`
`(12)
`
`United States Patent
`Bromberg et al.
`
`(10) Patent No.:
`(45) Date of Patent:
`
`US 7,225,787 B2
`Jun. 5, 2007
`
`(54) OPTIMIZED FUEL MANAGEMENT SYSTEM
`FOR DIRECT NUECTION ETHANOL
`ENHANCEMENT OF GASOLINE ENGINES
`
`(75) Inventors: Leslie Bromberg, Sharon, MA (US);
`Daniel R. Cohn, Cambridge, MA (US);
`John B. Heywood, Newton, MA (US)
`
`(73) Assignee: Massachusetts Institute of
`Technology, Cambirdge, MA (US)
`
`(*) Notice:
`
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`U.S.C. 154(b) by 31 days.
`
`(21) Appl. No.: 11/100,026
`(22) Filed:
`Apr. 6, 2005
`
`(65)
`
`Prior Publication Data
`US 2006/O1 O2136 A1
`May 18, 2006
`Related U.S. Application Data
`(63) Continuation-in-part of application No. 10/991,774,
`filed on Nov. 18, 2004.
`
`6, 2000 Wulff et al.
`6,076.487 A
`9/2001 Wulff et al.
`6,287.351 B1
`10/2001 Huff et al.
`6,298,838 B1
`12/2001 Ilyama et al.
`6,332,448 B1
`1/2003 Suhre et al.
`6,508,233 B1
`6,513,505 B2 * 2/2003 Watanabe et al. ........... 123/ 525
`6,543,423 B2
`4/2003 Dobryden et al.
`6,575,147 B2
`6, 2003 Wulff et al.
`6,668,804 B2 12/2003 Dobryden et al.
`6,990,956 B2
`1/2006 Niimi .................... 123,406.47
`7,021,277 B2 * 4/2006 Kuo et al. .................. 123,299
`2002fOO14226 A1
`2/2002 Wulff et al.
`
`OTHER PUBLICATIONS
`A. Modak and L.S. Caretto, Engine Cooling by Direct Injection of
`Cooling Water, Society of Automotive Engineers, Inc. 700887.
`
`(Continued)
`Primary Examiner Stephen K. Cronin
`AC, F.C. E. Ali
`R "SCIES' or Firm—Sam Pastemack; Choate
`
`(57)
`
`ABSTRACT
`
`(2006.01)
`(51) E. ,/04
`(52) U.S. Cl. .........r irrir. 123/198A
`(58) Field of Classification Search ............ 123/198 A.
`123/406.29,406.47, 435, 559.1, 25 C
`See application file for complete search history.
`References Cited
`
`(56)
`
`U.S. PATENT DOCUMENTS
`
`Fuel management system for enhanced operation of a spark
`ignition gasoline engine. Injectors inject an anti-knock agent
`such as ethanol directly into a cylinder. It is preferred that the
`direct injection occur after the inlet valve is closed. It is also
`preferred that stoichiometric operation with a three way
`catalyst be used to minimize emissions. In addition, it is also
`preferred that the anti-knock agents have a heat of vapor
`ization per unit of combustion energy that is at least three
`times that of gasoline.
`
`4.480,616 A *
`5.937,799 A *
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`11/1984 Takeda ..................
`8, 1999 Binion
`
`123,406.52
`
`- - - - - - - - - - - - - - - - - - - - - 123, 25 C
`
`9 Claims, 3 Drawing Sheets
`
`14
`
`ethanol
`injector
`
`gasoline
`injector
`
`
`
`O2 sensor
`
`FORD Ex. 1140, page 1
` IPR2020-00013
`
`
`
`US 7,225,787 B2
`Page 2
`
`OTHER PUBLICATIONS
`Julian A. LoRusso and Harry A. Cikanek, Direct Injection Ignition
`Assisted Alcohol Engine, Society of Automotive Engineers, Inc.
`880495, International Congress and Exposition in Detroit Michigan
`(Feb. 29-Mar. 4, 1998).
