The PatVVD is a CVVD (Continuously Variable Valve Duration) system: it varies continuously the valve opening duration.
The above animation shows the 24 "equivalent" camlobes when the "control" (explained in the following) makes a complete turn in steps of 15 degrees (15*24=360).
Here it is shown the working (the actual) cam lobe:
wherefrom the "equivalent" cam lobes result. It is from an old (non CVVD) Hyundai engine and has roller cam follower.
Here it is shown the above working and two "equivalent" cam lobes:
Explaining the patVVD:
The cyan part is the cam shaft whereon a cam lobe is pivotally mounted; the cyan cam shaft rotates at constant angular velocity.
The patVVD mechanism (not shown here) causes a fluctuation of the angular velocity of the cam lobe along a camshaft rotation.
The animation has a slide per 6 degrees of camshaft rotation (or per 12 degrees of crankshaft rotation).
At a first position of the control pin (not shown) the camlobe (shown red in this case) rotates slowly when its nose passes over the "more at: www.pattakon.com" label (at bottom), and quickly when its nose passes over the "patVVD : pattakon CVVD" label (at top).
At a second position of the control pin, the camlobe (shown blue in this case) rotates quickly when its nose passes over the "more at: www.pattakon.com" label (at bottom), and slowly when its nose passes over the "patVVD : pattakon CVVD" label (at top).
If the valve actuator were where the bottom label is, the valve duration with the control pin at its first position (cam lobe red) would be more than 70 camshaft degrees longer than with the control pin at its second position (cam lobe blue).
Here the PatVVD in the cylinder head of a 4-in-line plane-crankshaft 4-stroke engine:
In the following animation, with the sprocket immovable (i.e. with the engine stopped) the (eccentric) control pin (cyan) is displaced (by means of the cyan lever) at six different angles about the camshaft rotation axis; each angle of the control pin gives a different "valve lift profile", i.e. a different motion the valves will follow during engine operation:
An interesting application of the above mechanism would be on a V-twin Desmo Ducati Panigale: the engine remains as desmodromic as before, however its valve opening duration can widely increase (and decrease) on-the-fly.
As for the required "hardware" modification, it has to do only with the "connection" of the sprocket with the camshaf, say as shown above.
The resulting "Equivalent Cam Lobes" when the (cyan) control lever turns for only 60 degrees in steps of 15 degrees:
Some groups of available valve lift profiles, each for 60 degrees of control pin (or control lever) rotation:
Some of the available valve lift profiles;
at top: complete turn of the control pin,
at bottom: linear displacement of the control pin:
And some design details:
The PatVVD.exe windows "exe" program demonstrates the basics of the mechanism:
Introducing a variable "Waving" (along a camshaft rotation) phase difference between the camshaft and the crankshaft, the actual cam lobe behaves - for the engine - as a different cam lobe.
Like stretching the horizontal axis of a valve lift profile, compressing some regions and extending some other regions.
In comparison to the closest prior art (Rover VVC and Hyundai CVVD), the PatVVD:
- avoids the sliding keys, the sliding key slots in pins and the sliding friction between cooperating / heavily loaded parts,
- uses only constant length links pivotally or rotatably mounted,
- for the support of the mechanism on a cylinder head it does not require bearings other than a set of coaxial bearings as those supporting a conventional camshaft on a cylinder head,
- it does not require additional structure(s) to bear and displace the control parts,
- it requires only small modifications in order to fit in existing cylinder heads as add-on.
Valve lift, valve velocity, valve acceleration and valve-piston clearance.
Click here or here for two controllable windows "exe" programs wherefrom the slides of the following "gif" animation were taken:
The animated plot shows, among others, the valve lift, the valve velocity and the valve acceleration (red, orange and purple lines respectively for the case of a "tuned" Suzuki GSXR conventional valve train; blue, cyan and green/blue lines respectively for the case of the same valve train modified to PatVVD).
The beige curve (top half of the plot) is the piston position; its minimum vertical distance from the "valve lift" curves (red or blue) "defines" the valve-piston clearance.
Instructions for running the above two programs (i.e. which key does what) are on their window borders: for instance, pressing the "comma key" shorter durations are selected; for instance, pressing the SpaceBar key the valve lift profile is the one that the engine runs as not being modified to PatVVD.
In the second program as "basis" it was used a conventional single-mode valve-lift profile, while in the first program the conventional valve lift profile was slightly smoothed (corrected) just before the valve "landing").
From the animation it is clear that the PatVVD mechanism needs not VVTs (Variable Valve Timing mechanisms). It is also clear that the PatVVD either maintains or increases the valve - piston clearance around the TDC (Top Dead Center).
