Behavioral Psychology for the CAE Analyst

INTERACTIONS IN FINITE ELEMENT ANALYSIS: HOW PARTS COMMUNICATE

A Companion Reader

Generated by Joe McFadden with Claude and ElevenLabs

An Educational Audiobook Companion

McFaddenCAE.com

© 2026 Joe McFadden. All rights reserved.

Introduction: The Loneliest Parts in the World

Here is something that trips up even experienced analysts, and it is so fundamental that we almost never talk about it directly.

Your parts do not know each other.

Think about that for a moment. You have spent hours building a beautiful assembly. Every component meshes cleanly. The geometry looks perfect on screen. You hit submit, and the solver either blows up, or worse—it runs to completion and gives you answers that are quietly, dangerously wrong.

The reason? You never introduced your parts to each other. You never told them the rules of engagement.

Unlike us humans, finite element parts do not automatically build a model of their environment and relationships. We walk into a room and instantly begin reading the space. Who is standing close together. Who is avoiding eye contact. Who is in charge. We process dozens of social cues without thinking about it. Our brains build a relationship map in seconds.

Parts in a finite element model have no such instinct. They are born alone, each one a collection of nodes and elements floating in space, aware of nothing beyond their own mesh. Two parts can occupy the same physical space, share the same nodes, or sit a fraction of a millimeter apart, and unless you explicitly define how they relate to each other, the solver treats them as strangers. Ships passing in the night. Each one solving its own equations in perfect isolation.

The Intelligence Gap

Now, you might think this is just a software limitation. Something the developers will fix in the next release. But it is much more fundamental than that, and to understand why, we need to think about what makes living things different from engineered ones.

Recent work in biology and bioelectricity has revealed something remarkable. Intelligence is not centralized. It is not a brain thing. It goes all the way down. A single cell, with no nervous system, no brain, no central processor, can sense a chemical gradient, decide which direction to move, and solve problems about its environment in real time. A cell has agency. It does not wait for instructions from above. It reads its surroundings and acts.

Even a bacterium navigating toward a food source is doing something our most sophisticated finite element parts cannot do. It is perceiving, processing, and responding. It has, at some fundamental level, a competency that is built into the living system itself. The intelligence is not external. It is intrinsic.

Now look at your finite element model. Your parts have material properties, geometry, section definitions, mesh. They have what you might call the structural DNA of the assembly. And just like biological DNA, those properties define the potential for behavior. Steel can yield at 250 megapascals. Rubber can stretch to five times its length. Copper conducts heat beautifully. That information is all there, encoded in the material card and the section definition, like a blueprint sitting on a workbench.

But DNA alone does not build an organism. DNA is the parts list and the assembly manual. It defines what is possible. What brings it to life is the cellular machinery that reads the instructions, interprets them in context, and executes them with purpose. The DNA says this gene codes for a protein. The cell decides when to express it, how much to make, and in response to what signals. That decision-making, that agency, is what turns a pile of molecules into a living thing.

Your finite element parts have all of that—the blueprint, the specifications, everything needed to define behavior. What they do not have is a reader. There is no agent inside the model picking up those instructions and acting on them. The material card says steel has a Young’s modulus of 200 gigapascals. But the part does not know what to do with that information until you, the analyst, tell it how to interact with its neighbors. The intelligence is not in the model. The intelligence is you.

And here is where it gets even more interesting. In biology, the same DNA gets expressed differently depending on context. Inherited epigenetic tags, chemical signals from neighboring cells, mechanical forces from the surrounding tissue—all of these change how the same genetic code plays out in practice. One joint in your body ends up stiffer than another, not because the base code is different, but because the expression is different. The environment shaped the outcome.

We actually have something analogous in finite element analysis. The same material card applied to two different parts can produce completely different behavior depending on the interaction definition. Steel tied rigidly to its neighbor behaves nothing like steel in frictional sliding contact with that same neighbor. The DNA is identical. The context changed everything. But in biology, that contextual adaptation happens from within. The cell senses and responds. In our models, it must be imposed from without. Every interaction, every interface property, every contact definition is a decision that you, the analyst, must make and deliver to parts that have zero capacity to figure it out on their own.

That is the fundamental gap. Living systems have distributed intelligence. Your model has none. The analyst is the only intelligence in the system.

And it goes deeper than that. You need to define not only how parts interact with each other, but how they interact within themselves. A part needs to know its own material properties, its own stiffness, its own density. Without that self-knowledge, it cannot participate in any relationship at all. You cannot have a meaningful conversation if you do not know who you are.

