An Open and Shut Case - Andrew Noakes - Motoring Writer
Published in European Automotive Design 2008

Virtually all four-stroke internal combustion (IC) engines control the flow of gas into and out of their combustion chambers in broadly the same way. Poppet valves are actuated by a camshaft which is driven from the engine’s crankshaft. It’s a well-proven system, yet one with such fundamental drawbacks that many strenuous efforts have been made to eliminate it. Years of research have gone into camless valvetrains, yet so far the only successful applications outside the research laboratory have been in very large, very low-speed marine and industrial diesels. Automotive applications have been promised again and again, only to be delayed, abandoned or quietly forgotten. Now, at long last, it seems the wait might almost be over.

Conventional camshaft-driven valvetrains suffer from parasitic losses in the camshaft drive mechanism (be it a cogged belt, chain or geartrain), the rockers and the valve springs which can amount to a quarter of an engine’s internal friction at low speeds. High-quality lubrication is essential to avoid wear, and though basic reliability is very good it can be easily compromised by small design or manufacturing errors (as some vehicle manufacturers have found out the hard way) or inadequate maintenance. But the biggest drawback to conventional valvetrains is that they couple the timing of valve events directly to engine rotation, which immediately limits the optimization of valve timing. A cam profile which is ideal for one region of the engine’s operating envelope will be wrong for every other area of the speed and load range. As a result, powertrain engineers have always been forced into unpleasant compromises when choosing cam profiles. Usually the application will demand a blend of top-end power, low-speed torque, fuel economy, flexibility and emissions performance, and a cam profile must be chosen which is acceptable under all circumstances.

In recent years, the introduction of variable valve train (VVT) systems has helped to minimize the compromises. The simplest VVTs are variable cam phasing systems which rotate the whole camshaft relative to its drive sprocket – and therefore relative to the crankshaft – to advance or retard all the valve events. Simpler systems offered a choice of two discrete positions, while more complex versions offered infinitely variable control within a given range. On engines with twin camshafts per cylinder bank, each camshaft could be given its own variator and independent control of intake and exhaust valve timing could be achieved – but the lift and duration of each valve event remained fixed. Honda’s VTEC system was one method of varying valve lift and opening duration, using two different cam profiles which the engine management system could choose between. Later, fully-variable timing systems were developed, for instance by using cams which varied in profile across their width. A different profile could be chosen by moving the camshaft axially so the cam follower rode on a different part of the cam.

Even these systems still relied on a basic timing relationship between the valvetrain and the crankshaft, with only a certain degree of variation. So why not dispense with the mechanical connection to the crankshaft, and control each valve individually? That’s what camless valvetrains – variously known as active, adaptive, electronic or fully variable valvetrains – aim to do. Currently vehicle manufacturers and Tier 1 suppliers are concentrating on two main technologies, electro-magnetic valvetrains and electro-hydraulic valvetrains.

Valeo’s e-Valve is an electro-magnetic system. Each valve is controlled by a pair of opposed coil springs and a pair of electro-magnets. When the valve is in its closed position, an armature plate attached to the valve stem is held by the upper magnet and the upper spring is compressed. When the upper magnet releases the armature plate, the energy stored in the spring is released and the valve opens, compressing the lower spring, until the armature is ‘caught’ by the lower magnet, which holds the valve open. The same process in reverse closes the valve. Valeo says the system monitors the valve using a very precise position sensor, and controls the speed of the valve near the extremes of its movement to cut noise. To generate sufficient power the magnets are powered by a 42-volt supply generated inside the system control unit. One drawback, however, is that the present system does not offer any control of valve lift, which would demand movable springs and magnets.

Another electro-mechanical system, developed by Siemens VDO before its takeover by Continental, used a current-carrying coil instead of an armature plate. The operating magnets were placed beside the valve, which was moved by the interaction of the fields of the magnets and the coil. Fine control over the magnetic field coils were said to deliver precise control, this time providing the option of variable valve lift and also a ‘soft landing’ to reduce noise and valve seat wear.

Lotus Engineering’s Active Valve Train (AVT) comes from the opposing electro-hydraulic camp, and offers control of both valve timing and valve lift. AVT was developed in 1992 during research into port and valve deactivation, mechanical variable valvetrain systems and early inlet valve closure. Since then it has been widely used for prototype camshaft simulation on experimental engines, giving engineers a quick and easy way of altering valve lift profiles during combustion chamber development. Lotus is now working with Eaton to develop a production version of the system.

AVT uses a high-pressure hydraulic circuit which delivers hydraulic fluid to a double-acting hydraulic actuator for each valve. The actuator is mounted co-axial with the valve stem, and the actuator piston is attached to the valve. Electro-hydraulic servo valves proportionally control the flow of hydraulic fluid to either the top or bottom of the actuator piston to control the velocity, timing and lift of the valve. Opening and closing points can be varied individually in one-degree increments, and the maximum lift is adjustable from 0.01mm up to the maximum travel of the actuator. A crankshaft encoder is used to measure the crank position and each valve is monitored using an LVDT displacement transducer. The measured values are compared to the demanded profile approximately every 100 microseconds, as part of the feedback loop controlling the servo valves. AVT continuously checks for malfunctions such as hydraulic pressure loss, crank or valve position signal loss, and potential valve to piston or valve to valve contact.

