Date: Fri Sep 6 23:21:34 1991   
To: (z-car list)

>I recently read in rec.autos.tech how someone "cleaned" the cylinders of his 
>mustang by spraying water into the intake while the motor ran at 2k rpms.
>Anyone tried this?  Does it work?

yes and no.  Water injection over the long term WILL clean an engine of
deposits but gunking a quart or two of water down the intake will do 
little other than maybe stall the engine.

John

From: John De Armond
X-Source: The Hotrod Mailing list
Date: Jun 1992
Subject: Re: Water injection

>Just joined the list a few days ago, and this is my first post here...
>T'was wondering, has enyone ever tinkered with water injection,
>particularly on a port injected turbo motor?

Little bit here and there. :-)

>I'm trying to track down
>data on what the percentage of water should be to fuel volume. Are
>there some rough guidelines?

There really are none because the application varies so much.  If
your engine is just on the edge of detonation without WI or if
you simply want to lean the mix a bit for economy, a fine mist
is appropriate.  If you're going for an outer limit engine
as I've been known to do, it may require as much water as gas.
This is strictly a parameter you have to determine experimentally.
It is much better done if you have a chassis dyno but it can be
done on the road.  A knock indicator is very valuable.

>Does water injection have an impact on
>the fuel-air ratio needed to promote maximum power and safe combustion
>temps. I'm guessing that you might be able to run a tad leaner at high
>power because the water is providing some of the cooling effect that
>a richer than stoch mixture normally does. Is this reasonable?

Actually you can run a lot leaner.  You can bring the mixture back
to near stoch and use the water for cooling.

>Water injection sounds like a good idea to me for two reasons.
>1) Enhanced charge cooling and detonation resistance
>2) It keeps the combustion chamber and in/outlet ports free of
>flow disrupting carbon deposits.

Yup.

>There must be some drawbacks to water injection (besides the havoc that
>would result if you let it run out). What are they ???

There are no operational drawbacks but there are some logistical
problems.  I've seen reports that WI shortens engine life but I have
a few engines to prove the counterpoint.  I've put over 60k miles on
a 25 psi boost Datsun turbomotor.  The logistical problems include
having to remember to fill the tank, locating the tank somewhere in
the car, freezing in the winter and if you use methanol in the water,
the cost.

>It seems that an electronic injection would be fairly easily adapted
>to a water injection setup. I was thinking of simply using an injector
>pulse to (indirectly) drive another injector selenoid to squirt water
>directly into the intake, immediately after the throttle body. The
>tricky part would be finding a way to provide a pressurized water
>source. Maybe compressed air over water?

Normal fuel injectors rust in the presence of water in milliseconds.
This is the voice of experience talking.  The simplest arrangement is
to simply pressurize the water container with boost pressure and
pipe the water to the intake upstream of the compressor.  This setup
provides a proportional flow.  Some means of holding the flow off
until a minimum boost is reached is required.  I've used a valve
originally used in the de-icing system of the B-29 but they're
no longer available (no shit!)  This valve is very simple and could
be duplicated easily.  It was a simple right angle plug valve but
with a spring between the stem and the plug.  The flow goes so that
the pressure lifts the plug against the spring pressure.
Another option is a simple pressure switch and a solonoid valve.
In any event, a typical cutin pressure is around 5 psi.  I thread
the B-29 valve to accept a holly carb jet that sets the injection
flow.

I've experimented with a number of exotic injection schemes and the only
thing I've found that improves this setup is to hook the solonoid valve
up to a circuit that reads a knock sensor to provide the injection gate.

The worst problem you'll face is keeping the engine clean.  I lost
an engine at 160 mph one day because of a flek of rust that found its
way through the filter and stopped up the Holly jet.  Instant piston
kit.

John

From: John De Armond
X-Source: The Hotrod Mailing list
Date: Jun 1992
Subject: Re: water injection

> I finally got around to trying water injection on the Yamaha.  I used a
>small plastic bottle tucked under the side cover, one line coming from
>the carb plenum, and one line going to the inlet runner after the air
>cleaner.  Made sure the bottle was lower than the end of the outlet hose
>to prevent siphoning, filled it with tap water, and went for a ride.  I
>expected to need to calibrate the water flow, but I didn't expect the
>.100 diameter orifice to be *too* far wrong.
>
> WRONG.  I can see John laughing...

:-)

>  The pressurized bottle type of water injection works on *pressure*.
>Pressure is dependent on *area*.  The .1 psi or so I blew was multiplied
>many times by the area of the surface of the water in the bottle, and
>squirted FAR MORE water than I had figured.  Duh...

