THE UN I V E R S I T Y OF MICHIGAN COLLEGE OF ENGINEERING Department of Mechanical Engineering Final Report STRATIFIED AND LEAN FUEL-AIR RATIO ENGINES AND THEIR POSSIBILITIES E. T. Vincent Kamalakar Rao ORA Project 05841 under contract with: U. S. ARMY TANK-AUTOMOTIVE CENTER PROPULSION SYSTEMS LABORATORY CONTRACT NOo DA-20-018-AMC-0729-T WARREN, MICHIGAN administered through: OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR November 1966

TABLE OF CONTENTS Page LIST OF TABLES v LIST OF FIGURES vii ABSTRACT ix I. PURPOSE 1 IIo INTRODUCTION 2 A. Lean Fuel-Air Ratio Engine 2 Bo Stratified-Charge Engine 3 III. METHODS OF CALCULATION 4 A. Lean Fuel-Air Ratio Engine 4 B. Stratified-Charge Engine 4 C. Compression 5 D. Combustion 11 Eo Expansion 11 IV. SAMPLE CALCULATION FOR THE STRATIFICATION SYSTEM 13 V. RESULTS 19 VI. DISCUSSION 24 A. Stratified-Charge Engine 25 B. Lean Fuel-Air Ratio Engine 27 VII. PRESENT STATUS OF RESEARCH 35 VIII. CONCLUSIONS 36 IX. REFERENCES 37 iii

LIST OF TABLES Table Page I. Variation of Ratio as x Varies 9 II. Performance of Ideal Stratified-Charge Engine with Rich Mixture When x = 1.0 19 III. Performance of Ideal Stratified-Charge Engine with Lean Mixtures When x = 1.0 21 IV. Ideal Lean Fuel-Air Ratio L-141 Engine 23 v

LIST OF FIGURES Figure Page 1. Diagrammatic sketch of imaginary cylinder for stratification. 5 2. Gas constants for various mixture ratios. 8 3. Compression ratios for air (Ra) and mixture (Rg), overall ratio = 7.5:1. 10 4. Diagram of conditions at BDC of expansion stroke. 11 5. Indicator diagram when x = 0.9. 13 6. Ideal performance of a stratified-charge engine. 20 7. Ideal performance of a lean fuel-air ratio engine. 22 8. Combined performance chart. 26 9. Engine efficiency for stratified-charge engine. 28 10. Engine efficiency for lean fuel-air ratio engine. 28 11. Comparison with L-141 engine. 30 12. Net BHP performance at 3600 rpm, including air cleaner, generator, cooling fan, and exhaust system. 31 135 Comparison of cycles on BHP basis. 33 vii

ABSTRACT This report compares hybrid engines with conventional spark-ignition carbureted engines to determine any possible gains in fuel economy. Two combustion systems, the stratified-charge and the lean fuel-air ratio engines, are examined, and ideal cycle estimates for these engines are compared with those for the standard L-141 engine. These estimates are then converted to net BHP output for a practical engine comparison. ix

I. PURPOSE In this study we investigated the possibilities of improving the cycle efficiency of gasoline engines operating at part loads by eliminating the throttling losses of the conventional engine. The fuel supply was reduced; the air supply remained about constant at all speeds. For many years experimental work on gasoline engine efficiency has been done by various authorities and research organizations; however, to date no satisfactory method of increasing engine efficiency has been found. In the past few years one or two systems have shown promise, but developing them into acceptable engines requires much additional work and expense. In this study the theoretical possibilities of engine control systems were carefully analyzed to obtain an ideal yardstick of ultimate possibilities, and to justify the expenditure of additional funds where necessary. 1

II. INTRODUCTION The overall combustion process in mixtures of gasoline and air in sparkignition engines is well understood today. In this process the mixture ratio is maintained within well defined limits to produce the flame velocity needed to complete combustion in the time available for the desired cylinder size. If the fuel supply in an engine is reduced while the air supply remains constant, the mixture becomes lean and the flame velocity is lowered. Eventually the mixture will become too lean to ignite. Therefore, in practical operation the load of the conventional spark-ignition engine is regulated by varying the mass of the mixture contained in the cylinder rather than by varying the quality of the mixture. Maintaining a relatively constant ratio of fuel to air keeps the combustion rate at acceptable levels. Controlling an engine by varying the mass of the mixture produces considerable negative work at part load during the inlet phase of the cycle, and the more or less constant high temperature of the combustion process in the engine tends to keep heat losses high. Both of these factors reduce engine efficiency. Researchers have attempted to overcome these disadvantages by using lean fuel-air ratios and adjusting conditions in the cylinder for their successful combustion. The problem has received sporadic attention since about 1920, and there have been some notable advances, such as the well known Texaco combustion process. Examination of the present work supported by A.T.A.C. together with other information reveals that two basic methods of approaching this problem are in use today. These methods, which blend slightly in some of the proposed designs, are described below. A. LEAN FUEL-AIR RATIO ENGINE A variable fuel-air ratio from rich to lean is used to regulate the load. The mixture is relatively uniform, filling the cylinder at all loads without throttling. To date no method for achieving combustion as the mixture passes out of the normal ignitable ratios has proved satisfactory. One possible ignition method uses. the combustion of a small quantity of rich mixture in a separate cell, or equivalent, connected to the main chamber by an orifice. This rich mixture is readily combustible, and the pressure resulting from its combustion discharges a multitude of ignition sources into the lean mixture in the main cylinder, producing a large number of combustion zones. 2

