Cosworth’s XH LMP1 Sportscar Engine
© Alan Lis

Since 2004 Cosworth, the Northampton, UK-based race engine manufacturer, has been developing a turbocharged V8 power unit aimed the LMP1 sportscar racing market. Alan Lis spoke with Bruce Wood, Cosworth’s American racing chief engineer about the project.

What was the motivation for producing a Cosworth sportscar racing engine?
“I have always wanted to do Le Mans and Sebring. Cosworth has never been very well represented in sportscar racing. In the 1970s there were teams using de-tuned Formula One DFVs and in the 1980s there were lots of DFLs and DFVs around but none were that successful. It’s a discipline where Cosworth hasn’t really been represented and should be, but under the current climate it’s a difficult market to get into.

“For the last few years I had personally felt that Cosworth should be involved in sports prototype racing. It was something that we had discussed as a company and we decided that it if we wanted to be in it, we probably needed to have a product to sell rather than to try and persuade someone to put up the money to develop a product. That’s seems to be the way that sportscar racing works at the customer level. At the end of 2004 it became clear that our IRL programme with Chevrolet would be finishing after the 2005 season. The Champcar programme would continue, but from an engineering point of view that programme doesn’t require a lot of work. So it was a desire to be back in sportscar and the coinciding fact that we were losing one of our major programmes that led us to seriously look at doing this engine.

“We had also seen a window of opportunity demonstrated by race results in 2005, which suggested that whilst the Audi R8 had been a fantastic car, it had begun to show its age. It was also clear that it was not going to be modified to fit the new ACO regulations. The Pescarolo Judd won the 2005 LMES title and showed that the Audi, with the restrictions placed on it, wasn’t necessarily the ideal tool to win race that it had been in previous years. At that point, when we started looking at the design of a sportscar engine at the end of 2004, there had been no word from Audi as to what they would do in the future. We saw an opportunity for a good engine packaged in a Lola, Courage or other customer chassis and felt that the availability of customer engines had not kept up with the availability of chassis. So we decided to go ahead with and produce what we thought would be a top line P1 engine. At the same time we of course had to recognise that in sportscar racing, outside of the manufacturer teams, there was not tens of millions of dollars floating around, so we were going to have to do it from our own budget.

“That meant that it wasn’t going to be sensible to start completely from scratch. So we looked at what was available to us and the IRL engine seemed to be the perfect base for a sportscar racing engine. They need to be durable rather than lightweight flyers and the IRL engine had been built to rules specifying a minimum weight limit and certain minimum dimensions. It wasn’t going to be a case of trying to get an F1 engine to run for 24 hours (as Cosworth did with its DFV-derivative sportscar racing engines in the 1970s and 1980s).

Why choose the IRL engine (above) as a base for a sportscar power unit?
“The IRL engine is a substantial beast and would fit pretty well in the chassis that are available. So we took that as the basis for the XH and bored it out from 3-litres to 3.6-litres and added twin turbochargers. From the outset our intention was to take on Audi and so, recognising that there was a potential fuel economy benefit to be had from direct injection, we decided that from the start of the project the XH would be available with conventional port injection or direct injection. Rather than trying to reinvent the wheel by developing our own system, we chose to partner with Bosch who had provided the direct injection system that Audi used: in 2005 they made that system available on a commercial basis for the first time. Obviously the injector nozzle configuration and the spray arrangement are tailored to specific engine designs, but the fundamental mechanics of the system are the same as those for the Audi FSI system.

“The direction injection system was a very big investment for Cosworth, that technology doesn’t come cheap. It isn’t just a case of asking Bosch to send a system, you sign a deal with them where they provide the engineering support. The cost is why we offer the engine in two versions, with DI if you can afford it, with port injection if you can’t.

In the IRL, engines are rev-limited to 10,500rpm, with a turbocharged sportscar power unit you were presumably running slower than that?
“Yes, the LMP1 rules have an air restrictor so there is no point in running at high revs. From memory the engine became choked at around 6500-6800rpm and at that point the power curve flattens. We ran on to 7500rpm to give the driver somewhere to go before gear shifting, and set the limit there. Beyond 7500 the power begins to fall away. In endurance racing the slower you can run the engine the better.

What changes were made to the IRL engine to turn it into a sportscar racing engine?
“Having been designed to run flat out for 500 miles in IRL races we knew that the XG was a durable engine. The experience we gained from the XF-E Champcar programme, where we turned that engine from a 550-mile sprint engine to a 1200-mile endurance spec. engine, was also very useful. We knew which bits to look at within the auxiliaries and drive systems to make them last longer.

