‘Combustion is the hidden principle behind every artefact we create.’
WG Sebald – The Rings of Saturn
It is at once dispiriting and encouraging to realise the four cycle gasoline engine, the 150-year old device which has held so many of us in its thrall since the early days of the horse-less carriage, remains woefully inefficient in its combustion and mechanics, even today managing to turn less than one-third of the energy embodied in its fuel into motive force.
The positive view of this realisation is that there is considerable potential for improvement, building on extraordinary advances in technology achieved over the last quarter-century, confounding prophets of doom, and meeting and dismissing all challengers to its supremacy as the prime mover for private transport.
Variable valve timing is but one of these advances, but its ability to extend the capability and overall efficiency of an engine could make it the most important of all the new technologies.
The history of the quest for an effective variable valve timing system dates from a time long before the Otto cycle and workable internal combustion, with systems in common use on steam engines in the nineteenth century. Internal combustion petrol engines had been in large-scale production for well over half a century before a workable example of variable valve timing was demonstrated.
The Lycoming R-7755 aircraft engine, a 36-cylinder 127-litre monster, developed between 1943 and 1946 featured a system of cam lobe switching controlled by the pilot to optimise the engine for maximum power on take-off or efficient high altitude cruising. The principle behind the Lycoming system was identical to Honda’s inaugural VTEC, although execution and operation were quite different.
Interest in a workable method of variable valve timing for production car engines was demonstrated by work at Fiat in the 1960s and a GM patent filed in 1975. However, Alfa Romeo appear to have been first to offer a system for sale in 1980.
Their mechanical VVT system was fitted to the 2.0-litre version of the eight valve twin-cam engine and used in conjunction with Spica fuel-injection. Sold only in the USA, it seems to have escaped the notice of the automotive world. Japan bid for automotive technological leadership from the mid-1980s and the first production application of electronically-controlled VVT technology appeared in Nissan’s VG30DE V6, opening the floodgates for a host of systems from the major Japanese manufacturers.
Honda’s VTEC – Variable valve timing enters the mainstream
Honda are arguably the masters of translating advanced combustion technology into production reality, as demonstrated by numerous motorcycle engines, the astonishingly advanced CVCC engine and early adoption of multi-valve cylinder heads.
Honda were not the first to offer variable valve timing but it came as no surprise that, when the day came, what was offered was quite exceptional. In April 1989, Honda introduced a new iteration of the Integra, a neat but rather staid saloon and coupe based on the Civic family platform. The high performance versions featured the first result of the company’s ‘New Concept’ programme, a 157.8bhp (160PS) 1.6-litre version of the B-Series engine family, the numerically significant specific output being achieved by means of the first use of a variable valve control system on a four wheeled series production Honda.
With peak power developed at 7600rpm and a 8000rpm redline the Honda engine could too easily have been dismissed as a piece of ephemeral technological showboating, a bid for credibility with teenage hashiriya, of no relevance beyond Japan where specifications were fetishised in impenetrable acronyms and strategically placed decals listing litanies of outputs, revolutions and parts lists. Nothing could be further from the truth.
What Honda had achieved was bringing to cost-effective mass production the solution to one of the fundamental limitations of the reciprocating four stroke petrol engine, the inability of the fixed timing camshaft and poppet valves to permit the optimum valve lift and duration across the engine’s speed range.
The short durations and low lifts required for smooth running and torque delivery at low engine speeds result in the engine being starved of air flow at high revolutions, thereby robbing the driver of the full benefit of the engine’s cylinder capacity and valve area.
Previous VVT designs, and indeed the majority currently in production, were of the ‘cam-phasing’ type, where a mechanism located between the camshaft drive sprocket and the camshaft itself allowed its timing to be advanced or retarded when actuated by a signal from the engine management system. The system developed by Honda Engineer Ikuo Kajitani achieved cam lobe switching with remarkable elegance and, more importantly, near-absolute reliability.
Imagine the process of dismantling the top end of the engine, and replacing the ‘soft’ road camshafts with full race profiles and then putting the entire mechanism back together.
Honda’s system accomplished exactly this with no driver input other than the application of a little extra pressure on the accelerator pedal. The means by which several hours of painstaking workshop activity could be achieved in a digitally mapped, electro-hydraulically activated, fraction of a second is possibly easier to understand than the reasons why such a mechanism was so desirable in the first instance.
