Cosworth 4WD

They met with widely differing success, but the man with the self-belief to build his own F1 car in a lock-up knows where Cosworth were coming from with this flawed Tour De Force

Motor racing fans measure success in terms of wins and that’s not a bad thing. The Cosworth GP car never even entered a race, let alone won one. So why choose it? The truth is, I’ve never been a big fan of motor racing; it’s the engineering that did it for me. I’d have loved to have worked with Robin Herd while he put this design together, at what must have been the most exciting place to be in the late 1960s Cosworth Engineering. What a lucky chap he was.

The timing of the project, however, was very unlucky. Keith Duckworth’s 1967 DFV engine meant Formula One chassis had more power than they could handle. Traction was at a premium, and four-wheel drive was reckoned to be the future.

This was Cosworth’s first racing car, don’t forget, and they reasoned there was no point starting from scratch with rear-wheel drive. No, if they went four-wheel drive from the beginning, they felt they would be on a par with the rest in terms of this new trend, perhaps even enjoy a head start. The project began in early 1968. Trouble was, F1 cars sprouted wings that season, and the downforce cured the traction problem in a far simpler, much lighter way, rendering 4WD obsolete.

The project had lost much of its impetus even before then. Duckworth thought the world of Jim Clark, and I am sure he hoped his company was building a car that the Scot would eventually drive to victory after victory. But Jimmy was killed at Hockenheim that April, and thereafter, the project was relegated to the shed behind the factory in Northampton.

They called this shed ‘The Toyshop’. Some toy, eh? John Thompson built the car there. He had an enviable reputation for superb work; maybe just being his apprentice would have been enough, without ever designing the thing.

The car featured sponsors between the wheels, with a stressed sheet floor and bulkheads in mag nesium. The front bulkhead was double-skinned, effectively forming a box in which the front differential was housed. Its DFV was fitted back to front so that the clutch was immediately behind the driver’s seat, and the output shafts to and from the centre duff were sited to the right of the car, the driver being offset to the left to make way.

It was a striking-looking machine, and it would be interesting to compare Cd figures with other cars. I bet there’d be a few surprises. Herd’s background was in applied mathematics and aerodynamics, so I’m sure it was more considered than most designs of the day. It never struck me as being a slippery shape, more as an efficient body designed to be produced very quicky in terms of tooling and pattern-making. It’s clear the car was built as a one-off. If you make a glassfibre nosecone, you must first form a wooden

buck, then create the mould from that.

This takes up 90 per cent of the whole process of producing the item. On the Cosworth, someone simply spent time folding aluminium sheet It would have taken some drawings, but not much else.

But there was something that caught my eye more than the car’s radical shape: its bare aluminium. There’s something about this material that really grabs my imagination. It’s lovely to work with, easy to file and shape, and you can get a beautiful finish on it. A little wire wool, moistened with paraffin, a few wipes, and there’s a shine that catches the light like no other. The glint from the thread when you remove a tap from a casting is something that fires dreams for some people.

Yes, yes, you say, but what about the reality? How did the car go? Not great.

On paper, it was a purist’s dream: same suspension geometry front and rear; same tyre sizes front and rear; a 50:50 weight split; no bolt-on aerodynamic aids. On the track, though, it was a driver’s nightmare. When you look at photographs of it being tested during the spring of 1969, there are a lot of grumpy faces. Apparently, the big problem was imbalanced traction through the front wheels. The car weaved alarmingly, even on the straights. This was probably due to the limited-slip action of the front differential.

Today’s technology would have helped. An LSD using a modem viscous coupling can be tuned to give a very predictable shearing performance. They have been developed and perfected over the last 20 years or so to the point where a wide range of control is possible.

Things improved when the front duff was altered so that it didn’t work on the overrun. But the ‘ biggest jump was made when an aerodynamic aid, a rear wing, was bolted on… It was entered for Trevor Taylor to drive in the 1969 British GP, but it never arrived. Duckworth shelved the car. He realised they were throwing good money after bad: it had already cost 130,000 a lot of money back then. Cosworth had given it their all. They went to the trouble of casting a DFV block in magnesium to compensate for the weight of the drivetrain, and that in itself was a potential development nightmare. But even the talents of Herd, Duckworth and Mike Costin could not make it work. Circumstances had overtaken them.

But what a project! As Clive Sinclair said, “Those who operate at the cutting edge of technology shall be sacrificed upon it.” If nobody tried something new, something apparently crazy, where would we be? Still getting about by horse and cart, I suppose.

Peter Connew was talking to Paul Fearnley