US Urban Highways’ Condition Worsening; Repair Funding Waning
Bunge in Joint Venture with Acciona for Biodiesel Plants in Spain

Concept: A Cam-Based Infinitely Variable Transmission

The cam-based transmission has six primary components. Click to enlarge.

Engineers at the Virginia Polytechnic Institute and State University have developed a continuously variable transmission based on a cam and a planetary gearset that might increase fuel efficiency 3-5%.

Much work has gone into the field of continuously variable transmissions, which can be categorized as traction, belt, ratcheting and hydrostatic drives. The new cam-based system is a ratcheting drive type. Ratcheting drives take rotational input, convert it to an oscillating motion of varying amplitude, and then rectify this motion through a number of one-way clutches to a rotational output.

ne of the distinctive attributes of the new transmission developed by Derek Lahr and Dennis Hong compared to other ratcheting drive transmissions is that it not only produces uniform output given uniform input, but its output can be matched to nearly any periodic waveform.

As a result, the developers believe that the IVT would be ideally suited for application in wind turbines as well as in automotive and other industrial applications.

The cam. Click to enlarge.

The core of the transmission is the centrally located cam. A number of cam followers are rotably mounted to a carrier plate; each of the followers is fixed to a planet gear. An equal number of sun gears—each meshed to one of the planet gears.—connect to the output shaft.

By stacking an infinite series of profiles along the length of the cam to make a three dimensional camoid, an infinite number of transmission ratios can be selected. By varying the position of the cam followers in relation to the cam, the particular profile they follow can be changed. This affects the magnitude of the follower’s oscillations and therefore the transmission output.

The transmission. Note the quarter-circle pully (purple) at the mid-right.

Instead of gears with teeth, the transmission uses quarter-circle pullies and kevlar cord—this does away with the friction loss of one gear on another. (See diagram at right.) A flexible nylon ring to move the followers across the 3-D cam to change gears.

The developers have built an initial prototype as a test bed for design concepts such as the shifting mechanism, and to verify the underlying the kinematic equations.

A second prototype is in design, and will be tested for efficiency, durability, stress levels, and overall operability. The second design will also incorporate several improvements to increase the available gear range and torque capacity.

Among those improvements is a major change in cam implementation. The next prototype will have a barrel cam with the cam shape on the inside of a cylinder. The followers thus will be inside the cam, and held against the cam surface by centrifugal force. Sprag clutches are used so energy loss with the ratcheting is nearly zero.



allen Z would also help acceleration too.


Nice. Once CVT technology is adopted in cars, we would no longer have any gear shifts. Stepping on gas would produce a steady rate of acceleration, and no gear transitions.

My first experience with CVT is on a Prius taxi cab. I was impressed with how smooth the ride was, as compared with regular cars.

Bike Commuter Dude

You almost could have missed it: "...the IVT would be ideally suited for application in wind turbines as well as in automotive and other industrial applications."

Imagine how efficient you could make wind turbines (even in a retrofit) if they were always able to produce some electricity and would never have to be furled to avoid over-revving the turbine. Better still, you could optimize a turbine to operate at a specific RPM, and have the IVT figure out the rest.


The Prius doesn't have a typical CVT, it basically uses it's large electric motor as the main drive unit directly connected to the wheels and the engine output is mixed in and split between a generator and the output. They call it an eCVT (electronic CVT) but it really doesn't have much in common with any of the CVTs mentioned here, rather it's closer to an EV.

Bud Johns

Thanks Eric, I wish Toyota had never listed it as a CVT. You described it right. Anyway, CVTs in regular cars offer another advantage, the engine can be optimized to operate at certain rpms, and the redline needn't be above the peak power, so rotating parts can be lighter and valve springs weaker, leading to even greater efficency.


As promising as CVTs SEEM, in practice they have so far failed to yield any improvment in efficiency or acceleration.

The most fuel efficient vehicle is still a 5spd manual Honda Insight, the CVT version lags behind.

The same is true for all vehicles that have an option beteen CVT and Manual

Manual > CVT > Automatic

Rafael Seidl

Erick, Bud -

viewed from th outside as a black box, Toyota's HSD exhibits a continuum of torque conversion ratios and is therefore a traction CVT. The same is true of series electric hybrids. Seems to me you have got too caught up in the marketing literature which typically referes to belt-type CVTs, which are purely mechanical.

The articles categorization omits rhelogical CVTs. These feature disks mounted on a planetary carrier, whose own axis of rotation can be adjusted to a continuum of angles with the carrier's axis of rotation. The rims of these disks are in pre-stressed contact with a sun and an outer wheel; the cross-section of each represents a quarter-circle section of a single torus. Tilting the planetary wheels changes the radii of the contact points along the torus and hence, the speed ratio of input and output.

Torque is transmitted by rolling friction, which is rheologically enhanced by the use of thin films of special thixotropic (sticky) fluids. Using tight tolerances in mfg (= expensive), the required torque can be divided more or less equally among the available planetery wheels. This leads to high global torque ratings and compact dimensions, relative to belt type CVTs. As with all mechanical CVTs, noise is an issue.

For small loads, a simpler arrangement of two parallel shafts can be used, one fitted with a concave, the other with a convex conical element. Between them is a single, relatively large diameter ring that is held in place by guides in sliding contact. These may be rotated about an axis perpendicular to the plane defined by the input and output shafts. As before, the setup is pre-stressed and torque is transmitted by rheologically enhanced rolling friction forces.

