The Evolution of Modern Small Car Engines
(Continued from the May issue)
Standard-Triumph. The second paper at last year's I.A.E. Automotive Division Symposium was delivered by D. C. Eley, New Projects Engineer, Standard-Triumph Ltd., and dealt with the conception and design detail's of their 803-c.c. engine, its redesign to 948 c.c., the design of the twin-carburetter version and a further redesign to 1,147 c.c., in both single and twin-carburetter versions.
Before engine design was started a careful study was made of the vehicle as a whole and, this decided upon, a power-required curve was drawn, based on frontal area, aerodynamic coefficient, weights, tyre sizes, etc. This determined performance characteristics with the most suitable gearing.
Mr. Eley commenced his paper by outlining his preference for a water-cooled engine, after enumerating the many known advantages of air-cooling—and adding another, viz, that such engines are undoubtedly attractive "in territory where water may not be readily obtainable"!
The point was made that there are few examples of air-cooled units with four or more cylinders in present production. Standard-Triumph went for water-cooling as giving much greater refinement from very much better sound absorption, improved fuel economy on account of a small fan absorbing considerably less power, and simple, smell-free interior heating. The greater power outputs obtainable were regarded as of no moment when dealing with normal mass-production engines.
Two-stroke engines were apparently considered but "those available at the time" did not compare with four-stroke units and noise would have been a big problem. Four cylinders instead of two were chosen for the obvious improvement in refinement, and an in-line arrangement was considered preferable to a vee-four or flat-four as it was cheaper to make and there was ample space between the front wheels to accommodate it. The flat-four was regarded as taking longer to assemble than an in-line engine, having more parts, and as the labour force on the assembly lines cannot be reduced, as in the machine shops, by transfer machinery, this increases cost. The lengthy pipe-work and hot-spot complications were also against the flat-four.
It was realised that an o.h.v. engine was essential to get sufficient power from an economical size of package but to off-set the considerably greater machine-tool and assembly costs vertical valves in bath-tub chambers were chosen, instead of inclined valves in a wedge-shaped chamber.
To obtain a genuine 60 m.p.h., 23 b.h.p. at the road wheels was required, or a test-bed figure of 26 b.h.p. With the fuel available at the time, a 7-to-1 c.r. and Standard-Triumph's knowledge of the type of combustion chamber used, an estimated power curve showed 26. b.h.p. at 4,500 r.p.m. from 800 c.c.
Bore and stroke were dictated by the desirability of using an existing set of new and expensive cylinder boring and honing machinery of the fixed-centres multi-spindle type. A bore of 58 mm., with water all round the barrels, was a practical production casting and, with a stroke of 76 mm., gave a swept volume of 803 c.c. The stroke/bore ratio of 1.3 to 1 was high by modern standards but enabled adequate size valves to be used.
It was known that the design would remain in production for an extended period and the engine was deliberately made robust, particularly the crankshaft, rods and block. Careful consideration was given to accessibility of components, minimum number of parts, simple machining and reduction in cost. All auxiliaries were put on the camshaft side, leaving the other side free for manifolds, carburetter, air cleaner and silencer.
The block was of B.S. 1452/17 cast-iron, extended below the crankshaft, because with the split line on the crankshaft centre-line the only way of obtaining the required block stiffness and gearbox mounting would have been to use the more expensive and vulnerable cast sump. The extended block also allowed for bridge-pieces at front and rear over the bearing caps, thus presenting a continuous flat surface for a simple sump pressing.
The original design allowed for integral front and rear flanges to eliminate pressed engine mountings. This improved beam strength of the block/gearbox unit, although weight and casting prices increased. Unfortunately the almost new foundry moulding boxes were not large enough to accept flanges and rather than waste many thousands of pounds, pressed mounting plates were adopted.
The design catered for a 3-bearing crankshaft, the bearing caps located by a tenon in the block and 7/16-in. bolts with spring washers. The bolt centres varied one side to the other to ensure correct assembly of the caps.
The crankshaft was of manganese-molybdenum En. 16T with 277-311 Brinell hardness. The camshaft was of chilled cast-iron running direct in the block, driven by a 3/8 in.-pitch endless chain, of the Simplex type, tensioned by the Company's patented spring-steel blade with chromium-plated surface. Bucket tappets were used, as giving a better bearing extending nearer the nose of the cam than a mushroom tappet and because they can be fitted and removed with the camshaft in position.
