Hi all,

Can CFX model heat transfer from a heated wall into the flow for inviscid flow in a pipe?

Thanks, Dave

Hi,

I don't know why you would bother modelling heat transfer for an inviscid flow. The flow feature which strongly determines the heat transfer - the boundary layer - is not modelled in inviscid flow so the results would be very inaccurate.

If you are turning heat transfer on then better properly model the viscosity also.

Glenn Horrocks

Thanks Glenn,

I believe this is the case and it's only a matter of interest to me, but how would one model frictionless flow with heat transfer (Rayleigh flow) in a pipe????

In the strictest sense, this type of flow can't exist but in theory it can, so is it possible to model?

Or is it acceptable to use a turbulent model and make the heat transfer so large as to be driving force in the flow, thus neglecting wall friction?

What do you think?

Hi,

If the flow is inviscid then you have no boundary layer and no turbulence model. Therefore the only heat transfer at the walls will be conduction. This is numerically possible but as I said the heat transfer will be so far wrong for most cases that I would not bother.

Glenn Horrocks

Cheers