Hi!

I've just write down the code in Matlab for a supersonic flow, using the Time-Dependent technique I hope to reach a steady solution. The problem is I have no idea about which are the usual values for the Courant Number, Time Step and Artificial Viscosity coefficients. For instance, how much time took you to reach the convergence? (I mean real time t, of N-S equations, not the computing time).

Thank you very much.

Either I cannot see your answer in my computer, or you haven't help so much with that invisible words.

Hello

for the time step you can used the following equation:

Deltat=C*DeltaX/(a+V) where C is the courant number and C <= 1 is o.k for hyperbolic equations( supersonic flows) DeltaX is your mesh,or grid size. a=(kappa*R.T)

take kappa as 1.4 for ideal gas about R, i am not sure about value. I have taken 8.314 J/mole°K but in my model velocity rate of change with that R value give some strange result. T is temperature. V is velocity.

By the way dou you know how to treat B.C

The reason you are getting strange result is because you need to use gas constant for air and not universal gas constant while computing acoustic speed a. To obtain gas constant for air you have to divide universal gas constant by molecular weight of air.

I'm using the CFD book of Anderson as a reference. The problem I have is I haven't got any global idea of how much time my program takes away to converge.I'm noticing a stiffness in my velocity field. And I don't know if that stiffness is provoked either by a very small time step, or the program doesn't work. I'm using Re=100, Ma=4.5 and obtain a time step of 1e-12 approximately with a courant number of 1e-4.

It's the first time I do something like this. What problems could I have when simulating this?. Have you got any experience in this.?

I also use same book as a reference.I go out of the office now.let me look at for your problem and we talk tomorrow.

yesterday i have also posted a problem, and i thought that my time size is small.I doubt after i saw your step size.it is very very small.Any way what is the step time? and why did you selected Courant number so small? In the book Mr. Anderson point that for lava flow 1200 time steps are good enough to see steady state.

[Numz]

please give me your code
nomi16425@gmail.com