Still 0.1 psi on the orfice.  The force on the bottom of the jar is
proportional to the area but not the PSI.  That orfice was just
flat too big :-)

>  I'll definitely need some type of metering system.  I remember seeing
>some inline needle valve rigs for adjusting air flow for aquarium
>aerators; if I can find one I'll put it in and try it.

Even better, thread a piece of brass or whatever to accept Holly jets.
The setup in my Z uses a #50 jet.  That's a LOT of water.  If a
Holly jet is too big, consider a Mikuni jet.  I forget what those threads
are, probably about 5 mm coarse.

>  It'd take one heckuva washer pump to keep up with the volume you get
>from a pressurized tank.

Yup.

Speaking of turbocharging, I've been meaning to post this.  Recently
picked up the latest revision of Hugh MacInnes's "Turbochargers".
In the section on lubrication, he discusses low mount turbos and
recommends a couple of scavenging pumps.  One is from Martin Schneider
Designed Systems, 9063 West Washington Rd., Culver City, CA 90230
213 559 0020. This is a real neat looking little motor driven pump.
He also lists a couple of brands of belt driven scavenging pumps
but you're probably not interested in that.

John

From: John De Armond
X-Source: The Hotrod Mailing list
Date: Jun 1992
Subject: Re: water injection

>-> Even better, thread a piece of brass or whatever to accept Holly
>-> jets. The setup in my Z uses a #50 jet.  That's a LOT of water.  If a
>
> Er... nope.  Not at eight bucks a pair, local price.  I'd rather use
>some sort of adjustable orifice.

How much!?!?!?!  Can I sell 'em a truckload?  I'd probably be willing
to dig a few approrpiate sizes from my collection and sell 'em for oh,
a buck apiece.

> Best as I can tell, the turbo actually starts developing boost
>somewhere around 60mph; the bike uses a reed valve in the airbox to
>bypass the compressor when off-boost, and I get the surgies just before
>the boost guage picks up around 65mph.

You will want to hold off on the injection until right before knock
starts.  I've previously described a holdoff valve from a B-29 bomber
deicer I have used.  I've also made them from 90 degree 1/4" industrial
instrumentation plug valves.  You could buy a brass Swagelock Whitey valve
for $60 but I would not want to.  I see 'em frequently surplus.
Just cut the stem behind the plug, trim some stem for clearance,
braze or silver solder a small washer just under the packing for a spring
to rest on and insert a spring between this shoulder and the plug.  The
remainder of the stem serves as a spring guide.  Flow comes from
the axial port and exits the radial one.  The water must build enough
pressure to overcome the spring before flow starts.  The knob is now
a setpoint adjustment.   Another method is a checkvalve equipped with
a larger spring.  The disadvantage of that is you have to disassemble
the thing to adjust it.  Since water injection is needed earlier in
hot weather, it's a good idea to have the adjustment available.

John

From: John De Armond
X-Source: The Hotrod Mailing list
Date: Jun 1992
Subject: Re: water injection

>$8 a pop for jets? Yow. $3.95 last time I bought some - a couple months ago.
>You could try a length of small tubing and get your resistance from length
>instead of just orifice. Cut or extend to tune over some range.
>
>Sounds like you want a backwards power valve. float needle-n-seat with a
>soft spring, maybe an adjustment screw/keeper?

Small point.  If you want to tune in a methodical manner with repeatable
results, something adjustable and repeatable.  Particularly in the
area of flow metering, carburator jets are very useful because they
are calibrated in terms of FLOW instead of orfice diameter.

>If you leave the bottle pressurized, won't it keep dribbling excess water
>until the pressure bleeds down?

I missed that point the first go-round.  The bottle must communicate
freely with the intake manifold.  Otherwise the delivery won't match
the pressure curve.  There are two schools of thought.  One (mine)
is that the bottle should be connected to the manifold with a relatively
large diameter tube.  The bottle must be arranged so water cannot
slosh back into the tube.  A baffle arrangement is useful.  The other
school of thought is typified by the old Crown system.  There is a
tube from the pressurized manifold to the bottle but with a checkvalve
in it.  Another tube with an orfice runs from the top of the bottle to
a non-pressurized part of the manifold, typically in the air cleaner.
A third tube contains an orfice and injects the water.  The idea is
the air orfice can control the rise and fall of the water pressure
and minimize how much water can be sloshed if that ever happens.
One area where this system may be good is in blow-through throttle
applications.  The pressure delay will cause water to continue to be
injected for awhile after the throttle is closed.  Closing the throttle
causes the compressed air to heat from compressor surge.  This added  heat
can cause trailing throttle detonation and the extra water can
address this.  For my suck-through applications, I found the Crown system
to be excessivly complicated, particularly regarding tuning.

