Peugeot went Prototype racing when naturally aspirated 3.5 litre engines were imposed, replacing the freedom of earlier Group C years. In the 1980s, anything had gone within the context of a limit on fuel consumption (writes Ian Bamsey). In our previous issue (RET 134 September/October 2021) we saw how the all-conquering production-based Mercedes 5.0 litre V8 twin turbo had to be replaced by a pure race 3.5 litre V12 atmo for 1991, aside from at Le Mans.
Peugeot designed a clean-sheet-of-paper pure race 3.5 litre V10 atmo for its 905 Sports-Prototype. That car proved quick at Le Mans in 1991 against the handicapped, old-generation Group C cars but lacked reliability in what was its first full season of racing. In 1992, it won Le Mans and the FIA Sportscar World Championship titles for Drivers and Manufacturers. It then took first, second and third at Le Mans in 1993, by which time the marque was readying itself for Formula One.
For the 24 hours at Le Mans in 1993, Peugeot’s 11,500 rpm was adequate but in Formula One that year even Cosworth’s lone V8 had a peak power speed (PPS) of 13,000 rpm. The V10s – Renault, Ilmor, Yamaha, Mugen and Hart – had a PPS of at least 13,300 rpm. Renault had an estimated PPS of 13,900, just 100 rpm short of Ferrari’s V12 (the other V12, from Lamborghini, was less impressive in this respect).
Cylinder head showing locations for the PVRS cylinders
To enter Formula One in 1994, Peugeot engine designer Jean-Pierre Boudy created a brand new, higher-rpm derivative of the existing V10. This was again an all-aluminium production but the bank angle was reduced from the 80º of the previous Sports-Prototype unit to an even-fire 72º, smoothing the quest for higher rpm.
At the time, Boudy experimented with five rather than four valves per cylinder and with different bore sizes. The sportscar engine had a 91 mm bore with a 53.8 mm stroke for a displacement of 3499 cc, whereas for Formula One there was also the option of 93 mm with 51.4 mm stroke for 3492 cc.
In designing his Formula One engine, Boudy was assisted by Jean-Claude Fayard, a former Elf fuel engineer who in that role had worked closely with the Renault V10. At this stage, increasing crankshaft speed and fuel development were the keys to progress in Formula One engine performance. The 1994 Peugeot V10 was developed to produce an estimated 720 bhp at 13.500 rpm versus the estimated 770 bhp at 14,500 rpm of the rival Renault. Those estimates are based on a poll of Formula One engine insiders conducted that year by RET’s editor.
Peugeot had intended to produce its own Formula One car but that plan was scrapped in favour of an engine supply deal with McLaren. For the first two races of 1994, McLaren used the A4 version of the Peugeot V10, but thereafter it used the A6, which powered McLaren to a third place on its debut at Imola and a runner-up spot next time out at Monaco.
There were a further six podiums from the subsequent 24 Grands Prix starts but only one more above the third step. Dogged by engine unreliability, 1994 became the first season since 1980 that McLaren failed to win a single Grand Prix. McLaren switched to the Ilmor Mercedes V10 for 1995, when Peugeot supplied Jordan instead.
For 1995, the rules demanded a displacement reduction to 3.0 litres, prompting a reduction in stroke for the A10 version of 1995. Four valves per cylinder were now standardised (and five valves per cylinder had never been raced).
PVRS cylinder and ‘retainer’ (piston)
Under Boudy’s direction, Peugeot developed its 3.0 litre V10 well enough that by mid-1996 it was judged as good as any rival engine, running to 16,000 rpm. Alas, thereafter its relative performance started to decline. Its final guise was as the Asiatech V10 of 2001 and 2002, as described in RET 4 (Spring 2004).
