Analysis solutions incorrect

[quinnsriley]

Hi all,

I am trying to do a simple static structural analysis in Fusion 360, but after thorough troubleshooting I am finding Fusion 360 incapable of producing an accurate result after being compared to other FEA software and hand calculations.

I am trying to calculate the max stresses in a 10inx5inx1in solid rectangular prism with a .75 diameter hole in the middle. On one short skinny side is a fixed support and the other a 100 lbf load facing away from the center of the prism, perpendicular to the plane.

I ran simulation various times with fine mesh, and consistently got ~ 30 psi.

The hand calculations and Ansys both confirm in this case that the max stress in the stress concentrating feature of the hole should be ~60 psi.

I have been very thorough and I am at a loss. I am surprised there would be a discrepancy this large in values between identical setups. I am weary to using this program for further product development and analysis if a simple box cannot be analyzed with reasonable accuracy. 

Any revelations regarding this issue would be greatly appreciated, as I would like to stick with the fusion platform.

[mohammed.aliTDPXK · Accepted solution]

hello @quinnsriley ,

What I see here is you have compared Fusion's results with some other software.

I would like to highlight few points here,

The accuracy of the FEA result depends on :

• The mesh size : The thumb rule that, the finer the mesh the more accurate the results are and also it depends on whether it is linear mesh or parabolic mesh, Parabolic meshes are more accurate compared Linear
• The type of element : Fusion has tetrahedral elements and you are comparing with the results with hex elements, the accuracy is always higher for hex mesh than let
• Mesh refinement at critical areas : the critical area such as the hole in the centre of your model, needs to be refined or mesh flow needs to be finer to capture the stress variation smoothly

Considering the above points I request you to do a convergence study :

• Start by decreasing the element size by 10 %
• Refine the mesh at the hole (Use local mesh control option)
• Make sure parabolic mesh is selected in mesh settings

Repeat this by again decreasing the element size until you arrive at a point where the difference between the previous result is about 5 to 10%.

Thanks!

Let me know if this helps and do like the post and accept solution if this answers your query.

[dsouzasujay · Accepted solution]

Thank you @mohammed.aliTDPXK for providing details.

I can see you have used Hex elements in Ansys, We cant compare hexahedral with tetrahedral Elements
Hexahedral elements result in more accurate results at lower element counts than tetrahedral elements. 

(Adam has Already covered)

Below is the study gave me  ~ 60 psi with the same boundary conditions.
(Can you please share the *.f3d file for which you got ~ 30 psi Result)

[quinnsriley · Accepted solution]

Thank you for your timely responses

I understand the need for refining the mesh, this is something I have done in past iterations with the same result, even reaching convergence with a very high node count. 

It seems that you all have the solution. Please take a look at the attached f3d file for review

Let me know what seems to be the issue so I may better use this program in the future

EDIT: I revisited the mesh, as that is what you guys mentioned to go over. That was the issue. After looking at the mesh settings I was able to replicate the results getting 60 psi. I guess I need to be more thorough with my mesh and make sure it flows.

ALSO now I understand the difference between hexahedral and tetrahedral elements. Very enlightening.

Like always the user, not the tool, was at fault. Thank you very much for the help.