What's it worth?

air inlet
Source: Sport Rider, August 1995

Why are Kawasaki's so damn fast? Anyone who's experienced the giggling hard-core buzz of a ZX's awesome acceleration from the pilot's seat comes away shaking his head in disbelief.
Time after time the big K's machines run faster at the strip than rivals which put out notionally identical horsepower. So where do the extra horses come from?
This is an account of the first attempt to scientifically measure the real effect of ram air.
The results may surprise you.

Who has it?

The ZX-11, ZX-9R, later versions of the ZX-6 and ZX-7 and the new ZX-6R Ninja all utilize the ram air system, and now Honda's latest CBR600F3 has followed along the same path - a sure sign that it works.
The difficulty is in measuring the effect, simulating the results of a 150-mph airflow, while the motorcycle is harnessed to a static dyno. But that is exactly what we set out to do with the help of Steve Burns, noted builder of very special turbocharged trick-framed motorcycles, sometime dragracer, endurance-racing-team boss and the owner of a Dynojet Model 100 dyno. As an innocent patient we had one meticulously prepared Kawasaki ZX-9R Ninja.

The theory

In essence, the theory behind forced-induction systems like Kawasaki's is quite simple and not that far removed from turbocharging, just at a less extreme level. A motorcycle traveling at high speed is pushing a slug of pressurized air ahead of it. If the air-inlet is placed in the correct position then air entering the airbox will be greater than atmospheric pressure. The resulting intake charge will be denser and cooler and contain more oxygen and fuel, thus causing a bigger bang and hence -Hallelujah!- more power.
There are limitations. The amount of mixture you can force through a motor is finite. Imagine strapping a ZX-9R on top of a jet aircraft and starting the motor; the plane will rapidly reach a velocity where the motor would be incapable of utilizing the volume of mixture being forced into it.
Next, at very high speeds -over say 150 mph, where theory says that ram air should be working most effectively- the nature of air drag means that large increases in horsepower are needed to produce relatively small increases in speed. Finally, compared to turbocharging systems, the increase in pressure are quite low. How low? Before we began testing, Burns predicted, "I don't think we can get more than one psi in there."

The ZX-9R system

Kawasaki's ZX-9R uses a relatively straightforward system compared to the 1995 Honda CBR600F3. Twin vents mounted beneath the headlight channel air via ducts running over the frame beams and int a sealed airbox. Look closely, and you can see two smaller nozzles behind the grilles which connect to the carburetor float bowls. Their function is to equalize the pressure between float bowls and airbox; without them the higher pressure of the incoming charge would upset the carburetion, potentially blowing fuel out of the bowls and tending to push fuel back down the jets, causing mixture leanness. Kawasaki uses much the same system in all its ram-air machines, though the ZX-7 and earlier ZX-11s have a single inlet only.



To reproduce the effects of high-speed running on a static dyno, Burns' intention was to use a fan capable of producing relatively small volumes of air, but at high pressure. The fan was connected via custom-made tubing and coupling to the intake vents of the big Kawasaki. The joint was carefully sealed with high-density foam.
So we could measure the pressure generated in the airbox as we pumped air up the ZX-9R's nostrils, a manometer, or pressure gauge, was plumbed into it. With the manometer we would be able to measure pressure up to 30 millibars above atmospheric pressure. A bar is roughly equivalent to atmosperic pressure; one millibar (mb) is just one thousandth -0.001- of a bar. Not a lot compared to tire pressures, but Steve's experience with varying boost levels on his 250-bhp turbo -which churns out approximately an extra five horsepower for every 70-millibar (one psi) increase in boost ore intake pressure- suggested that if it were possible to create one psi of pressure in the airbox, we could be looking at an increase of 5 to 6 bhp. Note that pressure in the context of this article is pressure above atmospheric pressure.

Testing theory

Burns' first thought was to set the air pressure at a certain level, say 15mb, and then measure the power at a steady throttle at 1000-rpm intervals. This was abandoned when we realized the results would be meaningless using CV carbs, which wouldn't necessarily be at full lift.
The second problem was that as the slides lift and the motor drags in air, the pressure in the box drops off. Observation suggested that if the airbox pressure was set to 10mb at idle, then at the redline, the manometer would show just 4mb. Obviously this bears little realation to real life, as the only way the airbox would be pressurized at idle would be if the bike were freewheeling down the highway.
Most importantly, the level of intake pressure on the road would be realative to the velocity of the motorcycle. If the airbox were pressurized to 20mb at 150 mph, it would be correspondingly less pressurized at 120 mph and still less at 70 mph. We had no way of reporducing this effect on the dyno, but if we could show that an air pressure of, say, 20mb gave a boost of 3 bhp at a certain point in the rev range and could then relate that to real road conditions, we'd have a fair idea of what the actual power output on the road would be.
In the longer term, Burns hopes to be able to use an interface between fan and dyno to take account of increasing air speed and thus simulate the effect of road speed on a static dyno.

