I have a little problem with the methodology used for this myth especially with the first section with the gage hanging on the window. Having spent a good portion of my misspent youth, as an auto mechanic there is a problem with their setup. In a fuel injected car the system is a closed loop with a fuel pressure regulator that maintains a set pressure at the injectors in relationship to intake vacuum. All fuel over the amount needed to maintain that pressure is returned to the fuel tank to be reused. By creating their own little tank on the window to introduce the fuel without that fuel return they only measured the fuel that went through the system not what was actually used by the engine as some was returned to the fuel tank in the car. And while this is a very simple component of the system it can be affected by external sources. Which brings me to a second point that the modern fuel injection systems are an extremely complex system that responses to a variety of internal and external inputs. And to try to extrapolate fuel mileage from less than gallon used in a 5 mile run with all the other variables that can come into play is too small a sample to form a real conclusion.

phydeoux, you should probably discount what your high school physics teacher taught you about fluid dynamics. He either did not know what he was talking about or oversimplified to the point of confusing things.

Dimples on a golf ball "trip" the boundary layer from laminar to turbulent. Turbulent boundary layers have more drag than laminar ones, but do not separate as easily as a laminar layer when the shape curves away; separated flows have much more drag than either turbulent or laminar attached flows. On a large object such as a car (as Anthem obliquely referred to), the boundary layer trips so early that there is little point in rippling the surface.

That said, there are ways to energize the boundary layer. Many aircraft wings have vortelators forward of the flaps or ailerons to assist with keeping the flow attached. These are usually done less for drag reduction and more for stability and control reasons, since energizing the boundary layer, after a point, costs more energy than the payoff.

quote:

Originally posted by Charlie5824:
I have a little problem with the methodology used for this myth especially with the first section with the gage hanging on the window. Having spent a good portion of my misspent youth, as an auto mechanic there is a problem with their setup. In a fuel injected car the system is a closed loop with a fuel pressure regulator that maintains a set pressure at the injectors in relationship to intake vacuum. All fuel over the amount needed to maintain that pressure is returned to the fuel tank to be reused. By creating their own little tank on the window to introduce the fuel without that fuel return they only measured the fuel that went through the system not what was actually used by the engine as some was returned to the fuel tank in the car. And while this is a very simple component of the system it can be affected by external sources. Which brings me to a second point that the modern fuel injection systems are an extremely complex system that responses to a variety of internal and external inputs. And to try to extrapolate fuel mileage from less than gallon used in a 5 mile run with all the other variables that can come into play is too small a sample to form a real conclusion.

fuel in a tank isn't pressurized, it's pressurized by the fuel pump. Tank > filter > pump > cylinder. They were good on this point. you can even see the filter in one cut.

The entire episode was great! Loved every minute of it.

OK, When are the dimple body Prius conversion kits gonna be on the market? :-).

A question for those more familiar with laminar flow than I:

One reason a golf ball is completely dimpled is that it spins in flight--you don't know which direction it will face at any given instance.

Would it have made a difference if the dimples were only placed on the back half of the car? If the dimpling was only on the trailing edge, would this be even more efficient?

I agree with mdunlap24 that a normally dirty car should have been used. A car with splotches of dirt randomly created on the car would have caused the "dimpling" effect. It probably would have a significantly lesser effect than the clay created dimples, but I think that is the gist of the myth. Adding a level layer of dirt all around produces nothing but extra weight. No one is surprised though that Adam and Jamie would take a simple concept and RAMP IT UP :-)

Lokifan: I wondered the exact same thing as I watched the show and created this account just to ask that question

Anyone know?

This has to be one of the best episodes ever!

After watching this episode I did some searching on dimpled objects and their aerodynamics. Some suggest, cars will be more aerodynamic if the dimples were present only in the bottom of the car.

