EL PRINT and Von Mises equivalent stress

General Questions
1

Hi,

I feel stupid to ask for the following : how to “output” the Von Mises stress (using EL PRINT / History output)? I also had a look in the Calculix doc, but I’ve found nothing (may i’ve missed). I was thinking into the “Mises” keyword but i was wrong.

An intermediate solution is to calculate “by hand” the scalar using the output stress tensor components, but the (nodal) values have ever been extrapolated/smoothed from the integration points = less accurate results.

Paul

2

CalculiX doesn’t calculate von Mises stress even for field output - it’s up to postprocessors to do that and PrePoMax does it as well. In the case of history output, you would have to calculate von Mises stress manually using a proper formula. In PrePoMax, you can use Results → History Output → Create → From History Output by Equation for this purpose. Since *EL PRINT gives you integration point stresses, they should be sufficiently accurate.

3

Actually, the history output in PrePoMax should also compute the Mises stress, but there was a bug preventing that. I fixed it, and it will be available in a couple of days when I release a new developer version.

4

cxx does not differ from other FEA solvers:

  • it calculates the stress quantities at the gauss points and then extrapolates the values at the nodes using smooth functions
  • it is explained into the ccx doc.

A basic example using a C3D20 (full integrated) elements outputs 20 values and not 27 (see attachement); in other word, we are calculating the equivalent stress on values ever extrapolated (then not so accurate than the ones at the integration points).

Note this limitation comes from ccx which does not calculate VM stress at the Gauss points prior to do the extrapolation, and not prepomax which reads the frd file (i guess)

Thanks all for the feedback and the hints

output
output1113×597 31.9 KB

cube.pmx (53.3 KB)