1966 BRM P75 3-litre H16
1966 BRM P75 3-litre H16: BRM’s staff at Bourne in Lincolnshire were given the green light to plan new cars for the 1966 3-litre Formula1. In charge of the effort was Tony Rudd, who had succeeded Peter Berthon as BRM’S technical director. Chief engine designer was Geoff Johnson. A 12-cylinder was considered, Rudd’s choice fell on a radical 16-cylinder engine composed of 2-horizontally opposed 8’s placed one atop the other and geared together. Estimates were that it would weigh 176kg and produce 500bhp – figures which, in the event were destined to be changed.
Rudd’s reasoning was that the H16 could use the proven and effective porting, valve gear, combustion chamber, pistons, con-rods and cylinder liner from the successful 1-1/2-litre V8. In the last V8 the inlet ports had been resited between the camshafts to flow down from the top of the cylinder head, so they were at the sides of the H16. The exhaust ports were at the top and bottom of the engine.
The initial plan was to place one flat-eight atop the other in such a way that they would share a single inlet camshaft along each side. However, a need for increased clearance between the crankshafts blocked this scheme. If a single inlet camshaft had performed double duty it would have forced an angle between the valve stems that was excessively large. In fact Rudd and Johnson then chose narrower valve angles that those of the V8, 23 degrees for the inlets and 29 degrees for the exhausts, which required a slightly larger bore to accommodate the valve heads. The H16’s cylinder dimensions were 69.85x48.89mm compared to 68.5x50.8mm for the V8. With this move the interchangeability of pistons and liners with the V8 was forfeited. And weight rose with 8 instead of 6 camshafts. Each side of the engine was capped by a single aluminum-alloy head – a fantastically intricate casting.
Weight rose further with the inclusion of some features in the bottom-end design that made the 16 easier to assemble. Akin to a nearly square box cast of LM8 aluminum alloy, the crankcase was split vertically 63.5mm to the left side of its centerline to permit both crankshafts to be installed in the right-hand or primary half with separate caps for the 5 57.2mm-main bearings, then the smaller secondary half of the crankcase was attached with separate through-bolts.
The 2 cranks were geared together by a separate pack of 3 19mm-wide spur gears at the rear of the engine, mounted in ball and roller bearings. The roller bearings were inset into the rear panel of the crankcase and the ball bearings were carried by a separate bolted-on magnesium housing. Each crankshaft was connected to its gear by a short splined torsion shaft made of EN40 steel.
The cranks were initially of simple flat 4-cylinder design with side by side 115.6mm con-rods on a common journal, located at 90 degrees to each other by the output gearing. 2-cylinders fired together in the upper 8-cylinders, then 2 in the bottom 8 90-degrees later. All 8 camshafts were driven from the nose of the lower crankshaft only the 2-stage gear packs originally created for the V16 drove the lower camshaft from their crank. The gear on the lower inlet camshaft drove the upper inlet cam and through an idler gear, the upper exhaust camshaft. This introduced a long gear train between the lower crank and the topmost camshaft and an even longer train from the upper crankshaft, which was linked to the lower shaft by the output gear pack.
Lucas transistor ignition trigger and its distributors for the whole engine were driven directly from the front of the upper crankshaft at half engine speed. 2 water pumps, each powered a completely independent circulation and radiator system for its heads and 2 fuel-injection metering units were driven from intermediate valve-train gears at front of the H16. The oil pressure and scavenge pumps were gear-driven from the lower crank. Small pumps driven by the lower exhaust camshafts scavenged the lower camboxes.
