The 4 1/2-Litre Bentley
Exactly a year ago we published an authoritative and much-appreciated article on the famous 3-litre Bentley car, which was compiled at our request by the Bentley Drivers Club. So popular did it prove that it was even translated into Norwegian for publication in the official motor paper of Norway. Consequently we have much pleasure in presenting similar data on the 3-litre’s immediate successor, the 4 1/2-litre. This article is also the work of the B.D.C. and certain of their members who were closely associated with these cars, both in unblown and “blower” form, have kindly “vetted” the manuscript and have provided between them much inside information. — Ed
When the article dealing with the 8-litre Bentley, contributed by the Bentley Drivers Club, appeared in the February, 1947, issue of this journal, it was visualised that it would be followed by another article similar in character covering the development of its big brother, the 4 1/2-litre. Much research has resulted in the compilation of the following information, which may be regarded as authoritative. The principal reasons for the entry of the world-famous 4 1/2-litre Bentley into the motoring world of 1927 were twofold :—
1. An ever-increasing demand by the “Bentley-minded” motoring public for just that extra reserve of everything which they had come to value so highly when driving the three-litre.
2. It was considered that the ever-increasing pressure experienced in the field of International Competition would extend the three-litre beyond its capacity and render it less capable of dealing with its opponents; or that, in attempting to do so, the great reserve and reliability for which the marque was even at that time world-renowned, would of necessity be sacrificed.
The first 4 1/2-litre car to be produced was the famous old No. 1 “team” chassis, ST 3001, immortalised in motoring history as a participant in the epic crash at White House Corner on the Sarthe Circuit in the course of the Grand Prix d’Endurance at Le Mans in 1927. It will be remembered that hopes of a runaway victory, on this occasion, were dashed to the ground by the unfortunate triple crash involving the entire Bentley team.
Incidentally, some confusion has arisen due to the similarity of the radiator of this particular 4 1/2 to that of the standard three-litre and, indeed, the radiator-shell from trunnions to fillercap was, for all practical purposes, identical in shape and contour to its predecessor. The difference was to be found in the overall length of the shell, the design of the mounting, and the capacity of the matrix. The first nine chassis produced had radiators the filler-caps of which were smaller in diameter than those later standardised, although nearly all of them we’re subsequently modified.
Development of the 4 1/2-litre followed the development of the new six-cylinder 6 1/2-litre model, the latter then being in full production, and as might be expected, the two chassis have many features in common, with the result that such components as pistons, connecting-rods, valve-rockers, valves, valve-stem collars, valve-springs, etc., were interchangeable.
A noteworthy point is that pistons which gave a 4.4:1 compression ratio as employed in the 6 1/2 resulted in a compression ratio of 4.8:1, if fitted to a 4 1/2-litre. This increase was due to a difference of 8 mm. in the dimension combustion head to cylinder foot of the two cylinder-blocks concerned (6 1/2-litre: 270 mm., 4 1/2-litre: 262 mm.). A word of timely advice to enthusiasts with plans for machining large quantities of metal from the block in search of extra horses. A reduction of 3 mm., i.e. a dimension of 259 mm., is the absolute safety limit to which the mean height of the dimension combustion head to cylinder foot should be reduced. The effect of this reduction is that the cylinder foot thickness is reduced to an average thickness of 9 mm., below which it is inadvisable to go.
From a general design point of view, the 4 1/2-litre was similar in character to the 3-litre. Valve gear, magneto and water pump drive, also sump design, were basically the same, differing only in dimensional detail. A brief specification of the 1927 standard 4 1/2-litre chassis may, therefore, prove of interest:—
Bore 100 mm. Stroke 140 mm. Cu. cap. 4398.24-c.c. Direct metalled con. rods type G.5. Twin S.U. sloping carburetters. Duralumin rockers. Belt-driven fan. Rear axle ratio 3.53:1. Wheelbase 10 ft. 10 ins. Hobson ” Telegauge ” petrol level indicator. Calculated road speed at 3,500 r.p.m. with 3.53 axle and 5.25 by 21 ins. tyres 92 m.p.h. B.H.B. pistons. Compression ratio 4.8:1. Ball-end tappet adjusters. Cone clutch. Wellbase rims; 16 gallon petrol tank. Smith double-pole lighting and starting system. “C” type gearbox. Tyres 5 25 by 21 ins. Low-geared worm and worm wheel steering. Autovac fuel feed.
