Yamaha developed its novel five-valve technology for specific
gains. First on the list was a compact, nearly flat combustion
chamber of minimum surface area. In a two-valve design, adequate
valve area comes only by tilting the valves away from each other
and making the head somewhat hemispherical, but with five valves
the poppets can set into an almost flat
chamber. The Yamaha's head is only slightly domed, its piston slightly
concave. The resulting lens-shaped chamber concentrates the charge tightly
around the central spark plug, and this means most of the charge is quickly
inflamed shortly after the spark. The resulting short total combustion time
cuts energy loss through heat to the cooler metal of the piston and head,
and that saved energy is applied to the job of pushing the pistons down.
Being nearly flat, both piston and head offer minimum surface area, and
this further cuts combustion heat loss, again translated into power gains.
Detonation—engine knock—sets the upper limit on compression ratio, but the
five-valve's rapid combustion can consume the charge before detonation has
time to occur, permitting an unusually high compression ratio of 11.2:1
(Honda's VFR750 is good for 10.5:1, Suzuki's GSX-R750 10.6:1). This not only
gives the Yamaha 750 and 1000 more punch across the powerband, it also
increases fuel economy.
The paired exhausts and trebled intakes bring more advantages. To explain
one, we'll use a two-stroke analogy. Imagine two cylinder-wall ports, one
wide and short, another narrow and tall. Both 'have the same area when fully
open. Clearly, as the piston falls, the wide port will expose flow area
faster—because the narrow port is taller, it will take longer to open fully.
Yet when both are fully open, they have identical area.
Now for the four-stroke equivalent. Imagine two engines, one built with a
single, large intake valve, the other with three much smaller intakes of
identical total head area to that of the large one. Imagine that we equip
these engines with cams that accelerate the valves at identical rates. Which
design will expose flow area more quickly? In analogy with the two-stroke
case, our flow area will be the "width" of the port multiplied by the
distance it is opened. For the four-stroke, the width equates to the
perimeter of the intake valve or valves. The height is the valve lift—the
same for both engines because the valves are opening at the same rate.
Consider specific cases; the distance around a single 37mm intake valve is
pi times 37, or about 116mm. Three valves of the same total head area would
be 21.4mm diameter each, and the distance around all three will be pi times
three, times 21.4, or 202mm. Our three-intake-valve design exposes flow area
1.74 times faster (202 ± 116) than a single-valve design. Work the figures
for the twin intakes of a four-valve setup and you find the five-valve
Yamaha concept has a 22 percent advantage in rate of area exposure.
Here's a third benefit: Rapid opening gets the valve(s) out of the way of
the flow quickly, keeping the loss-producing restriction between valve and
seat to a minimum. Unfortunately, getting the valve open fast means serious
acceleration levels—up to 3000 times the force of gravity in some racing
engines. High valve opening and closing rates bring problems—like seat
hammering, cam and tappet scuffing, seat recession or loosening, or outright
valve breakage. The standard ways of limiting valve acceleration are to
reduce the lift and/or extend duration. Both have drawbacks: cutting the
lift cuts the flow, and extending the duration invites reverse flow from the
cylinder to the intake pipe; either cuts power.
The Manx Norton road racer had a radical 340-degree intake duration,
thought by many to be its key to high performance, but much of that
impressive timing existed because the designer couldn't get those big, heavy
valves up off their seats in anything less without breaking them. The Manx
could actually have made more power, and over a wider range, had it been
able to run less intake timing. These compromises were cut perilously close
in many cases; the great 1960s MV road racers would toss their valves if
overrevved by only 300 rpm!
Ideally, as the piston nears the bottom of its intake stroke at high
revs, the fuel/air charge is rushing towards the valve at something over 300
feet per second, and this velocity doesn't disappear just because the piston
stops at BDC and reverses direction. It's desirable to keep the intake(s)
open past BDC long enough to let this fortune in intake kinetic energy—a
kind of free supercharging—spend itself against the rising piston, forcing
in extra mixture to make extra power. At the instant that intake flow piles
to a stop against the rising pressure in the cylinder, the intake(s) should
snap shut, trapping these goodies. But as we have observed, valves and
springs can only take so much acceleration, and hence two-valve designs
suffer under a severe compromise between what is best for airflow and power
and what is possible mechanically. Again, the answer is smaller valves and
more of them.
Scale a part down in dimensions and it loses weight faster than it loses
strength—weight is proportional to roughly the cube of the linear dimension,
while the strength is related to a lesser power. This means small valves can
stand higher acceleration rates than can large ones. Consequently, not only
do many small valves expose perimeter area faster than a single one of equal
total area, but they can also be opened faster to redouble the effect.
