The Marles steering box was fitted to all production Allards including the pre war Allard Specials, having been supplied by the Adamant Engineering Company of Luton. The Marles design was, and is still known as the ‘hour glass worm’, a patented principle which is quite different from the other types of steering gears used. With differing opinions by Allard enthusiasts on the divided axle, the subject of the Marles box probably comes next !

Back in 1949 the Company had to reject quite a large proportion of the newly supplied boxes because there was tightness on one steering lock and too much play on the other. The problem was caused by an initial assembly error on the worm, which was quite easily resolved. The Marles box has a ratio of 14:1 which produced between 2.5 to 3 turns lock to lock for most Allard models. The J models had the option of a slightly longer drop arm, which gave a slightly reduced number of turns lock to lock. Some 60 years on, and with its use by other UK manufacturers spare parts and maintenance is still available, in the UK, through David Cornwallis of Leominster, Herefordshire and Steering Services of Dorking, Surrey.

The lubricating oil specification was originally EP 90 and it was recommended to be checked every 1000 miles, a figure which holds good to this day, however there is a need to keep a regular check of the oil level for leaks. The clamp casting which secures the stator tube, and uses two of the four end cover bolts can be a source of oil leaks as are the 2 remaining BSF set bolts. On some steering box casings, the two lower tapped holes are not blind holes and can be a source of leaks – fitting studs to these holes could potentially reduce any oil leaks. A useful thicker grade oil can help minimise leaks, being supplied by Millers Classic Worm Steering oil, and Penrite Steering Lube – under NO circumstances should grease be used ! The suggestion of using liberal amounts of well cured sealant prior to assembly is, in my view, extremely bad practice.

1. Adjustments and examination of components
It is not my intention to go into the detail of how to adjust the box, but to highlight some of the problems which can be encountered. In my view there is only one adjustment that can be carried out while the box is installed in the car, and that is the removal of any end play in the rocker shaft. This end play can be reduced by tightening the adjusting screw (320200 & lock nut (320300). In a well used box and if the play is still excessive further adjustments or replacement parts may be necessary. While still in situ it has been the practice to remove shims from one side of the cam in an attempt to remove excessive free play. This may appear to be beneficial however, it is likely that free play on one lock may be improved but the penalty is the other lock will have an increase free play. The reality is the cam needs to be central to the rocker shaft which is not possible to do while the box is in situ.

At this point, there is, in my view a need to remove the complete assembly from the car and strip it to its component parts. One of the major problems is to remove the drop arm from the rocker shaft, this may need the use of a hydraulic press. Once stripped a close examination of the parts is needed before deciding what action is required i.e. replacement parts or retain those showing signs of minimal wear that can be re-used after adjustments are carried out.

On a well used box the cam, and its bearings are likely to show varying degrees of wear, requiring at minimum new ball bearings and tracks. The cam bearing tracks are part of the cam and may well also show evidence of pitting (see above photo) which will require a replacement cam. While one solution has been to machine the cam to accept a ball race assembly it ignores the wear in the cam worm. I would suggest that the relative costs of machining the worm against a replacement parts are not great, plus in the long term it is preferable to only use standard parts.

Evidence of the wear to the cam faces (see below photo) which can be the result of long term wear or evidence that the cam has been not been centrally positioned. This level of wear will almost certainly require that the roller, along with its bearings, will need to be replaced.  Depending on the wear on the rocker shaft it may also need replacement bushes along with oil seal (992500).

My rebuild of the Steyr single-seater started with a close look at what needed to be done to return the 1949 Hillclimb Championship winning car to it’s former glory. When Sydney sold the car, it had a four wheel drive system, along with rather ungamely nose cone – no nicely shaped front with a small version of the Allard grille. At the Valence hill climb in 1964 there was still a lot of work to do, however the engine was running well and the opportunity to ‘have a go’ could not be missed. The car sported a new nose cone complete with a modified M type grille and a bonnet air scoop. The centre body section was not the correct shape, and on the top of the cockpit shroud was an external rev. counter. When Sydney was competing with the car, he never bothered with a rev. counter, arguing he would not have time to look at it, plus it was an unnecessary weight penalty !

The tail section was as original, still baring the ‘scars’ of previous off-course excursions. Likewise, the original 15″ inch Palm Beach wire wheels were still fitted to the correct axle beams, leaf springs, stub axles and friction type shock absorbers. The stub exhausts had been replaced with the original type exhaust system which had flex pipes to each cylinder head, with a long tail pipe finishing at the rear. Finally twin rear radius rods had been replaced with the original single radius rod each side. The original deDion axle had square section coil springs mounted as the J2, along with Armstrong lever shock absorbers – at this time suspension units were still fitted, however it was my plan to make a new dead axle tube of it’s original design along with coil springs and lever type shock absorbers.

