Did you learn CFD in college/university?

[Q_Goest]

Back in 1988 when I graduated, I can't even remember there being a course in computational fluid dynamics (CFD) let alone taking one. If you're in college/university today or graduated recently, did you take a CFD course? I'm wondering how common it is now.

Also, if you took a course in CFD, tell me what kind of things you modeled and what software you used. Did you have a large project that took most of a semester to accomplish? What did you model and did you do anything to verify the CFD predictions experimentally such as by comparing your results to some kind of documented phenomena? I'd be especially interested in your experience in modeling things like Rayleigh-Benard convection (Benard cells) or similar phenomena.

 

[Q_Goest]

Doesn't anyone learn CFD in college?

 

[xxChrisxx]

Back in 1988 when I graduated, I can't even remember there being a course in computational fluid dynamics (CFD) let alone taking one. If you're in college/university today or graduated recently, did you take a CFD course? I'm wondering how common it is now.

Also, if you took a course in CFD, tell me what kind of things you modeled and what software you used. Did you have a large project that took most of a semester to accomplish? What did you model and did you do anything to verify the CFD predictions experimentally such as by comparing your results to some kind of documented phenomena? I'd be especially interested in your experience in modeling things like Rayleigh-Benard convection (Benard cells) or similar phenomena.

Yep, and I modeled a pipe, and a stepped pipe and a diffuser.
The idea wasn't to do anything complicated, it's to learn how the problem is modeled. Near wall models, turbulence models, solver convergence etc.
Half of it was fluid dynamics, the other half was smashing my face against fluent.

I hate both fluids and CFD.

 

[FAlonso]

We are taking a course "Computer Aided Engineering" and we only touch the basics of FEM and CFD in this course. This course is clearly not a UG course and people study it on Post graduate level mainly.

 

[Q_Goest]

Thanks guys. I'm kinda surprised but then maybe I shouldn't be. It's been my experience in industry that folks that are doing CFD (or FEA for that matter) typically specialze in it, and only a few percent of the engineering staff actually use it. I took a course on Fluent about 10 years after I graduated but never used it. I think if you don't use the software regularly, it becomes very difficult or impossible to work with.

 

[FeX32]

I have used Comsol Multiphysics at uni for CFD. Was a lot of fun actually. Anyone looking for a new software I would recommend it. I went into the "solids" side of engineering after that so I haven't touched it since.
Cheers,

 

[Q_Goest]

Thanks FeX. What was the most complicated model you made in school and how long ago was that?

 

[Astronuc]

Back in 1988 when I graduated, I can't even remember there being a course in computational fluid dynamics (CFD) let alone taking one. If you're in college/university today or graduated recently, did you take a CFD course? I'm wondering how common it is now.

Also, if you took a course in CFD, tell me what kind of things you modeled and what software you used. Did you have a large project that took most of a semester to accomplish? What did you model and did you do anything to verify the CFD predictions experimentally such as by comparing your results to some kind of documented phenomena? I'd be especially interested in your experience in modeling things like Rayleigh-Benard convection (Benard cells) or similar phenomena.

I took a class in computational fluid dynamics somewhere around 1987 while in grad school. We wrote our own code in FORTRAN.

We used Suhas Patankar's text Numerical Heat Transfer and Fluid Flow (Hemisphere Series on Computational Methods in Mechanics and Thermal Science).

Since then, the field has taken off. It was primarily a tool used by aerospace and marine engineers, but it has burgeoned into other fields.

We compared results with analytical models. The professor, who taught the course, and his grad students have compared predictions with measurements. The professor is now widely published, particularly in the area of turbulence simulation.

I think today, most students use commercial programs, and may not develop the numerical methods from scratch.

 

[FAlonso]

I took a class in computational fluid dynamics somewhere around 1987 while in grad school. We wrote our own code in FORTRAN.

We used Suhas Patankar's text Numerical Heat Transfer and Fluid Flow (Hemisphere Series on Computational Methods in Mechanics and Thermal Science).

Since then, the field has taken off. It was primarily a tool used by aerospace and marine engineers, but it has burgeoned into other fields.

We compared results with analytical models. The professor, who taught the course, and his grad students have compared predictions with measurements. The professor is now widely published, particularly in the area of turbulence simulation.

I think today, most students use commercial programs, and may not develop the numerical methods from scratch.

You have mentioned a big problem. I am about to graduate this year and we no more pay heed to the roots of the analysis softwares like ANSYS, which creates a problem in identifying garbage data when an analysis results appear. Analysis are now nothing more than setting loads, boundary conditions etc. The real procedure has somewhat totally disappeared. The end result is that analysis today is a fast job, but highly unreliable

 

[FeX32]

Thanks FeX. What was the most complicated model you made in school and how long ago was that?

