Engine Variable valve timing
In internal combustion engines, Variable valve timing (VVT), also known as Variable valve actuation (VVA), is any mechanism or method that can alter the shape or timing of a valve lift event within an internal combustion engine. VVT allows the lift, duration, or timing (in various combinations) of the intake and/or exhaust valves to be changed while the engine is in operation. Two-stroke engines use a power valve system to get similar results to VVT. There are many ways in which this can be achieved, ranging from mechanical devices to electro-hydraulic and camless systems.
The valves within an internal combustion engine are used to control the flow of the intake and exhaust gases into and out of the combustion chamber. The timing, duration and lift of these valve events has a significant impact on engine performance. In a standard engine, the valve events are fixed, so performance at different loads and speeds is always a compromise between driveability (power and torque), fuel economy and emissions. An engine equipped with a variable valve actuation system is freed from this constraint, allowing performance to be improved over the engine operating range.
Strictly speaking, the history of the search for a method of variable valve opening duration goes back to the age of steam engines when the valve opening duration was referred to as “steam cut-off”. Almost all steam engines had some form of variable cut-off, the mechanism for which is the valve gear such as for example Stephenson valve gear or Walschaerts valve gear. That they are not in wide use is a reflection that they are all lacking in some aspect of variable valve actuation.
The desirability of being able to vary the valve opening duration to match an engine’s rotational speed first became apparent in the 1920s when maximum allowable RPM limits were generally starting to rise. Until about this time an engine’s idle RPM and its operating RPM were very similar, meaning that there was little need for variable valve duration.
It was in the 1920s that the first patents for variable duration valve opening started appearing – for example United States patent U.S. Patent 1,527,456. A surprising fact is that from these first patents until the appearance of the helical camshaft there has never been a really practical and useful variable duration camshaft.
Piston engines normally use poppet valves for intake and exhaust. These are driven (directly or indirectly) by cams on a camshaft. The cams open the valves (lift) for a certain amount of time (duration) during each intake and exhaust cycle. The timing of the valve opening and closing is also important. The camshaft is driven by the crankshaft through timing belts, gears or chains.
The profile, or position and shape of the cam lobes on the shaft, is optimized for a certain engine revolutions per minute (RPM), and this tradeoff normally limits low-end torque, or high-end power. VVT allows the cam timing to change, which results in greater efficiency and power, over a wider range of engine RPM.
An engine requires large amounts of air when operating at high speeds. However, the intake valves may close before enough air has entered each combustion chamber, reducing performance. On the other hand, if the camshaft keeps the valves open for longer periods of time, as with a racing cam, problems start to occur at the lower engine speeds. This will cause unburnt fuel to exit the engine since the valves are still open. This leads to lower engine performance and increased emissions. For this reason, pure racing engines which are designed to idle at speeds close to 2,000 rpm, cannot idle well at the lower speeds (around 800 rpm) expected of a road car.
Pressure to meet environmental goals and fuel efficiency standards is forcing car manufacturers to use VVT as a solution. Most simple VVT systems advance or retard the timing of the intake or exhaust valves (such as BMW´s early VANOS). Others (like Honda's VTEC) switch between two sets of cam lobes at a certain engine RPM. Furthermore Honda's i-VTEC can alter intake valve timing continuously.
The first variable valve timing systems came into existence in the nineteenth century on steam engines. Stephenson valve gear, as used on early steam locomotives, supported variable cutoff, that is, changes to the time at which the admission of steam to the cylinders is cut off during the power stroke. Early approaches to variable cutoff coupled variations in admission cutoff with variations in exhaust cutoff. Admission and exhaust cutoff were decoupled with the development of the Corliss valve. These were widely used in constant speed variable load stationary engines, with admission cutoff, and therefore torque, mechanically controlled by a centrifugal governor and trip valves. As poppet valves came into use, simplified valve gear using a camshaft came into use. With such engines, variable cutoff could be achieved with variable profile cams that were shifted along the camshaft by the governor. This is now coming in system.
