[evan247]

Hi all,

Could anyone advise or provide reference on how to calculate BL displacement/momentum thickness for internal flow (say inside an axi-symmetric curved duct)?

In particular, I'm not sure how to choose the free stream velocity and density, as for internal flow the flow normally has some gradient along the span.

Thanks!

[LuckyTran]

From a fluids and turbulence standpoint, an internal flow is which in which boundary layers takes up the whole duct and there is no more a freestream velocity. Displacement and momentum thickness are not very meaningful here.

If BL does not take up the whole duct, then you have a developing BL and it acts like an external flow (with curvature) and only here do you have a freestream velocity. This freestream velocity has the same meaning as usual although the freestream velocity can accelerate/decelerate under an inviscid pressure gradient which you can calculate very easily by application of Bernoulli's equation for example.

[evan247]

Hi LuckyTran,

Thanks for your reply. Given my duct is curved and the flow (pressure for instance) has spanwise variation as well, should I perform an inviscid CFD for the same geometry and extract V_freestream from the result? If so my follow up questions is how to calculate this freestream velocity then? Is it at the mid span?

Originally Posted by LuckyTran

From a fluids and turbulence standpoint, an internal flow is which in which boundary layers takes up the whole duct and there is no more a freestream velocity. Displacement and momentum thickness are not very meaningful here.

If BL does not take up the whole duct, then you have a developing BL and it acts like an external flow (with curvature) and only here do you have a freestream velocity. This freestream velocity has the same meaning as usual although the freestream velocity can accelerate/decelerate under an inviscid pressure gradient which you can calculate very easily by application of Bernoulli's equation for example.

[FMDenaro]

No at all, it makes no sense an inviscid fluid model for your case...