A simpler variation on this theme uses identical, straight conical elements on the two parallel shafts. Between them is a relatively small disk whose axis of rotation lies in the same plane but tilted to one-half of the cone angle. The disk's contact points are pre-stressed as before. It freewheels on a bearing whose inner ring features a coarse thread on the inside. The axial position of this assembly is defined by the angular position of the threaded bolt on which it is mounted.

Roger Pham

Developing a new transmission is a very expensive investment for auto mfg's. The last dual-mode hybrid transmission cost billions of dollars and required collaboration of several large auto mfg's. This new tranny design must show how and why it is much better than Nissan's current belt-type of CVT which appears to be much simpler, or Toyota's new two-stage gear HSD, or GM et al's dual-mode hybrid. They claim 3-5% improvement in fuel economy, but improved over what type of transmission? Hydramatic? Mannual?


I heard tell, from a gear box specialist, that most new automatic Nissan's actually have a CVT system, but it is 'notched' so as to give the impression that it is a staged auto for the sole reason that the driving public arent ready to accept the 'weirdness' of shiftless driving...


The IVT is nearly impossible to drive a wind turbine generator due to the incredible torque demands. Imagine that a 2MW Wind Turbine has a rotor of approx. 40tons that turns with approx. 17r.p.m...and at the other side a big generator (2MW) which is trying to 'slow down' the rotor. Now, how can the sliding parts of the IVT find traction against such torque?
Wind Turbine Gearboxes literally burn down due to the high loads, the vibrations and the stochastic input due to wind turbulence...IVT has to deal with all these, and this is not easy...


Would this transmission be suitable for a bicycle as well or would there be power loss compared to a derailleur system?

cam ellis

Hi, I am curious about a pulsing effect that may occur on the output shaft due to the nature of the cam and sprag arrangement. I have included some links for 3 designs that we worked on in the past using a simalair concept that maybe of interest to you.
The method of drive from the cam arrangement to the output shaft could be gears as shown in 5 & 6 or arms as shown in version 4. The internal cam arrangement on 5 has a completely flat output curve meaning zero pulse due to the shape of the cam. It uses 60% of the cycle as drive and has therefore 6 driven sprag clutches.

cam ellis

Sorry, the first link should read:


Regarding the toroidal type CVT developed by Nissan and other companies, do you think it is very reliable?
I agree with Ash: it seems that for now, manual transmissions are more efficient than CVTs.
And I think that the dual-clutch DSG is more efficient than any CVT.
Do you think that the Multitronic -Audi CVT- (steel chain) is better than the V-belt CVTs ?

kent beuchert

The problem caused by highwinds for turbines don't have
anything to do with overrevving the generator.
While CVTs have been around for a very long time
(from around 1970, as I recall), they have had practically
zero impact , and at this point, where the most important
move is into electric, they seem almost quaint and irrelevant. Nice to see my Alma Mater involved, though.


I agree with many comments in this string. CVT's sound like a great idea for cars, but actually have nil impact on MPG. The issue is that A) gearboxes are quite efficient, and most CVT's actually have more friction. B) There is no big MPG advantage to running the engine at constant speed. A 5 or 6 speed gearbox has relatively small rpm ranges. There is nothing in the operation of an internal combustion engine which makes efficiency much higher at one exact rpm vs. a small range. However...for a highly tuned racing engine, where the power band is a small rpm range - there would be a horsepower / performance advantage... but that isn't what the "everyman's" car needs.

One last observation - since this CVT design relies on oscillating cams and sprague cluthes - it seems that it will be quite limited in max rpm. Have they made any claims regarding this ?


If you want to learn more about IVT then go here.


The reason why Nissan's toroidal CVT won't come to the US market is the fact the special fluid need for that CVT to work doesn't work correctly at lower temperatures. However, the NuVinci Continuous Variable Planetary (CVP) overcomes this issue, and a number of automotive companies and automotive transmission makers express very serious interest in the design, especially since this transmission can be scaled up for even high-torque turbodiesel engines.


I would just like to comment on 2 things, the indexer clutch and the cam, if this is gonna be used in cars.

I kinda wonder if the engaging and disengaging of the indexer clutch will replicate frictional losses similar to that of changing gears in a manual gearchange vehicle. Imagine if the cam has 2 peaks per revolution, the resultant propulsion should simulate a driver in a manual gearchange car stepping the clutch 2 times per revolution of the engine flywheel, i can imagine the vehicle vibrating violently. Why not use something like a bicycle rear sprocket which only transmits torque in only 1 direction of rotation, or use a turbine torque converter (something like those turbosomething devices made by Voith Turbo AG). i believe the energy losses from these 2 devices will be less than using a clutch, and the ride would be a whole lot smoother.

Also, there is constant acceleration and deceleration of the sun gear in which energy is lost in the process. and to reduce this losses, i can imagine a complicated flywheel balancing system that is going to be required, because we can't possibly let the engine or worse the wheels do the jerking motions. In a nutshell, if the engine has a heavy flywheel, the wheels jerk and vice versa...

Thus I don't think this would be suitable for automotive applications, seems more suited to some other applications where vibration would not be a problem. The whole mechanism seems too complicated, but I sincerely wish the team all the best in their future R&D as it definitely an innovative design.

Therefore, I'll still swear by the toroidal CVT as still the best compromise available today. (I'm still afraid the belts will break)

The comments to this entry are closed.