John

From: Dan Malek <emory!widener!westford.ccur.com!dan>
X-Source: The Hotrod Mailing list
Date: Mar 1993
Subject: Re: Do gas and water mix? 
X-Sequence: 4311

> Date:  Sun, 28 Feb 93 22:55 EST
> From:  Dave W. and JGD
> 
>

I will add my few cents worth of comments based upon past experimentation,
and the little I remember from obtaining a chemistry degree.

	<...Stuff deleted...>

> The literature tells us the engine will run with just about any
>amount of water short of hydraulic locking the thing.  However, for a
>normal engine, it doesn't get you much.  The water displaces fuel at low
>flows, which isn't good for power.  When you turn the hose on full, it
>starts displacing air too.  Boo, hiss.

My experience tells me the same.  I used to spray water into the top of the
carb on my old, old Mustang.  It is amazing how much you can push into an
engine, and how poorly is runs just before it locks up.  This is also a test
of piston/rod/crank strength.  My rods always lost :-(.

> So far, a dud.  Now let's look at yer generic high boost turbocharged
>engine, like, for example, John's Z.  John has mentioned the car's
>full-boost thirst for water more than once.  As far as I know, he's
>simply shooting water somewhere upstream and letting it go where it
>will, which isn't all that bad.  

>[Yes, I just spray the water stream directly at the turbo fan through
>a tap under the throttle and let the turbo fold, spindle, mutillate
>and otherwise homogenize the mix.  JGD]

I have since graduated from carb class and have now entered Turbo/EFI school.
I have learned that water does bad things to fuel injectors and Mass Air
sensors.  I am now experimenting with injecting the water into the air
stream after the Mass Air sensor and before the turbo (for the same reasons
that John mentioned).  Of course, this messes up the A/F computations, but
it is nothing that a water flow sensor and a couple thousand lines of code
won't cure :-)!  My sole purpose for the water injection is to run more
boost without detonation.

>However, by getting the water and fuel
>together and (probably more importantly) using detergent to minimize
>recombination from vapor into droplets, it's possible he might (1) be
>able to use the water more efficiently and therefore (2) use less of it.
>Now, if this gas-over-water "egg" theory is correct, it might not help
>the anti-detonant properties much, but then it'd be worth an experiment
>to find out.

The only way you can get a gas/water "mix" is to use a detergent.  Without it,
gas and water will separate faster than money and your wallet when you are
building performance vehicles.  If you want to put them into the same
container, and keep the gas/water ratio fairly constant, this is the only
way.

>[I have not ruled out the value of water addition, whether injected or
>emulsified.  After all, water injected engines DO make more power
>when tuned to take advantage of the water.  There may be a good sound
>physics basis for what water does, I'm just not aware of it at the
>moment.  My major concern is that we avoid the kind of perpetual motion
>machine posts that makes r.a.* so noisy.  

>It seems to me like the best way to make a stable gas/water emulsion
>would be to use the same homogenation process used in milk processing.
>That is, force the gas and water through tiny openings under high pressure
>in order to create droplets too small to coalesce.  I wonder if a 
>high pressure car wash pump like the one I have in my shop would
>generate enough pressure.  Hmmmmm.  A stable emulsion might just address
>most of the negatives of conventional water injection, things like
>the necessity of two tanks and support systems and the fact that water
>freezes. Another thought is a water rich working fluid might be better
>than just combustion products.  Anyone up to doing the calculations?
>JGD]
>----------
>Posted by: emory!chaos.lrk.ar.us!dave.williams (Dave Williams)
 


I don't think your homogenation process will work, but don't let me stop
you!  There is something about compressing gasoline to very high pressures
(high hundreds or low thousands of PSI found in the car wash pump) that is
not appealing to me :-).  The gas and water will soon separate, since the
molecules don't want to be next to each other.  I also like injecting water
separate of the gasoline mixture.  I can control the mixture ratio to meet
the engine requirements (water lasts longer), and to keep it from freezing,
just mix in some ethanol.  Now you have the advantage of oxygenating the
mixture.

