I was running a simulation with success including gravity and applied loads. I was checking both stresses and deflections. Both were in line with what I was expecting. The next day I went to change a few dimensions and run it again. The stresses and deflections were significantly less than expected (I did not tweak the dimensions very much). I changed my dimensions, remeshed and ran it again. It worked fine and returned expected results. I continued in this manner tweaking the size of my box. Finally I went for one last run.... it didn't work again. It returned the significantly less than expected results. A bit of investigating revealed that it is only applying the gravity loads during the simulation. All of the applied loads are being disregarded (however, they show up in the report as if they are being applied.) If I supress the gravity my stress and deflections return 0 values. I have tried taking my loads off and reapplying them. I have tried removing the gravity load. I have remeshed. I have inputted my loads and the gravity in different orders. I cannot get my simulation to calculate using my applied loads. It was working fine.... what happened?
Molly Wood
Inventor Professional 2012
Hi Molly,
I haven't encountered this issue before. Would it be possible to post the dataset to take a closer look?
Thanks, -Hugh
Hugh,
Attached is the simulation we are trying to run with the correct live loads and gravity applied. I am expecting a deflection between 0.6 and 0.7 inches when all of the loads are applied correctly. When only gravity is applied I've been seeing a deflection around 0.14in.
Please let me know if you have any more questions or if my Pack n' Go didn't work.
Molly
Hugh,
Here is the report of the last simulation I ran. As you can see each of the applied loads are listed as well as gravity but the deflection shows only 0.14inches.....
I hope this helps.
Molly
Hi Molly,
Actually, I have seen this issue before. The small split faces are not being included in the mesh, which makes the Force loads not take effect. Development is actively looking into the solution.
There is a workaround for now, of splitting the long part into multiple bodies (updated model files attached). Bonded contacts will automatically be made between the bodies, and the results should be equivalent to a simulation without the splits.
I hope this helps and we apologize for the inconvenience.
Sincerely, -Hugh
Thank you. I appreciate the insight into how it is functioning. I will try breaking it up. I wonder why it worked for so many simulation before it "broke"?....
Thanks.
Molly