The new H16 was also to be supplied to team Lotus for its Grand Prix cars and in special 4.2-litre version for a 1966 Indianapolis effort. In the USA the engine displayed an awe-inspiring repertoire of failure modes. The cogged rubber drive belts to the Lucas fuel-metering units broke, as did the lower inlet camshaft where its drive gear was attached. Outrigger bearings were made for these, and flywheels were put on the drives to the metering units to damp their torsional vibrations, gear and center bearing failure in the output train between the crankshafts inevitably had expensive results, with the upper crankshaft’s valves being timed from the lower crank nose. Severe torsion vibrations affected the output gear train’s centre gear and bearings, which received twisting impacts from both crankshafts at alternating intervals and in opposed directions, the mass of the 4 of the crank counterweights was increased by 0.9kg apiece by bolting and welding a steel inertia ring to each one. This modification, crude though it was proved effective, and it was then possible to install the engine in a car.
It was also possible to obtain meaningful power figures for the first time. The highest output ever measured by BRM from its H16 was recorded in June 1966 from engine number 7504, a unit destined for delivery to Lotus. It returned 402bhp@10500rpm. In the first year of the 3-litre formula 1, that was a lot of horsepower. At Watkins Glen late in 1966 was engine number 7502, a heavily salvaged unit with metal plates and patches of Araldite holding its crankcase together. Giving 375bhp, this was rushed into service to win the US Grand Prix – the only victory ever credited to the H16.
Over the 1966-67 winter intensive development attention was given to the H16’s many problems. A new steel/steel material combination stopped excessive wear of the camshafts and tappets. Overheating was tackled by fitting a circulation system that prevented vapor locks in the complex heads from stalling the water pumps. Con-rods cracks were intercepted and cylinder-head studs failures were overcome, as was piston-ring breakage.
A major change allowed the H16 to run as a sequential 16, with one cylinder firing every 45 degrees of output shaft rotation. This required new crankshafts with 8 individual throws to give a firing order that allowed tuned scavenging exhaust pipes to be used. For these ‘8-pin’ mark-2 engine the diameter of the big-end journal was increased to 47.6mm to enlarge the area of overlap between the main and rod journals. New crankcase castings were also needed because the offset between facing banks of cylinders had to be increased from 0.65 to 1.05 inch to accommodate the required counterbalances, and new con-rods were required.
The weight of the mark-2 engine rose to no less than 236kg, but its output showed no advance on the typical level of 380bhp. In spite of all these improvements the mark-2 type 75 was still prone to occasional catastrophic failure. Valve breakage was identified as the culprit. The torsional flexibility allowed by the camshafts and their drive gears permitted the valve timing to drift enough to allow the valves to touch during overlap. Eventually a head broke off, usually the inlet valve. The piston hammered the head against the cylinder head, buckling the con-rod and shattering the cast-iron cylinder liner. Steel liners were fitted, they were batter able to contain the damage. A slight reduction in valve sizes and closer attention to valve timing cured this fault that had wrecked so many engines.
The last race entry by the BRM H16 was in the first event of 1968 South African GP, in which Mike Spence retired after 8 laps, as part of a complete BRM car, the engine had done no better during its career than a second place at Spa in the Belgian GP in 1967, with Jackie Stewart at the wheel.
After its initial tests the 4.2-litre Indy H16 engine posed so many problems that it had to be abandoned. It is an incredible but a documented fact of motor racing history that a single organization, BRM, suffered 2 total debacles in its flirtation with the undeniably-attractive 16-cylinder racing engine. The first 16 was build and run against all the financial odds. The second 16 was possible only because there were few constraints on finance. Both carried flaws of design and execution that kept them from developing their projected power and endowed them with a formidable capacity for self-destruction.
Stroke/bore ratio 0.70:1.
Compression ratio 10.5:1.
Con-Rod length 115.6mm.
Rod/crank radius ratio 4.7:1.
Main bearing journal 57.2mm.
Rod journal 47.6mm.
Inlet valve 39.7mm.
Exhaust valve 30.5mm.
Inlet pressure 1.0Atm.
Engine weight 236kg.
Peak power 380bhp@10500rpm.
Piston speed corrected 20.1 m/s.
Engine BHP/lite 126.8 BHP/litre.
Engine weight per BHP = 0.62kg/bhp.