Continuously Variable Valve Actuation system.
Provides a continuous range of valve lifts, from zero to maximum.
It is simple, reliable, light, short, efficient and capable for really high revs.
There is no need for any additional spring (the normal valve springs proved, in practice, adequate for the entire train).
There is no need of any electromotor to rotate the Control Shaft(s), i.e. no need for "drive by wire". The only "wire" involved is the string previously rotating the throttle valve, which now is employed to rotate the Control Shaft(s), and this is all.
Easy in assembly, involves a mere couple of components per valve (actually one and a "half"), plus a control shaft per row of valves.
The system mounted on the top of a Renault 19 Energy 1400 cc engine (without catalytic converter). Spot on the control shaft levers and the string at the right side. The butterfly of the carburetor is locked wide open.
As viewed from the timing belt side.
The angular position of the control shaft defines the lift of the relevant row of valves.
The camshaft acts on the valve by means of a control lever swiveling on the control shaft, and
a valve lever (just a needle) swiveling on both, control lever and valve actuator (bucket lifter or rocker arm or valve stem or �).
The system controls independently intake and exhaust valves, changing the valve lift while keeping unchanged the valve clearance.
Unlike the typical Variable Valve Timing ( VVT ) systems which control the angular overlap at TDC but spoil relatively the status around BDC, the Variable Valve Actuation system ( VVA ) changes not the angular (typical) but the actual overlap at TDC (namely, the "time-valve area" when both intake and exhaust valves are opened together) improving at the same time the status around BDC.
"Drive by wire" is simply an option, not a necessity. Some people might prefer a "drive by wire", even though the mechanism operates as simply as if turning the throttle valve of the conventional engine.
The prototype engine (Renault 19 GTS Energy, 1400 cc, 8 Valves, carburetor, not catalytic, 1989 model) was intended in the first place only to increase fuel economy and reduce emissions, however in the course of the events it turned out capable of providing other qualities as well. Besides being much more fuel efficient and with more than thirty times less Carbon Monoxide in the exhaust gas at idling (relative to the original engine), it proved itself not only quite more powerful, which had been anticipated on the grounds of the wilder camshafts, but surprisingly torquie and reliable, smooth at idling, able to handle extremely lean mixtures, excellent at partial loads, driver friendly with unexpectedly immediate response in transient conditions, etc.
The Carbon Monoxide at idling was officially measured at 0.05% with the engine normally warmed up. The CO is less than 0.1% with the engine completely cold at starting. The normal engine had 1.8% CO in exhaust gas, and this only after the warming up period.
The engine can effectively handle extreme lean mixtures by making them fine "mincemeat", that is homogeneous as it should. On the other hand, the response is immediate, crisp, even with cold engine and without any acceleration pump involved.
The engine can idle at 350 rpm.
The engine has been tested repeatedly at only 7000 rpm and full load, because more revs will damage the conventional parts of the engine (pistons, valves and springs, connecting rods and crankshaft, as the red starts at 5500 rpm and the dark red at 6000 rpm).
The torque at extreme low revs resembles that of a truck, and at high revs that of a motorcycle.
In the following the first and unique prototype is outlined and specified. Though it is actually a hand made mechanism, nevertheless it is surprisingly reliable and the car is currently used for any task. It was (and still it is) tested for thousands of kilometers in town traffic (Athens' traffic), open roads and extreme uphill. The specific engine (with a Weber carburetor and just 2 valves per cylinder) has been selected simply because it was available at the time.
The intake and exhaust control shafts, the control levers (the intake are assembled in their control shaft), and the valve levers (indeed valve needles). Double bearings on control shafts for keeping the control levers: at right side complete, at left partial for inserting the relevant control lever. There are no additional securing means for keeping the control or valve levers in their position.
The previous from the opposite side.
The system schematically at four different control shaft angles (same camshaft angle).
The Control Lever (stereoscopic representation*).
Case of four valves per cylinder (stereoscopic representation*).
A part of a Control Shaft (stereoscopic representation*).
Schematically the mechanism mounted on Renault 19 Energy engine (stereoscopic representation*).
Case of straight four 16 valves (stereoscopic representation*).
*To see stereoscopically the above slides, just hide the left image from your left eye and the right image from your right eye by your palms and then try to concentrate your sight on a small object located at the intersection of the line from your left eye to the right figure and the line from your right eye to the left image. The "software" is already into the brain waiting for activation. The result is stunning.
The stereoscopically viewed objects are formed in the space in front of observer and not on the screen.
The moving diamonds ("Diamonds" at the last page of the site) "leave" the screen and fly in the space, some moments close to observer and some moments away, behind the screen.
