COMBUSTION AND PERFORMANCE CHARACTERISTICS OF METHANOL PLUS HIGHER ALCOHOLS IN A SPARK IGNITION ENGINE

Transcription

1 ILENR/AE-87/03 COMBUSTION AND PERFORMANCE CHARACTERISTICS OF METHANOL PLUS HIGHER ALCOHOLS IN A SPARK IGNITION ENGINE DEPOSITORY FEB UNIVtKSITY OF ILLINOIS AT URBANA-C"J<VMPA1GN Illinis Department f Energy and Natural Resurces OFFICE OF SOLID WASTE & RENEWABLE RESOURCES

2 UNIVERSITY OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN BOOKSTACKS

3 ILENR/AE-87/03 Printed: Nvember 1987 Cntract 7ALT COMBUSTION AND PERFORMANCE CHARACTERISTICS OF METTHANOL PLUS HIGHER ALCOHOLS IN A SPARK IGNITION EN3INE Final Reprt Prepared by: University f Illinis at Chicag Department f Mechanical Engineering Chicag, II Principal Investigatr ::-*y Dr. K.S. Patel Pr^ared Fr: Illinis Department f Energy and Natural Resurces Office f Slid Waste and Renewable Resurces 325 West Adams Street, Rm 300 Springfield, Illinis James R. Thmpsn, Gvernr State f Illinis Dn Etchisn, Directr Illinis Department f Energy and Natural Resurces

4 11 Additinal cpies are available by calling the ENR Clearinghuse at 800/ (within Illinis) r 217/ (utside Illinis) Statements and cmments expressed herein d nt necessarily reflect the view f the Department. Additinal infrmatin is available frm Manager, Alternative Energy Develpment Prgram at 217/ This dcument printed with U.S. DOE funds.

5 lii FOREWORD The transprtatin fuel industry is experiencing a time f change. Depletin f fssil petrleum reserves and increased csts f petrleum-derived fuels are frcing the industry tward alternative fuels. There is als an urgent need t develp cleaner-burning fuels t relieve the pressures f air pllutin n ur envirnment. The best current ptin fr a future transprtatin fuel base appears t be methanl, a simple alchl that can be abundantly prduced frm natural gas, cal r bimass. When synthesized frm cal r bimass, the result is i mixture f methanl with small quantities f higher alchls, such as ethanl, butanl and prpanl. This methanl plus higher alchls mixture (MPHA), in cmbinatin with gasline and additinal ethanl, is likely t becme the transprtatin fuel f the future. With sme mdificatins, existing autmbile engines can use methanl mre efficiently than gasline and simultaneusly release fewer emissins. This prject develped base-line data n the physical and cmbustin characteristics f MPHA in mixtures with gasline. The data shuld prve invaluable t autmtive engineers fr designing mre-efficient and cleaner perating engines capable f using wide ranges f fuel mixtures. ENR is. pleased t have spnsred this research and hpes that the infrmatin prvides a useful base fr redefining ur transprtatin energy resurces. Dn Etchisn, Directr Illinis Department f Energy and Natural Resurces

6 S; ''' ^r-i-

7 / V > TABLE OF CONTENTS CHAPTER Page N. I Intrductin 1 II Experimental Setup 4 III Data Analysis and Sample f Calculatin 18 IV Theretical Analysis 28 V Results and Discussin 47 ( VI Cnclusins 103 VIII References 105

8 V.-'. f'-'.-'i"

9 CHAPTER I INTRODUCTION Methanl received wide attentin as a fuel fr the spark ignitin engine in pure frm r as a blend with gasline. Methanl tday can be prduced frm the natin's abundant resurces, cal, waste prducts frm farms and municipalities, as pssible additinal f eedstck[l-10]*. It is prbable that the synthetically prduced methanl fuel f the future will, in fact, be a mixture f methanl and higher alchls, rather than pure methanl. In the lw pressure methanl-prduct in prcess, the purity f methanl is achieved by cnventinal distillatin methds. The ther cmpnents are higher alchls which can be extracted at lwer levels f the distillatin clumn [12]. The use f methanl plus higher alchls instead f pure methanl reduces the prductin cst f the fuel, in general. The cmpsitin f alchl fuels derived frm cal and ther surces depends upn the manufacturing prcess, prcess perating cnditins, and prperties f the raw material [11]. The alchl blend which is representative f the prduct which culd be btained frm a methanl synthesis prcess using a nn-selective catalyst cntains by vlume, 75% methanl, 5% ethanl, 7.5% 1-prpanl, and 12.5% 2-methyl-l-prpanl [13]. This cmpsitin f alchls was used as the Methanl Plus Higher Alchls fuel. Because f different cmcsitins cf indlene-methanl blends and indciene-mpha blends, * Numbers in brackets refer t references in Chapter VII.

10 the physical prperties and perfrmance parameters need t be examined at varius perating cnditins. The fuel specificatins, are listed in Table 1.1. One f the mst imprtant prperties f methanl and MPHA is the high ctane number, which allws its use at higher cmpressin ratis, resulting in an increase in the thermal efficiency f the spark ignitin engine. The bjective f this wrk is t investigate the physical and chemical prperties f MPHAindlene blends, and the cmbustin characteristics. The experiments were perfrmed at varius methanl and MPHA cncentratins (frm t 100%) in the blend. The cmpressin rati was varied frm 5:1 t Knck Limited Cmpressin Rati (KLCR). The fuel prperties, perfrmance and cmbustin characteristics were als evaluated. In Chapter II, the experimental facility has been described. In Chapters III, IV and V, the data analysis, theretical analysis, and results are presented, alng with a discussin. Chapter VI cntains the cnclusins and recmmendatins.

11 TABLE 1.1 Puel In»p«ctin Data Prperty XadlMM Itthaal Bthal TwuprnMOl Imbbatmnl Fraula CH3OH C2H3OH C3H7OH CH3CH(CH3)CH20H Specific gravity 60»F/60*F Biling t«ap. Or Flash Pint Autignitin teap. 495 Vapr pressure at 68 *F. HMHG Flaaability limits lvar, % Vapr density (air - 1)

12 CHAPTER II EXPERIMENTAL SETUP Engine and Instrumentatin A standard single cylinder CFR spark ignitin engine with variable cmpressin rati, manufactured by Waukesha Mtr Cmpany was used in these experiments. The cmbustin chamber, with a flat cylinder head, is shwn in Fig One minr mdificatin was made t facilitate measurement f the flame speed. A spark plug with extended electrdes was placed in the magnetic pick-up plug psitin, and an inizatin prbe was placed in at the spark plug psitin. This was dne in rder t allw deep prbing int the cmbustin chamber, withut interfering with the pistn mtin. This engine was cupled t a General Electric CD-77 dynammeter, which can absrb 30 hrsepwer and deliver 25 hrsepwer. Fig shws the verall layut f the experimental setup. A Bendix breakerless ignitin system is used in this engine. This system cnsists f an ignitin cil, a cntrl unit, a spark plug cable, and a spark plug. An AC spark plug with extended electrdes was cut t allign it with the center line f the travelling prbe, as in Fig The intake air system cnsists f an air-filter, a laminar flw element (Meriam 50MW20-2), and an air heater. The abslute intake air pressure is measured at the intake f the laminar flw element, with an inlet abslute pressure gage (Meriam 310EF10). The air flw