The speed of combustion is increased by having a number of sources from which combustion spreads. Normal combustion uses only one flame front. B. STRATIFIED-CHARGE ENGINE As in the first method, the cylinder is filled without using a throttle. Here the main charge is air. A pocket of fuel-air mixture of the normal combustible strength for rapid flame travel is produced in one of a number of ways, and the size of this pocket is varied according to the engine load. The aim is to produce a complete cylinder charge of the rich mixture at maximum load as in a normal spark-ignition engine, and then to reduce the amount of rich mixture as the load is reduced. Thus at half load the cylinder is filled with half combustible mixture and half air, and when the cylinder is idling, it contains a very small pocket of combustible mixture surrounded with a large volume of air. The Texaco combustion process uses this method of operation. It is, of course, difficult to produce these two processes without some overlap, but we believe that these two simple methods represent the ideals for the various researchers. We examined the engine performance possibilities for these two ideal cases: (1) Lean Fuel-Air Ratio Engine: The cylinder is filled with a charge consisting of a uniform mixture of fuel and air, varying from F/A = 0.00, no fuel mixture at zero load, to F/A = 0.0782, a rich mixture at full load. (2) Stratified-Charge Engine: The charge consists of a cylinder full of air in which a pocket of combustible mixture is produced, varying in size from a minimum of mixture, approaching zero at no load, to a completely filled cylinder of mixture, at full load. 5

III. METHODS OF CALCULATION A. LEAN FUEL-AIR RATIO ENGINE The conventional method of theoretical cycle prediction was used for all cases with a normal uniform mixture of fuel and air. Charts are available for the various combustion mixtures over the combustible range, from approximately F/A = o.o6 to F/A = 0.0782 1 These charts provide for varying specific heat, chemical dissociation, etco For leaner ratios such charts are not available, but the temperature of the cycle and dissociation are known to decrease rapidly. The varying specific heat then becomes the major factor. Lean fuel-air ratio operation was therefore examined by using charts, whenever possible, and by using specific heat data,2 at the leaner ratios. Bo STRATIFIED-CHARGE ENGINE The stratified-charge engine produced another set of problems. In order to have the small combustible pocket in contact with the spark plug at the end of compression, the pocket must not mix with the main bulk of the air charge. Many inventions are concerned with the various processes for achieving this. An ideal engine cycle of this type would be one in which the combustible mixture was completely isolated from the remaining cylinder contents at all times and could be fired at will, with little or no mixing of the combustion products with the remaining air. Although this is an unattainable ideal, it would illustrate the optimum performance for the stratification system, This cycle was examined by keeping the air and mixture separated but in pressure communication with one another at all times, producing uniform gas pressure throughout the engine cylinder at all points of the cycle, Thus two gases with differing properties were to be compressed and expanded simultaneously with pressure equilibrium in the same cylinder. The stratified-charge engine can be visualized as an opposed piston engine in which the individual pistons can move as desired but the overall compression ratio remains constant. In such a cylinder a floating divider, leakproof but movable, maintains the same pressure throughout the gases, as represented in Figure 1, where V1 = volume of (1 + F/A) lb of mixture in cu ft x = volume of mixture in the cylinder 4

1 - x = volume of air in the cylinder F/A = fuel-air ratio Therefore, when x = 1, the cylinder volume is V1 cu ft under maximum power conditions containing (1 + F/A) lb of mixture. Under these conditions the ideal cycle can be determined directly from the combustion charts. Air Mixture Air / - / Mixture Piston x ---- V1 cu ft Figure 1. Diagrammatic sketch of imaginary cylinder for stratification. At part load the cylinder will contain x (1 + F/A) lb of mixture occupying x (V1) cu ft at a pressure and temperature that is the same as that for a full load. In addition the cylinder will contain a volume of (1 - x) V1 cu ft of air at a density determined by the ambient inlet pressure and temperature. These two elements of the charge are divided by the partition. C. COMPRESSION It follows that when 0 < x < 1.0, there are two separate gases present, each having its own thermodynamic properties. Air compression will follow the air tables2 using variable specific heat, and fuel-air mixture compression will follow the compression chart1 for that mixture ratio. At top dead center the pressure reached will be the same throughout the total clearance volume, but the compression ratio of each gas will be somewhat different, depending upon the individual gas properties. The condition to be fulfilled is na na Pi V = P2 Va (1) or n n PI Vg^ = P2 Vg (2) 5

where a = air g = gas mixture Pi = pressure at end of inlet P2 = pressure at end of compression na = index of compression (average) for air ng = index of compression (average) for gas The compression ratio for air is Val Ra () Va2 The compression ratio for gas is Vg1 Rg = v (4) Vg2 The compression ratio of the L-141 engine, which is being used in the AoToAoC. tests, is 7.5:1. This value will be employed in this study, though it can be changed. For this engine: R = overall compression ratio for the engine 7~5 Va + Vg1 Va2 + Vg2 The average values to be employed for na and ng for the compression stroke were determined from the air tables and combustion charts for a ratio of 705:1 with average ambient inlet pressure and temperature. The applicable values were na = 1.58 6

ng = 1.355 At the same time the gas constants for air and the various mixture ratios were obtained2; they are plotted in Figure 2. From Eqso (1) through (4) the following is obtained: Pi (Ra)1 38 = Pi (Rg)10355 g -RP1.02 Rg = R a (5) If the cylinder volume V1 is that volume occupied by 1 lb of air plus fueL at inlet conditions, i.e., x = 1.0 or full load, the clearance volume is For a rich mixture (F/A = 0.0782), an inlet pressure of 14.7 psia, and a termperLature of 600~R, V1 is 15.86 cu ft, and V2, the total clearance volume, is 2,115 cu ft. IfL b-oth air and mixture are present, i.e., 0 < x < 1.0, then: Total charge = volume of mixture + volume of air = 15.86 x + 15.86 (1 - x) cu ft Px+F/A) 1 x 144 x 15.86 (L - x) = x (1 + F/A) + - ---- Ib 553.34 x Ti whe.re P1 = 14.7 and T1 = 600~Ro At the end of the compression stroke Volume of air = 5.86 (1 - x) Ra Volume of mixture 15.86 x Rg 7

54.0 -X U x 52.0! I 0.0 0.02 0.o4 o.o6 o.o8 F/A Ratio Figure 2. Gas constants for various mixture ratios.

