To many rally fans, Peugeot’s 205 T16 and Lancia’s Delta S4 stand as the greatest rally cars of all time. But which was better? Martin Sharp takes a technical overview of both to find the answer
The pall of smoke over the Col d’Ominanda ballooned up into the dark cloud that would end the active life of the most exciting rally cars ever.
Just after 3pm on May 2, 1986, the charred, melted remains of Henri Toivonen’s Lancia Delta S4 were strewn around a downhill bank on the outside of a fast, tightening left-hander on the Col, 5km from Corte, the start of the 27km-long SS18 of the Tour of Corsica. Henri had been fastest on no less than 12 of the previous twisty Tarmac stages, and had a commanding lead. Nobody will ever know why his Delta left the road. Flipping in the air, the Lancia hit trees, fell 15ft and burst into a violent conflagration that obliterated any opportunity of conclusive investigation. Toivonen and co-driver Sergio Cresto perished immediately.
At the rally HQ in Ajaccio, unspoken glances expressed a fear of the worst. The Delta was built to conform to the Group B regulations; that meant that volatile ‘rocket’ fuel was allowed – and, for the extra-long Corsican stages, it was placed in tanks under the squabs of the crew’s seats. And these were not even safety tanks. Also, many of the car’s components, transmission casings, wheels and others, were cast using magnesium-based alloys.
The governing body FISA sensed a portent. It abandoned its cherished rules stability, and decreed that for the rest of 1986 no more ‘evolution’ homologations in Group B or A would be allowed.
More than that, the formula itself was to become extinct. FISA’s president Jean-Marie Balestre announced that, for 1987, larger capacity GpB rally cars were banned outright. He also put a stop, right there and then in Corsica, to its planned successor, Group S. It was the end of a magnificent age.
The sight and sound of GpB cars as they developed from 1982 had captured the imagination of many. Here was a category offering lightweight (890kg minimum), purpose-built, high-powered cars with unrestricted forced induction (600bhp was possible) and comparatively unsophisticated four-wheel drive systems. They were super-fast and difficult to drive, demanding dramatic sideways progress to screw the speed out of them.
It was an exciting time, not just for the drivers, but also for the engineers, who worked to rules offering comparative carte blanche. By 1986, the ultimate expression of this freedom was to be found in Peugeot-Talbot Sport’s awesome second evolution [E2] 205 Turbo 16, battling head to head against Lancia’s equally impressive Delta S4. They are remembered as the most extreme rally cars ever produced. But which was better?
The Peugeot’s successes are greater in number, primarily because its lifespan was that much longer than the S4’s. Peugeot’s marketing masterstroke in February 1983 was to launch the Gpb Turbo 16 rally car on the same day as its new 205 hatchback road car. This meant that, by the ’84 season, the T16 was competing, Ari Vatanen winning with it on three occasions.
The following year was its first serious world championship attempt, for which Peugeot developed the first evolution of its contender. But just as that car was configured, FISA altered its homologation rules. Now, only 20 of the 200 minimum road cars needed to be changed to qualify for the modifications. So to remain competitive Peugeot had to go back and start again: the T16 E2 would have to be substantially different. Development of this car began in July 1984, and it would be ready in time for the Tour of Corsica in ’85.
It is a commonly held belief that Lancia’s Delta S4 was a hasty reaction to the dominance of four-wheel drive rally cars. Sure, Lancia soon learned that its two-wheel-drive 037 Rally would be outdated once four-wheel-drive rally cars began to properly exploit their potential, but the Latin reaction was far from hasty. The Delta S4 project began in April 1983 – just weeks after the T16 appeared – and its first tests in ’84 showed that this hi-tech machine was already 4sec faster than the 037 around the 2.2km La Mandria test track. But production and homologation problems were to delay its launch, and it did not make its debut until the 1985 RAC Rally, where Toivonen and Markku Alén finished one-two.
The S4’s engine was all-new. With an all-alloy crankcase and a twin-cam 16-valve cylinder head, the unit was located longitudinally and slightly to the right in the rear-mid of the car, ahead of the rear axle and driving forward to a dog-engagement gearbox and viscous coupling centre differential between the driver and co-driver. Its most interesting feature, however, was the combination of a crankshaft-driven mechanical supercharger to enhance low-rev throttle response and torque, and an exhaust-driven turbocharger to provide boost at high revs. The supercharger operated up to 5000rpm, when a bypass valve short-cut the mechanical blower and the turbo kicked in.
Early on, the development of the S4 engine unit concentrated on its response, resulting in the fitting of a larger capacity blower that better matched the large KKK turbo to smooth the transition between the compressor devices. Under 2.4 bar boost the first engines produced 420bhp at 8200rpm and 296lb ft of torque at 5000rpm. Engines had been tested at 2.9 bar without failure.
The T16’s engine was also an all-aluminium twin-cam 16-valve, arranged transversely behind the crew along with its synchromesh gearbox. The motor was tilted back 20 degrees and biased heavily to the right-hand side of the car. This enabled comparatively compact exterior dimensions to follow the form of the 205 road car.
