the
following:

1. Press down the Alt key and select any input pin. All input pins are selected on the diagram.
2. Right-click the selection and click Show Type to clear the selection.
3. Repeat the steps above to hide the output pin types.

60 | MagicGrid® Book of Knowledge

Use the same method to specify the following:
• The Ambient Temperature signal flowing between the Measure Temperature and Display Temperature
actions. Remember the flow between the :Temperature Sensor and :VCCU part properties of the
Vehicle In Use system context.
• The Air block being conveyed from the Provide Air action to the Reach Desired Temperature action.
Remember the flow between the :Cabin and :VCCU part properties of the Vehicle In Use system
context.
• The Comfortable Air block being conveyed from the Reach Desired Temperature action to the Get
Comfortable Air action. Remember the flow between the :VCCU and :Cabin part properties of the
Vehicle In Use system context.
• The Information signal flowing between the Display Temperature and Evaluate Comfort actions.
Remember the flow between the :VCCU and :Vehicle Occupant part properties of the Vehicle In Use
system context.

After you finish, your diagram should look similar to the one below. You may notice that all the item flows
here are the same as in the Vehicle In Use ibd. Always remember that a precise and complete model
requires item flow synchronization between system structure and behavior. For this reason, all exchange
items identified in the structural model must be considered and included in the behavioral model as well.

Now you can initiate and run the simulation session to execute the model and validate the specified
behavior.

Step 8. Supplementing the use case scenario with a parallel flow

The Provide Comfortable Temperature scenario still needs a few updates to be fully completed. We also need
to specify that the Ambient Temperature signal is obtained by the Reach Desired Temperature action.
Therefore, we have to split the flow going out of the Measure Temperature action into two. For this, the fork
node can be utilized.

MagicGrid® Book of Knowledge | 61

To split the object flow going out of the Measure Temperature action into two

1. Click the small black triangle next to the Fork Horizontal button on the diagram palette and
then select Fork Vertical.
2. Click the object flow between the Measure Temperature and Display Temperature actions. The vertical
fork node is inserted.
3. Click the vertical fork node and then click the Object Flow button on its smart manipulator toolbar.
Select the Reach Desired Temperature action. The object flow and input pin for that action are created.
The input pin is automatically typed by the Ambient Temperature signal. You do not need to change
anything.
4. Select the object flow and click the Item Flow Manager button from its smart manipulator toolbar.
5. In the open dialog, click to select the single item flow that conveys the Ambient Temperature signal.
The signal becomes the conveyed item of the object flow.
6. Click Close to close the dialog.

The use case scenario can now be considered as complete. You can initiate and run the simulation session
again, to execute the model and validate the specified behavior.

Use Cases done. What's next?

Having created one or more use case scenarios, you are ready for functional analysis. Steps that are
allocated to the SoI (in SysML terms, they appear in the swimlane that represents the SoI) can
be decomposed to specify the internal behavior expected from the SoI (see Chapter Functional Analysis).

62 | MagicGrid® Book of Knowledge

Measures of Effectiveness
What is it?

In this cell, non-functional stakeholder needs are refined with measures of effectiveness (MoEs), which
capture quantifiable characteristics of the SoI in a numerical format. MoEs is a traditional term widely used
in systems engineering, describing how well a system carries out a task within a specific context. In the
problem domain model, they serve as high level key performance indicators that would be automatically
checked within the solution domain model: with MoEs, you can specify design constraints that can be
used to verify the system design.

The same set of MoEs can be used in different models to specify numerical characteristics of other
systems of interest.

Who is responsible?

MoEs can be identified by the Systems Analyst or Systems Engineer.

How to model?

To capture MoEs in the modeling tool, a SysML block definition diagram (bdd) can be used. To define a
reusable set of MoEs, a separate block should be created and set as a super-type of the block that stands
for the SoI. In real-world projects, blocks with reusable sets of MoEs can be stored in external models, also
known as libraries. MoEs are captured as value properties with «moe» stereotypes applied. The SoI block
inherits them from the super-type block. A mechanism of redefinition in the modeling tool allows you to
define different default values and refine different requirements by every single MoE.

Tutorial
• Step 1. Organizing the model for MoEs
• Step 2. Creating a block for capturing MoEs
• Step 3. Capturing MoEs for the SoI

MagicGrid® Book of Knowledge | 63

Step 1. Organizing the model for MoEs
Following the structure of the MagicGrid framework, model elements that capture the measures of
effectiveness should be stored in a separate package inside the 1 Black Box package. We recommend
naming the package after the cell: 4 Measures of Effectiveness.

To organize the model for measures of effectiveness

1. Right-click the 1 Black Box package and select Create Element.

2. In the search box, type pa (the first two letters of the element type Package) and press Enter.
3. Type 4 Measures of Effectiveness to specify the name of the new package and press Enter.

Step 2. Creating a block for capturing MoEs

To capture MoEs of the Vehicle Climate Control Unit, you first need to create a bdd and a block to set as a
super-type of the VCCU block. Remember that if it was a real-world project, the MoEs holder could be
created in an external model, also known as a library.

To create a diagram and a block for capturing MoEs

1. Right-click the 4 Measures of Effectiveness package and select Create Diagram.

2. In the search box, type bdd (the acronym of the SysML block definition diagram) and then double-
click Enter. The diagram is created.

 Note that the diagram is named after the package where is it stored. This name also suits the
diagram. You need only remove the sequence number from its name.

3. Click the Block button on the diagram palette and then click the empty space in that diagram
pane. An unnamed block is created.
4. Type MoEs Holder to specify the name of this block and press Enter.

64 | MagicGrid® Book of Knowledge

 5. In the Model Browser, select the VCCU block, using the quick find capability:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type VC.
c. When you see the VCCU block selected in the search results list (see the following figure), press
Enter. The VCCU block is selected in the Model Browser.

6. Drag the VCCU block to the diagram pane. The shape of the VCCU block appears on the diagram.
7. Click the Generalization button on the smart manipulator toolbar of the MoEs Holder block and
then click the shape of the VCCU block. The MoEs Holder block becomes the super-type of
the VCCU block.

Step 3. Capturing MoEs for the SoI

Now we will analyze non-functional stakeholder needs SN-1.1.1, SN-1.1.4, and SN-1.2.1. They prompt
you to create the MoEs for the sound level, total mass, and total energy consumption. Let’s capture these
MoEs in your model.

MagicGrid® Book of Knowledge | 65

To capture MoEs of the Vehicle Climate Control Unit

1. Select the shape of the MoEs Holder block and click the Create Element button > Value Property.
2. Type soundLevel to specify the name of the new value property and press Enter.

 You can type "/" in front of the property name, for example, /soundLevel, to indicate that it is
derivable. Otherwise, to set the Is Derived property value to true, you need to right-click the
property and then select the Is Derived check box.

3. Right-click the value property and select Stereotype.

4. In the search box, type moe and press Ctrl + Spacebar to select the «moe» stereotype.
5. Press Enter. The «moe» stereotype is applied to the soundLevel value property.
6. Repeat the steps above to capture the totalMass and totalEnergyConsumption MoEs.

MoEs may have units, which can be captured in the model as value types. The majority of the commonly
used value types are stored in the ISO 80000 library of units, which is provided by the modeling tool.

To set a standard unit (from the library) as MoE type

1. Select any of the value properties and click ISO on the smart manipulator toolbar. Wait until the library
of units is loaded.
2. Select the totalMass value property and click the Specify Type button on the smart manipulator
toolbar.
3. Type kilog to find the mass[kilogram] value type and select it. The type is specified.

66 | MagicGrid® Book of Knowledge

  The list of possible value types can be narrowed down. For this, click to select the Apply Filter
check box and then make sure the Filter By Package Imports check box (under the Filter
Options button ) is selected.

Be default, value types are filtered to show basic units only. However, the list may not include
the required value types, and you may need to extend it. For this, the Basic Units matrix should
be updated. For more information, refer to the latest documentation of the modeling tool.

4. Repeat steps 2 and 3 to find work[megajoules] and set it as the type for the totalEnergyConsumption
value property.

Custom units can be created manually.

To set a custom unit as MoE type

1. Select the soundLevel value property and click the Specify Type button on the smart manipulator
toolbar.
2. Type dBA and press Enter. The custom value type is created in the model, directly under the package 4
Measures of Effectiveness, and specified as the value type of the soundLevel value property.

If you want to display MoEs on the shape of the VCCU block (see the following figure), you need to update
the symbol properties of that block. For this, right-click the VCCU block, select Symbol Properties, and
then in the open dialog, find the Show Inherited property (make sure the All mode is switched on in the
dialog, or you will not be able to find it). Once you do this, set the property value to true.

MagicGrid® Book of Knowledge | 67

  The set of MoEs captured as the MoEs Holder block can later be used in different models to assign
the same quantifiable characteristics to other systems of interest.

If you want to make changes to the inherited MoEs of the SoI, you need to redefine them. Let's say you
want to change their names to make them more specific.

To redefine the inherited value properties

1. Double-click the VCCU block to open its Specification.

2. On the left side of the dialog, click Properties.
3. On the right side, select the totalEnergyConsumption value property and click the Redefine button
below.

68 | MagicGrid® Book of Knowledge

 4. In the Name cell, type the new name of the property (for example, totalEnergyConsVCCU).
5. Press Enter to confirm the changes.
6. Repeat steps 3 to 5 to redefine other MoEs, as shown in the following figure.

MoEs done. What's next?

Now that you have measures of effectiveness, you can specify design constraints, which can be used to
verify whether the system design is correct or not (see Chapter System Parameters).

MagicGrid® Book of Knowledge | 69

White-box perspective
Once you have finished the problem definition from the black-box perspective (i.e., the operational analysis
of the SoI is completed), you can switch to analyzing the SoI from the white-box perspective. This helps to
understand the expected behavior of the SoI before producing its logical solution architecture.

In the second phase of the problem domain definition, you should perform deeper analysis of system
functions in order to identify the conceptual subsystems (referred to as functional blocks in some
methodologies) of the SoI. Conceptual subsystems are the first step towards the logical solution
architecture of the SoI.

Deeper analysis of system functions is called functional analysis. As the following figure shows, every
function performed by the SoI, as identified in the Use Case cell (layer 2), is decomposed in the Functional
Analysis cell (layer 3). Functions of the detailed behavior are then allocated to conceptual subsystems (aka
functional blocks) to convey that each conceptual subsystem (aka functional block) is responsible for
performing one or more functions.

The functional decomposition process can have as many iterations as needed to achieve the relevant
granularity of the problem domain definition. The following figure shows that functions at each layer of
detail can be decomposed into even more detailed behavior. Conceptual subsystems are accordingly
decomposed into more elementary structures. It is important to note that the granularity of system
behavior and structure must be consistent in each level of detail.

After functional subsystems are identified, interactions between them can be specified. Every subsystem
can have its own quantitative characteristics (i.e., measures of effectiveness). The functions of the SoI, its

70 | MagicGrid® Book of Knowledge

functional subsystems, and MoEs together establish the basis for specifying system requirements (see
Chapter System Requirements).

Read the following pages to learn how to create the problem domain from the white-box perspective.

Functional Analysis
What is it?

In this cell, you should more deeply analyze the expected system behavior by decomposing every function
identified during the black-box analysis (see Chapter Use Cases) and performed by the SoI.

Decomposing each function into more detailed system behavior gives stimulus to identify the conceptual
subsystems (aka functional blocks) that are responsible for performing these functions. Conceptual
subsystems are captured in the next cell of the white-box analysis (see Chapter Conceptual Subsystems).

It is important to note that with deeper analysis, the subfunctions of a decomposed function can be
decomposed as well. Decomposition can be performed as many times as you need to achieve the relevant
granularity of the problem you need to address. Every decomposed subfunction is regarded as a function
from the standpoint of its subfunctions.

The result of the Functional Analysis cell is the model of the expected system behavior. An activity
decomposition map starting from the high-level goals of the SoI and ending with the expected functions
of it can be extracted from that model.

Who is responsible?

As a continuation of the use case analysis, functional analysis can be performed by the Systems Analyst
or Systems Engineer.

MagicGrid® Book of Knowledge | 71

How to model?

Functional analysis is a continuity of use case scenario refinements using SysML activity diagrams. A new
SysML activity diagram should be created for every function allocated to SoI (see Chapter Use Cases).
Though there are two or even more swimlane partitions, you should choose only those functions that are
nested under the partition representing the usage of block that captures your SoI. The following figure
shows the white-box scenario of the Reach Desired Temperature function (which is a part of the Provide
Comfortable Temperature use case scenario) performed by the VCCU SoI.

72 | MagicGrid® Book of Knowledge

Tutorial
• Step 1. Organizing the model for functional analysis
• Step 2. Creating an activity diagram to decompose a function
• Step 3. Specifying the white-box scenario

Step 1. Organizing the model for functional analysis

According to the design of the MagicGrid framework, model artifacts that capture decomposed system
functions should be stored under the structure of packages, as displayed in the following figure. As you
can see, the top-level package represents the domain, the medium-level package represents the
perspective, and the bottom-level package represents the cell.

To organize the model for the functional analysis

1. Right-click the 1 Problem Domain package and select Create Element.

2. In the search box, type pa, the first two letters of the element type Package, and press Enter.
3. Type 2 White Box to specify the name of the new package and press Enter.
4. Right-click the 2 White Box package and select Create Element.
5. Repeat steps 2 and 3 to create the package named 3 Functional Analysis.

Step 2. Creating an activity diagram to decompose a function

Let’s say you want to decompose the function captured in your model as the Reach Desired
Temperature activity. You will need to create a SysML activity diagram inside that activity.

Additionally, you need to change the location of the Reach Desired Temperature activity in your model by
moving it to the 3 Functional Analysis package, a sub-package of the 2 White Box package (created in the
previous step of this cell tutorial).

MagicGrid® Book of Knowledge | 73

To create a SysML activity diagram inside the Reach Desired Temperature activity

1. Open the Provide Comfortable Temperature activity diagram:

a. Press Ctrl + Alt + F to open the Quick Find dialog.

b. Click Diagram to search for diagrams only and type Prov.
c. When you see the Provide Comfortable Temperature activity diagram selected in the search results
list (see the following figure), press Enter. The diagram is opened.

2. Select the shape of the Reach Desired Temperature action and click the SysML Activity Diagram button
on its smart manipulator toolbar (see the following figure). A newly created activity diagram
opens.

 If you go back to the Provide Comfortable Temperature activity diagram (simply by clicking
on the open diagram toolbar), you can see that the shape of the Reach Desired Temperature
action is decorated with the rake ( ) icon, indicating that the action contains an internal
diagram, which opens by double-clicking the shape of that action.

74 | MagicGrid® Book of Knowledge

 3. In the Model Browser, select the Reach Desired Temperature activity:

a. Right-click the shape of the Reach Desired Temperature action on the diagram pane.
b. Select Go To > Behavior Reach Desired Temperature. The activity is selected in the Model
Browser.

4. Drag the Reach Desired Temperature activity to the package 3 Functional Analysis (within the 2 White
Box package). The related activity diagram is moved in together.

When you finish the white-box analysis in a real-world project, all the activities that capture the functions
of the SoI (they appear within the :VCCU swimlane in the Provide Comfortable Temperature activity diagram)
must be stored in the 3 Functional Analysis package.

MagicGrid® Book of Knowledge | 75

Step 3. Specifying the white-box scenario
Now we will specify the white-box scenario of the function captured as the Reach Desired
Temperature activity.

The activity diagram has a few inputs and an output. In terms of SysML, these are activity parameter
nodes and they correspond the input and output pins of the Reach Desired Temperature action in the Provide
Comfortable Temperature activity diagram. Inputs and outputs are a good starting point and should be
considered when specifying the scenario. Additionally, you can remove their names to remove
unnecessary information and have more space on the diagram pane.

To remove the name of the activity parameter node

1. Click the node once. When it is selected, click its name. The name is selected.
2. Press the Delete button on the keyboard.

This time before starting to specify the scenario, you don't need to create any swimlanes in the diagram.
You can do this later, as soon as you have conceptual subsystems (aka functional blocks) captured in your
model.

To specify the white-box scenario of the Reach Required Temperature activity

1. Be sure the Behavior Creation Mode is enabled in the diagram. If it is not enabled, activities will not be
created and assigned as behaviors to call behavior actions. For this, click the Behavior Creation Mode
button .
2. Click the Initial Node button on the diagram palette and then click an empty place on the diagram
pane. The initial node is created.
3. Click the Control Flow button on the smart manipulator of the initial node shape, and then click an
empty place on the diagram pane. An unnamed action with a new activity assigned as its behavior is
created.
4. Type Read Desired Temperature and press Enter.

 Be sure you type the name after the colon (":"), or the name will be given to the action, but not
to the referred activity.

5. Select the :Desired Temperature node and click on its smart manipulator toolbar.

76 | MagicGrid® Book of Knowledge

6. Click the Read Desired Temperature action. The object flow is created, as well as the input pin for the
action. The type of that pin is automatically set by the type of the input parameter node.

 Remember that to hide the input pin name and show its type, you need to right-click that pin,
click the Show Name check-box to clear the selection, and then click to select the Show
Type check box.

7. Capture other items to build the complete scenario, which is displayed in the following figure using
the following guidelines:

• To capture new actions, click the Control Flow or Object Flow on the smart manipulator
toolbar. When you create the object flow, do not forget to right-click an empty place in the
diagram pane and then select Call Behavior Action (otherwise the Central Buffer Node may
appear on the diagram).
• To connect two actions, click the Control Flow or Object Flow on the smart manipulator
toolbar of the source action, then click the target action.
• To set the type for a pin (unless it is not automatically set), drag the appropriate element (block or
signal) from the Model Browser.
• To create the decision or merge node, right-click an empty place in the diagram pane and then
select Decision or Merge.
• To specify a guard, select the relevant control flow, type the square bracket, and then specify the
condition, such as [Desired Temperature < Ambient Temperature. Keep in mind that you do not need
to close the brackets manually. This is done automatically once you press Enter.
• To create an output pin and an object flow to the output parameter node, click the Object Flow
on the smart manipulator toolbar of that action and then click the output parameter node you
want to relate. The output pin is created in tandem and its type is set automatically by the type
of the related output parameter node.
• To change the type of the element you want to create at the arrow end of the new control flow
(for example, from action to decision or from decision to merge), right-click an empty place in the
diagram pane and select that type.

MagicGrid® Book of Knowledge | 77

Note that there are a few pins that are typed by the CU Command element, where CU stands for Computing
Unit. This is actually a signal. It can be stored in the 3 Exchange Items package, together with other signals
and blocks that capture exchange items.

You can run the simulation to validate the specified behavior and see if your model is correct. To start the
simulation session, click the Run button on the Reach Desired Temperature activity diagram toolbar.
When in the Simulation panel, click the Start button to start the model execution.
This scenario enables us to presume that the Vehicle Climate Control Unit may consist of four subsystems:
one for receiving data from outside the system, one for cooling the air, one for heating it, and one for
controlling other systems. The next cell of the MagicGrid framework describes how to capture them.

Step 4. Creating an activity decomposition map

Now we will create an activity decomposition map for the Provide Comfortable Temperature activity and see
which black- and white-box functions it was decomposed to. The activity decomposition map is one of the
predefined relation maps available in the modeling tool.

78 | MagicGrid® Book of Knowledge

To create the activity decomposition map for the Provide Comfortable Temperature activity

1. Select the activity in the Model Browser:

a. Press Ctrl + Alt + F.

b. In the Quick Find dialog:

i. Make sure the Type option is selected.

ii. Type Provide C.

c. When you see the Provide Comfortable Temperature activity in the search results list, select it and
press Enter. The activity is selected in the Model Browser.

2. Right-click the activity and select Create Diagram.

3. In the search box, type adm, the acronym of the predefined Activity Decomposition Map, and then
press Enter. The diagram is created.

 If there are no results in the diagram types list, click the Expert button below it. The list of
available diagrams expands in the Expert mode.

4. Press Enter again.

 Note that the diagram is named after the enclosing activity. This name suits the diagram as
well, so no changes are needed.

 If only the fragments of long element names are displayed, you can change the map settings
to display the complete element names. For this, click the Options button on the relation
map toolbar and click to unselect Cut Element Names.

Functional Analysis done. What's next?

White-box scenarios stimulate identification of conceptual subsystems of the SoI; thus, the next cell is
Conceptual Subsystems.

MagicGrid® Book of Knowledge | 79

Conceptual Subsystems
What is it?

While functional analysis (see Chapter Functional Analysis) helps identify conceptual subsystems (aka
functional blocks), this cell captures them in the model. The conceptual subsystems should be considered
as a group of interconnected and interactive parts that are responsible for performing one or more
expected functions of the SoI.

It’s important to note that every conceptual subsystem can be decomposed into a more elementary
structure. The number of iterations of such decomposition depends on the granularity level of the
conceptual architecture you want to address. Every subsystem is regarded as a system from the standpoint
of its internal parts.

Once the conceptual subsystems are captured in the model, it is time to specify which system functions
they perform. For this, functions are allocated to conceptual subsystems. Remember that the granularity
of expected system behavior and structure must be consistent in each level of detail; thus, you cannot
assign the function of the subsystem to its component or function of that component to the subsystem.
Deeper analysis is only possible after you have completely finished the current level of granularity. The
recommended correspondence between the granularity levels of behavioral and structural decomposition
is displayed in Chapter White-box perspective.

Inputs and outputs of the SoI are also considered in this cell. They are identified by analyzing the system
context and use case scenarios, while others are determined considering the results of the functional
analysis.

Overall, this cell produces:

Inputs and outputs of the SoI
Definitions of conceptual subsystems of the SoI
Interactions:
• Between conceptual subsystems and the outside of the SoI
• Among conceptual subsystems
Allocation of expected system functions to the conceptual structure of the SoI
Structure decomposition map

 It should be kept in mind that it is not always necessary to specify the conceptual architecture of
the SoI. For example, when MagicGrid is used to modernize an already existing system, the
Conceptual Subsystems cell can be skipped to define the logical architecture of the SOI
straightaway.

Who is responsible?

The structural decomposition of the system can be performed and its internal communications can be
specified by the Systems Analyst or Systems Engineer.

80 | MagicGrid® Book of Knowledge

How to model?

To capture inputs and outputs of the SoI or any of its conceptual subsystems, you can use the
infrastructure of the bdd. These inputs and outputs can be specified as SysML flow properties. As defined
by SysML, these can be grouped by interface blocks that are stored somewhere in the model or even in
some external library. Proxy ports are used to relate interface blocks to the blocks that capture the
structure of the SoI. The following figure shows the inputs and outputs of the Vehicle Climate Control Unit
and one of its conceptual subsystems, the Processing subsystem.

To define the conceptual subsystems of the SoI and the interactions between them, you can utilize the
infrastructure of the ibd created for the block that captures the SoI in your model. Conceptual subsystems,
identified while performing the functional analysis (see Chapter Functional Analysis), can be specified as
part properties of the block that represents the SoI. Connectors with one or more item flows between
these part properties indicate interactions between appropriate conceptual subsystems. Connectors with
item flows between part properties and the diagram frame represent interactions between conceptual
subsystems and the outside of the SoI. Connections should be established via proxy ports. The following
figure shows the conceptual subsystems of the Vehicle Climate Control Unit as well as their
communication with the outside and in between.

MagicGrid® Book of Knowledge | 81

After you have captured all the conceptual subsystems in the model, you can go back to the white-box
scenario and allocate the white-box functions of the SoI to its conceptual subsystems in order to convey
that each conceptual subsystem (aka functional block) is responsible for performing one or more white-
box functions. Conceptual subsystems can be represented in the activity diagram as swimlanes. The
following figure displays the subfunctions of the Reach Desired Temperature function, allocated to the
conceptual subsystems of the SoI.

82 | MagicGrid® Book of Knowledge

Tutorial
• Step 1. Organizing the model for conceptual subsystems communication
• Step 2. Creating a bdd for specifying conceptual interfaces
• Step 3. Capturing conceptual interfaces
• Step 4. Creating an ibd for capturing conceptual subsystems
• Step 5. Capturing conceptual subsystems
• Step 6. Specifying interactions between conceptual subsystems and the outside of the SoI
• Step 7. Specifying interactions among conceptual subsystems
• Step 8. Allocating functions by conceptual subsystems and synchronizing item flows

Step 1. Organizing the model for conceptual subsystems communication

Following the structure of the MagicGrid framework, model artifacts that capture conceptual subsystems
communication should be stored in a separate package inside the 2 White Box package. We recommend
naming the package 2 Conceptual Subsystems Communication.

To organize the model for the conceptual subsystems communication

1. Right-click the 2 White Box package and select Create Element.

2. In the search box, type pa (the first two letters of the element type Package) and press Enter.
3. Type 2 Conceptual Subsystems Communication to specify the name of the new package and press
Enter.

Step 2. Creating a bdd for specifying conceptual interfaces

To begin specifying conceptual interfaces, which capture inputs and outputs of the SoI, you first need to
create a bdd and display the block that represents the SoI on the diagram pane. Before this, however, you
should change the location of that block in your model by moving it to the package 2 Conceptual
Subsystems Communication. Since we've started the white-box analysis, it is no longer a black box.

MagicGrid® Book of Knowledge | 83

To create and prepare a bdd for specifying SoI interfaces

1. Create the bdd:

a. Right-click the 2 Conceptual Subsystems Communication package and select Create Diagram.
b. In the search box, type bdd, the acronym of the SysML block definition diagram, and then press
Enter. The diagram is created.
c. Type VCCU Interfaces to specify the name of the new diagram and press Enter again.

2. In the Model Browser, select the VCCU block using the quick find capability:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type VC.
c. When you see the VCCU block selected in the search results list (see the following figure), press
Enter. The VCCU block is selected in the Model Browser.

3. Drag the VCCU block to the newly created package 2 Conceptual Subsystems Communication (see step 1
of this cell tutorial). If a message appears, asking if you want to move block's relationships together,
click Yes.
4. Drag the VCCU block to the newly created diagram pane. The shape of the block appears on the
diagram.

84 | MagicGrid® Book of Knowledge

Step 3. Capturing conceptual interfaces
By analyzing the system context (see Chapter System Context), we can conclude the following:
• The Vehicle Climate Control Unit is controlled by the Vehicle Occupant (driver or passenger), who
can Turn On and Turn Off the system, as well as specify the Desired Temperature.
• The Vehicle Climate Control Unit takes the Air from the Cabin and provides back the Comfortable Air.
• The Temperature Sensor of the Vehicle provides the Ambient Temperature of the Cabin Vehicle
Climate Control Unit.
• The Vehicle Climate Control Unit displays the Information (for example, status information,
temperature in the cabin).

This information serves as input to specify conceptual interfaces of the SoI, for defining its communication
with the outside: supposed users and other entities of the Vehicle In Use system context. Conceptual
interfaces for internal communication will be considered and captured later, after we define conceptual
subsystems (also known as functional blocks) of the SoI.

Remember that inputs and outputs can be captured as flow properties. Flow properties with in direction
indicate inputs, and flow properties with out direction indicate outputs. Flow property names are not as
relevant as their types. Blocks and signals that capture exchange items of the Vehicle In Use system context
become flow property types.

A single flow property always belongs to some interface block. A proxy port is a usage of that interface
block on the block, which captures the SoI. Proxy port names are relevant, if more than one proxy port is
typed by the same interface block; otherwise, their names are not important.

To create the interface block for providing manual control and relate it to the VCCU block

1. Open the VCCU Interfaces diagram, if not opened yet.

2. Click the Interface Block button on the diagram palette and then click an empty place on the diagram
pane. An unnamed interface block is created in the model, directly under the 2 Conceptual Subsystems
Communication package, and displayed on the diagram.
3. Type Manual Control to name the interface block and press Enter.
4. Click the Create Property button on the interface block shape and select Flow Property.

5. Directly on the shape of the interface block, do the following:

a. Remove the out part of the flow property direction indicator to specify that the direction of the
flow property is in.
b. Type uc to specify the flow property name.
c. Type :Use and pres Ctrl + Spacebar to open the list of possible flow property types. Once you see
the User Command signal in that list, select it. The type is set.

MagicGrid® Book of Knowledge | 85

 6. Repeat steps 4 and 5 to create another flow property you see in the following figure.

7. Select the Manual Control interface block on the diagram and drag it onto the VCCU block. As a result,
a new proxy port, typed by the Manual Control interface block, is created for the VCCU block. As can
see, it is decorated with an arrow indicating that this port is for inputs to the system of interest.

 If you hover over the proxy port shape, the tooltip shows which inputs can be provided to the
system over this proxy port.

Following the procedure above, create the interface blocks with flow properties (shown in the following
figure) and then relate them to the VCCU block by creating proxy ports.

86 | MagicGrid® Book of Knowledge

  Once finished, the VCCU Interfaces diagram can be marked as complete. A complete diagram
excludes unwanted layout changes that may occur as your model evolves. Model updates that
concern the complete diagram (for example, new properties of the block that resize its shape and
make it overlap with other shapes) are suspended there until you allow them to be displayed.
To mark the VCCU Interfaces diagram as complete, right-click anywhere on the diagram pane and
then select Complete. The notification indicating that the diagram is suspended to display changes
in the model appears at the top of it. For more information about this feature, refer to the latest
documentation of the modeling tool.

The following figure displays the Blackbox ICD Table, an alternative view of the Vehicle Climate Control
Unit interfaces. This tutorial doesn’t explain how to build the table, but you can find this information in
the latest documentation of the modeling tool.

To keep the model well-structured, we recommend storing all the interface blocks in a separate package.
For this, you need to create the 1 Conceptual Interfaces package directly under the 2 Conceptual Subsystems
Communication package and then drag all the interface blocks to it.

Step 4. Creating an ibd for capturing conceptual subsystems

Conceptual subsystems of the Vehicle Climate Control Unit can be specified in an ibd, which is owned by
the VCCU block. According to SysML, the frame of this diagram represents the borders of the SoI and thus
allows you to specify interactions between the internal parts of the SoI and the outside of the system via
proxy ports displayed on that frame (see step 3 of this cell tutorial).

MagicGrid® Book of Knowledge | 87

To create an ibd for capturing conceptual subsystems of the SoI

1. On the VCCU Interfaces diagram, select the shape of the VCCU block and click the SysML Internal Block
Diagram button on its smart manipulator. A new ibd named after the enclosing block is created,
and the Display Parts/Ports dialog opens.
2. Make sure all the proxy ports of the VCCU block are selected and click OK. The proxy ports are
displayed on the newly created diagram frame.
3. Type VCCU Conceptual Subsystems Communication to specify the name of the new diagram and press
Enter again.

The final view of your diagram should be similar to the one in the following figure.

 While the VCCU Interfaces bdd (see step 3 of this cell tutorial) represents the SoI as a black box, the
idb is created to represent the same system as a white box: the VCCU block is opened to see its
internal structure. That’s why proxy ports displayed on the borders of the VCCU block in the bdd are
displayed on the diagram frame in the ibd.

88 | MagicGrid® Book of Knowledge

Step 5. Capturing conceptual subsystems
As the functional analysis has revealed (see Chapter Functional Analysis) the Vehicle Climate Control Unit
consists of these conceptual subsystems:
• Cooling
• Heating
• Processing
• Human-Machine Interacting

They can be specified as part properties of the VCCU block. Part properties don’t need names, but they
should be typed by blocks that capture the conceptual subsystems listed above.

To specify the conceptual subsystems

1. Open the VCCU Conceptual Subsystems Communication ibd, created in step 4 of this cell tutorial, if
not yet opened.
2. Make sure the Type Selection Mode is on in the diagram. Otherwise, parts created in this diagram will
not be typed by blocks.

3. Click the Part Property button on the diagram palette and then click an empty place on the diagram
pane. An unnamed part property is created, and the list of existing blocks to type it is offered.
4. Type Cooling next to the colon (":") directly on the part property shape and press Enter. The
Cooling block to type the just-created part property is created in the Model Browser.
5. Repeat steps 3 and 4 to create another conceptual subsystem from the list above.

When you’re done, your VCCU Conceptual Subsystems Communication ibd should look very similar to the
one in the following figure.

MagicGrid® Book of Knowledge | 89

To keep the model well-structured, we recommend storing all the blocks that capture conceptual
subsystems in a separate package. For this, you need to create the 2 Conceptual Subsystems package
directly under the 2 Conceptual Subsystems Communication package and then drag all the blocks to it.

 If the functional analysis requires decomposing one or more subsystems of the SoI, we highly
recommend using the same model structure pattern for each subsystem as you used for the SoI.
Repeating the structure created at this level of detail for every lower level helps to establish a good
and easily readable recursive structure of the model.

Step 6. Specifying interactions between conceptual subsystems and the outside

of the SoI
In this step, we focus on specifying interactions between conceptual subsystems (aka functional blocks) of
the Vehicle Climate Control Unit and outside it. An interaction between the subsystem and the outside of
the SoI can be specified as a connector with one or more item flows, linking the appropriate part property
and the diagram frame (remember, the diagram frame in the ibd represents the borders of the SoI). In this
cell, unlike the System Context cell (see Chapter System Context), connectors are established over
compatible proxy ports. Information, matter, or energy exchanged over these connectors can be specified
as items conveyed by item flows.

90 | MagicGrid® Book of Knowledge

Let's start with specifying that the Cooling subsystem takes the Air from the outside of the SoI and
provides Comfortable Air back.

To specify the interactions between the Cooling subsystem and the outside of the SoI

1. Select the proxy port typed by the Cabin Air Inlet interface block and click Connector button on its
smart manipulator toolbar.

2. Click the :Cooling part property and select New Proxy Port (see the following figure). A new
compatible proxy port is created. As shown below, it is typed by the same interface block.

3. In the Model Browser, select the Air block. For this, use the quick find capability:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type Air.
c. When you see the Air block selected in the search results list (see the following figure), press
Enter. The Air block is selected in the Model Browser.

4. Drag the Air block to the newly created connector on the open diagram pane.

MagicGrid® Book of Knowledge | 91

  As defined by SysML, the element you want to specify as the item flowing over the connector
must be the type (or sub-type) of the flow property that determines the input of the part
property on one connector end and the output of the part property on another, and vice versa.
In this case, the Air block is the flow property type and can be set as the item flowing over the
relevant connector.

5. Don’t make any changes in the open dialog; click Finish. The Air block is specified as the item flowing
over the newly created connector.

6. Repeat steps 1 and 2 to create a connector between the proxy port typed by the Cabin Air Outlet
interface block and the :Cooling part property.
7. Repeat step 3 to select the Comfortable Air block in the Model Browser.
8. Drag the block to the newly created connector.
9. In the open dialog, change the direction of the flow to From Cooling System To VCCU and click
Finish. The Comfortable Air block is specified as the item flowing over the newly created connector.

92 | MagicGrid® Book of Knowledge

Interactions between the Heating subsystem and the outside are very similar. They can be specified in an
appropriate way. Once you're done, your diagram should look like this.

 As you can see, the proxy port names are hidden in this diagram. This greatly increases the
readability of this diagram. To hide the names of proxy ports, do the following:
1. Press Alt and hold it down while you select any proxy port; all proxy ports in that diagram are
selected.
2. Right-click the selection and then click to clear the Show Name check box.

Now specify that the Processing subsystem takes the Ambient Temperature from the outside, and the
Handling Human-Machine Interactions subsystem accepts the User Commands and Desired Temperature
from there. The latter provides various Information to the outside, too.

After you specify this by following the instructions above, your VCCU Conceptual Subsystems
Communication diagram should look very similar to the one below.

MagicGrid® Book of Knowledge | 93

Step 7. Specifying interactions among conceptual subsystems
In this step, we will specify interactions among conceptual subsystems of the SoI. As in the previous step,
these interactions can be specified as connectors established between appropriate part properties over
compatible proxy ports. Information, matter, or energy exchanged over these connectors can be specified
as items conveyed by item flows.

The functional analysis enables us to identify the following interactions:

• The Processing subsystem takes the Desired Temperature from the Human-Machine Interacting
subsystem.
• If the Processing subsystem identifies that the Desired Temperature is lower than the Ambient
Temperature, it initiates the Cooling subsystem.
• If the Processing subsystem identifies that the Desired Temperature is higher than the Ambient
Temperature, it initiates the Heating subsystem.
• The Human-Machine Interacting subsystem takes various Information produced by the Processing
subsystem to provide it to the outside.

The list above requires you to capture a few more interface blocks. You can see them highlighted in the
following figure, which displays the Processing Subsystem Interfaces diagram.

94 | MagicGrid® Book of Knowledge

We assume that you create this diagram and rearrange the model as it is displayed in the following figure,
on your own.

As you can see, the CU Command signal has two sub-types: the Cool and Heat signal. With this hierarchy of
signals, you can specify exactly which signal is flowing through the same conceptual interface in each
particular case. It is the Cool signal between the :Processing and the :Cooling part properties, and the Heat
signal between the :Processing and the :Heating part properties.

