First a story of my little water injection history that you may find interesting.

In1974 I bought a used 72 Alfa 2000 GTV with all the early emission controls.
The cars at that time were very slow. I decided a turbo might help.

The problem was at that time everyone up to then used suck-through
carbs to limit boost and I was going to try to blow-through the mechanical injection.
I put a .84 AR turbine housing on a Corvair turbocharger.
( I think it had a something like a 1.3 AR housing stock, also to limit boost).

It was my first turbo kit .... followed by making my first waste-gate…
then my first blow off valve ( I called it a “surge and lag pressure relief” because
I didn’t know that they existed).
I even fooled the Spica 400 psi mechanical injection rack to give its full flow,
but it was not enough. It had reached the fuel enrichment limit and I proceeded
to blow the lands off of my forged Mahle pistons...... Detonation!

My dad came to the rescue. He told me about the studies done in WWII which
enabled our bombers and fighters to achieve high boost using water injection.
(He was a lead mechanic on the engine line for B-24s in Ann Arbor at the Willow Run Ford Plant).

Water Injection transformed this car. This allowed performance exceeding a
911S, the fastest car at the time.


Back to the present..... this is one of the papers I used as reference.


If you study Report No. 815 under NACA papers from WWII
the tests show the limits of the different injected coolants on a
CFR test engine (which features an adjustable compression ratio).
The report is very long but this chart summarizes it best for me.


The chart's limits were established with a conservative pressure
safety margin of 7% (backed off from incipient knock pressure).

Indicated Mean Effective Pressure (IMEP) is listed on the left column
and I added boost in in.HgA and PSI on the right hand side.

They also used Fuel-Air Ratio so I converted it to AFR for our
convienence (some may prefere Lambda).


Knock limit for rich fuel- IMEP 285 lb/sq in. (about 14 psi)
For water IMEP 325 lb/sq in. (about 18 psi)
For water/meth- IMEP 460 lb/sq in. (about 30 psi)
My tests so far show that Rotaries behave similarly.

 

At what compression ratio?

Also, if I read this correctly, 14.2:1 afr is in theory as safe as 11:1 at 30 psi boost with water meth without detonation?

 

Neit jnf

Notice the amount of fluid they were using for the test.

1/2 pound fluid for one pound fuel. So add 50% to your injected weight of fuel.

We seem to run the bare minimum and sometimes we get away with it.

The scary part is that they could run the meth/water to 22 AFR at 35 pounds of boost!

Barry

 

In another study (NACA Report No. 756) a comparison of different percentages of just water was done, namely 0, 20, 40, and 60%.

This shows that on average we are not running enough injectorant.
Or put another way, we should run as much as our ignition can handle.

Barry

 

Compare the knock limit pressures when the inlet air temp goes from 150*F to 250*F.

Water stays about the same but the the meth/water curve drops below water when lean.

Remember this when designing safe guards to cut boost if the injectorant pressure is lost.

Also consider the consequences of not running a cold air inlet!

 

"72 Alfa 2000 GTV"

one of the most beautiful sedans... was running in the Under 2.5 Trans-Am in the early days. Horst Kwech... i was building a car for the series which vanished along w all factory racing just as i finished... so i switched to B Sedan SCCA and later GT3 and later to Mazda (1983)

oh, your thread...

should be required reading for all FD pilottos.

i wonder re the base fuel being 100/130 octane... as to the boost levels being transposed to our 93 octane world.

news flash!

from 1944.

howard

 

great post! and +1 on the Alfa...

 

Safe isn't the right world. These guys simply didn't consider something as too lean or too rich or dangerous. They just picked what worked best for their application.

Several things to consider, methanol energy was neglected from FAR calculations in this testing, so actual measured Lambda would be tad richer.
Knock limited power apears to be same, but overall engine temperature would be probably lower with richer mixture, which would reflect as less load on cooling system.

Barry, I'm surprized that you didn't post liquid-air ratio charts. I found them much more interesting than fuel-air charts. They clearly show, that at given liquid-air ratio, combination with less fuel but more water allows higher knock limited mean effective pressure. It can be well observed at high water to fuel ratios (0.5, 1, 1,5).

What also is very interesting is test at constant air inlet pressure of 35 inHg abs with three water-fuel ratios (0.2, 0.4, 0.6). At FAR from about 0.065 to 0.09 (15.4 - 11.1 AFR) IMEP and BSFC remains about the same as without wi, what indicates that even these "high" water flow rates have no ill effect on combustion efficiency or power production.

But this makes me wonder if most people are using such weak ignition systems, when they are seeing quite big loss of power at much lower water-fuel ratios. It seems that general thinking concludes, that ignition is sufficient when it doesn't misfire. This is probably not the case.