Also, Va2 + Vg2 = 2.115 15.86 (1 - x) 15.86 x 2 + = 2.115 (6) Ra Rg and 1.02 R R Rg a Substituting for Rg in Eqo (6) and simplifying, we get 7.5 (1 - x) RO 02 + 7.5 x = Ra 02 R (7) Values satisfying Eq. (7) are shown in Table I and Figure 3. TABLE I VARIATION OF RATIO AS x VARIES x Ra Rg 1.0 7.21 7.5 0.9 7.24 7.53 0.8 7.265 7056 0.6 7.30 7.62 0.4 7.36 7.682 0.2 7~42 7.74 0.0 7.5 7.81 -With the aid of Figure 3 it is now possible to obtain the change of state, work done, etc., on both the air and mixture contents of the cylinder, assuming they are separated at all times. 9

7.8 ~ 7.6 *g~~~~~~~~~~~~~~~ I ^gxture at, 7.40 co.01 Air R 7.0 o;.0 0.. 0.0 1 0.6 Ion Of Mixture 0. 0.5 0.40, ^^e,,, *~0-5 0.2 (*4 Figure 0- 0 anamixtu~e Com.Pre 0"a s s I ^8^ ~^,ura los for air (R ure (I~g, overal ra-0 7..5: ia

D. COMBUSTION At the end of the compression stroke, combustion occurs in the x fraction of the mixture, which expands under the heat addition, compressing the air. The final pressure P5 is reached at top dead center. This pressure P3 can be obtained by a trial and eirror process using the combustion charts. (See example below.) The values for P3, T3, etc., are then available for both the burnt mixture and the air. E. EXPANSION The gases now expand, forcing the pistons back to bottom dead center. The final state, with a uniform pressure P4 psi, is shown in Figure 4. ExP4 P4 a4 Vg4 15.86 --- Figure 4. Diagram of conditions at BDC of expansion stroke. panding the air and mixture separately, in a manner similar to that of compression, yields P Va Va (8) 5 a3 P4Va4 and Pw V a= Pv 4 Vg (9) where na and ng are two new average values for the index, appropriate for the changed temperature and pressure conditions. As before, the relation between the ratio of mixture and air becomes 11

Rg = (Ra) g (10) Since the overall expansion ratio is still 7.5:1, Va4 + Vg4 7.5 = -- Va3 + Vg3 and 15.86 = Va Ra + V Rg 15.86 = Va3 Ra + Vg3 (Ra) ng (11) an equation in Ra and known values for Va,5 etc. The magnitudes of na and ng will differ from those of Eq. (5) because of the change in the temperature and composition of the gases. Calculations indicate that na = 1.356 and ng = 1.244, making na/ng = 1.091. These values cover the expected range with sufficient accuracy. It follows that Eq. (11) can be expressed as 15.86 = Va5 R + Vg (Ra) 091 (12) Therefore, when Va5 and Vg3 have been obtained after the combustion process, the value for the ratio of expansion of the air can be obtained, and the ratio of expansion of the gases can be obtained from Eq. (10). The work of expansion can be obtained from the combustion charts for the gases and from calculations for the air by using the equations: P3V - P4V Work of air during 3 a 4 l (13) == -' a- ft lb (13) expansion na - 1 and na na P Va P4 Va14) 12

IV. SAMPLE CALCULATION FOR THE STRATIFICATION SYSTEM The following typical case represents the detailed calculations involved. The cycle will be examined for x = 0.9, with a manifold pressure of 14.7 psi, a temperature of 600~R, and a rich fuel-air ratio, F/A = 0.0782. The volume of the containing cylinder is maintained at 15.86 cu ft as required for x = 1. This case, therefore, represents a reduced engine load at rich mixture. The indicator diagram of Figure 5 represents the overall cycle. c) 2 U) Volume Figure 5. Indicator diagram when x = 0.9. State 1 P1 = 14.7, T1 = 6000R, x = 0.9, V1 = 15.86, F/A = 0.0782 Volume of mixture = 0.9 x 15.86 = 14.274 cu ft Volume of air = 0.1 x 15.86 = 1.586 cu ft Wt of mixture = 0.9 x 1.0782 = 0.9704 lb Wt of air = 14.7 x 144 x 1.586 = 0.1048 lb 53.34 x 600 From Figure 3 Ra = 7.24 and Rg = 7.53. 13

Compression of Air Pa = Pal (Ra)1'38 = 14.7 x 15.40 = 226 psi PaX 1/1.38 Va2 = Va ( = 1.586 x7.24 1 ^Va2 7.24 = 0.219 cu ft Vg1 V -- -= 1.896 cu ft g2 7.53 Total clearance volume = 1.896 + 0.219 = 2.115 cu ft The calculated clearance volume must equal the designed clearance volume. Checking shows that Designed clearance volume = 1.86 = 2.115 cu ft 7.5 Compression of Mixture (by charts) P = 14.7, V1 = 15.86/lb of air, T1 = 600~R, E = 15.0 Vg = 15.86 =2.108/lb of air = 1.896 (x = 0.9) 2 7.53 S2 = 0.088, E2 = 1.37, P2 = 224 psi, T2 = 1146 Pressure of air = 226 Pressure of mixture = 224 An average of the compression values for the air and the mixture of 225 psi is used for this operation. 14

