Using Drag and Downforce to Tailor Stock Cars to Different Tracks

Aerodynamics is complicated. Let’s just get that out of the way. But it’s not so complicated that we can’t understand what’s going on with just a little patience.

Why 3D?

Every wonder why they call it three dimensions? The reason it’s three is because I (or you) can denote any point in space with only three numbers. For example: a latitude, a longitude and an altitude. Since we’re dealing with much more limited spaces, a simple Cartesian Coordinate system, like this one, usually suffices.


The line that goes out to the point P is a three-dimensional vector. It’s got parts going in the x, y and z directions. By specifying that there is so much in the x-direction, so much in the y-direction and so much in the z-direction, I’ve told you everything you need to reconstruct that vector.

Forces in 3D

A force (a push or a pull) can act in any direction, but in order to understand the effect of that force, it’s easier if we break it down into its components – how much of the force acts in the x-direction, how much acts in the y-direction, etc.

When we do this on a race car, we give the different directions their own fancy names – just to make us seem extra clever. Basically, any force that acts in the opposite direction the car is going is called drag. Any force that pushes the car into the track is called downforce.

When the force acts up instead of down, it’s called lift. Yes, I know it should be called ‘upforce’, but the people who study aeroplanes named it.

Not shown in the picture below is side force, which would be in or out of the page.


Spoiler Alert!

Let’s think about the air acting on the spoiler. Because the spoiler is at an angle, the force on the spoiler is at the same angle (it’s perpendicular to the surface). So some of the force on the spoiler points down and some of the force points horizontally.

Which means some of the air hitting the spoiler creates downforce, and some of the air hitting the spoiler creates drag.


The more area there is for the air molecules to hit, the larger the force. A tall spoiler creates  more force than a short spoiler – but because of what I said up above, the angle of the spoiler is absolutely critical.

The more upright the spoiler, the more of the force is drag and the less of the force is downforce.  If the spoiler were horizontal, you’d get all downforce. If the spoiler were perfectly upright (vertical), all the force would be drag.

Why a Different Package for Indy vs. Kentucky?

In Kentucky, NASCAR went with a shorter spoiler to reduce the downforce. Passing has been a persistent problem at 1.5 mile tracks and the idea was that if the cars weren’t quite so dependent on aerodynamic forces, then the loss of those forces when you get close to another car wouldn’t have such a great impact.

And that strategy seems to have paid off well.

But Indianapolis and Michigan are very different kinds of tracks. At 2.5 miles and 2 miles respectively, they  are closer to superspeedways than they are to 1.5 mile tracks. At Indy and Michigan, the cars get going very fast down the straightaways, which lets the leading car get away from its pursuers. And it’s pretty tough to pass a car if you’re two lengths behind it going into the corner.

So the goal at these almost-superspeedway tracks is to slow down the maximum speeds along the straightaways so that a car can’t get away so easily. This is a little different than the goal at the intermediate tracks.

There’s a couple of ways to slow down a car: the two most obvious are

  • Decreasing horsepower
  • Increasing drag.

Decreasing horsepower introduces its own challenges, as we know from restrictor plate racing, so NASCAR is using the increased drag approach at Indy. And they’re doing that by setting the spoiler height at a pretty astounding 9 inches tall. At Kentucky, the spoiler had been reduced to 3.25 inches.

The best way to understand how much of a difference this is comes from a tweet from JGR Racing, which actually shows you the difference. Extra points for having gotten the product placement in there!


That’s a pretty big honking spoiler, eh?

But, you’re thinking (at least I hope you’re thinking) wait a moment… If they increase the spoiler height to increase the drag, aren’t they also increasing the downforce?

Yep. They are. It would be lovely to have a knob that you could turn and independently change the amount of front and rear downforce, and the amount of drag. But real life isn’t that simple.

Those Poor Engineers… NOT

The spoiler isn’t the only thing that’s changed. The changes in toto are…

  • 9″ spoiler
  • 1″ wickerbill (aka Gurney flap)
  • 2″ splitter
  • 43″ radiator pan width
  • speedway extension on the quarterpanels and rear bumper – the same ones run at the superspeedways.

So you’re thinking – my goodness, pity the engineers. All these changes.

Lemme tell you – the engineers are not upset. They love the opportunity to get ahead of the other teams by being smarter and figuring stuff out before someone else does. This is a chance for a team to get a win simply by understanding the set ups better than anyone else.

And something else to think about. In my column about Kentucky, I showed the changes in the spoiler and radiator pan sizes as a function of time. Well, I’ve updated those.

