PATTAKON

Greece

Variable Valve Actuation system.

A general view.

If the VVA is still a mystery, a look at the 'VVA mechanism' and at the 'equivalent cam lobe' drawings below may help.

The VVA mechanism can be realized in many ways, as shown in www.pattakon.com .

It uses only one cam lobe to activate a valve or a group of valves (for instance the three intake valves of a cylinder).

The VVA is nothing but a linkage between the cam lobe and the valve, as shown in the drawing. As the cam lobe rotates the reciprocation of the valve varies continuously from actually no motion to wildly opened valve, according the angular position of the control shaft.

The control shaft is not shown. It is coaxial to the roller cam follower when the last is in contact to the basic circle of the cam lobe.

The duration and the angles of valve opening and valve closing are constant. This means that the angular overlap is unchanged, as well as the closing of the intake valves and the opening of the exhaust valves. Nevertheless the actual overlap (or time valve area during overlap) changes from almost null to wild, depending on the valve lift selected. Also the actual opening of the exhaust valves and the actual closing of the intake valves is drastically affected by the selected valve lift.

VVA mechanism : a linkage between the cam lobe and the valve

The VVA is like having a continuous sequence of cam shafts, like those shown below, and use each moment the chosen one.

In the simplest realization the driver's foot presses the gas pedal, the cable from the gas pedal rotates for some degrees the control shaft and the engine operates (feels) like having a totally different camshaft.

The engine with the VVA can eliminate the throttle valve as the intake valves do the throttling.

The comparison of the three following plots of 'Valve Time Area' is illustrative of the operational differences between the conventional and the VVA engine.

The Valve Time Area, in case of conventional, for any load

 The Valve Time Area, in case of the VVA at full load

The Valve Time Area, in case of the VVA at 4.000 rpm and partial loads

In the conventional engine the time-valve area offered to the air or mixture at low revs for entering and for leaving the cylinder is huge compared to high revs. It is also independent from the load.

On the contrary, the time-valve area offered to the air or mixture for entering and leaving the cylinder in the case of the VVA is about constant versus revs, and it decreases strongly as the load becomes lighter.

As regards the overlap, the conventional needs additional VVT (Variable Valve Timing) system to limit the overlap at low revs and partial loads, otherwise the actual overlap (or the valve-time area during overlap) becomes huge at very low revs compared to high revs. On the contrary,  the VVA engine has strict control on the actual overlap.

If the VVA has at high revs with full load the same wild actual overlap with a conventional engine, then at low revs the overlap of the VVA becomes many times lower than the conventional, and many-many times lower in case of partial load.

To understand the time-valve area plot of the VVA, just think that at 7.500 rpm the mixture passes through the completely opened intake valves (having, for instance, 10 mm valve lift), and during the about 0.005 sec ((60/7500)*(2/3)) of the suction process, a good quantity of mixture enters into the cylinder. When the engine operates at only 750 rpm (one tenth of 7.500 rpm) the 0.005 sec of the suction process at 7.500 rpm, becomes 0.05 sec, that is ten times more, making an intake valve lift of about 1 mm adequate, at full load. And if the load of the VVA is one quarter, then the necessary intake valve lift at 750 rpm is around 0.25 mm.

Roughly speaking: for valve lifts lower than a quarter of the valve diameter, the flow coefficient is linearly proportional to the lift. In other words, with the same pressure difference before and after a valve, the flow is linearly proportional to the valve lift.

From the time-valve area plots it is also apparent that in the VVA at partial loads and at medium to low revs (where the engine will spend almost all its 'life') the valves move softly compared to conventional.

It is also apparent that in the VVA at normal operation the valve springs are only partially compressed (note that the restoring force from a typical valve spring with the valve closed is about 2.5 to 3 times less than the restoring force with the valve at its maximum lift).

