Having Problem solving 2D supersonic flow around a plug nozzle
 

Hi,

I am trying to solve a supersonic compressible flow around a "aerospike plug nozzle" at a particular altitude in fluent. Here are the conditions being used:

Model:
- Energy: On
- Viscous: Inviscid

Materials - Air
- Density: Ideal Gas
- Cp: 1286.68 j/kg-k
- Molecular Weight: 37.23 kg/kgmol

The mesh is divided in 4 fluids fluid zones, first 3 existing till the end of the plug and the 4th one being the rest of the domain.

Boundary Conditions:

- Inlet(Mass flow Inlet)
- Mass flow rate: 3.25757 kg/s
- initial gauge pressure: 2045430 N/m^2
- Total Stagnation pressure (P0): 1577.826k
- Direction specification: Normal to Boundary

There exist 3 far field pressures having the following conditions:
- Gauge Pressure: 64434 pascal
- Mach Number: 2.804
- Axial flow direction: 1
- Radial flow direction: 0
- Temperature(K): 264.4k

- Pressure Outlet:
- Gauge Pressure: 64434 pascal
- Backflow direction specification method: Normal To Boundary
- Backflow total temperature(T0): 482.689

Operating Condition: 0 pascal

Solution Method used:
- Implicit
- Second Order Upwind
- Flux type: Roe-FDS

Solution Controls:
- Courant Number: 1
- Limits have been increased for temperature and pressure

Solution Initialization:
- Standard Initialization
- Compute from mass flow inlet
- Reference frame: Absolute

THE PROBLEM:

The main problem am facing is.. the mesh file is huge (triangular mesh - 100,000+ cells)

There is divergence after about 15k + iterations. What message it shows is that due to sudden temperature change the time-step is reduced and courant number is also being reduced to 0.000xxx something and it iterates.

After one iteration it shows the same message again but now it further reduces the courant number.

I have tried using Explicit solver with courant number at 0.1 and it yields the same result

i have also tried both implicit and explicit with ASUM solver but the same result

RESULT INTERPRETATION: What i feel is that the sudden temp change across a shockwave is too much for fluent to compute (i maybe wrong)

but can anyone point out what the problem might be ?? or what can be done for it??

Thanks a Lot.

Hi chrislloyd,
I think one quick way of doing is to use ramp_up approach. you can start your computation with smaller mach number and then after let us say whatever 500 iterations changed to slightly higher Mach number until you reached to your required Mach number.

The second one is to use a very small Courant number before even starting the computation may be 0.05 or similar. Start with 1st order AUSM upwind scheme and then after some 100 or 1000 iterations switch to 2nd order AUSM upwind.

Hope this helps.

regards and good luck.

dear
 

in some reference book of CFd, we can see this recommendation that:
Don't use mass flow inlet condition for ideal gas.
try to set pressure inlet and pressure outlet.
use gas dynamics handbook to satisfy your mass flow rate with difference of inlet and outlet pressure.

Thanks for your suggestion @taxalian, i reduced the outlet pressure and the mach number and used your approach and its worked properly

but @64344 pressure and 2.8 mach number i have tried different variations with the ramp up approach, but every time (even after increasing the limits) it says "time step reduced in xxx cells due to excessive temperature change"

i read online that if this happens for high speed flow you should reduce the "positivity rate limit" to 0.05 or 0.02. but even then it diverges.

Any idea what can be done?

@farzadpourfattah - unfortunatly the problem is given such that i can't change the inlet from mass flow to pressure inlet? could you think of any other method to the solution?

Thank You

Dear
 

In our cfd group, we cannot converge solution ideal gas with mass flow inlet condition, If you can run with mass flow boundary condition for ideal gas please tell me.

Sorry for the late reply, what would "run with mass flow boundary condition for ideal gas" mean?? i am still working on the problem

Hello chrislloyd

I am working on simillar design of Aerospike nozzle in Ansys Fluent.
my profile is based on MOC and pressure inlet and oulet boundary conditions.
But i am still facing simillar problems as your..
Did you figure out , what was the problem.
Requesting to help me out.

Regards
Mehlam