Atalanta Special

[When we first saw Major R. G. Shattock’s “Special” outside Atalanta’s works we mistook it for a post-war Atalanta. Later, coming upon it at the Phœnix Green Garage, we were shown the bonnet full of blown, sixteen-valve, twin-cam Brooke racing-marine-engine and the ingenious electric traramission, and there and then asked the owner for its story. Below we print that story — which is a far too modest account of how a miscellany of damaged or unfinished components were assembled in a home garage into the most modern high-performance, road-equipped special it has been our privilege to inspect for a very long time. — Ed.

Looking back to October, 1944, I sometimes wonder whether I should have undertaken the building of the “Special” if I had known exactly what I was up against! Of course, the work is by no means finished yet, but I think it has been well worth while.

It started with the offer of a 1 1/2-litre, twin o.h.c., sixteen-valve Brooke marine engine, which appeared to have possibilities, although the general condition was far from perfect; masses of parts were missing, but I decided to chance it, influenced, perhaps, by the almost-perfect No. 9 Cozette supercharger offered with the engine. The history of this engine is obscure, but so much is certain. It was hand-made by Brooke Marine of Oulton Broad about 1926 for speed-boat racing and was, I think, a contemporary of the 1 1/2-litre Sunbeam used for the same purpose. Brooke’s only made very few of these engines, starting with the unblown 16-valve engine reputed to develop about 65 b.h.p. at 4,000 r.p.m. This was followed by an 8-valve engine, basically the same and with most parts interchangeable, but with a counter-balanced crankshaft extended to drive, direct from the front, a Roots supercharger. This engine was reputed to develop 100 b.h.p. and to run up to 6,000 r.p.m.

The engine is definitely “dated” by its wide bearings and long, comparatively “spidery” crankshaft, with 1 5/8-in. dia. journals. The bore and stroke is 68 by 102 mm., giving a decidedly long bore/ stroke ratio of 1 to 1.5. The head is of cast iron, with four valves per cylinder at an included angle of approx. 80 deg. The block is detachable from head and crankcase and consists merely of top and bottom plates joined by the four bores and surrounded by the water jacket, with side inspection plates, all in cast iron. The water jacket sideplates are typically marine, but were later replaced by my own, heavily ribbed alloy type.

The crankcase is aluminium, as is the very robust sump with its heavy flange, which acts as the mounting for the engine. The three white metal main bearings are in bronze shells and two studs in each bearing run right through the crankcase and also hold the block, thereby adding considerably to the rigidity of the whole. The rods are very long H section, drilled and machined all over. Each cap is held by 4 bolts. The little ends are fully floating, each gudgeon-pin running in bronze bushes and having aluminium soft end-caps. The pistons are the original ones as far as I can discover. They are steeply pent-house in shape, have only 2 rings above the gudgeon pin and internal fins run down from the crown to the pin bosses; they are rather long but surprisingly light. To Specialloid’s credit these pistons have stood up to exceptionally hard conditions, including the melting of two exhaust valves, without trouble. Not bad, as they were designed for unblown engines in 1926!

The twin overhead camshafts run in three bearings, two plain and one ball, and are provided with a ball thrust-bearing in addition. They are driven by skew gears from a vertical shaft, driven by bevels from the front of the crankshaft. The valves run in huge guides and are operated by the cams direct, each being in two pieces, an adjustable tappet post screwing into the hollow valve stem, and being locked automatically by the double-valve springs themselves. This adjustment has proved very difficult in spite of extremely weak valve springs. The valves are consequently fairly heavy and how these springs prevent bounce I shall never know! At its base the vertical drive passes through a box-shape casing which dispenses right-angle drives to magneto, oil pumps (near side), water pump (off side) and rev.-counter; the latter I have not used, instead taking a drive from the rear of the near-side camshaft to get a shorter cable length.

The magneto runs at engine-speed and is a Scintilla Polar Inductor type, firing a Lodge plug in each head; these plugs are most satisfactory. The oil-pumps operate a dry-sump system. The oil and water pumps are driven at only about a quarter engine-speed.

