CFD vs. reality

bcj · March 10, 2012, 22:23

I am looking into CFD now. What I think is important to know is how CFD simulations compare to real life experiments. It seems that journals and reports compare these things rarely. How do you know that your simulation is made correctly? It may be too big of a question, but are there any books you recommend in this case?

It would be nice to get experimental data from real simulations to compare with the CFD program I am trying to learn. Where can I find such data? Are there any collection of cases like these in any book or other source?

lava12005 · March 11, 2012, 00:49

You can go to Wiki > Validation Cases. Then work from there.

pete · March 11, 2012, 03:03

On CFD Online you can start from these two places:

http://www.cfd-online.com/Links/refs.html#validation

http://www.cfd-online.com/Wiki/Valid...and_test_cases

JBeilke · March 11, 2012, 04:24

What is a "real life experiment"?

You have at least 3 different things:

1) Reality
2) Experiment
3) Simulation

If you have an experiment everybody belives in the results except the person who is doing the experiment. For a CFD calculation nobody belives in the result except the guy who runs the simulation :-)

scipy · March 11, 2012, 05:30

Quote:

Originally Posted by JBeilke

If you have an experiment everybody belives in the results except the person who is doing the experiment. For a CFD calculation nobody belives in the result except the guy who runs the simulation :-)

Haha, I see this is a pretty popular one

They say this at college a LOT. Anyone knows the origin of the saying?

sbaffini · March 11, 2012, 08:03

Well, if you get into experiments a little bit (after being into CFD) you will surely know the origin of the saying.

FMDenaro · March 11, 2012, 14:12

Quote:

Originally Posted by sbaffini

Well, if you get into experiments a little bit (after being into CFD) you will surely know the origin of the saying.

yeah

first time I've heard this saying was about 24 years ago ...

bcj · March 12, 2012, 00:49

How are the results from CFD dealt with professionally? Do you take it for granted that the results are somewhat near reality? Do you compare with different solvers? Do you validate the results somehow?

JBeilke · March 12, 2012, 02:28

The most important task is to ask the right questions to your experiment/simulation. Make sure you know what you are interested in.

You can compare this with an oracle (... greece mythology , you ask a question, get an answer and a more confused than before). If you ask the false question you get an answer/result but don't know what it means. You just have to be very clever :-)

Read the book "Surely You're Joking, Mr. Feynman!" serval times very carefully. It tells you most of the stuff you need to become a good scientist.

Your quesion about the solversettings: once you know that the complete setup is suitable and covers all the important physical effect, the right setting for the solver will improve the solution.

sbaffini · March 12, 2012, 16:22

There is this huge field which is called Verification and Validation (V&V) which is all about the verification and validation of your numerical results. Then we have the Uncertainty Quantification (UQ) field (which is exploding in the last few years) and it is all about quantifying the uncertainty in the results due to your input parameters. Then you go in some experimental Lab and they barely know where they are and what time is it (i'm joking of course but still)

March 10, 2012, 22:23	CFD vs. reality	#1
bcj New Member Join Date: Mar 2012 Posts: 2 Rep Power: 0	I am looking into CFD now. What I think is important to know is how CFD simulations compare to real life experiments. It seems that journals and reports compare these things rarely. How do you know that your simulation is made correctly? It may be too big of a question, but are there any books you recommend in this case? It would be nice to get experimental data from real simulations to compare with the CFD program I am trying to learn. Where can I find such data? Are there any collection of cases like these in any book or other source?

March 11, 2012, 03:03		#3
pete Administrator Peter Jones Join Date: Jan 2009 Posts: 682 Rep Power: 10	On CFD Online you can start from these two places: http://www.cfd-online.com/Links/refs.html#validation http://www.cfd-online.com/Wiki/Valid...and_test_cases bcj likes this.

March 11, 2012, 04:24		#4
JBeilke Senior Member Joern Beilke Join Date: Mar 2009 Location: Dresden Posts: 498 Rep Power: 20	What is a "real life experiment"? You have at least 3 different things: 1) Reality 2) Experiment 3) Simulation If you have an experiment everybody belives in the results except the person who is doing the experiment. For a CFD calculation nobody belives in the result except the guy who runs the simulation :-) bcj likes this.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
comparison CFD results with reality	chouer	Phoenics	1	November 1, 2005 04:37
STAR-Works : Mainstream CAD with CFD	CD adapco Group Marketing	Siemens	0	February 13, 2002 12:23
ASME CFD Symposium, Atlanta, July 2001	Chris R. Kleijn	Main CFD Forum	0	August 21, 2000 04:49
PC vs. Workstation	Tim Franke	Main CFD Forum	5	September 29, 1999 15:01
Which is better to develop in-house CFD code or to buy a available CFD package.	Tareq Al-shaalan	Main CFD Forum	10	June 12, 1999 23:27

March 11, 2012, 00:49		#2
lava12005 Senior Member KHB Join Date: Aug 2010 Location: Singapore Posts: 118 Rep Power: 15	You can go to Wiki > Validation Cases. Then work from there.

March 11, 2012, 08:03		#6
sbaffini Senior Member Paolo Lampitella Join Date: Mar 2009 Location: Italy Posts: 2,152 Blog Entries: 29 Rep Power: 39	Well, if you get into experiments a little bit (after being into CFD) you will surely know the origin of the saying.

March 12, 2012, 00:49		#8
bcj New Member Join Date: Mar 2012 Posts: 2 Rep Power: 0	How are the results from CFD dealt with professionally? Do you take it for granted that the results are somewhat near reality? Do you compare with different solvers? Do you validate the results somehow?

March 12, 2012, 02:28		#9
JBeilke Senior Member Joern Beilke Join Date: Mar 2009 Location: Dresden Posts: 498 Rep Power: 20	The most important task is to ask the right questions to your experiment/simulation. Make sure you know what you are interested in. You can compare this with an oracle (... greece mythology , you ask a question, get an answer and a more confused than before). If you ask the false question you get an answer/result but don't know what it means. You just have to be very clever :-) Read the book "Surely You're Joking, Mr. Feynman!" serval times very carefully. It tells you most of the stuff you need to become a good scientist. Your quesion about the solversettings: once you know that the complete setup is suitable and covers all the important physical effect, the right setting for the solver will improve the solution.

March 12, 2012, 16:22		#10
sbaffini Senior Member Paolo Lampitella Join Date: Mar 2009 Location: Italy Posts: 2,152 Blog Entries: 29 Rep Power: 39	There is this huge field which is called Verification and Validation (V&V) which is all about the verification and validation of your numerical results. Then we have the Uncertainty Quantification (UQ) field (which is exploding in the last few years) and it is all about quantifying the uncertainty in the results due to your input parameters. Then you go in some experimental Lab and they barely know where they are and what time is it (i'm joking of course but still)