Incidentally, although the 10 ft. 10 ins. wheelbase was adopted as standard, eight chassis were built with a wheelbase of 9 ft. 9 1/2 ins., these being to special order.
The first modification of any importance made to the chassis was the incorporation of thermostatic water temperature control in the late autumn of 1927 (at chassis No. SL 3071). This was followed by the change over from the rigid spoke type steering wheel to the spring wheel which had proved so effective under rigorous competitions (at chassis No. RN 3035).
1927 saw the 4 1/2-litre Bentley entering international competition in no uncertain manner. “Old Number 1” was entered, as previously mentioned, at Le Mans in 1927 and shortly before the crash gained the lap record at 78.40 m.p.h. It was driven on that occasion by Clement and Callingham. It appeared again at the Grand Prix de Paris on August 19th and despite extremely bad weather conditions, won the race at 52.54 m.p.h., handled on this occasion by Clement and Duller.
The winter of 1927-28 saw the adoption of the Ki-gass injector and the long range type E.R.4 M.L. magneto as standard equipment. These two improvements considerably reduced the bogey of difficult starting and the consequent drain on batteries, usually experienced with large bore engines. These changes in specification became standard at the end of February, 1928 (at chassis No. HF 3192).
In an endeavour to reduce the lubrication problems of owner-drivers and, at the same time, to reduce the wear and tear experienced with the “plunging joint” type of cardan or propellor shaft with its blocks and slippers, much development work — the direct result of high speed track experience — resulted in the adoption of the enclosed type, balanced splined propellor shaft as standard for all models. All chassis from No. NT 3131 were fitted with the Hardy Spicer propellor shafts.
The next real step in the development of the 4 1/2-litre chassis was the design, development, and production of the now famous “D” type close ratio gearbox. Up to this time all chassis had been fitted with the “C” type general purpose gearbox, giving the following ratios:
In view of the popularity of the close ratio “A” type box as fitted to the 3-litre speed model and the absolute necessity for these ratios in competition work, an altogether new design of gearbox made its appearance. This, of course, was the “D” type and although completely re-designed to cope with the increased torque of the 4 1/2-litre engine, it retained the well-tried indirects of the “A” type box, whilst eliminating those danger points which had given trouble from time to time in the form of rapid wear in gears, main shaft keys, speedometer drives, etc.
It is a noteworthy tribute to the foresight and skill of the designers of the first truly sporting car to be produced after the 1914-18 war that the ratios calculated to be correct for the first 3-litres were re-introduced in 1928; these are as follows :—
In the “D”-type gearbox the 7 pitch gear superseded the 7/9 stub-toothed gear of its prototype the “A” type box; 5 pitch gears may be found on a few “A” type boxes and 5 pitch gears were also used in the Birkin Blower 4 1/2-litres. These special purpose gears were few in number, the standard 7 pitch gear being used in the majority of cars for competition work and for all standard chassis incorporating this box. Other features of the “I) ” type box were a ribbed casing of sturdy construction, multi-splined main and lay shafts of reduced overall length to avoid distortion under maximum loads, a speedometer drive re-designed and located similarly to the “C” type box, on the near side of the easing, and an external oil filler.
The whole design proved itself in practice to be most reliable and, for a “crash type” or non-synchromesh box, retained its silence in operation under competition conditions for extremely long periods, being absolutely trouble-free as a unit throughout its history. Chassis No. HF 3196 inaugurated the “D” type gearbox as a standard unit for all 4 1/2-litre chassis.
Around this period several minor alterations in design were introduced, amongst which were the adoption of the .44 carbon steel brake drums in place of the hitherto standard low carbon steel units, in an endeavour to increase the wear-resisting properties of the drum against the loads imposed upon them by braking from high speeds.