So what does the water really do?  First, it cools the intake mixture.
Does the temperature drop increase the air density?  I don't know, this is
probably the first calculation since you have to account for the additional
H2Os running around.  I know I do inject more fuel, so this would indicate
a more dense air mixture.  The water will serve as a combustion "contaminate"
and will slow down the flame, further reducing detonation.  If liquid water
is entering the combustion chamber, then some of the heat from the combustion
is used to vaporize the water.  Since this lowers the combustion temperature
it would seem that power output would decrease, but then you have to account
for the expansion of the water from liquid to gas.

Lot's of computing and measuring.  Guess I have another project for the dyno!
I don't have the ability to do the computations anymore, but I will seek out
an old chemistry friend (that still does it for a living) for some help.

Like John, I know this water injection works, in an engine designed to
take advantage of it.  But you have to be careful with corrosion of the new
EFI stuff, and too much water is not a good thing.


	-- Dan

From: emory!mlb.semi.harris.com!jws (James W. Swonger)
X-Source: The Hotrod Mailing list
Date: Mar 1993
Subject: Re: Do gas and water mix?
X-Sequence: 4321

 Does water injection at cruise (medium-high vacuum, low load) provide
the same MPG benefit as EGR? Seems that a vacuum-demand water shooter
would let the cylinders fill up with inert gas just like EGR. The
discussion so far seems aimed at the high end; is there benefit to
full-time water injection, or maybe even a two-circuit setup? One
for high vacuum and one for high demand?

 Maybe somebody can rig up water injection that runs standalone using
a knock sensor to close the metering loop? Might be a good DIY article
for PE...


[Good old fashioned air works well and doesn't require any storage space.
I posted an article a month or so ago to rec.autos.tech about an air
induction system I designed for my El Camino over 10 years ago that pushed
the smallblock mileage up over 25 mpg on the highway.  I can dig it out
and post it here if anyone is interested.

I've thought about a closed loop water injection system but I've not yet
come up with a good metering unit.  Conventional fuel injectors lock up
so fast you can almost hear it in the presence of water.  conventional
solenoid valves are too slow.  Any suggestions?  JGD]

From: GREG GRANVILLE <emory!arlvax.psu.edu!GAG>
X-Source: The Hotrod Mailing list
Date: Mar 1993
Subject: Re: Do gas and water mix?
X-Sequence: 4333

>I've thought about a closed loop water injection system but I've not yet
>come up with a good metering unit.  Conventional fuel injectors lock up
>so fast you can almost hear it in the presence of water.  conventional
>solenoid valves are too slow.  Any suggestions?  JGD]

How fast does the solenoid have to be? I was thinking of pulse width
modulating a solenoid at about 18 cps. Haven't done much testing yet,
but I seem to be able to get a fair amount of flow rate control.
Was considering using boost pressure (approx 15PSI) to pump the
fluid thru a .040 or so carb jet - directly into the compressor inlet
as you've described in the past.
Is it necessary to do the PWM at a faster rate than this?

...Greg

[That rate should do it.  What kind of valve?  I'm VERY interested.  JGD]

From: "KEN MOSHER" <emory!imd.sterling.com!KEN_MOSHER>
X-Source: The Hotrod Mailing list
Date: Mar 1993
Subject: Ok, how do I do water injec
X-Sequence: 4351

                          SUBJECT:  Ok, how do I do water injection?
All the talk about water injection here has got me thinking ... I wonder how
hard it would be to come up with a setup for my Buick turbo V6?  I would want
to use the setup mostly at the track, but it'd be fun to turn up the wick on
street gas also!

Here's a couple of my preliminary thoughts:

1.  The injection point for the water mist would be just ahead of the turbo ...
I was thinking of the Turbo inlet bell,  right where the valve cover oil
breather line used to go.

2.  Maybe an early version would be manually operated by a momentary contact
switch ... I was thinking of something like a windshield washer pump or
something like that to pump the fluid.

3.  I was thinking of using an NOS fogger nozzle with pretty small jets to be
used for the injector.  Very simple and should be fairly resistant to moisture.

4.  Mount another windshield washer fluid tank next to the existing one for the
injection fluid ... probably distilled water with maybe a little alcohol mixed
in.

Eventually I'd want some kind of boost actuated control, or maybe RPM related?
A simple type of control could read the RPM from the green tach connector (next
to the fuel pump test connector) and actuate the pump at a desired RPM.

Maybe a selenoid could be used to tap into manifold pressure to pump the water?