`Börje Grandin, Hans-Erik Angström, Per Stålhammar and Eric
`Olofsson, Knock Suppression in a Turbocharged SI Engine by
`Using Cooled EGR, Society of Automotive Engineers, Inc. 982476,
`International Fall Fuels and Lubricants Meeting and Exposition in
`San Francisco, California (Oct. 19-22, 1998).
`
`Börje Grandin, Hans-Erik Angström, Replacing Fuel Enrichment in
`a Turbo Charged SI Engine: Lean Burn or Cooled EGR, Society of
`Automotive Engineers, Inc. 199-01-3505.
`C. Stan, R. Troeger, S. Guenther, A. Stanciu, L. Martorano, C.
`Tarantino and R. Lensi, Internal Mixture Formation and
`Combustion—from Gasoline to Ethanol, Society of Automotive
`Engineers, Inc. 2001-01-1207.
`
`* cited by examiner
`
`FORD Ex. 1140, page 2
` IPR2020-00013
`
`
`
`U.S. Patent
`
`Jun. 5, 2007
`
`Sheet 1 of 3
`
`US 7.225,787 B2
`
`OOAS
`
`
`
`75%
`
`SOA
`
`25
`
`l
`
`3.
`2.5
`2
`1.5
`Inlet manifold pressure (bar)
`-I G. 4
`
`3.5
`
`FORD Ex. 1140, page 3
` IPR2020-00013
`
`
`
`U.S. Patent
`
`Jun. 5, 2007
`
`Sheet 2 of 3
`
`US 7.225,787 B2
`
`300
`
`250
`
`2 O O
`
`15 O
`
`1 O O
`
`50
`
`180
`
`230
`
`0.82 ethanol
`(energy fraction)
`
`Autoignition
`
`w
`/
`1 0.83 ethanoln's
`28O
`330
`380
`6, 2- (R - Crank angle
`
`430
`
`480
`
`33OO
`
`2800
`
`an
`a 2300
`
`1800
`3.
`5 1300
`800
`300
`18O
`
`
`
`O.82 ethanol
`(energy fraction)
`
`Autoignition
`N
`1 Y - 983 ethanol
`
`we
`
`230
`
`280
`330
`ft 6.2b Crank angle
`
`380
`
`430
`
`480
`
`Exhaus
`
`FORD Ex. 1140, page 4
` IPR2020-00013
`
`
`
`U.S. Patent
`
`Jun. 5, 2007
`
`Sheet 3 of 3
`
`US 7,225,787 B2
`
`
`
`gasoline
`tank
`
`
`
`
`
`gasoline
`pump
`
`Ethanol
`separator
`
`To engine
`
`(6 - 4/
`
`
`
`Ethanol
`pump
`
`To engine
`
`
`
`
`
`
`
`gasoline
`tank
`
`Ethanol
`separator
`
`gasoline
`pump
`
`To engine
`
`
`
`
`
`Ethanol
`tank
`
`FI6. Ah (b)
`
`To engine
`
`
`
`Movabled deformable wall
`
`
`
`Fuel tank
`
`Ethanol fill
`
`Gasoline fill
`
`Gasoline fuel line
`
`
`
`Ethanol fuel line
`
`- Lee S
`
`FORD Ex. 1140, page 5
` IPR2020-00013
`
`
`
`US 7,225,787 B2
`
`1.
`OPTIMIZED FUEL MANAGEMENT SYSTEM
`FOR DIRECT NUECTION ETHANOL
`ENHANCEMENT OF GASOLINE ENGINES
`
`This application is a continuation-in-part of U.S. patent
`application Ser. No. 10/991,774 filed Nov. 18, 2004 entitled,
`“Fuel Management System for Variable Ethanol Octane
`Enhancement of Gasoline Engines' the contents of which
`are incorporated herein by reference in their entirety.
`
`BACKGROUND OF THE INVENTION
`
`This invention relates to an optimized fuel management
`system for use with spark ignition gasoline engines in which
`an anti-knock agent which is a fuel is directly injected into
`a cylinder of the engine.
`There are a number of important additional approaches
`for optimizing direct injection ethanol enhanced knock
`Suppression so as to maximize the increase in engine effi
`ciency and to minimize emissions of air pollutants beyond
`the technology disclosed in parent application Ser. No.
