[osuguy]

Hey all,

I'm trying to model a Mach 2 flow over a missile. However I keep getting errors in the middle of my simulation saying "time step reduced in 3 cells due to excessive temperature change". I get a similar error for pressure as well. I attached a pic of my mesh and the error (along with residuals) for reference.

I feel this problem is due to wrong boundary conditions and I'm new to supersonic flows in Fluent. Here are my settings:

Density solver
Transient
Inviscid with energy equation on
Density as ideal gas
Constant specific heat
Boundary condition: -On the far boundaries (inlet and far from the missile) chose a pressure farfield setting with Gauge=101325Pa, Mach 2, 300K and the flow direction
-On the outlet chose pressure outlet with Gauge=0Pa normal to boundary
-Operating pressure=0Pa
Solution initialization: compute from inlet
Solution Methods: Implicit, AUSM flux, Gradient=Green Gauss Cell Based, Flow=3rd order MUSCL
CFL=2
Residuals at 1e-4
Time step=.1
Max Iterations/step=20

Let me know what you think. Thanks.
fluent error.jpg

mesh.PNG

[flotus1]

Compressible supersonic flows are not my strong point, but are you sure it is a good idea to set both the gauge pressure and the operating pressure to zero?
The time step size seems may be way to high and I would not start the first simulation with a third order scheme.
It might aswell have to do with bad mesh quality, but unfortunately there is nothing to see in the picture you posted.
Does the simulation really have to be transient?

[osuguy]

I read somewhere on these forums that supersonic flows needed 0Pa for the operating pressure. As far as the gauge pressure it's only 0 on the outlet, but 101325Pa in the farfield and inlet, but thinking about that twice I feel that they should be consistent (maybe I'll try both at 101325Pa).
I don't believe is is the mesh quality, at it has roughly 8.5 million cells (although I'm not sure about the orthogonality of the mesh). The mesh radius is roughly 20 diameters away.
I chose transient due to that it is slightly easier for Fluent to solve (as opposed to steady where some damping of the solution occurs). I was going to run transient until a steady state is approached. I don't really have a reasoning for the time step size though, I'll try some lower.

[osuguy]

I should also add that I get this same "time step in 3 cells changes due to excessive temperature change, reducing the CFL to something lower" regardless if I use a steady OR transient solver.

[flotus1]

Originally Posted by osuguy

I don't believe is is the mesh quality, at it has roughly 8.5 million cells

Mesh quality and cell count are not related.
Check the mesh quality explicitly.

Operating and gauge pressure set to zero result in an absolute pressure of zero (as far as I know) which could explain one of the error messages you get.

[yemai]

Hello, I recently met this problem, very like your first picture, and I also make a model about a Mach 2 flow in the combustion chamber. So, could you please give my some advises?