Before joining Peugeot Sport in 1983 to head its competition engine programme, Boudy had worked at Renault on its Formula One project. As head of r&d he had been instrumental in a switch from wire coil valve return springs to a pneumatic valve return system (PVRS) – the first such in Grand Prix racing. A PVRS smoothed the passage to ever-higher crankshaft speeds. The empirical evidence of the early 1990s was that a PVRS was particularly beneficial as Formula One engine speeds climbed well beyond 13,000 rpm.
Where his sportscar Peugeot V10s had used wire coil valve return springs, Boudy specified a PVRS for his Formula One engines. The engine we are focusing on here is a 93 mm bore A6 version of the 1994 3.5 litre V10 as raced by McLaren. It has recently been refurbished by Simone Cisotto’s SC-8 race engine consultancy.
In essence, at its heart the 1994 Peugeot PVRS comprises a cylinder in which runs a piston with the valve stem attached to it, this ‘retainer’ compressing the air underneath it as the valve opens. The bucket-like pneumatic cylinder is partly recessed into the cylinder head, the valve stem passing up through a hole in its base, with a rubber seal to contain the air pressure. The retainer running within it has an inverted tulip shape, with the valve stem attached to it via a pair of collets and again with the use of a rubber seal to contain air pressure.
Between the aluminium retainer and the superfinished bore of the heat-treated steel pneumatic cylinder is a rubber quad ring (X-type) seal. Again designed to maintain air pressure, this seal has to permit freedom of movement of the retainer. In addition, there is an O-ring seal between the base of the pneumatic cylinder and the cylinder head.
Neither the retainer nor the pneumatic cylinder bore are coated. However, there is DLC on the inside of the inverted bucket tappet, which encapsulates the pneumatic cylinder where it projects above the head casting in which it is mounted. The tappet acts on the top of the valve stem through a steel shim that provides adjustment of tappet clearance. The tappet is also DLC-coated externally so that the surface against which the cam lobe operates has the best possible characteristics.
Key seals – for the retainer and the valve stem
Cisotto notes that, compared to a mechanical spring, an air spring is considered to be practically without mass, “therefore it has advantages from the vibrational point of view compared to the former system”. He notes that by changing the operating pressure of the PVRS it is possible to reduce friction and set the spring stiffness according to the engine revs, camshaft profile and maximum valve lift. “This allows more freedom in case of changes, and there is no need to change any component, as there is in the case of mechanical springs,” he says.
The Peugeot PVRS under the spotlight here nominally operates at an air pressure of 15-18 bar (according to cam profile and the rpm attained). Cisotto adds, “In the past, there were experiments even up to 25 bar on V10 engines that ran at 20,000 rpm”.
This PVRS is of the ‘open circuit’ type, meaning there are calibrated holes rather than control valves in the air supply system. Cisotto remarks that this offers a slight frictional advantage over a ‘closed-circuit’ system, as is commonly used these days. “A closed-circuit system has dynamic control valves to regulate the minimum pressure – refill of air – and the maximum pressure [any overpressure caused for example by oil entering in the cylinder housing],” he says.
He adds that each approach has its pros and cons. “An open circuit has a slight advantage in terms of frictional losses, especially at low revs. But it has the drawback of high air consumption compared to the closed-circuit approach.
“The closed circuit has the possibility to have a better regulation of the pressure than the open circuit – it has two special valves to regulate the minimum and maximum pressure, whereas the open circuit has only one orifice for each cylinder calibrated as necessary.
“To avoid excessive air consumption, with the closed-circuit approach it is also possible to recirculate pressurised air in case of overpressure; this is not practicable in the case of an open circuit. It is also possible to set up a system for the closed circuit to separate the oil from the air.”
A view of the cylinder, revealing its superfinished interior
Cisotto reports that in refurbishing the PVRS system for this 1994 Peugeot A6 an initial check was made of each cylinder head to ensure there was no porosity in the oil or water passages or the PVRS’ air supply channels. He notes that such porosity is a typical problem with an aged cylinder head, particularly in the case of Formula One engines such as this, which will have been highly stressed, in part because of very high coolant temperatures.