Gudgeon-pins of the floating type were superseded by circlip-secured pins, but in view of the trouble experienced with scored cylinder-bores by reason of the clips coming adrift, an early return was made to the original design of floating type pins. Incidentally, the whole secret of success with the type of gudgeon-pin that employs the conical aluminium alloy pad method of location — apart from the use of the correct alloy for the pad — lies in the maintenance of point-contact with the cylinder wall by the conical pad. As soon as the area of contact with the cylinder wall exceeds for all practical purposes “point-contact,” there is a tendency for the pad to rotate in the gudgeon-pin bore with disastrous results to the cylinder walls — as many amateurs experienced to their cost. It is, therefore, of paramount importance in design that the angle of the conical buffer or rubbing pad is such that it precludes the possibility of the outer edges of the cone coming into contact with the cylinder walls, until the wear-factor on the tip of the cone is such that replacement of the pad becomes essential. Further points worthy of note, if readers fit replacements, are :—
1. A small air-release hole should be drilled in one pad to prevent pressure-generation in the tubular gudgeon-pin and a consequent forcing from position of the pad.
2. The bore of the gudgeon-pin should be most carefully finished to ensure the correct interference fit between pad diameter and gudgeon-pin bore.
3. The interference fit should not be excessive if distortion and local swelling of the ends of the gudgeon-pins is to be avoided.
4. The length of the plain portion of the pad should not be less than 15 mm. for this diameter gudgeon-pin.
Le Mans 1928 heralded the reappearance of the old veteran chassis No. ST 3001, victim of the White House crash in the previous year, this time driven by Barnato and Rubin, but supported by two new 4 1/2-litre chassis Nos. KM 3077 and KM 3088. These were fitted with distinctive Vanden Plas bodies designed to conform to the stringent Le Mans body regulations in force at the time. They were easily recognisable by the position of the tank and spare wheel carrier at the rear of the body — and were to become international celebrities during the course of their racing careers. It was No. KM 3077 in which Tim Birkin, in partnership with the veteran Jean Chassagne, experienced the burst tyre episode which relegated them to the fifth place in this event; there was nevertheless a degree of consolation in the fact that they broke the lap record with the creditable speed of 79.73 m.p.h.
KM 3088, driven by Clement and Benjafield, retired through a water leak. It is not generally known, however, that the chassis frame was broken and this, if not the direct cause, was a contributory factor in the elimination of the car from the race because of a split or displaced hose joint.
ST 3001 won the race at an average speed of 69.11 m.p.h. This car, too, had a broken frame and covered the last lap without water, a hose having pulled off! Heavier frames were incorporated the following year.
A brief specification of these cars will accordingly be of interest and was as follows:—
Compression ratio 6.1:1. Hour-glass pistons. 8 mm. removed from foot of block. BM-3480 camshaft. .019 tappet clearance. Twin S.U. carburetters. Cone clutches in two chassis, plate-clutch in KM 3088. Close ratio gearboxes. Direct-metalled, heavy-section con-rods : rear axle ratio, 3.3:1. Straight-tooth gears. 81 ins. by 5.35 Dunlop road racing covers and tubes. Road-speed at 8,500 r.p.m. with 8.3:1 rear end ratio and 81 ins. by 5.35 tyres, 100 m.p.h.
The following important changes in design received their baptism of fire during the 1928 racing season:—
(a) The self-energising front-brake shoes, commonly known as “servo-shoes.”
(b) The single-plate clutch.
(c) The vertical type S.U. carburetter, Type HV G.5.
These units were the outcome of development over a considerable period, carried out in order to cope with the exacting conditions of international competition.
It was during the 1928 season, too, that the reinforced H-section front axle-beam was introduced. The sectional thickness between spring-pad and swivel-pin eye, in this axle, was increased in order to counteract the increased brake-torque occasioned by the new servo-shoes. These re-designed axles did not, however, become standardised until May 1928, at chassis No. KM 3092.
