Solidification Modeling of Iron Castings Using SOLIDCast

Piyapong Muenprasertdee

Thesis submitted to
The College of Engineering and Mineral Resources
at West Virginia University
in partial fulfillment of the requirements
for the degree of

Master of Science
in
Industrial and Management Systems Engineering

R. C. Creese, Ph.D., Co-Chair

B. Gopalakrishnan, Ph.D., Co-Chair
X. B. Liu, Ph.D.

Department of Industrial and Management Systems Engineering

Morgantown, West Virginia

2007

Keywords: Casting, Simulation, SOLIDCast, Finite Difference Method, Cast Iron

Copyright 2007 Piyapong Muenprasertdee

www.Iran-mavad.com

 ABSTRACT

Solidification Modeling of Iron Castings Using SOLIDCast

Piyapong Muenprasertdee

SOLIDCast is a casting simulation software program which can simulate thermal

changes and heat transfer in the solidification process of a casting. It assists the user to
visualize the solidification process of a particular casting. The program offers functions
to help guide a user in producing gating and riser designs and also have functions which
produce visual outputs showing possible problem areas and defects which may occur in a
casting. It can help shorten the lead time and reduce the loss in the trial casting stage.

This thesis consists of studies to determine the important capabilities and

limitations of the software, the usefulness of the riser and gating design wizard programs
offered in SOLIDCast to help a user design the riser and gating system, and the
simulation results and analysis of a casting design of a locomotive piston made from gray
iron.

www.Iran-mavad.com

 iii

ACKNOWLEDGEMENTS

I would like to sincerely thank Dr. R. C. Creese and Dr. B. Gopalakrishnan, both
of my co-chairs of for teaching, advising and taking care of me throughout my three years
at West Virginia University. Dr. Creese was my graduate advisor and teacher and Dr.
Gopalahrishnan was my research advisor and teacher as well. My life in West Virginia
University has become easier because of both of them.

I would like to thank Dr. X. B. Liu for being in the committee and offering his
help in my thesis.

I thank Mr. J. D. Schmidt, manager of the HK Engine Components Casting

Division, HK Engine Components, LLC. for letting me use one of their products as a case
study and also for helping me with what I needed for the thesis. I would also like to
thank Mr. Brandon Lamoncha, CNC programmer from Humtown Products for helping
me with the CAD designs of the casting design.

I thank Mr. David Schmidt of Finite Solutions Inc. for helping me and giving me
valuable advices about SOLIDCast.

I would like to thank Mr. Sawai and everyone from BIS Pipe Fitting Industry Co.,
Ltd., Thailand for letting me study the casting processes at your company along with all
the help for writing my thesis.

I would like to thank Mr. Navin Vongchankit and family, and everyone at Thai
Alloy and Associates Co., Ltd., Thailand, for helping me learn the casting process and the
use of casting simulations in your company.

I would like to specially thank Dr. Sombun Charoenvilaisiri from King

Mongkuts University of Technology Thonburi, Thailand, and Dr. Suwanchai
Pongsukijwat from Chulalongkorn University, Thailand for teaching me about casting
simulations and advising me about my thesis.

I would also like to thank Siam Castech Co. and its sister company for letting me
study the casting process at their facilities.

www.Iran-mavad.com

 iv

TABLE OF CONTENTS
ABSTRACT....................................................................................................................... ii
ACKNOWLEDGEMENTS ............................................................................................ iii
TABLE OF CONTENTS ................................................................................................ iv
LIST OF TABLES ........................................................................................................... ix
LIST OF FIGURES ......................................................................................................... xi

Chapter 1: Introduction ................................................................................................... 1

1.1 Casting ................................................................................................................ 1
1.2 Advantages and disadvantages of casting........................................................... 1
1.3 Types and Methods of Casting ........................................................................... 2
1.4 The Casting Process............................................................................................ 4
1.4.1 New Casting Development ................................................................................. 6
1.4.1.1 Product Design.................................................................................................... 6
1.4.1.2 Tooling Development ......................................................................................... 6
1.4.1.3 Foundry Trials..................................................................................................... 7
1.4.2 The Next Step in Casting Development.............................................................. 8
1.4.3 The Ten Rules of Casting ................................................................................... 9
1.5 Casting Simulation Software ............................................................................ 11
1.5.1 The Usefulness of Casting Simulation Software .............................................. 12
1.5.2 Casting Simulation Methods............................................................................. 13
1.5.3 Finite Difference Method (FDM) and Finite Element Method (FEM)............. 13
1.5.4 Limitations of Casting Simulation Software..................................................... 15
1.5.5 Depths in Casting Simulation Involvement ...................................................... 15
1.6 Rapid Prototyping ............................................................................................. 16
1.7 Cast Iron............................................................................................................ 17
1.7.1 Cast Irons .......................................................................................................... 17
1.7.2 Gray iron ........................................................................................................... 20
1.7.3 Carbon Equivalent ............................................................................................ 21
1.7.4 Other Types of Cast Irons ................................................................................. 23
1.7.4.1 Ductile (Nodular) Iron ...................................................................................... 23

www.Iran-mavad.com

 v

1.7.4.2 White Iron ......................................................................................................... 23

1.7.4.3 Malleable Iron................................................................................................... 24
1.7.4.4 Compacted Graphite Iron.................................................................................. 24
1.8 Cast Iron Alloys Appearing in SOLIDCast ...................................................... 27
1.9 Problems Occurred in Casting .......................................................................... 28
1.9.1 Solidification Shrinkage Defects ...................................................................... 29
1.9.2 Defects Occurred in Cast Iron........................................................................... 30
1.10 Feeders or Risers............................................................................................... 31
1.10.1 The Seven Feeding Rules.................................................................................. 31
1.10.2 Riser Location and Shape ................................................................................. 32
1.10.3 Riser and Neck Design...................................................................................... 34
1.11 Gating Design ................................................................................................... 37
1.12 Introduction to SOLIDCast............................................................................... 41
1.12.1 Output Criteria .................................................................................................. 41
1.12.2 Plotting the Results of Output Criteria.............................................................. 47
1.12.3 Analyzing the Results ....................................................................................... 48
1.12.4 Limitations in Predicting Defects in SOLIDCast ............................................. 48
1.13 The Basic Steps and Functions in SOLIDCast for Sand Casting Cast Irons .... 49
1.14 Extension Programs of SOLIDCast .................................................................. 50
1.14.1 FLOWCast ........................................................................................................ 50
1.14.2 OPTICast........................................................................................................... 51
1.15 Commercial Software for Computer Simulation in Metal Casting .................. 52
1.16 Problem Statement ............................................................................................ 54

Chapter 2: Literature Review........................................................................................ 56

2.1 The History of Metal Casting ........................................................................... 56
2.2 Studies of Casting Technology and Simulation Software for Casting ............. 57
2.2.1 Casting Techniques and Technologies.............................................................. 57
2.2.2 Studies Useful for Casting Simulation Software .............................................. 58
2.2.3 Studies About Casting Simulation Software..................................................... 60
2.2.4 Implemented Casting Simulation Software Case-studies ................................. 61

www.Iran-mavad.com

 vi

2.2.5 Studies of SOLIDCast....................................................................................... 62

Chapter 3: The Use of SOLIDCast................................................................................ 63

3.1 Introduction....................................................................................................... 63
3.2 Starting SOLIDCast .......................................................................................... 63
3.2.1 Alloy Curves Tab.............................................................................................. 63
3.2.2 Model Colors Tab ............................................................................................. 64
3.2.3 Model & Sim Tab ............................................................................................. 65
3.2.4 Directories Tab.................................................................................................. 66
3.3 Adding or Loading Shapes to the Model .......................................................... 66
3.4 Selecting Casting Material................................................................................ 67
3.4.1 Casting Tab ....................................................................................................... 67
3.4.2 Mold Tab........................................................................................................... 69
3.4.3 Curves Tab ........................................................................................................ 71
3.4.4 Heat Transfer Coefficients Tab......................................................................... 76
3.5 Testing The Use of SOLIDCast........................................................................ 76
3.5.1 Basic Functions for Simulating Cast Irons ....................................................... 77
3.5.1.1 The Casting Tab and the Curves Tab................................................................ 77
3.5.1.2 The Gray Iron and Ductile Iron Calculator Buttons ......................................... 77
3.5.1.3 The VDG Iron Properties Calculator ................................................................ 79
3.5.1.4 Conclusions of Tests ......................................................................................... 81
3.5.2 Modifying the Curves Tab ................................................................................ 81
3.5.3 Other Tests ........................................................................................................ 82
3.6 Conclusions....................................................................................................... 84

Chapter 4: Using SOLIDCast: Simple Model .............................................................. 85

4.1 Drawing the Design .......................................................................................... 85
4.2 Riser Design and Analysis ................................................................................ 86
4.2.1 Analysis of the Riser Design Wizard Input Values .......................................... 91
4.2.2 Analysis of the Riser Design Wizard Output Values........................................ 93
4.2.3 The Results of the Riser Design........................................................................ 98

www.Iran-mavad.com

 vii

4.3 Gating Design and Analysis............................................................................ 101

4.3.1 Analysis of the Gating Design Wizard ........................................................... 101
4.3.2 The Results of the Gating Design ................................................................... 111
4.4 The Resulting Dimensions for the Simple Casting Design ............................ 114
4.4.1 Dimensions from Riser Design Wizard .......................................................... 115
4.4.2 Dimensions from Gating Design Wizard........................................................ 116
4.5 Simulation Results and Analysis of Simple Casting Design .......................... 117
4.6 Conclusions..................................................................................................... 125

Chapter 5: Using SOLIDCast: Case Study ................................................................ 127

5.1 Company Profile ............................................................................................. 127
5.2 Casting Design ................................................................................................ 127
5.2.1 Pattern and Mold............................................................................................. 128
5.2.2 Core System .................................................................................................... 129
5.2.3 Assembled Mold ............................................................................................. 130
5.3 Dimensions of Mold Cavity............................................................................ 131
5.4 Discarding a Part............................................................................................. 133
5.5 Modeling the Casting Design in SOLIDCast.................................................. 134
5.5.1 Modeling of Shapes ........................................................................................ 134
5.5.2 Casting Material.............................................................................................. 136
5.5.3 Setting Up the Materials List Window ........................................................... 139
5.6 Verifying the Simulation Results with the Casting......................................... 142
5.6.1 Simulation Results of Base Case .................................................................... 143
5.6.2 Conclusions of Base Case............................................................................... 158
5.7 Pouring Temperature Comparison.................................................................. 158
5.8 Modifying the Casting Design ........................................................................ 160
5.8.1 Remove Top Riser .......................................................................................... 160
5.8.2 Reduce Size of Neck of Top Riser.................................................................. 162
5.8.3 Add a Circular Fin Around Outer Surface of Upper Ring Hot Spot .............. 162
5.8.4 Add Circular Fins Around Outer and Inner Surface of Upper Ring Hot Spot 164
5.8.5 Add Fin at Outer Surface of Upper Ring Hot Spot Across from Top Riser... 165

www.Iran-mavad.com

 viii

5.8.6 Comparing 3.975 and 3.70 Percent Carbon Equivalent to Base Case ............ 166
5.8.7 Add Hemispheric Bottoms to Side Risers ...................................................... 168
5.8.8 Move Side Risers Down ................................................................................. 169
5.8.9 Conclusions of Modifying the Casting Design............................................... 169
5.9 Using the Riser and Gating Design Wizard Programs.................................... 170
5.9.1 Riser Design Wizard ....................................................................................... 170
5.9.1.1 Riser Design Wizard: Simulation Tab ............................................................ 170
5.9.1.2 Riser Design Wizard: VDG Iron Properties Calculator.................................. 183
5.9.2 Gating Design Wizard..................................................................................... 190
5.10 Simulation Results of Re-designed Casting Design ....................................... 195
5.10.1 Simulation Results: Test 1 .............................................................................. 195
5.10.2 Simulation Results: Test 2 .............................................................................. 199
5.10.3 Simulation Results: Test 3 .............................................................................. 204
5.10.4 Conclusions of Simulation Results ................................................................. 206
5.11 Conclusions..................................................................................................... 208

Chapter 6: Summary and Conclusions ....................................................................... 209

6.1 Basic Capabilities and Limitations of the Simulation Function ..................... 209
6.2 Usefulness of the Riser Design Wizard and Gating Design Wizard Programs
213
6.3 Further Capabilities and Limitations Found from the Case Study ................. 217
6.4 Future Recommendations ............................................................................... 217
6.5 Conclusions..................................................................................................... 218

REFERENCES.............................................................................................................. 219

www.Iran-mavad.com

 ix

LIST OF TABLES
Chapter 1: Introduction
Table 1.1: The typical chemical composition of each type of cast iron (Prasertsakul). ... 25
Table 1.2: Selected sprue-runner-gate area ratios (Heine, 1995)...................................... 40
Table 1.3: Different types of materials and their critical Niyama value........................... 44

Chapter 3: The Use of SOLIDCast

Table 3.1: The weight percent carbon and silicon and mold wall movement of each cast
iron alloy found by trial and error............................................................................. 78
Table 3.2: The relationships of the input and output values in the VDG Iron Properties
calculator................................................................................................................... 80

Chapter 4: Using SOLIDCast: Simple Model

Table 4.1: Resulting values when type of sleeve and option of gating through the riser is
selected in the Sleeve Type box................................................................................ 92

Chapter 5: Using SOLIDCast: Case Study

Table 5.1: Default values of each attribute of each gray iron alloy in the Casting tab... 137
Table 5.2: The results from the VDG Iron Properties Calculator of each pouring
temperature. ............................................................................................................ 143
Table 5.3: Values retrieved from the Simulation Status windows.................................. 158
Table 5.4: Comparing the severity of the problems predicted by each criterion of the three
pouring temperatures. ............................................................................................. 159
Table 5.5: Comparing the results from the VDG Iron Properties Calculator of 3.70, 3.85
and 3.975 percent Carbon Equivalent..................................................................... 167
Table 5.6: Comparing the results from the Simulation Status window of 3.70, 3.85 and
3.975 percent Carbon Equivalent............................................................................ 167
Table 5.7: The inputs needed and outputs provided by each calculator button in the
calculator box.......................................................................................................... 176
Table 5.8: Riser efficiency factor values resulting from different casting modulus
values .................................................................................................................. 177

www.Iran-mavad.com

 x

Table 5.9: Comparing the actual riser volumes and the calculated riser volume required
for each feeding area............................................................................................... 178
Table 5.10: The actual and calculated surface areas of the necks connecting the three
risers to the casting.................................................................................................. 182
Table 5.11: Dimensions for designing Riser 3................................................................ 183
Table 5.12: The modulus and volumes of the three feeding areas of the casting. .......... 184
Table 5.13: The calculated diameters, heights and volumes of the risers and the neck
dimensions of Riser 1, 2 and 3................................................................................ 190
Table 5.14: Dimensions of the current gating system and the proposed gating system. 194
Table 5.15: The percent decrease in casting material weight compared to the base
case.................................................................................................................. 207
Table 5.16: The casting yield of each case including the base case. .............................. 207

www.Iran-mavad.com

 xi

LIST OF FIGURES
Chapter 1: Introduction
Figure 1.1: Main procedures of casting, modified from the casting process procedure of
the Thai Alloy and Associates Co., Ltd. and Creese (1999)....................................... 5
Figure 1.2: Cost and impact of product development phases (modified from Ravi, 2005).8
Figure 1.3: Iron-carbon phase diagram............................................................................. 18
Figure 1.4: Iron-iron carbide (dotted lines) and true equilibrium phase diagram (bold
lines).......................................................................................................................... 19
Figure 1.5: The relationships between the thickness or diameter of the casting,
mechanical properties and the liquidus temperature of the cast iron with the percent
Carbon Equivalent (Reproduced from Prasertsakul) ................................................ 22
Figure 1.6: Types of cast iron which result from different carbon composition ranges and
schematic microstructures which result from a variety of heat treatments............... 26
Figure reproduced from (Callister, 2006) ......................................................................... 26
Figure 1.7: Simple guidelines in designing a side riser and a top riser with a height per
diameter ratio of 1.0 (Modified from Heine (1995)). ............................................... 36

Chapter 3: The Use of SOLIDCast

Figure 3.1: The Alloy Curves tab in the Systems Parameters window. ........................... 64
Figure 3.2: The Model Colors in the System Parameters window. .................................. 65
Figure 3.3: The Model & Sim tab in the System Parameters window. ............................ 65
Figure 3.4: The Directories tab in the System Parameters window.................................. 66
Figure 3.5: The Casting tab in the Materials List window. .............................................. 69
Figure 3.6: The Mold tab in the Materials List window. .................................................. 70
Figure 3.7: The Curves tab in the Materials List window. ............................................... 72
Figure 3.8: The iron properties calculator from pressing the Gray Iron or Ductile Iron
button. ....................................................................................................................... 73
Figure 3.9: The VDG Iron Properties Calculator located in the Tools menu. .................. 75
Figure 3.10: The corresponding shrinkage curve (black line) using the results from Figure
3.9.......................................................................................................................... 75
Figure 3.11: The Materials List window showing the HT Coefficients tab. .................... 76

www.Iran-mavad.com

 xii

Figure 3.12: Temperature curve developed by the Ductile Iron calculator button........... 79
Figure 3.13: Temperature curve developed by the Gray Iron calculator button............... 79
Figure 3.14: An example of the VDG nomograms from the German Iron Society using
metric units................................................................................................................ 81

Chapter 4: Using SOLIDCast: Simple Model

Figure 4.1: Figure of design drawn in SOLIDCast........................................................... 85
Figure 4.2: Casting tab with properties of material used in test. ...................................... 86
Figure 4.3: Setting up the meshing parameters................................................................. 87
Figure 4.4: The weights of the model. .............................................................................. 87
Figure 4.5: Finding the casting modulus from the Riser Design Wizard. ........................ 88
Figure 4.6: Modulus plot showing one feeding area plotted with the suggested value,
0.543.......................................................................................................................... 89
Figure 4.7: The VDG Iron Properties Calculator.............................................................. 90
Figure 4.8: The Riser Design Wizard window accessed from the VDG Iron Properties
Calculator.................................................................................................................. 90
Figure 4.10: The final calculation of the Riser Design Wizard ........................................ 98
Figure 4.11: Calculated dimensions of the riser from the Riser Design Wizard. ............. 99
Figure 4.12: Modulus plot showing one feeding area plotted with the neck modulus,
0.5285503 inch........................................................................................................ 100
Figure 4.13: Calculating the optimal fill time for the casting......................................... 102
Figure 4.14: Gate Position window for top gating system.............................................. 103
Figure 4.15: Gate Position window for bottom gating system. ...................................... 104
Figure 4.16: Gate Position window for parting line gating system. ............................... 104
Figure 4.17: Calculated sprue data window, continued from Figure 4.15...................... 106
Figure 4.18: The step down calculation of the runner................................................. 109
Figure 4.19: The runner and gate design window........................................................... 110
Figure 4.20: Calculating the fill time.............................................................................. 112
Figure 4.21: The resulting Gate Position window. ......................................................... 113
Figure 4.22: The resulting sprue data window................................................................ 113
Figure 4.23: The resulting runner and gate design window............................................ 114

www.Iran-mavad.com

 xiii

Figure 4.24: The sequence of how liquid metal poured from the ladle enters the mold. 115
Figure 4.25: Casting model in SOLIDCast using all calculated and user-defined
dimensions. ............................................................................................................. 118
Figure 4.26: The Casting tab set with values representing the casting model................ 119
Figure 4.27: The Curves tab set with values representing the casting model................. 119
Figure 4.28: The mesh creating window......................................................................... 120
Figure 4.29: The calculated weights of the materials in the model after meshing. ........ 121
Figure 4.30: Critical Fraction Solid Time plots plotted at 10, 10.5, 11, 11.5, 12 and 12.5
minutes after pouring. ............................................................................................. 122
Figure 4.31: Cut-plane plots of the Critical Fraction Solid Time criterion. ................... 123
Figure 4.32: Material Density Function plots plotted at 0.995. ...................................... 123
Figure 4.33: FCC Criterion plots. ................................................................................... 124
Figure 4.34: Hot Spot Criterion calculated from Critical Fraction Solid Time. ............. 124
Figure 4.35: Hot Spot Criterion calculated from Solidification Time. ........................... 125

Chapter 5: Using SOLIDCast: Case Study

Figure 5.1: Locomotive piston produced by HKEC. ...................................................... 128
Figure 5.2: The cope pattern (left) and drag pattern (right)............................................ 128
Figure 5.3: The cope mold (left) and drag mold (right).................................................. 129
Figure 5.4: Core system. ................................................................................................. 130
Figure 5.5: Assembled mold with liquid metal and magnesium based insulators covering
top risers.................................................................................................................. 130
Figure 5.6: Dimensions of top and side risers of casting................................................ 131
Figure 5.7: Dimensions of the mold cavity and gating system....................................... 132
Figure 5.8: Half of the model created in SOLIDCast with Plane of Symmetry
function ........................................................................................................... 135
Figure 5.9: The Casting tab............................................................................................. 140
Figure 5.10: The Mold tab showing the attributes of the magnesium based exothermic
material. .................................................................................................................. 140
Figure 5.11: The VDG Iron Properties Calculator.......................................................... 141
Figure 5.12: The modified Curves tab. ........................................................................... 142

www.Iran-mavad.com

 xiv

Figure 5.13: Left: Progress of solidification before model re-alignment. Right: Progress
of solidification after model re-alignment. ............................................................. 145
Figure 5.14: A shaken out casting placed upside-down. ................................................ 146
Figure 5.15: Critical Fraction Solid Time criterion plotted at 0, 1.0, 1.5 and 2 minutes
after mold is filled................................................................................................... 146
Figure 5.16: Critical Fraction Solid Time criterion plotted at 0, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5,
4.0, 4.5, 5.0, 5.5 and 6.0 minutes after mold is filled. ............................................ 147
Figure 5.17: Critical Fraction Solid Time criterion plotted at 6.5, 7.0, 7.5, 8.0, 8.5, 9.0,
9.5 and 10.0 minutes after mold is filled. ............................................................... 148
Figure 5.18: Material Density Function plots plotted at 1.0, 0.995 and 0.99 respectively
from top to bottom. ................................................................................................. 150
Figure 5.19: Front view of casting before fettling or knock-off. .................................... 151
Figure 5.20: Top view of casting before fettling or knock-off. ...................................... 151
Figure 5.21: Left; blind single riser. Right; blind shared riser........................................ 152
Figure 5.22: Material Density Function cut-plane plot plotting areas with density values
from 0 to 1............................................................................................................... 152
Figure 5.23: FCC Criterion plots. ................................................................................... 153
Figure 5.24: Hot Spot Criterion (Solidification Time) plotted at 1.1. ............................ 154
Figure 5.25: Hot Spot Criterion (Critical Fraction Solid Time) plotted at 1.1. .............. 155
Figure 5.26: Hot Spot (Solidification Time) cut-planes plot plotted between 1.0 and
1.1........................................................................................................................ 155
Figure 5.27: Hot Spot (Critical Fraction Solid Time) cut-planes plot plotted between 1.0
and 1.1..................................................................................................................... 156
Figure 5.28: Hot Spot (Solidification Time) plotted with 1.0......................................... 156
Figure 5.29: Hot Spot (Solidification Time) plotted with 1.01....................................... 157
Figure 5.30: Hot Spot (Critical Fraction Solid Time) plotted with 1.0........................... 157
Figure 5.31: Hot Spot (Critical Fraction Solid Time) plotted with 1.01......................... 157
Figure 5.32: Iso-surface plots of the Material Density Function plotted at 0.995. ......... 161
Figure 5.33: Cut-plane plots of the Material Density Function plotted from 0 to 1. ...... 161
Figure 5.34: Solidification Time plot plotted at 0 minute. ............................................. 163
Figure 5.35: Cut-plane plots of Hot Spot (Solidification Time) Criterion. .................... 164

www.Iran-mavad.com

 xv

Figure 5.36: Cut-plane plots of Hot Spot (Solidification Time) Criterion. .................... 165
Figure 5.37: Material Density Function plots plotted at 0.995. ...................................... 166
Figure 5.38: Options of Calculate and Display Casting Modulus and Design
Risers................................................................................................................. 171
Figure 5.39: Modulus plot set up window. ..................................................................... 171
Figure 5.40: The resulting plots from the modulus plot set up window in Figure 5.39. 172
Figure 5.41: Calculated feeding areas............................................................................. 173
Figure 5.42: Plotted feeding areas; from left to right, Feed Area 1, Feed Area 2 and Feed
Area 3...................................................................................................................... 173
Figure 5.43: Plots plotted from feeding area identification modulus value of 0.39 showing
3 feeding areas. ....................................................................................................... 174
Figure 5.44: Riser calculator for Feed Area 1................................................................. 175
Figure 5.45: Riser calculator for Feed Area 2................................................................. 175
Figure 5.46: Riser calculator for Feed Area 3................................................................. 176
Figure 5.47: Calculated dimensions for Riser 3; the top riser. ....................................... 179
Figure 5.48: Resulting calculations for Riser 1 in Riser Design Wizard........................ 185
Figure 5.49: Resulting calculations for Riser 2 in Riser Design Wizard........................ 186
Figure 5.50: Resulting calculations for Riser 2 feeding two feeding areas in Riser Design
Wizard..................................................................................................................... 187
Figure 5.51: Resulting calculations for Riser 3 in Riser Design Wizard with normal
neck. .. ..................................................................................................................... 189
Figure 5.52: Resulting calculations for Riser 3 in Riser Design Wizard with short
neck ................................................................................................................. 189
Figure 5.53: Mesh weights of base case. ........................................................................ 191
Figure 5.54: Entering the pour weight and fill time........................................................ 191
Figure 5.55: The Gate Position window. ........................................................................ 192
Figure 5.56: Calculated sprue dimensions. ..................................................................... 193
Figure 5.57: Calculated gate and runner areas and dimensions...................................... 193
Figure 5.58: Critical Fraction Solid Time plots plotted at 4.5 minutes show possible
isolated areas........................................................................................................... 197
Figure 5.59: Material Density Function plots plotted at 0.995. ...................................... 197

www.Iran-mavad.com

 xvi

Figure 5.60: FCC Criterion plots showing no highlighted areas. ................................... 197
Figure 5.61: The weights of the materials in Test 1. ...................................................... 198
Figure 5.62: Casting model built from calculated dimensions for Test 2....................... 200
Figure 5.63: The weights of the materials in Test 2. ...................................................... 201
Figure 5.64: Critical Fraction Solid Time plot plotted at 7.5 minutes............................ 201
Figure 5.65: Material Density Function plots plotted at 0.995. ...................................... 202
Figure 5.66: FCC Criterion plots. ................................................................................... 202
Figure 5.67: Hot Spot (Solidification Time) plotted at 1.1............................................. 203
Figure 5.68: Hot Spot (Critical Fraction Solid Time) plotted at 1.1............................... 203
Figure 5.69: Casting model built from calculated dimensions for Test 3....................... 204
Figure 5.70: The weights of the materials in Test 3. ...................................................... 204
Figure 5.71: Material Density Function plots plotted at 1.0. .......................................... 205
Figure 5.72: Material Density Function plots plotted at 0.995. ...................................... 205
Figure 5.73: Hot Spot (Solidification Time) plotted at 1.1............................................. 206
Figure 5.74: Hot Spot (Critical Fraction Solid Time) plotted at 1.1............................... 206

www.Iran-mavad.com

 Chapter 1
Introduction
1.1 Casting
Casting is the process of producing metal into a determined shape by melting
solid metal into liquid form, pouring it into a mold and letting it solidify into the desired
shape. The mold is a negative copy of the shape of the casting. There are many other
methods of shaping metals, such as machining, forging, welding, stamping and hot
working. Casting has many advantages over the other methods of metal shaping for
producing some particular shapes of metal and types of metal.

1.2 Advantages and disadvantages of casting

Casting can simplify production by casting a single complex shape piece instead
of manufacturing a product that requires assembling several pieces together. Mass
production of products can be done by casting; a large number of products in a single
mold or in other cases by reusing the mold. Very large heavy objects can be cast which
would be extremely difficult or economically impossible to produce by other methods.
Some engineering properties in casting are better than objects produced by other methods.
For example, uniform properties throughout the casting can be achieved if properly cast.
Casting can give an economic advantage resulting from one or more of the advantages
shown which may help in the competition against other types of manufacturing. (Heine et.
al., 1995)
There are situations where other shaping methods are more suitable. For example,
machining can achieve better surface finish and dimensional accuracy not achievable by
casting; welding can join metal objects which may be produced by wrought or casting
into more complex structures, stamping produces light weight sheet metal parts; and
forging helps improve the strength and toughness of steel, etc. An engineer with
knowledge of the possibilities of each shaping method may select a method or a
combination of methods which best suits his or her work achieving high quality, low
priced products (Heine et. al., 1995).

1
www.Iran-mavad.com

1.3 Types and Methods of Casting
There are many processes for producing a casting depending on size, type of
metal, complexity, dimensional allowance, quality and whatsoever. One of the oldest
methods known is sand casting. Sand is mixed with binders and water so the sand grains
hold shape, compacted in a flask which can be separated into two or more pieces with a
pattern in the middle. This pattern will have approximately the same shape and size (or
may be slightly different depending on the dimensional tolerance, shrinkage/expansion of
the metal or machining allowances) of the desired casting. The mold is then parted, the
pattern is removed and the compacted sand will have a negative shape of the pattern. The
mold may be assembled with cores to give the casting hollow shapes, and the gating
system, including runners, ingates, risers, sprues, etc., will be made in the mold. The
liquid metal would then be poured into the mold. After the casting has cooled down, the
mold is broken to remove the casting. The castings gating system and risers would be
broken off, the casting cleaned and then machined into the desired product.
Sand casting may be separated into green-sand mold casting, no-bake sand mold
casting and shell-mold casting. Green-sand mold casting is silica sand and clay mixed
with water, no-bake sand mold is silica sand mixed with resin that hardens within
minutes after shaping and shell-mold casting is silica sand mixed with resin, shaped and
baked to form the wanted mold shape (Society of Manufacturing Engineers).
Sand casting may be classified as Expendable Mold/Reusable Pattern process.
Other Expendable Mold/Reusable Pattern process casting methods are the plaster-mold
casting process, and ceramic-mold casting process (Society of Manufacturing Engineers).
The lost wax casting method or investment casting process was used by Asian
Indians to make sculptures of gods and goddesses for hundreds of years. The wax
models were carved and carefully pasted over and covered with natural clay obtained
from river banks, after wetting with water. After being sun dried for 3 to 4 weeks, the
wax was melted with heat, forming a hollow mold. Molten metal was poured into the
mold to make the casting (Ravi, 2005). The modern lost wax casting method is called
investment casting. Wax is injected into a metal mold to make patterns, which are
connected to a common sprue to form a tree. The tree is repeatedly dipped in ceramic
slurry and dried. The wax is removed by heat leaving a mold. The ceramic shell is

2
www.Iran-mavad.com

preheated and filled with molten metal, then broken after the casting has cooled down.
This method is suitable for casting any metal with small to medium intricate shapes and
thin walls.
Full mold casting or evaporative-foam casting is done by packing loose silica
sand around an expanded polystyrene (EPS) pattern. Molten metal is then poured into
the mold through the gating system, burning out the foam pattern as it fills the mold. It is
called full mold casting because the pattern is not removed, hence the name full mold.
This method is suitable for even intricate castings and requires less labor and skill
compared to sand casting. A mold must be used to make the EPS patterns and the mold
should properly vent out the gases generated from burning the foam. The replicast
process is a method similar to wax investment casting except EPS is used instead of wax.
Another variation is vacuum casting, which also uses loose sand but held between two
thin plastic sheets by vacuum applied to a pattern with a number of vent holes (Ravi,
2005).
Investment casting and evaporative-foam casting can be classified as Expendable
Mold/Expendable Patterns Process. Another classification of casting process is
Permanent Mold/No Patterns Process. The casting processes in this classification are
permanent mold casting, die casting and centrifugal casting (Society of Manufacturing
Engineers).
Gravity die casting (also called the permanent mold process) is a method which
molten metal is poured into a cast iron mold coated with a ceramic mold wash. Cores can
be made from sand or metal. After the casting has cooled down, the mold is parted and
the casting is removed. This method is suitable for non-ferrous metals with medium sizes
and moderate complexity and thickness (Ravi, 2005).
Pressure die casting is a process which molten metal is injected into a hardened
steel die under pressure. Usually this type of die is water-cooled and metal cores must be
used instead of sand cores. The casting is removed by parting one half of the die and the
casting is removed by ejector pins. This process is suitable for non-ferrous castings with
small to medium size, varying complexity and thin walls (Ravi, 2005).
Centrifugal casting is a process which molten metal is poured into a horizontal
rotating mold where the centrifugal force would push the molten metal to the mold wall.

3
www.Iran-mavad.com

This method can produce pipes of tubes without using cores (Ravi, 2005) and the
thickness of the casting wall depends on the amount of molten metal poured into the
mold (Society of Manufacturing Engineers). An alternative to this process is the semi-
centrifugal casting process which axis-symmetrical castings, like pulleys, gears and rotors
can be produced while rotating about a vertical axis rotating mold. Another variation
called the centrifuge casting is a process which mold cavities are arranged around a
central axis. This method uses the rotation of the mold to get better filling characteristics
(Ravi, 2005).
Squeeze casting or semi-solid casting is a process which semi-solid metal is
forced under pressure into a metal mold. This method would give a casting fine
microstructure free from dendrites. The mechanical properties of these castings are close
to those of forgings. This method is useful for non-ferrous metals and composites and is
also applied for aerospace and automotive parts (Ravi, 2005).

1.4 The Casting Process

The casting process starts from receiving an order from a customer which may
include the design, dimension, physical properties, etc., then the foundry must plan how
to make the castings, what methods must be used, then produce a prototype of the casting,
modify the casting methods to get rid of the defects, produce the product, and last of all,
the send the final product to the customer. Figure 1.1 shows the main procedures of a
casting process, but the procedures in each casting facility may differ in detail.

4
www.Iran-mavad.com

 Order

Core Design Pattern Design Production Plan

Core Base Design Pattern Manufacture

Customer
Casting Method and
Material
Core Base Manufacturing Pattern Inspection

Inspection of
Core Inspection Casting Trial Trial Casting Casting Design

Sand Metallic Raw

Raw Materials Materials

Sand Test Mold Assembly Furnace Charge

Preparation

Pouring
Sand Reclamation

Pattern Core Pattern Melting

Making Making
Shake Out

Molding Core Making

Ingredient
Analysis
Shot Blasting
Charge
Mold Test Core Test Material
Cutting Risers

Engraving/Gouging Gating Returns

Heat treatment Scrap

Microscopic
Structure Inspection

Finishing

Initial Inspection

Painting

Final Inspection/
Repair Reject
Non-destructive Test

Packaging

Shipment

Figure 1.1: Main procedures of casting, modified from the casting process procedure of
the Thai Alloy and Associates Co., Ltd. and Creese (1999).

5
www.Iran-mavad.com

1.4.1 New Casting Development
From Figure 1.1, the area in the shaded box could be called the development of a
new casting, which in detail, might be separated into three stages, product design, tooling
development and foundry trials (Ravi, 2005).

1.4.1.1 Product Design

This first stage is the most essential because it virtually influences all other
decisions and activities in the product life cycle, and eventually the technical and
economical value of the product.
This stage consists of three types of requirements. First is the functional
requirements, driven by geometry, dimensions, relative location and orientation. The
next type of requirements is property requirements, including thermal-physical properties,
mechanical, and chemical properties of the product. These properties are mainly driven
by the material composition and structure. The third type is production and quality
requirements, which include surface finish, tolerance, internal soundness, order quantity
and lead time, etc.
The above requirements are developed through three steps of the product design,
which are conceptual design which focuses on the geometries of the product to achieve
the required functions, detailed design involves selecting the material(s), defining the
geometry and its tolerances and prototyping which basically is producing a prototype to
test the form, fit and function of the product. Iterations may be done to these production
design steps to achieve optimality of functional requirements, quality and cost (Ravi,
2005).

1.4.1.2 Tooling Development

Tooling consists of patterns and core boxes for sand casting or dies for die casting
or investment casting. It can also be further classified as design of cavity or the pattern
for producing the cavity or cavities and its accessories.
This stage involves setting the best orientation of the casting and the
determination of the parting line or parting lines if there has to be more than two
segments of molds to produce the casting. Also, some castings might have multiple

6
www.Iran-mavad.com

cavities instead of just one. It also involves the internal cavities such as holes and
undercuts produced by cores. This require identification of core features, design of cores,
which also includes their supports, called prints in sand casting, and core boxes for
producing the cores.
Proper allowances must be incorporated in the mold cavity and cores considering
part shrinkage, distortion and machining which may be done later in the process. The
cores or dies must be easy to remove from the part. Other cavities include feeders or
risers (number, location, shape, and dimensions), sprues, runners, the gating system
which leads the molten metal into the mold. Other accessories include cooling, guiding
and ejection systems (for die casting). The method for manufacturing the tooling
depends on its material, complexity, quality and time/cost considerations (Ravi, 2005).

1.4.1.3 Foundry Trials

Trial castings are made to observe the flaws and defects that might happen in a
casting which may occur from the two previous stages. Castings are inspected using
destructive and non-destructive methods for finding external and internal defects. The
most common destructive method is to section the casting in the places of interest and
find if defects are present. Macroporosities and shrinkages may be seen by the naked eye
while microporosities and microstructural defects would require seeing through a
microscope. Non-destructive methods include radiography, ultrasound, magnetic partical,
dye penetrant, and eddy current testing. Based on the results, the tooling, usually the
gating and riser system may be modified, and the process parameters, usually the pouring
temperature, pouring time (and pressure variation in the case of die casting), etc., may
also be modified to improve the casting to a desirable level of quality. If the defects
cannot be eliminated by modifying the process parameters or tooling design, it is
necessary that the product design be modified, which is very expensive and time
consuming at this late stage. Figure 1.2 shows the relationships of management time
spent, the ability to influence the production process, the cost of rectifying mistakes in the
process and the accumulated costs to each product development phase (modified from
Ravi, 2005).

