## Physical Models In an Age of Computers 78

Posted
by
timothy

from the bah-all-models-should-be-life-sized dept.

from the bah-all-models-should-be-life-sized dept.

Harperdog points out this article

*"about the Bay Model in Sausalito, California, which was built in 1959 to study a (terrible) plan to dam up San Francisco Bay. The model was at the forefront of research and testing on water issues that affected all of California; its research contributions have been rendered obsolete by computer testing, but there are many who think it could contribute still. Now used for education and tourism, the model is over 1 1/2 acres and replicates a 24-hour tidal cycle in just 14 minutes. Good stuff."*
## Been there (Score:3, Interesting)

My wife and I stumbled across the Bay Model while on our honeymoon.

We stopped in and walked around for a while.

Very cool place!!

## Re:Been there (Score:5, Interesting)

What's interesting is that real life models sometimes still yields a better result than the computer simulated, but both have advantages and disadvantages.

Especially streaming water is a tricky thing to model entirely right and you can get surprises when you have a complex situation and try it in a physical model compared to a computer simulation.

## Re:Been there (Score:5, Interesting)

Any model that is inherently chaotic (read: almost all of them) cannot be simulated on a computer accurately. There will be a cumulative error which will grow extremely rapidly (this is known as the "Butterfly Effect" after a well-meaning but ultimately damaging title of a research paper on the subject). Fluid dynamics is a major headache when it comes to chaotic systems, since the equations are violently unstable at most points most of the time. The only reason anyone can get any work done at all with CFD is that most research groups simplify the equations to one very narrow, very specific range of conditions. Even then, CFD is treated as a first approximation. A physical model of some sort is almost always built, somewhere along the line.

Indeed, one of the unsolved Millenium Challenges is to find out just how bad the Navier-Stokes equations really are. I'd have thought a more interesting challenge would be to find an alternative way to model chaotic systems. (Chaos means that the system is deterministic but not predictable. Which is no different from saying that you have a superposition of probablistic outcomes where the actual outcome is only knowable by observing the system. A field that is of interest to all kinds of people, including physicists and sci-fi writers. A standardized way of dealing with such systems that can yield any additional information at all - doesn't matter what or how little - would have a lot of appeal.)

## Re:Been there (Score:5, Interesting)

Any model that is inherently chaotic (read: almost all of them) cannot be simulated on a computer accurately...Indeed, one of the unsolved Millenium Challenges is to find out just how bad the Navier-Stokes equations really are.

Is that really a chaotic system problem, or an incomplete model problem? Taking the Navier-Stokes equations as an example, they actually are surprisingly accurate at describing turbulence, even though, as you've mention, we haven't proven that they describe turbulent flow properly in all cases. So typically problems with fluid flow models come from making too many simplifications, having incorrect initial conditions, or making bad assumptions, not a problem with the Navier-Stokes equations per se.

In fact, as far as chaotic systems go, wouldn't a physical model have the same problems as well? The challenge there is a very high sensitivity to certain variables, so your physical model of the system may turn out to not scale at all.

## Re:Been there (Score:5, Informative)

The Navier-Stokes equations are definitely chaotic, but turbulence is itself chaotic. It's actually a wonderful example of it.

A physical model would have the same problems, yes, since you can't scale atoms and the sensitivity to initial conditions means that given a long enough run (which is going to depend on the exact nature of the system) the cumulative error will swamp the system. The advantage of physical systems (for now) is that both the step size and the particle size under consideration are considerably smaller and modeled to a far greater level of precision. Physical models are still imperfect and can lead to all kinds of false assumptions if relied upon too heavily, but as the step before the full-scale system, it's the best we currently have.

