Dear all,
I'm working with Simvascular for steady flow but after many times trying I couldn't get a good nonlinear residual number, every time I got higher than 1*10-3 as in the guide is mentioned that it's not a good sign.
I put here the details of my simulation:
Unit in MGS
Mesh
-
Number of Nodes: 182073
Number of Elems: 1069575
Number of Edges: 89568
Number of Faces: 59712
-
Inlet and Outlet BCs
-
BC Type: Prescribed Velocities
Analytic shape: parabolic
Point Number: 201
Fourier Modes: 10
Flow rate(from File):
0.0 -507163.7
1.0 -507163.7
Period: 1
Outlet, Resistance: 0.02366
-
Wall Properties
Type: Rigid
-
Solver Parameters
-
Time Step Parameters
-
Number of Timesteps: 1000
Time Step Size: 0.01
-
Output Control
-
Number of Timesteps between Restarts: 20
Output Surface Stress: True
Force Calculation Method: Velocity Based
Print Average Solution: True
Print Error Indicators: False
-
Step Construction: 15
-
Advanced Parameters
Pressure Coupling: Implicit
Backflow Stabilization Coefficient: 0.2
-
Non-linear Iteration Control
-
Residual Control: True
Residual Criteria: 0.01
Minimum Required Iterations: 3
-
Linear Solver
-
svLS Type: NS
Number of Krylov Vectors per GMRES Sweep: 100
Number of Solves per Left-hand-side Formation: 1
Tolerance on Momentum Equations: 0.005
Tolerance on Continuity Equations: 0.04
Tolerance on svLS NS Solver: 0.04
Maximum Number of Iterations for svLS NS Solver: 100
Maximum Number of Iterations for svLS Momentum Loop: 20
Maximum Number of Iterations for svLS Continuity Loop: 4000
-
Discretization Control
-
Time Integration Rule: Second order
Time Integration Rho Infinity: 0.5
Flow Advection Form: Convective
Quadrature Rule on Interior: 2
Quadrature Rule on Boundary: 3
-
Actually, I got good residual when I decreased the inflow to 10 [mm3/s] but I'd like to know that is it high inflow(507163.7[mm3/s]) for Simvascular? what is the maximum inflow that I can use in Simvascular?
I tried to change my solver parameters to have a good result but I wonder if you could possibly explain me to how can I deal with this problem?
Do you have any idea about good solver parameters?
Thank you in advance.
Kind regards,
Muhamad
Convergence problem
- Weiguang Yang
- Posts: 110
- Joined: Mon Apr 07, 2008 2:17 pm
Re: Convergence problem
You can change the following parameters in solver.inp to see if there is any improvement.
1. reduce your time step dt, say 0.005
2. decrease residual criteria, say 0.001
3. increase nonlinear iterations, add "Step Construction 0 1 0 1 0 1 0 1 0 1 0 1 0 1" to solver.inp
I am not sure how many iterations you use. Usually 7 iterations are sufficient for a rigid case.
1. reduce your time step dt, say 0.005
2. decrease residual criteria, say 0.001
3. increase nonlinear iterations, add "Step Construction 0 1 0 1 0 1 0 1 0 1 0 1 0 1" to solver.inp
I am not sure how many iterations you use. Usually 7 iterations are sufficient for a rigid case.
- mark cowell
- Posts: 1
- Joined: Sun Sep 15, 2019 4:48 am
Re: Convergence problem
I'm facing same problem. And I've tried everything you said.. But nothing helped me!wgyang wrote: ↑Thu Dec 20, 2018 1:41 pmYou can change the following parameters in solver.inp to see if there is any improvement. 1. reduce your time step best washing machine dt, say 0.0052. decrease residual criteria, say 0.001 3. increase nonlinear iterations, add "Step Construction 0 1 0 1 0 1 0 1 0 1 0 1 0 1" to solver.inp I am not sure how many iterations you use. Usually 7 iterations are sufficient for a rigid case.
Last edited by mark cowell on Tue Sep 17, 2019 6:51 am, edited 1 time in total.
- Vijay Vedula
- Posts: 63
- Joined: Mon Feb 09, 2015 1:27 pm
Re: Convergence problem
Could you give us more information about the geometry and vessel dimensions? Are you trying to solve flow in a cylinder? What is the flow Reynolds number?