7
www.Iran-mavad.com

 Ability to influence
(impact of decisions) Accumulated costs

Management time spent

Cost of (rectifying) mistakes

Concept Design Tooling Production Delivery Service

Product Development Phase

Figure 1.2: Cost and impact of product development phases (modified from Ravi, 2005).

1.4.2 The Next Step in Casting Development

Over the last decade, many casting buyers and assemblers adopted the Just-In-
Time philosophy to reduce their inventory costs, which requires foundries to deliver
smaller lots and more frequently, while maintaining a strict schedule. There has been a
steady increase in the requirements in terms of quality assurance, shorter lead-time,
smaller lot sizes and competitive pricing. Assemblers also are eliminating inspection of
incoming goods and expect the suppliers to be responsible for casting quality and want
the foundries to continuously reduce their costs by adopting better technologies and
methodologies. There is also pressure from energy conservation, environment and
operational safety regulations.
Casting buyers, especially in the automotive industry, are moving towards long-
term partnerships with a few foundries instead of short-term cost-based purchasing
agreements with a number of foundries as in the past. This is also one of the methods in
the Just-In-Time methodology. For a foundry to survive and compete in the market, it
must produce dimensionally stable and sound castings (Quality), have reliable on-time
small-lot delivery (Delivery) and continuously lower their prices (Cost) (Ravi, 2005 and
Pongsukijwat, 2006).

8
www.Iran-mavad.com

 Usually, the average lead time for developing a new casting might take several
weeks and tooling development and foundry trials account for nearly 70 percent of that
time. The lead-times can be reduced more than half, especially for intricate castings
when using computer-aided systems for product design, tooling development and process
optimization. Casting simulation software can help predict internal defects and the
process parameters, tooling design and part model can be modified and verified to
achieve the desired quality before pouring a trial casting for physical verification. Rapid
prototyping can greatly reduce the lead time by eliminating the tooling design and
requirements.

1.4.3 The Ten Rules of Casting

A good casting is a casting that meets or exceeds a customers specification.
Campbell (2004) has laid out ten rules, which if followed, will produce good and near
defect-free castings.

1) Rule 1
A good quality melt is very significant to the quality of the casting. The liquid
metal should be prepared, checked and treated to achieve an acceptable minimum
standard for casting a certain product.

2) Rule 2
Avoid turbulent entrainment of the surface file on the liquid. The liquid metal
should not go too fast or fall more than the critical height.

3) Rule 3
Avoid laminar entrainment of the surface film on the liquid. No part of the liquid
metal front should come to a stop before the mold cavity is completely filled. The filling
should progress only uphill in a continuous uninterrupted upward advance.

9
www.Iran-mavad.com

4) Rule 4
No bubbles of air entrainment should be allowed to pass through the liquid metal
into the mold cavity. Using a properly designed sprue, avoiding the use of sprue wells or
other volume-increasing systems, using small volume runners, using ceramic filters close
to the sprue or runner junctions and bubble traps, etc., are some ways of preventing
bubbles entering the casting. Also, the pouring should not be interrupted.

5) Rule 5
Core blows or bubbles from the out-gassing of cores must be avoided. Cores
must have low gas content or must be vented to prevent the bubbles from the cores
entering the casting. Clay-based cores or mold repair pastes should be fully dried before
use.

6) Rule 6
Avoid shrinkages by not feeding uphill in larger section thickness castings
because of unreliable pressure gradients and complications which are occurs from
convection. Feeders or risers should be used for counteracting the shrinkages. A riser
with appropriate dimensions should be added to the casting at the appropriate position.
The rules for designing a good riser will be explained later. Also, the level of flash at the
mold and core joints, the mold coat thickness, if any, and the temperatures of the metal
and mold should be controlled.

7) Rule 7
Avoid convection damage. Convection is a flow phenomenon resulting from
density differences in a fluid. The density differences may result from temperature
differences. The time needed for a convection to start may be between one and two
minutes. By three or more minutes, convection can become well established, causing
extensive re-melting and a major redistribution of heat in a casting. Thin castings which
take less than one minute to solidify or thick castings which may take more than perhaps
ten minutes to solidify would not have convection problems because in thin castings, the
viscous restraint of the walls make convection difficult and the rapid cooling allows less

10
www.Iran-mavad.com

time for convection to develop. For thick castings, the liquid metal would have plenty of
time to convect and re-organized itself. Intermediate size castings are most prone to
convection damage. The problem may be taken care by redesigning the casting and
rigging system, avoid feeding uphill or by using the tilt casting process.

8) Rule 8
At regions of the casting which local cooling rates change, either from section
change or at a chill or a riser, the change in composition of the casting will occur. This
problem is called segregation. Sometimes they are unavoidable, so, there should be a
specification limit agreement with the customer.

9) Rule 9
Reduce residual stress by not quenching into water.

10) Rule 10
All castings should be provided with agreed location points so the dimensions
may be measured accurately and the work done by the toolmaker to the foundryman
down to the machining operator could be smooth.

1.5 Casting Simulation Software

In this day and age, customers, especially in the automotive industry, would be
more likely to request castings with high quality (Q), quick delivery (D) and at a low cost
(C). A tool that foundries may use to achieve the three goals previously mentioned is to
apply Computer Aided Engineering into their process, in this case is by using computer
simulation software for casting.
The procedure of using casting simulation software may be explained as follows
(Pongsukijwat, 2006):
1) Build a model of the casting design including the gating system and all other
material used with the casting, such as chills, cores, sleeves, etc. This step may be
done by using a CAD (Computer Aided Design) system.

11
www.Iran-mavad.com

 2) Input required data needed for computation, such as the physical, mechanical and
heat properties of the metal, properties of the mold or die, pouring temperature,
pouring time, pressure, etc.
3) Computation of the simulation, which different casting simulation programs may
have different approaches in simulating the results. Some well known approaches,
for example are, the numerical simulation approaches (Finite Element and Finite
Difference Methods), the geometrical approach (Sarfaraz, 1988), the meshless
method (Lewis et. al., 2004), etc.
4) Simulated results and interpretation of results. The results from the simulation
program may be shown in the form of graphs or colored figures with numerical
results depending on what criterion is used, such as the temperatures in each
section of the casting at a given time, solidification times, hot spots, material
density, etc. These results must be translated into useful information to evaluate
if a casting is sound or not, or what must be done to improve the casting design
and start from step 1 once again.

1.5.1 The Usefulness of Casting Simulation Software

In the past the foundryman have strived for ways to improve the casting process
and eliminate the defects that occurred in the castings by trial and error and past
experiences. The time needed to produce a particular product is a time-consuming
process. Problems occurred in the casting may only be solved through trial and error.
Scientists throughout the years have studied the science of casting and metallurgy
and developed theories and mathematical models to explain the properties of metals
while going through the solidification process. Simulation programs were developed
from these methods which are useful in predicting how the casting will come out.
Defects and problems can be discovered before the actual casting is cast avoiding costly
tests to prevent the problems.
The process of manufacturing a new casting design in a foundry starts from
receiving a design from a customer, which would include all dimensions and tolerances,
what kind of material and additives, and may also include the strength, hardness or
surface finish, etc. Then the foundryman or the foundry engineer would design the gating

12
www.Iran-mavad.com

and risering system for the casting. The time used in designing and re-designing the
gating and risering system might take a few days or up to several weeks before good
castings can be made (as mentioned earlier in section 1.4.1 and 1.4.2) depending on the
casting complexity and the skill of the foundryman or the engineer. Casting simulation
software can predict where and what defects might occur in a casting and the time and
material used in the trial stage may be reduced significantly.

1.5.2 Casting Simulation Methods

The casting simulation programs have different approaches in calculating and
predicting the outcome of a casting. Each method hold advantages and disadvantages
compared to another. Some casting simulation methods may be shown below:
1) Numerical Approach
- Finite Element Method (FEM)
- Finite Difference Method (FDM)
2) Geometrical Approach
- K-Contour Method (Safaraz, 1988)
3) Computer Wave Front Analysis
- Pour-out Method
- Cubic Spine Functions
4) Meshless Method (Lewis, 2004)
5) Grid-based simulation system (Pan et. al., 2004)
This study used one program, SOLIDCast, which uses the Finite Difference
Method.

1.5.3 Finite Difference Method (FDM) and Finite Element Method (FEM)
The Finite Difference Method and Finite Element Method are the two most
common methods used by casting simulation software developers. The differences
between these two methods from a users point of view are presented.
If a three dimensional picture of a casting model mesh in a casting simulation
program is composed of several small cubic boxes or blocks or cells, the simulation
program used uses the Finite Difference Method. This method would take a three

13
www.Iran-mavad.com

dimensional casting design and break it into small elements, small square boxes along the
X, Y and Z axes. The calculation method has each box be uniform throughout itself and
have interaction with the adjacent boxes. The calculations are made to the closest dx, dy,
dz values and calculated at nodes which are located at the centers of the elements. The
smaller the boxes, the smoother the picture of the mesh becomes and the more accurate
the calculations are. But the smaller the boxes, the more boxes are needed and more
memory space on the computer is needed.
The Finite Element Method mesh would look like small triangles joined together.
Actually they are nodes connected with lines. The skin of the three dimensional picture
would look smoother than the mesh from Finite Difference Method.
Comparing these two methods, the meshing of Finite Difference Method is easier,
because when a user orders the program to mesh the design with Finite Element Method,
some nodes may not be properly connected. The user must check thoroughly throughout
the design to see that all nodes are connected, which is a very stressful job. If there is a
gap, the simulation may not be done and errors would occur which the user must find the
gap and re-mesh the design again with a different node size. The temperatures are
calculated at nodes which are located at the corners of the elements. The Finite Element
Method uses more computer memory and takes more time in calculating results, but the
primary advantage of Finite Element Method over Finite Difference Method is that it
gives more accurate results.
If there are thin sections in the casting, the Finite Element Method user can assign
smaller meshes to the thin sections and larger meshes in thicker sections to save memory
usage for calculation. For Finite Difference Method, there can only be one size of cells
throughout the casting. The size of the cells must be appropriate compared to the
smallest section of the casting or the calculations will become inaccurate. Less intricate
castings can be calculated quite accurately with larger cells resulting in less computer
memory usage in calculating for results.
The biggest difference in the two methods regards the axis alignment of the model.
Finite Difference Method breaks down the model into small boxes aligned to the X, Y
and Z axes and the simulation results would differ if aligned differently, but the results
would be the same in any direction using the Finite Element Method.

14
www.Iran-mavad.com

1.5.4 Limitations of Casting Simulation Software
Even though casting simulation software may sound like a savior from casting
heaven, it may not solve every problem. The limitations of each casting simulation
software program are different.
The most accurate simulation results require that all input parameters be set
carefully to match the real casting, including material, chemical composition, mold
material, pouring temperature, pouring time, heat transfer coefficient values, cooling
curves, expansion and shrinkage rates, etc. Casting simulation software cannot predict all
types of defects that may occur in a casting, such as processing defects, human error,
additional chemical elements added to the molten metal, etc. A foundry engineer must be
able to notice what kinds of defects occur and what causes them to occur so the problem
can be treated as casting simulation software can only simulate the solidification process.

1.5.5 Depths in Casting Simulation Involvement

After visiting foundries to interview the foundry managers and engineers and
from interviewing a few professors in the field of metal casting in Thailand, it was found
that the involvements in casting simulation software, if any, are different in each foundry.
The depths in the simulation of castings may be separated into four levels, as
follows:
1) The foundry engineer designs all gating system, runners, risers, sprues, tooling,
process, etc., and starts the trial casting stage without using casting simulation
software. This is what most foundries still practice. It really depends on the
experience of the engineer in getting a good casting in a short period of time.
2) The foundry engineer designs and verifies the casting design with a casting
simulation software program himself/herself.
3) The foundry engineer would send the design (along with the tooling design if needed)
and information about the casting to a casting simulation specialist and the specialist
would design all gating system, runners, risers, sprues, etc., with the simulation
program. The specialist would send back the results along with suggestions about the
casting design. The engineer may start the trial castings or make some modifications
if required before starting the trial castings.

15
www.Iran-mavad.com

4) The foundry engineer designs all of the gating system, including runners, risers,
sprues and tooling and writes the model into a 3D CAD program. The engineer then
sends the design along with all needed information, such as material name, material
grade, additives, chemical composition, process parameters, etc., to a simulation
specialist to verify the design. The specialist may suggest how the design should be
changed and then sends back the results of the simulation. If no problems occur in
the simulation, the engineer may start the trial castings, but if there are any, the
engineer would re-design the casting before starting the trial castings. The engineer
may re-design the casting on the 3D CAD program and verify the design once more
with the specialist. This is starting to be more common in the Thai foundry society
because the Ministry of Industry of Thailand is supporting and encouraging Thai
foundries to learn and make use of casting simulation software to improve their
efficiencies in order to compete in the modern market.

1.6 Rapid Prototyping

Rapid prototyping is the automatic construction of physical objects using solid
freeform fabrication (SFF) (http://en.wikipedia.org/wiki/Rapid_prototyping) which is a
technique for manufacturing solid objects by the sequential delivery of energy and/or
material to specified points in space to produce that solid
(http://en.wikipedia.org/wiki/Solid_freeform_fabrication).
Basically, rapid prototyping takes virtual designs from a Computer Aided Design
(CAD) model and creates a physical model of the design. The object created would
become a prototype so it may be analyzed by the designer to check for the form and
functions of the product. It would be easier for a designer to find ways of improvement
in the design of the product when there is a physical object. Rapid prototyping can
greatly reduce the lead time by reducing or eliminating the tooling design time in
producing a new product.
The physical properties of a prototype produced by rapid prototyping may or may
not be the same as a real product from a casting process. The rapid prototyping process
can produce sand molds for casting, but metal molds are usually not made by rapid
prototyping.

16
www.Iran-mavad.com

1.7 Cast Iron
The interest of this study focuses is on sand casting and cast irons which are
included in the SOLIDCast programs list of materials. A review of cast irons is
presented in the following sections.

1.7.1 Cast Irons

Cast irons are a class of ferrous alloys with carbon contents above 2.0 weight
percent, but in practice, most cast irons contain between 3.0 to 4.5 weight percent carbon.
In some exceptional cases, alloys with less than 2.0 weight percent carbon may be
considered in the cast iron family because silicon and some other alloying elements may
change the maximum solubility of carbon in austenite resulting in solidifying with a
eutectic structure (http://www.key-to-steel.com/Articles/Art63.htm). It can be seen in the
iron-carbon phase diagram Figure 1.3 how close they are to the eutectic composition.
This equilibrium diagram actually represents the metastable equilibrium between iron and
iron carbide (cementite or Fe3C). At 6.67 weight percent carbon, there is 100 percent
iron carbide (Fe3C) (Callister, 2006).
Iron carbide, also called cementite (Fe3C), is a metastable compound. It can
decompose to form alpha () ferrite and graphite if heated over long time periods
(decades) according to the reaction in equation (1.1) (Callister, 2006):

Fe3C 3Fe () + C (graphite) (1.1)

17
www.Iran-mavad.com

 Temperature
(F) = Austenite
= Ferrite
+L
2802
= Delta iron
2720

CM = Cementite

2552 CM begins
+ +L to solidify
2400 L

2200
L + Fe3C

2066

Fe3C
1670 + Fe3C

+
1333

+ Fe3C

0.83% 2% 4.3% 5% 6.67%

Hypoeutectoid Hypereutectoid Hypoeutectic Hypereutectic

Steel Cast Iron

Figure 1.3: Iron-carbon phase diagram.

Modified from http://www.msm.cam.ac.uk/phase-trans/2001/adi/cast.iron.html

Figure 1.4 shows the iron-iron carbide phase diagram (dotted lines) and the true
equilibrium iron-carbon phase diagram (bold line) with graphite instead of cementite as a
stable phase together. Because the decomposition rate of cementite into graphite is
extremely slow, virtually all the carbon formed at the eutectoid temperatures will be in
the form of Fe3C. The carbon formed between the eutectic and eutectoid temperatures
will tend to be in the form of graphite flakes or nodules. So, the iron-iron carbide phase
diagram can be used practically. But if silicon is added to cast irons, normally more than
one weight percent (Callister, 2006), it will slow the cooling rate of the casting and
accelerate the cementite decomposition reaction to form graphite. Slower cooling rates

18
www.Iran-mavad.com

during solidification also help the formation of graphite. Mostly, cast irons will have
carbon in the form of graphite which affects the microstructures and mechanical
properties.

Atomic Percent Carbon

0 5 10 15 20 25

1600

1400

1252C

1200 1153C 4.2

2.1
2.14 1147C 4.3
Temperature C

1000

Fe3C
800 0.65
0.021 740C

0.022 0.76 727C

600

0 1 2 3 4 5 6 6.67

Weight Percent Carbon

Figure 1.4: Iron-iron carbide (dotted lines) and true equilibrium phase diagram (bold
lines)
Modified from:
http://www.mrl.ucsb.edu/~edkramer/LectureVGsMat100B/99Lecture14VGs/FeCPhaseDi
agramVG.html

19
www.Iran-mavad.com

 Sulfur in cast irons is known to favor the formation of the graphite flakes. Adding
a small quantity of calcium carbide can help remove the sulfur from the melt which will
encourage graphite growing into a spheroidal shape. This is followed by a minute
addition of cerium or magnesium in the form of an alloy with iron and silicon rather than
pure magnesium will result in spheroids of graphite. Spheroidal graphite cast iron,
known as ductile iron or nodular iron, has excellent toughness and is used widely
(http://www.msm.cam.ac.uk/phase-trans/2001/adi/cast.iron.html).
Cast irons alloys become completely liquid at temperatures between
approximately 2,100 and 2,350 degrees Fahrenheit, and are easier to melt and cast than
steel. The most common known cast irons are gray iron, nodular or ductile iron, white
iron, malleable iron and compacted graphite iron (Callister, 2006). This thesis has
focused on gray iron as the type of material used in all examples and in the case study.

1.7.2 Gray iron

Gray iron has a carbon content varying between 2.5 to 4.0 weight percent and a
silicon content of around 1.0 to 3.0 weight percent. The graphite in gray iron would exist
in the form of flakes, similar to the shape of a corn flake, normally surrounded by alpha
()-ferrite or pearlite matrix. These graphite flakes make the surface appear gray, hence
its name (Callister, 2006).
The flakes of graphite cause good machinability, because the graphite acts as a
chip-breaker and lubricates the cutting tools. In applications involving wear, the graphite
is beneficial because it helps retain lubricants. But on the other hand, the graphite flakes
are also stress concentrators which lead to poor toughness and low strength. Gray iron is
considered weak and brittle in tension because of its microstructure. The tips of the
graphite flakes are sharp and pointed which serve as points of stress concentration when
there is external tensile strength applied, but it has higher strength and ductility under
compressive loads (Callister, 2006).
Gray iron also has the ability to damp mechanical and sound vibrations, so it is
used as base structures for machines and heavy equipment. It has high fluidity and
castability at casting temperature, makes it easy to cast complicated detailed shapes, has

20
www.Iran-mavad.com

little casting shrinkage and is among the least expensive of all metallic materials
(Callister, 2006).
Chemical elements other than Fe, C and Si may be added in order for gray iron to
improve its mechanical properties. Gray iron castings may be used as-cast without any
treatment, may be annealed for better machinability, or hardened for better corrosion
resistance or to improve strength and ductility.
The metal expands slightly on solidifying as the graphite precipitates. The
amount, size and shape of graphite during the solidification of the liquid metal are
affected by the chemical composition, how the liquid metal is treated and the
solidification rate of the casting in the mold. The matrix structure is affected by the
cooling rate after the casting has solidified until it reaches the room temperature. So, the
mechanical properties of gray iron are affected by the chemical composition, by the
thickness and solidification rate of the casting and by the cooling rate after solidifying.
Specific mechanical properties of the casting can be estimated by determining the
chemical composition. The applications of a type of gray iron must consider its
mechanical properties, which may be tested from pouring liquid metal into a standard
mold apart from the product (Prasertsakul).

1.7.3 Carbon Equivalent

If there were only carbon in the cast irons, there would be very little chance that
graphite would form. But if there were 0.05 to 3.5 weight percent of silicon in the liquid
metal, the graphite is able to form while the liquid metal solidifies. There may also be
phosphorus in the liquid metal up to about 1.0 weight percent which will be in the form
of impurities. Normally, gray irons that require high strength will allow less than 0.12
weight percent phosphorus and also 0.06 to 0.12 weight percent sulfur. If the cast iron is
melted with a cupola, both the sulfur and phosphorus will come from the coke burned.
Manganese is added to the liquid metal to lessen the effect of sulfur. Usually, manganese
would be added at a level of five times the amount of sulfur that is to be removed from
the melt. The manganese left over in the cast iron after solidification will be around 0.05
to 1.5 weight percent, which will help enforce the matrixs strength (Prasertsakul).

21
www.Iran-mavad.com

 The Carbon Equivalent (CE) is a way of measuring the effects of silicon and
phosphorus in the liquid metal which tells the qualities of the liquid metal and the
mechanical properties of the metal after it has solidified. The formula for finding the
percent Carbon Equivalent is shown in equations (1.2) and (1.3):

%CE = %C + 1/4 %Si + 1/2 %P (Prasertsakul) (1.2)

or %CE = %C + 1/3 (%Si + %P) which is more widely used (1.3)

The closer the percent Carbon Equivalent gets the metal composition to reach the
eutectic point, the lower the liquidus temperature will become, fluidity will increase and
will have less tensile strength when solidified. The relationships between the thickness or
diameter of the casting, mechanical properties and the liquidus temperature of the cast
iron with the percent Carbon Equivalent is shown in Figure 1.5.

Liquidus Temperature (F)

2,100 2,150 2,200 2,250 2,300 2,350

250 50 19.0

Section 0.875 in. dia

235 45 18.5

Modulus of elasticity (million psi.)

Tensile strength, (1,000 psi.)
Brinell hardness number

1.2 in. dia

220 40 17.5

2.0 in. dia

205 35 16.5
4.0 in. dia
8.0 in. dia
189 30 15.5

10.0 in. dia

170 25 14.5

150 20 12.5
4.4 4.3 4.2 4.1 4.0 3.9 3.8 3.7 3.6 3.5 3.4 3.3 3.2 3.1

Carbon Equivalent %CE = %C + 1/3 (%Si + %P)

Figure 1.5: The relationships between the thickness or diameter of the casting,
mechanical properties and the liquidus temperature of the cast iron with the percent
Carbon Equivalent (Reproduced from Prasertsakul).

The liquidus temperature is very sensitive to the percent Carbon Equivalent. The
cooling curve is a method to find the liquidus temperature, which is used to find the

22
www.Iran-mavad.com

percent Carbon Equivalent value. The wedge test and the chill test can also find the
percent Carbon Equivalent of the liquid metal but are less accurate (Prasertsakul).
The Carbon Equivalent of cast irons helps to distinguish the gray irons which cool
into a microstructure containing graphite from the white irons where the carbon is
normally in the cementite form. A high cooling rate and low Carbon Equivalent favors
the formation of white iron whereas a low cooling rate and a high Carbon Equivalent
promotes gray iron.

1.7.4 Other Types of Cast Irons

1.7.4.1 Ductile (Nodular) Iron
Ductile iron is made by adding a small amount of magnesium and/or cerium to the
gray iron before casting. This produces different microstructure and mechanical
properties. Graphite forms in nodular or sphere-like particles instead of flakes and is
called nodular or ductile iron. Depending on heat treatment and alloy composition, the
matrix phase surrounding the graphite is either pearlite, ferrite, austenite or martensite.
Normally the matrix is pearlite for an as-cast piece. Heat treating at 1,300 degrees
Fahrenheit for several hours will yield a ferrite matrix. The ferrite or pearlite types have
mechanical characteristics very close to steel (Callister, 2006).

1.7.4.2 White Iron

Most of the carbons in low-silicon cast irons (irons which contain less than 1.0
weight percent Si) and have rapid cooling rates exists as cementite or martensite instead
of graphite. White iron exhibits a white, crystalline fracture surface because fracture
occurs along the iron carbide pallets giving the surface a white appearance, hence its
name. White iron is extremely hard but very brittle and is virtually un-machinable
(Callister, 2006). Normally white iron occurs from rapid cooling; either from being a
very thin casting or the chemical composition may help accelerate the cooling rate.
Thick castings may have a surface layer of white iron while gray iron may form in the
interior. This is called chilled casting, which benefits are having a hard surface and a
somewhat tougher interior.

23
www.Iran-mavad.com

 White iron can also be made by mixing a high percentage of chromium into the
iron. Cr is a strong carbide-forming element, so if the percentage of Cr is high, the
precipitation of graphite out of the iron may be suppressed, so even a large white iron
casting can be achieved. But this type of white iron is not able to be heat treated in order
to decompose the Cr27C3 at all (Prasertsakul).

1.7.4.3 Malleable Iron

White iron, if heated at temperatures between 1,470 and 1,650 degrees Fahrenheit
for a prolonged period of time in a neutral atmosphere (to prevent oxidation) would cause
the cementite to decompose and form graphite, which will exist in the form of clusters or
rosettes surrounded by a ferrite or pearlite matrix depending on the cooling rate (Callister,
2006). This is called graphitization annealing. This would make this cast iron
machinable and more ductile (Prasertsakul). The microstructure of malleable iron is very
close to of ductile iron, which accounts for higher strength, appreciable ductility or
malleability (Callister, 2006).

1.7.4.4 Compacted Graphite Iron

This type of cast iron is actually new to the family. The same as gray, ductile and
malleable irons, carbon in compacted graphite iron (CGI) exists in the form of graphite,
which formation is promoted by the existence of silicon. Silicon ranges between 1.7 and
3.0 weight percent and carbon ranges between 3.1 and 4.0 weight percent.
The microstructure of graphite in compacted graphite iron is worm-like or
vermicular shaped. In a sense, this microstructure has a look in-between flake shape of
gray iron and nodular shape of ductile iron and a small amount (less than 20 percent of
the graphite) may be nodular shaped. Fracture and fatigue resistance has improved as the
sharp edges of the flakes are reduced. Magnesium and/or cerium is also added but in a
lesser amount than of ductile iron (Callister, 2006). Figure 1.6 shows the types of cast
iron which result from different carbon composition ranges and schematic
microstructures which result from a variety of heat treatments. Table 1.1 shows the
typical chemical composition of each type of cast iron (Prasertsakul).

24
www.Iran-mavad.com

Table 1.1: The typical chemical composition of each type of cast iron (Prasertsakul).
Type of Cast Iron wt% C wt% Si wt% Mn wt% P wt% S wt% Cr

Gray cast iron 2.5 4.0 3.0 1.0 0.5 1.4 0.05 0.20 < 0.2 -

Ductile (nodular) Mg
2.5 4.5 4.0 1.2 0.3 0.8 < 0.05 < 0.03
cast iron 0.02 0.07

Blackheart
2.0 2.9 1.5 0.8 < 0.4 < 0.2 < 0.2 -
malleable cast iron

Whiteheart
2.8 3.2 1.11 0.60 < 0.5 < 0.2 < 0.3 < 0.15
malleable cast iron

Pearlitic malleable
2.0 2.6 1.5 1.0 0.2 1.1 < 0.2 < 0.2 < 0.08
cast iron

The effects of each element in the chemical composition of cast iron may be
explained as follows:

Carbon mostly found in the form of graphite in cast irons, it would give a cast iron
different qualities, mechanical properties, strength, ductility, etc., depending on its
structure, which are different in each kind of casting. Hypoeutectic alloys when
solidifying, after reaching its eutectic point, the graphite would begin to participate and
causes the alloy to expand. Hypereutectic alloys would expand immediately upon
solidification or when reaching its liquidus point.

Silicon would help the graphite formation by accelerating the cementite decomposition
reaction and slowing down the cooling rate.

Magnesium or Cerium would help graphite form in a nodular shape. They may be
added to gray iron to produce ductile iron by reacting with the sulfur. Magnesium tends
to encourage the precipitation of cementite, so normally silicon will be added as an
inoculant to ensure the formation of graphite.

25
www.Iran-mavad.com

 Temperature

Commercial cast iron range Fe3C C

Mg/Ce

Fast cool Moderate Slow cool Moderate Slow cool

P + Fe3C P + Gf + Gf P + Gn + Gn

White Pearlitic gray Ferritic gray Pearlitic Ferritic

cast iron cast iron cast iron ductile ductile
Reheat: hold at cast iron cast iron
~1,300F for 30+ hrs
From the iron-carbon phase diagram, composition

Fast cool Slow cool ranges for commercial cast irons. Also shown are
P + Gr + Gr schematic microstructures that result from a
variety of heat treatments.
Gf Flake graphite
Gr Graphite rosettes
Gn Graphite nodules
P Pearlitic
Pearlitic Ferritic
malleable malleable Ferrite

Figure 1.6: Types of cast iron which result from different carbon composition ranges and
schematic microstructures which result from a variety of heat treatments.
Figure reproduced from (Callister, 2006)

26
www.Iran-mavad.com

Phosphorus is one of the elements apart from carbon and silicon that helps the cast
iron solidify closer to the eutectic point in the Carbon Equivalent (CE) formula.

Sulfur favors the formation of graphite flakes. Calcium carbide can be added to get rid
of the sulfur in the melt. Manganese can also be added to lessen the effect of sulfur.

Chromium can be added at a high percentage to suppress the precipitation of graphite

to be able to cast a large white iron casting. This type of white iron may not be treated to
eliminate Cr27C3 at all.

1.8 Cast Iron Alloys Appearing in SOLIDCast

From the previous section, in could be said that there are many types of cast irons
in the cast iron industry. There are also some types of cast iron that were not mentioned
as well. In the casting industry, alloys may be separated further into different types and
classes. Only the alloys appeared in SOLIDCast will be mentioned.
In the following, a list of cast iron alloys in SOLIDCast is shown:
1) CI A532 IA
2) CI A532 IB
3) CI A532 IC
4) CI A532 ID
5) CI A532 IIA
6) CI A532 IIB
7) CI A532 IIC
8) CI A532 IID
9) CI A532 IIE
10) CI A532 IIIA
11) CI DI Ferr
12) CI DI Pearl
13) CI GI 3.5 CE
14) CI GI 4.0 CE
15) CI GI 4.4 CE

27
www.Iran-mavad.com

 16) CI Malleable
List numbers 1 through 10 are wear-resistant cast iron alloys or white irons which
are classified by the American Society for Testing and Materials (ASTM). List numbers
11 and 12 are ductile irons with ferritic and pearlitic matrices respectively. List numbers
13, 14 and 15 are gray irons with 3.5, 4.0 and 4.4 Carbon Equivalent respectively. The
program would let the user adjust the parameter values in the Materials List for these
gray iron alloys. List number 16 is malleable iron which actually solidifies as white iron.

1.9 Problems Occurred in Casting

Before one may use a casting simulation program, he/she should know what
defects may occur in a casting and also the cause of each kind of defect, because casting
simulation programs may only forecast defects that occur from the solidification process.
Simulation cannot forecast defects that occur from human error or errors in the
production process. The details in using simulation software will be explained later on in
this thesis.
A casting that does not fulfill the requirements of a customer is an unacceptable
one. It either becomes scrap or is repaired and becomes a good casting (Pearce, 2001).
Generally, what customers look for in a good casting are its;
1) Physical features, which are the dimensional tolerances, the surface condition and
integrity.
2) Mechanical properties and other properties, including machineability, which may
depend on its:
- Chemical composition
- Macrostructure and microstructure
- Segregation
- Inclusions
Defects that occur in castings may be separated into;
1) Overall shape defects
- Oversized, distortion and misruns
2) Surface defects
- Burn-ons and scabs

28
www.Iran-mavad.com

 3) Internal defects
- Shrinkages, sub-surface holes, pores and cracks
4) Structural defects
- Undesirable microstructural constituents
- Excess inclusion level and oversized eutectic grains or cells, etc.
The overall shape defects and surface defects are the simplest defects to notice.
They may be found after knockout, after cutting or breaking off the gating system
(fettling) or after cleaning the casting. These defects mostly are caused from the mold
and/or while filling the molten metal. Internal defects can be found by using an
appropriate non-destructive test method or may be found after the machining process.
These defects are hard to detect and also hard to prevent. Structural defects mostly can
be found through mechanical tests, such as hardness or tensile strength or can be seen
through a microscope (Pearce, 2001).
Because one kind of defect may be called by many names, the International
Committee of Foundry Technical Associations (CIATF) has standardized how each
casting defect can be grouped and named, as follows:
1) Metallic projections such as flashes, excess metal on surface, mold material,
erosions, core misplacement, etc.
2) Cavities and porosities such as blow holes, macro or micro-porosities, pores,
centerline shrinkage, corner shrinkage, etc.
3) Discontinuities such as hot tearing, cold tearing, cold shuts, cold breakage, etc.
4) Defective surface
5) Incomplete casting
6) Incorrect dimensions and shapes
7) Inclusions or structural anomalies such as alien objects which may accidentally
enter the mold, such as loose sand, inoculants, inoculant products, etc.

1.9.1 Solidification Shrinkage Defects

Shrinkage in castings may be separated roughly according to size, macroporosity
and microporosity shrinkages.

29
www.Iran-mavad.com

 Macroporosity may appear as an area consisting of many shrinkage holes or a
single cavity with a rough surface. Macroporosity mostly occurs at isolated hot spots in
short freezing range alloys, which typical locations are the middle of thick sections,
junctions, corners and areas between two or more cores. Macroporosity could be
detected by cutting open the area or by using non-destructive tests, such as X-ray,
ultrasound and magnetic methods (Ravi, 2005). It is named macroporosity because this
type of porosity is large enough for the naked eye to see.
Microporosity appears as small holes of rough surface and usually detected during
machining. It mostly occurs in castings of long freezing range alloys and sometimes in
thick sections of short freezing alloys. They may be hardly seen with the naked eye,
hence its name, but it affects the strength of critical sections. In long thick sections, it
may appear as a dotted line called as centerline shrinkage (Ravi, 2005).
Areas in a casting most prone to shrinkage defects would be characterized by high
temperature, coupled with low gradient and high cooling rate. High temperatures
indicate fewer directions from where liquid metal can flow in to fill up an area where
shrinkage occurs. Low gradient would imply that even if liquid metal is available at a
neighboring region, there may be insufficient thermal pressure for the flow to actually
take place. For high cooling rate, it would imply that even though there is sufficient
liquid metal and gradients, the metal may freeze too quickly before reaching the hot spot
(Ravi, 2005).
Cracks or hot tears may occur in locations with high temperature, high gradient
and high cooling rate, coupled with a sharp corner. High temperature would contribute to
low strength while high gradient and high cooling rate would contribute to stress. A
sharp corner would contribute to crack initiation (Ravi, 2005).

1.9.2 Defects Occurred in Cast Iron

Many types and kinds of defects may occur in cast irons. Physical defects that
could be found in cast iron may occur from shrinkage, chemical composition, inclusions
or from the process. This thesis would only focus on the defects that can be detected by
SOLIDCast; defects occurred from the solidification process.

30
www.Iran-mavad.com

1.10 Feeders or Risers
Feeders or risers are designed to compensate the shrinkage in the solidification of
castings to eliminate porosities (Ravi, 2005). A feeder or riser would act like a reservoir
of liquid metal which would feed into an area of a casting which needs extra metal. Since
its action is to feed the casting, it seems to be more appropriate to call it a feeder than a
riser. Actually, the word riser is an American term which is not descriptive in a helpful
way (Campbell, 2003). It probably comes from the fact that feeders are located on the
surface or above the casting; that is they rise above the casting surface (Creese, 2007).
But since SOLIDCast uses the term riser, it shall be used throughout the thesis.

1.10.1 The Seven Feeding Rules

It can be considered very difficult adding a riser or multiple risers at the right
positions and with the right dimensions. There are seven rules that could be used to help
in the systematic approach to a solution (Campbell, 2003), which are explained in the
following:

1) Rule 1
Do not feed (unless absolutely necessary) (Campbell, 2003). There are several
reasons why risers should not be placed on a casting. One of the most obvious ones is
cost. It needs extra money to add a riser and more in removing it after the casting is done.
Some castings do not necessarily need risers, such as in some small and medium size
castings. Gray iron is a material which would expand from graphite growth that would
counteract shrinkage and may not need any feeding in many cases (Webster, 1980).
Many castings may be impaired from inappropriate placement of the riser and from using
an appropriate riser size (Campbell, 2003).

2) Rule 2
The riser must solidify at the same time as or later than the casting.

31
www.Iran-mavad.com

3) Rule 3
The riser must contain sufficient amount of liquid to compensate the shrinkage in
the casting.

4) Rule 4
The neck connecting the riser to the casting should not contain a hot spot or
should not solidify later than either the casting or the riser.

5) Rule 5
There must be a path for the liquid metal from the riser to be able to reach the
regions in the casting which need it.

6) Rule 6
There must be sufficient pressure from the riser to push the liquid metal to flow
into the requiring areas.

7) Rule 7
There must be a sufficient amount of pressure in all points of the casting to help
suppress the growth of porosities and prevent them from forming.

All rules must be fulfilled in order to achieve a sound casting. Also, there may be
other methods that would eliminate the use of risers and still achieve a sound casting, for
example, applying chills and cooling fins, etc. Also, Rules 2, 3 and 4 can be gathered
together into one criterion that the thermal center of the total casting, including all risers
should be in the risers. Or it can be simply said that the thermal hot spots should be in
the risers only (Campbell, 2003).

