95Racer Development

Well after some delays, the bottom end of the newest spec engine is ready to go together. The cylinder head is getting some additional work done including further porting work and dual valve springs fitted. It should be ready in another week or so.

In the mean time, I’ll be busy assemblying the bottom with the improved parts.

Improved Crank:
The balancer shaft gear has been removed and the crank lightened again and balanced to a new balance factor to move the vibrations to a more appropriate part of the rev range. As you can see from in the picture below, it’s a far cry from the stock part. It also weighs about 1.5 kg less!

This picture shows the shortened end of the crank with a slot cut into it for the external starter drive. Removal of the internal starter gears allowed us to shorten the crank and move the flywheel mass closer to the centre. As you can read further down, the flywheel is a fair bit lighter as well, as is the racing generator is now incorporates.

Reworked Stock Conrods:
The standard rods have so far proved strong enough. Plus we couldn’t find any lighter ones (although maybe stronger) on the aftermarket. Of course, there are always titanium rods. We concentrated on perfectly balancing them and strengthening the neck area with polishing and shot peening afterwards. It’s amazing how much hand work is just in rods alone.

We also found out that while originally Kawasaki had two types of rods available, J and K rods, they now have obsoleted the J rods. It turns out the the K rods are slightly thicker in the neck and small end area. It seems that the K rods were chosen maybe for slightly greater strength. They weigh roughly 8 grams more than the J rods. We now only use K rods.

After acquiring another spare engine with very few miles on it, a new winter project is taking shape. We knew from this year’s track and dyno testing that the valve train wasn’t performing at its best. The springs are likely not ideal probably resulting in some valve bounce and float during the closing cycle of the intake and exhaust valve. With the lightened conrods and pistons together with a re-balanced crank, raising the rev limit to 12,000 RPM is another goal. This will only exacerbate any valve train issues.

We felt the best way forward was to perform some CAE analysis to determine what an ideal valve train would look like, and then go back to the dyno for more testing.

The first round of analysis showed clearly the amount of valve bounce on the intake side. The following sequence (one inlet cam rotation) is at 10000 RPM with our race cams. All other valve train components are standard, including the valve springs.

You can clearly see the valve tappet (beige rectangle) separating from the cam and the valve head bouncing off the seat at the closing point of the sequence. To quantify this further, the following graph shows valve bounce versus RPM and you can see that there is almost 0.8 mm of bounce! All is fine until about 8500 RPM. It’s not so good thereafter.

Needless to say, this doesn’t help! It also explains some of the wear issues we’ve seen on the race engine. So, the next step is to use the simulation software to design an appropriate valve spring to eliminate the bounce and float. Given these results, we feel that the 95 RWHP goal is realistic for next season. Afterall, we’re only 3 HP short now!

What else do we have planned for this engine? We considered increasing the displacement to just under 700cc, but decided against it as it would somehow go against the grain. It would feel like cheating somehow.

We also considered using aftermarket rods. The only ones available off-the-shelf are from Falicon. However, these are considerably heavier than stock. Therefore, we’ll carry on with the standard rod albeit lightened, polished, and shot peened. So far, we have had no issues with the standard, so modified rods.

We’ve moved to an electric water pump. This will allow us to remove the balancer shaft completely. Up to now, we had run a plain shaft with the balancer weights machined off. This plain shaft’s only function, thus, was to drive the standard mechanical water pump. After weighing all the components, the electric waterpump weighs roughly the same as the mechanical pump it replaces. This creates a further weight saving from removal of the balancer shaft in excess of 1 kg.

A few people asked for more details on the auxilliary starter we built after having a lot of difficulty using rear wheel roller type starters. For one, the slipper clutch does exactly what it’s supposed to do. It slips! A lot! Because it’s a wet clutch, it’s not easily modified to accept a lock-out pin like Ducati used to use on their WSBK bikes. Even with the standard clutch (non slipper) installed, it was still very hard to start with the rear wheel either slipping on the rollers or just not getting the cranking speed high enough to where the engine would ‘catch’.

So, instead, we decided to build a fairly rudimentary prototype to test the concept using all the standard starter system parts, i.e. the starter, the reduction gearing, and the sprague clutch. It also uses the original bike’s battery. Funily, it took us by surprise how easy the bike is to start now. The crank based starter easily deals with the high compression engine and fires up after cranking for only a fraction of a second. This is great new and we are now building a second, more refined version that provides for an on-board battery mount and encloses all the gears for safety. We also need to change the end-of-crank engagement mechanism, most likely to spline type, and finish it off with ‘top hat’ that slightly recesses the spinning parts.

