100 replies to this topic

[Wuzak]

Does water injection improve the efficiency of turbocharged petrol engines?

 

Or does it simply increase the density of the intake air, allowing for more fuel to be added to give more power?

[2ms]

A cooler charge permits higher compression ratio which means higher efficiency.

[gruntguru]

Detonation supression permits higher CR which means higher efficiency or higher boost which means more power. Some of the detonation supression comes from the cooler charge but mostly because water is an anti-detonant.

[bigleagueslider]

As gruntguru noted, with boosted SI engines the water injection normally serves to supress detonation.  When injected into the intake airflow that has become heated due to compression work, the water vaporization provides cooling due to the latent heat effect.  While the cooler air charge is more dense, you must also remember that the water vapor is basically inert and does not contribute to the chemical combustion reaction between the fuel and oxygen.  The inert water vapor will absorb some of the heat released by combustion similar to how other inert gases present in the intake charge do (such as CO2 or N2), and then it will transfer much of that energy back to the piston during the expansion phase of the engine cycle.  But while injecting water into the hot charge air ahead of the intake valves can increase the air density, the water mass injected also displaces some O2.  And with an SI engine you cannot produce more power by adding more fuel unless you also add sufficient free oxygen.

 

Engine design is basically an excercise in compromise.  In general, the thermal efficiency of an internal combustion engine will improve with increased cycle pressure ratios and temperatures.  In practice, there are trade-offs that must be considered in terms of complexity, reliability, cost, etc.  So the answer to your question as to whether water injection improves engine efficiency is .......maybe.

[Magoo]

If you do a tech forum search on the topic using the terms water injection, efficiency, thermal efficiency, methanol, etc. a couple of worthwhile discussions will pop up. 

 

For me, water injection can only increase efficiency if the engine's efficiency is in some way wounded and the H2O can serve as a bandaid. Any way you look at it, you are using gasoline to heat water. 

[kikiturbo2]

 Any way you look at it, you are using gasoline to heat water. 

 

and in a "old school" turbo engine you are using gasoline to cool air..

[bigleagueslider]

For me, water injection can only increase efficiency if the engine's efficiency is in some way wounded and the H2O can serve as a bandaid. Any way you look at it, you are using gasoline to heat water. 

 

I suppose you could say the same thing about using water for liquid cooling of the engine.  Since substantial amounts of the heat energy released from combustion of the gasoline fuel simply ends up as heat in the water coolant, would water cooling of the engine also qualify as a "bandaid"?

[saudoso]

http://en.wikipedia.org/wiki/Elsbett

[indigoid]

I suppose you could say the same thing about using water for liquid cooling of the engine.  Since substantial amounts of the heat energy released from combustion of the gasoline fuel simply ends up as heat in the water coolant, would water cooling of the engine also qualify as a "bandaid"?

 

 

Pretty much every time a manufacturer ditches their "traditional" aircooled designs there is much wailing and gnashing of teeth (see: BMW, Porsche) and then people get over it, upgrade and quietly hope their fellow enthusiasts will forget what they said about never, ever, ever buying a water-cooled 911. Amusing/sad

[Magoo]

I suppose you could say the same thing about using water for liquid cooling of the engine.  Since substantial amounts of the heat energy released from combustion of the gasoline fuel simply ends up as heat in the water coolant, would water cooling of the engine also qualify as a "bandaid"?

 

 

I think we lost the plot. While we would certainly agree that water cooling is a needful and useful feature, it would be quite a feat of pretzel logic to hold that it increases efficiency, since it's a major source of operating loss. Increases efficiency? Compared to what?  And that is the proposition here: that H2O injection increases efficiency. 

 

Water injection can certainly increase output, but it cannot increase total efficiency. These are two different things. We cannot assume that increases in efficiency result from increases in output. It is generally safer to assume the reverse.  

 

Now here is the crucial difference between water cooling and water injection: With the latter, H2O is introduced into the fuel/air charge where it both occupies volume and absorbs heat, but it has no caloric value. 

Edited by Magoo, 03 September 2013 - 02:10.

[gruntguru]

http://www.autospeed...13/article.html

This article summarises some of the research published on the issue. Important points to note are:

 

- The water needs to be in vapor form during combustion to gain the full anti-detonant benefit. This is one reason that WI is more prevalent/beneficiial in supercharged engines. (The more obvious being - SC engine more likely to be knock limited, SC engine can gain an easy power increase with WI via increasing boost, intercooling effect, etc)

- On a NA engine the efficiency benefit could be as high as 2% (eg increase from 30% to 30.6%). The benefit would be more noticeable with low octane fuel.

[Wuzak]

I think we lost the plot. While we would certainly agree that water cooling is a needful and useful feature, it would be quite a feat of pretzel logic to hold that it increases efficiency, since it's a major source of operating loss. Increases efficiency? Compared to what?  And that is the proposition here: that H2O injection increases efficiency. 

 

I guess it would be compared to air cooling. 

 

I would think that water cooling would allow for tighter clearances and higher compression ratios, reducing losses and improving efficiency  - when comapred to air cooled motors.

[Kelpiecross]

Does water injection improve the efficiency of turbocharged petrol engines?