13 TO^VIfW pntsivnt TMAMSOUCCll ^lomlzation MOSCS tpajik PLUG KLtcmc lonrzatlon MOBCS TOP AND SlOe VtCVS OW CMt COyfUSDON CHAMttn Fig. 2.1 Tp and Side Views f C.F.R. Cmbustir. Chamber

14 Air-fuel mixture Instrumentatin Systems Intake manifld Trigger Inizatin pyccny** transducer 10 C.A. 1 C.A Fly wheel Cylinder head Engine K) Cmpressin rati indicatr RPM indicatr Dynammeter Magnetic pickup lad pickup LOAD indicatr Enclsure Speed variatr Trque variatr Dry type transfrmei Swich Pwer surce Fig. 2*2 Overall engine and instrumentatin systems

15 ' : : /:. < - SlQg View lonrzatlon PROBES Fig. 2.3 Spark plug and its alignment with th^ prbe

16 rate is cmputed frm the pressure differential acrss the laminar flw element, with a differential pressure indicatr (Meriam 40HE35). The inlet air temperature is kept cnstant at 120 F, by an electric heater. Fig shws the layut f the fuel-air intake system. The engine speed was measures by a shaft encder (BEI- H A-7460R-EM16), and a frequency cunter. The engine lad was measured with a lad cell (Interface SSMAJ250) and a digital transducer-indicatr (Newprt Q2000SGR3). The lad cell is used t measure the engine trque and t find the minimum spark advance fr best trque (MBT). A high frequency respnse AVL pressure transducer type KQP300CVK was used t measure the cylinder pressure. This transducer is water cled, and is vibratin and temperature- cmpensated. It is used in cnjunctin with a Kistler 504E Charge Amplifier. Prir t measuring cylinder pressure, the transducer was cated with silicne, and calibrated alng with its mating amplifier. The fuel system cnsists f a fuel reservir, a fuel put.p, a fuel cntainer, a weighing scale, an electrnic timer, a fuel fuel filter, and an injectin nzzle. Clear indlene, methanl, ethanl, prpanl, and butanl (industrial grade) vere used in this study. The rate f fuel cnsumptin is measured by a weighing scale and timer. Flame prpagatin time is measured with the inizatin prbe. It cnsists f a water cled prbe, and an electrnic amplifier-

17 1 9 Fuel Reservir Fuel fil ter 1 Fuel rntaine: Scale ZY \ Electric timer Weight Pressure (y/ gauge Injectr Heater T r^ *. intake manifld Air tank Pr<Bssure gauge '' T Cylinder-head Air tank Regulatr Air Valve Pipe line filter Valve Fig, 2.4 Intake fuel-air system

18 10 cunter. Fig shws the layut f the inizatin prbe. This prbe was designed, and manufactured here, at U.I.C. The prbe can traverse alng its center line, inside the cmbustin chamber, and can be held at any desired distance frm the spark plug. The electrdes f the inizatin prbe cnsist f tw thermcuple wires- a platinum ne and a platinum-13%rhdium wire. They are separated by MgO insulatin, and enclsed in an incnel sheath. The diameter f the wires is in. The uter ends f the electrdes are cnnected t the electrnic amplifier-cunter, and t an scillscpe, the signal frm which is input t the data acquisitin system. Data acquisitin and recrding system The data acquisitin and recrding system cnsists f 1 CAD at 2000 rpm acquisitin rate, 2 A t D channels, 1 interval timer, 1 TDC detectr, 1 parallel data latch, an IBM XT PC, and an FX85 printer. Fig shws the layut f the data acquisitin system. The fllwing data were recrded and mnitred n the scillscpe: 1. Instantaneus cylinder pressure signal. 2. Spark signal which indicates the spark timing. 3. Inizatin prbe signal. 4. Shaft encder signal. In additin, mass flw rate f air and fuel, trque, tempe atures f il, clant, and exhaust gas were als recrded t btain the perfrmace parameters.

19 11 _!_ 1 rv Z- ' ^-r ^ UJ CD -^ N -J UJ u 2 Ui > < ^5 u. O UJ ^ Z c! s is! (9 s tt: Ui < ^ IxJ Fig. 2.5 Water cled traveling prbe

20 12 Oscillscpe A/D nverer Interlace Timer Cunter IBM- PC- XT Micr-Cmputer Main Cmputer Flppy Disc Drive Memry Vide Disc Drive Printer Fig. 2.6 Data Prcessing System

21 13 Equipment Preparatin The engine was given a cmplete verhaul, including a cleaning f the pistn, cylinder head, intake and exhaust valves, intake and exhaust maniflds, cylinder pressure transducer, injectin pump and cling system. The pressure transducer was calibrated and munted flush with the cylinder head. The lad cell-indicatr system was als calibrated. The intake and exhaust valves were carefully lapped t ensure prper seating befre installatin in the engine. Calculatins were made t determine the stichimetric air t fuel ratis fr all the fuel blends, as described in the appendix. The data recrding sheets were develped as seen in the sample calculatin. The experiments were divided int tw parts. The first part was t btain the Minimum spark advance fr Best Trque (MBT), and the Knck Limited Cmpressin rati (KLCR) with indlene-methanl and indlene-mpha blends. The secnd part was t btain detailed cmbustin and perfrmance characteristics f all fuel blends. The engine perating cnditins and test matrix are given in Table Experimental Prcedure The fllwing experimental prcedure was used: 1. Prir t running experiments, the engine and ther equipments were prepared as stated in the previus sectin. 2. The checklist shwn in the appendix was fllwed t mtr and fire the engine.