In 1958 Porsche made application for a German Patent, also applied for and published as British Patent GB861369 in 1959. The Porsche patent used an oscillating cam driven via a push/pull rod from an eccentric shaft or swashplate. The cam was Desmodromic having opening and closing cam surfaces which operated the valve by a bifurcated rocker and ball joint. Being Desmodromic meant there was no valve spring. The cam pivot was adjustable for height, as the push/pull rod length was constant this rotated the cam so the lift and duration increased. A compensating link moved the rocker pivot to match the cam's position. The adjustment of the cam pivot could be by mechanical linkage to a screw thread, hydraulic from engine driven pump with spill valve or from an engine speed governor. At present it is unknown if any working prototype was ever made.
Fiat was the first auto manufacturer to patent a functional automotive variable valve timing system which included variable lift. Developed by Giovanni Torazza in the late 1960s, the system used hydraulic pressure to vary the fulcrum of the cam followers (US Patent 3,641,988). The hydraulic pressure changed according to engine speed and intake pressure. The typical opening variation was 37%.
In September 1975, General Motors (GM) patented a system intended to vary valve lift. GM was interested in throttling the intake valves in order to reduce emissions. This was done by minimizing the amount of lift at low load to keep the intake velocity higher, thereby atomizing the intake charge. GM encountered problems running at very low lift, and abandoned the project.
Alfa Romeo was the first manufacturer to use a variable valve timing system in production cars (US Patent 4,231,330). The 1980 Alfa Romeo Spider 2.0 L had a mechanical VVT system in SPICA fuel-injected cars sold in the United States. Later this was also used in the 1983 Alfetta 2.0 Quadrifoglio Oro models as well as other cars. The system was engineered by Ing Giampaolo Garcea in the 1970s.
Honda's REV motorcycle engine employed on the Japanese market-only Honda CBR400F in 1983 provided a technology base for VTEC.
In 1986,Nissan developed their own form of VVT with the VG30DE(TT) engine for their MID4 Concept. Nissan chose to focus their NVCS (Nissan Valve-Timing Control System) mainly on torque production at low to medium engine speeds, because, the vast majority of the time, automobile engines will not be operated at extremely high speeds. The NVCS system can produce a smooth idle and high amounts of torque at low to medium engine speeds. The VG30DE engine was first used in the 300ZX (Z31) 300ZR model in 1987. It was the first production car to use electronically controlled VVT technology.In 1987 Nissan also sold the Gloria, Leopard, and Cedric, all of which could come powered by the VG20DET engine which also utilized Nissans NVCS valve timing system.
The next step was taken in 1989 by Honda with the VTEC system. Honda had started production of a system that gives an engine the ability to operate on two completely different cam profiles, eliminating a major compromise in engine design. One profile designed to operate the valves at low engine speeds provides good road manners, low fuel consumption and low emissions output. The second is a high lift, long duration profile and comes into operation at high engine speeds to provide an increase in power output. The VTEC system was also further developed to provide other functions in engines designed primarily for low fuel consumption. The first VTEC engine Honda produced was the B16A which was installed in the Integra, CRX, and Civic hatchback available in Japan and Europe. In 1991 the Acura NSX powered by the C30A became the first VTEC equipped vehicle available in the US. VTEC can be considered the first "cam switching" system and is also one of only a few currently in production.
In 1991, Clemson University researchers patented the Clemson Camshaft which was designed to provide continuously variable valve timing independently for both the intake and exhaust valves on a single camshaft assembly. This ability makes it suitable for both pushrod and overhead cam engine applications.
In 1992, Porsche first introduced VarioCam in the Porsche 968 which provided continuously variable valve timing for the intake valves.
In 1992, BMW introduced the VANOS system. Like the Nissan NVCS system it could provide timing variation for the intake cam in steps (or phases), the VANOS system differed in that it could provide one additional step for a total of three. Then in 1996 the Double Vanos system was introduced which significantly enhances emission management, increases output and torque, and offers better idling quality and fuel economy. Double Vanos was the first system which could provide electronically controlled, continuous timing variation for both the intake and exhaust valves.