A 1:1000 model essentially treats a block of 1000 molecules on the full scale as being the same as 1 molecule on the small scale. This means that at 0'C and 1 atmosphere, such a model would consider 2.687 x 10^22 (remember, you're considering 1000 at a time) [wikipedia.org] molecules of an ideal gas per cubic metre of gas under consideration. In comparison, the world's fastest supercomputer can perform around 1 x 10^16 FLOPS and the cartesian form of Navier-Stokes [wikipedia.org] looks to me like you're going to need to perform around 24-25 floating point operations per iteration per molecule. So you're looking at around 600 million seconds per cubic meter of simulated gas flow to get CFD equal to a physical model, if no simplifying assumptions can be made, and that's only true if thee world's fastest supercomputer's FLOPS rating is with a level of floating-point precision great enough not to introduce rounding errors within those 600 million seconds.

## Re: (Score:3)

Acceptable simulation performance is the part of the whole thing I hadn't considered, I understand your points now.

Thanks for taking the time to draft that example, made the whole thing a lot clearer for me.

## Re:Been there (Score:4, Informative)

Neither. It's a computational capacity problem. The complexity and non-linearity of the equations means they cannot be feasibly solved in continuum for anything but the most simple of cases. That means you're left with an iterative, discretized version of them. In order to solve a flow directly using discrete Navier-Stokes, your computational grid must be sufficiently fine to resolve the smallest scales of turbulent mixing. Even with several decades of exponential growth in performance, we are still nowhere near what is needed to solve any real world solution with Direct Navier-Stokes (DNS).

That leaves us with generalization. Turbulence models designed to reproduce the results of physical simulations are applied, rather than direct solutions, easing the computational load for those small scales. More complex models can better model more complex flows, but in turn are more computationally intensive, giving us a sliding scale between solution quality, and solution speed.

Computer simulation was never intended to replace physical simulation, merely supplement it. Physical simulations will always be limited in the environments they can produce, the duration of testing, and the data that can be measured. Computer simulation works on inference. If the simulation matches the physical test's data points at certain test conditions, the rest of the flow field not physically measured and be considered reasonably accurate. The same simulation can be performed at different test conditions, and as long as they are bracketed by the testing, can also be considered reasonably accurate. The simulation can be performed at test conditions outside the parameters of the physical testing, extrapolating behavior with reduced accuracy.

## Re: (Score:2)

"So typically problems with fluid flow models come from making too many simplifications, having incorrect initial conditions, or making bad assumptions"

Sounds like what is plaguing economic thinking as well...

## Re: (Score:2)

Any model that is inherently chaotic (read: almost all of them) cannot be simulated on a computer accurately. There will be a cumulative error which will grow extremely rapidly (this is known as the "Butterfly Effect" after a well-meaning but ultimately damaging title of a research paper on the subject). Fluid dynamics is a major headache when it comes to chaotic systems, since the equations are violently unstable at most points most of the time. The only reason anyone can get any work done at all with CFD is that most research groups simplify the equations to one very narrow, very specific range of conditions. Even then, CFD is treated as a first approximation. A physical model of some sort is almost always built, somewhere along the line.

So I'd like to take a few exceptions to your statements. First to say "cannot be simulated on a computer accurately" is both entirely true and entirely deceptive... For an example I take weather, which, while it cannot be accurately simulated in the general case, is, though a better understanding of measurements of initial conditions and better computational models, more accurately able to be simulated for short term results . . . weather predictions out to even the 5 day forecast are getting remarkably a

## Re: (Score:2)

There's nothing wrong with the Navier-Stokes equations, except that they're unsolvable. From a theoretic standpoint however, they're perfectly accurate. Unless you want to claim that mass, momentum, and/or energy are not conserved--because that's all the Navier-Stokes Equations are. Each one of the Navier-Stokes equations is a particular representation of the corresponding conservation law. Except for relativistic systems, or systems where quantum effects become relevant, the equations are perfectly acc

## Re: (Score:2)

First, Quantum Mechanics is chaotic. Second, probability waves are a product of Information Theory.

## Re: (Score:2)

This.

Anyone who uses computer simulations to model physical things knows that computers just plain can't do everything for you. ie. In the case of a physical part, you still need to do hands-on testing and at least verify your simulation results.