`10/991,774 set out above. There are also additional
`approaches to protect the engine and exhaust system during
`high load operation by ethanol rich operation; and to mini
`mize cost, ethanol fuel use and ethanol fuel storage require
`ments. This disclosure describes these approaches.
`These approaches are based in part on more refined
`calculations of the effects of variable ethanol octane
`enhancement using a new computer model that we have
`developed. The model determines the effect of direct injec
`tion of ethanol on the occurrence of knock for different times
`of injection and mixtures with port fuel injected gasoline. It
`determines the beneficial effect of evaporative cooling of the
`direct ethanol injection upon knock Suppression.
`
`10
`
`15
`
`25
`
`30
`
`35
`
`SUMMARY OF THE INVENTION
`
`In one aspect, the invention is a fuel management system
`for operation of a spark ignition gasoline engine including a
`gasoline engine and a source of an anti-knock agent which
`is a fuel. The use of the anti-knock agent provides gasoline
`savings both by facilitating increased engine efficiency over
`a drive cycle and by Substitution for gasoline as a fuel. An
`injector is provided for direct injection of the anti-knock
`agent into a cylinder of the engine and a fuel management
`control system controls injection of the anti-knockagent into
`the cylinder to control knock. The injection of the antiknock
`agent can be initiated by a signal from a knock sensor. It can
`also be initiated when the engine torque is above a selected
`value or fraction of the maximum torque where the value or
`fraction of the maximum torque is a function of the engine
`speed. In a preferred embodiment, the injector injects the
`anti-knock agent after inlet valve?valves are closed. It is
`preferred that the anti-knock agent have a heat of vaporiza
`tion that is at least twice that of gasoline or a heat of
`vaporization per unit of combustion energy that is at least
`three times that of gasoline. A preferred anti-knock agent is
`ethanol. In a preferred embodiment of this aspect of the
`invention, part of the fuel is port injected and the port
`injected fuel is gasoline. The directly injected ethanol can be
`mixed with gasoline or with methanol. It is also preferred
`that the engine be capable of operating at a manifold
`pressure at least twice that pressure at which knock would
`occur if the engine were to be operated with naturally
`aspirated gasoline. A Suitable maximum ethanol fraction
`during a drive cycle when knock Suppression is desired is
`between 30% and 100% by energy. It is also preferred that
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`2
`the compression ratio be at least 10. With the higher mani
`fold pressure, the engine can be downsized by a factor of two
`and the efficiency under driving conditions increased by
`30%.
`It is preferred that the engine is operated at a substantially
`stoichiometric air/fuel ratio during part or all of the time that
`the anti-knock agent Such as ethanol is injected. In this case,
`a three-way catalyst can be used to reduce the exhaust
`emissions from the engine. The fuel management system
`may operate in open or closed loop modes.
`In some embodiments, non-uniform ethanol injection is
`employed. Ethanol injection may be delayed relative to
`bottom dead center when non-uniform ethanol distribution is
`desired.
`Many other embodiments of the invention are set forth in
`detail in the remainder of this application.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`FIG. 1 is a graph of ethanol fraction (by energy) required
`to avoid knock as a function of inlet manifold pressure. The
`ethanol fraction is shown for various values of B, the ratio
`of the change in temperature in the air cylinder charge due
`to turbocharging (and aftercooling if used) to the adiabatic
`temperature increase of the air due to the turbocharger.
`FIG. 2a is a graph of cylinder pressure as a function of
`crank angle for a three bar manifold pressure.
`FIG.2b is a graph of charge temperature as a function of
`crank angle for a three bar manifold pressure.
`FIG. 3 is a schematic diagram of an embodiment of the
`fuel management system disclosed herein for maintaining
`Stoichiometric conditions with metering/control of ethanol,
`gasoline, and air flows into an engine.
`FIGS. 4a and 4b are schematic illustrations relating to the
`separation of ethanol from ethanol/gasoline blends.
`FIG. 5 is a cross-sectional view of a flexible fuel tank for
`a vehicle using ethanol boosting of a gasoline engine.
