Smaller element size means higher secondary stress?
Smaller element size means higher secondary stress?
(OP)
Hi,
I got a question about element size and stress. We know that in the area of geometry discontiguous, boundary condition applying and node force applying, there usually gets secondary stress.According to ASME, if we use Sm for primary stress criteria, then we should use 1.5Sm for secondary stress, which means we use 1.5Sm to check the finite element results.
Here is the problem, I found that the smaller size of element, the higher stress we will get, until it goes to a stable value. In engineering, if we use this stable value to be the maximum stress, to make sure it is below 1.5Sm, we usually use much more thick plates than the reality need. I wonder what is the suitable element size?
Regards!
Rock
I got a question about element size and stress. We know that in the area of geometry discontiguous, boundary condition applying and node force applying, there usually gets secondary stress.According to ASME, if we use Sm for primary stress criteria, then we should use 1.5Sm for secondary stress, which means we use 1.5Sm to check the finite element results.
Here is the problem, I found that the smaller size of element, the higher stress we will get, until it goes to a stable value. In engineering, if we use this stable value to be the maximum stress, to make sure it is below 1.5Sm, we usually use much more thick plates than the reality need. I wonder what is the suitable element size?
Regards!
Rock





RE: Smaller element size means higher secondary stress?
Question 2 - you will want to refine your mesh until you have achieved a solution that is independent of the mesh.
More questions to follow once you answer these questions.
RE: Smaller element size means higher secondary stress?
I agree with TGS4's response to question 2. If you see that the value is becoming stable as you reduce the mesh size, then that is the solution.
RE: Smaller element size means higher secondary stress?
I used ASME Volumn VIII, the chapter " Design based on stress". For the primary stress, I calculated them with the function STRESS = E x STRAIN, and it has allowable stress Sm,which is the yield stress divided by safety factor. In the results of FEM, I can use the displacement of nodes to get the primary stress. For the stresses got from FEM directly, should we have to check them with 1.5Sm? If the results are higher than 1.5Sm, we must increase the thickness of plates or do sth to make them to be checked?
Regards!
Rock
RE: Smaller element size means higher secondary stress?
RE: Smaller element size means higher secondary stress?
5% plastic strain.