Chapter 1: The Human Analogy

Let us stay with the human analogy for a moment, because it maps beautifully onto the physics.

Think about what determines the quality of any human interaction. It is not just about proximity. Two people can stand shoulder to shoulder on a crowded train and never exchange a word. Or two people on opposite sides of a conference table can have an interaction that changes the direction of a company. The outcome depends on something deeper.

It depends on the nature of each participant.

Proximity: The Search Tolerance

Are they approachable, or too far away to interact? In finite element terms, this is the search tolerance—the initial gap or overclosure. If two surfaces are too far apart, the solver will never detect that they should be talking to each other. Too close, and you may have an initial penetration that creates artificial forces from the very first increment. Getting the distance right is the first prerequisite for any interaction.

Friction: The Personality of the Interface

Is there friction between the participants? We use this phrase in everyday life without realizing how precisely it maps to contact mechanics. When two people have friction between them, their interaction generates heat, resistance, and opposing forces. When two surfaces have friction, the tangential response creates shear stress that resists sliding. And just like human friction, contact friction dissipates energy. It turns kinetic energy into heat. The coefficient of friction is not just a number in a textbook. It is the personality of the interface.

Stubbornness: Penalty Stiffness

Are they stubborn or amenable? In human terms, a stubborn person resists change in their position. They hold firm. An amenable person adjusts, accommodates, yields. In contact mechanics, this maps directly to the penalty stiffness and the enforcement method. A hard contact formulation is stubborn. It says: no, you will not penetrate this surface. The penalty method is slightly more amenable. It allows a tiny amount of penetration, then pushes back proportionally—like a firm but flexible negotiator who gives a little to keep the conversation going.

Compliance: Material Stiffness

Are they compliant or hard? A compliant person absorbs your energy. They listen, they flex, they adapt their shape to accommodate yours. A hard person reflects your energy right back at you. In structural terms, this is material stiffness. A soft rubber gasket between two steel flanges will deform dramatically to conform to surface irregularities. The steel barely notices. The gasket does all the accommodating. This is not a flaw. That compliance is the gasket’s entire purpose.

Chapter 2: Who Leads the Dance

Every interaction has a power dynamic. In human relationships, sometimes this is explicit—a manager and a direct report, a teacher and a student. Other times it is subtler, negotiated moment by moment based on expertise, confidence, and context.

In Abaqus contact, this power dynamic is formalized through the master-slave relationship, or as it is increasingly called in modern solvers, the primary-secondary relationship. And the choice of who leads matters enormously.

The master surface is the one that dictates the geometric reference. It defines the dance floor. The slave surface must conform to it, must not penetrate it. The solver enforces the constraint from the slave side—it checks whether slave nodes have violated the master surface and applies corrective forces to push them back.

Why Assignment Matters

If you get the assignment backwards, the physics breaks down. The general rule is that the stiffer, coarser-meshed surface should be the master. Think of it this way: a rigid steel plate does not conform to a rubber pad. The rubber conforms to the steel. The steel leads. The rubber follows. If you reverse this assignment, you may get convergence failures as the solver tries to enforce an unphysical constraint, or worse, you get a solution where the stiffer surface punches through the softer one without the solver catching it.

The mesh density matters too. A finely meshed slave surface can conform to a coarsely meshed master surface reasonably well. But a coarsely meshed slave surface on a finely meshed master will miss features between its widely spaced nodes. Imagine a person with poor vision trying to follow a partner doing intricate footwork. They simply cannot track the detail.

And just like in human relationships, the best interactions happen when neither party is dramatically mismatched. When both surfaces have comparable mesh density and stiffness, the master-slave assignment becomes less critical because both parties can track each other’s movements with similar fidelity. The dance becomes a true partnership.

Chapter 3: The Types of Relationships

Not all part interactions are the same, and using the wrong type can be as damaging as using none at all.

Tie Constraints

This is marriage. A permanent, rigid bond between two surfaces. Once tied, the surfaces move as one. No relative sliding. No separation. No flexibility at the interface. A tie constraint is the simplest and most stable interaction type, and it is also the most common source of artificial stiffness in models.

When you tie two surfaces together, you are telling the solver that this joint is infinitely rigid. In reality, almost no joint is truly rigid. Bolted connections have compliance. Adhesive bonds have finite stiffness. Welded joints have a heat-affected zone with different properties. Every time you use a tie constraint, you should ask yourself: am I comfortable with the assumption that this interface is perfectly rigid? If the answer is no, you need a different approach.

