snowmobile clutch basics

Today's primary clutches are easily divided into two categories; those that use flyweights with a curvature working against a roller, or those with an arm with a roller attached to it working against a curved stationary cam.

The first category includes the present Arctic Cat, Polaris, Yamaha and Comet primary clutches. The second category includes the present Ski-Doo TRA and the older Arctic Hex clutch and the short-lived John Deere TR-800 unit.

In this story we'll look at the first category.

Primary history
The industry took awhile to settle on these two configurations. Early primaries often had simple weight blocks or even garter springs sliding against outside covers. This extremely high friction condition worked adequately when snowmobiles used small industrial engines with wide power bands. As engines improved and power increased the whole belt transmission came under sharp scrutiny.

Kidney weight flyweights were the next development, and had the advantage of being tunable by adding bolts, washers, and nuts, but a tunable curvature to calibrate the belt force was limited with this design. The moveable sheave on most of the early clutches also ran on a spline on the shaft, and this resulted in high friction as the torque load was high at the small radius.

When Polaris and Comet came out with newer designs in the late 1960s, they had two big advantages. First, the torque transfer point was moved away from the spline on the shaft and out to plastic buttons mounted on the spider. This cut potential friction by as much as 80% and improved response to changes in load conditions, resulting in quicker response to up and downshifts.

The new flyweight system consisted of curved weights acting against rollers mounted on the spider. By controlling the mass and curvature, the shift force could be matched to the engine power curve. This system with a number of updates and refinements is now used by Polaris, Arctic, Yamaha and Comet. They all use a curved flyweight pushing against a roller and torque transfer points on the outer perimeters.

The force balance working through the moveable sheave and shifting the belt through the ratios is complicated, and the present flyweight systems have evolved to meet these very special conditions. With the introduction of torque sensing secondary clutches using torsion springs and helix-cams to control the belt pressure, the force required to shift out the transmission actually declines as the belt moves from a 3:1 low ratio and out to a 25% overdrive position.
When the belt starts out in a 3:1 low ratio, the engine's torque is tripled at the secondary shaft. This requires high side pressure on the belt to transfer the power without slippage. On the other hand, in a 25% overdrive condition, the torque at the secondary shaft is 25% less than the engine torque. To match these side force requirements, the primary sheave needs to push against the belt many times harder in a low ratio than when the transmission is in overdrive.

This is accomplished in two ways.
First the flyweight pushes against a primary spring. As the transmission moves into the higher ratio, the spring force gets stronger depending on the spring rate, and this force is subtracted from the flyweight force. The centrifugal force acting on the flyweights is transferred into horizontal shift force through the interaction of the flyweight curvature with the stationary roller mounted on the spider. The contact angle in the horizontal plane determines the amount of force acting on the moveable sheave through the flyweight pivot pin mounting in the sheave itself.

The present flyweight configuration is well suited to this purpose. The flyweights start out hanging down in a vertical 6 o'clock position and this gives it a large horizontal force component with respect to the roller. As the flyweight swings out when the transmission shifts into higher ratios, it moves toward an 8 o'clock position. Due to its decreasing angle to the horizontal plane, the shift force on the belt is reduced, just as needed.

Clutching forces examined
By now this whole force component picture confuses most people, because we are not used to thinking in terms of diminishing forces in order to shift out into a higher ratio. Intuition would try to convince you that an increasing force would be needed to shift the belt into a higher ratio.

This was the case when we only had a pressure spring to control the belt pressure in the secondary, but with the introduction of torque sensing helixes the whole picture changed. Since the side force on a belt in low ratio is three times higher due to torque being multiplied threefold, the flyweight system has to produce a higher force to overcome the side pressure on the belt from the torque feedback through the helix and move it out into a higher ratio. As the belt moves out into higher ratios and less torque multiplication and feedback occurs, less force is required from the flyweights to shift into higher ratios. If the force from the flyweights had increased, the rpm of the engine would drop as less centrifugal force from the flyweights was needed to shift the belt out, and the transmission would not hold the engine speed at the power peak.