This, then, was the engine plus a layer or two of paint outside and a coating of congealed oil-cum-sea-water sludge, minus sundry caps and covers, exhaust manifold, flywheel and divers accessories. It was delivered to my garage (let me stress, an ordinary private-house garage of quite inadequate size) by a certain bearded type in a Bentley. Although in the Army at the time, I was lucky enough to be stationed near home and with spare time in the evenings.

I looked around for a suitable chassis on which to start operating and visualised picking up a more or less complete chassis, discarding the engine and fitting the Brooke, finally building a body — a not inconsiderable but perfectly feasible task. I chanced to be passing the Atalanta works in Staines one day and enquired had they a chassis. I was much heartened by the sight of two or three frames standing in a corner, by close inspection of a V12 Atalanta in the showroom and by a promise that they would supply all bits possible towards a chassis, but which, they warned me, would be incomplete and in many cases unmachined. I cannot say enough to praise the generosity of Atalantas, who, for apparently no reason, showered parts on me, a complete stranger.

In those days the drilling of a hole was quite an obstacle and when the frame and pile of bits were landed on the earthen floor of my garage, the awful truth of what lay ahead began to dawn. A period of feverish activity started during which the project seemed to advance not at all. The garden tools, and odd bits of furniture from a recent move were in the way, so perforce a garden shed had to be built. I gathered my aged and assorted hand tools together from their scattered wartime retreats, realised how few I had, and with an anxious glance at the bank statement, plunged for a portable electric drill as the first essential. The lounge was littered with components as I wrestled on the drawing board with wheelbase, frame, engine layouts and so on and so forth. Days and weeks went by and the jig-saw was insoluble for the good reason that so many parts were missing. I could not settle down to any one job and complete it. I arrived wild-eyed at work each day and the Pundits shook their heads sadly. Spectators came and went and I could see that they thought I should eventually be led away muttering about understeer and chasing x with a slide rule, the “Special” still a mere heap of parts!

It became quite clear that I must make a plan of campaign and stick to it absolutely. In those wartime days there was no “basic” petrol (do I hear hollow laughter?), so the search for parts, tools, people willing to help with machining, even finding nuts and bolts, was a formidable task in itself. Money, too, was definitely limited, so the accumulation of vital equipment took a long time and was acquired in this order: Portable electric drill, 3-in, motor-driven bench lathe, grinder ditto, gas-welding kit, together with hand tools from time to time, as required.

The plan was to organise the garage and then strip the engine to find out what major components would need replacing. No other work of any sort was to be done on the engine at this stage.

Drawings of general arrangement were done roughly, to enable chassis mods. to be carried out. Drawings of all details were to be done when necessary to enable suspension, steering and brakes to be put in hand. The chassis was to be set upon its wheels, but only roughly assembled. The frame was then to be drilled, cleaned, painted and its components finished. Then back to the drawing board for work on the engine layout and design of missing components, mountings, etc. Next, the engine and accessories were to be completed in the rough, followed by bulkheads, floor-pans, instruments, controls, under-pans, seats, temporary wings, etc. Finally, the engine was to be stripped down for final work and adjustments, etc., so that the moment the engine was running it would be ready to take for a test run. Bodywork was to be left until the end and completed only when the chassis had passed preliminary tests.

The object of the plan was to ensure that when the most interesting stage was reached, i.e., the engine running, no work of major consequence would remain to be done, thus obviating a tendency for “skimping” due to the desire to get on the road. I kept to it almost to the letter, although I fear that there are still many improvements to be made.

The chassis was shortened by 14 in., bringing the wheelbase to 8 ft. 10 in., the track remaining at 4 ft. 5 in. front, 4 ft. 6 in. rear. The chassis is of box-section, its side-members deepening in the centre to 7 in., with 4 in. depth front and rear; the gauge is rather heavy and, with a box-section cruciform, five cross-tubes and a cross-channel, gives the utmost rigidity.

The original Atalanta had 16-in. dia. brake drums, with both drums and backplates of light alloy. I chose a set of Girling brakes with cast-iron drums and backplates of 11 in. effective diameter, because some trouble had been reported with the larger diameter owing to high surface speeds, while this size limited me to a minimum wheel diameter of 18 in. and I wanted to be able to fit a 16-in, or 17-in. wheel. Also, the sets of alloy brakes were unmachined and getting liners shrunk in, in those days, would have been a formidable business. In any case 11 in. diameter drums would give very good braking area considering that the weight of the car was to be considerably less than standard.

