Michelin's Chris Baker, Director of UHP and Motorsports Business Development for North America, tells of an engineer hired from the U.S. Navy's nuclear submarine program who was shocked to find that security at the tire company was far more stringent than at his old job. Secrets at Michelin are secret.
That's still the case, but glasnost has taken hold in some areas. One place that has opened up dramatically in the last few years is Michelin's Laurens Proving Grounds in Laurens County, S. C. It's a test facility for all of North America, and in recent years has been used for more outside testing and public demonstrations. For example, Laurens is about 40 minutes from BMW's Spartanburg assembly plant, making it convenient to use the Michelin facility for driving schools, future-product testing and press launches.
Opened in 1976, Laurens Proving Grounds' 3,000 acres are largely undeveloped forest. It's so big that even the facility manager can't point exactly to the boundaries on the ground. One crosses a nearby swamp, and there's an early-19th-century cemetery on the premises. In the middle of it all, and protected by a very serious security force, are 13 separate test tracks, most of which are multi-configurable. The various tracks are built with areas dedicated to wet and dry handling, inside and outside noise, comfort, wet and dry braking, hydroplaning, dry and wet off-road capability, curb impacts, wet tread photography and dry and wet road holding. Some courses are rarely or no longer used, and the purpose of some has changed. The max dry handling course, for example, was originally designed as a heavy truck tire durability test. It would be difficult to add up the mileage of every loop, but it's well into the double digits.
european car spent two educational days with Chris Baker, Mac DeMere, and other Michelin test drivers and engineers while they answered the question, "What would you say it is you do here?" Mac, a refugee from a decade at Motor Trend, has been a major element behind the opening of the proving grounds.
Michelin's Laurens Proving...
Michelin's Laurens Proving Grounds. Most of my two days were spent on T1 (noise and comfort), T2 (dry handling), T3 (wet handling), T8 (vehicle dynamics area, the "black lake") and B1 (small wet skidpad).
Michelin is a $20 billion company worldwide, and Chris Baker said it spends a larger share of its revenue on research and development than any other company its size. A tire line used to require only 15 to 20 sizes to cover 90% of the market, but now it may require 60 to 80 sizes. It's possible to test only six to ten sizes of any tire in detail, so the rest are computer modeled and tested in the lab. The correlation of real-world testing performed at places like the Laurens Proving Grounds (LPG) to laboratory tests is crucial. Tire companies have driven the technology and power of finite element analysis (FEA) more than any other industry. Beyond its sheer complexity, a tire is a modeler's nightmare in that it is both non-linear and visco-elastic. Michelin claims to have excellent consistency, with very few dogs (where one particular size of an otherwise good tire line is a poor performer), which it states points to the quality of its engineering.
Driving
My two days gave me a good feel for the most-used parts of LPG, focusing on tracks 1, 2, 3 and 8. It started with driver training on the smaller wet skidpad, moving to the wet handling course, a technology session, a dry autocross and, finally, the noise and comfort loop.
Michelin's Chris Baker checks...
Michelin's Chris Baker checks my times in the Porsche with Pilot Sport Cup tires.
Driver training sometimes goes by the term "skills development," an LPG test driver's code for screwing off. Some of Michelin's test drivers are qualified to be top-level race drivers, but are experienced enough in life to want to go home to their houses and families most nights. These guys are good. The first thing Mac did was pass some of the basic skills along to me. I had to learn to drive sideways in the wet. The outer skidpad was being widened while I was there, and I provided cheap entertainment for the workers.
The car was an Infiniti Q45. It's big, easy to keep up with and rwd, like most of the vehicles at LPG. Driving sideways was completely foreign to my usual cautious approach of sneaking up on the car's limits. Get some speed up, just to the point the car wants to understeer, then stomp on the throttle to kick the tail out. As it's going, add throttle and wind the steering out. You have to be aggressive and hold the throttle down. Driving this car was a challenge; the automatic dulled the torquey V8's throttle response and, without a limited slip, the revs would climb without losing rear grip, because one wheel was spinning while the other gripped.
I haven't been to Tim O'Neil's rally school yet, and this was my first experience with an instructor wanting me to go sideways. Most normal programs chase speed, focusing on smoothness and not exceeding the limits of traction. That's what I did when learning to drive my own car on a track. I've learned it's important for a driver to understand in the seat of his pants that exceeding the limits of traction doesn't mean the end of control. That's far more important to begin with than focusing on speed. Mac DeMere said, "Spinning doesn't teach you anything." He also had this bit of advice: "If you know you're going off, go straight. Maybe you'll hit the little tree instead of the big one, or even miss them both."