22 14 TABLE 2.1 TEST MATRIX FOR INDOLENE -METHANOL BLENDS CR 100%I 70%I+ 30%M 50%I+ 50%M 30%I+ 70%M 100%M

23 ' <> t*-!-v.-. :. : TABLE 2.2 TEST MATRIX FOR INDOLENE-MPHA BLENDS CR 100%I 70%I + 50%I+ 30%I + 100%MPHA 30%MPHA 50%MPHA 70%MPHA

24 16 Table 2.3 GENERAL OPERATING CONDITIONS Spark plug gap (in.) Spark timing Valve clearance (in.) MBT (ht) Oil (crankcase) Synthetic il design fr methanl Intake air temperature ( 'F) Equivalence rati Speed (RPM) Cmpressin Fuel rati indlene- blends t KLCR indle ne, alchls Injectin pressure (psi) (manifld 1200 Intake air pressure WOT

25 17 3. The cmpressin rati was adjusted and varied t btain the KLCR. 4. The engine speed was set at a cnstant value f abut 1000 rpm. 5. The equivalence rati was set t abut 1.0 by using the calculatin sheet shwn in the appendix. 6. The engine was run fr a perid f an hur, at the perating cnditins given in Table 2.3 t ensure steady state peratin befre recrding the data. 7. In the first part, the spark advance was changed frm 20 deg.btdc t 50 deg.btdc in increments f ne degree and the lad cell reading was recrded. The spark advance at which the lad cell indicated maximum trque is called the MBT spark advance. 8. The cmpressin rati was varied frm 5:1 t KLCR in increments f 1. Perfrmance Parameters The parameters which have been studied in this wrk are; 1. brake pwer, 2. brake specific fuel cnsumptin and 3. brake thermal efficiency. The tests started with pure indlene. Methanl was then ;idded by 30%, 50% and 70%. Tables 2.1 and 2.2 shw the experimental test matrix. The cmpressin rati was varied frm 5:1 t KLCR, i.e., when engine knck became audible. The knck was bserved frm the pressure trace. Burning velc i ty The parameters which have been studied in this part f the study are: 1. Flame speed, and 2. Burning velcity. The vlume fractin f methanl and MPHA were varied by 0.1, 0.2, 0.3, and 0.4. The cmpressin rati was kept cnstant at 5:1. The flame prpagatin time was measured at eight lcatins during flame prpagatin.

26 . 18 CHAPTER III: DATA ANALYSIS AND SAMPLE CALCULATION In this chapter, the data analysis and a sample f the calculatin will be presented. The data analysis cnsists f tw parts: 1. Flame travel time analysis. 2. Pressure-Time histry analysis. Flame Travel Time Analysis The flame travelling time is taken as the time interval between the spark timing and the arrival f the flame at the tip f the inizatin prbe. A high speed time cunter is used t measure the flame travelling time in millisecnds. The spark and inizatin signals are als mnitred n the scillscpe. Fr each run, 512 data pints are temprarily stred in the micrcmputer memry and later transferred t files. After determining the 512 data pints fr the flame prpagatin time fr each psitin f the inizatin prbe, an arithmetic average is perfrmed n the data by the prgram TIMEAVG. This prgram als checks and cmpares the mean with all the data pints, excludes bad data pints, and recalculates the average flame travel time (Bad data "Jints were defined as abve 1.5 times r belw 0.5 times the mean) This prcess was repeated fr all the different prbe psitins between the spark plug and the chamber wall. The time average data

27 were then curve fitted using a fifth degree plynmial t give the distance vs. time relatinship fr the flame travel prcess. In rder t imprve the accuracy f the curve fitting rutine, additinal pints were inserted near the end pints. The apparent flame speed is cmputed frm the slpe f the distance-time curve by a cmputer prgram APPBVEL. This prgram calculates the apparent flame speed at every degree crank angle by using the plynmial cefficients as input. Pressure-Time Histry Analysis The cycle-t-cycle variatin in pressure traces is a cmmn prblem in spark ignitin engines. The cmbustin characteristics calculated frm the mdel are very much dependant upn the pressure, and its derivatives. An attempt has been made in the literature t use the average pressure f many cycles. Pattersns [18 ] has recmmended that 2000 pressure cycles be used fr a representative average cycle, and Peter[19] has recmmended 300 cycles. Recently Mattavi et al [20] used 96 cnsecutive cycles t btain the average pressure data. Fllwing was the prcedure used in this study t btain the average pressure data. The high speed data acquisitin system was used t cllect 80 cnsecutive cycles fr each run. The pressure was recrded at every 1/2 deg. crank angle. The arithmetic average f these cycles was taken t btain ne average pressure cycle.

28 20 Sample Calculatin Flame Speed The average time f 512 pints fr eight different prbe psitins are shwn belw: 'M\ Flame Travel Time Distance E E E E E E E E E E E E E E E E O.OOOOOE Imprved Flame Travel Time E E E E E E E E E-02 Fig. 3.1 is a plt f the distance vs. flame travel time. The circular symbls are the inserted data pints t increase the J ^ accuracy f the plynmial curve fitting. After curve fitting, the cefficients are used t calculate the flame speed at every crank angle. The appendix cntains plts f flame speed vs. crank angle calculated frm APPBVEL prgram. Mass Fractin Burne<' As shwn in Fig. 3.2, ignitin delay and cmbustin interval are calculated frm the mass fractin vs. degree crank angle curve. These data are btained frm the cmputer simulatin prgram SPMIX.

29 21 DISTRNCE VS FLAME TRRVEL TIME AT CR-5.0t l.indolene 0. ' A5.a TI)1E^lD*^+2 (SECOND) i^sts'3qrtr Fig. 3.1 Distance-Time curve

30 22 MASS FRACTION BURNED VS DEGREE Cfl DEGREE CRANK RNGLE MfiSS FRfiCT, BURNED Fiq. 3.2 Mass fractin burned vs **CA

31 Determinat in f Cmbustin Characteristics The fundamental quantities frm which cmbustin can be characterized are, actual burning velcity, ignitin delay, and cmbustin interval. The fllwing is an utline f the prcedure used in calculating these fundamental quantities. Ignitin Delay The perid f time frm spark t the pint when heat release is detected, is called ignitin delay. This perid can be calculated frm the p-t diagram where the pressure increases drastically* Alternatively, this perid can be btained frm the mass fractin burned vs. degree crank angle curve. Ignitin delay can be defined as the interval between the spark and the time when 5% f the mass is burned. This definitin is used in this study t calculate the ignitin delay perid. Frm Fig. 3.2, it can be seen clearly that after 5% f the mass is burned, the rate f burning rises sharply. Cmbustin Interval After the ignitin delay perid, the fully develped turbulent flame prpagates acrss the cmbustin chamber. Cmbustin interval is the perid f time frm the end f the ignitin delay perid t the end f cmbustin. Keeping in mind the accuracy f the mdel, the cmbustin interval is defined as the time between 5% t 90% mass burned. During this interval, the flame prpagates with a finite thickness thrugh the unburned mixture. Fig. 3.2 shws the definitin f cmbustin interval frm the mass fractin burned vs.

