In the previous two editions of ‘How to Improve an FEA Model’, we discussed how model simplification and proper mesh generation can help facilitate an improved finite element analysis (FEA).
In the third edition of the series we will discuss the importance of choosing accurate load applications as it relates to the real-world environment the object will likely encounter in its lifecycle.
Determining proper load applications is an extremely important step in an FEA. Load applications are the model inputs that the object is being tested for, such as a specific event like a thermal cycle, shock from a drop, vibration, or static flexure. Understanding the nuances of how to apply the loads are essential in order to simulate an event that the object will face in a real-world environment.
One common example is determining whether loads applied should be applied as static or transient. For example, if an engineer is simulating the flexure of a structure during assembly, it may be acceptable to model the load as a static displacement since strain rates are likely to be much slower and results time-independent. However, if an engineer is modeling a similar deflection caused by dropping the same assembly, they would likely need to use a transient model to capture the associated inertial effects, since the application time of the load is much faster and time-dependent effects must be captured.
In the electronics simulation world, we often deal with a similar case when simulating thermal cycling. For example, when investigating the thermal expansion and associated stresses on an entire printed circuit board assembly (PCBA) during thermal cycling, the material properties in the analysis can likely all be linear approximations, and static ramps to the minimum and maximum temperatures to investigate those stress states with no dependence on time can be reasonable. This is acceptable when board-level displacement and elastic stresses/strains are the focus of an analysis rather than creep strains/energies. However, when investigating component-level solder fatigue, solder creep properties must be included and it becomes important to accurately apply the ramp and dwell times of the thermal cycle. Creep models include time-dependent properties, so the simulated cycles must be modeled in their entirety to most accurately calculate creep strain/energy results that are used to make solder fatigue predictions.
The same real-world event is not always equal in the FEA world depending on the desired outcome of the analysis. It’s important to always keep in mind the real-world stressors the object will likely face and how those stressors could affect the component of interest. Inputting these nuances properly will result in an analysis that is accurate, valid, and actionable.