Learning Forum

[Eddie2324]

Hello, I have a non-linear static analysis model which includes brackets assembled by spot weld( I defined bonded contact from circular faces to simulate that). The thing is I got very wildly different results when I checked the von misses stress with varying options of display (Such as elemental mean and averaged) and different positions( Such as top/bottom and mid plane). I checked manual and did some research online but I confused and could not be sure about how to evaluate this solutions, I am trying to check that part is beyond the yield or how far beyond the yield but I can not be sure.

for example this is midplane- averaged ;

And this is mid plane elemental mean

And this is the top/bottom average

I have to decide about this part going beyond the yield or not and for bottom/top surface its beyond the yield and for mid plane it is not beyond the yield. I usually expect to see lower stress values in mid plane so that's why I was always deciding this kind of situations by check results in averaged, top/bottom surface. But I saw higher higher stress values in mid-plane for different force values than the screenshots I shared, the parts are not thick they re about to 1.5-3 mm thickness. How should I decide how to I check? Also I know how elemental mean and average display options works but I was not expecting see this much different results, is it better to check parts from elemental mean to ignore false stress which happens in 1-2 node? I am really lost, please help and I will be glad if you could advise me some training pdf, video or book about this topic to learn it deeply and make %100 sure about it.

Thanks

[Ashish Khemka]

Hi Eddie,

 

You can use the higher magnitude results to be on the conservative side. On different options for stress result display, please see if the following link helps:

 

Averaged Vs. Unaveraged Contour Results | Ansys Mechanical (simutechgroup.com)

 

Regards,

Ashish Khemka

[Eddie2324]

Hello, thanks for the answer, so you are saying I should check the other planes to see which one is higher for my future works too, by the way I also wanted to ask something, I am reading my max stress from one-two nodes which is in the fillet points and I used mesh convergence, the stress is finite so there is no singularity but it is just in one node. Is it might be artificial stress or should I consider this stress even it is in just one one to evaluate part will fail or not? ( It is not possible to change fillet radius since this radius is used for design the part)

[Ashish Khemka]

Hi Eddie,

 

The stress at one node might be an artificial one. You can try refinement and the stress will further increase. Generally, stress at fillet is not a singularity. Share the loads and boundary conditions - just to check if there is any load applied just next to the fillet. For the top and bottom layers with shell elements, you may expect similar stress magnitude and they may be higher with respect to the mid-plane level (an analogy is the bending of a cantilever beam where the top and bottom layers will have the same stress magnitude). 

Regards,

Ashish Khemka