First test

The initial step was to run the Kawasaki at atmosperic pressure -no boost- to get a baseline figure. The ZX-9R, like others tested on the same facility, gave 123 bhp at its power peak. The induction fan was then connected, and the bike was run with the intake pressure set to 10mb at idle. The process was then repeated with 20 and 30mb of pressure. In each case the intake pressure fell by approximately 6mb at peak revs when the slides were fully up and the engine was gulping down great gobs of mixture.
The results were gratifyingly clear. At peak power the ZX-9R was producing an extra 2.6 bhp for every extra 10mb of pressure fed into it by the fan. Peak power was up from 123 bhp to 131 bhp, an extra 8 bhp over atmosperic pressure. A secondary bonus was that the bike also hung on to its peak better, which would translate into a more forgiving motor on the road, which would be less sensitive to gearing and thus more likely to be able to take advantage of following winds or favorable gradients to give a higher maximum speed. Because of the testing procedure we'd been forced to use, the graphs also showed similar increase right through the rev range, but this was obviously deceptive. There was no way that the levels of boost measured at low speed on the dyno could be reproduced on the road. At this point we suspected that boost would be insignificant at speeds below 100 mph.

air inlet

Interim conclusions

So far so good. The first part of the experiment was a success. We'd shown that pressurizing the ZX-9R's airbox definitely produced power increases. We'd established that the system has the potential to work, but what we didn't yet know was how the pressure we'd managed to generate on the dyno -a maximum of 30mb at idle, or 24mb at peak revs- related to real road conditions.
Phase two was to attempt to establish what sort of pressure are actually generated in the Kawasaki's intake system at speed and relate them to the dyno results.

Phase two

Had we been NASA or a top G-team, the next step would have been easy. Strap a datalogger to the bike, rent a private test strip and go play for a couple of days. We weren't, so the manometer was cunningly strapped to the gas tank, green food coloring added to the fluid for added visibility, and a portable datalogger -yours truly- mounted to the bars.
Just riding from the dyno facility to the strip was illuminating. We'd reckoned on needing 90 mph before boost would register, but at an indicated 70 mph the manometer already showed 8mb of boost.
At the strip we were able to give the big Kwakker its head, with one eye on the slowly rising column of green fluid and the other on the rapidly rising speedo. At the end of each run we logged boost pressure agains indicated speed.
The results were even better than we'd hoped for. At lower speeds (under 120 mph) the gauge was easy to read and the result quite consistent: at 70 mph pressure was 8mb; at 80 mph, 10mb; at 100 mph, 12mb; at 110 mph, 14mb. From this point things really took off. At 120 mph (indicated) the airbox pressure was approximately 19mb, 130 mph about 23mb, at 140 mph, 26mb and at an indicated 150 mph, the gauge was beginning to pump out green liquid as it bubbled over the 30mb limit.
At a real speed of 167 mph, past experience shows that the ZX-9R speedo indicates 181 mph; there was obviously even more to come, perhaps as much as 30 mph worth of additional air pressure. Plotting the air pressure figures against speed for a rough representation of the way the air pressure increases suggests that the progression isn't linear.
This is as we'd expeted. Air drag doens't increase at a linear rate but relative to the square of the speed. At above 25 mph, air resistance builds proportion to the square of the air speed over the motorcycle: twice the speed, four times the resistance. The faster the bike goes, the greater should be the increase in pressure and thus intake pressure. When we plotted the rough course of the pressure increase on a graph and continued it upward we came up with a projected 44 mb (or more) of pressure at an indicated 180 mph, when the bike would actually be traveling at its real top speed of 167 mph.


So what does it mean?

The maximum pressure we were able to generate on the dyno was approximately 30mb, which gave a peak of 131 bhp from a ZX-9R as compared to the 123 bhp measured at rest. In other words, each 10mb increase in inlet pressure is worth approximately 2.6 bhp at peak on a derestricted 9R.
At an indicated 150 mph on the road, the inlet pressure had already neared the 30mb figure. We can therefor say with confidence that the ZX-9R is producing at least 131 bhp at the rear wheel in real world conditions - 8 bhp more than at rest on the dyno.
Flat out, however, the Ninja indicates another 30 mph on the speedo. If boost at this speed was, as seems likely, 40mb, then the gain over atmospheric pressure would be approximately 11.5 bhp, giving a peak figure of 134.5 bhp. If inlet pressure reached 45mb, which it might well do, then the increase would be as much as 12 bhp, of a peak of 135 bhp.
In other words, 123 bhp measured normally on a static Dynojet rolling road dynamometer could translate to as much as 135 bhp or more on the street. Ram air works.


An extra 12 bhp sounds like an extraordinary power gain for nothing except a bit of wind, but it's important to remember that at lower speeds the increases won't be as significant. Up to 120 mph when the boost hits 20mb, we're only talking about the odd bhp. From then on it gets progressively stronger. As the effect is speed relative, it's most pronounced at very high velocities; the faster you go, the stronger the boost. But, hey, how many of you ride at 150 mph on the street? Never mind, don't answer that.
Having said that, the effects of even small amounts of boost on throttle response haven't really been investigated and may help to explain some of the surging acceleration typical of big Kawasakis.(JDR: see my dynojet page)
It does, however, clarify the impressive figures that Kawasakis deliver at the strip and explain why a ZX-7 putting out the same power as a GSX-R on a static dyno will romp away under speed testing. It also begs the question of when someone is going to bring out a fully functional aftermarket pressurized-intake system.
Finally, it explains why Honda's jewellike CBR600 has finally gone the ram-air route in its quest to head off the ZX-6R.

By: Jon Doran (UK-based freelance journalist)
Email: jon@nozzer.demon.co.uk