Compression State 2 Air Mixture Pa = 225 Pg2 = 225 Va2 = 0.219 Vg2 = 1.896 Ta = 1273 Tg2 = 1146 Wa = 0.1048 lb Wg = 0.9704 lb Combustion State 3 Combustion now occurs and a pressure of P3 is obtained. Then Pa = Pg = P It follows that for the air VH 1.356 a2 Pa a2 Va ) therefore Va5 = (225 x 0.219 a 35.7Pa 38 11.83 (15) p0.737 Pa Various values for Pa are assumed until equilibrium with the gases at the end of combustion is obtained. The final value for Pa3 is 950 psi, which will be used to illustrate the method. When Pa = 950 psi, Va = 0.0759 cu ft,by Eq. (15). Therefore, Vg = 2.115 - 0.0759 = 2.0591 cu ft for the 0.9 lb of air in the combustible mixture. The chart volume (for 1 lb of air) then must be 2.267 cu ft/lb of air. The volume of the mixture must match the above 2.267 cu ft to assure that the correct pressure has been assumed. The mixture expands as combustion compresses the air and does work dW 15

on it. The heat added by combustion is the calorific value of the fuel, 1507 Btu/lb of air. By the first law of thermodynamics dQ = dE + dW Pa Va - Pa2 Va2 144 1 dQ = dE + x - x - n-1 778 x 1507 = dE + 950 x 0.0759 - 225 x 0.219 144 1 0.356 778 0.9 dE = 1507 - 13.0 = 1494.0 Btu/lb of air therefore E = E + 1494 g3 g2 = 137 + 1494 = 1631 Btu The intersection of Pa3 = 950 and Eg5 = 1631 on the rich mixture charts shows that the volume of the mixture must be 2.27 cu ft, which checks closely with the desired 2.267 cu ft; 950 psi is accepted as the pressure at point (3) of the cycle. Expansion The gases and air now expand from P3 to P4. During this operation Eq. (12) must be satisfied. 15.86 = 0.0759 Ra + 2.0391 (Ra)1.091 (12) 16

Equation (10) shows that Ra = 6.37 and Rg = 7.53. It follows that Va4 = 6.37 x 0.0759 = 0.485 cu ft Vg4 = 2.0391 x 7.53 = 15.365 = 17.06 cu ft/lb of air Total volume = 15.36 + 0.485 = 15.850 cu ft Checking the required 15.86 closely shows that Pa 15 = 77.8 psi al - a ) 12.2 The mixture values on the chart are extended from state 3 to state 4 to obtain the properties: Pg5 = 950, Eg = 1632, Sg = 0.571, V = 2.27 therefore, S4 = 0.571, Vg4 = 17.06, Pg4 = 77.0, Egg = 982 The average P4 of 77.4 psi is used. Net Work of Cycle Work of compression = work on air + work on mixture 225 x 0.219 - 14.7 x 1.586 144 0.38......' x1-7 + (137- 15) 0.9 = 12.4 + 109.9 17

= 122,3 Btu Work of expansion = work on air + work on mixture 950 x 0.0759 - 77.4x:o.485 144 x - + (1632 - 982) 0,9 = 17.9 + 585 = 602.9 Btu Net work of cycle = 602.9 - 122.3 480o.6 Btu 480.6 x 100 Cycle efficiency 4 6 x 1,507 x 0.9 35.4% IMEP _ — ^work change of volume 480.6 x 778 (15.86 - 2.115) x 144 = 189.5 psi IHP 480o6 x 778 550 679.5 hp/0.9 lb air/sec 755 hp/lb air/sec Sp^ 0o.0782 x 3600 x 0.9 679.5 - 0o 5753 Ib/IHP/hr 18

V. RESULTS Repeating this calculation for all of the desired values of x produces the data in Table II, plotted in Figure 6. TABLE II PERFORMANCE OF IDEAL STRATIFIED-CHARGE ENGINE WITH RICH MIXTURE WHEN x = 1.0 (Compression Ratio 7.5:1, F/A = 0.0782) Mixture, Air, Pmax, Net IMEP, IHP Thermal SFC X lb b psi Work, psi lb air/sec Eff. lb/IHP/hr Btu 1.0 1.0782 0.0 1040 552.0 209.0 752.0 -- 0.374 0.9 0.9704 0.1048 950 480.6 189.5 679.5 35.6 0.373 0.8 0.8625 0.2097 890 430.1 169.5 609.0 35.4 0.370 0.6 0.6469 0.412 730 418.2 125.2 450.0 35.6 0.375 0.4 0.431 0.617 575 214.0 84.2 303.0 35.2 0.373 0.2 0.2153 0.823 410 117.8 46.3 166.5 35.4 0.338 0.0 0.0 1.03 240 - - - 39.1 - Compression Expansion Corrected Data Corrected Data x Ratio Ratio Mixture Air Mixture Air IMEP IHP Eff. SFC 1.0 7.50 -- 7.5 - 209.0 752.0 35.6 0.374 0.9 7.53 7.24 7.53 6.37 188.0 675.0 35.3 0.375 0.8 7.56 2.27 7.6 6.4 167.0 599.0 35.4 0.376 0.6 7.62 7.3 7.72 6.53 125.2 450.0 35.2 0.376 0.4 7.682 7.36 7.88 6.62 83.5 300.0 35.1 0.375 0.2 7.74 7.42 8.18 6.86 41.7 149.8 35.2 0.376 0.0 -- 7.5 -- 7.5 - All of the data in Table II are for a constant F/A ratio of 0.0782, a rich mixture, ensuring positive ignition and high flame speeds at all conditions of operationeven if turbulence should cause some dilution of the charge. Lean mixtures of o.o6 fuel to air also ignite readily in the spark-ignition engine, producing improved economy. Charts were available for fuel-air 19