 

Or does it simply increase the density of the intake air, allowing for more fuel to be added to give more power?

 

  Another thing to consider with water injection and its effect on thermal efficiency is whether converting the water injected into high pressure steam would increase the TE.  For example if an engine has an exhaust gas temperature of 700 degrees C (or thereabouts) why couldn't some of this waste heat be converted to extra cylinder pressure by injecting water into the cylinder?  To answer my own question - this is an old idea and it seems that there is no particular gain in power in doing this.  Presumably the cooling effect on the combustion gases is pretty much balanced by the added steam pressure so overall the effective cylinder pressure stays about the same. 

 

  However if the hot exhaust gases are used after leaving the cylinder to produce high pressure steam then quite substantial gains in power (and TE) can be made.   Examples of such systems are the Still engine of the 1930's and the BMW system of recent years (and possibly Bruce Crower's 6-stroke engine also of recent years). 

 

  Even though logic would indicate that there is plenty of energy to be recovered from the exhaust gases, overall it would seem that with any method of using waste exhaust heat to produce steam the added complexity of the system is not worth any possible small gains in power/efficiency. 

 

 As a point of interest - the temperature of a SI engine's exhaust gases is about three times the temperature of the typical "steam age"  reciprocating  steam engine.    

[Kelpiecross]

I think we lost the plot. While we would certainly agree that water cooling is a needful and useful feature, it would be quite a feat of pretzel logic to hold that it increases efficiency, since it's a major source of operating loss. Increases efficiency? Compared to what?  And that is the proposition here: that H2O injection increases efficiency. 

 

Water injection can certainly increase output, but it cannot increase total efficiency. These are two different things. We cannot assume that increases in efficiency result from increases in output. It is generally safer to assume the reverse.  

 

Now here is the crucial difference between water cooling and water injection: With the latter, H2O is introduced into the fuel/air charge where it both occupies volume and absorbs heat, but it has no caloric value. 

 

 

 Increases in output for the same  amount of fuel consumed in the same engine with the same CR etc.  would indicate an increase in TE.   Whether or not the efficiency is increased (or decreased)  could probably be determined by careful dyno testing of the same engine with and without the  water injection operating.    I would not like to predict whether the TE would increase or decrease slightly - it could be either.         

 

 

 

 

[gruntguru]

 Even though logic would indicate that there is plenty of energy to be recovered from the exhaust gases, overall it would seem that with any method of using waste exhaust heat to produce steam the added complexity of the system is not worth any possible small gains in power/efficiency. 

 

 As a point of interest - the temperature of a SI engine's exhaust gases is about three times the temperature of the typical "steam age"  reciprocating  steam engine.    

Using the old rule of thumb 1/3 fuel energy heats the cooling system, 1/3 exhaust heat and 1/3 useful crankshaft work. At best a low temperature steam cycle would recover 10% of exhaust waste heat ie 3.33% of fuel energy raising TE from 33.33% to 36.66%. A good exhaust turbine would do the same job.

 

Large scale combined cycle systems recover a lot more from NG gas turbines, raising the TE from about 40% to 60%. The key is "Large Scale", higher exhaust temperatures and far greater heat rejection to the exhaust (no cooling system).

[Magoo]

http://www.autospeed...13/article.html

This article summarises some of the research published on the issue. Important points to note are:

 

- The water needs to be in vapor form during combustion to gain the full anti-detonant benefit. This is one reason that WI is more prevalent/beneficiial in supercharged engines. (The more obvious being - SC engine more likely to be knock limited, SC engine can gain an easy power increase with WI via increasing boost, intercooling effect, etc)

- On a NA engine the efficiency benefit could be as high as 2% (eg increase from 30% to 30.6%). The benefit would be more noticeable with low octane fuel.

 

The article you cite does not support your claim. 

 

 

 

If water injection truly increased efficiency in standard IC engines, we would see it in production cars. But we don't because it doesn't. All SI gasoline engines are ultimately knock limited, boosted or not. Increases in efficiency are expensive but H2O is free. Water injection is easy and cheap to execute and simple to failsafe. So where are all the water injection systems? 

[gruntguru]

 

 

 Increases in output for the same  amount of fuel consumed in the same engine with the same CR etc.  would indicate an increase in TE.   Whether or not the efficiency is increased (or decreased)  could probably be determined by careful dyno testing of the same engine with and without the  water injection operating.    I would not like to predict whether the TE would increase or decrease slightly - it could be either.         

 

It is very unlikely you would see any TE improvement by "switching on" the water. You would need to raise the CR.

[gruntguru]

If water injection truly increased efficiency in standard IC engines, we would see it in production cars. But we don't because it doesn't. All SI gasoline engines are ultimately knock limited, boosted or not. Increases in efficiency are expensive but H2O is free. Water injection is easy and cheap to execute and simple to failsafe. So where are all the water injection systems? 

There are many reasons we don't see water injection in production cars.

1. CR is already high in modern engines and TE benefits diminish rapidly above 10:1 (ie if fuel quality and engine design and control were not as good as they already are, the benefits of WE might be sufficient to justify the downsides.)

