bolted metal plate strips - wrong stress on shell

Dear ladies and gentleman.

I have a problem with my structural analysis.

I want to simulate 2 metal plates which are bolted together, as you can see in the following picture.

The one end is fixed. On the other end there is a force into the direction of the plate strip (5000N). There is also a force in the direction of the bolt (5000N) on the upper und on the bottom side of the plate.

At first I made a hand calculation. The section of the plate is 40mmx5mm.

So I get a normal stress on the upper side of the left plate of about -125N/mm^2 and a normal stress on the bottom side of about 175 N/mm^2.

The right plate shoud have stress about 25 N/mm^2.

I have modelled this at first with 3D elements and got also these results.

Now I wanted to model it with shells and a beam instead of the bolt. But now I get results, which can not be true.

You can see my geometry at the picture bellow:

I connected the circular ring shell, wich presents the bolthead with the plate with a bonded contact. Also the 2 rectangular plates which presents one plate are bonded together. The areas of the 2 plates which are bolted together are connected with a frictional contact. The endpoint of the beam is connected with the "bolthead" with a fixed joint.

At the first step I obtained the bolt pretension and on the second step I obtained the forces on the plate.

So I get normal stresses as you can see bellow:

The results on the right plates are correct. But I think my results for the left plate are incorrect. But I have no idea what I did wrong.

Maybe I need more integration points for the shell? But unfortunately I did not find, where I could change this in ANSYS?

I am really looking forward for answers and ideas.

Thank you very much



  • peteroznewmanpeteroznewman Member
    edited November 10


    Open the Geometry in SpaceClaim, go to the Workbench tab and use the Share Button. That will make the circular edge on the outside of the face used to represent the bolt head to share nodes with the rest of the surface of the plate. That will let you delete all the Bonded Contact, which will help simplify your model.

    Did you watch my video series?

  • Thank you very much for your answer.

    The deformation looks a lot smoother, as you can see in my following picture. Alos the stress peaks at the bonded sections are gone

    But there is still a big difference to the normal stress of the 3D modell as you can see in the following picture.

    I have turned lager deflection off in both cases, so I can better compare it to my hand calculation.

    The direction of my forces are definded by components.

    Yes, I have already seen your videos.

    Thank you therefore, they are great!! I built my modell like yours.

    Could it be possible, that I need more integration points in shell thickness direction?

    Thank you very much for your help

    Best regards


  • @bokaJ

    The solid model is built along the X axis while the shell model is built along the Z axis. Can I trust that the geometry, loads and supports are identical?

    Once two models are laid out with the same conditions, the difference between the stress in the shell elements should be comparable to the solid elements.

  • Thank you for your answer sir.

    I am sorry, that it is a little bit confusing because the coordinate system is not the same in the two examples.

    So i did I did it again and used the same coordiante system therefor.

    Above you can see the dimensions of the model. I prooved it again, the dimensions of the two models are the same.

    At first you can see my settings for the solid model:

    I did not share the topology in the geometry, So I can give every sinlge subsection of a part a mesh which fits best.


  • _____________________________________________________________________________________________________________________________________________________________________________________

    shell and beam model:


    beam settings:


    In this case I shared the topology in the geometry tool:


  • forces and boundary conditions:

    analysis settings:

    normal stress in z-direction:

  • The result of the 3D analysies is close to my hand calculation(25N/mm^2, 175 N/mm^2, -125 N7mm^2). But the result of the beam and shell model is far away from the others.

    If I turn on large deflection I get the following results. In this case these two reluts are in my opinion also too fare away from each other.

  • I hope that I provided all the information to you which you need to help me.

    If you need further information, please ask me.

    Than you very much sir!

  • mesh:

    forces and boundary conditions:

    analysis settings:

    normal stress in z-direction:

  • @bokaJ

    On the Shell model, try plotting Top and Bottom separately. When you plot Top/Bottom, how do you know which side of the plate is being plotted.

  • thank you for your answer sir.

    unfortunately it does not change my result. The stress is still the same

Sign In or Register to comment.