Monday, April 26, 2010

Talladega Pas de Deux: why two is the magic number at 'Dega

The official Talladega mascot should be:

Pas de Deux (pronounced as "pah-day-due")

Merriam Webster dictionary defines "pas de deux" as

1. a dance or figure for two performers

2. an intricate relationship or activity involving two parties or things

A funny thing happened in Alabama on Sunday. Suddenly for the first time in memory two cars in a pas de deux outran an entire string of cars at "the big track." Until now a long string of cars-- 5, 10, 15 or more cars-- always had the advantage,…a huge advantage because a string of cars always had less overall drag than one or two alone. But on Sunday suddenly we saw two cars outrunning the freight train; how is that possible when it never worked before?

As many TV commentators remarked repeatedly, "the closing rate of the two cars over the rest of the field is incredible," as much as 15 mph or 22 ft per second, a car length per second. When you're only 3 feet away from another car, they're going 180 mph and you're doing 195, closing on them at 22 ft per second makes the slower car appear as if it is sitting still.

Several factors explain the two car run away: Talladega was repaved in 2006 and the surface is unusually smooth, allowing cars to get very close to each, nose to tail, and stay that way for long periods of time. The geometry of the CoT has the nose and tail of the car at essentially the same height so the trailing car can literally push up against the lead car and shove it. Previously with the car dubbed the Twisted Sister if the trailing car pushed this hard on the lead car they'd lift the rear wheels of the lead car off the ground and they'd both wreck.

Also the CoT with a spoiler on it has a very different wake than with the wing (red is highest speed air flow down to blue which is the lowest speed air flow).

There's a very distinctly larger area of low speed wake behind the race car with the spoiler compared to the race car with the wing. The result is that the two cars can actually touch and stay in contact which reduces the overall drag of the two car duet, considerably, by perhaps as much as 30%.

Power consumption increases with the cube of speed (speed x speed x speed) so the observed difference of two cars together at 195 mph vs one car or a string of cars at 180 mph should have required (7.415/5.832) 27% more power to go that much faster, but actually what happens is that the two cars linked together in a pas de deux have 27% less drag as shown in these studies.

While the two cars have a huge speed advantage there is very little air flow reaching the front of the trailing car, which means that the engine of the trailing car will quickly overheat because cool fresh air is not getting into the radiator. This limits how long the two cars can continue their pas de deux, and from watching the race one would deduce that a couple of minutes is about as long as two cars can stay in contact (NASCAR Cup cars can run 2.2 laps at Talladega in two minutes).

Harvick played the final lap pas de deux perfectly, and then at the last possible moment swung around his dance partner (McMurray) to win the race by a mere 0.011 seconds.

After the race Jamie McMurray was quoted, saying, "it's hard to explain to you guys that aren't in cars, but when there's someone directly behind you and they pull their car out of line really fast, it's like you pull a parachute in your car. It literally feels like you lose 3 or 5 mph immediately, and when that happens, the car that's doing the passing just has the momentum."

It doesn't just feel like the car slows down, it actually does slow down. The aerodynamic drag on a car is the difference in pressure on the front of the car, minus the pressure on the back of the car, and that difference of pressure applied over the frontal area of the car is the drag force (pressure multiplied by area equals force), perhaps 700 lbs or so on a car going 190 mph.

When the cars are nose to tail the low pressure area behind the lead car is filled in by the high pressure area on the front of the second car.

But when the trailing car gets out of line to go around the lead car, suddenly there's a low pressure area behind the lead car and the drag force on the front car suddenly increases by as much as 400 lbs. This has the same effect as throwing an anchor out, it slows the lead car by as much as 5 mph almost instantly.

These CFD studies in SolidWorks by FastTrack Racing Challenges for participating junior high and high school teams are possible due to the generous support of SolidWorks (visit their web site and look at the full software package at SolidWorks dot com) and in particular thanks to Marie Planchard, their director of worldwide education markets.


  1. I'd hardly call the two car breakaways new to Sunday's race. We saw the same thing at Talladega last year. Think about when Carl when airborne, Brad had pushed him away from the field with another two car breakaway just a second or so back which had Newman and someone else (Junior?).

  2. This is awesome. Thanks for posting. It's been too long since your last post.

    Glad to see you're back.

  3. Jordon, good point, however, the extent and the speed effect of the two car run away this Sunday is like nothing we've seen before, hence the record number of lead position changes.

  4. 36 Races, thank you, glad you enjoy my musing.

  5. Maybe I'm not getting it, but in the pictures of the COT with wing and spoiler, the wake looks to be larger behind the wing version vs. the spoiler. The explanation next to the pictures states that is should be the other way around. Am I missing something?

  6. Anon, no, you're getting precisely the right image; the spoiler creates a larger wake which is why the pas de deux has such a dramatic drag reduction effect relative to the other cars. The high pressure area on the nose of the trailing car fills in the low pressure void behind the lead car making a bigger relative difference with the spoiler CoT than with the wing CoT.
    Good eye, good observation.

