Ford Cosworth DFV: the greatest racing engine ever made

Racing tech

Cosworth's DFV engine dominated Formula 1 for more than a decade, also finding success at Le Mans and in IndyCar. Lawrence Butcher examines what made it unbeatable

Lotus 49 of Jim Clark at 1967 Mexican Grand Prix

Jim Clark, Lotus 49 and Cosworth DFV: a defining combination

Bernard Cahier/Getty Images

135 years of innovation: making of the internal combustion engine, Part 6

Just as the Coventry Climax racing effort was fading in the mid-1960s, a new contender was waiting in the wings, which would seal its place in racing folklore: Cosworth. The company name was a combination of engineers Mike Costin and Keith Duckworth’s, both of whom worked together at Lotus and shared a background in aviation. Their most famous creation, the DFV 3.0-litre V8 engine, would form one part of an iconic motor racing combination in 1967, when paired with the Lotus 49 and Jim Clark. It would go on to be the engine of choice in grand prix racing for over a decade.

Much like Miller and Offenhauser before them, Costin and Duckworth were both quite different in their approaches to engineering. Duckworth was a first principles type, capable of thinking his way around problems, while Costin was more the hands on type; between them they would make an almost unstoppable combination. There were, however, at least two other key players in the DFV story, Colin Chapman and the Ford Motor Company, via Walter Hayes, without whom, it is unlikely the young Cosworth company would have secured funding to build its three litre Formula 1 engine.

From the archive

Having employed both Costin and Duckworth, Chapman was well aware of their talents and when the pair set up on their own in the late ‘50s (when Costin was in fact still Lotus technical director), a continued relationship with Lotus was logical. As the 1960s rolled around, Cosworth busied itself building engines for Formula Juniors, Formula 3 and Formula 2. This would see Duckworth start to refine his thinking around cylinder head design, later put to good use in the DFV, with the development of the SCA Formula 2 engine. That engine used the Ford Anglia 105E engine as a basis, retrofitted with an overhead cam head and gear drive on the front of the block. He focused on upping the volumetric efficiency of the engine, adopting steeply angled ports to reduce shrouding of the valves. By the mid ’60s the SCA was the engine to beat, but Cosworth had bigger plans and Duckworth was working on a design for a four-valve engine, in anticipation of new Formula 2 rules due in 1967 which upped capacity to 1600cc (from 1.0-litre).

The arrival of the 3.0-litre Formula 1 rules took the various British manufacturers somewhat by surprise. With big-banger sports cars making rather a mockery of F1’s 1.5-litre minnows, there was a desire from both organisers and teams to make the class faster. The British outfits hoped for a 2.0-litre formula, which would allow for the use of enlarged versions of the engines they already had. However, at a meeting with the FIA in 1963, which sought the opinion of teams before rules for 1966 were agreed, the British contingent – led by Chapman – opened the bidding at 3.0-litres; the FIA promptly agreed. Cue a dash to find new engines and with Coventry Climax soon to be out of the picture, there were not too many obvious options.

Ferrari and Maserati dusted off V12 designs from the 1950s, BRM decided it could take its 1.5-litre V8, stack two on top of each other having flattened the vee out and make an H16, Harry Weslake readied a V12 for Dan Gurney’s Eagles and so on. Never one for compromise, Chapman knew he needed a bespoke engine built to the formula and Cosworth would be the ones to build it. There were just two problems: time, and money.

Developing an all-new engine has never been a cheap endeavour and Chapman went off to try and raise funds, but pleas for government backing fell on deaf ears while Ford did not wish to enter Formula 1, despite its existing ties to Lotus. As a stopgap, Chapman made a deal with BRM to use its H16 while he continued the search for cash. Enter Walter Hayes, then head of public affairs at Ford of Britain, who knew Chapman well. He would bring Ford back into the game, convincing the UK operation to stump up the £100,000 needed at the end of 1965. The deal was that Cosworth would build a 3.0-litre F1 V8 and the 4-cylinder F2 motor. Duckworth and Costin had not been idle and by mid-’65 work was already underway on the latter, called the FVA, which would enter competition in 1966.

Cosworth DFV design team of Bill Brown Keith Duckworth Mike Costin and Ben Rood

DFV design and development team, from left: Bill Brown, Keith Duckworth, Mike Costin and Ben Rood

Grand Prix Photo

However, creating the DFV would not simply be a case of doubling up the FVA, which was still based around a production block. Duckworth had never developed an entire engine before and the DFV would require a bottom-up design; the result was not bad for a first try.

The engine would act as a stressed chassis member, so Duckworth had to account for the rear suspension loads in its structure, and the crankcase, cylinder heads and rocker covers were all designed to carry the forces through to the main monocoque. He opted for a 90-degree V-angle, keeping the engine narrow enough to not protrude beyond the monocoque sides of the car it was destined to power, the Lotus 49, and it was also impressively short. The use of a flat plane crank, which gave an even firing order, greatly simplified the exhaust layout.


Cosworth DFV: built entirely for racing, with no production constraints

Throughout 1966, Duckworth worked on the myriad of detail engineering problems, chivvied along by Costin when his machinations took too long. Freed from the constraints of a production cylinder block, such as a fixed bore spacing, he was able to take some of his earlier ideas to a new level. For example, the cylinder heads, while based on the FVA, had an included angle of 32-degrees (compared to the 4-cylinder’s 40 degrees). These worked in conjunction with a refined combustion chamber with a pent-roof rather than pure hemispherical design, Duckworth pursuing an idea of creating ‘swirl’ within mixture entering the chamber. This would better mix the air and fuel, improving combustion and thus power.  The results proved Chapman’s faith in Cosworth, during initial dyno running, the engine produced 408 bhp and its low weight and compact form perfectly suited the Lotus 49, which had been designed specifically to accept it.

When the DFV was unveiled to Ford and the media, it was not quite the finished product. There were a few teething issues. The engine made its competition debut at Zandvoort in 1967, Graham Hill put the 49 on pole, but during the race, an issue – already known to Cosworth – reared its head and two teeth on the cam drive gears sheared, seizing the engine. Clark came through to take the win, but reliability issues would dog the engine through that first season. Clark took secured three wins, but only wound up third in the championship.

1971 Tyrrell Ford of Jackie Stewart at Monaco

All three of Jackie Stewart's world championships were DFV-powered, including in 1971

Grand Prix Photo

Gulf GR 8 of Derek Bell and Jacky Ickx at Le Mans 1975

Jacky Ickx and Derek Bell won Le Mans in '75 with DFV-powered Gulf GR8


Cosworth would have to wait until 1968 before it could truly stretch the DFV’s legs, by which point, Ford had decreed that Lotus no longer had exclusive use of the engine. It would be a season tinged with tragedy, Hill took the driver’s championship but teammate Clark was killed early in the year driving an F2 Lotus at Hockenheim. Out of a total of 12 world championship rounds, Ford Cosworth powered entries won all but one, shared between Lotus, McLaren and Matra; with only Jacky Ickx salvaging some small honour for Ferrari with a win at the French Grand Prix.

It would not be until 1975 that a non-DFV powered car would win another championship and only the arrival of turbo power in the early 80s ended its dominance. The design’s success was not limited to F1, despite reservations about its suitability for endurance racing, it powered two Le Mans winners and in smaller capacity, turbo DFX form, it netted 10 Indy 500 wins. No engine before or since has ever shown such domination across as wide a breadth of racing series and for that, the DFV has a strong case as the greatest racing engine ever.

135 years of innovation: making of the internal combustion engine