I am trying to calculate flow parameters of the multi stage subsonic axial compressor using Ansys CFX. But it turned out that the solver cannot find the solution. In all cases an "OVERFLOW" error has appear. I am trying to solve with following boundary conditions: 1) total pressure and temperature at the inlet and massflow rate at the outlet; 2) total pressure and temperature at the inlet at the inlet and static pressure at the outlet. On the other hand when i performing calculations of the separate stages, the Ansys CFX solver easily find the solution. As I understand it, when i try calculate all stages in one calculation, the initial parameters in the solver are too wrong and because the solver cannot find the solution.
What can I try do to find a solution?
As I understand your question, your solution does not converge. Several information are missing for a specific answer. After how many iterations you get the "OVERFLOW"? Is it steady or unsteady? What are the Reynolds and the Mach numbers? What is the mathematical model? Where do you get the "OVERFLOW"?
In general, assuming your BC and mesh are good, you can reduce the CFL number or use a low order method (first order method) to try to get an initial solution. You could also try to reduce the Reynolds number (increase viscosity) to get an initial solution and then use this solution to start a calculation at the appropriate Reynolds number.
Probably, the high pressure stages are not converging because the initial solution for these stages are far away from the correct solution. You may try to initialize the solution from a good starting point or you may try to under-relax the convergence of the solution using an under-relaxation factor of a value less than one. Under-relaxation might help the solution to converge even for bad initial conditions but it will slow down the convergence rate.
François Morency , Big thank you for that answer ! As you say i am submit the missing information.
1.I get "OVERFLOW"after 15 iterations
2.I am performing Steady analysis
3. The solver (before crash) writes that maximum Mach number more than 100
4. The mathematical model is 8 stage axial turbo compressor.Mesh generated via Turbogrid program (fig 1). Working fluid - Air Ideal Gas.
5. I dont know where OVERFLOW is ,because solver is crashed and i dont have results fils.
I remembered your advices, but in this case as Farouk Owis advise i add into the each domain initialization. I add averaged velocities ,static temperatures and static pressures which i know from axisymmetric calculation which i do in another program . After that CFX solver could find the solution.
Unfortunately i found that my CFD calculation is bad accuracy (the total pressure at the outlet is higher real value on 7% ( i know real value of the total pressure at the outlet) . What would cause this?
- Badges
- Science topic