Mixability Test
When I could hang the tail out as long as the car would support it, and spins became acceptably infrequent, we moved to the wet handling course. The Q45 was set up for a mixability test, with an old and prototype new version of a Fisk Classic--one of Michelin's commodity brands. A mixability test is subtle, something most civilians wouldn't think of, but it's done any time a production change is made to a tire. Perhaps the engineers find a way to make a tire better with no added cost, or moving production to a plant in a less expensive country or the purchase of better manufacturing equipment lowers costs, allowing improvements to be made. That would appear on the face of it to be an all-good situation, but if a tire is too much better, it may be incompatible with the tires already in the distribution network. The test is basically to decide whether it's safe to mix new and old tires.
Which leads to a consumer warning: It is best to replace tires four at a time, but if replacing only two, always put the new ones on the back, whether the car is fwd or rwd. With less tread depth on the rear, the rear will hydroplane first in water. If that happens in a turn it results in oversteer so sudden and dramatic that Mac estimated, "Maybe 15 of the starters on the grid at Indy would catch it without being warned.""
The mixability test is done to determine what happens if the better tire is put on the front of a car. If old and new tires are mixed in inventory, even replacing a set of four could lead to this situation. We began on the short loop of Track 3, the wet handling course, with the old-version Fisk Classics on the front and the new on the rear. There was not much water-depth sensitivity, they were pretty good at putting the power down, and I didn't over brake them once. Turn-in was poor, but understeer was comfortable. Sprinklers across the windshield were a major distraction at first, but it was "one of the challenges of this course" and I eventually grew used to it.
With the tires rotated front-to-rear, water-depth sensitivity was about the same and braking was just slightly better. Turn-in was almost too good, tending to lead me into a pilot-induced oscillation, and acceleration and rear breakaway were less than acceptable. Because I was still learning, my impression was the latter setup was easier to drive; I was growing in my ability to catch and slow the vehicle, and stuff was happening slowly. The official result (which had nothing to do with my input) was that mixing these versions of the Fisk Classic would be unacceptable. LPG hands that verdict back to Michelin, where business types decide how to handle it. Options include blowing out the current inventory before introducing the new tire, or simply not making the improvements to the tire until it is replaced with a new model.
Wet-handling Comparison
I've been to many manufacturer-arranged tire tests, but the wet handling comparison was the clearest demonstration of differences under a set of conditions by tires in the same segment I've ever seen. Again we used Track 3, but this time in a Mustang GT five-speed on 245/45R17 tires. The tires were Michelin Pilot Sport A/S (all season) and Pirelli P7000 SuperSport. The car was a big pig but fun once I stopped judging the car itself and just drove it. Very few people would ever choose to drive this way in the rain, but it does usefully capture emergency/accident avoidance capability in the real world.
Pirelli's P7000 SuperSport shares its tread design with the intermediate tarmac rally tires used to win the 1995 Monte Carlo Rally. The production tire features many lightweight features of a racing tire, such as an aramid-reinforced carcass, civilized for street use. In the past, european car testers have found this tire to provide good dry grip while remaining highly civilized and quiet. The SuperSport version of the P7000, Pirelli has reported, adds year-round traction, even in light snow.
Michelin introduced the Pilot Sport A/S, an ultra-high-performance all-season tire, in 2001. It was engineered to take advantage of, and in fact requires, a new, flexible manufacturing technology Michelin calls C3M. Nobody I talked to would admit to knowing what C3M is and how it works, but it has many advantages. Fundamentally, C3M yields the most uniform tire Michelin has ever made, reducing vibration and wear. C3M allows Michelin to build shorter runs of tires with greater precision for flexibility in production and more rapid product development cycles and to explore new architectures.
Michelin states the Pilot Sport A/S is the first tire in history to have three tread compounds on the surface: The center rib of the tire contains 100% silica for wet-weather traction and stability and is backed up by a narrow band of polyester cords. On either side, in the meat of the angled tread pattern, is a softer compound to give traction in winter weather and at colder temperatures. The shoulder is a carbon-black compound for ultimate lateral performance. For more details on this tire, see european car's Ultra-High Performance Tire Buyer's Guide in the June 2002 issue. Basically, this test boiled down to the best ultra-high-performance, all-season tire of 1995 against the best of 2001.
On the P7000 SuperSports, braking was good, the tail didn't step out too far under lift-throttle conditions, and acceleration led to understeer at first, with a transition to power oversteer with too much throttle. Steering response was not super-impressive, but steady-state balance was good. I thought that breakaway and recovery characteristics were both good, not too sudden to catch.
On the Pilot Sport A/S, braking was much better, lift-throttle oversteer was nonexistent and acceleration was very good. There wasn't much water-depth sensitivity, and the steering was much better. I called steady-state balance excellent. The most amazing thing was the increase in progressivity of both breakaway and recovery. I noted simply, "Wow." My extended notes read, "Steering much better, even in deep water. Breakaway and recovery much smoother; easier to play with it, to be confident and smooth. Much better power delivery. Slides nicely."