32 24 degree crank angle curve. Burning Velcity The burning velcity can be calculated frm the flame speed and the pressure-time data, the flame frnt area at every crank angle is needed fr these calculatins. Flame frnt area was calculated frm the spherical flame prpagatin mdel. Fig. 3.4 is a plt f the flame frnt area vs. flame radius. Fig. 3.5 shws the curves f flame speed, expansin velcity, and actual burning velcity as functins f crank angle degree, fr indlene.

33 lv-'/ ^;:^^ MASS AND VOL. FRflCT, BURNED VS DEGREE Cfl ^.': DEGREE CRANK ANGLE fo t MRSS FRflCT. BURNEJ VOL. FPK^r. SSSS Fig. 3.3 Plt f mass and vlume fractin burned vs degree crank angle at base cnditins.

34 26 FLflME RffilUS VS SPHERICPL FLflME FRONT firer AT CR-5 "^.00 Z FLflME RADIUS ( CM) Fig. 3.4 Flame frnt area vs flame radius

35 27 VELOCITY VS DEGREE Cfl FOR INDOLENE, CR=5 <N s -2D fi. 87 DEGREE CRPNK ANGLE e- flpp. SJRNING VEL. -e- flctjpl 8URNIN6 ya. Fig, 3.5 Burning Velcity vs 'CA

36 .ives fairly gd results and is relatively simple t use 28 CHAPTER IV THEORETICAL ANALYSIS ^''^ Mdels f Cmbust inn MaeUn, f t.e cc.ustin phenmena.ns.de the spar, xgnitin engine is relatively si^pie if cmpared ith.deling f diesel cmbustin. in spar, ignitin engine, the fixture f air and f.el " - P-i-a ana can.e assumed as an hmgeneus.i.. ture. The cmbustin prcess is treated as a cnstant pressure cmbustin and immediately fllwed by an iscthermal mixing between newly burned an.^ ^k u y uurnea and the burned gases. As prpsed by Bracc i-h^ ^ y Bracc, the engine cmbustin mdels can be classified as fllws: 1. Zer-dimensinal mdels 2. Quasi-dimensinal mdels 3. Multi-dimensinal mdels The abve mdels are dxscu-ssed earuer in Chapter 2 2er-dimensinal mdel with the chsen three.nes cntrl vlume is emplyed n this analysis because it

37 D erivatin f the Energy Related burning Velcity Equatin The burning velcity equatin is derived by tw analytical methds: Thermdynamics analysis and cmbustin mechanism analysis. Assumptins : 1. During cmbustin, the cylinder cntents are made up f three distinctive regins, ne f burned part, ne f unburned part and the flame (reactive regin) in between separates the tw regins abve. 2. The mass burning rate f the fuel air mixture is due t the chemical reactin nly and is equal t the rate f mass gained t the cntrl vlume f the burned part. 3. Pressure surrunding the cntrl vlume f burned regin is unifrm thrughut the cmbustin prcess. 4. There are n effect f the ptential and kinetic energy in the cmbustin. 5. At initial stage, there is a very small (r nne) quatity f the burned mass. 6. The burned gases cnsist f a unifrm mi"ture f prducts f cmbustin. 7. The unburned mixture is made up f a hmgeneus mixture f air, fuel vapr and residual prducts f cmbustin.

38 JU 8. The systems behave accrding t ideal gas law. 9. Heat transfer is cnfined t cnvective exchange between the gasses and the walls. 10. The energy required t prpagate the flame frm initial t final psitin is dented as the flame wrk and is defined as: W =W -W -W, +W^ cv exp cmp lp f (4.1) Where W = Ttal net wrk fr a given system cv W = Expansin wrk exp ^ W = Cmpressin wrk cmp W, = Lp wrk lp ^ W-; = Flame wrk Thermdynamics Analysis Fig illustrates the initial and final states f an arbitrary chsen cntrl vlume f the burned gases. Fig. A-.Z shws the system after the cmbustin f an charge element ^m and the unifrm pressure surrunding the cntrl vlume. The cnservatin f mass^ tgether with First law f thermdynamics is applied t the system as fllws. * Cnservatin f Mass At any instant during the prcess, the cntinuity

39 31 equatin is written fr the system as: Pif'cv ^^ * i^"e^^ - dt (4.2) f^^i Fr a perid f time t, equatin (4.2) becmes ^^ 2 - '^l^cv + ^^e - ^"^i = (4.3) where m., m-, = Mass rate f the burned gases at initial stage and final stage respectively. rfi., r = Mass burning^ rate enter t and esit frm i' e Frm the assumptins 2 and 5, the cntrl vlume respectively. m^ = m^ = e after arranging and let m, dentes the mass burning rate, equatin (4.3) wuld becme: m- = m =. m, (4,4 2 lb ) * First Law f Thermdynamics Writting the First law fr the unifrm state, unifrm flw (USUF) prcess, at any instant f time fr the cntrl vlume f the burned part, during the prcess, we have:

40 Kv.'"-' '-v i- 32 Initial P. 1 T. Final "'Ill Figure 4.1: Initial and final state f the system.

41 ^i--^':<^ 33 Figure 4.2: Unifrm pressure surrunding the system

42 34 Q + Z m. (h. + ^ c V 11 V m (h + e e + gz. ) = ( ) + w 1,^ cv cv at V ^- 2 + gz/) (4.5) 7 -e Def f erentiate equatin (4.1) w = w - w - W, + W- cv exp cmp lp r 4.6 Substitude (4.6) int (4.5) and slve fr W^ vvj^i:* W^ = Q^^ + Zmih. - ( de dt - ) W + W + W, cv exp cmp lp (4.7) V'-lvN'-v jilt where W^ = Rate f flame wrk dne by the system. 6 = Rate f heat transfer in t the system, ^cv Z m.h. = Rate f ttal enthalpy input. (^) = Rate f energy accumulated in the system cv dt W = Rate f expansin wrk dne by the system, exp W = Rate f cmpressin wrk dne n the system. cmp W, = Rate f lp wrk. lp Let de-,/dt dentes the/^ight hand side f equatin (4.7) and multiply bth side by dt: ct W^ = m^^w^dt = de^ (4.8) Frm Fig.A.^, pressure surrunding the cntrl vlume is unifrm, the wrk dne is:

43 35 (p) (A) (dr^) = (F/A) (A) (dr^^) = F(dr. (4.9) At t = 0; r^^q = ^bo " ^bt ^f = ^^^bt^ (4.10) Subsitude (4.10) int (4.8) /w^ = de^ = mj^(f)(rj^^)dt (4.11) Frm definitin f mmentum, per Kg mass f burned gases, we have: F = ( dm, dt ) (velcity ) /dm, velcity dt _ u dt (4.12) Substitude (4.12) int (4.11) integrating frm zer t time t and rearranging: Su = V^'^b^^^b^ (4.13) ^y;<vv.' ^ = W^/(m^)(r^) Equatin (4.13) represents the burning velcity as a functin f the flame wrk, mass burning rate and the burned part. This equatin will be evaluated term by term and the