The task of modifying the 5-stud-fixing brake drums and brake plates to suit Rudge-type hubs was quite tricky but was greatly assisted by Girlings, who supplied drawings and much helpful advice. The hubs I had to machine on my 3-in. lathe (accompanied by much holding of breath). The front drums and brake-plates have 1 in. dia. holes drilled in them, covered by perforated discs in the case of the former and a copper gauze in the latter. The revolving drum, having the larger total opening, tends to draw air through the brakes. The rod operation had to be changed to cable in order to cope with independent suspension. Front and rear cables run to compensating devices located on the chassis near the wheels, and from there rods connect to foot and handbrake levers, thus keeping the cables as short as possible.

The rear axle housings had to be modified to give correct set out for the back-plates and then steel sleeves shrunk into the light alloy suspension arms had to be removed and machined on the 3-in. lathe! If anyone sees my wheels describing an eccentric path they will now know why!

The Atalanta rear suspension consists of horizontal coil springs laid inside the frame. A high-tensile rod runs forward from the suspension arm, through the main cross-tube and through the coil spring itself, and picks up by means of split taper collets on the forward end of the spring. The rod itself is supported in a rubber-mounted alloy bearing on the frame cruciform. As a wheel rises the spring is compressed against its rear seat on the main cross-tube. The Atalanta front suspension is Porsche in principle, with parallel trailing-links sprung on vertical coil springs received in housings at the cross-tube from which the front of the chassis is hung on light stays. The steel trailing-links are 7 in. long and are pivoted from the front main cross-tube on hardened steel sleeves and lead-bronze bushes. These carry the stub axle carrier, on which the spring platform located. The carrier is steel, machined from the solid, being well proportioned and hollow, and to it are welded 1/8-in. thick plates forming the platform and link-pin bearers.

Armstrong double-piston hydraulic shock-absorbers are used front and rear, and so far have worked well.

Suspension and wheel bearings completed, the frame at last stood upon four wheels — quite a moment, and some little time was lost in admiration.

A Bishop Cam steering-box was then mounted on the top side of the suitably reinforced frame, and held in a very simple bronze clamp. A long drop arm and push-pull rod connects to a central cranked lever mounted on the underside of the main front cross-tube. This cranked lever is suitably braced by two tubes in tension running up to either side of the frame. Two equal length track-rods run to the steering levers. These steering levers are cranked downwards sharply to prevent the rods hitting the underside of the frame in “bump.”

The “steering wheel” was for a long time an adjustable spanner, superseded later by an 18 in. dia., 4-spring-spoke wheel. The covering was stripped from the spokes, which were decoratively drilled; or dare I say it was for lightness?

The differential was next tackled; this was acquired in a broken casing and had to be rebuilt into a new casing which was incomplete and not fully machined. The gears are E.N.V. spiral-bevel crown and pinion, with the differential assembly held in taper-roller bearings and the halfshafts, with flanged ends, spigoted one into the other and allowed to slide fractionally along their axis to take the slight movement caused by the rise and fall of the i.s. wheels. The whole assembly is held in a light alloy casing mounted on two cross-tubes in “Silentbloc” bushes. The drive to the wheels is by a pair of Hardy-Spicer shafts universally-jointed at each end. The wheel-shafts are not floating, and run in two large taper rollers, the hubs being tapered and keyed to the shaft itself.

This point in construction was reached after about six months and marked the end of stage 1 of my plan. All was stripped, cleaned, painted, etc., and final assembly completed, including polishing all steering parts.

The engine layout was complicated by the fact that there was no enclosed-flywheel housing on which to bolt a gearbox. Indeed, there was no flywheel, no starter motor, no dynamo and no blower drive, so I was starting from scratch. After much head-scratching and many “offerings up,” I came to the conclusion that there was no other way than to make a sub-frame to which would be bolted the engine and all its attachments. This meant more weight as the frame was obviously going to be stressed quite highly and it finished up at about 30 lb. or so.