What Yamaha gets in return for its extra parts is an unusually wide and
strong powerband. A two-valve or four-valve engine could be made to give as
much peak power, or as much low-end and mid-range, but not both. The Yamaha
makes its numbers with grace, not with extremes of materials or design.
Next comes the matter of 'valve springs. From your place on a tall stool
in an air-conditioned drafting room, logic tells you that two revolutions of
the crank equals one valve-spring fatigue cycle. From the hot dyno cell or
race track, the springs see things differently: at high crank speeds the
rapid acceleration imparted by the cam lobe approximates a hammer blow. This
can make the coils of valve springs "ring" or vibrate end-to-end. This
ringing vibration may have a characteristic frequency of hundreds of cycles
per second, so it can, if excited at high speed, add up fatigue cycles so
fast that springs break prematurely. This spring surge can also cause
irregular actions at the valve—float, bounce, etc.—that deteriorate other
parts as well.
Designers like "soft" rate springs—those with little difference between
their seat pressure and their open pressure. Why? Too much spring pressure
can overload the oil film between cam and tappet, leading to scuffing.
Unfortunately, such springs also tend to have low natural frequencies.
Standard texts on valve-gear design suggest the spring frequency should be
at least eleven times the camshaft speed, but it is difficult to provide for
a large single spring or spring pack sufficient to close a single large
intake valve in a high-rpm engine. Such high-revvers need high-rate springs
with very few coils, operating at extreme stress levels, manufactured with
special processing and many inspections. Expensive, and difficult to make.
On the other hand, three tiny valves eliminate most spring problems. Tiny
springs are now all you need to handle the job, and such small springs
provide high natural frequencies without high-tech manufacturing and
expense. The single springs Yamaha uses are dualrate—the coils wound with
two pitches, a fine and a coarse. With the valve closed, all the coils are
in action; as the spring compresses during valve lift, the fine-pitch
section coil-binds, leaving only the coarse coils in action. This in effect
gives the spring two natural frequencies instead of one: a lower frequency
when the valve is closed, a higher one when it is open. This "confuses"
spring surge by favoring first one and then the other frequency, and tending
to suppress others in between.
Using one large intake, the designer must save all the weight he possibly
can by making the valve's stem skinny and short, and thinning down the head.
Such compromise valves usually employ stems whose diameter is only 18
percent of the valve-head diameter. Such valves, while light, affect both
durability and performance. They cramp the intake port into a hunched-over
position, huddled close under the valve spring seat and making a sudden
90-degree turn to enter the cylinder. This forces designers to use a short,
unsupportive valve guide that soon wears out, leaks, and forces the valve to
leak. Second, the sudden 90-degree turn flings most of the airflow to the
outside of the bend, so it enters the cylinder through only half of the
valve's circumference. These losses show up on a torque curve, making
foothills out of what might have been mountains.
Yamaha's three small intake valves can afford stem diameters a full 25
percent of their head diameter, and their length is more than four times
their head diameter—like the best racing designs. This allows excellent,
long-lasting support from an adequate valve guide that doesn't intrude into
the port, and also provides room for a nearly straight downdraft intake of
excellent airflow qualities.
Yamaha chose to operate all these valves in racing fashion, using one cam
lobe and inverted-bucket-type tappet per valve. Why not cut manufacturing
costs and ease maintenance by incorporating some form of forked rocker arms,
with clearance adjustment by screws and lock-nuts? What was gained in
valve-acceleration tolerance by using small poppets could easily be thrown
away by introducing a flexible element into the system—a rocker arm loaded
in bending. A rocker arm is effectively a high-rate spring, inserted between
cam lobe and valve. When the lobe accelerates the tappet, the spring first
winds up, and only then begins to lift the valve. When the cam contour calls
for the valve to slow for peak lift and then reverse, the spring unwinds,
then continues to oscillate for the rest of the valve event. If the
rocker-arm "spring" is again unwinding as the valve approaches its seat, the
valve may hit the seat with not only the seating velocity built into the cam
contour but also with the extra velocity resulting from rocker-arm
unwinding. If the rpm is up, and the designed-in seating velocity is already
on the high side, the result will be seat hammering, recession, or
loosening. With the rocker arm oscillating like this, it too can pile up
fatigue cycles like a surging valve spring until it breaks as well.
To go with their high-rpm, rockerless valve gear, Yamaha chose the most
reliable method of valve clearance adjustment—selective-fit lash caps on the
valve stem ends. Unlike clearance discs (shims) set into recesses on the
tops of the bucket tappets, these cannot come adrift during valve float,
free to wreck the top end. Adjusting clearance with this bulletproof system
does require removing the cams, but Yamaha has used hardened cam lobes to
extend the service interval.
And what about gross flow? Do three valves flow more air than one or two?