I agree that most people are injecting far too low amount of water. All these charts indicate, that 20% water to fuel ratio (which is still more than what most people use), with all other variables held constant, allows only 13% increase in IMEP at about 11.1 AFR. On the other hand, 50% water to fuel ratio allows 62% increase in IMEP (torque) at same AFR.

If same applies to rotaries, and I don't see any reason why it shouldn't, ridiculous power can be produced with right amount of water injection and pump gas.

 

Thanks for the thoughts,
Barry

 

I really like the fact, that they recovered water from exhaust gases(one of the main combustion products), so they wouldn't have to carry water onboard, albeit recovery aparatus had its own weight and bulk.


Exhaust valve area was hotter due to much higher mass flow, as it was done at knock limited IMEP. They mention that same power level with gasoline only would show much higher temperatures, so its pretty simple, higher power, higher temperature.

Well, as I said, I really like the fact, that this research engine practically didn't lost any power by running water injection in range from 20% to 60% of water to main fuel without any adjustment of spark advance or air-fuel mixture, except at very lean and very rich mixtures.

Very interesting thing also is, that highest levels of knock limited power were produced at very lean mixtures - at the expense of higher air consumption as given air mass with leaner than stoich mixture produces much less combustion heat - energy than same air mass with richer than stoich mixtures.
And I'm wondering what causes this mechanism. When we work with model of constant water to fuel ratio, rich mixture contains both more fuel and water and vica versa, so what causes, that really lean mixtures are able to operate at higher power level?

Under influence of latest research papers about engine knock, what really is knock and what part of mixture is most prone to spontaneous ignition, I would say that its vital to have smallest possible amount of partially burned fuel in combustion chamber. And I can imagine that too rich mixtures are recycling much more of these unstable combustion species than lean mixtures. But that's just my theory


Probably yes, but no one should forget, that both air-fuel ratio and water-fuel ratio determines knock limited power and looking at most of these charts, you can see that fuel only allows highest levels of knock limited power at very rich mixtures, 10.5 AFR and bellow, and any leaning is causing rapid drop in knock limited IMEP.

I really think that this point is moot, as both leaning of AFR and reducing of water flow rate compromises possible gains of water injection to suit for lack of ignition power. This is what Peter talked about, tuning engine at parameters that we want, not at what weak ignition allows us. All hail to Magneto

 

so just to back up a step (and or get the math right!) 50% water to fuel ratio would mean if i ran a 1000cc injector for fuel i should have a 500cc injector for water?

 

Well, I know that these test were comparing fluids on mass basis. So 1000cc/min. of gasoline would be about 720 grams. 720*0.5=360 grams of water - 360cc/min.

Bottom line is, that most people have really weak ignition systems, it may fire modest amount of air-fuel-water mix, but it doesn't have enough energy to initiate strong flame front so whole combustion process is delayed and efficiency and power drops.

I've done bit of reading and it seems that water doesn't slow down burn rate once the strong flame kernel is created, it lies in that initial phase, when large heat capacity of all these fluids absorbs heat from the space between spark plug electrodes, and delays or prohibits creation of plasma channel.

Why the people spend thousands on unnecessary engine modifications but aren't willing to step up with vital ignition power is beyond me

 

after playing with the P port i agree, just stock + an MSD and i'm happily running mixtures that are much richer than would be possible on the stock ignition on a stock motor!

which means i can run the mixture the engine actually wants, vs what the ignition will fire....

 

I'm just going to quote a couple posts I made in 2009 in another thread. It's been a few years since I did the research so it's not as fresh in my mind, but the graphs should be helpful to the discussion. They were modified to make them easier to read. one thing that's interesting is the amount of power they made with 70/30 methanol with an AFR of 25:1. I am guessing the methanol was able to make additional power since methanol is a fuel source and at those levels the amount of additional fuel they added was tiny or non existent. So for water and no AI the engine essentially had no fuel to burn. That's my take on it anyway. Another consideration is the quality of ignition systems they had at the time? The one thing that can be taken away from these WII studies is that AUX injection system definitely works.

 

Originally Posted by Dudemaaanownsanrx7
Not to take anything away from alcohol, it does have significant cooling abilities to the air temps, but those planes would be flying so high that keeping the air temps down probably wasn't too much of an issue.

Brent, you are using 25 psi on water but their conservative limit for water is 20 psi.
At first the alcohol was added for its antifreeze qualities but when tested it was found to allow a conservative 30 psi boost.

Notice that these conservative limits closely correspond to the commonly held opinions (and experience)
for the limits of: rich fuel (15psi), water (20psi), and water/meth (30psi).

Notice too on your colored chart how much of an increase the green highlighted Meth/water has over water.

