1D COMSOL Domain Implementation

  • Thread starter Thread starter womfalcs3
  • Start date Start date
  • Tags Tags
    1d Comsol Domain
Click For Summary
SUMMARY

This discussion focuses on implementing a 1D COMSOL model with multiple subdomains, specifically addressing the challenges of solving coupled partial differential equations (PDEs) across three distinct subdomains. The user is working with a stationary model in COMSOL, utilizing a coefficient form for the PDEs, and is encountering issues with singular matrices during finite element assembly due to the complexity of the equations. The proposed solutions include using overlapping subdomains or separate models for each subdomain to manage the equations effectively.

PREREQUISITES
  • Familiarity with COMSOL Multiphysics, specifically version 5.6 or later.
  • Understanding of partial differential equations (PDEs) and their stationary forms.
  • Knowledge of finite element analysis (FEA) principles.
  • Experience with boundary conditions (BCs) in multi-domain simulations.
NEXT STEPS
  • Explore the use of overlapping subdomains in COMSOL for managing complex PDE systems.
  • Research techniques for solving coupled PDEs in COMSOL, focusing on linear and homogeneous systems.
  • Learn about managing singular matrices in finite element assembly within COMSOL.
  • Investigate the implementation of multiple PDE models within a single COMSOL project.
USEFUL FOR

This discussion is beneficial for engineers and researchers working with COMSOL Multiphysics, particularly those dealing with complex multi-domain simulations involving coupled PDEs and finite element analysis.

womfalcs3
Messages
60
Reaction score
5
Hello,

This is the first time I'm using COMSOL on a problem where I have multiple subdomains within the domain. Previously, I have only been using one domain/one set of equations/Two global sets of BC's.

I've been having a hard time thinking of a solution to my problem.

I have 3 subdomains, and all have their own set of ODE's and BC's. (They're really PDE's, but I'm going to assume stead-state for the time-being until I get it to work. I want to see the steady state solutions anyway as well.)

The problem comes in that in the first and third subdomains, I have 3 equations each to solve. In the middle subdomain, I have 7 equations to solve.

They are all heavily coupled, but the system is homogeneous and linear. So I have a total of 13 dependent variables.

Is there a way to either:

- Use overlapping subdomains so that I can have 3 subdomains in the middle rather than just one?
- Have the first and third subdomains have a model for 3 equations each, and the middle one have a model for 7 equations?

Because I can't do all 13 at once, as that gives me singular matrices during the finite element assembly.



I am currently using a 1D PDE stationary model in the coefficient form with all the transient term coefficients set to 0 for the time being.
 
Engineering news on Phys.org
I think I may have figured it out. I'll use it and see how the results go.

I'm trying to have 3 separate PDE models, and activating each one in one of the subdomains.
 

Similar threads

  • · Replies 8 ·
Replies
8
Views
4K
Replies
1
Views
2K
  • · Replies 1 ·
Replies
1
Views
4K
Replies
2
Views
7K
  • · Replies 21 ·
Replies
21
Views
17K
  • · Replies 3 ·
Replies
3
Views
12K
Replies
5
Views
4K
  • · Replies 5 ·
Replies
5
Views
2K
  • · Replies 6 ·
Replies
6
Views
2K
Replies
18
Views
3K