Valve Control 101: Pneumatics

Part 3 of 3.

Even with the most careful design, there comes a time when the only way to reach higher rpm and, hence, make more power, is to accept the limits of springs and work around them. This takes into consideration waves reflecting continually in a valve spring and all other parts of the valve train (the cam itself is flexible, as are the bearings supporting it, along with any levers interposed between cam lobe and valve).

Why not just keep making the spring stronger? That has been a workable solution in drag racing, where so much can be changed between runs. But it works less well in any kind of distance racing, which requires extended parts durability. To get the spring force needed to reverse the valve’s motion at the top of its lift, the spring load when the valve is on its seat is often high enough to cause the exhaust valve stem to stretch.

Working around the limits means living with limited spring force, and that in turn means: 1) giving the spring longer to do its work (extending cam timing) and/or reducing the distance through which the valve is lifted; and 2) not lifting the valve as far.

When this became Renault Formula 1’s situation during the ’80s turbo era, there was an additional problem. Making valve timing longer to give enough time for springs to control valve motion allowed the turbo to blow a lot of fresh charge out the exhaust during valve overlap (the period around TDC at the end of the exhaust stroke when the exhaust valves have not yet closed but the intakes have begun to open). When some aspect of an engine nears its limits, other problems tend to pile up against it:

1) More spring pressure was needed “over the nose” but there could not be a similar increase in seat pressure. This implied a need for a more progressive spring (one whose rate, in pounds-per-inch of lift, increased with lift).

2) Engines could not deliver acceptable powerbands if their valve timing had to be extended for every increase in rpm. The later the intake valves were closed, the more mixture would be back-pumped at lower revs and the more light-switch-like and harder to use the torque curve would become.

3) As if that weren’t enough, in fuel-limited forms of racing, long valve overlap from long valve timing increased charge loss from intake to exhaust, wasting fuel.

4) Reducing valve lift in an effort to help metal springs control valve motion placed a “partial throttle” in the way of the intake process. Reaching higher revs only to fill cylinders less well was a step forward followed by a step back—no gain.

“But every motocrosser knows that air, used as a spring, is not like that; its “rate” stiffens with compression.

A metal spring with constant coil pitch and diameter has a constant rate. That means that if it takes 10 pounds to compress it by one millimeter, it will take 20 pounds to compress it two millimeters, and so on. But every motocrosser knows that air, used as a spring, is not like that; its “rate” stiffens with compression. Within limits, metal springs can be wound to have a rising rate—we have all seen fork springs wound with closer coils at one end than at the other. When the closer coils coil-bind, the spring’s rate stiffens because now fewer coils are carrying the load. But springs for high-speed engine use have very few coils, limiting the extent to which they can be made progressive.

So air became the key—a way to make a virtually mass-less spring that did not “ring” or surge and was as progressive as you needed to make it. We all use air springs every time we open the hood or cargo door of our cars; the spring medium in the black struts holding them open is compressed gas.

Pneumatic valve springs are no more exotic than those black gas struts. It is just as though you started with an overhead-cam engine, operating its valves through inverted bucket tappets, and you threw away the metal springs and instead put gas pressure under each inverted bucket.

Naturally, you would have to seal the inverted bucket to its bore and seal the valve stem where it entered its valve guide. Those seals would have to survive rapid motion. There would have to be a pressure make-up system to replace any gas lost in operation. And, as it turns out, there also has to be a kind of “sniftering valve” to release oil that the motions of the parts carry into the air chamber under the inverted bucket. Because hot oil and oxygen are a fire risk, the gas used to pressurize a pneumatic spring system is common inert nitrogen.

“A common belief about pneumatic springs is that they are useful only for reaching extreme rpm. In fact, their real usefulness lies in their ability to achieve high valve accelerations.

Engines for long-distance racing have carried a small compressor to make up pneumatic-valve leakage losses, but for GP use, a small compressed-gas bottle and pressure regulator are enough. The pressures involved are nothing special—of the order of 150 psi. Common-rail diesel injection systems routinely employ 250 times more pressure than do pneumatic valve-spring systems.

A common belief about pneumatic springs is that they are useful only for reaching extreme rpm. In fact, their real usefulness lies in their ability to achieve high valve accelerations. High rpm is indeed one situation requiring such high accelerations, but short valve timing and high valve lift are another. Thus, another attraction of pneumatic springs is their ability to broaden a powerband by combining short timing (which reduces low-rpm torque loss caused by back-pumping) with high lift (which enhances flow by getting the valves out of its way quickly).

In the final days of metal springs in MotoGP, teams were sacrificing powerband by using long valve timings, suffering throttling by cutting valve lift by 25 percent, and having to change valve springs every day because the high wire stress they were forced to use quickly fatigued the springs.

When I visited Del West Engineering some years ago, its engineers pointed out the frictional waste involved in giving 1.6-liter econo-car engines 7,600 rpm worth of metal valve springs, only to spend 98 percent of their time at 2,800 rpm or less. With a gas spring system, spring pressure could vary with rpm, generating much less friction (especially at idle and low revs, where high spring pressure threatens to break down the oil film between cam and tappet, generating extra friction from mixed lubrication). So far, though, no takers.

As a fraction of total engine friction, valve drive is quite small, but losses at freeway speeds have gained the attention of engineers seeking to please their stern taskmasters at the EPA; some auto engines now employ roller tappets to escape the rising cam/tappet friction at low speeds.

“The gains being sought are real: to be able to smoothly transition between Harley-Davidson-like bottom-end torque to Ducati Diavel midrange and then to no-holds-barred race-engine top-end.

We constantly hear that the holy grail of valve control—infinitely variable timing and lift—is about to be achieved. Regiments of futurist arm-wavers assure us that this will be achieved with solenoids or hydraulics, but so far, only low-to-moderate engine speeds have been achieved by such means. The gains being sought are real: to be able to smoothly transition between Harley-Davidson-like bottom-end torque to Ducati Diavel midrange (remember, only 11 degrees of overlap!) and then to no-holds-barred race-engine top-end.

Getting there, however, is not so easy. A Del West 38mm titanium valve weighs 36.5 grams (0.08 lb.), so to accelerate it at a race-level 3000 g would require a force of 240 lb. (remember that, in F1, peak piston acceleration has been as high as 10,000 g). That force requires a large actuator, which must also accurately control the deceleration of the valve back onto its seat at survivable speed; valves cannot survive "snapping shut." A profile capable of doing all this is ground into each cam lobe, but making an electrical or hydraulic actuator do the same for a million cycles (a little over two hours' operation in an F1 engine) is a very different challenge. The top F1 teams have had annual budgets as high as $300 million, so they are capable of tackling such research. Renault F1, the original inventors of the pneumatic spring, have reportedly made extensive experiments with flexible timing schemes but, so far, have not deployed such a thing.

Such technologies seek to give the piston internal-combustion engine the flat torque and controllability of an electric motor. Then why not make the leap to electric cars and forget about all the back-and-forthing of valves and pistons? We will, starting with the appearance of a compact, affordable, safe 300-mile-range battery that can be charged in a few minutes.

Slot: div-gpt-ad-leaderboard_sticky
Slot: div-gpt-ad-leaderboard_middle1
Slot: div-gpt-ad-leaderboard_middle2
Slot: div-gpt-ad-leaderboard_middle3
Slot: div-gpt-ad-leaderboard_bottom