X-Ray spec: Gulf Mirage M6
A brand-new endurance engine and a special low-drag body looked like the perfect mix for Mirage's Le Mans effort. Wrong, as Keith Howard finds
It all seemed such a good idea initially. When Stuart Turner (Ford's motorsport boss in the UK), GT40 designer Len Bailey and engine maker Harry Weslake hatched a plan after the British GP in 1970 for Weslake to build a Ford-badged V12 for sportscar racing, it looked as if everyone would benefit.
Ford was being pestered to supply the Cosworth DFV, but as its designer Keith Duckworth complained in characteristically forthright manner, it had been designed for Formula One, not the significantly different conditions of sportscar racing. The Ford-Weslake unit would satisfy that requirement while also understudying the DFV should its runaway success in F1 begin to show signs of faltering.
Harry Weslake, meanwhile, who had nurtured an ambition to return to F1 ever since supplying the V12 for Dan Gurney's Eagle, would get to build an engine which, if good enough, might fulfil his dream.
Ford's Walter Hayes struck a deal for the new engine similar to that for the DFV. Just as Lotus initially had exclusive use of Duckworth's design, so the Gulf racing team, which had done such a good job for Ford racing the GT40 with Weslake heads, would get first crack at Rye's new V12 to fit into its Mirage M6. Gulf already had access to the DFV, so it built two versions of the M6 chassis, one for each engine. Essentially they were identical except for a small increase in wheelbase to accommodate the slightly longer, slightly heavier V12.
Initial testing was promising. Running at Goodwood in 1972 with engine 001, the V12 car's lap time was within half a second of the DFV car. But it soon became apparent that Dan Gurney's warning to John Wyer that Weslake couldn't build two engines the same was a prescient one. Engines 002 and 003 just weren't as good, and as this became clear the relationship between Gulf and Ford on the one hand, and Weslake on the other began to sour.
The Weslake-engined M6 never ran in competition. In private pre-race testing at Daytona in 1973, fitted with engine 002, it was 2.4 sec a lap slower than the DFV car, albeit 5mph faster in top speed. All three cars — one V12 and two V8s — were sidelined during the test by seizures in their Hewland transaxles, and only the Cosworth cars were entered for the race.
For the critical Le Mans test in April, the Weslake car was fitted with a low-drag coupé body hurriedly designed by Bailey specifically for the 24-hour race, where the high drag coefficient of the open car would be a major disadvantage along the four-mile Mulsanne Straight. Although this closed body helped the V12 match the Cosworth car's speed down Mulsanne, engine 003 was so down on power that Derek Bell's best lap time was almost 16sec shy of Howden Ganley's with the DFV.
Gulf Research's managing director John Horsman made the inevitable decision not to run the Weslake car at Le Mans, and that was effectively the end of the Ford-Weslake V12 project. Following the oil crisis of that year, Ford cut back drastically on its motorsport budget for 1974 and, to Turner's relief, the Weslake engine was one of the casualties. Gulf refused to accept the remaining three of the six engines Weslake had built, as a result of which it found itself heading for the High Court. Gulf settled for an undisclosed amount.
Although, ironically, the DFV was to prove a weak point in the Mirage M6 at Le Mans in 1973, Horsman — who talks overleaf about aspects of the car other than the ill-fated V12 (shown in the cutaway) — is adamant that it was a vital stepping stone on the way to Gulf's triumph in 1975. That year, the Weslake episode long forgotten, the Mirage GR8s, powered by DFVs, finished first and third.
Going with the flow
Echoing Keith Duckworth's warning that the DFV wasn't designed for sportscar racing, the Cosworth engine suffered cooling problems in the M6s at Le Mans. "The Cosworth had a single water pump which fed both sides of the block, the left bank via a cast-in transfer passage. After a number of hours running, cylinders 5, 6, 7 and 8 seized in one car as a result of running at over 10,000rpm all down Mulsanne. There was insufficient water flow from the single pump. In the second car Schuppan, who was using a higher gear to try to save the engine, ran wide onto the marbles and rolled over the Armco. For years he thought he'd let us down but when we dismantled the engine we knew it wouldn't have lasted anyway. For 1974 we overcame the problem by welding a second water pump on the other side. That solved the problem, and Cosworth then went to a two-pump set-up too."
"Len Bailey's chassis was aluminium with steel cross-members" says John Horsman. "The front one, with all the holes visible, had vertical members either side fixed to the aluminium tub. The rear bulkhead had the engine and gearbox bolted to it, although there was a tubular steel frame as additional engine support. A similar frame on the GR8 saved us at Le Mans '75, when an engine bearer broke on one side on the winning Ickx/Bell car and on both sides on the Schuppan/Jaussaud car. The engine, transmission and suspension were only held on by the frame --- without it the cars would have broken in half."
"The drivers loved the handling of the M6 on the whole. The only complaint they had was the high speed turn-in wasn't precise enough --- there was a slight delay before the car responded. I wanted to go to a bigger wheel and tyre on the front to give a larger contact patch, but Firestone advised against it, so we never did it. In hindsight I kick myself for not having tried it. We should have seen if a 14- or even a 15-inch tyre would have made a difference, instead of trying to get more downforce on the front by extending the nose. That helped, but the underside then rubbed on the track.
"The open body version was really a short circuit car. It worked fairly well but had a drag coefficient of about 0.53. That's why we developed the low-drag coupé, although it was so late we should have kept it for 1974. I can't find the drag figure for it but it was about the same as a GR8, which was 0.35. In retrospect it was unfortunate that I chose to put the coupé body on the Weslake chassis. Had I put it on the Cosworth, we'd have had the full benefit of the Le Mans trial. In those days, with the long straight without chicanes, Le mans was so special that it was too risky to go there with an untried body. So we ran the two open cars, which I still think was the right decision."
"There was a time in 1972 when the Ferrari engineers --- with no official backing --- talked to us about putting a Ferrari engine in a Mirage. They said their car didn't handle very well and their drivers had seen that the Mirage did. We were very interested but then they got their car sorted out and it went no further. That was a 12-cylinder we could have lived with. If we'd had that, or a Weslake that worked, we could have spent our development time much more profitably instead of fighting breakages caused by the DFV's vibration."
In Formula One at this time, designers were just starting to use custom-designed aerofoils, but the M6's rear wing used a standard aircraft section. "It was the NACA 4412" recalls John Horsman. "John Wyer had a friend in the aeronautic industry who recommended that wing. First of all, on the M6 it was all-aluminium in construction. On the GR7 for 1974 we used a hot wire to cut the section out of polystyrene foam and clad it in a magnesium skin, to try to save some weight --- which it did. But the Cosworth shook the polystyrene to pieces, so the wings were throwaways after a race, although we re-used the mag skins." Not shown in the cutaway are the small canards ahead of the front wheels, added to improve downforce on both the open and coupé cars.
Gulf's transmission woes at Daytona were caused by pinion bearing failures in the Hewland DG300 transaxle. "Nobody had used it at those speeds and oil was centrifuging away on the banking. Hewland improved the oil feed and we had no more trouble. But we didn't run the Hewland at Le Mans, we ran the ZF. We couldn't get the Hewland to last. We had trouble with dogs chipping at the corners, then you couldn't get a gear. We had dome-profile dogs made and they lasted much longer, but it still wasn't a 24-hour gearbox." Even the ZF was not entirely trouble-free. "We broke an input shaft in testing. After we told ZF about the DFV's high vibration levels, they said they'd make smaller shafts, which was puzzling, but they were right --- it was torsional vibration that caused the breakage. The smaller diameter shaft flexed and didn't break ever again.