Hi,

I'm trying to model a simply convergent nozzle, but I have not good results. In particular I have reversed flow at pressure inlet b.c.

Pressure inlet is 1.5 bar Pressure outlet is 1 bar

I'm using coupled implicit solver and RNG non equilibrium wall functions.

Could you help me? Thank you Giacomo

How steep are the angle of the walls? If the angles are steep, stagnation points may exist, and then you may have recirculation near the walls.

But I doubt you will have recirculation, and therefor I agree, you should not have reversed flow at the boundary conditions.

What does your pressure gradient look like. And What about the velocity gradient. Can you identify the boundary layer.

Did you try to run the solution with all solution controls to 1st order, standard, simple, and let the solution converge after lets say several hundred iterations?

You can advance the solution controls when you are comfortable with the solution, but you want to increase the accuracy of cell to cell calculations.

What is your Reynolds number? What viscous model do you use?

Thank you for your interest,

fortunately I have succeeded in reducing the problem of reverse flow at pressure inlet by a better solution initialisation. I have a problem with static pressure of the flow, that is the flow (at the inlet) has a velocity much greater than that analytically calculated.

I use coupled axisimm k-e RNG two-layer. Total pressure inlet is 151325 Pa and temperature is 300k. Pressure outlet is 101325 and Temperature is 293k.

The angles are first 8° and before throat 12°.

See your e-mail, I'm sending to you an image of the nozzle.

Thank you Giacomo