I have written the 2D SIMPLER code successfully. Now,I want to develop it to LES code.Am I correct for my concept below:

1. 2D model is add Z-direction to become 3D model.

2. Add unsteady term du/dt

3. Add Smagorinsky viscous term to the original equation

4. Use SIMPLER to deal with special term and Runge-kutta or Crank-Nicolson to handle temperial term.

Am I right for my plan?

You may have difficulty with LES under the frame of SIMPLER, unless you invent develop some techniques to change the dissippative feature. As far as I know, a too dissippative scheme of accuracy lower then second order would result in too much error in calculating the nonlinear terms.

I do not understand how come SIMPLER is dissipative.. As far I as understand SIMPLER is an algorithm of how to properly solve a set of coupled equations in an segregated manner. The equations can be discretized which any method : finite volume, finite element, etc.

I can only guess that you have associated the Power Law interpolation scheme from Patankar with the SIMPLER algorithm.. I have run SIMPLER with central differencing scheme (which is second order) and with power law as well on not LES problems. For sure, power law was more dissipative than CD..

For LES, you may also want to try a second order time discretization scheme. The additional terms will end up on the right hand side vector anyways..

SIMPLER suffers from grid size and LES requires very fine uniform meshes.. The finer the mesh, the more iterations it will take to converge for a given time step. It is a hell of a lot faster than SIMPLE, but much slower than a coupled algorithm..

Good luck

[amir2920]

Hello

could you please send me discretization of momentum and energy equation in cylindrical coordinate or a paper that contains this? Thank you in advance. my email is: amir2920@gmail.com