At this stage you are probably even more confused, but take heart, it took me several years and a lot of testing to become familiar with this force balance, and the only thing that will make you more comfortable with it is a lot of experimenting.

The basic force produced by the flyweight depends on three things. First is the weight or mass of the flyweight itself. Second is the location of the flyweight's center of gravity and its radius from the center of the shaft. Third is the engine speed at which the flyweight rotates.

Centrifugal force increases proportionately with an increase in mass, and proportionately with an increase in radius, but increases to the square of the increase in rpm. This means that if you double the engine speed, the centrifugal force goes up by four times. The total amount of centrifugal force transferred into the moveable sheave is then dependent on the angle between the flyweight and the roller, which is determined by the curvature.

But wait, we are only beginning here.
Once the force from the flyweight is fed into the moveable sheave, it has to overcome the pretension of the pressure spring before the sheave even moves to contact the belt, this is commonly known as engagement rpm. Once the belt has been grabbed by the moveable sheave, the belt will stay in low ratio until the engine speed increases enough to produce the additional force required to overcome the side force on the belt in the secondary. Only then will the belt move out and start shifting into higher ratios. Hopefully this will occur at peak power rpm, and if all the forces balance out correctly as the transmission shifts out, the clutch will keep the engine at the peak power rpm. This is known as a "straight" shift, and is what clutch tuners strive for to obtain maximum performance.

3 important changes
In practical terms there are only a couple of things you can change to improve the tuning.
By going to lighter flyweights you can increase rpm. By going to heavier weights you can decrease the rpm.

By changing pressure springs you can affect the shift curve. A spring with a higher spring rate will increase rpm toward the higher ratios, while a lower spring rate will drop the rpm as the transmission shifts out. As a result you can tune a clutch to "shift straight" by changing spring rates. Obtaining a straight shift is the No. 1 priority, because it ensures that maximum power is always transferred through the transmission to the track.

Priority No. 2 is the engagement speed. This can be changed in several ways. With a higher pretention of the pressure spring, the engagement speed will increase. Engagement speed also can be changed by grinding notches or flats in the flyweights, effectively reducing the angle with the roller and requiring more rpm to overcome the pressure spring.

Engagement speed also can be increased by "tucking" the flyweights under the pivot point, but too much of this makes the system very unstable as the center of gravity gets closer to the pivot bolt. I advise doing this only with great caution.

I do not recommend grinding the flyweight's curvature as changes as small as .020 inch off the curve can raise havoc with your ability to obtain a straight shift.

In reality, it comes down to playing with pressure springs, changing higher or lower weights and grinding flats or notches.
Flyweights usually are available in 2 gram increments, but they are steadily getting more expensive, so a good supply can set you back a large amount. The aftermarket has jumped on this and there are a lot of flyweights available where you can increase mass by adding bolts, nuts, washers, rivets or set screws.

Yamaha flyweights come with several holes where rivets from 1 to 5 grams can be added. There is now a large selection of flyweights from Polaris, Arctic Cat and Comet too, and they are usually interchangeable, so chances are you will find one with the correct weight, profile and engagement notch to match your requirements.

Yamaha flyweights do not interchange with the other brands, but Yamaha has a good selection of weights. Yamaha also has an extra tuning variable by offering three different sizes of rollers. The larger rollers usually give a more aggressive shift force, lowering the shift rpm.
Obtaining the best performing clutch calibration for your particular requirement, being it trail riding or racing, also means you need some good test procedures and instrumentation.

In the next issue we will take a good look at the testing procedures and the Ski-Doo TRA clutch components.
Olav Aaen is a long-time contributor to AmSnow.
As a mechanical engineer and president of Aaen Performance, Olav has been heavily involved with snowmobile performance since 1968. Aaen Performance is best known for pioneering performance pipes and introducing the roller clutch to the snowmobile market. Accompanying diagrams from Olav Aaen's "Clutch Tuning Handbook," available at www.aaenperformance.com