The Honda domination in Grand Prix racing is almost taken for granted. It is a yardstick against which all others engines are measured, including Ferrari and Renault. While its dominance has remained in normally aspirated form it was its crushing superiority in the last year of the turbocharged era that it really made its mark. To put today’s engine into perspective, David Hebb analyses that turbocharged unit.
The general principle that insecurity breeds secrecy and conversely that confidence fosters openness is as applicable to Grand Prix racing as it is to politics. For over twenty years Honda has been racing, however, during all this time the company has released little detailed information on its racing cars or engines. For the period 1964-8, a desire for secrecy is understandable since the company’s record was anything but glorious. Honda built its first Grand Prix car in 1964, but this machine, the R270/1, won only one event in two seasons of racing. The car’s 1.5-litre engine, designated the RA272, produced adequate power —allegedly it developed 5% more horsepower than the competing BRM or Coventry Climax engines — but any horsepower advantage was more than offset by its weight which was at least 20% greater than its competitors. This lack of success may be excused on the grounds that this was Honda’s first effort and the company’s engineers were just learning.
Honda’s next effort, however, the 3-litre R273V-12, was by some standards even less successful than its predecessor. It also won only one race, the 1967 Italian Grand Prix, but even this was achieved only after abandoning the Honda-built chassis for one developed by Lola engineers. Again Honda produced a V12 engine that had sufficient power, though this time its horsepower advantage over rivals such as Eagle-Weslake and Cosworth DFV was 2% at most and again the Japanese motor was considerably larger and heavier, weighing over 100 pounds more than the DFV.
In an attempt to build a more compact and lighter powerplant, Honda next built an air-cooled V8, the RA302. It proved to be more of an embarrassment to Honda weighing as much as the DFV but producing 10% less power. Not surprisingly, it was quietly dropped shortly before the firm withdrew from racing at the end of 1968.
There followed a gap of over fifteen years during which Honda’s engineers concentrated on road cars before the Japanese company re-entered Grand Prix racing in 1985 with a 1.5-litre supercharged V6 engine. Although Honda’s return began inauspiciously, progress was rapid and relentless, and soon, success was at hand. By 1988 the Honda engine was unquestionably the best, and consequently, Honda powered cars won every race but one. With this record of success behind them, Honda’s R&D engineers, Otobe, Miyano, Kawamoto, Aoki, and Ogawa felt confident enough to reveal to the world details about their engine and how they achieved such unparalleled success.
That Honda’s winning engine, the RA 168E, has a familiar look about it is not surprising since this engine has a long history which can be traced back to the RA263, a naturally aspirated 2-litre V6 that was used at the very beginning of the 1980s to power Formula 2 cars. The most notable feature of that engine was its wide bore and short stroke. According to Kawamoto, the chief engineer responsible for this engine, his team began by looking at the competition, especially the four cylinder BMW engine, which “had a 92mm bore, so we decided to do the same but with six cylinders, so theoretically we could get one and a half times the power”. The extreme over-square nature of the RA263 was carried over to Honda’s first 1.5-litre turbo-supercharged engines, the RA165E, RA166E, and RA167E. However, when severe boost and fuel restrictions came into effect in 1988, Honda was forced to alter radically its turbocharged Grand Prix engine. Above all else, this re-design effort was dominated by a need to improve fuel consumption. To achieve this aim, Honda developed a more compact combustion chamber and raised the higher compression ratio substantially. Instead of using a 7.2:1 compression ratio, as on the RA167E, the new engine was designed to run at 9.4:1. This was accomplished by reducing the bore to 79mm and increasing the stroke to 50.8mm, leaving the new engine, still over-square but less so than its predecessors or the competing TAG-Porsche, Renault, Ferrari, or Motori Moderni engines.
More than anything else, the highly conservative nature of Honda’s design stands out. The RA168E displays no devilishly ingenious features, nor any parts fashioned from ‘unobtanium’ or exotic materials, as had been rumoured. In fact, the block is made from plain old cast iron. The attractive stiffness and ductile characteristics of this material appealed to the Honda team, and, though it is heavier than aluminium alloy, for example, the use of thinwall casting methods reduced wall thickness to a range of 2 – 3.5mm so that in the end the weight of the block was not very great.
