[kaegen]

Hi,

I am a student at NTU. Presently I am doing a Fluent simulation on choked flow in orifice to investiagate how discharge coefficient(Cd)of co2 varies with pressure difference. To get Cd, I need the mass flow rate. The geometry is a circular pipe 60mm in diameter and 142mm long, with a orifice(13.8mm) in the middle.

I have tried pressure-based solver with k-epsilon model and attained convergence. I use constant density. But it seemed not able to simulate choked flow as my mass flow rate keeps on increasing with pressure difference. I use inlet pressure of 1000kpa. The outlet pressure varies from 900kpa to 100kpa.

Next, I tried density-based solver, implicit and explicit, ROE-FDS and was not able to get any convergence. I use ideal gas as density. I have tried decreasing courant number and relaxation factor which fail to converge as well. Most of the time I get floating point error.

Questions
1. Do I have to play with adapt/gradient settings? If so, how to I determine the threshold and the max/min no. of cells.

2. I use turbulent intensity % and hydraulic diameter settings for inlet and outlet. Is the H.diameter the actual diameter of the inlet and outlet pipe diameter or are they the diameters of the orifice in the pipe?

3. I understand that density changes with pressure. Is it inaccurate that I use const density for co2 for my prssure-based simulation? If I use ideal gas as density, the mass flow rate attained is way too high.

Any help from anybody is greatly appreciated!!

[kozav]

Hi,

I have no advice, but the same problem. Since the February, didn't you come accross something useful?

Thanks,

kozav

[pertupd]

hello both,
firs of all you should read something about thermodynamics theory of 1d compressible flow, otherwise you would not have even tried to choke constant density flow (some text on cfd numerics would be good too).

To get chocked:
1; use ideal gas flow
2; use density, implicit solver
3; prescribe sub critical pressure ratio, for air lower than 0.52... (dedicated texts and google knows critical pressure ration for CO, or CO2)
4; start with first order upwind discretization
5; start with moderate value of CFL number (1)
6; increase CFL continuously (say up to 12, depends on grid and pressure ratio)
7; switch to second order, start with low CFL and little by little increase as BC propagates very slowly
8; get converged solution (based on e.g. mass flow rate balance or some downstream property - temperature or Ma, but not based on residuals)
9; try to read even more theory and comprehend the results.

If you found some 1d model how to simulate Cd and thermodynamics properties downstream the orifice after pressure recovery for chocked flow regimes, I would be very interested.

thanks and good luck,