Material Behavior and CAE: An Introduction
This series of topical overviews focuses on the crucial, yet often overlooked, influence of material behavior on the accuracy and effectiveness of finite element analysis (FEA) and computer-aided engineering (CAE) more generally. As computational tools and modeling techniques have advanced, the spotlight has increasingly turned to material models — the mathematical descriptions of how materials respond under various physical conditions — and their critical role in ensuring realistic simulations.
To learn more regarding material testing and characterization in support of material models for CAE, see our TestPaks or contact us to talk with our materials testing experts.
Understanding the Role of Material Behavior in FEA
Historically, engineering simulations have relied on simplified assumptions about material properties. These assumptions, largely stemming from traditional metallurgical approaches, treat materials as if they have fixed, consistent properties. While this may work adequately for metals in limited contexts, the emergence of modern materials — such as polymers, rubbers, foams, and composites — demands a more nuanced understanding. These materials often exhibit significant variation in behavior depending on temperature, loading rate, environment, and other factors.
To account for these complexities, FEA must utilize accurate material models that reflect how a material will behave in real-world service conditions. For instance, a rubber boot in an automotive constant velocity (CV) joint might be exposed to wide temperature swings, mechanical strain, cyclic loading, and chemical exposure all at once. Creating a simulation that meaningfully predicts performance in this context would require an intricate model that integrates hyperelasticity, temperature dependence, and possibly environmental degradation.
Because fully characterizing such material behavior is often impractical, engineers usually adopt a pragmatic strategy: use the simplest material model that captures the most critical performance aspects of the component. This balance helps avoid overly complex simulations while still producing useful, actionable results. Early, thoughtful selection of a suitable material model can dramatically improve simulation accuracy and reduce the need for costly design iterations.
Material Models: A Mathematical Foundation for Simulation
Material models are built from constitutive equations — mathematical expressions that describe how a material will behave under certain applied conditions, such as loading. FEA software packages typically offer a variety of built-in models, each suited for specific material types and conditions. When using a model, the analyst must supply material-specific parameters, some of which are readily available in datasheets or handbooks for well-characterized materials. More often, however, these parameters must be derived from laboratory testing.
The relationship between materials testing and FEA modeling is not always straightforward. Many FEA analysts are not deeply versed in materials science, while testing laboratories may not be familiar with the requirements of material modeling. This disconnect can result in miscommunication, inconsistent terminology, and ambiguity in model creation and interpretation. Two analysts working with the same material could easily arrive at different conclusions based on how they interpret the model requirements or test data.
Evolution and Limitations of Material Modeling in FEA
The development of nonlinear material models has lagged behind the growing use of complex materials. Traditional FEA tools were initially designed for linear materials such as metals and concrete. As non-metals grew in popularity, software vendors incrementally expanded capabilities to include plasticity, hyperelasticity, and viscoelasticity. Rubber, due to its highly nonlinear elastic behavior, was one of the first materials to receive specialized modeling treatment.
Despite advancements in theoretical models, many FEA users continue to rely on simplified models due to the perceived complexity of more advanced ones. The result can be misleading simulations and potentially risky design assumptions.
For example, a view found among some FEA users is that all nonlinear materials should be modeled using hyperelastic models. While hyperelasticity is appropriate for rubber, other materials may require entirely different approaches. Similarly, some engineers assume that user-defined material models (UMATs) are the only way to handle nonstandard material behavior. Although UMATs offer unmatched flexibility, they require deep expertise, extensive validation, and often costly, complex testing — limiting their widespread use.
Terminology and Fidelity: Barriers to Effective Material Modeling
One of the significant obstacles to accurate material modeling is inconsistent terminology. Materials engineers and simulation experts frequently use different language to describe the same properties. Testing standards such as ISO 472 and IUPAC recommendations attempt to standardize this nomenclature, but FEA software documentation and model definitions often deviate from these standards. The result is confusion, errors in data interpretation, and improper application of material properties.
Beyond terminology, the fidelity of the material model — how accurately it captures true material behavior — is a key concern. While some theoretical models match specific materials well (e.g., Arruda-Boyce for natural rubber), others may fail to capture complex behaviors. For instance, an elastic-plastic model may not adequately describe polymers, which display nonlinear elasticity prior to yield. In these cases, misapplication of standard models can lead to poor simulation accuracy.
Calibration and Data Conversion Challenges
Another barrier to effective material modeling is the process of parameter conversion — transforming raw test data into usable inputs for a material model. This process often requires nonlinear curve fitting and careful selection of parameter bounds. Many simulation codes provide tools to assist with fitting, but expertise is still required to avoid common pitfalls like local minima and instability in model behavior. Improper translation of characterization data to model inputs can lead to slow simulations, non-convergence, or even misleading results. In addition, most material models in FEA software include hidden parameters, flags, and configuration settings unrelated to the raw material data but which must be set correctly as a precondition for simulation success.
Material Model Input File Complexity
Material models can be submitted to the simulation through user interfaces or specialized input files. When dealing with complex models with numerous parameters, file-based input is often preferred — though it introduces challenges of its own. Each simulation code has its own file format and unit conventions, and preparing these files manually or through scripts opens the door to errors. These small complications add up, which may discourage analysts from using the most suitable models and leading them instead to rely on familiar but oversimplified ones.
Demystifying Material Models for Better FEA
The accurate representation of material behavior in CAE simulation is both critical and underutilized. This stems from a combination of factors — inconsistent terminology, limited knowledge of available models, complexities in calibration, and lack of communication between material experts and FEA users.
To overcome these challenges, it is essential to build a better bridge between materials testing and simulation, promote shared terminology, and simplify the process of material model development and integration. By doing so, engineers can significantly improve simulation accuracy, design reliability, and product performance while minimizing trial-and-error cycles and costly redesigns. The ultimate goal is to empower analysts to use the most appropriate models confidently — not just the most familiar ones.
To explore the topics discussed on this page further, see Hubert Lobo (Founder, DatapointLabs) and Brian Croop (CEO, DatapointLabs), Determination and Use of Material Properties for Finite Element Analysis (NAFEMS, 2016), Ch.1.
