Traditionally valves on four stroke reciprocating piston engines have been opened with camshafts in any one of four different configurations. Some recent engines have used complex variable valve timing systems to better match the opening and closing mostly of the intake valves to the requirements of the engine over a broad range of engine speeds, and higher levels of sophistication are also possible.
Faster is Better
Lifter Limitations
Lighter is Faster
Overhead Camshafts
Roller Camshafts
Varialbe Valve Timing
Camless Engines
The gradual opening and closing of intake and exhaust valves means that the valves are fully open for a short duration where maximum flow can take place and that the valves stay open a small amount over a long duration. At high engine speed the short duration of maximum valve lift hurts performance because the intake air has less time to get into the engine and the exhaust has less time to get out of the engine. At low engine speed the long duration where the valves are open a small amount tends to interfere with the operation of the engine. Particularly the intake valve staying open long into the compression stroke severely cuts into the compression pressure of the engine when starting and running at low engine speed. Opening and closing the valves faster alleviates these problems and allows an engine to run well over a wider range of engine speeds. Getting the valves to open and close faster though is a challenge at high engine speeds. Fast opening and closing valves benefit greatly from a lightweight valvetrain, and there are limitations as to how fast a particular type of camshaft can open and close the valves regardless of the weight of the valvetrain.
The most common type of traditional camshaft used lifters, cylindrical barrels that ride on the camshaft lobes and are contained in lifter bores. The reason that lifters are a problem is that they severely limit the rate of lift for a particular diameter camshaft. For direct acting lifters, as are used on flathead engines and direct acting overhead valves, the maximum rate of lift is proportional to the diameter of the camshaft. This is why modern engines with direct acting overhead camshafts use rather large diameter camshafts. On a pushrod engine that has lifters and rocker arms a smaller diameter camshaft can open the valves faster and farther with the use of ratio rockers. Ratio rockers open the valves approximately 1.5 or 1.7 times faster than the lifters lift. The problem with this arraignment with ratio rockers is that the small diameter camshaft is then heavily loaded and very aggressive camshafts tend to wear out quickly. Just as with direct acting camshafts a larger diameter alleviates these problems. A larger diameter camshaft however requires larger diameter lifters which are heavier and require stiffer springs that in turn require stronger pushrods. The heavier springs and heavier pushrods then put more stress on the camshaft meaning that a milder cam profile is required. A vicious cycle with no way out. The best solution for pushrod engines are auxiliary springs that act on the lifters themselves. These auxiliary springs have traditionally been known as a rev kit, since they never came stock on production engines.
Anything that can be done to reduce the weight of the valvetrain of an engine generally is highly beneficial for getting the engine to run well over a wide range of engine speeds. The number one thing that can be done to reduce the weight of a valvetrain is to eliminate the pushrods. This is a tricky distinction because the pushrods themselves are not all that heavy. Removing the pushrods however lightens a valvetrain much more than just the weight of the pushrods themselves. The reason for this is that a pushrod engine requires both lifers and rocker arms, where an engine without pushrods requires only lifters or rocker arms. Direct acting overhead camshafts have only lifters without rocker arms and work quite well, but they do need to be quite large in diameter which adds weight and bulk up high on an engine just where it is normally not wanted. A better solution is rocker arms that ride directly on an overhead camshaft. This is about as light of a valvetrain as can be attained, but more importantly it is more efficient and more adaptable than direct acting overhead camshafts.
Rocker arms ridding on an overhead camshaft obviously allows a smaller and lighter camshaft to provide a large amount of lift with fast opening and closing valves. This is not only because of the ratio provided by the rocker arms, but is also due to the fact that the followers are able to ride on somewhat more aggressive cam profiles than lifters can. Another advantage of an overhead camshaft with rockers is that any type of valve configuration can be driven off of a single camshaft for each head. The single overhead cam can drive two canted valves per cylinder, four parallel valves per cylinder or even four canted valves per cylinder. A single direct acting camshaft on the other hand can only drive two parallel valves per cylinder. Canted valves or four valves per cylinder require two direct acting overhead camshafts per head. For more on canted versus parallel valves see Canted Valve Engines.
