Be very, very careful what you put into that head, because you will never, ever get it out. Thomas Cardinal Wolsey (1471-1530)
Tolerance. In the immortal words of Matha Stewart, “It’s a good thing”. In this case, we’re actually talking about engineering tolerance, although tolerance of other things — like people who don’t know what they’re talking about and persist in talking about it anyway — remains a virtue to strive for.
Engineering tolerance is “the permissible limit of variation in a physical dimension or other property”. In English, that means how far you are allowed to miss the target number. The idea of an engineering tolerance is that you are trying to achieve the exact number, but you recognize that getting it that precise is unlikely because of random factors beyond your control. There is a difference between precision – how close you are to the target – and accuracy, which has more to do with how consistent you are.
Let’s use the analogy of building a car with hitting a bulls eye with an arrow. The goal is to hit the bullseye and let’s say you are given five shots (i.e. five distinct cars). The top left picture below shows the ideal. All five arrows are in the bulls eye. Now, in reality, you’re more likely to have something like figure B, where the average of the shots is in the bulls eye, but the individual shots are scattered a little bit. NASCAR cars are made primarily by hand and you have to expect a little variation. NASCAR would write a rule saying that a shot would be legal if it fell anywhere within the third circle. The goal is the bullseye, but we’ll give you a little breathing room. That’s tolerance. (Actually, NASCAR would probably write a rule that they owned the target and the bullseye had to be a particular shade of red and you had to pay them royalties if you wanted to reproduce it.) Hold Figure C for a moment while I explain why this archery analogy is at all relevant to NASCAR.
After the Dover inspection, crew chiefs Chad Knauss and Alan Gustafson were told “not to bring the cars back”. In NASCAR history, those words have a very particular meaning becuase they usually are the words that are used when a team figures out how to do something that violates the spirit, but not the letter, of the law. NASCAR can’t actually nail them for it, but they want to make sure they know that they won’t stand for it. In other words, OK, you got away with it once, but don’t try it again.
That isn’t what’s happening here. First, there was nothing illegal about either car either week. John Darby, Sprint Cup Series Director made that verey clear. This was a friendly reminder to the HMS teams that they were closer to the tolerance limits than NASCAR would like them to be.
How close is close? If I remember right (and I’m on the road away from my records), the tolerance on the body is one-eighth of an inch. If something is supposed to be one-inch long with a 1/8″ tolerance, it could range from 7/8″ to 1-1/8″ long. An eighth of an inch is 0.125″ or 125 thousands of an inch. The 5 car from Dover apparently was seven thousandths of an inch (0.007″) away from the target plus the tolerance. A ream of 20 lb paper is just about 2 inches thick. There are 500 sheets in the ream, so the thickness of one sheet is 2″/500 = 0.004″, or four thousandths of an inch thick. The 5 car was about two sheets of paper away from being illegal – and note that the emphasis is on ‘away from being legal’. No rules were broken here. So why is NASCAR making such a big point about it?
Well, now go back to the drawing above and look at Figure C. In figure C, all of the shots are within the three-ring limit; howevver, they are all in exactly the same place. The point of tolerances is that you are supposed to be trying to achive the actual number, not getting as close as possible to the absolute limit.
Now, that’s really sort of a silly thing to expect teams to abide by. There is a 5 mph tolerance on the Pit Road speed limit. No one is trying to go the Pit Road speed limit. If the Pit Road speed limit is 45 mph, everyone is trying for 49 mph or 49.5 mph or 49.99 mph. When Juablo got nailed for speeding, he was 0.06 mph over the limit. Or was he 5.06 mph over the limit? Trust me, every single team in the garage is playing the same game. It’s how you win. There’s nothing illegal about doing it, but NASCAR wanted them to know that they were playing with fire.
We’re talking about thousands of inches here. Measuring things is not as straightforward as it seems it should be. If I give you the same piece and asks you to measure it fifty times, you will come up with a range of measurements. That’s the nature of measuring things. So it is entirely possible that a car leaves the shop being perfectly within the rules according to the measurements there, but when it’s measured at the R&D Center, it’s illegal. Don’t forget that the car runs a race and (more often than not) the car hits things and that will affect the positions of parts as well.