Ford began using Variable Cam Timing in 1998 for the Ford Sigma engine and the Ford Zetec engine.
In 1999, Porsche introduced VarioCam Plus on its 911 Turbo which combined continuous valve timing and two stage valve lift on the intake valves.
In 2001, BMW introduced the Valvetronic system. The Valvetronic system can continuously and precisely vary intake valve lift, and in addition, the independent Double VANOS system can concurrently vary the timing for both the intake and exhaust valves. The precise control the system has over the intake valves allows for the intake charge to be controlled entirely by the intake valves, eliminating the need for a throttle valve and greatly reducing pumping loss. The reduction of pumping loss accounts for a 10-15% increase in power output and fuel economy.
Ford became the first manufacturer to use variable valve timing in a pickup-truck, with the top-selling Ford F-series in the 2004 model year. The engine used was the 5.4 L 3-valve Triton.
In 2005, General Motors offered the first Variable Valve timing system for pushrod V6 engines, LZE and LZ4.
In 2007, DaimlerChrysler became the first manufacturer to produce a cam-in-block engine with independent control of exhaust cam timing relative to the intake. The 2008 Dodge Viper uses Mechadyne's concentric camshaft assembly to help boost power output to 600 bhp (450 kW).
In 2009, Fiat Powertrain Technologies introduced the Multiair system in Geneva Motor Show. The Multiair is a hydraulically actuated variable valve timing system, which gives full control over valve lift and timing. The new technology is available in Alfa Romeo MiTo starting from September 2009.
In 2009, Porsche introduced an enhanced version of VarioCam Plus on its 911 GT3 including the previous variable valve timing and two stage valve lift on the intake valves but with additional variable timing of the exhaust valves.
In 2012, General Motors introduced Intake Valve Lift Control (IVLC) on the Ecotec Gen III engine to enable variable intake valve lift, duration and timing. The rocker arm can switch between low-lift and high-lift intake cam profiles, actuated through a valve by a hydraulic lash adjuster.
In 2010, Mitsubishi developed and started mass production of its 4N13 1.8 L DOHC I4 world's first passenger car diesel engine that features a variable valve timing system
Note that the types briefly described below relate to the ability to vary primarily the valve opening duration in a continuous fashion by mechanical means. To be able to alter the valve opening duration in a continuous fashion by mechanical means has proven to be a surprisingly difficult task. To achieve this aim has been a “holy grail” for many years for inventors and car company research departments alike. There have only ever been two systems of this type that have been successfully used in cars sold for normal road use – BMW’s Valvetronic (and some others operating on this principle, but BMW was the first) and the Rover VVC system. Essentially there are only a very limited number of basic classes (and subclasses) of mechanism known at present that can be employed (in a reasonably practical sense) to vary the valve opening duration of an engine. There are, of course, other proposed schemes and patents for realising continuous variable valve opening duration by mechanical methods. Generally speaking none of these other methods are usable in any practical sense. Those described below are essentially the only workable types.
