Honda’s second F1 engine was the RA273 V12 of 1966-68, the opening years of the new 3-litre formula. It was remarkable in several ways, though it never received the attention it merited. Even today, only a few basic facts are known, and to my knowledge, no drawings have been made public, nor any technical description.
To start with, the angle between the banks of cylinders was 90 degrees, not the usual 60 degrees that remains the first choice even today (for very sound technical reasons). Another unusual feature was the output drive, taken from the middle of the crankshaft (between the third and fourth cylinders of either bank) rather than from its rear end.
At that time, only three recent racing engines had featured the exhaust ports between the cylinder banks, inside the V, and inlet ports outside (Repco, BRM and Ford Indy, each possibly for rather different reasons). The Honda RA273 was also arranged that way, for reasons that are not readily apparent.
According to the “History of the Grand Prix Car 1966-85”, the wide angle was chosen because the output shaft, running below the crankshaft, caused the entire engine to be raised up. The wide angle brought the cylinder heads lower down again, lowering the CoG in consequence, so as to compensate for the rise, at least in part.
But why take the drive from there, rather than from the usual place? For a transverse engine (Honda RA272, Bugatti T251, various motorcycles), it makes architectural sense, quite apart from any engineering reasons that might also apply. For a longitudinal engine, there must be pressing engineering demands that override the architectural disadvantages.
Perhaps the best precedent for this was the Mercedes Benz M196 Grand Prix engine of 1954-55. A straight-8, it too had the output, along with all auxiliary drives, taken from between the fourth and fifth cylinders. Although I have not seen an explanation for this, it is likely that it was to overcome torsional vibration, which tends to afflict long crankshafts. Two short (4-crank), shafts are around twice as good as one long, (8-crank) shaft, in this context. Even so, the M196 crankshaft has a torsional damper at each end; logically therefore, the full-length shaft was otherwise beyond help.
On the other hand, recent and contemporary racing and high performance V12s, from other makers, showed little or no evidence of such difficulties (Ferrari, Maserati, Eagle etc). But Honda had considerable experience of crankshafts driven, at very high RPM, by multiple small capacity cylinders, and had successfully overcome all the associated problems, torsional included. Is it likely, then, that the RA273 engine incorporated that experience, possibly as a precaution, rather than as the solution to an observed problem?
The 90-degree angle between the cylinder banks rather complicates the situation. Assume, in the absence of any data, that the crankshaft was arranged in the usual manner of the 60-degree V12, with six cranks, spaced at 120 degrees, each carrying one big end from each bank. With a “conventional” V12, this would give equal firing intervals of 60 degrees for all cylinders, resulting in the very smooth torque, even at low RPM, that V12s normally produce. But with that 90-degree angle, even firing is no longer possible, in fact it will occur at intervals of 30, 90, 30, 90…. degrees, which is a major deviation from 60, 60, 60…. degrees.
The engine torque output over one cycle (2 revolutions) will now exhibit six lofty and lumpy peaks, separated by six bumpy hollows. Even this might possibly be noticeable to the driver only at rather low RPM, but its effect on torsional vibration could be a very different matter. Impossible to tell without detailed analysis, but there is scope for trouble, as the magnitude of each harmonic of the cycle frequency will be radically different, and its effect on the crankshaft as a whole will change accordingly. (Of course, it could change for the better, not necessarily for the worse).
So the question is: did the 90-degree angle come first, and the centre drive later, or was the centre drive the cause of the 90-degree angle?
The exhaust ports within the V are not easy to account for, because they raise the CoG, already a problem as described above. Could it be that it was done so as to give enough space on either side of the engine for substantial “wheelbarrow arms” on the monocoque, (as also favoured by BRM for the P261), space that the exhaust system would otherwise intrude upon? Or to tuck the exhaust headers behind the drivers head, hopefully to minimise aerodynamic drag?