The deliberate confusion on the last image of this page clears "magically" and becomes readable only when viewed stereoscopically.
The above system as a unit located on top of the valves,
and looked at from the timing belt side, at zero lift.
What the ideal Variable Valve Actuation system offers are constantly the optimum and practically the same conditions to the mixture (intake manifold pressure, swirl, turbulence, scavenging of the cylinder during overlap, velocity around valve seats, mixture homogeny etc) for entering the cylinder, in order to get burnt as effectively as it gets, and then scavenge the cylinder at all revs and every load.
The rule seems general and quite simple: the necessary valve lift is about linearly proportional to both revs and load.
For the mixture what counts is its own kinematics and thermodynamics, not the speed of the crankshaft, the cylinder wall temperature, the load etc.
In an ordinary engine tuned at 6000 rpm and full load but operating at 1500 rpm and one third of the full load, the entry speed of the mixture into the cylinder is not 2 or 3, but more than 10 times lower. The overlap from blessing at 6000 rpm becomes curse at 1500 rpm. Needless to say more.
The above plot depicts the original valve lifts of the engine (bold orange for the exhaust and bold green for the intake valves) and a group of available lift curves with the VVA system mounted. The vertical axis is the valve lift in mm and the horizontal the rotation angle of the crankshaft, in degrees. The red curves are the exhaust valve lifts for 65, 33, 18, 8, 3 and 1 degrees of exhaust control shaft rotation, while the blue curves are the intake valve lifts. Any red curve can be combined to any blue. Spot on actual (not simply the conventional angular) overlap: at high lifts it becomes more than double of the conventional engine while at low lifts it becomes negligible.
An ordinary VVT system reduces the overlap at low revs and partial loads, retarding intake valves opening and closing, and advancing exhaust valves opening and closing. The retarded intake closing and the advanced exhaust opening are the side effects, which make the situation around Bottom Dead Center completely wrong, yet the need for reduction of the overlap at Top Dead Center prevails.
An intelligent VVT system has to reduce, at partial loads and low revs, the overlap at Top Dead Center, and at the same time it has to retard the exhaust valves opening and advance the intake valves closing. It can be done, but it is complicated and expensive.
A VVT could of course be combined to the presented system, however, as the Variable Valve Actuation mechanism controls effectively the actual overlap too (where the term actual overlap defines collectively the lift - time area) it does not need a VVT system, at all.
In fact, the specific VVA system is an intelligent VVT too, since it controls properly the overlap at Top Dead Center and, at the same time, improves the conditions around Bottom Dead Center, closing, at low lifts, the intake valves substantially advanced and opening the exhaust valves substantially retarded (even if a valve is completely closed at a specific angle, it is in fact almost closed during many degrees before that typical closing, i.e. when the instant lift of the valve becomes adequately short).
The operation further improves thanks to the substantially constant pressure in intake manifold. As there is no vacuum in intake manifold to pull back (suction) the gases from cylinder and exhaust, the quality of the mixture becomes more controllable resulting in lower emissions, larger allowable overlap and enhanced quality of operation.
Applying this Variable Valve Actuation System the behavior of all engines (racing or ordinary) can be improved at all revs:
- the engine can idle at very low revs steadily, with low noise, vibrations, consumption and emissions,
- the engine can produce, roughly speaking, constant torque at all revs, from extreme low to extreme high,
- the engine's response becomes instantaneous, immediate, and crisp,
- the engine can work perfectly at partial loads,
- the valve train's expected life becomes significantly longer,
- the engine can efficiently burn lean mixtures,
- the consumption and the emissions can be drastically reduced, in particular at low to medium revs, and finally,
- the engine can deliver even more power at high revs using wilder camshafts, simply because the Variable Valve Lift System provides the drivability and friendliness at all revs and loads (the wildness of the camshafts show up only at maximum lifts).
The lightweight of the constituent parts, the diminished internal friction and the "as it should be" loading make the system suitable for most kinds of services, from lorries to racing.
Once mastered the mechanism, it becomes obvious that:
It is simple in operation and manufacturing, made up of very few parts.
There are no wear concentration or stress concentration points. Absolutely none.
The inertia loads from the reciprocating (or oscillating) parts are very low (lower than the loads, in the valve train of the conventional engine, produced from the replaced parts, i.e. bucket lifters or rocker arms).
The swivel joints -valve lever to control lever, valve lever to valve adjuster and control lever to control shaft- are mechanically correct, simply because there is surface contact (and not line or point contact which reduce the expected life and make inevitable often adjustments).
