Optimization of threads in a bolt-nut assembly

[agb253]

I am trying to minimize stress in a screw pullout test by optimizing the geometry (pitch, thread profile, etc.) of a bolt-nut assembly where the threads of the nut are made (using the component operation "Cut Out") by the bolt. The FEA runs fine but the optimization study gives the initial values as the optimal values every time for each geometrical characteristic being optimized.

I DO think the nut is being updated when changes are made to the bolt for these reasons:
when the cut out was performed, if I select "no" for "Support associative placement for the feature," I get warnings of cracks between the nut and the bolt during FEA;
if I select "yes" for "Support associative placement for the feature," I DON'T get warnings of cracks between the nut and the bolt during FEA. (So now I always click "yes")
Both result in the initial values as the optimum. The values tried by ProMechanica don't vary much from the initial values (not close to the limits set).

Please help

 

[grimmy65]

Instead of an optimization, try running a sesitivity study and expand your limits to above and below the current "optimized" value. This will clearly tell you if this value is indeed the correctly optimised value.

 

[agb253]

Thanks grimmy65. The sensitivity study works and shows the initial value is not the optimum. I still don't know why the optimization study is not working; I might just run many sensitivity studies to find the optimum.

 

[grimmy65]

I prefer, and have heard many others say the same, sensitivity studies because they allow you to select the results. This is very useful since it is possible to have many local maximums near the true maximum.