LES
 

hi all,

has anybody investigated the effects of varying cell size or shape within the domain in an LES calculation, i.e. skewing or grouping of cells.

if so i would be interested to know the main findings of the experiment. what effects did it have?

what would theory say the likely effects would be on an LES calculation in particular, and are there any good references?

Thanks Peter

Re: LES
 

Hi, I have worked with a channel flow config. with variable cell size (only one case). i found that anisotrpic correction to smagorinsky model proposed by Scotti and Meneveau is giving better results compare to simple smaganoski model. But my results are with implicit filtering approach.

Re: LES
 

What do yuo mean by "implicit filtering approach"??

Re: LES
 

HI it means i have filtered my NS equations implicitly.

Re: LES
 

thanks for reply.

i suppose the anistropic model will work better for any grid that is not equal in each direction, whether that be skewed or simply stretched. i may consider applying this model.

i would be interested to know though, how did the channel flow results compare - experimental against varied cell size against constant cell size? did you run these tests?

what i really want to know is, what theory would have us believe the implications of skewed (varied) mesh will have on an LES calculation?!? (say for example in channel flow).

peter