[bharath]

Hi to all,

I am simulating a de laval supersonic nozzle simulation.
Material as Steam using IAPWS table (CFX).Turbulence model as SST.

Nozzle parameters are Nozzle Pressure ratio = 0.201,Throat to exit area ratio = 1.3687,Gama =1.321

Target Mach number at nozzle exit is 1.7 usingIsentropic flow relations.
But i am getting exit mach number from CFX output 1.58.

I did Mesh sensitivity studies (3Million nodes to 12Million nodes).But nothing helps me to reach the targetted Mach number of 1.7

Could anybody tell me some advice on this.Thanks in advance

[Opaque]

Not sure how you setup the problem, but the isentropic relationships assume no heat transfer (no even conduction in the fluid), no losses (no friction on the walls), and one dimensional flow (no transversal flow as the flow is squeezed through the nozzle) .

How close does your setup satisfy those assumptions ?

[ghorrocks]

Also IAPWS is a real gas model, not an ideal gas model. You would expect a small difference between an ideal gas model and IAPWS.

[bharath]

Thanks ghorrocks and Opaque

I understood the difference between Isentropic,Ideal gas and Real flow simulation.

My setup assumptions
1) No heat transfer and Isentropic
2) Frictional losses
3) Very little two dimensional traverse flow (as the flow is squeezed and expanded)

Is these variations will affcet the Output this much (Exit Mach Number 1.7 to 1.58)? Please clarify

[ghorrocks]

How can you assume isentropic flow and frictional losses? Isentropic means you are assuming no friction.

It is best if you work out whether your result is reasonable yourself. Do some experiments and find what happens. What happens if you use an ideal gas instead of IAPWS? What happens if you make the walls free slip? Then you can find for yourself whether these things make a difference or not. Also the ideal gas, free slip wall simulation should match the simple isentropic flow calculation pretty accurately.

[bharath]

Thanks ghorrocks

There is very little losses due to wall friction.By neglecting this friction loss, flow is considered as Isentropic.

I tried Water Ideal gas with No slip wall matches the Exit mach number of 1.7

Due to Mach waves at exit, real gas simulation results shows diffenrce with design calculation.

Thanks for the reply

[Opaque]

Just for clarification. If you use no slip boundary condition, there will always be friction losses. Small or large depends on the Reynolds number you are running your simulation.

On the heat transfer side, if you activated the heat transfer model regardless of boundary conditions, you already have heat transfer unless you set the thermal conductivity to exactly 0.

My point (and I extrapolate here Glenn's) is that the isentropic flow equations are and idealization that do not match your setup; therefore, there will be minor differences even for ideal gases. Top that with a real gas model, and you are in for a set of results interpretation and sensitivity analysis to understand if the results make sense.

[bharath]

Thanks Opaque

1) "My point (and I extrapolate here Glenn's) is that the isentropic flow equations are and idealization that do not match your setup; therefore, there will be minor differences even for ideal gases"

Yes, there is small differences (for ideal gas, slip wall) with isentropic calculation.But unlike Real gas simulation these differences are negligible.

2) "Top that with a real gas model, and you are in for a set of results interpretation and sensitivity analysis to understand if the results make sense"

Yes results are making sense with real gas model as Mach waves at exit is cause for the different output.