44 36 mass burning rate: ""fi = "'f " "^""fl (4.14) Apply the First law t the flame element, neglecting the ptential and kinetic engergy, we have: cv * 1 1 de cv dt + 2ni h + W e e cv (4.15) and Frm m = e w=w -w -w, +w^ cv exp cmp lp f equatin (4.15), after slving fr Vv', wuld becme W^ = Q + ^m^,h ^ - de /dt - W f cv ^ fl uo cv exp + W + W- cmp lp (4.16) Let de /dt dentes the right hand size f equatin 4.16) and frm definitin f the flame wrk, "f = Pfi"^^fi' = ''fl W f 1 Pfi^^fi^^^i - ^0^ = ^f^^i - ^0^ (4.17) and w f 1 <f w f 1 dt (4.18) Burning velcity is defined as the rati f:

45 37 cmputatin f the burning velcity will be given in sample calculatin in Appendix Cmbustin Mechanism Analysis Cmbustin prcess in the S.I. engine is assumed t prgress in a stepwise maner as prpsed by Pattersn. It is described belw with the assumptin f n mass lss due t the mvement f exhaust valve, i.e ttal mass cntained in the cyclinder remains cnstant, during the prcess. The whle system cnsists f three distinctive regins, ne f burned part, ne f unburned part and the flame between separates the tw regins abve. The cmbustin is cnsidered t take place in the intermediate prcesses as fllws: Prcess 0-1 Cnstant pressure burning f an increment f the' charge f thickness dr = r^ ~ ^ ' 1-2 Isthermal mixing between newly burned and burned gases Isentrpic expansin f the burned gases. Cnstant vlume heat transfer. Plytrpic change due t pistn mtin. Table I summarizes the parameters f the whle system at every stage f the prcess.

46 38 Initial nil, After the intake and isentrpic cmpressin prcesses in the cycle, the whle system is assumed t have a negligible quatity f the burned mass. Initial thermdynamic i!«l prperties f the subsystems are given as: Burned 81 j! -ml Mass Vlume "bo = Temperature Pressure Nt N.A applicable Flame Mass Vlume "fo = Temperature Pressure Nt N.A applicable Unburned Mass Vlume Temperature m uo uo uo

47 ; 39 Pressure uo Step 1 Cnstant Pressure Burning An increment f mass, thickness f dr = r. - r^, is burned at cnstant pressure. The flame temperature wuld be at its adiabatic flame temperature. Flame element will increase its vlume, thus cmpressing the unburned part. The burned gases (if any) state is assumed remain unchange. The whle system is illustrated in the abve figure. Let's cnsider each f the subsystem as fllws: Burned State remained unchange. Flame The flame prperties are changing frm the initial cnditin t the end f this step and will be determined by applying the First law t the flame element. Prperties at this stage are: Pressure: P^- Vlume:

48 ' ).<.<'.-Ij 40 S = u dr dt ^1 - ^0 dt (4.19) Substitude (4.19) int (4.18) and slve fr S^, we have. S = W., / F. u r I r (4.20) frm F^ = m,- (a^ ^ (4.21 ) and acceleratin f the flame, a^^, is defined as: a = (Velcity at 1 - velcity at 0)/dt ^^1 - ^0^ dt dt (4.22) At initial cnditin, the flame radius Tq = 0. Substitude (4.22) int (4.21) then (4.20) and nting that ^fl ~ m-^ / dt we wuld have: Su = W^ / (m^^)(r^) (4.23) The flame wrk, W^ will be determined by the equatin (4.16) by the fllwed prcesses. Step 2: Isthermal Mixing Between Newly Burned and Unburned Gases.

49 41 '>-y.-'.' B A / / / u ^1 Burned Gases The vlume and pressure f this system is changing, the final vlume is: ^b2 = ^bl ^^fl =^fl " '" Temperature f the burned gases remained cnstant and is equal t the adiabatic flame temperature. '^b2 ^ "^fl " "^adia The mass f the system is equal t the mass entering ' the burned gas. Since the initial mass m^^ = m^ ^ =,! was assumed. The mass f the system after the mixing prcess is. "^b2 = "'fl = '"uo - ^u1 ^""'^^^ Applying the equatin f state fr an ideal gas, ther prperties f burned system can be calculated. Flame After the isthermal mixin g prcess, the flame regin

50 : 42 is reduced in its size with a negilible thickness and is assumed t be statinary and nn reactive, until the end f the whle prcess, then it can start the cycle again. Unburned Mixture The state f this system remained unchange, since it is assumed that the mixing prcess ccurred between the burned and newly burned systems nly. Step 3 Isentrpic Expansin f the Burned Gases.!i!r Burned Gases : Expansin «: The burned system underges thrugh an isentrpic expansin prcess. The burned mass remained cnstant. With the pressure rapidly increased frm the previus prcess, the burned gases wuld cmpress the remalng unburned mixture and causing the pistn t mve dwnward. Prperties f buned gases can be calculated as fllw M. b2 (K - 1 )/K b3 p P (4.26) b2

51 43 b3 b2 b2, p b3 (4.27) Where K_ = :. vp - i pi C - R PP (4.28) The expansin wrk can be determined by applying the First law: ^exp " "b23 = Ub2^^b2) - "b3^\3) (4.29) Unburned Mixture The left-ver unburned mixture, in cnstrast, wuld be cmpressed by the burned gas. The unburned mass is reir.ained cnstant; by similiar applicatin f the First law and equatin f state fr an ideal mixture, the prperties f this system can be calculated. With the assumptin f the cmbustin prcess is cmpleted, then there is a very little amunt f unburned T.ixture presented. Step 4; Cnstant Vlume Heat Transfer The verall heat transfer frm the burned t unburned and t the walls, plus heat lss frm unburned t the wall can be cmputed using Wschni's crrelatin. Because the

52 'k. y 44 B 'bu U "uw bw amunt f heat transferred t the unburned mixture is just ff -set by that frm the unburned gas t the walls, as prpsed by Pattersn. The heat transferred frm burned t unburned therefre can be ignred, the cnvective heat transfer can be calculated as fllw: T ) Qui = h (A. ) (T. - u ui ui w and u = ZQ (4.30) ui 'bi ^b^^i^^^bi - ^w^ %i (4.31 ) Cnvective heat transfer cefficients are cmputed frm Wschni's crrelatin and is given as: h = n.0.8 ^0.2^^0. 53j Pi 1 (4.32) The meaning f each term in this equatin is given in Appendix 2.