1.10.2 Riser Location and Shape

The direction of solidification inside a casting would start from an end region and
then to the intermediate regions into the region where metal would solidify last. The feed
metal would flow in an opposite direction from the hottest region to the regions which

32
www.Iran-mavad.com

solidify first. A feed path is the entire path starting from a local hot spot to an end region.
Any intermediate point on a feed path should have only one adjacent point with a higher
temperature. The hot spot should feed all regions along the feed path effectively, so it
should be located in the riser or risers for the casting to be defect free (Ravi, 2005).
If there is only one major hot spot in a casting, the riser should be located on the
face closest to the hot spot. If there is more than one isolated hot spot, multiple risers
may be required for each hot spot. For isolated hot spots with different solidification
times, the riser for the hotter hot spot should be designed and analyzed first, and then
verified if the same riser could feed any other hot spot or not to reduce the number of
risers. Then another riser should be designed for the next largest hot spot and so on. For
minor hot spots, chills may be used instead (Ravi, 2005).
Risers may be classified as top and side, hence their names, top risers are placed
above a hot spot while side risers are placed at the side of the hot spot, usually at the
parting line. Top risers are more effective because of the additional effect of gravity.
Side risers can be fed with liquid metal directly from the ingates, which would keep the
riser hotter for longer, implying that a smaller riser can be used (Ravi, 2005). Risers may
also be classified as open and blind. Open risers have their tops exposed to the
atmosphere showing the liquid metal rising as the mold is filled, while blind risers are
hidden inside the mold. For sand castings, open risers would lose heat faster than blind
risers, but on other hand, blind risers in metal molds would lose heat faster than open
risers since heat transfer by conduction through the metal mold is greater than heat
transfer by convection and radiation through air (Ravi, 2005).
A spherical shape would have the lowest surface area compared to its volume or
the highest modulus compared to any other shape, therefore would have the longest
solidification time and would be ideal for a riser, but in practice, other shapes are used
because of how the formation of shrinkage pipe (which may extend into the casting) and
molding constraints (mainly undercuts) (Ravi, 2005). Cylindrical shape side risers have a
height per diameter ratio of 2 or more for steel castings, about 1.5 for iron castings and 3
to 4 for aluminum castings. Cylindrical risers are widely used with small castings. For
larger castings, cylindrical risers with a spherical bottoms for side risers or spherical tops
for top risers are widely used (Ravi, 2005).

33
www.Iran-mavad.com

1.10.3 Riser and Neck Design
The most widely used neck shapes are rectangular for side risers and cylindrical
for top risers. The neck may be tapered down to the casting which would help in fettling
or knocking off the risers from the casting. This would not affect the neck modulus or its
solidification time because of low heat transfer from the sharp reentrant corner. A riser
designed for a given hot spot must satisfy three major requirements, which are the
solidification time, feed path and feed metal volume requirements.
The riser must solidify later, or have more solidification time than the nearest hot
spot, so, it must have a higher modulus compared to the casting region around the hot
spot. This may be represented by equation (1.4) shown below:

Mr = kr Mh (1.4)
Where,
Mr = Modulus of riser
Mh = Modulus around hot spot
kr = Riser design factor

The riser design factor is usually set at more than 1.0. It may be over 1.1 for
ductile iron castings and over 1.2 for aluminum and steel castings. If there is an
intermediate portion of the casting between the riser and the hot spot, the riser design
factor may be raised up to 1.4 or more (Ravi, 2005).
A clear feed path should be available between the riser and the hot spot. This
means that there should be sufficient thermal gradients for the liquid metal to flow from
the riser into the thermal hot spot. So, in order to achieve this, the modulus must
gradually increase from the hot spot to the intermediate section to the neck to the riser,
which should have the highest value. Similar to equation (1.4), the equations may be
represented by equations (1.5) shown below:

Mr = k1 Mn
Mn = k2 M i
Mi = k3 Mh (1.5)

34
www.Iran-mavad.com

Where,
Mr = Modulus of riser
Mn = Modulus of neck
Mi = Modulus of intermediate section
Mh = Modulus around hot spot
kn = 1,2,3 = Design factors

For metals which exhibit volumetric shrinkage during solidification, all kn should
be more than 1, such as aluminum and steel castings should have a minimum value of kn
of 1.1. So, the resulting modulus can be arranged as Mr > Mn > Mi > Mh. However, in
the cases of some gray irons and low grade ductile iron which would expand towards the
end of solidification, the value of k2 should be less than 1.0 to prevent reverse feeding of
liquid metal from the casting into the riser through the neck (Ravi, 2005).
The third requirement, the riser volume must be able to compensate the shrinkage
of the hot spot. Equation (1.6) shown below may represent this criterion:

r Vf = (Vh + Vr) (1.6)

Where,
r = Riser efficiency
= Volumetric shrinkage of the casting material
Vf = Riser volume
Vh = Volume of the region around the hot spot

The riser efficiency depends on the feeder shape, type (open or blind) and
application feed-aids, such as insulators or exothermic sleeves. It has to be considered
because the riser itself is also solidifying as it is feeding liquid metal to the casting. For
an open riser with a height per diameter ratio of 1.5, the efficiency is around 14 percent.
It can be increased to about 50 percent if insulated with sleeves or exothermic sleeves.
The net volumetric shrinkage may range from 1 to 2 percent for gray irons, 3 to 4 percent
for steel, 4 to 6 percent for copper and 5 to 7 percent for aluminum alloys. This equation

35
www.Iran-mavad.com

may fail in the cases of large castings with thin sections and molds with risers connected
to more than one casting (Ravi, 2005). Figure 1.7 shows simple guidelines in designing a
side riser and a top riser with a height per diameter ratio of 1.0 (Heine, 1995).

Casting
Riser
D
Casting
DN Riser D
DN
D/2 D/2

LN LN

LN D/2 DN = 1.2 LN + 0.1 D

a) Side riser

D D

Riser
D
Riser

LN
Casting
Casting
DN
DN

DN = LN + 0.2 D
LN D/2
b) Top riser

Figure 1.7: Simple guidelines in designing a side riser and a top riser with a height per
diameter ratio of 1.0 (Modified from Heine (1995)).

36
www.Iran-mavad.com

 The size of a riser should not be too large or too small. A riser too large would be
a waste of casting material, and being large, it would have a slow cooling rate and might
keep the area of the casting connected to it a hot spot which may be the cause of
porosities. A riser too small would not be able to feed enough molten metal into the
casting and shrinkage would occur. Also, the size of a neck should not be too large or too
small as well. A neck too large would result in a high modulus and the neck would cool
down too slowly causing it to be a hot spot. A neck too small would result in low
modulus and it may cool down too rapidly preventing enough liquid metal to be fed into
the casting.

1.11 Gating Design

A well-designed gating system would ensure smooth, uniform and complete
filling of liquid metal into the mold cavity resulting in a clean casting free of
discontinuities, solid inclusions and voids. Clean metal implies no slag or inclusions
entering the mold and with minimal surface turbulence, smooth filling implies minimal
bulk turbulence, uniform filling would mean that all portions of the casting are filled at
the same time and complete filling would mean that the liquid metal is led to the end and
thin sections with minimal resistance (Ravi, 2005).
The first step of designing the gating system is selecting the type of gating system
and the layout of the gating channels, which are the orientation and position of the sprue,
the runner and the ingate or ingates into the casting. There are also occasions when a
riser or risers may be positioned in the gating system, for example, the ingate from the
runner is led into a riser, then into the casting cavity through its neck. The major
elements of the gating system are the pouring basin, the sprue, the sprue well, the runner
or runners and the ingate or ingates.
The gating systems can be classified as horizontal and vertical gating systems.
Horizontal gating systems are those with a horizontal parting plane containing the runners
and ingates with a vertical sprue perpendicular to the parting plane. This type of gating
system is suitable for flat castings filled under gravity, such as green sand and gravity die
casting. Vertical gating systems are those with a vertical parting plane containing the
runners and ingates. For gravity filling processes, such as high pressure sand molding,

37
www.Iran-mavad.com

shell molding and gravity die casting, the sprue is vertical, which is suitable for tall
castings. For pressure die casting, the sprue may be horizontal, perpendicular to the
parting plane.
Also, depending on the position of the ingate or ingates, the gating system may
also be classified as top, bottom and parting line gating systems. Top gating systems are
for identifying gating systems with an ingate or ingates entering from the top of the
casting cavity, promoting directional solidification from the bottom to the top of the
casting. This type of gating system is suitable for flat castings to limit the damage
occurred from the liquid metal free falling into the mold during initial filling. Bottom
gating systems would have the ingate or ingates entering the mold from the bottom
minimizing turbulence. It is recommended for tall castings which free falling of molten
metal should be avoided. Parting line gating systems would have the ingate or ingates at
the parting line, usually at the middle, combining the characteristics of both top and
bottom gating systems. They are also easier to produce and to modify during trial runs
(Ravi, 2005).
The most critical design decision is finding the ideal filling time which is based
on how the gating channels are designed. A casting that is filled too slowly can have
discontinuities, such as cold shuts and misruns, and premature freezing in thin sections
before completion of filling. But if it is filled too fast, inclusions and blow holes may
occur from surface turbulence. The correct filling time would lie somewhere in between,
and is a function of cast metal, weight, minimum section thickness and pouring
temperature (Ravi, 2005). Several empirical equations for determining the appropriate
filling time were developed by many researchers. Equations (1.7) (Ravi, 2005), (1.8) and
(1.9) (Heine, 1995) are some examples of equations used for calculating the optimal
filling time.

tf = K0 (Kf Lf/1,000) x (Ks + Kt t/20) x (Kw W)P (1.7)

Where,
tf = Optimal filling time, seconds
K0 = Overall coefficient
Kf = Coefficient for fluidity

38
www.Iran-mavad.com

 Lf = Fluidity, millimeters
Ks = Coefficient for size
Kt = Coefficient for thickness
t = Thickness, millimeters
Kw = Coefficient in weight
W = Weight, kilograms
P = Coefficient of pouring temperature

For gray iron castings with sizes between 100 and 1,000 millimeters and wall
thicknesses up to 10 millimeters, the following values may be used:

K0 = 1.0
Kf = 1.0
Ks = 1.1
Kt = 1.4
Kw = 1.0
P = 0.4

Equations (1.8) and (1.9) are for calculating the optimal filling times for gray iron
castings weighing no more than 1,000 pounds and weighing more than 1,000 pounds
respectively (Heine, 1995).

tf = (F/40) x (0.95 + T/0.853) x W where W 1,000 (1.8)

tf = (F/40) x (0.95 + T/0.853) x W where W > 1,000 (1.9)
Where,
tf = Optimal filling time, seconds
F = Fluidity of iron, inches
T = Average thickness, inches
W = Weight, pounds

39
www.Iran-mavad.com

 There are several ways to calculate this filling time. The method SOLIDCast uses
in calculating the optimal filling time is different from the shown equations and shall be
mentioned later in Chapter 4.

Table 1.2: Selected sprue-runner-gate area ratios (Heine, 1995).

Casting Sprue : Runner : Gate ratio
Steel 1:2:1.5
1:3:3
1:1:0.7
1:2:2
- Fin-gated 1:1:1
Gray cast iron 1:4:4
- Pressurized system 1:1.3:1.1
Ductile iron, dry sand molds 10:9:8
- Shell molded, vertical pouring 1:2:2
- Pressure system 4:8:3
- Reverse choke 1.2:1:2
Aluminum 1:2:4
- Pressurized system 1:2:1
- Un-pressurized system 1:3:3
Brass 1:1:1 1:1:3

The sprue well should be designed to minimize turbulence as the liquid metal free
falls down the sprue. The recommended shape of a sprue well is cylindrical with a
diameter twice the diameter of the sprue exit and a depth of 1.5 times the depth of the
runner (Ravi, 2005) or twice the depth of the runner (Campbell, 2004). For the runner, its
main function is to slow down the liquid metal as it speeds up falling down the sprue and
take it to the ingates. The runners must fill completely before letting the molten metal
enter the ingates. The cross-sectional area of the runner should be reduced by the cross-
sectional area of the ingate it passes to ensure uniform flow through the ingates. The
ingates would lead the molten metal from the runners into the mold cavity. They should

40
www.Iran-mavad.com

be easy to fettle or knock off, should not lead to a local hot spot and its cross-sectional
area must be proportional to the volume of the connected casting region it is feeding.
The gating ratio is given by the cross-sectional area ratio of the sprue exit per
runner(s) per ingate(s). In the case of multiple runners and ingates, their total cross-
sectional areas are considered. A larger cross-sectional area would affect the velocity of
liquid metal to slow down as it passes. Different ratios may be used for different cast
metals and systems. Table 1.2 shows some ratios used in different casting situations.
The smallest value in the gating ratio is considered to be the choke, which controls the
flow rate of molten metal.

1.12 Introduction to SOLIDCast

SOLIDcast is a computer simulation software program for castings created and
owned by the Finite Solutions Inc. of USA. The version used in this thesis is Version
6.3.39. It is capable of simulating sand casting, investment casting and permanent mold
casting models with many types of metal.
SOLIDCast is a program which helps a user in predicting defects that might occur
in a casting from the solidification process of a casting. It uses the Finite Difference
Method (FDM) in calculating the thermodynamics of each element in the casting and
records the temperature changes. The temperature changes collected will be used in
plotting graphs of different criterions in predicting different types of defects in a casting.
This program has many functions and features, but only the topics of interest will
be presented.

1.12.1 Output Criteria

The output criteria which can be displayed in SOLIDCast for users to analyze are
as follows:

1) Solidification Time
Solidification time shows the time in minutes for each part of the casting to
become completely solid, i.e., to cool down to the solidus point from the time the pouring
ends. The solidification time plots shows the progressions of solidification through the

41
www.Iran-mavad.com

casting, which can help locate isolated areas of molten metal within the casting and get a
general idea of progressive solidification in various areas of the casting.

2) Critical Fraction Solid Time

Critical fraction solid time is the time in minutes for a part of the casting to reach
the critical fraction solid point, the point at which the molten metal loses its ability to
flow liquid feed metal from the time the pouring ends. It is a better indicator for looking
at progressive solidification than solidification time. It also helps find if there are any
isolated areas formed in the casting that cannot be fed by risers. It gives a good
indication of whether any contraction that forms will be able to be fed by liquid feed
metal with the risers or feeders. This plot is interpreted in the same manner as
solidification time, that is, a casting should show a good progressive cooling, from the
edges of the casting towards the risers.

3) Material Density Function

This criterion is a result of a calculation in which the contraction of the casting
and resulting flow of liquid metal is taken into account during solidification. Areas
having metal removed due to feeding liquid metal to other areas of the casting will have
lower material density numbers. In this way, potential shrinkage can be predicted.
The material density function is a number varying between 0 and 1; 0 meaning
that the metal has been completely drained from that part of the casting while 1 indicates
completely sound metal. It is found that, in general, values in the range between 0.995 or
0.990 and below are areas of detectable shrinkage porosity in castings.
The material density function is a measure of macroporosity and is most useful in
ferrous castings. In non-ferrous alloys, especially aluminum, it can be used to predict
gross macroporosity, but structural microporosity due to poor or stagnant solidification
can best be predicted by other criteria such as the Niyama or the FCC custom criterion.
In ferrous castings, if the Material Density Function is plotted at 1, the plot would
highlight any area that may experience any loss of density. This would overstate the
shape and size of a hole, but would indicate all the locations that might lose any density
whatsoever (David Schmidt of Finite Solutions Inc.).

42
www.Iran-mavad.com

 4) Temperature Gradient
Temperature gradient is a measure of how much change in temperature there is as
you go from point to point within a casting. A high temperature gradient means there is a
large temperature change within a short distance within the casting while a low
temperature gradient means there is a small change in temperature from one point to the
next.
The temperature gradient is calculated at each node within the casting as that
point reaches the Niyama point on the cooling curve. The temperature gradient is always
calculated in degrees Celsius per centimeter.
It can be used to get an idea of whether there is good or poor directional
solidification or not. Usually, higher temperature gradients are good, as steeper
temperature gradients mean a greater driving force for solidification.

5) Cooling Rate
The cooling rate is a measure of how quickly a casting cools down in degrees
Celsius per minute which is measured at each point in the casting as it reaches the
Niyama point on the cooling curve. Cooling rate can be an indication of material quality,
because areas of a casting that cools down rapidly generally has a more favorable grain
structure.

6) The Niyama Criterion

The Niyama criterion is a function based on the temperature gradient and cooling
rate criterions. This criterion is developed by Dr. Niyama, a Japanese researcher studying
shrinkage prediction in steel. He found that the temperature gradient divided by the
square root of the cooling rate, as shown in equation (1.10) corresponded to the presence
of shrinkage porosity in steel castings, which the lower the value, the higher probability
of shrinkage.

Niyama Criterion = Temperature Gradient Cooling Rate (1.10)

43
www.Iran-mavad.com

 The Niyama criterion is basically a prediction of directional solidification. Also,
it can be used to identify potentiality for centerline shrinkage. But Niyama plots need
interpretation, because they give low values in many areas where temperature gradients
are low but shrinkage might not be likely, such as the centerline of thin-wall areas and the
dividing line between the feeding areas of two risers. Also, Niyama is based only on heat
transfer and does not take the effect of gravity into account, which the Material Density
Function does and is a better indicator than Niyama for most ferrous materials. However,
the Niyama criterion is still used heavily in many non-ferrous applications, typically in
aluminum alloys, but rarely used in ferrous alloys anymore according to Mr. David
Schmidt of Finite Solutions Inc.
For plotted values, 0 represents poor directional solidification, good directional
solidification by higher values. The values that are critical values separating between
poor and good directional for each type of material are different. Table 1.3 shows the
critical Niyama values which would be used for plotting Niyama plots recommended by
the SOLIDCast manual (Finite Solutions Inc., 2005).

Table 1.3: Different types of materials and their critical Niyama value.
Material Critical Niyama Value
Steels 1
Cast Irons 0.75
Aluminum 0.30
Copper Base 1.30

It could be said that the Niyama Criterion is a shrinkage potential plot. The
lower the value plotted or the closer to 0, the higher the potential for shrinkage actually
being present. It is not a definite indication of shrinkage, nor does it show the size and
shape of shrinkage (David Schmidt of Finite Solutions Inc.).

7) Hot Spots
Hot spot plotting is a function in SOLIDCast that locates thermal centers or hot
spots within the casting by comparing the solidification times or critical fraction solid

44
www.Iran-mavad.com

times of each metal node to its neighbors. If it froze later than its neighbors, it is an
isolated area, or a hot spot. The range of values is between 0 to 10 which 0 means the
area is most isolated, 1 is stagnant, that is, froze at the same time as its neighbors and 10
is most directional. The method of calculation is to divide a nodes neighbors
solidification time or critical fraction solid time with its own and use the highest value to
represent its hot spot value. To see the hot spots in the casting clearly, plotting at 1.1 or
1.2 would cover a greater volume. 1.1 shall be the value used as the critical plotting
value for the Hot Spot Criterions in all simulation tests in this thesis.
The hot spot plots do not give an indication of the severity of the defect, as it does
not take the contraction or expansion into account, but it does give a good indication of
which areas are prone to have problems.

8) Custom Criterion
The custom criterion is a function that allows the user to plot data from standalone
programs in SOLIDCast, which are available from Finite Solutions Inc. (2005), or the
user can develop custom criteria functions using one of the standalone utilities in the
program.
One custom function called the FCC Criterion is built-in the SOLIDCast software.
It calculates the FCC criterion, which was developed by Franco Chiesa of the Collge de
Trois-Rivires in Quebec (Finite Solutions Inc., 2005). The FCC criterion function was
developed as an indicator of the total evolved microporosity in aluminum castings, which
has also been some success applying to other alloys as well. It also has been proven
useful in identifying thermal centers within castings, for example, the location of possible
secondary shrinkage in iron castings might be displayed.
The FCC criterion can be calculated in the program from selecting the FCC
calculation tab in the menu tab. To display the results, the user must use the Custom-
High criterion to plot.
This FCC criterion is based on the solidus wavefront velocity, meaning the speed
of which the solidus wavefront moves through the casting at each point of the casting,
and the local solidification time, which is the time to cool from liquidus to solidus.

45
www.Iran-mavad.com

 The range of numbers of the FCC criterion developed for any given casting
depends on the alloy and the geometry of the casting, the higher the number, the more
probability of microporosity formation. But it has been suggested to use a 40 percent of
the total range generated by the program as the critical value in starting a plot. Numbers
higher than this will indicate more severe areas and numbers lower indicate less severe
areas.
SOLIDCast allows the user to create a customized formula for plotting a
calculated output from the results of a simulation. The formula can be created with the
form shown in equation (1.11):

Result = K x Te1 x ST e2 x CFS e3 x G e4 x r e5 x LT e6 x LST e7 x V e8 (1.11)

Where
K = Constant
T = Temperature
ST = Solidification Time (min.)
CFS = Critical Fraction Solid Time (min.)
G = Temperature Gradient (C/cm)
r = Cooling Rate (C/min.)
LT = Liquidus Time (min.)
LST = Local Solidification Time (ST LT)
V = Solidus Velocity (cm/min.)
e1 through e8 = exponents applied to each of the above factors
= 0 means that that factor has no significance to the interested output

For example, a formula which has been developed for DAS (Dendrite Arm
Spacing) (Finite Solutions Inc., 2005) in steel is shown in equation (1.12):

DAS = 100 x LST 0.41 (1.12)

This can be evaluated for a simulation by entering the constant and exponents as
shown in equation (1.13):

46
www.Iran-mavad.com

 Result = 100 x T0 x ST 0 x CFS 0 x G 0 x r0 x LT 0 x LST0.41 x V 0 (1.13)

After evaluating the equation, the user could then plot the results in SOLIDCast to
see the calculated DAS in the casting.

9) Other Criterions
Other than the 8 criterions above, there are also other criterions that the user may
use to analyze the simulation, such as the Temperature criterion which the user can
select either the temperature throughout the whole model, the casting and riser(s) only or
the mold only at the time the pouring is finished. And the Liquidus criterion can be
used for seeing how much time until a particular area reaches the liquidus point from the
time the pouring is finished.

1.12.2 Plotting the Results of Output Criteria

From all output criteria mentioned, there are three methods to plot for results:

1) Plotting Results Using Iso-surfaces

An iso-surface plot is a three dimensional model with transparent surface
which displays highlighted areas with a particular value. Only one value entered
by the user can be plotted at a time.

2) Plotting Results Using Cut Planes

Cut plane plotting shows simulation results on a two dimensional plane
cut through the casting. The plot would consist of shades of color representing
the values of simulation results. Only three plane orientations can be observed;
XY-plane, XZ-plane and YZ plane. The planes could not be tilted into any other
orientation.

3) Plotting Results Using CastPic

CastPic is a function in SOLIDCast that can create a detailed three
dimensional image of the casting with shades of color representing the values of

47
www.Iran-mavad.com

 simulation results. A plane could be cut through the model to reveal the
simulation results under the surface of the casting. It can be said it is a
combination of iso-surfaces and cut planes. The plane orientations are the same
as the cut plane plots.

1.12.3 Analyzing the Results

As already discussed in section 1.12.1 and 1.12.2 about each output criteria and
what kind of defects or results may happen in a casting by interpreting the results from
the plots, a typical sequence for analyzing a casting may be done as shown in the
following:
1) Observe the sequence of solidification using the Critical Fraction Solid Time or
Solidification Time Criterion to check for directional solidification and locate
isolated freezing areas. Isolated freezing areas are most prone to casting defects.
2) Next, the Material Density Function is used for finding areas of potential
macroporosity shrinkage.
3) The FCC Criterion is used for finding the areas prone to microporosity in the
casting.
4) The Hot Spot Criterions are used for finding potential hot spot areas.
5) The Niyama Criterion is used for checking the possibility of centerline shrinkage.
The Niyama Criterion is used widely for non-ferrous castings but not a good
criterion for ferrous castings. So, the Niyama Criterion would not be needed for
analyzing a ferrous casting.
These steps may use any method of plotting from section 1.12.2 depending on
whichever plots gives out the most useful information to the user. By following these
steps, the user would learn about the typical characteristics of the simulated casting
design. The user then may also analyze the casting simulation results further with other
output criterions in order to help predict the outcome of the casting design.

1.12.4 Limitations in Predicting Defects in SOLIDCast

SOLIDCast simulates a casting from the pouring of molten metal into the mold.
The program calculates an approximate flow and heat loss during the flow but it cannot

48
www.Iran-mavad.com

calculate the turbulence from the flow. The program simulates and collects only
temperature and time data, starting from when the metal is poured until the last node of
casting material is completely solidified.
Casting simulation software currently do not predict the porosities occurred from
first principles, i.e., the pressure drop which may occur in various parts of the casting.
But many researchers found that there are other parameters which are easier to calculate
and can be used for assessing the porosity formation (Campbell, 2003). From all
criterions available in SOLIDCast and the custom criterion, it was found that all output
criterions are functions of temperature, time and distance, except for the Material Density
Function that also takes gravity into consideration. SOLIDCast only simulates these
values during the solidification.
There are many types of defects that may occur in a casting but there are limited
methods to detect those defects in a casting simulation program. For SOLIDCast, the
predictions of the outcome of a casting design are limited to the output criterions
mentioned earlier, but the user may also create custom criterions to predict certain defects.
However, the formula created for that custom criterion must consist only of functions of
temperature, time and distance.
Also, the time plots, such as the Critical Fraction Solid Time Criterion or
Solidification Time Criterion, etc., the particles may drop under the critical fraction solid
point or solidus point before the pouring is finished causing areas with no highlights at
time zero. It is because the plotting time starts from when the pouring is finished, so in
some occasions, the user may not be able to see the solidification process in some certain
areas.

1.13 The Basic Steps and Functions in SOLIDCast for Sand Casting Cast Irons
This study is interested in the use of SOLIDCast in cast irons cast by the sand
casting method. This section shall show the basic steps of using the program, along with
explanations of the programs functions associated with cast iron casting. The detailed
procedure of using SOLIDCast with cast irons will be shown in Chapters 3, 4 and 5.
The first step is building the model. The casting design can be created by using
SOLIDCast but it can not produce complicated geometrical shapes. A model can be

49
www.Iran-mavad.com

created in other 3D CAD programs and converted into a .stl file format which is the
type of file SOLIDCast is able to read and use.
The next step is to select the casting material, the mold material and other
materials which will be used in or with the mold, such as cores, chills, sleeves, etc. The
default characteristic values of the materials are included in the program, but can be
adjusted if needed. Those characteristics may include thermodynamic characteristics,
densities, cooling curves, etc.
There are also special functions to adjust the temperature curve and the shrinkage
curve of a cast iron. As mentioned in section 1.7, cast irons with different Carbon
Equivalent values would have different liquidus and solidus temperatures according to
the phase diagram. The temperature and shrinkage curves can be adjusted automatically
or semi-automatic by entering the weight percent of carbon, silicon and phosphorus
values to make the simulation as accurate as possible.
There are also functions that may help the user design risers, runners and gating
system, called the Riser Design Wizard and the Gating Design Wizard, so, the user
may design risers, runners and the gating system in a shorter time.
The weight of the casting, risers and gating system can be calculated to help the
user calculate the yield. The optimal filling time for the casting can also be calculated.
When the simulation is done, the user can analyze the results using the different
output criteria and plots. If there are problems in the casting, the user may re-design the
model and re-simulate the model again.

1.14 Extension Programs of SOLIDCast

There are two extensional programs or modules for SOLIDCast created by Finite
Solutions Inc., which are FLOWCast and OPTICast. Even though they are not in the
focus of this thesis, a brief explanation of the two modules shall be explained in the
following:

1.14.1 FLOWCast
FLOWCast is an extensional program that covers up the limitations of calculating
the speeds of the liquid metal particles which may result in defects occurred from

50
www.Iran-mavad.com

turbulence in SOLIDCast. It is a flow simulation module which models the filling of the
gating and mold cavity. It models progressive temperature and velocity of the liquid
metal.
The module can show flow pathlines, or particle trace lines which helps the user
visualize the flow of the metal and determine whether any excessive vortexing or
turbulence may be occurring. This can also be used to predict the likely end location of
foreign particles (oxides or dirt) entrained in the metal stream. Also, the velocity of the
liquid metal at any point can be shown as well.
This module can help the user visualize how the gating system will function and
how the mold will fill. Flow-related defects such as misruns or cold shuts can be
identified as well. It also provides a more realistic temperature distribution in the casting
and mold for a subsequent solidification analysis. The user using only SOLIDCast must
design a gating design which will create minimal turbulence in the pouring of liquid
metal.

1.14.2 OPTICast
This module will help the user optimize the casting, riser and gating system
designs, which shall mean a design that will result in better yield, lower costs, lower
production time, etc.
Optimization is a mathematical method for finding the best solution to a given
problem. It automates the search for a design solution, frees the engineers time and
provides a more thorough and repeatable design process.
The basic steps required for optimization may start from developing an initial
design, defining design variables, constraints and the objective function then launch the
optimization. The design variables are the elements allowed to vary for finding the
optimized design, such as the height and diameter of a riser, the size of a feature of a
casting, pouring temperature, etc. Constraints are aspects of a design whether that design
is acceptable or not. Typical constraints may be the macroporosity level, microporosity
level, yield percentage, minimum cooling rate, minimum thermal gradient, etc. And the
objective function would be the single result which the user would want to either
maximize or minimize. For example, maximize yield percentage, minimize

51
www.Iran-mavad.com

macroporosity level, minimize microporosity level, maximize the cooling rate, maximize
directional solidification, etc.
An experienced foundry engineer would know how to optimize or improve a
casting design. He or she may use SOLIDCast in verifying the improved design if it
would meet the required qualifications. But having OPTICast may help the user to
reduce the time used in optimizing the casting design.

1.15 Commercial Software for Computer Simulation in Metal Casting

There are many casting simulation software programs in the market today. Each
program may have different approaches in creating the model or calculation and may
have different advantages and disadvantages. Some programs shall be reviewed in the
following:
CastCAE is a casting simulation program which uses the Finite Difference
Method approach created by CASTech, Inc., Finland. It claims to be the first in the
world to bring simulation to the Windows environment. It can simulate mold filling,
solidification, iron properties and casting defects for steel, iron, light alloys, copper alloys,
superalloys and precious metals, and casting methods of sand casting, replicast,
investment casting, croning process, gravity die casting, low-pressure and high-pressure
die casting. (www.castech.fi)
CAPCAST was created by EKK, Inc., founded in 1991 and based in Michigan,
USA. The CAPCAST software package includes mesh generation tools, solidification
simulation, porosity simulation, flow simulation and stress simulation, which all of the
tools are based on the Finite Element Method. CAPCAST could model high-pressure die
casting with or without a vacuum, high-pressure die castings with shot sleeve analysis,
various low pressure processes, semi-solid, thixomolding processes, permanent and semi-
permanent casting processes, squeeze castings, green sand castings, hard sand castings,
centrifugal castings, structural investment castings and many other options. It can also
simulate various cast alloys and mold materials. Even though it uses the Finite Element
Method as its approach, it claims to simulate at a very good speed and with high accuracy
and efficiency. (www.ekkinc.com)

52
www.Iran-mavad.com

 MAGMASOFT was created by MAGMA, a German company founded in 1988.
This Finite Element Method program may help the user avoid gating and feeding
problems, predict casting quality, aids permanent mold design and reduces fettling costs.
It contains many modules for calculating specific processes, such as high-pressure die
casting, low-pressure die casting, lost foam, low-pressure die casting for wheel castings,
etc., optimization modules, and modules that could predict thermoelastic/thermoplastic
stresses, residual stresses and strain in castings and molds, modules that can predict the
solidification sequence of cast iron alloys using microscopic kinetic growth models, heat
treatment modules, etc. (www.magmasoft.com)
MAVIS-FLOW is a software program for simulating the mold filling and
solidification of cast alloys, developed by Eidawn Software Limited at the University of
Wales Swansea, UK. It uses the Finite Difference Method approach and includes a
Navier-Stokes mold filling simulator. It includes a database of commonly used casting
alloys and mold materials and the user can also edit and create their own material
properties data files if wished. It can simulate any conventional gravity process, low-
pressure and high-pressure die casting. It has been successfully applied to steel, gray iron,
SG iron or ductile iron, aluminum, copper, zinc and brass alloys. It also has program for
simulating grain growth as an aid to understanding grain structure evolution.
(www.alphacast-software.co.uk)
NovaFlow & Solid was created by NovaCast Technologies founded in 1981 in
southern Sweden. It is a complete mold filling and solidification simulation package
based on advanced fluid flow and heat transfer theories. It also has an add-on program
called NovaStress that calculates the stresses that occur in the casting during
solidification of the material. It uses the Finite Difference Method approach.
(www.novacast.se)
PASSAGE/PowerCast was created by Technalysis, Inc. which is based in
Indianapolis, Indiana, USA. This casting simulation software can be applied to sand
castings, permanent mold castings, die castings, lost foam castings, automotive parts and
appliances. Technalysis, Inc. also developed many other flow and heat transfer
simulation programs apart from casting simulation. (www.technalysis.com)

53
www.Iran-mavad.com

 ProCast was created by Calcom ESI, founded in 1991, an engineering company
located in Lausanne Switzerland, which is also involved in the development and
marketing of the metallurgical simulation software PAM-QUIKCAST and CALCOSOFT.
ProCast is based on the Finite Element Method approach. It covers a wide range of
casting processes and alloy systems, including, high and low-pressure die casting, sand
casting, gravity die casting and tilt pouring, investment casting, shell casting, lost foam
and centrifugal casting. CALCOSOFT is dedicated to the simulation of continuous
casting processes, including, horizontal and vertical continuous and semi-continuous
casting, direct chill casting, strip casting, twin-roll casting and Hazlett process.
PAMQUIKCAST is a software package which allows the user to simulate the entire
casting process from filling to solidification including defects prediction. It uses the
Finite Difference Method approach. (www.calcom.ch)
SIMTEC was founded under the parent company, RWP of Germany founded in
1984. SIMTEC USA opened its first office in Michigan in 1996. Its mission is to supply
engineering teams with simulation software that enables them to develop processes that
produces the highest quality castings in the shortest time at the lowest possible cost. It
was the first leader in Finite Element Method simulation. SIMTEC software provides a
full range of casting simulations. SIMTECs Finite Element Method approached
simulation enables computer-aided calculation of mold filling and solidification for a full
range of casting types, including, common casting processes, such as high and low-
pressure die casting, sand casting, lost foam, permanent mold (conventional/tilt pour),
investment casting, shell casting, and special casting processes, such as, continuous
casting, centrifugal casting, squeeze casting, semi-solid casting. It can also calculate
thermal stress and distortion and strength. (www.simtec-inc.com)

1.16 Problem Statement

There are various types of defects that may occur in a casting. There are also
many casting simulation software programs in the market nowadays. Different casting
simulation software programs and modules may predict some types of defects that may
occur in the casting. The accuracy of the simulation depends both on the program and
the input values it receives from the user. The user must also know how to interpret the

54
www.Iran-mavad.com

results from the simulation in order to know what defects may occur and where they
might take place.
SOLIDCast is one of the most popular PC-based casting simulation systems in the
market (http://www.moderncasting.com/MoreInfo/0804/MoreInfo_06_0804.pdf), and is
also the focus of this thesis. The reasons why SOLIDCast was chosen are, first of all, the
Department of Industrial and Management Systems Engineering of West Virginia
University has already purchased this program for educating their students, so it was
already available for use. This program is based on the Finite Difference Method, so it is
considered to be rather faster than the Finite Element Method and easier to use as it
would not produce meshing gaps as mentioned earlier in section 1.5.3. Also, the user
interface is easy to understand and user-friendly.
Next, SOLIDCast has a built-in flow calculation module which would calculate
the temperature drop during the filling of the mold. Although it does not calculate the
speed of the particles and the turbulence of the flow, it is considered to be very useful as
many other casting simulation programs might not be able to do the same. A separate
program or an extension program is needed to calculate the flow and to transfer the data
to the main casting solidification program. If the user would want detailed calculations
of the filling of the mold, FLOWCast, which is an extension program of SOLIDCast, can
be added. Another extension program for SOLIDCast is OPTICast which can help the
user optimize a casting design.
The focus of this thesis is to study the capabilities and limitations of the program
by going through simple models and the usefulness of SOLIDCast in a real casting
environment with a case study.

55
www.Iran-mavad.com

 Chapter 2
Literature Review
2.1 The History of Metal Casting
Casting technology dates back to around 5,500 B.C. Gold was the first metal
prehistoric man used because of how shiny it was and how easy to be shaped. Then
silver was found and treated similarly. Later mankind found copper appearing in the
ashes of camp fires from copper-bearing ore that lined the fire pits. It was found that
copper was a lot harder than gold and silver and couldnt be bent easily. Copper products
were used as tools and weaponry.
Pottery, made from clay, was found earlier than metal. Man learned how to make
pottery by shaping clay into bowls and hardening by fire. Food was cooked in these pots.
Man learned that gold could be melted with heat in clay pots, and if there was sufficient
heat, copper could be melted, too. These were the beginnings of the Casting Industry
(Danzur).
The oldest known casting in existence probably made around 3,200 B.C. is a
copper frog cast in Mesopotamia, todays modern Iraq (www.metal-technologies.com
and Ravi, 2005). The frogs complexity indicates that it was preceded by other simpler
casting efforts. In Eurasia, copper and bronze tools, decorations and weapons became
common around 3,000 B.C. Bronze refers to a broad range of copper alloys, with tin as
the main additive which makes it stronger and tougher than copper.
Because at that period of time, bronze was the most advanced metal working
technique, it has become the name of that period, The Bronze Age. The Bronze Age
may have started around 3,500 B.C. in the Near East area, or around modern Turkey and
around the Middle East. The Indian Bronze Age began around 3,300 B.C. while Chinas
Bronze Age started around 2,100 B.C. There are many areas in the world which Bronze
Ages starts and ends at different periods (Pernicka, 2003).
The Iron Age was the last period of the three-age system for classifying pre-
historic societies (Stone Age, Bronze Age and Iron Age). The Iron Age corresponds to
the stage at which iron production was the most advanced form of metal working. The
first use of iron probably came from the Sumerians and Egyptians around 4,000 B.C.
(Wikipedia). China started to have iron castings around 1,000 B.C. India made steel

56
www.Iran-mavad.com

around 500 B.C. (Danzur). The casting technology went to Europe from India and the
Middle East through Portuguese explorers in the 14th century (Ravi, 2005).