Contact Pairs

This is the full spectrum of human interaction, from a casual handshake to a wrestling match. Contact pairs allow surfaces to come together, separate, slide against each other, and generate friction forces. The solver tracks the gap between surfaces at every increment, determines which nodes are in contact, and applies the appropriate normal and tangential forces.

Contact is computationally expensive because it is inherently nonlinear. The stiffness of the system changes every time a node comes into or out of contact. Each status change is like a new person joining or leaving a conversation—the entire dynamic shifts. This is why contact problems often require more increments, tighter convergence tolerances, and more patience than bonded analyses.

General Contact

Think of this as walking into a crowded room where everyone can potentially interact with everyone else. In Abaqus, general contact automatically defines contact between all exterior surfaces in the model. You do not need to manually specify each pair. The solver handles the bookkeeping.

General contact is powerful for large assemblies where you may not know in advance which parts will come into contact—drop test simulations, crash analyses, assembly processes. But that convenience comes with a cost. The solver must search for potential contact across all surfaces at every increment, which increases computation time. And because everything can touch everything, you may get unexpected contact between surfaces that should never interact, requiring exclusions to clean up.

Self-Contact

Every interaction type discussed so far involves two different parts. But sometimes a part must interact with itself. This is the mechanical equivalent of self-awareness, and forgetting about it is one of the most common oversights in large-deformation analysis.

Think of a folding airbag. As the bag inflates and unfurls, different regions of the same fabric come into contact with each other. Without self-contact, the solver lets those regions pass through themselves freely. The elements overlap, the geometry inverts, and you get either a crashed analysis or a result that looks like origami gone wrong.

Self-contact appears anywhere a part undergoes enough deformation that distant regions of its own geometry approach each other: sheet metal stamping, cable routing, rubber seals compressed into grooves, buckling panels. In every case, the part needs to know its own boundaries well enough to prevent self-intersection.

In Abaqus, self-contact is typically included through general contact with the appropriate self-contact inclusions. The solver treats it like any other contact—the same search algorithms, the same enforcement methods—but applied within a single body rather than between two. The added computational cost is real but necessary. Without it, you are not just getting the wrong answer. You are getting a physically impossible one.

Coupling Constraints

These are the interpreters and translators of the interaction world. A coupling constraint connects a reference point to a surface and transfers loads or constraints between them.

Kinematic coupling (RBE2 in Nastran terminology) is the authoritarian approach. It forces all coupled nodes to move rigidly with the reference point. The surface cannot deform independently. This adds artificial stiffness and should only be used when the physical connection is genuinely rigid relative to the surrounding structure.

Distributing coupling (RBE3 in Nastran) is the democratic approach. It distributes loads from the reference point to the coupled nodes without adding stiffness. The surface remains free to deform naturally. This is physically more appropriate for most load application scenarios.

Multi-Point Constraints and Equations

These are the custom contracts of the interaction world. They let you define arbitrary mathematical relationships between degrees of freedom at different nodes. Need one node’s rotation to drive another node’s translation? An equation constraint can do that. Need a set of nodes to maintain a specific geometric relationship? An MPC can enforce it.

These are powerful tools, but they require careful thought because they directly modify the equations the solver is working with. A poorly conceived constraint can create a singular stiffness matrix, prevent rigid body modes from being properly captured, or introduce artificial loads that have no physical basis.

Chapter 4: Material Self-Knowledge

Before a part can interact with anything else, it must know itself. And in finite element analysis, self-knowledge means material properties.

Think about the human parallel again. A person’s response to any interaction depends fundamentally on their own nature. Are they resilient or brittle? Do they bend under pressure or snap? Can they absorb energy or do they transfer it? Do they behave the same way under gentle pressure as they do under extreme stress?

Every one of these human qualities has a direct material property counterpart.

Young’s modulus is confidence. It determines how much a material resists deformation under load. A high modulus material like steel stands firm. It takes enormous force to change its shape. A low modulus material like rubber yields readily. Neither is better. They simply have different roles in the assembly.

Poisson’s ratio is empathy. When you compress a material in one direction, it expands in the perpendicular directions. Poisson’s ratio quantifies how much a material’s response in one axis is affected by what is happening in another. A material with a Poisson’s ratio near 0.5, like rubber, is highly empathetic—squeeze it here, it bulges there. A material like cork, with a Poisson’s ratio near zero, is stoic. Compress it and the deformation stays local.