The Information signal has sub-types, too. These are the Ambient Temperature, Desired Temperature, and
Status signals. We need them if we want to be more specific in specifying which kind of information can
go through the proxy ports typed by the Output Information or Internal Data Transfer interface blocks.

Sub-types can be specified in a separate bdd, the Exchange Items diagram. You already know how to create
a new bdd, display existing and create new elements there, and establish generalization
relationships between them. Therefore, we assume that you do this on your own. To display the contents
of the Exchange Items diagram within the Processing Subsystem Interfaces diagram, you need to drag
the Exchange Items diagram from the Model Browser onto the pane of the Processing Subsystem Interfaces
diagram. Then, on the smart manipulator toolbar of the Exchange Items diagram shape, click the Show
Diagram Overview Content button .
Now let's move on and specify the interactions among conceptual subsystems of the Vehicle Climate
Control Unit.

MagicGrid® Book of Knowledge | 95

To specify interactions between the Handling Human-Machine Interactions and the Processing
subsystems

1. Open the VCCU Conceptual Subsystems Communication ibd, if not yet opened.
2. Select the :Processing part property and click Display All Ports button on its smart manipulator
toolbar.

Two more proxy ports are displayed.

3. Select the proxy port typed by the Internal Data Transfer interface block and click the Connector
button , its smart manipulator toolbar.
4. Click the :Human-Machine Interacting part property and select New Proxy Port. A new compatible
proxy port is created. As you can see, it is typed by the same interface block.
5. In the Model Browser, under the 3 Exchange Items package, select the Desired Temperature signal and
drag it to the newly created connector.
6. In the open dialog, change the direction of the flow to From Human-Machine Interacting to
Processing and click Finish. The Desired Temperature signal is specified as the item flowing over the
newly created connector.
7. In the Model Browser, under the 3 Exchange Items package, select the Ambient Temperature signal and
drag it to the newly created connector.
8. Don't change anything in the open dialog and click Finish. The Ambient Temperature signal is specified
as the item flowing over the newly created connector.
9. Repeat steps 7 and 8, to specify the Status signal as the item flowing over the newly created
connector.

96 | MagicGrid® Book of Knowledge

In the same manner, specify the interactions between the Processing subsystem and the Heating
subsystem as well as between the former and the Cooling subsystem. Once you're done, your diagram
should look very similar to the one below.

To be more precise, you can also specify that the Comfortable Air flowing out of the Cooling subsystem is
Cooled Air, and the one flowing out of the Heating subsystem is Heated Air. Appropriate blocks should be

MagicGrid® Book of Knowledge | 97

created as sub-types of the Comfortable Air block in the Exchange Items diagram: the Cooled Air block and
the Heated Air block.

Since you have two sub-types of the Comfortable Air block, you can update the items flowing over the
connectors between the :Cooling part property and the diagram frame as well as between the :Heating part
property and the diagram frame. You already know how to do this. As you no longer need the Comfortable
Air block flowing over the connector, you can simply get rid of it.

To remove the item flow

1. Select the relevant connector and click the Item Flow Manager button on its smart manipulator
toolbar.
2. In the open dialog, click to clear the selection of the row where the conveyed item is the Comfortable
Air block.

3. Close the dialog.

98 | MagicGrid® Book of Knowledge

Once you're done, your ibd should look like the one below.

 The following figure displays an alternate view of the VCCU Conceptual Subsystems
Communication ibd. This tutorial doesn’t explain how to build the Whitebox ICD Table, but
you can find this information in the latest documentation of the modeling tool.

Step 8. Allocating functions by conceptual subsystems and synchronizing

item flows
Now it’s time to update the Reach Desired Temperature diagram (see Chapter Functional Analysis) with
swimlanes, which represent conceptual subsystems of the Vehicle Climate Control Unit. In sequel, actions
should be allocated to these swimlanes in order to specify which conceptual subsystems are responsible
for performing each function.

These allocations enable you to update the Reach Desired Temperature diagram with item flows by
synchronizing them from the relevant ibd. This time, we will synchronize from the VCCU Conceptual
Subsystems Communication diagram.

MagicGrid® Book of Knowledge | 99

To allocate functions to conceptual subsystems

1. Open the Reach Desired Temperature diagram:

a. Press Ctrl + Alt + F to open the Quick Find dialog.

b. Click Diagram to search for diagrams only and type Rea.
c. When you see the Reach Desired Temperature diagram selected in the search results list (see the
following figure), press Enter. The diagram is opened.

2. Click the Vertical Swimlanes button on the activity diagram palette and then click an empty space on
that diagram pane. Swimlanes with blank headers are created on the diagram.
3. Right-click any of the swimlanes and select Insert Swimlane > Vertical to the Left. A new swimlane
with a blank header is created.
4. Repeat step 3 to create another swimlane.
5. In the Model Browser, select the VCCU block. For this, use the quick find capability:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Click Type to search for types only and type VCCU.
c. When you see the VCCU block selected in the search results list, press Enter. The VCCU block is
selected in the Model Browser.

6. Expand the VCCU block to see its properties.

7. Select the :Human-Machine Interacting part property and drag it onto the header of the first swimlane.
The swimlane becomes the representation of that part property.
8. Repeat step 7 for the :Cooling, :Processing, and :Heating part properties in succession. Once you're
done, your swimlanes should look like the following figure.

9. Drag the initial node as well as the Read Desired Temperature and Send Data to Computing Unit actions
to the :Human-Machine Interacting swimlane.
10. Drag the Get Desired Temperature, Get Ambient Temperature, Process Data, Initiate Cooling, and Initiate
Heating actions, as well as decision, merge, and activity final nodes to the :Processing swimlane.
11. Drag one of the Draw Air Out from Cabin actions as well as the Cool Down Temperature to
Desired and Blow Cooled Air Back to Cabin actions to the :Cooling swimlane.

100 | MagicGrid® Book of Knowledge

12. Drag another Draw Air Out from Cabin action as well as the Heat Up Temperature to Desired and Blow
Heated Air Back to Cabin actions to the :Heating partition.
When you’re done, your Reach Desired Temperature diagram should look very similar to the one below.

Once all the actions are allocated, it's time to specify item flows among them. Item flows that cross the
swimlane boundary can be easily synchronized from the appropriate ibd (in this case the VCCU Conceptual
Subsystems Communication diagram). Let's start from the item flow between the Send Data to Computing
Unit and Get Desired Temperature actions. The former is allocated to the :Human-Machine Interacting part
property and the latter is allocated to the :Processing part property. If you look at the VCCU Conceptual
Subsystems Communication diagram, you can see that the Desired Temperature signal is flowing between
those part properties. Therefore, you can conclude that the same signal is flowing between the actions
they allocate. The modeling tool helps you specify the item flows without switching between diagrams,
as outlined below.

MagicGrid® Book of Knowledge | 101

To synchronize the item flow between the Send Data to Computing Unit and Get Desired Temperature actions
from the relevant ibd

1. Select the object flow between these actions.

2. Click the Item Flow Manager button on its smart manipulator toolbar.
3. In the open dialog, click to select the only item flow in the list. It conveys the Desired Temperature
signal.

 The list always includes the item flows that are possible between two actions. These item
flows are taken from the relevant ibd.

4. Close the dialog. As a result, the item flow conveying Desired Temperature signal displays in the
diagram.

In the same manner, specify that:

• The Cool signal is flowing from the Initialize Cooling action to the Draw Air Out from Cabin action.
• The Heat signal is flowing from the Initialize Heating action to the Draw Air Out from Cabin action.

102 | MagicGrid® Book of Knowledge

After you're done, your diagram should look very similar to the one below.

Next you can specify the item flows within the same swimlane, starting from the item flow between the
Read Desired Temperature and Send Data to Computing Unit actions.

To specify the item flow conveying the Desired Temperature signal between the Read Desired Temperature
and Send Data to Computing Unit actions

1. Select the object flow between these actions.

2. Click the New Item Flow button on its smart manipulator toolbar.
3. In the open dialog, make sure that the Desired Temperature signal is selected as the conveyed item and
click Finish (see the following figure). As a result, the Desired Temperature signal becomes conveyed by
the item flow between the relevant actions.

MagicGrid® Book of Knowledge | 103

Following the previous procedure, specify the rest of the item flows you see in the figure below. Keep in
mind that you can be more specific when specifying the item flow between the Cool Down Temperature to
Desired and Blow Cooled Air Back to Cabin actions. For this, you should select the Cooled Air block instead of
the Comfortable Air block in the Create / Edit Item Flow dialog. Remember that you can also be more
specific when specifying the item flow between the Heat Up Temperature to Desired and Blow Heated Air
Back to Cabin actions.

104 | MagicGrid® Book of Knowledge

  Pins don't display their types; only item flows are shown on the diagram. This makes the diagram
less overcrowded and easier to read. If you want to get rid of pin types, do the following:
1. Press the Alt key and select any input pin. All input pins are selected on the diagram.
2. Right-click the selection and click Show Type to clear the selection.
3. Repeat the steps above to hide the output pin types.

Although more detailed than the use case scenario (see Chapter Use Cases), this one is still a high-level
scenario and may require decomposition as well. Further structural decomposition stimulated by the
functional decomposition is performed in parallel. The functional and structural breakdown may have as
many iterations as you need to understand the problem domain.

Conceptual Subsystems done. What's next?

• System functions and its conceptual subsystems, with or without quantifiable characteristics of the
SoI, refine stakeholder needs. Therefore, once you have finished it, you can establish the
traceability relationships to stakeholder needs (see Chapter Traceability to Stakeholder Needs).
• You might also need to specify quantifiable characteristics for one or more conceptual subsystems,
which would be more specific than those defined for the whole SoI (see Chapter Measures of
Effectiveness). In that case, you should move to the MoEs for Subsystems cell.

MagicGrid® Book of Knowledge | 105

MoEs for Subsystems
What is it?

In this cell, you should specify the MoEs for one or more conceptual subsystems (aka functional blocks) of
the SoI to make further refinements of non-functional stakeholder needs. This cell is optional, as you
might not need to specify MoEs in addition to those defined for the entire SoI (see Chapter Measures of
Effectiveness).

Who is responsible?

MoEs for conceptual subsystems can be identified by the Systems Analyst or Systems Engineer.

How to model?

To capture MoEs for conceptual subsystems in the modeling tool, you can utilize the infrastructure of the
SysML bdd – the same way you used it for capturing MoEs of the SoI in the Measures of
Effectiveness cell. Just like in that cell, measures of effectiveness for a conceptual subsystem can be
captured in the model as value properties of that subsystem, with the «moe» stereotype applied.

Tutorial
• Step 1. Organizing the model for MoEs for subsystems
• Step 2. Capturing MoEs for subsystems

Step 1. Organizing the model for MoEs for subsystems

Following the structure of the MagicGrid framework, model elements that capture MoEs for subsystems
should be stored in a separate package inside the 2 White Box package. We recommend naming the
package after the cell, that is, 4 MoEs for Subsystems.

106 | MagicGrid® Book of Knowledge

To organize the model for MoEs for subsystems

1. Right-click the 2 White Box package and select Create Element.

2. In the search box, type pa (the first two letters of the element type Package) and press Enter.
3. Type 4 MoEs for Subsystems to specify the name of the new package and press Enter.

Step 2. Capturing MoEs for subsystems

Let's specify the MoEs for the conceptual subsystem of the VCCU, captured in the model as the Cooling
block. Just like MoEs of the entire SoI, MoEs of that conceptual subsystem refine the same non-functional
stakeholder needs: SN-1.1.1, SN-1.1.4, and SN-1.2.1. Therefore, the Cooling block can inherit all the MoEs
from the same MoEs holder introduced in the Measures of Effectiveness cell. Later, the inherited value
properties can be redefined if there is a need to change their properties (such as the name or type). In this
case, we will change only names, as types are relevant.

To capture MoEs for the Cooling subsystem

1. Create a bdd. For this:

a. Right-click the 4 MoEs for Subsystems package and select Create Diagram.
b. In the search box, type bdd (the acronym of the SysML block definition diagram), and then
double-press Enter. The diagram is created.

 Note that the diagram is named after the package where is it stored. This name perfectly
suits the diagram, too. You need only remove the sequence number from its name.

2. Display the MoEs Holder block on the newly created diagram. For this:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type MoEs.
c. When you see the MoEs Holder block selected in the search results list (see the following figure),
press Enter. The MoEs Holder block is selected in the Model Browser.

MagicGrid® Book of Knowledge | 107

 d. Drag the block to the diagram pane.

3. In the same way, display the Cooling block on the diagram as well.
4. Select the Cooling block and click the Generalization button on its smart manipulator toolbar.
5. Select the MoEs Holder block. As a result, the Cooling block inherits all the properties of the MoEs
Holder block.

6. Display inherited properties on the shape of the Cooling block. For this:

a. Right-click the block and select Symbol Properties.

b. On the top-right corner of the open dialog, select All from the drop-down list to switch the dialog
to the All Properties mode.
c. In the search box below the properties list, type inh.
d. When you see the Show Inherited symbol property displayed on the top of the list, click its value
cell to set it to true.

108 | MagicGrid® Book of Knowledge

 e. Close the dialog.

7. Redefine the totalEnergyConsumption value property. For this:

a. Double-click the Cooling block to open its Specification.

b. On the left side of the dialog, click Properties.
c. On the right side, select the totalEnergyConsumption value property and click the Redefine button
below.
d. In the Name cell, type the new name of the property, for example, energyConsCool.
e. Press Enter to confirm the changes.

8. In the same way, redefine other properties.

9. Close the dialog.

MagicGrid® Book of Knowledge | 109

When you're finished, your diagram should look like the following figure.

After you capture the MoEs for all conceptual subsystems of the Vehicle Climate Control Unit, your
diagram should look similar to the one below.

 As you can see, the values compartments are cut to display the full length of value property rows.
For this, do the following:

1. Select all the blocks that capture conceptual subsystems of the Vehicle Climate Control Unit.
Press Shift for multiple selection.
2. Right-click the selection and choose Symbol Properties.
3. In the open dialog, find the Shrinkable Attributes and Operations Compartment property and
set its value to true.

MoEs for Subsystems done. What's next?

When you have the MoEs for subsystems, you can specify what stakeholder needs they refine (see
Chapter Traceability to Stakeholder Needs).

110 | MagicGrid® Book of Knowledge

Traceability to Stakeholder Needs
What is it?

In order to have a complete problem domain model, you need to establish traceability relationships
between the elements that capture stakeholder needs and the elements that capture the white-box
functions, conceptual subsystems, conceptual interfaces, and non-functional characteristics of the SoI.
Since the latter are more concrete than stakeholder needs, we recommend establishing refine
relationships. These relationships should point from the elements of the lowest level of detail, such as
actions that capture leaf functions of the functional decomposition model. When you're done with
traceability relationships here, every item of the stakeholder needs must be refined by one or more
elements of the other SysML types. The opposite means that not all stakeholder needs are considered in
the problem domain model, which may lead to incorrect system requirements specification (see Chapter
System Requirements), resulting in developing the wrong system.

Who is responsible?

Traceability relationships can be established by the Systems Analyst or Systems Engineer.

How to model?

Neither of the SysML diagrams is suitable for creating a mass of cross-cutting relationships, such as refine.
For this, a matrix or a map is more suitable. The modeling tool offers a wide variety of predefined matrices
for specifying cross-cutting relationships. To capture refine relationships, a Refine Requirement Matrix can
be used.

MagicGrid® Book of Knowledge | 111

112 | MagicGrid® Book of Knowledge
Tutorial
• Step 1. Organizing the model for traceability
• Step 2. Creating a matrix for capturing refine relationships
• Step 3. Capturing refine relationships

Step 1. Organizing the model for traceability

Traceability relationships and the views where they are displayed can be stored in a separate package,
which is created directly under the 1 Problem Domain package as the last of its internal packages to
represent the modeling workflow.

To organize the model for traceability

1. Right-click the 1 Problem Domain package and select Create Element.

2. In the search box, type pa, the first two letters of the element type Package, and press Enter.
3. Type 4 Traceability to specify the name of the new package and press Enter.

Step 2. Creating a matrix for capturing refine relationships

To begin capturing refine relationships, you should create a Refine Requirement Matrix first. The matrix is
predefined to represent requirements in its columns and refine relationships in its cells. You only need to
specify that the rows of the matrix display the following:
• Call behavior actions, which capture functions of the SoI
• Part properties, which capture conceptual subsystems of the SoI
• Proxy ports, which capture the interfaces of the SoI
• Value properties with the «moe» stereotype applied as they capture MoEs of the SoI or its
subsystems

Remember, stakeholder needs must be refined by the problem domain elements from the lowest level of
detail. Therefore, the column elements must be taken from the package that contains the white-box
analysis results.

MagicGrid® Book of Knowledge | 113

To create a Refine Requirement Matrix

1. Right-click the 4 Traceability package and select Create Diagram.

2. In the open dialog, click the Expert button to turn on the appropriate mode.
3. In the search box, type rrm, the acronym of the Refine Requirement Matrix, and then press Enter. The
matrix is created.
4. Type Refine Stakeholder Needs to specify the name of the new matrix and press Enter again.
5. Specify the column element scope:

a. In the Model Browser, select the SN-1 Stakeholder Needs requirement.

b. Drag it onto the Column Scope box in the Criteria area above the matrix contents. The
selected package becomes the scope of the matrix columns.

6. Specify row element types:

a. In the Model Browser, select any call behavior action, such as the one typed by the Initialize
Cooling activity (which appears under the Reach Desired Temperature activity).

b. Drag the selection onto the Row Element Type box in the Criteria area above the matrix contents.
The rows of the matrix are set to display the elements of the call behavior action type.
c. In the Model Browser, select any part property; for example, the one typed by the Cooling block.

d. Hold down the Ctrl key and drag the selection onto the Row Element Type box in the Criteria area
above the matrix contents. The rows of the matrix are set to display the elements of the part
property type as well.
e. Repeat steps c and d to enable the rows of the matrix to display the proxy ports and value
properties with the «moe» stereotype.

7. Specify the row element scope:

a. In the Model Browser, select the 2 White Box package.

b. Drag the selection onto the Row Scope box in the Criteria area above the matrix contents. The
selected package becomes the scope of the matrix rows.

114 | MagicGrid® Book of Knowledge

The following figure displays the Criteria area of the matrix, with highlights on steps 5, 6, and 7 related
criteria.

 If the matrix displays interface blocks, you can easily get rid of them by modifying its display
settings. For this, click the ... button next to the Row Element Type box and in the open dialog, click
to clear the Include Subtypes check box. Don't forget to refresh the matrix.

Step 3. Capturing refine relationships

Now you’re ready to capture refine relationships. Let’s create the one from the Cool Down Temperature to
Desired action to the SN-1.1.6 Desired Temperature stakeholder need (captured as a requirement) to convey
that the action refines the need.

To specify the refine relationship in the matrix

• Double-click the cell at the intersection of the row that displays the Cool Down Temperature to
Desired action and the column that displays the SN-1.1.6 Desired Temperature requirement. The
derivation relationship is created between the appropriate items and displayed in the cell.

 To save space, the contents of the 2 Conceptual Subsystems Communication package were
hidden in the following figure.

MagicGrid® Book of Knowledge | 115

After all refine relationships are captured, your Refine Stakeholder Needs matrix should look like the one in
the following figure. Remember that when you're done, every item of the stakeholder needs must be
refined by one or more elements from the problem domain analysis model.

116 | MagicGrid® Book of Knowledge

As you might have noticed, the matrix displays implied refine relationships (dotted arrows) as well. Once
the leaf requirement is refined, all upper level requirements of the same branch are also considered
refined; this is the logic of establishing implied refine relationships. For information on how to create and
use implied relationships, refer to the latest documentation of the modeling tool.

Problem Domain done. What's next?

The problem domain analysis is complete when at least 90 percent of stakeholder needs (this threshold
may vary in different organizations) are refined by one or more elements from the problem domain model.
For monitoring the status of the problem domain analysis project, you can utilize the metrics feature
supported by the modeling tool. The metric table in the following figure calculates the metrics specified
by the Requirement Refinement metric suite. For more information about using metrics, see the latest
documentation of the modeling tool.

The complete problem domain analysis model, including the stakeholder needs for the SoI, the expected
functions and conceptual structure of the SoI, either with or without the MoEs, can be considered as initial
system requirements specification in the model-based form. If the system requirements specification is
required in textual form (which is not a rare case), the problem domain analysis model can be transformed
(either manually or automatically by utilizing the transformation algorithms) into the SysML requirements
model and provided to system architects as input for developing the logical architecture of the SoI in the
solution domain. This is described in Chapter Solution domain.

MagicGrid® Book of Knowledge | 117

SOLUTION DOMAIN
Once the problem domain analysis is completed and the stakeholder needs for the SoI are transferred into
the SysML model, it’s time to start thinking about the solution domain model of that SoI. The solution
domain model defines a precise cross-discipline logical architecture of the SoI to solve the problem defined
in the problem domain analysis model. It should not be mixed with detailed design models, which are
created in later phases of the SoI development and are not part of the MagicGrid framework.

As the same problem definition can have several solutions, a trade-off analysis can be performed to choose
the optimal solution for implementing the system (see the following figure).

It should be noted that a problem domain can be specified by one organization, and a solution domain by
another. Furthermore, diverse solutions can be provided by separate organizations.

As you can see in the following table, creating the solution domain model consists of specifying
requirements, structure, behavior, and parameters of the SoI. Furthermore, the task of building a logical
architecture usually includes more than one iteration, going down from the system-level to the
subsystem-level, from the subsystem-level to the component-level architecture and even deeper, if there is
a need. The precision of the system architecture model depends on the number of iterations.

118 | MagicGrid® Book of Knowledge

Deploying the solution domain

The logical architecture of the SoI can be specified in a single model and even in the same model with the
problem domain definition. However, in most cases, the scope and complexity of the solution domain
model exceeds the scope and complexity of the problem domain definition, especially when
it includes more than one solution. Specifying the solution domain in two or more separate models helps
to manage the information complexity. Moreover, such model partitioning enables a more granular model
ownership, change analysis and control, access control, concurrent modifications, and model reuse.

The following material describes a method of deploying the solution domain across multiple models.

As stated before, more than one solution can be provided for the same problem. To explain our approach,
we introduce the case of three solution domain models: Solution Domain 1, Solution Domain 2, and Solution
Domain 3. Though the contents of the Solution Domain 1 and Solution Domain 2 are suppressed, you should
imagine them similar to the contents of the Solution Domain 3. The Logical System Architecture (LSA) model
is the core of each solution domain. Based on the results of the problem domain analysis, it defines the
logical subsystems of the SoI in a single-level hierarchy. As you can see in the following figure, there are
three logical subsystems in this case: Subsystem 1, Subsystem 2, and Subsystem 3. It's important to note
that each subsystem is considered a black box in the LSA model, which also defines logical interfaces for
each subsystem to determine how they shall operate with one another and integrate into the whole
according to system requirements. Interfaces are not included in the figure below, to avoid clutter. The
owner of the LSA model is the systems architect, who has “a big picture view” of the SoI. In other words,
the systems architect is like the conductor of an orchestra: he/she ensures everyone is playing the same
melody at the same time.

Definition of logical subsystems in the LSA model helps to identify work packages and appoint them to
separate engineering teams. Every team develops one or more logical architectures for the appointed
subsystem. It may happen that they work in parallel. As seen in the following figure, each of the three
subsystems has several solutions. Logical Subsystem X Architecture 1, where X stands for 1, 2, or 3, is visible
on the top and the others are hidden underneath. The generalization relationship (solid line with hollow,
triangular arrowhead) between the Subsystem 1 and Subsystem 1 Architecture 1 blocks means that the
engineering team working on the Logical Subsystem 1 Architecture 1 model knows (or, in terms of SysML,
inherits) the design constraints (such as logical interfaces) of the logical subsystem defined in
the LSA model, and must follow them.

Other generalizations convey appropriate information. Architectures of different logical subsystems and
even different architectures of the same logical subsystem may include a diverse number of components.
A solution domain model of a logical subsystem defines the requirements, structure, behavior, and
parameters of that logical subsystem.

After engineering teams finish with the logical architectures of all subsystems defined in the LSA model,
the systems architect can integrate them into a single model. He/she picks up the preferred one of each
logical subsystem and builds the integrated logical architecture of the whole system. Note that more than
one solution can be proposed. As you can see in the following figure, we call them System Configuration
1, System Configuration 2, etc. Just like the logical architecture of every subsystem, the integrated logical
architecture of the whole SoI defines the structure, behavior, and parameters, and must meet system
requirements specification.

Having the full model of the SoI, the systems architect can perform a trade-off analysis and pick up the
optimal system configuration; having optimal system configurations of several solution domains, the
systems architect can pick up one optimal solution above all. This is illustrated by the Trade Off analysis in
the following figure.

MagicGrid® Book of Knowledge | 119

It is important to mention that trade offs can be done not only between solution domains, but at every
level of granularity in the logical architecture of the SoI.

Re-applying MagicGrid from scratch in the solution domain

It commonly happens that subsystem requirements are not adequate to build the logical architecture of
the subsystem, especially when it is developed by a supplier (contractor), as the Original Equipment
Manufacturer (OEM, contracting authority) usually is not willing to share the problem domain model.
Thus, the supplier may decide to perform the problem domain analysis of that logical subsystem
beforehand. The figure below illustrates re-applying the MagicGrid framework from scratch, at the logical
subsystem level, where the logical subsystem is considered the SoI.

120 | MagicGrid® Book of Knowledge

Building the logical architecture of the SoI
In this phase of specifying the solution domain, you should imagine yourself being the systems architect
and working to deliver the logical architecture of the SoI (i.e., building the logical system architecture (LSA)
model). The essential goal of building this model is to define the logical subsystems of the SoI, which leads
to identification of work packages.

As mentioned before, the logical architecture of the SoI can be stored in a separate model, apart from the
problem domain definition. The highlighted area in the following figure shows exactly which part of the
solution domain we are talking about.

As you may conclude looking at the design layout of the MagicGrid framework (see the highlighted area in
the following figure), the system-level solution domain model specifies the system requirements,
structure, behavior, and parameters. Additionally, every SysML element you create in the LSA model must
satisfy at least one system requirement. Otherwise, there is no motivation for having it in the LSA model.
Therefore, you have to start and end the system-level solution domain model with system requirements:
first specify them and, when you have the entire LSA model, establish traceability relationships between
them and logical system architecture elements.

MagicGrid® Book of Knowledge | 121

System Requirements
What is it?

To specify a logical system architecture, you first need to have the system requirements specification in
order to follow it. As you may remember, the results of the problem domain analysis from both black- and
white-box perspectives can be considered as initial system requirements specifications presented in
the model-based form. The main task of this cell is to manually transform the problem domain analysis
model into the SysML requirements model, which enables you to present the system requirements
specification in a textual form.

After the transformation, the following traceability relationships must be established to cross the
boundary between the solution and problem domain models:

• From system requirements to stakeholder needs – to assert which system requirements are derived
from which stakeholder needs
• From system requirements to the rest of the problem domain model – to assert which elements are
refined by which system requirement

System requirements specification can be constantly updated while working on the logical architecture of
the SoI. Moreover, requirements for the logical architecture of subsystems, components, and even more
precise items are derived from the system requirements. This is described in Chapter Subsystem
Requirements.

Unlike stakeholder needs, system requirements are verifiable: after you finish with the LSA model, satisfy
relationships must be established from the elements capturing the logical system architecture to the
system requirements in order to assert which of the former fulfills which of the latter. This is described in
Chapter Traceability to System Requirements.

122 | MagicGrid® Book of Knowledge

Who is responsible?

System requirements specification must be produced by the Systems Analyst or Systems Engineer who
has performed the problem domain analysis.

How to model?

In the modeling tool, an item of system requirements specification can be stored as the SysML
requirement, which has a unique identification, name, and textual specification. As opposed to the textual
specifications of stakeholder needs, system requirements must be formal and written down following
certain guidelines. For example, write in short and clear sentences only, avoid conditional keywords (such
as if, except, as well), use the shall keyword, and include tables. For more information on how to write good
textual requirements, refer to the INCOSE Systems Engineering Handbook.

System requirements can be organized into one or more hierarchy levels. For this, use the containment
relationships among them. The SysML requirement referring to the system requirements specification
must be at the top of the hierarchy. It may have no text.

The SysML requirement table or diagram can help to visualize the hierarchy of system requirements (see
the following figure). The table or diagram can also be included in the requirements report.

Each system requirement must be derived from a certain item of stakeholder needs and/or refine one or
more elements from the problem domain model. To assert a derivation, you should establish the
«deriveReqt» relationship pointing from the system requirement to the appropriate item of stakeholder
needs (the following figure shows these relationships displayed in the matrix). To assert a refinement, the
«refine» relationship pointing from the system requirement to some part property, call behavior action, or
MoE defined in the problem domain model, should be established.

MagicGrid® Book of Knowledge | 123

.

Tutorial
• Step 1. Creating and organizing a model for system requirements
• Step 2. Creating a diagram for system requirements
• Step 3. Specifying system requirements
• Step 4. Establishing traceability to stakeholder needs
• Step 5. Establishing traceability to the rest of the problem domain model

Step 1. Creating and organizing a model for system requirements

To get ready for specifying system requirements, you need to establish an appropriate structure of the
model first. Following the design of the MagicGrid framework, model artifacts that capture system
requirements should be stored under the structure of packages, as displayed in the following figure. As
you can see, the upper-level package represents the domain, and the lower-level package stands for
the cell.

However, in real-world projects it is quite common to keep the solution domain model separated from the
problem domain model. Following this practice, we recommend specifying the system requirements and
the logical architecture of the Vehicle Climate Control System (VCCS) in another model. The contents of the
problem domain model should be available in that model, because the result of the problem domain
analysis is considered as an initial system requirements specification that should be transformed from the

124 | MagicGrid® Book of Knowledge

model-based to the textual form. Speaking in terms of the modeling tool, the problem domain model
should be used in the solution domain model. For this, we utilize the relevant capabilities of the modeling
tool.

To create and organize the model for system requirements specification

1. Start sharing the problem domain model.

 For more information on this topic and the topics mentioned in steps 2 and 3, refer to
the latest documentation of your modeling tool.

2. Create a new model. It can be named VehicleCCS Solution.mdzip.

3. Use the problem domain model as read-only in the solution domain model. This is the solution domain
model of the system.

 Names in grey indicate that elements cannot be modified. This is because the problem domain
model is used in the solution domain model as read-only.

4. Right-click the Model package (this is the default name of the root package) and select Create
Element.
5. In the search box, type pa (the first two letters of the element type Package) and press Enter.
6. Type 2 Solution Domain to specify the name of the new package and press Enter.
7. Right-click the 2 Solution Domain package and select Create Element.
8. Repeat step 5.
9. Type 1 System Requirements to specify the name of the new package and press Enter.

MagicGrid® Book of Knowledge | 125

Step 2. Creating a diagram for system requirements
For capturing and displaying system requirements, we recommend using the SysML requirement diagram
or table. While using the table is introduced in Chapter Stakeholder Needs, using the diagram is described
below.

To create the SysML requirement diagram for specifying system requirements

1. Right-click the 1 System Requirements package and select Create Diagram.

2. In the search box, type rd, where r stands for requirement and d for diagram, and then double-press
Enter. The diagram is created.

 Note that the diagram is named after the package where it is stored. This name works for the
diagram, too. You may only remove the sequence number from its name.

Step 3. Specifying system requirements

The white-box analysis of the Vehicle Climate Control Unit reveals the following system requirements:

# Name Text
1 Setting Desired Temperature The user shall be able to specify the temperature he/she desires in the cabin.
2 Human-Machine The system shall be able to accept data and commands from the user.
Interactions
3 Getting Ambient The system shall take the ambient temperature from the vehicle temperature
Temperature sensor.
4 Cabin Air The system shall be able to draw air out of the cabin and blow comfortable air
back.
5 Heating The system shall be able to heat the air, if the user wishes to increase the
temperature in the cabin.
6 Cooling The system shall be able to cool the air, if the user desires to decrease the
temperature in the cabin.
7 Central Processing The system shall "decide" whether to cool or to heat the air in the cabin. For
this, it shall take the ambient temperature and compare it to the desired
temperature set by the user.
8 Output Information Display The system status and ambient temperature in the cabin shall be always
displayed in some information output device.
9 Max Energy Consumption The system shall consume maximum 3.518 kWh.

126 | MagicGrid® Book of Knowledge

To capture the system requirements in your model

1. Open the System Requirements diagram, if not opened yet.

2. Click the Requirement button on the diagram palette and then click a free space on the diagram pane.
An empty requirement is created.
3. Type the name of the first requirement from the table above directly on the shape and press Enter.
4. Click the Text property value on the shape once and then again. The property switches to the edit
mode.
5. Type the text of the first requirement from the table.
6. Click somewhere outside the shape when you’re done.
7. Repeat steps 2 to 6 to capture the rest of the requirements.

After you’re done, the contents of the System Requirements diagram should look very similar to the one in
the following figure.

In addition, you need a grouping requirement to represent the entire requirements specification. The
easiest way of creating a grouping requirement is described in step 4 of Chapter Stakeholder
Needs tutorial. Follow the procedure to create the SRS for VCCS grouping requirement.

To display the requirements hierarchy in the System Requirements diagram

1. Drag the SRS for VCCS requirement to that diagram pane.

2. Right-click the shape of that requirement and select Display > Display Paths. Containment
relationships among requirements are displayed on the diagram pane.
3. On the diagram toolbar (located on the top of the diagram pane), click the Quick Diagram Layout
button .

In order to distinguish system requirements from stakeholder needs, subsystem and component
requirements, and detailed requirements for the system implementation, we recommend adding prefixes
to their numbers. Follow the procedure described in step 5 of Chapter Stakeholder Needs tutorial to add
the prefix SR- to your system requirements. Subsequently, these system requirements can be refactored to
more specific ones. For example, Setting Desired Temperature can become an interface requirement and Max
Energy Consumption a physical requirement. For this, follow the procedure described in step 6 of
Chapter Stakeholder Needs tutorial. After you finish numbering and categorization, the contents of
your System Requirements diagram should look very similar to the one in the following figure.

MagicGrid® Book of Knowledge | 127

Step 4. Establishing traceability to stakeholder needs
Traceability relationships from system requirements to stakeholder needs can be established either in a
SysML requirements diagram or a matrix. We recommend using the matrix for large scope, because it
provides a more compact view of the model than the diagram does.

In this case, you need to assert what system requirements (more concrete) are derived from what
stakeholder needs (less concrete), that is, establish derivation relationships between them. Therefore, we
recommend utilizing the infrastructure of the Derive Requirement Matrix, one of the predefined matrices
in the modeling tool. As opposed to a common matrix, it has certain criteria predefined to speed up
building the view.

Just like in the problem domain model, traceability views and relationships can be stored in a separate
package. As you already know how to create packages, we assume that you do this on your own. For
keeping your model well-organized according to the MagicGrid framework, that package should always be
the last sub-package of the 2 Solution Domain package. Therefore, its name must begin with number 5, for
example, 5 Traceability.

To create a Derive Requirement Matrix

1. In the Model Browser, right-click the 5 Traceability package and select Create Diagram.
2. In the search box, type drm (the acronym for the predefined Derive Requirement Matrix), and press
Enter.

 If you don’t see any results, click the Expert button below the search results list. The list of
available diagrams expands in the Expert mode.

3. Type System Requirements to Stakeholder Needs to specify the name of the new matrix and press Enter
again.
4. In the Model Browser, select the SR-1 SRS for VCCS requirement and drag it onto the Row Scope box in
the Criteria area above the matrix contents. The SR-1 SRS for VCCS requirement becomes the scope of
the matrix rows.

128 | MagicGrid® Book of Knowledge

 5. In the Model Browser, select the SN-1 Stakeholder Needs requirement (you might need to expand the 1
Problem Domain, 1 Black Box, and 1 Stakeholder Needs packages first) and drag it onto the Column
Scope box in the Criteria area above the matrix contents. The SN-1 Stakeholder Needs
requirement becomes the scope of the matrix columns.
6. Click the Direction box and select Row to Column, as you need to establish derivation relationships
pointing from system requirements (row elements), to stakeholder needs (column elements).
7. Click the Refresh hyperlink in the notification box below the Criteria area. The contents of the matrix
are updated. All the cells are empty at the moment.

The following figure displays the Criteria area of the matrix, with highlights on step 4, 5, and 6 related
criteria.

Now you’re ready to establish derivation relationships. Let’s create one between SR-1.1 Setting Desired
Temperature and SN-1.6 Desired Temperature to convey that SR-1.1 Setting Desired Temperature is derived
from SN-1.6 Desired Temperature or, in other words, that this system requirement is created because of that
stakeholder need.

To specify the derivation relationship in the matrix

• Double-click the cell at the intersection of the row that displays SR-1.1 Setting Desired
Temperature and the column that displays SN-1.6 Desired Temperature. The derivation relationship is
created between the appropriate items and displayed in the cell.

