feeling pressure

AmSnow.com is now SnoWest.com

The continuous belt drive transmission has been the key to practical snowmobiling. A snowmobile needs the continuous supply of power to the track to stay atop the snow and not get bogged down.

Back in the '50s, Carl Eliason's first production snowmobile had a Salsbury belt drive transmission, but, with relativity low power, belt problems were not a major concern. As soon as power increased and manufacturers started racing to gain market share, industrial-based clutch systems quickly reached their limit. Blown belts often sidelined racers, and the pressure to come up with transmissions that could take the power, and finish a race, became a priority. Efforts got divided into two areas.

Better clutches, better belts
Manufacturers worked on making better clutches, and belt suppliers began making stronger belts. Several clutch design changes improved belt life. Most important was the introduction of a torque-sensing helix in the secondary clutch that adjusted the side pressure of the belt proportionally with torque being transmitted.

Before the helix was introduced, side load was supplied by a straight spring. The problem was that the preload had to be enough to transfer torque in low gear when the load was the highest. As the clutch shifted out, the side pressure from the spring became larger than what was necessary to transfer the reduced torque at higher ratios, and belts blew.

The torque sensing helix reversed this. By feeding the torque load back as side load, and using a softer torsion spring, the side load on the belt now followed the torque transmitted through the secondary clutch. The result was that almost twice the power could be transmitted through the belt, since it was not overloaded in higher ratios.

Additional changes
Other changes, such as aluminum clutch sheaves, for better cooling, and torque transfer buttons, moved to the end of the spiders to reduce shift friction, also contributed to better belt life and improved reaction to up and down shifts.

On the belt side, suppliers, such as Dayco, devised improved cord material. Early belts had used glass fiber for load-bearing cords. Dayco introduced "Fiber B," its trade name for a Kevlar-based cord. This new cord, together with improvements to belt material, greatly improved belts' strength. The crisis seemed over.

Combine these advances with a voluntary agreement by manufacturers to drop anything bigger than 440 race engines, and transmission problems slipped to a lower rung on priority lists. Truth is, the chassis technology could not keep up with the power from the 800cc triple free-air engines on oval racetracks.

Aftermarket firms had no such agreement though, and with drag racing gaining popularity, 150-horse triples were soon the standard.

Manufacturers were slower in increasing the horsepower, but as IFS suspension improved handling, engine sizes started to grow even on stock sleds. It seemed that again we were headed down the road where clutches and belts may not keep up with the horsepower race.

Developing belts
In the mid-'80s, I contacted Jim Lewis, who headed development at Dayco. Jim also was concerned and wanted to test new belt designs on higher horsepower engines. But there were no such sled engines available yet, except for in drag racing, and duty cycles of less than 10 seconds a run hardly gave the feedback needed to make the belt work on normal uses.

I had contacted Dayco because I'd started racing a car in the D-Sports class in SCCA road racing. I was trying to make a CVT belt transmission work in the car, but was not doing well. Belts were blowing after only 5 laps, and a race lasted 20. The car weighed 1,100 lbs. with driver, the wheels were racing slicks with no wheel spin, and track temperature in summer could hit more than 100°F.

This was far removed from winter conditions, where sleds weighed half as much, tracks were spinning and the air temperature was below freezing. Yet, Jim took interest in our racecar because the extreme loads would quickly show improvements in belt design. The project lasted 10+ years, and developments were incremental, but at the end we were looking at a 50% increase in power transmission capacity.

Two major advances came from the racecar project. Dayco developed a new belt design for our car and we reduced friction in the drive system.

We measured temperature changes on the belt at every pit stop. When we got some top cog belts from Jim, we saw an immediate decrease of 20 degrees in belt surface temperature. To run ribs on the top of the belt to improve cooling, power transfer had to be concentrated in stronger cords in the center. This not only helped cool the belt, but also made it more flexible and efficient.

Teflon coating the helix helped us reduce friction in the drive systems. Lap times often deteriorated during the race, as dirt built up on the helix and the clutches got hotter. Teflon coating produced immediate improvement in both lap times and belt life.

Next we introduced a roller system to the secondary system, something that didn't exist for snowmobiles. Our first attempt was to convert a Polaris clutch. Buttons were machined away, and cam follower bearings installed instead. The steel needle bearings quickly chewed up the aluminum helix, so hardened steel helixes had to be made to last the race. With the prototype roller secondary, 1.50-inch wide top cog belt and a generous air duct to cool the clutches, we finally finished races.

The quicker shift action made the clutches run cooler because the side load was more constant and the belt slip was greatly reduced. Naturally we took advantage of our development, and we were the first to introduce top cog belts and secondary roller clutches to snowmobiles.

Top cog belts and secondary roller clutches are now standard on most sleds. Jim's foresight made it possible for Dayco to offer an improved belt design when the horsepower war heated up. Maybe Dayco's success, and the improvements made possible with roller clutch designs, made many take clutch performance for granted.

We had used triple engines in the early racecar program, but halfway through, we used our V-4. The smooth power delivery of the 90-degree firing V-4 layout contributed to improved belt performance.

Today much of the trend in snowmobile design is in the opposite direction. Instead of smoother power delivery, we see large long-stroke twin 1000cc engines.

The high peak torque load of a big twin is hard on the belt, hard on the crank and hard on the chassis. Balance shafts may reduce vibration, but the large power pulses are still there. This makes it hard to grip the belt consistently, and belt problems are rearing their ugly heads again.

Polaris had this problem recently and it lead to the axing of its 900 Fusion. Ski-Doo also had problems with the early Mach Z models until it introduced a secondary roller clutch. Ski-Doo has a primary clutch with a torsion dampener, which takes up some of the peak impact.

Large peak loads are not easy to handle, and this is now the engineering challenge for the factories with the introduction of large 2- and 4-stroke engines. Yamaha uses a rubber hub cushion in its drivetrain. Larger flywheels will even out torque pulses, and electric start ring gears are showing up where it is apparent that weight is added rather than reduced, to give an increased flywheel effect.

We found on our racer, that a well-designed system of torque links that allows the engine to move straight back under load, rather than twisting, not only makes the shifting better by allowing the belt to move out smoothly, but also takes peak loads off crank components.

The present snowmobile system, with a rubber bumper to stop the engine, only makes the situation worse for the crank, belt and chassis. The rubber bumper is cheaper than the good old torque links, but the links are sure to appear again to control engine motion in future sleds.

By controlling engine motion with links, softer engine mounts can be used, and vibration kept away from the chassis. Polaris wisely stepped back and is focusing on more manageable power sources.

Dayco's sled belt unit was bought by Carlisle and they now have a strong competitor in Yamaha's supplier, MLB, which has made a strong effort in developing belts for the 4-stroke market.

Chris Vogt's 1200cc triple (Tech Notes2, AmSnow Jan. '06) has proved reliable with 250 horses delivered through a primary 4-Star roller clutch, a secondary roller clutch with steel helix and needle roller bearings, and Kris swears by Yamaha's Mitsubashi-made belt.

History is repeating itself, and new engine designs are now pushing the limits of the CVT belt transmission. Will new solutions be found, or will we see a change back to engine designs that are easier on belts and clutches?

We'll probably see some changes in both.