These are all preliminary thoughts ... Any brillant engineering ideas on this?
I'm looking for some advice from those of you that have done this ... lessons
learned are the best kind! <grin>

-- Ken Mosher
-- Buick Grand National:  A *BOOST* of Buick Performance

[The systems I build for turbo-Z engines are quite similar to the one
described as the Dick Griffin system on page 135 of McInnes's book.  He
doesn't use revision numbers (grrr) but this is the version with the
glowing McLaren installation on the rear cover.  This system uses boost
pressure supplied to the water resevoir to provide the driving force for
the injection, thus, the injection flow is metered according to boost.
I change the design a bit to make it work better.  First, I use very large
bore tubing to connect the manifold pressure to the tank.  This is
important so that the tubing and the air volume of the tank does not
form an RC network that delays the start of injection.  Next I insert
a pressure holdoff valve in the water line.  This valve functions to
hold off the injection until the boost pressure reaches a certain level.
I originally used deicing valves from B-17 airplanes I got from Turbo
Toms but those are all gone now so I build my own.  This consists of
modifying a right angle Parker needle valve  by cutting the shaft to
the needle and inserting a spring between the stem and the needle.
Flow is arranged so that pressure attempts to lift the needle and is
opposed by the spring.  The knob sets the setpoint.  If you use a
valve with a large diameter needle and seat, the slope is pretty
steep, a desirable quality.

In the first installations, I used "pony" beer kegs for the water tanks
but the shape is not too hot for underhood installation.  For the last
few I've custom fabricated fiberglass tanks.  This is pretty easy to
do.  Simply make a mould of the tank to fit whatever space is available
using cardboard and/or posterboard and hotmelt glue.  Lay up the fiberglass
over this and dissolve out the cardboard with a lye solution.  The hotmelt
will drop out.  Embed steel or brass fittings and include internal struts to
brace flat surfaces.  JGD]

From: GREG GRANVILLE <emory!arlvax.psu.edu!GAG>
X-Source: The Hotrod Mailing list
Date: Mar 1993
Subject: Re: Do gas and water mix?
X-Sequence: 4352

[That rate should do it.  What kind of valve?  I'm VERY interested.  JGD]

The one that I was playing with was made by KIP Inc. (1-800-722-5KIP)
They have a wide variety of solenoids for fluid control... But then,
a lot of other companies do too.
I'll do some more exact testing sometime soon, and will report the
results. They claim a response time between 6 and 12ms, depending on
several variables... one of which is the medium being controlled.
There are several tricks that can be used to speed up selenoid response
time also, which I'm sure you're well aware of.
I'll measure the range of flow control at various PW rates, and at
different head pressures. Is there anything else I should be looking at?

--------------------------------------------------------------------------
| Greg Granville       |     99 Blanchard St.   |  gag@arlvax.psu.edu    |
| 814-865-3310 (work)  |    Osceola Mills, PA   |  Applied Research Lab  |
| 814-339-7244 (home)  |         16666          |  University Park, PA   |
--------------------------------------------------------------------------
|          "I say we take off, and nuke the site from orbit.             |
|          It's the only way to be sure" -Lt. Ripley - Aliens            |
--------------------------------------------------------------------------

From: Dan Malek <emory!widener!westford.ccur.com!dan>
X-Source: The Hotrod Mailing list
Date: Mar 1993
Subject: Re: Do gas and water mix? 
X-Sequence: 4358

> Date:  Tue, 2 Mar 93 14:50 EST
> From:  hotrod@dixie.com (The Hotrod List)
> 
> Does water injection at cruise (medium-high vacuum, low load) provide 
>the same MPG benefit as EGR? Seems that a vacuum-demand water shooter

Hmmm....I never thought of EGR as providing a MPG benefit :-).  Is there
something more I should know?

>[Good old fashioned air works well and doesn't require any storage space.
>I posted an article a month or so ago to rec.autos.tech about an air
                                          ^^^^^^^^^^^^^^

WHAT!!  I stopped reading that when this list was started.  This implies that
there is still some useful information over there AND John is contributing
(or maybe it is useful because John is contributing :-).  Yes, I would like
to see that information posted here.  Thanks.

[Still a few nuggets floating around out there.  Besides, I gotta prowl
around there occasionally just to let 'em know about this list :-)  JGD]

>induction system I designed for my El Camino over 10 years ago that pushed
>the smallblock mileage up over 25 mpg on the highway.  I can dig it out
>and post it here if anyone is interested.