53 . : 1 45 The wall area which expsed t the burned gas can be calculated as given in the belw equatin: A = (TC) Bre [ strke ^^ ^ ^^^q. + Rd( / _ sin Q Strke' y 4 Rd^ ) ] (4.33) The Heat transfer area fr burned gases is calculated frm the spherical flame frnt mdel. Area fr the unburned gases is the difference between ttal area (f chamber walls, pistn and head) and the area in cntact with the burned gases The wall temperature is measured with a thermcuple, m this temperature is assumed t be the same everywhere n the wall surface temperature variatins are quite small cmpared t the ttal temperature difference between the burned gases and the surfaces. Step 5 Plytrpic Change Due t Pistn Mtin This final prcess is due t the pistn mtin, where-

54 46 ther it is plytrpic expansin r cmpressin, the fllwing relatin ship can be applied t the burned and unburned systems: V k-1 = ( 5 ) k (4.34) After this prcess is cmpleted, the new cycle in the cylinder wuld take place with the same steps. NS<

55 , CHAPTER V RESULTS AND DISCUSSION This chapter describes the results btained frm the theretical and experimental wrk. Physical and Chemical Prperties When evaluating an alternative fuel, knwledge f its physical and chemical prperties is essential fr a thrugh understanding f hw the prpsed fuel shuld perfrm in an I.C. engine. While there are sme prperties which are generally attributed nly t either S.I. r C.I. engine fuels, many prperties are cmmn t bth, as well as t turbine engine fuels. These prperties include density, specific gravity, abslute viscsity, surface tensin, distillatin temperatures, flash pint, and heating value. The fuel prperties which have the greatest effect n S.I. engine peratin are the distillatin characteristics, flash pint, heating value, and cmpsitin. In the present wrk, the fllwing prperties were measured r cmputed: 1.Di st i Hat in characteristic, 2. Heating value, 3. Flash pint, 4. Water tlerance, 5. API gravity, and 6. Stichimetric air-t-fuel rati. Mst f the prperties were measured using standard ASTM prcedure and equipment. Distillatin Characteristics The distillatin temperatures fr the 15 fuels in this study are tabulated in Table 5.1. The data fr indlene, indlene-methanl

56 - «48 in in m O m m 91 Oi»» 0«Ok Ok Ok Ok e u OS OS CD O" CD OS OS CO OS OS * Ok Ok Ok Ok * e in n in CD «D Ol n *n la» * O r«««* m CO» cc re re m CD O m 10 M CO n re O) CO CO fkl s «e e r m m m Ok CM OS > r* r* O ' 00 M in 10 Ok CO m M 10 «0 m m 1^ CM in CO in ko \0 r ko 9> CO CD in in m in 01 «^ *^ s O CO m O m r * m a> r^ m m f"" ^ CO m 00 < ^ n rk irt CO m «0 m *^ m Ob «^ << ««««in 'O ^ n 00 ««fl M O 2^ 8* M O 2* M < as 4^ << < a < 4- a ^0 ^ < a < O X r»» z

57 . 49 and indlene-mpha at the same blend levels are pltted in Figs A substantial variatin in distillatin characteristics is evident. Fig. 5.1 shws the distillatin curves f pure indlene, methanl and MPHA. Because f a blend f many different hydrcarbns present in indlene, it shws a brad range f biling pints. Since methanl is a single cmpnent fuel, it has a cnstant biling pint temperature, and therefre, des nt prduce a distillatin curve. MPHA, en the ther hand, is a blend f different alchls, and prduces a curve similar t that f indlene. It cntains 75% by vlume methanl, and a remaining 25% f ther alchls. Therefre, the distillatin curve is straight line fr abut 75% distilled and then varies with temperature. The shape f the distillatin curve will change as the cmpsitin f MPHA. Fr example, if extra ethanl r ther higher is added t MPHA befre mixing with indlene, the distillatin curve will be different frm that shwn here Figs shw the distillatin characteristics f indlene, indlene-n-.rthanl and indlene-mpha fuels. The entire distillatin curve is affected by the presence f alchls. The cnstant temperature biling due t methanl, which ccurs slightly belw 150 T^ is shwn clearly in all figures. At all blend levels, the indlene-mpha blend is clser t indlene than the indlene-methanl blend. Althugh distillatin is a cntinuus prcess, nly a few pints

58 50 DISTILLATION CHARACTERISTICS INDO., METH. & MPHA I LL LEGEND =100% INDO. = 100% METH. = = 100% MPHA "O UJ en Q. v; ILl PERCENT DISTILLED Fig. 5.1 Distillatin Prperties f Pure Fuels

59 3) D D U c I C u L DISTILLATION CHARACTERISTICS INDO.-METH. & INDO.-MPHA BLENDS PERCENT DISTILLED Fig. 5.2 Distillatin Prperties f 20% Blends, \

60 52 DISTILLATION CHARACTERISTICS INDO.-METH. & INDO.-MPHA BLENDS > D HI CC LEGEND 100% INDO. 60%INDO.-f-40%METH. 60%INDO.-f40%MPHA \s CC LU Q. LU PERCENT DISTILLED Fig. 5.3 Distillatin Prperties f 40% Blends.

61 3) D D J c D C c LI L i,i.>< -w-^/' *.<'. ;... \..: >: 53 DISTILLATION CHARACTERISTICS INDO.-METH. & INDO.-MPHA BLENDS PERCENT DISTILLED Fig. 5.4 Distillatin Prperties f 60% Blends.

62 54 DISTILLATION CHARACTERISTICS INDO.-METH. & INDO.-MPHA BLENDS M: lis IV LL d) O "D LU a:? LU Q. 111 I LEGEND 100% INDO. 80%INDO.4-20%METH. 80%INDO.+20%MPHA PERCENT DISTILLED Fig. 5.5 Distillatin Prperties f 80% Blends.

63 55 alng the curve are necessary t characterize a fuel, namely, the initial biling pint, 10 percent maximum, 50 percent ir.inimum, 50 percent maximum, 90 percent maximum, and the end pint. Depending upn the vlatility class, the maximum 10 percent limit varies frm 122F t 158F, minimum 50 percent is 170F, maximum 50 percent varies frm 230F t 250F, maximum 90 percent varies frm 365F t 374F and end pint is 437F. Since the 50 percent minimum limit is 20 t 25 degrees abve the cnstant biling pint temperature, that part f the distillatin specificatin will be difficult t meet using indlene-methanl blends. As shwn in Fig. 5. 2, the 20 percent alchl blend is within the specified limit. The 40 percent blend, Fig. 5. 3, exceeds the limit. Hwever, by mdifying the frmulatin f MPHA, it may be be pssible t meet this part f the specificatin upt 70 percent blend level. The shape and level f the distillatin curve affects gasline engine perfrmance. The characteristic indicating the cld starting quality f a fuel is represented by t 20 percent evaprdted. A highly vlatile t 20 percent fuel will aid cld starting, but may tend t cause vapr lck at high teperatures. The warm-up characteristics are related t the temperature at which 50 percent f the fuel is vaprized. A lw 50 percent pint yields better perfrmance, but with an increase in the tendency t cause carburetr icing. Fuel ecnmy relates t heating value, and density, which may als crrelate smewhat with the 90 percent pint. Lw back-end vlatility may lead t crankcase dilutin, depsits, gum, and spark plug fuling.

