CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   Main CFD Forum (http://www.cfd-online.com/Forums/main/)
-   -   I am surprised by a CFD program (http://www.cfd-online.com/Forums/main/15576-i-am-surprised-cfd-program.html)

Jimmy August 15, 2008 17:24

I am surprised by a CFD program
 
It takes within a minute only for a cfd program to complete the plenum simulation of a data center. How can it be so fast?

Robin August 15, 2008 18:28

Re: I am surprised by a CFD program
 
What was the cfd program called?

Jimmy August 15, 2008 18:48

Tileflow
 
do u think it can have a converged result in one minute? or does this program really do a decent mesh?

Ahmed August 15, 2008 23:24

Re: I am surprised by a CFD program
 
That should not surprise you, solving air problems at the conditions described by you does not require the solution of the compressible Navier Stokes equations, just the incompressible ones, and if you add to that a Cartesian mesh generator then the mystery is out. You need to compare apples with apples, have you tried to solve the same problems using AirPak, just to mention one competing programme

Humm August 16, 2008 00:29

Re: I am surprised by a CFD program
 
and here i am spending hours for solving incompressible flow problems using fluent. Silly me.

Jimmy August 17, 2008 00:17

one more question
 
are there any options like setting the problem as compressible or incompressible in the commercial CFD codes?

Ahmed August 19, 2008 20:40

Re: one more question
 
When I read your first posting, the impression I got is that you are a new sales engineer trying to call the attention of this forum's readers to your product ( do not worry about that, others try to do the same), but when I read your last posting and the one that precedes it, I realize that you have an utterly serious defficiency in your academic preparation (sorry to say that on a public forum) The difference between compressible and incompressible flows is very well explained in all texts and references about fluid mechanics. Once you know the difference, you will find out that all codes provide mechanisms for all flow situations. As you see, implementing the condition for constant density is straightforward, so check the documents of your particular code or give your support engineer a phone call.

nose poker August 19, 2008 22:40

Re: one more question
 
I realize that you have an utterly serious defficiency in your academic preparation (sorry to say that on a public forum) The difference between compressible and incompressible flows is very well explained in all texts and references about fluid mechanics.

I am sorry to poke my nose in here but it seems that the only person who has deficiency in academics is you.

Could you point me to any article that says that if the flow is incompressible and if one is using Cartesian grid, flow problem could be solved in a minute, no matter how large the size of problem.

Even if navier stokes is written in incompressible form (which makes it harder to solve by the way), you can not write analytical solution to it. So the way to solve is to use iterative solution. This is why Fluent, starCCM+, CFX take hours to converge to some solution. If you have magic formula than you could make millions. Go ahead.

And yes, do not insult people if you do not understand question. It kinda shows your intelligence (or lack of it).

So you have anything useful to add, then please write.


nose poker August 19, 2008 23:15

Re: I am surprised by a CFD program
 
Jimmy by looking at one of the papers in technical paper section downloads. It seems they are using some formula to approaximate pressure drop. So the flow domain is devided into tiles or cartesian block. Then based on air flow coming into it, they calculate the pressure drop (by some formula) then move to next tile and so on. This way by the time they finished all the tiles, they have flow distribution and pressure drops. Giving you velocity and pressure distribution.

(this is what i gathered from my 5 minutes look at tileflow's site). _May be more to it, but i do not have much time to look at it now.


Jimmy August 20, 2008 11:57

Thanks much for your comments *NM*
 


All times are GMT -4. The time now is 02:14.