BMW’s Valvetronic, the Nissan VVEL and Toyota Valvematic all belong to the “oscillating cam” class of mechanism. This is a class that dates back to the days of steam engines where the amount of steam entering the cylinder was regulated by the steam “cut-off” point (although the valve gear of most steam engines was not of this type). The existing systems vary widely in their appearance but all have the same underlying principle. This involves a conventional cam lobe (Valvetronic and Valvematic) or eccentric and connecting rod (VVEL) to generate an oscillating or rocking motion in a part cam lobe which then acts on a follower and the valve. The part cam lobe consists of a section of base circle and a section of lobe flank. The position of the part lobe can be adjusted so that the follower is exposed to varying amounts of base circle or lobe flank. At one extreme is all base circle resulting in zero lift and zero duration, at the other extreme is all flank giving maximum lift and maximum duration. There is a continuous range of lift/duration settings between the two limits but the duration and lift are always directly proportional to each other. The Valvetronic has been widely used by BMW and has proved to be very reliable in service. However it has a major flaw in that the amount of duration is linked in direct proportion to the amount of lift. This has the effect of lessening the effectiveness of the potentially very wide duration range. The Valvetronic has only ever been used on the intake valves. A main strength of Valvetronic – the ability to produce valve openings of very short duration and very low (down to zero) lifts has little useful relevance for the exhaust valve. Because of the linked duration/lift situation the Valvetronic (and the other similar systems) act essentially in their behaviour as a variable lift systems rather than genuinely variable duration arrangement. Despite BMW’s claims, the Valvetronic has only a relatively minor beneficial effect on engine power and fuel consumption – something of the order of 10%. More recent developments of the basic oscillating cam principle have been used in the Nissan VVEL and the Toyota Valvematic. These are neater and more compact in size than the Valvetronic but still have similar limitations in their performance. Mention should be made of the Honda AVTEC system. This system is notable and interesting in that, despite its appearance, it is variable lift only - the duration remains fixed. This fact is made quite clear in the text of the Honda patent.
Eccentric Cam Drive
There is little question that this is an extremely ingenious system. Despite its cleverness it has serious basic design problems in practice. Only the Rover company ever used this system. The system operates through an eccentric disc mechanism which slows and speeds up the angular speed of the cam lobe during its rotation. Arranging the lobe to slow during its open period is equivalent to lengthening its duration. However it has a fundamental drawback in that each individual valve (or each pair of valves in a “four-valve” engine) needs its own eccentric drive and its own controller. This means that to have all the valves in a four-cylinder engine to be able to be varied in duration, eight separate drives and eight separate controllers would be needed. Economics and complexity make this almost impossible. Although the Rover VVC is a somewhat flawed system it is important in that it is the only genuinely variable duration system that has ever been available commercially for a road car – which is a little surprising considering the long history of the internal combustion engine.
Three-Dimensional Cam Lobe
This system consists of a cam lobe that is somewhat elongated axially and has at one end a short duration/reduced lift profile, and at the other end a longer duration/greater lift profile. The two profiles progressively vary in a continuous manner between the two extremes. By shifting the cam lobe axially a stationary follower is exposed to a varying lobe profile to produce different amounts of lift and duration. Although this seems to be a simple and effective idea it has a fundamental (and apparently insoluble) problem which has meant that it has never been used commercially. The varying profile, by its very nature, means that the lobe flanks of the cam lobe are not parallel to the axis of rotation of the camshaft. However the lobe surfaces on the base circle region are parallel to the cam’s axis of rotation. This means that a flat-faced follower has to tilt in various directions as the lobe rotates. Even if the follower has freedom to tilt freely it still does not act totally correctly in a geometrical sense on the lobe surface. The only type of follower which is geometrically correct is a spherical or ball-ended type. The resultant point contact stress makes this arrangement generally impractical. Although in popular automotive culture this system is generally attributed to Ferrari, it is concept that dates from the 1930s, or possibly even earlier. It is still being suggested in recent times in various patents (USPTO 7341032 for example). Ferrari has never used this system commercially in any of their road cars.
Combined Cam Lobe Profile Types
This class is divided into three basic subclasses.