MagicGrid® Book of Knowledge | 129

  If you want to read the requirement text before setting the relationship, just move the mouse over
that requirement name and rest it there for a few seconds. The text appears in the tool tip.

After all derivation relationships are established, your System Requirements to Stakeholder Needs
matrix should resemble the one in the following figure. You may remember that every system requirement
must be derived from one or more stakeholder needs. Otherwise, the system requirements specification
must be revised and updated. Some stakeholder needs may be left uncovered, if it was decided not to
address them in the system requirements specification.

Afterwards, a Requirement Derivation Map can be created, with blue arrows representing the relationships
between requirements from different domains.

130 | MagicGrid® Book of Knowledge

Step 5. Establishing traceability to the rest of the problem domain model
Traceability relationships from system requirements to the rest of the problem domain model can be
established either in a SysML requirements diagram or in a matrix. We recommend using the matrix for
large scope, because it provides a more compact view of the model than the diagram does.

In this case, you need to assert which problem domain elements are refined by which system
requirements; that is, establish refine relationships between them. Therefore, we recommend utilizing the
infrastructure of the Refine Requirement Matrix, one of the predefined matrices in the modeling tool. As
opposed to a common matrix, it has certain criteria predefined to speed up building the view.

This matrix and appropriate relationships can be stored in the 5 Traceability package you created in the
previous step.

To create a Refine Requirement Matrix

1. In the Model Browser, right-click the 5 Traceability package and select Create Diagram.
2. In the search box, type rrm, the acronym for the predefined Refine Requirement Matrix, and press
Enter.

 If you don’t see any results in the diagram types list, click the Expert button below it. The list
of available diagrams expands in the Expert mode.

3. Type System Requirements to Problem Domain to specify the name of the new matrix and press Enter
again.
4. On the matrix toolbar, click the Change Axes button to swap rows of the matrix with columns.
Requirement becomes the type of the matrix rows.
5. In the Model Browser, select the SR-1 SRS for VCCS requirement (you might need to expand
the 1 System Requirements package first) and drag it onto the Row Scope box in the Criteria area above
the matrix contents. The SR-1 SRS for VCCS requirement becomes the scope of the matrix rows.
6. Select column element types:

a. In the Model Browser, select any call behavior action; for example, the one typed by the Initialize
Cooling activity (it appears under the Reach Desired Temperature activity which can be found
within the 3 Functional Analysis package in the problem domain analysis model).
b. Drag the selection onto the Column Element Type box in the Criteria area above the matrix
contents. The columns of the matrix are set to display the elements of the call behavior action
type.
c. In the Model Browser, select any part property, such as the one typed by the Cooling block. It
appears under the VCCU block, which can be found within the 2 Logical Architecture package in the
problem domain analysis model.
d. Hold down the Ctrl key and drag the selection onto the Column Element Type box in
the Criteria area above the matrix contents. The columns of the matrix are set to display the
elements of the part property type as well.
e. Repeat steps c and d to enable the columns of the matrix to display the proxy ports and value
properties with the «moe» stereotype.

7. Specify the column element scope:

a. In the Model Browser, select the 2 White Box package in the problem domain analysis model.
b. Drag the selection onto the Column Scope box in the Criteria area above the matrix contents. The
selected package becomes the scope of the matrix columns.

MagicGrid® Book of Knowledge | 131

 8. Click the Direction box and select Row to Column, as you need to establish refine relationships
pointing from system requirements (displayed as row elements), to the artifacts that capture problem
domain concepts (displayed as column elements).
9. Click the Refresh hyperlink in the notification box below the Criteria area. The contents of the matrix
are updated. All cells are empty at the moment.
The following figure displays the Criteria area of the matrix with highlights on steps 5, 6, 7, and 8 related
criteria.

Now you’re ready to establish refine relationships. Let’s create one between the SR-1.1 Setting Desired
Temperature requirement and the p1 proxy port of the Human-Machine Interacting block to assert that the
former refines the latter, or, in other words, that this system requirement is created to refine that proxy
port.

To specify the refine relationship in the matrix

• Double-click the cell at the intersection of the row that displays SR-1.1 Setting Desired
Temperature and the column that displays the p1 proxy port . The refine relationship is created
between appropriate items and displayed in the cell.

After all refine relationships are specified, your System Requirements to Problem Domain matrix should
resemble the one in the following figure. Keep in mind that every system requirement must refine one or
more elements captured in the problem domain analysis model. Otherwise, the system requirements
specification must be revised and updated. Some elements can be left uncovered, if it was decided not to
address them in system requirements specification (such as MoEs for the sound level and total weight).

132 | MagicGrid® Book of Knowledge

System Requirements done. What's next?

• When you have the system requirements specification captured in the SysML requirements model,
you can start building the logical architecture of the SoI . This means you can move to the System
Structure cell.
• System requirements specification can be updated any time while building the logical architecture
of the SoI. The updates should stop only after switching to the subsystem-level solution domain
model.

MagicGrid® Book of Knowledge | 133

System Structure
What is it?

After the system requirements specification is completed, you can start building the logical architecture of
the SoI. Usually it starts with capturing the logical subsystems of the SoI. The LSA model also specifies
interfaces, to determine how these subsystems shall inter-operate with one another and integrate into the
whole. Subsystems of the solution domain may differ from the subsystems identified in the problem
domain. This happens quite often, as the problem domain definition is not as precise as the solution
architecture. For the purpose of tracing from the subsystems of the solution architecture to the
subsystems of the conceptual architecture in the problem domain model, cross-cutting relationships can
be established.

Defining the logical subsystems serves for the purpose of establishing the work-breakdown structure
(WBS) in the systems engineering project: each logical subsystem identifies a single work package, which
is immediately appointed to some engineering team. That team develops one or more logical architectures
for the appointed subsystem. Separate teams usually work in parallel to develop logical architectures of
different subsystems. Modeling the logical architecture of a single subsystem is described in Chapter
Building the logical architecture of subsystems.

Who is responsible?

System structure in the LSA model can be captured by the Systems Architect or Systems Engineer. He/she
must be the head of the whole systems engineering project, the one, who has “a big picture view” of the
SoI and is responsible for the smooth integration of the subsystems and their components into the whole.

How to model?

As you can remember, the LSA captures the system of interest, its logical subsystems at one level down,
and the interfaces to determine how these subsystems inter-operate, that is, what data, matter, or energy
they exchange. The LSA model can be created and displayed in a SysML bdd. Subsystems can be captured
as blocks, and interfaces as proxy ports typed by appropriate interface blocks.

In the following figure, you can see a sample LSA model of the Vehicle Climate Control System (VCCS). It
displays the blocks that capture logical subsystems of the VCCS, interface blocks that own flow properties
which capture inputs and outputs for both external and internal communications as well as various
exchange items which type relevant flow properties.

134 | MagicGrid® Book of Knowledge

The «abstraction» relationship can be used to declare what logical subsystems in the LSA are derived from
what conceptual subsystems (functional blocks) in the problem domain analysis model.

Tutorial
• Step 1. Organizing the model for the LSA
• Step 2. Creating a bdd for capturing the LSA
• Step 3. Capturing logical subsystems
• Step 4. Establishing traceability to the problem domain model
• Step 5. Capturing logical interfaces

MagicGrid® Book of Knowledge | 135

Step 1. Organizing the model for the LSA
The LSA can be captured in the solution domain model you created in Chapter System Requirements. To
get ready for modeling in this cell, you need to establish an appropriate structure of the model beforehand.
Following the design of the MagicGrid framework, relevant model artifacts should be stored under the
structure of packages, as displayed in the following figure.

To organize the model for the LSA

1. Open the solution domain model (the VehicleCCS Solution.mdzip file) you created in Chapter System
Requirements, if not opened yet.
2. Right-click the 2 Solution Domain package and select Create Element.
3. In the search box, type pa (the first two letters of the element type Package), and press Enter.
4. Type 2 System Structure to specify the name of the new package and press Enter.

For the purpose of keeping the inner structure of the 2 System Structure package well organized, you need
to create a few more packages. These are:
• 1 Logical Interfaces
• 2 Exchange Items
• 3 Logical Subsystems

After you create them (on your own by following the previous procedure), the Model Browser of your
solution domain model should look the same as the following figure.

136 | MagicGrid® Book of Knowledge

Step 2. Creating a bdd for capturing the LSA
To get ready for capturing the logical architecture of the VCCS, you need to create a bdd first.

To create a bdd for capturing the logical architecture of the VCCS

1. Right-click the 2 System Structure package and select Create Diagram.

2. In the search box, type bdd, the acronym for SysML block definition diagram, and then press
Enter. The diagram is created.
3. Type Logical Architecture of VCCS to specify the name of the new diagram and press Enter again.

Step 3. Capturing logical subsystems

As the system requirements specification reveals, the VCCS consists of:
• HMI System, where HMI stands for Human-Machine Interface
• CP System, where CP stands for Central Processing
• Cooling System
• Heating System

Logical subsystems can be captured as blocks in the bdd you created in the previous step. As already
mentioned, the LSA model always defines logical subsystems at one level down, and the logical
architecture of each subsystem should be subsequently produced by separate engineering teams in
isolated models (see Chapter Subsystem Structure).

Note that the names of the logical subsystems correspond to the names of the conceptual subsystems in
the problem domain model, such as Cooling System to Cooling, or Heating System to Heating. However,
these concepts are not the same; therefore, be sure you haven't selected the block from the problem
domain model instead of creating a new one in the solution domain model, when specifying the logical
subsystems of the VCCS. To prevent this mistake, you can narrow down the name auto-completion and
type selection scope beforehand. The following figure displays the auto-completion before (on the left)
and after (on the right), narrowing the scope to the solution domain model only.

MagicGrid® Book of Knowledge | 137

To narrow the auto-completion and type selection scope down to the solution domain model

1. Create a package diagram:

a. Right-click the 2 System Structure package and select Create Diagram.

b. In the search box, type pd, where p stands for package and d for diagram, and then press
Enter. The diagram is created.
c. Type Scope for Auto-completion and Type Selection to specify the name of the new diagram and
press Enter again.

2. In the Model Browser, select the 2 Solution Domain and 2 System Structure packages (the latter is a
sub-package of the former) and drag them to the pane of the newly created diagram.
3. Select the shape of the 2 System Structure package and click the Package Import button on its
smart manipulator toolbar.
4. Select the shape of the 2 Solution Domain package. The package import relationship is created
between the packages.

From now on, when you model in the 2 System Structure package, elements suggested as names and types
will be taken from the 2 Solution Domain package only, while elements stored in the 1 Problem Domain
package (from the problem domain model) will be left aside. This means it won't be suggested that you
choose from conceptual subsystems and interfaces as well as exchange items captured in the problem
domain model.

If you still see those elements, click the Filter Options button below the list and then click to select the
Filter By Package Imports check box (see the following figure). For more information about the package
import feature, refer to the latest documentation of the modeling tool.

138 | MagicGrid® Book of Knowledge

To capture logical subsystems of the SoI

1. Open the Logical Architecture of VCCS diagram, if not opened yet.

2. Capture the SoI in the solution domain model:

a. Click the Block button on the diagram palette and then click on the diagram pane.
b. Type VCCS to name the block and press Enter.

3. Capture the logical subsystems:

a. Click the Directed Composition button on the smart manipulator toolbar of the VCCS block and
then click an empty place on the diagram pane.
b. Type HMI System to name the block and press Enter.
c. Repeat steps 3.a and 3.b to capture other subsystems from the list above.

4. On the diagram toolbar, click the Quick Diagram Layout button .

 Note that the part property names (displayed at the arrow end of the relationship) differ from
the default. They can be changed manually. For this, you need to select the part property you
want to rename on the diagram and then click it again to enter the name edit mode (see the
following figure). You are able to modify it immediately.

5. In the Model Browser, select all blocks of the subsystems and drag them to the 3 Logical Subsystems
package you created in step 1 of this cell tutorial. The blocks are moved to the 3 Logical Subsystems
package.

MagicGrid® Book of Knowledge | 139

Step 4. Establishing traceability to the problem domain model
Traceability relationships enable you to specify which conceptual subsystems (aka functional blocks)
identified in the problem domain model determine the creation of what logical subsystems in the LSA
model. We use the abstraction relationships to trace from the solution domain to the problem domain in
this case.

You already know that the most effective view for establishing traceability relationships is a matrix. Let's
create one.

To create the matrix for traceability relationships to the problem domain model

1. In the Model Browser, right-click the 5 Traceability package, a subpackage of the 2 Solution Domain
package, and select Create Diagram.
2. In the search box, type dm (the acronym for the Dependency Matrix), and press Enter.

 If you don’t see any results, click the Expert button below the search results list. The list of
available diagrams expands in the Expert mode.

3. Type LSA to Problem Domain to specify the name of the new matrix and press Enter again.
4. Specify the row and column element type:

a. In the Model Browser, select any block, such as the VCCS block.
b. Drag it onto the Row Element Type box in the Criteria area above the matrix contents. Rows and
columns of the matrix are both set to display the elements of the block type (see the figure
below).
c. Click the button next to the Column Element Type box.
d. In the open dialog, click to clear the Include Subtypes check-box. Otherwise, the columns display
a few interface blocks, which you don't need here.
e. Click OK to close the dialog.

5. Specify the row element scope:

a. In the Model Browser, select the VCCS block.

b. Drag the selection onto the Row Scope box in the Criteria area above the matrix contents. The
selected block becomes the scope of the matrix rows.
c. In the Model Browser, select the 3 Logical Subsystems package, which appears under the same
package with the VCCS block.
d. Hold down the Ctrl key and drag the selection onto the Row Scope box, too. The 3 Logical
Subsystems package becomes the scope of the matrix rows, together with the VCCS block (see the
following figure).

6. Specify the column element scope:

a. In the Model Browser, select the VCCU block (for this, you might need to expand the contents of
the following read-only packages: 1 Problem Domain > 2 White Box > 2 Conceptual Subsystems
Communication).
b. Drag the selection onto the Column Scope box in the Criteria area above the matrix contents. The
selected block becomes the scope of the matrix columns.
c. In the Model Browser, select the 2 Conceptual Subsystems package, which appears under the same
package with the VCCU block.

140 | MagicGrid® Book of Knowledge

 d. Hold down the Ctrl key and drag the selection onto the Column Scope box, too. The 2 Conceptual
Subsystems package becomes the scope of the matrix columns, together with the VCCU block (see
the following figure).

7. Specify the dependency criteria:

a. Click the button next to the Dependency Criteria box.

b. In the open dialog, find the abstraction relationship. For this:

i. Type abs in the search box below the criteria list.

ii. Click to select the check box next to the abstraction relationship.
iii. Click OK. The abstraction relationship becomes the dependency criteria of the matrix (see the
following figure).

8. Click the Direction box and select Row to Column, as you need to establish the abstraction
relationships pointing from logical subsystem (row elements), to conceptual subsystems (column
elements).
9. Click the Refresh hyperlink in the notification box below the Criteria area. The contents of the matrix
are updated. All the cells are empty at the moment.

The following figure displays the Criteria area of the matrix with highlights on step 4, 5, 6, 7, and 8 related
criteria.

Now you’re ready to specify abstraction relationships. Double-click the cell at the intersection of relevant
row and column (for example, between the row that represents the Cooling System block and the column
that represents the Cooling block).

MagicGrid® Book of Knowledge | 141

After all abstraction relationships are specified, the LSA to Problem Domain matrix in your model should
resemble the one in the following figure.

Step 5. Capturing logical interfaces

As the system requirements specification reveals, the VCCS shall be able to receive the following from the
outside:
• Desired temperature specified by the user
• Ambient temperature measured by the temperature sensor of the vehicle
• Cabin air for cooling or heating

These statements indicate the need for the logical interfaces that you see in the following figure. These
interfaces are more precise than conceptual interfaces determined in the problem domain analysis model:
the DesiredTmpSet and AmbientTmpRead signals have properties for holding specific values of Integer type.
These properties can even have default values, such as DesiredTmp=22.

142 | MagicGrid® Book of Knowledge

If you've gone through the tutorial in Chapter Conceptual Subsystems, you should already know how to
create interface blocks with flow properties and how to bind these interface blocks to system architecture
blocks. If you've skipped it, here is another chance for you to learn.

To capture the interface block for providing the desired temperature and relate it to the VCCS block

1. Open the Logical Architecture of VCCS diagram, if not opened yet.

2. Click the Interface Block button on the diagram palette and then click an empty place on the diagram
pane. An unnamed interface block is created in the model, directly under the 2 System
Structure package, and displayed on the diagram.
3. Type UserDataInput to name the interface block and press Enter.
4. Click the Create Property button on the interface block shape and select Flow Property.

MagicGrid® Book of Knowledge | 143

 5. Directly on the shape of the interface block, do the following:

a. Remove the out part of the flow property direction indicator to specify that the direction of the
flow property is in.
b. Type dts to specify the flow property name. Don't press Enter!
c. Type :DesiredTmpSet and press Enter. The DesiredTmpSet signal is created in the model directly
under the UserDataInput interface block and immediately set as its only flow property type.

6. Arrange your model:

a. Make sure the UserDataInput interface block is selected on the diagram and press Alt+B. The
element is selected in the Model Browser.
b. Click near the UserDataInput interface block to expand its contents.
c. Select the DesiredTmpSet signal and drag it to the 3 Exchange Items package.
d. Select the UserDataInput interface block again and drag it to the 1 Logical Interfaces package, a
sub-package of the 2 System Structure package.

7. Select the UserDataInput interface block on the diagram and drag it onto the VCCS block. As a result, a
new proxy port, typed by the UserDataInput interface block, is created for the VCCS block. Note that it
is decorated with an arrow indicating that this port is for inputs to the system of interest.

8. Create a diagram for displaying exchange items:

a. In the Model Browser, right-click the 3 Exchange Items package and select Create Diagram.
b. In the search box, type bdd (the acronym for SysML block definition diagram), and then press
Enter. The diagram is created.
c. Remove the sequence number from the diagram name press Enter again.

9. Create a property for the DesiredTmpSet signal:

a. In the Model Browser, select that signal and drag it to the Exchange Items diagram pane.
b. Click the Create Property button on the signal shape.
c. Type DesiredTmp to name the new property.
d. Type :Int press Ctrl + Spacebar to open the list of possible types. Once you see the Integer value
type (the pink one) in the search results list, select it.

144 | MagicGrid® Book of Knowledge

 e. Type =22 and press Enter. The property of the Integer type with 22 as the default value is created
for the signal.

Following the previous procedure, capture another two interface blocks (with all the related information)
and bind them to the VCCS block. Once you're done, the Logical Architecture of VCCS diagram in your model
should look very similar to the first image of this tutorial step.

The system requirements specification also reveals that the VCCS shall be able to provide the following to
the outside:
• Comfortable (cooled or heated) air back to the cabin
• Output information about the VCCS status for the user

These statements indicate that the VCCS needs a few more interfaces. You can see them in the following
figure. We assume that you update the model on your own.

MagicGrid® Book of Knowledge | 145

Now it's time to start thinking about the logical interfaces of logical subsystems of the VCCS. Just like the
SoI itself, each logical subsystem can have a number of logical interfaces. Some of them can be for external
communication. Therefore, they are inherited from the SoI. Some are for internal communication with
other subsystems of the VCCS. These are specified in the LSA model from scratch.

Let's start from specifying logical interfaces for external communication. As already mentioned, you don't
need to create new interface blocks: simply drag the selected interface block onto the block that captures
the relevant logical subsystem (the same as you did for the VCCS block). For example, you can drag the
InsideAirInlet interface block onto the Cooling System block. As a result, a proxy port typed by the
InsideAirInlet interface block is created for the Cooling System block.

After you specify all the interfaces for external communication, the Logical Architecture of the VCCS bdd in
your model should look very similar to the one in the following figure.

146 | MagicGrid® Book of Knowledge

Now let's think about the logical interfaces for internal communication between logical subsystems.

It is obvious that the CP System shall be able to send commands to other subsystems. These might include
Start, Stop, Idle, Cool, Heat, etc. They can be identified later, when modeling the logical architecture of
that subsystem. This requires creating a new interface block and a few signals (you can see them
highlighted in the following figure). Also, new subsystem requirement must added, but let's leave it for
modeling in other cells.

MagicGrid® Book of Knowledge | 147

Note that the cmd flow property direction of the newly created interface block is out. This is because it
was specified from the point of view of the CP System, which shall be able to send commands to other
subsystems of the VCCS.

The next step is to specify that the CP System block is able to send the Command signal or any of its sub-
types out, and the Cooling System, Heating System, and HMI System blocks are able to accept them. For
this, you should create a proxy port for each block that captures the logical subsystem and then make the
proxy ports of the Cooling System, Heating System, and HMI System blocks conjugated. As a result, names
of the interface block typing these proxy ports appear decorated with a tilde (“~”), and its flow property
directions become reversed for these proxy ports, which conveys that the Cooling System, Heating System,
and HMI System blocks are able to accept the Command signal or any of its sub-types.

To make the proxy port conjugated

• Right-click that proxy port and then click to select Is Conjugated.

 It's important to note here that even though this way of specifying compatible proxy ports is easier,
it is not always best. It is more common in real-world projects to define a couple of compatible
interface blocks instead of defining a single interface block and then making one of the proxy ports
conjugated.

There are two more interfaces, the InternalResponse and InternalDataTransferer, that can be specified for
internal communication. You can see them in the following figure. We assume that you capture all the
highlighted SysML elements in your model.

148 | MagicGrid® Book of Knowledge

System Structure done. What's next?

• Having logical subsystems and knowing how they will communicate, you can start working on their
logical architecture models. Therefore, you could move to Chapter Building the logical architecture
of subsystems.
• However, you still have two cells left for creating the logical system architecture model. Therefore,
you should proceed to the System Behavior and System Parameters cells.

MagicGrid® Book of Knowledge | 149

System Behavior
What is it?

The LSA is not only about the system structure; the systems architect can capture the behavior of the
whole SoI in the LSA model, too. However, this is not a common practice. Since the precise logical
architecture of the SoI is not yet defined, the system behavior model appears too abstract and cannot say
much about the system behavior. Moreover, it may conflict with the system behavior in the integrated
model of the selected system configuration (see Chapter Building system configuration model). That's
why this cell is often skipped when building the LSA of the SoI in real-world projects.

Unlike that in the LSA model, the system behavior in the integrated model of the selected system
configuration is crucially important. Since the integrated model includes the architectures of all logical
subsystems of the SoI, it can be used to analyze the integrated behavior of the entire system and perform
the system requirements verification.

Who is responsible?

System behavior in the LSA model can be captured by the Systems Architect or Systems Engineer. He/she
must be the head of the whole systems engineering project, the one who has “a big picture view” of the
SoI and is responsible for the smooth integration of the subsystems and their components into the whole.

How to model?

The system behavior in the LSA model can be captured by utilizing the infrastructure of the SysML state
machine and activity or sequence diagrams in combination. The SysML state machine allows capturing
states of the SoI and transitions between them in response to event occurrences over time. The activity or
sequence diagrams should be used to specify the entry, do, and exit behaviors of these states or transition
effects.

Though the diagram in the following figure captures the behavior of the Cooling System, which is one of
subsystems, the behavior of the SoI in the LSA model looks very much like this.

150 | MagicGrid® Book of Knowledge

  To prevent conflicts with the system behavior in the integrated model of the selected system
configuration, the system behavior defined in the LSA model must be unassigned as the classifier
behavior from the block, which captures the SoI. For this, do the following:

1. Right-click the VCCS block and select Specification.

2. In the open dialog, from the Properties drop-down list, select Expert to extend the list of
properties.
3. Click the cell of the Classifier Behavior property value and select <UNSPECIFIED> from the
drop-down list.

Tutorial

To follow the common practice, we skip this cell in the LSA model and get back to it in the system
configuration model (see Chapter System Configuration Behavior).

What’s next?

The next cell to visit is System Parameters.

MagicGrid® Book of Knowledge | 151

System Parameters
What is it?

You can go to this cell as soon as the system structure is captured in the model, even as abstract as in the
LSA model.

System parameters capture quantitative characteristics of the system of interest, that you can measure.
They are used to calculate MoEs identified during the problem domain analysis (see Chapter Measures of
Effectiveness). MoEs captured in the solution domain model satisfy non-functional quantitative system
requirements.

Mathematical expressions for MoEs derivation from system parameters should also be specified in this
cell. Knowing how to calculate MoEs enables early system requirements verification. Using the capabilities
of the modeling tool, the model can be executed to calculate these values and show whether quantitative
requirements are satisfied or not.

It's important to note here that not all MoEs in the solution domain must be based on the MoEs in the
problem domain model. Some of them can be identified only when building the logical architecture of the
SoI.

Who is responsible?

System parameters in the LSA model can be captured by the Systems Architect or Systems Engineer. He/
she must be the head of the whole systems engineering project, the one, who has “a big picture view” of
the SoI and is responsible for the smooth integration of the subsystems and their components into the
whole.

How to model?

As you should already know, measures of effectiveness can be captured in the model as value properties
with the «moe» stereotype applied. These value properties must belong to the block, which represents the
system of interest. You must also know that measures of effectiveness can be inherited from the reusable
sets of MoEs, called MoEs holders, instead of specifying them from scratch.

The formula for calculating the MoE can be captured as a constraint expression of some constraint block.
The infrastructure of the SysML block definition diagram can be utilized for capturing MoEs, constraint
expressions, and even required system parameters. The following figure displays the diagram that captures
the totalEnergyConsVCCS measure of effectiveness and the formula for calculating this MoE as the
constraint expression of the Total Energy Consumption constraint block.

152 | MagicGrid® Book of Knowledge

To empower the constraint expression to perform calculations, you have to specify what system
parameters should be consumed as variables of that constraint. Just like MoEs, system parameters can be
captured in the model as value properties of the block that captures the SoI or any of its subsystems, only
without any stereotype. The SysML parametrics diagram can be used to relate the system parameters and
MoEs to the constraint variables, with system parameters as inputs and MoEs as outputs. Binding
connectors (represented as solid lines in the following figure) should be used for this.

The simulation capabilities of the modeling tool allows you to calculate MoEs with given inputs. With the
MoE calculated, you can verify the appropriate non-functional system requirement and show whether it is
satisfied or not. The modeling tool enables you to perform this verification automatically. To get ready for
the automated requirements verification, you need to create a satisfy relationship from the value property,
which captures the MoE, to the appropriate system requirement. The following figure displays an example
of the satisfy relationship between the totalEnergyConsVCCS value property and the Max Energy
Consumption non-functional requirement, which means that the requirement is satisfied when the value of
the totalEnergyConsVCCS value property is not greater than 3.518 kWh. As you can see, the natural
language expression in the requirement text is refined by another constraint expression of the Max Energy
Consumption constraint block with the x <= 3.518 inequality.

MagicGrid® Book of Knowledge | 153

Although both satisfy and refine relationships are represented in the SysML requirements diagram, the
bdd can also be used for this purpose.

Tutorial

• Step 1. Organizing the model for system parameters

• Step 2. Specifying the MoE for capturing the total energy consumption
• Step 3. Defining a method for calculating the total energy consumption
• Step 4. Specifying system parameters for calculating the total energy consumption
• Step 5. Binding constraint parameters to corresponding value properties
• Step 6. Performing early system requirements verification
• Step 7. Storing values in the model

Step 1. Organizing the model for system parameters

As the system requirements specification defines, we need to specify the mathematical expression for
calculating the total energy consumed by the Vehicle Climate Control System. The constraint block which
captures this expression can be stored in a separate package within the 2 Solution Domain package.

154 | MagicGrid® Book of Knowledge

To organize the model for storing constraint blocks

1. Open the solution domain model (the VehicleCCS Solution.mdzip file) you created in Chapter System
Requirements, if not opened yet.
2. Right-click the 2 Solution Domain package and select Create Element.
3. In the search box, type pa (the first two letters of the element type Package), and press Enter.
4. Type 4 System Parameters to specify the name of the new package and press Enter.

 Even though we don't have the 3 System Behavior package (remember that it's not necessary to
model the system behavior in the LSA model), we skip “3” in package numbering to
correspond to the layout of the MagicGrid framework.

Step 2. Specifying the MoE for capturing the total energy consumption
As you can see in the following figure, the system requirements specification (see Chapter System
Requirements) and the appropriate MoE (see Chapter Measures of Effectiveness) in the problem domain
model determine the specification of the MoE that captures the total energy consumed by the VCCS, in the
solution domain model.

It was already mentioned that MoEs can be specified from scratch or inherited from the MoEs holder. The
following procedure describes inheritance from the MoEs holder. The inherited MoE is redefined to change
its name and type. Redefinition is necessary for automated requirements verification, which is performed
in one of the following steps.

MagicGrid® Book of Knowledge | 155

To specify the MoE for capturing the total energy consumed by the VCCS

1. Create a bdd for capturing system parameters:

a. Right-click the 4 System Parameters package you created in step 1 of this cell tutorial and
select Create Diagram.
b. In the search box, type bdd (the acronym of the SysML block definition diagram), and then
double-press Enter. The diagram is created.
c. Type System Parameters for Total Energy Consumption to specify the name of the new diagram and
press Enter again.

2. Display the MoEs Holder block on the diagram:

a. Press Ctrl + Alt + F to open the Quick Find dialog.

b. In the dialog, type moes.
c. When you see the MoEs Holder block selected in the search results list, press Enter. The block is
selected in the Model Browser.
d. Drag the MoEs Holder block to the diagram pane. The shape of the block is displayed on the
diagram.

3. Display the VCCS block on the diagram:

a. Press Ctrl + Alt + F to open the Quick Find dialog.

b. In the dialog, type vccs.
c. When you see the VCCS block selected in the search results list, press Enter. The block is selected
in the Model Browser.
d. Drag the VCCS block to the diagram pane. The shape of the block is displayed on the diagram, too.

156 | MagicGrid® Book of Knowledge

4. Establish the generalization between the VCCS block and the MoEs Holder block:

a. Select the shape of the VCCS block and click the Generalization button on its smart
manipulator toolbar.
b. Select the shape of the MoEs Holder block. The MoEs Holder block becomes the super-type of
the VCCS block, which means that the VCCS block inherits all MoEs from the MoEs Holder block.

5. Redefine the totalEnergyConsumption value property for the VCCS block:

a. Double click the VCCS block to open its Specification.

b. On the left of the open dialog, click Properties.
c. Select the totalEnergyConsumption value property and click the Redefine button below.

d. In the Type cell, immediately type kWh and press Enter. The kWh value type is created in the
model and set as type for the redefined value property.

MagicGrid® Book of Knowledge | 157

 e. In the Name cell, type totalEnergyConsVCCS to rename the redefined value property press Enter.

f. Click Close.
When you're done, the System Parameters for Total Energy Consumption bdd in your model should be
similar to the one below.

Step 3. Defining a method for calculating the total energy consumption

The next thing you need to do is capture the mathematical expression which specifies how to calculate a
few minutes ago specified MoE. It can be captured as the constraint expression of some constraint block
stored in your model.

Let’s say the total energy consumption of the VCCS equals a sum of energies consumed by all its
subsystems. The constraint expression can be defined as follows:

totalEnergyCons = energyCons1 + energyCons2 + energyCons3 + energyCons4

where:

1. totalEnergyCons stands for the total energy consumed by the VCCS

2. energyCons1 may stand for the energy consumed by the Cooling System
3. energyCons2 may stand for the energy consumed by the CP System
4. energyCons3 may stand for the energy consumed by the Heating System
5. energyCons4 may stand for the energy consumed by the HMI System
In terms of SysML, totalEnergyCons, energyCons1, ..., and energyCons4 are generally referred to as
constraint parameters and later should be bound to value properties.

158 | MagicGrid® Book of Knowledge

A bdd can be utilized for capturing constraint blocks. This constraint block is not an exception; you can use
the bdd created in step 2 of this cell tutorial.

To capture the method for calculating the total energy consumed by the VCCS

1. Create a constraint block:

a. Open the System Parameters for Total Energy Consumption diagram, if not opened yet.
b. Click the Constraint Block button on the diagram palette and then click an empty space in that
diagram pane. An unnamed constraint block is created.
c. Type Total Energy Consumption directly on the shape of that element to specify its name and press
Enter.

2. Specify a constraint expression to capture the above defined formula:

a. Click the symbol of the empty constraint expression on the Total Energy Consumption constraint
block.

b. Click the selection again. The empty constraint expression switches to the edit mode.

c. Type totalEnergyCons=energyCons1+energyCons2+energyCons3+energyCons4 and then press Enter.

3. Specify constraint parameters:

a. Select the shape of the Total Energy Consumption constraint block and click the Parse and Create
Parameters button on its smart manipulator toolbar (see the following figure). Constraint
parameters are automatically extracted from the constraint expression and displayed in
the parameters compartment box on the shape.

MagicGrid® Book of Knowledge | 159

 b. Select parameters one by one and change their types from the default Real to kWh directly on the
shape of the constraint block:

i. Double-click the type to select it, type kWh, and press Ctrl + Spacebar.
ii. Select the kWh type from the list below.
iii. Press Enter.

Once you’re finished with the last one, the shape of the Total Energy Consumption constraint block should
look like the one in the following figure.

Step 4. Specifying system parameters for calculating the total energy

consumption
With the constraint expression, you can easily determine what system or (in this case) subsystem
parameters are necessary for calculating the total energy consumed by the VCCS. Input parameters of the
Total Energy Consumption constraint block can help you with this task.

Remember that it has four input parameters, one each for the energy consumption of a single subsystem
of the VCCS, and requires four subsystem parameters in your model. Appropriate value properties should
be specified for each block that captures one of the subsystems of the VCCS.

160 | MagicGrid® Book of Knowledge

To specify subsystem parameters for capturing the energy consumed by each subsystem of the VCCS

1. Open the System Parameters for Total Energy Consumption diagram, if not opened yet.
2. Display the blocks capturing the logical subsystems of the VCCS block:

a. Right-click the VCCS block and select Display > Display Related Elements.
b. In the open dialog, click Outgoing and then click to clear the Abstraction check box.

c. Click OK to close the dialog. As a result, all the blocks representing the subsystems of the VCCS
are displayed on the diagram pane.

3. Create the energyCons value property for the Cooling System block:

a. Select the shape of the Cooling System block, click the Create Element button , and then select
Value Property.
b. Type energyCons:kW and press Ctrl + Spacebar to select the suitable value type.
c. Select the kWh value type from the list below and press Enter.

MagicGrid® Book of Knowledge | 161

 d. Press Enter again.

4. Repeat step 3 to create value properties for the CP System, Heating System, and HMI System blocks.

Step 5. Binding constraint parameters to corresponding value properties

Once you have all the necessary value properties in the model, it’s time to bind them to the appropriate
constraint parameters. Remember that it is the binding which empowers the constraint expression to
perform calculations on given inputs.

Bindings between value properties and constraint parameters can be established using binding connectors.
This type of SysML relationship can be created by utilizing the infrastructure of the SysML parametrics
diagram created for the VCCS block. However, creating the SysML parametrics diagram is not enough to
get started. Establishing binding connectors is possible only after the constraint property typed by
the Total Energy Consumption constraint block is created for the VCCS block. Be aware that the same
constraint block can be used in multiple contexts, and binding connectors are context-specific
relationships.

162 | MagicGrid® Book of Knowledge

To bind the constraint parameters to the corresponding value properties

1. Select the VCCS block in the Model Browser:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type vccs.
c. When you see the VCCS block selected in the search results list, press Enter.

2. Create a SysML parametrics diagram for the VCCS block:

a. Press Ctrl + N. The Create Diagram dialog opens.

b. In the search box, type par (the acronym for the SysML parametrics diagram), and press Enter.
The blank SysML parametrics diagram is created, and the Display Parameters/Properties dialog
opens.
c. Besides the totalEnergyConsVCCS value property (which is selected by default), select to display
four more value properties, each named energyCons and owned by the block which captures a
single subsystem of the VCCS.

 If you cannot see them, you might need to expand the related part properties.

d. Click OK. Selected value properties are displayed on the diagram pane, and the diagram is
selected in the Model Browser with the name edit mode switched on.

MagicGrid® Book of Knowledge | 163

  As you can see in the figure above, the deeply nested value properties are displayed in the
dot notation. If you want to switch to the nested notation, select all the shapes, except the
shape for the totalEnergyConsVCCS value property, then right-click them and select
Refactor > Convert to Nested Parts.

e. Type Total Energy Consumption of VCCS to name the diagram and press Enter.

3. Find the Total Energy Consumption constraint block:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type total en.
c. When you see the Total Energy Consumption constraint block selected in the search results list,
press Enter. The constraint block is selected in the Model Browser.

