80 replies to this topic

[Greg Locock]

A concept regularly pooh-poohed by the experts. The Mythbusters instead went and did some good work.

If you think you can spot a hole in their methodology, make sure you watch the whole thing, it was a good experiment and produced a 10% improvement in fuel consumption at 65 mph.

So, I expect to see all LMPs to be dimpled from now on.

http://www.autoblog....e-dimpling-mpg/

The water tank test showed that the separation bubble behind the car was smaller if the car was dimpled (not that that is necessarily the be all and end all).

It does strike me that perhaps ripples would be more effective than dimples, as the flow over a car is in a defined direction.

[Canuck]

This topic...Deja Vu... Other thread...

[Greg Locock]

bah. I hadn't seen the episode then. I still think it was a great experiment.

[Canuck]

Still haven't caught it unfortunately. Funny how it takes a couple of entertaining tv personalities to actually try it, to prove the professionals wrong in the process. This sort of thing is why people continue to attempt the create/modify/improve concepts the "professionals" swear will never work.

So...has anyone actually tried a pebbled surface akin to a dolphin's skin before it was dismissed, or is that another thing that we simply know doesn't work either ;)

[Greg Locock]

Sharkskin finishes for yachts are available, but haven't set the world on fire, probably because of fouling.

What they don't show you on the show is the failed experiments they tried along the way, and specifically I wonder how they came up with that size of dimple. I also wonder if the edges of the dimple need to be sharp.

This'd be dead easy to proto from a high density foam sheet, using PVA to stick it to the car. I bet 3M could make it on a roll just for laughs. I'm going to be having a little chat with the aero people at work, I'd be an idiot not to.

[Catalina Park]

Maybe try a vinyl roof first to see if it reduces lift.

[cheapracer]

So...has anyone actually tried a pebbled surface akin to a dolphin's skin before it was dismissed, or is that another thing that we simply know doesn't work either ;)

My twin beam car is under way, getting drawings together for it slowly - probably won't work just like many other things people tell me won't work

I read an article once about a Drag Rail who's team coated the car with Johnsons Baby Powder before each run swearing it was worth 1/10th of a second, a lot at their level.

[McGuire]

I don't like the methodology. I am less interested in the dye patterns in the water tank than in the drag numbers. Since they didn't share any numbers, we may assume they never got any, or they were inconclusive. Neither would be surprising as the color visualization was somewhat inconclusive itself.

I didn't care for the fuel measurement methods either. Both systems were pretty hinky: an un-graduated cylinder (?) and a parallel fuel tank with switch. Five one-mile runs, five measurement errors.

Since it doesn't seem likely that a surface treatment could produce an 11 percent decrease in aero drag at 65 mph (that's a lot) it seems even more doubtful that the surface treatment could produce an eleven percent decrease in fuel consumption. That suggests a reduction in drag even greater than 11 percent.

There is an easy way to test the proposition: measure road hp at 65 mph with and without the surface treatment. I forecast no measurable difference, or perhaps a slight reduction for the knobbly version.

It is not the reduction in wake separation that makes a dimpled golf ball travel further. The dimples impart backspin, which imparts lift via Magnus Effect and also gives the ball stability. An undimpled golf ball will fly like a wounded duck -- or a football (American) thrown without a spiral. A clean golf ball will drive further than a dirty one -- that's what the ball washers are for.

[Greg Locock]

One of the fun things about working for the Falcon car company is that we are very good at cribbing improvements in highway fuel economy, whereas improvements in urban fuel economy cost big gobs of hard cash.

This is another one of the cheap highway fixes. I suspect that treating the front half of the car is a waste of time, the trick is to identify which bit was important, and how to make it look OK. I'm guessing the B pillar area , across the roof. But I don't need to guess, that's why we have propeller-heads.

Edited by Greg Locock, 14 February 2010 - 11:10.