At short valve durations the increased peak acceleration seems as a limitation; however the shorter durations are only for the middle-low rev range where the valve acceleration is substantially smaller (the acceleration drops with revs square: at 10,000rpm the valve acceleration is half than at 14,000rpm); even in the case the engine is forced to high rev at short durations (by shifting into the wrong gear, for instance), the larger valve - piston clearance at the shorter durations (animated plot) protects the engine from piston valve collision.
At the longer valve-durations the maximum valve acceleration decreases; for instance, at the rev limit the PatVVD valve train (set at a "longer" valve duration) runs more reliably than the conventional valve train because the valve accelerations (and the resulting loads) are reduced.
In the animated plot the transition from the shortest to the longest valve-duration "takes" 145 degrees of rotation of the control pin (the eccentric cyan pin in the following animation):
Case study
Advancing or retarding, by a VVT, the "red curve" of the animated plot, some limitations arise as the following dyno reveals (the Harley-Davidson is using four VVTs one per camshaft, while the BMW R1250GS, instead of VVTs, is using a two-mode VVA (called ShiftCam) similar in functionality and limitations with Honda's old VTEC; while the old VTEC is controlling all valves (intake and exhaust), BMW's ShiftCam is controlling only the intake valves):
With same capacity, same number of cylinders and same number of valves, the Harley-Davidson has 11% smaller peak torque, while the BMW has 13% smaller peak power.
Despite its impressively flat torque curve, the torque of the Harley-Davidson is weak everywhere (as if the engine is of smaller capacity). Its architecture (constant valve duration combined with variable timing) proves not good enough in practice. The H-D engine is a compromise oriented to the peak power. Softer cams can improve the middle range torque in expense of the peak power; wilder camshafts can increase the peak power, reducing even more the middle range torque.
Despite its strong middle-range torque, the BMW obviously lacks power. And if, in order to increase the peak power, the "high-rpm" camlobes of BMW are replaced by wilder ones, the now shallow torque-hole (just before 6,000rpm) will turn into a deep torque-hole at the most usable revs. With the two-mode ShiftCam on the intake valves (and without VVTs) the BMW engine is a different compromise that focuses on the middle-range torque.
Replacing the VVTs of the Harley-Davidson Pan America 1250 Special by PatVVDs (there is plenty of space available), its torque curve (the red dot line) can shift above that of the BMW R1250GS (blue dot line) at all revs; also its peak power (right end of the red line) can shift upwards, and not slightly.
Another case study
The Honda CRF250R DOHC engine:
can rev reliably up to 14,000rpm. However, above 12,000rpm it runs out of air and needs "more time" to fill its cylinder, i.e. it needs what the PatVVD provides by keeping the valves open for longer.
Wilder camshafts of longer duration cannot help spoiling the middle-low rev torque, which is already compromised.
If one only PatVVD is to be used, it goes to the intake.
With a PatVVD the decompressor is optional because at its longest duration the PatVVD reduces substantially the actual compression ratio; without the decompressor, a lot of space is released "in front" of the exhaust sprocket for the exhaust-PatVVD.
The space "in front" of the intake sprocket is barely adequate for a "slim" PatVVD.
A CRF250R engine modified to PatVVD can provide at least as much power and as much torque as those provided by the original engine. This is so because at a specific angle of the control pin the valve lift profiles can be identical to those of the original engine.
With the control pin turning a few degrees to provide smaller overlap and shorter valve-duration, the torque can increase at the middle-low rev range.
And with the control pin turning for a few degrees the opposite direction to provide larger overlap and longer duration (i.e. larger valve-time-area), the power can increase at the high rev range.
And another case study
Yamaha 150cc, 4-valve, single cylinder, single camshaft engine:
In the above "animation" the "angular distance" between the exhaust camlobe and the intake camlobe is at its minimum at left, and at its maximum at right.
With the control pin (yellow) turning a few degrees to provide smaller overlap and shorter valve-duration, the torque can increase at the middle-low rev range.
And with the control pin turning for a few degrees the opposite direction to provide larger overlap and longer duration (i.e. larger valve-time-area), the power can increase at the high rev range.
The mechanism for the two settings of the previous animation, at "operation":
Click here to enlarge or click here for slow motion.
At "slow motion" the different motion of the parts at the two selected modes is more clear.
Not only the overlap is substantially larger at left, but also the intake valves close substantially later at left (one slide per 15 camshaft degrees or per 30 crankshaft degrees).
While the exhaust camlobe (red) rotates at constant angular velocity, the angular velocity of the intake camlobe (blue) varies substantially along a rotation.
Click here to enlarge or click here for slow motion.
The decompressor is optional because at its longest duration the PatVVD reduces substantially the actual compression ratio.