2. Vaporising all the water before combustion is difficult in a NA engine.

3. TE increase is restricted to certain (knock limited) operating modes.

4. User resistance (is it worth regularly topping this tank up for a measly 1% fuel saving?)

5. Initial cost of an accurate, finely atomising system would not be insignificant.

[MatsNorway]

I am surprised it is not more common on pikes peak and such.

 

Those who has it seems to go like mad.

 

http://www.youtube.c...ZJX8NuNzW0#t=14

[TDIMeister]

From a simplistic, fundamental thermodynamic point-of-view, water injection should reduce the thermal efficiency because:

1) the specific heat capacity of the working fluid increases > cyclic peak gas temperatures decline > lower second-law efficiency;

2) the specific heat ratio (Cp/Cv, a.k.a. isentropic exponent, polytropic coefficient, n, k, kappa, gamma, etc.) declines with increasing water in the working fluid (see Fig. 3-2 in Heywood, 1988), thus reducing the thermal efficiency from its definition for an air-standard Otto cycle that eta=1-1/(r_c^(k-1)).

 

On the other hand:

a) to 1) above, lower charge inlet temperatures will reduce the minimum cycle temperature, which also increases the cycle efficiency as will an increase in peak temperature, eta=1-T_L/T_H;

b) lower peak gas temperatures could reduce wall heat transfer losses (eta rises as a consequence);

c) anything that inhibits detonation and therefore inhibits ignition timing ****** and/or fuel enrichment from what would otherwise occur will directly improve the thermal efficiency/BSFC of the engine.

[gruntguru]

From a simplistic, fundamental thermodynamic point-of-view, water injection should reduce the thermal efficiency because:

1) the specific heat capacity of the working fluid increases > cyclic peak gas temperatures decline > lower second-law efficiency;

2) the specific heat ratio (Cp/Cv, a.k.a. isentropic exponent, polytropic coefficient, n, k, kappa, gamma, etc.) declines with increasing water in the working fluid (see Fig. 3-2 in Heywood, 1988), thus reducing the thermal efficiency from its definition for an air-standard Otto cycle that eta=1-1/(r_c^(k-1)).

 

On the other hand:

a) to 1) above, lower charge inlet temperatures will reduce the minimum cycle temperature, which also increases the cycle efficiency as will an increase in peak temperature, eta=1-T_L/T_H;

b) lower peak gas temperatures could reduce wall heat transfer losses (eta rises as a consequence);

c) anything that inhibits detonation and therefore inhibits ignition timing ****** and/or fuel enrichment from what would otherwise occur will directly improve the thermal efficiency/BSFC of the engine.

Although it is true that in general terms cycle efficiency is maximised by adding heat at the highest possible cycle temperature and rejecting heat at the lowest possible cycle temperature, your expression "eta=1-T_L/T_H" is an over-simplification. Going back to Otto cycle efficiency expressed in terms of cycle temperatures we have eta = 1-(T4-T1)/(T3-T2) so to maximise cycle efficiency we need to maximise T1 and T3 and minimise T4 and T2. Although water injection will reduce T3 (peak temp at end of combustion) it will also reduce T2 (Temp at the end of compression and beginning of combustion).

 

I agree with your statement c) above but would add that an optimised engine should not be operating with ignition ******, however the benefit of WI would be to allow higher CR.

[TDIMeister]

Agreed.  It is an oversimplification; I'll expound.  Nobody will (or should) use air-standard and Carnot efficiencies - the latter is what eta=1-T_L/T_H expresses - to discuss engine efficiencies, although some will try, notably purveyors of some "revolutionary" new discovery or engine design.  The Carnot efficiency sets up the boundary by which nothing operating in a thermodynamic cycle, within the same minimum and maximum temperatures, can exceed.  A typical, contemporary internal combustion engine can have cyclic temperatures that range between a peak flame temperature of about 2700 kelvins to ambient of about 300 K, which means the Carnot - or maximum attainable - efficiency is 88.9%.  That some method, e.g. water injection, can alter those temperatures in any sensible order of magnitude is superfluous in an engine whose efficiencies typically do not exceed 35%.  A change that could reduce the Carnot efficiency could still give the opposite directional effect to the brake thermal efficiency, as can some change in properties (compression ratio or specific heat ratio k) alter the air-standard efficiency figure but have the opposite effect in the real efficiency.

 

My point was to point out that basic thermodynamic theory would argue for one set of outcomes, but greater understanding thereof - its assumptions, simplications and limitations - as well as accounting for other factors, could result in a contrary conclusion.  By far, c) in my first post above has a largest effect, and there I'm not talking about an OEM calibration with optimized ignition timing, compression ratio, etc. for the available fuel type, but rather someone who has thrown on a turbocharger to a naturally aspirated one, or who has turned up the boost from stock, which puts detonation at a much higher likelihood and thereby opting for water injection to the exclusion of other efficiency-killing means like ******** timing, reduced CR and fuel enrichment that would otherwise be necessary.

[bigleagueslider]

I guess it would be compared to air cooling. 

 

I would think that water cooling would allow for tighter clearances and higher compression ratios, reducing losses and improving efficiency  - when comapred to air cooled motors.

