David_Wallis

shock horror mark posts a pic of part of his engine

David

Pavlo

I think I know why your engine isn't running Mark, you have no battery!

On a serious note, charge temperature does have a direct infulence on knock sensitivity, certainly as far as pre-ignition goes.

Inlet temp will reduce cylinder temps, which is what causes bad knock. Any reduction in inlet temp will be carried through to the point of combustion, what happens after depends on more things again.

Basic calc ignoring effects of heating water and the partial pressure of the water vapour:

Latent heat of water is 2500 J/g
Specific heat of air is 1.005 J/g
Assume ERL 138cc/min nozzle
Assume 2 litre engine at 1.5bar and 85% VE running at 6000rpm

Air consumption per min:
2000/2 *2.5(absolute pressure) *0.85 *6000 = 12750000 cc/min = 12.75m^3/min

Mass flow rate:
12.75 * 1.024kg/m^3 = 13kg/minute = 216g/s

Energy for evapouration:
2500 * 138 = 345000 J/min = 5.75Kw

Reduction in temperature of air
5750/1.005/216=26.49 kelvin

So at full chat, a WI system using 1 138cc/min nozzle could reduce the inlet charge by 26.5 degrees, which results in a 9% reduction in inlet pressure. If you were running at 1.5bar, that means you could get the same power for 1.365 bar at the manifold.

So a reduction in inlet temperature and inlet pressure, seems like a good result, and both would serve to reduce knock.

Paul

David_Wallis

paul... mark uses an optima yellow top battery to give him more room so he could have the pumps there..

David

Adam M

david,

mark "uses" is a matter of debate! "used" is perhaps better, as we are both without engine, but I am also without battery thanks to his light fingers!

Mark had a battery in there underneath the elaborate water injection and it was working fine long before he went for MY optima. [img]images/smilies/mad.gif[/img]

David_Wallis

lol...

I noticed it as I used to use them when I was into car audio in a big way...

how long before you are both back with engines?

Adam M

paul, you are missing a /kelvin other wise units dont add up.


There is also something puzzling me that I am hoping you can explain.

Your calculations are based on the assumption that only the heat in the charge prior to combustion provides the total energy for evaporation of all the water.

I.e. all the water has been turned to steam before the fuel is ignited.

I am almost certain that this isnt the case. I dont think you can simply ignore partial pressure, as the main purpose of water injection is to reduce likelihood of det by altering flame propogation speed, and this is reliant on there being a non homogenous charge/steam steam mixture at the point of combustion (according to pat anyway).

It is fair to say that there is no way all (if any) of the water will have evaporated prior to arriving in the cylinder.
We should assume that the bulk of the water is going to evaporate in the cylinder.

At this point you have compression forcing the temperature to rise, fighting against pressure increase which will prevent the water from evaporating. Whether the water is steam or not exactly prior to combustion I dont know, and I am not going to bother to calculate it, but either way, a temperature reduction figure of 27 kelvin is misleading to those who dont know what is going on.

The mechanism by which water injection works has almost nothing in common with that of a front mounted intercooler.

The discussion is academic anyway, as there is no question that water injection does what it is supposed to do, I just disagree with the blanket expression people choose to use of reducing inlet charge temp to reduce det as that is not what it is about.

It also confuses people into thinking that if they have rediculously low charge temps due to a great intercooler (itself not a good thing as atomisation means combustion would be worse) then they also believe that they can gain nothing from using water injection which again simply isnt true.




[Edited by Adam M - 10/15/2002 12:34:00 PM]

R19KET

Adam,

Salvage mate, salvage

With the set up in the pic', I actually used a lower battery, form Halfords, and then used the SS plate the pumps are mounted onto, as the "clamp", to hold it all down.

Worked very well.

Mark.

MorayMackenzie

Pavlo,

Your calculations seem to avoid dealing with amount of fuel/air mix that is displaced by the water. You reason that you will get X power from Y less boost because of an inlet temp drop of Z... you also have to factor in that the water displaces fuel and air, which ultimately means less power, assuming perfect mixes are maintained.

Moray

Pavlo

Yes moray, I think I was quite specific on ignoring the partial pressure of the water vapour.

You're quite correct about the units, strictly speaing the specific heats are per Kelvin, but I do tend to assume such things, rather lazy.

To say that water injection works by affecting the fuel and slowing the flame speed is rather crude. If you were to get incomplete evapouration this would indeed be the case, in effect you contaminate the fuel. However, if this were the main means by which it worked, you wouldn't really get much of a power increase, due to reducing the Calorific Value of the fuel. The same effect could be had my mixing water with the fuel in the tank.

