Water Injection

[you]

Hey all..been a while since ive posted but as the winter moves along im starting to look toward spring projects....ill be moving the van over to the new house soon and will finally have a good place to work on it (never had concrete underneath me before, or a carport overhead)...first things will be plumbing in the smc vacuum switch to control load enrichment and hopefully source a couple 2.0l afms to play with...really looking forward to dialing this thing in and the engine loosening up....but i dont expect that to take all that long so im toying with the idea of water injection....i have done very little research so far but im thinking that i will be able to get away with running lower octane fuel at higher afrs, all while maintaining lower chts..sounds like a pipe dream but i think its doable....i remember once driving through the nc highlands and doggin the van up a pretty good grade at maybe 55 er so...chts were climbing and then i got into the fog and presto they dropped 50 degrees....proof enough for me to try it out....off the cuff im thinking i may try doin a very simple setup (tank>atomizer>egr valve, no pumps/switches/etc)....anyone have experience with this er water/meth injection (for the winter) and care to share their wisdom?

[andy198712]

so are you after a constant water injection system? instead of one that kicks in under load to cool for boost reasons? very cool, i dabbled with this kinda thing very basically and briefly!

[you]

no boost...running lobotamized cis-e per piledrivers setup(i have a write up somewhar in the forum)....if i were to run it via the egr valve it would be based on throttle position...im looking for cool heads and added mpgs would be a bonus...assuming power isnt sacrificed

im curious if manifold vacuum is enough to draw water through an atomizer....set up this way it would allow for super lean light load cruise..at full load when vac is low the smc would richen it up to take care of power and heating...maybe a balancing point can be reached to make it work....just in the brainstorming stage now

[gearheadgreg]

Yes, I have done this and played with it. There's a few ways to go about it.

You do NOT want to set it up to have manifold vacuum draw it in. The most vacuum (which means the most injection with that type of system) is on overrun, and next at idle. The least injection volume is at WOT, which is zero vacuum.

You need a system that injects according to airflow, which is possible. I have done it with a carburetor - drilling through the venturi and epoxying a very small brass tube (hobby/model store) that extends into the carburetor barrel a bit. This way, it delivers water something roughly equivalent to airflow, being drawn in by the engine vacuum signal through the venturi. Additional parts that are good, are a solenoid valve to prevent waterflow when parked on an incline (connected to the coil so solenoid shuts off when ignition off) and an adjustable valve to help determine overall flow range (other than brass tube diameter), and a reservoir that is slightly below the tube height in the engine bay.

On a CIS system, there are some throttle bodies that have the EGR restrictor or venturi that does the same thing as on the carburetor for a vacuum signal. I have also used electronic systems that injected variably according to a vacuum signal.

David Vizard wrote several articles about water injection in the 70s and 80s, and was able to run stupid high compression on very low octane fuels and get very good mileage. Variable injection according to engine load is what you want.

[you]

thanks for the input greg, ive been reading some of your other posts in wi threads here, great info....i guess i should be more specific bout things...running a t4 in a vanagon westy....i have built/rebuilt several engines for it and have always noted that upon teardown the valves and seats always had the most damage on cylinder 1...i finally got a 4 channel cht gauge and on two setups i noticed that head temps were consistently 40+ degrees hotter on that cylinder....for the longest while i couldnt wrap my head around it until i looked inside a plenum and noted the design of the egr piping inside...i hooked it up to the vacuum and sho nuff it favors cylinder 1...so i come to the conclusion that i either have to tune each cylinder individually which seems like a pain plus buying egt probes etc, reinstall some sort of egr, redesign the intake, or replace the egr with water/meth injection....wi makes the most sense to me at this point...utilizing the mechanical egr valve would essentially cut off flow at an upper and lower limit i choose (took one apart and it looks straigh forward)....if i go with a pressurized system the pump could be tied to a throttle position switch...or i could do exactly as you describe and plumb it to the ported vacuum on the tb...