The valve adjuster pushes the valve stem only along the valve guide direction, so the valve guide can long live.
The valve adjuster slides along a "hole/slider". At high lifts the thrust force is negligible as the valve needle reciprocates nearly parallel to the "hole/slider" axis. At low lifts the thrust is small partly because the inertia loads are small, partly thanks to the only-partially compressed valve spring, and partly due to the small angle between valve needle and "hole/slider" axis.
As typical engines operate most of their life at partial loads and low to medium revs (i.e. calling only for low to medium valve lifts) the mechanical friction falls and the expected life of the valve train system becomes much longer than conventional, where the strength of the valve springs -necessary to return the valves to their seats at maximum revs- loads excessively everything (cam lobes, cam followers, valves, valve guides, valve seats, timing belt) at all revs and loads, even at idling.
The mechanism is simpler in manufacturing and operation than "two step" systems offered nowadays from car industries (valve deactivation systems included), as it has no need for tinny pins and holes, as it does not necessitates complicated camshafts with many cam lobes, as it has no need for even a single additional spring and involves absolutely no hydraulic system, piping, pumps etc. Despite its simplicity it offers not only two, but infinite lifts, starting from zero.
On the other hand, compared to the "continuous variable valve lift" systems known today (which are the real competitors) in terms of:
number of parts involved,
number of interfering joints conveying the motion from cam lobe (or eccentric pin) to valve stem,
necessary accuracy of constituent parts (to allow the manufacturing even in ordinary machine-shop),
wear concentration points,
additional inertia loads and friction,
easy of control,
control on both, intake and exhaust valve lifts,
applicability on both, new and existing engines,
completeness, i.e. ability to work without supporting mechanisms (like "variable valve timing", special controllers, etc),
applicability on all engines, from cheap single cylinder, to luxury multicylinder, to racing,
effective manipulation of all loads (full load included) from extreme low revs,
ability for really high revs, in order to make full use of the theoretical VVA's advantage, i.e. improved behavior from extreme low revs to the limit imposed only by the strength of the underneath power train (pistons, connecting rods, crankshaft),
the system seems to stand in a class of its own.
The above plot is the valve lift (vertical axis, mm) of exhaust (red) and intake (blue) valves of the mechanism for 65, 33, 18, 8, 3 and 1 degrees of control shaft rotation. The horizontal axis is the crankshaft rotation. The bold curves are the relevant lifts of exhaust (orange) and intake (green) valves of the normal engine.
As shown, the overlap, not the angular but the effective or actual one, is linearly proportional to the selected valve lift for intake and exhaust.
The substantially constant pressure at intake manifold, combined with the variable actual overlap, changes dramatically the behavior compared to that of ordinary engines. Since there is no vacuum in intake manifold to pull back (suction) the gases from cylinder and exhaust manifold, the behavior of the whole system, from intake to exhaust, is quite different. This explains the surprisingly instant response to the acceleration pedal in all conditions, despite the elimination of the carburetor's acceleration pump (due to constantly full open throttle) and the lean mixture.
The immediate constant idling well below 500 rpm with the engine really cold (after a night at 5 degrees centigrade), without any assistance from chock, is just an indication of the capabilities of the variable valve lift system at low revs.
The resistance for the rotation of the camshafts, as compared to that one of the normal engine, is negligible at idling: the small finger of a child can rotate it, giving another indication of the smoothness at idling this variable valve lift system provides. This negligible resistance proves the elimination of the loads and hence the fatigue, friction and wear of the cam lobes, of the timing belt, of the valve springs and valve seats etc.
The valve lift which at idling is a couple of tenths of a mm makes the turbulence of the mixture suctioned in the cylinder as optimum as full lift makes it at optimum load and revs and its homogeny too, making the idling quiet, constant, with clean exhaust gasses and very low consumption. The engine needs long time to warm up at idling, with less than 500 rpm. Yet, why to wait for the warming of the engine, since the driver can lock the throttle at wide open and start immediately without any response problem?
An interesting matter is the low noise from the exhaust valves at very short lifts. At idling and after removing the exhaust manifold, the noise from the exhaust ports on cylinder head reminds more a calm whistling than the unaffordable tab, tab, tab, tab of the conventional engine.
At idling the exhaust gas was checked. The official result was 0.05 % in volume with normally warmed up engine. The CO was checked with completely cold engine at less than 0.1% in volume. The CO was also metered at full load in the region from 1000 to 6500 rpm and the result was about zero. When the original (without variable valve lift) engine was metered at idling, the reading for CO was 1.8 %, which makes for more than 30 times higher. And this 1.8 % value for the conventional engine was achieved only after the warming up period.
