Hi to all masters of cfd, i am working on a steady-state analysis of pressure coefficient of rotary-wing.In my simulation i made up a rotating domain which has fluid of ideal gas of 296K temperature. I selected SST as turbulence model and total energy as heat transfer model of fluid thatdosen't include viscous work term. Mach number of the rotarywing tip is 0.89M and the adjacent of external domain boundary has 1.8 Mach. That is to say, flow of arround the wing(or blade) is subsonic and flow of near the external domain boundary has supersonic behaviour. I built inlet, wall for blade and opening boundaries. High Resolution was selected as Advection scheme in Solver Control menu.

I wonder correct values of the following expert parameter settings. I think these are very important in my case.

Expert Parameters in Discretisation. pressure diffusion scheme : 2 scalar diffusion scheme : 5 stress diffusion scheme : 2 tbulk for htc : 296.0 K

Expert Parameters in Linear Solver. mg solver option : 2

Expert Parameters in Converge Control. model coefficient relaxation: 1.0 relax mass: 0.75

Expert Parameters in Physical Models. tef numerics option : 0

best regards.

charles.

Dear Charles,

In general, unless adviced by ANSYS CFX support specialists, you should not need to modify any of the expert parameters default values..

Did support lead you in that direction?

Opaque

Hi Opaque, i have not support of Ansys CFX specialist. i am studing on the final work of master of science. I changed two of them.these are mg solver option and tbulk for htc and others were staying in cfx default values. mg solver option has 1 as default in cfx release. I thinked as i will analysis pressure coefficient of blade i should choose mg solver option as 2 that is solves fluid using anizotropic pressure coeeficent based mg solver. The 2 value is more suitable than geometric based mg solver. tbulk for htc has 300K as default and i i changed it to 296K that will not create any problem.

thanks for your warning and interests.

best regards.

charles.

Dear Charles,

The mg solver option parameter will not change your results.. It changes the way multigrid is done which can only affect performance (number of cycles required), and/or robustness.. If your case runs fine with the default option, changing the value is a way to get into trouble instead.

the scalar diffusion scheme does not affect the mass and momemtum equations.. This one may change your results slightly, but again.. Are you having problems with the default values?

Opaque

Hi Opaque, my runs were converging so slowly and i wondered that cfx defaults of expert parameters are suitable or is there the better? i adjusted y+ value on the blade surface is less than 1 for sst turb model. it has 0.94 average value on the blade but that created so huge aspect ratio in mesh. max aspect ratio is nearly 72100. and now i think that aspect ratios makes slow converge speed. i will reduce aspect ratio values with increasing y+ value. i am estimating y+=20 is sufficient. if there is a comment from you i am looking forward it.

thank you for tips.

best regards.

charles.