Wuzak-

 

The benefit of water cooling is somewhat analogous to that of water injection in terms of providing a net improvement in SI engine thermal efficiency.  Both allow safe operation at higher cycle pressures/temperatures.  Water cooling of a cylinder head is more efficient than air cooling at transferring heat away from the combustion chamber & quench surfaces, thus increasing the engine's knock threshold.

 

On a somewhat related note, here is a press release from Ricardo Labs that describes their "Cryo Power" diesel engine concept.  It injects LN₂ (an inert liquid like H₂O) during compression to control temperature rise and increase mass. They predict thermal efficiency of 60%.  Pretty impressive if it proves practical.

Edited by bigleagueslider, 05 September 2013 - 04:03.

[Kelpiecross]

It is very unlikely you would see any TE improvement by "switching on" the water. You would need to raise the CR.

 

  I should point out that I am not actually advocating injecting water to increase cylinder pressure etc. and TE etc.    I have in the past tried the water injection on/off experiment in a normal non-turbo engine and I found that there is no noticeable difference in the power output with or without water injection - and I had it arranged so I could switch the injection pump on and off while I was driving.   The only noticeable result is that it very quickly buggers the engine oil. 

 

  However there are possible arrangements that may have an effect.   With the basic TE of an being related to the peak combustion temp and the exhaust temp - injected water in the intake charge would lower both the peak and lower temp points thus resulting in probably a reduction in the possible TE.   But what if the water spray were injected at some point after combustion was complete?   In this case the peak temp would still be the same but the exhaust temp would be greatly reduced (depending on the amount of water sprayed in).

 In this case would the possible TE be related still to the peak and lower temps in the cylinder?   Clearly if you considered these two extremes of the temp range by conventional thermodynamic theory the possible TE should be higher.  If it is not - how would you explain this by conventional thermodynamic thinking?

 

  I have not tried this experiment (it would be tricky to test the idea) but I suspect there would still be no improvement in the TE etc.

 

  Does anyone else have a problem with most of the links (like Magoo's link to Ricardo) not working and bringing up an red exclamation mark in a yellow star?       

[gruntguru]

  Does anyone else have a problem with most of the links (like Magoo's link to Ricardo) not working and bringing up an red exclamation mark in a yellow star?       

Yes. Hold the Ctrl key while clicking the link.

[gruntguru]

 However there are possible arrangements that may have an effect.   With the basic TE of an being related to the peak combustion temp and the exhaust temp - injected water in the intake charge would lower both the peak and lower temp points thus resulting in probably a reduction in the possible TE.   But what if the water spray were injected at some point after combustion was complete?   In this case the peak temp would still be the same but the exhaust temp would be greatly reduced (depending on the amount of water sprayed in).

 In this case would the possible TE be related still to the peak and lower temps in the cylinder?   Clearly if you considered these two extremes of the temp range by conventional thermodynamic theory the possible TE should be higher.  If it is not - how would you explain this by conventional thermodynamic thinking?

The low Temp (heat rejection temp of the Otto cycle) is not really relevant. This is because the Otto cycle assumptions include a "Closed" system ie the working fluid stays in the cylinder and heat is added at TDC and removed at BDC. In a real engine, the BDC air temp at the beginning of the compression stroke is fixed by the intake air temp and the BDC temperature at the end of the power stroke is a function of the temperature at the beginning of the power stroke and the expansion ratio (usually equal to the compression ratio).

[malbear]

The low Temp (heat rejection temp of the Otto cycle) is not really relevant. This is because the Otto cycle assumptions include a "Closed" system ie the working fluid stays in the cylinder and heat is added at TDC and removed at BDC. In a real engine, the BDC air temp at the beginning of the compression stroke is fixed by the intake air temp and the BDC temperature at the end of the power stroke is a function of the temperature at the beginning of the power stroke and the expansion ratio (usually equal to the compression ratio).

therefore an engine with the atkins cycle with an expansion ratio larger than the compression ratio will have a lower exhaust temperature?

[Kelpiecross]

therefore an engine with the atkins cycle with an expansion ratio larger than the compression ratio will have a lower exhaust temperature?

 

  Yes - with some Atkinson engines it is very noticeable that the exhaust/exhaust manifold  is much cooler - the exhaust is also quieter.   I have seen one such engine that ran on 9:1 CR and 18:1 ER where at idle the exhaust manifold was cool enough to touch.    

 

  The general Atkinson principle could probably be extended up to an 40:1 (or so) ER where the exhaust would be almost at room temperature.   Such extremes of ER are at the expense of overall power output - 40:1 ER would probably be of novelty interest only.  I have always thought that a stationary or industrial engine running at 9:1/18:1 would be very practical - especially running on the cheaper LPG fuel.   

[Magoo]

 

  Does anyone else have a problem with most of the links (like Magoo's link to Ricardo) not working and bringing up an red exclamation mark in a yellow star?       

 

 

Whoa, that's not my link. That link goes to a press release about an experimental diesel engine that has absolutely nothing to do with the question at hand here. 

[Kelpiecross]

The low Temp (heat rejection temp of the Otto cycle) is not really relevant. This is because the Otto cycle assumptions include a "Closed" system ie the working fluid stays in the cylinder and heat is added at TDC and removed at BDC. In a real engine, the BDC air temp at the beginning of the compression stroke is fixed by the intake air temp and the BDC temperature at the end of the power stroke is a function of the temperature at the beginning of the power stroke and the expansion ratio (usually equal to the compression ratio).

