S4R elements

Fatmac

I have been working thru a few shell examples from this document https://hub.hku.hk/bitstream/10722/54269/2/92945.pdf to get a better feel for the shell elements used in Calucix - particularly with non-linear analysis. I have set up a solid model as a control and have been playing around with different shell elements. With S4 elements the solution looks reasonable compared to the abaqus results presented in the paper

But, from what I understand, S4R elements were used in the study, . When I try to use S4R the deflections are significantly less -perhaps I am not setting it up correctly?
any ideas

Fatmac

here is a picture of the set up, the papers results and the S4 element results

disla

No problem with your set up. This is a known limitation of the calculix S4R formulation. There is a strong shear MODE III present in the benchmark which S4R (Expanded C3D8R) is not suitable for.

Fatmac

Disla - Thank you for the answer. My knowledge is so limited I had to look up what MODE 111 meant - if my source was correct then
Mode I = opening

Mode II = in‑plane shear

Mode III = out‑of‑plane shear / tearing

the fact that the calculix S4 elements compared well to the published results but the S4R elements appeared next to useless on (what appeared to me to be a reasonably dense mesh) gives me little confidence in S4R elements all round - should I generally try to avoid using them in calculix, or are there simple ways to explain what type of analysis they will work ok

disla

I don't use S4R at all. I have also found discrepancies with reference benchmarks when used for Buckling analysis.

Fatmac

thanks again for the heads up - sounds like i will cross them off my list of elements to use

disla