`
`DESCRIPTION OF THE PREFERRED
`EMBODIMENT
`
`Ethanol has a heat of vaporization that is more than twice
`that of gasoline, a heat of combustion per kg which is about
`60% of that of gasoline, and a heat of vaporization per unit
`of combustion energy that is close to four times that of
`gasoline. Thus the evaporative cooling of the cylinder air/
`fuel charge can be very large with appropriate direct injec
`tion of this antiknock agent. The computer model referenced
`below shows that evaporative cooling can have a very
`beneficial effect on knock suppression. It indicates that the
`beneficial effect can be maximized by injection of the
`ethanol after the inlet valve that admits the air and gasoline
`into the cylinder is closed. This late injection of the ethanol
`enables significantly higher pressure operation without
`knock and thus higher efficiency engine operation than
`would be the case with early injection. It is thus preferred to
`the conventional approach of early injection which is used
`because it provides good mixing. The model also provides
`information that can be used for open loop (i.e., a control
`system that uses predetermined information rather than
`feedback) fuel management control algorithms.
`The increase in gasoline engine efficiency that can be
`obtained from direct injection of ethanol is maximized by
`having the capability for highest possible knock Suppression
`enhancement. This capability allows the highest possible
`amount of torque when needed and thereby facilitates the
`largest engine downsizing for a given compression ratio.
`
`FORD Ex. 1140, page 6
` IPR2020-00013
`
`
`
`3
`Maximum knock suppression is obtained with 100% or
`close to 100% use of direct injection of ethanol. A small
`amount of port injection of gasoline may be useful in order
`to obtain combustion stability by providing a more homo
`geneous mixture. Port fuel injection of gasoline also
`removes the need for a second direct fuel system or a more
`complicated system which uses one set of injectors for both
`fuels. This can be useful in minimizing costs.
`The maximum fraction of ethanol used during a drive
`cycle will depend upon the engine system design and the
`desired level of maximum torque at different engine speeds.
`A representative range for the maximum ethanol fraction by
`energy is between 20% and 100%.
`In order to obtain the highest possible octane enhance
`ment while still maintaining combustion stability, it may be
`useful for 100% of the fuel to come from ethanol with a
`fraction being port injected, as an alternative to a small
`fraction of the port-fueled gasoline.
`The initial determination of the knock suppression by
`direct injection of ethanol into a gasoline engine has been
`refined by the development of a computer model for the
`onset of knock under various conditions. The computer
`modeling provides more accurate information for use in fuel
`management control. It also shows the potential for larger
`octane enhancements than our earlier projections. Larger
`octane enhancements can increase the efficiency gain
`through greater downsizing and higher compression ratio
`operation. They can also reduce the amount of ethanol use
`for a given efficiency increase.
`The computer model combines physical models of the
`ethanol vaporization effects and the effects of piston motion
`of the ethanol/gasoline/air mixtures with a state of the art
`calculational code for combustion kinetics. The calcula
`tional code for combustion kinetics was the engine module
`in the CHEMKIN 4.0 code R. J. Kee, F. M. Rupley, J. A.
`Miller, M. E. Coltrin, J. F. Grcar, E. Meeks, H. K. Moffat, A.
`E. Lutz, G. Dixon-Lewis, M. D. Smooke, J. Warnatz, G. H.
`Evans, R. S. Larson, R. E. Mitchell, L. R. Petzold, W. C.
`Reynolds, M. Caracotsios, W. E. Stewart, P. Glarborg, C.
`Wang, O. Adigun, W. G. Houf, C. P. Chou, S. F. Miller, P.
`Ho, and D. J. Young, CHEMKIN Release 4.0, Reaction
`Design, Inc., San Diego, Calif. (2004). The CHEMKIN
`code is a Software tool for Solving complex chemical kinet
`ics problems. This new model uses chemical rates informa
`tion based upon the Primary Reference gasoline Fuel (PRF)
`mechanism from Curran et al. Curran, H. J., Gaffuri, P.
`Pitz, W. J., and Westbrook, C. K. “A Comprehensive Mod
`eling Study of iso-Octane Oxidation.” Combustion and
`Flame 129:253–280 (2002) to represent onset of autoigni
`tion.