The benefits of this layout were threefold. Firstly, it released space above the gearbox for ancillaries. It also enabled the engineers comparatively easy access to the belts and equipment on the engine’s front face in the car’s right-hand flank; and also to the torque-splitting equipment and viscous coupling centre differential on the end of the gearbox in the other flank.
The E2 version’s engine was ready before the chassis work began. Porting and manifolding work enabled the E1’s KKK turbocharger to be replaced by a much bigger Garrett unit. A minimum of 435bhp was available from the 1775cc engine at 2.5 bar boost pressure, but 2.8 bar was possible, yielding around 500bhp. Torque was improved and available over a wider band; at 2.6 bar boost, max torque was 383lb ft.
From birth, the T16 engine incorporated a system first seen in Renault’s turbo Formula One engines, designed by Jean-Pierre Boudy, who had moved to Peugeot-Talbot Sport. His DPV (Dispositif Pre-Rotation Variable) system was essentially an early form of anti-lag, providing a variable-section hole on the inlet side of the turbocharger which, in the E1, was linked mechanically to the throttle pedal. In the E2 it operated electronically, the hole size depending on throttle angle and charge gas temperature as it entered the cylinder head.
For GpB cars the supercharging equivalence formula was 1.4:1 [today it is 1.7:1]; both the T16 and S4 equated to under 2500cc, thereby enabling these 500bhp machines to compete at a minimum weight of 890kg. They were the only two GpB cars allowed to compete at that weight; all their rivals were calculated to over 2.5 litres and thus had to be heavier.
The two cars’ engines were veritable powerhouses, using equally innovative but widely differing methods of reducing turbo lag and improving low rpm performance. If one ascribes to the ‘simpler is better’ principle, then perhaps Peugeot’s option was the wiser, but FISA had not begun regulating fuel by then: octane rating was a hush-hush subject in service areas, and the evident power benefits from ‘rocket fuel’ could perhaps be more easily harnessed by the more complex Lancia unit. Suffice to say that standing by a ticking-over S4 for more than a couple of minutes would bring tears to the eyes.
Peugeot’s top engineer Jean-Claude Vaucard freely admits that the T16 E1 would have been different if they had known of FISA’s homologation relaxation.
The T16 E2 that emerged midway through 1985 gained a lighter and more accessible structure with increased torsional rigidity through the replacement of its monocoque extensions from the passenger shell with a tubular frame. This brought the T16’s basic structure in line with that of the S4. However, with its transverse engine hanging over the right-hand rear, the Peugeot had a significant weight bias to that corner. Gearbox, engine and rear-axle oil coolers went to the rear-left of the E2 car, and engine coolant radiator and air/water intercooler were relocated to the front of the car. Even so, a dry E2 car was 50kg heavier on the right-hand side, but with fuel tanks under each seat, the left-hand one carrying 55 litres most of the time, the realistic right-hand weight bias was only 20kg.
The laden front-to-rear weight balance of the 205 T16 was two percentage points closer to 50/50 than that of the Delta, but the S4’s much more equal right-to-left balance must carry the structure/balance argument in the Lancia’s favour.
An additional problem derived from the transverse engine was that the T16 did not fly properly. As the car became airborne, naturally the rear wheels would be the last to leave terra firma; then, as the engine, driveshafts and wheels speeded up, the car’s nose would rise dramatically. So the driver would lift the throttle slightly, rpm dropped and the car rotated around its centre of gravity, aiming it into a nosedive.
The Delta was better. With its longitudinal powertrain producing a twisting force in that direction, a more equal right-to-left weight balance, and the transverse torque produced by its wheels and driveshafts to balance the longitudinal torque, it was significantly more manageable.
Another weakness of the T16 was the heat produced by its upper wishbone-mounted, coil-over dampers, particularly on its heavier right-rear corner. A water-cooled damper system, activated by the co-driver’s foot, was introduced for the E2. A dashboard read-out provided the temperature information from a sensor on the bottom of the ‘problem’ shock absorber: once that became critical, he hit the pedal to energise the electric pump, which squirted the damper exteriors. With more power and increased down-force, the E2’s springs were made 50 percent stiffer to counteract pitch on braking and acceleration.
The 205 T16 was modified for Argentina in August 1986, when the amount of torque delivered to the front axle was increased from 34 to 45 percent. The result was that the magnesium front axle housing split straight down the middle. The problem was solved come the 1000 Lakes a month later when an aluminium front axle housing proved unburstable. Inlet camshaft-driven hydraulic power steering was then introduced at San Remo, reducing the increased steering effort required by the extra grunt going through the front axle.
The lack of development in ’86, induced by FISA’s decision to can GpB, brought many T16 transmission components to the limits of their durability. The team was reluctant to dial in an extra 0.2 bar boost to obtain some 490bhp when 45 percent of that was directed at the front axle. A further threat to reliability had been the introduction of a sixth gear on the Safari in March. This remained the rally specification to the end of the car’s life and enabled a shorter final drive ratio, thereby increasing torque multiplication in the lower gears.