Even if a valvetrain is light enough to easily operate at the required speeds with the desired rate of lift of the valves there are still the geometric limitations of how steep the lobes can be on the camshaft. Larger diameter camshafts and ratio rocker arms can circumvent these limitations, but there is another way to allow steeper cam lobes as well. Instead of a lifter or follower simply ridding on the cam lobes a roller lifter or roller follower can instead be used. Roller lifters and roller followers add weight to the valvetrain, but the advantages of lower friction and lower wear are generally worth the extra weight.
When roller lifters are used on a pushrod engine the extra weight and faster opening and closing of the valves makes a rev kit highly desirable. With the use of a rev kit and roller lifters a pushrod engine then has a valvetrain that is almost as good as an overhead camshaft with rockers. The only disadvantage then to the pushrod engine is the extra weight of the pushrods and the higher load on the camshaft from the heavy roller lifters and the rev kit springs to keep them in check. If a rev kit is not used roller lifters can still be an advantage for a pushrod engine, but catastrophic valvetrain failure at very high engine speeds becomes a significant concern.
A better rollerized valvetrain uses roller followers on an overhead camshaft. This valvetrain remains very lightweight and has all the advantages of lower friction and little limitation on how steep the camshaft lobes can be made. Heavy duty rockers for fast opening and closing valves at very high engine speed might seem to add a lot of weight to the valvetrain, but the weight of the rockers is less significant the closer to the pivot it is. The weight of the rockers out at the ends is just as significant as the weight of the valves, springs, retainers and locks, but the weight halfway out on the rockers is only half as significant. Roller follower valvetrains can do amazingly better with fast opening and closing valves at high engine speeds than other types of valvetrains. So much so that a roller follower valvetrain tends to look good enough for just about any type of engine, no matter how wide the range of engine speeds that is required.
On an engine with slow opening and closing valves the best thing that can be done for good performance is to match the timing of the closing of the intake valve to the speed of the engine. At lower engine speeds the intake valve needs to fully close early enough that full compression pressure can be attained with no backflow out of the intake valve during the early part of the compression stroke. At higher engine speeds the intake valve needs to stay open longer to allow time for the delayed movement of the intake air to flow into the cylinder. A faster closing intake valve alleviates these problems, but an even better solution is to actually change the timing of the intake valves for different engine speeds. Gasoline racing engines typically have quite long duration camshafts and often have intake valve closing times so late that maximum torque production at lower engine speeds is severely compromised. Diesel engines and many older automotive gasoline engines on the other hand have such early closing intake valves that the engine has a hard time flowing at more than about 5,000RPM.
The modern sophisticated solution used on many recent automotive engines is a variable valve timing system that actually changes the timing of the intake valve as the engine speed changes. The two relatively simple ways of accomplishing this were either with the use of an electro-hydraulic shifting mechanism inside the drive gear for the intake camshaft or with the use of a second intake lobe and a deactivating rocker arm.
The variable timing mechanism on intake camshafts is fairly straight forward, but it does require a high enough capacity hydraulic cylinder built into the cam drive to shift the timing and hold it there while the engine is running at fairly high engine speed. Shifting the entire timing of the intake cam so that the intake valve stays open longer at higher engine speeds is beneficial, but opening the intake valve later does not accomplish anything. A better way to get the intake valve to stay open longer at high engine speeds is to actually increase the intake duration.
An additional intake lobe and a second intake rocker arm allows the intake duration to be increased for higher engine speed operation. This is accomplished with a hinged rocker arm that is capable of turning on and off, usually with a hydraulic lock of some kind. Engaging a second longer duration, and usually slightly higher lift, intake lobe at high speed means that the intake valve does not have to open later but can stay open much longer into the compression stroke.
Variable valve timing systems on the exhaust camshafts are also often used, and provide some benefits for getting an engine to run as well as it can over a wide range of engine speeds. Just as the size of exhaust valves is not terribly critical on normally aspirated engines the timing of the exhaust valves is less critical than the timing of the intake valves. Generally the same thing is true of exhaust valves that is true of intake valves, they need to close early for low speed operation. Closing the exhaust valves a bit early for high speed operation however does not cut into power production potential all that much. A variable valve timing system on the exhaust is however beneficial for getting an engine to run as efficiently as it can over a wide range of engine speeds. Particularly on forced induction engines leaving the exhaust valve open a bit longer at high engine speed means that less power is required to force the hot exhaust gases out, making the engine more efficient at maximum power output.