[Wuzak]

It is not the reduction in wake separation that makes a dimpled golf ball travel further. The dimples impart backspin, which imparts lift via Magnus Effect and also gives the ball stability. An undimpled golf ball will fly like a wounded duck -- or a football (American) thrown without a spiral. A clean golf ball will drive further than a dirty one -- that's what the ball washers are for.

I always thought that the dimples worked because of backspin.

I guess, also, that the dimple effect is not as good for more streamlined shapes - like aircraft wings.

[Tony Matthews]

I always thought that the dimples worked because of backspin.

I guess, also, that the dimple effect is not as good for more streamlined shapes - like aircraft wings.

Not dimples, I know, but I have often wondered what effect the early construction methods used on aircraft like the Junkers J52, and possibly Ford Trimotor, certainly many others, with chord-wise corrugations, had on efficiency. Better, much worse or neutral? And, of course, Citroen 2CV van!

[Rob]

The box of clay which they put in the back altered the weight distribution and angle of attack of the car.

[mariner]

I am glad somebody as well as me enjoys Mythbusters because they have a sense of fun and also are prepared to look afresh at things. In the UK it follows American Hotrod on one of those "cheapy" channels ( Quest) which come and go from the station line up over time. I haven't sen the dimple episode yet but it sounds worth a look.

On a similar vein here is a link to some research somebody did on the lumps on a whale fluke which it is thought prevent stall during manouvers.

http://dsc.discovery...bine-whale.html

Amazing what nature can do, now what I want to know is why I cannot find any animals or birds with a Kamm tail? If a Kamm tail saves weight and does not increase drag why do birds not have them? Is it because at the speed of most birds the Reynolds number is different to a road or race car?

[SteveCanyon]

Not dimples, I know, but I have often wondered what effect the early construction methods used on aircraft like the Junkers J52, and possibly Ford Trimotor, certainly many others, with chord-wise corrugations, had on efficiency. Better, much worse or neutral? And, of course, Citroen 2CV van!

I've wondered that as well - I suspect it may help a little with the spanwise flow.

[Tony Matthews]

Amazing what nature can do, now what I want to know is why I cannot find any animals or birds with a Kamm tail? If a Kamm tail saves weight and does not increase drag why do birds not have them? Is it because at the speed of most birds the Reynolds number is different to a road or race car?

I think probably because the tail is so important in maneuvering, if you watch a kestrel hovering the tail is costantly on the move , sort of twisting longitudinally. Anyway, what about the Lynx and the Manx cat?

[Tony Matthews]

I've wondered that as well - I suspect it may help a little with the spanwise flow.

I like the look of it, whenever I see it my model-making fingers start to twitch! There is obviously more surface area for a given span, but these were slow flyers - it is intriguing to know what was happening with the airflow. I imagine it was done purely for reasons of strength. You still see a lot of Nissen huts about...

[JtP1]

I like the look of it, whenever I see it my model-making fingers start to twitch! There is obviously more surface area for a given span, but these were slow flyers - it is intriguing to know what was happening with the airflow. I imagine it was done purely for reasons of strength. You still see a lot of Nissen huts about...

The bringing it to the model bit, leads straight into would dimples work or not. Now the people who compete with model aircraft will try anything as it is not the end of their world if it does not work and is not finacially out of reach. In some areas technology is led from the remnants of the Soviet aerospace industry, they can sell their talents for $s to the rest of the world. Now these fellows will try anything for 0.1 secs/ Km and yet none of them sells or uses dimpled planes.

On a car, from the reasons it works on a golf ball, it would possibly make a difference on something like the tail of a C type Jag.

[Tony Matthews]

In some areas technology is led from the remnants of the Soviet aerospace industry, they can sell their talents for $s to the rest of the world. Now these fellows will try anything for 0.1 secs/ Km and yet none of them sells or uses dimpled planes.

From some close-up shots I've seen of a Sukhoi su-27 it looked fairly dimpled to me - much like an early Ferrari body before the filler and paint...

Edited by Tony Matthews, 14 February 2010 - 16:56.