Three 4 1/2-litre Bentleys were entered for the Six Hours Race. at Brooklands on May 18th that year. Their race-numbers were 4, 5 and 6, and they took eighth, third and sixth places respectively, carrying off the Team Prize in the process. No. 4 (Old No. 1) was driven by Barnato and Clement and, despite some brake trouble, finished at 70.29 m.p.h., No. 5 by Tim Birkin (72.27 m.p.h.) and No. 6 by Benjafield and Rubin (71.26 m.p.h.). Some aspects of Le Mans 1928 have already been discussed and it will additionally be recalled that Barnato and Rubin brought “Old No. 1” over the line in first place at 4 p.m. on June 18th.
Up to this period the compression ratio of the standard engine had been set at 4.8:1 but, with the introduction of the BM 3622 piston at engine No. KM 3081, this was raised to 5.3:1. By employing a 13 S.W.G. compression plate, however, this ratio could be reduced to 5:1, thus providing readily available alternative ratios.
The plate clutch, servo-type front brakes and vertical S.U.s were not standardised until the advent of the XR, series chassis at No. XR 3332 during the autumn of 1928.
The policy of regarding the racetrack as the most suitable test-bench possible was pursued with greater intensity than ever and two 4 1/2s, in the hands of Tim Birkin and H. Cook, finished 5th and 7th in the T.T. on August 23rd. Birkin appeared again with a 4 1/2 in the Georges Boillot Cup in September and finished 5th, after beating all records for the race with a speed of 73.16 m.p.h. He even took one to Nurburg, a course to which it was by no means suited, and yet managed to achieve 8th place after a drive which drew much applause and favourable comment from those present.
Another modification introduced about this time, was the use of stainless steel in place of aluminium sheet for the cylinder water-jacket plate. This modification, coupled with the use of a corrosion-resisting aluminium alloy for the water pump body, reduced to a very great extent the sludge content of the radiator matrix, which consisted, for the greater part, of the products of decomposed aluminium. The hitherto prevalent bogey of the perforated water-jacket plate was thus eliminated.
In the early autumn of 1928, it was evident that the demand would persist for a 4 1/2-litre chassis fitted with closed coachwork and it was therefore necessary to revert to the “C” type or general purpose gearbox and to fit an “anti-boom” type double silencer to these chassis. Another feature was also inaugurated with the installation of a single-pole wiring layout and this was duly standardised in time for the Motor Show.
Early in 1929, the magnesium alloy known as “Electron” made its debut and replaced, at intervals through the year, components formerly manufactured with the standard aluminium alloys.
Throughout the spring of 1929 various minor modifications were standardised, culminating in the preparation for Le Mans, wherein heavier section frames made their initial appearance.
This preparation followed very much the same line as in the previous year. The engines used fiat-topped inlet valves in conjunction with stronger valve-springs.
The remainder of the specification was substantially the same as for 1928.
The latter two cars were added to the team to cover the withdrawal of the supercharged cars and they received somewhat hurried preparation. Work on KM 3088 was carried out by night and day in England and was watched over by the eagle eye of “T.K.W.” (now Major T. K. Williams), who was personally responsible for its delivery at the start on time. KM 3077, however, was transferred straight from Montlhery, where it had just broken the International Class “C” records for 24 hours, 3,000 kms, and 2,000 miles in the able hands of Mrs. Victor Bruce, and it was taken to Le Mans without any major preparations whatsoever.
The first “Speed Six,” incidentally, won this race. It was handled by Tim Birkin and Woolf Barnato — and its speed was 73.62 m.p.h., but to digress further upon this point would be to go outside the province of this article.
It was during this race that a degree of trouble was experienced with the headlamp bulb-holders and focusing device.
The finish of Le Mans 1929 provided the finest spectacle in British motor-racing history to date as four green Bentleys thundered triumphantly past the chequered flag in the first four places, and the 4 1/2s came 2nd, 3rd and 4th (9, 10 and 8). Clement lost valuable time during the course of the race owing to his lead ballast shifting.