`The compression on the fuel/air mixture end-gas was
`modeled using the artifact of an engine compression ratio of
`21 to represent the conditions of the end gas in an engine
`with an actual compression ratio of 10. The end gas is
`defined as the un-combusted air/fuel mixture remaining after
`75% (by mass) of the fuel has combusted. It is the end gas
`that is most prone to autoignition (knock). The larger com
`pression ratio includes the effect of the increase in pressure
`in the cylinder due to the energy released in the combustion
`of 75% of the fuel that is not in the end gas region. The effect
`of direct ethanol vaporization on temperature was modeled
`by consideration of the effects of the latent heat of vapor
`ization on temperature depending upon the time of the
`injection.
`The effect of temperature increase due to turbocharging
`was also included. The increase in temperature with turbo
`charging was calculated using an adiabatic compression
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`US 7,225,787 B2
`
`10
`
`15
`
`4
`model of air. It is assumed that thermal transfer in the piping
`or in an intercooler results in a smaller temperature increase.
`The effect is modeled by assuming that the increase in
`temperature of the air charge into the cylinder AT
`is
`AT
`faT
`were AT
`is the temperature increase
`tairbo
`tairbo
`after th compressor due to boosting and beta is a constant.
`Values of B of 0.3, 0.4 and 0.6 have been used in the
`modeling. It is assumed that the temperature of the charge
`would be 380 K for a naturally aspirated engine with port
`fuel injection gasoline.
`FIG. 1 shows the predictions of the above-referenced
`computer model for the minimum ethanol fraction required
`to prevent knock as a function of the pressure in the inlet
`manifold, for various values of B. In FIG. 1 it is assumed that
`the direct injection of the ethanol is late (i.e. after the inlet
`valve that admits air and gasoline to the cylinder is closed)
`and a 87 octane PRF (Primary Reference Fuel) to represent
`regular gasoline. The corresponding calculations for the
`manifold temperature are shown in Table 1 for the case of a
`pressure in the inlet manifold of up to 3 bar for an engine
`with a conventional compression ratio of 10. The tempera
`ture of the charge varies with the amount of ethanol directly
`injected and is self-consistently calculated in Table 1 and
`FIG. 1. The engine speed used in these calculations is 1000
`rpm.
`
`TABLE 1.
`
`Computer model calculations of temperature and ethanol fraction
`required for knock prevention for an inlet manifold pressure
`of 3 bar for an engine with a compression ratio of 10, for
`various values off (ratio of change of the cylinder air
`charge temperature due to turbocharging to the adiabatic temperature
`increase due to turbocharging ATchase = fB ATs).
`The engine Speed is 1000 rpm.
`
`T charge init
`Delta T turbo
`Delta T after intercooler
`Delta T due to DI ethanol and gasoline
`T init equivalent charge
`Gasoline octane
`Ethanol fraction (by energy) needed
`to prevent knock
`
`-B-
`
`O.3
`
`0.4
`
`O.6
`
`380
`K
`18O
`K
`S4
`K
`K -103
`K
`331
`87
`74%.
`
`380
`18O
`72
`-111
`341
`87
`82%
`
`380
`18O
`108
`-132
`356
`87
`97%
`
`Direct fuel injection is normally performed early, before
`the inlet valve is closed in order to obtain good mixing of the
`fuel and air. However, our computer calculations indicate a
`substantial benefit from injection after the inlet valve is
`closed.
`The amount of air is constant in the case of injection after
`the inlet valve has closed. Therefore the temperature change
`is calculated using the heat capacity of air at constant
`volume (c.). The case of early injection where the valve that
`admits air and fuel to the cylinder is still open is modeled
`with a constant-pressure heat capacity (c.). The constant
`Volume case results in a larger evaporation induced decrease
`in charge temperature than in the case for constant pressure,
`by approximately 30%. The better evaporative cooling can
`allow operation at higher manifold pressure (corresponding
`to a greater octane enhancement) without knock that would
`be the case of early injection by a difference of more than 1
`bar. The increase in the evaporative cooling effect at con
`stant Volume relative to that at constant pressure is Substan
`
`FORD Ex. 1140, page 7
` IPR2020-00013
`
`
`
`US 7,225,787 B2
`
`10
`
`15
`
`25
`
`30
`
`5
`tially higher for the case of direct injection of fuels such as
`ethanol and methanol than is the case for direct injection of
`gasoline.
