Learning Forum

[Zeska ZESKA]

I want to simulate the following problem:

I want to study the mixture of two gases, but without combustion. the flows are supersonic and there are a total of three inlets, two of which with oxidizer and one with fuel and the gases are not premixed with the other one. I only know their pressure, which is 6 Pa for oxidizer and 12 Pa for fuel.

I have the following procedure: 

Sover: Density-based, Absolute velocity formulation, steady,  2D planar, no gravity

Energy: on

Viscous model: Spalart-Allmaras

Speices Model: Species Transport (Ractions (volumetic) -uncheked ; Options( diffusions)-uncheked;)

           Mixture Properties:  Mixture species: air and h2

           denisty: ideal-gas

           cp- mixing law

           thermal conductivity- constant 0.0454

           viscosity- constant 1.72e-05

          mass diffusivity-constant dilute approx 2.88e-05

 

Boundary conditions: 

inlet air: pressure inlet: gauge tot. press. 600000

supersonic/initial gauge press: 600000

thermal: 300

species: h2: 0

 

inlet h2: pressure inlet: gauge tot. press. 1200000

supersonic/initial gauge press: 1200000

thermal: 290

species: h2: 1

 

outlet: gauge pressure: 100000

thermal: 300

species: h2: 0

 

operating pressure: 0

 

Solution methods: 

Formulation: implicit

Flux-type: roe-fds

Spatial discretization: gradient: least squares cell based

         flow: second order upwind

         modified turbulent viscoisty: second order upwind

         pseudo time method: off

 

Initialization: hybrid (no patch)

Iterations steps: 10000

 

when I run the simulation, the residuals go high (higher than e+04) and the simulation stops. 

what have I done wrong? What do you thinnk of my procedure? Is it plausible?

[Rob]

Check the pressure field after the initialisation process. Also, what stops the mixing region dropping to 100,000 Pa very quickly once it's running? The pressure based coupled solver is good to about 3M so you may not need the Density based solver. 

[Zeska ZESKA]

 

 

Hi Rob

Thank you very much for your reply!

I changed the solver to pressure based and the method to Coupled. The rest is completly the same. After initialization I plotted the pressure and density contours and they looked good. The pressure and density 

at the boundaries show the values ​​that I prescribed. 

I run the simulation once again for 10000 time steps, but it stoped after around 900 time steps and it said that the solution converged. Do you think that's ok/normal?

Moreover, when I plot the final pressure contour, the values at the boundaries are not the one that I had at the begging, but values that are slightly lower, like 1 or 2 bar lower. What causes that?

 

 

 

 

[Rob]

It shouldn't stop at 900 time steps; 900 iterations is plausible. If you go to Monitors>Advanced you can stop checking convergence and then run on. 

Re the boundary values. Assuming the solution is converged (not just residual checks) check the mesh and also that you're looking at the correct pressure (static, total etc).