rigid body convergence issue
Hi,
I have problems with the convergence of my rigid body. Although, the flow parameters are converged. Can anybody specify what the problems might be?? http://pichost.de/images/2012/09/26/GEwkf.png But look at the flow parameters, they do converge. http://pichost.de/images/2012/09/26/YlKO0.png |
I applied residuals for torque, force, and mesh motion. Still divergent.
http://pichost.de/images/2012/09/26/qoIX.png http://pichost.de/images/2012/09/26/p4A1d.png |
|
|
For a combination of above settings turned on these were the results.
http://pichost.de/images/2012/09/26/1UK8k.png http://pichost.de/images/2012/09/26/VNxGW.png |
Can you describe what you are modelling?
|
I am simulating a rigid body drop in fluid. Using 6 dof Rigid body solver and the rigid body defined as an immersed solid.
The problem is that the fluid parameters u, v, and w show convergence but the rigid body does not. Also in rigid body convergence curves overshoots occur. |
These are the best results that I could obtain, but it has 100 iterations for the stagger/coupling loop.
http://pichost.de/images/2012/09/26/iM0lF.png http://pichost.de/images/2012/09/26/ypq9.png http://pichost.de/images/2012/09/26/wOVzp.png Still unsatisfactory!! |
Hi,
I have explained my case. A rigid body falls under gravity in a column of fluid. As I have shown that the rigid body convergence diverges. However, in ANSYS help I found this. Force Convergence It is defined as ratio of difference of forces for previous and present time step divided by the greater of the two forces. Now, in my case the rigid body falls under gravity. At each time step the force acting on the body differs as the rigid body rotates and translates. Is that the explanation of the curve behavior??? Also torque convergence is defined similar to force convergence but for torque. |
Are you using immersed solids or moving mesh approach?
|
the immersed approach
|
Have you read section 1.2.10.2 of the documentation listing the limitations of this approach?
Would explain what you are seeing. Also have a look at section 11.7. Also |
Yes, I have. My case is incompressible fluid. However, since the rigid body falls with initially at rest. This limitation applies to it.
"For low Reynolds number turbulence cases, the immersed solids turbulence model cannot accurately predict the pressure fields near the immersed boundaries". All clear for rigid body limitations |
Yes, that restriction may cause problems. What is the Re number of your bodies motion at the time you are having problems?
So your simulation is single phase? |
Yes, it is single phase = water
Rigid body has max. velocity = 2 m/s. Re = [(rho =1000kg/m^3)*(v=2m/s)*(d=0.02m)]/(mu=1e-3) = 40k. I am extremely sorry, it isn't low reynolds. |
Agreed, that does not sound like low Re. But when the motion first starts it might be. What is the Re at the time of convergence difficulties?
|
well if you see the curves above. I would say pretty every where.
|
Can you try putting the body in a duct with a cross flow in it? Then the body will be starting from a reasonable Re right from the start. I suggest this just to see if the problem is convergence at low Re.
|
you mean that I should give some velocity to the fluid i.e. 2 m/s, as max. velocity of rigid body in previous case was 2m/s.
Also, should I make the rigid body stationary?? I will do that, using 6 dof rigid body solver! |
Hi,
Here is a brief explanation of what I am trying to do. A rigid body defined as an immersed solid and solved by using 6 dof Rigid body solver is dropped from rest into a fluid domain. The fluid domain is single phase. Opening BCs are given to the top and bottom of the fluid domain. Wall BCs the sides. |
Hi,
guys is there no expert on this. At least anyone tell me how to reduce dispersion in the residuals. What are its reasons. I need basic CFD knowledge on this. |
I have already told you an experiment which will tell you whether your problem is convergence problems at low Re when the body first starts moving.
Have you put your body in a strong cross flow (to make sure it is high Re right from the start) and checked whether that converges well? If you are unhappy with the help you are getting on the forum then contact CFX support. |
All times are GMT -4. The time now is 08:57. |