Hi I'm working on cell scaffolds and l used Darcy and solid physics for modeling and I'm using Livelink COMSOL with MATLAB for material update (Young module) and I used the following code to set the previous solution as the initial solution .
model.sol('sol1').feature('v1').set('initmethod', 'sol'); model.sol('sol1').feature('v1').set('initsol', 'sol1');
But I got the following error: Messages:
The following feature has encountered a problem: - Feature: Time-Dependent Solver 1 (sol1/t1) Nonlinear solver did not converge
Maximum number of segregated iterations reached Time: 0 s. Last time step is not converged.
Thank you for your help.
Assalamuwalaiqum Hani Soleimani
Hopefully this will help you jazakallah khairan.
Hello Hani Soleimani,
I would like to ask if the time of the previous solution you are trying to set as initial step is the last time step of the previous study, if so it is recommended to add a line with ".set('solnum', 'last');". Below you will find an example with this :
model.study('std1').feature('time').set('useinitsol', true);
model.study('std1').feature('time').set('initmethod', 'sol');
model.study('std1').feature('time').set('initstudy', 'std1');
model.study('std1').feature('time').set('solnum', 'last');
Still, I don't think this is the problem, as I can see from the error messages it seems like its more of a convergence problem. It could be due to the boundary conditions and physics setup and coupling. However, you can try to increase the number of iterations hopefully this solves it.
To increase the number of iterations, an example below for the fully coupled and segregated steps is shown:
For fully coupled:
model.sol('sol1').feature('t1').feature('fc1').set('maxiter', 25);
For segregated steps:
model.sol('sol1').feature('t1').feature('se1').set('maxsegiter', 25);
Note here that i have changed the maximum iterations to 25, you can try this number. However, i would not recommend to increase the maximum iterations too much as it could take more time to simulate your solution and the problem with convergence could be elsewhere.
I hope you find the information above helpful.
Best regards,
- Badges
- Science topic