backtop


Print 72 comment(s) - last by spread.. on Oct 4 at 4:29 PM


MIT Professor Tom Peacock works in the lab where his team verified the new fluid theory.  (Source: Donna Coveney)

MIT visiting professor George Haller cracked the greatest fluid dynamics mystery which eluded Ludwig Prandtl and other researchers for the last century.  (Source: MIT)

This year brought the biggest breakthough with researchers extending the unsteady state fluid flow equation to three dimensions and verifying the results.  (Source: Amit Surana, Gustaaf Jacobs and George Haller, MIT)
New insight into fluid behavior may allow for better fuel economy, better golf balls

Fluids -- liquids and gases -- are such an integral part of our everyday lives that we often don't even give them a second thought.  However, for engineers, the behavior of fluids bears great importance.  Fluid flow affects everything from the miles per gallon we get in our car, to how well a golf ball flies through the air.

Now a key breakthrough in understanding of fluid modeling has been achieved.  For years the key equation in the world of fluid mechanics was the Prandtl equation, developed by Ludwig Prandtl, which described how air and water flowed over objects.  Despite its brilliance in 1904 when it was conceived, it had serious limitations -- it only worked for steady flow, such as a car traveling at low speeds, and it only applied to idealized 2 dimensional problems.  For decades researchers tried to improve the equation to little avail.

Solutions obtained often diverged greatly from real world mechanics.  For example, the air flow around a car making a hairpin turn, would often fall off, unable to keep up -- something the Prandtl equation could not explain.

This was a sizable problem as optimizing fuel flow is extremely important to many applications.  One perfect example is Speedo's quest for the perfect swimsuit, which was showcased in its new designs which made their official mark on the Beijing Olympics.  Another example is in car aerodynamics.  Cars are sculpted to try to make airflow less unsteady.  In an optimal scenario air would just glide across the car's surface and reform into a steady stream.  In the real world air flows off the car in a turbulent stream akin to a boat wake, and separates from the surface as it passes over the car.  By minimizing these effects, fuel economy can be improved.

However, nothing could explain exactly how these unsteady-state behaviors worked -- until now.  MIT's George Haller, a visiting professor in the Department of Mechanical Engineering, developed a theory which applies to 3 dimensional unsteady state flows. This was confirmed with the help of Thomas Peacock, the Atlantic Richfield Career Development Associate Professor in the same department, who led experimental efforts to validate the results.

The new work -- if it survives the extensive peer review that is to come -- will likely go down as the greatest scientific advance of the decade. The research has already survived a strenuous initial round of peer review.  Papers on the theory and experiments will be published in the Journal of Fluid Mechanics and in the September issue of Physics of Fluids, respectively.

Professor Haller's quest began when in 2004 he devised an equation for unsteady state in two dimensions.  Having remedied half the shortcomings of Prandtl's equation, he set to work trying to extend the equation into three dimensions.  Four years later, his dream has finally been achieved.  Assisting Professor Haller in his research and coauthoring the paper were Amit Surana, now at United Technologies; MIT student Oliver Grunberg; and Gustaaf Jacobs, now on the faculty at San Diego State University.

Professor Peacock says the experimental verification is equally important, though, stating, "While we fully trust George's new mathematical results, the engineering community is usually skeptical until they also see experimental results."

Professor Haller concurs, stating, "While giving a beautiful validation of the 2D theory, Tom's work also gives strong experimental backing to our new 3D theory."

The experimental work was coauthored by Haller, Jacobs, Matthew Weldon, and Moneer Helu.

Having reached a solution, scientists can now begin to use it to optimize their systems.  The equation will forever change the face of advanced fluid dynamics and will have a profound impact on many industries, including the aerospace and automotive industries.  Professor Peacock states, "This is the tip of the iceberg, but we've shown that this theory works."

The research received initial funding from MIT's Ferry Fund.  It is now funded by the Air Force Office of Scientific Research and the National Science Foundation.



Comments     Threshold


This article is over a month old, voting and posting comments is disabled

Examples?
By InvertMe on 9/26/2008 10:43:22 AM , Rating: 3
Could I get a little more detail here? What are some real world applications of where this would be used, how it would make the current process better and how complex is this to actually put into "production".