2.2 Studies of Casting Technology and Simulation Software for Casting

Metal casting has evolved throughout the ages. The techniques have been passed
on and improved through generations. Metal casting, although having a history of
thousands of years, still hasnt stopped evolving. The developments and findings of new
casting techniques and technologies are made everyday.

2.2.1 Casting Techniques and Technologies

The casting techniques and technologies are continuously improved, whether they
be casting materials, mold materials or tooling, etc. Zanchuk (2006) has done a study
about a permanent graphite mold that may become an economical alternative to CNC
(Computer Numerically Controlled) machining, die casting, sand casting and investment
casting for producing parts from ZA-12, a zinc-aluminum alloy that is harder, stronger
and more durable than aluminum, brass, bronze or plastic. Furthermore, the surface
finish of ZA-12 parts virtually eliminates additional finishing steps required with other
casting techniques due to high accuracy and fine, shiny surface, which will result in the
reduction of overall production times and costs.
Other than direct techniques in casting, there may be indirect techniques and
methods in improvement in the casting field, such as, casting management. Apart from
the managements of planning, leading, organizing and controlling, another critical factor
to a die-casters survival in a highly competitive industry is the ability to manage the
simple basics of the casting process. Because each of the casting alloys and die casting
dies behave differently towards temperature, planning should also be applied to
temperature management to optimize die casting quality and throughput (Andersen,
2006).
Herman (2004) has done a study that in die casting, solidification and cooling
rates can be forced by the design of the die cast. If the water lines are properly placed
and at the correct length and the die operated at the correct cycle rate, the maximization

57
www.Iran-mavad.com

of the casting cycle rate could be achieved and the casting production could be
maximized as well.
Guharaja et. al. (2006) have done a study about obtaining the optimal settings of
green sand casting to yield the optimum quality characteristics of spheroidal or ductile
cast iron by using Taguchis method. The process parameters considered were green
strength, moisture content, permeability and mold hardness. The effect of the selected
process parameters and its levels on the casting defects and the subsequent optimal
settings of the parameters were accomplished using Taguchis parameter design approach
and verified by confirming with practical experiments.

2.2.2 Studies Useful for Casting Simulation Software

In order for a casting simulation software program to predict the results of a
casting, there must be studies and researches about the characteristics of each component
in each process. Liu et. al. (2005) studied the influences of casting pressure, the loading
time and the piston position of pressure intensification on the variation of pressure and
the quality of casting because casting pressure conditions in die casting have great
influences on the casting defects, such as gas porosity, shrinkage porosity and gas holes.
It was found that casting pressure, the loading time and the piston position of pressure
intensification have great influences on the pressure variations in the mold, the quality
and the performance of casting.
Normally metals shrink when they lose heat, which is the same as a casting would
shrink when it solidifies in a mold, but sometimes the casting may start to expand after it
cools down to a certain temperature according to what material it is made of. The heat
transfer rate would depend on the heat transfer coefficient (h) between two types of
surface, in the case of casting are the casting and the mold, but since a casting may shrink
in the solidification process, the h value may change because of the gap of air formed by
the shrinkage of the casting. Wang et. al. (2003) has conducted a study to measure the
interfacial heat transfer coefficient (h) between high temperature casting alloys and molds
during the casting due to gap formation. It was found that a high value of interfacial heat
transfer coefficient is generally obtained at the start of the casting, then the value drops
abruptly and then rises to a certain value, and then the value gradually decreases. It was

58
www.Iran-mavad.com

also observed that the heat transfer coefficient (h) value is not considerably affected by
the casting alloys but rather by the mold material; castings with ceramic molds would
have an h value between 22W/m2-K and 350W/m2-K while sand molds are between
40W/m2-K and 90W/m2-K.
De Looze (2005) studied how the operating parameters of a low pressure die cast
(LPDC) machine and the quality level of the aluminum melt affected the casting cooling
rate and/or the microstructure of the aluminum. The formation and distribution of
microporosity in the castings was used as an indicator of casting quality and solidification
conditions, and experimental evidence for the operation of burst feeding in low pressure
die casting was detected. There were significant improvements to the directional
solidification and microstructural refinement were achieved with die cooling.
Wong (2004) has done research on applying Campbells 10 casting rules
(Campbell, 2004) to develop high quality aluminum castings. Campbells rules were
adopted to design the runner and gating system which the proper designs of runner and
gating system included bottom filling, low filling rate and good design of pouring basin,
riser and venting. The CAE software was then used to confirm the designs. In this
research, three castings were made and tested by X-ray examinations and fluorescent
penetrant inspections and all passed. The castings were a CVD heater cast by sand
casting, a gate valve body cast by permanent mold die casting and an aerospace housing
cast by the Quickcast process.
Yang (2003) studied the effect of casting temperature on the properties of gravity
casting and squeeze cast of aluminum alloy with 13.5 percent silicon and zinc alloy with
4.6 percent aluminum and found out that casting temperature had an effect on the
mechanical properties of both the types of casting. The aluminum alloy was tested at
temperatures 660, 690 and 720 degrees Celsius while the zinc alloy was tested at
temperatures of 440, 460 and 480 degrees Celsius. A top-loading crucible furnace was
used to melt the alloys and the die-preheating temperatures used were 200 to 220 degrees
Celsius for the aluminum alloy and 150 to 165 degrees Celsius for the zinc alloy. It was
found that the best temperature to gravity cast the aluminum alloy was 720 degrees
Celsius and for zinc alloy was 460 degrees Celsius. For squeeze casting the aluminum
alloy, either 690 or 660 degrees Celsius would be the best temperature for which the

59
www.Iran-mavad.com

former would give better property at the top of the casting while the latter, at the bottom.
However, the best temperature for squeeze casting the zinc alloy was again 460 degrees
Celsius.
Filipic (2004) has done a study about implementing an optimization tool
consisting in an optimization algorithm and casting process simulator. It was applied to
an industrial casting machine where spray coolant flows were optimized. The manual
setting of coolant flows were significantly improved.

2.2.3 Studies About Casting Simulation Software

Many a casting and casting related simulation software programs are created in
order to achieve the most accurate predictions. Some may have more strength in some
areas than the other. Jakumeit (2003) has done a study of coupling two simulation
programs, the CFD (Computational Fluid Dynamics) program FLUENT and casting
simulation tool CASTS, for simulating a mold filling and solidification for an aerospace
investment casting. It is because of the various physical processes and the geometry. It
is a combination of the expertise of FLUENTs flow simulation and the strength of
CASTS for casting simulation in order to achieve an accurate prediction of this
complexity level of casting.
Also, a paper by Moreira and Ribeiro (2003) discusses the advantages and
limitations of the use of two software packages, FLOW-3D based upon the Finite
Element Method and SOLIDCast based upon the Finite Difference Method. Both
software packages were applied to a specific need for the development of a feeding
system of a real casting produced into a Portuguese sand foundry using equivalent sets of
thermo and physical constants. The conclusion was the development of feeding systems
applied to sand castings are equally powerful for both the programs, but SOLIDCast
retrieved the results in a much shorter time for the same amount of cells.
The Finite Difference and Finite Difference Methods are the two most well
known approaches in casting simulation software. Both methods are meshed based
simulation programs which may have some disadvantages in predicting hot spots and
simulating the jetting/splashing effects during mold filling (Lewis, 2004). From
discussions with experienced foundrymen, Lewis (2004) found alternative techniques for

60
www.Iran-mavad.com

predicting hot spots by using a technique which uses a geometric transformation method
known as the medial axis transformation and a technique based on meshless methods
used for simulating the mold filling process. The medial axis transformation reduces the
dimensionality of the problem by unity, e.g. the medial object of a three-dimensional
object is a two-dimensional surface whereas the medial axis for a two-dimensional object
is simply a curve. The saving in computational time becomes significant if the analysis is
coupled with optimization techniques. Finite Element based optimization techniques are
inherently computationally expensive and hence the objective of this research was to
develop techniques which can speed up the optimization process without significant loss
of accuracy. For the meshless methods, they do not need meshes or grids in their
formulation. Since these particle methods involve only a number of nodal points, or
particles, they are totally free from mesh entanglement and distortion, which may occur
in the computational simulation of large deformation problems using traditional mesh-
based methods. Hence, the advent of mesh-free methods has led to the opening of new
avenues in the numerical computational field. Particle-based methods have emerged as
an attractive alternative for modeling mold filling simulation in casting processes.

2.2.4 Implemented Casting Simulation Software Case-studies

Many casting manufacturers have implemented casting simulation software to
their production and have been successful. Wright (2004) conducted two successful case
studies of implementing casting simulation technology within the company, Walker Die
Casting, which is a producer of complex aluminum castings.
In 1995, a foundry named Raahen Tersvalimo Oy in Finland was casting various
valve components for Neles Controls. The foundry was experiencing some defects in a
particular stainless steel that resulted in repair welding. The foundry considered
investing in casting simulation software and this component was selected as a test case.
The foundry used CastCAE to simulate the test model and the defects were predicted
exactly as what the foundry experienced. A circular chill and insulating sleeves were
added in the system and resulted in a sound casting. Later, the real casting was made
according to the new design and resulted in sound castings (www.castech.fi/).

61
www.Iran-mavad.com

2.2.5 Studies of SOLIDCast
A paper by Moreira (2003) discusses the comparison of two sets of software; one
based upon the Finite Element Method, FLOW-3D, and the other based on the Finite
Difference Method, SOLIDCast. FLOW-3D is considered much more powerful than
SOLIDCast but more complicated to be used. Both packages were applied in developing
a feeding system of a real casting produced in a Portuguese sand foundry. It was found
that for developing feeding systems applied to sand castings, SOLIDCast was equally
powerful and quicker in retrieving results.
A research by Alonso and Franco (2005) used SOLIDCast along with OPTICast
to increase the yield of vertical gating systems in the investment casting process of a
jewelry workshop to improve productivity, quality and promote new product designs.
SOLIDCast was used for verifying the casting models while OPTICast was used for
optimizing the simulation of the vertical gating system for this investment casting of this
jewelry workshop. It was found that these two modules showed a great potential from
improving the design of the filling systems.

62
www.Iran-mavad.com

 Chapter 3
The Use of SOLIDCast
3.1 Introduction
The SOLIDCast program is able to simulate numerous types of casting alloys, but
cast iron seems to be the only type of material which has special functions to calculate
and predict the characteristics of a certain cast iron alloy. For example, the Gray Iron and
Ductile Iron calculator buttons in the Curves tab for adjusting the temperature curve, the
VDG Iron Properties Calculator for adjusting the shrinkage curve and the set points for
the critical fraction solid point and Niyama point. Even an additional Riser Design
Wizard program was built into the VDG Iron Properties Calculator for designing risers
for a cast iron casting.
This thesis is interested in modeling cast iron sand castings using SOLIDCast.
Sand casting is one of the simplest methods in casting and cast iron is one of the most
popular casting alloys in the casting industry. Sand casting needs the least effort in
setting up for simulation compared to other casting methods. The casting design chosen
for the case study is a locomotive piston made from gray iron by sand casting.

3.2 Starting SOLIDCast

Before drawing or loading any shape into the program, the user must set the
system parameters in order to get the correct results and prevent any problems which may
occur from these settings. The system parameter setting is located in the Tools menu.
There are 5 tabs, which are; Alloys Curves, Model Colors, Model & Sim,
Directories and FLOWCast.

3.2.1 Alloy Curves Tab

In the Alloy Curves tab, the user would set the default values for the default
critical fraction solid percentage, usually at 60 percent, default Niyama point percentage,
usually at 65 percent (the percentages in both cases mean the percent of solidification in
the curves graph) and default solidification shrinkage, usually at -7 percent. The negative
sign means volumetric contraction of the metal. This value is actually the net
expansion/contraction of the metal. In the curves graph, the volumetric shrinkage line

63
www.Iran-mavad.com

would linearly incline or decline from 0 percent to the net value at 100 percent of
solidification.
The critical fraction solid point is the point assumed to be the end of feeding or
metal movement for a cast alloy. It means if liquid metal reaches this point, it will no
longer have the ability to flow anymore even though it has not completely solidified. The
Niyama point in the program is the point which the Niyama value is calculated. The
Niyama value of a single node is the temperature gradient divided by the square root of
the cooling rate when it reaches its Niyama point in the cooling curve. It is a value used
as an indicator of the degree of directional solidification and the likelihood of shrinkage
formation; the lower the value, the greater potential for shrinkage formation. It is
normally set at 5 percent after the critical fraction solid point. Figure 3.1 shows the Alloy
Curves tab in the Systems Parameters window.

Figure 3.1: The Alloy Curves tab in the Systems Parameters window.

3.2.2 Model Colors Tab

The Model Colors tab lets the user set up the colors for each material which will
be shown in the model, such as the casting, riser, mold, and fill material and the selected
shape(s). Figure 3.2 shows the Model Colors tab in the System Parameters window.

64
www.Iran-mavad.com

Figure 3.2: The Model Colors in the System Parameters window.

3.2.3 Model & Sim Tab

The Model & Sim tab contains the defaults for measurement system, snap to grid
and the time step interval for volumetric calculations. It also has entries for the
maximum number of facets displayed in an iso-surface plot, a toggle for displaying data
graphically during simulation, and the redraw interval for the graphic display. Figure 3.3
shows the Model & Sim tab in the System Parameters window.

Figure 3.3: The Model & Sim tab in the System Parameters window.

65
www.Iran-mavad.com

 The user must be careful to select what unit of measurement shall be used.
Normally, the default would use English units (inches, degrees Fahrenheit, etc.), but if the
user would want to design or import a model file using Metric measurements (millimeters,
degrees Celsius, etc.), he or she must select the Use Metric Measurements checkbox.

3.2.4 Directories Tab

The Directories tab lets the user set up the default locations for Projects and
Import Files, shown in Figure 3.4. The user should set up the Projects path to an existing
directory, such as C:\Projects. So, when the user creates a new project, the system would
create a new directory underneath this one. The Import Files path is where the user
would be keeping the .stl CAD files. This is also where the system would store the .avi
movie files created by the movie functions in SOLIDCast.

Figure 3.4: The Directories tab in the System Parameters window.

3.3 Adding or Loading Shapes to the Model

SOLIDCast allows the user to design and build the model in the program itself or
import models from other CAD programs as well. Imported files must be in the .stl
format. Although most 3D CAD systems can export .stl files of solid models, users may
have difficulty exporting the files if the solid model was not combined into one single
solid body, etc.

66
www.Iran-mavad.com

 The user must make sure what unit of measurement was used in the designing of
the model and check in the correct box in the Model & Sim tab before loading the
imported file. The user may also draw a design in SOLIDCast. The geometric shapes
provided in the program are quite basic. If the user would like to draw a more complex
design, he or she may use some other 3D CAD program in creating it. The basic
geometric shapes provided in SOLIDCast are:
1) Rectangular solid
2) Solid cylinder aligned with the X axis
3) Solid cylinder aligned with the Y axis
4) Solid cylinder aligned with the Z axis
5) Hollow cylinder aligned with the X axis
6) Hollow cylinder aligned with the Y axis
7) Hollow cylinder aligned with the Z axis
8) Sphere
The shapes can be edited and modified by the user, such as moving the shape or
modifying the radius of a cylinder or changing an elements properties, etc., also, there
are the other commands such as the Revolve and Extrude commands to create more
complicated shapes.

3.4 Selecting Casting Material

Selecting Model and then Materials List from the menu tab, a new window
would appear on the screen. This window will have four tabs for the user to set the
primary parameters for the simulation, which are the Casting Tab, Mold Tab, Curves Tab
and Heat Transfer Coefficients Tab.

3.4.1 Casting Tab

The user will use this tab to select the casting material along with setting the
values of parameters which will be used in the simulation. First, the user may select
alloys from the SOLIDCast database by pressing the From DB button. Many grades
of alloys may be selected, but the scope of this study would focus only on cast irons. The
list of cast irons in SOLIDCast is shown as follows:

67
www.Iran-mavad.com

 1) CI A532 IA
2) CI A532 IB
3) CI A532 IC
4) CI A532 ID
5) CI A532 IIA
6) CI A532 IIB
7) CI A532 IIC
8) CI A532 IID
9) CI A532 IIE
10) CI A532 IIIA
11) CI DI Ferr
12) CI DI Pearl
13) CI GI 3.5 CE
14) CI GI 4.0 CE
15) CI GI 4.4 CE
16) CI Malleable
From list number 1 to 10, are cast iron grades defined by the ASTM (American
Society for Testing and Materials). List number 11 and 12 are ductile cast irons with
ferritic and pearlitic matrices, respectively. List number 13 to 15 would represent gray
iron with 3.5, 4.0 and 4.4 percent Carbon Equivalent (CE). List number 16 is malleable
cast iron. Figure 3.5 shows the Casting tab in the Materials List window.
The user may select a material and modify values of its attributes or create a new
material and save it to the programs database. The attributes of a casting material in the
Casting tab along with its units (English units) may be shown in the following:
1) Alloy Name this is the casting material selected by the user.
2) Thermal Conductivity (Btu/hr-ft-F).
3) Specific Heat (Btu/lbm-F).
4) Density (lbm/cu.ft.).
5) Initial Temperature normally, this temperature would be the pouring
temperature (F).
6) Solidification Temperature can also be called the solidus temperature (F).

68
www.Iran-mavad.com

 7) Freezing Range is simply the difference of the liquidus and solidus temperatures
(F).
8) Latent Heat of Fusion (Btu/lbm).
9) Fill Time is the amount of time used in filling the casting with liquid metal or
could be called the pouring time (seconds).

Figure 3.5: The Casting tab in the Materials List window.

3.4.2 Mold Tab

In this tab, the user would be able to select the mold material(s) which will be
used in the simulation. The mold material in this case does not only mean the material
used for making the mold, but would include materials other than the casting used in the
mold. It would include different sands, cores, chills and insulator materials that may be
added to the model. The list of materials in the database is shown in the following:
1) Air
2) Cast Iron
3) Chromite Sand
4) Copper Chill

69
www.Iran-mavad.com

 5) Exothermic Sleeve
6) Graphite
7) Insulation Sleeve
8) Investment Shell
9) Iron Chill
10) Olivine Sand
11) Silica Sand
12) Steel
13) Stainless Steel
14) Zircon Sand
These materials may be selected and added to the materials list which will be used
in the model. The ambient temperature can also be set in this page.

Figure 3.6: The Mold tab in the Materials List window.

From Figure 3.6, the mold materials attributes are shown and values may be
changed by the user. The attributes are as follows:

70
www.Iran-mavad.com

 1) Name the name of mold material.
2) Type the type of mold, which the user can select one of five types of mold
materials that can be simulated in SOLIDCast, which are:
a. Normal mold is a material that changes temperature as the simulation
progresses.
b. Exothermic mold is a material that burns for a period of time at a
constant temperature and then changes temperature. It may have an
elevated ignition temperature.
c. P.M. core is a special material used to indicate consumable cores used in
permanent mold simulation.
d. Constant temperature is a material which remains fixed at the given
temperature during the simulation.
e. Cooling channel is a material which activates a cooling (or heating)
channel, under the control of either a timer or a thermocouple inserted in
the mold.
3) Initial Temperature of the mold (F).
4) Thermal Conductivity (Btu/hr-ft-F).
5) Specific Heat (Btu/lbm-F).
6) Density (lbm/cu.ft.).
7) Burn Temperature for exothermic molds only (F).
8) Burn Time for exothermic molds only (minutes).
9) Ignition Temperature for exothermic molds only (F).

3.4.3 Curves Tab

This tab shows the temperature and shrinkage curves of the casting material along
with the critical fraction solid point line and the Niyama point line.
The Curves tab is shown in Figure 3.7. The white line represents the temperature
curve comparing with the values on the left vertical axis while the black line represents
the volumetric shrinkage curve comparing with the values on the right vertical axis. The
line with L on top is the liquidus line (100 percent liquid, 0 percent solid) and the line

71
www.Iran-mavad.com

with S on top is the solidus line (100 percent solid, 0 percent liquid). The lines with C
and N below them are the critical fraction solid and Niyama point lines respectively.
These curves have no relationship with time, so the steepness of the temperature
curve does not mean that it has rapid cooling. The information shown in this graph is the
relationship between temperature and volumetric shrinkage.
The curves shown in this window come from the default values from the program.
On the bottom of this window, there are six buttons which the user may use to modify the
curves. The Set Niyama Pt. and Set CF Solid Pt. are buttons used for setting the
Niyama and critical fraction solid points on the graph respectively while the Draw Sol.
(solidification) Curve and Draw Shrk (shrinkage) Curve buttons can be used for
modifying the temperature and shrinkage curves respectively.

Figure 3.7: The Curves tab in the Materials List window.

For gray and ductile irons, there are two buttons to help the user set the curves
automatically. Three inputs are taken from the user, which are the weight percent carbon,
weight percent silicon and percent mold wall movement. Lower values for mold wall
movement represent more rigid molds, such as high-pressure automatic molding and

72
www.Iran-mavad.com

small chemically-bonded sand molds, while higher values represent softer molds such as
squeezer or floor-molded green sands. Normally, green sands would have values of one
to two percent.

Figure 3.8: The iron properties calculator from pressing the Gray Iron or Ductile Iron
button.

Normally, if the shrinkage curve shows hypoeutectic cooling, it is recommended

in the SOLIDCast manual (Finite Solutions Inc., 2005) that the critical fraction solid
point be reset on the eutectic point (the point where the curve starts to incline again after
declining) and the Niyama point set at 5 percent from the critical fraction solid point. But
if the shrinkage curve shows hypereutectic cooling, the critical fraction solid point should
be reset to the range of 20 to 25 percent solid and the Niyama point set at 5 percent from
the critical fraction solid point.
Pressing the reset button would reset the curve back to the default values set in the
system parameters mentioned earlier in section 3.2.1, the Alloy Curves tab, which are the
values of the critical fraction solid and Niyama points and the net percent shrinkage.
Also, the temperature curve would incline linearly from the liquidus line to the solidus
line.
The VDG Iron Properties Calculator located in the Tool menu is a new sub-
program added to SOLIDCast, called vdg.exe, to modify the shrinkage curve, the critical
fraction solid point and Niyama point for cast irons to better represent the characteristics
of the alloy. The VDG Iron Properties Calculator is based on the VDG Nomograms as
published by the German Iron Society (Finite Solutions Inc., 2005). By setting the

73
www.Iran-mavad.com

weight percent carbon, silicon and phosphorus, casting modulus and temperature in the
mold, the values for setting up the shrinkage curve and the critical fraction solid point can
be found. According to the SOLIDCast manual, the temperature in the mold would mean
the average temperature of the metal in the mold after pouring is finished, not the pouring
temperature. This temperature is difficult to measure. So, considering that the metal
does not lose much heat in the process of pouring, the pouring temperature may be used
instead.
When all five inputs in the calculator are set, pressing the Calculate Iron
Properties button would give the results of the shrinkage time in percent of solidification
time and the net expansion or contraction. The shrinkage curve would be modified to
start shrinking from the pouring temperature linearly to the point equal to the calculated
shrinkage time in percent of solidification time with a net shrinkage of the calculated
contraction (or expansion) added the percent mold wall movement. After this point, 0.5
percent expansion would be added for every ten percent of solidification time. This point
where contraction ends and expansion begins could also be called the eutectic point. The
critical fraction solid point should be set at five percent of solidification time beyond the
eutectic point and the Niyama point set at five percent of solidification time beyond the
critical fraction solid point.
Figure 3.9 is an example of the VDG Iron Properties Calculator and Figure 3.10
shows the corresponding shrinkage curve using the results from the calculator in Figure
3.9 with a mold wall movement of 1.5 percent. As could be seen in Figure 3.10, the
eutectic point is set at 90 percent of solidification time with a net shrinkage of 5.6 percent
(4.1 + 1.5 = 5.6). On the bottom of Figure 3.9, there is a button called Riser Design
which may be used for helping the user design riser systems. The Riser Design button
shall be mentioned in detail in Chapter 4.
Regarding an e-mail from Mr. David Schmidt of the Finite Solutions Inc., the
method in generating the shrinkage curve mentioned above is for a good quality cast iron.
For poorer quality cast irons, he recommended that the critical fraction solid point should
be moved more to the left to represent the less amount of expansion to counteract the
shrinkage of the cast iron.

74
www.Iran-mavad.com

Figure 3.9: The VDG Iron Properties Calculator located in the Tools menu.

Figure 3.10: The corresponding shrinkage curve (black line) using the results from Figure
3.9.

75
www.Iran-mavad.com

3.4.4 Heat Transfer Coefficients Tab
Heat transfer coefficients indicate how well or how poorly heat is transferred
between two surfaces. Normally in sand casting, the Use Internal HT Coefficients
checkbox as shown in Figure 3.11 would be left unchecked for the system to consider
only external surface heat transfer coefficients which would normally be set at 1.5 Btu/hr-
sqft-F. But in cases, for example, in permanent mold applications, the Use Internal HT
Coefficients checkbox would be checked and the heat transfer coefficients between all
materials in the model must be specified.

3.5 Testing The Use of SOLIDCast

Tests were conducted to study some of the basic functions in the SOLIDCast
program that are used for simulating a cast iron casting model. Several capabilities and
limitations of the program were found in these tests. The tests and observations were
conducted and the results are summarized.

Figure 3.11: The Materials List window showing the HT Coefficients tab.

76
www.Iran-mavad.com

3.5.1 Basic Functions for Simulating Cast Irons
This set of tests was conducted to observe the input settings in the Materials List
window along with the VDG Iron Properties Calculator. The tests and their brief
conclusions are presented.

3.5.1.1 The Casting Tab and the Curves Tab

First, the cast irons listed in the program and their default values in the Casting
tab and the default graphs in the Curves Tab were observed. Next, a test was conducted
to find the relationships between the each tab in the Materials list window. It was found
that only the Casting tab and the Curves tab have some relationships between each other.
The Mold tab and the Heat Transfer Coefficient tab did not affect any values of the other
tabs in the Material List window when their input parameters were altered.
The input parameters in the Casting tab that affected the graph in the Curves tab
were the specific heat, the initial temperature, the solidification temperature, the freezing
range and the latent heat of fusion. These parameters mentioned would only affect the
temperature curve in the Curves tab. The input parameters in the Casting tab which did
not affect the graph in the Curves tab were the thermal conductivity and the density
values.

3.5.1.2 The Gray Iron and Ductile Iron Calculator Buttons

In the Curves tab, there are buttons labeled Gray Iron and Ductile Iron.
These buttons are used for modifying the graph in the Curves tab in the cases which the
casting used in the model was either made of gray iron or ductile iron respectively. The
input values for these calculators are the weight percent carbon and silicon and the
percent mold wall movement. A test was conducted by observing the graph in the Curves
tab while altering the values of the three input parameters in the Ductile Iron calculator
button using CI DI Ferr or ductile iron with a ferrite matrix from the SOLIDcast database
as the subject.
The main findings in this test was that the relationship between the Carbon
Equivalent (CE) values calculated from the weight percent carbon and silicon and the
liquidus temperatures follows an iron-carbon phase diagram quite accurately. For

77
www.Iran-mavad.com

example, the liquidus temperature would decrease as the Carbon Equivalent value
increases closer to the eutectic point and increase as it passes the eutectic point in an iron-
carbon phase diagram. Also, the test showed hypoeutectic and hypereutectic cooling as
the Carbon Equivalent value increased to the eutectic point and passed it respectively.
The conclusion to this test is that the resulting graph in the Curves tab from using the
calculator buttons were derived from metallurgy knowledge applied into the program.
It was also found in another test that the weight percent carbon and silicon and the
mold wall movement of each ductile and gray iron from the SOLIDCast database could
be found. By trial and error, the weight percent carbon and silicon and the mold wall
movement were adjusted until the graph in the Curves tab would match the default graph
given by the program of a selected material, shown in Table 3.1.

Table 3.1: The weight percent carbon and silicon and mold wall movement of each cast
iron alloy found by trial and error.

%Mold Wall
No. Name %C %Si
Movement

1 CI DI Ferr 3.5 2.6 2

2 CI DI Pearl 3.5 2.6 2
3 CI GI 3.5 CE 2.9 1.5 2
4 CI GI 4.0 CE 3.45 1.4 2
5 CI GI 4.4 CE 3.85 1.4 2

A test was conducted to find the differences between the Gray Iron and Ductile
Iron calculator buttons. It was found that no matter which gray iron or ductile iron was
selected from the database, the same graph would result in the Curves tab. But if
observed closely, the temperature curve from using the Gray Iron calculator button, as it
inclines, would be curved as the steepness of the curve changes, similar to an S shape.
The temperature curve resulted from using the Ductile Iron calculator button would not
curve as the steepness of the temperature curve changes, so the shape of the curve can be
compared similar to a Z shape. Figures 3.12 and 3.13 shows examples of a temperature
curve developed by the Ductile Iron and Gray Iron calculator buttons respectively.

78
www.Iran-mavad.com

Figure 3.12: Sketch of temperature curve developed by the Ductile Iron calculator button.

Figure 3.13: Sketch of temperature curve developed by the Gray Iron calculator button.

3.5.1.3 The VDG Iron Properties Calculator

As mentioned earlier in section 3.4.3 about the Curves tab, the VDG Iron
Properties calculator in the Tools menu would take five inputs; the weight percent carbon,
silicon and phosphorus, casting modulus and temperature in the mold and give two
outputs; the shrinkage time in percent of solidification time and the expansion or
contraction percentage for the user to modify the shrinkage curve and set the critical
fraction solid line and the Niyama line. It does not automatically modify the graph in the

79
www.Iran-mavad.com

Curves tab like the Ductile Iron and Gray Iron calculator buttons do. This calculator is a
stand-alone program that has no relationship with any other functions in the program.
A test was conducted to find the relationships between the input and output values
of the VDG Iron Properties calculator. Table 3.2 shows how the each output value varied
when an input value was increased and decreased.
From the SOLIDCast manual (Finite Solutions Inc., 2005), this VDG Iron
Properties calculator is a stand-alone program in SOLIDCast called vdg.exe which is a
program that gives an estimate of the total amount of contraction (or expansion) for a cast
iron and the point at which this contraction (or expansion) ends based on the VDG
nomograms as published by the German Iron Society. An example of this VDG
nomogram is shown in Figure 3.14 and a test was conducted and proved that the values
calculated by the VDG Iron Properties calculator matches the VDG nomograms. It is
assumed that SOLIDCast has a database of the nomograms for retrieving results when
values are entered into the VDG Iron Properties calculator.

Table 3.2: The relationships of the input and output values in the VDG Iron Properties
calculator.
Outputs Shrinkage Time in Percent
Percent shrinkage
Inputs of Solidification Time
Increase Decrease Decrease
Carbon
Decrease Increase Increase
Increase Decrease Decrease
Silicon
Decrease Increase Increase
Increase Decrease Decrease
Phosphorus
Decrease Increase Increase
Casting Increase Decrease Decrease
Modulus Decrease Increase Increase
Temperature Increase Increase Increase
in Mold Decrease Decrease Decrease

80
www.Iran-mavad.com

Figure 3.14: An example of the VDG nomograms from the German Iron Society using
metric units.

3.5.1.4 Conclusions of Tests

The conclusion to this set of tests is that the information in the database of the
SOLIDCast program and the calculators built into the program were derived from
metallurgy knowledge which were arranged and applied into the program. So, in order
for a simulation to be able to simulate as accurately as possible, the user must enter the
correct data collected from the materials used in the casting process, i.e., the weight
percent carbon, silicon and phosphorus in the casting material, the pouring temperature,
the approximate percent mold wall movement of the mold material used, etc.

3.5.2 Modifying the Curves Tab

In SOLIDCast, the amount of shrinkage of metal at any time would equal to the
volumetric difference of the mold cavity and casting. So, the shrinkage curve in the

81
www.Iran-mavad.com

Curves tab is actually the shrinkage of the metal added the shrinkage occurred from the
expansion of the mold or the percent mold wall movement. A test was conducted and
was found that modifying the shrinkage curve would not affect the weight of the casting
because the weight of the casting is calculated from the size of the nodes, the total
number of nodes and the density of the material. The calculation of the weight of a
casting could only be found from meshing and not from simulation. It was also found
that SOLIDCast does not simulate the dimensional shrinkage or expansion of the casting,
including deformations.
It was found in another test that modifying the shrinkage curve would not affect
any output criterion except the Material Density Function output plots. It is because all
other output criterions would only use temperature data and/or time and/or distance for
calculation. It was also found that in a cylindrical shape casting, which in this case is
representing a riser, the shrinkage would normally consist of two shapes; an upside-down
cone-shape or a funnel shape and a flat cylindrical shape with a diameter slightly smaller
than the cylindrical shape casting or the riser. The test showed that the cylindrical shaped
shrinkage is the shrinkage which occurs during the time when the liquid metal cools
down to its liquidus point and the funnel shape occurs during the time when the metal
cools down from its liquidus point to its critical fraction solid point. The size of the
shrinkage depends on the degree of the shrinkage slope and where the shrinkage ends or
the metal starts expanding. If the shrinkage ends early in the percent of solidification
time, the development of the cone-shape shrinkage would stop abruptly showing very
little or no growth of the cone-shape shrinkage at all. The shrinkage or expansion may
still occur after the metal has reached its critical fraction solid point but in very small
volumes. It was also found that modifying the Niyama line would only affect the Niyama
Criterion output plots as it will only calculate the Niyama value when a particular node
reaches its Niyama point.

3.5.3 Other Tests

A test was conducted to find out if the SOLIDCast program would be able to
simulate a misrun or not. It was found that in the simulation of the flow of metal, the
simulation would allow the molten metal to flow until it reaches its solidification point or

82
www.Iran-mavad.com

solidus point, which in reality the molten metal should lose its ability to flow after it
reaches its critical fraction solid point. This may conclude that the user must use his or
her judgment to interpret where the misrun would actually occur. From this test, it may
be assumed that the program would also be able to detect coldshuts as well if they may
occur.
The maximum number of nodes which could be created by this version of
SOLIDCast is 19,046,664 nodes which could be found from the mesh creating window
shown in Figure 4.3. The Planes of Symmetry function is a useful function which can
help reduce the use of nodes in a simulation. It can reduce the time used for simulating a
casting model using the same node size or achieve more detailed simulation results using
a smaller node size for a symmetrical casting design.
The Planes of Symmetry function was tested and the results showed that the
results from simulating a symmetrical casting using this function may have results
slightly different from simulating a symmetrical casting without using this function. The
reason that the results may be different is because the model does not start the mesh from
the symmetric plane, so, the size of the casting at that symmetrical parting line may be
one node size larger or smaller than the model without using the Planes of Symmetry
function. It was also found that if there is no casting material at the symmetrical parting
plane, the program would assume that the closest point where the casting material is to
the symmetric plane is where the casting model and its symmetrical counter part is
connected, instead of considering them as two objects with a distance twice the distance
between the casting model and the symmetrical plane.
A test was conducted to observe the meshing function. It was found that the
program would start meshing from the edge of mold along the X, Y and Z axes. A
portion of a material would be considered a whole node if it is larger than half the size of
the node size and neglected if smaller, and it was also found that the node size cannot
exceed the size of the thinnest section of any material in the mold. Also, the weights of
each material in the model would be calculated from the number of nodes, the node size
and the density of the material, so, the weights of a model may be different according to
the node size. So, the only situation that the calculated weights of the model would be
exactly correct is the model intersects perfectly with the nodes. For intricate shaped

83
www.Iran-mavad.com

models, the more detailed the meshes could represent the model or the smaller the nodes
could possibly be, the better approximation of the weights of the model. There are also
situations which a symmetrical hollow casting may not be symmetrical after meshing
with a difference of one node size thickness due to the meshing pattern.

3.6 Conclusions
The characteristics of a material in SOLIDCast were derived from metallurgy
knowledge, so, a user must input precise data of the actual alloy into the simulation
program to receive the most accurate results. Some limitations of SOLIDCast were also
found. Being based on the Finite Difference Method, the shape of a meshed model may
differ significantly from the original design according to the node size and alignment of
the model. Problems that may occur, for example, are the thickness of two areas with the
same thickness in a casting design may have a difference in the number of nodes
representing those thicknesses, the weights of the model would only be precise if the
nodes intersect the design perfectly, and because the nodes are square blocks, weights of
casting designs with curves or intricate designs would be more accurate if smaller nodes
are used. Misruns and coldshuts cannot be predicted precisely because the program
would allow the liquid metal to flow until it reaches the solidification temperature, not
the critical fraction temperature which liquid metal should lose the ability to flow.

84
www.Iran-mavad.com

 Chapter 4
Using SOLIDCast: Simple Model
A test was conducted to learn how a user would develop a riser and gating system
for a simple shaped casting model by using the Riser Design Wizard and the Gating
Design Wizard programs built-into the SOLIDCast program. The calculation methods
behind the two programs and the simulation results from using the calculated riser and
gating dimensions were analyzed.

