This is an article that
SteveD CB500F scanned for us. The article from 1977, by
LJK Setright, appeared in the English magazine Bike and discusses the differences in the engine design between the venerable CB400/4 and the CB400 twin that replaced it in the 1978 model year.
Two, Four Or Five ?The name of Soichiro Irimajiri is not very well known here; indeed, he says that it is not even a name that is easy to pronounce, although I thought he did it rather well. It is the name of the director of Honda's Research and Development division, the name of the man who designed the racing (and championship winning) 50cc twin, 125cc five cylinder, and 250cc six-cylinder motorcycle engines, not to mention the 90 degree V12 3 litre Grand Prix car engine. On the strength of that background, Mr Irimajiri must be reckoned one of the most accomplished engine designers that the world has yet seen; but the world has never seen much of him. He is by vocation a back-room boy, it seems — so it was a rare and surprising treat for us when Honda persuaded him to come to Britain to talk to us about the new 400cc twins that they recently introduced to the British market.
The Grand Prix car engine made its debut late in the 1966 season, and at the time (and for some considerable time thereafter) it was the most powerful engine in Formula 1 cars. It was hampered by a chassis that was excessively heavy, despite which it was prodigiously quick and could induce wheelspin (on tyres more than a foot wide) at higher speeds than any of its rivals, most of which were 30 per cent lighter. The five-cylinder 125 was making 23,000rpm noises around the racing circuits a year earlier, and the 50cc twin upon which it was based was enlivening the races for two-wheeled tiddlers earlier still, so their designer must by now be well into his forties. However, he looks still a young man; and like so many of those mature Japanese whose faces reveal nothing of the passage of the years, his face is equally unexpressive of doubt, insincerity, or any of those other emotional weaknesses that we occidentals so readily betray above our collars. Is it therefore possible that he was not telling us what he truly believed, but rather what Honda wished us to believe, when he set himself to explain the design and morphology of the new twins that are to supplant the beloved 400 four.
It must be possible; but I would not like to say that it is likely. More probably we felt that he ought to have been apologising for the fact that the new 400 is nothing like as romantic as that beautiful little four that we have come to love so much, and that the Americans have evidently failed even to understand. The Americans are to blame: they did not buy the 400F in the quantities that they should, even if we bought more than we were meant to — and thus did Honda's UK profits no good at all, since there was less profit for them from the sale of a 400F than they hoped they would make from the bigger 'bikes that they had expected us to buy. And because we love the little four, we felt that he ought to have been apologising for its replacement, especially its replacement by a mere twin, with or without its counterbalancing, wobbleknocking shafts. He was on the defensive, even though there was no hint of it in the quiet authority of his address and his replies to our questions immediately afterwards. Later, however, I was able to question him further in private discussion, and it soon emerged that he felt no need to be on the defensive at all: the simple fact of the matter was that, for the sort of power appropriate to the 400cc market, there was simply no need for the mechanical elaboration of the four-cylinder engine.
When Irimajiri made multi-cylinder engines (assuming that multi implies more than two) he did so in order to extort real power, corresponding to specific outputs three times higher than anything claimed by Honda or anybody else for a roadgoing motorcycle. He did it by a combination of revs, of volumetric efficiency (good breathing, in other words) and carefully controlled combustion. It needs only a small proportion of the racer's finesse in breathing and burning to produce more than adequate power from a street machine of far simpler mechanical layout than those to which we have become accustomed. In contenting ourselves with two cylinders, we need not suffer much more vibration than we did from four, since the twin can incorporate the balancing shafts already mentioned while the four continues to be as bad as ever in its liberation of secondary forces tending to make it shake up and down. On the other side of the profit and loss account is the fact that mechanical friction losses are less in a twin than in a four: Mr Irimajiri told me that the most significant source of these losses is the main bearings of the crankshaft (surprisingly more important than piston friction) and there are only three of these in the 400T, compared with at least five in any four-cylinder motorcycle engine current. Since any reduction in friction not only increases the power but also offers an alternative fuel economy, such an advantage is not lightly to be dismissed.
Piston friction is evidently not disregarded, for Mr Irimajiri insisted that piston velocity was more important than piston acceleration in weighing the factors for and against high revolutions. This was why he was advocating large cylinder bores and short piston strokes — but this was also where his arguments began to assume an air of special pleading. The plain fact is that the stroke of the 400T piston is virtually the same (actually 0.6mm or 1.2% more) than that of the 400F, so at any given rate of crankshaft rpm its mean piston velocity is roughly the same as in the four. Since the capacities are roughly equal, it follows that the area of the twin's two big-bore pistons is roughly the same as that of the four's little ones: it is in fact 4 1/2% less, which largely accounts for the twin having a displacement of 395cc and the four mustering 408.
What is important about piston area is simply and only that it allows a proportional valve area: the bigger the bores, the bigger the valves that can be fitted into the space, and the bigger the valves the greater should be the mass flow of air and fuel mixture to be breathed and burned to yield power. Now note this: the piston area of the 400T is practically the same as that of the 400F, and the same is true of the two engines' inlet valve areas. The four has one inlet valve to each cylinder, the twin has two, and so each engine has the same number of inlet valves which once again are practically the same size!
The twin gives more power than the four, nevertheless; figures from the same Honda source show an increase of 16% in DIN bhp. How does this come about? First because the twin (despite the professed worries about mean piston velocity) is run nearly 12% faster, which it can do because of its superb and breakerless ignition system, not because of anything to do with its stroke dimension or its valve apparatus. The rest of the improvement is attributable to better combustion (itself largely due to the new spark characteristics) and better breathing. The 400F is slightly strangled by its tiny carburettors, whereas the 400T has a pair of whopping 32mm instruments whose gross throat area is 28% greater. Moreover, the 400F has a less than perfect exhaust system with unequal-length pipes and a sadly compromised collector box, while the twin's system is admirably efficient, so a proportion of the better volumetric efficiency of the 400T is accounted for downstream of the exhaust valves. Those valves are most interesting of all: the twin has two big ones, as big as I would care to see valves in a 10,000 rpm engine, and (by coincidence?) their area is about the same as the area of the four little valves which exhaust the 400F. Isn't it amazing?
And isn't it a pity that Mr Irimajiri's pet four-valve configuration, which is what made Honda's racing engines so exceptional in the early 1960s that all the world copied them (even the car racing fraternity, Coventry-Climax and Ferrari and Ford and all), could not be adopted in the new engines, or could not have been adapted to the older ones that are now due to lapse? The new twins have only three valves per cylinder, because if they had four they would have to have two overhead camshafts if the spark plugs were to remain reasonably located. Indeed the new three-valve layout only works with twins, for the plugs would be displaced yet again if the engine were extended laterally to make it a four. With two valves per cylinder, one overhead camshaft is enough, whatever Kawasaki or Suzuki may claim to the contrary. Two overhead camshafts may be more than enough without being necessarily too much, until we come to increasing the number of valves per cylinder: with four valves per pot (or even with three valves per pot if those pots are more than two), two overhead camshafts are necessary. Are they desirable? It depends what you want; what I want is a 400F (or thereabouts — I would not mind it growing a little bigger) with sixteen valves. For a start, a 400F with the new ignition system and better exhaust and carburation would be more than satisfactory; but I suppose the Americans would not buy it. What a shame: after a few years' evolution, and a proper effort made to popularise the real Irimajiri philosophy, I could imagine the 400 four becoming a 500 five — and that is something I would really like to have!