A joy to drive
I had previously entered the AOC Eelmoor Plain sprint meeting which was an ideal site to try it in ‘anger’ – apart from the need to bed in the brakes, it was a joy to drive and a challenge to adapt to the ‘reversed’ gear change – first and top were forward, but once in second gear that problem was solved.

One feature of starting the Steyr was to ensure that the high level carburetor float chambers did not flood. To ensure this didn’t happen the engine was spun over with the plugs removed and the pressurised fuel system turned off. Once any surplus fuel had been expelled platinum pointed plugs were fitted, the magneto and fuel switched on – with its own starter, and external battery, it fired up without any trouble. It was however, very important to turn the fuel off a few seconds before stopping the engine.

My runs at Valence in 1964 were made without any drama, managing to round the final sharp right hand bend without getting too close to the banks on either side.

(Above left) – The Steyr, Sydney and the team at Shelsley Walsh. (Above right) – Jim Mac fitting the hot platinum plugs as Sydney looks on. (Title photograph) – David Hooper and The Steyr Allard at the Valence hill climb in 1964.

03.12.12

Allard front suspension

A production Allard would not be an Allard unless it had independent front suspension, and apart from the last few Mk 2 Palm Beach, it would of course be a divided axle. There are two variants of the divided axle:  the angled and the parallel axes type. While Lesley Ballamy was responsible for converting the pre war Allard Specials to his version of the divided axle there was a fierce debate  between the Company and Lesley that the Allard design had infringed his patents – however those issues are not relevant in the context of this article.

Geometric genius
Apart from a small number of M & P types, all of the Allard J1 & J2, L, K1,and K 2 types were fitted with the angled (26 degree) axle beam & radius rod pivot design, while the Mk 1 Palm Beach, J2X, JR, K3, P2 all had parallel axes axle beam pivots. The parallel design (X type) does have a much improved steering geometry, and as a result an increase in cockpit leg room on the J2X. A Cadillac powered J2 road test by the UK Motor magazine in 1951 quoted a kerb weight of 2352 pounds with a front to rear weight distribution of 47/53. Currently a well known modified J2 with Chevy engine has a 50/50 weight distribution, while the Autosport road test of the T.T Chrysler J2X kerb weight was quoted as 2688 pounds, but no weight distribution figures were provided. The limiting factor in moving the engine forward was the position of the steering idler and its clearance to the engine sump. One of the other advantages was that it was possible to move the Marles steering box back well away from the exhaust system – an annoying problem on the J2 because the steering box was very close to the exhaust system.

Good enough to beat Fangio…
While many people voiced concerns on Sydney Allard’s insistence that the divided axle should be retained, his justification was that it was a simple independent system which was extremely robust and cost effective to produce. Clearly Sydney’s International racing performances, with a third place finish in the1950 Le Mans, winning the Monte Carlo Rally in 1952 and the 1949 RAC Hillclimb Championship supported his views which were summed up by – ‘I do not have a problem’. If the subject came up on a Friday afternoon after returning from a good lunch he had been known to relate the tale of passing Fangio in a single seater Cooper Bristol at an Easter Goodwood meeting!! I believe Sydney was well aware of the limitations of the divided axle but was concerned about the potential development costs to replace it. Also he probably always drove cars which were kept at a high level of maintenance avoiding the inherent problems which could happen with worn suspension components.

Wishbone development
I had designed an experimental McPherson strut system which was fitted to the works P2 Safari (MXA 555) and I believe it showed that it could be produced at a reasonable cost, and that there was the potential to replace the divided axle without major costs. Sydney was gaining experience driving a Ford Zephyr in International Rally’s and the manner in which the Ford Zephyr stood up to some fairly severe front end damage convinced him to explore use a modified
Ford Consul McPherson strut on the two seater Steyr engined sports car. It is often forgotten, or little known, that the Robert Clerk design for the light weight Hill Climb car was planned to be fitted with double wishbone front suspension design – The twin engined 4WD sprint car (upper left) of the late 50’s used some of the these components.

The MK2 Palm Beach (lower left) was shown at the 1956 Earls Court Motor Show with a further version of the McPherson front suspension, using for the first time laminated torsion bars to replace coil springs. Suspension loads were taken by a link from the strut to the torsion bar, however Sydney eventually conceded the steering was too heavy, and a revised design was implemented – the suspension loads were then taken through the lower wishbone, which is shown in the schematic drawing of the production layout.