This was contained in steel tubes and, when it became displaced (owing to the stretching of the security-bolts) it distorted the body-frame and fouled part of the front brake-operating mechanism. Kidston’s oil-filler overflow-gear came adrift but, apart from that, he had a clear run, while Benjafield experienced electrical trouble when the inter-cell connecting lug fractured on the off-side 6-volt battery. The 4 1/2 allotted to Howe and Rubin very obviously resented the apparent lack of gratitude for her excellent performance at Montlhery, evidenced by the lack of pre-race attention to her various needs and requirements, and stripped the fibre distributor-gear on one of her mags. When the magneto was duly replaced, however, a certain oil-retaining paper gasket was omitted and that was too much; so in high dudgeon she seized up her magneto cross-shaft and broke it — and that was that.
Available records reveal the interesting fact that oil-pressures (hot) which at the start of the race stood at. 40-45 lb. per sq. in., had dropped, after 20 hours, to between 10 and 15 lb. at 3,500 r.p.m. — some indication of the gruelling nature of a race of this description. Valve-rocker failures, roller-trouble and valve-spring breakage were conspicuous by their complete absence.
From a competition point of view, it can be noted that “cavitation” whilst cornering due to low sump oil level was experienced in all 4 1/2-litre cars precisely 2 hours 30 minutes after refuelling — although the reserve oil-tanks covered this contingency with plenty in hand.
Towards the close of the race KM 3088 suffered from a bad water leak from the pump gland when the engine was stationary at refuelling stops. Much time would have been saved at the pits had it been realised that the leak would have been very considerably reduced — if not entirely sealed off — by the action of the pump thrust when the engine was running.
Although worn down to reline limit, with the special brake adjuster used to its limit of travel, “anchors” on all cars were still effective at the conclusion of the event.
Despite these sundry oddments which caused irritation during the race, the Bentley team, by a performance unparalleled in the history of the motor industry, blazed its way into the Valhalla of motoring on June 15/16th, 1929 . . . Nor should the other competition activities of the 4 1/2-litre Bentleys in that year be overlooked. Davis and Gunter finished 2nd in the Double-12 at 81.39 m.p.h., while Cook finished 3rd in the Six Hours Race at Brooklands. They roared over the line 3rd, 4th, 5th, 7th and 8th in the Irish Grand Prix, in the hands of Birkin, Wood, Cook, Scott and Rubin (those of Birkin and Rubin, however, being blown versions) and they were 1st and 5th in the classic “500” at Brooklands on October 12th, the winning car being driven by Clement and Barclay at the creditable speed of 107.32 m.p.h.
This article would not be complete without reference to “W.O.” riding as Tim Birkin’s mechanic in the T.T. of this year. It was caused by a certain amount or nattering among the mechanics who considered that the course was a very dangerous one and in order to give a lead “W.O.” decided to go as Tim’s mechanic. This gesture symbolises the whole spirit which went into the original Bentley make-up.
Thus finished a memorable racing season, full of high hopes, disappointments in profusion, but above all, an invincible emergence into prominence.
During the summer of 1929, considerable modifications were made to the standard chassis. These included connecting-rods using separate steel shell white metal-lined big-end bearings (from engine No. NX 3456 in June). These connecting rods were designed and developed in an endeavour to arrest the tendency to cracking of the white metal, a fault evident in the direct metalled rods. In this Bentleys were entirely successful, but many racing enthusiasts to this day prefer the direct metalled rod with the .010 inch to .015 inch white metal lining.
Main bearing lubrication was modified as a result of experience gained in the field of competition — and the rocker-casing was altered to oil-bath the rocker-rollers.
The Pulswell Silencer replaced the double silencer at chassis No. XF 3514 in October, 1929, and the re-designed and strengthened half-shafts became standard at the same chassis.
It was not until the autumn of 1929 that electron castings completely replaced the somewhat weightier but hitherto successful aluminium alloys as standard practice — the KL and PB series engines incorporating these changeovers.
One final major alteration in design, a direct result of the development of the supercharged 4 1/2 engine, became standardised during February, 1930, in the adoption of the heavier blower-type crankshaft with its attendant crankcase and cylinder-block at engine No. AD 3651.