No problem.
The most complicated model was a fluid-structure-interaction problem. It was a model of a human artery with a stent in place. The trick was to get the forces caused by solving the Navier-Stokes equations to cause the deflection of the artery wall.

Cheers,

 

[FeX32]

I forgot to mention that it was only a couple years ago. And that particular project was in grad school.

 

[BTown]

I just graduated in May 2011 from Purdue University and we had a few courses specifically on CFD. None of us understood what we were doing when it came to Fluent (except for the few people that read up like crazy on it and spent all their time in the lab), but on paper we all have "CFD experience".

The project I remember working on was an airfoil for a commercial aircraft.

 

[Q_Goest]

I took a class in computational fluid dynamics somewhere around 1987 while in grad school. We wrote our own code in FORTRAN.

We used Suhas Patankar's text Numerical Heat Transfer and Fluid Flow (Hemisphere Series on Computational Methods in Mechanics and Thermal Science).

Since then, the field has taken off. It was primarily a tool used by aerospace and marine engineers, but it has burgeoned into other fields.

We compared results with analytical models. The professor, who taught the course, and his grad students have compared predictions with measurements. The professor is now widely published, particularly in the area of turbulence simulation.

I think today, most students use commercial programs, and may not develop the numerical methods from scratch.

Hi Astronuc. The part about comparing predictions with measurements is interesting. Any suggestions for some really good papers that do this kind of comparison for highly non-linear phenomena? Anything would be good but especially Rayleigh Benard convection. I've seen some outputs on YouTube but I'd love to see a paper that compares CFD analysis to physical measurements!

 

[Q_Goest]

No problem.
The most complicated model was a fluid-structure-interaction problem. It was a model of a human artery with a stent in place. The trick was to get the forces caused by solving the Navier-Stokes equations to cause the deflection of the artery wall.

Cheers,

Thanks FeX. I don't suppose you ever saw papers that compared CFD results with physical measurements for these stents have you?

 

[Q_Goest]

I just graduated in May 2011 from Purdue University and we had a few courses specifically on CFD. None of us understood what we were doing when it came to Fluent (except for the few people that read up like crazy on it and spent all their time in the lab), but on paper we all have "CFD experience".

The project I remember working on was an airfoil for a commercial aircraft.

Just curious if that was a fluid-structure interaction problem. I've seen some interesting papers written on CFD analysis of wing flutter.

 

[Astronuc]

Hi Astronuc. The part about comparing predictions with measurements is interesting. Any suggestions for some really good papers that do this kind of comparison for highly non-linear phenomena? Anything would be good but especially Rayleigh Benard convection. I've seen some outputs on YouTube but I'd love to see a paper that compares CFD analysis to physical measurements!

I'll have to dig around. Really good measurements and predictions are likely proprietary, but there may be research reports and papers from NASA or DOE labs. I do know that work has been done with measurements from laser doppler anemometry with comparison to CFD predictions. I'm sure Boeing, Lockheed-Martin, Northrup-Grumman and others have done significant work in this area with respect to air flow, and other companies have done it in fluid (liquid) flow.

One area of current interest for me is conjugate heat transfer and fluid-structure interaction.

The big push now if for integrated (and fully coupled) multi-scale, multi-physics simulation in a broad range of industries and applications.

Some interesting papers, particularly the one by Smith, which gives some background on CFD.

http://www.iaea.org/NuclearPower/Technology/Meetings/2010-Dec-14-16-TM.html
http://www.oecd-nea.org/nsd/csni/cfd/workshops/XCFD4NRS/

 

[FeX32]

Thanks FeX. I don't suppose you ever saw papers that compared CFD results with physical measurements for these stents have you?

I never saw any experimental stent vorticity analysis of a real artery. I presume that this type of experiment would be difficult to attain on a actual human for reasons we can understand.
There was however, another student working on building a system to experimentally simulate this sort of thing. I'm not sure if he published anything on that experiment or not. The work I did with Comsol was for a grad class, not part of my thesis.

In general, I'm not really in that field. After my MSc I went into robotics (still grad work) so my specialty is in a different realm right now.

 

[Q_Goest]

Thanks Astro, FeX. If you stumble across any good papers just drop the name here. :)

 

[FeX32]

Will do!

 

[BTown]

Just curious if that was a fluid-structure interaction problem. I've seen some interesting papers written on CFD analysis of wing flutter.

No, we didn't go that in depth with our mesh. It was simply to optimize lift based on different dihedral angles and chamfer of the wing.

 

[OldEngr63]

Just for counterpoint, when I graduated back in 1968, ANSYS had not been invented, FEA was not being taught other than as a research method, although we did study the use of beam elements only as derived from beam theory, and CFD was only a twinkle in the eyes of a few fanatics. Obviously much has changed in the intervening years.