I want to say I'm a bit surprised that physical models aren't used more, but I'm really not. I just pray it's because people are trying to trim costs and are rolling the dice that the simulations are good enough and not because they actually believe the simulation

## Re: (Score:2)

Note that if they are always generally chaotic there would be no such thing as climate predictions for one.

## Re: (Score:2)

Absolutely, they are still used. I'm certainly not saying they aren't. I wish they were used MORE. I'm just pointing out the fact that way too many manager types think computer simulations are good enough. Whether it's because of TV or some kind of strange trust in salesmen, I don't know, but it's becoming a bigger problem in some industries.

## Re: (Score:1)

at least fit that shit to the song. i was singing along then fell on my face because you got no rhythm. ...or is that part of the troll?

## Re: (Score:2)

at least fit that shit to the song. i was singing along then fell on my face because you got no rhythm. ...or is that part of the troll?

Of course it's part of the troll, it manages to annoy both people who love Queen and people who loathe Queen.

## Don't miss the Mississippi (Score:5, Interesting)

## Re:Don't miss the Mississippi (Score:4, Interesting)

Or check this out: Mississippi River Bason Model in decay [blogspot.com]

## Re:Don't miss the Mississippi (Score:5, Informative)

More (much more) info [designobserver.com]

## Re:Don't miss the Mississippi (Score:4, Interesting)

I worked for Brown and Butler a consulting firm in Baton Rouge that did beta testing of the first PC port of the Corps HEC-1 mainframe hydraulic simulation software. John Haestad of Haestad Methods developed the software which is currently available from Bentley Systems at bentley.com. In the course of my duties I visited the COE Mississippi River model in the mid 1980's.

We were using the HEC software on a Leading Edge 8088 PC, 640K ram and 30mb hard drive, researching the effects of a proposed reservoir on the Amite River near Darlington, Mississipi to prevent annual flooding downstream in increasingly populated areas. The COE eventually decided against the B and B proposed reservoir and instead build a diversion canal to the Mississippi River, a project which is still in progress. The project decided upon still has not been completed 40 years after the initial studies began and is still facing opposition to its completion.

Here are the coordinates of one of the control structures:

3037'14.26"N 9114'24.11"W

## Re: (Score:2)

## Railroad scale (Score:2)

What model railroad scale is the closest? I have no interest in CA, so I don't know if 1.5 acres makes that bigger than G scale or smaller than Z scale or something in between. The live steamers might want to turn it into a live steam park, if allowed. Around here, the live steam parks are not quite as elaborate as this sounds.

## You have to see this (Score:5, Interesting)

The model cover about a 3000 sq mile area. The Golden Gate Bridge is about 6 inches long IIRC.

You have to see the Bay Model if you visit the Bay Area. There is a fully restored Nike missile silo nearby that is also a must-see of Bay Area nerdy sights.

## Re: (Score:1)

Why a shoe company thought it was a good idea to buy the naming rights to an AA missle project is beyond me.

## Re: (Score:3)

I realize you're likely joking, but I bet Nike (the shoe company) is named after Nike, the goddess of victory. ( http://en.wikipedia.org/wiki/Nike_(mythology) [wikipedia.org] )

## Re: (Score:2)

Nike Missile project : circa 1945

Nike Athletics : First named Blue Ribbon Sports, renamed Nike in 1978

According to Wikipedia, BOTH were named after the goddess of victory....

## Re: (Score:2)

I hope they remembered to pay the estate of Emperor Norton royalties (imperialties?) for the bridge.

## Mythbusters (Score:3, Informative)

Isn't this the same model that was featured on the Mythbusters episode [wikipedia.org] about escaping from Alcatraz?

