Hi, i'm a student working on my thesis about transonic fluid flow through a Laval nozzle of a spiral jet mill. The fluid is air. My CFD-Software is FIRE(AVL List) and COMET(ICCM). Can anyone help me do find the appropiate boundary conditions for the inlet and outlet for such a problem ? Are there any test cases for such a problem ?

the inlet to a laval nozzle is usual subsonic and so it is necessary to specify 3 boundary values. usually the total pressure and temperature (or enthalpy) are specified along with either the flow angle (usually 0) or the mass flow rate. if the exit is supersonic then no boundary values need be specified if it is subsonic then one boundary value usually the pressure is specified. also you may try using nonreflecting exit conditions if they are available in the codes you are using. hope this helps

Thank you very much, clifford. Did you know any code which has nonreflecting exit conditions included.

When you specify pressures at both boundaries, is it not the case that the pressure drop between your inlet and outlet deterimes your mass flow rate? Consequently, your problem may be ill-posed if you impose mass flow boundary condition at the inlet as well. What do you think?

(1). Yes, you are right. (2). Because I am using the Pressure conditions everyday. (3). I specify inlet total pressure, inlet total temperature, inlet velocity direction components (sine, cosine of the flow angle), and the k-epsilon turbulence levels. (4). At the exit, I specify the static pressure. (5). This set of boundary conditions has been used for the subsonic flows, and the transonic flows at the exit. The problem I am solving is 3-D turbine nozzle flows (stator)

yes sorry i made a goof. the specification of inlet bc depends on the idea you have for the upstream conditions. the typically used To, Po and flow angle constraints are applied when you assume the flow is coming from a large upstream reservoir and the mass flow is free. however if the flow upstream is throttled (choked) then the upstream mass flow is specified and you replace the Po with it. the specification of flow angle is convenient although it may be arbitrary. it is also possible to specify upstream vorticity in place of it

i can't really say because my experience with commercial codes is limited. you can consult with the vendors of your codes if they have this capability (it might be under another name) also your codes may be pressure based in which case i'm not sure about the avialbility of such bc. for more information you can read Kevin Thompson's papers on this subject in Journal of Computational Physics (i referenced them in some of my earlier posts which you can find with the search feature on this list.