Two Shaft Combined Profile
This arrangement consists of two closely spaced parallel camshafts of conventional layout. A pivoting follower spans both camshaft and is acted on by two lobes simultaneously. Each camshaft has a phasing mechanism which allows its angular position relative to the engine’s crankshaft to be adjusted. The basic idea behind this system is that one lobe controls the opening of a valve and the other controls the closing of the same valve – the spacing of the two events being a period of variable duration. The system is quite workable in that valve clearance, smooth actuation of the valve and other practical issues are not a problem. However there is something of a basic flaw in that at long duration settings one lobe may be starting to reduce its lift as the other is still increasing – this has the effect of lessening the overall lift and possibly causing dynamic problems. There are also more obvious problems in the overall size of layout due to the parallel shafts, the necessarily large follower etc. Surprisingly, although workable in many aspects, it is not known to have ever been used commercially. The Mechadyne company (from the U.K) has been developing a version of this basic idea having the two shafts arranged coaxially, the lobes attached to the inner and outer shafts being able to be phased separately, a “summation linkage” replacing the conventional follower. This goes someway towards solving the space and follower size problems. Mechadyne also claim to have solved the uneven rate of opening of the valve problem to some extent thus allowing long duration at full lift.
Coaxial Two Shaft Combined Profile
It should be noted here that despite the very similar description to the Mechadyne arrangement the mechanism described here is of a basically different principle. This system is distinctly different to the Mechadyne system in that, although similarly coaxial with pairs of lobes acting on the one follower etc. the actual lobe profile has a distinctly different profile and acts on a conventional (although maybe somewhat wider than normal) follower. The profile used has a characteristic “snub-nosed” appearance with usually about a twenty degree angular region of true radius about the axis of rotation of the camshaft on the lobe nose. This type of profile is actually very little different from some conventional lobe profiles that act on valve lift multiplying followers of a fairly high ratio of lobe lift to valve lift. The essential operating principle is that the one follower spans the pair of closely spaced lobes. Up to the angular limit of the nose radius the follower “sees” the combined surface of the two lobes as a continuous, smooth surface. When the lobes are exactly aligned the duration is at a minimum (and equal to that of each lobe alone) and when at the extreme extent of their misalignment the duration is at a maximum. The basic limitation of the scheme is that only a duration variation equal to that of the lobe nose true radius (in camshaft degrees or double this value in crankshaft degrees) is possible. In practice this type of variable cam has a maximum range of duration variation of about forty crankshaft degrees. Any attempt to drastically increase this range (by increasing the angular extent of the true radius lobe nose region) soon gives rise to problems with excessive rates of valve acceleration etc. This range limits the usefulness of the cam to some extent. Prototypes of the cam system have been run and have proved to be quite reliable and to provide useful peak power increases of the order of thirty to forty percent. Once again it is a little surprising that the general idea has not been used commercially. This system is also notable in being quite often the idea that inventors come up with, at least initially. It is also the principle behind what seems to be the very first variable cam suggestion appearing in the USPTO patent files in 1925 (1527456). The “Clemson Camshaft” referred to elsewhere in this article is of this type. There are literally hundreds of patents in the USPTO files on this general theme alone. A more modern and workable design of this basic principle (with a notably practical and rugged internal structure and neat self-contained centrifugal controller/actuator) is USPTO 7007652.
Combined Two Shaft Coaxial Combined Profile with Helical Movement
This system is usually referred to more simply as the “Helical Camshaft”. It has a very similar principle to the previous type described using virtually (or sometimes exactly) the same base duration lobe profile. However instead of a strictly two-dimensional rotation (that is the rotation is in a single plane) the adjustment is both axial and rotational giving a helical or three-dimensional aspect to its movement. This unusual (and difficult to describe in words) movement overcomes the restricted duration range in the previous type. The duration range is theoretically unlimited but typically would be of the order of one hundred crankshaft degrees. A range like this would be sufficient for all high-performance needs and more novel matters like engine load control by LIVC (late inlet valve closing) or HCCI (homogenous charge compression ignition). The cam is reportedly somewhat tricky and expensive to make requiring very accurate helical machining and careful assembly. The helical camshaft does appear to be, of the all the basic classes and subclasses, the one system that could be regarded as, for instance, a car enthusiast’s idea of what an ideal variable duration camshaft should be like – a cam that literally has a usable range of variable duration operating through followers of conventional design. It is slightly surprising that it has not been used commercially to date – at least in some sort of specialised application like racing or engine research.