Yield strength is patience. It is the threshold beyond which a material’s behavior permanently changes. Below yield, the material is elastic. It springs back. It forgives. Above yield, the material plastically deforms. It remembers the insult. The damage is done. In human terms, everyone has a point beyond which the relationship changes permanently. Knowing where that threshold lies—for both the material and the person—is critical to predicting outcomes.

Density is presence. It determines how much mass occupies a given volume, and in dynamic analysis, mass is everything. A dense part has inertia. It resists acceleration. It dominates vibratory response. In an assembly, the densest parts often drive the modal behavior. They are the heaviest voices in the room.

Chapter 5: The Quality of the Interface

The quality of any interaction—human or mechanical—depends not just on the nature of each participant, but on the properties of the interface between them. In finite element terms, this is the surface interaction property definition, and it controls everything about how forces are transmitted across the boundary.

Normal behavior defines what happens when two surfaces push against each other. Hard contact means no penetration is allowed. Period. The surfaces are impenetrable walls. Softened contact, or pressure-overclosure relationships, allow the interface to have its own compliance. This is physically more realistic for many situations—a gasket, an adhesive layer, surface roughness, coating thickness—all of these create a zone at the interface that has its own force-displacement character.

Tangential behavior is where friction lives. The friction coefficient is not just a single number. It can vary with contact pressure, sliding velocity, temperature, and accumulated slip distance. Static friction—the resistance to initial motion—is typically higher than kinetic friction, the resistance during sliding. This difference is what causes stick-slip vibration: the squeal of brakes, the chatter of a machine tool, the creak of a door hinge.

Thermal conductance across the contact interface determines how heat flows between parts. Parts in firm contact transfer heat efficiently. Parts barely touching create a thermal bottleneck. This is why thermal grease exists in electronics. The air gap between a processor and a heat sink, even if it is only a few microns, has terrible thermal conductivity. The grease fills that gap and dramatically improves the thermal interface.

Cohesive behavior models adhesion. This is the interaction type for bonds that can form and break. Unlike friction, which only resists relative motion, cohesive behavior can resist separation. It can model glue joints, solder connections, delamination in composites, and interface cracking. Cohesive surfaces have both a strength—the maximum force they can sustain—and a toughness, the total energy required to fully separate them. A strong but brittle bond snaps suddenly. A weaker but tougher bond peels gradually.

Chapter 6: Getting It Wrong

Because interactions are so foundational and yet so easy to overlook, the failure modes deserve their own discussion.

The most dangerous interaction error is not a crashed solver. It is a clean run with wrong answers. And the most common cause of wrong answers is missing or incorrect contact definitions.

Consider a bolted assembly where you forgot to define contact between two flanges. The solver will happily let the flanges pass through each other. The bolt loads will distribute incorrectly. The stress field will be wrong everywhere downstream. And if you are only looking at peak stress in a region far from the joint, you might never notice that the contact was missing. The numbers look reasonable. They just are not right.

Another common mistake is using tie constraints where contact should be defined. This is the over-committed relationship. You have permanently bonded surfaces that, in reality, can separate or slide. The model becomes too stiff. Natural frequencies shift upward. Stress concentrations appear at the edges of the tied region that do not exist in the real hardware. And under load reversal, the tied joint cannot open, so you miss a critical failure mode entirely.

The opposite mistake is equally dangerous. Using contact where a tie would be appropriate adds unnecessary nonlinearity to the model. The solver must track contact status at every increment, convergence becomes harder, and run times increase dramatically—all for an interface that never separates or slides in reality. This is engineering conservatism misapplied. You are not making the model more accurate. You are making it more expensive and potentially less stable.

Incorrect master-slave assignment shows up as either convergence failures or penetration that the solver does not detect. If the slave surface is too coarse, its nodes may pass between the master surface’s facets without triggering the contact algorithm. This is the equivalent of two people trying to shake hands in the dark—they keep missing each other, and the solver reports everything is fine.

Chapter 7: Best Practices

After decades of working with contact and interaction definitions across consumer electronics, portable devices, and complex assemblies, here are the practices that consistently produce reliable results.

First, always verify your interactions visually before submitting a job. Display contact pairs. Check surface normals. Confirm that master and slave assignments make physical sense. Five minutes of visual inspection saves hours of debugging wrong results.