4.1 Drawing the Design

A hollow rectangular box was created in SOLIDCast with outer dimensions of 10
by 15 by 4 inches and inner dimensions of 7 by 12 by 4 inches along the X, Y and Z axes.
So, the walls of this box was 1.5 inches thick with a height of 4 inches. A rectangular
box with dimensions of the outer dimensions of the hollow box was created first using
the casting material as the material. Then, a smaller rectangular box with dimensions of
the inner dimensions of the hollow box was created using void as the material. This
void material is used in SOLIDCast modeling to create a voided shape in any material the
user would like to become the primary mold material. This method is also much easier
than modeling four rectangular shapes that would form into the final shape of this hollow
rectangular box. Figure 4.1 shows how the hollow box appeared in the model.

Figure 4.1: Figure of design drawn in SOLIDCast.

85
www.Iran-mavad.com

4.2 Riser Design and Analysis
The Riser Design Wizard can be accessed from either the Simulation tab in the
menu or from the VDG Iron Properties Calculator from the Tools tab in the menu. In this
test, the Riser Design Wizard from the VDG Iron Properties Calculator was used because
the information calculated from the VDG Iron Properties can be drawn directly to this
Riser Design Wizard and used for designing the riser, and it was specially used for
designing single risers.
The important information the Riser Design Wizard would need were the casting
modulus and the casting weight. The weight could be found from meshing the model
using the selected casting material the user would select from the Casting tab in the
Materials List window while the casting modulus could be found from simulation. The
casting material selected is gray iron with properties similar to the material which will be
used in the case study in Chapter 5. Figure 4.2 shows the Casting tab with the properties
of the material used in this test.

Figure 4.2: Casting tab with properties of material used in test.

86
www.Iran-mavad.com

 The node size used in the meshing of this model was 0.25 inches, so the nodes
would intersect with the model completely and represent the weights of the model and the
casting modulus most accurately. A rectangular shaped mold with silica sand used as the
mold material with one inch thickness surrounding the casting was set in the meshing
function, as shown in Figure 4.3. The weight of the casting material and silica sand was
calculated to be 70.434 pounds and 52.778 pounds respectively, as shown in Figure 4.4.

Figure 4.3: Setting up the meshing parameters.

Figure 4.4: The weights of the model.

87
www.Iran-mavad.com

 Next, the Riser Design Wizard from the Simulation menu was accessed to find the
representative modulus of the casting was in the VDG Iron Properties Calculator for
modifying the volumetric shrinkage curve in the Curves tab. Selecting the Calculate and
Display Casting Modulus option, the maximum modulus value of the casting shall be
used as the representative casting modulus in the VDG Iron Properties Calculator as
recommended by the SOLIDCast manual (Finite Solutions Inc., 2005). The modulus
values calculated from the Riser Design Wizard actually represents the solidification rate
of a node, not the actual modulus of the whole casting, so, the nodes which have higher
modulus values mean that they take more time to solidify. The resulting window
showing the calculated range of modulus values is shown in Figure 4.5 and the casting
modulus value which will be used in the VDG Iron Properties Calculator is 0.6004735
inch. The actual modulus of the whole casting was calculated and shown in equation
(4.1).

Figure 4.5: Finding the casting modulus from the Riser Design Wizard.

Casting modulus = Volume Surface area

= [4 x {(10 x 15) (7 x 12)}] [2 x {(15 x 4) + (10 x 4)
+ ((10 x 15) (7 x 12)) + (12 x 4) + (7 x 4)}]
= 264 cubic inches 484 square inches
= 0.545 inch (4.1)

88
www.Iran-mavad.com

 The actual casting modulus is 0.545 inch, which is very close to the value in the
box, 0.543, as shown in Figure 4.5. But the value in this box is actually a suggested
plotting value for finding approximate feeding areas for the casting, not the actual casting
modulus. For a more accurate number of feeding areas, the neck modulus value is
needed for plotting instead of this suggested value, which could be found later in the
process. Plotting with a larger value would show the areas with higher modulus values or
more solidification time. The highlighted areas would be smaller and closer to the
centerline of the thickness of the casting as the plotting value is increased, meaning that
the centerline of the casting takes a longer time to solidify than the rest of the casting.
Figure 4.6 shows the resulting plot using the recommended value which shows one
feeding area.

Figure 4.6: Modulus plot showing one feeding area plotted with the suggested value,
0.543.

Now, the VDG Iron Properties Calculator shall be accessed from the Tools menu.
The weight percent carbon, silicon and phosphorus, casting modulus and temperature in
mold shall be entered. The weight percent carbon, silicon and phosphorus and
temperature in mold information are similar to the case study in chapter 5. Clicking on
the Calculate Iron Properties button, the shrinkage time in percent of solidification time
and the expansion or contraction will be calculated and shown, as in Figure 4.7. The
Riser Design Wizard can now be accessed by clicking on the Riser Design button, as
shown in Figure 4.7, and the Riser Design window would appear, as shown in Figure 4.8.

89
www.Iran-mavad.com

Figure 4.7: The VDG Iron Properties Calculator.

Figure 4.8: The Riser Design Wizard window accessed from the VDG Iron Properties
Calculator.

90
www.Iran-mavad.com

 In the Riser Design Wizard window in Figure 4.8, the casting modulus value will
be transferred automatically from the VDG Iron Properties Calculator into this window.
This calculation window may only be activated if the weight of volume of the casting and
the riser height per diameter ratio are entered. An analysis about the input and calculated
values for this Riser Design Wizard was done and could be summarized in the following:

4.2.1 Analysis of the Riser Design Wizard Input Values

1) Enter Casting Weight or Volume
Either the casting weight or volume could be entered into this window. But it was
found that the density of the cast iron material the program uses to calculate between the
weight and volume is constant at around 0.25926 pound per cubic inch or 448 pounds per
cubic foot. This sub-program calculator does not take the density of the material set in
the Casting tab nor consider the input values in the VDG Iron Properties Calculator to
calculate for a new density. But since the VDG Iron Properties Calculator is only used
for cast irons, the weights of cast iron alloys are not significantly different, so, this
constant weight may be used as an average weight for cast irons.

2) Percent Mold Dilation

The mold dilation is the same as percent mold wall movement. In calculating the
riser dimensions, the calculated expansion or contraction rate from the VDG Iron
Properties Calculator would be taken into account, added with this percent of mold
dilation. The percentage value shown under the slider bar has only one decimal, but the
actual number used in calculation seems to have more. First of all, one percentage value
seems to consist of many snap grids and each snap grid produces slightly different
calculated values. Also, clicking on the slider arrows would slightly change the
calculated dimensions every time, showing that there are more detailed values in between
the snap grid values. But since the dimensions a user would need are just approximated
dimensions, and the dimensions of objects in SOLIDCast can only be created to the
detailed level of three decimals of a measuring unit, the slight difference of the
calculation would not be a significant problem.

91
www.Iran-mavad.com

3) Sleeve Type
There are three types of riser sleeves (SOLIDCast uses the term sleeve for both
insulating sleeves and exothermic toppings) and the option of having no sleeve, and an
option to set the gating through the riser. When the user selects the type of sleeve and
gating type, the program would automatically give out the proportion of liquid metal
removed from the riser into the casting and the modulus increase factor. It seems that
these values are kept in the program database, which were derived from casting and
metallurgy knowledge. They could also be modified if needed. Table 4.1 shows the
resulting values as the options are chosen.

Table 4.1: Resulting values when type of sleeve and option of gating through the riser is
selected in the Sleeve Type box.
Not gated through riser Gated through riser
Proportion of Proportion of
Sleeve type Modulus Modulus
liquid metal liquid metal
increase factor increase factor
removed (%) removed (%)

None 14 1 20 1

Insulating 32 1.25 33 1.25

Exothermic 35 1.33 35 1.33

Exothermic
70 1.33 70 1.33
Mini-Riser

4) Riser Height Per Diameter Ratio

The user may enter the riser height per diameter ratio of the risers needed for the
casting in this space. The user may enter a riser height per diameter of 1.5 for an iron
casting (Ravi, 2005).

92
www.Iran-mavad.com

5) Riser Type
The user may select either a side or a top riser for the casting model. The
SOLIDCast Training Overview (Finite Solutions Inc., 2005) recommends users to use a
hot riser, i.e., to design the gate into the riser if using a side riser, so that the amount of
heat in the riser and its ability to feed the liquid metal into the casting is maximized. For
top risers, it is difficult to design a gate into this type of riser, so, insulating or exothermic
sleeves are recommended.

6) Short Neck Check Box

In the case the user would want to use a short neck for the riser, the short neck
check box could be checked so the program may calculate the dimensions of a riser with
a short neck. A short neck would be able to cool down slower than a normal length neck
because of the heat from the casting and the riser. The modulus of the neck would be
reduced by multiplying a factor 0.6 to the normal neck modulus resulting in a smaller
neck dimension. In order to be considered as a short neck, the distance between the
casting and the riser should be less than the minimum dimension of the connecting
surface of the riser neck on the casting. The neck dimensions connected to the riser can
be larger and tapered down to the side connected to the casting to allow easier fettling or
break off of the riser.

4.2.2 Analysis of the Riser Design Wizard Output Values

There are actually two methods in calculating for the riser design; calculation
based on volume requirement and calculation based on the riser modulus being 20
percent more than the calculated neck modulus. The latter method of calculation would
only be used if the calculation based on volume requirement does not satisfy the rule that
the riser modulus has to be at least 20 percent larger than the calculated neck modulus,
which from testing would mean the casting size may be too small compared to the casting
modulus. In the following shows the calculation methods based on volume requirement.

93
www.Iran-mavad.com

1) Riser Diameter and Riser Height
The diameter and height of the calculated riser dimensions would be shown in this
space. They would follow the ratio set in the riser height per diameter ratio input space.
For a side riser, the proportion of liquid removed from this cylindrical shaped riser would
equal to the total shrunken volume of the casting. Equation (4.2) would explain this
calculation.

Liquid removed from riser = Total shrunken volume of casting

[ x (D2/4) x (D x r)] x %p = Vc x (%VDG + %MD) (4.2)
Where,
D = Diameter of side riser, inch
r = Riser height per diameter ratio, no unit
Vc = Volume of casting, cubic inch
%p = Proportion of liquid metal removed from riser in percent
%VDG = Percent shrinkage calculated from the VDG Iron Properties Calculator
%MD = Mold dilation, percent

For a top riser, it shall show the dimensions of a cylindrical shape with 20 percent
more volume than a cylindrical shape with the dimensions calculated for a side riser, as
shown in equation (4.3) and Figure 4.9. It is because the program assumes that the side
riser would have a hemispheric bottom, so, the true height of the side riser is more than
the calculated height. For a height per diameter ratio of 1.5, it is approximated that the
volume of the hemispheric bottom is around 20 percent of the cylindrical shape. So, even
though the calculated riser diameter and height for both types of risers are different, the
total volumes of both riser types are approximately the same, as shown in equation (4.4).

Volume of top riser = 1.2 x Volume of cylindrical shape of side riser

Vtop = 1.2 x ( x Dside2 x Hside 4) (4.3)
Where,
Vtop = Volume of top riser, cubic inch
Dside = Diameter of calculated side riser, inch

94
www.Iran-mavad.com

 Hside = Height of calculated side riser, inch

VH VC = [(2/3) x x (D/2)] [ x D2 x (1.5 D) 4]

= 0.222
20% (4.4)
Where,
VH = Volume of hemispheric bottom, cubic inch
VC = Volume of cylindrical shape, cubic inch
D = Diameter of calculated side riser, inch

Dtop
Dside

Vtop = 1.2 Vside Htop

Vside Hside

Neck
Neck

Side Riser Top Riser

Figure 4.9: Comparing the volumes of a side and top riser.

But in cases which the riser height per diameter ratio is not 1.5, the rule of 20
percent would not be true. From equation (4.1), it seems that the effective feeding
volume is the volume of the cylindrical shape of the side riser, excluding the hemispheric
bottom. So, the 20 percent increase in the size of the top riser may be to compensate for
heat loss if it is considered as an open top riser.

95
www.Iran-mavad.com

2) Neck Dimensions
The calculated neck dimensions are shown as either the diameter of a round neck,
a side of a square neck, or if the user would want to use a rectangular shaped neck, there
is a calculator which would calculate the dimensions of a rectangular shaped neck which
would have the same cross-sectional area as a square neck.
The neck dimensions calculated in this Riser Design Wizard considers the neck to
have a cubic shape with four surfaces excluding the surfaces connected to the riser and
casting. The formula for calculating the length of one side of this cubic shaped neck is
shown in equation (4.5). The formula for calculating the neck modulus is shown in
equation (4.6).

Mn = Volume of cubic shaped neck Four surface areas

Mn = L3 (4 x L2)
=L4
L = 4 Mn (4.5)
Where,
Mn = Neck modulus, inches
L = Length of one side of the cubic neck, inches

Normally, a neck would have a rectangular shape, not a cubic shape. If the user
uses the rectangular neck dimensions calculator, the resulting neck would have the same
connecting area as a square neck, but the actual neck modulus would decrease. So, if the
user would like to have a neck with a modulus equal to the calculated modulus, he or she
must recalculate a new neck.

3) Riser Modulus
The riser modulus value in the Riser Modulus box is calculated by multiplying
the modulus increase factor with the riser modulus. The modulus is the volume divided
by the surface area of an object. The side riser type is assumed to have a cylindrical
shape with a diameter and height of the calculated riser diameter and height added a
hemisphere shaped bottom which has an equal diameter as the cylindrical shape. The

96
www.Iran-mavad.com

volume used for calculating the modulus is the volume of the cylindrical shape added the
hemisphere bottom. The area used for calculating the modulus is the total surface area of
this riser less the area of the cross-sectional area of the calculated square-shaped neck.
In the case of a top riser, it is assumed to have a cylindrical shape with a diameter
and height of the calculated riser diameter and riser height. The volume used for
calculating the modulus is the volume of the cylindrical shape. The area used for
calculating the modulus is the total surface area of the cylindrical shape less the area of
the cross-sectional area of the calculated square-shaped neck.

4) Neck Modulus
The neck modulus is calculated from the square root of the shrinkage time in
percent of solidification time from the VDG Iron Properties Calculator multiplied by the
casting modulus. This formula is shown in equation (4.6).

Mn = Mc x %ST (4.6)
Where,
Mn = Neck modulus, inch
Mc = Casting modulus, inch
%ST = Shrinkage time in percent of solidification time from VDG Iron
Properties Calculator, percent

5) Ratio of Riser Modulus per Casting Modulus

This is the value of the riser modulus divided by the casting modulus. It should
be more than 1 showing that the riser should solidify later than the casting.

6) Ratio of Neck Modulus per Casting Modulus

This is the value of the neck modulus divided by the casting modulus. It should
be less than 1 showing that the neck should solidify before the casting does to prevent
liquid metal being pushed back into the riser due to expansion of graphite.

97
www.Iran-mavad.com

4.2.3 The Results of the Riser Design
The final calculation of the Riser Design Wizard is shown in Figure 4.10. The
sample model with a weight of 70.434 pounds and casting modulus of 0.6004735 inch
cast in a sand mold with one percent mold dilation having a side riser with a height per
diameter ratio of 1.5 without any sleeves and with gating through it would have a
calculated diameter of around 3.786 inches, a height of 5.679 inches and a cubic shaped
neck with 2.114 inches on all sides. Figure 4.11 shows the calculated dimensions.

Figure 4.10: The final calculation of the Riser Design Wizard

It could be seen in Figure 4.11 that the total height of the riser divided by the
diameter is more than 1.5 due to the calculations behind the program. If a user would
like to have a side riser with a true height per diameter ratio of 1.5, the user must set the
riser height per diameter ratio at 1.0 instead.

98
www.Iran-mavad.com

 D = 3.786

Riser

H = 5.679
Casting

Casting
DN Riser D
DN
D/2 D/2

LN LN

LN and DN = 2.114

Figure 4.11: Calculated dimensions of the riser from the Riser Design Wizard.

The VDG Iron Properties Calculators Riser Design Wizard was designed
especially for designing risers for cast irons. Designing risers for gray iron and ductile
iron castings involves one primary consideration above all else, which is control of
expansion pressure (Finite Solutions Inc., 2005). This means allowing a riser neck to
remain open long enough to feed liquid metal into the casting but ensuring it would
freeze in time to pressure the casting during expansion so the liquid metal would not flow
back into the riser causing the formation of shrinkage porosity. As could be seen from
Figure 4.10, the riser has the highest modulus, then the casting, and the neck has the least
modulus so it may freeze first.
Only one riser should be used for each feed zone or feeding area within a casting
(Finite Solutions Inc., 2005). Feed zones can be visualized by plotting the calculated
neck modulus in the Calculate and Display Casting Modulus option in the Riser Design
Wizard from the Tools menu. This neck modulus can also be called the transfer
modulus. If more than one riser is used for a single feed zone, only one riser would pipe
and the other riser would create a thermal hot spot which some shrinkage porosity will
likely be able to form underneath that riser. Figure 4.12 shows the single feed zone

99
www.Iran-mavad.com

which was plotted with the neck modulus. Also, for gray and ductile iron castings, the
tops of the risers should be above the highest point of the casting by at least the minimum
section thickness of the casting (Finite Solutions Inc., 2005).

Figure 4.12: Modulus plot showing one feeding area plotted with the neck modulus,
0.5285503 inch.

100
www.Iran-mavad.com

4.3 Gating Design and Analysis
4.3.1 Analysis of the Gating Design Wizard
To start the Gating Design wizard, first double-click the Simulation tab in the
Project Tree. The Simulation Status window would appear. Next, close the window.
This would load the information from the simulation results to the design wizard program.
The Gating Design Wizard can be accessed from the Simulation tab in the menu. The
Gating Design Wizard window would appear and would let the user choose between
designing a horizontal gating system or vertical gating system. In this case, a horizontal
gating system is selected for the casting.
Next, the wizard program would let the user select the alloy sensitivity, the
pouring weight and the critical section thickness to calculate the optimal fill time for the
casting, as shown in Figure 4.13. The alloy sensitivity slider bar refers to the sensitivity
of alloys to form oxides during the pouring process. Alloys which are less sensitive (low
value of this factor) may be poured more quickly. Alloys which are more sensitive to
oxide formation should be poured more slowly to avoid excessive turbulence which may
expose more of the metal surface to oxygen, thus forming and entraining the oxides as
inclusions in the finished casting (Finite Solutions Inc., 2005).
From Figure 4.13, the pour weight for a horizontal gating system would include
the casting along with rigging; the riser and gating system. The pour weight for a vertical
gating system would not include the rigging weight. The user may estimate this weight
from past experience. The critical section thickness is the minimum thickness in the
casting that is most likely to misrun. The calculated fill time or optimal fill time is
calculated with the formula shown in equation (4.7), but the user may also manually enter
the fill time in the case the fill time is already known or estimated by past experience.
The formula used for calculating the optimal fill time was found to neither consider the
type nor the density of the casting material.

t = 2.26 x 3W x 3T x S (4.7)
Where,
t = Optimal filling time or fill time, second
W = Weight of casting, pound

101
www.Iran-mavad.com

 T = Critical section thickness, inch
S = Alloy sensitivity number, no unit

Figure 4.13: Calculating the optimal fill time for the casting.

It was found that the material density used for calculating the volume would
change whenever the material density information in the Casting tab in the Materials List
window was changed. It was assumed that the material density in the Casting tab in the
Materials List window would be automatically transferred to the Gating Design Wizard.
Next, the wizard program would let the user select a sprue type between round
tapered, round straight and square tapered sprue shape. Each sprue type would have
different efficiencies accounting from friction losses associated with flowing through
each sprue shape. It was found that a round tapered shape sprue, a round straight shape

102
www.Iran-mavad.com

sprue and a square tapered shape sprue would have efficiencies of around 88.2, 47.0 and
74.0 percent respectively. The flow areas calculated would be increased by a factor of
the inverse of the efficiency of the sprue type to account for the energy losses associated
with flowing through it, as shown in equation (4.11). In this case, a round tapered sprue
is chosen as the sprue type as it is the most common type used. Then the user may
choose a gating type between top gating, bottom gating and parting line gating. Figure
4.14, 4.15 and 4.16 shows the appearance of the window when top gating, bottom gating
and parting line gating are selected respectively.

Figure 4.14: Gate Position window for top gating system.

103
www.Iran-mavad.com

Figure 4.15: Gate Position window for bottom gating system.

Figure 4.16: Gate Position window for parting line gating system.

104
www.Iran-mavad.com

 From Figures 4.14, 4.15 and 4.16, the user must enter the height of liquid metal in
the pouring basin from the bottom of the sprue during filling (h) for all three types of
gating, the total height of the casting (c) for the bottom and parting line gating, and the
height of the casting from the parting line (p) for parting line gating. The depth of the
metal in the basin is basically the depth of liquid metal when the metal is poured into the
pouring basin, so it may be equal to or less than the depth of the basin. This pouring
basin is used for absorbing the initial surge of liquid and minimizes splashing. The
effective sprue height (ESH) is a value calculated by formulas shown in equations (4.8),
(4.9) and (4.10) for top, bottom and parting line gating respectively. In the case that
liquid metal is poured directly into the sprue and not into a pouring basin, the additional
height of the ladle above the top of the mold should be added to the calculated effective
sprue height (ESH).

ESH =h (4.8)
ESH = h (c/2) (4.9)
ESH = h [(0.5/c) x p2] (4.10)
Where,
ESH = Effective sprue height, inch
h = Height of liquid metal in pouring basin from bottom of sprue, inch
c = Total height of the casting, inch
p = Height of the casting from the middle of the gate, inch

Next, the gating ratio, which is the ratio of the area of flow at the sprue, the runner
and the gates needs to be defined from the user. For example, a ratio of 4:8:3 is used
with pressurized ductile iron casting systems in sand molds, and a ratio of 1:1.3:1.1 is
used with pressurized gray iron castings (Heine, 1995). The choke is defined as the
location within the gating system of the minimum cross-sectional area. For example, the
4:8:3 ratio would have the choke at the gate and the 1:1.3:1.1 ratio would have the choke
at the sprue exit. Last of all, the user must fill in the number of runners and gates.

105
www.Iran-mavad.com

 Next, the window would show the calculated sprue data. This window is shown
in Figure 4.17 and the formulas used for calculating these values are shown in equations
(4.11) through (4.15).

Figure 4.17: Calculated sprue data window, continued from Figure 4.15.

Ac = (1/Effsprue) x [(W ) {t x (2 x g x ESH)}] (4.11)

Where,
Ac = Choke area, square inch
Effsprue = Efficiency of sprue shape due to friction loss, percent
W = Weight of casting, pound
= Density of casting, pound per cubic inch
t = Fill time, second
g = Acceleration due to gravity, 32.2 feet per second square or 386.4 inches
per second square
ESH = Effective sprue height, inch

106
www.Iran-mavad.com

 V =W (4.12)
Where,
V = Volume of casting, cubic inch
W = Weight of casting, pound
= Density of casting, pound per cubic inch

ABS = Ac x RAS
AR = Ac x RAR
AG = Ac x RAG (4.13)
Where,
Ac = Area of calculated choke, square inch
ABS = Area at bottom of sprue, square inch
AR = Total runner area, square inch
AG = Required gate area, square inch
From the gating ratio,
RAS = Ratio of the area of flow at the sprue
RAR = Ratio of the area of flow at the runner
RAG = Ratio of the area of flow at the gates

ATS DB = ABS ESH (4.14)

Where,
ATS = Area at top of sprue, square inch
DB = Depth of metal in basin, inch
ABS = Area at bottom of sprue, square inch
ESH = Effective sprue height, inch

V t = Ac x c = ABS x BS = ATS x TS (4.15)

Where,
V = Volume of casting
t = Fill time, second
Ac = Area of calculated choke, square inch

107
www.Iran-mavad.com

 c = Velocity of liquid metal through choke area, inch per second
ABS = Area at bottom of sprue, square inch
BS = Velocity of liquid metal through bottom of sprue, inch per second
ATS = Area at top of sprue, square inch
TS = Velocity of liquid metal through top of sprue, inch per second

From equation (4.13), the choke area should equal to the lowest section of the
gating ratio, but the calculated section areas are multiplications of the choke area by the
ratio value. The method for calculating for the gating sections would be wrong if none of
the three section ratio values are equal to one, for example, 1:1.3:1.1, etc. If the gating
ratio is 4:8:3, the choke area would be at the ingate, but its area would be three times the
area of the calculated choke area and the velocity of the liquid metal passing it would be
only one third of the calculated choke velocity. The program also does not consider if the
velocity of the liquid metal would be too high or not. The recommended limit of velocity
of liquid metal is approximately half a meter per second for aluminum (Ravi, 2005) and
also for most metals (Campbell, 2004), and one meter per second for iron (Ravi, 2005).
Next, the user may input the required number of runners, the friction loss factor
and option of flow distribution. The friction loss factor will default to 5 percent when the
program is first run, but may be modified by the user. The user may select the flow
distribution to be either to equalize the flow through the gates, meaning all gates have
equal flow, or equalize the flow in the runners, meaning all runners have equal flow
regardless of the number of gates they feed. The more common practice is to equalize
the flow through the gates (Finite Solutions Inc., 2005).
For equalizing the flow through the gates, in the case of multiple runners with one
ingate per runner, the cross-sectional area of each runner would be the calculated total
runner area (AR) divided by the number of runners. The required gate area would be the
calculated required gate area (AG) divided by the number of runners.
For runners which feed multiple gates, in order to equalize flow in the gates, it is
normal to step down the runner, which is reducing the cross-sectional area of the
runner after reaching each gate by the area of the gate to try to equalize the flow through

108
www.Iran-mavad.com

all of the gates fed by the runner. The friction loss factor would be considered in
calculating for each runner and ingate section to compensate for friction loss.
The cross-sectional area of the runner before the first ingate is the runner area of a
runner with one ingate multiplied by the number of ingates for that runner. The first
ingate would have a required gate area of the calculated required gate area (AG) divided
by the number of runners. The runner area after passing an ingate would equal to the
runner area less the ingate area and added with the friction loss factor. This step down
calculation is summarized in Figure 4.18 and equation (4.16). The following window
after the window in Figure 4.17 is shown in Figure 4.19, which is the runner and gate
design window.

AG1 AG2 AG3 AG4

AR4
AR2 AR3
AR1

Figure 4.18: The step down calculation of the runner.

AR2 = (AR1 AG1) x (1 + FLF)

AG2 = AG1 x (1 + FLF)
AR3 = (AR2 AG2) x (1 + FLF)
AG3 = AG2 x (1 + FLF)
AR4 = (AR3 AG3) x (1 + FLF)
AG4 = AG3 x (1 + FLF) (4.16)
Where,
FLF = Friction loss factor, percent
AR1 = Cross-sectional area of runner in section number 1, square inch
AR2 = Cross-sectional area of runner in section number 2, square inch
AR3 = Cross-sectional area of runner in section number 3, square inch

109
www.Iran-mavad.com

 AR4 = Cross-sectional area of runner in section number 4, square inch
AG1 = Cross-sectional area of gate number 1, square inch
AG2 = Cross-sectional area of gate number 2, square inch
AG3 = Cross-sectional area of gate number 3, square inch
AG4 = Cross-sectional area of gate number 4, square inch

Figure 4.19: The runner and gate design window.

In Figure 4.19, it could be seen that there are two calculator buttons which can be
used by the user to calculate a rectangular shaped runner or gate area with the same areas
as the runner area to feed this gate and required gate area respectively. Gating should be
designed to freeze relatively quickly after the liquid metal has filled the mold cavity. The
cross-sectional dimensions of the gate is recommended to have a width per thickness
ratio of 5:1 to ensure relatively quick freezing so that expansion pressure can be
contained (Finite Solutions Inc., 2005). Equation (4.17) shows the formula to calculate
the thickness of the gate.

110
www.Iran-mavad.com

 t = (A/5) (4.17)
Where,
t = Thickness of rectangular gate with width per thickness ratio of 5:1, inch
A = Area of gate, square inch

4.3.2 The Results of the Gating Design

The weight that shall be entered into the Gating Design Wizard for horizontal
gating system must include the weight of the rigging of the casting design. The weight of
the calculated riser and its neck can be calculated as shown in equations (4.18) through
(4.22) shown below:

Volume of cylindrical shape = ( x 3.7862 4) x 5.679

= 63.933 cubic inches (4.18)

Volume of hemispheric shape = [(4/3) x x (3.786/2)3] 2

= 14.207 cubic inches (4.19)

Volume of neck = 2.1143

= 9.447 cubic inches (4.20)

The total volume = 63.933 + 14.207 + 9.447

= 87.587 cubic inches (4.21)

The total weight = 87.587 cubic inches x 0.25926 pound per cubic inch
= 22.708 pounds (4.22)

From the casting weight being 70.434 pounds and the riser with its neck weight
being 22.708 pounds, the total weight would equal to 93.142 pounds. But this weight
does not include the sprue and gating weight. So, the approximated weight that seems
appropriate to be entered into this calculation window may be around 100 pounds.

111
www.Iran-mavad.com

 In the fill time calculation window, the pour weight was entered as 100 pounds
and the critical section thickness was entered as 1.5 inches. Clicking on the Calculation
Fill Time button, the calculated fill time is 24 seconds, as shown in Figure 4.20.

Figure 4.20: Calculating the fill time.

Next, a round tapered sprue with parting line gating was selected in the gate
position window. Since the riser height from the middle of the casting was calculated to
be 5.679 inches, it seems quite appropriate to have the variable h equal to 7 inches, so
the riser may be a hidden riser with more than an inch of mold material covering on top
of it. The total height of the casting or variable c is 4 inches and the height of the
casting measured from the center of the gating or variable p is 2 inches. The calculated
effective sprue height became 6.5. The depth of metal in the basin was approximated to
be 1.8 inches for a pouring basin with a depth of 2 inches. The gating ratio was selected
to be 1:1.3:1.1 as should be used for a pressurized gray iron casting (Heine, 1995). The
number of runners and total number of gates were both set to be equal to one. The
resulting Gate Position window is shown in Figure 4.21. The resulting sprue data and
runner and gate design windows are shown in Figure 4.22 and Figure 4.23 respectively.

112
www.Iran-mavad.com

Figure 4.21: The resulting Gate Position window.

Figure 4.22: The resulting sprue data window.

113
www.Iran-mavad.com

Figure 4.23: The resulting runner and gate design window.

4.4 The Resulting Dimensions for the Simple Casting Design

The calculated dimensions for this simple casting design are for a casting design
which liquid metal would first enter the mold from the pouring basin down the sprue into
the sprue well, into the ingate, into the riser, into the neck and into the casting
respectively. The riser was located next to the sprue instead of being located on the
opposite end because it was recommended in the SOLIDCast manual (Finite Solutions
Inc., 2005) that gating through the riser would increase the proportion of liquid metal that
would feed the casting. The program did not define where the neck of the riser shall be
placed, so, it was placed on the width side of the casting, which is similar to an example
in the SOLIDCast training material (Finite Solutions Inc., 2005). Since the liquid metal
will directly enter the riser from the sprue well, a runner is not needed. Figure 4.24
shows the sequence of the liquid metal as it is poured from the ladle into the mold.

114
www.Iran-mavad.com

Figure 4.24: Sketch of the sequence of how liquid metal poured from the ladle enters the
mold.

Almost all the required dimensions can be calculated by the Riser Design Wizard
and the Gating Design Wizard, but there are some dimensions which need to be
determined by the user. The following sections show the dimensions calculated and the
dimensions determined by the user.

4.4.1 Dimensions from Riser Design Wizard

1) Riser Dimensions
A side riser having a cylindrical shape with a diameter of 3.786 inches and a
height of 5.679 inches with a hemispheric bottom with the same diameter as the
cylindrical shape.

115
www.Iran-mavad.com

2) Neck Dimensions
A cubic shaped neck with 2.114 inches on each side. But since the surface
connecting the riser is not flat, 2.114 inches shall be the length of the neck measured from
the casting surface to the closest position of the riser, assuming the modulus would not
have significant change. The centerline of the neck would be placed at the centerline of
the castings thickness.

4.4.2 Dimensions from Gating Design Wizard

1) Pouring Basin Dimensions
The depth of the metal in the basin during pouring is 1.8 inches assuming that the
pouring basin is 2 inches deep. The width and length of the pouring basin must be
determined by the user. The width and length of the pouring basin was set at 2 and 3
inches respectively. It was aligned with the length on the X-axis to minimize the total
mold size. A rectangular box shape with dimensions 3 by 2 by 1.8 inches was assigned
as fill material which would represent a full pouring basin at the start of simulation.
The term fill material in SOLIDCast is basically the same kind of material as the
casting material but would be defined in the simulation as the position where the liquid
metal enters the mold. In this case, the pouring basin made from fill material would
signify the simulation would start with a full pouring basin.

2) Sprue Dimensions
The calculated diameter for the top of the sprue is 0.778 inch and the calculated
diameter for the bottom of the sprue is 0.565 inch. The height of the sprue is the height
of liquid metal in the pouring basin from the bottom of the sprue during filling less the
depth of metal in the pouring basin, which is 5.2 inches (7 inch 1.8 inch).

3) Sprue Well
The dimensions of the sprue well could not be found from the Gating Design
Wizard, so, it must be calculated by the user. The recommended shape of a sprue well is
cylindrical with a diameter twice the diameter of the sprue exit and a depth of 1.5 times
(Ravi, 2005) or twice (Campbell, 2004) the depth of the runner. In this particular case,

116
www.Iran-mavad.com

the runner was neglected, so the ingate becomes the runner. The calculated diameter and
depth of this sprue well should be 1.13 inch (2 x 0.565 inch) and 0.353 inch (1.5 x 0.235
inch) respectively, but these dimensions do not seem appropriate for this particular
casting as the calculated diameter of the sprue well is smaller than the width of the ingate.
So, the diameter of the sprue well was adjusted to match the width of the ingate instead,
which is 1.174 inch. For the depth of the sprue well, it seems acceptable to use the
calculated runner area for calculation. If the runner would have had a square shaped
cross-sectional, each side would have measured 0.571 inch (0.326 square inch1/2). So,
the depth of the sprue well may be calculated as 1.5 times the calculated length of one
side of calculated runner area, which is 0.856 inch (1.5 x 0.571).

4) Ingate to Riser
The calculated gate connection area has a thickness of 0.235 inch and a width of
1.174 inch. The length of the gate from the sprue well to the riser is unknown and must
be determined by the user. Normally, the ingate should have a length of at least the width
or thickness whichever side is longer. In this case, the length of the ingate is equal to the
calculated width.

4.5 Simulation Results and Analysis of Simple Casting Design

After drawing the casting model with the calculated dimensions, both from the
Riser Design Wizard and Gating Design Wizard programs and the user-defined
dimensions, the finished casting model is shown in Figure 4.25. The pouring basin was
made from fill material and the risering and gating system were made from riser
material. Fill material is made of casting material indicating the position where liquid
metal enters the mold. Riser material indicates shapes made of casting material but not
part of the casting.

117
www.Iran-mavad.com

Figure 4.25: Casting model in SOLIDCast using all calculated and user-defined
dimensions.

Next, the parameters in the Material List window must be set to represent the
casting model. The parameter values in the Casting tab are similar to the settings in the
Case Study in Chapter 5, except for the fill time which was calculated to be 24 seconds,
as shown in Figure 4.26. In the Mold tab, only silica sand was selected for the mold
material to be used in the model. The Curves tab was set to represent this particular
casting model using values from the VDG Iron Properties Calculator which was used
earlier. The values from the calculator were 77.48 percent of shrinkage time in percent of
solidification and 3.72 percent contraction. So, the shrinkage curve was set to start
shrinking from pouring down to 4.7 percent contraction (3.7 percent contraction added
one percent due to mold dilation) at 77 percent of shrinkage time in percent of
solidification, then start expanding with a rate of 0.5 percent expansion for every 10
percent of shrinkage time in percent of solidification. The critical fraction solid point and
Niyama point were set at 82 and 87 percent of shrinkage time in percent of solidification
respectively. The Curves tab setting is shown in Figure 4.27.

118
www.Iran-mavad.com

Figure 4.26: The Casting tab set with values representing the casting model.

Figure 4.27: The Curves tab set with values representing the casting model.

119
www.Iran-mavad.com

 The thinnest section of the casting design is the ingate, which is 0.235 inch. There
should be at least 5 nodes representing a casting section, so, the primary selected node
size is 0.047 inch (0.235 5 = 0.047). But due to the limitations of the program, the
number of nodes could not exceed the recommended maximum number of nodes, so, the
node size achieved was actually 0.0546 inch. But even though the meshed model was
achieved with a node size of 0.0546 inch, the simulation failed twice, so, the final node
size achieved in the simulation was 0.078 inch, which gives 3 nodes representing the
thinnest section of the casting. Figure 4.28 shows initial mesh creating window and
Figure 4.29 shows the calculated weights of the materials in the model after final
meshing. The total weight of the casting material is 97.713 pounds from 70.601 pounds
of the casting, 24.273 pounds of the riser material and 2.839 pounds of the fill
material, which is very close to the estimated 100 pounds of casting material used in the
calculation of the optimal fill time or fill time in the Gating Design Wizard.

Figure 4.28: The mesh creating window.

120
www.Iran-mavad.com

Figure 4.29: The calculated weights of the materials in the model after meshing.

The simulation output criterions that shall be used for analyzing the simulation
results for this casting model are the Critical Fraction Solid Time, Material Density
Function, FCC Criterion and Hot Spot Criterions. It was found in the Critical Fraction
Solid Time plots that the casting started reaching the critical fraction solid temperature at
0.469 minute after the mold is completely filled or pouring. The sprue and the ingate
dropped lower than the critical fraction solid temperature at around 2 minutes after
pouring. The casting then slowly solidified towards the riser leaving some isolated spots
in the inner corners of the far side of the casting across from the sprue and riser. Figure
4.30 shows the Critical Fraction Solid Time plots plotted at 10, 10.5, 11, 11.5, 12 and
12.5 minutes after pouring. Figure 4.31 shows the cut-plane plots of the Critical Fraction
Solid Time criterion. It could be seen that the thermal center of the casting is in the riser.