 

 Thank you for the link "trick" - I would never have guessed that. 

 

  I think another way of looking at it is that the combustion gas expands as normal until the water is injected and cools things down.   Injecting water after combustion is complete but still at constant volume (which doesn't usually happen in real life) would be equivalent to injecting water into the intake charge. 

[Kelpiecross]

Whoa, that's not my link. That link goes to a press release about an experimental diesel engine that has absolutely nothing to do with the question at hand here. 

 

  I thought it didn't seem very relevant - but I put that down to my gross stupidity.   

[TDIMeister]

Thermodynamic calculations using equations derived after a long line of assumptions and algebra break down when you introduce anything through the system boundary under consideration into the closed control volume.  The first law derivation for efficiencies of closed engine cycles are always based on the premise of internal energy conservation and additionally that the fluid is an ideal gas (Except Rankine and other phase-change cycles).  From the latter, one applies the property that U=m*C_v*(T₂-T₁) and when one goes through calculating efficiencies of the form eta=W_net/Q_in or eta=(Q_in-Q_out)/Q_in using ideal gas and polytropic relations, many terms like masses and specific heats cancel out to arrive at the headline equation that's usually only a function of the ratios of state temperatures, volumes or pressures.  However, as soon as you introduce something with different properties at any time through the system during the closed part of the cycle, e.g. water, the assumption of constant mass and specific heats no longer hold.

 

One of the biggest transgressors of this was the Crower six-stroke engine.  Its capabilities were hyped from some profound ignorance of basic thermodynamics.  Literature about this engine claimed that the engine harnessed "free" energy from the spontaneous "explosive" expansion of liquid water injected into the cylinder during the hot "re-expansion" stroke after combustion into steam - strike one: there's no such thing as "free" energy; strike two: the expansion of water into steam sucks up a lot of energy - precisely the latent heat of evaporation - from the surrounding gas such that the surrounding gas would cool and contract, with the resulting overall result rather less explosive than claimed.  Also, as "proof" of the engine's incredible and revolutionary efficiency, the purveyors claimed that the exhaust from a running engine was cool enough to touch - strike 3: that's quite an imprecise metric, since I can hold the exhaust manifold of a just started and idling TDI engine for quite a while before it gets too hot to the touch because of the thermal inertia of the manifold and the fact that the exhaust of an idling Diesel is also quite relatively "cool"; strike 4 and the nail in the coffin: we're talking energy here, specifically internal energy or enthalpy where applicable and not temperatures per se. Recalling that U=m*C_v*(T₂-T₁) and H=m*C_p*(T₂-T₁),  If a normal engine rejects its exhaust, say at 1000 K, I can quite easily inject water (increase mass and C_p or C_v) so that my exit temperature is near ambient but in reality the amount of U or H has remained constant so I've really done nothing to the efficiency.

Edited by TDIMeister, 05 September 2013 - 14:30.

[TDIMeister]

  The general Atkinson principle could probably be extended up to an 40:1 (or so) ER where the exhaust would be almost at room temperature.   Such extremes of ER are at the expense of overall power output - 40:1 ER would probably be of novelty interest only.  I have always thought that a stationary or industrial engine running at 9:1/18:1 would be very practical - especially running on the cheaper LPG fuel.   

In theory, you could have an Atkinson cycle reject almost at room temperature  - ignoring that you'd have to expand to a near perfect vacuum... - but long before then you come to several practical limitations.

1. The cylinder pressure will become basically atmospheric while the temperature is still well above this.  This is one of the results of the fact that no engine cycle can be as efficient as Carnot.  If you expand to at or below atmospheric pressure, exhaust scavenging will "suck" because of the lack of blow-down and gas dynamics effects, requiring piston displacement on the upstroke (PdV work, increasing pumping losses) to do all of the scavenging. Once the exhaust valves open with cylinder pressure below atmospheric, you will have a reverse flow back into the cylinder;
2. You get no positive PdV work once the pressure drops below atmospheric but you have atmospheric pressure working against you on the other side of the piston, achieving similar results as very early historic engines that harnessed NOT the expansion force from combustion, but rather the work from the aftermath partial vacuum;
3. You'll need a very complex kinematic mechanism to get an appreciably large difference between CR and ER.

[Magoo]

Thermodynamic calculations using equations derived after a long line of assumptions and algebra break down when you introduce anything through the system boundary under consideration into the closed control volume.  The first law derivation for efficiencies of closed engine cycles are always based on the premise of internal energy conservation and additionally that the fluid is an ideal gas (Except Rankine and other phase-change cycles).  From the latter, one applies the property that U=m*C_v*(T₂-T₁) and when one goes through calculating efficiencies of the form eta=W_net/Q_in or eta=(Q_in-Q_out)/Q_in using ideal gas and polytropic relations, many terms like masses and specific heats cancel out to arrive at the headline equation that's usually only a function of the ratios of state temperatures, volumes or pressures.  However, as soon as you introduce something with different properties at any time through the system during the closed part of the cycle, e.g. water, the assumption of constant mass and specific heats no longer hold.

