Mesh changing resulted in stress change

[amirrezaaj]

HI
I have a problem in ansys that made me confused...I have modeled a conceret dam and meshed it with solid 45 elements...after that I have assigned the available surface pressure on the dam to the model...in another modeling, I changed the mesh size only and analyzed the the model with the same loading conditions...but the results have been changed? why the results are depended to the mesh size in this situation? is there any relation with surface pressure loading and mesh number...?! what should i do...? with kind regards...

 

[Skrambles]

That is normal for a finite element model. Changing the mesh density can help you figure out if you have any singularities. As you increase the mesh density the result should change less and less each time as it converges on a solution. If the results never stop increasing or decreasing, then there is a singularity at one of the nodes and the solution is divergent. There should be plenty of papers on the web over "convergence analysis."

 

[amirrezaaj]

Thanks for your regards dear friend...But there is a point that the mesh change that I mentioned, affects the results considerabley...for example when the number of the elements in the model increase about 2 times in the dam ansys model,the stresses in the model increases about 4 times... Also, when some elements in the model refined in mesh size in one location, the results increase 2 time...I don't understand why this happen in my model since the loading condition is the same in both models...this amount of incease in the stress s value is not logical...

 

[xxChrisxx]

It's probably a stress singularity as it get worse with mesh refinement, or a stress concentration near a boundary condition that is breaking the model. So it may or may not be important.

Take a convergence plot at several nominal locations away from boundaries, see how those results are affected.

 

[amirrezaaj]

Thanks a lot chris...
I had a section in my model that was wncountered with stress concentration and i excactly want to control and read the stress in this section...this section has a low thickness and is connected to another part that is considered as a constrain for the mentioned section...i check this section and see that changing element size in this section resulted in stress increase...now how can remove the effect of stress concentration in this section chirs...? is there any special method or trick in ansys to reduce the effect of stress concentration...?? thank u some much for your kind regards...

 

[Unrest]

I had a section in my model that was wncountered with stress concentration and i excactly want to control and read the stress in this section
...
is there any special method or trick in ansys to reduce the effect of stress concentration...??

Since you want to know the stress at the concentration, you can't use any tricks to get rid of it. You have to locally refine the mesh until the stress stops increasing much.

However if you have a singularity, that's not physical and you must get rid of it to see the real stress. For re-entrant corners, use their actual radius, don't make them sharp. Also watch out for other discontinuities like the edges of fixed faces - these may need to be radiussed too.

Be aware that singularities aren't a fault in FEA, they're the software trying to model something that cannot exist in the real world. If you approach it in the real world, the stress really will approach infinite, just like the model says.

 

[amirrezaaj]

Thank u so much for your kind regards...