4. Drag the constraint block to the Total Energy Consumption of VCCS diagram. An unnamed constraint
property, typed by the Total Energy Consumption constraint block, is created for the VCCS block.
The Parametric Equation Wizard dialog opens concurrently.
5. Select constraint parameters (on the left side of the dialog) one by one and drag each to the
corresponding value property (on the right side of the dialog):

• energyCons1 to the energyCons value property of the Cooling System block

• energyCons2 to the energyCons value property of the CP System block
• energyCons3 to the energyCons value property of the Heating System block
• energyCons4 to the energyCons value property of the HMI System block

 • You may need to expand the structure on the right side of the dialog to see the value
property you need. Remember that they belong to the subsystem blocks, but not to
the system block directly.
• Once bound, the constraint parameter changes its color from red to grey.

164 | MagicGrid® Book of Knowledge

6. After you bind all the items, click OK. The dialog closes and all the bindings are displayed on the
diagram.

MagicGrid® Book of Knowledge | 165

Binding connectors can also be created from the smart manipulator toolbar of the value property (see the
following figure) or constraint parameter.

Now you can execute the model, provide input values, and see the result of the calculation. To start the
execution, click the Run button in the toolbar above the Total Energy Consumption of VCCS diagram.
When in the Variables tab of the Simulation panel (by default, it is on the right side of the Simulation
panel), you can specify the input values and see the output value (result of the calculation).

Step 6. Performing early system requirements verification

Having the concrete value of the value property that captures the MoE, you can verify the relevant system
requirement and show whether or not it is satisfied. The modeling tool enables you to perform this
verification automatically. To get ready for the automated requirements verification, you need to do the
following:

1. Formalize the constraint captured as the natural language phrase in the text of the particular system
requirement.
2. Indicate the value property whose value determines whether or not the system requirement is
satisfied.

In this case, you should formalize the Max Energy Consumption system requirement, which indicates that
“The system shall consume maximum 3.518 kWh." This requires creating another constraint block with a
constraint expression (this time the x <= 3.518 inequality), and making it refine the relevant system
requirement.

166 | MagicGrid® Book of Knowledge

To formalize the Max Energy Consumption system requirement

1. Create a separate diagram for the system requirements formalization:

a. Right-click the 1 System Requirements package and select Create Diagram.

b. In the search box, type rd (the acronym for the SysML requirements diagram), and then press
Enter. The diagram is created.
c. Type System Requirements Formalized to specify the name of the new diagram and press Enter
again.

2. Display the Max Energy Consumption requirement on the diagram pane:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type max en.
c. When you see the Max Energy Consumption requirement selected in the search results list, press
Enter. The requirement is selected in the Model Browser.
d. Drag the requirement to the newly created diagram pane.

3. Display the SRS for VCCS requirements on the diagram pane:

a. Right-click the Max Energy Consumption requirement and select Display > Display Related
Elements.
b. In the open dialog, click Outgoing, click to clear the DeriveReqt and Refine check boxes, and then
click to select the Containment check box.

c. Click OK.

MagicGrid® Book of Knowledge | 167

 4. Create the Max Energy Consumption constraint block:

a. In the diagram palette, click the Constraint Block button.

b. Click an empty place on the diagram pane.
c. Immediately type Max Energy Consumption to name the newly created element and then press
Enter.
d. Specify the x <= 3.518 inequality as the constraint expression:

i. Click the symbol of the empty constraint expression on the constraint block.
ii. Click the selection again. The empty constraint expression switches to the edit mode.
iii. Type x <= 3.518 and then press Enter.

e. Select the shape of the Max Energy Consumption constraint block and click the Parse and Create
Parameters button on its smart manipulator toolbar. The x constraint parameter is
automatically extracted from the constraint expression and displayed in
the parameters compartment box on the shape.

f. Select the parameters and change their types from the default Real to kWh directly on the shape
of the constraint block:

i. Double-click the type to select it, type kWh, and press Ctrl + Spacebar.
ii. Select the kWh type from the list below it.
iii. Press Enter.

5. Specify the refine relationship between the Max Energy Consumption constraint block and the relevant
system requirement:

a. Select the shape of the constraint block and click the Refine button on its smart manipulator
toolbar.
b. Click the shape of the Max Energy Consumption system requirement. The refine relationship
conveying that the textual system requirement is formalized by the Max Energy
Consumption constraint block is created.

168 | MagicGrid® Book of Knowledge

  The modeling tool enables you to create the constraint block automatically by parsing the
requirement's text. First, you must update the project options.

To update project options

1. From the main menu, select Options > Project.

2. In search box of the the Project Options dialog, type glossary.
3. Click to set the Use Requirement Terms Glossary option to true.
4. Click OK to close the dialog.

As a result, the word maximum is underlined on the shape of the Max Energy Consumption
system requirement. If you hover the mouse over that text, you can see the inequality operator,
which is automatically parsed from the word maximum.

The x <= 3.518 inequality is stored in the model as the constraint expression of the virtual constraint
block, which can be extracted on demand, as this is not necessary for automated requirements
verification; the virtual constraint expression is enough.

To extract the constraint block from the Max Energy Consumption system requirement

1. Right-click the requirement and select Tools > Extract Constraint Block From Requirement. The
new constraint block is created in the model and automatically related to the Max Energy
Consumption system requirement.
2. Drag the Max Energy Consumption constraint block onto the diagram pane. The shape of the
constraint block and the refine relationship pointing from that constraint block to the Max
Energy Consumption system requirement are displayed on the diagram.

Next, specify that values of the totalEnergyConsVCCS value property can be used to determine whether or
not the Max Energy Consumption system requirement is satisfied. For this, you need to bind the
x parameter of the Max Energy Consumption constraint block to the totalEnergyConsVCCS value property.
This can be done in the Total Energy Consumption of VCCS parametrics diagram.

MagicGrid® Book of Knowledge | 169

To bind the x parameter of the Max Energy Consumption constraint block to the totalEnergyConsVCCS value
property

1. Open the Total Energy Consumption of VCCS parametrics diagram.

2. Drag the Max Energy Consumption constraint block to the diagram pane.
3. In the open dialog, select the x constraint parameter on the left and drag it to the totalEnergyConsVCCS
value property on the right.

4. Click OK. The Total Energy Consumption of VCCS parametrics diagram is updated with the new
constraint property for verification of the Max Energy Consumption system requirement.

170 | MagicGrid® Book of Knowledge

  If the Use Requirement Terms Glossary option is set to true in your project, you don't need to
update the Total Energy Consumption of VCCS parametrics diagram. It will be enough to establish a
satisfy relationship pointing from the totalEnergyConsVCCS value property to the Max Energy
Consumption system requirement.

To create a satisfy relationship from the totalEnergyConsVCCS value property to the Max Energy
Consumption system requirement

1. Display the VCCS block on the System Requirements Formalized diagram:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type vccs.
c. When you see the VCCS block selected in the search results list, press Enter. The block is
selected in the Model Browser.
d. Drag the block to the diagram pane.

2. Select the totalEnergyConsVCCS value property (not the entire block!).

3. Click the Satisfy button on the smart manipulator toolbar of that value property.
4. Click the Max Energy Consumption system requirement.

Now, if you hover the mouse over the underlined word maximum on the shape of the Max Energy
Consumption system requirement, you can see the complete condition that must be fulfilled to
satisfy that system requirement.

If you go back to the Total Energy Consumption of VCCS diagram and execute the model again, you should
note that the totalEnergyConsVCCS value property is highlighted in green, indicating that the Max Energy
Consumption system requirement is satisfied. Try to change the input values and see how the highlight

MagicGrid® Book of Knowledge | 171

color changes when the requirement is not satisfied. This is how you perform the automated requirements
verification. Moreover, this way you can perform manual or automatic (it requires an algorithm to be
defined) parameters optimization in order to find the optimal system configuration.

Step 7. Storing values in the model

System or subsystem parameter values you specify in the Variables tab of the Simulation panel, as well as
the calculated MoE value you see displayed there, can be stored in your model. SysML instance
specifications and their slot values can be used for this. Once you have the combination of inputs and the
output you would like to keep, you can save them as slot values of the instance of the relevant system
architecture blocks (this time, the blocks of the VCCS and its subsystems). Keep in mind that each time you
save the instance specifications of these blocks, you are capturing a snapshot of the system at run-time.
You can take as many snapshots as you want at different points in time.

For keeping your model well organized, it is recommended to store the instance specifications and their
slots in a separate package. They can be displayed in a special type of table: the instance table. It provides
a convenient layout for reviewing slot values and recalculating them. Moreover, new instances with
corresponding slots can be created here.

172 | MagicGrid® Book of Knowledge

To create an instance of the VCCS block

1. In the Variables tab of the Simulation panel, select the root element; that is, the VCCS block.

2. Click the Export to New Instance button in the toolbar above.

3. In the Select Owner dialog, select the 4 System Parameters package (you may need to expand the 2
Solution Domain package beforehand).
4. Click the Create Owner button at the bottom of the dialog and then select Package.

 If you don't see the Create Owner button, click to toggle the Creation Mode button. As a
result, the Create Owner button appears in the dialog.

5. In the Specification of the new package, type 1 Parameter Values to give it a name.
6. Click Close in the Specification of that package.
7. Click OK in the Select Owner dialog. The instance of the VCCS block with slots holding concrete values
of relevant value properties is created.

MagicGrid® Book of Knowledge | 173

As you can see in the preceding figure, the Model Browser is not the best place for reviewing instances,
especially when you have more than a few. In that case, the infrastructure of the instance table should be
utilized.

To display the instance of the VCCS block in the instance table

1. Create the instance table:

a. In the Model Browser, right-click the 1 Parameter Values package and select Create Diagram.
b. In the search box, type ins. When you see the instance table selected in the search results, press
Enter. The table is created.
c. Type Total Energy Consumption Values to specify the name of the new table and press Enter again.

2. In the Model Browser, under the same package, select the root instance; that is, the one typed by the
VCCS block.

3. Drag the selected instance to the instance table on the right. As a result:

a. The instance of the VCCS block is displayed in the table.

b. The VCCS block, as it types the selected instance, is automatically set as a classifier of the
instance table. You can see it in the Criteria area above the table.

4. Hide all the columns except the Name:

a. Right-click any column.

b. Select Hide.
c. Repeat steps a and b to hide other unneeded columns.

5. Display the columns to show relevant slot values:

a. On the toolbar above the instance table, click the Columns button and select Select Columns.
b. In the open dialog, click to select:

174 | MagicGrid® Book of Knowledge

 i. All the energyCons value properties (you may need to expand the appropriate part
properties beforehand). These are the slots' input values.
ii. The totalEnergyConsVCCS value property. This is the slot for the output value.

c. Click OK to close the dialog.

When you're done, the instance table in your model should look the same as the one in the following
figure.

Now you can create new rows to display other instances with slots for diverse input and output values.

To create a new instance specification of the VCCS block in the instance table

1. On the toolbar above the instance table, click the Create (Insert) button and then select Create
with Parts (see the following figure). This is necessary because you also need to create slots of value
properties that belong to the parts of the VCCS block, but not to this block directly. A row of the new
instance with all slot values 0 is created.

2. In the newly created row, specify the input values.

MagicGrid® Book of Knowledge | 175

 3. When you're done, on the toolbar above the instance table, click the Run button and then select
Evaluate Selected Rows. The result is calculated.

 Note that output values violating the Max Energy Consumption system requirement are
highlighted. If you cannot see the highlights in your model, make sure the
SimulationProfile.mdzip is used in your model. For more information on how to use another
project or profile in your project, refer to the documentation of the modeling tool.

The contents of the instance table can be exported to the MS Excel spreadsheet to share it internally
or externally.

System Parameters done. What's next?

With this cell, the task of building the LSA model of the SoI is almost finished. All that is left to do is
establish traceability relationships from the LSA model to system requirements specifications (see Chapter
Traceability to System Requirements).

176 | MagicGrid® Book of Knowledge

Traceability to System Requirements
What is it?

To have a complete LSA model, you need to establish traceability relationships from the elements that
capture the LSA of the SoI to the elements that capture the system requirements. As determined by
SysML, these are satisfy relationships, meaning that the element from the LSA model fulfills one or more
of the system requirements. For example, a satisfy relationship can be established from the part property
typed by the block that captures some logical subsystem to a certain system requirement. In the complete
LSA model, all system requirements are satisfied by one or more elements of the logical system
architecture.

Who is responsible?

Traceability relationships in the LSA model can be established by the Systems Architect or Systems
Engineer. He/she must be the head of the whole systems engineering project, the one, who has “a big
picture view” of the SoI and is responsible for the smooth integration of the subsystems and their
components into the whole.

How to model?

Neither of the SysML diagrams is suitable for creating a mass of cross-cutting relationships, such
as satisfy. For this, a matrix is more suitable. The modeling tool offers a wide variety of predefined
matrices for specifying cross-cutting relationships. To capture satisfy relationships, a Satisfy Requirement
Matrix can be used.

Tutorial
• Step 1. Creating a matrix for capturing satisfy relationships
• Step 2. Capturing satisfy relationships

MagicGrid® Book of Knowledge | 177

Step 1. Creating a matrix for capturing satisfy relationships
In this step, we create a Satisfy Requirement Matrix and specify its criteria. System requirements can be
displayed in the rows of the matrix, and the elements that capture the LSA of the SoI can be displayed in
its columns.

To create a Satisfy Requirement Matrix

1. In the Model Browser, right-click the 5 Traceability package (you created it in Chapter System
Requirements) and select Create Diagram.
2. In the search box, type srm, the acronym of the predefined Satisfy Requirement Matrix, and press
Enter.

 If you don’t see any results, click the Expert button below the search results list. The list of
available diagrams expands in the Expert mode.

3. Type LSA to System Requirements to specify the name of the new matrix and press Enter again.
4. On the matrix toolbar, click the Change Axes button to swap rows of the matrix with columns.
Requirement becomes the type of the matrix rows.
5. Select the SR-1 SRS for VCCS requirement (for this you might need to expand the 1 System
Requirements package first) and drag it onto the Row Scope box in the Criteria area above the matrix
contents. The SR-1 SRS for VCCS requirement becomes the scope of the matrix rows (see the following
figure).
6. Select column element types:

a. In the Model Browser, select any part property, proxy port, and value property.

 To select the set of non-adjacent items in the tree, click the first one, press Ctrl, and while
holding it down, select other items one by one.

b. Drag the selection onto the Column Type box in the Criteria area above the matrix contents. The
columns of the matrix are set to display all the part properties, proxy ports, and value
properties (see the following figure) within the scope you select in the next step.

7. Select column element scope:

a. In the Model Browser, select the VCCS block (you might need to expand the 1 System
Structure package beforehand).
b. Drag the selection onto the Column Scope box in the Criteria area above the matrix contents. The
selected block becomes the scope of the matrix rows (see the following figure).

178 | MagicGrid® Book of Knowledge

Once you're done with the criteria, the contents of the matrix is updated. All the cells are empty at the
moment, except the one at the intersection of the totalEnergyConsVCCS value property and the SR-1.9 Max
Energy Consumption requirement, if you have created the satisfy relationship in step 6 of the System
Parameters cell tutorial.

Step 2. Capturing satisfy relationships

Now you’re ready to specify other satisfy relationships. In the matrix, this can be done very easily.

To specify a satisfy relationship from the p1 : UserDataInput proxy port to the SR-1.1 Setting Desired
Temperature requirement

• Double-click the cell at the intersection of the column that displays the proxy port and the row that
displays the requirement. The satisfy relationship is created between the appropriate items and
displayed in the cell.

MagicGrid® Book of Knowledge | 179

We assume that you continue filling in the matrix yourself. If you need a hint for which elements can be
related, look at the following figure.

LSA done. What's next?

Once you're done with the LSA, you have the following:
• Logical subsystems
• Logical interfaces
• Exchange items

Having this information, you are ready to specify the solution architecture of each logical subsystem.

180 | MagicGrid® Book of Knowledge

Building the logical architecture of subsystems
Once the systems architect identifies the logical subsystems of the SoI, he/she can distribute them to
separate engineering teams to begin working on logical subsystem architectures (LSSA). Therefore, this
phase of creating the solution domain model is devoted to building the architectures of all logical
subsystems of the SoI. In this phase you should imagine being a member of the engineering team working
on the LSSA of a single logical subsystem of the VCCS.

As noted previously, an LSSA of each logical subsystem should be stored in a separate model, apart from
the problem domain definition and LSA model. The highlighted area in the following figure shows the part
of the solution domain we are talking about. As you can see, these models can be developed in parallel.

A complete LSSA model consists of the specification of subsystem requirements, structure, behavior, and
parameters. Additionally, every SysML element you create in this model must satisfy at least one
subsystem requirement; otherwise, there is no motivation for having it in the LSSA model. Therefore, you
have to start and end the LSSA model with subsystem requirements: first specify them and, when your
model is complete, establish traceability relationships between them and SysML elements.

MagicGrid® Book of Knowledge | 181

A logical subsystem in this phase of building the solution domain model can be referred to as the
subsystem of interest. Moreover, from the point of view of the appointed engineering team, it is referred
to as the system of interest.

Note that the following material mainly focuses on building the logical architecture of a single subsystem
of interest. This is the Cooling System of the VCCS. You should keep in mind that while you build the
logical architecture for the Cooling System, other engineers / engineering teams will work in parallel to
build the logical architectures of the CP System, Heating System, and HMI System.

Subsystem Requirements
What is it?

To provide the solution architecture of the logical subsystem, the responsible engineering team needs to
analyze the relevant subsystem requirements specification, which is produced when having the complete
LSA model. Each subsystem requirement from the leaf level of the specification, should point either to a
particular system requirement or an element from the LSA model (e.g., value property or proxy port of the
block which represents the relevant logical subsystem). This proves the necessity of the subsystem
requirement (just like system requirements point to stakeholder needs or problem domain model
elements).

After you are done with the LSSA, you should establish satisfy relationships from the LSSA elements to
subsystem requirements in order to assert which LSSA elements fulfill which subsystem
requirements. This is described in Chapter Traceability to Subsystem Requirements.

Who is responsible?

The subsystem requirements specification can be produced by the Systems Architect or Systems Engineer,
who has built the LSA model.

182 | MagicGrid® Book of Knowledge

How to model?

In the modeling tool, a single subsystem requirement can be captured as the SysML requirement. Just like
a system requirement, the requirement for a subsystem must have a unique identification, name, and
textual specification, which must be formal and written by following certain guidelines (examples of such
guidelines are provided in Chapter System Requirements).

As with system requirements, subsystem requirements can be categorized and organized into hierarchies.
A SysML requirement diagram or table can be used to capture and display them.

Matrices and maps can be used to establish and review the cross-cutting relationships between the
subsystem requirements and system requirements or LSA model elements.

MagicGrid® Book of Knowledge | 183

Tutorial
• Step 1. Creating and organizing a model for subsystem requirements
• Step 2. Creating a diagram for subsystem requirements
• Step 3. Specifying subsystem requirements
• Step 4. Specifying traceability relationships

Step 1. Creating and organizing a model for subsystem requirements

To begin specifying subsystem requirements, you need to establish an appropriate structure of the
model. Following the layout of the MagicGrid framework, model artifacts that capture subsystem
requirements should be stored under the structure of packages, as displayed in the figure below. The
upper-level package represents the domain, and the lower-level package stands for the cell.

184 | MagicGrid® Book of Knowledge

However, this is not exactly what we are going to create. In real-world projects, the 2.1 Cooling System
Solution Domain package together with its subpackage (and all subsequently added ones) would appear in
another model separate from the LSA model. This is necessary because the engineering team appointed to
build the LSSA model should only be able to read it, and not make any changes in the LSA model.

Therefore, before organizing packages for building the LSSA of the Cooling System, you need to create a
new model and then, in terms of the modeling tool, use the LSA model in it. It should be noted here, that
an LSSA model should be created for each logical subsystem identified in the LSA.

To create a new model for the LSSA of the Cooling System and organize it to capture the subsystem
requirements specification

1. Start sharing the LSA model. Once you share it, the problem domain analysis model used in the LSA
model becomes shared, too.

 For more information on this topic and the topics mentioned in steps 2 and 3, refer to
the latest documentation of your modeling tool.

2. Create a new model. This is the solution architecture model of the Cooling System, so it can be
named Cooling System Solution.mdzip.
3. Use the LSA model as read-only in the LSSA model of the Cooling System. The problem domain
analysis model is automatically used as well.

 Names in grey indicate elements that cannot be modified, as the problem domain model is
used in the solution domain model as read-only.

4. Right-click the Model package (this is the default name of the root package) and select Create
Element.
5. In the search box, type pa, the first two letters of the element type Package, and press Enter.
6. Type 2.1 Cooling System Solution Domain to specify the name of the new package and press Enter.

 Let's agree that packages for storing the solution architectures of other logical subsystems
(such as the CP System and Heating System, among others) should include 2.2, 2.3, and so on
in their names.

7. Right-click the 2.1 Cooling System Solution Domain package and select Create Element.
8. In the search box, type pa, the first two letters of the element type Package, and press Enter.

MagicGrid® Book of Knowledge | 185

 9. Type 1 Subsystem Requirements to specify the name of the new package and press Enter.

Step 2. Creating a diagram for subsystem requirements

For capturing and displaying subsystem requirements, we recommend using the SysML requirement
diagram or table. Let's create a diagram this time.

To create the SysML requirement diagram for specifying subsystem requirements

1. Right-click the 1 Subystem Requirements package and select Create Diagram.

2. In the search box, type rd, where r stands for requirement and d for diagram, and then double-press
Enter. The diagram is created.

 Note that the diagram is named after the package where it is stored. This name also works for
the diagram. You may only remove the sequence number from its name.

Step 3. Specifying subsystem requirements

Let's say that the engineering team working on the solution architecture of the Cooling System receives
from the System Architect the following subsystem requirements:

# Name Text
1 Air for Cooling The Cooling System shall be able to draw the air out of the cabin and blow back the
cooled air.

186 | MagicGrid® Book of Knowledge

 # Name Text
2 Cooling Approach The Cooling System shall be able to compress, condense, and evaporate the
refrigerant.
3 Cooling System The main working states of the Cooling System shall be Off and On. Being in the On
Working States state, the Cooling System shall visit the Initializing, Idling, and Operating sub-states.
4 Max Energy The Cooling System shall consume maximum 1.2 kWh.
Consumption for
Cooling
5 Internal Communication The Cooling System shall get control signals from the Central Processing (CP) System
send back the response.
6 Initialization The Cooling System shall be initialized by the CP System. If the initialization
succeeds, the Cooling System shall transit to the Idling state and wait there for the
signal to start cooling.
7 Idling The Cooling System shall go to the Idling state either after the successful
initialization or when the CP System requires it to stop cooling.

8 Operating While working the Cooling System shall cool down the cabin air until the CP System
requires to stop.

As you already know how to capture requirements in the diagram, we assume that you can do this on your
own. If you need to refresh your knowledge, refer to the procedure in step 3 of the System
Requirements tutorial. When you are done, the Subsystem Requirements diagram in your model should look
very similar to the one below.

You also need to group, number, and categorize the subsystem requirements. These topics are described
in step 4, step 5, and step 6 of the Stakeholder Needs tutorial, so we assume that you can do this on your
own, too. As you can see in the following figure, subsystem requirements should be:
• Grouped under the SRS for Cooling System grouping requirement (if you need a reminder on how to
display relationships on the diagram, refer to the procedure in step 3 of the System
Requirements tutorial)
• Numbered using the SSR- prefix
• Categorized into interface, functional, non-functional, and physical requirements

MagicGrid® Book of Knowledge | 187

This flat structure of subsystem requirements can be organized into a deeper hierarchical structure, as
shown in the figure below.

Step 4. Specifying traceability relationships

As mentioned before, the necessity of any subsystem requirement must be proved. That is, the subsystem
requirement can be created either to derive a system requirement or refine an LSA model element. In the

188 | MagicGrid® Book of Knowledge

SysML model, the deriveReqt relationship is used to assert the derivation, and the refine relationship to
assert the refinement. Both types of relationships can be specified either in a SysML requirements
diagram or a matrix. We recommend using the matrix for a larger scope, because it provides a more
compact view of the model than the diagram does.

In this case, we recommend utilizing the infrastructure of one of the predefined matrices in the modeling
tool: either the Derive Requirement Matrix or the Refine Requirement Matrix. Whatever matrix you choose,
unlike in other cases, you need to update the dependency criteria to set up the matrix criteria. Hence, if it
is the Derive Requirement Matrix, you need to set the refine relationship as dependency criteria in addition
to the deriveReqt relationship, and if it is the Refine Requirement Matrix, you need to set the deriveReqt
relationship as dependency criteria in addition to the refine relationship. The following figure displays how
the dependency criteria in your matrix should look like.

The following procedure shows how to create only one of the matrices, the Derive Requirement Matrix,
but you can create the alternative one on your own and see which one suits better for you.

As in other models, traceability views and relationships should be stored in a separate package. Because
you already know how to create packages, we assume that you can do this on your own. For keeping your
model well-organized according to the MagicGrid framework, that package should always be the last sub-
package of the 2.1 Cooling System Solution Domain package. Therefore, its name must begin with number
5 (for example, 5 Traceability).

To create a Derive Requirement Matrix

1. In the Model Browser, right-click the 5 Traceability package and select Create Diagram.
2. In the search box, type drm, the acronym for the predefined Derive Requirement Matrix, and press
Enter.

 If you don’t see any result, click the Expert button below the search results list. The list of
available diagrams expands in the Expert mode.

3. Type Subsystem Requirements to System Requirements and LSA to specify the name of the new matrix
and press Enter again.
4. In the Model Browser, select the SSR-1 SRS for Cooling System requirement and drag it onto the Row
Scope box in the Criteria area above the matrix contents. The SSR-1 SRS for Cooling
System requirement becomes the scope of the matrix rows.
5. Select the column element types:

a. Holding down the Ctrl key, in the Model Browser, select any part property, proxy port, and value
property.
b. Drag the selection onto the Column Type box in the Criteria area above the matrix contents. The
columns of the matrix are set to display all the part properties, proxy ports, and value properties
in addition to requirements, which is set by default, within the scope you select in the next step.

6. Select column element scope:

a. In the Model Browser, select the SR-1 SRS for VCCS requirement (for this you might need to
expand the following packages beforehand: 2 Solution Domain and 1 System Requirements).

MagicGrid® Book of Knowledge | 189

 b. Press the Ctrl key and holding it down, select the Cooling System block (for this you might need to
expand the 1 System Structure package first).
c. Drag the selection onto the Column Scope box in the Criteria area above the matrix contents.
The SR-1 SRS for VCCS requirement and the Cooling System block become the scope of the matrix
columns.

7. Set the refine relationship as dependency criteria:

a. Click the button next to the Dependency Criteria box.

b. On the left side of the open dialog, select Simple Navigation.
c. In the search box at the bottom of the dialog, type refine.
d. Select the Refine [Abstraction] (SysML) and click to set the Is Applied check box value to true.

 It is also worth to change the style of the Refine relationship so it differs from the
DeriveReqt relationship in the matrix (see the following figure). For more information, refer
to the documentation of the modeling tool.

e. Click OK to close the dialog. Together with the deriveReqt relationship, the refine relationship is
set as dependency criteria for the matrix.

8. Click the Direction box and select Row to Column to enable the creation of relationships pointing from
subsystem requirements, which are row elements, to system requirements of LSA elements, which
are column elements.
9. Click the Refresh hyperlink in the notification box below the Criteria area. The contents of the matrix
are updated. All the cells are empty at the moment.
The following figure displays the Criteria area of the matrix, with highlights on step 4, 5, 6, 7, and 8
related criteria.

Now you’re ready to establish both derivation and refine relationships. Let’s create the derivation between
SSR-1.1.1 Air for Cooling and SR-1.4 Cabin Air. This will convey that the SSR-1.1.1 Air for Cooling subsystem
requirement is created because of the SR-1.4 Cabin Air system requirement.

190 | MagicGrid® Book of Knowledge

To specify the derivation relationship in the matrix

1. Double-click the cell at the intersection of the row that displays SSR-1.1.1 Air for Cooling and the
column that displays SR-1.4 Cabin Air.

 If you want to read the requirement text before setting the relationship, just hover the mouse
over that requirement name. The text will appear in the tool tip.

2. Select DeriveReqt to create a derivation relationship. The derivation relationship is created between
the appropriate items and displayed in the cell.

After all derivation relationships are established, the Subsystem Requirements to System Requirements and
LSA matrix in your model should resemble the one in the following figure.

As you can see, the SSR-1.2.1 Internal Communication subsystem requirement has no derivation to any
system requirement. This is because the SSR-1.2.1 Internal Communication subsystem requirement was

MagicGrid® Book of Knowledge | 191

created to refine the p3 and p4 proxy ports of the Cooling System block. Let's specify appropriate
relationships in the matrix.

To specify the refine relationship in the matrix

• Double-click the cell at the intersection of the column that displays the p3 or p4 proxy port and the
row that displays the SSR-1.2.1 Internal Communication subsystem requirement. The refine
relationship is created between the appropriate items and displayed in the cell.

Although derived from the system requirement, the SSR-1.1.1 Air for Cooling subsystem requirement also
refines the p1 and p2 proxy ports of the Cooling System block from the LSA model. We assume, you create
appropriate relationships on your own.

192 | MagicGrid® Book of Knowledge

Afterwards, a Requirement Derivation Map can be created, with blue arrows representing the derivation
relationships between requirements from different domains and green arrows the containment.

Subsystem Requirements done. What's next?

When you have the requirements for a particular logical subsystem of the SoI, you can build the logical
solution architecture of that subsystem. It begins with specifying the subsystem structure, which means
you can move to the Subsystem Structure cell.

MagicGrid® Book of Knowledge | 193

Subsystem Structure
What is it?

You can begin building the solution architecture of a subsystem by defining the internal structure of that
subsystem and specifying how its components operate together, and how they interact with other
subsystems of the same SoI.

As you may recall, the interfaces for interacting with other subsystems are defined in the LSA model by
the system architect. These are the constraints the engineering team appointed to build the solution
architecture of a single subsystem have to deal with. This ensures the model integrity of the whole SoI.

If changes are required (for example, an additional interface is necessary, or an existing interface is
acknowledged as not needed) the engineering team applies to the systems architect, who considers the
request. If the request is accepted, other engineering teams must be informed about the change to ensure
integrity at the system level.

Who is responsible?

The structure of the appointed logical subsystem can be captured by the Systems Architect, or the
Systems Engineer who belongs to the engineering team responsible for building the solution architecture
of that logical subsystem.

How to model?

In the modeling tool, the components of the subsystem can be defined by using the infrastructure of the
SysML block definition diagram (bdd). Components themselves can be captured as blocks, and interfaces
as proxy ports typed by appropriate interface blocks.

194 | MagicGrid® Book of Knowledge

To specify how the parts of the subsystem operate together and outside of the subsystem, as well as what
material they exchange, the infrastructure of the SysML internal block diagram can be used. The internal
structure and interactions between different parts of the Cooling System in the form of the SysML internal
block diagram is displayed in the following figure.

Tutorial
• Step 1. Organizing the model for subsystem structure
• Step 2. Getting ready to capture the subsystem structure
• Step 3. Capturing components of the Cooling System
• Step 4. Creating an ibd for specifying interactions
• Step 5. Specifying interactions with the outside
• Step 6. Specifying interactions within the Cooling System

Step 1. Organizing the model for subsystem structure

The LSSA of the Cooling System of the VCCS can be captured in the model you created in
Chapter Subsystem Requirements. To begin modeling the subsystem structure, you need to establish an
appropriate structure of the model beforehand. Following the design of the MagicGrid framework, relevant
model artifacts should be stored under the structure of packages, as displayed in the following figure.

MagicGrid® Book of Knowledge | 195

To organize the model for subsystem structure

1. Open the solution domain model of the Cooling System (the Cooling System Solution.mdzip file) you
created in step 1 of the Chapter Subsystem Requirements tutorial, if it is not opened yet.
2. Right-click the 2.1 Cooling System Solution Domain package and select Create Element.
3. In the search box, type pa, the first two letters of the element type Package, and press Enter.
4. Type 2 Subsystem Structure to specify the name of the new package and press Enter.

To keep the inner structure of the 3 System Structure package well organized, you need to create a few
more packages. These are:
• 1 Logical Interfaces
• 2 Exchange Items
• 3 Logical Components

When you create them on your own (by following the previous procedure), the Model Browser of your
solution domain model of the Cooling System should look the same as the following figure.

Step 2. Getting ready to capture the subsystem structure

Remember that the engineering team of the appointed logical subsystem must take into consideration all
the interfaces of that subsystem, as determined by the system architect in the LSA model. Therefore,
those interfaces must exist in the solution domain model of that logical subsystem.

Although the engineering team of the subsystem can read the LSA model, they cannot make any changes
there. The first thing they should do to get ready to model the subsystem structure is create a block that
captures the particular subsystem of interest (in this case, the Cooling System) and enable it to inherit all
the relevant interfaces from the LSA model.

The inheritance can be achieved by establishing a generalization relationship between two blocks
representing the Cooling System: the one in the LSA model (as the super-type) and the other in the LSSA
model of the Cooling System (as the sub-type). As a result, the sub-type block inherits all the
specifications (including proxy ports typed by appropriate interfaces) from the super-type block. It's
important to note here that inherited proxy ports should be redefined; otherwise, certain tasks of building
the LSSA cannot be accomplished; for example, specification of satisfy relationships from proxy ports to
relevant subsystem requirements.

The generalization relationship can be created by utilizing the infrastructure of a bdd, which can be created
under the 2 Subsystem Structure package within the solution domain model of the Cooling System (created
in step 1 of this cell tutorial).

196 | MagicGrid® Book of Knowledge

To get ready to capture the structure of the Cooling System

1. Create the bdd:

a. Right-click the 2 Subsystem Structure package and select Create Diagram.

b. In the search box, type bdd (the acronym for SysML block definition diagram), and then press
Enter. The diagram is created.
c. Type Cooling System Structure to specify the name of the new diagram and press Enter again.

2. Create the block of the subsystem of interest:

a. Click the Block button on the diagram palette and then click on the diagram pane.
b. Type Cooling System Architecture to name the block and press Enter.

3. Display the Cooling System block from the LSA model on the diagram:

a. Select the Cooling Systemblock in the Model Browser (you may need to subsequently expand the2
Solution Domain, 2 System Structure, and the 3 Logical Subsystemspackagesfirst).

b. Drag the selection to the diagram pane. The shape of the Cooling System block is displayed on the
diagram.

4. Draw a generalization between the Cooling System block (super-type) and the Cooling System
Architecture block (sub-type):

a. Select the shape of the Cooling System Architecture block.

b. Click the Generalization button on the smart manipulator toolbar of the selected block.

MagicGrid® Book of Knowledge | 197

 c. Select the shape of the Cooling System block.

5. Redefine the inherited proxy ports:

a. Double-click the shape of the Cooling System Architecture block on the diagram.
b. On the left of the open dialog, click Ports/Interfaces.
c. Select the proxy port in the first row and click the Redefine button below.

As a result, the selected proxy port is redefined and since that moment no longer decorated with
the caret symbol (“^”), which indicates the inheritance. Don't change anything.

198 | MagicGrid® Book of Knowledge

 d. Row by row, redefine the rest (the order of items may differ from what you see in the following
figure).

e. Click Close.

6. Display the proxy ports and interfaces of the Cooling System Architecture block on its shape:

a. Select the shape of the Cooling System Architecture block on the diagram pane.
b. Click the Display All Ports button on its smart manipulator toolbar.

The proxy ports and interfaces are displayed.

MagicGrid® Book of Knowledge | 199

Step 3. Capturing components of the Cooling System
As defined by the subsystem requirements, the Cooling System shall be composed of:
• Compressor
• Condenser
• Evaporator

Based on their experience, the appointed engineering team identifies a few more components:
• Receiver Dryer
• Metering Device

All these items can be captured using either bdd or ibd, although we use the Cooling System Structure bdd
(created in step 2 of this cell tutorial). The ibd is useful later, when you need to represent interactions
between the specified components.

To capture the components of the Cooling System

1. Open the Cooling System Structure bdd, created in step 2 of this cell tutorial, if not opened yet.
2. Select the Cooling System Architecture block, click the Directed Composition button on its smart
manipulator toolbar (see the following figure), and then click a free space on the diagram pane.

Another block is created in the Model Browser and its shape is selected on the diagram pane.

200 | MagicGrid® Book of Knowledge

3. Type Compressor directly on the selected shape to specify the name of the appropriate component of
the Cooling System, and press Enter.

4. Repeat steps 2 and 3 as many times as you need to create the Condenser, Evaporator, Receiver Dryer,
and Metering Device blocks.

MagicGrid® Book of Knowledge | 201

 5. In the Model Browser, select all the blocks of components and drag them to the 3 Logical Components
package you created in step 1 of this cell tutorial. The blocks are moved to that package.