F/A 0.0782 o.o665 1000 500 o. 0o6o5 PX - 800 Pmax 4001 -- 4oo |4ooe~~ /Output Btu/ SFC F/A = 0.0782 200 4.38 — 3oo He ~ H 0 - 0.34 SFC F/A =.0.0665 o 300 0.3 0 SFC F/A = 0.0605 // 00 Fuel Flow lb/hr 0.0782 200 m/Fe lo o5"^^H 00,^..-4>a4^'s^ IMEP psi r0.2 0.4 o.6 0.8 1.0 Fraction of Mixture Figure 6. Ideal performance of a stratified-charge engine. 20

mixture ratios of 0.0665, the correct mixture, and 0.0605, a lean mixture. Calculations were performed when x = 1.0, using each of these mixtures, with the results given in Table III. TABLE III PERFORMANCE OF IDEAL STRATIFIED-CHARGE ENGINE WITH LEAN MIXTURE WHEN x = 1.0 F/A = 0.0605 F/A = 0.0665 Output 495 Btu 527 Btu IMEP 194.5 psi 207 psi IHP 700/lb air/min 746/lb air/min Fuel Flow 0.311 lb/IHP/hr 0.321 lb/IHP/hr SFC 960 psi 1000 psi Pmax 218 lb/hr 239 lb/hr The straight-line relationships obtained for the rich mixture as x varies are expected to apply to the lean mixture conditions also. Therefore, joining the points of Table III to zero by straight lines in Figure 6 described the performance to be expected from these lean mixtures for various values of x. A small triangular area is produced inside of which satisfactory engine performance can be expected. For a stoichiometric ratio of 0.0665, the differences in IMEP and output for the rich and the correct mixtures were too small to be shown on Figure 6 without greatly expanding the scale. For most practical purposes the F/A ratio values of 0.0782 and 0.0665 give identical data for these two parameters, also shown in Figure 7. If x = 1.0, 5532 and 527 Btu are released in the form of work for F/A = 0.0782 and 0.0665, respectively. The principal change resulting from the reduced F/A is a reduction of SFC. Figure 6 thus records the ideal performance for a stratified-charge engine using conventional values for F/A. The uniform mixture calculations for the lean fuel-air ratio engine were made by using the gas charts1 for mixture ratios of 0.0782, 0.0665, and 0.0605. At lower values for F/A dissociation was neglected, and variable specific heat data were employed. The results obtained are recorded in Table IV. The cylinder displacement employed was constant at the same volume as that used for the stratified-charge engine, i.e., 15.86 cu ft per lb of air, when F/A = 0.0782 with an inlet pressure of 14.7 psi and a temperature of 600~R. Tables II and IV contain corrected data obtained by using the values from the curve drawn through the plotted points to eliminate the effects of the trial-and-error process. 21

x = 1.0 1000 500 P- / X800o. Po 600 ~ 0. 38 f Output Btu/ / 400 -0./ 36 400o 0-.52 f 7 - 200 P *SFC b/IHP/hr 5300 -0.28 oo1 300..P Fuel Flow lb/hr. | 200 / 200 $ X / X^ IMEP psi 100 /- 100 0.01 0.02 0.03 0.04 0.05 o.o6 0.07 0.08 F/A Ratio Figure 7. Ideal performance of a lean fuel-air ratio engine. 22

TABLE IV IDEAL LEAN FUEL-AIR RATIO L-141 ENGINE (Compression Ratio 7.5 1) Net Thermal F/A IMEP, IHP, SFC, Pmax. Corrected Data Work Eff. Ratio Btu psi lb air/sec lb/IHP/hr p IMEP IHP Eff. SFC 0.0782 532.0 209.0 752 0.374 35.6 1040 209.0 752 35.6 0.374 o.o665 527.0 207.0 746 0.321 41.2 1000 205.0 739 40.8 0.324 o.o605 495.0 194.0 700 0.311 42.5 960 194.6 703 42.7 0.310 0.05 403.8 159.0 570 0.316 42.0 890 165.5 592 43.7 0.304 0.0333 289.0 114.0 409 0.293 45.1 677 114.0 409 45.1 0.293 0.025 241.0 94.8 341 0.264 50.2 621 85.4 308 45.3 0.292 0.0167 150.0 59.0 212 0.284 46.9 496 57.8 208 45.9 0.290 0.0 - - -- 240 -

VI. DISCUSSION The data of Figures 6 and 7 show the relative effects of engine operation under the two systems examined. We will now discuss the implications for practical applications of these systems. For the system for the stratified-charge engine we assume that a definite quantity of mixture, of a known combustible ratio, is available at the spark plug. The initiation of combustion is assured under all conditions, and the load is controlled by varying the quantity of this mixture in the pocket from a very small amount at idle to a full cylinder volume at full load and speed. The mixture ratio remains constant in the case examined; however, this ratio can be leaned out until ignition becomes critical. In the lean fuel-air ratio engine the mixture ratio of the charge is changed continually from a very lean mixture at engine idle to a rich one at full load. It is well known that lean mixtures do not ignite and burn fast enough to be useful in a practical engine. Therefore, successful use of this system requires a means of accelerating the combustion process so that its time to completion is acceptable. We have assumed that such a process will be developed to a satisfactory degree. The torch ignition process is one example. Figure 6 shows that for the stratified-charge engine the mixture ratio used when x = 1.0, i.e., at a full charge of combustible mixture, was varied over the well known combustible range from rich, F/A = 0.0782, to lean F/A = 0.0605. Actually fuel-air mixtures of 0.05 to 0.055 can be ignited readily, but they are not generally used in an engine because of the critical nature of their ignition and the unequal distribution of fuel between the individual cylinders of a multi-cylinder unit. The data of Figure 6 indicate that although x, the proportion of mixture, varies, the SFC remains practically constant when the F/A is constant. The SFC lines for each F/A indicate that there is no gain in fuel economy on an IHP basis with a stratified-charge system. Less fuel is consumed in such a system, however, as the F/A is leaned out and x = 1.0, just as in a conventional engine. Since the power is controlled by varying the mixture and not by using a throttle, the pumping losses of a throttled engine do not exist, and an overall gain on a BHP basis is made by eliminating the negative work of engine charging, resulting from a considerable reduction in frictional losses at light loads. The magnitude of the gain would depend upon how lean the mixture could be made without sacrificing satisfactory combustion. This will be examined in detail later. Similarly, the lean fuel-air ratio engine uses no throttle, and thus achieves a gain equivalent to that of the stratified-charge engine. Figure 7 24

