I try to simulate a tensile test, but I dont really get it how CalculiX deals with the *PLASTIC card. I’ve plotted the stress strain curves how it should look like and how it does look with my ccx simulation (‘stress_strain.png’). My input data for the material are points of the real curve (‘material_input.png’). The curve I created with CalculiX data reaches a stress of 370 MPa and then remains almost constant. Maybe someone could explain how CalculiX handles with the input data.
And another thing is, that there are two critical points in my simulation (‘cgx_output.png’). I expected maximum of stress only in the middle of the specimen…
Also, make sure the mesh is refined enough. Too coarse a mesh can cause the stress/strain to not match the input data as well as necking to be in the wrong place.
Thank you all. Using the example of Calc_em I got a good result.
But there’s already a new problem. I tested simulations for different rectangular profiles and different displacements as loads. The only good result is with a profile of 4/12.5 and a displacement of 25% as load (see pictures). Why are the results so different? I’ve already tried using a fine mesh.
Advice: Each time you see Strain in a book, result, manual,…… ask your self ¿Which strain are you talking about? . Finite Element Analysis has many (more than 10) different Strain measures.
¿What does “displacement” of 25% means?.
¿What does “strain” shown in your graph means?.
¿What kind of strain are you providing below the *PLASTIC card?
¿What kind of strain are you reading in your results file?
I meant I plotted the Mises Stress and the disp1 in load direction as CalculiX results and it should be the same as cauchy stress and logarithmic strain if I understood it right.
Sorry, it was late yesterday. Actually I plotted the logarithmic strain with the Cauchy stress in “test”. The other curves in the diagrams are representing the strain ε = ΔL/L (ΔL is the output of CalculiX/disp in direction 1, L=specimen length) and the Mises stress. In the first diagram I used different displacements as loads (disp25 means ε = ΔL/L = 0.25 → Input = 0.25L → The specimen is pulled 0.25L in direction 1).
The second diagram shows the different profiles I used (4.0/12.5 → 4 mm/12.5 mm).
I don’t think the error is in the plot, because the results for disp25 and 4.0/12.5 agree very well with the test.
I’ll try to be more careful about what I post.
But the different profiles still do different results. After reaching the necking smaller profiles ensure that the stresses drop. Is it maybe because of the hardening of CalculiX? I didn’t define any hardening parameter.
They all go worst after necking and you only plot up to 25%…mmmh
¿Are you measuring the Strain or are you linearly interpolating according to the time step?
EDIT: Sorry the question is not properly formulated.
¿Have you request ccx to compute for you the Strain or are you linearly interpolating according to the imposed displacement and time step?
