|
| |
WATER INJECTION
By Randy Norian
An old trick, injecting water into the
exhaust pipes is a very effective way to cool the pipes and lower the operating range of
an expansion chamber. I obtained a copy of an SAE paper 931506, expanding the torque curve
of a Two Stroke Motorcycle race engine by exhaust water injection", by Robert Fleck
at QUB. In this paper, they built a simple system to inject water into the headpipe of a
125cc test motor and recorded power gains of up to 24% in the lower midrange. I decided to
copy their setup and was able to reproduce their results on the dyno. After that, I
decided to built a usable system that would function smoothly on my streetlracebike.
But first, some numbers.
Fleck recorded pipe temps at several points along the exhaust pipe. In the diffuser,
without water injection, he recorded gas temps rising from 425 deg Cat 9000rpm, to 510 deg
Cat 10800 rpm.
With the water injection active, the same sensor recorded temps of 150 degrees Cat 9K RPM,
rising to 420 deg Cat 10800 rpm.
Lets go back to the equation for tuned length.
If we consider a stock RG500, this distance is about 84cm, (33 inches) and Eo is 188
degrees. Using Vs of 1700 fps, this formula predicts a peak power RPM of 9684 RPM. This is
a pretty good estimate, as my bike peaked at 9500 rpm in stock form.
Now we consider Vs with water injection active, let’s say we have a mean 200 degreee
C drop in gas temps. If we use Vs of 1700 fps with no water, this figures to a new Vs of
about 1430 fps with water. Plug that into the equation for tuned length, and our same
exhaust pipe is now tuned for a peak of 8150RPM. So we have been able to shift our peak
power down almost 1500 rpm, by injecting water. Obviously, this will reduce the peak rpms
of our motor, so the trick here is to turn off the water at some point and let the engine
rev out normally on top.
If we were able to cool the pipes all the way down to room temperature, the stock Gamma
would have a peaking RPM of just 6270 RPM. Clearly, pipe temperature has a huge role in
determining the rev range of a pipe/port combination.
Controlling pipe temperature is an effective way to alter the tune of a 2 stroke.
This is how I have the water injection system
installed on my bike:
I use an NOS rpm switch, which senses the tach signal on the blu/white wire, and
‘arms’ the system between 6000 and 8400 rpm. In order to inject water, the motor
must be between those rpms and also there are 2 switches that control the system, one
senses when the powervalve is in the high rpm position, and the other is a throttle
position switch. So, half+ throttle, revs btwn 6k and 8400 rpm, and valves
"shut" will inject water. The entire setup is turned on or off by a master
switch on the dash.
I use a variable exhaust valve controller made by BDK, and the dial for this is also on
the dash. so, wherever I set the exhaust valve, the water will come on in synch with the
valve.
A 3 qt tank is mounted under the tail. this feeds a generic windshield washer pump (thru a
filter). The pump puts out 5 psi. This injects into the headpipes thru #70 round head
Mikuni main jets. Bungs welded onto the pipes accept brass fittings, which have jets
tapped into them where they go into the pipe, and have 6" of 1/8 steel brake line
‘leaders’ coming out to keep from melting the plastic water lines. I have 3 psi
check valves installed just upstream of the jets to prevent water from siphoning into the
pipes. I will get you the correct size of the brass fittings-The 3 qt tank has a low level
indicator, which lights up the sidestand indicator when the water is low.
This setup will provide about 180 seconds of injection to 4 pipes. That doesn’t sound
like a lot, but it lasted hours in frequent use in mountain ridinq. On the track, 1 qt
lasted me thru a 20 minute practice session.
The only real problem I have anymore is the switch on the powervalve, which is cheap and
keeps failing one way or another. I guess I need to drive PAST radio shack, and get a real
switch.
When the water kicks in, there is not a violent
burst of power, more of a quick whoooosh, like a turbo spooling up rapidly. It has been
very manageable so far, it only remains to fine tune the throttle opening/rpm combination
to get the water dialled in exactly. If I were to get trickier, I might pulse the water
system, or find a way to taper off water flow so that we might extend the advantage to the
highest possible rpm range. As it is, I simply ran the bike on the ~yno with and without water,
saw where the crossover points were, and switched the system there. More sophistication is
possible, but at this point I feel I am getting 90% of the benefits of a water system.
I’m not sure its a good use of my time to chase the last 10% of this setup!
This Dyno graph shows, same
engine, back-to-back, with water and without.
N ext
set of internal stingers will just have to have some water bleed outlets, or something, I
dunno, but I aim to make the water injection a part of daily life with the TZ barrels and
their comparative lack of midrange. I am changing to a 2-stage water injection setup,
prolly with 1/3 - 2/3
split of the water, 1/3 system in early and out late, half throttle or maybe 1/3 I dunno,,
2/3 in the middle of the used range, and at larger throttle openings.
The idea is to taper the water in, and out, to extend thu useful range, as it rather kills
your top end.
Also, at part throttle you dont have enough heat in the pipes to vaporize the full volume
of water, this ought to really extend the range on a tank as well, during street riding.
Injectors
|
Pipe Fittings
|
Washer Pump
|
This second Graph shows one of the curves when i first started to dabble.
Unfortunately all my other graphs were lost when the computer got stolen
from my dyno operator (and he did not take a back-up)
My set-up was much the same as what Randy described above, except i used
a 3-stage programmable shift light system to interface with the pump.
Since then the set-up has changed completely because i now use a
Wolf 3D CDI to control the water, this can be programmed via a computer
to control when the water comes in and also change ignition timing.
|