[McGuire]

On the Junkers and Ford aircraft, the corrugated skin had no aerodynamic purpose. The inner structure was designed to take the bending loads while the skin took the torsional loads, so the external panels were corrugated to increase their strength -- like a cardboard box. Detail aerodynamics for reduced drag were not really a concern on these aircraft. On the Ford for example, note the external control wires, unshrouded engines, and fixed, naked landing gear.

Apropos of nothing at all, the former Ford plant where the Tri-Motor was built is today part of Ford's Dearborn Proving Grounds, recently renamed the Dearborn Development Center. Some of the Tri-Motor complex is still standing and still in use. The experimental engine lab is in there, among other things. (Nowadays that means fuel cells etc.) The proving grounds is on the same acreage previously occupied by the former Ford Airport, said to have the first two concrete runways in the world. (And the world's tallest private dirigible mast.) Part of the proving ground roadway follows the path of the former runways, though they have long since been paved over. The well-known hotel just across Oakwood Boulevard from the Proving Grounds, the Dearborn Inn, was originally built to service the airport, which finally ceased operations just after WWII.

[Tony Matthews]

On the Junkers and Ford aircraft, the corrugated skin had no aerodynamic purpose. The inner structure was designed to take the bending loads while the skin took the torsional loads, so the external panels were corrugated to increase their strength -- like a cardboard box. Detail aerodynamics for reduced drag were not really a concern on these aircraft. On the Ford for example, note the external control wires, unshrouded engines, and fixed, naked landing gear.

I realized that it was corrugated for structural purposes, I just wondered if there was an unintentional aero effect, either negative or positive. Still like the look of it! The weird thing is, I would have thought it more complex to fabricate curved, ribbed skins than to make more substantial ribs and stringers to suport a flat, easily affixed skin and achieve the necessary torsional strength. I write, as usual, as someone who knows nothing about the internal structure of the JU52 wing, and it may be that early aluminium sheet was not of sufficiently high quality to use without strengthening corrugations. Everything I ponder on seems to generate more imponderables.

[NeilR]

The bringing it to the model bit, leads straight into would dimples work or not. Now the people who compete with model aircraft will try anything as it is not the end of their world if it does not work and is not finacially out of reach. In some areas technology is led from the remnants of the Soviet aerospace industry, they can sell their talents for $s to the rest of the world. Now these fellows will try anything for 0.1 secs/ Km and yet none of them sells or uses dimpled planes.

On a car, from the reasons it works on a golf ball, it would possibly make a difference on something like the tail of a C type Jag.

Unfortunately the scale effect (viscosity) may upset this reasoning.

[JtP1]

Unfortunately the scale effect (viscosity) may upset this reasoning.

Since most of the model aircraft I am talking about are right in the size and speed range for the dimples to work, if they in fact did.

[NeilR]

They would need to be a damn big model. Even high tech model applications such as the F3 classes and quarter midget will have scale effect issues and I'm not sure I can see the cross over in their Re to a car application. How Greg is describing it in modifying boundary layer attachment also has no parallel to a model plane as far as I can see. I'm happy to be corrected if you can explain it further.

[GreenMachine]

... I would have thought it more complex to fabricate curved, ribbed skins than to make more substantial ribs and stringers to suport a flat, easily affixed skin and achieve the necessary torsional strength...

In Australia, corrurgated metal rainwater tanks once were ubiquitous in rural areas, and even in urban areas - overtaken by concrete for larger tanks and now by plastic for the smaller ones. Metal rollers did the job quickly and cheaply. I don't know the precision required, but these mostly required both horizontal and vertical joints... In an aircraft application, changes in chord and/or airfoil section would complicate things, and perhaps the ability of the supplier to provide the metal to tolerances close to specification. I assume the (aircraft) manufacturer rolled both the corrugations and the curvature, separate operations IIRC.

Would this have been an early application of 'stressed skin'?

[Dmitriy_Guller]

The box of clay which they put in the back altered the weight distribution and angle of attack of the car.

