Discussion Groups

Simulation CFD

Reply
Active Member
adesk360
Posts: 6
Registered: ‎10-22-2012

Flow over immersed body

147 Views, 3 Replies
12-19-2013 10:57 AM

Hi everyone,

Just wanted to pick your brains on a problem I have been dealing with lately: I have an internal low reynolds flow that faces a cylindrical resistance and becomes turbulent downstream. Image attached shows the described problem.

Which flow regime would you chose to simulate the flow in SIM 360? 

 

Thanks,

Jay

 

Please use plain text.
*Expert Elite*
OmkarJ
Posts: 434
Registered: ‎10-02-2012

Re: Flow over immersed body

12-20-2013 01:07 AM in reply to: adesk360

Probably the low Re model (with fine mesh near walls) would be best for this? You don't want a turbulence model for a laminar flow as it would introduce unnecessary diffusion. And laminar flow won't be able to capture turbulent scales.  Although, this seems to be a case of transition model which SimCFD doesn't have, to my knowledge. 

 

 

Please use plain text.
Active Member
adesk360
Posts: 6
Registered: ‎10-22-2012

Re: Flow over immersed body

12-20-2013 08:33 AM in reply to: OmkarJ

Thanks Okmar,

That was my thought too. I just don't fully understand the Turb/Lam ratio and the basics of this turbulent regime solution in SIM360; is it somewhat like the two-layered zonal model available in Fluent that can recognize where the laminar flow becomes turbulent?

Please use plain text.
*Expert Elite*
OmkarJ
Posts: 434
Registered: ‎10-02-2012

Re: Flow over immersed body

12-23-2013 01:33 AM in reply to: adesk360

From what I know, FLUENT has k-kl-w and SST transition  model, with former solving for turbulent k, laminar kinetic energy kl and w; and the latter solves k and w coupled with intermittancy. In essence, the transition from laminar to turbulence is addressed. SimCFD only has 2 eqn models, from which the low Re model is meant for low Re flows. However, it is generally believed that the name "low Re" mostly refers to its advantage in intergrating upto wall (region of low Re) over  normal k-eps model which has to use wall functions, and not its ability to predict the intermittancy of  transition.  It is based on k-eps framework with use of damping functions in definitions of of eps and eddy viscosity to avoid any extreme results near wall when the boundary layer is resolved unlike normal k-eps models. This seems the best bet for your application. 

Please use plain text.