121
www.Iran-mavad.com

Figure 4.30: Critical Fraction Solid Time plots plotted at 10, 10.5, 11, 11.5, 12 and 12.5
minutes after pouring.

The Material Density Function plots show no signs of macroporosity in the

casting at all, as shown in Figure 4.32, but the FCC Criterion plots show possible
microporosities or centerline shrinkage in inner corners of the far side of the casting
across from the sprue and riser, the centerline of the far side of the casting and the
centerline of the sides of the castings, as shown in Figure 4.33. The shrinkages
highlighted in the FCC Criterion plots possibly occurred from local hot spots as the
highlighted areas are very similar to the Hot Spot Criterions plots shown in Figures 4.34
and 4.35, especially the Hot Spot Criterion calculated from Critical Fraction Solid Time
shown in Figure 4.34.

122
www.Iran-mavad.com

Figure 4.31: Cut-plane plots of the Critical Fraction Solid Time criterion.

Figure 4.32: Material Density Function plots plotted at 0.995.

123
www.Iran-mavad.com

Figure 4.33: FCC Criterion plots.

Figure 4.34: Hot Spot Criterion calculated from Critical Fraction Solid Time.

124
www.Iran-mavad.com

Figure 4.35: Hot Spot Criterion calculated from Solidification Time.

4.6 Conclusions
The Riser Design Wizard and the Gating Design Wizard programs are both very
useful in designing the initial riser and gating dimensions for a casting, but they do not
always provide a flawless casting design, for example, there may be some predicted
macroporosity or microporosity problems. The user must use his or her own judgment in
improving the casting design, for example, by changing the riser type or its position,
redesigning the gating alignment or gating type, adding chills or altering the casting
design to not have sharp corners, etc. So, instead of a user having to calculate for the
riser and gating sizes by hand for an initial riser and gating design, these two built-in
programs can assist the user in designing the initial riser and gating systems. They are
also useful in the cases that the user does not know the formulas to calculate for
appropriate riser and gating systems dimensions. Not all dimensions for designing the
riser and gating systems are given by the program and must be defined by the user, such
as the sprue well dimensions, the appropriate sprue height, the dimensions of the pouring
basin, the length of the ingate, etc. Positions of laying down the risering or gating system

125
www.Iran-mavad.com

are also not provided by the program. The user must decide on the layout of the casting
and rigging system with his or her knowledge and past experience.
The method of calculating for the gating section areas from the choke area is
wrong. The gating section areas calculated in the program are multiplications of the
gating ratio values by the calculated choke area. In cases which none of the gating ratio
values are equal to one, the calculated section areas would become wrong. The program
also does not consider if the velocity of the liquid metal would be too high or not. The
recommended limit of velocity of liquid metal is approximately half a meter per second
for aluminum (Ravi, 2005) and also for most metals (Campbell, 2004), and one meter per
second for iron (Ravi, 2005).

126
www.Iran-mavad.com

 Chapter 5
Using SOLIDCast: Case Study
A case study was conducted to verify the use of SOLIDCast in an industry casting
environment. The selected case study was a locomotive piston design made from gray
iron by the HK Engine Components Casting Division, HK Engine Components, LLC.
(HKEC).
The tests conducted in this case study used SOLIDCast to find potential defects in
the casting model with different criterions and compare them with the real casting design
and to find possible outcomes to modifications attempted to improve the existing casting
design. The Riser Design Wizard and Gating Design Wizard programs in SOLIDCast
were also used to try to improve the existing casting design and develop riser and gating
systems for the casting from scratch and compare the calculated dimensions and resulting
simulations with the existing casting design.

5.1 Company Profile

HK Engine Components Casting Division, HK Engine Components, LLC.
(HKEC) is located at Weston, West Virginia, founded in 1989. Casting metals produced
include ASTM A 48 and 278, Class 30 through 50 gray iron and ASTM A 536 and 395
ductile iron (www.hkec.com).

5.2 Casting Design

The casting design studied in this case study is a locomotive piston made from
ASTM A48-76 Class 40 gray iron. Figure 5.1 shows a figure of the actual finished piston
with machined ring grooves, a figure of a cut open piston from a similar model to the
case study and a figure of a cut open view modeled with SOLIDCast.

127
www.Iran-mavad.com

a) Machined piston b) Sectioned piston c) Model of section casting
Figure 5.1: Locomotive piston produced by HKEC.

5.2.1 Pattern and Mold

Currently, the casting company already has a gating and riser design and is in use
for producing this particular locomotive piston casting. The figures of the patterns for the
cope and drag are shown in Figure 5.2 and the finished cope and drag molds are shown in
Figure 5.3.

Figure 5.2: The cope pattern (left) and drag pattern (right).

128
www.Iran-mavad.com

Figure 5.3: The cope mold (left) and drag mold (right).

The cope and drag are flow coated with alcohol based graphite wash before
assembly. The blind risers cavities in the cope would be drilled up to the surface of the
cope with a diameter around half an inch so the air inside the risers would be able to
escape.

5.2.2 Core System

The core system that is used in the casting is actually two cores assembled
together with a glue paste. The assembled core is shown in Figure 5.4. From Figure 5.4,
the upper core part is made from chromite sand and the lower core part is made from
silica sand. They are glued to together and sprayed by water base graphite wash, then the
upper core part will be painted with terillium wash which would help chill the casting and
facilitate the formation of white cast iron in that particular area. The core would be
torched to dry the coating and then carefully aligned and glued up-side-down onto the
cope.

129
www.Iran-mavad.com

Figure 5.4: Core system.

5.2.3 Assembled Mold

After gluing the assembled core onto the cope, the cope will be placed on top of
the drag and then the molten metal will be poured into the mold through the sprue. The
two top risers are open risers, but were covered with magnesium based exothermic tops
to keep the risers hot to feed the casting for an extended period of time. Figure 5.5 shows
an assembled mold filled with liquid metal and top risers covered with the exothermic
insulation. The exothermic insulation would be approximately 1.25 inches thick.

Figure 5.5: Assembled mold with liquid metal and magnesium based insulators covering
top risers.

130
www.Iran-mavad.com

5.3 Dimensions of Mold Cavity
The approximate dimensions of the mold cavity measured from the patterns and
the mold are presented in Figures 5.6 and 5.7.
4 1/4

4 1/2

3/4 2

Contact Pad 2 x x

3 1/2

10 1/4

Cope

Drag 3/4

2
1
3 1/4

Contact Pad 1 x x
9 1/2

Figure 5.6: Dimensions of top and side risers of casting.

131
www.Iran-mavad.com

 3 1/2
3 1/4

1
1 1/2

1 1/4

9 1/2

1
1 1/4
1/2

1 1/2

1
3/4 4 3
5/8 deep
3/8 deep

1
8 1
23 1/2
28 1/2
3 1/2
5 1/2
2 1/8

9 1/2

1 3/4 deep (7/8 wide) 2

2 1/2 23 1/2 2 1/2

Figure 5.7: Dimensions of the mold cavity and gating system.

132
www.Iran-mavad.com

 In the upper portion of Figure 5.7, the white circles represent the casting cavities
and riser cavities which appear in both the cope and drag molds, the gray shapes with
bold lines would only appear in the drag while the transparent gray shapes with dotted
lines only appear in the cope. The lower figure shows the dimensions of the sprue and its
runner.

5.4 Discarding a Part

The defects that would force the company to scrap a part are found mostly in the
ring groove areas; more likely in the upper thick ring area of the casting than the lower
thick ring area and across the top riser than under it. The company claims that the lower
thick ring area must have cooled down more rapidly than the upper thick ring area due to
the terillium wash applied to the core placed inside the casting cavity. Also, there may be
insufficient feeding from the top riser causing porosities to occur in the thick ring area
across from the riser.
Another type of defect that would force the company to scrap a part is surface
porosity occurring in the thin section of the casting. The porosities are very small and
occasionally disappear between the rough cut and finish cut. The company would not
scrap the part if there are no visual porosities showing after the final machining. There
may be some internal porosity left in this area not discovered, but since they are
considered to not affect the physical properties of the part, and the customer accepts parts
with no cosmetic defects, the company finds no need to X-ray or check for internal
defects.
The cylindrical surface of the casting would be machined off by almost half an
inch, which is near the mid-point of the thin area of the casting. Also, the grooves made
by the milling process would approximately reach the mid-point of the thick ring areas
and the porosities may be found in these areas. The company believes that these
porosities may have occurred from centerline shrinkage.
The rate of scrap is around 10 percent post-machining and the scrap would be re-
melted and used for producing new castings. Since the scrap rate is not high, the defects
detected by SOLIDCast should not be high.

133
www.Iran-mavad.com

5.5 Modeling the Casting Design in SOLIDCast
A user must model a casting in SOLIDCast which best represents the casting
design. Many types of materials were used in this casting design; the gray iron used as
the casting material, chemically bonded silica sand used as the primary mold material, the
core pieces made from silica sand and chromite sand, graphite and terillium washes, and
magnesium based exothermic insulators. Most of the materials in this casting design
could be found in SOLIDCast. The gray iron material in SOLIDCast was modified to
match the properties of the gray iron used as the casting material, silica and chromite
sand were used directly without modifications, the exothermic material in SOLIDCast
was adjusted to represent the magnesium based exothermic insulator material, but the
graphite and terillium washes were not able to be added to the model due to node sizing
limitations.
At first, it seemed possible to increase the thermal conductivity of the silica and
chromite sands so they would represent the more rapid cooling effects from the graphite
and terillium washes, but it was not appropriate to randomly adjust the thermal properties
of a material. After consulting with Mr. David Schmidt of the Finite Solutions Inc., heat
transfer is affected mostly by the properties of the sand, so, the graphite wash and
terillium wash coatings were neglected in the model. If a model must have a wash
applied to it, the node size must be small enough to represent the very thin layer of the
wash and the heat transfer coefficient values between each material must be set in the
Heat Transfer Coefficients tab in the Materials List window.

5.5.1 Modeling of Shapes

The model of the casting consisted of three .stl CAD files; the casting cavity in
the cope and drag, and the two core pieces made from silica and chromite sand.
Combined together, they formed the complete casting cavity. The files were provided by
the Humtown Products of the HK Engine Components, LLC. The casting was made
from the ASTM A48-76 Class 40 gray iron, the upper core piece was made from silica
sand and the bottom core piece was made from chromite sand. Each CAD file was added
to the model with the appropriate materials and properties.

134
www.Iran-mavad.com

Figure 5.8: Half of the model created in SOLIDCast with Plane of Symmetry function.

After adding the casting and its core parts, the gating and riser systems were
added to the model using the dimensions measured at the casting facility, shown in
Figure 5.6 and 5.7, added the drilled holes through the blind risers. But since the
program had a function called the Planes of Symmetry in the Options tab of the Model
menu, this case study used this function to build half of the model. Another advantage to
using this function was that smaller nodes could be used, so, the model would be more
detailed and simulation results could be more accurate. Figure 5.8 shows the half model
created in SOLIDCast. There were slight differences between the model and the casting
design, such as, the side risers were tapered at the parting line for easy removal, the
ingates were tapered and the total height of metal in the sprue was increased. Modeling
tapered risers and ingates in SOLIDCast were difficult and they were considered to not
affect the simulation results. The exothermic top was added on top of the top riser, but
adding the exothermic top to the model, the total height of the mold would increase due
to the meshing method of the program. Extra casting material was added to the top of the
sprue to represent the top of the sprue is open to the atmosphere. The extra casting
material was modeled as fill material which would represent where liquid metal would
enter the mold. Since SOLIDCast is not able to simulate pressure and the extra metal is
considered minimal, it was considered that the added height of metal in the sprue would
not affect the simulation results.

135
www.Iran-mavad.com

5.5.2 Casting Material
The next important step to do in a simulation is to set the primary parameters
properly, which are the parameters in the Materials List window. The attributes of the
casting material in the Casting tab, the mold materials in the Mold tab, the temperature
and shrinkage curves along with the CFS and Niyama lines in the Curves tab and the heat
transfer coefficients in the Heat Transfer Coefficients tab in the Materials List window
were set up. The characteristics of the casting material were researched to find the
appropriate values to set the attribute values in the Casting tab in the Materials List
window and the graph in the Curves tab.
The casting alloy used was ASTM A48-76 Class 40 gray iron which
specifications for the casting alloy for this product in weight percent are:

Total Carbon between 3.00 to 3.50 percent

Silicon between 1.80 to 2.40 percent
Manganese between 0.60 0.90 percent
Sulfur 0.15 percent maximum
Phosphorous 0.15 percent maximum
Nickel 0.50 to 1.20 percent maximum
Chromium 0.50 percent maximum
Molybdenum 0.40 to 0.70 percent
Copper 0.20 to 0.80 percent
Vanadium 0.10 percent maximum

The grades of gray iron in the SOLIDCast database were CI GI 3.5, 4.0 and 4.4
CE. These alloys were preset to approximately represent cast iron alloys with 3.5, 4.0
and 4.4 percent Carbon Equivalent. Normally, the user should start with selecting the
gray iron grade with the Carbon Equivalent closest to the alloy desired. But from tests
conducted, the three grades of gray iron had the same temperature curve if input with the
same weight percent carbon and silicon using the Gray Iron Calculator button in the
Curves tab, so, the temperature curve of the alloy can actually be set with any of the three
alloys. The shrinkage curve, critical fraction solid point and Niyama point can be set by

136
www.Iran-mavad.com

using the VDG Iron Properties Calculator. As for other attributes in the Casting tab of
the Materials List window, the default values of all three gray irons were observed.
Table 5.1 shows the default values of each gray iron alloy in the Casting tab in the
Materials List window.

Table 5.1: Default values of each attribute of each gray iron alloy in the Casting tab.
Thermal Specific Latent
Initial Solidification Freezing Liquidus
Alloy Cond. Heat Density Heat
Temp. Temperature Range Temperature
Name (Btu/hr- (Btu/hr- (lbm/cu.ft.) of Fusion
(F) (F) (F) (F)
ft-F) ft-F) (Btu/lbm)
CI GI 3.5 CE 21 0.11 448 2,600 2,052.667 229.315 2,282 99
CI GI 4.0 CE 25 0.11 466.6 2,500 2,051.965 136.285 2,188 99
CI GI 4.4 CE 33 0.11 448 2,400 2,051.38 75 2,126 99

It can be seen from Table 5.1 that the values of the specific heat and latent heat of
fusion of all three grades of gray iron are the same. It was possible to assume that all
gray irons have approximately the same specific heat and latent heat of fusion.
The appropriate values must be assigned to the parameters in order to achieve the
most accurate simulation results. In the specifications of the alloy in this case study, the
total carbon and silicon percentages are between 3.00 and 3.50 percent and 1.80 and 2.40
percent respectively. The values used to represent the casting alloy were the mean values
of the ranges in the specifications, 3.25 weight percent carbon and 2.1 weight percent
silicon. From the specifications, the maximum weight percent for phosphorus was 0.15
percent and 0.075 weight percent phosphorus was used to represent the casting alloy.
The Carbon Equivalent value of this alloy was calculated as shown in equation
(5.1):

CE = 3.25% + (2.1%/3) + (0.075%/3)

= 3.975% (5.1)

The calculated Carbon Equivalent value is very close to 4.0 percent Carbon
Equivalent, so, the CI GI 4.0 CE material was used to represent the casting alloy. It was
found that the pouring temperature used for this particular casting is between 2,425 and

137
www.Iran-mavad.com

2,575 degrees Fahrenheit. A simulation with the actual dimensions and a pouring
temperature of 2,500 degrees Fahrenheit was conducted as the base case. Simulations
with pouring temperatures of 2,425 and 2,575 degrees Fahrenheit were conducted to test
the differences in the results of different pouring temperatures. Simulations with
modified dimensions were all conducted with a pouring temperature of 2,500 degrees
Fahrenheit.
From a later visit to the company, it was learned that the Carbon Equivalent value
of the casting alloy was reduced to approximately 3.85 percent. The weight percent of
carbon and silicon were reduced but kept the same weight percent value for phosphorus
to achieve this lowered percent Carbon Equivalent value. The new weight percent values
of carbon and silicon were calculated as shown in equations (5.2) through (5.6).

3.975 = 3.25 + 1/3 (2.1 + 0.075)

3.975 0.025 = 3.25 + (2.1/3)
3.95 = 3.25 + (2.1/3) (5.2)

3.85 = C + 1/3 (Si + 0.075)

3.85 0.025 = C + (Si/3)
3.825 = C + (Si/3) (5.3)

C = 3.25 x (3.825/3.95)
= 3.147% (5.4)

Si = 2.1 x (3.825/3.95)
= 2.034% (5.5)

CE = 3.147 + 1/3 (2.034 + 0.075)

= 3.85% (5.6)

Since the percent Carbon Equivalent value had changed, the density of the
material changed as well. The calculations used in calculating the new approximated

138
www.Iran-mavad.com

density of the casting alloy is shown in equation (5.7), assuming the density changes
linearly.

448 + [(466.6 448) x (3.85 3.5)/(4.0 3.5)] = 461.02 pounds per cubic foot (5.7)

For the thermal conductivity, it was calculated in the same manner as equation
(5.7), which is shown in equation (5.8).

21 + [(25 21) x (3.85 3.5)/(4.0 3.5)] = 23.8 Btu per hour-foot- Fahrenheit (5.8)

The filling time of the casting was claimed by the company to be between 20 and
24 seconds per casting mold. The average time was used in every simulation, which was
22 seconds.

5.5.3 Setting Up the Materials List Window

The Materials List Window consists of four input parameter tabs; the Casting,
Mold, Curves and Heat Transfer Coefficient tabs. The Casting tab was set up first by
selecting the CI GI 4.0 CE as the casting alloy, then modifying the attribute values as
calculated in the previous section. The attributes modified in the Casting tab were the
thermal conductivity, density, initial temperature and fill time. Figure 5.9 shows the
modified Casting tab. The modified alloy and its attributes could be saved with a new
name into the database.
Next, the Mold tab in the Materials List window was used for selecting the mold
materials in the model. Silica sand and chromite sand were selected without being
modified. The Exothermic Sleeve (the term for exothermic topping used in
SOLIDCast) material in the database was selected and modified according to Mr. David
Schmidt of the Finite Solutions Inc., to represent the magnesium based exothermic
material used, shown in Figure 5.10.

139
www.Iran-mavad.com

Figure 5.9: The Casting tab.

Figure 5.10: The Mold tab showing the attributes of the magnesium based exothermic
material.

140
www.Iran-mavad.com

 The values set in the Curves tab were the temperature curve, the shrinkage curve,
the critical fraction solid point and the Niyama point. The temperature curve for this
model was set by clicking on the Gray Iron calculator button, and the weight percent
carbon and silicon were entered. The mold wall movement was neglected because it
would only affect the shrinkage curve which would need another method to modify. The
shrinkage curve, critical fraction solid point and Niyama point were set by using the
VDG Iron Properties Calculator. In the VDG Iron Properties calculator, the casting
modulus was found from a prior simulation using the Riser Design Wizard. One and
one-half percent shrinkage was added to the net contraction at the eutectic point to
represent the mold wall movement of a chemically bonded silica sand mold. This
percentage value was recommended by Mr. David Schmidt of the Finite Solutions Inc.
The VDG Iron Properties Calculator is shown in Figure 5.11 and the completed Curves
tab is shown in Figure 5.12. No modification was needed for the Heat Transfer
Coefficients tab.

Figure 5.11: The VDG Iron Properties Calculator.

141
www.Iran-mavad.com

Figure 5.12: The modified Curves tab.

5.6 Verifying the Simulation Results with the Casting

The simulation results must be compared to the casting results to determine if the
predictions from the program are accurate. The pouring temperature for this particular
casting is between 2,425 and 2,575 degrees Fahrenheit, so, three simulations with pouring
temperatures of 2,425, 2,500 and 2,575 degrees Fahrenheit were conducted. The settings
for each pouring temperature were almost exactly the same. The differences between
them were; the pouring temperature and the resulting shrinkage curve, critical fraction
solid point, and Niyama point in the Curves tab from the VDG Iron Properties calculator.
The results from the VDG Iron Properties Calculator of each pouring temperature are
shown in Table 5.2.

142
www.Iran-mavad.com

Table 5.2: The results from the VDG Iron Properties Calculator of each pouring
temperature.
Inputs Outputs
Weight Percent Shrinkage
Casting Time in
Temperature Percent
Modulus Percent of
in Mold (F) Carbon Silicon Phosphorus Contraction
(inch) Solidification
Time
2,425 0.4407624 3.147 2.034 0.075 89.79 3.2
2,500 0.4407624 3.147 2.034 0.075 90 4.1
2,575 0.4407624 3.147 2.034 0.075 90 4.96

The Temperature in Mold used in the VDG Iron Properties Calculator should be
the average temperature of the liquid metal in the mold after the mold is completely filled.
The SOLIDCast manual (Finite Solutions Inc., 2005) recommends using the pouring
temperature less 75 to 100 degrees Fahrenheit as the average temperature. It seems
inappropriate to guess for the average temperature. SOLIDCast should develop a
calculator to find an appropriate temperature in mold for each casting by taking inputs
from the user, such as, the casting weight or volume, the casting modulus, the pouring
temperature, etc. It was decided that the pouring temperature be directly used as the
temperature in mold for all tests.
From Table 5.2, the shrinkage time in percent of solidification time values are
almost the same. As mentioned earlier in Chapter 3, the maximum value for the
shrinkage time in percent of solidification time is 90 percent. This was due to the low
value of the casting modulus. The percent contractions between each pouring
temperature are different due to the difference in the temperatures in the mold (the
pouring temperatures).

5.6.1 Simulation Results of Base Case

The output criterions selected to analyze the casting model were the Critical
Fraction Solid Time, Material Density Function, FCC Criterion and Hot Spot Criterions

143
www.Iran-mavad.com

(Solidification Time and Critical Fraction Solid Time). Even though the company highly
suspected that the defects occured from centerline shrinkage, the Niyama Criterion is
rarely used for predicting cast irons. The best criterion for predicting macroporosity
shrinkage in cast irons is the Material Density Function. For predicting microporosity
shrinkage, the FCC Criterion was used. Microporosities are small holes of rough surface
and usually detected during machining and appearing as a dotted line in the centerlines of
long freezing range alloys and occasionally in thick castings of short freezing range
alloys. Thus, also were called centerline shrinkage (Ravi, 2005). So, the FCC Criterion
may be a good candidate for predicting centerline shrinkage for cast irons. From
previous test runs, the highlighted areas in the Niyama Criterion plots were quite large,
and probably too large because the defects in the casting were much smaller. It was also
found that the highlighted areas of the Hot Spot Criterions and the Niyama Criterion were
almost alike in size and positions, so, the Hot Spot Criterions may be used as an aid for
predicting areas which may have centerline shrinkages.
The simulation model with original dimensions and average pouring temperature,
2,500 degrees Fahrenheit, was chosen to be the base case. It was compared with the
maximum and minimum pouring temperatures and also the modified casting design
models. The base case analyses with the selected output criterions are presented.

1) Critical Fraction Solid Time

This output criterion was used for observing the progression of solidification;
showing the areas which reach their critical fraction solid point at a particular time. The
Solidification Time output criterion is another criterion used for observing solidification
progression, but since the critical fraction solid point and the solidification point are near
to each other on the percent of solidification time scale in the Curves tab, and the plots
showed very little difference in the simulation runs conducted for comparison, the
Critical Fraction Solid Time criterion was the only criterion used for observing the
progression of solidification.
In early runs, it was found that there was an un-natural solidification pattern
showing in the thick ring area in the bottom part of the casting. Instead of the liquid
metal pool between the blind risers separating in the middle from the area between them

144
www.Iran-mavad.com

and solidifying towards each riser, the liquid metal pool seemed to separate at the X-axis
and solidified towards the risers resulting in a longer path and some isolated areas. So,
the model was revolved from the original set up with the runners aligned to the X and Y
axes to the position which the blind risers are aligned to the Y-axis, or 58.211 degree
counterclockwise. The solidification progression appeared more natural with the liquid
metal pool separating in the middle from the area between the blind risers. From the
calculated mesh weights, the weight of the casting was missing by almost nine pounds
(80.645 pounds 72.138 pounds = 8.507 pounds) with no explainable reason when the
whole model was revolved. The weights of the runners and risers were the same. The
casting model along with its core pieces were then revolved back to its original alignment
and the same weights were achieved. It seemed to be that SOLIDCast is axis dependent,
due to it being a Finite Difference Method calculation based casting simulation program.
Figure 5.13 shows the Critical Fraction Solid Time plots showing how the liquid metal in
the bottom part of the casting solidify towards the blind risers.

Figure 5.13: Left: Progress of solidification before model re-alignment. Right: Progress
of solidification after model re-alignment.

The plots of the Critical Fraction Solid Time Criterion of the base case were
plotted from time zero, which is immediately after the mold is filled with molten metal.
It was found that at time zero, some areas in the bottom of the casting have already
reached their critical fraction solid point. The drilled holes through the blind risers were
completely filled, which should be frozen by the time it reaches a height of 4.5 to 6

145
www.Iran-mavad.com

inches from the top of the risers in the actual casting. The settings in the Materials List
window may not adequately represent the casting material. So, the drill holes in the
model were changed to be 5 inches long. It was found that the length of these drill holes
in the model had no effect to the simulation results as a whole. Figure 5.14 shows a
solidified casting placed upside-down showing the front and back with the misrun drill
holes of the single risers and shared riser.

Figure 5.14: A shaken out casting placed upside-down.

The Critical Fraction Solid Time plots showed that the liquid metal in the drill
holes cooled down to its critical fraction solid temperature within a little over one minute,
as shown in Figure 5.16, and the areas surrounding the core and the bottom of the casting
within two minutes after the mold is filled. The knob at the bottom of the casting was the
last area to solidify compared to the areas surrounding it. Figure 5.15 shows the
progression of cooling in the areas around the core pieces and the bottom of the casting
from a top view.

Figure 5.15: Critical Fraction Solid Time criterion plotted at 0, 1.0, 1.5 and 2 minutes
after mold is filled.

146
www.Iran-mavad.com

Figure 5.16: Critical Fraction Solid Time criterion plotted at 0, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5,
4.0, 4.5, 5.0, 5.5 and 6.0 minutes after mold is filled.

147
www.Iran-mavad.com

Figure 5.17: Critical Fraction Solid Time criterion plotted at 6.5, 7.0, 7.5, 8.0, 8.5, 9.0,
9.5 and 10.0 minutes after mold is filled.

The thin wall of the casting was next to solidify, starting from the side across
from the top riser towards the area under it, separating the liquid metal pools in the top
and bottom parts of the casting. The heat from the top riser kept the area under it hotter
for a longer time. All other parts of the casting formed shells with liquid metal inside.
The thin wall solidified completely separating the top and bottom pools by around 3
minutes after the mold was filled.
The whole casting cooled down further until the upper thick ring solidifed
towards the top riser, breaking the ringed pool from the area across the top riser at a little
over 4 minutes after the mold is filled. The bottom pool started to separate into the side
risers at approximately 6 minutes after the mold was filled starting from the side across
from the top riser. It was believed that the side under the top riser solidified slightly later
because the temperature of the liquid metal in this area was heated by the top riser. By
8.5 minutes after the mold was filled, the pool of liquid metal under the top riser had

148
www.Iran-mavad.com

shrunken into the riser completely out of the casting. At around 10 minutes after the
mold was filled, the pools of liquid metal at the bottom ring of the casting had shrunken
into both the side risers. Figures 5.16 and 5.17 are the Critical Fraction Solid Time plots
showing the progression of solidification.

2) Material Density Function

This output criterion was used for finding possible areas where macroporosity
may occur. The plotting values are between 0 and 1; 0 meaning 0 percent metal density
and 1 meaning 100 percent metal density. From the SOLIDCast manual (Finite Solutions
Inc., 2005), the critical plotting value may be around 0.995 or 0.99, which would show
the most likely outcome of a casting. If the plotting value is 0.99, the plot would
highlight all areas with less than 99 percent density. Plotting with a value of 1 would
show all areas which may be prone to microporosity even though there might not be any.
Figure 5.18 shows Material Density Function plots plotted at 1.0, 0.995 and 0.99
respectively from top to bottom.
In all three sets of plots, the common area highlighted in the casting is the bottom
of the casting. The plots plotted at 0.995 and 0.99 were almost exactly the same. For the
plot plotted with 1.0, an area in the ring between the core pieces was highlighted. It
showed a very small possibility of macroporosity occurring in this area.
The company does not experience any macroporosity problems at all. It may be
because the highlighted areas at the bottom of the casting are quite small and may have
quite a high density percentage value, showing that the macroporosity problem may not
be serious and probably does not show on the surface. Also, the bottom of the casting
would only be machined off slightly. This type of problem was never detected.
Figure 5.19 and 5.20 shows the front and top views of the casting before fettling
of knock-off. The top risers and sprue show some signs of piping quite similar to the
plots, but the blind risers do not show piping. Figure 5.21 shows the blind single riser
and the blind shared riser.

149
www.Iran-mavad.com

Figure 5.18: Material Density Function plots plotted at 1.0, 0.995 and 0.99 respectively
from top to bottom.

150
www.Iran-mavad.com

Figure 5.19: Front view of casting before fettling or knock-off.

Figure 5.20: Top view of casting before fettling or knock-off.

151
www.Iran-mavad.com

Figure 5.21: Left; blind single riser. Right; blind shared riser.

A Material Density Function cut-plane plot plotting areas with density values
from 0 to 1 was plotted as shown in Figure 5.22. It shows a more representative
shrinkage figure and the blind risers must have created a shell covering the internal
rough-hole porosities represented by plotted areas having darker shades of color showing
higher density.

Figure 5.22: Material Density Function cut-plane plot plotting areas with density values
from 0 to 1.

152
www.Iran-mavad.com

 The plots show that the drilled holes from the blind risers were all filled up. As
already mentioned in the Critical Fraction Solid Time criterion, the settings in the
Materials List may not represent the casting material as there were no misruns shown in
the plots of the drilled holes.

3) FCC Criterion
The FCC Criterion is a custom criterion provided by SOLIDCast for predicting
areas prone to microporosity. The FCC Calculator calculates the values of each node in
the casting design through a formula, starting from 0 to the value of the node with the
highest calculated value. The recommended plotting value is 40 percent of the maximum
calculated value.
As shown in Figure 5.23, the thick ring area at the lower part of the casting shows
some possible microporosity occurring near the side risers necks close to the outer
parting line of the two core pieces. The thick ring area at the upper part of the casting
also shows some sign of microporosity, mostly in the areas above the side risers and a
very small sign in the area across the top riser.

Figure 5.23: FCC Criterion plots.

The company claims that the porosities found in the castings in the upper ring
grooves are likely to be found in the area opposite the top riser. This does not match the
simulation results. Also, the highlighted areas seem to only be in the areas along the X
and Y axes. No highlighted areas were found in between. This may also be because of

153
www.Iran-mavad.com

how SOLIDCast is axis dependent due to being a Finite Difference Method approached
casting simulation program.

4) Hot Spot Criterions

The Hot Spot Criterion plot can be calculated by two methods, comparing the
solidification time or the critical fraction solid time of a particular node with its neighbors.
A plotted hot spot does not always mean that it will have any problems. It may mean that
a particular area is prone to problems, but in some cases, other criterions should be
considered as well. For example, the hot spot plot indicating hot spots in the thin wall
area plotted all around the casting has a very slight chance of having shrinkage porosities
because of rapid cooling occurred from the section being very thin.
The hot spot plots plotted with the solidification times compared to plots plotted
with the critical fraction solid times seems slightly larger and has additional highlighted
areas. Because the critical fraction solid temperature and the solidification temperature
are considered very close to each other, the plots would be almost identical. The
differences in the plots may have resulted as the calculated ratios slightly increase.
Figure 5.24 shows the hot spot plots plotted with the solidification times and Figure 5.25
shows the hot spot plots plotted with the critical fraction solid times, both plotted at 1.1.

Figure 5.24: Hot Spot Criterion (Solidification Time) plotted at 1.1.

154
www.Iran-mavad.com

Figure 5.25: Hot Spot Criterion (Critical Fraction Solid Time) plotted at 1.1.

Figure 5.26 and 5.27 are cut-plane plots plotted with values between 1.0 and 1.1
and it was found that the thick ring areas both in the upper and lower part of the casting
shows more severity on the inner side. So, the porosities found after milling the ring
grooves are probably because they were milled close to the hot spots.

Figure 5.26: Hot Spot (Solidification Time) cut-planes plot plotted between 1.0 and 1.1.

155
www.Iran-mavad.com

Figure 5.27: Hot Spot (Critical Fraction Solid Time) cut-planes plot plotted between 1.0
and 1.1.

Plotting a Hot Spot Criterion plot with a smaller value showed the areas with
more possible severity. Figure 5.28, 5.29, 5.30 and 5.31 shows hot spot plots plotted with
1.0 and 1.01 showing that the area around the core pieces is the most severe area, then the
inner side of the upper thick ring area, the thin wall and the lower thick ring area
respectively. But the thin wall and thin fins in the cores may be considered as not severe
because of its high cooling rate.

Figure 5.28: Hot Spot (Solidification Time) plotted with 1.0.

156
www.Iran-mavad.com

Figure 5.29: Hot Spot (Solidification Time) plotted with 1.01.

Figure 5.30: Hot Spot (Critical Fraction Solid Time) plotted with 1.0.

Figure 5.31: Hot Spot (Critical Fraction Solid Time) plotted with 1.01.

157
www.Iran-mavad.com

5.6.2 Conclusions of Base Case
The simulation results showed many similarities to the casting, such as the
microporosities or centerline shrinkage predicted by the FCC Criterion were in the same
areas where centerline shrinkage was suspected, Hot Spot Criterions plots showing areas
of potential centerline shrinkage and the Material Density Function plots showing no
macroporosity problems in the machined surfaces. The biggest difference between the
simulation results and the casting was that the simulation did not show any misrun in the
drilled holes. It may be because the temperature and shrinkage curves calculated by the
program do not represent the characteristics of the casting material well enough.

5.7 Pouring Temperature Comparison

The simulation results of the maximum and minimum pouring temperatures were
compared with the base case at the average pouring temperature. All three cases were set
with the same node size and mold wall thickness, so, the weights of the materials
calculated were the same. The three cases having different pouring temperatures resulted
in the difference in the Curves tab, as shown earlier in Table 5.2 (the pouring
temperatures were presented as Temperature in Mold). The values retrieved from the
Simulation Status windows are summarized in Table 5.3.

Table 5.3: Values retrieved from the Simulation Status windows.

Pouring Simulated Max Min
Max Mold Min Mold Time
Temp Time Casting Casting
Temp (F) Temp (F) Steps
(F) (min.) Temp (F) Temp (F)
2,425 19.739 2,057.483 214.122 2,029.466 80 21,575
2,500 21.41 2,055.294 206.452 2,025.805 80 23,400
2,575 23.287 2,058.777 199.093 2,032.052 80 25,450

From Table 5.3, it was seen that the higher the pouring temperature, the longer it
took the casting to solidify completely (simulated time) and the longer it took to finish
the simulation (time steps). The maximum casting temperature was the temperature of
the casting material node which solidified last, so, all three cases have nearly the same

158
www.Iran-mavad.com

temperatures, and are very close to the solidification temperature, 2,058.798 degrees
Fahrenheit from Figure 5.9. The minimum casting temperature of the highest pouring
temperature is lowest because the section of the casting which solidified first had more
time to cool and wait for the last section of the casting to solidify to finish simulating. As
for the maximum mold temperature, the higher the pouring temperature, the higher the
maximum temperature of the mold should be.
The results of the simulations were compared by the Critical Fraction Solid Time,
Material Density Function, FCC Criterion and the Hot Spot Criterions. The Critical
Fraction Solid Time plots of all three cases show the same progressions, but with
different speeds. With a difference of 75 degrees Fahrenheit between each step, it was
found that the step with a higher pouring temperature compared to the adjacent step
would take around less than half a minute later to reach the same progression in
solidification in the early stages, and around one minute when the whole casting
excluding the risers, runners and sprue passes the critical fraction solid temperature.

Table 5.4: Comparing the severity of the problems predicted by each criterion of the three
pouring temperatures.
Material Density
Hot Spot
Pouring Function FCC
Temp Criterion CFS Solidification
1.0 0.995 0.99
Time Time
2,425 1 1 1 1 1 1
2,500 2 2 2 1 1 2
2,575 3 3 3 3 3 3
1 = least severe, 2 = moderate, 3 = most severe

The highlighted areas in the Material Density Function, FCC Criterion and the
Hot Spot Criterions plots of all three cases were almost at the same positions with slightly
different intensities. It seemed possible to compare the severity of the problems predicted
by each criterion by the volume and size of the plots and additional highlighted areas.
The larger the highlighted area or the more highlighted areas, the more severe the

159
www.Iran-mavad.com

problems predicted would be. Table 5.4 shows a summary table of the comparison of the
three cases with 1 meaning least severe, 2 meaning moderate and 3 meaning most severe
compared between the three cases.
It was seen that the higher the pouring temperature, the more possibility of
macroporosity may occur. This could be seen from the calculations from the VDG Iron
Properties Calculator which set up the shrinkage curve in the Curves tab that the higher
the pouring temperature, the more shrinkage would occur. For the FCC Criterion and
Hot Spot Criterions, the plots were almost exactly the same with very little difference in
the level of severity. The levels of severity shown in Table 5.4 for the FCC Criterion and
Hot Spot Criterions among each case were actually very alike and should not be
considered significant.

5.8 Modifying the Casting Design

The casting design was modified to find ways to improve the casting results and
to study some particular outcomes from the modifications. All cases used a pouring
temperature of 2,500 degrees Fahrenheit or the mean temperature.

