[LuckyTran]

Originally Posted by FMDenaro

I do not agree....why? Think a similar flow, the backward facing step with a small height. If you set a laminar profile the onset of turbulence from the expansion will be different from that you would have with a real turbulent profile. The stress at the wall is different. The reattachment lenght will be different. And having a laminar profile after a long pipe is not physically reasonable.
Therefore, you are right when you consider the flow trasport properties advected (apart from the elliptic behavior of pressure) downstream but, to simulate accurately the transient, you have to prescribe a correlated profile. The only way to do an accurate representation of this problem is to have an upward pipe sufficiently long to let the laminar profile develop in a turbulent profile before the expansion.

You are assuming that the flow is turbulent at the inlet and that the flow in the long pipe is also turbulent, simply because somebody mentioned the word DNS. I.e. this could be a unstable flow problem, or hydrodynamic instability problem. As long as you resolve all the scales in those problems, it's still DNS.

[FMDenaro]

Originally Posted by LuckyTran

You are assuming that the flow is turbulent at the inlet and that the flow in the long pipe is also turbulent, simply because somebody mentioned the word DNS. I.e. this could be a unstable flow problem, or hydrodynamic instability problem. As long as you resolve all the scales in those problems, it's still DNS.

No, I just assume by a physical point of view that the flow motion in a long pipe is developing a turbulent profile at a certain Re_x. Turbulence is the common nature of the flow in channels and pipe, laminar conditions are limitated to very small region. Of course I am not talking of fluid at high viscosity.

[diggee]

When I started this thread, I was only interested in knowing if I could post process anything useful with a partial simulation but seeing the general direction this thread is going in, I think I should give some physical clarifications.

The Re based on the inlet diameter is only 2000. The expansion ratio is 2 so the Re after expansion decreases to 1/4th of 2000. Transitional flow is triggered partly by the geometry (diverging c/s) and partly due to the perturbation I am imposing on the inlet BC. Since I am reproducing an article, I am giving all BCs as the author gave; hence the inlet paraboloid UDF with perturbation and so on.

What LuckyTran also says about the usage of the word DNS is right; this is a spatial localized problem and that is why I have a very fine mesh only in the expansion area. Of course, it can be questioned as to how did I decide till where to make the mesh fine but then again since I am reproducing an article, I already know till where approximately the mesh needs to be fine.

[FMDenaro]

Could you post a link to such paper?

[FMDenaro]

Originally Posted by diggee

The Re based on the inlet diameter is only 2000. The expansion ratio is 2 so the Re after expansion decreases to 1/4th of 2000. expansion area.

Increasing the diameter after expansion theRe number increases ....

[LuckyTran]

Originally Posted by FMDenaro

No, I just assume by a physical point of view that the flow motion in a long pipe is developing a turbulent profile at a certain Re_x. Turbulence is the common nature of the flow in channels and pipe, laminar conditions are limitated to very small region. Of course I am not talking of fluid at high viscosity.

Re_x doesn't guarantee turbulent flow in a pipe, even if the pipe is very long. You have also consider the channel Reynolds number, based on the hydraulic diameter and you must exceed this critical value to get turbulent flow. For circular pipes the critical Re is 2300. You can have an infinitely long pipe and no turbulence if your channel Reynolds number is less than 2300.

So the inlet is not turbulent and parabolic velocity profile is appropriate. Inside the domain then, you solve for everything so there are no worries.

[FMDenaro]

Originally Posted by LuckyTran

Re_x doesn't guarantee turbulent flow in a pipe, even if the pipe is very long. You have also consider the channel Reynolds number, based on the hydraulic diameter and you must exceed this critical value to get turbulent flow. For circular pipes the critical Re is 2300. You can have an infinitely long pipe and no turbulence if your channel Reynolds number is less than 2300.

So the inlet is not turbulent and parabolic velocity profile is appropriate. Inside the domain then, you solve for everything so there are no worries.

Yes but the theory assumes the Re_D for a pipe developing infinitly downward... Conversely, You should consider here a pipe ending at a certain x due to the expansion that changes the pressure condition...

[LuckyTran]

Originally Posted by FMDenaro

Yes but the theory assumes the Re_D for a pipe developing infinitly downward... Conversely, You should consider here a pipe ending at a certain x due to the expansion that changes the pressure condition...

The inlet comes from a long pipe from which you can assume it has developed infinitely and that any possibility of turbulence has been damped out by the walls. Inside the domain is DNS so there are no worries. You can assume that the inlet is parabolic.

[FMDenaro]

Originally Posted by LuckyTran

The inlet comes from a long pipe from which you can assume it has developed infinitely and that any possibility of turbulence has been damped out by the walls. Inside the domain is DNS so there are no worries. You can assume that the inlet is parabolic.

My idea is that for Re_D not more than 2000 it could be still reasonable to set a parabolic profile at the numerical inlet but considering thereafter at least 10-15 diameters of pipe lenght in the computational domain to let the flow developping according to the pressure change in the expansion.
With the correct grid resolution the simulation could be an acceptable reproduction of the onset of turbulence.

A similar problem was simulated in this old paper:

https://engineering.jhu.edu/fsag/wp-...6/JFM_1999.pdf

in their case the authors assumed a turbulent inflow