FEA problem solving- always begin from something simple

There are two kinds of problem in the world of FEA:

easy ones and difficult ones.

Haha…I’m joking. My point is that the difficulty to solve a particular problem depends on lots of factors that beginners in FEA may not know. And it makes them difficult to estimate the difficulty of their problems or even lead them to a totally wrong way. So what are the factors which play important roles on the level of difficulty?

 

1) The types of FEA Analysis

The type of analysis to perform depends on the results we want to obtain. If you want to determine the deformation due to a force, the stress due to temperature or the energy released after an impact the method and the type of analysis employed will be different of course!

questions useful to ask yourself are:

Am I studying a linear problem or a non-linear problem?

Consider a problem as linear is a big simplification that makes everything easier for the analysis. The problem? the world is not linear… everything is nonlinear. So what are the hypothesis of a linear problem?

– The relation between stress and strains should be linear… to speak with simpler words, stress and strains should be “proportional” and this proportional constant is the Young’s modulus (in case of 1-dimensional model)

– Deformations should remain very small (otherwise stress-strain relation cannot be linear)

– The direction of loads applied remain the same during the analysis

if you have all that, you are good for linear analysis ! Otherwise, you will have to plunge into the nonlinear world… more interesting, but also much more difficult to understand and perform correctly analysis for a beginner.

In nonlinear Analysis, you can consider large deformations and strains, nonlinear materials and nonlinear contacts. It will be the subject of another article on this blog later.

Am I studying a static problem or a dynamic problem?

static problems are always more simple to solve than dynamic ones because in static problems the time can be neglected and is basically ignored in the analysis. On the contrary, if you realize that time has to be considered (transient analysis, impact analysis, see my previous blog post), then you will need to perform a dynamic analysis, which usually requires more theoretical knowledge than simple static problems and also more…time ;-)

2) The detail and quality of the CAD model

CAD designers tend to believe that more the model is detailed and better it will look… and yes, it will LOOK better, but it will be also more difficult to simulate because the FE Analyst will have to simplify the model afterwards to get an accurate “mathematical model”.

In addition, more the model has small details and more mesh you will need to simulate it. When you increase the number of mesh, the calculation speed falls and the analysis becomes difficult to perform.

So, as a rule of thumb: simplify, simplify and…simplify.

Another important factor is the “quality” of the CAD model. If your CAD model has some small edges and small faces, it will become very difficult to mesh, so again, simplification is the rule (see the previous blog post on CAD model simplification)

In the next blog post, I will explain the workflow to solve efficiently an FEA analysis problem considering all that.