Power to Win
The challenge was formidable. Although Mika Hakkinen used Mercedes Ilmor engines to win consecutive F1 championships in 1998-99, the company went from winning nine races and the 1997 CART engine manufacturers Championship to going winless in the 2000 campaign. That result added little to the Mercedes reputation for engineering excellence.
Ilmor is a dedicated racing engine producer
based in Brixworth, England, formed by two former engineers of Cosworth
Engineering, producer of Ford racing engines since the 1960s. Mario Illien
(the "Il") and Paul Morgan (the "Mor"), approached
Roger Penske about forming their own company in 1983 and, with General
Motors as a 25 percent partner, Ilmor Engineering was born just weeks
But challenges from other manufacturers increased
the competition - so much that cost escalation caused Chevrolet to relinquish
its share in the company. In 1994, Penske's network of contacts brought
Mercedes in as a replacement, an enthusiastic partner with sights on Formula
1 as well as Champ car racing.
The biggest problem we face today is that the
life of an engine is becoming shorter and shorter," said Illien,
a native of Switzerland. "With competition (from Ford Cosworth, Honda
and Toyota in the CART series, from Ferrari, BMW, Renault, Jaguar Cosworth,
Peugeot, Honda and, in 2002, Toyota in F1) you have to update it much
faster than you used
"For instance, the original CART 265/A
engine was alive for five, six years. That's impossible today. You need
a new engine every year. That is quite tough on the building side, the
design side, the manufacturing side."
To meet the challenge, Ilmor drastically changed its structure. It nearly doubled the workforce from 155 to 406 employees and expanded facilities from what began as a single 6,000-square foot building to a complex comprising six permanent buildings and nearly a dozen temporary construction trailers totaling nearly 70,000 square feet. Much of that growth came after Mercedes executives explained their goals for Ilmor.
"When we first became involved
with Mercedes," Morgan said, "it was the first time we could
fund both programs properly, but the difficulty is that they demand a
first-quality program that is going to win races. So we had to change
our company from the way it was operating at the time. We needed another
150 people to be able to accomplish more research and development and
to have a greater capacity for manufacturing pieces.
"It was done to improve both time and quality. We've always tried to have the facility to make any part of the engine in a fast time. A good example is the crankshaft manufacturing shop where previously our supplier used to take 22 weeks to manufacture a crankshaft.
"Sixty months ago we were down to 10 weeks internally, 30 months ago we were down to eight and now we're down to six. We want to get lower than that, possibly to as low as four. By taking on many projects we had previously assigned to outside suppliers, we have more work, but we also have more control on time and quality.
Mercedes gained more than just new engine pieces.
Raising the Bar
The way the company came up with new ideas and introduced them could help Mercedes, said Illien.
"It's the process itself - the way of thinking, of looking at the whole engine and the way we go about putting through a complete engine in 6-7 months from first line to running on a dyno. It's more the methods of how you attack a new program. I think that can be a good lesson for production cars because you are pushed by the competition and have to come up with new things faster and faster."
To achieve greater efficiencies, Ilmor substantially changed its corporate structure. Morgan and Illien, who now delegate much responsibility, try to meet with small employee groups every three or four months to review what the company has recently achieved and to discuss what is anticipated over the next six to 12 months.
Planning New Powerplants
For 2001, that direction applies to a new versions of the FO110 V10 used to win the last two F1 championships. Interestingly, although V10s are relatively unusual in production cars, they are the most popular configuration in F1.
"The V10 is pretty much where everybody is," Morgan said. "With the V12, the cylinders are nice and small, but the friction losses tend to be too high. The V10 produces optimum performance. With the V-angles less than 90 degrees, they do vibrate like hell, but we have got used to it and it isn't an issue.
"When you're used to building a flat-plane V8 and you go and look at a 90-degree engine like in NASCAR, you just can't believe that a magneto on a big stalk or any of the other components are going to stay in place for five minutes without falling off. You have to tie things onto a flat-plane V8 very strongly.
"We don't use a flat-plane design because the exhaust tuning is such a very important part of getting the engine working, but that may change if we're asked to help the Dodge NASCAR program. With the V10, you also have to have equal firing intervals down one bank - that's the only way to do it."
