I am trying to modell a simple axissymmetric turbulent jet (air to air at Re = 11000) at the exit of the jet. But my results look terrible and I believe that it depends on the boundary condition and the size of the domain. I would really appreciate advices in this type of modelling with no "walls". Regards Ulf

I believe you are simulating a confined jet. Could you tell us how terrible are your results to try to find out why it is terrible.

Hi Ahmed The computational domain consist of several recirculation zones!!! I use inlet, pressure and symplane as boundary conditions. I try to model the jet as a steady state problem, but it seems more propely to use a transient-solution. Regards Ulf

Hi Ulf,is your jet impinging somewhere? where is the outlet? what is the Re?

Hi again!

The jet is not impinging and the Re=11000, based on the jet-velocity and the jet nozzle diameter. I seem to find a (good) solution with a transient approach but I don´t understand why I don't get a solution with a steady state approach, Regards Ulf

If it is not impinging somewhere, how do you define the outlet boundary, this could be one of the reasons.

Hi again!! I define my "outlet" boundary as a pressure boundary. And I have to make a sufficient long domain in the axial direction, in order to make sure that the flow is undisturbedly. The domain is 100*Dj in the r-direction and 400*Dj in the axial direction, where Dj is the jet-nozzle diamter.

Ulf, try to make a confinment from two sides (the nozzle plan and the side parallel to the axisymetric) and see if you will reach steady state. I think with your boundary it is too difficult to reach steady state. Also you may try to use less values instead of 100 Dj use 10 Dj and instead of 400 Dj use 40 Dj and see if it makes any difference.

Hi! The pressure field of a normal jet decays as R^-2, where R is the spherical radius (Unconfined jets have spherical, or hemispherical rather than cylindrical symmetry.) Hence, pressure boundary conditions can be used when imposed sufficiently far away. However, it is well known that you do get better results if you use absorbing outflow conditions which ensure that you get the correct mass flow. There are some references in my article in J. Fluids Eng. March 1999, vol 121, pp 102--105, and I know that Cercignani and Moin (independently) have used these conditions subsequently, but I do not know where their results are published. From your name I suspect that you are Swedish, so send me an email (carste@frt.frontec.se) and I will give you my phone number in case you want to discuss the matter any further.

Yours sincerely, Carl Fredrik Stein

What is the Mach no?