Mac had me switch back and forth several times. My initial best time for the course on the Pirellis was 51.30 sec., which dropped to 46.69 sec. on my first lap with the Pilot Sport A/S. I immediately felt a huge increase in confidence. My eventual best times were 49.62 sec. on the Pirellis and 45.55 sec. on the Michelins, a dramatic difference. Michelin states on some vehicles, the Pilot Sport A/S is very close to the Pilot Sport in the dry. If I was to choose a tire for One Lap of America, I would pick the Pilot Sport A/S and pray for rain.
Dry Handling
This was the official, traditional "fun part." One of many standard autocross courses was set up on Track 8's "black lake" section, which measures 1,400 x 400 ft. A patience-teaching slalom was set up on a straight, with a series of five tight turns linking it to a fast, increasing-radius turn, ending in a tight left hander into a short slalom and a sharp left onto the main straight.
This course was used to compare two Michelins, the Pilot Sport and the Pilot Sport Cup. Both were factory size on the Porsche 911, 225/40R18 front and 285/30-18 rear.
The Pilot Sport is the ultimate street tire in Michelin's lineup. Its design objectives were to maintain the peak dry performance of the Pilot MXX3 it replaced while increasing wet traction and treadwear characteristics. It had to be good enough to be original equipment on the Porsche 911 GT3, a serious attempt at the ultimate road-legal handling machine. Progressivity at the limits of traction was held to be especially important. Much of the Pilot Sport's technology can be traced to Michelin's European racing efforts in the early 1990s, leading to dramatic improvements in high-speed water evacuation and tread life. (For more info, see the UHP guide in european car, June 2002.)
The appearance of the Pilot Sport Cup has nothing to do with the decision for BFGoodrich to pull out of the club racing scene with the Comp T/A and g-Force R1 lines, but instead grew out of an agreement with Porsche to be that company's exclusive motorsports partner. A Michelin engineer previously told me that what this tire really brings to the table is useful life. It beat one competitor's off-the-shelf slick tire, he said, and compared favorably against other DOT-legal R-compound tires. It was only half-worn after 156 laps of a 40-sec. course, and it has no requirements for heat cycling or shaving prior to use. It was designed to be driven to and from the track for club races, so it is much more streetable than any other track tire, according to Michelin, with no decrease in performance over time. Internally, the Pilot Sport Cup is much like the Pilot Sport, but uses thinner, ultra-high-tensile wire, making the carcass both stiffer and more flexible in the ways those are desirable. It is a dual-compound tire, with mixes derived from the Michelin's GT race-tire family.
I drove back to Greenville on the Pilot Sport Cups the evening before this test. They seemed very sensitive to pavement roughness for noise, but were less harsh on impact than european car's long-term 996. I never felt any loss of traction in street driving on wet pavement, even through light rain.
The test drivers' rating sheet includes spaces for corner placement and under the heading "stability/balance" the following: front/rear, lift-throttle, trail braking, progressivity, power wheelspin, lateral firmness, lateral grip and going away.
On the autocross course, the Porsche on Pilot Sports was, to use Chris Baker's description, "a good dance partner." It had a very smooth transition from grip to slip to slide and back. It tended to push in slower corners, but toward the top of second gear in the fast sweeper, I could make the car do whatever I wanted. It was totally benign, an awesome street car.
The course really drove home the truth of an Emerson Fittipaldi quote, "The difference between a fast driver and a slow driver is that a slow driver goes fast in the slow parts and slow in the fast parts." The correct way is to "survive in the slow parts and set up for the fast parts." If you can learn to do this, you will be fast.
When I switched to the Pilot Sport Cup, I was giggling. Wow. There was tons more grip. It took me some time just to get used to going as fast as I then could, and not get behind on the rhythm in the turns. I could brake in a straight line harder and deeper, and turn-in was better. It was so much better, in fact, that I had to learn to turn later. Initially, I sometimes took out cones on the inside of what I had planned to be a late-apex corner entrance because they went under the nose. It took a lot more to make the Pilot Sport Cups slide, so adjusting the line with the tail took more work and commitment, but was still the fast way. My eventual best times were 40.76 sec. with the Pilot Sport and 38.55 sec. with the Pilot Sport Cup. Extrapolate that difference to a longer course, where 1:30s or 1:50s are expected of a fast car, and the value of the Pilot Sport Cup becomes even clearer.
Interestingly, the course used two days later at M School with the M3s was a faster, less technical version of this one, giving a same tire, same surface comparison of the M3 and the 911. The Porsche is smaller, lighter, torquier and much better balanced. The M3 pushed more in the tight stuff, and in the fast sweeper, lifting the throttle moved the balance from understeer to less understeer. It could be rotated with throttle oversteer in tight turns, but that was rarely necessary or even possible with the rear-engine car.