RE: Examples?
By JasonMick (blog) on 9/26/2008 10:51:27 AM , Rating: 4
One example might be by correctly modeling aerodynamics, we can accurately see the wind resistance of car design 1 v. car design 2. To do this, you would model the car in 3D, then submit it with its physical characteristics into a physics simulator, which would now properly model unsteady wind flow. More scenarios could be tested, such as highway speeds and sharp turns, thanks to the new eqns. In the end, this will a large effect cumulatively, though it may only allow cars to increase a few mpg. Basically it will have a small impact on a whole lot of fields.

It will, however likely be computationally intense to run such 3D simulations, and they're inherently more complex than 2D simplifications. It will likely take a little time to properly apply this, but the greatest obstacle to its full potential is computing power.


RE: Examples?
By Amiga500 on 9/26/2008 11:34:17 AM , Rating: 3
May I suggest you google Direct Numerical Simulations (DNS)?

We've been modelling the full turbulence field for some time now.


RE: Examples?
By Suntan on 9/26/2008 11:59:42 AM , Rating: 2
Yes, we have.

But whenever the analysis is complete, all of us look at each other and the first question that gets asked is, “Ok, are these results even close to what we see in real life?”

I don’t think you will find too many people, working on CFD in the real world, that would argue we have a long way to go in this type of analysis.

Unfortunately, CFD is extremely dependent on knowing the boundary conditions of a system. Guessing those incorrectly often leads to much more divergent results than the underlying calculations.

-Suntan


RE: Examples?
By Amiga500 on 9/26/2008 12:14:17 PM , Rating: 4
Erm...

Are you getting mixed up between DNS and RANS based turbulence modelling?

DNS is real life. Its the full simulation, no modelling made.

We've a very long way to go in improving turbulence modelling, be that through DES or LES based methods, but if we could improve computing power by a factor of 10,000 overnight, everyone would just start using DNS and then do the time averaging automatically through a post-processor.


RE: Examples?
By Jimbo1234 on 9/26/2008 1:43:03 PM , Rating: 2
As the old saying goes (usually when working with ANSYS) "garbage in, garbage out."


RE: Examples?
By MrDiSante on 9/26/08, Rating: 0
RE: Examples?
By vandalizmo on 9/29/2008 7:26:11 AM , Rating: 2
Wiki:
quote:
A direct numerical simulation (DNS) is a simulation in computational fluid dynamics in which the Navier-Stokes equations are numerically solved *without* any turbulence model.


Hmm...


RE: Examples?
By johnadams on 9/26/2008 12:57:49 PM , Rating: 1
Maybe they could hook up a cluster of Playstation 3s to do the modeling.


RE: Examples?
By b534202 on 9/26/2008 1:51:54 PM , Rating: 5
I bet Formula 1 teams are licking their lips on the possibilities. Aero simulation over a whole track and even with cars in dirty air going into corners.


RE: Examples?
By Sunrise089 on 9/26/2008 2:36:57 PM , Rating: 5
I was thinking the same thing. Still, the current modeling techniques can't be all bad, since teams still invest in some of the world's most powerful computers rather than more wind tunnel time.


RE: Examples?
By cingkrab on 9/27/2008 11:57:27 AM , Rating: 2
Well, that's because the FIA has set a limit on wind tunnel use for 2008. CFD limits are coming as well.

http://www.autosport.com/news/report.php/id/64312


RE: Examples?
By Regs on 9/29/2008 9:45:09 AM , Rating: 1
Imagine air liners. Wonder how much jet fuel they can save? The funny thing is they'll likely use the cost savings of fuel by cramming more passengers into the plane.


RE: Examples?
By stromgald30 on 9/30/2008 2:25:52 AM , Rating: 2
I have to agree. Like others have said, this modeling may be great and continue to minimize/eliminate the need for wind tunnel testing, but overall, I don't see this as big of an impact as reported.

Through wind-tunnel testing and empirical data, engineers have already gotten alot of inefficiency out of fluid flow designs. Although we haven't had the theory to back it up, the empirical data has been enough to know how to modify/improve designs. IMO, this will probably only net a 5-10% boost, if that.


