Under the skin of the Porsche 956: May the downforce be with you

Porsche’s first ground-effect racer, the 956, set the endurance standard for much of the Eighties, and its influence resonates still. We take a look beneath its skin

Endurance racing has seen several ‘boom and bust’ cycles over the years, usually driven by regulation changes. High points included the glorious Ford versus Ferrari showdowns of the mid-1960s and the gladiatorial battles between the Porsche 917 and Ferrari 512 in 1970 and ’71.

For many enthusiasts, however, the greatest era in endurance racing began in 1982, with the introduction of FIA Group C regulations. These were designed to encourage manufacturers by allowing a wide range of engine types and rewarding fuel efficiency. Some mocked the regulations as an ’economy run’, but Group C was immensely successful in attracting manufacturer and privateer participation. The 1982 to 1991 period produced the fastest cars ever to race at Le Mans and, in addition to great racing, Group C inspired some of the most spectacular and visually stunning racing cars ever built.

The 1982 campaign also heralded the birth of the ultra-successful Porsche 956/962. The Porsche 956 won the World Sports Car Championship and the Le Mans 24 Hours for four consecutive years from 1982 to 1985 and the almost identical 962C won Le Mans in 1986 and 1987. The 956/962 saw off very serious challenges from Ford, Lancia, Jaguar, Sauber- Mercedes, Nissan, Mazda and many others, until finally outgunned by the Jaguar XJR-9 in 1988. Success was not confined to Europe, either: the 962 was highly successful in Japan and in the IMSA GTP series in the USA.

The 956 was the first ground-effect Porsche – it is estimated to have developed more than three times the downforce of its forebear the 917 – and the example featured here will form part of a themed Motor Sport display at Race Retro 2018, which runs from February 23-25 at Stoneleigh Park, Warks. It has been arranged to commemorate the 40th anniversary of Lotus’s F1 title success with the 79 – the first such success for a ground-effect car – and extends to embrace other period aero concepts (for tickets, see raceretro.com).

So, let’s take a detailed technical look at the Porsche 956/962 and try to determine just why it was so successful. We will then look at the Jaguars, from XJR-5 through XJR-12 to the IMSA GTP XJR-16, to establish how Coventry toppled Stuttgart

First, we need to understand the background.

The Group C regulations for 1982 were unusual in that they did not stipulate maximum engine capacity/swept volume. Instead, they imposed a fuel tank capacity of 100 litres and a maximum number of fuel stops: five were permitted for a 1000-kilometre race, 25 for a 24-hour race. The cars had to achieve an average of 1.67km/litre, or 4.71mpg. Minimum weight was 800kg dry, but most manufacturers did not get close to that. Maximum wheel/tyre width was limited to 16in. When these regulations were drafted, in late 1981, ground-effect aerodynamics dominated Formula 1. Mindful of this, the Group C regulations required a 1m x 800mm flat-floor section and stated that no component other than tyres and wheels could protrude below the plane of the flat floor. This still left plenty of room for ground-effect venturi diffuser tunnels, but meant that the skirts used in F1 to seal the venturi tunnels could not be used in Group C. Both these clauses were designed to reduce the designers’ ability to generate downforce. Maximum width was restricted to 2m, length 4.8m. Fuel was restricted to petrol, maximum 102RON.

These regulations allowed manufacturers to compete using their own choice of turbocharged, normally aspirated or rotary engines of any capacity – the challenge being to deliver the maximum possible performance from the limited fuel supply.

The Porsche design team, headed by Norbert Singer, plumped for a 2.65-litre twin- turbocharged flat six featuring four camshafts, four valves per cylinder, water-cooled cylinder heads and air-cooled barrels. It initially delivered 620-635bhp at 8200rpm using 1.2 bar boost and Bosch mechanical fuel injection. Drive was transmitted through Porsche’s own five-speed, full synchromesh transaxle. Final drive options were 9:36, 9:37 and 9:38.

When GSD RaceDyn analysed an early 956, we were surprised to discover that the car – a low-drag Langheck example in Le Mans configuration – used a ‘spool’, i e there was no differential but a solid connection between the rear wheels. This was used at Le Mans and at most other circuits.

