New research tells of W O Bentley’s final sportscar design and its frustrating still-birth
It is well-known that W O Bentley was employed by Rolls-Royce after they bought the old Bentley company and that after assisting with the roadtesting of the 3 1/2-litre and 4 1/4-litre Derby Bentleys, he designed other great cars. W O had gained much esteem from 1919 to 1931 for the design and production of the 3-litre to 8-litre Cricklewood built Bentleys with their great run of racing successes, notably at Le Mans, and before that for his revision in war-time of the Clerget rotary aero-engine into the BR1 and BR2 designs.
In 1947 he joined Lagonda and was responsible for the outstanding V12 Lagonda after revising the LG 4 1/2-litre Lagondas, and for the post-war 2 1/2litre and 3-litre twin-cam cars, two of which were used by HRH the Duke of Edinburgh. (Had Rolls-Royce not outbid Napiers for the old Bentley Motors concern, W O might have had the design of an advanced pre-war Napier six-cylinder ohc 24valve sportscar to his credit.) He was frustrated again at Lagonda’s by being legally prevented from linking his name to the cars he designed for the Staines company, but usually these cars are seen as his last technical achievement; indeed, Georgano says they were W O’s “last motorcar design”. However, he had one more ace to play, for Armstrong -Siddeley.
It is usually thought that this project never got off the ground, but this is not so. Thanks to the research done by Mr W E Smith of the Armstrong-Siddeley OC, who lectured the W O Bentley Society on the subject, we know there was a final project to W O’s credit, an Armstrong-Siddeley sportscar which was actually built and run. I am greatly indebted to Mr Smith for allowing me to summarise this most valuable lecture, the findings of which will, I am sure, be of much interest to all followers of Walter Owen Bentley’s illustrious career. It may not tell the full story, as Mr Smith is continuing his research into the last W O-designed sportscar, but the fresh ground covered in this remarkably interesting lecture is most intriguing.
All the more so because with the exception of the splendid Siddelely Special, which was something of a breakaway, Armstrong-Siddeley was always associated with rather staid, comfortable, even pedestrian products those “cars of aircraft quality”, as befitted a branch of the Hawker-Siddeley Aircraft group and never mind that a private owner managed to make one of the massive 1921 Armstrong-Siddeley 30s win a Brooklands race. There was a sports-tourer version of the nice little Armstrong-Siddeley Twelve Six, one of which was used as his personal car by H S Linfield, The Autocar’s pre-war road-test driver, but this was sporting mainly in the body style sense. The A-S Twenty had quite a brisk performance, and the company took part in the major long-distance competition events, including the Monte Carlo Rally, with Col Siddelely as a driver (one remembers the trio of red, white and blue AS Twelves which ran as a rally team in the 1930s) but in no way could the Armstrong-Siddeley, prewar or post-war, be regarded as a sportscar until W O Bentley was commissioned to develop one.
He signed a three-year contract to act as car consultant and designer to Armstrong-Siddeley Motors, in conjunction with Donald Bastow, and to perform such other duties as the directors might require. The usual confidentiality clauses were inserted to prevent industrial espionage. W O was allowed to have not less than four assistants or draughtsmen, whose salaries he was to pay, to work solely on A-S projects, and W O had to provide his own office premises, again at his own expense, but travelling outgoings would be met and drawing materials supplied. W O’s salary was £6500 a year, and the offices were at Kingston-on-Thames, convenient to Bentley’s home at Shamley Green. (Armstrong-Siddeley Motors had tried to buy the Lagonda company, but backed off; this may be how they became associated with W O.) The contract was signed in October 1947 by T M Sopwith, Harold Chapman and the ASM Secretary W Johnson.
Armstrong-Siddeley had been among the first car makers to bring out a new model after the end of WWII, these having a 2.3-litre six-cylinder pushrod engine and being named after the group’s famous war-time aeroplanes, Hurricane, Lancaster, Typhoon and later the Whitley. The 1991cc engine developed 70bhp at 4200rpm, good for 75mph. In 1948 it was bored out to 2309cc and then gave 75bhp at the same engine speed. What ASM was seeking was a more powerful car than these 16hp and 18hp models, even a sportscar. W O was provided with an Armstrong-Siddeley car and set to work… Apparently it was intended to lay down two cars, the first with a coupe body, the other with a sports body, and a set of parts for another car. Tooling costs were extimated at £42,000 but when the project was run down these amounted to only £1197.