 

One of the biggest transgressors of this was the Crower six-stroke engine.  Its capabilities were hyped from some profound ignorance of basic thermodynamics.  Literature about this engine claimed that the engine harnessed "free" energy from the spontaneous "explosive" expansion of liquid water injected into the cylinder during the hot "re-expansion" stroke after combustion into steam - strike one: there's no such thing as "free" energy; strike two: the expansion of water into steam sucks up a lot of energy - precisely the latent heat of evaporation - from the surrounding gas such that the surrounding gas would cool and contract, with the resulting overall result rather less explosive than claimed.  Also, as "proof" of the engine's incredible and revolutionary efficiency, the purveyors claimed that the exhaust from a running engine was cool enough to touch - strike 3: that's quite an imprecise metric, since I can hold the exhaust manifold of a just started and idling TDI engine for quite a while before it gets too hot to the touch because of the thermal inertia of the manifold and the fact that the exhaust of an idling Diesel is also quite relatively "cool"; strike 4 and the nail in the coffin: we're talking energy here, specifically internal energy or enthalpy where applicable and not temperatures per se. Recalling that U=m*C_v*(T₂-T₁) and H=m*C_p*(T₂-T₁),  If a normal engine rejects its exhaust, say at 1000 K, I can quite easily inject water (increase mass and C_p or C_v) so that my exit temperature is near ambient but in reality the amount of U or H has remained constant so I've really done nothing to the efficiency.

 

 

[TDIMeister]

Interesting if funny reading.

http://www.autoweek....=73238057519707

 

As of publication in 2006:

 

 

The engine has yet to operate against a load on a dyno

Yeah, no kidding! 

Edited by TDIMeister, 05 September 2013 - 21:52.

[Kelpiecross]

 

In theory, you could have an Atkinson cycle reject almost at room temperature  - ignoring that you'd have to expand to a near perfect vacuum... - but long before then you come to several practical limitations.

1. The cylinder pressure will become basically atmospheric while the temperature is still well above this.  This is one of the results of the fact that no engine cycle can be as efficient as Carnot.  If you expand to at or below atmospheric pressure, exhaust scavenging will "suck" because of the lack of blow-down and gas dynamics effects, requiring piston displacement on the upstroke (PdV work, increasing pumping losses) to do all of the scavenging. Once the exhaust valves open with cylinder pressure below atmospheric, you will have a reverse flow back into the cylinder;
2. You get no positive PdV work once the pressure drops below atmospheric but you have atmospheric pressure working against you on the other side of the piston, achieving similar results as very early historic engines that harnessed NOT the expansion force from combustion, but rather the work from the aftermath partial vacuum;
3. You'll need a very complex kinematic mechanism to get an appreciably large difference between CR and ER.

 

 

  Meister - thank you for that. 

 

    A couple of points - if by "kinematic" you mean "mechanical"  - as you know Atkinson Cycle engines these days are done by building an engine with a "geometrical" CR of  whatever you want the ER to be and then limiting the CR (and the cylinder pressure after compression) to some "normal" figure by whatever amount of late closing of the inlet valve is needed.   So essentially building an Atkinson Cycle engine with even an extremely (or even ridiculously) large difference in CR and ER is basically no more difficult than a conventional engine. 

 

  My example of a 40:1 engine was just as a novelty - it would be interesting to see it run (if it ran at all).   But I still think that more "reasonable" Atkinson engines of, for example 10:1/20:1  CR/ER would have useful applications.   I once drove a  9:1/18:1  1300cc engine in a Morris 1100 - I found that it went very well and got very good fuel mileage.  The engines also appeared to have a much smoother and more even idle than a standard 1300cc A-Series engine.  If the engine had been fitted to the much lighter Morris Mini body I suspect it would have had quite a spritely performance while returning diesel-like fuel economy. 

 

 Finally - I would not be too unkind to Bruce Crower.  He is a very good and practical innovator and builder od "novel" engine ideas - and he does build and test engines rather than just proposing ideas.  

As far as I know when he found that his 6-Stroke engine didn't perform as he expected he did not continue with the patent application or the promotion of the idea - unlike a lot of inventors of similar schemes. 

 

. 

[bigleagueslider]

  I thought it didn't seem very relevant - but I put that down to my gross stupidity.   

The link was mine, and as I noted in my post there actually is some relevance between injecting liquid nitrogen into the cylinder during the compression phase of a CI engine and injecting liquid water into the intake airflow ahead of the cylinder in an SI engine, with regards to the overall net effect on engine thermal efficiency. Liquid nitrogen and liquid water are both inert and don't contribute chemical energy during the combustion process like fuel and oxygen do.  However, both N2 and H2O have an effect on the combustion cycle by absorbing heat energy during the compression and combustion phases and then later returning that energy back to the work output of the engine via the piston and exhaust turbine systems. Obviously, the specifics of how each fluid is employed is different. But the ultimate net result on engine thermal efficiency is still similar.  To put things in perspective, consider that the intake air in a conventional engine is composed of around 78% nitrogen.

 

And you're not ignorant Klepiecross.