The chassis was a sheet-aluminium monocoque with integrated light alloy roll cage and a triangulated tubular spaceframe supporting engine, transmission and rear suspension. The 962/962C variants had a steel roll cage and a 120mm longer chassis (and wheelbase), to move the driver’s feet behind the front axle centreline. These changes were made to meet IMSA GTP regulations and address safety concerns in Europe.

The front suspension was conventional, with dual unequal-length wishbones and interposed coil spring/damper units. Rear suspension was by lower wishbone and a tubular top rocker operating an angled coil spring/damper unit. This arrangement kept the springs and dampers out of the vital venturi diffuser tunnels. Rising-rate titanium springs, Bilstein gas dampers and conventional anti-roll bars with rotating blade adjustment were used at both front and rear. Steering was by rack and pinion with an 11:1 ratio. Surprisingly, power steering was not fitted.

The carbon fibre-reinforced glass fibre bodywork was developed in the wind tunnel, in two versions. The Langheck (long-tail) version, with a relatively small, low-mounted single-plane rear wing was used at Le Mans and some fast circuits. The short-tailed version had a much larger, higher mounted, heavily cambered single-plane rear wing and was used on slower, twistier tracks. Water and oil radiators and the air/water turbo intercooler were mounted in the sidepods, fed by large forward-facing ducts.

At the nose, the only openings were for brake ducts. The centre section of the nose was raised to feed air into the ground effect underbody, which featured a throat and short diffuser well forward. Near the front axle, a very unusual concave ‘hump’ fed in to a secondary throat, the mandatory flat-bottom section and two long, relatively deep diffuser tunnels that vented into the low-pressure area behind the car.

The flat six was not ideally suited to ground-effect cars, tending to restrict tunnel size and depth. However, the engine and gearbox were angled up slightly to allow wider, deeper tunnels. Regarding the unusual concave hump, GSD has not run wind tunnel tests or CFD, but we suspect that the forward throat helps to move the centre of pressure (aero balance) forward and the ‘hump’ helps to reduce pitch and ride height sensitivity. The Porsche 956 was the first sports-prototype effectively to use a ground-effect underbody. It has been said – reputedly by Porsche insiders – that Porsche did not really understand the ground-effect aerodynamics, but I suspect they’ll have had a pretty good understanding. They clearly understood the importance of low drag to achieve the necessary fuel efficiency and high downforce to deliver cornering speed and lap time. They also understood the need to minimise pitch and ride-height sensitivity for stability – and to make the driver’s job easier over a long stint.

When analysing the 956, we did not have access to wind tunnel data and could not carry out the usual airfield aerodynamic tests. Using basic theory, experience and published information, downforce for the ‘Langheck’ car in Le Mans configuration was estimated at 1050kg at 200mph, drag at 382kg at 200mph. With 635bhp, this gives a theoretical maximum speed of 218mph. The ratio of downforce to drag (L/D) – a measure of aerodynamic efficiency – is 2.75. We have not looked seriously at the high-downforce ‘sprint’ version, but we estimate that it would deliver more than twice the downforce of the Langheck, which equates to roughly 1280kg downforce and 320kg drag at 150mph (figures at 200mph are irrelevant for the high-downforce configuration – drag levels would not permit the car to reach that kind of speed). To put this into context, in sprint trim, downforce is six per cent higher than the best 1982 F1 car, drag 15 per cent less, power 19 per cent greater – seriously potent.

Moving on to mechanical design, chassis torsional stiffness is clearly adequate, but certainly not as stiff as the carbon-composite Jaguar XJR-6 and XJR-9. Anecdotal evidence from the 1980s tends to confirm this – IMSA 962s were sometimes referred to as ‘flexible flyers’. Both factory and customer teams fitted additional stiffening diaphragms.

Tyres were Dunlop radial slicks front and rear, on 16in rims. Goodyears and Yokohamas were also used in the USA and larger rims and tyres were used later, particularly at the rear.

Suspension geometry and kinematics are generally good. The roll axis is slightly higher than ideal at static ride height, causing jacking, but almost at ground level – which is ideal – when the suspension is compressed by downforce at speed. Pitch geometry gives appropriate anti-dive and anti-squat. Steering scrub radius is very small, therefore avoiding kickback over bumps. Rear roll camber correction is poor, but manageable provided roll is well contained.