This sub-frame consists of two channels running parallel for the length of the engine; to the rear end is welded a 1/4-in. thick plate and to the front end another plate, 3/16-in. thick, both plates being suitably braced. The rear plate carries the gearbox and also supports the engine, being extended right out to the chassis side-members to withstand torque reaction. The front plate carries a modified Riley Six dynamo, driven off an extension of the crankshaft via a “Layrub” coupling. Vertically above, in the same plane, is the Cozette super charger, mounted on a swivel plate for belt-tensioning — crude but effective, and less trouble than a jockey-tensioner. The drive is by two 1/2-in. Vee-belts, and the pulleys were machined from a 5-in. dural billet. The driving pulley acts as the front spider for the “Layrub” and takes the load in torsion off the dynamo shaft, but the dynamo shaft takes the radial load of the blower belts, through a substantial ball-race in the dynamo.

The Bugatti water pump is mounted on the sub-frame and driven through a rubber coupling. The starter motor was a problem because the flywheel size was limited by the width between the sub-frame channels. With a little carving of the webs of the crankcase a Lucas starter from an Austin Ten, mounted on a bracket on the sub-frame, came close enough to the flywheel, which is ex-2-litre Atalanta, machined to as light a disc as possible, the starter teeth being machined on the flywheel itself.

The gearbox, also from Atalanta’s, is a Series D.I. Cotal electro-magnetic epicyclic, driven by a three-point universal joint made up from Borg & Beck clutch plates. Borg & Beck were most helpful and sent me some of their blank discs from which I made the joint. This allows for slight misalignment axially and also allows spring in the drive via the usual B. & B. cushion plate. The selector-lever extends to the left below the steering wheel and a small central lever controls reverse; a tumbler safety switch is fitted on the dash. No clutch is used, as the clutch pedal actuates an electrical resistance which progressively steps up the voltage to the box, allowing its discs to act as clutch plates. The box weighs 112 lb., of which nearly 100 lb. revolves; it is beautifully made.

The making of the sub-frame presented quite a problem. I had no welding kit in those days, and so had to get someone to bring his plant round. The end plates had to be kept at right angles to the axis of the crankshaft, and the channels had to be kept flat, and, as anyone who welds knows, the distortion is sometimes embarrassing! When all was welded up we found that the front plate was nearly an inch out as the channels had pulled up badly in welding in the bracing pieces. By means of a precarious system of beams, blocks and struts, we converted the Morrison shelter I use for a bench into a hydraulic press, using a hydraulic jack as the prime mover. Even using heat on the affected zone it was touch and go whether I straightened the frame or bent the bench. We won in the end, after some anxious moments. Having got the frame within reasonable limits I had to find someone to machine it for me.

No one without a jig-borer would do it and it seemed that everyone with a jig-borer had more urgent jobs on hand. I eventually struck gold in the shape of J. L. Jameson, of Ewell, an enthusiast himself, who not only undertook the job but refused to discuss remuneration.

To ensure correct alignment the crankcase was dowelled to the sub-frame and the whole frame bored in line with the main bearings for the spigots of the gearbox and dynamo, the surfaces being faced up true at the same time.

The original input shaft of the Cotal box was too long and a new one had to be made, a job undertaken by Sykes Gears Ltd., of Staines, who also finished machining the extension to the crankshaft for me. Later they cut a new oil-pump drive gear, the original having been badly worn owing to misalignment of the oil pumps. The inlet manifold was on the engine and consisted of eight 1 in. inside dia. stub pipes running out from the ports into a 1 1/2-in. dia. pipe, with the inlet flange at the front. This I altered to bring the inlet into the centre from above, and fitted the blow-off valve on the elbow so formed.

The exhaust manifold I made up from scratch, using a 3/16-in. thick flange plate with 1 1/8-in. dia. tubes welded to the plate, the front one being only 2 in. long, the rear one about 14 in. long. These join up to the main pipe, which tapers from 1 1/8 in. to 1 3/4 in., being designed to give as near a constant gas flow as possible. The whole is in 18 gauge stainless steel and I had the most frightful difficulty in preventing everything from twisting hopelessly out of shape while welding. The exhaust is taken out through the bonnet and down into a Burgess silencer along the outside of the frame, thence to the rear via a duralumin tailpipe. The oil tank is below the radiator in front of the main cross-tube, oil being drawn via a flexible pipe to the pressure oil pump and fed through a Tecalemit full-flow filter to the various points of the engine through external pipes. Both pressure and scavenge pumps are kept primed by bringing the oil pipes above the level of the pumps themselves. Compression ratio was unaltered, and is a little above 5 1/2 to 1, and blower pressure reaches 12 1/2 lb./sq. in. Petrol/ benzole has been used up to now. An air-cleaner is used, to preserve the blower.

