The Next Suspension Revolution

by Kevin Cameron

We think we live in modern times and that the technologies we use are sophisticated. The truth is that the present is just a transition between what was, and what it will soon become. In this view, our technologies are just a snapshot of development at a given instant. In a year, or in five years, we'll look back and wonder how we managed with such crude ideas. It's very easy to accept the way things are, and much harder to see what they might become.

The dark ages of snowmobile suspension were the era of chrome trailer shocks and short travel. No one worried about compression and rebound damping because (a) we didn't know enough to need or even imagine adjustable damping, (b) the shocks were welded together and non-rebuildable and, (c) at least one of the shocks on a given sled was either rusted solid or had pumped out all its damping oil anyway. To work with short travel, the springs and damping rates had to be stiff, so the ride was harsh and choppy. Why didn't we object? We didn't care because it was so great just to be out hammering over the snow instead of stuck at home watching winter re-runs on TV.

Horsepower growth magnified the problems of stiff, short-travel suspension. During the 1980s people were forced to do something about it.

The first something was longer travel. Long travel hit auto racing in 1953, with the introduction of the Mercedes W196 Grand Prix car. Its nine inches of wheel travel were a revelation.

Suspension energy-absorption ability, other things being equal, is proportional to the square of the travel. If you double the travel, with the original spring and damping rates, you get four times the energy absorption ability. This allowed you to hit rough stuff much harder and not be flung into the trees. Or, you can double the travel, cut the spring and damping rates in half, and have a much easier riding vehicle with still twice as much energy absorption ability as before. This is win-win change.

Softer spring and damping rates meant that suspensions tracked surface irregularities better, and transmitted less upset to the chassis. Long travel transformed off-road motorcycling in the mid 1970s, and brought with it much improved suspension dampers. By the time long travel hit the snow, sleds had needed it for a long time. Long travel was a great equalizer because it allowed a long travel sled with even a moderate motor to keep up with a big-inch modified whose short travel track couldn't hook up and deliver all that power. With long travel you no longer needed a heavy triple or four to go fast - a hooked up twin was a stronger play than a high horsepower monster in a stiff riding, old-tech chassis.

We've been enjoying this transformation for a few years now and, human nature being what it is, dissatisfaction has set in again. Every medicine, no matter how much better it makes us feel, has side effects. Long travel was great because it improved the ride and allowed sleds to run through rough terrain faster without rider fatigue or control loss. But long travel requires tall ride heights, and that causes strong weight transfer during turning or accelerating. Riders loved the ride and the control, but they hated ski lifting and the loss of acceleration when headlights pointed at the sky. To turn tightly or accelerate really hard, the old-style low ride height and hard springing was best.
The faster snowmobilers went, the harder they hit terrain. Very sporting riders adopted a stand-up motocross style, and jumping became routine. Trouble was, the compression damping setting that could absorb the impact of jump landings was so stiff that over little ripples or stutter bumps the ride became almost rigid. It was like driving your car down a washboard road at that exactly wrong speed that has all four wheels off the ground most of the time. Control disappears and you have to slow down or end up in the ditch.

Every vehicle has special requirements. Springs and shocks that give a good ride in a car are too soft when you have to swerve, because they allow the body to roll. Suspension soft enough to follow rough pavement is too soft to handle the pull-out at the bottom of a hill without bottoming. To deal with such things, engineers come up with add-ons. Cars, for example, have anti-roll bars - torsion springs that resist body roll. One make has a system that senses steering wheel shaft speed. If the driver swerves suddenly, shock damping is increased to prevent delay from body roll.

Now let's ask a more basic question; why do we need suspension damping at all? Damping is just controlled friction. Doesn't adding friction make the suspension less compliant, rather than more so? The simple answer is that adding friction to suspension movement slows down and stops unwanted suspension movements. But wait -when people, driven by intellectual curiosity or a few drinks, decide to try riding their sleds with no damping at all, they find that the sled does some things better. Most notably, it handles small bumps and ripples better than it does with damping. Why is this? Why do we even need damping?