## Re: (Score:3)

The Golden Gate Bridge is about 6 inches long IIRC

Thats what I needed. I looked it up and the GGB is about nine thousand feet long, and a typical railcar is "about half a hundred feet" (there is no standard) so the GGB if it were a railcar bridge (maybe it has a rail deck, for all I know or care) then it would hold about 180 railcars, if I divided that in my head correctly (10K/50 = 200 minus 1K/50 = 20 means about 180, right?). I don't think there is any model railroad gauge where 180 railcars fits in 6 inches. Although that would look pretty cool if i

## Re: (Score:3)

## Re:Railroad scale (Score:5, Informative)

What model railroad scale is the closest? I have no interest in CA, so I don't know if 1.5 acres makes that bigger than G scale or smaller than Z scale or something in between. The live steamers might want to turn it into a live steam park, if allowed. Around here, the live steam parks are not quite as elaborate as this sounds.

You don't have to have an interest in CA to read the first few paragraphs of the article:

its 1.5 acres replicate a 1,600-square-mile area that runs from the Pacific Ocean to the Sacramento Delta

1600 mi^2 is 1024000 acres, so it's a 1.5:1024000 (or 1:682666) scale if you believe the article.

However, the bay model's webpage tells a different story:

http://www.spn.usace.army.mil/bmvc/bmjourney/the_model/facts.html [army.mil]

Model Scales (Model to the Bay)

Horizontal: 1 foot = 1000 feet

Vertical: 1 foot = 100 feet

Velocity: 1 foot/ second = 10 feet/second

So using their numbers, it's a 1:1000 scale.

I have no interest in model trains, but Wikipedia tells me that Z-scale is the smallest commercially available scale, and is 1:220, so this is a much smaller scale than any model train system.

## Re: (Score:3, Informative)

The 1:682666 and 1:1000 aren't really very far apart...you based the 1:682666 on area, but scale is normally based on linear dimension. The square root of 682666 is about 825, so the two aren't really that different. Since the model isn't square, and actually twists part of it to make it smaller, the two are pretty close...

## Re: (Score:2)

The 1:682666 and 1:1000 aren't really very far apart...you based the 1:682666 on area, but scale is normally based on linear dimension. The square root of 682666 is about 825, so the two aren't really that different. Since the model isn't square, and actually twists part of it to make it smaller, the two are pretty close...

Ahh yes, good point! I forgot to convert from area to linear scale!

## Re: (Score:2)

The 1:682666 and 1:1000 aren't really very far apart...you based the 1:682666 on area, but scale is normally based on linear dimension. The square root of 682666 is about 825, so the two aren't really that different. Since the model isn't square, and actually twists part of it to make it smaller, the two are pretty close...

Ahh yes, good point! I forgot to convert from area to linear scale!

Now you know how those NASA engineers felt...

## Re: (Score:3)

Linear and quadratic scales aren't the same thing - if you double a length scale, you quadruple the area. Thus 1:1000 in length -> 1:1000000 in area, which is closer to the number you got from the article, modulo rounding (1.5 acre could be anything from 1.25-1.75 if they just rounded to nearest half integer).

## Education (Score:5, Insightful)

## Re: (Score:2)

You can make a computer model fit a physical model.

Your physical model will never fit your computer model. At least, not if your physical model is supposed to be natural.

## Re: (Score:2)

Children, be sure to take home these permission slips and get your parents to sign them. We'll spend an ENTIRE day the next FOUR WEEKS touring various fluid simulation sites.

vs

Children, Please download the Fluid Simulation Software to your workstations or laptops. Please load, and view each of the 20 simulation demos

before our next classeither in the Lab, or AT HOME. Each situation faces a water problem. RE-ARRANGE THE LANDSCAPES in the simulation and see if you can solve the problems.Since whe

## Re: (Score:1)

You mean it's not a horrendous conglomeration of Flash ads, moving type, slideouts/slideins and inconsistent use of font sizes?

I'm glad they went with a simple, easy to use to design that allows one to get to what they want without having to wade through piles of cruft.

Quite obviously they didn't let a web designer design their site.

## Re: (Score:2)

Quite obviously they didn't let a web designer design their site.

What's the difference between a web designer and a telephone sanitiser?

About thirty thousand quid a year.