Mythbusters, please go one step further and test the bottom dimple myth as well.

quote:

Originally posted by bcfrombalto:
I agree with mdunlap24 that a normally dirty car should have been used. A car with splotches of dirt randomly created on the car would have caused the "dimpling" effect. It probably would have a significantly lesser effect than the clay created dimples, but I think that is the gist of the myth. Adding a level layer of dirt all around produces nothing but extra weight. No one is surprised though that Adam and Jamie would take a simple concept and RAMP IT UP :-)

I can tell you from personal experience that a clean car gets better mileage. I have used a sports car (92 Sunbird), a small sedan (97 Escort) a full size (87 Corsica), 2 different types ov minivans (97 Caravan/01 Caravan & 97 Winstar), over 15 years to make the same 600+ mile trip a few times a year between where I live and where my parents live. We heard about the "clean vs dirty" car theory 20 years ago and decided to try it.

I can also say that before trying it I was a constant mileage tracker, as I had to show mileage/consumption for work, so I got into the habit of doing it all the time and have continued to this day.

I can also confirm that I am a non car washer, at least for my personal car, as city driving mileage sucks anyway and washing the car does not change the mileage numbers enough.

I also drove mostly highway miles for many years and got the same results, to the annoyance of my boss as I drove a white company vehicle and was supposed to keep it always looking "like new".

What I got from both the show and my personal experience is: If you are doing city driving a clean car is aesthetic, but if you are going to do mostly high-way driving, then clean early and clean often.

quote:

Originally posted by green_phantom:
lexus did it first:

http://www.autoblog.com/galler...exus-ls460-l/full/#2

lol

Yes but where are the "Wind tunnel" and mileage test results from this car compared with the same non-dimpled car?

I would expect that this car was made this way as a publicity stunt only for the golf tournament and no real mpg/lpk testing was done.

Or is this also another reason that the foreign car companies have better cars and profitable car companies, save Ford somewhat? Does Detroit need to get out of the dark ages in car design?

I think this dirty dimple experiment was a well done set but like anything they do, time and medium really limits how deep they can travel down the bunny hole.

Truth is, though. We've known about dimpling on vehicles for a good while. The principle is usually applied to address wind noise issues on cars. Several modern vehicles use a raised bump or dimpling applique to flow out trouble spots. Those trouble spots are usually the result of most of the car being so flush while a certain section by necessity is not flush such as windows seams, A pillars, etc.

But the MB folks are missing out on a way to optimize dimpling further. Round dimples work great on a golf ball but the dimples are round because they are unidirectional. There is no "front" to a golf ball. But a car is directional and has a front. So it follows that a shaped dimple, something akin to a reverse NACA duct/round dimple hybrid distributed densely (many small dimples) thoughtfully placed would yield even better numbers.

Popular Science and Popular Mechanics ran this story few months ago.
Popular Mechanics Link

Few thoughts that I have:
1. This is a much "nicer looking" approach..
2. Can't imagine how many car washes required per week to keep the same efficiency (assuming dusts tend to be "trapped" inside the dimples).

Website for the "skin"
http://www.fastskinz.com/

quote:

Originally posted by cobaltHawk:
The entire episode was great! Loved every minute of it.

I 100% agree. It was the best episode in a long time, and the car was especially fascinating. My mind knew it had to work, yet I kept thinking that it was such a crazy concept that it mustn't. And of course, it was entertaining to watch Grant and Tory yet again make fools of themselves for science!

I think they should do it again with more dimples, the dimples are small on a golf ball because a golf ball is small. Shouldn't it work better with more dimples of a smaller size?

Great episode.

I was hoping they would break-out their cannon and try firing a cannon ball and a cannon ball with dimples, then compare the distances, just for fun. )

Ok, so next question would be- How small can the dimples be to give the effect? I ask because I had a '92 Chevy Lumina (3.4 l 6 cyl) that always got more mpg than feds said it should.
I had bought the car used because I got $1k off because it had got caught in a hailstorm while at the dealer's.
Now they weren't big dimples, in fact, you had to be close and looking at a low angle (10-20 degrees) to see them. But they were on all horizontal surfaces.