Step 4. Creating an ibd for specifying interactions

Remember that a bdd works best to capture components, while an ibd is helpful when you need to specify
interactions between these components. That’s why we need to create an ibd for the Cooling System
Architecture block. To begin specifying the interactions, you need to represent all proxy ports and
components of the Cooling System in that diagram. They have all been captured using the Cooling System
Structure bdd.

To create an ibd for specifying interactions of the Cooling System

1. In the bdd Cooling System Structure, select the Cooling System Architecture block and click the SysML
Internal Block Diagram button on its smart manipulator toolbar (see the following figure). The ibd
is created and the Display Parts/Ports dialog opens on top of it.

202 | MagicGrid® Book of Knowledge

2. In the open dialog, select all the part properties and the first two proxy ports (see the following
figure). Click the OK button.

 It is not necessary to capture all communications in the same diagram. Therefore, we will skip
showing a couple of proxy ports, as they are not relevant to the aspect we are going to model
in the newly created diagram.

3. In the Model Browser, where the newly created diagram is automatically selected in the edit mode,
type Cooling System Logical Components to specify its name and press Enter again.

MagicGrid® Book of Knowledge | 203

 4. Shorten the names of part properties to make their shapes more compact (don't remove the names;
you might need them later).
5. Rearrange the shapes on the diagram, as shown in the following figure.

Step 5. Specifying interactions with the outside

In this step, we focus on specifying interactions between components of the Cooling System and outside
it; that is, other subsystems of the VCCS or even other systems of the whole vehicle. As you already know,
interactions can be specified as connectors established via compatible proxy ports. Information, matter, or
energy exchanged throughout these interactions can be specified as items flowing over relevant
connectors.

We can start from specifying the interactions via the inherited interfaces (the ones identified in the LSA).
There are two of them in this diagram: one for letting the air into the Cooling System, and the other is for

204 | MagicGrid® Book of Knowledge

letting it out. Let's specify that the Compressor consumes the air from the cabin, and the Evaporator (its
fans actually) supplies the cooled air back into the cabin.

To specify that the Compressor consumes the air from the cabin

1. Select the p1 proxy port (typed by the insideAirInlet interface block) on the diagram frame and click
the Connector button on its smart manipulator toolbar.
2. Click the shape of the cmp part property (typed by the Compressor block).
3. Select New Proxy Port. A compatible proxy port is created for the Compressor block and the connector
is established between the diagram frame and the selected part property.

4. In the Model Browser, expand the contents of the following read-only packages: 2 Solution Domain > 2
System Structure > 2 Exchange Items.

 It is important that exchange items for communication with the outside are only taken from
the LSA model. Otherwise, the consistency and integrity of different logical subsystem models
is broken.

5. Select the Cabin Air block and drag it to the newly created connector on the open diagram pane.

 According to SysML, the element you want to specify as the item flowing over the connector
must be the type of at least one of the flow properties owned by the interface block that types
the connected proxy ports.

6. Don’t make any changes in the open dialog; click OK. The Cabin Air block is specified as the item
flowing over the newly created connector.

MagicGrid® Book of Knowledge | 205

In the same way, specify that the Evaporator supplies the cooled air back to the cabin (you only need to
change the flow direction in the last step). Once you're done, the ibd in your model should be similar to
the diagram in the following figure.

However, we are not finished yet. Let's say that the engineering team has identified two more interfaces
that are mentioned neither in the system requirements nor in the subsystem requirements specification:
one for letting the hot air out the vehicle, and the other for removing the moisture separated from the
refrigerant out of the VCCS. To capture this information in the model, you need to create two new interface
blocks and assign them to the VCCS block as proxy ports. Since it is system level information, this can only
be done in the LSA model. Therefore, the engineering team of the Cooling System must inform the
systems architect about two new interfaces and request for relevant changes the LSA model. When this is
done, the engineering team of the Cooling System can update their model to let the Cooling System
Architecture block inherit two new proxy ports from the LSA model. These changes may also impact the
work of other engineering teams; in that case, the systems architect must take the responsibility to inform
them.

We assume that you temporarily get back to the LSA model and capture the mentioned interfaces on your
own (this procedure is explained in Chapter System Structure). Once you're done, the Logical Architecture of
VCCS bdd in your model should look like the diagram in the following figure (the updates are highlighted).
It is also important to note that the subsystem and even system requirements must be reviewed and
updated accordingly. We will get back to this in Chapter Traceability to Subsystem Requirements.

206 | MagicGrid® Book of Knowledge

Now it's time to update the Cooling System Logical Components ibd in the solution domain model of the
Cooling System. First, you need to display newly created proxy ports on the diagram fame; second, you
need to specify the above mentioned interactions between the components of the Cooling System and
outside it.

To display the newly created proxy ports on the Cooling System Logical Components ibd frame

1. Make sure the solution domain model (file) is updated in the solution domain model of the Cooling
System.

 For more information on this topic, refer to the latest documentation of your modeling tool.

2. Open the Cooling System Logical Components diagram, if not opened yet.
3. Select the diagram frame and click the Display All Ports button on its smart manipulator (see the
following figure). All the proxy ports of the Cooling System block are displayed on the diagram frame.

MagicGrid® Book of Knowledge | 207

 4. You don't need the ones typed by the InternalControl and InternalResponse interface blocks. Select
them one by one and press Delete to remove them from the diagram (don't worry; they remain in the
model).

 As it was mentioned before, it is not necessary to capture all communications in the

same diagram. Therefore, we will skip showing a couple of proxy ports, as they are not
relevant to the aspect we are going to model in the newly created diagram.

After you capture that the Condenser releases the unneeded hot air out of the Cooling System, and the
Receiver Dryer lets out the moisture, your ibd should look the same as the diagram below.

Now if you open the Cooling System Structure bdd created in step 3 of this cell tutorial, you can update it
to display new proxy ports on the blocks that capture the Cooling System and some of its components.

208 | MagicGrid® Book of Knowledge

Step 6. Specifying interactions within the Cooling System
In this step, we proceed with specifying interactions, between different components of the Cooling
System this time. As in the previous step, interactions can be specified as connectors established via
compatible proxy ports. Information, matter, or energy exchanged throughout these interactions can be
specified as items flowing over relevant connectors.

Let’s say the engineering team identified the following interactions:

1. The Compressor consumes the refrigerant in the form of low pressure and low temperature gas and
transforms it into high pressure and high temperature gas (almost liquid).
2. Inside the Condenser, the refrigerant turns into liquid.
3. The Receiver Dryer separates the moisture from the refrigerant, but doesn't change its physical
characteristics.
4. The Metering Device makes the refrigerant turn into low pressure and low temperature liquid.
5. The Evaporator turns the refrigerant into gas and provides it to the Compressor to close the loop.

The list above determines one more interface of the Cooling System. It is for exchanging the refrigerant
and can be captured in the model as a new interface block. The engineering team doesn't need to inform
anybody about the newly identified interface; it is used for internal communication only and can be
captured in the solution domain model of the Cooling System.

To capture the interface block for exchanging the refrigerant

1. Open the Cooling System Structure diagram.

2. Click the Interface Block button on the diagram palette and then click an empty place on the diagram
pane. An unnamed interface block is created in the model, directly under the 2 Subsystem
Structure package, and displayed on the diagram.
3. Type RefrigerantExchange to name the interface block and press Enter.

MagicGrid® Book of Knowledge | 209

 4. Click the Create Element button on the interface block shape and select Flow Property.

5. Directly on the shape of the interface block, do the following:

a. Leave inout, the default direction of the flow property, as it is.
b. Type r to specify the flow property name. Do not press Enter!
c. Type :Refrigerant and press Enter. The Refrigerant signal is created in the model directly under
the RefrigerantExchange interface block and immediately set as its only flow property type.

6. Select the RefrigerantExchange interface block in the Model Browser:

a. Select the shape of this interface block on the diagram.
b. Press Alt + B.
7. Drag the selection to the nearest 1 Logical Interfaces package. The RefrigerantExchange interface block
is now stored in this package as it should be.
8. Expand the contents of the interface block and select the Refrigerant signal.
9. Drag the signal to the nearest 2 Exchange Items package. The Refrigerant signal is now stored in this
package as it should be.

210 | MagicGrid® Book of Knowledge

As you can see, the refrigerant was automatically captured in the model as a signal, which is not correct,
as the nature of the refrigerant implies that it should be defined as a block. This can be easily fixed by
refactoring the Refrigerant signal to a block. The next thing you need for specifying the above described
interactions is diverse physical conditions of the refrigerant. Each condition can be captured as a sub-type
of the Refrigerant block. For creating sub-types, the infrastructure of the bdd can be used.

To capture diverse physical conditions of the refrigerant

1. Create a bdd:
a. Right-click the 2 Exchange Items package and select Create Diagram.
b. In the search box, type bdd, the acronym for SysML block definition diagram, and then press
Enter. The diagram is created.
c. Type Refrigerant Conditions to specify the diagram name and press Enter again.
2. Drag the Refrigerant signal to the diagram pane. The shape of it is created on the diagram.
3. To refactor the Refrigerant signal to a block:
a. Right-click it and select Refactor > Convert To > Block.
b. Click OK to remove all properties that are incompatible between these two types.
4. Select the Refrigerant block and click the Generalization button on its smart manipulator toolbar.
5. Click an empty place on the diagram pane.
6. Type LPLTG directly on the new shape to name the block and press Enter.

 LP stands for Low Pressure, LT for Low Temperature, G for Gas.

MagicGrid® Book of Knowledge | 211

 7. Follow steps 4 to 6 as many times as necessary to create the HPHTG, HPHTL, and LPLTL blocks.

 HP stands for High Pressure, HT for High Temperature, L for Liquid.

When you're done, you can specify the interactions between the components of the Cooling System
(listed at the beginning of this step).

To specify that the Compressor provides the Refrigerant to the Condenser, in the form of high pressure and
high temperature gas

1. Make sure the Type Selection Mode is enabled in the Cooling System Logical Components diagram.
Otherwise, creation of a proxy port doesn't trigger selection of the interface block to type that proxy
port.
2. Select the shape of the part property typed by the Compressor block and click the Proxy Port button
on its smart manipulator toolbar. A new proxy port appears on the border of the part property
shape.
3. Type ref to find the RefrigerantExchange interface block in the Select Type list, press the Down Arrow
key to select it, and then press Enter.
4. Click the Connector button on the smart manipulator toolbar of the newly created proxy port.
5. Click the shape of the part property typed by the Condenser block.
6. Select New Proxy Port. A compatible proxy port is created for that part property, and the connector
between the couple of part properties is established.

212 | MagicGrid® Book of Knowledge

 7. In the Model Browser, expand the contents of the 2 Exchange Items package (if it is not expanded yet),
select the HPHTG block, and drag it to the newly created connector.

 According to SysML, the element you want to specify as the item flowing via the connector
must be the type of at least one of the flow properties owned by the interface block that types
the connected proxy ports. This also works with sub-types. That’s why the Refrigerant block as
well as any of its types (in this case, the HPHTG block), can be assigned to the connector.

8. Don’t make any changes in the open dialog; click OK. The HPHTG block is specified as the item
flowing via the newly created connector.

In the same way, specify the rest of the previously described interactions. Once you’re done, your ibd
should be similar to the one in the following figure.

 The connector colors can be changed by modifying their symbol properties. For this, right-click the
connector and select Symbol Properties. In the open dialog, select the value cell of the Pen
Color property and click the ... button to open the Color dialog. Then select blue, or any other color
you want, and close both dialogs.

MagicGrid® Book of Knowledge | 213

Now if you get back to the Cooling System Structure bdd created in step 3 of this cell tutorial, you can
update it with the newly created proxy ports.

Subsystem Structure done. What's next?

• Having logical components and knowing how they communicate, you can proceed with structure
modeling and go inside each logical component.
• However, there are still two unvisited cells of the solution architecture for the subsystem. That's
why you should proceed to Chapters Subsystem Behavior and Subsystem Parameters.

214 | MagicGrid® Book of Knowledge

Subsystem Behavior
What is it?

Once finished with the subsystem structure, the appointed engineering team can switch to modeling the
behavior of that subsystem. Behavior models can be relatively abstract and must satisfy functional
requirements of that subsystem (see Chapter Subsystem Requirements). Using the capabilities of
the modeling tool, these models can be executed against test cases to show whether or not these
functional requirements are satisfied.

When finished, the subsystem behavior should be integrated with other subsystem behaviors, into a single
model that represents the comprehensive behavior of the whole SoI. This can be achieved only if the
integrated models are compatible; that is, if they can communicate over interfaces and exchange items.
Remember that interfaces and exchange items are determined by the systems architect in the LSA model.

This cell can be skipped if the engineering team makes the decision to model the behavior of each logical
component captured in the structure model of that logical subsystem (see Chapter Subsystem Structure).

Note that the following material mainly focuses on building the behavioral model of the Cooling System of
the VCCS. In the final steps of this cell tutorial, the scope extends to include the behavioral model of the
CP System. This is necessary for learning how to model the communication between the two subsystems.
Behavioral models of other subsystems, like Heating System and HMI System, are not included;
however, you should imagine that appointed engineers / engineering teams work on them in parallel.

Who is responsible?

The behavior of the appointed logical subsystem can be captured by the Systems Architect, or the
Systems Engineer who belongs to the engineering team responsible for building the solution
architecture of that logical subsystem.

How to model?

The behavior model of the subsystem can be captured by utilizing the infrastructure of the SysML state
machine and activity or sequence diagrams in combination. The state machine allows for capturing states
of the subsystem and transitions between them in response to event occurrences over time. The activity or
sequence diagrams should be used to specify the entry, do, and exit behaviors of these states or transition
effects.

The following figure displays the SysML state machine diagram, identifying the main states of the Cooling
System. Transitions between states can be triggered by various types of events (i.e., a signal event such
as Start between the Off and On states). These signals are taken from the LSA model. This ensures the
integrity of behaviors models of diverse logical subsystems.

MagicGrid® Book of Knowledge | 215

States can have one or more internal behaviors that are specified in the form of the SysML activity diagram
created somewhere in the model. As you can see, the Operating state has the Cool Down Cabin Air activity
specified as the entry behavior. The Cool Down Cabin Air activity diagram is displayed as follows.

Tutorial
• Step 1. Organizing the model for subsystem behavior
• Step 2. Creating a diagram for capturing Cooling System states
• Step 3. Capturing states of the Cooling System
• Step 4. Specifying event occurrences on transitions
• Step 5. Specifying internal behaviors of the state
• Step 6. Synchronizing item flows from the structure model

216 | MagicGrid® Book of Knowledge

Step 1. Organizing the model for subsystem behavior
The states of the Cooling System can be captured in the solution domain model of that subsystem (the
one you created in Chapter Subsystem Requirements). According to SysML, states can only be stored
under the block whose behavior they represent. In this case, it is the Cooling System Architecture block;
thus, you don’t need to create any additional packages for storing states in the model. They should appear
directly under the Cooling System Architecture block, which is stored within the 2 System Structure package.

However, if you need to specify some behavior that the Cooling System performs while being in one or
another state or in transition from one state to another, you should specify that behavior in a separate
package. Following the design of the MagicGrid framework, this package should be stored under the
structure of packages displayed in the following figure.

To organize the model for the subsystem behavior

1. Open the solution domain model of the Cooling System (the Cooling System Solution.mdzip file) you
created in step 1 of the Chapter Subsystem Requirements tutorial, if not opened yet.
2. Right-click the 2.1 Cooling System Solution Domain package and select Create Element.
3. In the search box, type pa, the first two letters of the element type Package, and press Enter.
4. Type 3 Subsystem Behavior to specify the name of the new package and press Enter.

Step 2. Creating a diagram for capturing Cooling System states

The SysML state machine diagram that describes the states of the Cooling System should be created
directly under the block which captures that subsystem.

To create a SysML state machine diagram for capturing the Cooling System states

1. Open the solution domain model of the Cooling System you created in step 1 of the
Chapter Subsystem Requirements tutorial, if not opened yet.

MagicGrid® Book of Knowledge | 217

 2. In the Model Browser, select the Cooling System Architecture block. For this, use the quick find
capability:

a. Press Ctrl + Alt + F. The Quick Find dialog opens.

b. Type csd, c for Cooling, s for System, and a for Architecture.
c. When you see the Cooling System Architecture block selected in the search results list below, press
Enter. The Cooling System Architecture block is selected in the Model Browser.

3. Create the SysML state machine for the Cooling System Architecture block:

a. Right-click the Cooling System Architecture block and select Create Diagram.
b. In the search box, type smd (the acronym for the SysML state machine diagram) and then press
Enter. The diagram is created with the initial state and an unnamed one to jump-start the states
definition.

c. Type Cooling System States to specify the name of the new diagram and press Enter again.

Step 3. Capturing states of the Cooling System

Let’s say the engineering team working on the solution architecture of the Cooling System has identified
the following set of states in which the Cooling System can exist during the VCCS operation:

1. Off
2. On:
a. Initializing
b. Idling
c. Operating

The following rules describe transitions between those states:

• The first state of the subsystem is Off.
• The subsystem can move from the Off state to the On state and back.
• When in the On state, the subsystem goes to the Initializing state first.
• From the Initializing state, the subsystem can move either to the Idling state or to the Off state.
• From the Idling state, the subsystem can move to the Operating state.
• From the Operating state, the subsystem can move either back to the Operating or the Idling state.

Event occurrences that trigger transitions between states are discussed later in this tutorial. This step
mainly focuses on creating states and transitions between them.

218 | MagicGrid® Book of Knowledge

To specify states of the Cooling System

1. Open the Cooling System States diagram, if not opened yet. The diagram already contains the initial
state and an unnamed state.
2. Select the shape of the unnamed state and then click somewhere in the middle of that shape. The
shape switches to the name edit mode.
3. Type Off to specify the name and press Enter.
4. Select the shape of the Off state and click the Transition button on its smart manipulator toolbar.
5. Click an empty place on the diagram pane. Another state is created.
6. Type On directly on the shape of that state and press Enter.
7. Drag the corner of the newly created shape to enlarge it. This is necessary for creating internal states
within it.

8. Click the Initial button on the diagram palette and move the mouse over the shape of the On state.
9. When you see the blue border around the shape of the On state, click it. The initial state is created
within the shape of the On state, and the On state automatically becomes composite.
10. Select the shape of the initial state and click the Transition button on its smart manipulator
toolbar.
11. Right-click an empty place on the shape of the composite state and select State. Another internal
state of that composite state is created.
12. Type Initializing directly on the shape of that state and press Enter.

MagicGrid® Book of Knowledge | 219

13. Create the rest of the states from the list above following the given rules of transition. When you’re
done, your state machine diagram should be very similar to the one in the figure below.

Step 4. Specifying event occurrences on transitions

As defined by SysML, transitions between states can be triggered by various types of events, such as time,
change, signal events, etc. Transitions between states of the Cooling System are triggered by signal events
only. For example, the Cooling System moves from the Off to the On state after it receives the Start signal;
it moves back to the Off state when the Stop signal arrives.

It's important to note that these signals come from outside the Cooling System. They actually come from
the CP System, to be more precise. Therefore, the engineering team doesn't need to capture them in the
LSSA model of the Cooling System. These signals have already been captured in the LSA model and
therefore must be taken from there. The engineering team that works on the LSSA of the CP System uses
these signals in their model, too. The only difference is that the CP System sends them to the outside.

Let's start from specifying the trigger on the transition between the Off and On states.

To make the Start signal a trigger of the transition between the Off and On states of the Cooling System

1. Open the Cooling System States diagram, if not opened yet.

2. In the Model Browser, expand the contents of the following read-only packages: 2 Solution Domain > 2
System Structure > 2 Exchange Items.

220 | MagicGrid® Book of Knowledge

3. Select the Start signal and drag it to the transition between the Off and On states.

4. When you release the mouse, the Start signal becomes the signal event that triggers the transition
between the Off and On states of the Cooling System.

MagicGrid® Book of Knowledge | 221

Looking at the following figure, specify Stop, Cool, and Idle triggers.

When you're done, you can run the simulation to validate the specified behavior and see if your model is
correct. To start the simulation session, click the Run button on the Cooling System States diagram
toolbar. In the Simulation panel, click the Start button to start the machine execution. Note that the
state machine enters the Off state and rests there waiting for the Start signal to trigger it.

There are two places from where you can fire the transition:
• On the diagram

• On the Simulation panel

Try to fire all the transitions and see if all states can be visited. As you may see, a couple of transitions still
have no triggers. These will be specified later, following a certain logic of subsystems communication.

Now it's time to imagine yourself being the engineering team who is responsible for the LSSA of the CP
System, and create the state machine capturing the behavior of that logical subsystem. This is necessary

222 | MagicGrid® Book of Knowledge

to be able to model the communication between the two subsystems. Just like the LSSA of the Cooling
System, the one of the CP System should be kept in a separate model. We assume that you create the
model, use in it the LSA model, and establish the necessary package structure on your own (see step 1 of
the Subsystem Requirements tutorial, to refresh your memory). Afterwards, create a bdd for the logical
subsystem architecture and a block to capture the CP System. Then make it a sub-type of the block that
represents the CP System in LSA model (see step 2 of the Subsystem Structure tutorial) and finally
redefine inherited from it proxy ports.

When you have the model you see in the preceding figure, you can specify the behavior of the CP System.
Note that the Ignition On and Ignition Off signals are captured in the LSSA model of the CP System, while
the DesiredTmpSet and Idle signals are taken from the LSA model.

To make the Ignition On signal a trigger of the transition between the Off and On states of the CP System

1. Select the transition from the Off to On state.

2. Type Ignition On directly on the path and press Enter. The Ignition On signal is created in the Model
Browser and automatically assigned as the signal event to the transition from the Off state to
the On state.

MagicGrid® Book of Knowledge | 223

 3. Right-click the transition and select Go To > Signal Ignition On (see the following figure). The signal is
selected in the Model Browser.

4. Drag the signal to the nearest 2 Exchange Items package.

Follow the steps above to create the Ignition Off signal.

As you can see, the Operating state transition to self is triggered by another type of event: a relative time
event. It allows the modeler to specify that every 3 seconds the CP System moves back to the Operating
state. There will be an internal behavior assigned to this state, performed every time it enters the Operating
state.

To make the time event a trigger of the transition from the Operating state to itself

1. Select the transition.

2. Type after (3s) directly on the path and press Enter. The relative time event is created on the transition
to self.

224 | MagicGrid® Book of Knowledge

 3. Make sure the transition to self is still selected and press Enter. See the trigger information in the
Specification of the transition.

After you're done with states of the CP System, run the simulation and see if the behavior model is
correct.

You have now created two state machines: one for the Cooling System and the other for the CP System.
Moreover, you have specified that the Cooling System is waiting for signals from the CP System. Now it's
time to capture when and how the CP System sends them.

Step 5. Specifying internal behaviors of the state

Every logical subsystem of the SoI can perform one or more behaviors related to one or more of its states.
These behaviors are generally referred as internal behaviors of the state. The internal behavior can be
captured in the form of the SysML activity diagram and stored in the relevant LSSA model, under the 3
Subsystem Behavior package (see step 1 of this tutorial).

Let's say the appointed engineering team has to specify that when the CP System enters
the Initializing state, it invokes other logical subsystems of the VCCS. Therefore, it should send the Start
signal to all of them. Since the signal should be sent to outside the CP System, it is important to capture
that it is passed via the p2 proxy port, as shown in the following figure.

MagicGrid® Book of Knowledge | 225

To specify that the CP System sends out the Start signal on entering the Initializing state

1. Open the LSSA model of the CP System (the one you created in the previous step), if not opened yet.
2. Create a new SysML activity diagram:

a. In the Model Browser, select the 3 Subsystem Behavior package.

b. Right-click the package and select Create Diagram.
c. In the search box, type ad (the acronym for the SysML activity diagram), and then press Enter. A
SysML activity diagram is created. It is owned by the SysML activity.
d. Type Initializing Other Subsystems to specify the name of the new diagram and press Enter again.
The same name is given to the SysML activity, which owns that diagram.

3. Click the Initial Node button on the diagram palette and then an empty place on the diagram pane.
The initial node is created and displayed on the diagram.
4. In the Model Browser, select the Start signal:

a. Press Ctr + Alt + F.

b. Type start.
c. When you see the Start signal in the search results below, press Enter.

5. Drag the selected signal to the newly created diagram pane. The Start send signal action is created and
displayed in the diagram pane.
6. Establish a control flow between the two nodes:

a. Select the initial node and click the Control Flow button on its smart manipulator toolbar.
b. Click the Start send signal action. The control flow is created.

7. Select the Start send signal action and press Enter.

226 | MagicGrid® Book of Knowledge

 8. In the Specification of the send signal action, find the On Port property and select its value cell.
9. Click the button with three dots in this cell.
10. In the Select Port dialog, select the p2 proxy port that belongs to the CP System Architecture block in
the LSSA model, and close the dialog.

As a result, the selected proxy port name displays on the shape of the Start send signal action.

11. Open the CP System States diagram.

MagicGrid® Book of Knowledge | 227

12. Select the Initialize Other Subsystems activity in the Model Browser and drag it onto the Initializing
state.

 Be sure, you selected the SysML activity, not the diagram.

13. Select Entry.

As a result, the Initialize Other Subsystems activity is set as the entry behavior of the Initializing state.

The engineering team appointed to build the LSSA of the Cooling System has to specify the reaction of
that subsystem to the Start signal. Let's say the Cooling System in response sends the coolStatus signal
with the coolOK attribute value true, if it is ready to start, and false, if it is not. As you can see in the
following figure, the response signal is sent over the p4 proxy port. Reactions of other logical subsystems

228 | MagicGrid® Book of Knowledge

of the VCCS should be adequate, but we are not going to model their behavior. It's enough to have two
subsystem models to learn how to model the subsystems communication.

To specify that the Cooling System, when in the Initializing state, performs a self-check and sends out
the coolStatus signal with appropriate attribute value

1. Open the LSSA model of the Cooling System (the one you created in step 1 of the Subsystem
Requirements tutorial), if not opened yet.
2. Create a new SysML activity diagram:

a. In the Model Browser, select the 3 Subsystem Behavior package.

b. Right-click the package and select Create Diagram.
c. In the search box, type ad (the acronym for the SysML activity diagram), and then press Enter. A
SysML activity diagram is created. It is owned by the SysML activity.
d. Type Perform System Check to specify the name of the new diagram and press Enter again. The
same name is given to the SysML activity, which owns that diagram.

3. Click the Initial Node button on the diagram palette and then an empty place on the diagram pane.
The initial node is created and displayed on the diagram.

MagicGrid® Book of Knowledge | 229

 4. Create a decision node:

a. Click the Control Flow button on the smart manipulator toolbar of the newly created initial
node.
b. Right-click a free space below and select Decision.

5. Create an opaque action to set the coolOK attribute value to true:

a. Click the small black triangle next to the Action button on the diagram palette and then select
Opaque Action.
b. Click an empty place on the diagram pane. A new opaque action is created and displayed on the
diagram.
c. Type coolOK=true; directly on the shape.
d. Click somewhere on the diagram pane when you're done.

6. Establish a control flow between the decision node and the newly created opaque action, with 80
percent probability:

a. Select the decision node and click the Control Flow button on its smart manipulator toolbar.
b. Click the opaque action. The control flow between actions is created.
c. Make sure the control flow is still selected and press Enter.
d. In the Specification of the control flow, find the Probability property and type 0.8 in its value cell.

e. Close the dialog. The probability displays on the diagram.

7. Follow step 5 to create another opaque action. This time set the coolOK attribute value to false (see the
following figure).

230 | MagicGrid® Book of Knowledge

8. Follow step 6 to establish another control flow, this time with 20 percent probability (see the
following figure).
9. Create a send signal action to send the coolStatus signal out:

a. In the Model Browser, select the coolStatus signal:

i. Press Ctr + Alt + F.
ii. Type cools. When you see the coolStatus signal in the search results below, press Enter.

b. Drag the selected signal to the open diagram pane. The send signal action with the same name is
created and displayed on the diagram pane.

c. Keep it selected and press Enter.

d. In the Specification window, find the On Port property and select its value cell.
e. Click the button with three dots in this cell.
f. In the Select Port dialog, select the p4 proxy port that belongs to the Cooling System Architecture
block in the LSSA model, and close the dialog.

MagicGrid® Book of Knowledge | 231

 As a result, the selected proxy port name displays on the shape of the coolStatus send signal
action.

10. Establish object flows from each opaque action to the send signal action:

a. Select one of the opaque actions on the diagram pane and then click the Object Flow button
on its smart manipulator toolbar.
b. Click the input pin of the send signal action. The object flow is created, and a new output pin is
attached to the opaque action.
c. Click the pin label on the diagram pane. Then click it again to enter the name edit mode.
d. Type coolOK to change the pin name.

 Note that pins do not display their types, only their names, in the following figure. For
this, do the following:
i. Press down the Shift key and select both pins.
ii. Right-click the selection and click to select the Show Name check box.
iii. Then click Show Type to clear the selection.

232 | MagicGrid® Book of Knowledge

 e. Click the Merge button on the diagram palette and then the newly created object flow. The merge
node is created.
f. Select another opaque action on the diagram pane and click the Object Flow button on its
smart manipulator toolbar.
g. Click the merge node. One more object flow to the merge node is created, and a new output pin is
attached to the opaque action.
h. Click the pin label on the diagram pane. Then click it again to enter the name edit mode.
i. Type coolOK to change the pin name.

11. Open the Cooling System States diagram.

12. Select the Perform System Check activity in the Model Browser (be sure that you selected the SysML
activity, not the diagram!) and drag it onto the Initializing state.
13. Select Entry. As a result, the Perform System Check activity is set as the entry behavior of the
Initializing state.

When you're done with this state machine diagram, try to run the simulation again. Before starting the
model execution, be sure the Auto Open Diagrams mode is enabled (see the following figure). In this
mode, the tool automatically opens the diagram that is currently executed. Hence, when the model
execution switches from the Cooling System States diagram to the Perform System Check diagram (the one

MagicGrid® Book of Knowledge | 233

that specifies the entry behavior of the Initializing state), the Perform System Check diagram is opened
automatically.

The appointed engineering team also has to specify what the CP System does after it gets the responses
from other subsystems. It should evaluate the status of all subsystems according to this information and
send them signals triggering them to move either to the Idling, or to the Off state. As you can see in the
following figure, the appropriate signals are sent over the p2 proxy port.

Note that the SysOK and SysNotOK signals are not defined in any of your models, although they should
have been identified in the LSA model by the system architect many cells ago (see step 5 of the System
Structure tutorial). This means that the system architect must be informed about two new identified
signals. As soon as he/she updates the LSA model, the signals can be used in the LSSA models of all
logical subsystems. Here we expect you to update the LSA model appropriately on your own.

To specify that the CP System evaluates the status of all subsystems and sends them appropriate signals

1. Open the LSSA model of the CP System (the one you created in the previous step), if not opened yet.
2. Create a new SysML activity diagram:

a. In the Model Browser, select the 3 Subsystem Behavior package.

b. Right-click the package and select Create Diagram.
c. In the search box, type ad, the acronym for the SysML activity diagram, and then press Enter. A
SysML activity diagram is created. It is owned by the SysML activity.

234 | MagicGrid® Book of Knowledge

 d. Type Evaluate Status to specify the name of the new diagram and press Enter again. The same
name is given to the SysML activity, which owns that diagram.

3. Click the Initial Node button on the diagram palette and then an empty place on the diagram pane.
The initial node is created and displayed on the diagram.
4. Create an accept event action waiting for the coolStatus signal:

a. Click the Accept Event Action button on the diagram palette and then an empty place on the
diagram pane. The accept event action is created and displayed on the diagram.
b. In the Model Browser, select the coolStatus signal:

i. Press Ctr + Alt + F.

ii. Type cools.
iii. When you see the coolStatus signal in the search results below, press Enter.

c. Drag the selected signal onto the shape of the accept event action. It becomes the coolStatus
accept event action.
d. Double-click the accept event action.
e. In its Specification, find the Is Unmarshall property and set its value to true.
f. Close the dialog.
g. When the shape of the coolStatus accept event action is surrounded by the yellow border, select
the shape and click the button with the yellow triangle (see the following figure).
h. Select Synchronize Accept Event Action Output Pins by Signal Properties.

The output pin named and typed after the attribute of the coolStatus signal is attached to the
shape of the accept event action.

5. Create an empty opaque action:

a. Click the small black triangle next to the Action button on the diagram palette and then select
Opaque Action.

MagicGrid® Book of Knowledge | 235

 b. Click an empty place on the diagram pane. A new opaque action is created and displayed on the
diagram.
c. Click somewhere on the diagram pane when you're done.

6. Establish a control flow between the initial node and the opaque action:

a. Select the initial node and click the Control Flow button on its smart manipulator toolbar.
b. Click the opaque action. The control flow is created.

7. Establish an object flow between the coolStatus accept event action and the opaque action:

a. Select the output pin of the coolStatus accept event action and click the Object Flow button
on its smart manipulator toolbar.
b. Click the opaque action. The object flow is created along with the input pin attached to the
opaque action.
c. Click the label of the input pin on the diagram pane. Then click it again to enter the name edit
mode.
d. Type coolOK to rename the pin and press Enter.

8. Specify the opaque action expression:

a. Select the shape of the opaque action.

b. Type coolingReady=coolOK; directly on the shape.
c. Click somewhere on the diagram pane, when you're done.

9. Create the coolingReady value property for the CP System Architecture block:

a. In the Model Browser, right-click the block and then select Create Element.
b. In the search box, type vp, the acronym of the value property, and press Enter. The new value
property is created and appears in the name edit mode.

236 | MagicGrid® Book of Knowledge

 c. Type coolingReady:Bool and press Ctrl + Spacebar.
d. When you see the Boolean type in the list below, select it and press Enter.

10. Create a decision node:

a. Click the Control Flow button on the the smart manipulator toolbar of the opaque action.
b. Right-click a free space below and select Decision.

11. Create two of the SysOK send signal actions:

a. In the Model Browser, select the SysOK signal:

i. Press Ctr + Alt + F.
ii. Type syso.
iii. When you see the SysOK signal in the search results below, press Enter.

b. Drag the selected signal to the diagram pane twice. Two of the SysOK send signal actions are
created and displayed in the diagram pane.

12. Repeat previous step to create two more of the SysNotOK send signal actions.
13. Create control flows you see in the following figure on your own.
14. Specify the condition when all the subsystems transit to the Idling state:

a. Select the control flow on the right.

b. Type [coolingReady==true and press Enter. The guard is specified.

 Note that the closing bracket is added automatically. You don't need to do this.

15. Specify the condition when all the subsystems transit to the Off state:

a. Select the control flow on the left.

b. Type [else and press Enter.

MagicGrid® Book of Knowledge | 237

16. Specify that one of the SysOK send signal actions is sent via the p2 proxy port of the CP System
Architecture block on your own.
17. Specify that one of the SysNotOK send signal actions is sent via the p2 proxy port of the CP System
Architecture block on your own.

 The following diagram shows how the Evaluate Status diagram would look, if you specified that
the CP System takes into consideration the status of all logical subsystems.

18. Open the CP System States diagram.

19. Select the Evaluate Status activity in the Model Browser and drag it onto the Initializing state.
20. Select Do Activity.

238 | MagicGrid® Book of Knowledge

As a result, the Evaluate Status activity is set as the do behavior of the Initializing state.

Afterwards, state machine diagrams of both logical subsystems must be updated, by specifying the
signals which trigger the transitions to the Idling and Off states. You can do this on your own (see step 4 of
this cell tutorial, if you need to refresh your knowledge).

This is the final view of the CP System States diagram.

MagicGrid® Book of Knowledge | 239

And here is the final view of the Cooling System States diagram.

Step 6. Synchronizing item flows from the structure model

As in the problem domain model, the item flows you identified in the structure model of the logical
subsystem architecture can also be displayed in its behavior model. The following figure shows the item
flows in the Cool Down Cabin Air activity diagram (the HPHTG and HPHTL signals, among others) are
synchronized from the Cooling System Logical Components ibd you created in step 4 of the Subsystem
Structure tutorial. Remember, this can be easily done after you allocate actions to appropriate swimlanes.

240 | MagicGrid® Book of Knowledge

The Cool Down Cabin Air diagram defines the behavior of the Cooling System when it is in the Operating
state (see the following figure). As you already know how to create a SysML activity diagram and set the
selected SysML activity as an entry, do, or exit behavior of the state, it is not explained here how to do this.
However, the depicted diagrams are available in the Cooling System Solution model, which can be found in
<the installation folder of the modeling tool>\samples\MagicGrid (i.e., C:\Program Files\Cameo Systems
Modeler 2021x\samples\MagicGrid).

To synchronize the item flow from the corresponding ibd

1. Select the object flow between the Compress Refrigerant and Turn Gas Into Liquid actions.
2. Click the Realize All Item Flows button on the smart manipulator toolbar of the object flow.

