Comparison Between Linear and

Nonlinear Analysis in FEA Using ANSYS

Darshan R

Copyright2024 Darshan R
Linear and nonlinear analyses in ANSYS differ fundamentally in how they handle
structural behavior under applied loads. Here’s a comprehensive comparison of the two:

1. Basic Concept :

Aspect Linear Analysis Nonlinear Analysis

Definition Assumes that the Accounts for nonlinear

relationship between loads behavior in materials,
and responses geometry, and boundary
(displacements, stresses) conditions.
is linear.

Assumption Small deformations and Large deformations,

linear material properties. nonlinear materials, and
changing boundary
conditions.

2. Types of Nonlinearity:
Linear analysis ignores nonlinearity, while nonlinear analysis can account for:

● Material Nonlinearity: Plasticity, creep, hyperelasticity.

● Geometric Nonlinearity: Large deformations, buckling, or large rotations.
● Contact Nonlinearity: Surface interactions like sliding, friction, or separation.

1. Geometric Nonlinearity:
The changing geometric configuration due to large deformation of the structure causes
nonlinear behavior. Geometric nonlinearity is not only because of large deformation but
also due to large strain & large rotation too. The geometric nonlinearity causes a
change in geometry cross section due to large deformation. Geometry buckling is also a
cause of geometric nonlinearity, in case of buckling huge compressive load causes the
structural member to buckle resulting in higher displacement.

Copyright2024 Darshan R
Here the question comes, How to decide whether the analysis is linear or nonlinear?

In linear geometric analysis, the deformations and rotations are smaller like within 5 %
as generic rule, but in case of nonlinear geometric analysis, displacement and rotations
are large. Small changes in magnitude of force for nonlinear geometry can change
convergence behavior considerably. However the geometric nonlinearity is not only due
to large deformation/rotation but varies based on loading and situation too. Consider an
example of buckling of shells which happen with small initial displacements but under a
very nonlinear behavior. It happens in small displacements and deals with membrane or
axial loads. In FEA software geometry nonlinearity is solved simply by keeping
nonlinear zones on.

It is difficult to predict the case of geometric nonlinearity, as it needs a lot of experiments

and to judge the case by looking at the scenario.

2. Material Nonlinearity:
In the real world all engineering materials are inherently nonlinear but we can idealize
by accounting only certain effects which are important for analysis. Linear materials are
the most simplified material type. The other type is nonlinear elastic if the deformation is
recoverable and plastic if deformation is permanent.

Figure 2: Material-Nonlinearity

Figure 2 shows the stress strain curve for metal, non-metals & nonlinearity features. In
case of metals, material moves to plastic zone after elastic limit, while non-metals have
nonlinear stress strain curve (plastic, asbestos, fibres, etc.) from origin itself thus in both
the cases curve shows the source of nonlinearity. Nonlinear curve helps to input exact
stress vs strain after the yield point which gives the exact results. Furthermore creep,
also considered as a nonlinear phenomenon, at maximum temperature, even for small
magnitude loads, kept applied for a longer period of time, would cause failure. \

Copyright2024 Darshan R
Few Major nonlinear material (behavior) classification:

1. Nonlinear elastic

2. Hyperplastic (example gasket (rubber) material)

3. Linear elastic perfectly plastic

4. Elastic perfectly plastic

5. Elastic time dependent plastic (creep)

6. Strain & temperature dependent elasticity and plasticity.

Stress strain curve is usually obtained from tensile tests in labs and you might
experience some struggle to find out the actual strain stress curve for the material you
are analyzing. In that case you can refer to some research papers or free internet
sources like Matweb, etc. you can even assume the curve of equivalent material and
complete the simulation instead of giving up. In FEA you will put stress vs strain curve
as material input and then FEA calculates the deformation and calculates the strain for
those deformations. And then FEA uses material stress strain curve to translate stain
into stresses.