Today’s derivatives 
Some 40-50 years on, this original wishbone layout has been further developed by American historic racer Bob Girvin in his Chrysler powered GT – the laminated torsion bar having been replaced with a more conventional rod design, along with roll bars fitted to both the front and rear suspension. If further orders had been placed  for the MK2 PB and GT models there were further developments planned with the external sliding portion of the strut working inside the pillar, and the rubber mounted trunion being replaced by a spherical joint. The availability of adjustable shock absorbers made it advantageous to use these along with roll bars to front and rear axles.The lower threaded swivel would be replaced by a commercially available suspension joint. Sadly the last two MK2 Palm Beach’s did not sell for nearly two years by which time the Company decided to discontinue car production.

The divided axles on both J2, J2X and JR have also undergone development from those who compete in serious historic racing, with static wheel camber being substantially reduced along with substantial reductions in axle movements. Also castor angles have been reduced to nearer  2 degrees. It is often forgotten that the Company only supplied a standard spring rate setting – no alternative one for racing (excluding the JR). With the advancements in spring design and manufacture it is now possible obtain stiffer springs at a relatively low cost – something that Sydney would have very much liked !

The Allard Farrallac

The Farrallac as it is today came about when Don Farrell decided he wanted to improve the performance of his J2, and first resolve the poor pick up performance of the American Carter carburettors. The Company had fitted four dual choke Solex P11 carburettors to the Rupert Larrinaga JR which Sydney occasionally shared with him at Prescott and Shelsley, and no doubt recognised how much better the Solex performed against the Carters. Don discovered that two sets of Solex manifolds had been made and that the second set was stored in the collection of Allard spares housed on the upper floors of the Head Office at Clapham High Street. At this time we were able to supply a ZF differential which had been designed to fit the Allard de Dion differential unit.

During my discussions with Don he said his aim was to build a lightweight tubular chassis and use both front and rear Allard axle assemblies, plus alter the car’s front to rear balance by moving the engine further forward. I had been engaged in an exercise to explore a replacement for the Allard front swing axle, and suggested to Don he might like to consider using this McPherson type front axle design. Having looked at this option I concluded that a better proposition would be to design a double wishbone arrangement, using as many Allard parts as possible. I agreed with Sydney Allard that I would assist Don with his rebuild as my own project and that the Company would supply materials plus manufacture any of the special parts required.

Front Suspension Design
The wishbones, which were slightly unequal in length being tubular construction with phosphor bronze pivots for the chassis mountings and fabricated mountings for both lower and upper suspension joints. Consideration was given to use aircraft style Rose joints however the cost was excessive. (Current ‘rod end’ joints were not available at that time!) The outer suspension joints were made by Automotive Products and were used on the Ford 400E light van. The uprights were a steel fabrication having the stub axle shafts bolted into the upright, allowing the Allard splined hub complete with bearings to be used. Also it was possible to retain both Allard brake drums and brake back plates.

The MK2 Palm Beach was designed to have an optional 11.7” diameter front disc brake arrangement, and it was agreed that the calliper mountings would be incorporated in the upright fabrication. This addition would allow disc brakes to be fitted at a later stage; however it would require the MK2 Palm Beach hub to be used in conjunction with Girling sourced discs. One of the design criteria was to achieve minimal bump steer by locating the steering arms at their optimum relationship to a Jaguar 140 rack and pinion steering box. The steering arms were Ford Zephyr being bolted to a machined ‘U’ section which was part of the upright fabrication.

Rear axle
The rear axle was incorporated with two modifications, the first being to use the more accurate peg and slider for lateral location, a system used on the prototype J2’s the production J2’s used the less accurate panhard rod which avoided an unacceptable level of wear and noise. Secondly the J2 de Dion radius rods were replaced by the Allard JR type twin parallel radius rods.

Tubular chassis frame
I explored two options with Don, the first being the twin tube design as used by Allard on Palm Beach and JR models or a simple ladder type chassis. Don opted for the simpler ladder type which was easier to construct and almost certainly lighter and more importantly torsionaly stiffer. Tube size was I believe 4” diameter and 12 SWG seamless cold drawn mild steel of the era.The chassis rebuild was carried out by Don Farrell in his workshop in the Edgware Road, Hendon being completed in early March 1958. With my agreement my layout drawings and sketches for the front suspension were retained by the Allard Motor Company.

Some 55 years on the Allard Farrallac is still a very competitive car owned and driven by Tony and Pia Bianchi having a best Silverstone Club circuit lap time of 1:06.4secs.

Title image courtesy of Patrick Garcia.