MagicGrid® Book of Knowledge | 241

The item flow is synchronized from the Cooling System Logical Components ibd and displayed on the object
flow.

Subsystem behavior done. What's next?

• Once you define the supposed behaviors of all the logical subsystems, you can start integrating
their LSSA models into the whole and create diverse configurations of the SoI model (see Chapter
Building system configuration model).
• You can also go to the Subsystem Parameters cell and determine the method for calculating one or
more subsystem parameters.

242 | MagicGrid® Book of Knowledge

Subsystem Parameters
What is it?

You can go to this cell as soon as the subsystem structure is captured in the model.

Subsystem parameters capture quantitative characteristics of the relevant logical subsystem, that you can
measure. They are used to calculate MoEs for that logical subsystem, identified during the problem
domain analysis (see Chapter MoEs for Subsystems). MoEs for subsystems captured in the solution
domain model satisfy non-functional quantitative subsystem requirements.

Mathematical expressions for MoEs derivation from subsystem parameters should also be specified in this
cell. Knowing how to calculate MoEs enables early subsystem requirements verification. Using the
capabilities of the modeling tool, the model can be executed to calculate these values and give the verdict
on whether quantitative requirements are satisfied or not.

It's important to note here that not all MoEs in the solution domain must be based on the MoEs in the
problem domain model. Some of them can be identified only when building the logical architecture of the
logical subsystem.

Who is responsible?

Parameters of the appointed logical subsystem can be captured by the Systems Architect or Systems
Engineer, who belongs to the engineering team responsible for building the LSSA of that logical
subsystem.

How to model?

Just like MoEs for the whole system of interest, the MoEs for logical subsystems can be captured in the
model as value properties with the «moe» stereotype applied. These value properties must belong to the
block, which represents the relevant logical subsystem. You must also know that measures of
effectiveness can be inherited from the reusable sets of MoEs, so called MoEs holders, instead of
specifying them from scratch.

The formula for calculating the MoE can be captured as a constraint expression of some constraint block.
The infrastructure of the SysML block definition diagram can be utilized for capturing MoEs, constraint
expressions, and even required subsystem parameters. The following figure displays the bdd that captures
the totalEnergyConsCool measure of effectiveness and the formula for calculating this MoE as the
constraint expression of the Total Energy Consumption for Cooling constraint block.

MagicGrid® Book of Knowledge | 243

To empower the constraint expression to perform calculations, you have to specify what subsystem
parameters should be consumed as variables (or, in terms of SysML, constraint parameters) of that
constraint expression. Just like MoEs, subsystem parameters can be captured in the model as value
properties of the block that captures the logical subsystem or any of its components, only without any
stereotype. The SysML parametrics diagram can be used to relate the subsystem parameters as well as
MoEs to the constraint parameters: former as inputs, latter as outputs. Binding connectors represented as
solid lines in the following figure, should be used for this.

The simulation capabilities of the modeling tool allow you to calculate MoEs with given inputs. Having the
MoE calculated, you can verify the appropriate non-functional subsystem requirement and show whether
it is satisfied or not. The modeling tool enables you to perform this verification automatically. To get ready
for the automated requirements verification, you need to create a satisfy relationship from the value
property, which captures the MoE, to the appropriate subsystem requirement. The following figure
displays an example of the satisfy relationship between the totalEnergyConsCool value property and
the Max Energy Consumption for Cooling non-functional requirement, which means that the requirement is
satisfied when the value of the totalEnergyConsCool value property is not greater than 1.2 kWh. As you can
see, the natural language expression in the requirement text is refined by another constraint expression of
the Max Energy Consumption for Cooling constraint block with the x <= 1.2 inequality.

244 | MagicGrid® Book of Knowledge

Both satisfy and refine relationships are represented in the SysML requirements diagram, but the bdd can
also be used for this purpose.

Tutorial
• Step 1. Organizing the model for subsystem parameters
• Step 2. Specifying the MoE for capturing the total energy consumption for cooling
• Step 3. Defining a method to calculate the total energy consumption for cooling
• Step 4. Specifying subsystem parameters for calculating the total energy consumption for cooling
• Step 5. Binding constraint parameters to appropriate value properties
• Step 6. Performing early subsystem requirements verification
• Step 7. Recalculating subsystem parameters in different states

Step 1. Organizing the model for subsystem parameters

The constraint blocks that capture expressions for calculating subsystem parameters can be stored in a
separate package in the 2.1 Cooling System Solution Domain package. Following the layout of the
MagicGrid framework, its name must begin with number 4.

MagicGrid® Book of Knowledge | 245

To organize the model for the subsystem parameters

1. Open the solution domain model of the Cooling System (the Cooling System Solution.mdzip file) you
created in step 1 of the Subsystem Requirements tutorial, if not opened yet.
2. Right-click the 2.1 Cooling System Solution Domain package and select Create Element.
3. In the search box, type pa (the first two letters of the element type Package) and press Enter.
4. Type 4 Subsystem Parameters to specify the name of the new package and press Enter.

Step 2. Specifying the MoE for capturing the total energy consumption for
cooling
According to the subsystem requirements specification (see Chapter Subsystem Requirements) and the
MoE identified for the Cooling block in the problem domain model (see MoEs for Subsystems), the
appointed engineering team must specify the MoE for capturing the total energy consumed by the Cooling
System.

It was already mentioned that MoEs can be specified from scratch or inherited form the MoEs holder. The
following procedure describes the latter case. The inherited MoE is redefined to change its name and type.
Redefinition is necessary for automated requirements verification, which is performed in one of the
following steps.

To specify the MoE for capturing the total energy consumed by the Cooling System

1. Create a bdd for capturing system parameters:

a. Right-click on the 4 System Parameters package you created in previous step and select Create
Diagram.
b. In the search box, type bdd (the acronym of the SysML block definition diagram), and then
double-press Enter. The diagram is created.
c. Type Subsystem Parameters for Total Energy Consumption for Cooling to specify the name of the
new diagram and press Enter again.

2. Display the MoEs Holder block on the diagram:

a. Press Ctrl + Alt + F to open the Quick Find dialog.

b. In the dialog, type moes.
c. When you see the MoEs Holder block selected in the search results list, press Enter. The block is
selected in the Model Browser.
d. Drag the MoEs Holder block to the diagram pane. The shape of the block is displayed on the
diagram.

3. Display the Cooling System Architecture block on the diagram:

a. Press Ctrl + Alt + F to open the Quick Find dialog.

246 | MagicGrid® Book of Knowledge

 b. In the dialog:

i. Click Any Element to search for particular elements only.

ii. Type cooling system a.

c. When you see the Cooling System Architecture block selected in the search results list, press Enter.
The block is selected in the Model Browser.
d. Drag the Cooling System Architecture block to the diagram pane. The shape of the block is
displayed on the diagram.

4. Establish the generalization between the Cooling System Architecture block and the MoEs Holder block:

a. Select the shape of the Cooling System Architecture block and click the Generalization button
on its smart manipulator toolbar.
b. Select the shape of the MoEs Holder block. The MoEs Holder block becomes the super-type of the
Cooling System Architecture block, which means that the latter inherits all MoEs from the former.

5. Redefine the totalEnergyConsumption value property for the Cooling System Architecture block:

a. Right-click the Cooling System Architecture block and select Specification.

b. On the left of the open dialog, click Properties.
c. On the right, under the Value Properties category, select the totalEnergyConsumption value
property.
d. Click the Redefine button.
e. In the Type cell, immediately type kWh and press Enter. The kWh value type is created in the
model and set as type for the redefined value property.

MagicGrid® Book of Knowledge | 247

 f. In the Name cell, type totalEnergyConsCool to overwrite the current property name and press
Enter. The inherited value property under this new name now belongs to the Cooling System
Architecture block.
g. Click Close.

 As you can see, the inherited value property it redefines displays in the curly brackets.

Step 3. Defining a method to calculate the total energy consumption for cooling
The next thing you need to do is capture the mathematical expression which specifies how to calculate the
MoE, minutes ago specified for the Cooling System. It can be captured as the constraint expression of
some constraint block stored in your model.

Let’s say the total energy consumption of the Cooling System equals a sum of energies consumed by all
its components. The constraint expression can be defined as follows:

totalEnergyCons = energyCons1 + energyCons2 + energyCons3 + energyCons4 + energyCons5

where:

1. totalEnergyCons stands for the total energy consumed by the Cooling System
2. energyCons1 may stand for the energy consumed by the Compressor
3. energyCons2 may stand for the energy consumed by the Condenser
4. energyCons3 may stand for the energy consumed by the Receiver Dryer
5. energyCons4 may stand for the energy consumed by the Metering Device
6. energyCons5 may stand for the energy consumed by the Evaporator
In terms of SysML, totalEnergyCons, energyCons1, ..., and energyCons5 are generally referred to as
constraint parameters and later should be bound to corresponding value properties specified in the model.

A bdd can be utilized for capturing constraint blocks. This constraint block is not an exception; therefore,
you can use the bdd created in step 2 of this cell tutorial. The procedure of capturing the method for
calculating the total energy consumption of the Cooling System is adequate to the one described in step
3 of System Parameters tutorial. Once you capture the method, the Subsystem Parameters for Total Energy
Consumption for Cooling bdd should look like in the following figure.

248 | MagicGrid® Book of Knowledge

Step 4. Specifying subsystem parameters for calculating the total energy
consumption for cooling
Having the constraint expression, you can easily determine what subsystem or, as in this case, component
parameters are necessary for calculating the total energy consumed by the Cooling System. Input
parameters of the Total Energy Consumption for Cooling constraint block can help with this task.

The Total Energy Consumption for Cooling constraint block has five input parameters, each for the energy
consumption of a single component of the Cooling System, and therefore requires five component
parameters in your model. Appropriate value properties should be specified for each block that captures
one of the components of the Cooling System.

The procedure for capturing the above mentioned component parameters is adequate to the one described
in step 4 of System Parameters tutorial. Once you're done with this, the Subsystem Parameters for Total
Energy Consumption for Cooling bdd should look like the following figure.

Step 5. Binding constraint parameters to appropriate value properties

Once you have all the necessary value properties in the model, it’s time to bind them to the appropriate
constraint parameters. Remember that it is the binding which empowers the constraint expression to
perform calculations on given inputs.

MagicGrid® Book of Knowledge | 249

Bindings between value properties and constraint parameters can be established using binding connectors.
This type of SysML relationship can be created by utilizing the infrastructure of the SysML parametrics
diagram created for the Cooling System Architecture block. However, creating the SysML parametrics
diagram is not enough to get started. Establishing binding connectors is possible only after the constraint
property typed by the Total Energy Consumption for Cooling constraint block is created for the Cooling
System Architecture block. Be aware that the same constraint block can be used in multiple contexts, and
binding connectors are context-specific relationships.

If you name the SysML parametrics diagram Total Energy Consumption for Cooling, the procedure of
establishing the bindings is very similar to the one described in step 5 of the System Parameters tutorial.
Once you're finished with this, the Total Energy Consumption for Cooling parameters diagram in your model
should look like the one below.

Now you can execute the model, provide input values, and see the result of the calculation. To start the
execution, click the Run button in the toolbar above the Total Energy Consumption for Cooling diagram.
When in the Variables tab of the Simulation panel (by default, it is on the right side of
the Simulation panel), you can specify the input values and see the output value (result of the calculation).

250 | MagicGrid® Book of Knowledge

Step 6. Performing early subsystem requirements verification
Having the concrete value of the value property that captures the MoE, you can verify
the related subsystem requirement and show whether it is satisfied or not. In this case, you can perform
the verification of the Max Energy Consumption for Cooling subsystem requirement, which insists that “The
Cooling System shall consume maximum 1.2 kWh”. The procedure for getting ready for this verification is
adequate to the one described in step 6 of System Parameters tutorial. If you name the requirement
diagram Subsystem Requirements Formalized, it should look very similar to the one below when you're
done.

Now, if you go back to the Total Energy Consumption for Cooling diagram and execute the model again, you
should see that the totalEnergyConsCool value property is highlighted in green, which means that the Max
Energy Consumption system requirement is satisfied. Try to change the input values and see how the
highlight color changes when the requirement is not satisfied. This is how you perform the automated
requirements verification. Moreover, this way you can perform manual or automatic (it requires some
algorithm to be defined) parameters optimization in order to find the optimal system configuration.

These values can be stored in the model as SysML instance specifications and their slot values. If you
cannot remember how to do this, refer to step 7 of the System Parameters tutorial.

MagicGrid® Book of Knowledge | 251

Step 7. Recalculating subsystem parameters in different states
At times, values used for calculating a subsystem parameter may be dynamic and may vary in different
states of the subsystem. As a result, the system parameter values may vary, too.

To illustrate this, let's assume that the components of the Cooling System consume more energy in the
Initializing and Operating states, and less energy in the Idling state.

Values can be specified using Action Language Helper (ALH) expressions supported by the modeling
tool in:
• The body of the opaque behavior set as the entry or do behavior of the relevant state.

• The body of the opaque action which belongs to the entry or do behavior of the relevant state.

252 | MagicGrid® Book of Knowledge

As you can see, this ALH.setValue expression has a set of variables that require:

1. The name of the part property (highlighted in the following figure) that is typed by the block whose
value property you want to update. That part property must belong to the block which encloses the
state machine diagram; in this case, it is the Cooling System Architecture block.

2. The name of the value property you want to update (highlighted in the following figure), in quotes.
The value property must belong to the block which types the previously defined part property.

3. Concrete value you want to set for the previously defined value property (for example, 0.5).

To specify values in the body of the opaque behavior set as the entry behavior of the Idling state of the
Cooling System

1. Open the Cooling System States diagram:

a. Press Ctr + Alt + F.

b. Select Diagram and type cooling system s.
c. When you see the Cooling System States diagram in the search results below, press Enter.

2. Double-click the shape of the Idling state.

3. On the right side of the open dialog, under the Entry category, select the Behavior Type property.

MagicGrid® Book of Knowledge | 253

 4. Click the value cell of this property and from the drop-down list, select Opaque Behavior.

5. Select the nearest Body and Language property below.

6. Click the value cell of this property and type there:

ALH.setValue(cmp, "energyCons", 0.1);

ALH.setValue(cnd, "energyCons", 0.1);
ALH.setValue(evp, "energyCons", 0.1);
ALH.setValue(rd, "energyCons", 0.02);
ALH.setValue(md, "energyCons", 0.02);

7. Click Close. The shape of the Idling state displays the ALH expressions.

To specify the values in the body of the opaque action which belongs to the entry behavior of the
Initializing state

1. Open the Perform System Check diagram:

a. Press Ctr + Alt + F.

b. Select Diagram and type perform.
c. When you see the Perform System Check diagram in the search results below, press Enter.

254 | MagicGrid® Book of Knowledge

 2. Click the small black triangle next to the Action button on the diagram palette and then select Opaque
Action.
3. Click the control flow between the initial and decision nodes. A new opaque action is created and
displayed on the diagram.
4. Immediately type on the shape of the newly created opaque action:

ALH.setValue(cmp, "energyCons", 0.2);

ALH.setValue(cnd, "energyCons", 0.2);
ALH.setValue(evp, "energyCons", 0.08);
ALH.setValue(rd, "energyCons", 0.08);
ALH.setValue(md, "energyCons", 0.05);

5. Pres Enter.

Following the procedure above, you can update the Cool Down Cabin Air diagram. When finished, you can
run the simulation to see how the totalEnergyConsCool value property changes the value over various states
of the Cooling System. Make sure you run the simulation with the context (the Cooling System Design
block) which encloses the state machine diagram; otherwise, the subsystem parameter value is not
calculated. For this, click the small black triangle next to the Run button and select Run with Context.

Subsystem parameters done. What's next?

Finishing with the Subsystem Parameters cell means that you're done with the entire model, which
describes the solution architecture of the logical subsystem. The only task before you can integrate it into
the model of the whole SoI is establishing traceability relationships from the elements that capture the
LSSA to subsystem requirements.

MagicGrid® Book of Knowledge | 255

Traceability to Subsystem Requirements
What is it?

To have a complete LSSA model of a particular logical subsystem, you need to establish traceability
relationships from the elements that capture the LSSA to the elements that capture the subsystem
requirements. As determined by SysML, these are the satisfy relationships, meaning that one of the latter
fulfills one of the former; for example, a satisfy relationship can be established from the part property
typed by the block that captures some logical component to a certain subsystem requirement. In the
complete LSSA model, all subsystem requirements are satisfied by one or more architecture elements even
though these elements are not yet implemented.

Who is responsible?

Traceability relationships in the LSSA model of the appointed logical subsystem can be captured by the
Systems Architect or Systems Engineer who belongs to the engineering team responsible for building the
LSSA of that logical subsystem.

How to model?

Neither of the SysML diagrams is suitable for creating a mass of cross-cutting relationships, such
as satisfy. For this, a dependency matrix is more suitable. The modeling tool offers a wide variety of
predefined matrices for specifying cross-cutting relationships. To capture satisfy relationships, a Satisfy
Requirement Matrix can be used.

256 | MagicGrid® Book of Knowledge

Tutorial
• Step 1. Creating a dependency matrix for capturing satisfy relationships
• Step 2. Capturing satisfy relationships to subsystem requirements
• Step 3. Traceability to and revision of system requirements

Step 1. Creating a dependency matrix for capturing satisfy relationships

In this step, we create a Satisfy Requirement Matrix and specify its criteria. Subsystem requirements can
be displayed in the rows of the matrix, and the elements that capture the LSSA can be displayed in its
columns.

To create a Satisfy Requirement Matrix

1. In the Model Browser, right-click the 5 Traceability package (you created it in step 4 in the Subsystem
Requirements tutorial) and select Create Diagram.
2. In the search box, type srm (the acronym of the predefined Satisfy Requirement Matrix), and press
Enter.

 If you don’t see any result, click the Expert button below the search results list. The list of
available diagrams expands in the Expert mode.

3. Type LSSA to Subsystem Requirements to specify the name of the new matrix and press Enter again.
4. On the matrix toolbar, click the Change Axes button to swap rows of the matrix with columns.
Requirement becomes the type of the matrix rows.
5. In the Model Browser, select the SSR-1 SRS for Cooling System requirement (you might need to expand
the 1 Subsystem Requirements package first) and drag it onto the Row Scope box in the Criteria area
above the matrix contents. The SSR-1 SRS for Cooling System requirement becomes the scope of the
matrix rows (see the following figure).
6. Select column element types:

a. In the Model Browser, select any part property, proxy port, value property, state, and call
behavior action.

 To select the set of non-adjacent items in the tree, click the first one, press Ctrl, and while
holding it down, select other items one by one.

b. Drag the selection onto the Column Type box in the Criteria area above the matrix contents. The
columns of the matrix are set to display all the part properties, proxy ports, value properties,
states, and call behavior actions (see the following figure) within the scope you select in the next
step.

7. Select the column element scope:

a. In the Model Browser, select the Cooling System Architecture block (you might need to expand
the 1 Subsystem Structure package beforehand).
b. Drag the selection onto the Column Scope box in the Criteria area above the matrix contents. The
selected block becomes the scope of the matrix columns (see the following figure).

MagicGrid® Book of Knowledge | 257

Once you're done with the criteria, the contents of the matrix is updated. All the cells are empty at the
moment, except the one at the intersection of the totalEnergyConsCool value property and the SR-1.3 Max
Energy Consumption for Cooling requirement (as you may remember, this relationship was created in step
6 of the Subsystem Parameters tutorial).

Step 2. Capturing satisfy relationships to subsystem requirements

Now you’re ready to establish other satisfy relationships. As you already know how to do this, we assume
that you continue filling in the matrix yourself. If you need a hint as to what elements can be related, see
the following figure.

258 | MagicGrid® Book of Knowledge

  If functional requirements become marked as invalid, there is no need to worry; you can simply
ignore the warning.

As you can see, there are a few proxy ports that cannot trace to any subsystem requirement (as you can
remember, they were identified in Chapter Subsystem Structure). This requires the revision and update of
the subsystem requirements specification, and vice versa. This time, only the subsystem requirements
specification needs to be updated. You can see the new requirements highlighted in the following figure.

MagicGrid® Book of Knowledge | 259

Afterwards, the satisfy requirements matrix in your model can be completely filled in. As you can see in
the following figure, there are no hanging elements any more.

Step 3. Traceability to and revision of system requirements

If you look at the Subsystem Requirements to System Requirements and LSA matrix, you can see there a few
new subsystem requirements, which don't have traceability relationships. These are the subsystem
requirements that you created in the previous step: SSR-1.1.2 Hot Air Removal and SSR-1.1.3 Waste
Removal.

After you establish a few more refine relationships in this matrix, it should look like in the following figure.

260 | MagicGrid® Book of Knowledge

Now it's time to establish derivation relationships from these subsystem requirements to system
requirements. This requires you to capture new system requirements first. The SSR-1.1.2 Hot Air Removal
or SSR-1.1.3 Waste Removal requirements were created because of interfaces identified by the engineering
team working on the LSSA model of the Cooling System. Moreover, these interfaces belong to the entire
SoI, first of all. Therefore, we can state that at least two system requirements are currently missing.

Here is the action plan dealing with this situation:

1. As the systems architect, append the system requirements specification in the LSA model, with two
new interface requirements.

MagicGrid® Book of Knowledge | 261

 2. In the LSA to System Requirements matrix, establish satisfy relationships from relevant proxy ports
(created in step 5 of the Subsystem Structure tutorial) to these requirements.

3. In the System Requirements to Stakeholder Needs matrix, establish missing derivation relationships from
new system requirements to stakeholder needs.

262 | MagicGrid® Book of Knowledge

 4. As the appointed engineering team, establish missing derivation relationships between subsystem
and new system requirements in the Cooling System SA model.

When you're done, look at the From Stakeholder Needs to Subsystem Requirements map. It should also be
updated.

LSSA done. What's next?

The MagicGrid BoK does not detail how to model each component (i.e., Compressor, Condenser, and
Evaporator, etc.) or even more precise units as defined by the MagicGrid framework. It's enough to note
that the modeling workflow in these levels of detail is adequate to modeling in the subsystem level. That
is, you start with derivation of requirements at that level of detail and then work on the structure,
behavior, and parameters to satisfy them; once you've finished, you establish traceability relationships in
order to approve the suggested logical architecture.

As mentioned in the Preface, the component-level model of the solution domain can be both MBSE and
MBD, because this is where the logical architecture relates to the logical (high-level) design of the SoI, and
MBD starts. Therefore, besides SysML, other standards can be applied at the component-level. These

MagicGrid® Book of Knowledge | 263

include OMG® Unified Modeling Language (UML)® for designing software components, Modelica® for
designing mechanical, electrical, and electronic components, among others. Please be aware that for UML,
you don't need to switch software, as UML is supported by the modeling tool along with SysML.

When you have the LSSA models of all logical subsystems, including their components and even more
precise units, you can integrate them and have the entire logical architecture of the SoI in a single
model (see Chapter Building system configuration model).

264 | MagicGrid® Book of Knowledge

Building system configuration model
Building the system configuration model is the final phase of specifying the solution domain. It is
performed after engineering teams are done with solution architectures of all logical subsystems of the
SoI. The goal of this phase is to build the integrated model of the entire SoI and check if its
logical subsystems can successfully interact with each other by verifying the integrity of the model. This is
the task of the systems architect or systems integrator. Note that more than one solution can be proposed;
the systems architect/integrator would perform a trade-off study to pick a preferred one of each logical
subsystem.

The highlighted area in the following figure show the part of the solution domain we are talking about.

As you may conclude looking at the design layout of the MagicGrid framework (see the highlighted area in
the following figure), the system configuration model captures the overall structure, behavior, and
parameters of the SoI.

MagicGrid® Book of Knowledge | 265

System Configuration Structure
What is it?

Once all engineering teams deliver the LSSA models of each logical subsystem of the SoI (defined in the
LSA model), the systems architect/integrator is able to finish his/her job. That is, he/she can integrate all
the models into the whole and build the system configuration model.

The main purpose of building the system configuration model is to verify whether logical subsystems of
the system of interest can successfully communicate. The systems architect/integrator must check
whether interfaces of separate logical subsystems are compatible. If one or more incompatible interfaces
are detected, the systems architect/integrator requests the responsible engineering team to review and
update their model appropriately. After the engineering teams complete the request, the systems
architect/integrator can make another attempt to integrate the models.

Although the VCCS has four logical subsystems overall, the following material focuses on integrating only
two of them: the Cooling System and the CP System. The model with all logical subsystems integrated is
available in <the installation folder of the modeling tool>\samples\MagicGrid; for example, C:\Program
Files\Cameo Systems Modeler 2021x\samples\MagicGrid.

Who is responsible?

The system configuration structure can be captured by the Systems Architect or Systems Engineer, who is
the head of the whole systems engineering project and is responsible for the smooth integration of the
subsystems and their components into the whole, the same person who owns the LSA model of the
SoI. This task can also be performed by the Systems Integrator, a subordinate of the Systems Architect.

266 | MagicGrid® Book of Knowledge

How to model?

The integrated structure model of the system configuration can be captured by utilizing the infrastructure
of the SysML bdd.

The structure decomposition map enables users to review and update the integrated structure model in a
compact form.

The interface compatibility can be verified by specifying the logical subsystems communication in the
SysML ibd.

MagicGrid® Book of Knowledge | 267

.

Tutorial
• Step 1. Creating and organizing a model for system configuration structure
• Step 2. Getting ready to capture the structure of the system configuration
• Step 3. Capturing the integrated structure of the system configuration
• Step 4. Verifying interface compatibility

Step 1. Creating and organizing a model for system configuration structure

Since you already have the LSSA models of the Cooling System and CP System, you (in the role of the
systems architect/integrator) can integrate them into a single model that represents the whole VCCS.

For this, you must create another model. In general, this model can be called the system configuration
model. Note that there can be more than one system configuration model within a single solution domain;
however, we will focus on a single one.

To create a system configuration model of the VCCS

1. Start sharing the LSSA models of the Cooling System and the CP System.

 • The LSSA models of all subsystems of the VCCS should be shared in this step.
• For more information on this topic and the topics mentioned in steps 2 and 3, refer to
the latest documentation of your modeling tool.

2. Create a new model. It can be named VehicleCCS Configuration.mdzip. This is the configuration model
of the Vehicle Climate Control System.
3. Use the LSSA model of the Cooling System as read-only in the configuration model of the VCCS.
Because it has already been used in the LSSA model of the Cooling System, the problem domain

268 | MagicGrid® Book of Knowledge

 model and the solution domain model of the VCCS (LSA model) are automatically used as well.

 Grey color element names within the used model indicate that elements cannot be modified.
This is because the external model is used in the read-only mode.

4. Repeat step 3 to use the solution domain of the CP System.

5. Right-click the Model package (this is the default name of the root package) and select Create
Element.
6. In the search box, type pa (the first two letters of the element type Package), and press Enter.
7. Type 3 Configuration Domain to specify the name of the new package and press Enter (see the figure in
step 9).
8. Right-click the 3 Configuration Domain package and select Create Element.
9. Repeat step 6.
10. Type 2 System Structure to specify the name of the new package and press Enter.

Step 2. Getting ready to capture the structure of the system configuration

To get ready to capture the integrated structure of the whole system of interest or one of its
configurations, you first need to create a block which represents the system configuration. The block that
captures the VCCS configuration must inherit all the structural and behavioral features from the block that
captures the VCCS in the LSA model (see step 3 of the System Structure tutorial). Thus, the block that

MagicGrid® Book of Knowledge | 269

captures the VCCS configuration must be a sub-type of the the block that captures the VCCS in the LSA
model.

A bdd created under the 2 System Structure package can be used for capturing this block, as well as for
performing the next steps of this cell tutorial.

To get ready to capture the structure of the VCCS configuration

1. Create a bdd:

a. In the Model Browser, right-click the 2 System Structure package you created in step 1 of this cell
tutorial, and select Create Diagram.
b. In the search box, type bdd (the acronym for SysML block definition diagram), and then press
Enter. The diagram is created.
c. Type VCCS Configuration Structure to specify the name of the new diagram and press Enter again.

2. Display the VCCS block from the LSA model on the diagram:

a. In the Model Browser, select the VCCSblock (you might need to expand the2 Solution Domain
and 2 System Structure packagesfirst).

b. Drag the selection to the newly created diagram pane. The shape of the VCCS block is displayed
on the diagram.

270 | MagicGrid® Book of Knowledge

3. Create a block to capture the VCCS configuration:

a. Select the VCCS block, if not yet selected, and click the Generalization button on its smart
manipulator toolbar.

b. Click an empty place on the diagram below. A new unnamed block is created, and its shape is
displayed on the diagram pane.
c. Type VCCS Configuration to specify the name of the new block and press Enter.

4. Redefine the inherited proxy ports:

a. Double-click the shape of the VCCS Configuration block on the diagram.

b. On the left of the open dialog, click Ports/Interfaces.

MagicGrid® Book of Knowledge | 271

 c. Select the proxy port in the first row and click the Redefine button below.

As a result, the selected proxy port is redefined. From there, it is no longer decorated with
the caret symbol (“^”), which indicates the inheritance. Don't change anything.

d. Row by row, redefine the rest (the order of items may differ from what you see in the following
figure).

e. Click Close.

272 | MagicGrid® Book of Knowledge

 5. Click the Display All Ports button on the smart manipulator toolbar of the VCCS Configuration block.
The shapes of proxy ports are displayed on the VCCS Configuration block.

Step 3. Capturing the integrated structure of the system configuration

You can utilize the bdd created in the previous step to build the integrated structure model. All you need to
do in this step is define the logical subsystems that are included in the particular system configuration. If
(as in this case) every logical subsystem has only one solution architecture, the system configuration is
also singular.

When specifying the system configuration structure, the blocks which represent the logical subsystems
must be taken from the LSSA models; that is, the Cooling System Architecture block from the LSSA model of
the Cooling System (the Cooling System Solution.mdzip file), and the CP System Architecture block from the
LSSA of the CP System (the CP System Solution.mdzip file). Be sure you haven't taken the blocks
representing the logical subsystems from the LSA model. These blocks have neither internal structure no
behavior; remember, they were created to identify the work packages.

The modeling tool actually enables you to prevent the selection of inappropriate blocks. For this, you need
to narrow the scope for type selection down to one or more packages. In this case, these are the 2.1
Cooling System Solution Domain and 2.2 CP System Solution Domain packages from the used models, as
shown in the following figure. Since you are aware of how to narrow down the scope (see step 3 of
System Structure tutorial, if you need to refresh), we assume that you do this on your own.

MagicGrid® Book of Knowledge | 273

Now you're ready to capture the integrated structure of the VCCS configuration. Just don't be too quick to
create new part properties for the VCCS Configuration block. As a sub-type of the VCCS block, it already has
them, inherited from the super-type. Therefore, all you need to do is redefine the inherited part properties
of the VCCS Configuration block and select new types for them.

To capture the integrated structure of the VCCS configuration

1. Open the VCCS Configuration Structure bdd, if not opened yet.

2. Integrate the LSSA of the Cooling System into the VCCS configuration:

a. Double-click the VCCS Configuration block.

b. In the Specification of the VCCS Configuration block, click Properties on the left.
c. Select the cooling part property on the right and then click the Redefine button below.
d. In the Type column, select the Cooling System Architecture block.

274 | MagicGrid® Book of Knowledge

 3. Follow steps 2.c and 2.d to integrate the LSSA of the CP System into the VCCS configuration.

After you close the dialog, the VCCS Configuration Structure diagram should look very similar to the one
below.

MagicGrid® Book of Knowledge | 275

If you drag the parts, one by one, out of the enclosing block shape, the diagram should look like this.

If you decide to display the deeper structure (right-click the Cooling System Design block and select Display
> Display Related Elements), the diagram should look as follows.

276 | MagicGrid® Book of Knowledge

The integrated structure of the SoI configuration can also be displayed in the structure decomposition
map, one of the predefined relation maps of the modeling tool. It can display any structure in a compact
form, which is especially suitable for review.

To display the integrated structure of the VCCS configuration in a map

1. In the open diagram, select the shape of the VCCS Configuration block.
2. Right-click the selection and then select Create Diagram.
3. In the search box, type sdm (the acronym of the structure decomposition map), and then press
Enter. The map is created.

 If you don’t see any results, click the Expert button below the search results list. The list of
available diagrams expands in the Expert mode.

4. Type VCCS Configuration Structure to specify the name of the new map and press Enter again.

 If your modeling tool version is lower than 2021x, you should see the cooling and cp part
properties twice (see the following figure). This is because the structure decomposition map in
these versions displays the inherited part properties even if they are redefined.

To remove the inherited part properties from the map, select them and press Delete.

Step 4. Verifying interface compatibility

Now it's time to specify how logical subsystems communicate within the VCCS configuration.

Remember that communication is possible only if these logical subsystems have compatible interfaces. For
this reason, we can state that the systems architect performs the interface compatibility verification by
specifying the interactions within the particular configuration of the SoI.

You already know that interactions between the internal parts of the system can be specified in the SysML
ibd. So let's create one and check whether the Cooling System and CP System, two logical subsystems of
the VCCS, can successfully communicate.

MagicGrid® Book of Knowledge | 277

To specify interactions between the Cooling and CP Systems

1. Create an ibd for specifying interactions between the logical subsystems:

a. In the VCCS Configuration Structure diagram you created in step 2 of this cell tutorial, select
the shape of the VCCS Configuration block.
b. Click the SysML Internal Block Diagram button on its smart manipulator toolbar.
c. A new blank diagram is created.
d. In the Display Parts/Ports dialog, do the following:

i. Click Clear All on the right, to unselect all the items.

ii. Click to select the check boxes near the cooling : Cooling System Architecture and cp : CP
System Architecture part properties (see the following figure) and then click OK.

The selected part properties display on the diagram pane.

e. In the Model Browser, select the newly created diagram and press F2 to enter the name edit
mode.
f. Type Subsystems Communication in VCCS to change its name and press Enter.

278 | MagicGrid® Book of Knowledge

2. Display proxy ports for interactions between the Cooling System and the CP System:

a. Right-click the shape of the cooling property and select Display > Display Parts/Ports.
b. In the Display Parts/Ports dialog, do the following:

i. Click Clear All on the right, to unselect all the items.

ii. Click to select the check boxes near the p3 : ~InternalControl and p4 : InternalResponse proxy
ports (see the following figure), and then click OK.

The selected proxy ports display on the shape of the cooling part property.

MagicGrid® Book of Knowledge | 279

 c. Follow steps a and b to display the relevant proxy ports on the shape of the cp part property.

 As you can see, there are two pairs of compatible proxy ports: one pair typed by the
InternalControl interface block, and the other by the InternalResponse interface block. Each
pair has one proxy port for data input, and one for data output.

3. Create connectors between the part properties, over compatible proxy ports:

a. Select the p2 proxy port on the shape of the cp part property and click the Connector button
on its smart manipulator toolbar.
b. Click the p3 proxy port on the shape of the cooling part property.
c. Follow steps a and b to create a connector between the other pair of compatible proxy ports.

4. Specify the Start signal as the item flowing over the connector between the cp and cooling part
properties:

a. In the Model Browser, select the Start signal:

i. Press Ctrl + Alt + F. The Quick Find dialog opens.

ii. Type start and when you see the Start signal selected in the search results list below, press
Enter. The Start signal is selected in the Model Browser, under the following structure of
packages: 2 Solution Domain > 2 System Structure > 2 Exchange Items.

b. Drag the selection onto the connector between two proxy ports typed by the InternalControl
interface block.

280 | MagicGrid® Book of Knowledge

 c. In the open dialog, don't change anything and click Finish.

5. Follow step 4 to specify the SysOK and SysNotOK signals as items flowing over the same connector,
and the coolStatus signal as the item flowing over the connector between cooling and cp part
properties.

As you can see, no incompatible interfaces were identified, and we can state that the Cooling System and
CP System can successfully communicate within the VCCS.

After you specify the interactions of the logical subsystems with the outside, your Subsystems
Communication in VCCS diagram should look like the one below.

MagicGrid® Book of Knowledge | 281

System Configuration Structure done. What's next?

• The integrated model of the system configuration enables you to calculate more precise values of
system parameters than in the LSA model. Chapter System Configuration Parameters describes
how to do this.
• Moreover, the specification of logical subsystems communication enables their behavioral models to
exchange signals (so far, it was possible to send and receive signals only within the scope of a single
logical subsystem). Therefore, you can switch to analyzing the integrated behavior of the whole
SoI , which is described in Chapter System Configuration Behavior.

282 | MagicGrid® Book of Knowledge

System Configuration Behavior
What is it?

The behavior of the system of interest integrates the behaviors of its logical subsystems, identified in the
LSA model (see Chapter System Structure).

Once you build the integrated structure model of the entire SoI (see previous chapter), you have its
integrated behavior model as well. No modeling effort is required in this cell. Its main purpose is to analyze
the integrated system behavior model and verify it.

Who is responsible?