BSPEED_2015RulesChangesbyTrack_Spoiler BSPEED_2015RulesChangesbyTrack_RadiatorPan

The radiator pan is the exact same size at Indy as it was in 2014. The spoiler is only one inch taller than it was in 2014. And the teams have plenty of experience with the rear aerodynamic extensions from years of racing at Talladega and Daytona.

Yes, it does mean that they have to put those disparate elements together – which they haven’t done before – but the teams with the strongest technical staffs will be in the best position to take advantage of these just-in-time adjustments.

Personally, I’m psyched about track-specific packages. It gives the teams much more of a box to work in, which means they have that much more room to be creative. Looking forward to Indy!


The Kinetic Energy of Austin Dillon

Someone asked in the comments how much kinetic energy Austin Dillon had when he hit the catchfence at Daytona. I don’t know exactly how fast he was going  (probably was somewhere around 180 to 200 mph), so I figured I’d just make a graph and include a couple reference points in terms of kinetic energy scales.

The left axis is labeled in MegaJoules (MJ), which are millions of Joules. To give you an idea

  • 0.009 MJ = energy contained in a AA alkaline battery
  • 0.038 MJ = energy contained in one gram of fat
  • 0.04-0.05 MJ = energy contained in one gram of gasoline
  • 8.4 MJ =  daily recommended energy intake for a typical active woman  (2000 calories)
  • 110 MJ = energy expended by a typical rider in the Tour de France
  • 122 MJ = energy contained in a gallon of gasoline
  • 1000 MJ = energy of a typical lightening bolt

So without further comment, here’s the kinetic energy of Austin Dillon, where I used his NASCAR official weight of 185 lb and minimum car weight of 3300 lbs. Black cowboy hat not included.


Daytona, Catchfences and Flying Cars

As some of you know, I was in New York for the weekend celebrating my anniversary, so I’m just now catching up on the weekend’s accident at Daytona. Note: Some basic information about catchfences is modified (and updated) from a blog originally posted on 11/14/11.

And I bet there are some typos, still.

A Brief History of Barriers

Track barriers originally were erected to keep cars separated from spectators.  In addition to concrete walls to prevent the cars from driving off track, debris-spewing accidents necessitated fencing to contain airborne objects.

ChainlinkFenceCatchfences should have the same properties as walls, but they can’t block the view.  Chain link fence is a good compromise: It’s cheap, plentiful, easy to put up and surprisingly strong given its high visibility.

Chain-link fabric is an elastic metal mesh. It can give in two ways: gentle forces cause the mesh to deform.  The diamonds stretch out of shape, but when the force is removed, the fabric springs back to its original shape. The fence can also deform by stretching the wires that make up the mesh. A large-enough force will break the wire entirely.

The mesh must must supported, usually by poles and cables. How much the mesh can stretch depends on how it is supported.  If the frame is too big – meaning that there’s a very large area of mesh between supports — the mesh can stretch too much. How the poles are attached to the mesh is critical, because the attachments allow the load to be shared between the fabric and the poles.  The larger the forces, the more robust the links between the poles and the mesh must be.
Catchfence_Daytona2Race track fencing is stouter in just about every way.  The mesh is made of larger-diameter wire with higher tensile strength.  The links between the poles and the fabric are stronger:  In the picture at right (Daytona), steel cables run horizontally through the mesh and are fixed to the vertical poles using some massive turnbuckle-like fixtures.

Different tracks have different installations.  Some have metal tubing running horizontally as reinforcement. Catchfence improvements have primarily been via stronger mesh, stronger or a greater number of poles, or better links between the poles and the mesh.  But it’s basically the same fundamental design.

The chain-link fence is a motorsports institution, with different sanctioning bodies requiring different standards for debris fencing.  In the FIA test, a 760-kg  (1675 lb) test mass is shot into a fence at a speed of 65 km/h (40 mph) at heights of 1.6 m and 2.5 m (5.25 and 8.2 feet respectively).  While 40 mph seems very slow, they’re taking just about the entire mass of an Indy car and concentrating it in a relatively small sphere.  A real car would impact over a much larger area and spread out the force.


The photo at left shows a Geobrugg fence being tested:  The mesh deforms (a lot!) – but it does not break. Load is transferred to the poles, with the poles nearest the impact bending, but not breaking.  The emphasis, however, is pretty strictly on containment.

Geobrugg, one of the primary catchfence providers for motorsports (and many other things) made the video below that shows a car going into a standard vs. going into one of their fences.

The advances in catchfencing that have been made are huge; however, you are always going to have issues. As long as the fence is permeable, small pieces (and fluids) are going to get through the fence. Some of the people injured in the Carl Edwards crash in Talladega were burned by oil. Unless you use a solid fence, this is a hazard you will never eliminate.

