Problem Structural Analysis Singularity

[robbedalex]

Hello. I have a problem with singularity on Ansys Mechanical. Would someone help me? I'm trying to simulate two parts joined with a bayonet. on one part there is a pressure. I want to study the resistence of the bayonet. I do a structural analisys with frictionless contacts, and as the mesh is finer, the stress increase. Attached some images with a coarse mesh. The bottom part as a fixed support on the lower face. I need to use weak springs instead the model is underconstrained.

 

[L_K]

Please provide picture of the following: boundary conditions, loads, overall mesh, mesh and stress in the detail
where the singularity occurs.

Normally, you should apply enough constraints that weak springs are not needed. In this case you might have to
activate friction to achieve that.

 

[robbedalex]

Hello

 

[robbedalex]

Contact

 

[robbedalex]

the other body contact

 

[robbedalex]

pressure

 

[robbedalex]

mesh fixed body (patch conforming, tethaedron quadratic with refined edge)

 

[robbedalex]

mesh other body (same options)

 

[robbedalex]

results on the target body

 

[robbedalex]

results on the fixed body

 

[robbedalex]

zoom on the fixed body

 

[L_K]

Taking into account the location of max. stresses in both bodies, in my opinion frictional contacts are warranted in this analysis. The target body max stress is on the edge so it is singularity, the more mesh is refined, the higher the stress value. Fixed body max stress value should converge as mesh is refined.

 

[robbedalex]

I tried also with frictional contact, but the result is the same. The analysis is linear with isotropic elastic material. options on Analysis settings are Program Controlled. If you need more details please tell me. thank you in advance for your help.

 

[robbedalex]

Do you think I have to put a radius instead of the edge to solve the problem?

 

[L_K]

Other option is to read the stress a small distance away (from half element to element length) from the edge.

 

[robbedalex]

Yes, this is the way and I'm not worried about singularieties because of DSV principle, but I tried to have a good results without singularities. Thank you for your help

 

[robbedalex]

Tried again and again, Also with adding radius but the result is the same. Also with frictional contact. I think the problem is that One body is undercostrained.but it is only a cap. The only " constrain" is done by surfaces in contact. If I add a constrain in vertical direction the problem disappear, but It is not the reality. Someone have some idea?

 

[L_K]

What is the issue? Model is not converging? Stresses are too high compared to acceptance criteria?

 

[robbedalex]

Yes. The stress is too high and continue to increase depending of the mesh. I can't reach a correct solution because of singularity. The next step for me is fatigue analysis so it's important to eliminate singularity. The only way for now is to add some constrains on one body, but this is not the reality

 

[L_K]

Boundary conditions and connections between bodies are crucial to get accurate results, even more so for fatigue analysis.

But stress singularity is in my opinion a separate issue. Maximum stress value should converge to a value as mesh size is decreased.