`Typical results from the calculations are shown in FIG. 2.
`The figure shows the pressure (a) and the temperature (b) of 5
`the cylinder charge as a function of crank angle, for a
`manifold pressure of 3 bar and a value of B=0.4 Two values
`of the ethanol fraction are chosen, one that results in
`autoignition, and produces engine knock (0.82 ethanol frac
`tion by fuel energy), and the other one without autoignition,
`i.e., no knock (0.83 ethanol fraction). Autoignition is a
`threshold phenomenon, and in this case occurs between
`ethanol fractions of 0.82 and 0.83. For an ethanol energy
`fraction of 0.83, the pressure and temperature rise at 360°
`(top dead center) is due largely to the compression of the air
`fuel mixture by the piston. When the ethanol energy fraction
`is reduced to 0.82, the temperature and pressure spikes as a
`result of autoignition. Although the autoignition in FIG. 2
`occurs substantially after 360 degrees, the autoignition tim
`ing is very sensitive to the autoignition temperature (5 crank
`angle degrees change in autoignition timing for a change in
`the initial temperature of 1 K, or a change in the ethanol
`energy fraction of 1%).
`The effect of evaporative cooling from the antiknock
`agent (in this case, ethanol) is shown in Table 2, where three
`cases are compared. The first one is with port fuel injection
`of ethanol. In this case the vaporization of the ethanol on the
`walls of the manifold has a negligible impact on the tem
`perature of the charge to the cylinder because the walls of the
`manifold are cooled rather than the air charge. The second
`case assumes direct injection, but with the inlet valve open,
`with evaporation at constant pressure, where the cooling of
`the charge admits additional air to the cylinder. The third
`case assumes, as in the previous discussions, late injection
`after the inlet valve has closed. It is assumed stoichiometric
`operation, that the baseline temperature is 380 K, and that
`there is cooling in the manifold after the turbocharger with
`|B=0.4.
`
`35
`
`TABLE 2
`
`Knock-free operation of ethanol port fuel injection (assuming no charge
`cooling), and of direct injection before and after the inlet valve is
`closed. Compression ratio of 10, baseline charge temperature of 380 K,
`intercooler cooling post turbo with f = 0.4, Stoichiometric
`operation, gasoline with 87 RON. Engine Speed is 1000 rpm.
`
`40
`
`45
`
`No
`
`Evaporative cooling
`
`Evaporative
`Cooling
`
`After
`Before
`Valve Closing Valve Closing 50
`
`O.95
`
`1.OS
`
`1.OS
`
`O.9S
`
`2.4
`
`2.4
`
`O.95
`
`4.0
`
`3.0
`
`383
`
`360
`
`355
`
`55
`
`Ethanol fraction
`(by energy)
`Max manifold pressure
`(bar)
`Cylinder pressure
`after cooling (bar)
`Cylinder charge
`temperature after
`cooling (K)
`
`The results indicate the strong effect of the cooling. The
`maximum manifold pressure that prevents knock (without
`spark retard), with 0.95 ethanol fraction by energy in the
`case of port fuel injection is 1.05 bar. With direct injection
`of the ethanol, the maximum knock-free manifold and
`cylinder pressures are 2.4 bar, with a temperature decrease
`of the charge of ~7.5 K. The final case, with injection after
`inlet valve closing, allows a manifold pressure of 4 bar, a
`
`60
`
`65
`
`6
`cylinder pressure (after cooling) of 3 bar, and a charge
`temperature decrease of ~120 K. It should be noted that the
`torque of the late injection case after the valve has closed is
`actually higher than that of the early injection case, even
`though the early injection case allows for additional air (at
`constant pressure). For comparison, the model is also used
`to calculate the manifold pressure at which knock would
`occur for port fuel injection of 87 octane gasoline alone. This
`pressure is -0.8 bar assuming spark timing at MBT (Maxi
`mum Brake Torque). Conventional gasoline engines operate
`at 1 bar by retarding the timing at high torque regions where
`knock would otherwise occur. Thus the model indicates that
`evaporative cooling effect of direct injection of ethanol after
`the inlet valve has closed can be significantly greater than
`that of the higher octane number rating of ethanol relative to
`gasoline.