The biggies for valve timing are the times of closing of the intake and exhaust valves, the valves need to close earlier for low speed operation. Ultimately though the timing of both the opening and closing of both the intake and exhaust valves would have a certain ideal value for any engine speed and load combination. Shifts in the timing of opening and closing of the valves based on both the engine speed and the engine load can potentially allow an engine to run more efficiently over a wider range of engine speeds and loads. Providing the perfect valve timing for any engine speed and load combination requires a fully variable valve timing system. A combination of multiple cam lobes and multiple rocker arms for each valve along with a camshaft timing mechanism on the drive of each camshaft could potentially provide this fully variable valve train. The complexity of all of these different timing mechanisms is however substantial, and a more appealing fully variable valve timing system is the camless engine.
A camless engine does not use camshafts for opening the intake and exhaust valves, but rather relies on some form of hydraulic or electric drive mechanisms for the valves. Prototypes of both hydraulic and fully electric camless engines were built in the late 20th century, and the demonstrated advantages in efficiency and performance over a wide range of engine speeds and loads were significant. The limitation of the hydraulically actuated valves was that they just would not open and close fast enough for high speed engine operation. The hydraulically driven valves worked well on diesel engines that operated at less than 2,000RPM, but those rather slow engines never had much trouble with valve timing in the first place.
For improved high speed performance the camless engine needs electrically driven valves. The early electric valve drive systems used simple electric solenoids to open and close the valves. This worked well for getting good valve timing up to very high engine speeds, but the electrical consumption of the huge solenoids was extremely excessive.
A more efficient camless engine would use not simple solenoids to drive the valves but rather would use electric motors. This could be in the form of a rotational electric motor geared to the valve by an eccentric, but a better solution would be linear electric motors. A linear electric motor could act directly on the valve stem, or it might act through a ratio rocker arm so that the motor could be located off to the side. In all likelihood the best arrangement would be a direct acting linear electric motor built into the valve stem area. The key to getting a linear electric motor to develop high driving force is for it to be long with multiple magnets in the valve stem. With a large number of alternating magnetic poles along the valve stem the linear electric motor can smoothly generate a large driving force at any valve position.
A significant disadvantage to most camless engine designs is that the driving power is inevitably higher than for a camshaft acting against springs. As the cam forces the valve open there is a large driving power requirement, but once the valve begins to close some of the energy stored in the compressed spring is used to drive the camshaft forward. If friction is kept low with roller followers then the net driving power requirement of the valvetrain remains rather low even at high engine speeds. The prototype camless engines from the 20th century had no means of recovering power from the valve springs. Power was required to open the vales, and in the case of the electric solenoid valve trains, more power was required to pull the valves shut in the other direction.
An electric motor driven camless valvetrain could however work against valve springs and harvest the stored energy in the springs similarly to the way that a roller follower valvetrain works. The way that this would work would be for the linear electric motors to work as linear electric generators as the valves are closed by the valve springs. At high engine speed the linear electric motors would accelerate the valves off of the seats, and then the inertia of the moving valves would open them the rest of the way and some of the driving power required would be stored in the compressed springs. As the valves began to close the energy stored in the valve springs would accelerate the valves up to high speed, then as the valves neared their seats the linear electric motors would work as generators slowing the valves and recovering a portion of the energy that was stored in the springs.
There are however some severe problems with the implementation of this type of electric motor driven valvetrain. One is that the capacity of the linear electric motors would have to be rather large to rapidly accelerate and decelerate the valves at high engine speed. Another problem would be that holding the valves at maximum lift for an extended period would require that electrical power be wasted just holding the springs open.
Large capacity motors being required to rapidly open the valves at high engine speed is problematic from two different perspectives. For one it is simply a challenge to build such high power linear electric motors into the valve stem area of an engine without the valve stem sticking farther up above the engine. This problem can be partially circumvented simply by using a larger diameter valve stem with bigger and stronger permanent magnets riding inside larger higher capacity windings. The other problem with large capacity motors though is that they only run at maximum output for a short period of time as the valve is accelerated off of the seat, and then they run as high capacity generators again for only a short period of time as the valve slows towards it's seat. The rest of the time that the valve is moving and the linear electric motor/generators are not being used some power is being wasted by the permanent magnets inducing eddy currents in whatever they are passing by.
