Hi all!

I'm simulating 2D circular cylinder for both steady and unsteady. Steady case results are good. Cd values from my simulation are close to several references.

The problem is unsteady case simulation. I've tried these procedures: 1. start the simulation direct for unsteady 2. use the steady result, change the solver to unsteady, and run the simulation

the results that I get have these patterns: 1. if I use previous steady results, I can't get the vortex shedding process. Maybe because the time step that I use was too large, but I use the formula that Fluent gave.

2. if I use unsteady direct simulation, the velocity contours are not symmetric.

Thank you

time step maybe 1/100 of the period of vortex shedding.

Sheila, it is normal that unsteady flows give asymmetrical contours of velocity because of time dependent proces of periodic vertex generation at the outflow part of the cylinder (2-D - circle). I tried to perform steady simulation around a cylinder and got over 400% static pressure error at the back side of the cylinder. If you want to get correct results from unsteady case you must indicate some points within important area and perform monitoring for each of them. After calculation is finished you must elaborate these data which are different for any time step. Did you try to use k-omega turbulence model with enhanced wall treatment (Transitional Flows)? It's very important to execute appropriate mesh close to the wall (y+=1). I hope it helps you a little. Do you simulate laminar or turbulence flow? What is the Reynolds number? Good luck

why time step size equal to 1/100 of the period of vortex shedding. i can find the reference in where???

Dear First, i would like to introduce my shelf. I am a final year student of major Aeronautical engineering, Hanoi University of Technology, Vietnam. I am aslo doing study of the phenomenon of the flow over 2D circular cylinder in both steady and unsteady flow and compare to some experient results. I did with some Re number 4,40,100,1000,1E+5,3E+5, 2E+6. with Re = 4,40 the result i got similar to some reference, but with Re = 100, i can't get the result exactly (of course, compare to reference), i can't visual the von Karman vortex...and can you show me how to calculate the Time step size in unsteady solver. Help me now, i need your help. Thanks and best regard ! My email: mrquangngoc@gmail.com My yahoo nick: ngocbkat46

Dear friend,

some times in order to see if the solution converges you must do succesive studies refining timestep and grid. The solution of this problem is rather simple. See if you can find test case of featflow free code.

Regards Achilleas