5.8.1 Remove Top Riser

The top riser was removed from the base case model to observe the consequences
of not having a top riser to feed the thick ring area in the upper part of the casting. The
settings in the Materials List window were set exactly the same as the base case, except
for the fill time which should be changed due to the missing weight. If both the top risers
were removed, the total weight of the casting material would be approximately 45 pounds
less. The fill time used in the base case is 22 seconds and the casting weighing around
300 pounds. Back calculating the optimal fill time calculator in the Gating Design
Wizard from equation (4.7) would result in a critical section thickness of around three-
eights of an inch (3/8 inch), which is approximately the thickness of the ingates of the
casting. Using this critical section thickness, the new calculated fill time from the Gating
Design Wizard for this particular casting design with no top risers is approximately 19
seconds.

160
www.Iran-mavad.com

 The Critical Fraction Solid Time plots showed the upper thick ring area is an
isolated area. It was separate from the bottom pool of liquid metal when the thin wall
solidified. The rest of the casting solidified in the same manner as the base case. The
Material Density Function plots showed quite a high possibility of macroporosity
occurring inside the upper thick ring area from lack of feed metal, as shown in Figure
5.32 and 5.33.

Figure 5.32: Iso-surface plots of the Material Density Function plotted at 0.995.

Figure 5.33: Cut-plane plots of the Material Density Function plotted from 0 to 1.

161
www.Iran-mavad.com

 The FCC Criterion plots showed almost no signs of microporosity. Only some
small highlighted pixels showed, which may be considered insignificant. As for the Hot
Spot Criterions plots, they were very similar to the base case, except that normally there
would be a gap under the top riser, now the upper thick ring area has a full highlighted
ring. This test concluded that there should be a riser to feed the upper thick ring area or
else macroporosity would occur inside the ring area.

5.8.2 Reduce Size of Neck of Top Riser

In this test, the neck connecting the top riser to the casting was reduced to two-
thirds (2/3) of its original width in order to observe if some liquid metal could be trapped
in the casting cavity and solidify in isolation instead of directionally solidifying into the
riser as in the base case. This test was conducted assuming that if graphite precipitation
occurs and causing expansion, it may push the liquid metal back into the riser, and when
the metal shrinks, would cause shrinkage porosities. Actually, SOLIDCast would not be
able to calculate this type of situation as the nodes cannot move, but the user may be able
to see the solidification sequence and predict how the casting would solidify and find out
the best time the neck should freeze off in this type of situation.
The Critical Fraction Solid Time plots showed almost no difference between the
test and the base case. The pool of liquid in the upper thick ring area solidified into the
riser by around less than half a minute faster than the base case. The neck did not freeze
off separating the pool of liquid metal. It may have been because the neck was too short
or not small enough, so the heat from the casting kept it from freezing off.
The Material Density Function and FCC Criterion plots did not show significant
difference from the base case, but the Hot Spot Criterions plots showed that the size of
the hot spot directly under the neck in the casting increased. This may mean that the neck
trapped a small amount of liquid metal retarding the directional solidification towards the
riser even though it did not show in the Critical Fraction Solid Time plots.

5.8.3 Add a Circular Fin Around Outer Surface of Upper Ring Hot Spot
A circular fin with a thickness of 0.08 inch (or actually 0.079 inch due to the node
size restrictions) and a width of 1 inch was added to the upper thick ring area of the

162
www.Iran-mavad.com

casting exactly surrounding the ring shaped hot spot in the upper thick ring area. Even
though adding this fin may not be done in the real casting, the purpose of adding this fin
in the model was to see if it could cool this ring area faster and reduce the potential hot
spots in this area.
From the Critical Fraction Solid Time plots, the fin cooled down lower than the
critical fraction solid temperature before the pouring was finished. The Solidification
Time plots were also observed and it was found that the fin has totally solidified before
the pouring was finished as well, as shown in Figure 5.34.

Figure 5.34: Solidification Time plot plotted at 0 minute.

The Critical Fraction Solid Time plots showed that the fin increased the speed of
solidification and decreased the solidification time in that area by approximately one and
a half minutes. The other parts of the casting solidified at the same speed as the base case.
The Material Density Function plots did not show any difference at all compared to the
base case, but the FCC Criterion plots showed that adding the fin reduced the possibility
of microporosity occurring in this area. The intensity of the highlighted hot spots in the
Hot Spot Criterions plots were the same compared to the base case, but it seemed like
they were pushed towards the inner side of the ring area. Figure 5.35 shows cut-plane
plots of the Hot Spot Criterion calculated by the solidification time.

163
www.Iran-mavad.com

Figure 5.35: Cut-plane plots of Hot Spot (Solidification Time) Criterion.

5.8.4 Add Circular Fins Around Outer and Inner Surface of Upper Ring Hot Spot
In this case, an additional fin was added on the inner surface of the upper thick
ring area apart from the fin surrounding the outer surface added from the previous section.
The thickness and width of this additional fin is the same as the previous fin, only smaller,
as it is attached on the inside.
The Critical Fraction Solid Time plots showed that the upper thick ring area
cooled down faster than with only the outer ring fin by almost half a minute. The
Material Density Function and FCC Criterion plots showed no difference between the
two modifications, but the Hot Spot Criterions plots showed that the ringed hot spot
seemed to be pushed back into the middle and increased in size, as shown in Figure 5.36.
An additional test was conducted along with this test adding iron chills
surrounding the whole upper thick ring area to observe if the hot spots could anyway be
eliminated. It was found that they never disappeared because the method of calculating
the plots is comparing the time (critical fraction solid or solidification time) of a
particular node with its neighbors. So, no matter how rapid the cooling is, if the direction
of solidification is still the same, the hot spot plots shall still be in the same positions.
The hot spot plot does not give an indication of the severity of the defect, as it does not

164
www.Iran-mavad.com

take contraction or expansion into account. But it does give a good indication of what
areas may have problems (Unit 26 of the SOLIDCast manual). So, even if the hot spots
still showed in the plots, there may be no problems in the area at all.

Figure 5.36: Cut-plane plots of Hot Spot (Solidification Time) Criterion.

5.8.5 Add Fin at Outer Surface of Upper Ring Hot Spot Across from Top Riser
In this case, a fin with the same thickness as the previous cases, but with 2 inches
in width from the outer surface of the upper ring hot spot at the point across from the top
riser and tapered down to the mid-point instead of surrounding the whole casting was
added. Although this design may not actually be done in a real casting, but its purpose
was to see if the design could encourage better directional solidification. In the Critical
Fraction Solid Time plots, they showed that the solidification speed at the area across the
top riser was relatively faster than the base case, but after mid-point where there was no
fin, the solidification speed was the same compared to the base case. The Material
Density Function and FCC Criterion plots were at the same levels as the base case.
Figure 5.37 shows an example of the design plotted with the Material Density Function at
0.995. As for the Hot Spot Criterions plots, they were somewhat a mixture between the
base case and the case which the fin surrounded the outer surface of the hot spot. The hot

165
www.Iran-mavad.com

spot was pushed inside in the half across the top riser, and stayed the same on the side of
the top riser.

Figure 5.37: Material Density Function plots plotted at 0.995.

5.8.6 Comparing 3.975 and 3.70 Percent Carbon Equivalent to Base Case
In this test, the weight percent carbon, silicon and phosphorus were set back to the
average percentages of the chemical composition ranges, with 3.975 percent Carbon
Equivalent as was calculated in equation (5.1), and decreased to 3.70 percent Carbon
Equivalent. The weight percent of carbon, silicon and phosphorus used in this test was
3.25, 2.1 and 0.075 percent respectively for the 3.975 percent Carbon Equivalent and
3.024, 1.954 and 0.075 percent respectively for the 3.70 percent Carbon Equivalent. The
calculations for finding the weight percent of carbon and silicon are shown in equations
(5.9) through (5.13).

3.975 = 3.25 + 1/3 (2.1 + 0.075)

3.975 0.025 = 3.25 + (2.1/3)
3.95 = 3.25 + (2.1/3) (5.9)

3.70 = C + 1/3 (Si + 0.075)

3.70 0.025 = C + (Si/3)
3.675 = C + (Si/3) (5.10)

166
www.Iran-mavad.com

 C = 3.25 x (3.675/3.95)
= 3.024% (5.11)

Si = 2.1 x (3.675/3.95)
= 1.954% (5.12)

CE = 3.024 + (1.954/3) + (0.075/3)

= 3.700% (5.13)

Table 5.5: Comparing the results from the VDG Iron Properties Calculator of 3.70, 3.85
and 3.975 percent Carbon Equivalent.
Inputs Outputs
Weight Percent Shrinkage
Casting Time in
Temperature Percent
Modulus Percent of
in Mold (F) Carbon Silicon Phosphorus Contraction
(inch) Solidification
Time
2,500 0.4407624 3.024 1.954 0.075 90 4.21
2,500 0.4407624 3.147 2.034 0.075 90 4.1
2,500 0.4407624 3.25 2.1 0.075 87.4 4.01

Table 5.6: Comparing the results from the Simulation Status window of 3.70, 3.85 and
3.975 percent Carbon Equivalent.

Simulated Max Min

Max Mold Min Mold Time
CE (%) Time Casting Casting
Temp (F) Temp (F) Steps
(min.) Temp (F) Temp (F)
3.70 21.255 2,056.957 330.531 2,059.166 80 23,230
3.85 21.41 2,055.294 206.452 2,025.805 80 23,400
3.975 21.666 2,057.252 208.027 2,027.919 80 24,577

167
www.Iran-mavad.com

 Table 5.5 and 5.6 shows the results from the VDG Iron Properties Calculator and
from the Simulation Status window of 3.70, 3.85 and 3.975 percent Carbon Equivalent
respectively. The calculations do not seem to differ much from each other, except for the
minimum casting temperature and maximum mold temperature of the 3.70 percent
Carbon Equivalent. Actually the casting model built for this particular case had a
reduction of the drill holes lengths down to 5 inches, which is the length of what the
metal in the drill holes should be if misrun. Additional tests were conducted and showed
that the length of the drill hole does not affect the model in a whole. So, the minimum
casting temperature measured was probably from the metal in the drill hole, i.e., the
further away from the riser, the cooler the metal would be. But for the maximum mold
temperature, it normally should not exceed the maximum casting temperature, unless it
was measured from the exothermic top which was slowly cooling down. The clearest
trends which were considered were the percent contraction and the simulated time. It
seems that the higher the percent Carbon Equivalent, the smaller the net contraction, and
that more time was needed for the casting to solidify.
The Critical Fraction Solid Time plots showed no difference in the progression of
solidification, but slightly different in the timing, as a higher percent Carbon Equivalent
took a slightly longer time. The Material Density Function plots were considered to be
the same, but the FCC Criterion plots, although almost identical, showed that the case
with 3.85 percent Carbon Equivalent had the highest chance of having microporosities,
then the 3.70, and the 3.975 percent Carbon Equivalent respectively. As for the Hot Spot
Criterions plots, all three cases seemed to be the same.

5.8.7 Add Hemispheric Bottoms to Side Risers

The single and shared blind risers were added with hemispheric shaped bottoms.
The rectangular shaped mold was slightly enlarged due to the extended bottom resulting
in a slightly larger node size. The Critical Fraction Solid Time plots showed the same
solidification progression compared to the base case. The Material Density Function
plots showed slightly more possible macroporosities occurring in the area between the
core pieces and the bottom of the casting but in the FCC Criterion plots showed almost
no highlighted areas of possible microporosity. As for the Hot Spot Criterions, the hot

168
www.Iran-mavad.com

spot plots calculated by the solidification time seemed the same compared to the base
case but the hot spot plots calculated by the critical fraction solid time showed that the
hot spot under the top riser disappeared. The reason why the hemispheric shaped
bottoms were added was to see if the thermal centers in the risers showing in the hot spot
plots would move down towards the necks or not, which they did not. Actually, in a real
casting, the thermal center should move down because of the level of the liquid metal in
the risers should drop due to feeding, but because the nodes in SOLIDCast cannot move,
the thermal centers do not move as well.

5.8.8 Move Side Risers Down

The single and shared blind risers were moved lower until the middle of the risers
connected to the necks. The purpose was to move the thermal centers in the risers
towards the necks. Normally the thermal center should move down because of the level
of the liquid metal in the risers should drop due to feeding. The Critical Fraction Solid
Time plots showed the same solidification progression compared to the base case. The
Material Density Function plots showed slightly more possible macroporosities occurring
in the area between the core pieces and the bottom of the casting, and when plotted with
1.0, an area in the thin wall under the top riser was highlighted. The FCC Criterion plots
showed almost no highlighted areas of possible microporosity. The Hot Spot Criterions
plots were almost exactly the same as the model with risers with hemispheric shaped
bottoms.

5.8.9 Conclusions of Modifying the Casting Design

SOLIDCast is helpful for users to predict the outcomes of a new casting design.
A user may use SOLIDCast to help predict outcomes and potential problems in deciding
on an option of improving the design of a casting design before casting a real casting. It
would help prevent loss in the investment on trial castings. But the results from
SOLIDCast have limitations due to the software, such as axis alignment, node size
restrictions and node movement and limitations in the physics of metallurgy such as
gravity, head pressure, etc.

169
www.Iran-mavad.com

5.9 Using the Riser and Gating Design Wizard Programs
The Riser and Gating Design Wizard programs are sub-programs in SOLIDCast
which may help a user design the riser and gating systems for a particular casting design.
They may be used to improve an existing casting design or design the riser and gating
system from scratch. The following sections show the calculated dimensions from the
Riser Design Wizard accessed from the Simulation tab and accessed from the VDG Iron
Properties Calculator and the Gating Design Wizard. The calculated dimensions shall be
used for improving and re-designing the casting design.

5.9.1 Riser Design Wizard

There are two Riser Design Wizard programs in SOLIDCast; one can be accessed
from the VDG Iron Properties Calculator and the other can be accessed from the
Simulation tab in the menu. From past observations, the two programs may have
different approaches, but the methods of calculating for the riser dimensions are quite
similar. The Riser Design Wizard program from the Simulation tab shall be used for
improving the currently used casting design, and the Riser Design Wizard program from
the VDG Iron Properties Calculator shall be used for redesigning all the risers and their
necks from scratch.

5.9.1.1 Riser Design Wizard: Simulation Tab

The piston including the core pieces excluding all risers and gating system were
simulated once in order to use this Riser Design Wizard program. The settings in the
Materials List window were set exactly the same as the base case, except for the Curves
tab which uses the new calculated maximum modulus of the casting, from Figure 5.38 to
calculate for the shrinkage curve using the VDG Iron Properties Calculator. The net
shrinkage after adding a 1.5 percent mold wall movement at the eutectic point was 5.4
percent contraction. When entered the program, selecting the Calculate and Display
Casting Modulus option, as shown in Figure 5.38, would let the user plot the suggested
modulus for determining the number of feeding areas. The plots showed two feeding
areas, but the bottom feeding area shows a narrowing area which may be two separate

170
www.Iran-mavad.com

feeding areas. The modulus plot set up window is shown in Figure 5.39 and the resulting
plot is shown in Figure 5.40.

Figure 5.38: Options of Calculate and Display Casting Modulus and Design Risers.

Figure 5.39: Modulus plot set up window.

171
www.Iran-mavad.com

Figure 5.40: The resulting plots from the modulus plot set up window in Figure 5.39.

Next, returning to the Riser Design Wizard again, selecting the Design Risers
option would let the user select a single riser for the casting or a sensitivity level for
designing multiple risers, as shown in Figure 5.38. From observation, the numbers of
risers calculated for a medium sensitivity to the lowest sensitivity was three risers, but as
the level of sensitivity in the sliding bar increased, the calculated numbers of risers were
4, 5 and 6. When plotted the areas which each riser fed, the additional risers were placed
inside the casting or fed very small areas which did not seem practical, so, the sensitivity
level of the sliding bar was set at the middle and resulted in three feeding areas, as shown
in Figure 5.41. Selecting the option Plot feed area would plot the areas of the
calculated feeding areas, as shown in Figure 5.41. These areas were identified from a
calculated modulus value of 0.39, which is shown in Figure 5.42. This value was plotted
by going back to select the option Calculate and Display Casting Modulus in Figure
5.38, and the resulting plots are shown in Figure 5.43.

172
www.Iran-mavad.com

Figure 5.41: Calculated feeding areas.

Figure 5.42: Plotted feeding areas; from left to right, Feed Area 1, Feed Area 2 and Feed
Area 3.

173
www.Iran-mavad.com

Figure 5.43: Plots plotted from feeding area identification modulus value of 0.39 showing
3 feeding areas.

Selecting a feeding area and selecting the option Design Riser in Figure 5.41
would appear a window as shown in Figures 5.44, 5.45, 5.46 from selecting Feed Area 1,
2 and 3 respectively. They would automatically retrieve the modulus of that casting area,
the casting volume or actually the volume of the feed area from the simulation and
calculate the required riser modulus according to the riser per casting modulus ratio,
which default value is 1.2. This value means that the modulus of a riser must be 20
percent more than the casting area it feeds (Finite Solutions Inc., 2005).
The riser could be selected to have no sleeve, an insulting sleeve or an exothermic
sleeve which would have a riser modulus increase factor of 1, 1.25 and 1.33
respectively. The riser modulus increase factor would be multiplied by the calculated
riser modulus and displayed in the Actual Riser Modulus box. The shape of the riser in
this Riser Design Wizard is cylindrical. The modulus of a riser in this case is the volume
of the cylindrical shape divided by the surface area of the whole cylindrical shape. As
could be seen, this Riser Design Wizard does not calculate the neck dimensions.
In the calculator box, four calculator buttons with four calculation methods could
be chosen to calculate the risers dimensions, which could be summarized in Table 5.7.

174
www.Iran-mavad.com

Figure 5.44: Riser calculator for Feed Area 1.

Figure 5.45: Riser calculator for Feed Area 2.

175
www.Iran-mavad.com

Figure 5.46: Riser calculator for Feed Area 3.

Table 5.7: The inputs needed and outputs provided by each calculator button in the
calculator box.
Calculator Inputs Outputs
Height Diameter
1
Required riser modulus Actual riser volume
Diameter Height
2
Required riser modulus Actual riser volume
Height Actual riser modulus
3
Diameter Actual riser volume
Height per diameter ratio Diameter
4 Required riser modulus Height
Actual riser volume

176
www.Iran-mavad.com

 There is a checkbox shown in Figures 5.44, 5.45 and 5.46 which says Use
Wlodawer. Without checking the checkbox, the riser efficiency factor value would be
retrieved from a chart kept in the database of the SOLIDCast program. It would take the
riser per casting modulus ratio and riser type into consideration and produce the riser
efficiency factor value. This is referred as the AFS Method according to the SOLIDCast
manual (Finite Solutions Inc., 2005). But from various simulation runs on this piston
casting, even though the riser per casting modulus ratio was always set at 1.2 and the
same riser types selected, the riser efficiency factor value would change between
simulation runs. It seemed more likely that the riser efficiency factor value would
increase as the modulus of the whole casting increases, as shown in Table 5.8.

Table 5.8: Riser efficiency factor values resulting from different casting modulus values.
Casting Riser:Casting Riser Efficiency
Type of Riser
Modulus Modulus Factor (%)
0.423 1.2 No sleeve, no top 12.25
0.426 1.2 No sleeve, no top 12.45
0.441 1.2 No sleeve, no top 13.64

If the checkbox was checked, the riser efficiency factor value would be retrieved
from the Wlodawer method (Finite Solutions Inc., 2005). The riser efficiency factor
value would be 15 percent for a non-sleeved riser, 32 percent for a riser with an
insulating sleeve and 35 percent for a riser with an exothermic top. Both the methods
would follow equation (5.14). Rearranging the items in equation (5.14), the required
riser volume may be calculated with equation (5.15).

(Vc + Vr) x s = Vr x E (5.14)

Vr = Vc / (E/s 1) (5.15)
Where,
Vc = Volume of feed area, cubic inch
Vr = Required riser volume, cubic inch
E = Riser efficiency factor, percent

177
www.Iran-mavad.com

 s = Shrinkage of alloy at the critical fraction solid point, percent

The procedure this Riser Design Wizard follows may be summarized as follows:
1) Calculating the required riser modulus by multiplying the casting modulus
(modulus of the feeding area) by the riser per casting modulus ratio.
2) Calculating the required riser volume by using the formula in equation (5.15).
3) Enter the input values required for using the desired calculator button.
4) Calculating the actual riser modulus by multiplying riser modulus increase factor
with the riser modulus calculated from the volume and surface area of the
calculated riser dimensions.

The dimensions of the currently used risers were entered into the Riser Design
Wizard. Table 5.9 shows the actual volume of each riser compared to the calculated riser
volume required for each feeding area. For Riser 3, even though the actual riser has a
funnel shaped bottom, the effective volume of the riser would be very close to a
cylindrical riser with the same diameter and height.

Table 5.9: Comparing the actual riser volumes and the calculated riser volume required
for each feeding area.
Riser Dimensions Actual Required Riser Volume
Riser Type of Riser (inch)
Diameter Height
No. Riser Volume AFS Wlodawer
(inch) (inch)
(inch) Method Method
No sleeve,
1 3.5 8 76.969 69.591 49.78
no top
No sleeve,
2 3.5 8 76.969 76.17 54.487
no top
Exothermic
3 4.25 6.5 92.211 22.749 20.924
top
No sleeve,
3 4.25 6.5 92.211 89.806 64.242
no top

178
www.Iran-mavad.com

 It could be seen from Table 5.9 that the actual riser volumes of the currently used
risers are only slightly more than the required riser volumes calculated by the AFS
method considering all risers to have no insulating sleeves and no exothermic tops. For
Riser 3, or the top riser, the currently used riser is actually covered with an exothermic
top, so, it seemed that this riser was initially calculated to not have an exothermic top.
From the calculations, the top riser was over designed.
This Riser Design Wizard considers the mold to have a single casting, so, Riser 2,
the shared riser, may not be large enough to feed two feeding areas of two castings. But
according to the simulation results, there were no potential problems occurring from the
size of the shared riser.

Figure 5.47: Calculated dimensions for Riser 3; the top riser.

It seemed possible that the size of the top riser could be reduced to increase the
yield and use less casting material in casting the piston castings. From Table 5.9, the
riser sizes were very close to the sizes of the calculated risers calculated with the AFS

179
www.Iran-mavad.com

method, so, it seemed appropriate to calculate for a top riser with an exothermic top in the
same manner. Figure 5.47 shows the calculated riser dimensions with a 1.5 height per
diameter ratio for the exothermic top riser. The calculated riser is 2.75 inches by
diameter and 4.125 inches by height.
Even though the neck dimensions cannot be calculated by this Riser Design
Wizard, the neck modulus and the neck dimensions could be calculated by adopting
equations (4.6) and (4.5) respectively.
The length of the necks of Riser 1 and 2 were each 1 inch, which cannot be
considered as short necks, but the length of the neck of Riser 3, which was 0.25 inch, can
be considered as a short neck. So, 0.6 must be multiplied to the normal neck modulus
resulting in a smaller neck. Since this top riser was tapered, both neck dimensions shall
be calculated. The neck dimensions for Riser 1, 2 and 3 was calculated as shown in
equations (5.16) through (5.23).

Neck dimensions for Riser 1 were calculated as;

Mn = Mc x %ST from (4.6)

= 0.423 x 0.90
= 0.4013 inch (5.16)

L = 4 Mn from (4.5)
= 4 x 0.4013
= 1.605 inch (5.17)

Neck dimensions for Riser 2 were calculated as;

Mn = Mc x %ST from (4.6)

= 0.420 x 0.90
= 0.3984 inch (5.18)

180
www.Iran-mavad.com

 L = 4 Mn from (4.5)
= 4 x 0.3984
= 1.594 inch (5.19)

Neck dimensions of a short neck for Riser 3 were calculated as;

Mn = 0.6 x Mc x %ST from (4.6)

= 0.6 x 0.408 x 0.90
= 0.2322 inch (5.20)

L = 4 Mn from (4.5)
= 4 x 0.2322
= 0.929 inch (5.21)

Neck dimensions of a normal neck for Riser 3 were calculated as;

Mn = Mc x %ST from (4.6)

= 0.408 x 0.90
= 0.3871 inch (5.22)

L = 4 Mn from (4.5)
= 4 x 0.3871
= 1.548 inch (5.23)

The connecting surface areas of the necks of the three risers were calculated as
shown in equations (5.24) through (5.27).

Neck area for riser 1 = 1.6052

= 2.576 square inches (5.24)

181
www.Iran-mavad.com

 Neck area for riser 2 = 1.5942
= 2.541 square inches (5.25)

Short-neck neck area for riser 3 = 0.9292

= 0.863 square inches (5.26)

Normal-neck neck area for riser 3 = 1.5482

= 2.397 square inches (5.27)

From the casting design dimensions shown in Figures 5.6 and 5.7, the actual
surface areas of the necks of the three risers connected to the casting were calculated as
shown in equations (5.28) through (5.30). The calculated and actual surface areas of the
necks connecting the three risers to the casting are summarized and shown in Table 5.10.

Neck area of riser 1 = 0.75 x 1.25

= 0.9375 square inches (5.28)

Neck area of riser 2 = 0.75 x 1.25

= 0.9375 square inches (5.29)

Neck area of riser 3 = 0.75 x 2.75

= 2.0625 square inches (5.30)

Table 5.10: The actual and calculated surface areas of the necks connecting the three
risers to the casting.
Surface Areas of the Necks Connected the Casting (square inch)
Riser No. Calculated
Actual
Normal neck Short neck
1 0.9375 2.576 -
2 0.9375 2.571 -
3 2.0625 2.397 0.863

182
www.Iran-mavad.com

 The actual surface areas of the necks connected to the casting area were smaller
than the calculated neck dimensions, except for the short neck condition. Normally, a
small neck is preferred so it has a low modulus and will freeze off before the expansion
in the casting could push the liquid metal back into the riser. Since the original neck
dimensions were working well and changing the neck sizes would not significantly help
increase the yield, the neck dimensions were kept at their original sizes.
From Table 5.9, Riser 1 and 2 have approximately the same actual and calculated
riser sizes but Riser 3 was quite different when calculated with this Riser Design Wizard
accessed from the Simulation tab. Riser 3 was redesigned with the dimensions calculated
from this Riser Design Wizard. The dimensions of the redesigned riser can be found in
Figure 5.47. For the neck dimensions, the thickness and width was kept the same as the
original design, so, the neck was considered a short neck. The resulting surface area of
the connecting neck can be calculated as shown in equation (5.31). A summary of the
dimensions which shall be used in designing Riser 3 or the exothermic top riser are
shown in Table 5.11.

Connecting area of neck = 0.863 inch2

Length of neck = 0.863 0.75
= 1.151 inch (5.31)

Table 5.11: Dimensions for designing Riser 3.

Riser Dimensions (inch) Neck Dimensions (inch)
Diameter Height Width Length Thickness
2.75 4.125 0.75 1.151 0.25

5.9.1.2 Riser Design Wizard: VDG Iron Properties Calculator

The Riser Design Wizard accessed from the Simulation tab can design multiple
risers for a single casting but does not calculate for the dimensions of the necks. The
Riser Design Wizard accessed from the VDG Iron Properties Calculator can design one
riser for a single feeding area casting and the neck dimensions. From analyzing the
methods of calculating the riser dimensions of both the Wizard programs, they are very

183
www.Iran-mavad.com

similar to each other. It seemed appropriate that the Riser Design Wizard accessed from
the VDG Iron Properties Calculator can calculate the dimensions of a single riser feeding
a single feeding area by entering the volume and modulus of the feeding area instead of
the whole casting. The modulus and volumes of the feeding areas from Figures 5.44,
5.45 and 5.46 are summarized in Table 5.12.

Table 5.12: The modulus and volumes of the three feeding areas of the casting.
Riser No. Modulus of Feeding Area (inch) Volume of Feeding Area (inch)
1 0.423 90.069
2 0.420 98.586
3 0.408 116.235

1) Design For Riser 1

Riser 1 represents the single blind side riser in the actual casting design. The
gating system is gated through the riser. The mold dilation is 1.5 percent. If the riser
height per diameter ratio is 1.5, by using the formula in equation (4.2), the diameter of the
riser can be calculated as shown in equation (5.32) and the results in the Riser Design
Wizard shown in Figure 5.48.

[ x (D2/4) x (D x r)] x %p = Vc x (%VDG + %MD) from (4.2)

Where,
D = Diameter of side riser, inch
r = Riser height per diameter ratio, 1.5
Vc = Volume of casting (feeding area) from Table 5.12, 90.069 cubic inches
%p = Proportion of liquid metal removed from riser, 20 percent
%VDG = Percent shrinkage calculated from the VDG Iron Properties Calculator,
4.1 percent
%MD = Mold dilation, 1.5 percent

[ x (D2/4) x (D x 1.5)] x 0.2 = 90.069 x (0.041 + 0.015)

D = 2.777 inch (5.32)

184
www.Iran-mavad.com

Figure 5.48: Resulting calculations for Riser 1 in Riser Design Wizard.

2) Design For Riser 2

Riser 2 represents the shared blind side riser in the actual casting design. The
gating system is gated through the riser. The mold dilation is 1.5 percent. If the riser
height per diameter ratio is 1.5, by using the formula in equation (4.2), the diameter of the
riser can be calculated as shown in equation (5.33) and the results in the Riser Design
Wizard shown in Figure 5.49.

[ x (D2/4) x (D x r)] x %p = Vc x (%VDG + %MD) from (4.2)

Where,
D = Diameter of side riser, inch
r = Riser height per diameter ratio, 1.5
Vc = Volume of casting (feeding area) from Table 5.12, 98.586 cubic inches
%p = Proportion of liquid metal removed from riser, 20 percent
%VDG = Percent shrinkage calculated from the VDG Iron Properties Calculator,
4.1 percent

185
www.Iran-mavad.com

 %MD = Mold dilation, 1.5 percent

[ x (D2/4) x (D x 1.5)] x 0.2 = 98.586 x (0.041 + 0.015)

D = 2.862 inch (5.33)

Figure 5.49: Resulting calculations for Riser 2 in Riser Design Wizard.

Riser 2 is the shared riser in the casting design, so, actually it must feed two
feeding areas. A test shall be conducted by making the riser size large enough to feed
both feeding areas considering them to be one whole feeding area. The feeding areas
would have the same modulus even when combined ( (V + V)/(A + A) = V/A ), but since
the Riser Design Wizard was designed for risers feeding one feeding area with one
connecting neck, the dimensions needed to make two necks cannot be found. At first,
dividing the neck connecting area seems possible, but it would affect the neck modulus,
so, both necks shall have the dimensions of the calculated neck dimensions. The total
volume of the two feeding areas is 197.172 cubic inches (2 x 98.586). By using the

186
www.Iran-mavad.com

formula in equation (4.2), the diameter of the riser can be calculated as shown in equation
(5.34) and the results in the Riser Design Wizard shown in Figure 5.50.

[ x (D2/4) x (D x r)] x %p = Vc x (%VDG + %MD) from (4.2)

Where,
D = Diameter of side riser, inch
r = Riser height per diameter ratio, 1.5
Vc = Volume of casting (feeding area), 197.172 cubic inches
%p = Proportion of liquid metal removed from riser, 20 percent
%VDG = Percent shrinkage calculated from the VDG Iron Properties Calculator,
4.1 percent
%MD = Mold dilation, 1.5 percent

[ x (D2/4) x (D x 1.5)] x 0.2 = 197.172 x (0.041 + 0.015)

D = 3.606 inch (5.34)

Figure 5.50: Resulting calculations for Riser 2 feeding two feeding areas in Riser Design
Wizard.

187
www.Iran-mavad.com

3) Design For Riser 3
Riser 3 represents the top riser with exothermic topping in the actual casting
design. The gating system was not gated through the riser. The mold dilation is 1.5
percent. If the riser height per diameter ratio is 1.5, by using the formula in equation
(4.2), the diameter of a normal side riser can be calculated as shown in equation (5.35)
but the volume of a top riser must be 20 percent more than the volume of a side riser
without a hemispheric bottom, so, the diameter of this top riser must be calculated as
shown in equation (5.36).

[ x (D2/4) x (D x r)] x %p = Vc x (%VDG + %MD) from (4.2)

Where,
D = Diameter of side riser, inch
r = Riser height per diameter ratio, 1.5
Vc = Volume of casting (feeding area) from Table 5.12, 116.235 cubic inches
%p = Proportion of liquid metal removed from riser, 35 percent
%VDG = Percent shrinkage calculated from the VDG Iron Properties Calculator,
4.1 percent
%MD = Mold dilation, 1.5 percent

[ x (D2/4) x (D x 1.5)] x 0.35 = 116.235 x (0.041 + 0.015)

D = 2.509 inch (5.35)

x (D2/4) x (D x 1.5) = [ x (2.5092/4) x (2.509 x 1.5)] x 1.20

D = 2.666 inch (5.36)

Figure 5.51 shows the resulting dimensions for the top riser with exothermic
topping with normal neck dimensions. In the current casting design, the top riser has a
neck length of a quarter of an inch and is considered to be a short neck. The new riser
shall also have the same neck length of 0.25 inch and a connecting surface width of 0.75
inch. Figure 5.52 shows the resulting dimensions for a top riser with exothermic topping
with short neck dimensions.

188
www.Iran-mavad.com

Figure 5.51: Resulting calculations for Riser 3 in Riser Design Wizard with normal neck.

Figure 5.52: Resulting calculations for Riser 3 in Riser Design Wizard with short neck.

189
www.Iran-mavad.com

 The side risers would have a hemispheric bottom added while the top riser shall
be a normal cylindrical shape. The calculated diameters, heights and volumes of the riser
and the neck dimensions of Riser 1, 2 and 3 are summarized in Table 5.13.

Table 5.13: The calculated diameters, heights and volumes of the risers and the neck
dimensions of Riser 1, 2 and 3.
Riser Dimensions Neck Dimensions (inch)
Riser Number and Connecting
Diameter Height Volume Neck
Descriptions Surface
(inch) (inch) (inch3) Length
Width Length
Riser 1 2.777 4.166 30.839 1.605 1.605 1.605
Riser 2 with 1 feed area 2.862 4.293 33.755 1.594 1.594 1.594
Riser 2 with 2 feed areas 3.605 5.408 67.465 1.594 1.594 1.594
Riser 3 with normal neck 2.666 3.999 22.323 1.548 1.548 1.548
Riser 3 with short neck 2.666 3.999 22.323 0.75 1.151 0.25

5.9.2 Gating Design Wizard

The gating dimensions calculated from the Gating Design Wizard used the current
casting designs layout and dimensions as guidelines. The gating system would start
from the sprue down to the sprue well, and then three runners shall separate from the
sprue well. Two of the runners shall go left and right to the ingates into the single side
risers. The other runner shall lead into the shared riser with no ingate.
This casting design has a horizontal gating system. The castings including the
risers and gating system has a weight of approximately 300 pounds, found from adding
the mesh weights of all material which are casting material. The Mesh Weights window
of the base case is shown in Figure 5.53. The pouring time used for this casting design
was 22 seconds. This pouring time could be directly entered into the optimal fill time or
fill time calculation box in the Gating Design Wizard as shown in Figure 5.54. The
weight and pouring time was used to calculate the volumetric feed rate.

190
www.Iran-mavad.com

Figure 5.53: Mesh weights of base case.

Figure 5.54: Entering the pour weight and fill time.

191
www.Iran-mavad.com

 Next, the Gate Position window was set up. The casting design has a horizontal
gating system with parting line gating. It has a conical basin, so, the molten metal is
poured directly into the sprue. Conical basins are probably responsible for the production
of more casting scrap than any other single feature of the filling system and are not
recommended (Campbell, 2004). The metal enters the mold at an unknown, high and
unchecked velocity. Slag or dross may enter the mold cavity with the molten metal.
Entrainment of air may occur and may most likely form a vortex.
The height of the current sprue is 16.75 inches with no sprue well. The height of
the casting is 13.5 inches and the height of the casting measured from the parting line is
10.25 inches. The height of the ladle exit is approximated to be 6 inches from the top of
the sprue, which shall be added to the Effective Sprue Height, which was 12.859 inches,
so, the new Effective Sprue Height is 18.859 inches. The gating ratio was set at 1:1.3:1.1
for gray iron pressurized castings. The number of runners was 3; two to each single riser
and one to the shared riser. Even though the ingate into the shared riser was neglected,
the number of gates entered shall still be 3. The completed Gate Position window is
shown in Figure 5.55.

Figure 5.55: The Gate Position window.

192
www.Iran-mavad.com

Figure 5.56: Calculated sprue dimensions.

Figure 5.57: Calculated gate and runner areas and dimensions.

193
www.Iran-mavad.com

 Clicking the Next button in Figure 5.55, the window shown in Figure 5.56 would
appear, and clicking the Next button in Figure 5.56, the window shown in Figure 5.57
would appear. Figure 5.56 shows the calculated sprue dimensions and Figure 5.57 shows
the calculated gate and runner areas and dimensions. In Figure 5.56, the velocity of the
liquid metal seems to be very high compared to the critical maximum velocity for most
liquid metals of 0.5 meters per second (approximately 20 inches per second) (Campbell,
2004). In Figure 5.57, the runner areas for all runners and the required gate area for all
the gates are the same. The runners were set to have a width per height ratio of 0.5 and
the gates were set to have a width per length ratio of 0.2 for the connecting surface. The
calculator buttons were used for finding the length of the other side of the rectangular
shaped runner or gate with the same area. From Figures 5.56 and 5.57, the dimensions of
the sprue well can be calculated. The sprue well should have a diameter of twice the
diameter of the sprue exit, which is 1.566 inch, and a depth of 1.5 times the height of the
runner, which is 0.968 inch. Table 5.14 shows the dimensions of the currently used sprue,
runners and ingates compared to the calculated ones.