Rather unusual, the skew-gears for the distributor and oil-pump were both of cast-iron, but by providing good lubrication and using a 46° helix angle on the driving gear, ensuring correct gear proportions and surface treatment, copper-plating the driven gear and giving the shaft a Bonderite and Parkerizing treatment before dipping it in a Dag compound, completely satisfactory operation was achieved.
The oil pump was positioned above the sump tray and oil level to enable various sump shapes to be used, and is of internal and external rotor type. Due to its high level, oil was not fed back to the suction side. An angled con.-rod enabled withdrawal through the bore but necessitated accurate location of cap to rod by press-fitted hollow dowels in the cap engaging reamed holes in the rod. Strength for strength, this design is heavier than a straight-split rod. The pistons had two compression rings of 4K6 and a slotted scraper ring to D/26 proportions, of DTD 485.
It was decided that siamesed inlet ports would suffice, and save cost by lower weight, simpler castings and less machining, but the expense of a water distribution tube in the head was considered technically justified, water being diverted on plug bosses, valve-seat areas and down into the block. Additional water holes between head and block gave freedom from steam pockets. Normal cast-iron is considered quite suitable, even for tuned engines, providing even sections can be maintained and good directed water flow is provided.
A combined spindle and bearing water pump having been "a most unfortunate and expensive experience on a larger engine unit," separate ball bearings with a seal on their outer ends running on a stainless-steel spindle to which was attached the impeller, was specified. Push-rods were of steel wire of En 8R and hardened to VPN 600 to 750. The cast-iron B.S. 1452/7 rockers had a ratio of 1.5 to 1 and pads chilled to Rockwell C.48 Min. Cast-iron pedestals were later replaced by pressure die-cast alloy ones. Inlet valves were of Silchrome (En. 52) and exhaust valves of XB (En. 59), both ends flame hardened and the seats at 45°. Eleven 3/8-in. manganese-molybdenum steel (En. 16T) head studs were used; the original c. and a. gasket was replaced by a corrugated steel sheet gasket.
A Solex carburetter fed through a single cast-iron manifold unit. The engine gave its designed power, and 470 lb. in. torque at 2,500 r.p.m., on 74-octane fuel. The Standard Eight went into production in June 1953. Service complaints led to the exhaust pipe attachment being altered from 2- to 3-stud with cheaper steel in place of asbestos washer, spigot protected, to obviate leaks, and flywheel ring hardness had to be reduced from VPN (30kg.) 625-725 to VPN (10 kg.) 525-625 to bring the starter pinion life up to that of the ring.
When the Sales Dept. called for more performance a new cylinder block casting with 63 mm. bore gave a capacity of 948 c.c. Inlet valve throat diameter was increased from 0.94 to 1.06 in., exhaust valve diameter from 0.88 to 0.94 in., lift being unchanged. Some 30 b.h.p. at 4,500 r.p.m. was developed, with 555 lb./in. torque, equal to 120 lb./in.2 b.m.e.p., lifting car speed from 60/61 to 66/67 m.p.h.
To cater for enthusiasts, a twin-carburetter kit was introduced. An aluminium inlet manifold matched two 1 1/8-in. S.U. carburetters but was very short to enable the rear carburetter to clear the clutch master cylinder, and no air cleaning or silencing was possible. With 8.5-to-1 c.r., high-lift wider-overlap timing (lift, 0.312 in.; timing 18-18-58-58), double valve springs with special collars and cotters, and a separate iron exhaust manifold with free-flowing tracts of fair length, 46 b.h.p. was produced at 5,700 r.p.m. When the 948-c.c. engine went into production in May 1954, the head of the 803-c.c. engine was altered to allow a 1-in.-dia. inlet valve and port, giving a small power increase throughout the range. In April 1956 the steel gasket put the c.r. up to 7.5 to 1, the older c. and a. gasket giving a c.r. of 7 to 1.
With the advent of premium petrol in August 1957 the inlet valves and ports of the 803-c.c. unit were increased to 1.06 in. and the exhaust valve to 0.9375 in., using 948-c.c. valves, the c.r. was put up to 8.25 to 1, 30½ b.h.p. being realised at 5,000 r.p.m., with 510 lb./in. maximum torque. The 948-c.c. engine was given a 28-mm. instead of a 26-mm. carburetter, 12-12-52-52 timing instead of 10-10-50-50, and the c.r. put up to 8.0 to 1, with 0.281 in. valve lift. 36 b.h.p. was developed at 5,000 r.p.m., with 610 lb./in. torque. These engines were designated the "Gold Star" power units.