I have discussed the whole engine layout as one but, in fact, I constructed the body hoops, instrument panel, tanks, floor-pans, etc., during this period, while waiting for various parts to be made. One petrol tank at the rear is of conventional shape and the other is only five inches deep and slung below the seats, the two being connected together and filled through the rear tank. A hand pressure-pump is used to lift the petrol to the carburetter. The rear brake-rod runs through a tube in the midship petrol tank. Total capacity is about 19 gallons.

The spare-wheel mounting is held on a duralumin channel which rises from the rear cross-tube and runs over the differential casing, being supported by an inverted-V bracket which spans the propeller-shaft.

This channel also supports the floor of the luggage and tool space and holds the hand pressure-pump, which projects forward between the seats. The central floor pan is constructed of duralumin sheet, using the prop.-shaft tunnel as a stressed backbone to support the two seats, which are both of duralumin, one out of an aircraft and one made up on similar lines. The scuttle hoop is of rivetted duralumin sections and only weighs 5 lb. The dash panel is stressed and carries a corrected rev.-counter, boost, air and oil pressure gauges, three thermometers, ammeter, and the usual switches. A drilled advance and retard lever is mounted above the steering wheel.

The remaining floor pans are all duralumin, as are bulkhead, body hoops and under-pans. The controls, pipes, etc., are designed as far as possible for speedy removal and the bulkhead can be lifted off intact in a comparatively short time, to enable the engine to be lifted bodily out on its sub-frame. The “clutch” foot-controlled variable-resistance is mounted on the bulkhead and covered by a “hush-hush”-looking black box, both to fox spectators and to keep out the dust.

This brings me to the end of stage 2, completed in early December, 1946, about two years after commencing work. All that remained to be done was the strip and rebuild of the engine and final fitting of accessories. This proceeded apace. Boxing Day came and the engine was all buttoned up, and all that was left was the final strip and cleaning of the Cozette supercharger. This occupied the whole of the night until 4 a.m., and to the consternation of the household was carried out on the kitchen table, partly for cleanliness (of the blower!) and partly owing to the sensitivity of my immediate neighbours.

“D”-day was spent in last minute fitting of manifolds and controls, rigging up a test petrol tank, etc. Tea was brought out on a tray, and between mouthfuls of cake the essential wiring was applied, oil and water inserted, and vital fluid introduced to the carburetter. A last check over was prolonged by the conviction that something must be wrong, and the feeling of fatality once the starter button was pushed.

Unable to postpone the moment any longer I inserted myself into the seat with fingers crossed and, not daring to breathe, pressed the button! We were greeted by only the barest cough. My friend said, “What about choking it?” in a rather thin voice. I nodded, speechless, and again pressed the button — to be greeted by a roar that nearly lifted the roof off the garage. A runner at last!

I immediately decided to tax the car, greatly hindering stage 4 of construction, i.e., completion of the body! We weighed the chassis as it then was — 16 cwt. — and made a brief test while doing so. Obviously there was ample urge. Bad weather now intervened, the garage all but flooded, and it was the end of March before the under-pans were finished and temporary wings fitted. Some preliminary tests gave 0-60 m.p.h. in 10.2 sec., the average being 12.5 sec. Alas, on a 60-mile run the water jacket cracked, but not before 88 m.p.h. was logged. The block was welded, engine re-assembled. The car was then taken to C. A. Tanner, of Wood Lane, who made me a one-piece, dish-cover bonnet. The cockpit I constructed myself. Tanner later made the tail and rear wings; front wings, painting, etc. still remain to be finished. I had by now realised that the roadholding was exceptional, steering a little low-geared for Prescott but excellent on the road, and the brakes good, although they could have been better. The slight oversteering tendency of the rear suspension had been cured by slightly wider rear than front track, considerable front-wheel camber and weight well forward as well as over the driving wheels.

