High residual for extruder mesh
 

Dear All,

My problem is a cylinder oscillating(swaying) on free surface. I am comparing the horizontal force from star ccm+ with experimental values.
I am have tried to use 3 different motions to simulate the same problem. The motions used were translation(overset), morphing and rigid body(overset). My inital cell size was about 5.7 lacs. I got a good agreement with experimental and CFD values. I had used trimmer for all the 3 cases.

Now i was trying to simulate the same problem with just about 50-70k cells and using morpher. In this case i used extruder mesh. I had used fine cells even at areas(the first time) far away from the cylinder. Increasing the sizes of these cells helped me to decrease the total cell count.

When i run this mesh using morpher, i am facing a problem with residual levels. I do understand residuals are not the final say for a solution to have converged. When i check the horizontal force, it is again similar to my previous results, agreeing well with the experiments. I have attached the screen shot of the force, residuals and my mesh.

1. The residuals settle to a high value but they remain constant there.The solution is not diverging i guess. I believe this is because the cells far away from the domain have a very high aspect ratio and hence there is something going wrong with the numerical calculation in those cells. How do i ascertain this in STAR CCM+? Is it possible to plot individual cell residual for me to see if they are high in the cells which are of interest to me? Ya and when i simulated it previously i was getting residuals of order of 10^-5(the first time for all 3 motions).

2. How should i improve my mesh to get a lower value of residuals?

There is no flow, cylinder is on calm water and i have wall boundary all around.

Please guide if you know the answer

Thanks
Regards
Arun