As Adam points out, you possibly won't get complete evapouration (especially if you're using an FMIC) prior to the inlet charge entering the cylinder. However, as he correctly says, simply compressing the inlet charge will increase it's temperature and pressure, thereby increasing the dewpoint, allowing further evapouration in cylinder. In reality you're not going to get H20 condensing in the combustion chamber (will in the exhaust however).

It depends on the humidity of the air before WI, ambient temp, post IC temp. You will get about another 20 degree drop in inlet charge temp using petrol (over 100 using methanol), due to the lower boiling point, it's sensible to assume complete evapouration upon entring the cylinder, prior to compression. I would suspect with the older style slant IC you could inject prior to the IC and get complete evapouration.

I'll work out the effects of the partial pressure of the water and fuel. Giving it further thought, the effects could most certainly be measurable if you got complete evapouration prior to the inlet valves closing. So I suppose that aiming to get complete evapouration after the inlet valves close, but before the ignition begins would be the best all round. You would get the reduction in inlet charge temp, but with negligible partial pressure effects because the H20 would have been mostly water (ie not steam) in the inlet manifold.

My main aim was just to put some numbers to the concept that people could appreciate.

Paul

Adam M

paul, one thing still doesnt make sense.

from what you are saying, adding water injection out the box and not changign anything will give you a power increase as a result the lower temperature air being more dense and hence containing more oxygen.

in the same way that using water spray or running on a cold day gives you more power without any map changes as it forces the car to run a little leaner.

This is of course assuming you are running an aftermarket ecu which doesnt automatically advance the timing.

the thing is, water injection costs you about 15bhp without map changes, whereas water spray or a fmic do not.

It is only after advancing the ignition that you get back the power you have lost, but if water injection was working the way you say it does then this surely would not be the case.



if it works the way you say it does, then can you explain why it costs you power without adjusting the map whereas other charge cooling methods increase your power (until knock sets in) without adjusting the map.

MorayMackenzie

Pavlo,

Re "
quote:Basic calc ignoring effects of heating water and the partial pressure of the water vapour.

Yes moray, I think I was quite specific on ignoring the partial pressure of the water vapour.
"

You have got yourself mixed up somewhere. I asked nothing about partial pressure of water, possibly that was Adam? I asked about the fuel and air mix that is displaced by the water injected into the cylinder. I believe that my question is valid and still stands.

Cheers,

Moray ( For the sake of clarity: That's M O R A Y )

[Edited by MorayMackenzie - 10/17/2002 4:31:28 PM]

Pavlo

Adam, m00r4y,

I didn't work out the partial pressure effect of the water, was doing normal things. But I see what you are both saying.

If water turned to steam it would take up a lot of volume (partial pressure effect) so I'm guessing the effects are larger than I thought. I'll work it out, but I suppose it's not too much to expect from the steam to cancel gains in increased density. Obviously if you get the same sort of gains from spraying the intercooler, there will be no penality, which fits in with what Adam says.

More advance and higher boost without detonation is the ultimate aim of WI. I will do the numbers as promised, because it would suggest there is a point at which WI loses power, even if you don't get water drolets mixing with the fuel.

Paul

carlos_hiraoka

Does the use of 50% water / 50% methanol injection cause any bad effects on the engine internals ?????
This is commonly used on some older grp. A cars.

Carlos H.

Adam M

too much water and you will lose power no question, but my point is that initially if you dont advance the ignition but do inject water, although you will be less likely to det you will lose a significant amount of power.

This is how I know that water injection is not about charge cooling, as if it was power would go up due to leaner running.

R19KET

Adam,

"a significant amount of power", is a bit misleading.

If the right size jet is used, power loss will be minimal, circa 5~10bhp on the average Scoob.

However, there will be occasions, where you will probably gain more power, than you will lose, or to be more accurate, "re-claim power that has been lost" due to high intake temps, caused through tuning method, or high ambients.

One of the biggest problems, is that some people think that the bigger the jet, the better it will work. Not so. Too much water will certainly kill power, even if the ECU is re-mapped.

Mark.

Adam M

ok a max of 15 bhp.

john banks

I have been reading Aquamist's website and product information with some interest recently My ignition is more retarded than I want it and the FMIC has not given me back as much timing as I expected. Probably because I just turned the boost up to take advantage of the lower charge temps. I would imagine that WI would make more difference to timing than a P20 housing - agree?

Disappointing to hear about the lack of promise with self mapping. Continuous self learning in daily use through a wideband would be nice?

[Edited by john banks - 10/12/2002 9:40:43 PM]