The answer is that we need damping mainly to control motions of the suspension at specific frequencies. At other frequencies, damping just increases suspension resistance to motion and causes problems like poor performance over stutter bumps.
The specific frequencies that need control by damping are determined by the masses in motion and the spring forces that oppose that motion - so-called "simple harmonic oscillators". The classic example is a weight hanging from a spring. If we give it a poke, the weight bounces up and down at a frequency determined by the mass and spring stiffness. In the case of vehicles, up-and-down vertical bouncing motion on the springs is called the “heave mode”. You can see this one in action when an old car with worn-out shocks hits a dip in the road and continues to bounce up and down several times afterward. This needs damping because each extra cycle of the motion causes wide variation of load on the tires. In a corner, the result would be skidding and loss of control. The goal of suspension is even pressure of the wheels (or skis and track) on the terrain surface.

Another is the pitch frequency, in which the vehicle rocks forward and back. This is often excited by passing over whoops. Each ski and the track have what are called corner frequencies, which are determined by the mass of what moves and the rate of the spring controlling it. You see the corner frequencies excited when your car goes out of control on a washboard dirt road - the wheels bouncing up and down so extravagantly at their corner frequencies that they spend only a small part of each cycle actually on the ground. Result - no traction, and loss of control.

More complicated frequencies also exist, depending upon the flexure of parts of the vehicle we normally assume to be rigid. In the case of motorcycles, one of the most bothersome is rapid forward-and-back bending of the front fork, typically at 20-22 cycles per second. Rest assured that sleds have analogous flexure frequencies that create problems, thus the pursuit of making sub-frames and chassis components as rigid as feasable.

The Polaris PPS shock design uses fluid bypass in the shock's mid- stroke to relieve damping force for a smooth ride through the "stutter" and "chatter" bumps.

When one of these modes of oscillation is excited to large amplitude - as corner frequencies are in ski chatter, or heave and pitch are by whoops - controllability is endangered. With sufficient build-up of oscillation, the skis and/or track will spend part of each cycle off the snow and control will be lost. Ideally, suspension dampers would suppress just these particular frequencies and otherwise would just let the suspension move freely up and down over terrain irregularities.

Because we don't at present know how to damp these frequencies selectively, we must damp all frequencies. Damping increases the suspension's resistance to motion, so adding this damping makes the suspension less compliant (less able to follow terrain accurately). This is the price of preventing the build-up of unwanted oscillation and motion at the specific problem frequencies. This is the answer to why sleds, cars, and bikes do some things better when their suspensions have no damping.

Our entire supposedly sophisticated technology of damping by means of gas pressurized shocks with carefully chosen washerstacks inside them is in fact a crude, messy compromise, controlling the target frequencies by the shotgun technique of resisting all motion frequencies.
Computers do everything else for us. Why not suspension? Isn't it time we said goodbye to the messy compromise inherent in passive, no-brain systems, and found ways to break out of these compromises?

Some small steps in that direction are being taken now. A beginning was Polaris's Position Sensitive Damping, which uses multiple rings of bypass holes in the damper cylinder to relieve damping force in mid-travel, then stiffen it as suspension nears full compression. The effect is to improve and soften response to rapid motion in mid-travel (while running on the flat), yet preserve the ability to absorb jump landings by means of extra stiffness in the last third of travel. This has an historical precedent in the hydraulic recoil buffers of artillery, which use tapered damping rods to achieve something similar.

If this Polaris system seems simple, it's because it is. Any such add-on must meet two criteria; (1) it gives the customer an improvement he can feel and (2) is cheap enough that most or all customers see it as good value for money. Simple and cheap are good.
Now comes Arctic Cat's Smart Ride System. It uses a computer-controlled bypass valve on the damper to control its firmness, based upon a sensor that monitors the position and velocity of the damper rod. Thus the computer can recognize mid-travel stutter bumps and reduce damping to cut harshness. It can also recognize the rapid compression of a jump landing, and apply full compression damping to absorb it without bottoming. This system may also increase damping at the above mentioned "target frequencies" to suppress heave, pitch, and chatter oscillations, leaving the damping on a light, compliant setting at other times.

The control scheme for this system has been worked out by accumulating suspension data in the field and optimizing the system's response to it.