Anyway, here is a picture of the initial prototype. I did warn you that it was rudimentary.

And here is a link to a short video where you can see it in action. As you’ll see, it is a truly one man operation and the starter is very portable.

Well, it seems my updates are like the proverbial buses that only come in groups with lengthy gaps in between. It must be the chilly weather we are now having that keeps me from spending too much time in a cold garage.

In any event, I am pleased to have finally completed the required carbon fibre work a couple of weeks ago. I have to admit that had I known how much work this would require, I’d not had started it most likely. Still, I learnt lots and not least have a much greater appreciation for the people who do this professionally, and why they charge what they do for one off work.

The main features of the new airbox configuration are:

• much larger volume (~15 litres)
• true ram-air system with air passing directly through the head stock
• monitoring of airbox pressure (and vacuum)
• accommodation for ECU, PowerCommander, and exhaust gas sensor amplifier
• tunable bellmouths for shape and length
• reduction in weight
• semi-structural front air scoop to carry dash-logger and upper fairing

As usual, most changes have plenty of ramifications. As such, this change required modifications to the frame, mainly in the headstock area, and a new tank. Relocating the ECU also required modifications to the wiring harness. as we were changing the headstock, we also incorporated a means to change the yoke offset. This then required a new way to mount the steering damper that now lives just above the bottom yoke. Its mounting is adjustable to account for the range of available yoke offsets.

So, what started as a means to make the most of the tuned engine with a proper airbox, turned into an almost year long project. The outcome, however, is very positive and track testing will commence in January. Let’s see which bit falls of first!

This picture shows the lower part of the airbox in place and one type of the carbon fibre bell mouths. You can also see where the ECU is now located at the upper front part of the airbox.

Here you can see the entire airbox installed with the tank still off. The top is simply joined to the bottom with black racers tape (duct tape). The bottom self-locates at the front around the air scoop and mounts to the throttle bodies via the stock mounting plate which we retained.

And lastly, a couple of pictures with the tank in place (now painted as you can see. The cover is held on with a quarter turn fastener. It certainly makes it easy to connect a laptop for data downloading and engine mapping.

You earn bonus points for knowning which other racing bike, achieving fame in the mid 90’s, had the top of the airbox exposed through the tank. I am still kicking myself for not buying one when they still could be had for only 15,000-20,000 UK Pounds about 8 years ago!

One of the items we identified pretty early on was that we wanted to shorten the left side of the engine. This is where the starter gears, flywheel, and generator live. To save weight, we wanted to eliminate the starter and its associated gears and so went a little bit further.

The tuned engine now has a shorter crank with a new, one-off flywheel and a light weight generator. This allowed us to shorten the left side engine cover considerably. You can see the difference in the following picture with the stock one on the right.

Here is a picture with the new flywheel installed on the shortened crank. The remaining flywheel mass is now located much closer to the centre of the engine.

The hole in the middle of the modified cover is for access with the external starter.

The next set of pictures show the difference between the stock flwheel and the replacement we are now using. It has about a third of the inertia and also weighs a lot less. We also make one that has even less inertia but it makes the engine less happy.

So from left to right:

• This is the lightest version of the flywheels we’ve tried so far. It has probably just slightly too little inertia, although it may work well at certain tracks. We will track test it in 2008 more thoroughly. The black band around the centre are the new rotor magnets imbedded in carbon fibre epoxy. It has just under a quarter of the inertia compared to the stock one.
• This is a modified stock version which has the recess for the sprague clutch machined off. It is 600 grams lighter and has 70% of the inertia compared to the stock one.
• The left most one is the stock one. It weighs 2.2 kg and has an inertia of 3.25 kg.dm2. The starter gear and sprague clutch has been removed in this shot.

You can also see the difference in the generator stator. The replacement produces about 200W which is still plenty to keep the battery topped up and run the ECU, fuel injection, fuel pump, and dash logger.

Eliminating the starter system and running a smaller generator also allows us to run a much smaller battery. The picture below shows the 3 different batteries that we’ve used.

From right to left:

• stock battery, 4140 g wet weight
• what we used last year, still running an on-board starter and standard generator, 2660 g
• what we are using now with the light weight generator and no on-board starter, 890 g

That means just in battery weight we saved over 3 kg. Quite a difference! All in pursuit of more power and lighter weight.