 

[sareyes91]

I took a class in computational fluid dynamics somewhere around 1987 while in grad school. We wrote our own code in FORTRAN.

We used Suhas Patankar's text Numerical Heat Transfer and Fluid Flow (Hemisphere Series on Computational Methods in Mechanics and Thermal Science).

Since then, the field has taken off. It was primarily a tool used by aerospace and marine engineers, but it has burgeoned into other fields.

We compared results with analytical models. The professor, who taught the course, and his grad students have compared predictions with measurements. The professor is now widely published, particularly in the area of turbulence simulation.

I think today, most students use commercial programs, and may not develop the numerical methods from scratch.

Hi Astronuc,

What you said in regards to students nowadays not developing the numerical methods from scratch, is not entirely true. I am currently taking an introduction FEA course where the professor teaches us all the theory and background that programs like ANSYS and pro/e use. All of this while introducing us to ANSYS. In fact, one of our homeworks is to write a program that solves one and two dimensional structural mechanics problems. I'm still in the process of writing it.

Although it might just be our professor, he is very old fashioned.

 

[OldEngr63]

Although it might just be our professor, he is very old fashioned.

Sounds to me like you have a very good professor who thinks you need to learn more than simply how to push buttons. You will thank him for this later.

 

[Astronuc]

Hi Astronuc,

What you said in regards to students nowadays not developing the numerical methods from scratch, is not entirely true. I am currently taking an introduction FEA course where the professor teaches us all the theory and background that programs like ANSYS and pro/e use. All of this while introducing us to ANSYS. In fact, one of our homeworks is to write a program that solves one and two dimensional structural mechanics problems. I'm still in the process of writing it.

Although it might just be our professor, he is very old fashioned.

I fear this is an exception in our current educational system.

I've heard developers and sellers of code systems say that "it is not necessary for the user to know or understand the details of the code". I vehemently disagree. Usually, such a statement really means "we don't won't users to look under the hood such they know how to do this." It is a matter of controlling technology and IP, such that users have to keep paying license fees and buying technical support year after year.

It's very important to be able to write the equations, both analytically and numerically, and be able to solve them. Theoreticians write equations (and develop models), and developers write the numerical equations and solution technique/system, as well as develop models. Developers create the tool. Then it's possible that the user will use the tool, often times without knowing the 'guts' of the tool.

My experience is a bit unique in being developer/theoretician and user.

One also has to familiar with solution techniques, computational theory, and some degree of hardware contol, and how all that effects the results of the computation.

Computational multiphysics (particularly multiscale) is a rather complex and challenging field, but it's a lot of fun.

 

[Exocet]

I'm in first year at university in England doing ME, and while I haven't met any CFD yet I know that there is a course in second year which runs across both semesters which teaches FORTRAN and MATLAB programming and basic CFD and FEA using FLUENT and ANSYS. This continues with another compulsory module in modelling computation and simulation and an optional advanced CFD course in 3rd year.
I don't know about you, but that sounds like an awful lot of computing :P

 

[sareyes91]

I'm in first year at university in England doing ME, and while I haven't met any CFD yet I know that there is a course in second year which runs across both semesters which teaches FORTRAN and MATLAB programming and basic CFD and FEA using FLUENT and ANSYS. This continues with another compulsory module in modelling computation and simulation and an optional advanced CFD course in 3rd year.
I don't know about you, but that sounds like an awful lot of computing :P

That sounds awesome, I didn't even know they still taught Fortran. In my FEA course right now we had to write a program (in the language of our choosing) to solve one dimensional truss problems. I wrote it in MATLAB, so much easier

 

[hari_srt]

Yes, i undergone a cfd course and am in a job related to the same.i wish to do M.sc in the same field,i want to do it in Uk or in Aus,Can anyone of you help me.

i had worked on proj related to drill bits,valves etc.,

 

[Vadar2012]

Nice thread necro mate, but I can give you some advice.

I've done a buttload of CFD. I've did my undergrad and Masters theses on CFD. More specifically modelling the aerodynamic and thermodynamic conditions on reentry aircraft and scramjet combustion chamber research.

I think the best way to improve your CFD skills is do a Masters thesis in the field. Alot of Australian Universities are good at it. Such as UQ, UNSW, USYD, and ANU. I can't help you with Canada.

 

[Q_Goest]

Hi Vadar,
Are you aware of any papers that compare CFD analysis to experimental results in an effort to prove the accuracy of some CFD code? Anything would be good but especially Rayleigh Benard convection. I've seen some outputs on YouTube but I'd love to see a paper that compares CFD analysis to physical measurements!

 

[Vadar2012]

I have a tonne of journal articles that compare CFD simulations to experimental results. I have done 2, 200 page thesis papers on it. Although my field was mostly exploring the flow affects inside the combustion chamber of a scramjet and then comparing them to experimental results. This is far more interesting than it sounds, some freaky stuff goes on inside there. I proved 50 years worth of thinking wrong. Was quite fun.