64 56 Energy Density Fuel ecnmy is usually expressed in terms f BSFC fr engines. The BSFC is a functin f the energy-mass density, which affects the cnversin effiency f the chemical energy t useful mechanical wrk. The energy density, r heating value f a fuel is a measure f the amunt f energy released as a result f the cmplete cmbustin f the fuel. The higher heating value (water frmed by cmbustin is in vapr frm) is measured using an adiabatic bmb calrimeter. The H^H "ll '-^H lit "^1.1".^l^^l II ''^^911 Ik :H Ml ^ ^ ''"" ' ' ' heating values f indlene, indlene-methanl, and indlene-mpha blends were measured. Fig. 5.6 shws the heating values f the indlene-alchl blends. The measured heating value decreases as the cncentratin f alchl increases in the blend. The heating value f indlene-mpha blend is higher than indlene-methanl blend at all "c;^;--.-^:-- ^>;i;;>a' ^''.^s',,fi -I'll! blend levels. ii. 'Ill iirj The energy-vlume density, BTU/gal, is dirctly prprtinal t the prdust f the energy-mass density and the specific gravity. Since the specific gravity f MPHA is similar t that f methanl, the energy-vlume density f indlene-mpha blends als be great-er than that f indlene-methanl blends. The heating values f blends are generally predicted frm the pure cmpnent prperties by equatin 5.1. Figure 5.7 shc's the blend heating values predicted frm its pure cmpnents using this equatin, end the measured values, fr indclene-mpha blends. KVM ^ MI * HVI MA * HVA (5.1) -r where HVM = Heating value f mixture

65 I 57 MEASURED HIGHER HEATING VALUE INDO.-METH. A INDO.MPHA BLEND I I 4D0LENE 4D0LENE METHANC L BLEND MPHA BU!ND FRACTION OF ALCOHOLS 1.0 Fig. 5.6 Measured Higher Heating Values f Indlene-Alchls Bleni

66 . 58 MEASURED i CALCULATED HEATING VALUE INDOLENE-MPHA BLENDS S.5 IH f lij ''';^ '^^ m^^l ^^H ''^'^J -i^ ^^1 '^ ^H (D ^^1 HM > z H- :-ip!!'< /,:;: i UJ til V _: -,.;. -; c UJ. ^: z 9 C5 z ifl - ',<' * IIEASUREC VALUE CALCULATIED VALUE a FRACTION OF ALCOHOLS 1.0 Fig Msasured ( Olmlntiwl Higher Beating Valis f In3blaie-MPHA Blends*

67 59 MI = Mass fractin f indlene HVI = Heating value f indlene MA = Mass fractin f alchl HVA = Heating value f alchl There is a gd agreement between measured and predicted values fr indlene-mpha blends. Water Tlerance Fig shws the water tlerance f the indlene-alchl blends. The slubility f alchls in indlene when water is present even in minute quantities is limited at rm temperature. When small amunts f water added t blends f alchl with indlene, hydrgen bnds frm between the water and alchl mlecules, and the blends separate int tw phases. Paraffinic hydrcarbns predminate in the upper phase, while the lwer phase cnsists primarily f alchls and water. Water tlerance is als significantly decreased at lwer temperatures. At rm temperature, as the alchl cncentratin increases, water tlerance increases. The 90 percent methanl blend has a water tlerance f 8.25 percent by vlume, cmpared t 0.25 percent at 10 percent blend. The 90 percent MPHA blend has abut 50 percent higher water tlerance cmpared t the 90 percent methanl blend at rm temperature. Slubility f methanl and indlene in the presence f water can be ne f the mst difficult prblems attending the use f these blends as mtr fuels. The fuel distributin system, refinery, pipeline, and terminal tanks nw used are nt necessarily designed

68 60 WATER TOLERANCE LU O > >- m LU z< IT J,..l 'lltll 'Hi: ii' 111 O CC HI O.e FRACTION OF ALCOHOL Fig. 5.8 Water Tlerance

69 m^mm-m^m-^ t fully exclude rainwater frm cming int cntact with the fuel. Cndensatin in the strage tanks can cause additinal water cntaminatin. Using MPHA-gasline blends instead f methanlgasline blends increases the water tlerance, and may be used withut making majr mdificatins t the existing fuel distributin and delivery system. Flash Pint The flash pint is the lwest temperature at which the vaprs arising frm a fuel will mmentarily ignite as a flame is swept acrss its surface. It is used t assess the flammability f a fuel and t establish shipping and safety regulatins. The flash pints f methanl and MPHA fuels were measured using a Pensky-Martens clsed up tester, fllwing the ASTM D93 methd. The flash pint f pure methanl was 50F, and that f MPHA, 49F. API Gravity and Specific Gravity The API gravity is defined by an arbitrary scale whse units are degrees API, and is measured at 15.5 deg.c. The API gravity f indlene-methanl and indclene-mpha blends were measured using Fisher brand Petrleum hydrmeters. Specificatins fr API gravity hydrmeters have been established by ASTM. Fig. 5.9 shws the API gravity f indlene-methanl and indlene-mpha blends. In genera], as the alchl cntent increases, API gravity decreases. Indlenemethanl has a slighcly higher API gravity than indlene-mpha. The specific gravity is fund by dividing the density f the fuel

70 \ii, :P^ 62 API GRAVITY INDO.-METH. & INDO-MPHA BLENDS 90.0 'lit; LEGEND = INDO.-METH. =INDO.-MPHA ^?^ I'll' > 'i':i,?^$s;> CD K < /" FRACTION OF ALCOHOL Fig. 5.9 API Gravity