Each camshaft in the original 1989 VTEC engine has one additional lobe per cylinder. The middle lobe is profiled for high-speed operation, while the two outer ones are optimised for ‘normal’ engine speeds and actuate valves through finger rockers. A third rocker, driven by the centre lobe, is located between the outer rockers but is free-floating until the VTEC system is activated.
At this point, typically between 4800-5800 rpm, dependent on load, an electric solenoid forces oil into hydraulic channels, thereby pushing out pins which link the inner and outer rockers. The physically larger ‘high-speed’ cam profile then acts on both valves, changing their lift and opening and closing duration.
When engine revolutions drop back into the ‘low-speed’ range, the solenoid releases the oil, the pins linking the rockers retract, returning the centre rocker to free-floating mode and valve operation to the ‘normal running’ cam lobes. From the start the VTEC engine lived up to expectations, delivering unstinting power up to its 8000rpm redline, while pulling smoothly, if not strongly, from 500rpm in fifth gear.
The harshest criticism of the early VTEC engine was an understandable lack of linearity in its power delivery – perhaps Honda’s Engineers couldn’t resist the potential to build in a ‘Jekyll and Hyde’ character to demonstrate the principle of their system, which, unlike some cam-phasing rivals does not allow any means of continuously varying timing.
What astonished the Automotive Engineering world as much as its performance, was VTEC’s unimpeachable reliability, given the intricacy of its small components running at high speeds. The Integra’s B-Series engine found new homes in the Civic and CRX Si-R before the end of 1989, and the 1990 NSX supercar’s 3-litre V6 also featured the VTEC system under its red bright red cam covers.
The powerplant line up for the fifth-generation Civic range, which appeared in late 1991, left no doubt that VTEC was central to Honda’s future engine plans and that the system’s adaptability could yield benefits other than extending the upper limits of the performance envelope. For the new Civic range, the now-familiar B16A twin cam VTEC unit was joined by two versions of the long-stroke D-Series engine.
The first was a 1590cc single cam unit, with VTEC operation on the inlet side only, which effectively replaced the 130bhp D16A8 twin-cam, familiar from the Honda CRX and Rover 216GTI, although Honda continued to provide the non-VTEC engine for Rover applications until around 1996. Matching the catalyst equipped twin-cam in power output, the new engine was more flexible and fuel-efficient and, in all probability, cheaper to produce.
Although offered in the ESi mid-range trim, with few sporting pretensions, performance was comparable to all but the quickest two-litre European hot hatchbacks and, with a 7200 rpm red line, was delivered in a manner which completely belied a bore/stroke ratio of 0.83. The second 1991 VTEC D-Series variant was the 1493cc VTEC-E unit, a ‘function specific’ use of the technology which used the conventional two cam lobes to operate each pair of inlet valves.
One cam lobe was profiled to open its valve a minimal amount to promote swirl in the combustion chamber. At higher engine speeds the ‘low-lift’ valve’s rocker was linked to its ‘high-lift’ neighbour using the same sliding rocker-linking pin principle as the high performance engines and allowing both valves to open with the same lift and duration.
The effectiveness of the principle was demonstrated by diesel-like fuel consumption figures, bettering those of the conventional 1.5-litre engined Civics by 7-10mpg on the ECE cycles, while providing the same power output and greater torque, both at lower rpm.
VTEC marches on
The story of VTEC’s continuing development is a lengthy and complex one, the technology’s adaptability and cost-effectiveness being demonstrated by its adoption across Honda’s entire product range. Its development, into more advanced forms, has been unrelenting, the latest iteration being the ‘Advanced VTEC’ engine which combines continuously variable valve lift and timing control with continuously variable phasing control.
The engine, which went into large-scale production for the 2009 model year, was claimed to increase fuel efficiency by 13% with emission levels well below the most stringent US and Japanese regulations.
Rover goes it alone
Given that, by the early 1990s, Rover appeared to have developed a chronic dependency on Honda platforms, the arrival in March 1995 of an all-new, home-grown MG sports car with a mid-mounted engine featuring variable valve timing was a highly momentous event.
The MGF’s Variable Valve Control system owed nothing to the technologies developed by Rover’s former partner Honda, nor new owner BMW’s VANOS cam-phasing system developed in conjunction with German component supplier Continental Teves.
Instead, Rover’s system evolved from a design described in a 1973 patent filed by British piston manufacturer The Automotive Engineering Company Limited (AE), and which differed radically in its operation and capabilities from anything which had gone before. Its development had been gradual, as a variable valve timing option had been envisaged as part of the K-Series range from an early stage in its design.