also indicates a considerable reduction in SFC per IHP for the lean fuel-air ratio engine. This method of operation therefore seems superior to stratification. It must be remembered, however, that the results presented were based on the assumption that satisfactory combustion is maintained over the whole range of the mixture ratio employed. To achieve these results, lean fuel-air mixture ratios of about 0.02 to 0.025 must be made combustible. If this can be achieved, an SFC approaching that of the compression-ignition engine can be secured. Comparing the date of Figures 6 and 7 shows little difference in engine performance, except ini 3FC, between the two systems. The stratified-charge e.gine at F/A = 0.06Od delivers 200, 300,and 400 Btu for a Pmax of about 550, 680, and 820 psi; and the lean fuel-air ratio engine gives the same outputs for 600, 720, and 860 psi. However, the SFC is constant at 0.31 lb/IHP/hr for the stratified system and varies from 0.29 to 0.50 for the lean F/A system, Combining Figures 6 and 7 into one allows both. systems to be examined simultaneously. Such a plot is shown in Figure 8. This diagram is based upon the straight-line plot of Figure 6. When x is constant and equal to 1.0, power output is reduced and economy is improved as F/A is reduced. Therefore, in Figure 8 when x = 1.0, the data of Figure 7 can be plotted from F/A = 0,0782 to F/A = 0.0167. Joining these points for the lean ratios to the origin produces a plot equivalent to Figure 6 for each F/A. At the same time, by making the vertical axis one of relative power and IMEP, i.e., the ratio between the power or mean pressure at any desired F/A ratio to that power or pressure at stoichiometric ratio, this one diagram gives output, IMEP, and the SFC curve for both stratified and lean F/A systems. A. STRATIFIED-CHARGE ENGINE To use Figure 8 when x = 1.0 and F/A = oo0605, locate the point of intersection of x = 1.0 with F/A = 0.0605. A horizontal line drawn from this point to the vertical axis (shown as dotted) gives the SFC at the intersection of this line with the fuel curve, i.e., 0.312 lb/IHP/hr. Proceeding to the vertical axis gives the relative hp and IMEP of 0.952. The values for the stoichiometric mixture, i.e., F/A =.Oo665, are listed on the diagram as basic hp = 739 and IMEP = 205 psi. Therefore the conditions for the ideal cycle specified are IHP = 703/lb air/sec, IMEP = 195, and SFC = 0.312 Ib/IHP/hr. For a stratified system when x / 1o0, say x = 0.8, and F/A = 0.0605, Figure 8 gives Relative ratio = 0o762 SFC = 02,,98 25

F/A 1.1 0.0782 1.0 Xo.o665 - - 0 ___ / - - o.06o5 0.9 (a) Performance oA 0.8 U X 0.05 F/A 0. OO 300 0.0782 - (c) Fuel Flow 0.7 0.0665 -\ 0.6 omo6os ------- - -.- - 0.0888 0.2 06052 0-4 ~.6 o.8.0 Stratified Fraction of Mixture (x) 0.3 _ *_3O00^^^\\\S^( Mtr atified Fraction of Mixture (x)sure eCompression Ratio = 7pr5fo1 c t 0.5 2 o.o6 0.4- -00 -. 0~o o.9 — 900 \\\ o -05 0.261.O 0.8 0.6 0.4 0.2 0.2 0.4 0.6 08 Compression Ratio = 7.5:1 Basic IMEP = 205 0.4 — 4oo 0.6 — 600 -o.025 - 0.0533.0782 Figure 8. Combined performance chart.

Th?1u IHP = 565/lb air/sec, IMEP = 156.2, and SFC = 0,298 lb/IHP/hro B. LEAN FUEL-AIR RATIO ENGINE In this case x = lo0 at all times, since the cylinder is filled with a uniform chargeo To find the performance when a lean F/A of 0.033 is used, proceed from the intersection of x = loO and F/A = 0.033 to the intersection with the fuel curve at SFC = 0.293, and on to the vertical axis. where the relative performance is given as 0.555. Then IHP = 0.555 x 759 = 410/lb air/sec SFC = 0.293 lb/IHP/hr IMEP = 0.555 x 205 = 114 psi By combining two other graphs for the parameters Pmax psi and fuel flow in lb/hr with Figure 8, these parameters also become available, The essential performance factors for any ideal stratified or lean fuelair ratio engine operating at a 7~5:1 compression ratio are available when there is 100% volumetric efficiency and when there are no heat losses or residuaL gas. However, the specific heat may vary and dissociation can occur. It is interesting to compare these results with those obtained from the conventional ideal thermodynamic cycle efficiency equationo 1-k Ideal cycle efficiency = - R R = Compression ratio k = Ratio of specific heats Thbis equation uses variable values for k, already determined, based upon the average temperature for the cycle. As the mixture ratio varies, kc for the compression stroke varies from 1.55 to 1o3555 and ke for the expansion stroke changes from 1o25 to 1o28, In the thermodynamic case for stratification, F/A is held constant at 0o0782,^ and hence k also remains constant for the mixture pocket. At a constant compres-ion ratio of 7o5 the cycle efficiency was 45o7%. iigure 9 shows the efficiencies of Table IIIo 27