“The 3.6 litre displacement of the XH was achieved by increasing the bore and stroke of the XG, by the fitting of new liners and the crankshaft from the 3.5-litre version of the IRL engine.

“Part of the deal with Bosch for use of their DI system is that they specify the injector location and design bespoke versions of the injector to suit a particular engine. It’s definitely better to retain the ignition source in the centre of the combustion chamber, between the valves, and introduce the fuel underneath the inlet valves. That meant designing all new cylinder heads to fit the injectors in, but not a fundamental tear up like having to move the camshafts around.”

“From talking to a lot of sportscar teams and manufacturers it became apparent that we needed to move our alternator. On the IRL version of the engine it was down on left side, which is quite a common location for a single seater. In endurance racing, the teams want to be able to change the alternator at a pit stop if it’s failed, so we did some engineering to move it into the centre of the of the vee, and make it a quick release part.

How far was the sportscar engine developed?
“We took it to the point where, from all the information we could glean, it was a bit better than the Audi V8. We measured 617 bhp and 730Nm of torque, while from our information the Audi engine made around 610bhp and 700Nm.

“Those figures led us to believe that we had an engine that was a bit better than the benchmark LMP1 engine, in terms of performance, at that time. With regard to durability, we ran Cosworth’s first ever 24-hour endurance test using data from a qualifying lap at Le Mans, and ran a full race simulation at that pace on the dyno. We ran the equivalent of 14-15 lap stints, what we were expecting from a tank of fuel: at the end of each stint we simulated an in-lap, after which we stopped the engine for around a minute to simulate refuelling, before restarting it and doing another stint.

“We stopped the test after 24 hours, ticked that box on the engine test sheet and then ran it for another six hours to make sure that we had some ‘headroom’. We didn’t have a single problem throughout the test, which was very rewarding. That engine was running on port injection, so we felt that we had an engine that was probably better than any out there in terms of performance and durability.

Did the direct injection version of the engine get onto the dyno?
“A limited amount of dyno testing was carried out with the direct injection system. We probably gave it only half a dozen power tests to feel our way around before we decided to stop the programme. I’m entirely confident that it would work very well, although it would probably require some fine tuning of the combustion chamber profile.

“Performance and endurance development on the XH was initially done with port injection because that’s what we knew. Direct injection was a complete departure for Cosworth and the cost of it was going to have to be passed on to the customer, and at that point we weren’t at all sure that many would be able to afford it.

“By that time we knew we’d got the performance from the port injection version. We felt that there would be some economy to come from using direct injection, but we decided that we really needed to wait and see if we were going to get a customer before we invested any more money: that’s where the project stands at the moment. We approached pretty much every sportscar racing team you could care to mention, we had several people come and visit and have the engine demonstrated to them, and there was a lot of interest, but we are still at the same point of waiting for a customer. There was another flurry of interest at the beginning of 2006, but that hasn’t gone anywhere either.

“We did also consider an LMP2 engine. The way the rules are for that class it would be a naturally aspirated 3.4-litre engine and the IRL engine would lend itself pretty well to that use too. What made us decide against it was that we knew we were going to be at the expensive end of the customer scale, so if the teams couldn’t afford a P1 Cosworth engine they certainly would be able to afford us in P2. If the P1 engine had gone ahead and met with good take up and success, it was in the back of our minds to do a P2 engine.

Would a customer buy engines outright?
“The deal for the XH would be the same as all of our customer programmes, a lease arrangement with full technical support from Cosworth. We approached several manufacturers to see if anyone was interested in badging the engine and turning it into a semi-works deal. Obviously we could make it available to customers at a more attractive price if that happened. Again we had several conversations in that direction, but none that came to anything.

Where does the project stand now?
“After the dyno tests with the direct injection system we put the project on hold: we’d spent a lot of money on it and during the months of development we’d been looking for customers, but circumstances meant that we were not able find any. Within that period Porsche launched its LMP2 car, Peugeot committed to a diesel Le Mans programme and Audi announced that they would have new car in 2006, which turned out to be the R10. A lot of the teams that might have been potential customers were either holding out for works deals, which is perfectly reasonable, or perceived that with Audi, Peugeot and Porsche coming back, it was going to be difficult for customer teams. It seemed to us that they were thinking that without a works deal they weren’t going to win anyway, so they might as well go for the cheapest customer engine deal. We made no bones about the fact that the XH was the most expensive customer engine. At Cosworth we have a big overhead and the reason for that is that we have things like a metallurgy department with six staff, and equipment like scanning electron microscopes and all the other things you need to build Formula One engines, as well as engines for other classes, and that all has to be paid for.

“Although it comes with a price tag attached that is probably more than any of the other customer engines available, we feel that we were offering the best product available in terms of performance.”