`A manifold pressure of 4 bar is very aggressive. Table 2
`is indicative of the dramatically improved performance of
`the system with direct injection after the inlet valve has
`closed. The improved performance in this case can be traded
`for increased compression ratio or reduced use of the
`anti-knock agent.
`It should be noted that, as mentioned above, the calcula
`tions of autoignition (knock) are conservative, as autoigni
`tion for the case shown in FIG. 2 occurs relatively late in the
`cycle, and it is possible that the fuel has been combusted
`before it autoignites. Also it should be noted that the
`calculations in FIG. 2 break down after autoignition, as the
`pressure trace would be different from that assumed. Figures
`similar to FIG. 2 are used to determine conditions where
`autoignition would not occur, and those conditions are then
`used to provide the information for FIG. 1. The initial
`temperatures of the cases shown in FIG. 2 are 341 K for 0.82
`ethanol fraction, and 340 K for 0.83 ethanol fraction, a
`difference of 1 K (the difference due to the cooling effect of
`the ethanol).
`Because of the large heat of vaporization, there could be
`enough charge cooling with early injection so that the rate of
`vaporization of ethanol is substantially decreased. By
`instead injecting into the hot gases, which is the case with
`injection after the inlet valve has closed, the temperature at
`the end of full vaporization of the ethanol is substantially
`increased with respect to early injection, increasing the
`evaporation rate and minimizing wall wetting.
`The optimum timing of the injection for best mixing and
`a near homogeneous charge is soon after the inlet valve
`closes, provided that the charge is sufficiently warm for
`antiknock agent vaporization. If on the other hand, a non
`uniform mixture is desired in order to minimize ethanol
`requirements and improve ignition stability, then the injec
`tion should occur later than in the case where the best
`achievable mixing is the goal.
`Late injection of the ethanol after the inlet valve has
`closed can be optimized through the use of diesel-like
`injection schemes, such as injectors with multiple sprays. It
`is important to inject the fuel relatively quickly, and at
`velocities which minimize any cylinder wall wetting, which
`as described below could result in the removal of the
`lubrication oils from the cylinder liner. Multiple sprays from
`a nozzle that has multiple holes results in a distributed
`pattern of sprays, with relatively low injection velocities.
`This is particularly important for ethanol, because of the
`higher Volume throughputs (as compared with gasoline) of
`ethanol for equal energy content.
`Injection after the valve has closed may require that a
`modest fraction of the fuel (e.g. 25%) be port injected in
`order to achieve the desired combustion stability. A tumble
`
`FORD Ex. 1140, page 8
` IPR2020-00013
`
`
`
`US 7,225,787 B2
`
`5
`
`10
`
`25
`
`7
`like or swirl motion can be introduced to achieve the desired
`combustion stability. The port injected fuel can be either
`gasoline or ethanol.
`Use of the computer model for operation with gasoline
`alone gives results that are consistent with the observed
`occurrence of knock in gasoline engine vehicles, thereby
`buttressing the credibility of the projections for ethanol. The
`computer model indicates that for knock-free gasoline
`operation alone with a compression ratio of 10, knock
`imposes a severe constraint upon the allowed manifold
`pressure for a naturally aspirated gasoline engine and very
`limited (i.e., less than 1.2 bar) manifold pressure can be
`achieved even with direct injection of gasoline unless spark
`retard and/or rich operation is used. These changes, how
`ever, can reduce efficiency and increase emissions.
`15
`FIG. 1 shows that knock can be prevented at manifold
`pressures greater than 2 bar with direct injection of an
`ethanol fraction of between 40 and 80% in an engine with a
`compression ratio of 10. The manifold pressure can be at
`least 2.5 bar without engine knock. A pressure of 3 bar
`would allow the engine to be downsized to ~/3 of the
`naturally aspirated gasoline engine, while still producing the
`same maximum torque and power. The large boosting indi
`cated by the calculations above may require a multiple-stage
`turbocharger. In addition to a multiple stage turbocharger,
`the turbocharger may be of the twin-scroll turbo type to
`optimize the turbocharging and decrease the pressure fluc
`tuations in the inlet manifold generated by a small number
`of cylinders.