The conventional engine compared to the VVA seems inflexible and primitive as :

it uses constant restoring force on the valves, no matter if the engine revs at 7.500 rpm or at 500 rpm, resulting in high friction, wear and roughness,

it uses constant intake valve lift making, at medium to low revs, the entry speed low and the turbulence, swirl and homogeny of the mixture poor, which in turn means slow flame propagation, poorer efficiency and inability of processing lean mixtures,

it operates with completely improper actual overlap at medium to low revs and at partial loads, making necessary the presence of additional VVT system, etc.

From the 'VVA at partial loads' plot, it is apparent the way the VVA operates at medium and light loads. At 4.000 rpm the engine can operate at full load with a valve lift of  about 6 mm to produce its maximum torque there. As the valve lift decreases, the engine can still operate at 4.000 rpm producing only a part of the maximum possible torque. The actual overlap at 4.000 rpm and 1/5 of the full load becomes about five times smaller compared to the actual overlap at 4.000 rpm with full load.

According the 'equivalent cam lobe' drawing, with 8 degrees of control shaft angle, the valve lift is about 1.5 mm. This means that the VVA engine sees an extremely mild cam lobe (there is no camshaft in commerce for valve lift of only 1.5 mm), while the conventional has to use the same cam lobe at high revs and at medium and at low revs.

When the control shaft is rotated at 18 degrees, the resulting valve lift becomes about 3.5 mm. This valve lift seems small, but actually it is more than necessary for normal city traffic. Porsche's variocam plus uses 3 mm valve lift for the intake valves bellow about 4.500 rpm, and then it uses the 10 mm lift.

As the control shaft rotates at 33 degrees, the valve lift is about 6.0 mm, and when the control shaft rotates at 65 deg, the valve lift is about 10.5 mm.

These lifts are necessary only when the driver needs a lot of power. Even in racing cars the engine operates at maximum revs with full load only for a part of the total time.

Roughly speaking:

A force on the circumference of the sprocket makes a travel of about 50 mm (to rotate the camshaft for about 60 degrees) to move the valve for just 0.2 mm (idling), in VVA case.

The valve spring is significantly softer (about three times) at small lift compared to full lift.

If the valve spring force, with the valve closed, is 20 Kp, the necessary force on the sprocket (the force on the timing belt) to rotate the camshaft at idling is only 20 kp * 0.2 mm / 50mm, that is about 0.08 Kp.

For 10 mm valve lift the force on the sprocket becomes 50 * 3 = 150 times stronger (50=10mm/0.2 mm, and 3 because at maximum lift, the spring force is about triple compared to the spring force with the valve in contact to its seat), that is 12 Kp.

In both cases the friction into the transmission from the cam lobe to the valve was taken zero.

In case the friction loss into the valve train is proportional to the loads, the VVA at short lifts consumes many times less energy compared to conventional valve train. It is said that from the energy given to lift the valves, only 80% is returned (some insist it is less than 30%) as the valve springs expand after their compression, the rest is lost as friction.

The 'operational area' plot below (valve lift versus revs) has a maximum permitted valve lift curve (red line) and a minimum permitted valve lift curve (blue line).

It is common sense that 'increasing the valve lift, the duration and the overlap, more torque is produced', but it is wrong.

With its valves activated by the wild cam lobes, the engine of Honda S2000 produces at medium to low revs significantly less torque compared to the torque produced when the valves are activated by the mild cam lobes.

The mild cam lobes of S2000 have significantly lower valve lift (about 2/3) and less duration (that is less valve time area) and less overlap. Nevertheless with the mild cam lobes it is produced at medium to low revs more torque with better response, with lower consumption and with cleaner exhaust.

Back to the VVA.

At 4.500 rpm, the best torque is produced with a valve lift of about 7 mm. If the valve time area is increased selecting a valve lift of 9 or 10 mm, the engine produces less torque and operates worse.

Why?

More valve time area results in lower entry speed, in less homogeny of the mixture, in lower quantity of mixture finally trapped into the cylinder (because a part of the mixture has the time to escape back to the intake manifold when the piston moves upwards with the intake valves still opened), in less turbulence and swirl into the combustion chamber, in SLOWER PROPAGATION OF THE FLAME, in LOWER EFFICIENCY of the cycle, in hotter exhaust gas,  in more energy lost at exhaust (because the exhaust valves open widely sooner than necessary) and in more pollution.

