[DaIN]

Dear all,

I would like to know what is the general approach to impose a Reynolds number in a simulation of a channel flow with periodicity in the streamwise direction. I understand you have to consider a source term in the momentum equation which is given by friction forces, but how can I estimate this term?

Besides, should I start the simulation from a stagnated field (zero velocity) or would it be better to impose a velocity close to the target velocity (the one giving the Re I want)? Any advise to help me to start this experiment is highly appreciated.

Thanks in advance

[FMDenaro]

Originally Posted by DaIN

Dear all,

I would like to know what is the general approach to impose a Reynolds number in a simulation of a channel flow with periodicity in the streamwise direction. I understand you have to consider a source term in the momentum equation which is given by friction forces, but how can I estimate this term?

Besides, should I start the simulation from a stagnated field (zero velocity) or would it be better to impose a velocity close to the target velocity (the one giving the Re I want)? Any advise to help me to start this experiment is highly appreciated.

Thanks in advance

the source term can be expressed both in term of a pressure gradient and a mass flow. In the first case you can think of a constant gradient equivalent to the one you have in a Poiseuille flow. The additional pressure fluctuation is solved then by the elliptic equation.

[DaIN]

Thanks for your quick reply.

So what you mean is basically to run an initial simulation to get a Poiseuille flow at the target Reynolds and then to impose that pressure gradient as source term in the next simulation, is it right?

[FMDenaro]

Originally Posted by DaIN

Thanks for your quick reply.

So what you mean is basically to run an initial simulation to get a Poiseuille flow at the target Reynolds and then to impose that pressure gradient as source term in the next simulation, is it right?

but there is an analytical solution for the pressure gradient in Poiseuille flow ...

[sbaffini]

The value of the pressure gradient is different between the laminar and turbulent cases (in order to achieve the same mass flow). Check this reference:

http://www.lamc.ing.unibo.it/aimeta2.../MEM-273-0.pdf

[DaIN]

Oh excellent!

Thank you very much