So how do you get that detailed in your measurements?
Teams use templates at the shop; however, they rely much more on coordinate-measuring machines (CMM). CMMs consist of a probe (which may be mechanical, optical or other) and a reading device. Modern versions are attached to computers to collect the readings. Mechanical devices include the Romer and Faro arms, which are brand names of popular CMMs. These device looks like arms, with joints that mimic the elbow, wrist and fingers. Those joints allow motion along all three axes (up/down, left/right and back/forth), plus the ability to rotate about each of these axes. (Check out this interview with the inventor, Homer Eaton.) The arm is touched to the car in specific spots. The probe transmits its three-dimensional coordinates to the computer, thus forming a precise 3D map of the car inside the computer. The picture below shows a Faro arm (top) and a Romer arm in use at the NASCAR R&D Center (bottom). The Romer arm is used to certify the chassis (with over 100 distinct points tested) as well as measure the body position and sheet metal thicknesses.
One of the challenges using mechanical CMMs is measuring over a very large volume. The NASCAR system measures over a 13′ x 20′ area defined by a set-up plate. To improve the measuring accuracy, 5/8″ diameter seats are mounted in the plate every three feet. The placement of the seats is verified during installation using laser triangulation. Before measuring the car, the probe is touched to any three of the seats, which ensures that the probe uses the same origin every time. Triangulation is also the basis for the CMM. Remember all that geometry you learned in eighth grade? If you have an unknown length – the distance from the origin to the pointer – you make the unknown length one leg in a triangle. If you know the length of one side and two angles of your triangle, you automatically know the lengths of all sides and all angles.
On “This Week in NASCAR”, Chad Knaus noted that HMS only had one Faro arm available and would usually measure the cars right after construction and not much after that. He also reported that they bought a second machine ($60,000) so that both the 24/48 and the 5/88 shops would have one. I”ve also heard that there were some differences in how HMS was measuring the cars and how the NASCAR R&D center was measuring the cars. Differences in exactly how the measurement is done might make a measurement look fine at the shop, but be over tolerance at the R&D Center. That wouldn’t be a big deal if you were aiming for one inch and your limit was 1-1/8′, but if you’re getting as close as possible to the 1-1/8″ number, a couple of sheets of paper make a big difference.
The mechanical arms are really nifty pieces of technology, but laser scanning takes accuracy one step further because the only thing touching the car is light. A laser stripe is focused on the car, and a sensor analyzes the line on the surface of the car. A computer program uses triangulation to back-calculate what surface shape would cause the observed line to appear as it does. I’ve seen a couple of these systems in action at the major team shops. They are fast and accurate. They are also very expensive, but I know a couple of teams that laser scan every car before and after they hit the track, looking for subtle differences that might mean a few hundredths of a second per lap.
I’ve had a difficult time getting this confirmed, but what I’ve heard is that the area of the car being questioned by NASCAR is the rear end and its offset from the centerline. The old car was remarkably asymmetric, as shown below. See how it’s almost jelly bean shaped, curving to the left in the front and the rear? That assymetry is one reason the old car turned better than the new car. The new car (rear view below) is still a little asymmetric. Note the position of the wing with respect to the decklid to see the asymmetry; however, the asymmetry is much smaller in the new car. Apparently, a number of teams (HMS isn’t the first one to be warned) have been pushing the rear end as far as they can within the tolerances to help the car turn.
All NASCAR was doing was reminding the teams that that one reason for a tolerance is that there are bound to be discrepancies in measuring or fabricating and the tolerance is there to give you a safety cushion. The point is that you design for the specification and hope it is within tolerance. The teams, on the other hand, look at the tolerance as being “the grey area” and therefore legitimate for them to work with. Given everything that NASCAR prohibits, you can’t blame the teams for zeroing in on those things they can still innovate with.
So let’s just put the black helicopters away for the moment, remove the aluminum foil helmets and back slowly away from the digitizing arms. Remember that, in NASCAR, the best innovations are the ones that they we never hear about.