Unlike the block, the cylinder heads were made of aluminium alloy (Al Si6Cu4), while a few parts, such as cam covers and crankcase sump, were cast from a magnesium alloy. Complete with all accessories, the Honda engine weighs no more than 146 kgs, or about the same as a 3-litre Cosworth DFV.
In valve design Honda also took a conservative approach. Indeed, in this area this engine is less innovative than any earlier Honda Grand Prix engine; every one of its main features can be found on its first Grand Prix unit or those deployed soon thereafter. Moreover, valve design differs little from that used on rival engines. There are no desmodromic actions, no torsion bars nor pneumatic springs; four valves per cylinder are used with each valve being closed in the familiar manner by a pair of concentric springs. The cams act upon finger followers, a mechanism that Honda used on the RA302E in 1968 and seems to prefer to the more common inverted bucket type tappets used in most racing engines. Finger followers can reduce inertia loadings and thereby allow an engine to rev higher. This mechanism certainly works on the RA168E for the engine would run safely at 13,500 rpm. However, normally finger followers are more bulky than bucket type, but Honda has largely overcome this disadvantage by placing the follower pivots between the intake and exhaust valves. As a consequence, the cylinder heads of the Honda engine are no wider than its competitors. The cylinder heads are relatively tall affairs, however, partly because very long valve stems are used to obtain narrow angle inlet tracts but also because the finger followers are not flat, as is often the case, but ‘T’ shaped with the flat surface supported by a deep beam to supply extra rigidity.
Combustion chamber design is also conservative with compactness, (and hence good burning) and fuel economy is its a primary objective. Following the fashion set by Ferrari twenty years ago, the valves are rather small and arranged at an included angle of 32 degrees, the same angle that was first used on the Cosworth DFV two decades ago. Except for slight indentations to permit valve clearance, the pistons are flat and therefore light; the absence of a high crown also reduces surface area available to absorb heat and thereby permits a higher compression ratio to be used. From the point of view of valve gear and combustion chamber design then, the cylinder head of this engine is quite unremarkable. However, on closer examination, it becomes clear that considerable thought went into the design. Cooling, in particular, has received very careful attention since the presence of hot-spots would hinder progressive combustion and, ultimately, fuel economy.
Two water pumps are used on the RA 168E, each of which supplies coolant to its own side of the cylinder block. Moreover, water galleries are located on the outside walls of the cylinder block where they will absorb less heat. In the cylinder heads the opposite is true: the water galleries are on the inner walls, i.e., the intake or cooler side of the head. Thus, the coolant flows laterally from cool to hot to cool, thereby encouraging uniform thermal conditions around all cylinders and combustion chambers.
The induction system relies on two IHI (Ishikawajima-Harima Heavy Industries Co.) turbo-superchargers. Though working on the same principle as those found on road cars, these compressors, designated RXD6D by the factory, are very special racing appliances. They employ ceramic turbine wheels to reduce inertia, and hence ‘turbo lag’. Secondly, this silicon-carbide material is able to run safely at higher temperatures than conventional metal turbines. Ceramics are also used for the bearings that support the turbine and compressor not only because of the attractive high-temperature properties of this material, they are also a means for reducing friction and lubrication problems.
Naturally fuel injection is used. Though similar to the standard Bosch/Bendix indirect injection systems, that used by Honda was developed in-house and employs two injectors per cylinder, the second being needed to ensure that sufficient fuel is available in all operating regimes. Although meeting the fuel requirements set down by the CSI, i.e. petrol with a maximum of 102 octane, the fuel in the Honda RA168E is far removed from pump petrol. Produced by Elf France, this unleaded brew consisted of up to 84% Toluene, a heavy, slow-burning hydrocarbon that resists premature detonation, a real danger in the high-temperature, high-pressure conditions found in combustion chambers like those of the RA168E. The extent of Honda’s concern over fuel and fuel efficiency is evident in the engine’s provision to pre-heat the fuel, a procedure that was found after exhaustive tests to improve fuel economy quite dramatically by promoting better vaporization.