Second, start simple and add complexity. Begin with tie constraints to verify that loads are flowing through the assembly correctly. Then replace ties with contact where relative motion matters. Then add friction. Then add thermal conductance if needed. Each addition is a step up in complexity and computational cost. Make sure each step gives you the right behavior before adding the next.

Third, pay attention to units. A friction coefficient of 0.3 is dimensionless and works in any unit system. But thermal conductance, cohesive stiffness, and pressure-overclosure relationships all have units. A thermal conductance value that is correct in SI will be wrong in millimeter-tonne-second. And unlike a material property error that might produce obviously wrong stress values, an interaction property error can produce subtly wrong results that pass casual inspection.

Fourth, understand that every interaction choice is an engineering judgment, not just a software setting. When you assign a friction coefficient of 0.2, you are making a statement about surface finish, lubrication, temperature, and loading rate. When you choose hard contact over softened contact, you are making an assumption about interface compliance. When you tie instead of defining contact, you are accepting the consequences of a rigid joint assumption. Own those decisions. Document them. Revisit them when results do not match test data.

Fifth, validate against physical understanding. If two parts should be able to separate under load reversal, verify that they do. If a bolted joint should transfer shear through friction, confirm that friction forces in your model match the expected clamping force times the friction coefficient. If a gasket should compress under bolt preload, check that the contact pressure distribution is reasonable. The solver will give you answers regardless. It is your job to ensure those answers reflect reality.

Chapter 8: When the Relationship Needs Counseling

Chapter 6 covered the conceptual mistakes. This chapter is about what you actually see on your screen when something goes wrong, and what to do about it. Think of it as relationship counseling for your model.

The Solver Diverges Early

This is the awkward first encounter. The parts came in too hot. The most common cause is initial overclosure, where your geometry places surfaces slightly inside each other before the analysis even begins. The solver tries to push them apart, generates enormous forces, and the solution spirals out of control. The fix is to adjust the initial contact state. Let the solver resolve small overclosures automatically, or clean up the geometry so surfaces start just touching or with a controlled small gap. If you see massive contact forces in increment one, this is almost always the culprit.

Unrealistic Stress at Interface Boundaries

This is the overcommitted relationship, where you have used a tie constraint or a kinematic coupling, and the edge of the constrained region shows stress spikes that do not exist in test data. The problem is artificial stiffness. The perfectly rigid interface creates a discontinuity in compliance that the mesh resolves as a stress concentration. The fix depends on the situation. If the interface should have compliance, replace the tie with contact. If the coupling is the issue, switch from kinematic to distributing. If neither is practical, at minimum recognize that stress values within one or two element lengths of the constraint boundary are unreliable and should not be used for design decisions.

Parts Pass Through Each Other

This is the silent failure, and it is the scariest symptom because nothing in the solver output flags it. It happens when the slave surface mesh is too coarse relative to the master, when the master-slave assignment is backwards, or when contact was simply never defined for that pair. The only way to catch it is visual inspection of the deformed shape. Make it a habit. After every analysis, display the deformed model at full scale and look for interpenetration. Rotate the view. Zoom into joints. If two surfaces occupy the same space, you have a problem that the solver was perfectly happy to ignore.

Run Times Dramatically Longer Than Expected

Not every long run is a problem, but when an analysis that should take hours takes days, contact is often the reason. General contact with many surfaces means the solver is searching for potential interactions everywhere, every increment. The fix is specificity. Replace general contact with explicit contact pairs for the interfaces that matter. Exclude surfaces that will never interact. In large assemblies, this bookkeeping is tedious, but the payoff in run time can be enormous. The solver should be spending its time on physics, not searching for relationships that do not exist.

Convergence Oscillation

This is the on-again, off-again relationship. Nodes chatter between open and closed contact status, and each status change flips the stiffness matrix. The solver takes a step, contact opens, so it takes the step again, contact closes, so it tries again, and the cycle repeats. Damping helps. Contact stabilization applies a small viscous force that smooths the transition between open and closed states. But monitor the stabilization energy. If it becomes a significant fraction of the total strain energy, you are not stabilizing the solution. You are changing the physics.

Closing

Interactions are where the model becomes an assembly and where the assembly becomes a system. They are the relationships that determine whether your simulation captures the real physics or merely produces colorful pictures of something that never existed.

The parts in your model are not like us. They cannot sense their surroundings, read context, or adapt to unexpected situations. Every relationship must be explicitly defined. Every interface property must be deliberately chosen. Every master-slave assignment must be physically justified.