3. Contact Nonlinearity:
Contact nonlinearity occurs in structural and mechanical systems when two or more
components interact through physical contact, causing dynamic changes in boundary
conditions. This introduces nonlinearity, as surfaces can make contact, separate, or
slide relative to each other. Such interactions lead to complexities like friction, material
separation, and variable contact forces, which must be accurately accounted for during
analysis.
Key Features of Contact Nonlinearity:
1.Dynamic Contact Status: Surfaces may start separated, come into contact as load
increases, or separate again upon unloading.
2.Frictional Interaction: Many contact problems involve friction, which introduces
additional nonlinearity by affecting sliding and sticking behavior.
3.Penetration Constraints: The analysis must enforce a condition where surfaces do
not penetrate each other, requiring specialized solution techniques.
4.Large Deformations: Contact scenarios often involve large deformations, further
contributing to the system's nonlinearity.

Copyright2024 Darshan R
In contact nonlinearity abrupt change in stiffness may occur when bodies come into or
out of contact with each other. This type of nonlinearity is used to simulate the gap
between two parts. While defining contact between two bodies, you need to decide
whether it carries the friction or not. Friction coefficient can be used to decide the
amount of resistance between the contact bodies. The selection of friction coefficient
affects the contacting surfaces behaviour and hence the friction coefficient is selected
based on dry or lubricated friction, contacting material, etc. Sometimes boundary
conditions can also be treated as nonlinear analysis like elastic support.
Rubber boot seal is an example of nonlinear geometry (due to large strain and
large deformation), nonlinear material (rubber) and changing status nonlinearities
(contact).
Contact Algorithms:
To accurately model contact behavior, specialized algorithms are utilized to ensure that
contact conditions, such as preventing penetration or handling friction forces, are
enforced throughout the analysis. These algorithms operate iteratively, adapting as
contact conditions evolve. The most commonly used algorithms include:
1. Penalty Method:
•Concept: The penalty method introduces an artificial stiffness that imposes a force
when surfaces penetrate. The force is proportional to the depth of penetration and a
user-defined stiffness, which pushes the surfaces apart.
2. Lagrange Multiplier Method:
•Concept: This method uses Lagrange multipliers to rigorously enforce the contact
constraint without allowing any surface penetration. A multiplier (force) is introduced to
maintain the contact condition.
3. Augmented Lagrangian Method:
•Concept: A combination of the penalty and Lagrange multiplier methods, the
augmented Lagrangian approach applies a penalty term but refines it iteratively for
more precise constraint enforcement without requiring an excessive penalty stiffness.
Applications of Contact Nonlinearity:
Contact nonlinearity is prevalent in many engineering applications, such as:
•Mechanical Joints: Bolted, riveted, or press-fitted assemblies.
•Gear and Cam Systems: Involving repeated surface contact and sliding.
•Crash Simulations: For vehicle safety analysis where parts undergo contact and
deformation.
•Sealing Mechanisms: Gaskets and seals, where accurate contact modelling is
essential for performance.

Copyright2024 Darshan R
In summary, contact nonlinearity involves complex interactions due to the changing
contact conditions between surfaces. Algorithms such as the penalty method, Lagrange
multipliers, and the augmented Lagrangian approach are used to ensure accurate
modelling of these interactions, providing precise enforcement of contact forces and
constraints.

How to conquer contact penetration issues:

● Refine the mesh elements on the surfaces (especially contact side) which are
close to each other.
● Decrease the contact penetration tolerance, set it to factor to 0.05, to reduce the
tolerance to 5%.
● Increase the contact normal stiffness (keep in mind this might cause
convergence issues too, if the default value 1 does not work, use 0.1)
● Use normal lagrange formulation in contact, this will result in least penetration but
it can be a bit expensive.

3. Governing Equations:

Aspect Linear Analysis Nonlinear Analysis

Equations K⋅u=FK \cdot u = F Iterative solution to

(Stiffness matrix KK, account for changes in KK
displacements uu, forces or FF as deformation
FF) progresses.

