Hi,everybody,

I am a beginner in writting a CFD code for unsteady flow,I'm not sure whcih primitive variable algorithm(SIMPLE, MAC,ect.) is better(accurcate,or easy to program) for fullfil my task while using the multigrid method,people who have experence on unsteady flow and MG,please give me some advice.

Thank you very much and waiting your reply.

andy

Hello,

Mac method was developed by Harlow and Welch (1968) at las alamos laboratories. It is based on primitive variable formulation. It is based on Explicit scheme and pressure poisson equation approach. Simple method was developed by S.V. Patankar based on Semi-implicit scheme, pressure correction scheme. Simple algorithm is somewhat difficult comparitively to Mac solver. Now a days, people are using simple method. You can download the MAC algorithm from "www.lanl.gov".

bye.....

Dear Srinivasulu,

Thank you for your help,what you said is right,the MAC which uses explicit scheme is too much simple,but the explicit scheme maynot be work well for high Re number(10e5--10e6),in this case,I thank the implicit scheme must be used and may result in difficulty.compared with explicit scheme.

Additionally,my other care is which algorithm can efficient be used with MG,in SIMPLE or SIMPLER,the calculation on each time step are splitted into too much substeps,so it need too much computer resource and programming, I guess.

andy

You

can download the MAC algorithm from "www.lanl.gov".

<= Where can I download at www.lanl.gov?

I visited there, but I couldn't find it.

Hi Andy

I have my own codes for SIMPLE, SIMPLEC, SIMPLER. And the best algorithm for unsteady problems is modified SIMPLE. I write the code for fully explisit SIMPLE (it easy to do) and it is the quickest algorithm to calculate (you just have to use rather small time step).

Evgueny

Hello,

you search in that website at library online catalogue. Give words "MAC" for title. Hope, you will get it there.

bye.........

Hi Evgueny,

Accidently I read your message on "unsteady SIMPLE". I am carrying out unsteady calculations with a fully implicit time discretisation with a SIMPLE. If the time step becomes rather small, I have serious stability problems; I guess this is due to the fact that the "Rhie and Chow" coupling term becomes small (I use co-located variables). So what I am interested in is an explicit method, in particular I thought about MAC. So I am wondering:

- How does an "Explicit SIMPLE" work; by now I was sure that you need to use an implicit method with the SIMPLE

- How can you guarantee divergence free flow, if you solve the momentum equations explicitely, and not solve the Poisson equation for the pressure EXACTELY (this would be something like MAC)

- What is the difference between your Algorithm and MAC?

I would be very grateful for some hints.

Best regards Romuald