  7. What a fun article to read!

    I think its fascinating to watch the "Pas de Deux" two-car breakaway at the restrictor plate tracks. It is the most exciting part of watching a race at Talladega. I would have rather watched groups of two cars try an over take the lead for the last 10-20 laps than any of the Green-White-Checkered attempts.

  8. I thinnk we will see this Pas de Deux at the 2011 Daytona 500 since for then we will have the new pavement! That would be really exciting to see!!

  9. I'll completely agree with Jordan on this and say that this is not new and was the reason why NASCAR insitituted the no bumping in the corners last fall. The two car break away has nothing to do with the spoiler or the wing as was proven this weekend. It has to do with the fact as Larry MacReynolds pointed out, you can't have three cars pushing each other because three cars is more difficult to stay in line and if any of them get out of line, there's a wreck. Go back and watch the broadcast and listen to the real explination of why it works, not your uneducated opinion. If you're saying something is new when it's a year old and that it doesn't work with the wing when it did, then you prove you have no clue what you're talking about.

  10. Mike,
    What was different this year was the closing rate of the two cars compared to other configurations. With Track Pass from NASCAR you can watch lap times, two cars in a pas de deux had as much as a 3 second (per lap) advantage over the "freight train." There's never been this disparity in lap times in the 15 years I've watched NASCAR.
    The relative speed of the two cars vs other arrangements is entirely due to aerodynamics.
    As to the educational background of my opinions, I've been an engineer and scientist for more than 40 years, including working as the senior design engineer for Gary Eaker at the AeroDYN Wind Tunnel in Mooresville. This blog on drafting has been given great praise by several race car engineers (three in NASCAR Cup teams, two in Indy Car, among others), along with Dr Diandra Leslie-Pelecky who wrote "the Physics of NASCAR." My undergraduate degree is in Mechanical Engineering (Carnegie-Tech, Pittsburgh, PA), with graduate studies in Germany and Tokyo. For more than two decades I was chief scientist for a large corporation, and specialized in aerodynamic processes that often also included combustion and chemical reactions all at the same time. We currently run CFD studies and 1/10th scale wind tunnel models for many race teams.
    I raced Legend cars for a few years and often parked next to or near Larry when his son, Brandon, was running Bandolero cars, and we chatted often. Larry is a nice fellow with a high school diploma from Birmingham, Alabama.
    May I suggest that before you impugn someone's opinion you might want to check your spelling and grammar.

  11. Good answer. Larry speaks from experience and doesn't try to give the physics explanation for much of what he knows in a visceral sense.

    I enjoyed the Pas de Deux, Prof.
    Looking forward to more.

  12. Nova,
    thank you,
    A good engineer/scientist often finds extremely useful insight in the observations of people with a real intuitive feel for a subject; that then combined with mathematics can gain a real understanding of a something as complex as vehicle dynamics.

  13. Wow, that was pretty much the most interesting article I've ever read on or about NASCAR. Thank you so much for taking the time to explain that in the great amount of detail that you did, it really got me thinking about all the things that go on with the cars during the race.

    That's just awesome Prof, you've successfully made me think today. Maybe I'll actually pay attention in my college classes from now on...!

  14. Dear Anon,
    Thank you, I don't if I should laugh or cry to imagine that you've not been paying attention and thinking,.... please do both, it's a lot of fun

    prof pi

  15. I will honestly make an attempt to pay attention in my class haha. You've gotta admit though, some of my classes aren't nearly as stimulating as say, NASCAR. :-)

    Thanks again, I look forward to reading more of the awesome articles that help explain the physics of NASCAR!

  16. Mr. Butler,
    I don't know what you're studying but as I always told my students are Carnegie-Tech, math gives you X-ray vision. Only it isn't a joke, math really does allow you to "see" what is otherwise invisible. There are a couple of books I recommend very highly: "Chaos" by James Gleick; "A new kind of science" by Stephen Wolfram, and any decent text on mixed partial differential equations,... and you'll have a good start on developing X-ray vision.

  17. I'm still confused just like the other anon reader. It makes sense that a spoiler would produce more wake area than the wing, but like the other reader stated, the picture clearly shows the wing producing more wake than the spoiler. Are the pictures wrong?

  18. Anon 2,
    Thank you, it's nice to have sharp readers; perhaps the problem is my description. In Part 1 the illustration we're discussing is the VELOCITY distribution in the X direction ONLY, the direction the car is pointed. This was put in the paper to illustrate that some air flows between the truck lid and the wing, thus reducing the VOLUME of low pressure area behind the car. I think the confusion stems from the fact that the area of disturbed flow extends farther behind the wing car than the spoiler car. If you look in Part 2 where I have added the PRESSURE distribution around the car which illustrates that there is a lower pressure behind the spoiler car hence the relative effect of drafting is greater with the spoiler car. This graphic shows the pressure distribution in two dimensions, and the third is pressure at it is projected out of the plane right at the viewer and is color coded to be easier to read (think of a topographical map, only here elevation is pressure). Compared to the rest of the field of cars there's more to be gained in terms of reducing drag by being right up against a car in front of you with a spoiler car than with a wing car.
    Does this help?
    Sincerely, prof pi