Ricardo has also developed an electro-hydraulic camless valvetrain, which features on the 2/4SIGHT research project which the company is leading. Another system is being developed by Sturman Industries, a specialist in electro-magnetic and hydraulic valve technology. Its Hydraulic Valve Actuation system first appeared in 1996 and has been successfully demonstrated in large commercial vehicles and a VW Jetta passenger car.

Other approaches to camless operation have been suggested. British technology company Camcon is developing Intelligent Valve Actuation, based on its ‘binary actuating technology’, which is also used in a variety of industrial valve designs. Swedish engineering company Cargine, by contrast, has a pneumatic system it calls Free Valve Technology, which consumes less than 4kW on a four-cylinder engine running at 6000rpm – said to be a similar power consumption to electro-magnetic systems and less than electro-hydraulic systems. Piezo-electric operation has also been mooted, but no such system is close to a production application.

When camless valvetrains do start to appear in production vehicles – and that could be around 2010 – they are likely to operate only the intake valves, with a conventional camshaft-driven valvetrain for the exhaust valves. According to Valeo these ‘half camless’ engines provide most of the benefits of fully camless operation, at half the cost. That should mean half-camless systems can be implemented at a cost which is competitive with, or perhaps even better than, modern diesel engines.

Once valve timing is under the absolute control of an engine management ECU, rather than shackled to engine rotation, all sorts of exciting possibilities present themselves. The most obvious benefit is that once valve opening and closing is software controlled it is relatively simple to provide valve timing (and, where possible, lift) maps within the engine management ECU to provide optimum valve events at all points in the engine speed and load range. The current compromises inherent in cam profile design would be a thing of the past: an engine could operate with a low-lift, short-duration profile at idle and a high-lift, long-duration profile at high speed and load – giving excellent flexibility, low fuel consumption and also a high maximum power output.

The camless valvetrain could be operated with an early intake valve closure (EIVC) strategy for load control without the need for a throttle valve, giving a reduction in pumping losses. Further improvements in this area would result from faster valve opening. In a conventional valvetrain the rate at which the valves open is directly related to engine speed, so at low engine speeds the valves open relatively slowly. A camless system could snap the valves open at the same rate at all engine speeds, giving a near-trapezoidal lift curve. The result is an improvement in low-speed torque, and the potential to reduce idle speed. Manipulating the valve timing would also facilitate HCCI combustion and a variable expansion ratio.

Deactivation of valves and cylinders, for economy and emissions benefits and for additional limp-home capabilities, can be implemented with little more than additional software. In multi-valve engines just one intake valve could be opened at low speeds to keep gas velocity high and encourage swirl. Cylinders could be switched off by keeping all the valves closed so air in the cylinder acted as a gas spring, storing energy during compression and releasing it as the piston dropped. The result would be further fuel consumption and emissions benefits in low-load operation, particularly for multi-cylinder engines used in premium vehicle sectors. According to FEV, which is developing a system known as ‘Electromechanical Valvetrain’ or EMVT, a camless valvetrain with cylinder deactivation can potentially improve fuel consumption by about 18%.

Another approach, as Lotus Engineering has suggested, would be to use deactivated cylinders as an air pump to charge a pneumatic reservoir during braking. The compressed air would then be used to launch this ‘pneumatic hybrid’ vehicle from standstill.

The camless valvetrain can also give the engine management system the option of alternative combustion strategies to improve emissions performance and fuel economy under low-load conditions. The engine could rapidly switch to a high-efficiency Atkinson/Miller cycle with late inlet valve closure, or could change to six-stroke or eight-stroke operation on a cylinder-by-cylinder basis. One suggested six-stroke cycle uses a spark ignition combustion phase with rich combustion followed by second compression stroke and HCCI combustion, the two-stage process being said to deliver low NOx and low hydrocarbon emissions. Lotus has already revealed the results of simulation work which shows the intake and exhaust tuning advantages of using a ‘fully variable’ valvetrain to alter an engine’s firing order, a process limited only by the geometrical restrictions of the crankshaft design.

Clearly camless valvetrains offer enormous potential, though utilizing it all will take much more time and development once fully production-ready camless systems are available. When fully variable electro-magnetic or electro-hydraulic control of valve timing does take the place of mechanical control using camshafts, the door will be open for rapid improvements in fuel consumption, emissions and performance. In the current climate the industry may tend towards using those benefits to downsize engines while maintaining existing performance levels. In the longer term we can expect engines with a much greater variety of operating cycles and combustion strategies, leading to still greater benefits. The limitation might well be in the cost of development for the complex control ECUs and sensor systems which will be needed.

All that lies far in the future, but at long last it seems that reliable, affordable camless valvetrains are on the horizon. It will surely not be too much longer before virtually all automotive engines adopt them.

Share this page