[TDIMeister]

    A couple of points - if by "kinematic" you mean "mechanical"  - as you know Atkinson Cycle engines these days are done by building an engine with a "geometrical" CR of  whatever you want the ER to be and then limiting the CR (and the cylinder pressure after compression) to some "normal" figure by whatever amount of late closing of the inlet valve is needed.   So essentially building an Atkinson Cycle engine with even an extremely (or even ridiculously) large difference in CR and ER is basically no more difficult than a conventional engine. 

Fair enough, using valve timing is an effective tool to alter the effective CR/ER for Miller/Atkinson.  More - actually all - contemporary real world applications I have seen (Ford and Toyota hybrids; industrial and marine engines) use early intake valve closure (before BDC) rather than late, but that's just a small detail.  However, in either case, doing so you also lose effective displacement and power output by limiting the volumetric efficiency.

 

  Finally - I would not be too unkind to Bruce Crower.  He is a very good and practical innovator and builder od "novel" engine ideas - and he does build and test engines rather than just proposing ideas.  

As far as I know when he found that his 6-Stroke engine didn't perform as he expected he did not continue with the patent application or the promotion of the idea - unlike a lot of inventors of similar schemes. 

 

. 

 

I don't hold anything against Bruce Crower the man.  He's an accomplished businessman, racer and tuner by empiricism and intuition rather than analysis, but his Six-Stroke Engine highlights a lack of understanding of thermodynamics and his ilk (not intending to use the term derogatorily on himself but rather on those like him) continue to fill the internet with snake-oil products and engine designs claiming to solve the world's problems.  

[Greg Locock]

I'd have thunk the main objection to the six stroke was that steam cleaning the oil off the cylinders every cycle was not an easy problem to get round. I think it got the traction it did due to one hyperventilating journo, Crower himself seemed to be more of the attitude he was just having a look.

[malbear]

Fair enough, using valve timing is an effective tool to alter the effective CR/ER for Miller/Atkinson.  More - actually all - contemporary real world applications I have seen (Ford and Toyota hybrids; industrial and marine engines) use early intake valve closure (before BDC) rather than late, but that's just a small detail.  However, in either case, doing so you also lose effective displacement and power output by limiting the volumetric efficiency.

 

I don't hold anything against Bruce Crower the man.  He's an accomplished businessman, racer and tuner by empiricism and intuition rather than analysis, but his Six-Stroke Engine highlights a lack of understanding of thermodynamics and his ilk (not intending to use the term derogatorily on himself but rather on those like him) continue to fill the internet with snake-oil products and engine designs claiming to solve the world's problems.  

some continue to use good technology to sell shares and use the capital to support lifestyle  about 120K a yr.. It must be under the event horizon of a certain gov department as no action is taken

[Kelpiecross]

  Having just read about Ricardo's Cryo Power engine my first reaction is that it was a April Fool's Day joke by Ricardo - especially when applied to a truck engine.   If somebody on this forum had proposed such an idea I think there would be many unkind replies to the idea. 

 

  Presumably nobody else will criticize Ricardo's Cryo truck engine because of their status - I think it is plain bloody silly.    

[Powersteer]

Thermodynamic calculations using equations derived after a long line of assumptions and algebra break down when you introduce anything through the system boundary under consideration into the closed control volume.  The first law derivation for efficiencies of closed engine cycles are always based on the premise of internal energy conservation and additionally that the fluid is an ideal gas (Except Rankine and other phase-change cycles).  From the latter, one applies the property that U=m*C_v*(T₂-T₁) and when one goes through calculating efficiencies of the form eta=W_net/Q_in or eta=(Q_in-Q_out)/Q_in using ideal gas and polytropic relations, many terms like masses and specific heats cancel out to arrive at the headline equation that's usually only a function of the ratios of state temperatures, volumes or pressures.  However, as soon as you introduce something with different properties at any time through the system during the closed part of the cycle, e.g. water, the assumption of constant mass and specific heats no longer hold.

 

One of the biggest transgressors of this was the Crower six-stroke engine.  Its capabilities were hyped from some profound ignorance of basic thermodynamics.  Literature about this engine claimed that the engine harnessed "free" energy from the spontaneous "explosive" expansion of liquid water injected into the cylinder during the hot "re-expansion" stroke after combustion into steam - strike one: there's no such thing as "free" energy; strike two: the expansion of water into steam sucks up a lot of energy - precisely the latent heat of evaporation - from the surrounding gas such that the surrounding gas would cool and contract, with the resulting overall result rather less explosive than claimed.  Also, as "proof" of the engine's incredible and revolutionary efficiency, the purveyors claimed that the exhaust from a running engine was cool enough to touch - strike 3: that's quite an imprecise metric, since I can hold the exhaust manifold of a just started and idling TDI engine for quite a while before it gets too hot to the touch because of the thermal inertia of the manifold and the fact that the exhaust of an idling Diesel is also quite relatively "cool"; strike 4 and the nail in the coffin: we're talking energy here, specifically internal energy or enthalpy where applicable and not temperatures per se. Recalling that U=m*C_v*(T₂-T₁) and H=m*C_p*(T₂-T₁),  If a normal engine rejects its exhaust, say at 1000 K, I can quite easily inject water (increase mass and C_p or C_v) so that my exit temperature is near ambient but in reality the amount of U or H has remained constant so I've really done nothing to the efficiency.

when can we get the layman version?