From Concept to Build
The design team takes that loose set of ideas and then uses CAD (Computer Aided Design) to do most of the work to create an engine. The CART engines are simpler and more restricted than the Formula 1 engines. (Rules specify many dimensions, materials, and system operations, for example, gearshifts and throttles.)
For most components, the designers know what is practical and what isn't. All untried major components are reviewed prior to creation with the staff of the production engineering department. Development begins soon after the engine is run on the dyno. Teams start testing and data can be accumulated. Once on the track, competition with other engine manufacturers takes priority.
Getting it Right
"The keenness to get new parts is intense," said Andrew Hurley a 13-year Ilmor veteran who heads the development department. "Since we generally want to add another 2, 3, 400 rpm each year, it's a new challenge. You just never know where a problem may surface - springs, valves, pistons, rods, valve seats. But as an engineer, it's just a fabulous environment. I have absolute, complete engineering freedom to solve any problem. It's not a question of costs - you've just got to make it better."
Mercedes has contributed significantly to this aspect of the program.
"One of the most important things in performance improvements in motor racing is in materials technology," Morgan said.
"Mercedes has got an extremely good materials technology laboratory at Uhlm, just south of Stuttgart. Mario's been there two or three times and he's brought back some useful information that we've been able to put into the engines."
As a development engine is built, it finds its way to one of Ilmor's dynamometers. Three are conventional and measure torque and calculate power. A fourth dyno, used primarily for F1 engines, is another contribution from the three-pointed star.
"Mercedes has a pretty advanced transient dyno facility which we were able to somewhat duplicate," Morgan said. "In Formula 1, the throttle and gearshift are hydraulic so the driver inputs are basically electrical.
"You can simulate that with a data recorder in the car as it goes around the track and see how it performs. You then program that into the computer and you can have a complete race simulation on your dyno. With a gearbox attached to the engine, if you close your eyes, you can't tell whether you're at Monza (for the Italian Grand Prix) or in Brixworth.
"We can run race distances in here and we typically run tracks that are hardest on engines like Monza and Hockenheim (site of the German Grand Prix).
"We have quite a big flywheel, which simulates
the weight of the car, and a water dyno, which simulates the wind resistance.
This way, we can tune the rate of acceleration to be similar to a real
race car as it goes down the straight. In Stuttgart, where we also do
quite a lot of transient dyno testing, they have the facility to run the
gear box complete with its
Once a design is finalized, each of the more than 4,000 components is built and the engine is tested. If it runs properly, the components are taken to the inspection department to have every dimension carefully measured - sometimes to 1/5 of a micron or 1/250 of the diameter of a human hair! That data is input into a computer and downloaded into Ilmor's intricate collection of milling machines, which generally produce pieces in batches of 14 to optimize production capabilities.
"We have about eight different types of machines to do different tasks and about 40 machines in total," Morgan said.
"Virtually all them are five-axis machines - a turntable which can tilt forward and backward and the three linear axes.
"One of the problems we faced with making pieces with no one here is for components which might take only 20 minutes or so to make. How do you keep the machine from running out of work?
Burning the Midnight Oil - Alone
"We have a very good relationship with
Mori Seiki, the biggest machine tool manufacturer in Japan, and we discussed
this problem. Between us we came up with a solution. Most of the machines
have 200 palettes to hold different components, so we have as many as
65 different jobs programmed into the machine, which has over 200 cutting
tools. Of the 65
Using these CNC (Computer Network Control) machines, 14 examples of a complete set of castings can be made every four weeks. With improved planning, spare pieces are added to inventory quicker and in greater numbers.
In the sub-assembly area, all the water galleries are pressure tested, the sharp corners are taken off, crankcases are assembled and the main bearing tunnel can be line-bored. Then those pieces are inspected for consistency.
Cookie Cutter Accuracy
"After you make one powerful and reliable engine," Morgan said, "the key to success is in how accurately you duplicate the engine specifications. It's essential to get the replication accurate and that's the reason you need good quality machines. If every engine is exactly the same, they'll all produce the same power with reliability. Every engine we produce is within half of one percent of any other engine in performance."
The cylinders and inlet systems are built in
separate departments and then the core of the engine is built with pairs
of engine builders working on the engines. The other pieces come back
and are bolted on as complete units. Cylinder heads and oil and water
pumps come ready finished, but everything else is put together by the
two builders on each engine. Total
Once an engine takes to the track, the whole process begins again.
