supersonic inlets filling take with no outlets
 

Hi,

I'm having trouble with supersonic flow boundary conditions for an air-based simulation.

When I use CFX, I use Shear Stress Transport Model and Total Energy for heat transport.

My geometry contains two inlets that connect to a rectangular tank. They connect on opposite sides of the tank. These inlets have a constant pressure of 80 psi. They're supposed to represent a shock wave of air. The rectangular tank initially is at atmospheric pressure, but as the simulation runs it should increase in pressure because there is no air leaving the tank, just entering, and compressing.

I want to capture the pressure distribution time history in the tank. But when I specify supersonic inlet properties I come to a halt. My static pressure is a constant 80 psi, but my total pressure varies as the model runs because it's dependent on the variable pressure elsewhere in the tank.

Apparently you need to know something constant about the velocity at the inlet, but I can't do this, and I can't purchase a sensor to record such a time history at such high speeds. I also have no information on the total temperature, as it relates to the velocity value.

My solution doesn't even start when I assume a Mach number to calculate all these values, I receive an invalid number buffer overflow error.

I've also tried to use Flotran in APDL, but I it also fails before iterating, diverging in either ENKE or negative temperature values.


When I began simulations on a subsonic level, I saw Mach numbers as high as 7 outputted to the screen (in CFX). Subsonic levels do not output results, the tank reaches 80 psi in 0.7 ms which is not close to my experimental values of pressure.

Can Ansys perform this type of simulation with CFX or Flotran? In CFX, I can't just specify a static pressure of 80 psi, which is all I know about the model (well that and where the walls are). I'm stepping really slow, my time step is 1e-5.


Thanks!