one of my main concerns currently is corrosion, it seems "common knowledge" that its corrosive particularly to aluminum but i keep reading conflicting info...for example:

from methanol.org
"But you can't simply put pure methanol fuel in your car today. At high levels, methanol fuel is corrosive to certain materials commonly used in engines and fuel lines. In order to be able to operate on high-level blends like M-85 (a mixture of 85% methanol and 15% gasoline), small modifcations must be made to an engine to include methanol compatible componenets. These modifications generally cost less than 0.5% the cost of a new automobile. Low-level blends of methanol though do not cause adverse effects to a car's engine., and can be used in cars today where the fuel is available without any adverse effect Methanol is used as a denaturant for ethanol fuel in many countries, and some amounts of methanol are combusted with those blends everyday."

from methanex.com
"What materials are compatible with methanol?
Pure methanol is non-corrosive to most metals (not including lead, magnesium and platinum) at ambient temperatures. However, coatings of copper (or copper alloys), zinc (including galvanized steel) or aluminum are attacked at high temperatures and more slowly at ambient temperatures. Stainless steel and carbon steel are typically used in Methanex plants.

Plastics are generally not recommended for storage purposes due to long-term deterioration of the material and the subsequent risk of methanol contamination.

Many resins, nylons and rubbers, particularly nitrile (Buna-N), ethylene-propylene, teflon and neoprene are used satisfactorily as components of equipment in methanol service."



can anyone with wi/meth inj speak to head/valve/seat/intake longevity (with say a solution of 35% meth/mater, ran at ~15% of fluid to air ratio)?

[gearheadgreg]

The corrosion issue is a complete non-issue, IMO when handled correctly. There are lots and lots and lots of vehicles that have lots of miles, with zero metal corrosion due to water-methanol. The issues I have seen are relatively minor and usually avoidable (throttle body shaft seals, heating grids on diesels, etc). I have never seen interior engine damage due to WM.

The water in methanol cuts down the corrosive nature a huge amount, and I personally disagree with some of the quotes above.

Teflon is NOT methanol safe. It basically melts in contact with it. There was a flow-meter that was developed and the small turbine ran in Teflon bushings. The bushings lasted for about an hour in straight methanol, slightly longer with water in the mix, but you get the picture. Buna isn't methanol-proof, either, only resistant.

EPDM rubber, nylon, some other plastics are perfectly fine. Aluminum shouldn't be used for a long term storage container, but lots of people use aluminum reservoirs and it doesn't melt through like acid or anything, it just oxidizes the inside some.

The things you need to fashion your own WM injection setup -
1. Atomization
2. Delivery control (variable)
3. Safety

For atomization, you need pressure differential and a nozzle of some sort. If you do it like my simple carb setup, you are using the lower pressure of the venturi to draw the fluid out and shear it into smaller particles, just like it does for gasoline. In a pump-based system, the pump pressure pushes it through the nozzle and it comes out into lower pressure. Variable delivery, *according to what the engine needs* is important and what most people don't pay enough attention to (IMO).

People will work very hard to get accurate fuel delivery and ignition timing, etc. but then just have a WMI system that is on/off in delivery (shaking head). You can get some benefit that way, but it's leaving so much on the table it's silly.

Safety is just to keep the thing from emptying the reservoir into your engine when it isn't running - causing a hydrolock. Usually, a simple 12V fluid solenoid can handle this.

[you]

looking to get back into designing an h2o/meth setup again....im leaning towards a mechanical setup...anyone have ideas on a small pump that could be driven off the crank/pulley and work suitably well?...could be an airpump to pressurize a tank or a waterpump off anything: just looking for something thats cheap/easily sourced, and will package relatively well in the engine compartment...preciate any of yalls input

[gearheadgreg]

Mechanically driven pumps are not going to be cheap or easy, IMO. You could research some alcohol mechanical injection setups (drag race type setups, etc).

Using an air pump to pressurize a WM container is a good idea, but unfortunately, the reality is you will pump it up with pressure, and it will then spray and spray, even with the throttle shut, etc. Basically, you won't get correct or responsive injection.

