How to obtain correct viscous pressure drag using CFX?
I am investigating ship viscous resistance using a CFD code (CFX) and so far, I use ellipsoids body of revolution. In the absence of wave resistance, the total resistance consists of frictional resistance and viscous pressure resistance.
The grid is created using CFX-MESHBUILD, a structured grid generator, and the flow solver is CFX 4.2. Further, the grid quality and density seem to be OK.
As far as ITTC correlation line is concerned, frictional resistance seems to be correct. However, viscous pressure resistance looks to be overestimated. Does anyone have the explanations for this?
Re: How to obtain correct viscous pressure drag using CFX?
Helen from AEA Technology have checked the problem and gave me advised to introduce the so-called DEFERRED CORRECTION term under >>FLOW SOLVER. This Deferred correction is in fact a dummy correction i.e. k and epsilon start and finish at the same number of iteration. For example,
K START 1001 K END 1001 EPSILON START 1001 EPSILON END 1001
It helps me a lot and makes the solution quicker to converge. Furthermore, I also made grid refinement at both end (leading and trailing edges) and to allow me to have about 300,000 grid cells. So far, everything works well and I can reduce viscous pressure drag quite significantly and also the form factor which is now about 1.4. The form factor expected is about 1.2 and 1.3 and it means that CFX has given me a very good result.
Finally, many thanks to Helen.
|All times are GMT -4. The time now is 17:06.|