Discussion Forum

Hi,

I'd like to model a simple 2-D deformation of the rectangular solid material using the previous solution operator in transient structural mechanics module. Here are the I.C. : 1) zero displacement field; 2) zero structural velocity field. Here are the B.C. : 1) fixed constraint at the bottom; 2) roller on the sides; 3) boundary load on the top as a function of time.

Then I follow this blog www.comsol.com/blogs/using-the-previous-solution-operator-in-transient-modeling/ to implement the previous solution operator in transient modeling. I add a Domain ODEs and DAEs interface to the model, since I am interested in tracking the solution at the previous time step throughout the volume of the part. Then, I define the units of both the dependent variable and the source set to displacement field u_new =[u2,v2]. I don't know why but I dont have 'Discretization' section.

Then I specify the equation in Distributed ODE section: u-nojac(u2)=0 and v=nojac(v2) = 0. The nojac() operator is needed, since we do not want this equation to contribute to the Jacobian. Lastly, I need to specify that this equation should be evaluated at the previous time step. This is done within the Solver Configurations by adding Previous Solution operator. Once I add this feature, I simply select my dependent field variables u_new =[u2,v2] to be evaluated at the previous time step. I tried both Segregated Solver and the Fully Coupled solvers.

But I keep getting an error: Previous component solver failed, system matrix is zero, last time is not converged. I was wondering if there is anyone had an experience with Previous Solution Operator. I'd assume the error I'm getting is due to nojac() and the way I'm using it. I'd appreciate if you could give me any advice or suggestion how to fix this issue.

Sincerely,
Abdulla