Years ago, Harry Weslake, the famous English airflow pioneer, believed he
had proven one valve was best—it minimized wall-friction losses. True, but
that small gain ignored the huge gains that would soon come from the use of
multiple, long-stemmed valves and gently curved ports. It also ignored the
greatly increased safe rev limit and durability of multi-valve designs;
either of these advantages by itself is enough to make nonsense of any
putative extra flow through a single valve.
Is there any limit to the process of valve multiplication? Yamaha has
tried as many as seven valves—four intakes and three exhausts—in larger-bore
engines. Five valves seem to work best in motorcycle sizes; more tend not to
leave enough head material between seats and spark-plug holes. On the other
hand, the old process of forming all the valve seats and the spark-plug
threads as a single austentic iron insert set into the aluminum head might
offer a way around even that limitation.
The iron insert would have another advantage as well; small, big-bore
engines have a lot of combustion chamber surface area in relation to volume,
and that means rapid heat loss. Compared to aluminum, iron is an insulator
that has proven its ability to keep the heat where it belongs—in the
So five-valve engines are the ticket in—we knew that last year, and they
haven't changed much for 1987. Show us something new, you say? Right this
way. While you've seen aluminum chassis on the street before, you've never
seen one so close to the track as this one. Conventional motorcycle chassis
have almost always been made from tubes—bolted, brazed, or welded together.
If any tubing is made smaller and of heavier wall thickness, so the weight
per foot remains constant, the bending and torsional stiffness of the tubing
drops, reaching the lowest limit as the tube becomes a solid bar. Reverse
the process and the tube becomes stiffer roughly in proportion to the square
of the tube diameter until at the other extreme the likelihood of the now
very thin wall crumpling under load becomes greater than the possibility of
actual rupture or tearing of the material. Designers seek- ing a high
stiffness-to-weight ratio make their structures with the largest possible
diameter and the thinnest possible wall. A single-tube chassis represents
the conceptual ultimate in bending and torsional resistance; many
experimental frames have been built this way.
Ken Sprayson, a noted English frame specialist, built steel single-beam
chassis in the 1950s. In 1969 the Spanish OSSA firm fielded a welded-sheet
aluminum 250 road-racing chassis. Harry Hunt constructed one of riveted
aluminum sheet two years later. The erratic innovator Eric Offenstadt ran a
welded aluminum monocoque 750 at Daytona in 1972. These experiments
apparently showed only that aluminum could not long survive the vibration of
motorcycle service. We now know correct design procedures can produce
aluminum structures of any desired lifetime, even in a motorcycle chassis,
but there is a compromise between weight and life.
High-frequency engine vibration is deadly to thin aluminum, yet in 1979
Yamaha pioneered conventional multi-tube designs in a welded-aluminum
chassis with wall thicknesses of two to three millimeters. The light metal
allowed both the wall thickness and diameter to increase with no weight
penalty. Soon they were both stiffer and lighter than steel designs, and
durable enough to last more than one race.
How do you stiffen an existing twin-loop frame? The goals are clear: the
steering head shouldn't flex and should resist braking forces, the frame
must be stiff enough torsionally to prevent the wheels from straying from
their common plane, and untriangulated bays should be braced with diagonals.
The best way to do this is to deepen the top frame rails to better resist
bending and torsion, and to provide equally deep cross-members to make the
two work together. Make the load path direct between steering head and rear
fork pivot. Do this and watch the top frame rails grow and the lower loops
shrink into mere engine-hangers. The opening between the top rails remains
to provide clearance for engine upper structure, or for service access.
The 1982 racing season was a turning point for Yamaha. The company had
mastered the multi-tube aluminum chassis and began working towards something
else in very much the way described above. That something else was the OW61,
a motorcycle not in itself successful, but a necessary step towards the
future. If you looked at that chassis with 1982 hindsight it was just a
twin-loop design with its engine hanging from already-shrinking lower frame
members. The engine, too, was significant, its two cylinder pairs set close
to 90 degrees, a configuration that cancels major vibratory forces. The
engine was supported in rubber mounts since the chassis designer had finally
decided to make the chassis stiff enough to do its job unassisted and let
the engine provide only power. The concept of a load-bearing engine is
attractive, but such a system fatigues an aluminum frame's welds.