RE: Examples?
By Fnoob on 9/28/2008 12:41:14 PM , Rating: 2
I wonder if Formula 1 regulations will ever allow teams to use dynamic body surfaces (if they already do, I missed it). For example, a different spoiler configuration 'on the fly' for when the car is out in front, side by side, trailing, etc. Seems that would make a bigger difference than simply improving the CD one percent or so.


RE: Examples?
By snownpaint on 9/29/2008 10:38:49 AM , Rating: 2
F1 racing has pushed out the cutting edge trill of high speed racing.
Personally, I feel they should set a Dollar limit on car design and such.. 1/2 billion dollars. Make your car go as fast as possible, with the coolest, greatest technology. This will push developing technologies in the industry even farther and faster.. Right now the new rules have handicapped design. It might as well be Nascar. With great technology, comes big technology sponsors that want to attach their companies names to impressive machines.


RE: Examples?
By abhaxus on 9/30/2008 12:14:13 PM , Rating: 2
They do not allow for movable aerodynamic devices or active suspension anymore because the cars simply would be too fast. Without some serious G-suits the drivers would black out if the cars were pushed to their limits.


RE: Examples?
By SnakeBlitzken on 9/29/2008 10:23:20 AM , Rating: 1
Yes, but what will the car have to look like? If it turns out to be a 57 cadillac with large tail fins than that's OK. If it's a javeline with wheels, forget it.


RE: Examples?
By theapparition on 9/26/2008 11:35:34 AM , Rating: 5
I had the opportunity to engage Professor Haller at a recent CFD syposium. His work on the 2D Prandtl equations was incredible, as it improved upon existing equations inconsistancy on transient responses. He told me about his work with 3D equation, although the matrix computing required would be large, he belived a better understanding of the full equations could no only lead to better designs, but could also lead to systems that could optimize designs.......rather than they way we do it now, create a design, analyse it, then tweak and re-analyze. Upfront optimization could signifigantly improve how products are created.
However, we are far away in computing power and applications that can take advantage of this. But the first step is to create the math. I look forward to reviewing his work.


RE: Examples?
By derwin on 9/26/2008 6:15:29 PM , Rating: 2
It still sounds like you are solving a backwards problem, which often (depending on how good your guess really is) takes a very long time and can often find local mins as your solution.

I would imagine in all but the simplest objects (like a golf ball) the parameter space would just be too enourmous to even think of finding good solutions anywhere but almost exactly where your guesses were in the first place. So, I suppose you could optimize a particular design, but like people have said, that will only give marginal improvments.

Perhaps I am mistaken. I am a physics undergrad, so I never deal with any more complex situations than a helium atom, or a basic square well polymer grating, but as far as my knowledge goes, this sounds very much like a backwards problem with infinite parameter space.


RE: Examples?
By theapparition on 9/27/2008 12:21:53 PM , Rating: 2
No it doesn't have to be an infinite parameter space. You can reasonalby set boundary conditions and utilize symmetry to reduce the complexity.
For current CFD analysis, equations are solved based on iterative methods. The more inaccurate the equations, the more iterations required to balace the force and energy requirements.....and the more inaccurate the results. Improving releationships in a 3D space should also help to signifigantly reduce calculations.

If so, that leads us to the next phase. Previously unthinkable was the idea of design optimization, but with more accurate flow models, the thought of true computer simulation becomes a possibility. We still need 10000x computing power, but that better than the virtually impossible that was thought several years ago.


RE: Examples?
By Ratinator on 9/26/2008 2:42:49 PM , Rating: 3
I have little knowledge in the ares of aerodynamics etc....but one thought that comes to mind is the R&D areas. With a more accurate equation I would venture to guess R&D costs could come down in some areas.

With the inaccuracy of the previous equations I would venture to guess the cost R&D costs were higher as they would have continuously build re life examples of the product, test it, remodel it, rebuild it, test it, remodel it....and on and till they got it right. Now they could more accuractly produce the real world scenario at the computer modelling phase. This would save on costs of having to build full version models as often and reduce the testing times required. This could mean initial cost when a product is released would be cheaper.


RE: Examples?
By Ratinator on 9/29/2008 11:55:14 AM , Rating: 2
Me no type English well. I really need to proof read my posts before posting.


RE: Examples?
By MrPoletski on 9/27/2008 8:12:20 AM , Rating: 4
A little more detail?