Starting even with cold engine the response is immediate, direct, 1:1, in all conditions. As the throttle is kept constantly open there is no richness of the mixture at transient conditions. But strange as it appears the response is as crisp as if the acceleration pump were in full operation.
The engine is currently running at 7000 rpm with full load. The red line starts at 5500 rpm, and the dark red at 6000 rpm. More revs may damage the basic parts of the engine (connecting rods, pistons, valves and crankshaft).
The top red and top blue curves, in the diagram above, compared to the normal ones, explain why the engine can rev higher. The lift is more (about 15%), the duration is wider and the overlap is about double. At high revs the engine's behavior (response, power output and sound) reminds motorcycle.
At low and medium revs the most torque is taken with intermediate lifts, not the maximum. If the power pedal is pressed more than necessary the torque decreases, as more lift decreases the entry speed and the turb. Drive by wire could help. Drive by wire could also help in rotating the two control shafts at proper angles, according existing conditions. The selected relation of intake and exhaust control shaft angles in the prototype is very simple, as a string pulls both of them.
The proposed mechanism can achieve constant valve clearance in all valve lifts. It can also achieve slightly variable valve clearance, if it is desirable. For instance, if at very low lifts (idling, partial loads and low revs) a valve clearance of 0.1 mm is selected and at high valve lifts a clearance of 0.4 mm seems better selection, then by means of a small modification of the actual lengths of the constituent parts (valve lever, control lever etc) or by a small relocation of the swiveling point between control lever and valve lever, or by a slight modification of the cam follower, or . . . the desirable slightly variable valve clearance can be achieved.
The material of the aluminum plates is of "aviation" quality and cost, and was selected as the only available. The steel parts (control shafts, camshafts, control levers, valve levers, cylindrical push rods/adjusters) were made from cheap low carbon steel and a double nitride surface hardening was applied. The accuracy in dimensions and the quality of the surfaces of the various parts are in "tenths" of a millimeter, as most of them were finished by hand.
The system can be applied to engines having all valves in a single row. In so far as a proportional lift is selected for exhaust valves relative to the intake ones, the application of the method becomes easier, as it takes only one control shaft and only one camshaft.
The system fits to the Diesel engines too. The operation of the Diesel engine can be improved using the Variable Valve Lift System by increasing the turbulence of the air at all revs and by increasing the amount of air entered and trapped into the cylinder.
The system matches also to turbo engines, as the exhaust valve lift can effectively smooth out pulses, which otherwise impact turbine fins at low revs.
The system can easily be applied to engines having pushrods for transferring the cam lobe action to the valves.
The simplicity of the system makes it proper for motorcycle engines too.
Renault 19 Energy (1390 cc, bore 75.8 mm, stroke 77 mm, compression ratio 9.5:1).
The engine is a normal one with the addition, modification or replacement of the following parts:
1. The single camshaft remains in place idle (immovable), only for keeping the oil into the cylinder head covering its end bearings.
2. The rocker arms, their pivot shaft, and the cylinder head cover are removed.
3. In the place of the original cylinder-head cover it is secured (by the eight 6 mm screws previously securing the cover to the cylinder head and the five 8 mm screws previously securing the rocker arm pivot shaft to the cylinder head) a base aluminum plate completely covering the cylinder head. This base aluminum plate serves as the lower bearings framework for the two control shafts. This base plate has eight holes, one above each valve. Each hole has a diameter of 10 mm, coaxial to the relevant valve. Along each one of these holes slides a small cylindrical pin (acting as both, push rod and valve clearance adjuster) being of 10 mm diameter and about 14 mm height, having at its top end a hemispherical cavity of 6 mm diameter which acts as the bearing of the relevant valve lever (valve needle).
4. Two intermediate aluminum plates, one for intake and one for exhaust valves. Each of them acts as the top bearing for the relevant control shaft and also as the lower bearing for the relevant camshaft.
5. Two top aluminum plates, one for intake valves and one for exhaust valves. On each one of them the top bearings, for the relevant camshaft, are cut.
6. On the base aluminum plate it is secured by ten 8 mm screws the intermediate and the top plate of intake valves, and with another ten screws of 8 mm are secured the intermediate and the top plate for exhaust ones.
7. On the base aluminum plate there are holes for the "oil bath" type lubrication, of the involved parts. The central screw, for securing the base plate to the cylinder head, is properly drilled to deliver the pressed and clean oil. The oil level in the eight chambers formed into the aluminum plates is kept by a number of drain holes. When the oil exceeds the desired level, it is drained into the cylinder head through the holes and then into the crankcase. The aluminum plates form also the top cover and oil sealing means of the engine (see photos of real thing).