The system’s operation is best explained by diagrams, but its unique feature is its ‘Variable Duration’ principle. Unlike the Honda engine, the same cam lobes operate valves at all engine speeds. However, the Rover system offers a refinement not found in VTEC, or the simpler ‘cam-phasing’ mechanisms described above, in that the cam speed can be slowed during each revolution to open the valve earlier and close it later in its ‘high speed’ operation mode, giving a considerable advantage over cam-phasing which can only advance or retard the valve opening and closing points but does not have the desirable capability to extend duration overall.
The speed change is achieved by electro-hydraulic activation of an eccentric drive wheel between the camshaft drive sprocket and camshaft. The additional complexity imposed by the rotational speed variation principle is that, since each cylinder in the engine is at a different stage in the combustion cycle, each pair of cam lobes has to be controlled independently of the others.
The means by which this is achieved would do credit to a Swiss watchmaker in its mechanical ingenuity. What appears to be a pair of inlet camshafts is used, each serving two neighbouring cylinders. The front camshaft is driven from the main cam belt, while the rear inlet camshaft takes its drive from a belt driven from the end of the exhaust camshaft.
Independent operation of the cam lobes for each cylinder is achieved by the inlet camshaft being concentric, effectively a single, independently operable, camshaft per cylinder, giving a total of five including the fixed timing exhaust camshaft.
This undeniably complex arrangement allowed duration of the inlet camshaft to be varied between 220° and 295°, and valve overlap ranging between 21° and 58°. With an output of 143bhp at 7000rpm from 1796cc, the original VVC K-Series was tuned for driveability rather than epoch-making specific output figures, with, by Honda standards, a modest 7200rpm red-line.
The change-over point was set rather lower than the original Honda VTEC engine, at 4000rpm. The VVC K-Series engine’s progressive power delivery, low speed flexibility and responsive nature all found favour and it was to play an important part in Rover and MG Rover’s high-performance product line-up until the end of production at Longbridge.
Notable applications were the R3 Rover 200 Vi and BRM LE, late model R8 Coupes and Tourers and the MG ZR160. The later MG/Rover VVC engines benefited from an increased power output to 160PS at 6900rpm, but business uncertainties for most of the unit’s life meant resources were diverted away from development of the system’s potential, making it something of a cul de sac on variable valve timing’s development road map.
The mechanical solutions adopted are limiting in that they appear to be only readily applicable to banks of two or four cylinders. Despite daunting complexity, the Rover system had a great deal of unrealised potential, including the possibility of being combined with Honda-type cam lobe switching and doubling up the system to control the exhaust valves as well as inlets, although the mechanical complexity required to achieve this would be a formidable challenge.
Contemplation of the Rover VVC’s ingenious design and undoubted functional strengths leads easily to speculation as to what would have come out of a more collaborative approach with their former Japanese partner, whose one-time advertising slogan ‘Difficult is worth doing’ has proved again and again to be no empty boast.
The VVC Enigma
The enduring puzzle about the Rover VVC engine is not why the company went down the variable duration route, but why it felt the need to concentrate substantial funds and resources on developing any system of variable valve timing in the mid-1990s.
Early MGF development concentrated on forced induction versions of the 1400cc K-Series, which produced outputs sufficient to test the limits of the prototypes’ chassis and drivetrain. Turbocharging was being used effectively on the T-Series engine, while the smaller K-Series, even in fixed timing form, was acclaimed for superior output and efficiency to almost any of its contemporaries.
At the back of my mind is the notion that Rover wanted to show Honda that they were not the technological ‘poor relation’ in the joint venture. Honda sold Rover hundreds of thousands of engines during and beyond their 14-year alliance, none of these featured VTEC valve operation. Perhaps Rover’s bold foray was a defiant bid to demonstrate that they were worthy of more than Honda’s second-division powertrains.
Ultimately it was a vainglorious gesture. By the time the MGF appeared, Honda and Rover had parted and the British company’s future path envisaged a limited, yet very active, life for the K-Series in all its varieties.
As well as contributing to AR’s forums, he writes his own 5ivegearsinreverse blog which he describes as ‘automotive observations occasionally informed by an arcane body of obsessively-garnered knowledge, but more often simply by disbelief, bewilderment, and spite.’
Latest posts by Robert Leitch (see all)
- The cars : Argyll Turbo GT - 21 January 2019
- Concepts and prototypes : Austin Healey 4000 (ADO30) - 10 January 2019
- Engines : Rover M and T-Series - 17 March 2018