A mechanism to prevent the 'more than necessary' valve lift is quite simple

The min-max lever is rotated for a few degrees by a control motor, according present revs.

The system is quite different than the 'drive by wire' systems. The driver's foot presses the gas pedal and consequently rotates the control shafts, selecting directly the desirable valve lift. The min-max lever is there just to prevent the 'more than necessary' valve lift. In other words, the min-max mechanism is a 'brake', not a mover. If this min-max mechanism is removed, the engine still operates (in the worst case it operates as the conventional engine).

Without min-max control, the feedback (response) to the driver is enough for him to achieve the best operation, as happens in the Renault 19 prototype.

The same min-max lever can control the minimum permitted valve lift, as well.

The use of this minimum permitted valve lift is to allow the VVA mechanism to operate safely at high revs without additional restoring springs. For instance, at 10.000 rpm the restoring force on the VVA mechanism from the original valve springs is not adequate, when the valve lift is at only 0.2 mm, to keep the cam follower in permanent contact to the cam lobe.

But a valve lift of 2.5 mm at 10.000 rpm offers to the VVA mechanism the necessary restoring force, coming from the original valve springs, to keep the cam follower always in contact to the cam lobe.

Even without load, a valve lift of 2.5mm cannot increase the revs to 10.000. When the engine is revving without load at 10.000 rpm and the valve lift is decreased to 2.5 mm, the engine 'decelerates' because the friction exceeds the power produced at expansion cycle (the cylinder is only partially filled with mixture while the total friction is significantly increased).

On the other hand, if the VVA mechanism uses additional restoring springs, there is no need for such control of the minimum allowable valve lift. But the min-max mechanism is so simple that seems preferable.

Let's follow the typical operation of the VVA.

The engine idles at 400 rpm, where the min-max mechanism does not allow more than 0.8 mm valve lift.

As the driver presses the gas pedal, the revs increase rotating a little the min-max lever to permit more valve lift, and so on.

If the car starts moving on a strong uphill, with its revs constantly at 500 rpm, the engine uses 1 mm valve lift at most, because with this valve lift it is produced the maximum possible torque at 500 rpm.

If the valve lift is increased to 10 mm at 500 rpm, the engine becomes unable even to idle, not to produce torque to push the car.

When the revs increase at let say 1.800 rpm (very often case in city traffic) the min-max control lever does not allow more valve lift than 3.0 mm. It also does not allow less valve lift than 0.4 mm. The driver can select any valve lift between 0.4 to 3.0 mm. The min-max lever at 1.800 rpm does not allow any valve lift more than 3.0 mm or less than 0.4 mm. Using the 3.0 mm valve lift, the driver achieves the most possible torque at 1.800 rpm. Pressing the gas pedal less, the valve lift decreases to 1.5 mm and the engine produces about half of the maximum torque at 1.800 rpm. If the min-max mechanism is removed to allow full valve lift (10 mm) at 1.800 rpm, things get much worse, by any point of view.

When the revs increase to 4.500 rpm, the min-max lever allows valve lifts between 0.8 mm and 7 mm. The 7 mm is for the maximum possible torque at 4.500 rpm, the 0.8 mm is for providing the necessary restoring force from the valve springs to the VVA mechanism, in order not to use additional restoring springs to keep in contact the cam follower to the cam lobe.

When the revs increase above 6.000 rpm, the maximum valve lift is the maximum possible valve lift defined by the valve train design. So above 6.000 rpm, the min-max mechanism does not limit the maximum allowable valve lift. The min-max mechanism at more than 6.000 rpm controls only the minimum allowable valve lift for keeping the VVA mechanism free from additional restoring springs.

Above 6.000 rpm, the VVA mechanism seems to operate as a conventional engine. Actually the VVA prevails and in this high revs area, too, as it can use much wilder cam shafts than conventional, to achieve even more power. This is because a conventional using extra wild cam shafts cannot operate acceptably at medium to low revs.

