I have done an analysis on a solid beam, I have attached the analysis file here. The example which I have done in Prepomax is same as the FEM examples given in Freecad, I have attached the Freecad tutorial link. In the result, the Tresca stress is not comparable. In freecad also calculix is the solver used.
Thank you
test_file.pmx (1.9 MB)
https://wiki.freecadweb.org/FEM_CalculiX_Cantilever_3D
The difference comes from the definition. According to the source codes, FreeCAD defines Tresca stress as (SP1-SP3)/2 while PrePoMax calculates it as SP1-SP3 (where SP stands for principal stress). Unfortunately, I don’t have access to Abaqus in the next few days to check how it works there but (SP1-SP3)/2 seems to be the most common definition in the literature.
1 Like
(SP1-SP3)/2 , SP1-SP3… which is actually correct??
I’ve found a way to check this in Abaqus now. Interestingly, it uses the same definition as PrePoMax: SP1-SP3. But both FreeCAD and GraphiX use the other definition (with division by 2). So there’s no single correct formula, Tresca stress can be understood in more than one way.
Could you please check the tresca value in Abaqus for the analysis I have done…
With a coarser mesh, the max Tresca stress is:
- in Abaqus: 431.876 MPa
- in PrePoMax: 426.1 MPa
1 Like
I think we talked about that and decided to go with the Abaqus formulation where the Tresca stress is equal to 2 * max shear stress which can be directly compared to the tensile material strength.
1 Like
Matej,
Fully agree with that. There is only one definition of the Equivalent Von Mises Stress, and its value for the uniaxial tension test.
For the Tresca Yield criterion the definition is (σ1 – σ3) = Y = 2k for yielding in a tension test and relates the the yield stress to the shear stress at yielding.
It must be remembered that (SP1-SP3)/2 is a shear stress. Multiplying by 2 gives the stress value