On the next run the block cracked an inch above the former crack. This was re-welded but sea-water corrosion had taken its toll and a crack opened up in yet a third place. Going down to spectate at the first 1947 Prescott meeting I was continually troubled by the loss of water and finally the gasket blew. I decided to have the crack metal-sprayed but in the preceding shot-blasting process the bores gave way and ruined the block.

Fortunately I was able to buy the 8-valve engine from Brooke’s. I reassembled my engine using this newly-found block, but on test it cracked in the same place as had the first one. “Wonderweld” assisted in slowing up the leak but did not entirely cure it. The cooling system, I should explain, uses a small Atalanta radiator, sealed to 4 lb./ sq. in. pressure. This was rebuilt for me by Gallays, who also assisted me with drawings of the supercharger and many spare jets for the carburetter. The Bugatti pump is efficient enough but there is so little head of water that loss of only 3 or 4 pints causes trouble. Moreover, the engine ran abnormally hot, and the outlet from the head was by a single 1-in.-dia. pipe on the rear face, below the highest point. In a boat with cold sea-water pumped straight through this was probably sufficient, but in the car steam pockets formed. To overcome this I led off six 3/8-in.-dia. pipes opposite the exhaust valves, and three opposite the inlet valves, gathered into a manifold, from which a pipe at the rear led along the off side of the engine, lagged where it passed above the exhaust pipes. The old outlet was blanked off. This improved matters but as the leak continually reduced the meagre waterhead, I was always more or less in trouble.

Next, one of the forward/reverse planet pinions in the gearbox broke up and jammed the transmission solid. I stripped the box and obtained University Motors’ only spare pinion. It proved hopeless to attempt to re-assemble the gear trains without first mathematically calculating the correct settings and later I was told that Cotals regard the job as quite impossible without a special collapsible jig!

The Atalanta was entered for Gransden, the tail and rear wings being hastily added. In practice the exhaust valves in Nos. 1 and 2 cylinders burnt out, due to the continual overheating. We were towed home behind Buckler’s Jaguar, and with five days to the July Prescott, Laystall’s sportingly made up eight new KE 965 valves. The engine was reassembled right through the Friday night. We clocked 57.74 sec. on the hill, taking third place in our class.

On the run home I was “pushed along” by a Jaguar 3 1/2 “100” and attained 94-95 m.p.h., with something in hand, the LEAK ever-present in my mind. With 4.25 to 1 axle ratio and 6.50 by 16 tyres, 5,000 r.p.m. in top equals 99 m.p.h. I have had 5,500 r.p.m. by mistake in 1st gear and 5,000-5,200 several times in 2nd and 3rd, but never more than 4,700 in top. At Great Auclum we managed a second in the class and at the V.S.C.C. Prescott reduced the time to 56.63 to get another second. At the Chichester Speed Trial I tried a 3.6 to 1 axle ratio, but this was altogether too high, and the get-away on the clutchless Cotal was very tricky. The best I could do was 36.3 sec., losing over a second at the start but nevertheless 3rd in the class. On the way home the old trouble grew worse and water consumption was 4 m.p.g. Consequently I have now laid the car up until my new block is cast. I shall then reassemble it with the 8-valve head and the counterbalanced crankshaft. I shall fit a header tank up above the rear of the engine, its filler protruding through the scuttle and completely seal the radiator, letting the pump feed up through the block and directly out of the head to this new tank. The new engine should give more power, and I look forward to 120 b.h.p., with a weight of 16 cwt. by further mods., drilling of the chassis, etc. (the car, less occupants, at present weighs about 17 1/2 cwt.), a bit heavy for sprint events. Final details completed, I think I shall have a very desirable motorcar, possessing exceptional performance, very good roadholding and the quick-change Cotal transmission. In connection with the last-named, I have a scheme to assist the box during sprint get-aways — I expect everyone will tell me it won’t work, just as they did of my clutchless layout. With what amounts to a complete rebuild ahead of me I wonder will I ever get adequate sleep again!