Table 5.14: Dimensions of the current gating system and the proposed gating system.
Current Gating System Proposed Gating System
Gatings and Descriptions
Width Depth Area Width Depth Area
Runner from sprue to gates
0.875 1.75 1.53125 0.323 0.645 0.208
into single side risers
Two runners to middle 2 x 0.28125
0.75 0.375 - - -
runner = 0.5625
Middle runner into shared
1 0.625 0.625 0.323 0.645 0.208
riser

Ingate into a single side riser 0.75 0.375 0.28125 0.938 0.188 0.176

Top of sprue 2.125 - 3.547 0.783 - 0.481

Bottom of sprue 2 - 3.142 0.783 - 0.481

194
www.Iran-mavad.com

 The calculated diameter of the top and bottom of the sprue are the same (round
straight sprue), which is not practical and not recommended (Campbell, 2004). The sprue
should have a tapered shape, so, the diameter of the top of the sprue may be increased
and have a conical basin or maybe a pouring basin at the top of the sprue.
It was seen that the velocity of the liquid metal exiting the sprue exit is extremely
high (106.237 inches per second) compared to the critical maximum velocity of liquid
metal (20 inches per second) (Campbell, 2004). It seems that SOLIDCast does not
consider if the velocity of liquid metal may be too high and may damage the casting. An
experienced user may consider the dimensions calculated from the Gating Design Wizard
to be a guideline in designing the gating system and then modify the dimensions to
reduce the velocity of the liquid metal himself or herself.
SOLIDCast has an extension program called FLOWCast used for simulating the
flow of liquid metal during pouring. Since SOLIDCast cannot simulate the affect of the
speed of flow of the liquid metal, FLOWCast should be used along side SOLIDCast to
help the user modify the gating dimensions calculated from the Gating Design Wizard.
SOLIDCast may also need to be improved so it may recognized and warn the user if
velocities of liquid metal may be too high.

5.10 Simulation Results of Re-designed Casting Design

Three simulation tests were conducted. The first test was the casting design with
a smaller top riser calculated from the Riser Design Wizard accessed from the Simulation
tab. The second and third tests were the casting design with riser dimensions calculated
from the Riser Design Wizard from the VDG Iron Properties Calculator and gating
dimensions calculated from the Gating Design Wizard, but in the second test, the shared
riser was larger than the third test because it was considered to have a volume to feed two
feeding areas of the two castings inside the mold.

5.10.1 Simulation Results: Test 1

The casting design was a smaller top riser with a smaller neck calculated from the
Riser Design Wizard accessed from the Simulation tab. The single side risers and their
necks, the shared riser and its neck, and the gating system had the same dimensions as the

195
www.Iran-mavad.com

original design. The settings in the Materials List window was exactly the same as the
base case except for the fill time which was reduced from 22 seconds to 21 seconds due
to the new weight of total casting.
Because the top riser was designed with a height per diameter ratio of 1.5, the
height of the riser was reduced. The sprue was shortened to match the height of new top
riser with exothermic topping. This would result in a smaller flask, which is not practical.
In order to be able to use the same flask size, the height of the top riser should be kept the
same and the diameter should be reduced but having the same volume. For this test, the
dimensions were taken directly from the program to observe its true usefulness.
The Critical Fraction Solid Time plots showed that the thin wall and the upper
ring area solidified slightly faster than the base case. The plots also showed that some
areas in the upper ring area may have been isolated as shown in Figure 5.58. Since the
solidification rate in the ring is quite high, they can be considered to cool down almost at
the same time. The bottom ring area of the casting solidified at the same rate as the base
case, because the dimensions of Riser 1 and 2 were not modified and the bottom part of
the casting is isolated from the top. The Material Density Function plots showed slightly
larger highlighted areas in the areas between the core pieces and at the bottom of the
casting compared to the base case, as shown in Figure 5.59. But since the areas
highlighted are not the main problem and the level of porosity is not high, they were
considered to not affect the final casting results.
The FCC Criterion plots showed no highlighted areas in the casting at all, as
shown in Figure 5.60. This shows a remarkable improvement in the casting design. The
Hot Spot Criterions iso-surface plots were very much the same as the base case except
the area under the top riser seems to have a larger highlighted spot compared to the base
case, but since the solidification rate in that area was quite high, possible problems
occurring in that area may be low.

196
www.Iran-mavad.com

Figure 5.58: Critical Fraction Solid Time plots plotted at 4.5 minutes show possible
isolated areas.

Figure 5.59: Material Density Function plots plotted at 0.995.

Figure 5.60: FCC Criterion plots showing no highlighted areas.

197
www.Iran-mavad.com

 The simulation results showed a remarkable improvement in the level of predicted
microporosity. Other criterions may show slightly different predicted plots compared to
the base case but still at very much the same levels. Some minor adjustments may be
made to further improve the casting design, but if this design was used instead of the
currently used design, the casting yield may be increased. The weights of the materials in
the model are shown in the Mesh Weights window, shown in Figure 5.61. The total
weight of the casting material in this model can be calculated as shown in equation (5.37).

Figure 5.61: The weights of the materials in Test 1.

2 x (82.077 + 51.666 + 0.491) = 268.468 pounds (5.37)

The weights of the materials of the base case is shown in Figure 5.53 and the total
weight of the casting material in the base case can be calculated as shown in equation
(5.38).

2 x (82.098 + 70.163 + 1.353) = 307.228 pounds (5.38)

The weight of the casting material was reduced by 38.76 pounds per mold or
approximately 12.62 percent. This can be considered a significant savings in the
materials cost. The weight reduced was from the reduction in size of the top risers and

198
www.Iran-mavad.com

the sprue. If the height of the sprue was kept the same and the risers were modified to be
open risers to fit the same flask size, less casting material would actually be reduced.

5.10.2 Simulation Results: Test 2

The dimensions of the risers in this test were calculated from the Riser Design
Wizard accessed from the VDG Iron Properties Calculator and the dimensions of the
gating system were calculated from the Gating Design Wizard. The shared riser was
considered to have a volume which could feed two feeding areas in both the castings in
the mold. The top riser was considered to have a short neck. The dimensions used for
building this model was retrieved from Table 5.12 and 5.13. The model built with these
dimensions is shown in Figure 5.62.
The total height of the model was reduced due to the reduction in size of the top
riser. The sprue height was also modified to match the new top riser. Reducing the size
of the flask is not practical, so, the dimensions of the top riser may need to be modified to
use the same flask size. The sprue shape calculated from the Gating Design Wizard was
a round straight sprue, which is not pratical. The user may modify the sprue dimensions
for the real casting, but because SOLIDCast cannot calculate the flow and friction loss in
the simulation, and the sprue would be considered full during filling, the gating
dimensions shall be kept as is for the simulation.
A sprue well was added to the model. The gating design was slightly changed
due to the methods for calculating for the gating dimensions in the Gating Design Wizard.
The runner that runs to the shared riser would have two smaller runners connected from
the side runners in the original model. In this case, the runner would run directly from
the sprue well into the shared riser. If this design was applied to a real casting, there may
be defects occurring from slag and dirt. The side runners have the same length as the
original model.

199
www.Iran-mavad.com

Figure 5.62: Casting model built from calculated dimensions for Test 2.

The settings in the Materials List window were kept the same as the base case, but
the fill time was reduced to 20 seconds due to the reduced weight of the casting model.
The fill time used in the base case was 22 seconds and the casting weighing around 300
pounds. Back calculating the optimal fill time calculator in the Gating Design Wizard
from equation (4.7) would result in a critical section thickness of around three-eights of
an inch (3/8 inch), which is approximately the thickness of the ingates of the casting.
Using this critical section thickness, the new calculated fill time from the Gating Design
Wizard for this particular casting design was approximately 20 seconds. The weights of
the materials in the Mesh Weights window are shown in Figure 5.63. The total weight
of the casting material was 217.614 pounds. The weight of the casting material was
reduced by 89.614 pounds per mold or approximately 29.17 percent. The weight reduced
was from the reduction in size of all the risers and the gating system.

200
www.Iran-mavad.com

Figure 5.63: The weights of the materials in Test 2.

The solidification progression shown in the Critical Fraction Solid Time plots
showed that the upper ring area solidified at a faster rate than the base case. For the
bottom ring area, the shared riser seemed to feed a larger area of the bottom ring area
compared to the single riser, and also took a longer time to solidify towards the riser. An
example is shown in Figure 5.64. It took around 12 minutes for the casting to cool down
lower than the critical fraction solid temperature.

Figure 5.64: Critical Fraction Solid Time plot plotted at 7.5 minutes.

201
www.Iran-mavad.com

 The predicted macroporosity levels shown in the Material Density Function plots
were very similar to the base case, with slightly more highlight levels in the area between
the core pieces, as shown in Figure 5.65. The FCC Criterion plots showed the same
predicted microporosity level compared to the base case, but more dispersed, as shown in
Figure 5.66.

Figure 5.65: Material Density Function plots plotted at 0.995.

Figure 5.66: FCC Criterion plots.

The Hot Spot Criterions plots were quite similar to the base case, but with more
highlighted intensity, especially under the top riser and in the upper ring area, as shown

202
www.Iran-mavad.com

in Figure 5.67 and 5.68. There may not be any problems or may be very minor because
this area has a high cooling rate.

Figure 5.67: Hot Spot (Solidification Time) plotted at 1.1.

Figure 5.68: Hot Spot (Critical Fraction Solid Time) plotted at 1.1.

The simulation results of this model has slightly more predicted problems than the
base case, but the need for casting material has decreased by around 29.17 percent. From
the Critical Fraction Solid Time plots, it seems that the shared riser is too big and can be
reduced. In the base case, the shared riser has the same dimensions as the single riser, so,
it may be possible to consider as if the shared riser only needs to feed one feeding area.

203
www.Iran-mavad.com

5.10.3 Simulation Results: Test 3
This model is almost exactly the same as Test 3, but the size of the shared riser
was considered to feed only one feeding area of one casting. The size of the shared riser
was almost the same size as the single side riser, similar to the base case which the single
side riser and shared riser have the same dimensions. The model is shown in Figure 5.69.
The weights of the materials in the Mesh Weights window are shown in Figure 5.70. The
total weight of the casting material is 210.374 pounds. The weight of the casting material
was reduced by 96.854 pounds per mold or approximately 31.53 percent. The weight
reduced was from the reduction in size of all the risers and the gating system.

Figure 5.69: Casting model built from calculated dimensions for Test 3.

Figure 5.70: The weights of the materials in Test 3.

204
www.Iran-mavad.com

 The Critical Fraction Solid Time plots showed almost exactly the same
solidification progression compared to Test 2. The shared riser still seemed to feed a
larger portion of the lower ring area compared to the single riser and still took more time
for the liquid metal to solidify towards the shared riser than the single riser, but this test
took slightly less time compared to Test 2. In the Material Density Function plots plotted
with 1.0, the section at where the thin wall and the lower thick ring are connected is
highlighted, as shown in Figure 5.71, but when plotted with 0.995 would disappear, and
almost no predicted macroporosities are shown, as shown in Figure 5.72. The critical
value for plotting a Material Density Function plot is between 0.99 and 0.995. The
highlighted areas disappeared when plotted with 0.995 showing that there is less than 0.5
percent possibility of macroporosity occurring in the highlighted areas, which means
there may only be a very slight possibility to have this particular shrinkage.

Figure 5.71: Material Density Function plots plotted at 1.0.

Figure 5.72: Material Density Function plots plotted at 0.995.

205
www.Iran-mavad.com

 The FCC Criterion plots were almost the same as Test 2, only with less
highlighted areas in the upper ring area. For the Hot Spot Criterions plots, they showed a
high similarity compared to Test 2, but with less highlighted intensity, as shown in Figure
5.73 and 5.74. Compared to Test 2, Test 3 showed less predicted possible defects.

Figure 5.73: Hot Spot (Solidification Time) plotted at 1.1.

Figure 5.74: Hot Spot (Critical Fraction Solid Time) plotted at 1.1.

5.10.4 Conclusions of Simulation Results

The Riser and Gating Design Wizard programs were useful for designing the riser
and gating system for a casting design. The results showed that the reduction in riser
sizes and gating system would increase the yield. The percent decrease in casting

206
www.Iran-mavad.com

material weight compared to the base case are shown in Table 5.15 and the casting yield
of each case including the base case are shown in Table 5.16.

Table 5.15: The percent decrease in casting material weight compared to the base case.
Casting Calculated
Decrease in Decrease in
Material Mesh Casting
Test No. Casting Casting
Weight of Base Material Mesh
Material (lbs) Material (%)
Case (lbs) Weight (lbs)
Test 1 307.228 268.468 38.76 12.62
Test 2 307.228 217.614 89.614 29.17
Test 3 307.228 210.374 96.854 31.53

Table 5.16: The casting yield of each case including the base case.

Weight of Weight of riser

Simulation Total Weight Casting Yield
Casting and gating
Case of Model (lbs) (%)
Material (lbs) system (lbs)
Base Case 164.196 143.032 307.228 53.44
Test 1 164.154 104.314 268.468 61.14
Test 2 162.868 54.746 217.614 74.84
Test 3 163.796 46.578 210.374 77.86

Both Table 5.15 and 5.16 show that the current casting design could be modified
and a higher yield is possible to achieve. Even though the riser and gating dimensions
produced from the Wizard programs do not produce castings with zero possible defects,
or with acceptable liquid metal velocity levels, but they are a good starting point for
designing a casting design. The design may then be modified or additional materials may
be added with the knowledge and experience of the user to eliminate or minimize the
defects.

207
www.Iran-mavad.com

5.11 Conclusions
No castings were able to be made to prove if the modified designs would result in
the same way as predicted, but the simulation results are good estimations of how a
casting may come out. The calculations of an outcome could be predicted without
calculating by hand or just from experience. The calculations for dimensions for risers
and gating system in the Riser and Gating Design Wizard programs are based on casting
knowledge and could be useful even to a user who may not know the calculations behind
them. SOLIDCast was useful for modifying a castings riser and gating system
dimensions, for helping the user in designing the whole riser and gating system and for
predicting the design outcomes.

208
www.Iran-mavad.com

 Chapter 6
Summary and Conclusions
SOLIDCast is one of the casting simulation software programs that can be used to
help a user visualize the solidification of a casting and forecast the outcome of a casting
design. The focus of this thesis was to observe the usefulness, the capabilities and the
limitations of SOLIDCast in simulating cast iron castings cast by the sand casting method.
Tests were conducted on some functions of SOLIDCast which determined the important
capabilities and limitations of the software. A simple casting model was created to study
the usefulness of the Riser Design Wizard and Gating Design Wizard in helping the user
design the risering and gating systems for a casting. SOLIDCast was used in simulating
and analyzing the case study design, a casting design of a locomotive piston made from
gray iron. The Riser Design Wizard and Gating Design Wizard programs were also used
to improve the casting designs risering and gating designs respectively.

6.1 Basic Capabilities and Limitations of the Simulation Function

There are various types of defects that may occur in a casting. Different casting
simulation software programs and modules may predict some types of defects that may
occur in the casting. SOLIDCast is a casting simulation software program which can
simulate thermal changes and heat transfer in the solidification process of a casting. It
assists the user to visualize the solidification process of a particular casting. The program
offers functions to produce visual outputs showing possible problem areas and possible
defects that may occur in a casting. The output criterions available in SOLIDCast are:
1) Solidification Time
2) Critical Fraction Solid Time
3) Material Density Function
4) Temperature Gradient
5) Cooling Rate
6) Niyama Criterion
7) Hot Spots
- Solidification Time
- Critical Fraction Solid Time

209
www.Iran-mavad.com

 8) Custom Criterion
- FCC Custom Criterion
9) Temperature
10) Liquidus Time

The predictions are limited to the output criterions. SOLIDCast would simulate
and collect temperature and time data, starting from when the metal is poured into the
mold until the last node of casting material is completely solidified. The program could
calculate an approximate flow and the heat loss during the flow but it cannot calculate the
turbulence from the flow.
Casting simulation softwares currently do not predict the porosities occurred from
first principles, i.e., the pressure drop which may occur in various parts of the casting.
Many researchers found that there are other parameters which are easier to calculate and
can be used for assessing the porosity formation (Campbell, 2003). From all criterions
available in SOLIDCast and the custom criterion, it was found that all output criterions
are functions of temperature, time and distance, except for the Material Density Function
that also takes gravity into consideration.
The function of time in SOLIDCast would be restricted to the programs time
zero; the time at which pouring ends. For example, when pouring is still in progress,
some parts of the casting may lose heat and solidify before pouring is finished. There
may be areas which the solidification process may not be observed because of this
limitation, especially castings with very thin sections.
The accuracy of the predictions from the output criterion functions depends upon
the knowledge and opinions of the user. The plots of the output criterions may need
further inspection to achieve the most accurate predictions, for example, areas
highlighted in the Hot Spot Criterions plots may not have any problems if a high cooling
rate was achieved in that area, etc.
Tests and observations were conducted to study some of the basic functions in the
SOLIDCast program that are used for simulating a cast iron casting model. The tests and
their brief conclusions are presented.

210
www.Iran-mavad.com

1) Basic Functions for Casting Cast Irons
This set of tests consisted of observations on the Materials List window, the Gray
Iron and Ductile Iron Calculator buttons in the Curves tab and the VDG Iron Properties
Calculator. It was found that only the Casting tab and the Curves tab in the Materials List
window have some relationships between each other. This was found from altering the
parameter values in each tab and observing the changes occurred from the alterations.
Only the specific heat, the initial temperature, the solidification temperature, the freezing
range and the latent heat of fusion in the Casting tab would affect the temperature curve
in the Curves tab.
The Gray Iron and Ductile Iron buttons were found to achieve the same
temperature curves in the Curves tab when entered with the same weight percent carbon
and silicon. The slight difference found was that the temperature curve produced from
the Gray Iron Calculator button slightly curved when the steepness of the temperature
curve changed and had an S shape, whereas the Ductile Iron Calculator button did not
and had a Z shape. As for the VDG Iron Properties Calculator, it is based on the VDG
nomograms as published by the German Iron Society (Finite Solutions Inc., 2005).
The VDG Iron Properties Calculator takes five inputs; the weight percent carbon,
silicon and phosphorus, casting modulus and temperature in the mold and give two
outputs; the shrinkage time in solidification and the expansion or contraction rate for the
user to modify the shrinkage curve and set the critical fraction solid line and the Niyama
line. The temperature in mold or the average temperature in the mold was recommended
in the SOLIDCast manual (Finite Solutions, 2005) to be 75 to 100 degrees Fahrenheit
less than the pouring temperature. It seems inappropriate to presume the temperature.
The pouring temperature was used in all tests as the temperature in mold. SOLIDCast
should develop a calculator to find an appropriate temperature in mold for each particular
casting by taking inputs from the user, such as the casting weight or volume, the casting
modulus and the pouring temperature, etc.
The conclusions from this set of tests were that the information in the database of
the program and the calculators built into the program were derived from metallurgy
knowledge arranged and applied into the program. To achieve the most accurate results,

211
www.Iran-mavad.com

the user must enter the correct data collected from the materials used in the casting
process.

2) Modifying the Curves Tab

In SOLIDCast, the amount of shrinkage of metal at any time would equal to the
volumetric difference of the mold cavity and casting. The shrinkage curve in the Curves
tab is actually the shrinkage of the metal added the shrinkage occurred from the
expansion of the mold or the percent mold wall movement. Modifying the shrinkage
curve would not affect the weight of the casting because the weight of the casting is
calculated from the size of the nodes, the total number of nodes and the density of the
material. The calculation of the weight of a casting could only be found from meshing
and not from simulation. It was also found that SOLIDCast does not simulate the
dimensional shrinkage or expansion of the casting.
Modifying the shrinkage curve in the Curves tab only affected the Material
Density Function output criterion plots. It was also found that the shrinkage occurred
before the liquid metal reaches its liquidus point would form a flat shape pipe. The
shrinkage that occurred after the liquidus point would form a cone shape pipe.
Contraction or expansion was not significant after the liquid metal passes its critical
fraction solid point.

3) Other Tests
It was found that SOLIDCast would allow liquid metal to flow until it reaches its
solidification point instead of the critical fraction point where it should lose the ability to
flow. The user must use his or her own judgment in identifying misruns and coldshuts.
The program may need improvement to allow the liquid metal to flow until it reaches its
critical fraction solid point, not its solidification point.
The maximum number of nodes which could be created by this version of
SOLIDCast was 19,046,664 nodes. The Planes of Symmetry function was a useful
function which can help reduce the use of nodes in a simulation. It reduced the time used
for simulating a casting model using the same node size or could achieve more detailed
simulation results using a smaller node size for a symmetrical casting design. It was

212
www.Iran-mavad.com

found that the meshing starts from the edge of the casting model, so, the symmetrical
plane may not actually be at the symmetrical line dividing the symmetrical model. The
user must select a node size which would divide the model symmetrically or select the
smallest node size possible in order to achieve the most accurate results.
A test was also conducted to observe the meshing function. A portion of the
material would be considered a whole node if it is larger than half the size of the node
size and neglected if smaller. The node size could not exceed the size of the thinnest
section of any material used in the mold. Because the weights of each material in the
model were calculated from the number of nodes, the node size and the density of the
material, the only situation where the calculated weights of the model would be exactly
correct is if the model intersects perfectly with the nodes. For intricate shaped models,
the more detailed the meshes could represent the model or the smaller the nodes could
possibly be, the better approximation of the weights of the model.

6.2 Usefulness of the Riser Design Wizard and Gating Design Wizard Programs
The Riser Design Wizard and the Gating Design Wizard were programs built into
SOLIDCast to help a user develop risering and gating designs. Using the simple hollow
rectangular box shaped casting case and the case study, the calculation methods behind
the programs were analyzed and the formulas used in the programs were uncovered.
SOLIDCast has two separate Riser Design Wizard programs. One was accessed
from the VDG Iron Properties Calculator and the other from the Simulation tab in the
menu. The one accessed from the VDG Iron Properties Calculator would draw the
results from the calculator into the program to calculate for the riser and neck dimensions.
This Riser Design Wizard program would only be good for calculating a riser for a single
feeding area casting. The other Riser Design Wizard accessed from the Simulation tab in
the menu was able to calculate dimensions for multiple risers, but does not provide the
neck dimensions for the risers.
The Riser Design Wizard and the Gating Design Wizard programs provide only
some information in designing the risering and gating systems. The output dimensions
from the program and the dimensions which must be determined by the user are shown in
the following:

213
www.Iran-mavad.com

1) Riser Design Wizard Accessed from the VDG Iron Properties Calculator
This Riser Design Wizard used two methods to calculate the riser design; a
calculation based on riser volume requirement and a calculation based on the riser
modulus being at least 20 percent more than the calculated neck modulus. The latter
method of calculation would only be used if the calculation based on volume requirement
does not satisfy the rule that the riser modulus has to be at least 20 percent larger than the
calculated neck modulus. The output values from the wizard program are as follows:
Riser diameter and height
The program assumes the shape of a side riser would have a cylindrical
shape with the calculated diameter and height added a hemispheric bottom, so, the
true height per diameter ratio would not be the same as entered primarily. For a
top riser, the program would calculate a riser size with twenty percent more
volume than the side riser without the hemispheric bottom.
Neck dimensions
Diameter of neck
Neck dimensions as a cubic shape
Neck dimensions with a rectangular connecting face with same area as the
cubic neck
The neck dimensions were calculated from the calculated neck modulus
and had a cubic shape, i.e., the width, thickness and length of the calculate neck
had the same length. If the user would want to change the connecting surface to a
different shape, such as a rectangular shape, the true modulus of the neck would
change and the solidification rate of the neck would change as well. The program
provides a built-in calculator to calculate for a rectangular shape connecting
surface with the same area as the calculated square shape connector, but this
would change the modulus of the neck.

The Riser Design Wizard does not locate the position where the user should
attach the riser nor what type of riser should be used for the casting. The user must use
his or her own judgment to position the riser.

214
www.Iran-mavad.com

2) Riser Design Wizard Accessed from the Simulation Tab
This Riser Design Wizard program can help the user design multiple risers for a
casting with multiple feeding areas determined by the program. The calculations for the
riser dimensions are based on the method that the riser to casting modulus ratio is 1.2.
The output values from the wizard program are as follows:
Riser diameter and height
The program assumes that the shape of either a side or a top riser would
have a cylindrical shape, which is different from the other Riser Design Wizard.
The height per diameter ratio of the calculated riser dimensions would have the
same ratio as entered primarily.

This wizard program does not provide neck dimensions. They must be
determined by the user. This wizard program also does not provide the position where
the user should attach the riser nor what type of riser should be used for the casting.

3) Gating Design Wizard

When entered the Gating Design Wizard, the flow rate of the liquid metal must be
calculated. The user must enter the alloy sensitivity and the pour weight (including
casting and rigging). The alloy sensitivity value must be approximated by the user
without exact rules of which alloy must have what certain value. The critical section
thickness must be approximated by the user to calculate the fill time or the user may enter
the fill time from past experience. The approximate sprue height added the liquid metal
in the basin and the casting dimensions for the selected gating system must be determined
by the user. The user must also determine the gating ratio for the casting design from
casting handbooks according to the alloy used.
The Gating Design Wizard program would provide the user with the calculated
areas, velocities and diameters of the choke area, the area at the bottom of the sprue and
the area at the top of the sprue. It also provides the total runner area and required ingate
area. The areas of the bottom of the sprue, the runner and the ingate would fall in the
gating ratio set by the user, but their calculated values are the results of its ratio value
multiplied by the calculated choke area instead of the smallest of the ratio being the

215
www.Iran-mavad.com

choke. For example, the ingate area for a gating ratio of 4:8:3 would have three times the
area of the calculated choke area. This would permit the pouring time to be reduced up to
one third of its original value and thus this calculation needs further investigation.
It seems that SOLIDCast does not consider the velocity of liquid metal and it may
be higher than the critical maximum velocity of liquid metal (Campbell, 2004) and may
damage the casting. An experienced user must consider the dimensions calculated from
the Gating Design Wizard to be a guideline in designing the gating system and modify
the dimensions or otherwise to reduce the velocity of the liquid metal.
SOLIDCast has an extension program called FLOWCast used for simulating the
flow of liquid metal during pouring. Since SOLIDCast cannot simulate the affect of the
velocity of liquid metal, FLOWCast should be used along side SOLIDCast to help the
user modify the gating dimensions calculated from the Gating Design Wizard.
SOLIDCast may also need to be improved so it may recognized and warn the user if
velocities of liquid metal may be too high.
The program provides calculators for calculating dimensions for rectangular
runners and ingates. The program does not recommend what should the width per depth
ratio for the runners or width per thickness ratio for the ingate be. It is only
recommended in the SOLIDCast manual (Finite Solutions Inc., 2005) that the width per
thickness ratio of the ingate should be 5 to encourage high cooling rate so it may freeze
off quickly. As for the runners, a width per depth ratio of 0.5 is practical. The program
is also able to calculate the step down calculation of the runner, meaning the runner
areas after each ingate.
The program does not provide the length of the runners or the ingates, but the
length must be larger than the width or thickness of the ingate or runner. The user must
determine the dimensions of the pouring basin, the height of the sprue and calculate the
dimensions of the sprue well.

4) Conclusions for Riser Design Wizard and Gating Design Wizard

Designing the risering and gating systems for a casting design, the user must
design the layout. For example, the orientation of the casting, where the risers should be
attached to the casting and where the gating systems are laid, etc. The dimensions of the

216
www.Iran-mavad.com

risers and gating system calculated from Riser Design Wizard and Gating Design Wizard
programs may not produce the best solution or defect free casting, but would produce a
good starting point for the user to modify and achieve good castings.

6.3 Further Capabilities and Limitations Found from the Case Study
It was found early in the testing of the case study that the orientation of the
casting in the model significantly affected the simulation process. Because the program
is based on the Finite Difference Method, the program is axis critical. Revolving the
casting and using the same node size resulted in different casting weights and different
simulation results.
In the case study, the metal in the drill holes would freeze and misrun, but the
simulation showed that the drill holes were completely filled. This shows that the
calculation for the shrinkage curve may not be correct and the methods for calculating for
the shrinkage curve may need to be improved.
The simulation results showed very positively that the defects detected from the
simulation matched with the defects found in the casting. But the simulation results
seemed to show signs of axis dependence, such as the highlighted plots were denser near
the X and Y axes.
SOLIDCast is a useful tool for forecasting the outcomes of a modified casting
design. So, instead of trial casting modified designs, the modified design could be tested
by SOLIDCast to see its possible outcomes to reduce defects and improve yield.

6.4 Future Recommendations

Since SOLIDCast is a Finite Difference Method based casting simulation program,
the orientation of the casting is important in the solution. Guidelines must be developed
to indicate proper orientation to give better solutions more rapidly. The impact of the
orientation difference should be determined to determine preventatively the effect of
orientation.
The maximum number of nodes is a restriction reducing the use of an effective
node size for the thinnest part of the casting. Either the program should be able to have

217
www.Iran-mavad.com

more nodes along with faster calculation methods or the meshing method should be
improved.
The use of different node sizes, like the Finite Element Method, must be
investigated to reduce the simulation times. Thick sections must be able to use larger
nodes than thin sections. A possible suggestion method to achieve a larger node size is to
combine x3 number of nodes (8, 27, 64, etc.) into a large single node.

6.5 Conclusions
The accuracy of the simulation depends highly on the input parameters while the
accuracy of the predictions from the output criterion functions depends highly on the
knowledge and opinions of the user. Although, SOLIDCast may have many limitations,
it still is a good tool for verifying the solidification process in a casting. In the future,
metallurgists and software developers may produce more precise casting simulation
software programs that would be user friendly to users with a minimum metallurgical
background. Modules may be developed to find the best design for a casting for casting
quality, yield improvement, and scrap reduction.

218
www.Iran-mavad.com

 REFERENCES
1. Alonso, Arnaldo and Franco, Luis, Foundry Management and Technology, Penton
Publishing Co., v 133, n 10, October, 2005, p. 22 24.
2. Andersen, B., Temperature management [Casting Process], Die Casting Engineer, v.
50, n 1, January, 2006, p. 42 46.
3. Callister, William, D., Material Science and Engineering: An Introduction, 7th
Edition, John Wiley & Sons, Inc., 2006.
4. Campbell, John, The New Metallurgy of Cast Metals: Casting, 2nd Edition,
Butterworth-Heinemann, 2003.
5. Campbell, John, Castings Practice The 10 Rules of Castings, 2004.
6. Creese, Robert C., Introduction to Manufacturing Processes and Materials, 1999.
7. Charoenvilaisiri, Sombun, Ph.D., Faculty of Engineering, Production Engineering
Department, King Mongkuts University of Technology Thonburi, Bangkok,
Thailand, 2006, Personal communication.
8. Danzur, Wynn, The Casting and Molding Processes, Wynn Danzur Marketing,
www.wynndanzur.com.
9. De Looze, G., Effect of die temperature and melt quality on a commercial low
pressure die casting, Shape casting, The John Campbell Symposium, Proceedings of
a Symposium Sponsored by the Aluminum Committee of the Light Metals Division
(LMD) and the Solidification Committee of the Materials Processing and
Manufacturing Division of TMS, held at the 2005 TMS Annual Meeting, 2005, p.
423 432.
10. Finite Solutions Inc., SOLIDCast, Training Course Workbook, Copyright 2005,
Version 6.3.39.
11. Finite Solutions Inc., Training Overview 6.3.39, PowerPoint slides, January 15,
2005.
12. Guharaja, S., Noorul Haq, A. and Karuppannan, K. M., Optimization of Green Sand
Casting Process Parameters by Using Taguchis Method, International Journal of
Advanced Manufacturing Technology, v. 30, n 11 12, October, 2006, p. 1,040
1,048.

219
www.Iran-mavad.com

13. Heine, Richard W., Loper, Carl R. and Rosenthal, Philip C., Principles of Metal
Casting, TMH Edition 1976, Tata McGraw-Hill Publishing Company Limited, India,
1995.
14. Herman, Edmund, Casting cycle rate by design, Die Casting Engineer, v. 48, n 5,
September, 2004, p. 46 52.
15. Iron Castings Society, Iron Castings Handbook, 1981.
16. Jakumeit, J., Laqua, R., Iyas, T., Scheele, J., Braun, M., Mukhopadhya, A., Pelzer, M.,
Multidisciplinary coupled simulations of investment casting processes using
CASTS-FLUENT, Proceedings of Modeling, Control and Optimization in Ferrous
and Nonferrous Industry Symposium, 2003, p. 333 342.
17. Lewis, R. W., Ransing, R. S., Pao, W. K. S., Kulasegaram, K. and Bonet, J., Civil and
Computer Engineering Centre, University of Wales Swansea, UK, Alternative
Techniques for Casting Process Simulation, International Journal of Numerical
Methods for Heat & Fluid Flow, v. 14, n 2, 2004, p. 145 166.
18. Liu, Wenhui, Liu, Yangai, Xiong Shoumei, Liu, Baicheng, Matsumoto, Y., Murakami,
M., Influences of casting pressure conditions on the quality and properties of a
magnesium cylinder head cover die casting, Journal of Materials Science &
Technology, v. 21, n 2, March 2005, p. 170 174.
19. Moreira, Jose and Ribeiro, C.A. Silvia, Confrontation of two solidification
simulation software to a particular use into foundry industry, Material Science
Forum, v 455 456, Advanced Materials Forum II: Proceedings of the II
International Materials Symposium: Materials 2003 and XI Encontro Sociedade
Portugesa de Materials, 2003 ATERIAIS, 2004, p. 690 693.
20. Pan, Weihua, Shen, Yunfu and Zhang, Wu, Grid-based simulation system of casting
solidification process, 8th International Conference on Computer Supported
Cooperative Work in Design (IEEE Cat. No. 04EX709, 2004, pt. 1, 692 697 Vol. 1.
21. Pearce, John T. H., Dr., Tanasombat, Bancha, Dr. and the METALS Group, Casting
Defects, National Metal and Materials Technology Center (MTEC) of Thailand,
2001.

220
www.Iran-mavad.com

22. Pernicka, E., G.A. Wagner, et al. Early Bronze Age Metallurgy in the Northeast
Aegean., Troia and the troad: scientific approaches, Berlin, London: Springer; 2003,
p. 143 172.
23. Pongsukijwat, Suwanchai, Lortongkum, Kobbun, Nisaratanaporn, Aekasit,
Lueangwaranunt, Thachai, Supradit Na Ayutthaya, Mawin and Wisuthpitakkul,
Patama, Materials Science and Engineering: An Introduction, Top Publishing Co.,
Ltd., 2005.
24. Pongsukijwat, Suwanchai, Computer Simulation in Metal Forming Processes,
Research Group of Solidifications of Liquid Metals, Department of Metallurgy,
Faculty of Engineering, Chulalongkorn University, Thailand, 2006.
25. Pongsukijwat, Suwanchai, Department of Metallurgy, Faculty of Engineering,
Chulalongkorn University, Thailand, 2006, Personal communication.
26. Prasertsakul, Supachai, Assistant Professor, Foundry Technology Vol. 2 Standard
Edition, B & K Science & Engineering Publisher.
27. Ravi, B., Metal Casting Computer Aided Design and Analysis, Prentice Hall of
India Private Limited, New Delhi, 2005.
28. Sarfaraz, Ahmad Reza, Computer Simulation and Experimental Analysis of Fin
Effect in L Shaped Casting, Dissertation, College of Engineering of West Virginia
University, 1988.
29. Society of Manufacturing Engineers, Fundamental Manufacturing Processes: Near-
net Shape Casting.
30. Wang, Huei-Sen, Tseng, Hua-Ling, Hwang, Weng-Sing, Weng, Ruey-Jer, Yeh,
Jiunn-Lin, Measurements of interfacial heat transfer coefficients during the casting
of high temperature alloys in sand mold and ceramics mold, Modeling of Casting,
Welding and Advanced Solidification Process X. Proceedings of the Tenth
International Conference, 2003, p. 693 700.
31. Webster, P. D., Fundamentals of Foundry Technology, Portcullis Press Ltd., 1980.
32. Wong, Chen-Chieh, Applications of Campbells casting rules on high quality
aluminum castings, JOM, v. 56, n 11, November, 2004, p. 301 302.
33. Wright, Frank, Walker die casting finds new ways to compete, Die Casting
Engineer, v. 48, May, 2004, p. 32 38.

221
www.Iran-mavad.com

34. Yang, L. J., The effect of casting temperature on the properties of squeeze cast
aluminum and zinc alloys, Journal of Materials Processing Technology, v. 140, n 1
3 SPEC, September 22, 2003, p. 391 396.
35. Zanchuk, V., ZA-12 alloy casting with graphite molds: economical precision parts,
reduced time-to-market, Die Casting Engineer, North American Die Casting
Association, USA, v. 50, n 1, Jan 2006, p. 38 41.
36. www.alphacast-software.co.uk
37. www.calcom.ch
38. www.castech.fi
39. www.ekkinc.com
40. http://en.wikipedia.org/wiki/Rapid_prototyping
41. www.hkec.com
42. http://www.key-to-steel.com/Articles/Art63.htm.
43. www.magmasoft.com
44. www.metal-technologies.com, History of metal casting.
45. http://www.moderncasting.com/MoreInfo/0804/MoreInfo_06_0804.pdf
46. http://www.mrl.ucsb.edu/~edkramer/LectureVGsMat100B/99Lecture14VGs/FeCPha
seDiagramVG.html
47. http://www.msm.cam.ac.uk/phase-trans/2001/adi/cast.iron.html, University of
Cambridge.
48. www.novacast.se
49. www.technalysis.com
50. http://www.twi.co.uk/j32k/protected/band_3/jk25.html.
51. www.simtec-inc.com
52. Wikipedia, The Free Encyclopedia, http://en.wikipedia.org.

222
www.Iran-mavad.com