The DFV was the world’s most successful F1 engine – but it was much more than that. For more than 30 years it, and its derivatives, won in almost every branch of motorsport. Sports cars? For sure. Indycars? No question. But what about hillclimb cars, special saloons – and powerboats? Yep, those too.
Keith Duckworth designed the DFV to do one thing: win F1 races for Ford. Many times in later years he complained that his One Big Idea was being used in ways he didn’t like. But Cosworth built hundreds of DFVs – maybe over 1200 if you count all the offshoots – and it was impossible to keep track of them all. Some were delivered to shadowy overseas clients and never seen again. Others with well-known identities went on and on, were sold and resold, developed and redeveloped. Some were used to settle loans. Others were the basis of coffee tables and wine racks!
Except for projects which were never made public, either by Cosworth or by its clientele, this is a summary of the major users, and derivatives, of the DFV seen in the last 40 years.
DFW – Tasman Cup
This, the first and last version of the DFV to have a reduced cylinder capacity, was built so that Lotus’s Type 49 could contest the Tasman series of 1968 and 1969. Except for the substitution of a short-stroke (54mm) crank and longer conrods to match, this 358bhp/2491cc engine was no more than a re-prepared DFV. It won five races in 1968, and three in 1969. According to Cosworth, only four were built, and all were eventually reconverted to 3-litre DFV status.
Racing Sports cars
F3L: After Ford’s 1967 Le Mans win with its 7-litre MkIVs it was clear they would not race again, and little had been done to modernise the less specialised GT40s. Ford’s Walter Hayes, ‘Godfather of the DFV’, wanted to keep Ford in sports car racing. Accordingly, even though the DFV was still very new and in short supply – and in spite of Duckworth’s insistence that it was not meant for that purpose – he commissioned Alan Mann Racing to develop the F3L. It was designed by Len Bailey and covered in a deliciously styled two-seater coupé body shell. Four brand-new DFVs were supplied, and success was expected.
None came. It was a jinxed project. F3L never won a race in two seasons of trying, either in coupé (1968) or open roadster (1969) form.
C100: Karl Ludvigsen was Ford’s vice-president, governmental affairs and motorsport, from 1980 to 1983, which is when this two-seat coupé powered by the DFV (and derivatives) was engineered. The aim was to beat Porsche, but this was never achieved.
The 1981 car was engineered by Bailey. For 1982 the project was effectively handed over to German company Zakspeed, then for 1983, with success still over the horizon, Tony Southgate carried out yet another redesign.
All versions used what Cosworth knew as the DFL – L for long-distance – of which there were three types. The original was a 90mm bore/3298cc engine which produced 490bhp at 9500rpm, but suffered badly from vibrations. An interim version returned to the standard bore (85.67mm) but had a long 77.7mm stroke, which equated to 3583cc. The definitive naturally aspirated version combined the 90mm bore and 77.7mm stroke and its 3955cc (the largest works DFV capacity) produced 540bhp at 9250rpm.
Cosworth, in the meantime, was working on a single-turbo version, fitted with twin Lanchester balancer shafts. Projected power was 700bhp at 8500rpm, but these were not available before Ludvigsen left the company. When Stuart Turner took over the project was abruptly cancelled.
Other Sports cars
De Cadenet: DFV used from 1970, when a McLaren M8C was re-engineered, then with a much-modified Lola, and finally, redesigned by Len Bailey, the de Cadenet-Lolas. These won two world championship races in 1980. The original Ford C100 evolved from the latest de Cadenet proposal.
Gulf Mirages: A Le Mans win in 1975 vindicated the DFV sports car idea – see page 50. A GR8-DFV also finished second at Le Mans in ’76, run by Harley Cluxton’s GTC team.
Inaltera: Engineered by Jean Rondeau, they raced in 1976 and 1977, but were then replaced by Rondeau-badged cars.