Is Speed a Factor?

We’ve been hearing calls for slowing things down at Daytona. Is speed the issue? It is certainly true that speeds have been increasing.


Here’s the pole speeds at Daytona over its entire history. You have to be a little careful. The idea of group qualifying is very recent, and we know single car speeds are lower than those of cars in a pack, so it’s not quite fair to compare these one one one. But the pack qualifying pole speeds are a good 4mph below the peak single-car pole speed, which means we’re still probably 8-10 mph slower than the highest speeds that have been raced at Daytona.

So maybe it would be helpful to compare speeds at different tracks this year, since single-car speeds aren’t really all that different than race speeds at non-plate tracks.


I put triangles above the plate tracks. You’ll notice that speeds there aren’t that much higher than many of the other tracks – and Michigan was faster than either Talladega or Daytona.

Which led me to wonder about whether this year was just an anomaly. So here’s this year and last year.


Michigan is consistently faster than either Daytona or Talladega. I also wanted to look at the speeds for the Sprint Cup versus the XFINITY series, because the last big accident we had at Daytona was Kyle Larson running in the then-Nationwide series.


You can see that the lower-level series runs at significantly lower speeds (10-15 mph) – and still they’re getting in the air.

So I have a hard time believing that simple reducing speed is going to have a lot of effect on these types of crashes.

Pack Racing

So what’s the big different between Michigan – the fastest track – and Daytona/Talladega? It’s the restrictor plates. Restrictor plates produce a very different type of racing. If you watch the throttle/brake indicators during non-restrictor plate races, you’ll see the drivers easing off the throttle going into the turns, or even braking. The pole speed is an average speed, which means they’re generally traveling faster down the straightaways and slower in the turns.

Not at a plate track. The throttle is full open at all times. The cars are maxed out in terms of their engines.

Remember that, at 204 mph, a car goes a football field a second. Think about that. You’ve seen cars scatter to avoid accidents on other tracks. You can’t do that on a plate track because there is nowhere to go. You’re blocked in on all side.

An accident on a plate track is more likely to involve multiple cars. The last line here is the total number of cars involved in accidents

Michigan 2014 Daytona 2015 July Daytona 2015 Feb Las Vegas 2015 Atlanta 2015
6 1-car accidents
1 9-car accident
5 1-car accidents
1 7-car accident
1 9-car accident
1 11-car accident
2 1-car accidents
2 2-car accidents
2 7-car accidents
3 1-car accidents 2 1-car accidents
1 4-car accidents
1 6-car accidents
15 32* 20 3 12

* didn’t mention how many were involved in the end-of-race crash because it wasn’t technically a caution, so the number is much larger.

The numbers are small, so it’s hard to prove this, but my intuition, based on observations and the data we do have, is that more cars involved in an accident at close quarters mean

  • more likelihood of a car spinning (and cars are more likely to become airborne if they are not going straight)
  • more likelihood of a car launching off another car and getting airborne

When all this happens in conjunction with high speeds, you have all the elements for a catastrophic accident.

I am not suggesting we remove restrictor plates – that would be just plain stupid for both the drivers and the fans.

So Let’s Just Make Pack Racing Safer!

Everyone seems to assume that NASCAR will come up with a solution that will allow pack racing to continue the way we’re used to it happening. They’ve done an admirable job of dealing with past issues…


There are limits. I mean, if we could do anything, we’d have cured cancer and found a way to make sure everyone in the world has access to clean drinking water.


If there were an obvious solution to make pack racing safer, NASCAR would have already done it. It is possible that there isn’t a solution and that pack racing will always be inherently more dangerous than other types of racing.

Cars travel just as fast at other tracks and they don’t leave the ground nearly as often as they do at Talladega and Daytona.  Putting on a smaller restrictor plate to decrease speeds will not help. It’s not the speed. It’s the combination of the speed and the pack racing.

Perhaps the best that can be done is to protect spectators and let drivers take their chances. (If you’re wondering whether a Lexan ‘hockey-type wall’ would work, I addressed that elsewhere. (TL;DR: expensive and difficult, especially since you not only have to stop the car and parts from getting into the stands, you have to make sure that you don’t make it more dangerous for the drivers.)

Perhaps you have to make a radical change to the engine so that the drivers have to brake and accelerate around the track and you don’t get pack racing. This would make a lot of fans upset. There is nothing as breathtaking as standing in the infield watching the entire field take the turn.