## This is nothing new... (Score:2)

Wind tunnels and subscale models are still used in aerospace even with the availability of CFD and other computerized tools.

## Re: (Score:3)

I work in an old wind tunnel today! They use it as a server room now.

Unfortunately they don't use it to cool the servers, because that would be just too awesome.

## Re: (Score:2)

Skin drag of the air along the interior walls of the wind tunnel will generate heat and will probably destroy your servers.

I used to work in a closed circuit wind tunnel. It was about 30 ft tall, with the upper test section measuring about 6'x5'x4'. The top speed in the upper test section was about 200mph. After running tests for 4 hours at speeds no greater than 60mph, the temperature inside the test section was 105F. And no, it wasn't from the motor, that was actually outside the tunnel and a shaft pr

## Re: (Score:2)

## Re:This is nothing new... (Score:5, Interesting)

I imagine testing of scale (including full scale) models is still used in most industries. Computers help you reduce the number of mistakes and shorten the iterative development cycle so you make fewer elementary errors.

In Oil & Gas, almost all of the major manufacturers have been modeling all their new components in 3D for the last decade. You can have whole departments dedicated to running Finite Element Analysis on these 3D models. It's not as simple as putting in your constraints (loads, fixed and pinned reaction points, etc) and hitting "solve". You still have to make certain assumptions, tailor the matrix (the 3D lattice/approximation made of pins and beams) or make simplifications to the model if you want it to converge toward a solution. The computer solves the cases iteratively. Once you think you're close, you perform a test at scale and verify your assumptions. Then your match it to your computer model and hopefully you're not too far off. For example, things like incorrect friction factors between materials matter a lot when they are amplified by huge forces. The manufacturer of a particular coating may claim it's .07 (if they even tell you). You learn it's more like 0.08 through repeated testing. At this point there is no way of calculating things like that.

Of course, this assumes that you have the time, money, and server time to run such experiments. Most of the time you focus on the one or two critical areas of a design for FEA and use past experience or classic formulas on the rest. Customers still want final tests before product ships, so technically you have one last point to catch a failure/issue before it ships.

## Re: (Score:1)

That's exactly how it is in the aerospace field, too. Thank you for explaining it in better terms than I could.

I will say it does tick me off when (usually in response to flaws discovered in testing) people say things like "We have been using aluminum for a few decades, and we have computers now. There should be no excuses for any design that isn't perfect."

## both with 3D printing? (Score:2)

## Re:both with 3D printing? (Score:4, Funny)

As soon as we get a 3D printer that can properly print bedrock, dirt, forests, and

San Francisco.I thought that was what the magenta ink tank was for.

## Insomniac (Score:2)

## Question for experts (Score:2, Insightful)

Something I've always wondered about physical models is, how can you compare them to real situations at different scales? So many relationships in physics are non-linear. It seems like the model behavior must differ both quantitatively and qualitatively from the real behavior.

## Re:Question for experts (Score:5, Interesting)

Something I've always wondered about physical models is, how can you compare them to real situations at different scales?IIRC the folks who design America's Cup sailboats have (had) a rule of thumb that a hull model less than 1/3 scale would not accurately predict the drag of the full-size design. Off-topic: if you think water and fluid dynamics are tough to figure out, take a look at what little we know about how ice skating works -- or to be exact, just how the surface of ice behaves to allow gliding motion.

## Re: (Score:3)

## Re:Question for experts (Score:4, Interesting)

I learned some things in school but sometimes I wonder if it was worth stuffing my head with all the wrong things I learned.

## Re: (Score:2)

Funny...all these years I thought it had something to do with the drastically reduced coefficient of friction of ice (especially at just-below-freezing temps where there is a thin layer of liquid water present).

## Re: (Score:2)

I also learned that airplanes work because air travels a longer distance going over the top of the wing than the bottomI learned that too. I guess it was sincerely believed for a long time.

that movies work because of persistence of visionYeah. The brain-centric nature of motion perception was discovered in the early 20th Century, but I was still being taught that the

eyeis where image persistence became perceived motion. Kind of disappointing.and that the primary colors were Red, Yellow and Blue.They

## Re: (Score:2)

Somethings are different from different points of view (the primary colors are different in art and science.)