Barry

 

If you are wondering why some people seem to be able to run higher boost than these
published papers this is how it can be done.

The researches for these papers had different goals than we do.
Their idea was to carry the least weight while making the most dependable power.
We are less interested in fuel/liquid consumption or weight of said liquids.
We just want max safe power.


The maximums are usually stated in IMEP (indicated mean effective pressure)
and peak combustion pressure. All we need to do is to lower the peak
pressure, when it is heading toward detonation.

Things to reduce the possibility of detonation:
Lower the intake air temp.
Increase the octane.
Stabilize the burn (water).
or.... we simply retard the spark.


Footnotes-
1- the more retarded we run the higher the EGT.
2- maximize your IMEP for your individual requirements.

This is what it looks like as you optimize your ignition event.
The peak moves closer to 45 degrees.
Once we maximize the the IMEP we increase the boost.
If the peak pressure gets too high we then start to retard the spark.

Timing builds the leading edge of the curve as shown.
Boost will move the whole curvature upward.
Notice that the TDC pressure is about 300 psi. This would relate to about 12 psi boost.

 

I use the KISS method. Keep It Simple Stupid. I let the boost make the power.

As far as my ignition timing, I don't tune using a dyno most of the time. I tune on the street so optimal timing for power can't really be achieved as easily. Since there really are only a hand full of guys running high boost there's not a lot of maps to compare to. I basically use the 15* at 15 psi rule and then remove 1 degree per psi added. Which is a pretty common tuning technique for rotary engines. I also increase the timing a bit in the upper RPMs. My timing at 26 psi is around 6 degrees under peak torque (6.5k) increasing to 7 after 7k. I couldn't tell you if the timing is conservative or not. To me it seems about right. It makes gobs of power with no knock and that's all my goal ever was. I think my AFR is leaner than what rice or BDC ran, I usually tune for high 10's under full boost and I wouldn't be concerned if I saw 11:1 at 26 psi. I think they usually stuck to low 10's which personally I didn't see a reason to be that rich. I think tuning for 12's would be possible and make a lot more power, but for now I keep it where its working.

I do believe a 50/50 mix is ideal if your after maximum HP, but you will not see the major power differences like in the charts for the WW2 planes. I think most our engines with proper tuning would do just fine at 18 psi without AI maybe a bit on the edge but not extreme, and the addition of either water or methanol done properly 30-35 psi being more of an upper limit. I think with both water or methanol 25 psi is pretty conservative, again done properly and provided nothing goes wrong. I don't have any charts that show why I believe this, but I'm used to seeing it every time I get in my car so I have confidence from the experience of it. I do tend to find my own way of doing things so who knows really the best way, or what the limits are. They seem to get pushed further and further all the time. I've reached the level where the car scares me everytime I reach full boost, so I have no real reason to go for more power anytime soon.. The Power to weight ratio is close to my R1.

 

Brent I use the KISS method also.

But I also use the WAG (wild *** guess) and

my favorite, the SWAG method...... wait for it.....

......(scientific wild *** guess)

Sorry for the off-color work jokes,
Barry

 

Lol I like that

 

+1.

 

We were screwing around, trying to figure out what worked on the 100% methanol side. For me in particular, I was trying to see what varying (usually increasing, ergo the richer AFR's) amounts of alcohol injection would do with EGT's and Air temperatures for my hot-air setup.

I've re-thought my position that I've held (which was the same as your current one is) concerning methanol only injection and 25psi being conservative. I'd dial it back down and say 18-20psi is. There's just something odd going on in this engine once boost starts to hid low to mid 20's and beyond. It's some weird, odd, intermittent, out-of-nowhere-just-when-you-think-it-works thing that seems inexplicable. I've talked with Arghx and Howard several times about it and the theory is it has something to do with the odd chamber shape and ergo the way combustion occurs in this engine (somewhat rectangular shape, very long stroke, tendency of end gasses to clump up on the trailing edge of the rotor where all the knock seems to occur). The documents we've read indicates some sort of 'surface ignition temperature' that seems to be irrespective of auto-ignition temperature and is something that so many folks have fun into using straight methanol. I wonder if it's just something with alcohol use alone. I don't know.

B

 

Lol sounds like everyone is flip flopping their positions. Rice running alcohol mix and Bdc theorizing with water what has the world come to? I was never an advocate for either side, I remain open minded to whatever works, but I also don't mind playing devils advocate to see what could be on the other side of the coin. The information and research done in the ww2 era is helpful and encouraging, but it's not the tell all when it comes to rotary engines or modern technology with ignition and fuel delivery systems. I run water on my setup but I have encouraged customers to give 50/50 a wirl. I would like to experiment more like I used to, but I have too many jobs and hobbies as it is, plus a 2 year old who is going nuts as a type. Lol