Friend, you really do not want to see the equations involved in fluid dynamics. The reason they are not posted on news sites is for health and safety reasons. I am still recovering from the synapse hemorrhage from last time I tried to read them.


Really?
By Mitch101 on 9/26/2008 10:37:58 AM , Rating: 6
quote:
if it survives the extensive peer review that is to come -- will likely go down as the greatest scientific advance of the decade.

Sorry I believe it pales in comparison to the following research. ;)
http://www.style.org/unladenswallow/




RE: Really?
By novacthall on 9/26/2008 11:18:19 AM , Rating: 5
If his isn't a 6-worthy post, I don't know what is.


RE: Really?
By BladeVenom on 9/26/2008 11:55:41 AM , Rating: 2
But this discovery will allow us to make more aerodynamic Swallows and coconuts. Swallow will one day be able to carry coconuts.


RE: Really?
By quiksilvr on 9/26/2008 12:10:16 PM , Rating: 2
Haha...references...


RE: Really?
By Mojo the Monkey on 9/26/2008 12:27:26 PM , Rating: 4
It could grip it by the husk!


RE: Really?
By AnnihilatorX on 9/26/2008 1:31:27 PM , Rating: 5
MIT Discovers "Breakthrough of the Decade", Could Affect Everything From Swallows to Coconuts


RE: Really?
By gsellis on 9/29/2008 11:01:57 AM , Rating: 2
Now that is a title worthy of Fark.


RE: Really?
By DeepBlue1975 on 9/26/2008 4:45:03 PM , Rating: 2
Awesome!!

Though that info is actually a bit hard to swallow.


Affect vs. Effect
By Gondorff on 9/26/2008 10:29:52 AM , Rating: 5
...in the title. Please fix it before my eyes start bleeding.




RE: Affect vs. Effect
By JasonMick (blog) on 9/26/2008 10:42:38 AM , Rating: 1
You might want to get that looked at.

"Bleeding out of the eyes when viewing partial synonyms could be a sign of a rare but serious condition and should be reported to your doctor right away."


RE: Affect vs. Effect
By kickwormjoe on 9/26/2008 11:44:13 AM , Rating: 2
affect = verb
effect = noun


RE: Affect vs. Effect
By Hlafordlaes on 9/26/2008 1:23:02 PM , Rating: 5
affect = verb
effect = noun

Both can be used as verbs:

affect: influence
effect: make happen


RE: Affect vs. Effect
By Jimbo1234 on 9/26/2008 1:47:35 PM , Rating: 5
"Affect" is used correctly in the title.

Examples:
The effects (nouns) of sitting in your parents' basement all day long are staggering.

Living in your parents' basement can severely affect (verb) your chances to procreate.


RE: Affect vs. Effect
By mindless1 on 9/28/2008 6:45:05 PM , Rating: 2
Not really, both effect and affect are properly used as verbs but affect is best used only to describe an emotional or cognitive impact.

Gravity effects a car.
Love affects your heart.


RE: Affect vs. Effect
By foolsgambit11 on 9/26/2008 2:27:38 PM , Rating: 2
False. Examples:

Verbs:
To effect change, we must act quickly.
Gerrymandering was used to affect the results.

Nouns:
It was used to great effect.
Keyser Söze's limp was merely an affect.

In this case, the title is correct. 'Could affect' here means 'could influence'. If the title read, '[Breakthrough] could effect everything from cars to golf', then it probably wouldn't make any sense. But it would mean that the breakthrough would cause cars and golf to come into existence, or that cars and golf are the result of the breakthrough.


RE: Affect vs. Effect
By omnicronx on 10/1/2008 3:03:56 PM , Rating: 2
He changed the title..


This has been blown way out of proportion.
By Amiga500 on 9/26/2008 11:32:33 AM , Rating: 5
Yes, it is very good work. However, it is not going to revolutionise aerodynamics.

For instance, the technique used here constructs a time average of the fluctuating zero shear stress location on the surface to estimate a 'steady' separation point.

Now, try to extrapolate that to commericial aircraft and it won't quite work. The number 1 source of lift independent drag on current aircraft is shock-induced boundary layer separation on the wing, the shockwave typically oscillates significantly (can be over 10% chord length) along the wing, and the separation point moves accordingly. This is directly at odds with the quasi-'steady' separation point calculated using this work.