`With an increase in allowed manifold pressure in an
`engine by more than a factor of 2, the engine could be
`downsized by a factor of 2 (that is, the cylinder volume is
`decreased by a factor of 2 or more) and the compression
`ratio could be held constant or raised. For example, the
`performance of an eight cylinder engine is achieved by a
`four cylinder engine.
`The occurrence of knock at a given value of torque
`depends upon engine speed. In addition to providing Sub
`stantially more maximum torque and power, direct injection
`of ethanol can be used to provide a significant improvement
`in torque at low engine speeds (less than 1500 rpm) by
`decreasing or eliminating the spark retard. Spark retard is
`generally used with gasoline engines to prevent knock at low
`engine speeds where autoignition occurs at lower values of
`torque than is the case at high engine speeds.
`45
`FIG. 1 can also be used to determine the ethanol fraction
`required to prevent knock at different levels of torque and
`horsepower, which scale with manifold pressure in a given
`size engine. This information can be used in an open loop
`control system.
`The efficiency of a gasoline engine under driving condi
`tions using direct ethanol injection enhancement can be at
`least 20% and preferably at least 30% greater than that of a
`naturally aspirated gasoline engine with a compression ratio
`of 10. This increase results from the substantial engine
`boosting and downsizing to give the same power, and also
`the high compression ratio operation (compression ratio of
`11 or greater) that is enabled by a large octane enhancement.
`With more aggressive downsizing of more than 50% (where
`the same engine performance is obtained with less than
`one-half the displacement), the increase in efficiency could
`exceed 30%.
`Greater downsizing and higher efficiency may also be
`obtained by decreasing the octane requirement of the engine
`by using variable valve timing (VVT). Thus, at conditions of 65
`high torque, variable valve timing can be used to decrease
`the compression ratio by appropriately changing the open
`
`8
`ing/closing of the inlet and exhaust valves. The loss in
`efficiency at high torque has a small impact on the overall
`fuel economy because the engine seldom operates in these
`conditions.
`VVT can also be used to better scavenge the exhaust gases
`B. Lecointe and G. Monnier, “Downsizing a Gasoline
`Engine Using Turbocharging with Direct Injection SAE
`paper 2003-01-0542. Decreasing the exhaust gas decreases
`the air/fuel temperature. Keeping both the inlet and exhaust
`valves open, while the pressure in the inlet manifold is
`higher than in the exhaust, can be used to remove the exhaust
`gases from the combustion chamber. This effect, coupled
`with slightly rich operation in-cylinder, can result in
`increased knock avoidance while the exhaust is still sto
`ichiometric. Cooled EGR and spark timing adjustment can
`also be used to increase knock avoidance.
`Any delay in delivering high engine torque at low engine
`speeds can decrease drivability of the vehicle. Under these
`conditions, because of the Substantial engine downsizing,
`the vehicle would have insufficient acceleration at low
`engine speeds until the turbo produces high pressures. This
`delay can be removed through the use of direct injection of
`ethanol by reduction of the spark retard or ethanol/gasoline
`with rich operation and also with the use of variable valve
`timing.
`Another approach would be to use an electrically assisted
`turbo charger. Units that can generate the required boosting
`for short periods of time are available. The devices offer very
`fast response time, although they have Substantial power
`requirements.
`A multiple scroll turbocharger can be used to decrease the
`pressure fluctuations in the manifold that could result from
`the decreased number of cylinders in a downsized engine.
`The temperature of the air downstream from the turbo
`charger is increased by the compression process. Use of an
`intercooler can prevent this temperature increase from
`increasing the engine's octane requirement. In addition, in
`order to maximize the power available from the engine for
`a given turbocharging, cooling of the air charge results in
`increased mass of air into the cylinder, and thus higher
`power.
`In order to minimize emissions, the engine should be
`operated substantially all of the time, or most of the time,
`with a stoichiometric air/fuel ratio in order that a 3-way
`exhaust catalyst treatment can be used. FIG. 3 shows a