Cisotto adds that in the past he has even used helium to carry out porosity checks. The main tests are done with a penetrating dye that will reveal even very fine cracks. Then a leak test is applied to the channels that feed the PVRS.
“We plug the end of the calibrated hole, then pressurise the circuit to 15 bar,” he explains. “The head is then immersed in a special tank containing hot water – close to 100 C – to see if bubbles escape owing to any porosity. A special leakage machine measures the loss in the circuit over time.”
In this case, the PVRS cylinders were not replaced as they are made from steel and it was verified that the internal superfinishing was still good. Cisotto notes that more recent PVRSs use aluminium alloy cylinders, “with a special internal treatment, and that type must often be changed”.
The 1994 Peugeot V10 PVRS retainers are made from an aluminium alloy similar to 7075, and these were remanufactured. Also, all the seals were replaced. Cisotto says, “Once you have replaced all the seals you reassemble all the valves with their locks. Before doing that, a small amount of Vaseline oil [which has neutral characteristics and does not contain metals] is injected using a precision syringe. This is so as not to dry the rubber seals.”
Once everything had been reassembled, a second leak test was performed to see how long the system took to discharge, pressurising it at 15 bar. “Of course, if none of the pneumatic seals leak and the engine is stationary rather than running, the system will discharge only after several days,” remarks Cisotto. “A visual test of any leakage between the pneumatic cylinders and their retainers is also made, a test carried out by spraying with a few drops of low-viscosity engine oil and looking for the presence of any micro-bubbles.”
He explains that the air used to feed the pneumatic springs “is micro-filtered synthetic air, very similar to that used in hospitals. The car has a small reservoir to restore pneumatic air to the system during the race in case of leakage. This reservoir is charged to around 300 bar, which is pressure-regulated to the required static operating pressure.”
During engine operation, at maximum valve lift the operating pressure increases from the nominal 15-18 bar static pressure up to about 46-60 bar. Cisotto adds that normally the air consumption of the PVRS should not exceed 5 g/minute: “a value that can be acquired using a special pressure sensor on the air channel of the PVRS.
“Air consumption is caused by leakages that might occur at a seal between the valve stem and the retainer, at the retainer’s quad ring, through the valve guide seal and from any leakage from the connections or any porosity of the head.”
Another view of the retainer
When it comes to maintenance of a PVRS, Cisotto observes, “It seems strange but compared to the classic system of mechanical springs this type of system does not require difficult interventions. However, it is very difficult to find spare parts that are made to measure for each configuration and engine.
“In refurbishing a system such as this it is essential to replace all the seals, especially the quad ring between the retainer and the pneumatic cylinder, and to check the internal finish of each cylinder. In the case of a closed-circuit system it would be important to check the settings of minimum and maximum valves. Another important aspect is to install the gaskets in an aseptic environment to avoid soiling the seals [even with dust]; we always use specially treated air.”
THE SC-8 CONSULTANCY
Aside from the 1994 Peugeot V10 considered here, Simone Cisotto’s SC-8 race engine consultancy has worked on a number of other classic Formula One engines, including Cosworth DFV and DFR V8s and Ferrari V12s. Founded in 2010, the company has undertaken a variety of other historic race engine projects, ranging from a 1.0 litre Hillman Imp I4 to an
8.3 litre Viper V10.
Based about 50 km south of Milan, SC-8’s portfolio also embraces Lamborghini, Ferrari, BMW, Porsche, Jaguar, Bugatti, Chevrolet and Ford engines. Cisotto notes that he has worked a lot with NASCAR, GT40 and Cobra engines, along with many other notable makes, together with motorcycle race engines and even rebuilding piston aero engines.
He says, “Our main objective is tending to all the various types of IC engines, from pre-World War One engines to the most modern Formula One units, and from the smallest motorcycle engines to those with 12 or more cylinders.
“Through our work, we can provide numerous services ranging from the restoration of the most vintage engines to technical support for developing race engines destined for global competition.”