Dear Friends,

I am presently simulating a 2d underexpanded jet case using FVM with AUSM scheme. However, I land up in problems, presumably because the BCs are not very right. I suspect that the outflow(exit bc) is causing trouble. The jet is supersonic and I have tried both extrapolation and Riemann characteristic bc there. I do get some good results(before blow-up), but the code blows up early. I have tried out with Roe scheme and still the problem persists.

Any references with specific references to underexpanded jets and non-reflecting BCs could be particularly helpful

Kindly advise

Thanks in advance,

Kiran

Dear Kiran,

I've been working about simulation of highly underexpanded jets. Of course, the outlet boundary condition was very problematic, in particular if the nozzle pressure ratio is high and if your outlet section is located in the subsonic area downstream of the strong shock wave. In fact, a first idea was given by Prudhomme to help to prescribe a more adapted level of total pressure within this core subsonic zone. It consists in hypothesizing that the flow is nearly one-dimensional so that you can use the level of total pressure found within the supersonic shear layer and prescribe it in your subsonic core. In fact, in my case, it was insufficient because of the high levels of difference of total pressure between the supersonic shear layer and the subsonic coflow, which were leading to strong instabilities propagating upstream and inducing a kind of feedback loop. Finaly, I did not find any theoretical was to prescribe such adequate conditions. If you do not want to use a too long computational domain, I found nethertheless that it is possible to strongly coarsen the mesh just in a thin zone just upstream of the outlet in order to dissipate these artificial numerical waves. Of course, the solution is not valid in this thin zone but the perturbation on the solution found upstream of remains really marginal. Tell me if what I say is clear enough ?

Guillaume Lehnasch

[karthick]

hi halo,
this is karthick 4m MIT, i've been tryng to simulate underexpanded sonic jet, bu i ve had no success so far.. i am using a rectangular domain with a pipe at left edge as inlet .. my only prob, there is a rise in the total pressure immediately after the exit..but this is nt the case in underexpanded case..the total pressure wil never increase....hw do i correct it...i am using fluent 4 as solver with standard k epsilon model