There is also nothing as breathtaking as that gasp of fear, your heart skipping a beat and the feeling in the pit of your stomach as you whisper a prayer that the driver in the crunched up shell of a car just coming to a stop will climb out and wave and live to race another day.

Racing without Friction

Daytona is an enormous, sweeping track. Two-and-a-half miles, 31-degree banking and corner radii of a thousand feet. The infield by itself is 180 acres. If you’ve ever been there (or Talladega), it really does take your breath away when you first enter. Now, bigger tracks (or rather, tracks with bigger turns) automatically mean higher speeds.

There’s a formula for this that tells you how much force you need to make a car turn under specific conditions.


The way to think about this is that it is harder to turn (i.e. you need more force)

  • when you have a heavy car
  • when you’re going fast
  • when you’re trying to make a tight turn

So when you compare a thousand foot turn radius like at a superspeedway with the 250-foot turn radius of Bristol, it’s four times easier to turn at Daytona if all other things are held equal.

The equation above is the equation for centripetal force, which is the force that makes a car turn. The centripetal force tends to confuse people because of its direction. The centripetal force points toward the center if the car is moving in a circle. The way I think of it is if you swing a tennis ball around on a string in a horizontal circle over your head, the thing that keeps it going in a circle is the string – producing a force toward the center.

Well, it’s the exact same thing for a car, except instead of a string, you have tires. The force needed to turn a NASCAR stock car at 130 mph at Bristol is about six tons. Yep, tons.


Because Daytona is so much larger, you need about four times less force to turn at the same speed.

But why stop at 130 mph?

When Daytona was being planned in the fifties, Bill France knew he wanted high banks. Why?


That’s right, banking equals speed, too. Here’s why. Look at the car on a flat track first. I’ve drawn it so the car is moving away from you and it’s turning left (of course).



The force the track exerts on the car is always perpendicular to the track surface. So none of the force of the track on the car is in the direction that helps the car turn. All of the turning force has to come from the interaction between the tires and the track. If you don’t have enough friction, then you’re going to slide out toward the wall.

Banking helps us turn. Let’s give our track a little banking and see why.



Two things change. First, the friction between the tires and the track have also tilted. That means you’re not getting the full force from the tires that you did before; however, the force of the track on the car has also shifted direction.

Now the track is helping the car turn. The higher the banking, the more help the car gets from the track.

If you’ve never been on a track, it’s almost hard to appreciate banking. Here’s me filming for our Science of Speed webvideo series at Texas Motor Speedway. I had this great pair of boots I had planned to wear for this shoot, but it turns out you really can’t wear heels on 24 degrees of banking.

And no, the car is not moving. I am adventurous, but I am not (usually) stupid.



Turning on Ice

So one of the questions I sometimes get asks how important friction is in turning corners. So let’s play Einstein here…

Einstein thought up all kinds of very strange and mathematically intense ideas about how the universe – space and time, specifically, work. He couldn’t actually do experiments to test all of his ideas. (Plus, he was a theorist and it’s usually best not to trust them with anything more potentially dangerous than a sharp pencil.)

So he did what are called gedanken experiments. Gedanken is the German verb for ‘to think’. These are thought experiments – but they sound much more impressive if you call them gedanken experiments.

We’re going to imagine that a highly localized ice storm hits Daytona. So localized, in fact, that it just hits turns 3 and 4 of the Daytona International Speedway. It covers them with ice. What happens to the car hurtling in there?

There’s an equation – and if you’re the kind of person who breaks into a cold sweat at the sight of a radical (that’s a square root), then just grab your chair tight for a moment. (If you want to see the details, I suggest the wonderful Hyperphysics site.)



All this says is that it is possible to bank a track highly enough that you can take the turn without ANY FRICTION AT ALL.

So if we plug in the numbers for Daytona… we find that, in the absence of friction, you could go 139 mph around the turn.This shouldn’t be all that surprising – after all, Daytona could be viewed as an overgrown luge or bobsled track, right? Those tracks have very high banks because there’s a minimal amount of steering going on.

Being the mathematically OCD person I am, I graphed the maximum speed as a function of banking degree.


Remember that we not only have friction, we have lots of it from the tires interacting with the track – that’s why the cars go much faster than in our frictionless case here.

Interestingly, if a car doesn’t go fast enough around a banked turn, it will actually slide down the track.

This presents a major problem when you’re repaving a very banked track because, as a rule, heavy machinery doesn’t move very quickly. The video below shows the 2010 Daytona repaving (pictures, but mostly video).

You’ll see that the paving trucks are actually being held in place by other equipment because otherwise, they would slide (or worse, tumble), right down the track. And that would make for some pretty sloppy surfaces to race on.