The problem of large capacity electric motors being required cannot really be avoided, but the problem of electric power being wasted to hold the valves fully open for an extended period of time can be dealt with in two different ways. One solution is simply to keep the valve duration short at low engine speeds and low power output. In this way normal operation involves simply accelerating the valve up into the spring with no power being used to hold the valve open. This is however only a partial solution since high torque generation at lower engine speeds would require that the valves be held open for substantial lengths of time. One way that higher torque generation at low engine speed could be provided would be by bouncing the valves up and down against the spring pressure. To provide somewhat increased torque at low engine speed while keeping the valve train as efficient as possible the valves would be bounced from nearly closed up to somewhere near full lift without the valves touching the seats. To provide near maximum torque at low engine speed the valves would have to be held farther open, which would require bouncing the valves from about half lift to fully open and this would waste more electrical power.
Another way to avoid having to waste electrical power to hold the valves wide open against the valve springs would be simply to not use valve springs at all. This would mean that the valves would be accelerated with the linear electric motors and then decelerated towards maximum lift using the linear electric motors as generators. With no valve springs a minimum of electrical power would be required to hold the valves fully open at low engine speeds. This sounds great, but there is then the problem of providing sufficient seating pressure to get the valves to reliably stay sealed. During the compression and power strokes there is no problem with the valves staying seated as the pressure in the cylinder easily holds them shut. On the intake stroke though it would be necessary to waste electrical power in holding the exhaust valve closed against the intake vacuum. On a forced induction engine electrical power might also have to be wasted in holding the intake valves closed against the boost pressure during the later part of the exhaust stroke. The reason that the intake valves might have to be held closed against the boost pressure would be to avoid the cylinder filling with air early as that air would then just have to be forced out the open exhaust valve as the piston neared top dead center. In all likelihood this problem of keeping the intake valves closed would be much less of a power consumer than holding the exhaust valves closed against the intake vacuum. And of course the exhaust valves would not require much power be wasted in holding them closed when the engine was operating under boost. It could then be said that an electrically driven camless valvetrain has the potential to be significantly more efficient on a forced induction engine running under boost. On a normally aspirated engine it is unavoidable that some significant amount of electrical power be wasted in holding the valves either open or closed depending on whether valve springs are used or not.
On anything other than a forced induction engine running under substantial boost it seems unavoidable that a camless engine will waste a considerable amount of electrical power. If that electrical power is generated with an efficient permanent magnet generator the loss in overall efficiency might not be all that high. After all the power required to drive a valvetrain really is extremely small compared to other losses on a reciprocating piston engine.
The advantages of a camless engine are compelling, and the level of flexibility provided by eliminating the constraints of camshafts is nearly limitless. The really big advantage that this limitless flexibility could provide would be the ability to run an engine either as a two stroke or as a four stroke. All in head valve two stroke engines are problematic in that it is difficult to get the intake air in and the exhaust gas out through such small openings in such a short period of time. The flexibility of a camless engine though does make an all in head valve two stroke much more appealing. The main problem with two stroke engines is that they are extremely sensitive to valve timing, meaning that most two stroke engines will only run well over a narrow range of engine speeds. Fully variable valve timing would do a whole lot to get a two stroke to run over a wide range of engine speeds and would more than make up for the smaller valve sizes compared to huge cylinder ports.
Of course the really big thing that a camless valvetrain could do for an all in head valve two stroke would be to allow it to start and run at low power output as a four stroke engine. By being able to start as a four stroke the all in head valve two stroke could use just a turbocharger for boost production, and would have no need for a starting compressor. The ability to switch back and forth between two stroke and four stroke operation would also tend to help improve the efficiency of the electric motor driven valvetrain. When running as a four stroke the valves would open and close half as many times for a given engine speed, which would tend to cut the driving power of the valvetrain in half. Essentially the engine would be half as large for light load operation in four stroke mode without the engine speed having to be reduced so much that the piston speed would be radically too slow for efficient operation.