50 Thermodynamic Cycle w 45 cH 40 CH I _tg~)~ _ ~Engine 3 35 _ * - X-X-X — Constant F/A = 0.0782 0I I I I 0.2 0.4 0.6 0.b 1.0 Stratified Fraction Figure 9. Engine efficiency for stratified-charge engine. In addition Figure 9 shows an ideal cycle efficiency for the actual varying compression ratio, Rg, of the stratified cycle when combined with an average value of k for each cycle. The value for cycle efficiency shows only sliiht variation. The third curve for a stratified-charge engine cycle illustrates a difference between the thermodynamic cycle and engine efficiencies that is also relatively constant. This is to be expected, since the F/A and the compression ratio are almost constant. For lean F/A, as shown in Figure 10, the effects of dissociation, etc., at rich mixture settings are apparent, but the thermodynamic cycle and the practical engine cycle agree fairly closely when leaner mixtures are employed. 50.co ~~ ~~~~~~~~~~~~~Cycle 45 U X X' --.H U40 |.< Engine co U 35 0o I I I l 0.02 0.04 o.o06 0.08 F/A Ratio Figure 10. Engine efficiency for lean fuel-air ratio engine. 28

To compare these data with those for the L-141 engine, a best estimate of the frictional losses of the L-141 engine and of the SFC/IPH/hr were made. This estimate is shown in Figure 11, together with that for the two engines being examined. The solid-line L-141 plot is of an actual engine performance based upon gross IHP, and the solid-line curves for stratification and lean fuel are the ideal curves predicted for those engines. To make an approximate comparison the L-141 line was displaced to pass through the same SFC point, 752 hp/lb/sec, as the other two curves at full load. The dotted L-141 curve obtained is labeled ideal L-141. This curve follows the lean-fuel ideal curve fairly closely down to an F/A ratio of about 0.05, the leanest mixture usually employedo Below this point SFC increases rapidly as throttling occurs, and t:.T mixture is richen(a somewhat for smooth idling. The difference between the dotted curve for the ideal L-141 and the lean F/A curve indicates the approximate penalty due to throttling and the need for stable operation. Generally speaking, the best way to compare cycles is to compare indicated data as we have done so faro However, the BHP data, which are of major interest to a purchaser, had to be calculated for the stratified-charge and lean-fuel engines from results based upon known losses in net output for the L-141 engine. Thus the net performances of the L-141, stratified-charge, and lean F/A engines are compared on a BHP basis. Losses from the air cleaner, generator, cooling fan, and exhaust system are included for a speed of 3600 rpm. Figure 12 shows the BHP performance calculated for the ideal stratifiedcharge and lean-fuel cycles and the actual performance of the L-141 engine on the basis of the equivalent BHP for 1 lb air/seco These results are shown by the solid-line curves, If the L-141 engine results are assumed to be for a relative cycle efficiency of 80%, a value usually exceeded in most engines, the dotted line marked 80% would represent the ideal L-141 engine performance. If the full-load SFC of the L-141 engine is reduced to equal that of the ideal cycles examined, a relative efficiency of 70o8% is obtained. The second dotted line then represents the ideal L-141 cycle on roughly the same scale as the other cycles under consideration. A relative efficiency of 70.8%, therefore, must exist for the L-141, a rather low value for a modern engine; this low efficiency value could result from low volumetric efficiency, poor manifolding, or high manifold temperature. The SFC of a stratified-charge engine with a constant F/A of 0.0782, then is inferior to that of the normal throttled engine over an appreciable portion of its output range. Not until 1/3 load is reached is any gain in economy made. However, the lean F/A engine has superior performance over the whole load range. The F/A of 0.0782 was selected for the stratified-charge engine to give a mixture rich enough to assure ignition even with some dilution due to air turbulence. We assume that a leaner mixture could be used while maintaining ignition in the spark plug regiono If a ratio of 0.05 could be used, the strati29

0 L-141 0.44 Stratified Charge X Lean F/A 0.42 0.40 0.38 Stratified ^o~6_X -X X ~xx o.28 _ Figure \ Comp onwL-141 / 0.34 03 \ Ideal L-141 / F/A = 0.05 O/ 0.32 0 0.30 Lean F/A 0.28 100 200 300 4oo00 500 600 700 800 HP lb/air/sec Figure 11. Comparison with L-141 engine. 350

1.0 \ Stratified A Equivalent L-141 0.9 X Lean F/A 0.8 0.7 i-. \\\\ V / L-141 Engine -0.6 o \ 0.5 Stratifie /Rel. Effy. 80% Stratifiedl 0.4- ~y X Rel. Effy. 70.8 Lean F/A Cycle xx X F/A = 0.05 0.50.2 100 200 300 400 500 600 700 800 Net BHP/lb air/sec Figure 12. Net BHP performance at 3600 rpm, including air cleaner, generator, cooling fan, and exhaust system. 51