Some things are explained in only one way, while they can be done in different ways: a plane usually flies by Bernoulli's p

## Re: (Score:3, Insightful)

It depends on what effects are interesting to you. For example, in testing scale models of aircraft, reducing the model size generally means increasing the wind speed, to keep the Reynold's number (viscosity and kinematic effects) constant. However, that can jack up the Mach number (shock wave effects). Or, if you're doing dynamics, then the weight will be very important, which doesn't scale linearly with size. So it's usually a tradeoff.

## Mythbusters (Score:5, Interesting)

The accurate reconstruction of the tidal effects allowed them to convincingly show that a raft would be unable to reach Angel Island, and that a more plausible route would have been toward the Marin Headlands - before confirming the model's result experimentally.

## Dimensionless quantities (Score:2)

## Re: (Score:2)

They don't necessarily have to be of different scale, wind tunnels are often used to study 1:1 scale models.

## Re: (Score:2)

## Utility of physical modeling (Score:4, Informative)

The days of large physical models of tidal hydraulics in large estuarine systems are past because properly calibrated/validated numerical models provide good results at a fraction of the cost.

However, it is paramount the the numerical models are capable of simulating the correct physical processes without over-simplication. For example, the flow hydrodynamics near the Port of Anchorage in the Knik Arm of Cook Inlet are dominated by large gyres that are shed off prominent headlands. A large physical model of the Knik Arm constructed by the Corps of Engineers at their Vicksburg, MS, research facility reproduced the large gyres with a good match to measured field data, and local tug pilots agreed that the flows resembled what they experience daily. Initial attempts at numerical modeling the flow fields produced no gyres, and it was not until a very sophisticated adaptive turbulence closure scheme was added that gyres formed in the numerical model. Both the physical and numerical models required good boundary and initial conditions for success.

Physical models are still useful for simulating processes that are beyond our ability to describe mathematically (required for numerical modeling). Examples in the field of hydraulic engineering include some sediment transport processes, stability of rubble-mound structures such as jetties and breakwaters, erosion of cohesive sediments, wave forces on structures, and resiliency of levee grasses subjected to wave overtopping, just to name a few.

Numerical modeling in hydraulic engineering is making rapid advances, and often any unknown processes can be adequately represented in the model by empirical formulations that have been developed based on physical model tests. Whatever the skill of the numerical model, it is imperative that engineers who apply the numerical model to a problem have a good understanding of what physics are being simulated and what compromises have been made during model development. Failure to understand what the model does will assuredly lead to disaster.

Finally, physical models can be successful provided: (1) The dominant forcing in the real world is correctly represented in the scaled model, (2) any forcing not correctly represented has minor influence, (3) laboratory and scale effects can be minimized or some compensation can be applied, and (4) model results have been validated to the extent possible. A similar set of criteria applies to numerical models.

In the future, physical modeling will continue to be used to validate numerical models, they will provide physical understanding and empirical formulas for use in numerical models, and physical models will continue to address those engineering problems that cannot be formulated mathematically.

## Re: (Score:2)

1. If all you want is "good results", then computer models are fine.

2. it's still expensive to build a model of anything natural, and still can't be perfectly accurate, because you can always make it an even more detailed copy. shorelines are fractal, and so are all the textures and texture mixes. at a certain point, you might as well just run experiments on the actual thing.

## Physical models are returning (Score:3)

There is currently a huge research effort underway funded by tons of millions to make physical models of more complicated stuff: quantum simulators. With classical computers, there's basically no way to calculate properties of things like quantum magnets, quark-gluon plasmas, or high-temperature superconductors. But by using a different quantum system that we can precisely control (for example, ultracold atoms [quantenblog.net]), we can tune it to simulate the behavior of the material we are interested in.