Going from 15-inch to 16-inch cast-magnesium wheels provided more lateral acceleration and allowed for bigger brakes. The S4’s adjustable double wishbones in each corner offered long wheel travel, but with twin dampers (and a single central coil spring) on each side at the rear.
However, Lancia chief development engineer Giorgio Pianta and his team faced a steep learning curve with the S4’s four-wheel drive system, initially experimenting with a viscous coupling to replace the ZF mechanical slip limiter in the front axle. Poor straight-line directional stability on gravel also pointed to compliant plastic inserts in the suspension’s spherical joints, although differential compatibility was also not discounted.
Before the Argentinian event in 1986, Lancia found a major handling improvement for the S4 by using 195mm-wide wheels instead of 185s. The same 180mm tread Pirelli tyres were fitted front and rear, with 10mm shaved off front tread width. This improvement was “fantastic” because, according to Pianta, it meant more than one man could master it: “When the Delta first arrived, Henri Toivonen knew the car before any of the other drivers. When he drove this car it was immediately his. For Henri it was as if he was born in this car. Other drivers coming from the Lancia Rally  to the S4 needed more time to learn the car. They were not so fast because the car was difficult to drive. Now the car’s limits are the same, but more people are capable of arriving at its limit.”
If Balestre’s knee had not made that fateful jerk in May ’86, rally car engineering would have progressed much earlier than it did toward even more exciting and sophisticated cars. This was Vaucard talking about the ever-present turn-in problem of the T16 in 1968: “To get over the problem we were looking to work with a new generation of limited-slip differential – a ZF-type controlled by something like electronics. We were developing the system for both the axles and the centre unit, because the centre differential also badly influences the handling.” Plainly, if Vaucard had been allowed to introduce this development in that season, he would have been at least a decade ahead of similar developments.
A second evolution of Delta S4 never happened, of course. But a hint of what might have been was given in December 1986, when Fiat exhibited an ‘ECV concept car’ (Experimental Composite Vehicle). This was, in effect, just a very radical E2 version of the S4. We got no inkling at all of the E3 T16, apart from Vaucard confirming that it would most likely have retained a transverse engine – but with its crankshaft rotating in the opposite direction. GpS would have been a serious business.
Trying to judge objectively between two awe-inspiring cars, whose dramatic progress on stages stirred strong emotions, is nigh on impossible. The plain results make things little clearer.
In their only head-to-head championship, 1986, the first evolution S4 acquitted itself well against the developed second-evolution T16 (four wins to the 205’s six). Peugeot won the manufacturers’ crown, but the drivers’ title was only awarded to Juha Kankkunen, instead of Lancia’s Alén, in a French courtroom.
By 1986, the 205 T16 was a highly developed expression of immense Gallic attention to detail. If it was flawed it was because of the effect its transverse engine and gearbox assembly had on creating unwelcome turning torque on the car, producing those seriously nose-heavy landings if a driver backed off while jumping. Hardly a confidence aid for the pilot; indeed, many top drivers admitted that the characteristic scared them.
Was there a flaw in the high-tech, yet minimalist Latin-style design of the Delta S4?
Claudio Lombardi, the Lancia engineer responsible for its layout, had dialled existing four-wheel-drive knowledge into its chassis but, certainly in the early part of the ’86 season, the team struggled to understand the intricacies of this system in a rally car.
By the 1000 Lakes in August the GpB game was nearly over. Although improvements had been made, Pianta remained perplexed; every traditional solution he’d tried had failed to get the S4 to turn in neatly and accurately, which the T16 achieved reasonably well. Surprisingly, Vaucard and Pianta had never met so, late in the season, I asked Jean-Claude whether a chat with Giorgio might be useful. He echoed my thoughts: “Of course, it would be interesting; there will be no secrets now.”
Talking later to an exasperated Pianta I mentioned that Vaucard had nearly cracked the turn-in conundrum (predominantly using viscous coupling differentials), and would he like me to arrange a meeting between them? Giorgio said simply: “Interesting before, but there is no point now.”
One memory which will never leave me is of clambering up a Finnish tree overhanging a right-hander after a flat-out straight as Stig Blomqvist’s T16 E2 headed for ‘my’ bank without lifting. The Peugeot took the perfect, oversteering line through that corner faster than seemed physically possible, still flat, Stig feathering its comportment with his left foot on the brake. It was incredible to watch and experience. In contrast, in the same place, the S4s looked to be a bit of a handful.
Later, in 1987, I quizzed top rally engineers on their optimum layout for a GpS rally car, putting all the findings to the late Len Bailey (designer of the GT 40), evaluating the lot and coming up with a theoretical car to sketch out. Its layout was almost pure S4. It looked a lot like an 037.
The final analysis? It’s subjective, but the Lancia Delta S4’s concept was better; the Peugeot 205 T16 E2’s spectacle simply the best.