That sounds implausible on two counts. It's unlikely that the center of mass of the car is significantly away from the back seat, and it's unlikely that the box would significantly change the angle of attack.

Even ignoring both of those points, if the improvement came from the changed angle of attack, then it seems like a more miraculous improvement than the dimples. Putting large dimples on a car is hard. Changing the angle of attack is not.

[McGuire]

In Australia, corrurgated metal rainwater tanks once were ubiquitous in rural areas, and even in urban areas - overtaken by concrete for larger tanks and now by plastic for the smaller ones. Metal rollers did the job quickly and cheaply. I don't know the precision required, but these mostly required both horizontal and vertical joints... In an aircraft application, changes in chord and/or airfoil section would complicate things, and perhaps the ability of the supplier to provide the metal to tolerances close to specification. I assume the (aircraft) manufacturer rolled both the corrugations and the curvature, separate operations IIRC.


Would this have been an early application of 'stressed skin'?

That's exactly what it was. The material was Duralumin, the trade name for an aluminum alloy in the general category of 20xx today. Easily corrugated via roller forming, just as you say -- same general process as beading or flanging. The sheet is pulled through two rollers that transfer the desired profile. The material is strengthened by the ribbing and hardened by working. Corrugation was one of the first volume-production processes applied to metal sheet early in the 19th century.

The Ford Tri-Motor was designed by William B. Stout, who also designed the Stout Scarab automobiles. A sort of futurist-visionary in the mold of Buckminster Fuller or Norman Bel Geddes but not as well known today, Stout was also an accomplished practical engineer and an amateur poet who published several volumes. He originated the phrase, "simplicate and add more lightness." To finance the Stout Metal Aircraft Company, he sent letters to 1,000 industrialists, asking each one for $1,000 and promising they would never get their money back. Henry and Edsel Ford were among the initial investors and eventually bought the company.

[McGuire]

One of the fun things about working for the Falcon car company is that we are very good at cribbing improvements in highway fuel economy, whereas improvements in urban fuel economy cost big gobs of hard cash.

This is another one of the cheap highway fixes. I suspect that treating the front half of the car is a waste of time, the trick is to identify which bit was important, and how to make it look OK. I'm guessing the B pillar area , across the roof. But I don't need to guess, that's why we have propeller-heads.

Good luck with that but I don't believe that by itself, a dimpled or mottled surface can produce any measurable improvement in fuel consumption at highway speed.

[Terry Walker]

I also thought their method of measuring fuel consumption was a bit iffy, but the fact that the smooth clay covered car came in almost exactly the same as the plain metal car, and the dimpled clay came in quite substantially better, it may be that the result is broadly valid if not accurate.

The considerable extra weight of the smooth clay made no difference to the fuel consumption over the standard car, which was a useful control. It would have had the effect of increasing the frontal area, but also making the car sit lower.

Go down the street, buy two identical cars about five years old (so they're cheap and expendable), make a portable press which can dimple the sheet metal itself (or a dolly with a cup, plus a hammer), and do long-range tests side by side, AFTER a pre-dimpling long range control test.

Interesting, but the dimpled car would be a bummer to clean.

(The guys also tested the various gizmos which supposedly defeat the speed cameras: they all failed. And even very high speed failed.)

Edited by Terry Walker, 15 February 2010 - 07:05.

[cheapracer]

. Easily corrugated via roller forming, just as you say -- same general process as beading or flanging. The sheet is pulled through two rollers that transfer the desired profile. The material is strengthened by the ribbing and hardened by working. Corrugation was one of the first volume-production processes applied to metal sheet early in the 19th century.

Every town here has an industrial area and theres half dozen of these Guys on the street in those areas who will unwind you some sheet off a large roll and do it on the spot for you - corrugated and a few other shapes.

[SteveCanyon]

Go down the street, buy two identical cars about five years old (so they're cheap and expendable), make a portable press which can dimple the sheet metal itself (or a dolly with a cup, plus a hammer), and do long-range tests side by side, AFTER a pre-dimpling long range control test.