The autumn of 1930 saw the introduction of .55 carbon drums at chassis No. FS 3602: reinforced electron sump at engine No. FS 3611 and reinforced crankcase at engine No. XT 3267. In 1930, the 4 1/2 was still acquitting itself creditably in competitions. Kensington Moir Was 11th, and won his class, in the Tourist Trophy, at 68.78, and Eddie Hall was 12th at 68.36 m.p.h.
The Supercharged 4 1/2-litre
Fifty of these were produced as standard chassis, but several had their blowers removed at a later date.
The supercharger was the Amherst-Villiers Mk VI Roots-type with double-walls, situated between the front dumb-irons and driven at engine speed off the front end of the crankshaft. The radiator was re-designed to accommodate the blower. The lubrication system of the engine fed the gear-wheels. Two S.U. carburetters were fitted on the off side of the supercharger and a series of spring loaded safety valves were arranged in the inlet pipes to safeguard against damage should a blow-back occur.
The blower gave a pressure of 10 lb. per sq. in. from 25 m.p.h. in top gear and did not exceed 11 lb. per sq. in. at the highest engine speed.
Various modifications were incorporated from time to time :—
At chassis No. SM 3906: Steel alloy diff. centres (BM 5423). At chassis No. SM 3910: Reinforced sump. At chassis No. SM 3920: One side location of mag. cross-shaft. At chassis No. SM 3921: Reinforced crankcase. At chassis No. MS 3926: Ribbed rotor-casing 4773 in lieu of double walls, and .55 carbon steel brake-drums. At chassis No. MS 3941: 25 gallon fuel-tank.
These chassis gave excellent service in practice within the limits of the blown engine under traffic conditions. The greatest fault was the tendency of the mixture control to become stiff in operation by reason of its situation in relation to the instrument board control and also owing to its proximity to road-grit, etc. Apart from this one point, which would doubtless have been overcome by logical development had the marque continued in existence, its performance was absolutely trouble free. These blower 4 1/2’s behaved with complete normality and for their power were docile to a degree which had to be experienced to be believed.
Thus concluded the development of a world-famous chassis which ceased production in the early months of 1931. At this time it is interesting to note the standard 4 1/2-litre chassis cost £1,050, and fitted with a four-seater open sports body, cost £1,295. The supercharged version cost £100 more in each case.
It is not proposed to deal in this article with the special Birkin Blower >4 1/2s which he built at the Welwyn Garden works. These were developed entirely with a view to racing, and, though on the basis of standard blower cars, incorporated many features beyond the scope of ordinary production requirements. They were, however, most successful and added further laurels to British motor-racing prestige.
This year saw one of the finest performances standing to the credit of the marque and, incidentally, one of the high-lights of Tim Birkin’s career, when he finished 2nd in the French Grand Prix at Pau in a stripped blown 4 1/2-litre Bentley touring car against the official Bugatti team and other racing cars. Two other noteworthy successes in 1930 were Birkin’s 4th place in the Irish Grand Prix and Benjafield and Hall’s 2nd place in the “500”at 112.12 m.p.h. — both these being supercharged cars.
The single-seater 4 1/2-litre supercharged Bentley in the hands of Tim Birkin twice held the Brooklands Outer Circuit Lap Record and eventually held it in 1932 at 137.89 m.p.h. (This particular car is now being rebuilt as a road-car by Shortt for Peter Robertson-Rodger who has designed a new two-seater body not unlike the single-seater.)
In all 662 4 1/2-litre Bentleys were produced, of which 8 had a wheelbase of 9 ft. 9 1/2 ins., the remainder being 10 ft. 10 ins. Two of the latter were supplied to Birkin for supercharging (HB 3402 and HB 3403). In addition, there were the 50 supercharged chassis and Birkin built 2 supercharged cars (HP 3976 and HR 3977).
Six further 4 1/2-litre chassis were built in the early thirties by Bentley Motors (1931), Ltd., from parts taken over from the old company.