Before looking at springs, it is important to recognise that downforce varies with the square of speed – i e downforce at 200mph is four times greater than downforce at 100mph. Ground-effect cars like the Porsche thus need massively stiff springing to hold them off the ground, but that makes them very stiff at low speeds, which tends to reduce grip in slow corners.

Porsche offered a range of rising-rate titanium springs to suit low- and high-downforce configurations. The rising-rate springs were initially soft, giving good adhesion in low-speed corners, but stiffened as the car was compressed by aerodynamic downforce. This particular car had suitable front springs, but the rear springs were too soft. A car in high-downforce configuration would have required significantly stiffer springs or a dual-rate system with secondary spring assisters.

Very stiff anti-roll bars are also required to keep the underbody aerodynamics stable.

On this 956, the front anti-roll bar is four to five times softer than we would expect and the rear bar two to three times stiffer. This is because of the solid spool differential.

The advantages of the ‘spool’ were simplicity, reliability, straight-line stability and straight-line traction. The disadvantage is understeer in slow- and medium-speed corners. Le Mans had very few slow corners in 1982 (before the Mulsanne chicanes), so overall the spool was probably an advantage. To combat the spool-induced understeer it was necessary to run the front bar soft, the rear bar stiff and add castor to help unload the inside rear tyre and let the car turn.

The Porsche 956/962 was not necessarily outstanding in one particular area, but was absolutely competent in all of them. A superbly engineered all-rounder, the car delivered sufficient power, good fuel consumption (adjustable by reducing revs and boost), good downforce, low drag and above all, excellent reliability. It was also relatively easy for customer teams to run.

These cars were very physical and demanding to drive near the limit. Huge g forces, heavy steering and a cramped, hot cockpit made for a challenging ride. Launched in early 1982, the 956 had a difficult first race at the Silverstone Six Hours. The fuel allocation was for 1000 kilometres, but six hours at Silverstone meant 1100. A 956 qualified on pole, but was forced into low-boost fuel conservation mode in the race, finishing second to a Group 6 Lancia. Thereafter, matters improved, the Porsche winning five major races in 1982, including Le Mans.

For 1983, Lancia built a pukka Group C car, the Lancia LC2/83 with a Dallara chassis and a 2.6-litre twin-turbocharged V8. The Lancia was generally as fast as the Porsche, but suffered tyre problems and niggling reliability issues. It won only one race in 1983 and the marque never would realise its full potential.

Despite increasing competition from several manufacturers, Porsche domination of Group C continued through 1983, ’84 and ’85. Its success was built on the all-round technical competence of the 956/962 – and numbers. In addition to the factory team, strength in depth came from highly professional customer teams including Joest, Kremer, Lloyd, Fitzpatrick and Trust. These privateers introduced significant improvements of their own, while the factory continued with major developments including electronic fuel injection and engine management (which improved both power and fuel efficiency), water-cooled barrels and improved chassis stiffness. Porsche also developed the ground-breaking PDK double-clutch, seamless-shift gearbox on the 956. It was later used regularly on some of the factory 962Cs.

The most significant threat to Porsche dominance was launched in September 1984, when Jaguar commissioned Tom Walkinshaw Racing to design, build and race Group C cars. The original intention was to use the successful Bob Tullius/Group 44 Jaguar XJR-5 IMSA car, but it soon became clear that this wouldn’t be competitive. TWR contracted top designer Tony Southgate to pen a completely new car. The Jaguar XJR-6 used a carbon-composite monocoque and a 6.2-litre 24-valve V12 based on the XJS. The engine initially delivered 650bhp at 7000rpm and drove through a March/TWR five-speed transaxle with a ramp-and-plate LSD. The car featured very advanced aerodynamics with exceptionally deep, wide diffuser tunnels, a large dual-plane rear wing and a substantial front splitter. Both front and rear suspension used dual unequal-length wishbones. Front spring/damper units were operated by a pushrod and bell-crank system. The novel front roll bar was a large-diameter carbon-fibre ‘horn’. The rear was a conventional interposed arrangement with the near vertical spring/damper tucked inside the wheel rim to keep the diffuser tunnels clear. The XJR-6 was somewhat overweight, but it did win the 1986 Silverstone 1000Kms.