Stiffness Matrix Constant throughout the Updated iteratively to

analysis. reflect nonlinearity.

4. Iterative Solving:

Aspect Linear Analysis Nonlinear Analysis

Solution Approach Direct solution without Requires iterative methods

iterations. like Newton-Raphson or
arc-length techniques.

Convergence Not a concern; single-step Ensuring convergence is

solution. crucial; divergence
indicates incorrect setup.

Copyright2024 Darshan R
5. Computational Cost:

Aspect Linear Analysis Nonlinear Analysis

Time Faster due to single-step Slower due to iterative

solving. solving and frequent
stiffness matrix updates.

Memory Less memory-intensive Higher memory

requirements for storing
updated matrices and
results.

6. Accuracy and Applications:

Aspect Linear Analysis Nonlinear Analysis

Accuracy Sufficient for small Essential for high loads,

deformation problems or large deformations, or
low loads. material nonlinearities.

Applications Small-strain elasticity, Contact problems, crash

low-load stress analysis, analysis, buckling, or
modal analysis. hyperelastic materials.

7. Boundary Conditions:

Aspect Linear Analysis Nonlinear Analysis

Boundary Updates Remain constant Can change due to

throughout the analysis. contact, deformation, or
sliding effects.

8. Material Models:

Aspect Linear Analysis Nonlinear Analysis

Material Properties Elastic only. Includes plasticity, creep,

viscoelasticity,
hyperelasticity, etc.

Copyright2024 Darshan R
9. Stress-Strain Behavior:

Aspect Linear Analysis Nonlinear Analysis

Stress-Strain Curve Assumes a linear Incorporates nonlinear

relationship. stress-strain curves for
plasticity or hyperelasticity.

10. Solver Options in ANSYS:

Aspect Linear Analysis Nonlinear Analysis

Preferred Solver Direct solvers like sparse Nonlinear solvers using

matrix solvers. iterative methods like
Newton-Raphson.

Comparison Table Summary

Aspect Linear Analysis Nonlinear Analysis

Behavior Linear load-response Nonlinear load-response

relationship relationship.

Stiffness Matrix Constant. Continuously updated.

Computational Cost Low. High.

Accuracy Adequate for small Necessary for large

deformations. deformations.

Material Models Linear elasticity only. Includes plasticity, creep,

etc.

Applications Simple problems (static Complex problems

stress, modal). (contact, buckling).

Copyright2024 Darshan R
When to Use Each Analysis
● Linear Analysis: Use when:
○ Deformations are small.
○ Material behaves elastically.
○ Boundary conditions remain constant.
● Nonlinear Analysis: Use when:
○ Deformations are large.
○ Materials exhibit plasticity, creep, or other nonlinear behavior.
○ Contact or changing boundary conditions exist.

Nonlinear FEA Issues:

Achieving convergence:
● Obtaining convergence is the biggest challenge in nonlinear analysis.
● Trial and error is sometimes required.
● Complex problems might require more load increments, and many iterations at
each load step to achieve the convergence.
● Balancing expenses versus accuracy
● FEA involves expenses (Solution time, disk and memory requirements)
● More detail and a finer mesh generally lead to a more accurate solution, but
require more time and system resources.
● Nonlinear analysis needs additional iteration that affects both accuracy and
expenses.
● Difficult to verify the FEA results due to increased complexity of nonlinear
behaviour.
● The sensitivity studies (mesh convergence, increased mesh density, reduced
load increments, varying other model parameters) become more expensive.
● Results of the nonlinear analysis cannot be scaled.
● The structural behavior can be marked non-proportional to the applied load

Copyright2024 Darshan R
Conclusion:
Linear analysis is suitable for simple problems with low computational cost, while
nonlinear analysis is essential for accurately capturing real-world complex behaviors. In
ANSYS, the choice depends on the problem's physics, desired accuracy, and
computational resources.

Copyright2024 Darshan R