>I've thought about a closed loop water injection system but I've not yet
>come up with a good metering unit.  Conventional fuel injectors lock up
>so fast you can almost hear it in the presence of water.  conventional
>solenoid valves are too slow.  Any suggestions?  JGD]

Well, I have thought about trying to use (find, build, etc.) something similar
to the medical "pumps" that deliver controlled medications.  You know, the
kind that look like a flexible tube wrapped around some number (2,3,4, etc.)
of rollers connected to a motor.  As the rollers turn, they trap a fixed amount
of liquid in the tube, and push it out.  Connect the tube to a jet for
atomization, and control the speed of the motor with the engine controller.
Sounds easy...now to find the parts and the time.

>----------
>Posted by: emory!mlb.semi.harris.com!jws (James W. Swonger)
> 


	-- Dan

From: emory!zeus.tamu.edu!smc9782 (DEAD DOG)
X-Source: The Hotrod Mailing list
Date: Mar 1993
Subject: Re: Ok, how do I do water injec
X-Sequence: 4375

>
>[The systems I build for turbo-Z engines are quite similar to the one
>described as the Dick Griffin system on page 135 of McInnes's book.  He

Just curious, but what's the book's title?
Is it "How To Select and Install Turbochargers"?

[Just plain "Turbochargers.  HP book # 49  JGD]

>doesn't use revision numbers (grrr) but this is the version with the
>glowing McLaren installation on the rear cover.  This system uses boost
>pressure supplied to the water resevoir to provide the driving force for
>the injection, thus, the injection flow is metered according to boost.
>
>In the first installations, I used "pony" beer kegs for the water tanks
>but the shape is not too hot for underhood installation.  For the last
>few I've custom fabricated fiberglass tanks.  This is pretty easy to

My Father did something like this on his turbocharged Corvair back
in the eary 70's.  Only he used the windshield washer bottle out of
a VW bug for the pressurized water tank.  He picked the VW water bottle
because in the Bug the air from the spare tire is used to pressurize
the water bottle in order to squirt water on the windshield without the use of
a pump.  So it didn't take any effort to convert the water bottle for
water injection.  Only problem was that the capacity of the VW bottle was
kinda small.


Steve Cole 			'66 Corvair Convertible
				'65 Turbocharge autoxer (in construction)

[That seemed to be pretty popular back then on Corvairs.  That windshield
washer bottle would last oh, about 5 seconds at the rate I have to inject
for 25-28 psi of boost.  JGD]

From: emory!naucse.cse.nau.edu!met (Millam Tackitt)
X-Source: The Hotrod Mailing list
Date: May 1994
Subject: Water Injection
X-Sequence: 8524

I saw a post where someone was worried about injecting too much water
into their engine and ruining it, something that I think is basically
impossible.  I used to have a very simple setup on my old '71 Torino
w/302 auto that was just a ford windsheild washer box reservoir
w/internal pump.  The injector was just a piece of 1/8" copper tubing
that was crimped in the center of the orifice such that it left two .014
holes, drilled to size, of course.  A toggle switch to run it was all
the input it needed.

When I drag-raced it I mixed water and methanol 50/50, pumping a full
quart and a half in 13.235 seconds.  I gained a full tenth of a second
believe it or not!  After three runs I had to replace the plugs as both
of the electrodes were eaten away leaving a .090+ gap.  I opened the
top-end during the week after my first races.  the pistons and valves
were so clean that you'd swear that they were brand-spankin' new!  There
was no noticable deterioration of the pistons.

So I'd say to go ahead and inject as much water as my orifice setup will
allow with an adjustable metering needle and seat you can restrict the
flow to be just drowning the knock.  Unless you are foolish enough to
run some kind of large orifice flow, I can't see anything harming the
engine, if you manage to flow enough water to drown the flame you're
going to lose so much power that you won't run long enough to harm the
engine anyways.

Millam

[That pretty much agrees with my experience.  I've not put too much
research into this because water injection seems to tolerant to a
wide range of water flow.  You'll inject enough to put the fire out
long before you hydraulic the motor.  My little bit of tuning, as it
were, consisted of finding out approximately how little water I could
get by with so that I didn't have to fill the tank so often.  I've
never used water injection in a non-turbocharged engine so my
setup is dirt simple.  A pressure line from the intake applies pressure
to the water resevoir.  Water is injected into the intake before
the turbo and is metered with a Holly carb jet.  A homemade holdoff
valve holds off the water flow until an adjustable minimum pressure is
reached.  That's it.  Flow is roughly proportional to boost pressure.
JGD]