fied-charge engine efficiency would then follow the lean fuel-air ratio engine efficiency line down to F/A = 0.05, on to about half load, and along the dashdot line of Figure 12 for lower engine powers. There would then be little difference between the two imaginary cycles, and it would be immaterial whether the stratification or the lean fuel-air ratio method was used. The problem of igniting very lean fuel-air ratios has so far been ignored, since it is assumed that some method for this exists. The method in common use today is to inject a small pocket of rich mixture near the spark plug; combustion then produces many sources of ignition throughout the lean mixture, resulting in satisfactory combustion of all of the fuel. But this method has been only partially success;ul in practice, and some throttling at low loads is usually necessary. Some increase in SFC of the lean-mixture curve of Figure 12 is therefore expected in any practical engine. The difference between stratified and lean fuel-air conditions, then, will probably be less than this diagram shows; in fact, considerable increase in fuel flow for lean F/A engines will occur if much throttling is required. A comparison of the net BHP per lb of air, as shown in Figure 12, with the actual net BHP output of the L-141 engine shows there is 656 hp/lb of air in the former and 62-63 BHP for the latter. The L-141 engine scale is, therefore, almost 1/10 that of the ideal, providing a convenient reference to both conditions. The four important curves of Figure 12 are redrawn on a separate diagram in Figure 13 to summarize this investigation. Curve 1 shows the actual L-141 engine results for the assumed relative efficiency of 70.8% and for full-load conditions equal to those of the ideal cycles examined. Curve 2 shows the stratified-charge cycle for a rich mixture, 0.0782, throughout the whole power range. Curve 3 shows a stratified-charge cycle in which the load is reduced first by leaning the mixture from 0.0782 to 0.05, and then by stratification. For the cycle in curve 4 the load was controlled by using lean mixture ratio only. The usable portions of each of these curves must now be determined. Since the data for curve 1 were obtained experimentally from an actual L-141 engine test, the whole curve represents possible engine operation. The rich-mixture stratification results shown in curve 2 ensure ignition of the charge fraction containing the fuel at all times, unless turbulence should cause bad mixing of the air charge with the stratified charge. All of this curve could probably be used. The stratified-charge engine is controlled by varying the fuel mixture down to the 0.05 ratio and then by varying the magnitude of x. This cycle as shown in curve 3, is also possible. Tests have already confirmed that mixtures of 0.04 are ignitable under such conditions, and the extra margin given by F/A = 0.05 should ensure regular ignition even if some mixing with the air occurs. 32

1.0 1 0 LL-141 Engine _0.~9 k \ O0 Q Stratified Charge Rich Mixture 0.9 \O Stratified Charge Lean Mixture Q o.8 _ \ \ X Lean Fuel-Air Ratio 0.8 \ L-141 0.7 Str fi~ Stratified X 0.5 L-141 Engine Output 0.4 10 20 0 40 50 60 70 100 200 300 400 500 600 700 Net BHP/lb air/sec Figure 13, Comparison of cycles on BHP basis. 5533

The lean-fuel mixture, as shown in curve 4, cannot be completely controlled without some throttling. Fuel-air ratios of 0.04 to 0.035 do appear to be possible, judging from some rather early tests, but at and below this point some throttling was necessary. Therefore, curves 1, 2, and 3 represent possible operating conditions, and curve 4 can be employed down to about 300 net BHP per lb. We conclude that the area between curves 2 and 3 represents possible operation, with curve 3 being about the best we can expect. From full load to about 75% load, the gains in fuel economy from these ideal engines are quite small, in fact negligible, since each ideal engine under these conditions operates with a full cylinder charge at a reduced F/A. Below 75% load the stratified-charge engine shows some advantage. At about 25% load the SFC of the stratified-charge engine has been reduced to 68% of that of the throttled engine, a saving of 42% in fuel consumption. This saving will increase at lower loads. The engines considered in this report do not operate at constant load and speed; hence a complete map would have to be predicted to accurately determine the savings under typical operating conditions, When the low load predominates over any given time interval the savings could approach 25% to 30%; however, for heavy loads the savings would be smaller. Some engines have been designed and tested using higher compression ratios than those of the standard engine. Any fuel gains from this type of engine design would add to the savings mentioned above. 54

VII. PRESENT STATUS OF RESEARCH Each of the two methods of control examined has two systems that have shown some degree of practical success. For the stratification method the Texaco combustion process, which forms a continually changing pocket of combustible mixture that burns with a minimum of mixing, appears to be the most advanced system and shows the best practical possibilitieso Engines of this type have been run for a number of years. Costs are higher because of the need for both injection and ignition systems, but fuel savings, etc., have been demonstrated. Another stratification engine, the Southwest Research engine, does not form such a clean cut pocket with its present design, and some mixing and leanfuel burning seem unavoidable. The maximum load, speed range, and SFC of this engine are not well established, The Walker system, a lean fuel-air ratio engine tested by Continental Aviation and Engineering Corporation, recently gave extremely good results developing about 85% of maximum load of the L-141 engine while giving a SFC at 20% load of 74% of that engine. A limited speed range was used in these tests. If similar results can be obtained at other speeds, a satisfactory engine should result. No injection system is involved. The second lean fuel-air system, tested at the University of Rochester, has demonstrated the combustion of lean mixtures with some reservations. More research on both of these systems is needed before a satisfactory engine will be available. 35

VIII. CONCLUSIONS As a result of this investigation it can be concluded that: 1. A stratified-charge engine should possess the widest range of power outputs without throttling the air. 2. Theoretically a lean fuel-air ratio engine would give the lowest SFC provided that satisfactory ignition and combustion can be assured at all F/A ratios. 3. The SFC of a stratified-charge engine closely approaches that of the lean fuel-air ratio engine over most of the load range. Only at low loads does the SFC increase. Ignition of a lean F/A ratio engine in this low range, however, has not been definitely established. 4. Control of power by varying the mixture ratio only should permit a power output almost equal to that of a conventional engine. 5. Stratified-charge engines will probably give a maximum output slightly lower than that of present carbureted engines. 6. The two methods of control considered here could result in consuming as little fuel as the compression ignition engine when operating at low loads. 7. The main reasons for the high fuel consumption of conventional sparkignition engines are the throttling losses and the need for a rich mixture for smooth idling. 8. The stratified-charge engine requires more research and development, eliminating turbulent mixing as much as possible, in order to increase cycle efficiency. 9. Lean F/A engines ignitied by torch should consume a very small quantity of rich mixture for ignition. 56

IX. REFERENCES 1. C. F. Taylor and E. S. Taylor, The Internal Combustion Engine, International Textbook Co., Scranton, Pennsylvania, 1938. 2. J. H. Keenan and J. Kaye, Gas Tables, John Wiley and Sons, New York, 1948. 57

UNIVERSITY OF MICHIGAN 3 9015 03527 2916