The following year the upper rails grew again, the lower members
shrinking correspondingly. Yamaha repeated the process each succeeding year:
the current Yamaha YZR500 road racer's chassis is a delta-shaped twin
boxbeam, made as deep as the steering head and as wide as it must be to
clear the engine. It extends almost straight from head to swing-arm pivot,
and is made largely from special weldable aluminum sheet about two
millimeters thick. Almost without exception, previous designs have carried
steering-head bearings in the ends of a piece of tubing—the steering head
proper—and have joined the rest of the structure not to the bearing area,
but to the head tube, relatively far from the bearings. This sacrifices
strength by cantilevering the bearings out in space above and below the
points at which chassis loads are fed in. Yamaha put top and bottom bearings
into pieces of plate which extend rearward into the box structure, directly
carrying head-bearing loads into the frame. A tube separates the bearing
pair, but it is no longer loaded in bending. At the rear, plates at the side
pick up the rear fork pivot pin and the footpeg carriers. Between the
steering head and swing-arm pivot, the twin beams are gracefully shaped,
cross-members blended into them in organic-looking fashion.
Although the original OW61 used extruded frame tubes, Yamaha's present
racing chassis are fabricated from machined shapes and from special-purpose
pressings in sheet aluminum. Why do they last when frames before them
cracked? Using the longest possible welds cuts down the load per inch of
bead. In the Deltabox, the welds have the same dimensions as the frame
itself, and because the frame is largely continuous pressings, there is a
bare minimum of welds in the first place. The resulting structure probably
has about five times the torsional and bending stiffness of previous
In the new FZR Yamahas, we now have a Deltabox design for the street,
significant because this signifies the Deltabox design has passed rigorous
vibration, drop, and longevity testing. As in the competitors' aluminum
chassis, Yamaha uses high-quality castings for the steering-head and
swing-arm structures, a cost-cutting move. Castings have poor fatigue
properties as compared with wrought materials (rolled or extruded mill
forms) because traditionally most castings are full of voids or impurities,
both of which invite crack growth under stress cycling. On the other hand,
castings lend themselves to high-volume production where machining from
solid stock does not.
There is an answer, though. Traditional die-casting fills the die by
gravity flow, and dissolved gases in the metal pass out of solution during
solidification to form voids. Vacuum casting fills the die by drawing the
metal up from below: as the liquid emerges into the mold, the low pressure
there causes evolution of the gases in much the same manner as uncapping a
bottle of soda. The result is a casting with greatly improved fatigue
properties. Another approach is Hot Isostatic Pressing (HIP), which
submerges castings in a hot, high-pressure (15,00030,000 psi) bath, in
effect forging the part from all directions, closing the voids, and
therefore reducing the population of crack nucleation sites. The
formed-aluminum sections of a Deltabox chassis, pressed in the same fashion
as auto body parts, already lend themselves well to quantity production with
no sacrifice in strength.
Welding aluminum is tricky. Any welding process is really a continuous
casting in which the molten weld puddle freezes behind the moving torch arc.
Just as the freezing of sea water yields fresh-water ice and a slush of
concentrated salt water, the freezing of the weld puddle has some tendency
to produce a weld area of purified aluminum with a zone of concentrated
impurities and alloying elements down its centerline. Aluminum expands far
faster than steel; after welding its cooling contractions may tear the bead
apart, particularly in that sensitive, impurity-rich centerline zone. Welds
made with "high restraint"—on parts jigged so firmly they cannot move easily
during cooling—are especially subject to cracking in this fashion. Taking
care in the design not only of the chassis but of how its welds are
sequenced can make the difference between cracking and not cracking.
Yamaha uses robots to weld their aluminum frames, and began using robot
welders in 1974. The trend in robotics in the U.S. is toward many-jointed
arms that mimic human function; in Japan robots now tend to be specialized
for particular jobs. Yamaha builds its own robots for about half the cost of
commercially available machines, and in a typical operation, several
machines work with one or more human "stagers." The stager picks up frame
elements and locates them accurately into a fixture on .a rotary table—such
fixturing is very difficult for robots, but easy for humans. The table
indexes 180 degrees, carrying the fixtured part into the operating envelope
of the robots. A glare curtain sweeps across, protecting the worker's eyes
from the ultraviolet light as the welding begins, often performed by more
than one robot simultaneously. The worker removes the just-completed part
and fixtures another. He may also complete welds in areas difficult for the
machines to reach. Why robots? Can you repeatedly place an arc source to
within 0.004 inch anywhere in space, eight hours a day, Monday-morning
hangover or no?
What can Yamaha—or any other manufacturer for that matter—do for an
encore? The latest FZRs still look a step behind in the aerodynamic fight:
maybe slicker bodywork for '88. Or how about a reliable, responsive
fuel-injection system? And aluminum isn't the only stuff to make frames
from: a carbon-fiber chassis? A ceramic cylinder block? When will center-hub
steering come of two-wheeled age?
Yamaha's five-valve engines and Deltabox chassis are but a momentary stay
against the unending rush of technology, but together they represent a canny
balance between performance and handling, between technology and rider
ability. The featherbed Nortons did the same, and we remember and revere
such machines even today. Will these new Yamahas be worthy of such
immortality? We'll all soon know
Source Cycle Magazine