[Kelpiecross]

when can we get the layman version?

When it comes to thermodynamics I don't think that there really is a "layman version". However I think what Meister wrote slightly over-complicates the basic idea of thermodynamics when applied to heat engines. Heat engines (SI, diesel and steam) turn temperature in useful mechanical work - so the higher the initial or unexpanded combustion (or steam) temp is and the lower the exhaust temp is the more mechanical energy mechanical energy has been extracted from the burning fuel. The thermal efficiency is directly dependent on the initial and final temps.
Just what the lower final temp is of the burnt combustion gases depends on how much the gas is expanded - and this in turn depends on the expansion ratio (which is usually much the same as the CR). So the TE is directly dependent on the ER/CR.
I am not particularly familiar with the specific equations used Meister used - you personally get to use whatever equations (and various symbols etc.) you are familiar with.

So you need to know the basic equations linking initial/final temps and TE; TE and CR (or ER - usually the same as CR) and possibly the equation linking the compression or expansion of a gas and its final temp. This last equation depends on strange things like the ratio between specific heats etc. (Yer what?)- but this rarely comes into basic thermodynamic calculations. Also - for the ratio of specific heats etc. you always just use for internal combustion engines 0.4 - so don't have to worry where it comes from (I don't).

Sorry if you already know all this PS - but I am always trying to explain to myself, and understand, the basic application of thermodynamic equations.

I would be interested to see if anybody else has their own "words-of-one-syllable" explanation of everyday thermodynamic equations. (I might start using them instead).

Remember also the one basic law of the Universe from which I think all the laws of thermodynamics and science in general can be derived (with a bit of imagination) - "You don't get something for nothing".

You can have hours of nerdy fun working with these equations. Remember to use a "scientific" calculator when working with the 0.4 number. I remember in the days before calculators it took me two days to calculate 9 to the 0.4 using log tables etc. - you young people don't realize how times have changed.

[Kelpiecross]

I am pleased to see that I have "no warning points" - whatever that means. Nobody else seems to have this message. Do you expect cross-breed dogs to suddenly go "troppo"? (Always a chance I admit).

[RogerGraham]

Remember also the one basic law of the Universe from which I think all the laws of thermodynamics and science in general can be derived (with a bit of imagination) - "You don't get something for nothing".

 

Can you pass that on to Feliks?   

[TDIMeister]

http://www.roadandtr...-atkinson-cycle

[blkirk]

When it comes to thermodynamics I don't think that there really is a "layman version". However I think what Meister wrote slightly over-complicates the basic idea of thermodynamics when applied to heat engines. Heat engines (SI, diesel and steam) turn temperature in useful mechanical work - so the higher the initial or unexpanded combustion (or steam) temp is and the lower the exhaust temp is the more mechanical energy mechanical energy has been extracted from the burning fuel. The thermal efficiency is directly dependent on the initial and final temps.
Just what the lower final temp is of the burnt combustion gases depends on how much the gas is expanded - and this in turn depends on the expansion ratio (which is usually much the same as the CR). So the TE is directly dependent on the ER/CR.
I am not particularly familiar with the specific equations used Meister used - you personally get to use whatever equations (and various symbols etc.) you are familiar with.

So you need to know the basic equations linking initial/final temps and TE; TE and CR (or ER - usually the same as CR) and possibly the equation linking the compression or expansion of a gas and its final temp. This last equation depends on strange things like the ratio between specific heats etc. (Yer what?)- but this rarely comes into basic thermodynamic calculations. Also - for the ratio of specific heats etc. you always just use for internal combustion engines 0.4 - so don't have to worry where it comes from (I don't).

Sorry if you already know all this PS - but I am always trying to explain to myself, and understand, the basic application of thermodynamic equations.

I would be interested to see if anybody else has their own "words-of-one-syllable" explanation of everyday thermodynamic equations. (I might start using them instead).

Remember also the one basic law of the Universe from which I think all the laws of thermodynamics and science in general can be derived (with a bit of imagination) - "You don't get something for nothing".

You can have hours of nerdy fun working with these equations. Remember to use a "scientific" calculator when working with the 0.4 number. I remember in the days before calculators it took me two days to calculate 9 to the 0.4 using log tables etc. - you young people don't realize how times have changed.

Law #1 - You can't win.  You can only break even or lose.

Law #2 - You can only break even at absolute zero.

Law #3 - You can't get to absolute zero.

[Greg Locock]

9^0.4 in my head:  log 9 is about 0.95, .4*.95=.38, 10^.38 is a bit tricky, but log 2=.3, log 3=.5, so antilog .4 must be about 2.5

[TDIMeister]

Law #1 - You can't win.  You can only break even or lose.

Law #2 - You can only break even at absolute zero.

Law #3 - You can't get to absolute zero.

Couldn't have stated it better or more layman myself! 

[gruntguru]

9^0.4 in my head:  log 9 is about 0.95, .4*.95=.38, 10^.38 is a bit tricky, but log 2=.3, log 3=.5, so antilog .4 must be about 2.5

Awesome. You are one of a dying breed.