I watched the episode as well but I can't remember the reason (if any?) for the decision to pick the size & depth of the dimples. I know there's a lot of variables but if I had to decide I would have thought smaller ones would have worked better ..... say about 3mm - 4mm diameter and about 1mm depth.
It certainly bears further investigation though.

[cheapracer]

I watched the episode as well but I can't remember the reason (if any?) for the decision to pick the size & depth of the dimples. I know there's a lot of variables but if I had to decide I would have thought smaller ones would have worked better ..... say about 3mm - 4mm diameter and about 1mm depth.
It certainly bears further investigation though.

Biff Musters Bolf Gall episode

Biff Musters Bolf Gall Fart Two

Edited by cheapracer, 15 February 2010 - 13:45.

[Tony Matthews]

say about 3mm - 4mm diameter and about 1mm depth.

Yeah, the sort you could make with a small ball-peen hammer, a couple of hours to spare, and your neighbour's car...

[Canuck]

[McGuire]

Clearly, the dimples were scaled simply to make the car most resemble a golf ball. There is no more to it that I can see. We are giving these guys waay too much credit, IMO.

Some may take the view that since the gain in highway-speed fuel economy is 11 percent (which is enormous) there must be something to it somewhere, even if it's not the full 11 percent. Must be good for something obviously, etc. I take a different view. For me, the 11 percent gain demonstrates that in all likelihood the whole thing is hooey and there is no potential gain whatsoever.

I believe that if there is something to be had in exterior surface modeling it will be incremental and probably outside the reach of the casual test methodology employed here. Would this test regime be able to quantify the difference in aero drag/fuel economy between, say, a stock vehicle and the same vehicle with a great honking pair of travel-trailer mirrors hung on the doors? I doubt it. The rigor isn't there. I didn't see them paying any attention to wind direction or speed, for instance. Which are huge when you get down to the fine whittling in aero drag.

I am also bemused by the water tank test. We have to presume that the models were attached in the tank via a force scale of some kind. (Otherwise, why do it?) So where are those numbers? Maybe they just weren't that compelling in regard to the TV show's storyline.

[OfficeLinebacker]

...

I am also bemused by the water tank test. We have to presume that the models were attached in the tank via a force scale of some kind. (Otherwise, why do it?) So where are those numbers? Maybe they just weren't that compelling in regard to the TV show's storyline.

"Never let the truth get in the way of a good story."

[Terry Walker]

I thought the water tank test was purely to make visible, for the TV audience, the difference in turbulence behind the "car", not to measure it in any scientific way. And quicker, simpler and cheaper way of doing it than a wind tunnel.

[gruntguru]

Every town here has an industrial area and theres half dozen of these Guys on the street in those areas who will unwind you some sheet off a large roll and do it on the spot for you - corrugated and a few other shapes.

Will they dimple your car - "while you wait"?

[cheapracer]

Will they dimple your car - "while you wait"?

You must be talking about the panel beaters here.

The standards of panel beating here would be hilarious, especially the standards of paint matching (admittedly not an easy art), if I had not had my own car repaired previously.

The old "it will be the same colour in a few days when it dries"

[Dmitriy_Guller]

I watched the episode as well but I can't remember the reason (if any?) for the decision to pick the size & depth of the dimples. I know there's a lot of variables but if I had to decide I would have thought smaller ones would have worked better ..... say about 3mm - 4mm diameter and about 1mm depth.
It certainly bears further investigation though.

My guess is that two factors drove the dimple size; the fact that they worked with clay, and that they put the dimples in manually. I imagine you can't get as precise with wet clay as you can with metal, and there are only so many dimples you can put in the clay by hand before you think better of it.

Edited by Dmitriy_Guller, 16 February 2010 - 09:16.

[just me again]

I watched the episode as well but I can't remember the reason (if any?) for the decision to pick the size & depth of the dimples. I know there's a lot of variables but if I had to decide I would have thought smaller ones would have worked better ..... say about 3mm - 4mm diameter and about 1mm depth.
It certainly bears further investigation though.