The integrated behavior of the system configuration can be analyzed and verified by the Systems
Architect or Systems Engineer, who is the head of the whole systems engineering project and is
responsible for the smooth integration of the subsystems and their components into the whole, the same
person who owns the LSA model of the SoI. This task can also be performed by the Systems Integrator, a
subordinate of the Systems Architect.

How to model?

Remember that this cell is not about modeling. It is about integrated model analysis. For this, the
simulation capabilities of the modeling tool can be exploited. By executing the model, you can simulate
the behavior of the whole SoI and verify if it is as expected.

The simulation animation enables you to monitor the execution progress. It marks the current, last visited,
and all other visited element shapes. The following figure depicts the animated moment of model
execution, when the CP System Architecture block sends out the Start signal via the p2 proxy port and the
Cooling System Architecture block is about to receive it. Note that the Subsystems Communication in
VCCS diagram helps; it specifies where the Start signal is sent.

MagicGrid® Book of Knowledge | 283

Tutorial
• Step 1. Starting simulation of the system configuration block
• Step 2. Analyzing the integrated behavior of the system configuration

Step 1. Starting simulation of the system configuration block

Let's say you, as the systems architect, perform the analysis of the integrated system behavior to verify
the SR-1.2.1 Internal Communication, SSR-1.2.2 Initializing, and SSR-1.2.3 Idling requirements. You will need
to prove that the specified system behavior meets the following logic:

1. Once ignition starts, the CP System initializes all other subsystems of the VCCS (in this case, the
Cooling System only).
2. In response, the Cooling System performs the self-check and informs the CP System if it can start.
3. After the CP System receives the status from the Cooling System, it decides whether or not they can
both start.

To simulate the behavior of the whole configuration of the SoI, a relevant block must be indicated as the
simulation context. In this case, it is the VCCS Configuration block, shown in the following figure.

 If the system behavior has been defined in LSA model and it is still assigned as the classifier
behavior of the VCCS block (the super-type of the VCCS Configuration block), during the simulation it
will conflict with the integrated behavior of the VCCS Configuration block.
To prevent this conflict, go back to the LSA model and set the Classifier Behavior property value to
<UNASSIGNED> for the VCCS block (refer to Chapter System Behavior). If you cannot update the
LSA model (for example, you don't have relevant permission), you can define a fake system
behavior for the VCCS Configuration block (such as the state machine with a single default state). As
a result, the fake system behavior becomes the classifier behavior, which doesn't conflict with the
integrated one during the simulation.

284 | MagicGrid® Book of Knowledge

To start the simulation of the VCCS Configuration block

1. In the Model Browser, select the VCCS Configuration block.

2. Right-click the selection and then select Simulation > Run. The Simulation panel opens at the bottom
of the modeling tool window.

3. Click the Auto Open Diagrams button to enable the mode, in which currently executed
diagrams open automatically.
4. On the toolbar of the Simulation panel, click the Start button . The simulation sessions are created
for the state machines of each block that is a part of the VCCS Configuration block. In this case, these
are the state machines of the Cooling System Architecture and CP System Architecture blocks.

Step 2. Analyzing the integrated behavior of the system configuration

When the model execution begins, monitor the simulation progress until both state machines rest in the
Off states. The following figure shows what happens when you fire the transition between the Off and On
states in the CP System States diagram.

As a result of the manually fired transition, the CP System Architecture block enters the Initializing state. On
entering that state, it sends out the Start signal via the p2 proxy port (see the following figure).

As the Subsystems Communication in VCCS diagram defines, the Start signal is received by the Cooling
System Architecture block.

MagicGrid® Book of Knowledge | 285

Acceptance of the Start signal automatically fires the transition between the Off and On states displayed in
the Cooling System States diagram. As a result, the Cooling System Architecture block enters the Initializing
state. On entering the Initializing state, it performs the Perform System Check activity and sends out the
coolStatus signal via the p4 proxy port. Remember, the coolStatus signal has the coolOK attribute with the
true or false value.

286 | MagicGrid® Book of Knowledge

As the Subsystems Communication in VCCS diagram defines, the coolStatus signal is received by the CP
System Architecture block.

The CP System Architecture block accepts the coolOK signal, reads its attribute value, and decides whether
or not both subsystems can switch to the Idling state. For this, it performs the Evaluate Status activity.

As a result, the CP System Architecture block sends out either the SysOK or SysNotOK signal via the p2 proxy
port (see the following figure). As the figure shows, the same signal in each case is also sent internally,
this time to fire one of the transitions of the CP System Architecture block: either to the Idling or to the Off
state.

MagicGrid® Book of Knowledge | 287

As a result, both blocks capturing the logical subsystems move to the Idling states of their state machines
(see the following figure). Here they wait for the DesiredTmpSet signal to proceed.

288 | MagicGrid® Book of Knowledge

  If you want to display two or more diagrams on a single diagram pane, create a content diagram,
drag all the diagrams you want to see to its pane from the Model Browser, and convert each
diagram shape into the diagram overview shape (see the following figure). For more information,
see the latest documentation of the modeling tool.

This analysis proves that the integrated system behavior model meets the relevant system and subsystem
requirements.

System Configuration Behavior done. What's next?

After you finish the analysis of the integrated system behavior model, you can switch to the last task
of Building system configuration model and calculate more precise values of system parameters than in
the LSA model. Chapter System Configuration Parameters describes how to do this.

MagicGrid® Book of Knowledge | 289

System Configuration Parameters
What is it?

The system configuration model enables you to relate the calculation methods from different levels of
detail, such as one from the system with some from the subsystem level. You can define more specific
input values and get more precise calculation results than in the LSA model. For example, in the LSA
model, you had to specify the energy consumption of the Cooling System, while in the system
configuration model it can be derived by summing up the energy consumption of its logical components.

Who is responsible?

The system configuration parameters can be analyzed and verified by the Systems Architect or Systems
Engineer, who is the head of the entire systems engineering project and is responsible for the smooth
integration of the subsystems and their components into the whole; the same person who owns the LSA
model of the SoI. This task can also be performed by the Systems Integrator, a subordinate of the Systems
Architect.

How to model?

Usually you don't need to capture any calculation method in the system configuration model. You've done
that in the LSA model and also in the LSSA models. You've also defined the system, subsystem, and
component parameters there. You don't need to re-capture them in the system configuration model as
well.

However, you may need to update the bindings. For example, you might need to bind a constraint
parameter specified in the LSA model to the value property created in the LSSA model. In this case, the
infrastructure of the SysML parametric diagram helps a lot.

290 | MagicGrid® Book of Knowledge

Tutorial
• Step 1. Redefining the MoE
• Step 2. Creating a parametric diagram
• Step 3. Updating bindings

Step 1. Redefining the MoE

Let's say you want to check how the energy consumption of the logical components of the Cooling
System affects the total energy consumption of the whole VCCS. For this, you need to do a few things.

First of all, redefine the totalEnergyConsumptionVCCS value property inherited from the VCCS block. The
redefinition is not necessary for calculations, as you can perform them with an inherited value property.
However, you need to redefine the value property to be able to create a satisfy relationship from the value
property to the relevant system requirement.

To redefine the totalEnergyConsumptionVCCS value property

1. Open the VCCS Configuration Structure diagram you created in step 2 of this cell tutorial.
2. Right-click the VCCS Configuration block and select Specification.
3. On the left of the open dialog, click Properties.
4. Under the Value Properties category, select the totalEnergyConsumptionVCCS value property click
the Redefine button below.
5. As a result, the selected value property is redefined and is now no longer decorated with the caret
symbol (“^”), which indicates the inheritance. Don't change anything.

6. Click the Close button.

Now you're ready for the next step.

Step 2. Creating a parametric diagram

The next thing to do is relate the value property capturing the total energy consumption of the Cooling
System to the input parameter of the constraint property, which defines the method for calculating the
total energy consumption of the whole VCCS.

MagicGrid® Book of Knowledge | 291

Creating a SysML parametric diagram could be a good start to establish this relationship. Remember, this
type of diagram can only be created under a block. This time, it is the VCCS Configuration block.

To create s SysML parametric diagram for the VCCS Configuration block

1. In the Model Browser, right-click the block and select Create Diagram.
2. Select SysML Parametric Diagram in the Create Diagram dialog below. A new diagram is created, and
the Display Parameters/Properties dialog is opened.
3. In the dialog:

a. Leave the check box near the :Total Energy Consumption constraint property selected.

 This constraint property defines the method for calculating the total energy consumption
of the VCCS. You already specified it for the VCCS block in the LSA model. It belongs to the
VCCS Configuration block, too, because it is a sub-type of the VCCS block and inherits all its
features from the VCCS block, including the constraint properties.

b. Click the button near the cooling : Cooling System Architecture part property to expand its
contents.
c. Click to select the check box near the /totalEnergyConsCool : kWh value property.

 This subsystem parameter is derived from component parameters, by summing up the

energy consumed by the Compressor, Condenser, Evaporator, etc. When calculating the
total energy consumption of the VCCS, you need to replace the energyCons: kWh value
property specified in the LSA model with this value property.

d. Click the button near the cp : CP System Architecture part property to expand its contents.
e. Click to select the check box near the energyCons : kWh value property.

 As you don't have yet the method defining the energy consumption of the CP System,
you can take the value property defined in the LSA model. The same logic applies for other
logical subsystems.

f. Follow steps d and e to select the energyCons : kWh value properties of the heating : Heating
System and hmi : HMI System part properties.
g. Make sure the check box near the /totalEnergyConsVCCS : kWh value property is selected.
h. Make sure everything else is unselected.
i. Click OK to close the dialog.

4. In the Model Browser, type VCCS Configuration Energy Consumption to name the newly created
diagram.

Once you're done, the VCCS Configuration Energy Consumption diagram should look like the following
figure.

292 | MagicGrid® Book of Knowledge

  If you want to, you can change the dot notation of displaying value properties to the nested one.
For this, select all the shapes, except the one of the totalEnergyConsVCCS value property, then right-
click them and select Refactor > Convert to Nested Parts.

Step 3. Updating bindings

Now you can update the bindings and re-run the calculations.

To update the bindings in the parametric diagram

1. Select the constraint property and click the Display All Parameters button on its smart
manipulator toolbar. Parameters and binding connectors to corresponding value properties are
displayed on the diagram.

MagicGrid® Book of Knowledge | 293

 2. Select the shape of the /totalEnergyConsCool : kWh value property within the shape of the cooling :
Cooling System Architecture part property.
3. Click the Binding Connector button on its smart manipulator toolbar.
4. Click the shape of the energyCons1 parameter. The new binding connector is created, and the
parametric diagram is ready to calculate the total energy consumption of the VCCS encountering the
energy consumed by the components of the Cooling System.

5. Repeat steps 2 to 4 to establish another binding connector between the energyCons value property of
the CP System Architecture block and the energyCons2 parameter.
6. Repeat steps 2 to 4 to establish one more binding connector, this time between the
/totalEnergyConsVCCS value property of the VCCS Configuration block and the totalEnergyCons
parameter.

Now, if you run the simulation on this diagram, you can enter diverse values of energy consumption for
subsystems and components, and see the result. The result values can be stored in the model (see step
7 in the System Parameters tutorial).

294 | MagicGrid® Book of Knowledge

When you have the solution architectures of all logical subsystems integrated into the system
configuration model, you can update other bindings in this diagram.

System Configuration Parameters done. What's next?

Finishing the system configuration parameters means that you're almost done with the system
configuration model. Only traceability relationships to system requirements specification are left to deal
with.

MagicGrid® Book of Knowledge | 295

Traceability from System Configuration to System Requirements
What is it?

To ensure that the system configuration model is complete, and all system requirements are considered
and implemented, you need to perform the analysis and specify traceability relationships from the system
configuration model to the system requirements specification. These are satisfy relationships.

Who is responsible?

Traceability relationships can be specified by the Systems Architect or Systems Engineer, who is the head
of the entire systems engineering project and is responsible for the smooth integration of the subsystems
and their components into the whole.

How to model?

As you already know, a Satisfy Requirement Matrix provides the best infrastructure for specifying
satisfy relationships. The matrix is pretty much the same as the one created in the LSA model. Only the
column scope is different: here you should select the VCCS Configuration block instead of the VCCS block.

The matrix in the following figure shows that a few system requirements are not yet satisfied. For this,
you need to build the LSSA of the Heating and HMI Systems and then integrate them into the system
configuration model.

296 | MagicGrid® Book of Knowledge

Tutorial

Steps for specifying the traceability relationships are equivalent to steps of the Traceability to System
Requirements tutorial:
• Step 1. Creating a matrix for capturing satisfy relationships
• Step 2. Capturing satisfy relationships

Solution Domain done. What's next?

Having the complete system configuration model means that you're done with the entire Solution
domain, which indicates that you're ready to move to the Implementation domain and ready to start
considering your system real.

MagicGrid® Book of Knowledge | 297

IMPLEMENTATION DOMAIN
After the logical architecture and design of the whole system is defined and the optimal system
configuration is selected (by means of the trade-off analysis), the system is ready for implementation. At
this point, you should think of the system as real, and not abstract, as you did with the problem and
solution domains.

As you can see in the following table, the implementation domain is only partially covered by the
MagicGrid framework. The framework only defines how to specify implementation requirements for the
SoI. However, detailed physical design of the SoI is not a part the MagicGrid framework. The next steps of
the SoI development can be accomplished by using such software as electrical and mechanical CAD tools
or automated code generation tools, among others.

Implementation Requirements
What is it?

In order to implement the system of interest, you first need to have detailed requirements to follow while
producing the system's detailed physical design. These requirements are based on the logical architecture
of the system of interest, including the logical architectures of all its subsystems and even more precise
units.

The implementation requirements specification is given to the engineers of various disciplines, who are
responsible for designing different aspects of the system under implementation.

Who is responsible?

Implementation requirements specification can be produced by the Systems Designer.

298 | MagicGrid® Book of Knowledge

How to model?

In the modeling tool, an item of the implementation requirements specification can be stored as a SysML
requirement, which has a unique identification, name, and textual specification. Implementation
requirements must be formal and written down by following certain guidelines. These may include writing
in short and clear sentences only; avoiding conditional keywords, such as if, except, as well; use of
the shall keyword; and incorporating tables. For more information on writing good textual requirements,
refer to the INCOSE Systems Engineering Handbook.

Managing implementation requirements is similar to managing system, subsystem, and component

requirements. As with those requirements, implementation requirements can be categorized and
organized into groups, sub-groups, and so forth. SysML supports a wide variety of requirement categories,
such as physical, functional, and performance, etc. For grouping requirements, use the containment
relationships among them. The SysML requirements that capture grouping items might not contain any
text. As you can see in the following figure, which shows an example of implementation requirements for
the Cooling System, there are four groups of requirements: Electronics & Electric, Mechanics, Fluids, and
Generic Characteristics.

As shown in the following figure, each implementation requirement should refine one or more elements
from the system configuration model, or be derived from at least one requirement from the solution
domain. For this, a matrix for specifying both types of relationships can be created (similar to the one
in step 4 of Subsystem Requirements).

MagicGrid® Book of Knowledge | 299

As you can conclude by analyzing the matrix, the implementation requirements specification is not yet
completed. It will be considered to be complete when each element within the system configuration model
is refined by one or more implementation requirements and each solution domain requirement is covered
by at least one implementation requirement.

Having derivation relationships among requirements in different domains, you can easily trace from very
detailed requirements to very abstract requirements and see what items determine the creation of other
items. To perform this impact analysis (either downstream or upstream), we recommend utilizing the
infrastructure of the Requirement Derivation Map, one of the predefined relationship maps in
the modeling tool.

Tutorial

Steps of this cell tutorial are adequate to steps of the Subsystem Requirements tutorial:
• Step 1. Creating and organizing a model for subsystem requirements
• Step 2. Creating a diagram for subsystem requirements
• Step 3. Specifying subsystem requirements
• Step 4. Specifying traceability relationships

The difference is that in step 4 you establish traceability relationships to the subsystem requirements for
the Cooling System and to the relevant elements of the system configuration model.

What's next?

You are ready to develop the detailed physical design of the system. These activities are outside the scope
of the MagicGrid approach.

300 | MagicGrid® Book of Knowledge

ANNEX A: SAFETY & RELIABILITY
ANALYSIS
By Andrius Armonas, PhD, Elona Paulikaitytė, Osvaldas Jankauskas, Tomas Juknevičius

Reliability is the ability of a functional unit to perform a required function under given conditions for a
given time interval (ISO/IEC 2382:2015 Information Technology). Safety is freedom from unacceptable
risk (IEC 61508:2010 EEPE safety-related systems).

SysML, as one of the key MBSE components, has a good foundation for capturing requirements,
architecture, constraints, views, and viewpoints. However, SysML does not provide the necessary
constructs to capture safety and reliability information in the system model. The Cameo Safety and
Reliability Analyzer plugin adds support for those missing constructs. The following safety and reliability
standards are supported:
• The failure mode, effects, and criticality analysis (FMEA) according to IEC 60812:2006 standard
• Hazard analysis according to the following medical standards:
• IEC 62304
• ISO 14971:2007, Corrected version 2007-10-01 (Medical devices - Application of risk
management to medical devices)

The ISO 26262 Functional Safety plugin, used in conjunction with the Cameo Safety and Reliability
Analyzer plugin, supports ISO 26262 (Road vehicles – Functional safety) standard.

This annex describes how FMEA, supported by the Cameo Safety and Reliability Analyzer plugin, is used
to enrich a systems model created using the MagicGrid method with safety and reliability information.

The figure below depicts the general safety and reliability analysis workflow. Reliability and safety
analysis is performed against the systems model (stakeholder needs, system behavior, and structural
parts) and identified as separate items, which are called FMEA items when the FMEA method is used.
These items are then ranked and the most critical are addressed by introducing new safety and reliability
requirements. Subsequently, changes to the architecture and design of a system are done by introducing
new behaviors and altering the structure of the system to support the identified safety and reliability
requirements. Newly introduced modifications to the systems model need to undergo the same safety and
reliability assessment.

Safety and reliability analysis is either performed by a safety engineer or a systems engineer.

MagicGrid® Book of Knowledge | 301

Stakeholder needs: safety and reliability requirements
Safety and reliability requirements, as identified by stakeholders, can be captured as SysML requirements.
See Chapter Stakeholder Needs for more information on how to capture stakeholder needs in requirement
tables. Capturing safety and reliability requirements using SysML requirements does not require any
extensions of the SysML language; a grouping requirement with a corresponding name is sufficient. As
you can see in the following figure, the SN-1.3 Safety & Reliability requirement acts as a container for
safety and reliability requirements that originate from stakeholders.

Conceptual FMEA in Black Box view

The Black Box view should contain qualitative conceptual component and functional failure mode analysis
based on the use of the FMEA method. The purpose of the conceptual FMEA is to take actions to eliminate
or reduce failures that originate at the Black Box view. The reduction of failure is achieved by introducing
new safety and reliability requirements as a result of the conceptual FMEA analysis and then refining and
satisfying these requirements with the changes downstream.

Traditional FMEAs are done in a spreadsheet-like environment, supported by the Cameo Safety and
Reliability Analyzer plugin as a primary way to manage FMEAs. Each row in the table is called an FMEA
item which references (or semantically groups) classical FMEA elements, such as failure cause, failure
mode, local and final effects, and mitigation.

The Black Box conceptual FMEA table contains the following columns:

Column name Purpose

Id A unique ID assigned to each FMEA Item.
Name A user-friendly name of an FMEA Item.
Item The part undergoing analysis related to a particular FMEA Item.

Cause of Failure An element indicating the architecture weakness causing a Failure

Mode. An FMEA Item, e.g., a single row of an FMEA Table, can have
multiple Causes of Failure.

Failure Mode An element describing the specific manner in which a system could
potentially fail to meet the design intent.

302 | MagicGrid® Book of Knowledge

 Column name Purpose
Local Effect of Failure An element describing the effect that a Failure Mode has on the
system element under consideration. An FMEA Item, e.g, a single row
of an FMEA Table, can have multiple Local Effects of Failure.
Final Effect of Failure An element describing the effect that Failure Mode has on an end-user
or environment. You can specify multiple Final Effect of Failure values
for a single FMEA Item.

Refines The reference to a stakeholder requirement which the FMEA Item

refines.
Mitigation The reference to a requirement that mitigates a failure.

The following figure shows an FMEA table supported by the Cameo Safety and Reliability Analyzer plugin,
which represents the conceptual FMEA. The item is the :VCCU unit, which is the system being modeled.
Every FMEA item refines safety and reliability requirements from stakeholder needs. For example, the
F-2 FMEA item refines the SN-1.3.2.2 Biofouling stakeholder requirement. The refining FMEA item
identifies how the :VCCU unit can fail to meet the need for passengers of a car to not be exposed to any
toxic materials accumulated in the climate control unit:
• First, potential causes of failure are identified; in this particular case, direct contact of a passenger
with toxic materials accumulated in the climate control unit.
• Then, failure modes are identified; in this particular case, passengers not being able to operate
the :VCCU unit (since the unit produces toxic materials).
• Local and final effects of failure are either introduced or reused from the library as they tend to
repeat across product lines or different products.
• Finally, new safety or reliability requirements are introduced; in this particular case, a requirement
to have air filters so that toxic materials would not accumulate in the climate control unit.

As already mentioned, FMEA items can have a name. The name is a resumptive sentence that conveys the
essence of analysis performed at concrete FMEA items. The purpose of the FMEA item name is to optimize
the search of FMEA items in the model to promote the reuse and facilitate the management of FMEA
items.

Causes may not be specified at all if for any reason or cause within the created system a failure mode
occurs and only that fact is important to introduce a mitigation safety or reliability requirement (see the
F-1 FMEA item). The newly introduced safety and reliability requirements then need to be linked to the
newly created or updated behavioral or structural models in both Black Box, White Box, LSA, Subsystem,
or Component views.

Safety and reliability analysis workflow is repetitive, thus failures that originate at the Black Box view can
resonate through the whole SoI, from the problem domain down to implementation. In order to capture
the failure scenarios, FMEA cause-effect chains are used that can be depicted using Relation Map
diagrams. For example, the Biofouling Relation Map (see the following figure) depicts the model elements
and FMEA items that originate during FMEA analysis starting from the SN-1.3.2.2 Biofouling stakeholder
requirement.

MagicGrid® Book of Knowledge | 303

This Relation Map is going to be expanded gradually during the FMEA analysis whenever the failure
effects (of newly introduced changes in the system architecture) are identified leading to an unintended
scenario.

Functional FMEA in Black Box view

The functional FMEA focuses on behaviors of the system rather than refining stakeholder requirements.
The Black Box functional FMEA table contains the following additional columns comparing with
conceptual FMEA table:

Column name Purpose

Subsystem incoming and/or outgoing flow connecting action allocated to part defined
in the Item column.

Source (derived) source action of flow.

Target (derived) target action of flow.

The Black Box functional FMEA is depicted in the following figure. As you can see, the Item column is still
referencing :VCCU, just as in the conceptual FMEA. However, the Subsystem column represents object
flows that are analyzed for potential failures. The source for this analysis is the activity diagrams specified
for use cases (see the Use Cases section). The Provide Comfortable Temperature activity from the Use
Cases section is shown below the functional FMEA table. When performing a functional FMEA, all
incoming and outgoing flows into the swimlane representing :VCCU, as well as actions being performed by
it, are being analyzed for potential failures (e.g., how would the system perform if one of the incoming
flows that flows into the given action did not produce the expected output?).

In this particular example, we only focus on object flows related to the Check System and Reach Desired
Temperature actions, but the same can also be done for control flows and actions. The flows being
analyzed are color-coded in orange in the activity diagram below the FMEA table. The actions that these
flows connect are represented in the Source and Target columns in the functional FMEA table. The Refines
column represents the analyzed use case. The Mitigation column has the same purpose as in the
conceptual FMEA table.

The F-7 FMEA item analyzes the object flow coming from the Provide Air into the Reach Desired
Temperature action by assessing the possible impact of the failure to receive air to the system when it is
trying to reach the desired temperature, which is captured as the cause in the Cause of Failure column in
the FMEA table (VCCU cannot suck air from cabin). As a result, the failure mode Loss of function occurs,

304 | MagicGrid® Book of Knowledge

resulting in the local effect being VCCU not operational and the final effect on passengers being overheated
or undercooled. To mitigate this risk, a new safety requirement for automatically shutting the system
down in fail-safe mode is introduced (see the requirement 6). As in the conceptual FMEA, the newly
introduced requirements are addressed either at the problem level or downstream. Other functional FMEA
items are analyzed in a similar fashion, demonstrated with the F-7 FMEA item.

To demonstrate how safety and reliability requirements are addressed at this level or downstream, let's
have a look at the newly introduced Provide Limp Mode safety requirement. The original MagicGrid activity
diagram for the Provide Comfortable Temperature is shown as follows.

To address the Provide Limp Mode safety requirement, the new :VCCU function called Start Climate Control in
Fallback Mode @72F is introduced. This mode would provide the required default 72 °F temperature as long
as the :VCCU has not yet accepted the command to start or received the desired temperature.

MagicGrid® Book of Knowledge | 305

Functional FMEA coverage analysis
It is important to cover all incoming and outgoing flows to the :VCCU unit with the functional FMEA; thus,
a traceability matrix should be built (see the following figure). In this matrix it is clearly visible how many
of the flows still need to be covered by FMEA items.

306 | MagicGrid® Book of Knowledge

Addressing safety and reliability requirements in
Problem domain
All newly introduced safety and reliability requirements are contained in one package (see the following
figure).

The table below lists safety and reliability requirements. Requirements 8 and 9 are created as a result of
the FMEA analysis at the White Box level (see the sections below).

Safety and reliability requirements can be addressed at all levels, from the problem statement to the
component design. For example, the safety requirement 1 Use Flame-Resistant Materials cannot be
addressed at the Problem level since this level does not contain actual system design. This requirement
could be addressed at the logical subsystem level since specific materials used in the design of a system
are known at this level. Another example is the requirement 2 Air Filters, which can be addressed at the
White Box level and downstream as well.

The original MagicGrid activity diagram for the Reach Desired Temperature activity (at the White Box view) is
depicted below. The air is flowing directly to the actions for drawing air from the cabin.

MagicGrid® Book of Knowledge | 307

To address the 2 Air Filters safety requirement, the new :Filtering logical subsystem needs to be added with
an air filtering function and corresponding flows to Filter Air flowing into :Cooling and :Heating subsystems
(see the following figure).

308 | MagicGrid® Book of Knowledge

The Satisfy relationship is created from the Filter Air action and :Filtering subsystem to the 2 Air Filters
safety requirement.

After making the aforementioned changes to the model, the Biofouling Relation Map is updated with the
newly introduced model elements that satisfy the 2 Air Filters safety requirement.

MagicGrid® Book of Knowledge | 309

Let's look at another example where the 5 Shut Down the System in Fail-Safe Manner when Unable to Read
or Accept Ambient Temperature safety requirement is addressed. The new action called Fail Safe is allocated
to :Processing subsystem and the Fail Safe signal is emitted by the Fail Safe action to the outside of the SoI.
The Fail Safe signal is one of the sub-types of CU Command signal that could flow through an existing
interface, as shown in the following figure.

310 | MagicGrid® Book of Knowledge

FMEA at the White Box view
Functional and conceptual FMEA needs to be performed at the White Box level just as it is done at the
Black Box view: all functions, their incoming and outgoing flows, and conceptual subsystems need to be
examined, as well as new conceptual subsystems and functions introduced to address safety and reliability
requirements. For example, the :Filtering subsystem has to undergo the conceptual subsystems FMEA
analysis (see the following figure). As a result of this analysis, new safety features need to be installed to
address the safety of the :Filtering conceptual subsystem (since it can cause safety issues). For this, two
new safety requirements 8 and 9 are created.

For example, the F-9 FMEA item analyzes how the :Filtering subsystem can fail to protect the passenger
from contact with toxic materials that may accumulate in the :VCCU, resulting in possible death from fire or
poisoning from smoke due to a clog in the filter caused by external factors such as dust, leaves, or trash.
As a result, the failure mode Reduction of function occurs, meaning the reduced ability of :Filtering
subsystem to filter the air, resulting in local effects on the :VCCU (overloading or catching on fire), and final
effects on the passenger, such as being burned, poisoned by smoke, or having an accident while driving,
among others. To mitigate these risks a new safety requirement is created to detect any clog in the :Filterin
g subsystem.

The newly introduced safety requirements need to be addressed by making modifications to the systems
model. The following figure depicts the Test Air Flow function that addresses these two requirements by
testing incoming air flow. The diagram also contains the other safety requirements introduced at the Black
Box level that are addressed and satisfied by appropriate functions.

MagicGrid® Book of Knowledge | 311

After the :Filtering subsystem examination, additional failures were found following fresh changes in the
model. It all boils down to a sequence of possible failures that can affect lower-level system elements and
model elements that mitigate the risks. The following figure illustrates the enriched cause-effect chain
after the :Filtering subsystem conceptual subsystems FMEA analysis.

Addressing safety and reliability concerns in the LSA

At the solution level, safety and reliability requirements are derived from the problem domain. The
diagram below shows a new Safety and Reliability grouping requirement that has safety and reliability
requirements underneath.

312 | MagicGrid® Book of Knowledge

These safety and reliability requirements originate from the problem domain, as shown in a Derive
Requirement Matrix called System Requirements to Stakeholder needs, as shown in the following figure.

These requirements, just like safety and reliability requirements in the problem domain, need to be
addressed by making modifications to the model. For example, to address the SR-1.12.7 Air Filters

MagicGrid® Book of Knowledge | 313

requirement, a Filtering System block needs to be created with corresponding ports and parts in the LSA
level (see the following figures).

The «abstraction» relationship is used to declare that the Filtering System in the LSA is derived from the
conceptual subsystem called Filtering in the problem domain analysis model. This relationship is illustrated
in the LSA to Problem Domain matrix provided below.

314 | MagicGrid® Book of Knowledge

The total energy consumption of the VCCS equals a sum of energies consumed by all its subsystems,
meaning the total energy consumption expression has to be supplemented with the energy consumed by
the Filtering System as well (see the System Parameters section). The constraint expression can be
defined as follows:

totalEnergyCons = energyCons1 + energyCons2 + energyCons3 + energyCons4 + energyCons5

where energyCons5 constraint parameter may stand for the energy consumed by the Filtering System. The
energyCons value property needs to be created for the Filtering system and an additional energyCons5
constraint parameter specified for the Total Energy Consumption constraint block. Once defined, the binding
between the former value property and the constraint parameter needs to be established to enable the
constraint expression to perform calculations on the modified model.

After these changes are complete it is time to specify the solution architecture of Filtering System and
revisit other VCCS subsystems in order to ensure integrity at the system level.

Addressing safety and reliability concerns at the

subsystem level

Cooling Subsystem level

As mentioned in the Functional FMEA in the Black Box view section, the newly introduced requirements
can be addressed either at the problem level or downstream. The SR-1.12.1 Use Flame-Resistant Materials
safety requirement is one of the requirements that was introduced early at the problem domain level but
could not be addressed due to a lack of specific materials in the design of the SoI. However, the Cooling
subsystem architecture is detailed enough to be able to address this requirement.

The Cooling subsystem has already specified interactions for exchanging the refrigerant. The physical
conditions of the refrigerant are already captured (see following figure). However, the SR-1.12.1 Use Flame-

MagicGrid® Book of Knowledge | 315

Resistant Materials safety requirement introduces the Hazardous Materials Identification System (HMIS)
flammability class that the used material (in this case Refrigerant) shall comply with.

Additional Cooling Subsystem physical requirement called SSR-1.1.4 Use Flame-Resistant Refrigerant is
introduced to address the SR-1.12.1 Use Flame-Resistant Materials safety requirement.

316 | MagicGrid® Book of Knowledge

To satisfy this safety requirement the modification on the refrigerant type hierarchy needs to be
performed. Therefore, the concrete R-134a refrigerant type can be defined as a block which can be
captured as a sub-type of the each physical condition of the Refrigerant. The following figure contains the
changes that were performed and illustrates how the SSR-1.1.4 Use Flame-Resistant Materials requirement
is addressed.

Filtering Subsystem level

The engineering team has to begin working on logical subsystem architectures (LSSA) after the FMEA
analysis has taken place and the new :Filtering subsystem has been identified (see Chapter Building the
logical architecture of subsystems). As noted in the Building the logical architecture of subsystems
section, an LSSA of each logical subsystem should be stored in a separate model. Filtering System can also
have its own FMEA analysis.

Subsystem Requirements

Let's say that the engineering team working on the solution architecture of the Filtering System receives
the following subsystem requirements from the System Architect:

Matrices and maps can be used to establish and review the cross-cutting relationships between
the :Filtering subsystem requirements and system requirements. The matrix below captures which more
concrete :Filtering subsystem requirements were derived from what system requirements.

MagicGrid® Book of Knowledge | 317

A Relation map can be created, where blue arrows represent the relationships between requirements from
different domains, red arrows represent the relationships between failures and requirements that mitigate
the failures, and cyan arrows represent the refinement relationships between failures that were created to
refine requirements from different domains. All these relationships illustrate the FMEA cause-effect chains
from initial stakeholder requirements down to Filtering System requirements in accordance with safety
and reliability decisions made during FMEA analysis. The following figure shows a portion of the relation
map called from the SN-1.3.2.2 Biofouling stakeholder requirement down to the Filtering System.

Subsystem Structure

The internal structure of the Filtering System can be identified, then operations among its components,
and how logical components of the Filtering System interact with other subsystems of the same SoI can
be specified. To find the detailed steps refer to Chapter Subsystem Structure.

Capturing components of the Filtering system

As defined by the subsystem requirements and based on the engineering team experience, the Filtering
system shall be composed of:
• Fan
• Membrane Type Particle Filter
• Pressure Sensor
• Voltage Comparator With Threshold

318 | MagicGrid® Book of Knowledge

These can be specified as part properties of the Filtering System block. For this, the SysML block definition
diagram (bdd) is used, as shown in the following figure.

It should be noted that there is a need for the system to be constantly aware of filter status according to
LSA SSR 3.2.1 Filter Status requirements group. To be able to operate securely the Filtering System shall
monitor if the filter is ripped, clogged, and the severity of the clogging. Therefore, the pressure sensors are
installed before and after the air has blown through the filter. Then the pressure information is compared
by three dedicated voltage comparators to determine the status of the filter.

Specifying interactions within the Filtering System

In this step, we focus on specifying interactions between components of the Filtering System and outside
of the subsystem. To specify interactions between subsystem components the SysML internal block
diagram (ibd) is used as shown in the following figure. An ibd is created for the Filtering System
Architecture block. As the procedure in step 5 of the Subsystem Structure tutorial suggests, first, we start
specifying interactions via inherited ports:

1. One for letting the cabin air and its pressure flow into the Filtering System that is consumed by :
Membrane Type Particle Filter and before : Pressure Sensor components.
2. Second for letting the filtered cabin air out from the : Fan of the Filtering System.
3. Another dedicated to control the : Fan by passing over the signals: Fan1, Fan2, Fan3, Fan4, and Stop.
4. The last one is dedicated to emit the signals indicating the state of the Filtering System from voltage
comparators out of the subsystem.

MagicGrid® Book of Knowledge | 319

After interactions via inherited ports are captured the next step is to proceed with interactions among the
Filtering System logical components. For more details refer to the step 6 of the Subsystem
Structure tutorial.

Connectors are created between compatible ports to convey the following interactions:

1. The before : Pressure Sensor converts air pressure to an electrical signal (voltage) proportional to the air
pressure before filtering.
2. The : Membrane Type Particle Filter provides filtered air to : Fan and filtered air pressure to after :
Pressure Sensor.
3. The after : Pressure Sensor consumes filtered air pressure and produces the voltage signal proportional
to the air pressure after filtering.
4. The filtered air is consumed by : Fan in order to be circulated out of the system.
5. Pressure sensors produce the voltage signals to clogModerate : VoltageComparatorWithThreshold,
clogHigh : VoltageComparatorWithThreshold, and rip : VoltageComparatorWithThreshold components
where voltage signals of air are compared with the present voltage reference, called threshold before
and after the filtering.

After identifying interactions between the components and the outside of the SoI, the next step is
to update the Filtering System Structure bdd diagram to display new proxy ports on the blocks that capture
the Filtering System Architecture and its components (see the following figure).

320 | MagicGrid® Book of Knowledge

.

Subsystem Behavior

After identifying and capturing the structural architecture of the Filtering system, the next step is to
continue with behavioral architecture modeling. The SysML state machine and activity or sequence
diagrams are used to capture the behavior of the Filtering System.

Capturing states of the Filtering System

To capture the main states of the Filtering System the SysML state machine diagram is used, displayed
in the following figure analogously as the Cooling subsystem. The transitions between states are triggered
by a signal event, such as Start between the Off and On states or SysNotOK and SysOK signals between
Initializing, Off and Operating states.

