NASCAR had not one but two drivers on their roofs yesterday at Talladega. Ryan Newman’s wreck was by far the most spectacular, but I had to wince at Mark Martin looking a little shook up and disgusted during the post-infield-care-center interview. I don’t blame him – or Newman – for being angry. Yes, it’s really impressive to see three-wide around the turns ten rows back, but is that really great racing? Is it good enough racing to subject the drivers (and the fans) to the types of accidents that we keep seeing at Talladega and Daytona? Is it good racing when you can cruise around the back all afternoon and finish in the top ten with drivers that fought for the lead throughout the race?
We learned (or were reminded of) a couple important things at Talladega.
- Most people do not understand the fundamental difference between the physics of drafting and the physics of bump drafting. There is a big difference between prohibiting bump drafting in the corners and prohibiting drafting in the corners
- It appears to be much easier for the new car to become airborne than the old car
- When cars are traveling in a pack, any accident is highly likely to involve more than one car
Drafting vs. Bump Drafting
Air rushing around the back of the race car creates a high pressure region at the front of the car and a low-pressure wake at the rear of the car. The high pressure in the front of the car (the car ‘punching a hole in the air’) creates a force opposite the direction the car is moving. The pressure behind the car is lower, which also acts in the direction opposite the car’s motion. So the car is fighting against a force pushing it backward in the front and a force pulling it backward in the back. If there are two cars running some distance from each other, each is experiencing two forces slowing it down: one in the front and one in the back (as indicated by the large red arrows in the top figure). Between the two cars, there are four big red arrows.
When a second car gets very close behind the first car, the air rushes over the two cars as if they were one, which removes the force at the rear of the first car and at the front of the second car. In the lower picture, there are only two big red arrows, so there is less total force working against the cars and – voila – they go 3- 5 mph faster. This is the important part: the two cars don’t need to touch to make this happen. This is plain ole drafting. You can get some more information on drafting and bump drafting in the Science of Speed segment “Drag and Drafting“.
Bump drafting is totally different physics. The leading car is running at full throttle. The trailing car is being pulled along, which means that at full throttle, it can actually go faster than the leading car. The trailing car bumps into the leading car, transferring some momentum from the trailing car to the leading car. The leading car goes faster and pulls the trailing car along with it. ‘Bump’ is probably a misnomer. Brian Vickers says in the Drag and Drafting video that he’s come away from plate tracks with headaches because he got bumped so hard – but adds that he was happy to be hit that hard because that’s the way you go fast at a plate track.
In drafting, you’re essentially removing a force by driving within inches of each other. In bump drafting, you are applying a concentrated force from one car to another. Bump drafting takes significantly more skill. The black dot in the diagram below indicates the center of gravity of the leading car. Newton’s laws: If you apply a force, the car goes in the direction of the force. The two cars bump squarely in the top diagram. The force from the trailing car pushes the first car straight ahead.
The middle picture shows two cars in a turn. There’s no way to bump squarely because one car is rotated relative to the other. If you hit the car ahead of you, you create a torque, which is a force that makes things turn. Think of the leading car as a spinner, pinned by the dot in its center. If you hit it in a direction so that the force goes directly through the center, it won’t spin. If you hit it at an angle, the car will spin. This is why you don’t bump draft in the corners. It is very easy to hit someone on the side of the bumper, sending them into a spin and wiping out half the field.
You can cause a car to spin by bump drafting in a straightaway if the two cars are not fully aligned. The bottom diagram shows that hitting a car off-center – even when both cars are going straight – is pretty much equivalent to hitting a car in the corner.
The two techniques have one thing in common: the person in the trailing car is in control and the leading driver is really just along for the ride. The trailing driver decides when to push, where to push and how hard to push. A number of drivers’ have expressed discomfort with ‘being pushed’ (drafting) too hard in the corners because an overaggressive – or an inexperienced – driver can make your car unstable fairly easily. But there is a pretty significant difference between bumping and pushing. Requiring drivers to leave space going into the turns just caused an accordion effect with cars having to back up down the line.
One of the big reasons for the aerodynamic changes in the new car was to decrease the amount of wake behind the car. The large wake with the old car made it very difficult for one car to get up close behind another one because the swirling air from the wake didn’t provide enough downforce on the front wheels and the car got tight. The big difference between the old car and the new car in the rear is the spoiler on the old car vs. the inverted wing on the new. All of the air must go up and over a spoiler, creating a huge amount of turbulence behind the car. The wing allows air to flow on top of and below the wing, so there is less turbulence behind the car.