Mervyn Cutler was then Chief Engineer at Armstrong-Siddeley Motors of Parkside, Coventry (hereafter ASM or AS), and from 1951 also their Technical Manager. Under him W O and Bastow would work. A start was made by modifying a 2.6-litre Lagonda engine, the six-cylinder twin-cam 12-valve unit which had proved effective and smooth-running in the pre-war Lagonda. The stipulation seems to have been a faster, more accelerative car which would retain the reliability, smoothness, solidity, spaciousness and value-formoney for which the Coventry make was respected. (At this time ASM had on the stocks a 105bhp light-alloy engine of this size, but with push-rod valve-gear. It was abandoned in favour of the 1953 346 design.) W O designed a chassis with a conventional back axle, independent front suspension and a Cotal electric gearbox, but for cost reasons a three-speed syncromesh ‘box was substituted and experiments were made using a Wilson preselector box and Newton clutch.
Bentley and Bastow were hampered by ASM requiring a £500 car, and by having to re-draw their suspension, gearbox, axle and engine layouts, as well as a complete new engine. By August 1948 Chapman was complaining to Sid Thornett, Cutler’s Chief Designer, about the slow progress being made by Bastow in W O’s absence (he was minuted) which was perhaps the commencement of later dissention. Sid was told to take any action necessary to complete the 3-litre drawings. It was nine months since work had started, and at a meeting in Weybridge (at Hawker’s Brooklands factory perhaps?) Thornett was told to put more pressure on the two engineers to get the 3-litre chassis completed. But remember that W O and D B had only four assistants and were scarcely being left untroubled. It had been decided to go for a 3-litre six-cylinder all-aluminium twin-cam engine of 84x90mm (2993cc), with a seven-bearing crankshaft. Three units were built and tested in the Experimental Department; the 3-litre gave 125bhp at 5100rpm, the 2.3-litre 78bhp at 4400rpm.
Parkside compared the 3-litre twin-cam W O engine linered-down to 2322cc with their own 2310cc push-rod power unit and an American-type lower-revving engine. The W O engine gave a peak of 264lb/ft bmep, compared to 297lb/ft from a supposed 4.9-litre engine; W O’s design would peak at about 5000rpm, the big engine at 3600rpm, allowing the respective gearings to give 9820 and 10,400 ‘explosions’ per mile in top gear, the respective weights being 520lb (without aircleaner) and 650-700lb. The calculated weight of the W O car with driver and passenger was 1.708 tonnes, and with a Cotal gearbox 0-90mph would take 45sec, or 39.4sec with a three-speed manual. As piston speed would be 1932ft/sec from the 3-litre engine, 2127 in the American-engined car, the former would wear better. These were “paper” figures.
Meanwhile W O was revising his Lagonda engine for lower-cost production, changing the threaded-in crankshaft assembly for the conventional design; 10mm plugs replaced 14mm ones, changes were made to the water valley in the cylinder head, and other castings, as Armstrong-Siddeley used a different source than had Lagonda. The multiple-chain camshaft drive was abandoned, eliminating four sprockets, idler shaft, intermediate shaft, one blade and one tensioner. The chain-length was then 4ft 6in, but Renold thought it satisfactory and it made for a more compact chain-case. Other changes were to a vertical oil-pump drive spindle, a more accessible ignition distributor, and a new position for the water pump.
Tests were done with a Cotal, Wilson and normal gearboxes, using steering-column control for the two latter, and with Buick and Borg-Warner syncromesh, as well as with a fluid flywheel and Wilson centrifugal clutch, W O preferring the last named. Bastow, driving what may have been the prototype in a fog, found third gear better than second for stop-start driving, but this caused some blueing of the Newton & Bennett clutch plate.
By January 1949 there was pressure on W O to complete the transmission scheme for costing. (It is interesting that in 1952 F W Allard, who had designed the pre-war Siddeley Special, was still the Chief Designer under Mr Cutler.) W O and Bastow got out schemes for de Dion rear suspension, independent rear suspension, and normal leaf suspension; cost necessitated the last-named. With additional bracing to obviate front-end vibration, an Armstrong-Siddeley stipulation, the chassis was heavier than that of the Lagonda, so 6.5×16.00 instead of 6×16.00 tyres were specified. Other design exercises covered engines of different sizes, a four-cylinder and small-bore variant of the 3-litre, even a transverse-engined Cotal-transmission fwd car! On a test engine the 5/8-in auxiliary drive shaft in the timing chain was found to be faulty, Thornett asking for its revision.