Although this system requires a computer, it is basically simple because the only extra part in the dampers is the bypass valve. This is a sensible, cost limiting approach to compromise breaking. Think of the intelligent things you can do with this. For example, you can tell the system that if all four suspension units go to full extension, indicating that the sled is off the ground in mid-jump, then compression damping should go to maximum for the next single compression stroke. You can think of many more "if-then" kinds of programming to improve suspension performance in specific situations. This is a significant improvement over passive dampers with fixed damping curves.

What about Yamaha, who have yet to release new suspension technology of this kind? It is common for major companies to keep some technologies they have developed "in the bank" until an appropriate application is found for them. One such that Yamaha are holding was developed by their Swedish suspension partners Ohlins and is called C.E.S. This stands for Computerized Electronic Suspension, and describes a damper than contains oil and an electromagnetically controlled valve. There are no washer stacks, no check valves, no blow-offs. A sensor tells a control computer the position and velocity of the damper rod. The damping curve, relating damping force to damper velocity and direction, exists only in software, implemented by the rapidly changing resistance of the damping valve in the shock. It can therefore be instantly altered by uploading a new damping curve to it from a laptop. This is especially useful in a racing situation because it eliminates removing the shocks from the vehicle, and it eliminates disassembly-reassembly to change damping curves.

When a wheel, ski, or track hits a bump, the bump accelerates it upward to some velocity. What the vehicle feels is a sudden increase in spring pressure, plus the compression damping force appropriate to the motion. The vehicle would be less upset if it felt spring force only, with the damping applied only after upward bump acceleration had ceased. This is possible with a fast-acting system such as C.E.S. Once you start thinking in terms of systems that can respond in special ways to special circumstances, the suspension horizon becomes suddenly much larger.

C.E.S. resembles Arctic Cat's SmartRide system in its versatility and ability to identify and handle specific situations, but has a wider range because it can vary damping, not just between a low and a high setting, but over the whole range from zero resistance to maximum. C.E.S. was developed in the late 1980s and then tested in motorcycle GP racing in the early 1990s. It is unknown when, if ever, C.E.S. will become part of a production vehicle.

Computer controlled systems like these can adapt to specific conditions. For example, if the suspension bottoms twice in quick succession, the system can apply more compression damping to prevent further bottoming.

Ohlins C.E.S. and the Arctic Cat Smart System are properly called "semi-active" suspensions because they still use bump forces to move the suspension. Beyond this is true active suspension which has the most to offer, but is also burdened with side effects.

True active suspension has no springs and no dampers. Instead, each suspension corner is positioned by a hydraulic ram. The fluid is supplied from a high pressure accumulator, kept charged by a pump on the engine. At each corner is an accelerometer. When one of these detects upward or downward motion of the vehicle, the computer valves hydraulic fluid to that corner's ram to zero that motion, or to reduce it to a practical minimum. Thus, engine power moves the suspension in response to what the vehicle is doing. Naturally, this requires quite a lot of power - as much as 15% of the engine's maximum. Yet the range of effects that can be created is unlimited - for example the vehicle can lean inward as it rounds a turn, rather than outward, or it can squat down during hard acceleration or turning to prevent ski lifting. The hydraulic struts that actuate the suspension are controlled by proportional valves originally developed for aircraft control, and capable of operation at up to 40 cycles per second. This technology really works but is expensive and complex. It was banned from Formula One for those reasons, but is used on some high-end production cars.

Cost will keep true active suspension out of our reach, but sleds do need one thing that active does superbly - variable ride height. Active was developed in Formula One, for the specific purpose of maintaining ride height to make aerodynamic ground effect work optimally. Sleds need variable ride height because the same owner may want at different times both long travel for rough trails - with the high ride height that goes with it - and a low ride height with stiffer suspension for quick turning and hard acceleration. I believe active suspension is not the only way to achieve this, and expect to see something of this kind both on the snow and on two wheels soon.

Variable ride height will need not only a way to lower the sled on its suspension, but also a variable linkage to make that suspension stiffer and reduce its travel at the same time. Otherwise the original long travel will allow the sled to hit the ground with the ride height in the low position.

These problems are never ending. Solving one creates or reveals others we never considered before. This is good because they give all snowmobilers something important to think about during the off season. While the children paddle happily in the shallows and you digest cold fried chicken and potato salad under a beach umbrella, you can be secretly considering how to implement the next suspension revolution.