This method of valve control is often found in camless engine designs. A proponent of this technology is Valeo, which has indicated that its design will be utilized in volume production by 2009
In this design the valves are opened and closed and held open or closed by means of electromagnets.
Some of the problems which may be encountered with this methodology are:
- Deceleration of the valve once set in motion is difficult to accomplish. This difficulty is exacerbated by the greatly increased magnetic pull of the magnets as the valves approach the end of their travel (magnetic pull increases dramatically as the distance from the magnet decreases). Inadequate slowing down of the valve can cause significant deterioration of the valve seat and other parts. Utilizing springs to effect valve deceleration limits the engine to lower speeds and on its own will not effect a gentle landing of the valve on its seat at all engine speeds.
- Springs utilized in this type of system may require very careful balancing with the valve movement in order to achieve gentle valve seating at differing engine speeds. As the springs deteriorate or the engine speed changes, the valve and spring balance may be compromised and ultimately lead to failure.
- The electromagnets will draw a significant amount of electrical energy, which may require a higher capacity alternator, which will in turn reduce the potential fuel efficiency of the engine, though other sources of battery recharging such as regenerative braking may help in a more fuel-efficient manner.
- Solenoids do not respond well in high temperature environments, increasing their resistance by 0.39% for every 1°C increase in temperature. Exhaust valves get very hot with surface temperatures over 1000°C. Thus cooling the exhaust valve stem is a substantial problem.
This type of valve control has been advocated in the search for a camless engine. Sturman Industries, which incorporated its design into a large truck engine a number of years ago, is a proponent of this technology. (The truck did the hill climb at Pikes Peak)
Various methods have been explored to utilize hydraulic mechanisms to move the engine valves. Some claim to be successful at low engine speeds, but few claim to achieve that goal meaningfully at the higher RPM requirements of passenger vehicles.
Hydraulic systems suffer from two inherent problems :
- The faster a liquid is moved, the more it tends to act like a solid. A fast-acting hydraulic system to activate automotive valves at the speeds required in passenger vehicles could require immense pressures, with all the incumbent problems, including the additional energy requirements of the hydraulic pump. Even if higher engine speeds were achieved, valve movement would likely be abbreviated and not fully follow the desired or optimum lift schedule.
- Temperatures can vary seasonally over a wide range. The hydraulic medium could change viscosity as the temperatures change, which could cause variances in the system's performance which may be difficult to control.
Utilizing springs to assist the hydraulic system may also prevent the engine attaining higher speeds.
In order to achieve gentle valve seating, hydraulic systems must be carefully controlled. This control may require the use of powerful computers and very precise sensors.
This methodology of valve control has previously not been successful in camless engine design due to the limited RPM range inherent in the design.
The following link indicates that some measure of success has been achieved and a camless design for a medium to slow revving engine may be feasible.
Valves that open and close in fixed times cannot optimize engines running at differing speeds and importantly severely restrict engine speed. This is because the degrees of rotation for the valve events increase as the engine speed increases to the point where they are no longer practical.
Powertrain claim in their documentation at the above link that their device operates the valves “to 8,000 rev/min and beyond”. This figure was likely obtained under laboratory conditions. Real world engines using their devices with 7 ms valve open and close times may have useful speeds limited to well under 6,000 RPM, perhaps even under 5,000 RPM (7 ms at 8,000 RPM is 336 degrees of rotation which is currently unworkable in passenger cars). The engine may be able to run at higher RPM by limiting the lift and hence shortening the valve open close time but will most likely have a lower power output than is achieved at a lower RPM due to the lessening of breathing capability.
Systems utilizing pneumatics to drive the engine valves would in all probability not be feasible because of their complexity and the very large amount of energy required to compress the air.
Cargine Engineering AB, a Swedish Company, has produced pneumatic valves and has fitted them into several different engines. One of these test engines is running in a test Saab 9-5.
The first prototype of the Scuderi Engine uses Cargine pneumatic valves for the intake and exhaust valves.