"Development is a continuous process,"
said Paul Ray, an Ilmor Vice President. "In an ideal world, you'd
try everything out before you ever got to the race track, but this isn't
an ideal world and you discover things when you discover them. You can't
guarantee you'll discover everything before the first race and you certainly
don't stop looking once the season
Pushing the Limits
"Some gains can give you two or three horsepower, but they're simply not worth (the effort and expense) of manufacturing the parts. You wait until you find a combination of things which give you a reasonably large gain. But the level of competition is forcing us to push the limits and run the engines much harder than we would like."
Unfortunately, failures in racing tend to be very public. Ilmor's CART troubles can be traced to the failure of the IC108E engine, successor to the 1997 champion IC108D. The "E" appeared to be a two-thirds scale model of its 1997 predecessor - length was reduced by 17 percent, width by 12 percent and height by 7 percent. The result was an engine with a total volume some 31 percent smaller than the previous model, and the new package weighed just 220 pounds - 50 pounds less than the IC108D.
Changing the Rules
But in the 10 months that elapsed from the start of the design to the completed product, CART changed its chassis rules for safety reasons and the aerodynamic and weight advantages provided by the IC108E were eliminated. By making everything so small, the driveability - the power band and consistency of horsepower production - and durability of the IC108E suffered, contributing to the drop off in success.
The introduction of new technologies were also delayed because of the small size of the new engine, allowing rival manufacturers to catch up and then surpass the performance advantage Mercedes had enjoyed in 1997.
Teams running Mercedes engines had other difficulties. Some were contracted to run uncompetitive tires, some had drivers that were either too young and inexperienced or too old and unwilling to drive hard, some lost key crew members (weakening the fine tuning necessary for top performance). The result: was that every team running Mercedes engines in 1998 and 1999 had several serious problems - the power plants were just one of them.
When Penske Racing and Forsythe Racing switched
to Honda power for 2000, Mercedes was left with PacWest Racing (winless
since 1997 with two of the three oldest CART drivers) and Mo Nunn Racing
(a totally new single-car team created by the chief engineer to the last
four CART champions) as their best hopes. Ilmor snared a second place
As the season wore on, Mercedes-Benz executives made a difficult decision and called a news conference at the 15th race of the 20-race season.
"We have had to reevaluate our motorsport strategy from 2001 onwards in order to meet higher targets in the Formula 1 World Championship where eight manufacturers will be competing beginning in 2002," said Juergen Hubbert, Daimler Chrysler board member supervising Mercedes-Benz motorsport activities.
"We are also very keen to further develop the DTM Touring Car Series, especially in Germany, the home market of Mercedes-Benz. In order to concentrate our efforts on these two series we have to stop our CART program."
"It's a sad day," said Morgan. "I know Mercedes-Benz deliberated long and hard but, at the end of the day, the Formula 1 program took precedence. They (Mercedes) are going to be facing even greater competition (in F1) than before with another engine manufacturer coming in, and, at the end of the day, it's better to have one good program than two mediocre ones.
"Ten years ago, we were dominant in CART and nowhere in Formula 1. Who's to say what the future may hold? Someday we may very well be back."
Getting A Fresh Start
Mercedes, however, wasted no time in turning its full attention to F1. The new Mercedes engine for F1 designated FO 110K V10 is a completely new design. The first dyno testing was completed on November 14 last year, and to date has run 23,000 km (14,292 miles) in static testing.
The first actual test in the MP4-15K vehicle platform was initiated 20 days later and completed 10,000 km (6,210 miles) in the capable hands of drivers David Coulthard, Mika Hakkinen and Alexander Wurz.
New side impact safety technologies designed in accordance with new FIA (Federation Internationale de l'Automobile) regulations are being incorporated into the new car, designated MP4-16, as well. The first real world test is at the F1 season opener March 4 in Melbourne at the Australian Grand Prix, with Coulthard and Hakkinen driving and expectations high that Mercedes' refocusing of its motorsports developmental efforts, in conjunction with long-time partner McLaren, will do them proud.
Back to Front Page Now or return to Top or visit the Archives
Site Map ... Go