I think they just enlarge the holes so they were on the same size on the car as on the golfball. car is larger therefore bigger holes.

Bjørn

[Fat Boy]

that's what the ball washers are for.

audio for pic

[imaginesix]

I believe that if there is something to be had in exterior surface modeling it will be incremental and probably outside the reach of the casual test methodology employed here. Would this test regime be able to quantify the difference in aero drag/fuel economy between, say, a stock vehicle and the same vehicle with a great honking pair of travel-trailer mirrors hung on the doors? I doubt it. The rigor isn't there. I didn't see them paying any attention to wind direction or speed, for instance. Which are huge when you get down to the fine whittling in aero drag.

Agreed, there was no control over the testing environment. Plus, there was no indication of the level of precision of their acquired data, only averaged totals. Strictly speaking , their results are worthless however I don't believe they intended to deceive. If their data was all over the map they would have changed their methodology until it was reliable. The fact that they report scientifically justifiable results with the dirty/clean/smooth clayed tests leads me to wonder if in fact the dimpled clay test result might not also be accurate. It deserves to be further researched if for no other reason than to explain the error in their methodology. And even if they dn't find any fuel economy gains, aero engineers stand to learn a lesson or two on airflow if they perform this out-of-the-ordinary investigation.

[McGuire]

The fact that they report scientifically justifiable results with the dirty/clean/smooth clayed tests leads me to wonder if in fact the dimpled clay test result might not also be accurate. It deserves to be further researched if for no other reason than to explain the error in their methodology.

Best of luck with that. Shop Tip: when working with clay, apply a bit of vaseline to your hands.

[Fat Boy]

It's science. They had the finest data collection, storage and distilling methods available. Apparently, employing the finest techniques and methods of the University of East Anglia just isn't enough for some people.

[gruntguru]

Haven't seen the episode but seem to remember seeing a still frame or clip that showed a car with huge dimples applied. Not an expert in the science of aerodynamics but I am sure that if modifying skin friction is the purpose of golf ball dimples, they wouldn't need to be scaled up for a larger ball (or a car).

[smokingwheels]

It was an 11% gain in reducing fuel consumption, I think they just rounded it off for the viewers.
They only did 2/3 of the surface of the car.
What about under side of the car add some dimples on some sheet metal and bolt it on, I'm only guessing but there could be another 3 % in that (29.72 MPG).
Also I would have liked to see them do the test over a longer distance to reduce errors, but I'm happy with the results.

I have a similar pattern on the intake of my car engine and I have had sum surprising results with reducing fuel consumption and also increasing it.

[jatwarks]

Any thoughts here on using perforated body panels? The perforations would be vanishingly small; on the molecular scale.

Goretex fabric, used in waterproof garments, helps the wearer to remain dry but ventilated by including a membrane perforated with holes that are large enough to allow air to pass out through them, across a pressure gradient, but too small to allow water to penetrate to the interior.

A rigid body panel with perforations of similar size, if it is possible to construct such a panel, would allow air from within the panel to be drawn out by the slipstream and form a layer between the panel surface and passing airflow.

This is similar to the use of bubbles to reduce the drag of boat hulls.

Apart from construction problems, what else would make this a non-starter when considering drag and efficient airflow over a streamlined car body?

[meb58]

I think that water and air work in similar and not so similar ways...that water is non-compressible is clear.

I cannot remember the actually mechanism at work here but fish scales apparently prevent a portion of water from actually touching their surface...

[cheapracer]

Wow, I typed in "Mythbusters dimpled car pictures" into Google and got a picture of "Mythbusters dimpled car"! Outstanding....

Edited by cheapracer, 18 February 2010 - 19:29.

[cheapracer]

Any thoughts here on using perforated body panels? The perforations would be vanishingly small; on the molecular scale.

Does rust count?

No wonder the Government has a campaign against rusty cars, it's all clear now - it's a conspiracy from the oil companies to get people's cars that get better MPG off the road!