Would it be just the car or was I seeing the golf-ball effect?

I think there is a reason why golf balls have dimpling and other forms of moving objects don't. I believe the dimpling in a golf ball works as well as it does because the ball itself is spinning. (Like previously mentioned). (Remember the Pro Golfer.) Since the car isn't spinning, I don't think the effect will be as apparent (if at all) when applied to a car not undergoing any rotational movement.

As far as the MB experiment with the clay-dimpled car, a statistics expert would probably say that the data set of the MB's trials would be too small to come to a definitive conclusion about the mpg. As a general skeptic myself, I would like to see the experiment done on a larger scale, ie more cars, longer distances, more trials. (Not necessarily by the MB crew, I believe they did an incredible job with their resources)

IF by some strange reason there is a significant effect on the performance of a straight line motion kind of object, it would be neat to see all of our everyday things becoming dimpled. But then there's the question of marketability and production cost....

quote:

Originally posted by DIMrBobsir:
Ok, so next question would be- How small can the dimples be to give the effect? I ask because I had a '92 Chevy Lumina (3.4 l 6 cyl) that always got more mpg than feds said it should.
I had bought the car used because I got $1k off because it had got caught in a hailstorm while at the dealer's.
Now they weren't big dimples, in fact, you had to be close and looking at a low angle (10-20 degrees) to see them. But they were on all horizontal surfaces.

Would it be just the car or was I seeing the golf-ball effect?

I know this may astonish you but I am very familiar with the '92 Chevy Lumina, the 3.4 L V6 engine and the dimple effect and hail damage. I know... I specialize in the convergence of knowledge no one needs. The great MPG you experienced was actually characteristic of the Lumina and its powertrain. First of all, I can't go into why the EPA numbers are actually not accurate. But it is not the EPA's fault. They did right by the citizenry. The engine is also much more efficient (due to some relatively late changes to the engine management computer). The body design was also somewhat variable over the model year, so there are undocumented aerodynamic improvements.

GM is much aligned and much of the crap hurled their way is deserved but there is a little secret about their products... sometimes the vehicles they design to be only mediocre end up becoming exceptional. The 1992 Lumina actually exceeded expectations.

But I assure you that hail had no appreciable bearing to your MPG improvement.

quote:

Originally posted by webgeek:
Lokifan: I wondered the exact same thing as I watched the show and created this account just to ask that question

Anyone know?

I have a degree in Aerospace Engineering and agree with you both. The NASA Engineer specifically said the dimples are effective after the flow passes the apex of the ball. Before the apex of the car, they'll actually cause drag and would probably decrease fuel mileage. The golf ball must be dimpled all around because it spins, but we know which direction the car is going so just dimple it after the apex. In fact, maybe even try a half dirty car, with the dirt after the apex. Just an idea.

In geek speak, after the apex, there's an "adverse pressure gradient" which tends to separate a laminar boundary layer and increase drag. The dimples trip the laminar flow, and causes turbulence which naturally helps the boundary layer stay attached longer and reduce drag.

Another interesting idea has been brought up regarding the size of the dimples. I don't think it would work well to use dimples the same size as the golf ball because the ball and the car are travelling at different speeds. A good guess for the dimple size could be to incorporate the Reynolds number as Anthem suggested. Since both are traveling through the air, density and viscosity can be ignored. So (Speed x Diameter) of the golf ball dimples should equal (Speed x Diameter) of the car dimples. A golf ball can go about 150 mph which is about triple the speed of a car on the highway, so the diameter of the dimple on the car should be about three times the diameter of the golf ball for an equitable effect. Much smaller than the ones in the MB test. Of course, wind tunnel testing would be pretty simple and give the most accurate result.

And finally, it's been proven that hexagonal dimples work better than round dimples. Just another factor to consider.

What a fun episode though. Thanks to the MB for all the hard work.