When the Triumph Herald went into production it had the 948-c.c. engine, giving a top speed of 70/71 m.p.h. As the major components of the engine were made on fully automated machinery and the crankshaft machinery was highly specified, no change in stroke was possible when more performance was sought. So the bore was increased to 69.3 mm., to give 1,147 c.c. This satisfactorily met the performance requirements with a higher axle ratio, but the increased centres and bores interfered with head studs and transfer line locations on the r.h. side of block and head. The problem was solved by moving the bore centre-line 5/32 in. towards the camshaft, giving a désaxé condition. In an attempt to use the existing crankshaft, con.-rods with 5/32 in. offset to the big-end were designed. Alas, tests showed that the high-loading on one end of the big-end bearings produced breakdown of the surface and heavy wear of the crankpin, and there was evidence of lack of crankshaft stiffness.
A new design, eliminating big-end offset by moving the webs 1/16 in. from the cylinders, using copper-lead bearings and stiffening the crankshaft with a "flying web" of greater section, was prepared. When redesigning the con.-rod was made so that big- and little-ends could be machined to width at the same time, and subsequent operations eased because the surfaces are at the same height from the location point of view—an immense gain from the production angle.
Mr. Eley felt that full-flow oil filtration was essential with copper-lead bearings but the extra modifications to transfer plant and the added expense caused the experimental engines to be built with by-pass filters. Tests showed that if an engine could be built with the highest degree of cleanliness, bearings and crankshaft were entirely satisfactory. But in practice full-flow filtration was proved to be essential.
For the new 1,147-c.c. engine an up-to-date piston-ring layout was adopted, with DTD 485 top ring chromium-plated, second 4K6 ring taper faced and a DTD 485 slotted scraper ring all to D/24 proportions. Valve size went up to 1.1875 in. inlet, 1.031 in. exhaust, combustion chambers were re-shaped to suit the new bore size and a 30-mm. carburetter fitted. Output was 41 b.h.p. at 4,600 r.p.m., with 730 lb./in. torque at 2,400 r.p.m.
Final testing involved day-and-night running, alternating between motorways and cross-country routes, and even when the 1,147-c.c. engine had outlasted three 948-c.c. engines, it was still operating perfectly satisfactorily.
A twin-carburetter kit retaining the H.1 1 1/8-in. S.U.s of the 948-CC. kit but a high-lift 18-18-58-58 camshaft, gave some 56 b.h.p. at 5,700 r.p.m. on an 8.5-to-1 c.r. An example of the value of developing honed-up versions of ordinary engines was provided when development of this and subsequent twin-carburetter engines caused oil leakage from the rear crankshaft seal at high r.p.m. Reducing clearance between crankshaft and aluminium seal gave a considerable improvement but temperature checks showed greater running than static clearance to be to blame, and a full cure came with the introduction of cast-iron housings, first for the high-speed units, then for the whole engine range.
For the Triumph Spitfire a new inlet manifold was designed to take twin HS2 1¼-in. S.U.s, and individual pancake air cleaners enabled the length of the inlet manifold tract to be increased. A cast-iron exhaust manifold was retained, of better shape. With the high-lift camshaft and 9-to-1 c.r., 63 b.h.p. at 5,800 r.p.m. was developed, with 780 lb./in. torque at 3,500 r.p.m., giving some 93/94 m.p.h. The modest c.r. and timing were chosen after considerable test work, as they give a tractable engine with low idling speed. The exhaust valves are of 21/4NS. The header tank brackets and petrol pipe clips gave trouble until rubber-mounted.
A de luxe Herald model, the 12/50, was evolved using a free-flow exhaust manifold and the high-lift camshaft, and a 30-mm. d/d. carburetter of the strangler type. The front exhaust pipe was increased to match the manifold and an 8.5-to-1 c.r. used. The engine now gave 50 b.h.p. at 5,300 r.p.m., giving the Herald saloon a top speed of 80 m.p.h., and this direct descendant of the Standard Eight power unit delivers twice the power of the original engine, in comparative, single-carburetter form.—W. B.
(To be continued)