Mythbusters "tests" efficiency of dimples vs. smooth, and clean vs. muddy car

[pgfpro]

Originally Posted by Old Tele man

...no ABA tests.

...looked more like five each "flying-mile" runs through a 1-mile distance at 65 mph, with speed run-ups prior to the measured mile.

...in fact, it wasn't clear whether they reversed the run directions to cancel wind / terrain incline, etc. or not, because it seemed to appear they were making "backyard" loops over paved closed roadway (flying through road crossings is why).

WOW!!!!

That surprises me. They're usually much better then that!!!

It looks like its on tonight on Discovery at 9:00pm so I going to have to watch it.

[RobertSmalls]

Originally Posted by MadisonMPG

I think we should write in.

Yes, they have a forum for that sort of discussion.

[majestic]

they really screwed the test by picking a tauras with a rear spoiler.

the clay models spoiler is shaped (regardless of surface) a million times better than stock.

[Christ]

Originally Posted by bomber991

They clayed the whole car smooth, which added some weight, but then they did 5 runs and averaged them. Then they made the dimples and even put the little caps they carved out into the car so that the weight was the same as before, and then did 5 more runs and averaged them. The way they measured the fuel is probably more accurate than using a scangauge.

"26 and change" mpg for the smooth car
29.65mpg for the dimpled car

Now where's the 11% improvement? 29.65/26=1.14 That would mean it was 26.7mpg when smooth wouldn't it?

The problem is that it wasn't really accurate at all. At 65 MPH, Adam had like .1 seconds to react and flip that switch, both times. Human error is accountable here, as well. Also, there are far too many potential variations to count a 1 mile stretch as valid.

There is no way their "test" was more accurate than using a tuned scan gauge. Unfortunately, it's what I've come to expect from them.

[achang1]

if anyone wants to watch it its at
zSHARE - mythbusters.s07e14.hdtv.xvid-lmao.flv



edit: anyone wanna fabricate some dimpled body parts and test it out?

[Piwoslaw]

Originally Posted by Christ

The problem is that it wasn't really accurate at all. At 65 MPH, Adam had like .1 seconds to react and flip that switch, both times. Human error is accountable here, as well.

All the more reason for using a longer test track.
I wonder how long it took them to spray that clay on, etc. What I'm getting at is that in normal testing, the time interval between runs should be as short as possible, and the whole test should take the least amount of time. If the 'A' run was in the morning, then a coffee break, then they put the clay on, fooled around, waited for it to dry, lunch break, finally 'B' in the late afternoon... Well, that's not the kind of testing I would do. If there was another 'A' run later, then it'd be easier to see time-of-day effects. But getting that clay off would be at least another 1-2 hours. A-B-A-B would be ideal, but not possible in this set up.

As for dimples, they should have saved that for another episode and tested different sizes and spacing.

[JasonG]

Its amazing how some poo-poo anything new, regardless of the outcome.
yes I agree they needed to do longer runs to be conclusive, but they did due diligence in the time they had.
Note the clay was on the car for both runs, no weight changes.

This forum is about doing something, taping something on and testing it. Lets stop armchair-quarterbacking "FAIL" when we don't like the call.

[tasdrouille]

If you want to know if it's working, stick it in a wind tunnel with a scale. Else there's just too many variable to conclude anything.

[MadisonMPG]

Winfield1990, you mad brah?

Test has too many things wrong with it to be accurate, end of discussion.

[Christ]

Well, I watched the second run of it the same night that I posted. Other than by admission, I'm not sure how you could tell someone hadn't watched it anyway, so I'm going to dismiss your post as conjecture.