I might have a paper on that though, give me some time to have a look. I don't have them with me at work.

 

[hari_srt]

thanks you vadar, I didnt get correctly what masters thesis means,BTW i want to do a full time course in cfd and am confused to select the branch in it...
can you help me in figuring out the best modules or the best future scope of the branches etc.,
and in aus what will be the time period for completing the M.Sc.

Regds.
Hari

 

[bigfooted]

If you want to find CFD/experiment comparisons and your library does not have a subscription to sciencedirect or springer (most university libraries do however), go to the website of a university library that puts all their phd theses online and search for Rayleigh benard. It is one of the most studied fluid flow phenomena around, so I am surprised that you have a hard time finding something useful. Try it for instance here, I know it was studied in the lab where I worked:
http://repository.tudelft.nl/

 

[Q_Goest]

Hi bigfooted. I'm not in college. I graduated in 1988. I've looked around for papers through Google Scholar but haven't found anything compared a CFD analysis with experimentation. If you know of any good papers that provide that comparison, I'd really appreciate a lead. I can get papers but I haven't found anything, even on the website you provided.

 

[Vadar2012]

What do you have access to? If you got access to AIAA and the likes, maybe I can just give you some titles instead of spending ages uploading them.

 

[Q_Goest]

What do you have access to? If you got access to AIAA and the likes, maybe I can just give you some titles instead of spending ages uploading them.

If you have titles (and author/s), that would be terrific. Some of the journals I can get and my daughter is just starting college so she might also have access. At any rate, I should be able to find them if I have a name.

 

[Vadar2012]

I found a couple decent ones I somehow have with me at work:

ShockWave/Transitional Boundary-Layer Interactions in Hypersonic Flow
R. Benay, B. Chanetz, B. Mangin, and L. Vandomme
AIAA JOURNAL
Vol. 44, No. 6, June 2006

Separation length in high-enthalpy shock boundary-layer interaction
Jean-Paul Davisa and Bradford Sturtevant
Graduate Aeronautical Laboratories, California Institute of Technology, Pasadena, California 91125

The first one is especially good. I spent a year replicating its experiment and CFD results using a new NASA program. I was able to get much better CFD results, which allowed me to explain the weird stuff they were seeing in the experiment. It's also good for comparing newly compiled CFd program results to for verification.

 

[Vadar2012]

Hi Astronuc,

What you said in regards to students nowadays not developing the numerical methods from scratch, is not entirely true. I am currently taking an introduction FEA course where the professor teaches us all the theory and background that programs like ANSYS and pro/e use. All of this while introducing us to ANSYS. In fact, one of our homeworks is to write a program that solves one and two dimensional structural mechanics problems. I'm still in the process of writing it.

Although it might just be our professor, he is very old fashioned.

I just wanted to add some stuff to this. You learn pretty much nothing in these classes. I've done a class at uni (multiple) that say they teach you the background and how they work, but if you do any research in this field. You'll see they don't teach you anything in comparison.

 

[boneh3ad]

Addressing the idea of validating computations with experiments, this is a practice that is done constantly in the scientific community. In fact, in general a CFD solution can never be truly validated without either a direct comparison with either an analytical solution (very, very rare) or experiments. Naturally, experiments are the typical route here when possible.

Lately, Physics of Fluids has had a lot of work on Rayleigh-Bénard convection. Just browse through recent issues of that and you ought to find a pretty nice chunk of work which will, of course, reference other, more canonical works in the field.
http://pof.aip.org/search?key=PHFLE6&societykey=AIP&coden=PHFLE6&q=Rayleigh-B%C3%A9nard&displayid=AIP&sortby=newestdate&faceted=faceted&sortby=newestdate&CP_Style=false&alias=&searchzone=2

The thing about that particular phenomenon is that, at least today, there is not going to be a lot of experimental data on simple Rayleigh-Bénard convection because it is already a well-studied problem and in many cases it has an analytical solution for the interesting bits of the phenomenon such as the hydrodynamic stability. You may have better luck finding data for variations of the traditional Rayleigh-Bénard problem.

The other thing to keep in mind is that quite often, you will see a numerical work in a separate paper from its experimental counterpart, especially because oftentimes one inspires the other.

I will also add that during my undergrad, FEA and CFD courses were offered by my university as electives that attempted to do a mix of the background theory (e.g. Galerkin methods) while also providing familiarity with various software (generally ANSYS and Fluent). Of course it is a class so, as Vadar2012 pointed out, it will still never quite be the same as what you learn from reading a couple papers and then doing it. Then again, most undergraduates don't have that opportunity so the class is at least a good start.