Mercedes-llmor 500i

One engine. One race. One win. Mario Illien talks about the push-rod engine that was developed in secret, produced more than 1000bhp and won the Indy 500 for Penske in 1994

For Penske Racing, 1994 was an annus mirabilis. Its three drivers – Emerson Fittipaldi, Paul Tracy and Al Unser Jr – won 12 of the 16 Indycar races, 10 from pole position, and finished line astern in first, second and third places on five occasions. At Indianapolis, even through Tracy retired with turbo failure and Fittipaldi crashed trying to lap team-mate Unser, it famously won the race with a mighty Mercedes-Ilmor engine built specially for that one event.

The year before, USAC had relaxed its rules on pushrod engines for the race. Intended to encourage the entry of production-based powerplants, these had initially insisted that the block be a stock item from a road car, with the carrot that it could then have 3.43-litre capacity, 30 per cent greater than OHC competitors, and 22 per cent more turbo boost. What changed for 1994 was that a stock block was no longer mandatory, opening the door for a purpose-built pushrod engine that in all other respects was state of the art.

In great secrecy and during a period of only 10 months, Mario Illien of Ilmor Engineering, who talks about the engine below, designed the 500I and Penske took it testing. Initial reliability problems evaporated and at the Brickyard Penske’s drivers were able to overtake, as Illien recalls it, ‘like a hot knife through butter’. Only a year later, in one of motor racing’s cruel reversals of fortune, Penske failed even to qualify.


“Roger [Penske], Paul [Morgan] and I came up with the idea for the engine over a dinner in Phoenix in 1993, after that year’s Indy 500. Roger asked what we’d get out of it and I said 940bhp minimum. And I committed there and then to making the engine the same size for attachment to the existing chassis and gearbox, without having considered the design in any detail at that stage. I never regretted those promises but it was a bit tough to meet them. It wasn’t straightforward to get the length and other dimensions right. That’s why we chose a 72-degree bank angle rather than the 82 degrees used for the D engine of that year, for packaging reasons. It was the first pushrod engine I’d ever designed, and I had never designed an engine just for a single race before, and haven’t since.”


“I didn’t consider that we were exploiting a loophole. The Buick guys were continuously at USAC to relax the rules so that they could become competitive, and year after year they were given more freedom to develop their V6. But we were clear that the regulations would be changed as soon as we’d raced the engine and that we’d never get the opportunity again. We had one go at it, that was all. If what we were doing had leaked out there was a danger that some of the opposition would have tried to get it banned before the race, or others would have started a similar programme. That’s why speed was so essential, and secrecy. We went so far as putting odd descriptions on the drawings we sent to component suppliers so that it wasn’t obvious what we were designing, and Roger had a secret workshop in Pennsylvania where they only worked at night. We sent parts over by Concorde almost daily. Paul flew them down to Fairoaks airfield near Brooklands, and the shipper picked them up from there and took them straight over to Concorde at Heathrow.”


“We had reliability issues, mainly with the piston gudgeon pins – that was our biggest headache. The first test we did was at Nazareth in February, which is owned by Penske so there was security everywhere to make sure that nobody could get access. It came out later that Mario Andretti, who lived nearby, had heard the engine running and realised that it wasn’t a normal Indy engine. He was curious at the time but didn’t think any more about it! They had to clear snow from the track before we could run. Then at Michigan we again ran in sub-zero temperatures and again failed a gudgeon pin. We strengthened the pin, made it larger, did all sorts of things to fix the problem but it persisted. What we didn’t realise until later was that running at sub-zero temperatures the engine performance was significantly higher than it would be in normal temperatures at Indianapolis. The peak cylinder pressure was so high that the gudgeon pin couldn’t take it. It wasn’t a design issue – the pin just wasn’t made to tolerate that load. I think we probably also had a little bit of knock because of the cold weather and high cylinder filling. The first 500 miles we covered with that engine was on the Sunday before Indianapolis opened; we’d never managed 500 miles before that because of the gudgeon pin. The temperatures had risen, it was a nice spring day, and that fixed the issue. Early on we also had some valve spring failures but we fixed that with cam design and some other changes. Otherwise the valvetrain was absolutely bullet-proof.”


“I promised 940bhp, we achieved 1024bhp at 9800rpm and we could rev the engine to about 10,500rpm. I think we had about a 200bhp advantage over the nearest competitor. The engine’s torque was so high that initially the drivers would get wheelspin coming out of the pits and hit the rev limiter. That worried us, obviously – it’s not something you want. They also had to be very careful on a restart because if they nailed it in third gear they would also get wheelspin and risk spinning the car. It had a lot of torque, that engine. The drivers had to adapt but they got used to the additional power and torque pretty quickly. And then they wanted more, as always.”