Hi Dear friends,

In recent days, I have been simulated the behaviour of the flow behind a square cylinder. My code uses RANS and two K-Epsilon models (the standard and Low-Reynolds). With both the models, the two-vortex row does not persist in the wake. My question is: is this behaviour correct?

Thanks. Ferreira.

(1). Fill your bath tub with water. (2). Move the square cylinder stick through the surface of the water in a steady stroke. (3). Observe the wake behind the square cylinder. (4). Spread some baby powder on the surface to aid the observation. Estimate the Reynolds number and check the result. (5). Effect of Reynolds number? You will have to run a real test to find out.

Your RANS simulation is doing what it is designed to do with the k-eps. (an eddy viscosity) model - all unsteady turbulent motion, and the unsteady vortex shedding, is smeared out with the large eddy viscosities thrown into the simulation just behind your square through the use of the k-eps. model. As a result, the unsteadiness that occurs in real life can't be reproduced

At best, what you can expect from your RANS simulation using the turbulence model is something that is close to the long time average of the real flow.