Hello Bill, shell elements don't compute through thickness stress caused by forces normal to the shell element. They only compute stress from bending or in-plane forces. Shell elements are useful for overall structural response. When you want detailed through thickness stress in the plates caused by the pretension, you have to use solid elements. See this video: kzfaq.info/get/bejne/gtF9hZSZstPGlHk.html

@@ansysfeaexamples9351 Thanks for the response! The plate through thickness stress is not so important. My point is, since shell elements dont have shear deformation (in this case plate compression due to the preload), the elastic resiliences of the bolted joint are not correctly represented. So, why is modelling the plates of the bolted joint as shell elements a valid method to analyse a bolted joint? I suppose for just getting the overall response of the structure it is fine.

Apply separate Bolt Preloads to each bolt. Apply the same value of preload to each bolt in step 1. Apply some other load to the structure in step 2. Add a Bolt Probe to the Solution section of Mechanical. There you can output the Working Load for each bolt to see the individual loads carried by each bolt, which could be different. You could also put a Contact Tool in the Solution branch and plot the Contact Pressure in the interface between the parts. You may see higher pressure around some bolt holes and lower pressure around other bolt holes.

You could have a multi-step solution and apply the pretension in a pattern.

The bolt pretension force clamps the outer plates to the middle plate. Friction between the plates supports any axial loads that the assembly might see along the length of the structure.

I have a video using solid elements. kzfaq.info/get/bejne/gtF9hZSZstPGlHk.html If you have a question on your model, you can post a Discussion at forum.ansys.com/ where it is easier to provide guidance.

Riveted shell plates would be similar but without the bolt pretension.

@@ansysfeaexamples9351 thank you so much i will ask that i couldnt solve rivited plate with this method because of the fact that my computer is restricted and my model iş.very complicated include 198 rivited. Instead of the fact that, Can i modeled this rivits as a beam which has no contact to plates, just their behaviour from rijit to deformable. And this technique give near results without contacts and busy operation. Do you think that operation whic i told is true ?

@@hakansezen2019 Yes, using beams connecting one hole edge to another and no contact is a valid method to model a sheetmetal part that is riveted together.

@@ansysfeaexamples9351 thank you so much, i had needed that info

Yes. Go to forum.ansys.com/ where after you Register, you can post a question in the Structures category where I answer questions.