Four separate scavenging pumps are used in order to ensure that oil is collected and removed from the base chamber where its presence would otherwise cause pumping losses and thereby reduce power and waste fuel. With a pump at each corner of the dry-sump, oil will be evacuated no matter the direction of the centrifugal force imputed to the lubricant by the high-G manoeuvres of the car.
The accumulation of this attention to detail during the design process has resulted in an engine that was proven to be not only powerful and economical but reliable as well. Performance figures for the Honda engine go a long way in explaining the basis for McLaren’s string of victories in 1988. By the end of the season the RA168E was producing 685 (metric) bhp @ 12,500. This is equivalent to 457 hp/litre. Turbocharged engines in road cars are exceptional if they achieve 100 hp/litre. However, even this performance pales in comparison to the specific output achieved the previous year when restrictions were few. In 1987 the RA167E engine developed no less than 1,010 hp. @ 12,000 rpm in race trim (677hp/litre) and produced a torque peak of 490 ft. lbs. @ 9800 rpm.
Although the absolute output of the RA167E is certainly impressive, it is relatively less potent than the 1988 engine. The regulations in effect in 1987 permitted the RA167E to use a boost pressure of 4 bar (rather than only 2.5 bar as permitted the following year), and 195 litres of fuel (rather than 150 litres) that were allowed in 1988. Leaving aside the smaller amount of fuel available in 1988, the reduction in boost pressure alone would suggest that the RA168E should produce no more than 630 hp., rather than the 685 hp which the Honda’s engineers actually achieved.
Torque figures for the RA168E are also impressive with a maximum of 313 ft lb (a) 10,000 rpm, but the really astonishing aspect of the torque curve is not its peak figure but the incredible flatness and spread of power. Over 295 ft lb were available all the way from 8000 to 12,000 rpm. Driving a car with such a broad torque curve is less fatiguing because gear changing is less frequent and exacting: as a consequence, the driver can devote more time to steering, braking and overtaking slower cars during a race.
As explained above, fuel efficiency was one of Honda’s main objectives in developing the RA168E, and the abstemiousness of this engine is noteworthy. Specific fuel consumption (SFC) figures were as low as 272g/kw-h. (200g/hp-h.) indicating a degree of thermal efficiency of over 30%, a very respectable figure for a road car, let alone an all-out racing engine. Moreover, the SFC curves for the RA168E show that the trough was quite extensive with less than 280g/kw-h being consumed all the way from 10,500 to 12,500 rpm.
That fuel consumption should remain low at maximum power was one of Honda’s chief aims because the R&D engineers had calculated that their engine would be run at (or near) maximum power for much of a race. For example, at Imola in 1988, the RA168E in Alain Prost’s McLaren was operated at maximum power (approximately 90-100% throttle opening) for 63.6% of the race. Conversely, to use the Imola example again, when the engine was run at or near closed throttle (19.8% of the time) fuel consumption shot up to 340g/kg-h.
Anyone watching the 1988 Grand Prix season could observe the Honda’s combination of power and fuel efficiency at work, especially in the McLaren cars. The McLaren-Honda team cars did not run out of fuel nor were their drivers forced to back off the throttle or reduce boost in order to finish a race. Honda-engined cars had more power and could use that power throughout the race. Moreover, the power and efficiency of the Honda engine did not come accidentally nor was it achieved by some trick feature, rather, as their engineers have now revealed, it was the culmination of a design effort that began with a clear identification of objectives, concentration on those objectives through all phases of the design process, and painstaking devotion to detail in development. More than anything it was this focusing of all engineering effort that brought success to Honda. With this experience in mind, it is therefore not surprising that Honda engines have again dominated this year’s Grand Prix racing.
Although it has taken many years, the Honda R&D team have acquired, perhaps like only Mercedes-Benz in the past, those indefinable design and development skills needed to produce race-winning engines at will. That they are now prepared to discuss their work is evidence, if any were needed, that they are almost disdainfully confident of that ability. The dominance of the naturally-aspirated 3.5-litre RA109E in 1989 and again this year is but another demonstration of Honda’s engineering ability and one, hopefully, about which we will learn in detail soon when the R&D team next explain how they went about designing and developing their latest world-beating engine.