In summary, it is useful to us, but it isn't exactly going to turn the world of aerodynamics upside down. :-)




RE: This has been blown way out of proportion.
By Amiga500 on 9/26/2008 11:52:32 AM , Rating: 4
Footnote:
I've been reading the 2008 paper "Experimental and numerical investigation of the kinematic theory of unsteady separation".

This is substantially different to the 2006 paper "Exact theory of three-dimensional flow separation. Part 1. Steady separation", which I've just stumbled across.

The 2006 paper is substantially stronger than the (May) 2008 one.

However, the earlier point still stands - this is a very long way off turning fluid dynamics upside down. We've known the mechanism of transition from laminar to turbulent boundary layers for ages now... doesn't mean we can prevent it happening.


By LTG on 9/26/2008 2:28:41 PM , Rating: 5
Will you quit stating the obvious any 3rd grader knows that stuff.


RE: This has been blown way out of proportion.
By ElBrujo on 9/27/08, Rating: 0
By jskirwin on 9/27/2008 10:35:05 AM , Rating: 2
He's making a joke by being ironic. Amiga is obviously an expert - unlike some of us who weren't physics majors. It's a Gen-X thing.


WTF?
By mattclary on 9/26/08, Rating: 0
RE: WTF?
By phxfreddy on 9/26/08, Rating: -1
RE: WTF?
By Titanius on 9/26/08, Rating: 0
RE: WTF?
By BansheeX on 9/26/2008 12:00:02 PM , Rating: 5
Well, you'd be wrong since a some people have surpassed 120.


RE: WTF?
By spread on 10/4/2008 4:29:04 PM , Rating: 2
It should be:

There is a 100% chance of death... eventually.


RE: WTF?
By Chernobyl68 on 9/26/2008 12:08:55 PM , Rating: 3
I'll take that bet. I'd wager that one of the millions of kids born in the last year will live to 101.
Of course, I won't be around to collect... ;)


RE: WTF?
By niva on 9/26/2008 2:20:48 PM , Rating: 2
His prediction was that all people will die, not just the ones currently alive who might reach 100+ years. As the poster above said, the sky is falling!


RE: WTF?
By TheFace on 9/26/2008 4:00:19 PM , Rating: 2
The statement is colloquial. It could be seen as a reference to the whole of humanity, or just as a statement saying that no person shall achieve the age of 100.

Which is it? Apocalypse or longevity?


RE: WTF?
By Misty Dingos on 9/26/2008 12:00:45 PM , Rating: 2
Hey man why you got to hate on the cat? That aint right.


improved aerodynamics for cars
By bohhad on 9/26/2008 1:57:42 PM , Rating: 4
does this mean our cars are going to get even uglier?




RE: improved aerodynamics for cars
By TheFace on 9/26/2008 4:02:27 PM , Rating: 2
No, just yours. ZING!


By mindless1 on 9/28/2008 6:49:39 PM , Rating: 2
No it means you'll be attacked by tree huggers if you don't retrofit it with the Batman Tail Fin kit.


Geesus someone is queer for fluid dynamics
By phxfreddy on 9/26/2008 11:08:08 AM , Rating: 2
Even if proven out I'm not sure it will be the best thing in a decade. It is afterall just how the goo goes. How the fluid flows.




RE: Geesus someone is queer for fluid dynamics
By SiliconJon on 9/26/2008 11:30:33 AM , Rating: 1
I'm queer for fluid dynamics...in a contrasting kind of way. I mean, wouldn't it be nice to get some good looking aerodynamic cars on the road! Those things are so queer looking...but at 50-60mpg I have to look queer if I want efficiency, and the end of that is my contrasting queer for fluid dynamics stance.

PS: I'm totally gay for science, in a 30's kind of way.


By dagamer34 on 9/26/2008 12:15:53 PM , Rating: 1
Yep, back in 0030 AD, the Roman guys were definitely all OVER each other. :p


Interview
By LTG on 9/26/2008 2:33:14 PM , Rating: 2
"The research received initial funding from MIT's Ferry Fund. It is now funded by the Air Force Office of Scientific Research"

AliG would ask him why he would accept money from a Ferry Fund.




RE: Interview
By ElBrujo on 9/27/2008 2:07:13 AM , Rating: 2
High five!


