Hi, everyone,

There is a tank with compressed air in it. At some specific time, the valve of this tank is opened and then some air will escape from this tank quickly. Apparently, this is a transient problem. My question is: how to set the boundary conditions for this problem? It looks like very simple. When I used FLUENT 5.3 to solve this problem, I set the valve as the pressure-outlet at which the static pressure is equal to the ambient pressure. All the internal surfaces of the tank are treated as walls. My results are very strange. No matter how higher I set the initial pressure within the tank and how small the time step size is, the pressure at some point within that tank, which I chose as the indicator of the internal pressure, will dramatically (6~8 iterations)drop to the zero (ambient pressure). Even I constraint the size of the valve to be very small, I got the same results. Is there anything wrong with using FLUENT? Can FLUENT solve this type of problem(no inlet, the density in the tank will decrease with time)? How should I set the initial and boundary conditions?

Thank you very much!

Whiatker

Check the velocity at the valve? if it is the sonic speed (Mach=1), I believe you have reached the [shocking] (I'm not sure of the spelling)condition and the velocity at the valve will be constant (Mach=1) no matter the pressure difference is.

Also if you're not already using it, try the coupled solver.

Greg

Your right, but You have to check the ratio of external and internal pressure. If p_ex/p_int < 0,5283, the Mach number in the valve is 1. FLUENT can solve simple problems of compressible flows, check the maunals.