Ligier: A JS2-DFV took second place at Le Mans in 1975. The team then turned to F1.
Rondeau: First raced at Le Mans in 1978, won the French classic in 1980, and struggled on without much success until the mid-1980s.
The records show that DFV-engined sports cars won twice at Le Mans (1975 and 1980) and twice in the Monza 1000Km (1980 and 1982). They also won six-hour/1000km races at Spa (1973), Silverstone (1980), Enna (1981) and Brands Hatch (1981).
From the early/mid-1980s, several DFV-engined cars ran in the C2 class of the world sports car championship, including chassis from Ecosse, Tiga and – most successfully – the Spice SE88C.
Egged on by Bill Brown, one of Cosworth’s four directors at the time, DFVs occasionally found a home in powerboats. American Chris Hodges’ DFV-powered catamaran, driven by Jackie Wilson, set a world speed record of 123mph in the unlimited engine class, and later won endurance races by impressive margins.
DFX and DFS – Indycars
Although Duckworth originally set his heart against turbocharging the DFV for racing in North America – Roger Penske asked in 1972 but the project never reached the track – this did not deter the Vels Parnelli Jones team from doing their own thing; to meet USAC regulations, the very first DFV turbo was a 2.65-litre developed by VPJ. Cosworth soon adopted and re-engineered it and re-named it DFX.
The first USAC victory followed for a Parnelli VPJ6 at Pocono in June 1976. Two more wins followed later that year, after which there was an avalanche of DFX successes – not least 10 successive victories at Indianapolis, from 1978 to 1987 inclusive.
At peak, before boost restrictions were in place, a DFX produced 840bhp. It was not until 1987 that another engine got the better of it – this being the GM-financed Ilmor V8 design by ex-Cosworth man Mario Illien.
As a none-too-successful stopgap for 1988 and 1989 Cosworth produced the DFS, a redesigned DFX which drew on the DFR for many of its development details.
Although DFVs were generally thought too expensive and too rare for use in club racing, there were some impressive silhouette Special Saloons on the British scene.
In 1976 Colin Hawker produced an extremely smart VW1600 lookalike, which had Lola running gear and a Cosworth DFV engine under its GRP skin. Dubbed the DFVW, and backed by Toleman before it stepped up to F2 and eventually to F1, this was extremely successful for two seasons. When Toleman withdrew its support, it was retired, apparently broken up and returned to its Lola roots.
DFY – Formula 1
By the early 1980s, even when further developed outside the Cosworth factory – Williams, for instance, had engines modified by John Judd – the DFV was struggling to stay competitive with the 1.5-litre turbocharged F1 power units. In one final fling Cosworth set Mario Illien to make a two-stage redesign: one by combining the DFL’s wider 90mm bore with a new short-stroke crank (revised oil management inside the engine and different camshaft profiles also figured); the other being new cylinder heads with an ultra-narrow valve angle. Definitive DFYs had 22.5 degrees between lines of inlet and exhaust valves (with suitable porting changes to match), compared with 32 degrees for conventional DFVs.
First seen in 1983, this was the first DFV derivative to have a visually different cylinder head/camshaft cover top end. Although it produced about 520bhp this was still not enough to match the latest turbos, and it was short-lived.
Michele Alboreto’s DFY-powered Tyrrell 011 scored the last DFV-derived F1 victory: the US GP, appropriately held in Detroit, in June 1983.
When Illien and his friend Paul Morgan left Cosworth to set up Ilmor, the turbocharged V8 they designed for CART racing had a strong resemblance to a DFY.
With nothing more serious than publicity in mind, Ford commissioned a series of Transit-derived Supervans in the 1970s and 1980s.
The original was engineered – if that’s the word! – by Terry Drury, around a very simple mid-engined tubular chassis and the running gear of a GT40. It appeared in 1971. A replacement did not then break cover until 1984.