Let’s briefly review the aerodynamics of wings. As shown in the diagram below, air moving over the top of the wing moves faster than the air moving over the bottom. Faster-moving air exerts less pressure than slower-moving air, so a wing experiences more pressure and more force below than it does above. That’s what gives an airplane lift. The wing on the rear of the car is upside down, so there’s more force on the top than the bottom and the inverted wing on a NASCAR car provides downforce.
The key for the computational fluid dynamics simulations of the old vs. new cars (shown below, from a GM publication) is that red are areas of high pressure, meaning lots of downforce. Orange, yellow, green and blue show decreasing levels of pressure, so those colors mean less downforce. Note in the old car (on the left), there was a significant amount of downforce generated by the rear decklid, while in the new car (right), the vast majority of the downforce comes from the wing.
NASCAR cars have an aerodynamic instability problem when they get going too fast. They are fine as long as they are pointed forward, but when the car spins, the aerodynamics change significantly. The air moves very quickly along the rear window and the roof, and remember that fast-moving air doesn’t generate much pressure. The ‘shark fin’ on the side of the rear window and the roof flaps were designed to slow down the airflow because slower-moving air creates more pressure. The roof flaps and the sharksfin were designed (the story is in my book, The Physics of NASCAR) for the old car and I don’t know how much work was done to compare the effectiveness of the roof flaps on the new car versus on the old car.
A couple things I noted this morning watching the video of Newman’s crash in slow motion.
- Both roof flaps deployed at the same time. In theory, one is supposed to deploy first and then the other if it is needed. But things were happening very quickly.
- The roof flaps deployed when the car was at about 135-150 degrees from heading in the correct direction. (180 degrees would be facing backward.)
- The right rear wheel was already off the ground as the roof flaps were opening
- One the car got fully backward, it flipped over without any twisting. You would expect a car that was spinning and then got airborne to continue the rotational motion, but the 39 did a really clean black flip.
- Once the car made about a 60 degree angle with the track, the roof flaps went down again. I wouldn’t think that gravity would pull them down, so that suggests that the pressure had increased to a point where the roof flaps didn’t think they needed to be deployed.
A backward spoiler is still pretty much a spoiler. If you think about the inverted wing running backward, I wouldn’t be surprised if it were generating a lot of lift. Put yourself in the place of the people designing the car. Would you have thought to simulate the car going straight backward to see what happens? We may not even know enough about the aerodynamics of the cars to do an appropriate simulation. But between this incident and Carl Edwards’ takeoff, NASCAR needs put some serious resources into re-evaluating the aerodynamic behavior of the car at different speeds. Decreasing the restrictor plate holes, especially by fifteen thousandths of an inch (about four times the diameter of a human hair) is not going to affect safety much. The reduction took ten to fifteen mph off the cars, but it didn’t address the primary problem of plate racing, which is that the drivers are on the throttle wide open most of the time, which makes them run in a pack.
A number of drivers are suggesting that NASCAR needs to sit down with them and talk about how to make racing safer. It’s not the drivers NASCAR needs around the table. With all due respect to Ryan Newman’s engineering degree, it’s the top aerodynamics people at the race teams (some of whom actually have Ph.D.s), and people like Gary Nelson and Gary Eaker, who were largely responsible for designing the original roof flaps. NASCAR has many more times the number of people working marketing and licensing than they do on safety. The teams have some incredibly smart people in their aerodynamics departments who have spent the last three to four years trying to understand everything they can about the aerodynamics of this car. NASCAR needs to make use of those folks’ skill and talent because they simply don’t have the in-house resources necessary to do the job quickly and effectively.
NASCAR has a history of being a reactive organization – as Carl Edwards noted last spring, when he said “I guess we’ll do this until someone gets killed, and then we’ll change it”. For a few scary moments Sunday, I was afraid we’d reached that situation. Take the initiative and solve this problem before someone gets killed. Don’t tells us that your rules weren’t the cause of the problems. Slowing the cars down is not enough: It is time for a major change, whether that be repaving Talladega to decrease the banking (as the late David Poole suggested), a major re-design of the aerodynamic safety equipment on the car, and/or introducing a significantly less powerful engine that could be run without plates.