PCOMP to PSHELL model correlation?

[Ellzy]

Hi,

I have two models of laminate cone structure. One model is made of PCOMPS and one model is made of PSHELLS with laminate properties. Both models have identical elements (QUAD4s), boundary conditions and loads...only the element property and material cards are different. When I look at stress plots of the two models they do not seem to correlate well. The PCOMP stress plots/gradients make sense for the loading and BCs when maxes from all plies are plotted but the PSHELL stress plots/gradients do not make sense when plotting the max stresses at Z1 and Z2. The max stress values and locations are also different. I cannot figure out why this is happening because I expected good correlation. I checked the punch file and I'm pretty sure the PCOMPs are being properly converted to PSHELLs and MAT2s. If I add PARAM,NOCOMPS,-1 to the PCOMP model, the stress plots/gradients, and max values exactly match the PSHELL model. It seems good that the models match when using PARAM,NOCOMPS,-1 but like I said the plots using that or the PSHELL don't seem to make sense like the ones using the PCOMP do. Any idea what I could be missing? Thanks for any advice!

 

[SWComposites]

You are probably comparing ply stresses (from PCOMP) and smeared laminate/plate stresses (from PSHELL) which are NOT the same.

 

[Ellzy]

Thanks for the quick response SWComposites! That makes sense...I don't know why I had such a brain fart. I'll have to look at my results again a bit closer but that might be it.

 

[Ellzy]

The ply stress vs laminate stress was part of the difference but the biggest reason for the plots not correlating was because I was not plotting the stresses in the material coordinate system for the quad4 elements.