8. A control shaft for the intake valves and a control shaft for the exhaust valves are added. Each one is rotatably supported between the base aluminum plate and its relevant intermediate aluminum plate. Both control shafts exceed for about 15 mm from the aluminum plates, at the flywheel side. At this end of each control shaft it is secured a lever for its rotation. The control shaft has pairs of holes of 12 mm diameter for bearing the control levers at an eccentricity of 25 mm from its base axis. All these holes are as coaxial as could be managed.
9. For each valve there is a control lever swivelably coupled at the relevant 12 mm eccentric holes on the control shaft, abovementioned. On the control lever and at an eccentricity of 25 mm from the swiveling axis is formed a hemispherical cavity of 6 mm diameter. On the control lever there is also formed the cam follower, which is a cylindrical surface of 35 mm diameter (at an angle of about 120 degrees) with axis passing through the center of the relevant hemispherical cavity.
Click on the photo to download a 2.5MB video (QuickTime format) showing the Renault VVA-Rod-version engine increasing revs without load.
10. For each valve, a valve lever is provided. It is a 4 mm diameter needle ending at both ends in 6 mm diameter spherical surfaces. The centers of the two spherical surfaces are at 25 mm distance. The upper spherical surface of each valve needle is swivelably supported on the spherical bearing of the relevant control lever, while the lower spherical surface of each valve needle is swivelably supported on the spherical cavity of the relevant cylindrical pushrod.
11. One camshaft for intake valves and one for exhaust valves, each one having four cams, are provided. The cams have 30 mm base circle and 10 mm maximum eccentricity. The duration of the camshaft is significantly longer than the normal camshaft (diagram below). The intake camshaft drives also, at its flywheel side, the distributor. Each camshaft is rotatably supported between its relevant intermediate and top aluminum plates and at its opposite to the flywheel side has been secured its sprocket (the normal one, having 38 teeth).
12. The rotation angle for the control shaft from minimum valve lift (at idling) to maximum valve lift (at max revs and full load) is 65 degrees. The maximum valve lift available is 10.3 mm (wider than the 10 mm of the cam lobe size).
13. The timing belt is removed and another one, having 128 teeth, is used to drive the sprockets of the two camshafts from the existing sprocket on the crankshaft. The existing timing belt tensioner is kept in charge. As the normal width of the timing belt is 17 mm and the only available (Renault Laguna 1.8) 128 teeth timing belt of same form and module has 25 mm width, the last one was sliced in two belts, one 17 and one 8 mm width.
14. The distributor is the original one, but it is now mounted in a new aluminum base secured between the top and intermediate aluminum plates, at flywheel side.
15. The fuel pump is the normal one. It is moved 10 mm backwards, by the insertion of a proper gasket, to make room for the base aluminum plate. The fuel pump arm is driven by an eccentric cylindrical cam, cut on the intake camshaft, by means of a "J" shaped wire.
16. The carburetor is the normal one. The butterfly is fixed permanently wide open. The chock is idle (i.e. not used). The richness of the mixture is not modified and it seems that at very low revs it becomes too lean, as the carburetor is designed not for this kind of use (fully opened butterfly at very low air speed). The accelerating pump is idle too, as there is no butterfly motion, yet the response in all conditions is immediate, direct and instant. Of course the last thing that matches such a design is a poor Weber carburetor.
17. The spark is retarded as many as 7 degrees, by relocating the relevant sensor on the flywheel cover. The high voltage cables are the original as well as the air and oil filters.
18. Silicon gasket glue is used between the aluminum plates for oil sealing.
19. The same string, previously activating the throttle of the carburetor of the original engine, is now used to transfer the commands from the accelerator pedal to both control shafts. The intake and exhaust control shaft rotation angles needs not be necessarily equal. There is no returning spring, neither for the control shafts nor for the control levers since valve springs manage to take care for all of these.
20. The oil level stick is removed and from its metal pipe the crankcase gases, by means of an elastic pipe, are forwarded to the carburetor.
As there is no vacuum in the intake manifold, there is no servo in the braking system. The brakes pedal is too hard, yet it manages to stop the car, provided driver's foot is sturdy.
Depending on the revs of the engine, more torque can be obtained with partial valve lift instead of maximum. And this is exactly the driver's feeling. Only at high revs and full load the maximum valve lift seems necessary.
The camshafts are made deliberately wild in order to improve the power output at extreme revs, as the torque at any lower revs is guaranteed by the correct control of the valve lifts.
Move mouse above an image for content
Mouse click an image to download the relevant photo, video or animation.