Discussion Closed This discussion was created more than 6 months ago and has been closed. To start a new discussion with a link back to this one, click here.
Problem in Defining a New User-Defined Creep Law in COMSOL.
Posted Sep 8, 2023, 6:33 a.m. EDT General, Structural Mechanics, Material Models Version 6.0 3 Replies
Please login with a confirmed email address before reporting spam
Hi all,
I want to study the effect of creep on a pressure tube (PT). The required creep law contains many terms and can only be implemented by a user-defined option in COMSOL.
The creep law is given as-
ε ̇_tot= Aσ^n exp((-Q1)/T) + Bσ Φexp((-Q2)/T) + C Φexp((-Q3)/T)
where, ε ̇_tot is the creep strain rate.σ is the equivalent stress and Q1, Q2, Q3 are the constants and defined in the parameters. A is constant, and B, and C are the functions of z ( length of the PT) and are defined as an analytical function in the GUI.
I defined the above equation in the f(σ) in the user-defined creep setting by putting solid.lemm1.cmm1.sequ in place of σ.
Now my question is that if I consider only 1st and 2nd term (Aσ^n exp((-Q1)/T) + Bσ Φexp((-Q2)/T)) in the expression, my simulation converged, but if I add the 3rd term (C Φexp((-Q3)/T) also in the expression then it will not converge. (Note that the 3rd term not a function of σ)
Why this thing happen? the 3rd term is also important , how can I handle this problem. One thing I did multiplied the 3rd term by (solid.lemm1.cmm1.sequ/solid.lemm1.cmm1.sequ), but this also not worked.
Your suggestions are helpful for me. Thanks Shyam.
Please login with a confirmed email address before reporting spam
Have you tried placing a very small coefficient in front of C? If that converges, try increasing the coefficient step-by-step, and see what happens.
Also, try to evaluate all three terms separately. Do they have reasonable orders of magnitude?
-------------------Henrik Sönnerlind
COMSOL
Please login with a confirmed email address before reporting spam
Thanks Henrik for your valuable suggestion.
As per your suggestion I tried placing small coefficient in front of C. when I put 0 and 0.001 it converged but when I put 0.01, then it did not converge.......
How can I put the term separately in the setting of user-defined creep in f(σ) expersion? Yes there is difference in the magnitude of term, when I solve using analytical method using matlab then- 1st term-> in the order of 10^-17; 2nd term-> in the order of 0 to 120; 3rd term-> in order of 0 to -120;
If I consider only 1st and 2nd term then simulation run well. But If we add 3rd term the it creates problem.
If I use only 3rd term then also not converged.
Your any valuable suggestion for this type of problem will be helpful for me.
Please login with a confirmed email address before reporting spam
As I understand it, the third term is negative. The total creep rate must however be positive. So if the third term (in magnitude) exceeds the sum of the two first terms anywhere, the solution will fail.
-------------------Henrik Sönnerlind
COMSOL