Check potential energy of elements after FEA?

[bongirs]

Hello,

In my Finite Element Methods course I learned that the best way to check whether a FEA is precise or not based on the potential energy of the elements. We could do this on paper using the differential equations for simple beams.
But for complex objects, when I do FEA in a software how do I know whether my simulation is stable?
Currently I am using Hypermesh in which there is no option for potential energy.
What could be a better way to do this?

 

[ShellsPlatesMeshes]

The accuracy of the FEA is a broad topic. The best way to evaluate precision of a particular FEA software tool is to compare analytical and numerical solutions for models when analytical solutions are available. The fact is that not only simple beam problems allow such comparison: plate and shell models, sometimes quite complex, also quite often have good analytical solutions. Here are several examples, for plates and shells:

http://members.ozemail.com.au/~comecau/quad_plate.htm

http://members.ozemail.com.au/~comecau/quad_shell.htm

http://members.ozemail.com.au/~comecau/quad_plate_elastic_support.htm

Hope this helps.

 

[stressebookllc]

In general, you need to do basic hand calculation checks for whether the response makes sense, also use the 'epsilon' value, the strain energy error measure.

It needs to be a very small number (usually less than 1.0E-12)

The way I like to think of this is that the difference in energy you apply to the model and the strain energy that is induced in the model is very small indicating that the model is not flying away, or things are not falling apart.

Aerospace Stress Analysis and FEA Courses

http://www.stressebook.com

Stressing Stresslessly!