Let’s say the engineering team working on the solution architecture of the Filtering System has identified
the following set of states in which the Filtering System can exist during the VCCS operation:

1. Off
2. On:
a. Initializing
b. Operating

The following rules describe transitions between those states:

• The first state of the subsystem is Off.
• The subsystem can move from the Off state to the On state and back.
• When in the On state, the subsystem goes to the Initializing state first.
• From the Initializing state, the subsystem can move either to the Operating state or to the Off state.

MagicGrid® Book of Knowledge | 321

As in the case of the Cooling System, these signals come from outside the Filtering System as well. The CP
System sends these signals to invoke the system. Thus, the reaction of that subsystem to the Start signal
has to be specified. The Filtering System in response sends the filterStatus signal with the status attribute
value set to filterOK if it is ready to start, and one of the values of filterStatusCloggedM, filterStatusCloggedS
or filterRipped, if it is not. The following figure depicts the response signal that is sent over the p1 proxy
port.

 filterStatusCloggedM stands for filter Status Clogged Moderately, and filterStatusCloggedS stands for
filter Status Clogged Severely.

322 | MagicGrid® Book of Knowledge

The Membrane Particle Filter is the concrete component that can physically be damaged and cause the
Filtering system failure to initiate. Therefore, the Membrane Particle Filter can have its own state machine
indicating the states in which it can exist, such as:

1. Filtering Normally
2. Partially Clogged
3. Severely Clogged
4. Ripped

These can be triggered by such environmental conditions as ProlongedOperationInDustyEnviron-

ments, LargeDebreeInAirIntake or ExcessVibration that are depicted as signal events on the transitions in the
Membrane Type Particle Filter States diagram. When the Membrane Particle defect is removed it can
resume working normally. For this, the Servicing signal is created to indicate that component was repaired.

To address requirements under the SSR-3.2.2 Fan States grouping requirement, another state machine is
created for Fan component of the Filtering System (see the following figure). The Fan State diagram
depicts the power states in which Fan can exist during system operation, such as:

1. Off
2. Blowing1
3. Blowing2
4. Blowing3
5. Blowing4

These states are triggered by signal events Fan1, Fan2, Fan3, and Fan4 to change the power of the Fan.
The Fan also can be turned off any time despite its Blowing state. The Stop signal is used for this transition

MagicGrid® Book of Knowledge | 323

to occur. The Fan is controlled by the CP System; thus, all triggering signals are being sent from the CP
System.

 The number along the Blowing state name represents the power of the fan, the bigger number
means a more powerful mode of the fan. Signal names that trigger these states are named
analogously for each of the states.

Specifying internal behaviors of the state

As mentioned in the step 5 of the Subsystem Behavior tutorial, every logical subsystem of the SoI can
perform one or more behaviors related to one or more of its states. The internal behavior of the Filtering
System is captured using the SysML activity diagram. To indicate that the Filtering System performs a
self-check and sends out the filterStatus signal with an appropriate attribute value, an activity diagram is
created that is specified as a do behavior on the Initializing state.

To specify that the CP System evaluates the status of all subsystems and sends the SysOk and SysNotOK
signals the CP System Evaluate Status state machine diagram should be updated. The following figure
captures the final view of the Filtering System States state machine diagram.

The change has to be reconciled with other engineering teams to ensure integrity at the system level.

324 | MagicGrid® Book of Knowledge

Subsystem Parameters

Subsystem parameters are used to capture quantitative characteristics of the Filtering logical subsystem
that can be measured. They are used to calculate MoEs for that logical subsystem to satisfy non-functional
quantitative subsystem requirements. Using the capabilities of the modeling tool, the model can be
executed to calculate MoEs using mathematical expressions to determine whether quantitative
requirements are satisfied or not. (See the Subsystem Parameters section.)

Specifying subsystem parameters for calculating the total energy consumption for filtering

According to the subsystem requirements specification and the MoEs identified for the Filtering System
captured in the problem domain and downstream, the formula for calculating the total energy consumed
by the Filtering System needs to be modeled. The SysML block definition diagram (bdd), called System
Parameters for Total Energy Consumption for Filtering, is used to specify constraint expressions and
parameters necessary to define totalEnergyConsFilter value property.

The Total Energy Consumption for Filtering constraint block depicted in the following figure has seven input
parameters, each for the energy consumption of a single component of the Filtering System, and therefore
requires seven component parameters to be defined in the model. Value properties should be specified for
each block that represents components of the Filtering System.

Let’s say the total energy consumption of the Filtering System equals the sum of energies consumed by
all its components. The constraint expression can be defined as follows:

totalEnergyCons = energyCons1 + energyCons2 + energyCons3 + energyCons4 + energyCons5 + energyCons6 +

energyCons7

where the following constraint parameters:

1. totalEnergyCons stands for the total energy consumed by the Filtering System
2. energyCons1 may stand for the energy consumed by the Fan
3. energyCons2 may stand for the energy consumed by the Membrane Type Particle Filter
4. energyCons3 may stand for the energy consumed by the before : Pressure Sensor

MagicGrid® Book of Knowledge | 325

 5. energyCons4 may stand for the energy consumed by the after : Pressure Sensor
6. energyCons5 may stand for the energy consumed by the rip : VoltageComparatorWithThreshold
7. energyCons6 may stand for the energy consumed by the clogHigh : VoltageComparatorWithThreshold
8. energyCons7 may stand for the energy consumed by the clogModerate :
VoltageComparatorWithThreshold
should be bound to corresponding value properties specified in the model. For this, the SysML parametric
diagram called Total Energy Consumption for Filtering is used, which is illustrated in the following figure.
The Max Energy Consumption for Filtering subsystem requirement requires that “The Filtering System shall
consume maximum 1.2 kWh”. To satisfy this requirement the constraint block is created and bound with
the totalEnergyConsFilter value.

This is a simple sensor model, converting input pressure in the range 0-2 atmospheres (0-200000 Pa) into
the voltage output signal in 1-10V range, that is depicted using parametric diagrams presented in the
following figure.

The second figure illustrates the simplified model for fan energy consumption depending on the : Fan state
and the input pressure, which in turn depends on the pressure drop in the membrane filter permeability).
The consumption is 37 Watts times the : Fan blowing intensity (Blowing1-Blowing4) times the linear
correction factor depending on the input pressure drop.

326 | MagicGrid® Book of Knowledge

This is a simple membrane filter model, which models pressure drop through the filtering membrane
depending on the filter state (clogged/ripped/normal). The pressure drop then affects downstream
systems, such as the energy consumption of the : Fan.

This is a model for a voltage comparator with a threshold. When the voltage at input1 is greater than
input2 + threshold, an output signal is fired.

The next step is to establish traceability relationships from the elements that capture the LSSA of Filtering
System to subsystem requirements.

MagicGrid® Book of Knowledge | 327

.Traceability to Subsystem Requirements

According to the Traceability to Subsystem Requirements section the cross-cutting relationships (such as
satisfy) need to be defined in order for subsystem requirements to be fulfilled. As you can see in the
following figure, the satisfy requirement matrix called LSSA to Subsystem Requirements is completely
fulfilled, meaning that all subsystem requirements are satisfied by one or more architecture elements.

Logical subsystems FMEA at the subsystem level

As mentioned in the conceptual FMEA in the Black-Box view section, the safety and reliability analysis
workflow is repetitive, thus safety and reliability analysis can also be performed for logical components of
the VCCS. The following figure represents logical subsystems FMEA and safety and reliability analysis
performed for the FilteringSystem with focus on the before : Pressure Sensor and after : Pressure Sensor
components. The pressure sensors have to undergo safety and reliability analysis since the Filtering
System may no longer detect malfunctions in case of the failure, and, as a consequence, the subsystem
would no longer be able to prevent safety issues in the system.

For example, the F-11 FMEA item analyzes how before : Pressure Sensor can fail to protect the passenger
from accidents, such as burns from fire or poisoning due to adverse conditions formed in the VCCU, such as
a clogged filter or VCCU on fire, that are called local effects of failure. These conditions can be caused by
wire breakage in the before : Pressure Sensor, which would result in an inability to provide data about the air
pressure to other subsystem parts. As a result, the failure mode Loss of Function occurs, meaning the lost
ability to determine the state of the pf : Membrane Particle Filter. To mitigate the risk a new safety
requirement is introduced to handle unknown filter status (see the SSR-4 Handle Unknown Filter

328 | MagicGrid® Book of Knowledge

State requirement in the table view above). Other logical subsystems FMEA items are analyzed in the
same way as described with the F-11 FMEA item.

The Filtering System logical subsystems FMEA table contains the following additional column compared
with conceptual FMEA tables in the problem domain:

Column name Purpose

Classification The classification of failures (FMEA Items) by certain aspects of a
system.
The newly identified SSR-4 Handle Unknown Filter State safety requirement (depicted in the following
figure) can be addressed at the Filtering System.

Addressing safety and reliability concerns at the

configuration level
So, let’s jump to the final phase: adapting the configuration model according to the changes introduced in
the safety & reliability analysis. Since the filtering subsystem architecture is completed, it should be
integrated by ensuring successful interaction with already defined subsystems. For this reason,
the interface compatibility verification is performed by specifying the interactions within the particular
configuration of the SoI.

As you can see, no incompatible interfaces were identified, and we can state that the Filtering System can
successfully communicate within the VCCS.

MagicGrid® Book of Knowledge | 329

ANNEX B: FROM SYSTEM OF SYSTEMS
TO SYSTEM ARCHITECTURE
By Aistė Aleksandravičienė, Gintarė Kriščiūnienė, Aurelijus Morkevičius, PhD

Model-based engineering practices have recently evolved beyond engineering a single system. In this age
of innovation, system thinking can be applied in the system of systems (SoS) environment to deal with
new challenges, such as autonomous traffic control, industry 4.0, the internet of things, etc.

The SoS is a SoI, whose elements are systems themselves; multiple organizations are involved in design,
maintenance, and operation of the SoS.

Therefore, when designing a SoS it is important to establish a digital environment for multiple
organizations to interact, especially when transitioning from SoS to systems architectures (SAs) (see the
following figure). In most cases this is a collaboration between the contracting authority and a contractor,
or the Original Equipment Manufacturer (OEM) and a supplier. This annex provides a detailed approach for
ensuring the digital continuity from SoS to SAs in the model-based environment.

System of Systems Engineering

Knowledge of an SoS can be captured using the Unified Architecture Framework® (UAF®) v1.1, whose
layout (also known as a UAF grid) is organized into rows and columns, where rows are domains and
columns are model kinds. The intersection of a row and column is called a view specification or just a
viewpoint.

In this annex, we only focus on the UAF Resources domain, which captures logical SoS architectures and
how they serve in the achievement of SoS capabilities. However, the UAF does not specify ways and
means to define the architecture for a single system. For this, we utilize MagicGrid with SysML.

330 | MagicGrid® Book of Knowledge

Transitioning from SoS to SA
The UAF Resources domain captures how different resources, including human resources and systems,
interact to implement operational needs and achieve capabilities of the SoS. Resources are grouped into
logical containers called capability configurations. Each capability configuration traces to the capability it is
meant to achieve at a specific period in time. There might be multiple capability configurations achieving
the same capability in different periods of time. At the same time, there might be several alternative
capability configurations. It is important to identify which version of the capability configuration to use as
an input to analyze the system context of the SoI.

MagicGrid: System Context

The system context or operating environment determines the external view of the SoI. Therefore, the
collection of elements in a particular system context includes the SoI itself and external systems (natural or
artificial), as well as users (humans, organizations) that interact with the SoI to exchange data, matter,
energy, or human resources.

As defined by MagicGrid, the system context is captured in the model as a SysML block. The SoI, external
systems, and supposed users are also captured as SysML blocks. While composite association relationships
enable modelers to relate them to a particular system context, the SysML internal block diagram provides
the infrastructure to display the participants of the system context and their interactions.

MagicGrid® Book of Knowledge | 331

Let’s see how to retrieve the system context of a single SoI from the SoS model. The Resources Structure
and Resources Connectivity viewpoints serve as the main source of information in this case (see the
following figure).

The system context of the SoI is initially defined in the SoS (UAF) model as the Resources Structure
viewpoint, which captures that SoI along with other systems within the same SoS. The System Context
view in the SysML model can be considered a subset of the Resource Structure viewpoint (see the
following figure).

Therefore, the block that captures the system context of the particular SoI in the SA (SysML) model
becomes a subtype of the capability configuration in the SoS (UAF) model. As a consequence, that block
inherits all the parts of the related capability configuration. The parts that participate in the system
context of the SoI are redefined with SysML blocks that capture copies of the SoI, actors, or external
systems in the SysML model (see the following figure). The redefinition of parts is necessary for further
modeling: this enables a modeler to update the system context information without modifying the
SoS (UAF) model. For example, you may need to add new ports or more specific ones than those defined
in the SoS (UAF) model.

332 | MagicGrid® Book of Knowledge

If there are ports/interfaces and interactions captured in the Resources Connectivity viewpoint of the SoS
(UAF) model, they show up in the SA (SysML) model, too. As shown in the following figure, this task
includes: (1) creation of relevant SysML interface blocks; (2) creation of the abstraction relationships from
the SysML interface blocks to UAF resource interfaces; (3) redefinition of UAF resource ports with SysML
proxy ports; and (4) redefinition of UAF resource connectors with SysML connectors. UAF elements (e.g.,
Data Elements) that capture the items to flow over resource connectors can be reused in the SysML model,
unless it requires more specific information. In that case, a new SysML element is created (5), and the
abstraction relationship pointing from it to the relevant element in the UAF model is specified (6).

MagicGrid: Stakeholder Needs

Stakeholder needs may include primary user needs, system-related government regulations, industry
standards, policies, procedures, internal guidelines, etc. They can be captured as SysML requirements,
which enable modelers to store the textual information in the SysML model. Stakeholder needs can be
grouped by the source they came from and categorized into functional and non-functional needs. The
infrastructure of the SysML requirements diagram or table can be utilized to display their hierarchy.

Stakeholder needs can be captured by interviewing stakeholders, giving them questionnaires, discussing
needs in focus groups, or studying documents written in diverse formats. In addition to these information
sources, the UAF model of the SoS can serve as input to the stakeholder needs as well. To be more precise,
the Resource Processes, Resource States, and Resource Interaction Scenarios Viewpoints can be used for
identifying functional stakeholder needs, while the Resources Measurements Viewpoint can be used for
non-functional (see the following figure).

MagicGrid® Book of Knowledge | 333

The subset of functions defined in the Resource Processes Viewpoint that are allocated (using the Is
Capable To Perform relationship, as shown in the following figure) to the system which becomes the SoI in
the SysML model motivate the creation of one or more functional stakeholder needs for that SoI. This also
applies to the case of states and interactions allocated to the relevant system in the UAF model.

When functional stakeholder needs are captured, SysML refine relationships can be established from
SysML requirements that capture the stakeholder needs to the relevant functions, states, or interactions in
the UAF model (see the following figure).

Measurements that are owned by the system that becomes the SoI in the SysML model motivate the
creation of one or more non-functional quantitative stakeholder needs for that SoI. When these are
captured, SysML refine relationships can be established from them to the relevant measurements in the
UAF model (see the following figure).

When use cases and MoEs of the SoI are captured in the SysML model, SysML refine relationships can help
to convey which stakeholder needs these use cases and MoEs refine.

334 | MagicGrid® Book of Knowledge

AFTERWORD
Thank you for taking the time to read the second edition of the MagicGrid BoK. Your feedback on the book
is more than welcome. All comments will be considered before releasing the third edition of the book.
Additionally, you can sign-up as a reviewer for the next editions of the MagicGrid BoK. Please express your
interest by sending an email to [email protected].

MagicGrid® Book of Knowledge | 335

GLOSSARY
Action

A SysML element that can capture a basic unit of the system functionality within the activity. It
represents some form of processing or transformation that will occur when the activity is executed
during the system operation. An action is represented as a round-cornered rectangle, with the name
displayed as a string inside that rectangle.

Activity

A type of behavioral SysML element that can contain a set of model elements, such as actions and
flows, and the activity diagram that represents these elements.

Activity diagram

A type of behavioral SysML diagram that can be used to describe the flow of control or the flow of
inputs and outputs among actions. Activity diagrams can be used to specify the black- and white-box
scenarios of the expected system functionality, and in combination with state machine diagrams to
design the system behavior.

Actor

A SysML element that can capture a role played by a person, an organization, or another system
when it interacts with the system of interest in the particular context. However, the MagicGrid
approach recommends using a block instead of an actor to capture these concepts, for the reasons
described in Use Cases.

Association

A type of SysML relationship that enables you to specify what actors interact with the system of
interest in the particular system context to invoke or participate in a use case. The notation for the
association is a solid line.

Binding connector

A special type of SysML connector that is used to represent an equality relationship between a
constraint parameter and a value property in a parametric diagram, in order to convey that the
constraint parameter receives the values from that value property. The notation for the binding
connector is a solid line.

Black box / white box

The black box represents an external view of the system. During the system analysis from the black-
box perspective, the system is viewed in terms of its inputs and outputs, without any knowledge of
its internal structure and behavior.

The white box represents an internal view of the system. During the system analysis from the white-
box perspective, the functional decomposition of the system is performed in order to identi-
fy conceptual subsystems (also referred as functional blocks in other methodologies.

Block

A SysML element of definition that can capture a basic unit of the system structure. A block is
represented as a rectangle with a stereotype block preceding the name of the «block» in the name
compartment.

336 | MagicGrid® Book of Knowledge

Block definition diagram (bdd)

A type of structural SysML diagram that can be used to define features of blocks and relationships
between blocks, such as associations, generalizations, and dependencies. It captures the definition of
blocks in terms of properties and operations, and relationships such as a system hierarchy or a
system classification tree. Block definition diagrams can be used to specify the structure of a system
or subsystem, and to define the interfaces and constraints of that system or subsystem.

Call behavior action

A special type of SysML action that can invoke another behavior: an activity, a state machine, or an
interaction. Call behavior actions enable you to decompose a higher-level behavior into a set of lower-
level behaviors.

Composite association / Directed composition

A special type of SysML association that enables you to specify that one block is a part of another
block. The notation for the composite association is a solid line between two blocks with a black
diamond on the composite end and an open arrowhead on the part end. Drawing a composite
association between blocks creates a part property for the block that appears on the black diamond
end of the relationship; the part property can have a name and is typed by the block that appears on
the arrowhead end of the composite association.

Component

An elementary object of the system structure. Components are often grouped into subsystems of the
system; they are identified by decomposing these subsystems. A component can be captured as a
block in the model.

Connector

A type of SysML relationship that is used to represent communication between two part properties in
an ibd. The connector may be established between part properties through various kinds of ports (as
long as these ports are compatible); for example, two part properties connected through proxy ports
convey that they exchange matter, energy, or data that can flow between them over the connector,
through compatible proxy ports. The notation for the binding connector is a solid line.

Constraint block

A SysML element of definition that can capture a constraint expression (either an equation or an
inequality). The constraint expression enables you to constrain the value properties of blocks. The
variables of the constraint expression are called constraint parameters. These constraint parameters
receive their values from the value properties that are being constrained. The notation of the
constraint block is the same as for the block, but with the «constraint» stereotype before the name of
the element.

Constraint parameter

A variable that appears in a constraint expression of the constraint block. It is captured as a property
of the constraint block in the model, and is represented as a string in the parameters compartment of
that constraint block.

MagicGrid® Book of Knowledge | 337

Constraint property

A usage of a constraint block in the context of some block or, in other words, a property of a block
typed by a constraint block defined somewhere in your model. It can be represented as a string in the
constraints compartment of that block in a bdd, or as a round-cornered rectangle in a parametrics
diagram. The name of the constraint property in both cases is displayed as a string followed by a
colon and the name of the constraint block which types that property; for example, mass : totalMass.

Control flow

A type of SysML relationship between a pair of actions. Control flows enable you to express the order
in which actions are performed in the associated activity. Control flow is represented as a dashed line
with an open arrowhead.

Decision/Merge

The decision node marks the start of an alternative flow in an activity. It is represented as a hollow
diamond, and must have a single incoming flow and two or more outgoing flows.

The merge node marks the end of the alternative flow in an activity. It is represented as a hollow
diamond, and must have two or more incoming flows and a single outgoing flow.

Exchange item

A concept that defines a type of matter, energy, or data that can flow between two structural objects
within a system; for example, between two part properties that capture the subsystems of the
system in an ibd. An exchange item can be captured as a block or a signal and then assigned as an
item flow to the connector. In that case, the connector is decorated with a filled-in triangle.

Failure Mode & Effects Analysis (FMEA)

The process of reviewing as many components, assemblies, and subsystems as possible to identify
potential failure modes in a system and their causes and effects.

Flow property

A property of an interface block. It can have a direction and a type. The direction of the flow property
specifies whether the property represents an input or an output of the structural object within the
model.

Fork/Join

The fork node marks the start of concurrent flows in an activity. It is represented as a line segment
(either horizontal or vertical), and must have a single incoming flow and two or more outgoing flows.

The join node marks the end of concurrent flows in an activity. It is represented as a line segment
(either horizontal or vertical), and must have two or more incoming flows and a single outgoing flow.

Function

A task performed by a system, or a certain part of it, to transform inputs to outputs. A single function
can be captured as an activity in the model.

Functional block

A concept that is used to define a structural object responsible for performing one or more functions
of the system. Functional blocks are identified by performing the functional decomposition of
the system. Functional blocks serve as inputs for defining the structure of the system of interest.
In MagicGrid, they are called conceptual subsystems.

338 | MagicGrid® Book of Knowledge

Functional decomposition / Functional breakdown

A set of steps to break down the overall function of a system into its smaller parts, in order to
identify conceptual subsystems (also referred as functional blocks in other methodologies).

Graphical User Interface (GUI)

A type of user interface that allows the use of icons or other visual indicators to interact with
electronic devices, rather than using only text via the command line.

Logical System Architecture (LSA) model

The core of a solution domain model. Based on the results of the problem domain analysis, it defines
the logical subsystems of the system of interest (as a single-level hierarchy). The purpose of the LSA
model is to identify work packages, one for each subsystem, and appoint them to separate
engineering teams.

Logical Subsystem Architecture (LSSA) model

A solution domain model that contains the logical architecture of a single logical subsystem and the
high-level design of its components, as well as even more precise units.

Interface

A hardware or software component that connects two or more structural objects of the system or the
system and the objects from its environment, for the purpose of passing matter, energy, or data from
one to the other.

Interface block

A SysML element of definition that can contain a set of flow properties which define the inputs and
outputs of some structural object captured in the model as a block; for example, a system, a
subsystem, or a component. The notation of the interface block is the same as for the block, but with
the «interfaceBlock» stereotype before the name of the element.

Internal block diagram (ibd)

A type of structural SysML diagram that can be used to capture the internal structure of a block in
terms of properties and connectors between these properties. Block definition diagrams can be used
to specify the system context and the interactions within the system.

Measure of Effectiveness (MoE)

A characteristic of the SoI in numerical format, MoE is a traditional term widely used in systems
engineering, and describes how well a system carries out a task within a specific context.

Model Browser

An element of the modeling tool GUI which represents the contents of the model repository. It is by
default located at the left of the main window of the modeling tool.

Model-based design (MBD)

A mathematical and visual method of addressing problems associated with designing complex
control, signal processing, and communication systems.

MagicGrid® Book of Knowledge | 339

Model-Based Systems Engineering (MBSE)

The formalized application of modeling to support system requirements, design, analysis,

verification, and validation activities beginning in the conceptual design phase and continuing
throughout development and later life cycle phases. MBSE focuses on creating and exploiting models
as the primary means of information exchange between engineers, rather than a document-based
information exchange.

Modeling tool

CATIA Magic software for MBSE:

- Magic Cyber Systems Engineer / Cameo Systems Modeler along with the Cameo DataHub plugin
installed

- Magic System of Systems Architect / Cameo Enterprise Architecture along with the Cameo DataHub
plugin installed

- Magic Software Architect / MagicDraw with the SysML and Cameo DataHub plugins

Object flow

A type of SysML relationship between a pair of actions. Object flows enable you to express what
matter, energy, or data flows through an activity from one action to another when the activity
executes. Object flow is represented as a solid line with an open arrowhead.

Package

A kind of SysML element that can contain other elements. Packages can be used to organize the
elements into logically cohesive groups.

Parametric diagram

A special type of ibd that displays the internal structure of a block, focusing on bindings between the
value properties of that block and the constraint parameters of the applied constraint block.
Parametric diagrams can be used to calculate the values of system parameters (by utilizing the
simulation capabilities of the modeling tool).

Part property

A usage of a block in the context of another block; in other words, a property of a block typed by
another block defined somewhere in your model. It can be represented as a string in the parts
compartment of that block in a bdd, or as a rectangle in an ibd. Using part properties enables you to
specify the internal structure of the system captured in the model as a block. A simplified ibd
(without proxy ports) can be used to specify the interactions within a system context; in that case the
system context is captured as a block, with a part property typed by the block that captures the
system of interest.

Presentation artifact

A visual representation of some fragment or aspect of the system model. It can be a diagram, matrix,
map, or table. Presentation artifacts work as data inputs to the model or data editors.

Proxy port

A type of property of the block, typed by an interface block defined somewhere in your model.
It represents the point at the boundary of the structural object represented by that block, through
which external entities can interact with that structural object. In other words, it is a usage of an
interface block in the context of another block. A proxy port is always represented as a small
rectangle on the shape of the block or part property, usually decorated with an arrow to indicate

340 | MagicGrid® Book of Knowledge

 whether the proxy port is for accepting inputs to the system, subsystem, or component, or providing
outputs.

Requirement

A SysML element that can capture a text-based statement that translates or expresses a need. It can
be used to capture stakeholder needs, or to specify the system requirements and detailed physical
requirements in the model.

Reliability

Ability of a functional unit to perform a required function under given conditions for a given time
interval (ISO/IEC 2382:2015 Information Technology).

Safety

Freedom from unacceptable risk (IEC 61508:2010 EEPE safety-related systems).

Sequence diagram

A type of behavioral SysML diagram that can be used to describe how the parts of a block interact
with one another via operation calls and asynchronous signals.

Signal

A SysML element that you can use to capture an exchange item.

Smart manipulator toolbar

A piece of the modeling tool GUI. It appears when a symbol of the element is selected on the diagram
pane.

Stakeholder

An individual or organization having a right, share, claim, or interest in a system or in its possession
of characteristics that meet their needs and expectations.

Stakeholder needs

Information gathered from various stakeholders of the system of interest. This information includes
primary user needs, system-related government regulations, policies, procedures, and internal
guidelines, among others. A single stakeholder need can be captured as a requirement in the model.

State

An element that captures a state of the system, in which it can exist during the operation. A state is
represented as a round-cornered rectangle with the name displayed as a string inside that rectangle.

MagicGrid® Book of Knowledge | 341

State machine diagram

A type of behavioral SysML diagram that can be used to specify how a structure within a system
changes state in response to event occurrences over time. It can be used to specify the behavior of
the system or its parts.

Subsystem

A secondary or subordinate system within a larger system. A subsystem can be captured as a block in
the model.

System context

A concept that determines an external view of the system. It must introduce elements that do not
belong to the system, but do interact with it. Besides the system itself (considered to be a black box),
the collection of elements in the particular system context can include external systems and users
(humans, organizations) that interact with the SoI in that context. A single system context can be
captured as a block in the model.

System

A combination of interacting elements that are organized into a complex whole for one or more
common purposes.

System architecture (SA)

A more abstract, conceptualization-oriented, global architecture, focused to achieve the mission and
operational concept of the SoI, in comparison to system design. It focuses on high-level structure in
systems and system elements.

System of interest (SoI)

The system whose life cycle is under consideration. The system that is being analyzed first from the
black- and then from the white-box perspective in order to produce the system requirements
specification.

System of systems (SoS)

A SoI whose system elements are systems themselves. Typically, these entail large-scale
interdisciplinary problems with multiple, heterogeneous, distributed systems.

Systems Engineering (SE)

Interdisciplinary tasks that are required throughout a system’s life cycle to transform stakeholder
needs, requirements, and constraints into a system solution.

System requirements

Requirements to be met by the design of a system, subsystem, or component. System requirements

are derived from stakeholder needs, which are more abstract requirements.

System parameter

A numerical characteristic of the system, subsystem, or component. It can be specified as a value

property of the block that captures that system, subsystem, or component. System parameters are
derived from MoEs and can be verified against system requirements.

Systems Modeling Language (SysML)

A general-purpose modeling language for systems engineering applications. It supports the

specification, analysis, design, verification, and validation of a broad range of systems and systems-
of-systems.

342 | MagicGrid® Book of Knowledge

Trade-off

A situational decision that involves diminishing or losing one quality, quantity, or property of a set or
design in return for gains in other aspects. In simple terms, a trade-off is where one thing increases
and another must decrease.

Transition

A type of SysML relationship that represents a change from one state to another. It can be triggered
by an event concurrency.

Unified Architecture Framework (UAF)

Defines ways of representing an enterprise architecture that enables stakeholders to focus on specific
areas of interest in the enterprise while retaining sight of the big picture. UAF meets the specific
business, operational and systems of systems integration needs of commercial and industrial
enterprises as well as the U.S. Department of Defense (DoD), the UK Ministry of Defence (MOD), the
North Atlantic Treaty Organization (NATO) and other defense organizations.

Use case

A type of SysML element that can be used to capture what capabilities people expect from the
system and what they want to achieve by using it within the particular system context. The notation
for the use case is an ellipse with the name (usually a verb phrase) inside that ellipse.

Use case diagram

A type of behavioral SysML diagram that can be used to define a set of use cases performed within a
particular system context; it represents the black-box view of the system of interest. It also displays
and can be used for creating associations between the use cases and the participants of the system
context, to specify who/what is responsible for invoking or participating in what use case.

Value property

A property of a block which can represent a quantity, a Boolean, or a string. It is displayed as a string
in the values compartment of the block, and can be used to capture measurements of effectiveness of
the system of interest (the «moe» stereotype must also be applied).

MagicGrid® Book of Knowledge | 343

BIBLIOGRAPHY
1. Chami M., Oggier, P., Naas, O., Heinz, M. “Real World Application of MBSE at Bombardier
Transportation”, SWISSED, Zurich, 2015. Available at https://goo.gl/zf5uvp.
2. Dandashi, F., Hause, M. C., "UAF for system of systems modeling", 10th System of Systems
Engineering Conference (SoSE), pp. 199-204, San Antonio, TX, 2015.
3. Delp, C., Lam, D., Fosse, E., Cin-Young Lee, “Model based document and report generation for systems
engineering”, Aerospace Conference, IEEE, 2013. Available at http://www.omgsysml.org/
View_Paper-IEEE_2013.pdf.
4. Estefan, J. INCOSE Survey of MBSE Methodologies, INCOSE TD 2007-003-02, Seattle, WA, USA, 2008.
5. Friedenthal, S., et al. A Practical Guide to the SysML, 4th Edition: The Systems Modeling Language.
Boston: MK/OMG Press, 2011.
6. Friedland, B., Herrold, J., Ferguson, G., Malone, R. “Conducting a Model Based Systems Engineering
Tool Trade Study Using a Systems Engineering Approach”, 27th Annual INCOSE International
Symposium, pp. 1087-1099, Adelaide, 2017.
7. Hallqvist, J., Larsson, J. “Introducing MBSE by using Systems Engineering Principles”, 26th Annual
INCOSE International Symposium, pp. 512-525, Edinburgh, 2016.
8. Hause, M.C., Thom, F., "Bridging the Chasm – Tracing from Architectural Frameworks to SysML", 17th
Annual INCOSE International Symposium, pp. 1317-1332, 2007.
9. Hause, M. C., Thom, F., "MoDAF and SysML – A Winning Combination", INCOSE UK Chapter – Spring
Symposium 2005, 2005.
10. Hoffmann, H.-P. “Systems Engineering Best Practices with the Rational Solution for Systems and
Software Engineering”, Release 3.2.1, February 2011. Available at https://www.slideshare.net/
billduncan/ibm-rational-harmony-deskbook-rel-312.
11. ISO/IEC/IEEE 15288:2015 “Systems and Software Engineering–System Life Cycle Processes”, Geneva,
2015.
12. Mazeika, D., Morkevicius, A., Aleksandraviciene, A. “MBSE driven approach for defining problem
domain”, 11th Conference of System of Systems Engineering (SoSE), pp.1-6, Kongsberg, 2016.
13. Morkevicius, A., Aleksandraviciene, A., Armonas, A., Fanmuy, G. “Towards a Common Systems
Engineering Methodology to Cover a Complete System Development Process”, 30th Annual INCOSE
International Symposium, Cape Town, South Africa, July 2020.
14. Morkevicius, A., Aleksandraviciene, A., Mazeika, D., Bisikirskiene, L., Strolia, Z. “MBSE Grid: A
Simplified SysML‐Based Approach for Modeling Complex Systems”, 27th Annual INCOSE
International Symposium, pp. 136-150, Adelaide, 2017.
15. Morkevicius, A., Bisikirskiene, L., Jankevicius, N. “We Choose MBSE: What’s Next?”, Proceedings of the
Sixth International Conference on Complex Systems Design & Management, CSD&M, pp. 313, Paris,
2015.
16. Morkevicius, A., Jankevicius., N. “An approach: SysML-based automated requirements verification”,
IEEE International Symposium on Systems Engineering (ISSE), pp. 92-97, Rome, 2015.
17. Object Management Group OMG Systems Modeling Language (OMG SysML), v1.5, May 2017. Available
at https://www.omg.org/spec/SysML/1.5/.
18. Object Management Group OMG Unified Modeling Language (OMG UML), v2.5.1, December 2017.
Available at https://www.omg.org/spec/UML/2.5.1/.
19. Soegaard, S. E. “Adopting MBSE using SysML in System Development–Joint Strike Missile (JSM)”, No
Magic World Symposium, May 2016. Available at https://vimeopro.com/user28256466/no-magic-
world-symposium-2016-presentations/page/3.
20. Spangelo, S. C. et al. “Applying model based systems engineering (MBSE) to a standard CubeSat”,
Aerospace Conference IEEE, 2012.
21. Walden, David, ed. INCOSE Systems Engineering Handbook: A Guide for System Life Cycle Processes and
Activities, v.4. International Council on System Engineering (INCOSE), August 2015.
22. Weilkiens, T. Systems Engineering with SysML/UML: Modeling, Analysis, Design. MK/OMG Press, 2008.
23. Zachman, J. A. “A framework for information systems architecture”, IBM Syst. J., vol. 26, no. 3, pp.
276-292, 1987.

344 | MagicGrid® Book of Knowledge

ABOUT THE AUTHORS
Aistė Aleksandravičienė

Aistė is an OMG® Certified Systems Modeling Professional (OCSMP). Currently she holds the position
of industry business consultant for cyber systems engineering at Dassault Systèmes, CATIA brand.

Aistė takes responsibility for educating Dassault Systèmes clients on the three pillars of MBSE: language
(SysML), method (MagicGrid), and tool (CATIA Magic software and integrations). While working with
clients, such as Renault-Nissan, Leica Geosystems, Bombardier Transportation, and Kongsberg Defence &
Aerospace, she provides trainings, gives tool demonstrations, and participates in providing custom
solutions. Aistė also works on training material and writes papers (alone and in collaboration with her
colleagues) to spread the MBSE culture in the systems engineering community. She is a speaker at
multiple MBSE conferences and other public events.

Aistė holds a master’s degree in Information Systems Engineering from Kaunas University of Technology
(Lithuania).

Aurelijus Morkevičius

Aurelijus is an OMG® Certified UML, Systems Modeling, and BPM Professional. Currently he is a senior
manager of industry business consultants for cyber systems engineering at Dassault Systèmes, CATIA
brand.

He has expertise with enterprise, mission, model-based systems and software engineering (mostly based
on UAF, SysML, and UML accordingly) and defense architectures (DoDAF, MODAF, NAF). Aurelijus is a co-
chairman and the leading architect for the current OMG UAF (previously known as UPDM) standard
development group. He represents the CATIA brand at INCOSE and NATO.

Aurelijus holds a PhD in Informatics Engineering from Kaunas University and Technology since 2013. He
currently teaches an Enterprise Architecture course at the same university. Aurelijus is also an author of
multiple articles, and a speaker at multiple conferences.

MagicGrid® Book of Knowledge | 345

 Aistė Aleksandravičienė
Aurelijus Morkevičius, PH.D.

MagicGrid® BOOK OF KNOWLEDGE

A Practical Guide to Systems Modeling using
MagicGrid from Dassault Systèmes
2nd edition

Language editor Kim Papke

Cover pictures www.3DS.com
Designed by Arūnas Aleksandravičius

Published and prinded by

Vitae Litera, UAB
Savanoriu av. 137
LT-44146 Kaunas, Lithuania
www.tuka.lt | [email protected]
ISBN 978-609-454-554-2