The XJR-6 evolved into the XJR-8, which won three races and the WSC title in 1987. A further evolution, the XJR-9, stormed to six race wins – including Le Mans – in 1988. For 1989, the Metro 6R4-inspired Jaguar twin-turbo V6 powered the XJR-11, but this was not as successful as its forerunners.

The V12 – now delivering 730bhp at 7000rpm – returned for 1990, winning Le Mans again in the XJR-12, the final evolution of the XJR-6.

Jaguar decided that a car designed specifically for the IMSA GTP series was needed for 1991. The XJR-16 was based on the XJR 6-12 family, but used the turbocharged V6 from the XJR-11, a longer bellhousing and pushrod/bell-crank rear suspension. A conventional front anti-roll bar replaced the XJR-12’s carbon-fibre ‘horn’. Unlike Group C, the IMSA regulations did not include fuel restrictions, so turbo boost could be turned up to counter additional downforce – and drag. This was added in the form of a larger splitter and dive planes at the front. At the rear, the normal high-mounted dual-plane rear wing was augmented by a second low-mounted dual-plane wing which helped energise the diffusers.

GSD RaceDyn has analysed Jaguar XJR-5, XJR-12 and XJR-16 models. We will focus on the XJR-12 and XJR-16 to understand how the Jaguars finally derailed the Porsche 956/962 freight train. First, aerodynamics.

Aerodynamic tests on Richard Eyre’s XJR-16 at North Weald were not entirely satisfactory due to wind gusts, but gave a downforce of a staggering 2800kg at 200mph, the highest levels GSD has ever measured. The ‘drag coastdown’ tests showed much higher drag than expected, 720kg at 200mph. A low-downforce set-up was used for the 2014 Le Mans Classic. Although expected maximum was 196mph, before braking for the first Mulsanne chicane, the car pulled 208mph, meaning that drag was only 550kg at 200mph. As suspected, those gusts at North Weald had made a Horlicks of our drag measurement, and the higher top speed pushed ride height lower than planned, causing diffuser flow separation and porpoising. A small change to the spring set-up eased the problem and Richard finished fifth against very strong competition.

GSD has also analysed the Jaguar XJR-12 run by Moto Historics for Shaun Lynn, but has not yet been able to run aerodynamic tests. Based on theory, published data and experience with the XJR-16, we have estimated that at 200mph the XJR-12 delivers 1700kg downforce and 446kg drag in low-drag Le Mans configuration, 2700kg downforce and 632kg drag in high-downforce ‘sprint’ configuration.

Comparing the Jaguar XJR-12 with the Porsche 956/962, the Jaguar is about 90kg heavier but gives almost 15 per cent more power and huge torque – it still delivers 606bhp at 5500rpm, allowing rpm to be reduced to control fuel consumption. With both cars in Le Mans trim, we estimate that the Jaguar delivers 62 per cent more downforce for only 17 per cent more drag. The Jaguar chassis is significantly stiffer – and safer – than the Porsche. Suspension geometry and kinematics are both good. The Jaguar runs stiffer springs (because of the higher downforce) and a much stiffer front antiroll bar which gives better balance and traction.

By running simulations, we can compare the theoretical performance of the XJR-12 and 956/962 at the current Le Mans circuit. The Mulsanne chicanes were in place when the XJR-12 won in 1990, but other parts of the circuit have changed slightly, so we cannot compare directly with 1990 times. In order to run enough laps to win, fuel consumption of around 7.4 litres/lap was necessary in 1990.

For qualifying, Porsche could increase boost. Simulation suggests that the Porsche 962 should outqualify the Jaguar XJR-12 by some 1.2 to 1.8sec. In the race, Porsche would need to run reduced turbo boost and Jaguar reduced rpm. Simulation shows that with Porsche boost reduced and Jaguar rpm reduced so that both cars achieve the necessary consumption, the Jaguar XJR-12 is almost 5sec per lap faster than the Porsche.

Results from Le Mans 1990 bear this out. The best Porsche 962 qualified 3sec faster than the fastest XJR-12, but the XJR-12 won and the best 962 finished nine laps behind.

In conclusion, the Porsche 956/962 was superbly optimised to the Group C regulations and was dominant for an incredible five years. By 1988, however, the Jaguar could outperform the Porsche on most circuits, by virtue of a better chassis, more efficient aerodynamics and a simple, torquey, fuel-efficient engine.