Supervan 2 looked like a current-model Transit, though the vast air scoops in its flanks , big rear spoiler, and noise soon gave the game away. Based on the chassis and running gear of a just-cancelled C100 (an alloy-and-carbon fibre monocoque) its engine was placed where there would once have been a loading floor in a real Transit. The first engine fitted was a 3.9-litre DFL and was rated at 590bhp. Martin Brundle lapped Silverstone’s full circuit in 1min 38sec, recording 176mph down the Hangar Straight.
It wasn’t until 1977 that a DFV-powered Pilbeam, driven by Alister Douglas-Osborn, won the British Hillclimb Championship. After a decade in which small, turbocharged Hart engines ruled the roost, the DFR figured strongly in Martyn Griffiths’ Pilbeam in 1990 and 1991, after which Roy Lane, David Grace and Andy Priaulx used Pilbeams and Goulds powered by one or other of the DFV, DFL and DFR family to win many championships up until 2000.
When 2-litre F2 racing died out in the early 1980s, and F1 threatened to become a pure 1.5-litre turbo formula, a large number of late-model DFVs might have been scrapped. This inspired Bernie Ecclestone to encourage the launch of a new stepping stone single-seater formula – F3000 – in 1985. This was for normally aspirated 3-litre engines, and its mandatory 9000rpm rev limit made the DFV ideal for this purpose; it produced about 450bhp at that level. From 1985 to 1990 DFVs were the most powerful and reliable power units, after which Mugen-Honda produced a brand-new V8 which was at once more powerful and much more expensive.
Although no DFV-engined road cars were built, there was no shortage of ‘if only’ proposals. Models, plaster styling studies and paper projects were all that most had to show. These were the most prominent:
Albar’s UK importer annnounced it would build a DFV-powered road car in 1984 – starting from its Beetle-based Jet kit. A good way to use up redundant DFVs, it claimed.
Cheetah, a Swiss-based company, reputedly based a car on a March chassis in the mid-1980s.
That patron saint of white elephants, Frank Costin, built a model of a proposed DFV supercar in the early 1990s.
DFZ and DFR – Formula One
Long after the DFV had become an F1 makeweight, Cosworth produced two derivatives which briefly proved more competitive, though not outright winners, against the last turbocharged engines and the new, naturally-aspirated 3.5-litre Hondas of the late 1980s.
DFZ: A simple update of the DFV/DFY, this 560bhp/3495cc engine was intended as a stopgap, for use in the Jim Clark Cup/Colin Chapman Trophy competitions for naturally aspirated F1 cars in 1987. Tyrrell, March and AGS all bought DFZs for this purpose, and at the end of the year it was Jonathan Palmer and Tyrrell who took those awards.
In 1988 many of the back-of-the-grid teams took supplies of DFZs, but only as an interim measure, as Cosworth had better things in mind for Benetton.
DFR: Developed specifically for Benetton in 1988 – and for general supply in later years – the DFR was a root-and-branch redesign, in detail but not in general layout, of the DFV/DFZ family. It looked, and was, very different. Originally Ford encouraged Cosworth to use five-valve cylinder heads designed by Yamaha, but these never lived up to their promise, and there were delays.
Ford ditched Yamaha, Cosworth redeveloped everything, and eventually the DFR produced 600bhp at 11,000rpm for 1988 – a figure which Brian Hart (his company was taken over by Cosworth at this time) eventually massaged to at least 620-630bhp.
Although it never won an F1 race, the DFR was much respected until rendered obsolete in 1991 – the 25th year in which DFV-derived V8s had been seen in F1!
Cosworth might have stopped building complete engines years ago, but you can still buy new pieces to keep a DFV fresh. These days Cosworth will even cast and machine new cylinder blocks and heads, too. Like helicopters and commercial aircraft, of course, a DFV is never allowed to get old, for every active engine gets rebuilt at regular intervals. One thing is for sure – that because of the rise of Historic F1 racing, the DFV will probably never die.