[mehyan]

hi everybody,


i am modeling turbulence flow in a smooth pipe.

the y+ range is 11<y+<30.

1- can i use k-epsilon model?(please tell the reason in detail)

2- which wall functions should be considered? standard or scalable?

3- and how can i sure k-epsilon is the best model?
4- what is the range of linear and log law? and which wall function should be considered for each range?

tanx for all

[Red Ember]

50<y+<300 is better for k-e

[urosgrivc]

hi, I think SST is the wright way to go.

There are beter models for flow near wals than k-epsilon, k-omega is beter and
SST combines both and blends inbetween. So it will probably stil use k-epsilon just not near wals.

And the range of y+ you mentioned is good for k-omega

CFD.NINJA on youtube uses SST model for pipeline as well.
https://www.youtube.com/watch?v=da5BNAwy9Us
if this is something you want to model.

if you want to add some comlexitiy to the problem, you can use aditional gamma-theta transition model. This will add two additional equations and the computational time will be longer but the laminar-turbulent transition will be beter described, i do think you ned quite fime mesh to use it properly though.