The previous standard pump for commercial systems - available from Shurflo or even WM suppliers, is only around $40-50 I think. It's a great pump that can be adjusted to put out 75-150psi pretty easily, and runs very well with a PWM controller. The amp draw is way too much to be driven by a Megasquirt or similar, FYI.

I have made some UBER simple setups, based on DIY type articles from the 1970s, On a carburetor, I plumbed in a 1/16" tube through a hole I drilled in a carburetor venturi. Then ran hose to a reservoir that was just below that level (on the horizontal). The low pressure at the tube end drew in the higher pressure (atmospheric) from the reservoir, basically just like a carburetor with a float bowl.

Worked easily, and with an adjustable needle valve in the hose, made it adjustable. All you need to do is find a low-pressure area in your intake, relative to atmospheric. Otherwise, use a sensor (like a MAF or MAP) to drive a controller with a 0-5V signal to drive the pump with a PWM wave.

[you]

i have an smc vac gauge/switch i had planned to use for fuel enrichment and could program it to control the wi but im currently running mechanical fuel injection and wanted to try to run the meth that way as well...leaning towards finding one of those t4 pulley setups for a smog pump or ac unit to run a small pump mounted on the fan housing...run that to the egr port on the plenum...maybe use an injector similar to cis that has a cracking pressure so theres no leaks at shutdown....just shooting from the hip at this point

...i know some older john deeres ran wi but the pump units look bulky and appear more than what i need for my application but i know nothing about them...maybe there were smaller units or ?

i was sorta thinking just a pump from a small water cooled motor might fit the bill (atv, jetski, whathaveyou)...or maybe something from an industrial application, dental tool, aircraft, etc....looking to explore all the options and get the creative juices flowing....thanks for all your thoughts thus far..been very helpful

[gearheadgreg]

You want to remember that using a small smogpump as an air pump to pressurize a tank to push the water out of it, will act like a balloon or air compressor tank. With RPMs, it will build lots of pressure, but when you back off the throttle and your water injection need is zero, it will be injecting at max, or for a long leak-down time. Not what you want.

I know - it's hard to get a very simple, robust setup. If you find a simple water-pump that can be crank driven, let me know! I've got lots of leftovers on the shelf from partial systems, prototypes, etc as well as some vintage setups - one that's called Octa-gane, and fit on the top of a carburetor with it's own venturi and had it's own float bowl and reservoir, and simply injected according to airflow/pressure diff just like a carb.

[andy198712]

I hooked a bit of clear tubing up to my carb vac take off, then to a fish tank metering valve, then to a 2litre coke bottle with a hole in the lid and the tube going to the bottom.

This worked well and drew a lot of vacum, but as has been said, not at ideal times!!
It ran quieter I thought at cruise and seemed happier. But need to measure temps.

I've seen over methods where you have a container you can seal, and two holes I the lid, one has a spout which you pipe to vacum, and the other you pipe to an air stone under the water.
So as the engine pulls vacum I the jar, air is pulled in the other pipe and through the air stone causing bubbles and mist which is drawn into the engine...

These are all diy options

[you]

andy- i played around with hooking up a shopvac to the tb and plenum just to get an actual visual of how the water is drawn in different setups and while the option youre suggesting would be easiest, like you said, it doesnt seem ideal...currently working on other things with the van ive been putting off...need to get some free time to go scrounge the junkyards...currently im toying with the idea of snowmobile water pumps(still need to get my hands on one just to eyeball size, etc) or a chemical pump from some sort of industrial application..i browsed ebay a bit and there is a fair selection of these pumps but currently id like to find one at a ridiculously low pricepoint seeing as im not entirely sure it will work as i hope...right now the easiest setup seems to be a small pump plumbed to the plenum driven off a smog pump/ac pulley (need to source one of these too)...it wont be load sensitive but i believe it will still be effective at keeping head temps down and allow me to pull some fuel out....ill post pics once things start coming together

[gearheadgreg]

Nice - keep up the progress reports.