|
[Sponsors] |
August 25, 2006, 09:54 |
Fullly Coupled Solver
|
#1 |
Guest
Posts: n/a
|
Could anybody please brief anything about Fully Coupled Solver. How pressure correction is achieved in that. What I feel is it is unlike our SIMPLE,SIMPLER algorithms in Segregated solvers. Pl help
|
|
August 25, 2006, 20:27 |
Re: Fullly Coupled Solver
|
#2 |
Guest
Posts: n/a
|
As far as I know a solver can be considered fully coupled when you solve all degrees of freedom in the same linear system. Most of the NS solvers are based in some pressure correction scheme like SIMPLE, SIMPLER or SIMPLEC where the pressure and velocity are decoupled and solved in a segregated way sweeping between the equation until a converged solution is reached (the velocity condition satisfies the pressure solution or vice-versa). A typical NS solver structure appears like this:
TIME INTEGRATION .....NON-LINEAR SOLUTION ..........LINEAR SOLUTION (*) .....END NON-LINEAR SOLUTION END TIME INTEGRATION (*) here you can solve 4 dofs (ux, uy, uz, pres) at the same time (coupled) or the equations can be solved separetely in some way (segragated). Hope it helps Cheers Renato. ps.: Take a search on the CFD-online wiki page that I'm sure you'll find a better explanation ;o) |
|
August 26, 2006, 00:59 |
Re: Fullly Coupled Solver
|
#3 |
Guest
Posts: n/a
|
The book by Anderson, describes and develops a 2D coupled solver for the Navier Stokes equations
|
|
August 26, 2006, 09:40 |
Re: Fullly Coupled Solver
|
#4 |
Guest
Posts: n/a
|
It has given good insight. I could get some information on wiki page. I wanted to know more about pressure based fully coupled finite elment based incompressible NS solver. Any link available will be of great help. Thanks a lot
|
|
August 30, 2006, 15:15 |
Re: Fullly Coupled Solver
|
#5 |
Guest
Posts: n/a
|
Very interesting question.
To me, one usually speaks about two kinds of coupling at two different levels. At the physical level, one has or not to consider density coupled to other variables if the Mach number is high or low (incompressible or variable density flows). It has nothing to do with coupled solver where, effectively, one can solve the linear system of equations in a block instead of sequentially. Any comments? |
|
September 3, 2006, 22:48 |
Re: Fullly Coupled Solver
|
#6 |
Guest
Posts: n/a
|
You are right that if Mach No. is high, physically, one has to solve for density, the flow being compressible. We also have to solve for energy equation even for pure flow analysis. However, as far as I know, in Pressure coupled solver we do not solve for Ideal Gas equation and thus we do not consider density as a variable. We do not decouple pressure and velocity. We solve it in a block and do not use SIMPLE, SIMPLER etc for pressure correction. Now the question is how we correct the pressure if we do not consider Ideal Gas equation? I may be wrong somewhere, please correct me. Thanks.
|
|
September 5, 2006, 09:09 |
Re: Fullly Coupled Solver
|
#7 |
Guest
Posts: n/a
|
Pressure coupled incompressible codes are relatively seldom and I do not know them in details.
However, I do not see why, in principle, the Poisson equation and the momentum equations cannot be solved in block instead of sequentially. In practice, there might be some specific issues I do not know. |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
why CFL is only in coupled solver? | john | CFX | 6 | May 10, 2007 11:27 |
Can I use coupled solver for this problem | Frank | FLUENT | 0 | April 11, 2006 06:28 |
Coupled or Seggregated solver | Ejaz Ahmed | FLUENT | 4 | July 15, 2005 12:35 |
Mixing plane - coupled solver | Pawel | FLUENT | 0 | February 10, 2005 06:17 |
coupled solver / uncoupled solver | Jaan Unger | Main CFD Forum | 0 | September 3, 2002 08:30 |