Guidelines for Interpreting Results

In structural simulations you are usually interested in viewing plots of deformed shapes and stress distributions; for thermal simulations, temperature contour plots and heat flux vector plots; for frequency studies, fundamental frequencies and mode shapes.

The final step of the FEA process involves the visualization and interpretation of results. The FEA results should be qualified in terms of accuracy and correctness. It is important to verify that the results are accurate and fulfill the original design goals of the analysis.

Results accuracy relates to the convergence level and quality of the solution method. It is good practice to assess the correctness of results through the validation of all your modeling assumptions.

When viewing FEA results:

  • Check the displacement results. Is the order of magnitude of the deformations what you are expecting? Unexpected discrepancies could be caused by an inconsistent set of units and improper loading definitions. If the displacement magnitudes are so large that the linear assumption is violated, consider running a nonlinear analysis. Is the overall deformed shape in agreement with the applied boundary conditions and loading definitions? Animations of the displacement results from the model’s initial undeformed state to its fully deformed shape help you visualize the response of your model and detect any modeling errors.
  • Check the stress results. Are the magnitudes of the stress results in line with what you are expecting? Examine the regions of high stress concentration. Are these caused by “bad” quality mesh elements? Consider refining the mesh locally at areas of high stress concentration to resolve any convergence issues. Eliminating insignificant geometry features (such as narrow faces, sharp corners, or short edges) can also help to eliminate fictitious high stress values. Verify that the transition of stress values is smooth throughout the geometry.

    Stress results are generally used to predict yielding or guard against failure. Stress quantities are related to the failure criteria that best describe the particular material. The most popular stress quantity to display when evaluating the onset of yield for ductile materials is the von Mises stress. The material starts yielding when its von Mises stress reaches a critical value known as the yield strength.

  • Verify that the reactions at the supports balance the applied forces.
  • Review and qualify your modeling assumptions. Keep in mind the inherent assumptions of each input variable in your FEA model (loads, restraints, material properties, element formulation, solution method). When you interpret results, review these assumptions and try to quantify their effects on the solution. You must have a good understanding of the mechanics of materials, the potential failure modes of the products, and the product's actual operating environment.
    Note:

    The consideration of the proper restraints (fixtures, rollers, or hinges) that define how the model interacts with the surrounding environment is a challenging task, which carries significant modeling uncertainty. Over-simplification of the true support structure often leads to erroneous results. Small changes in the support conditions can cause large changes in the results.

  • Verify that the solution satisfies the original design intent. The setup of an FEA model serves a design purpose; for example, minimize deflection at a certain location, control oscillation amplitudes, reduce weight of a component, or predict possible failure modes. It is important to succeed in getting meaningful result quantities that best serve your original design goals.