Noise and Comfort, Soft Handling
We were loaned a BMW 745i equipped with Bridgestone Turanzas, the factory's 245/50R18 fitment, and compared it to Michelin's Pilot Primacy, a factory option, sized 245/45R19 front and 275/40R19 rear. Noise and comfort are separate factors in how civilized a tire is. Evaluating them requires a complete change of gear and change in attitude. Test drivers are no longer the kid in the candy store going as fast as they can, but are restrained, careful engineers. It is just as difficult to do well, perhaps even more difficult when differences are slight. Mac said it takes a couple years before a new test driver is considered really reliable at this. Engineers are always attempting to find ways to measure the differences with instrumentation, but none have yet been as good as the best humans.
Noise and comfort are tested on T1, the track with long straights and turnarounds at each end. The straights all consist of special pavement sections designed to excite the tires in certain ways that are known to generate certain kinds of noise or motion. Every individual test is run at a certain speed. This is one thing that was stressed throughout my time with Michelin: If a tire is not evaluated for noise and comfort on the same lane of the same stretch of road at the same speed as one it's being compared to, the test is meaningless.
Sound quality is as important as volume. The sound "pressure" can be high, but if it's spread out over a wide-frequency spectrum, there isn't any peak associated with a particular vehicle frequency. If a tire has noise peaks, there is a chance they could coincide with vehicle resonance peaks and become very obtrusive in the cabin.
Noise is tested with the ventilation fan and stereo off so the car will be as silent as possible. The first test, tread-pattern noise, is carried out on smooth pavement. The engine is cut and the car coasts from highway speed to a stop, the changes in volume and quality of the tire sound noted along the way. Ideally, the sound should progress smoothly, with no accentuated volume peaks along the way. Other noise qualities are tested at constant speeds and even throttle openings over special pavements: sizzle, impact noise, rolling boom, road noise, joint slap and cornering squeal. Each of these is very subtle, and simply learning to distinguish and recognize them accurately is a true challenge, even for a tire geek. Each quality is rated on a point scale. I was provided with Michelin's own tester's score sheet for both tires. On some, the Michelin was better; on some, it was worse. None of the differences were greater than half a point, and the total score was virtually the same.
The comfort testing is similar, only the challenge is to feel the tires without listening to them. Having the A/C on is good; some testers find a classical music station on the radio. Categories evaluated include impact harshness, vertical firmness, damping, smooth-road isolation and coarse-road isolation. As with noise, special pavement sections are designed to excite each characteristic. Michelin's own ratings place the 19-in. Pilot Primacy within 1/2 point of the 18-in. fitment; some better, some worse. The total scores were acceptably close.
We also drove the 745i on the dry handling loop. This is a beautiful circuit, with a flowing rhythm that's easy to learn. It has FIA-legal curbing on its apexes, but it is not a racetrack. It was designed as a truck-tire durability loop, with the assumption that drivers who went off the pavement would soon be finding new work. In many places, going off means tumbling down a hill into trees, and there are no barriers. This handling test was done not at the limit, but about 7/10 or 8/10, the way an enthusiast might reasonably drive on public highways. Major areas on the evaluation sheet include straight-line stability, on-center steering, off-center steering, transitional stability and cornering. Differences between the tire fitments were subtle and fine, but the 19-in. fitment did have a slight edge in the ways conventional wisdom would suggest.
Every tire is evaluated this way independently by multiple drivers, and Michelin's internal standards require their ratings correlate within 1/2 point at least 80% of the time. The purpose of this demonstration was not to show one tire being better than another, but that an extreme fitment like the 19-in. tire could be comparable to a less-dramatic 18-in. tire. Conventional wisdom says that making such a step will result in decreased comfort, but that's not necessarily the case.
What It's All About
That's a long answer to the question, "What would you say it is you do here?" In fact, there was a lot of learning about driving, tires and people that we just don't have space to include. Places like Michelin's Laurens Proving Grounds are important because, in the end, cars are driven by people, not by computers. While machine test data is critically important, and is sometimes capable of precision and clarity that a human's soft parts can't match, it has so far proven impossible to design instrumentation, or even define measurements to be made, that can capture the many subtle and complex characteristics of tires. The human element is the final test and is an absolute necessity in the development process. Every Michelin test driver I met is a raving motorhead to the core and cares deeply about the work. What european car learned while visiting LPG is only a small part of what happens there, but we still gained a lot of insight into how Michelin tires become good enough to be original equipment on some of our favorite vehicles. I'm hoping I've helped readers understand that there's a lot more to a tire's performance than how many lateral g's it can pull on dry pavement.