At this stage Col Cyril Siddeley took charge of communications with W O, not always to Bentley’s advantage. The 3-litre engine proved very smooth and refined, but did not give the perfomance required against the Armstrong-Siddeley 2.3-litre push-rod power unit. The prototype W O sportscar gave 95mph against 90mph and 0-50mph in 7.54sec against 8.92sec; tests were done on low-octane petrol. At this stage there was discussion between the Colonel and Bentley about details of the engine, parried by W O, although his camshaft drive involving nine chains and three sprockets was pared down to one with five sprockets and one chain. W O admitted chain ‘moan’ but this passed off with mileage. He gave quick solutions to Col Siddeley’s worries about the starter-motor solenoid and brake-fluid cylinder being too close to the exhaust manifold, and pointed out that the fan was now driven by the dynamo belt instead of by chain, so simplifying the timing chain arrangement. The water-pump was also repositioned to reduce engine length and make the pump accessible without having to remove the radiator.
Col Siddeley sent W O a memo thanking him for the work done on the de Dion rear end, though it was not to be used, but after that he seemed less interested in the sportscar, perhaps influenced by his desire that his cars should be quiet and have room for leg-stretching and good luggage space. There was even discussion about putting a limousine body on W O’s new chassis. He then objected to Bastow’s arrangement of mounting brackets, saying no stress must be taken by the chassis, yet he had previously sanctioned unitary Armstrong-Siddeley body/chassis construction! By May 1949 he was telling W O not to waste energy and time on his 3-litre engine, as the 2.3 was adequate and if a departure from it was made the public would think Armstrong-Siddeley did not know what to make. He asked W O’s advice about torsion-bar suspension, such as he had employed on his Lagondas. But the emphasis was now on “the well-trodden path”, and the sportscar idea receded.
From then on the colonel’s attitude to the great designer of Bentley cars was patronising, observes Bill Smith. He was annoyed when W O and Bastow went to Parkside without telling him or his staff and told W A Henley, the Commercial Manager, that he was in charge. However, the 3-litre engine was apparently put into an A-S 18hp chassis fitted with a nice Graber-style coupe body and tested for 5000 miles (no long Alpine tests as W O had driven with the then-new 3 1/2-litre Derby Bentley). The Cotal gearbox was in use, but a Wilson box was to replace it. The gearbox gave some trouble, but the car ran well on the track (MIRA?), and the cause of the fractured cast-iron tappets was discovered. However, before the car was shown to Sopwith it was necessary to lower the back springs by up to 2in and put a wood block at the front to attain the desired height.
Col Siddeley then memoed some criticisms to Bastow, clearly moving away from a sportscar which he felt was for the specialist motorist, not Armstrong-Siddeley customers. Although Sopwith had inspected the prototype, the project was now over. W O asked to see the Colonel but was told he was very busy… The contract was duly terminated and it says much for W O that he is so gracious about the matter in his memoirs. (One is reminded of how when Sir Henry Royce interviewed W O for an appointment with Rolls-Royce after they had bought his old company, it was suggested that he was a commercial man rather than an engineer. He replied that when Royce was a boy in the running sheds of the Great Northern Railway at Peterborough he was a premium apprentice at Doncaster!) But W O was apparently glad the Armstrong-Siddeley contract was over.
It is apparent that by now Armstrong-Siddeley had the Sapphire cars in mind (they were in production by 1953), with their 3.4-litre ‘square’ six-cylinder engines that had hemi-heads with push-rod valve-gear, obviating the need for overhead camshafts. The Sapphire and Star Sapphire were nice cars; in the twin-carburetter 4-litre version, Armstrong-Siddeley had their comfortable 100mph closed car 22 years after the W O 8-litre Bentley had been the first production saloon to exceed the ton.
The Bentley Armstrong-Siddeley Graber coupe with A-S chassis (JWK 722) sadly ended up as a factory hack, but survived to the end of Armstrong-Siddeley production in 1960. It is said that it was then dismantled, along with the Sapphire that had a W O twin-cam engine, and given to one of the new universities, there to vanish. The power-unit drawings are in the Armstrong-Siddeley OC’s archives. I am indebted to Mr W E Smith for allowing me to quote freely from the paper which formed the subject of his lecture to the W O Bentley Society, an organisation dedicated to research, whose secretary is Mr Ray Wiltshire, Southpark Farmhouse, Helmingham, Stowmarket, Suffolk IP14 6EP. As Bill Smith says: I think the Armstrong-Siddeley project leaves W O Bentley and Donald Bastow with their reputations enhanced, not diminished. WB