Fuel efficiency
By cokbun on 9/26/2008 11:11:07 PM , Rating: 2
Hey jason, if it's fuel efficiency you need, then you might wanna check this one out. It's about a device that saves fuel that's being developed. http://www.eurekalert.org/pub_releases/2008-09/tu-...
oh and also could you email me, i have another interesting article about fuel efficiency in a foreign language.




RE: Fuel efficiency
By Magnus Dredd on 9/29/2008 11:43:25 AM , Rating: 2
When something seems too good to be true...

http://www.sec.gov/litigation/litreleases/lr19469....

quote:
The Court found that, beginning in February 1999, Muller carried out a fraudulent worldwide promotional campaign to disseminate false and materially misleading information about STWA's product, the "Zero Emission Fuel Saver Device" ("ZEFS Device"), a device Muller claimed would reduce toxic exhaust emissions and improve fuel economy in motor vehicles.


One word: SCAM!

http://www.sec.gov/litigation/litreleases/lr18487....

quote:
Blackwelder was a freelance consultant who, in or about July 2000, became a marketing consultant for Save the World Air, Inc. ("STWA"), a company that purported to have successfully developed and marketed a pollution control device for automobiles called the "zero emission fuel saver device." The Complaint alleges that Blackwelder prepared and arranged to have issued at least one false press release announcing a major licensing deal for STWA that, in fact, did not exist.


Might be good for oil
By Hlafordlaes on 9/26/2008 1:28:55 PM , Rating: 2
Oil producers (eg PEMEX in Mexico) often use natural gas to pump oil thru pipelines. IIRC from my last visit to their HQ in 2001, they had spent several millions in various attempts to model the gas behaviour, which affects both amount of gas used/wasted and of course the crude flow. Wonder if these equations will help; hope so.




Does this mean....
By foolsgambit11 on 9/26/2008 2:34:30 PM , Rating: 2
...we're ready to build the SeaQuest DSV? If so, I'll get to work learning to talk to dolphins.




another "amazing" breakthrough?
By shin0bi272 on 9/26/2008 3:46:09 PM , Rating: 2
at least once a week we get these breakthrough posts and while some of them seem cool this one really doesnt have a lot of affect on our daily lives. You cant make a teardrop shaped golfball so that it cuts the wind better (and if you could it would be banned by the PGA)... You can model aerodynamics of planes and cars all you want but when you take the thing to a wind tunnel you can see it for yourself realtime. The most this will do is allow people to simulate the chaotic turbulence behind an object better. So it might cut down on development time but it wont really do anything in the long run.




Is this man really trustworthy?
By DOSGuy on 9/26/2008 6:09:04 PM , Rating: 2
quote:
Professor Peacock says the experimental verification is equally important, though, stating, "While we fully trust George's new mathematical results, the engineering community is usually skeptical until they also see experimental results."


Didn't Professor Peacock kill Mr. X in the ballroom with the candlestick? Sorry, I'm easily distracted.




Hurricane damage?
By UNCjigga on 9/30/2008 2:01:14 PM , Rating: 2
Can this be used to more accurately predict what kind of damage a hurricane's storm surge can cause, based on given measurements of hurricane strength, size, direction etc.? Not really sure if fluid dynamics are important, or if this relies purely on meteorological science.




By kontorotsui on 9/26/2008 12:31:14 PM , Rating: 1
From cars to golf?
Think how important improved fluid dynamics are for sex, reproduction and of course porn.




Navier-Stokes
By pool1892 on 9/26/2008 2:52:26 PM , Rating: 1
i am no expert in fluid dynamics, neither in theory nor in computation. all i know is navier-stokes (from partial differentialequations 3 or something back in the learning days), which should describe fluids perfectly (if general solutions were known). is this new equation just a better approximation of navier-stokes in a special case of 3dimensions and certain fluids or is it something completely different? i have to say: i am to lazy to read the papers, it is friday evening;-) and you guys seem to know this topic quite well.

on a different note: i think it is a little overstated to call this the breakthrough of the decade, perelman with poincare, lhc or craig venter, who did sequence the genom in 2000 might all argue with that.




"People Don't Respect Confidentiality in This Industry" -- Sony Computer Entertainment of America President and CEO Jack Tretton














botimage
Copyright 2014 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki