17 replies to this topic

[GSX-R]

Does water injection in current gasolines engines can increase their thermal efficiency ?
Volumetric ratio changes and separate fuel & water injection (indirect & direct) are permitted.

[Greg Locock]

It depends. The answer is demonstrably yes, in the right circumstances. One of the things I'll never get around to is to fit a misting system to the Corona, so that I can run advanced and lean without burning the exhaust valve on the freeway.

That's fine for me with no emissions to worry about.

[GSX-R]

Very interesting topic.

1) anti knocking vs thermal protection of the engine : Theorically, if we can imagine a massive production, do you think it would be more interesting to increase the volumetric ratio or to burn lean, or even a mix of the both ?

I admit i haven't tried to make calculations to know what could be ideal from the thermodynamic point of view, for the moment.

2) Water steam pressure effect at power stroke and/vs charge cooling effect : better to inject water (directly) just before ingnition (thus the steam pressure would not act against the compression stroke, but also less volumetric efficiency) or at inlet (opposites remarks)

4) Do you know the adiabatic index of steam ? I guess it should be less than 1.4 (due to more degrees of freedom for H2O).

---

Pratically, it's an evidence to say it's easier to burn leaner for an DIY testing.

4) What could be the exhaust valve temperature if you run (for the moment without water injection) for instance at AFR : 18


P.S : a university guy claims to get 2.7% of thermal efficiency in a LPG engine. But we don't know if he changed the AFR or he made no other thing that injecting water...
http://www.sciencedi...4bb305c9c9c4c09

[LandOfSnow]

Originally posted by GSX-R

4) Do you know the adiabatic index of steam ? I guess it should be less than 1.4 (due to more degrees of freedom for H2O).

It's 1.327

[GSX-R]

I guess you're right. I've found 1.31. So we can neglict this in the mixture.

P.S : strange on the Wiki H20 page i've reported :
for H2O gas
Cp= 1850 J/(kg·K)
Cv= 3724 J/(kg.K) :

[J. Edlund]

J.A.Harrington, "Water Addition to Gasoline - Effect on Combustion, Emissions, Performance, and Knock", SAE Technical Paper 820314

"With respect to gasoline, water-gasoline mixtures have slower burn rates, with both ignition delay periods and combustion intervals being longer. Minimum advance for best torque (MBT) spark timings are advanced relative to gasoline values. Knock limited spark advance is extended with water addition. Power at constant equivalence ratio and engine airflow is decreased while fuel and energy consumption are increased. Water has virtually no effect on CO concentration in exhaust gases and has a slight effect on HC concentrations. A small decrease in HC concentration is observed at water levels below the saturation limit of air but an increase occurs at higher levels. The concentration of NO is decreased with added water. Lean limits of operation are compromised."

Originally posted by GSX-R
2) Water steam pressure effect at power stroke and/vs charge cooling effect : better to inject water (directly) just before ingnition (thus the steam pressure would not act against the compression stroke, but also less volumetric efficiency) or at inlet (opposites remarks)

Since water will provide cooling during both compression and combustion I would expect water to reduce the in cylinder pressure. This would explain the powerloss described above. Since combustion temperatures are reduced, the carnot cycle will explain the loss in efficiency.

Water which vaporise before the inlet valve has closed would most likely decrease volumetric efficiency. Yes, water do provide cooling which makes the air denser but water vapor will also displace air which reduce the volumetric efficiency. Similar results have been observed with port injection of fuel.

[Greg Locock]

Sinc I'm concerned with part throttle fuel efficiency, I really don't care about the reduced power or so on. I wonder what the offset is between increased pumping losses, and reduced throttle work, at part throttle?

[J. Edlund]

Originally posted by Greg Locock
Sinc I'm concerned with part throttle fuel efficiency, I really don't care about the reduced power or so on. I wonder what the offset is between increased pumping losses, and reduced throttle work, at part throttle?

Since there aren't any knock issues at part load I can't see any reason to use water injection during that condition.

Both power and efficiency decrease at full load, and it's most likely going to have the same effect at part load.

[GSX-R]

Forgot one instant the knock issues. Back to thermodynamics and the piston engine :

Originally posted by J. Edlund
Since combustion temperatures are reduced, the carnot cycle will explain the loss in efficiency.
...

You're partially right for the peak temp. But since there's still enough heat in the exhaust pipe to vaporize water thus basically the cycle can still be improved using water/steam. But is it possible inside the classical 4 strokes piston engine cycle ?

(Even if this engine doesn't provide better BSFC, that's the idea in the 6 strokes engine for instance : http://www.autoweek..../THISWEEKSISSUE )

But is there any suitable moment to inject water to increase BSFC in a 4 strokes engine ? that's one of the interesing points. It seems the report you talked about was "just" about a mixture of fuel and water at the same time and location. I will try to read it if it's a free access.

I don't think that just cooling effect of water vaporization by itself can directly increase the thermal efficiency for thermodynamics reasons. The only way for water to help for BSFC in a pure thermodynamic way could be to translate residual heat into pressure.

Can the water indirectly (prevent knock or allow lean mixture) increase the BSFC, that's another question.

Sure you'll easily have the both phenomenons melted on a 4 stroke engines. That is, amha what makes the global question difficult to answer for the moment, at a first overview.

[Greg Locock]

JE - Don't understand. When we run closed loop spark it is knock limited at cruise. I think?

[Powersteer]

Originally posted by J. Edlund


Since there aren't any knock issues at part load I can't see any reason to use water injection during that condition.

Both power and efficiency decrease at full load, and it's most likely going to have the same effect at part load.

Same reason why overboost is a beautiful thing at mid rp/m range.

[J. Edlund]

Originally posted by GSX-R
Forgot one instant the knock issues. Back to thermodynamics and the piston engine :



You're partially right for the peak temp. But since there's still enough heat in the exhaust pipe to vaporize water thus basically the cycle can still be improved using water/steam. But is it possible inside the classical 4 strokes piston engine cycle ?

(Even if this engine doesn't provide better BSFC, that's the idea in the 6 strokes engine for instance : http://www.autoweek..../THISWEEKSISSUE )

But is there any suitable moment to inject water to increase BSFC in a 4 strokes engine ? that's one of the interesing points. It seems the report you talked about was "just" about a mixture of fuel and water at the same time and location. I will try to read it if it's a free access.

I don't think that just cooling effect of water vaporization by itself can directly increase the thermal efficiency for thermodynamics reasons. The only way for water to help for BSFC in a pure thermodynamic way could be to translate residual heat into pressure.

Can the water indirectly (prevent knock or allow lean mixture) increase the BSFC, that's another question.

Sure you'll easily have the both phenomenons melted on a 4 stroke engines. That is, amha what makes the global question difficult to answer for the moment, at a first overview.

Did a few simulations where I tested to add water to the incoming charge and the following was noticed:
1. Power and torque output decreased
2. Cylinder pressures decreased, especially peak pressures
3. Brake specific fuel consumption increased

So basically, when you inject water into an engine, water displaces air which reduce the volumetric efficiency, also, large amounts of energy are spent on vaporising water which will not be recoveded which results in a large heat loss.

If you want to recover lost energy by the use of water the best solution would be to use a heat exchanger, exhaust/water which powers some kind of turbine as found on some compound engines. On the other hand, if you want to prevent knocking and cool critical engine components water injection is a good solution. For example Saab did some experiments with the latter on a modern turbo engine in order to prevent the usually required fuel enrichment at high loads. But I think they stopped with thoese experiments, it caused some kind of trouble.
http://www.aquamist....ress/press.html

[GSX-R]

Originally posted by J. Edlund

Did a few simulations where I tested to add water to the incoming charge

Interesting. Can we know which software you used ? Wave ?

Can you simulate water injection with an increased volumetric ratio ? (normally water cooling effect should permit you to do that). negative big latent vaporization energy of water vs positive bigger compression ratio efficiency duel. I am rather pessimistic as for this antagonism. :

Is your simulation software able to simulate direct water injection beetween [just before TDC at the end of compression] and [90° after TDC] rather than inside the inlet port ?
If we inject water at the end of the combustion stroke, one can reasonably think it could protect the exhaust valve and the cat but also convert residual heat into pressure, that at least a turbocharger turbine could exploit

[McGuire]

Originally posted by GSX-R
Does water injection in current gasolines engines can increase their thermal efficiency ?

No. Not unless there is something wrong with the engine to begin with that water injection can mask. At the end of the day, no matter how you do it you are using gasoline to heat water instead of to turn the wheels.

[McGuire]

Originally posted by Greg Locock
JE - Don't understand. When we run closed loop spark it is knock limited at cruise. I think?

Every engine has a point of incipient knock but water injection does not raise or extend it, only keeps the cylinders at/under it.

[GSX-R]

:yawn:

[Christiaan]

About 10 years ago I saw an after market water injection system for the legendary Escort Cosworth. Did any such thing ever make into a production car? If so, how often would one need to top up the water?

[McGuire]

Only one I know of in a production car is 1962-63 Olds Jetfire. Used a methanol/water mix sold at Olds dealers as "Turbo-Rocket Fluid."

The system typically uses a throttle or vacuum switch of some kind, so how much fluid is consumed depends mainly on how much OT you use. If you just toddle around the bottle will last forever. In GM lore, the story goes that Jetfire owners didn't check their bottles often enough.

Water injection is the simplest thing on earth to rig up. All you need are a switch, a bottle, and a windshield washer pump. ..and a check valve in the line so manifold vacuum doesn't suck all the fluid out of the bottle at idle.