Hello,

I'm starting to use Fluent for some LES calculations (straight channel, backward facing step and similar) and i'm going to use the nita-fractional step method with the PREssureStaggeringOption and i have some doubts about the numerical implementation.

First of all, what is the particular Fractional Step Method implemented. That is, what of the following papers do i have to check first to find the one implemented in the solver:

S. Armseld and R. Street. The Fractional-Step Method for the Navier-Stokes Equations on Staggered Grids: Accuracy of Three Variations. Journal of Computational Physics, 1999

J. K. Dukowwicz and A. S. Dvinsky. Approximate Factorization as a High-Order Splitting for the Implicit Incompressible Flow Equations. Journal of Computational Physics, 1992.

H. M. Glaz J. B. Bell, P. Colella. Second-Order Projection Method for the Incompressible Navier-Stokes Equations. Journal of Computational Physics, 1989.

H. M. Glaz J. B. Bell, P. Colella. An Analysis of the Fractional-Step Method. Journal of Computational Physics, 1993.

R. I. Issa. Solution of Implicitly Discretized Fluid Flow Equations by Operator Splitting. J. Comput. Phys., 1986.

The user's manual is not clear about this.

The second question is about PRESTO. What i understood is that, whatever pressure-velocity coupling scheme i use, the pressure obtained from the pressure correction equation is still in the cell centers. At this time i should interpolate it to cell faces to implement it in the momentum equation (to make the pressure correction in a conservative way). The PRESTO, instead of doing this interpolation, acts in writing the continuity equation in a ficticious staggered cell (centered around the face where i need the pressure). Is this right?

Finally, if i'm right, is also right that the continuity equation, in being discretized in this new position, still needs that the velocity on the faces has to be calculated with the Rhie & Chow approach (Ferziger p.247 ?) to prevent the pressure checkerboarding?

Also, i would appreciate if anyone could tell me where, in the Patankar's book (referenced in the references), i can find something about the PRESTO.

The question about the LES is that, if there will not be stability problems, i'd like in future to implement the scale similarity approach via UDF. Specifically, to follow the approach suggested by Morinishi using a two parameter dynamic mixed model, i think that the define_source approach is not effective (because the momentum equations with their sources are solved before of the turbulence model), instead an execute_at_the_end should be the right choice. Does anyone has experience about this?

Finally, about the last question, could be useful the reading of the paper:

S.E. Kim. Large eddy simulation using unstructured meshes and dynamic subgridscale turbulence models. Technical Report AIAA-2004-2548, American Institute of Aeronautics and Astronautics, 34th Fluid Dynamics Conference and Exhibit, June 2004

suggested in the user's manual?

I know that it's a lot of stuff, but any kind of help would be appreciated. Thanks and sorry for my not so good english

Paolo

Check Issa, be warned NITA's time step is much more restrictive than Iterative... I've done reacting LES using NITA but it blows up frequently when the flow physics change faster and unpredictably. Cold flow is alright I guess in NITA, but the best way to judge is a pilot study - typically I can get three times to an order of magnitude increase of time-step using the iterative form...

Thanks Sandeep for your help, it is greatly appreciated.

I have found some of the other papers, and are not so clear; i think issa would be the right one.

If nothing will go wrong, computational time should not be an issue, at least not a big one. That is, the mine is going to be a pilot study on a cold flow.

It's interesting to know about this kind of instability; in fact i'm going to use unsteady source terms in the momentum equations. I will be careful.

Thanks again.