But here is the deeper lesson. The same principles that make human interactions succeed—awareness of each participant’s nature, appropriate proximity, clear rules of engagement, understanding who leads and who follows, and honest assessment of friction—apply with mathematical precision to finite element contact.

The analyst who understands interactions at this level does not just set up contact pairs. They engineer the communication pathways through which forces, heat, and constraints flow through an assembly. They understand that the interface between two parts is not empty space. It is a physical entity with its own properties, its own behavior, and its own potential to control the outcome of the simulation.

Get the interactions right, and your model talks to you. Get them wrong, and it lies to you.

And it does so very politely, with perfectly formatted output files.

Glossary

Cohesive Zone Model  A method for modeling interface debonding using traction-separation laws. Defines both the peak stress an interface can sustain and the total energy required for full separation.

Contact Pair  An explicit definition of two surfaces that may come into contact during an analysis. Allows separation, sliding, and friction between the surfaces.

Contact Stabilization  A small viscous damping force applied during contact to smooth transitions between open and closed contact states and improve convergence. Must be monitored to avoid altering the physics.

Distributing Coupling (RBE3)  A constraint that distributes loads from a reference point to coupled nodes without adding stiffness. The coupled surface remains free to deform naturally.

Friction Coefficient  A dimensionless value defining the ratio of tangential (friction) force to normal (contact) force at an interface. Varies with surface finish, lubrication, temperature, and sliding velocity.

General Contact  An automatic contact definition that includes all exterior surfaces in the model. Convenient for large assemblies but computationally expensive due to global contact searching.

Hard Contact  A normal contact enforcement that permits zero penetration. Mathematically strict but can cause convergence difficulty in problems with rapidly changing contact status.

Kinematic Coupling (RBE2)  A constraint that forces all coupled nodes to move rigidly with a reference point. Adds artificial stiffness; appropriate only when the physical connection is genuinely rigid.

Master Surface (Primary)  The surface in a contact pair that defines the geometric reference. Typically the stiffer or coarser-meshed surface. The solver enforces non-penetration from the slave side against this surface.

Multi-Point Constraint (MPC)  A mathematical relationship linking degrees of freedom at different nodes. Used to enforce geometric relationships or transfer motion between parts of a model.

Overclosure  Initial geometric interference where two surfaces overlap before the analysis begins. Can cause large artificial forces in the first increment if not resolved.

Penalty Stiffness  A numerical spring constant used in the penalty contact method. Allows a small, controlled amount of penetration proportional to the contact force. Too high causes instability; too low permits excessive overlap.

Pressure-Overclosure  A relationship defining contact pressure as a function of surface penetration. Used in softened contact formulations to model interface compliance such as gaskets, coatings, or surface roughness.

Search Tolerance  The maximum gap distance within which the solver will detect potential contact between surfaces. Surfaces farther apart than this distance are ignored by the contact algorithm.

Self-Contact  Contact defined within a single part, allowing different regions of the same body to interact during large deformation. Essential for folding, buckling, and stamping analyses.

Slave Surface (Secondary)  The surface in a contact pair whose nodes are checked for penetration against the master surface. Typically the softer or finer-meshed surface.

Softened Contact  A contact formulation that allows a controlled amount of penetration with a pressure-overclosure relationship, providing interface compliance rather than rigid enforcement.

Stick-Slip  Oscillation between static and kinetic friction states, causing vibration. Occurs when the static friction coefficient is higher than the kinetic friction coefficient.

Tie Constraint  A permanent, rigid bond between two surfaces. No relative motion or separation is permitted. Simplest interaction type but introduces artificial stiffness at the interface.

About This Series

This companion reader is part of the FEA Best Practices audiobook series from McFaddenCAE.com. The series covers the full arc of building and interpreting finite element models, from unit systems and material definitions through modal analysis, meshing decisions, and the engineering judgment that holds it all together.

Other titles in the series include Building the Model (Volumes 1–4), which covers unit systems, material assignment, mesh quality, and the common pitfalls that produce wrong answers from correct-looking models, and Modal Analysis in Structural Dynamics, which covers eigenvalue extraction, boundary condition sensitivity, coupling selection, joint stiffness calibration, and damping estimation.

All audiobooks and companion readers are available at McFaddenCAE.com.

Generated by Joe McFadden, The Holistic Analyst

with Claude and ElevenLabs

McFaddenCAE.com