Selecting properly the shape (profile) of the control arm and of the min-max lever, the desirable form of the red and the blue curves in the operational area' plot can result. Note that these maximum and minimum permitted valve lift curves are not 'critical'. 

The numbers and the plots used in the analysis, even close to reality, are just indicative.

Oil consumption

The VVA system operates without vacuum in the intake manifold, so the common problem of oil leakage through the clearance between intake valves and intake valve guides, especially at partial loads, is gone. No vacuum, no suction of oil, no oil into the combustion chamber, less deposits on piston and valves, no need for addition of oil between oil changes. Coates (rotary valves) claims that the oil into the combustion chamber reduces the maximum compression because at the oil droplets and deposits is where the pre-ignition starts.

Variable preloading of the valve springs

Lately there were announced systems for adjusting the strength of the valve springs, in order to reduce the friction into the valve train, the wear of the parts involved and the need for regular adjustments. The general idea is to move the low 'immovable' support of the valve spring (valve tappet) lower when it is necessary less restoring force for the valve (for instance at low revs, or when the mild cam lobe is used in VTEC-like systems). Actually, what is changed is the pre-loading of the valve spring.

The VVA system achieves more, in a simpler way : as long as the engine needs low or very low valve lift, the linkage between valve and cam lobe moves with very low resistance. At the minimum used lift (idling), the camshafts can be rotated easily by the finger of a small child.

The valve springs in conventional engines have to be hard enough to restore the valves at the maximum revs the engine can operate safely. When the car is stopped at traffic lights with its engine at idling, the compression of the valve springs and consequently the resulting force, loads and wear, are the same with those with the engine revving at red line. But at idling it is necessary less than 100 times that force to restore the valves (the necessary restoring force falls with the square of the revs). And 100 times lower restoring force means significantly lower resistance for the timing belt, less wear, smoother operation etc, etc.

The VVA uses the total valve lift only at high revs with full load. Only when you operate the engine above 6.000 rpm with full load, only then the loads on the valve train becomes similar to those of the conventional. In all other cases, the VVA operates with lower or extremely lower loads, with less friction, with less wear, with longer expected time between repairs and adjustments.

Fatigue is the main enemy of the valve springs, and  in VVA the valve springs operate at full stroke only occasionally.

No doubt, the VVA at normal use has several times lower friction than conventional valve train systems.

Driving the Renault 19 prototype in Athens' traffic, valve lift of more than 2.5 mm is almost never used.

If the friction reduction were its only advantage, even then this advantage alone suffices to make the VVA a better choice than conventional valve train systems.

Valve clearance and VVA

Due to the valve clearance, the angular duration of the suction cycle, of the exhaust cycle and of the overlap are reduced at low valve lifts.

For instance, if at full valve lift (10 mm):

the exhaust valves open 70 degrees before BDC and close 40 degrees after TDC,

and the intake valves open 40 degrees before TDC and close 70 degrees after BDC,

then at 1 mm valve lift and due to the valve clearance the timing could change like:

the exhaust valves open 45 degrees before BDC and close 15 degrees after TDC,

and the intake valves open 15 degrees before TDC and close 45 degrees after BDC,

reducing even more the actual overlap, extending the expansion cycle and trapping more mixture at suction cycle.

It is only a matter of proper design of the cam lobes and then the valve clearance improves the operation of VVA, especially at low revs and partial loads (or to say city traffic).

VVA: an intelligent VVT system

As regards the action of the VVA as a Variable Valve Timing (VVT) system, the plot below is illustrative.

The VVA engine can operate at high revs and full load with higher actual overlap than most conventional racing engines. The same VVA engine can operate at medium to low revs with actual overlap much lower than normal conventional engines.

At partial loads the actual overlap becomes even lower, in case of the VVA.

The typical VVT sacrifices the operation around BDC, because the control of the overlap at TDC prevails.

The VVA acts not as a typical but as an intelligent VVT system, improving the operation on both, TDC and BDC.

What do you think?

mailto:vva@pattakon.com

Thank you

PATTAKON

Greece