71 63 by the density f water. The specific gravity is related t API gravity as shwn in equatin 5.2. Degree API = ( /spec i f ic gravity) (5.2) Fig shws the specific gravity f indlene-methanl and indlene-mpha blends. As the alchl cncentratin increases, the specific gravity increases slightly. In general, indlene-mpha has a higher spefic gravity than indlene-methanl blend. Stichimetry Stichimetry and cmbustin are related t the H/C rati f the fuel. In general, fr indlene, H/C is 1.86, 4.0 fr methanl, and fr MPHA. The stichimetric air-fuel rati depends upn the H/C rati f the fuel. The stichimetric air-fuel rati is fr pure indlene, 6.43 fr methanl, and fr MPHA. Fig shws the variatin f stichimetric air-fuel rati as a functin f alchl cncentratin. As the alchl cncentratin Increases, stichimetric air-fuel rati decreases. The stichimetric air-fuel rati f indlene-mpha blends is higher than thse f indlenemethanl blends. PERFORMANCE PARAMETERS Effect f Alchl n MBT Spark Advance Figures shw the effect f varying the methanl and MPHA cncentratins n the MBT spark advance, at different cmpressin ratis. At any fixed cmpressin rati, as the alchl cncentratin increases, the MBT spark advance decreases, r shifts

72 64 i 1.6 SPECIFIC GRAVITY INDO.-METH. & INDO.-MPHA BLENDS I 1.4 LEGEND INDO.+METH. INDO.+MPHA >- 1.2 'Dili "111," ii LL HI Q. CO FRACTION OF ALCOHOL Fig Specific Gravity

73 I " t 1! i 1 1 \ STOICHIOMETRIC A/F RATIO INDO.-METH. & INDO.-MPHA BLENDS 19.0 t 17.0 LEGEND =INDO.+METH. -INDO.+MPHA i '^O^ ^^ 1 S 11.0 \: =^^-1.,^ iimm i ^1\^^^^ FRACTION OF ALCOHOL Fig Stchimetric A/F Rati

74 1 1 i 1 i!! MBT SPARK ADVANCE AT CR=5.0 INDO.-METH. & INDO.-MPHA BLENDS / u.u - 1 QHm ftn DU.Un L LEGEND =INDO.+METH. =1 MDO. +MPHA 1 i 1! j 1 1 i 1 j i i ' : i i i I'll III d) D UJ O D < 50.0-^ AH n ^ ru.u ^^^ 1 i ~! 1 w ^^^^^ i! DC < Ql FRACTION OF ALCOHOL i Fig MBT Spark Advance at CR=5:1

75 i 1 1 i 1 67 MBT SPARK ADVANCE AT CR=6.0 INDO.-METH. & INDO.-MPHA BLENDS 70.0 T 60.0 LEGEND =INDO.+METH. =indo.+mpha i 60.0 ' ^x^^; * (1 1 i ^^*^f * -1 _ >^ ^**^*< > _ ^ -" I FRACTION OF ALCOHOL 1.0 Fig MBT Spark Advance at CR=6:1

76 68 MBT SPARK ADVANCE AT CR=7.0 INDO.-METH. & INDO.-MPHA BLENDS 70.0 QH CQ d> 60.0 LEGEND INDO.+METH. INDO.+MPHA LU O z 60.0 Q < 40.0 < Q. CO FRACTION OF ALCOHOL Fig MBT Spark Advance at CR=7:1

77 MBT SPARK ADVANCE AT KLCR INDO.-METH. & INDO.-MPHA BLENDS LEGEND INDO.+METH. INDO.+MPHA , FRACTION OF ALCOHOL Fig MBT Spark Advance at CR=KLCR

78 70 twards the tp dead center. This is an indicatin f a higher burning velcity f indlene-alchl blends cmpared t purej indlene. At any cmpressin rati, the MBT spark advance f indlene-mpha blends is higher than that f indlene-methanl blends. The MBT spark advance is cntrlled by the thermdynamic prperties, such as temperature, pressure and burning velcity at ne specific engine perating cnditin. With the same intake air temperature, the temperature f the air-fuel mixture during the cmpressin prcess is related t the latent heat f the fuel. MPHA has a lwer latent heat, and a higher burning velcity than; methanl. Hwever, the cmbined effect n spark advance is a higher MBT spark advance in the case f indlene-mpha blends. Effect f Alchl n KLCR Table 5.2 shws the KLCR as a functin f methanl and MPHA cncentratin in the blend. As the methanl cncentratin increases frm t 40 percent, KLCR increases frm 7:1 t 10:1. Further increase in methanl cncentratin des nt cause any further increase in the KLCR. The variatin f KLCR with MPHA fractin is similar t that f methanl. Hwever, the kncking intensity with indlene-mpha was lwer than indlene-methanl blends. The maximum cmpressin rati that can be btained in this type f engine is 10:1. The KLCR f indlene-mpha is higher than indlene-methanl blend. The actual magnitude f the KLCR culd nt be btained, hwever, due t the limitatin f the engine.

79 71 TABLE 5.2 KNOCK LIMITED COMPRESSION RATIO Vlume Fx act in f Knck Limited Alchl in Indlene Cmpressin Rati 0.3 aethanl Methanl aethanl nethanl MPHA MPUA MPUA MPUA ^

80 72 Effect f Alchl n Brake Pwer Figures shw the brake pwer as a functin f fractin f alchls. In general, as the alchl cncentratin increases, brake pwer increases. At any cmpressin rati, r any fractin f alchl, the brake pwer f indlene-methanl is n different frm that f indlene-mpha. Effect f Alchl n Brake Specific Fuel Cnsumptin i i Figures shw the brake specific fuel cnsumptin as a functin f fractin f alchls. In general, as the alchl cncentratin increases, the brake specific fuel cnsumptin increases. At any cmpressin rati, indlene-methanl has a higher BSFC cmpared t indlene-mpha, specifically after 70 percent blend level. It is nted that BSFC is n a mass basis, (kg/kw-hr). With regards t BSFC, methanl is at a distinct disadvantage cmpared t MPHA because f its lwer heating value. Effect f Alchl n Brake Thermal Efficiency Figures shw the brake thermal efficiency as a functin f alchl fractin. At any cmpressin rati, the thermal efficiency f indlene-alchl blend is higher than that f indlene. In general, the indlene-methanl blends have a higher thermal efficiency than indlene-mpha blends. The brake thermal efficiency depends upn cmbustin perid, cling lsses, chemical equilibrium lsses, change in specific heat lsses, and latent heat f vaprizatin. The thermdynamic analysis f the entire cycle is

81 73 BRAKE POWER AT CR=5.0 INDO.-METH. & INDO.-MPHA BLENDS LEGEND =INDO.+METH. =INDO.+MPHA O.l FRACTION OF ALCOHOL i.o Fig Brake Pwer at CR=5:1

82 1 74 BRAKE POWER AT CR=6.0 INDO.-METH. & INDO.-MPHA BLENDS LEGEND INDO.+METH. INDO.-fMPHA 'It! '111! 'Hi I cc LU CL LU < CC CO FRACTION OF ALCOHOL 1.0 Fig Brake Pwer at CR=6 :

83 75 BRAKE POWER AT CR=7.0 INDO.-METH. & INDO.-MPHA BLENDS LEGEND INDO.+METH. INDO.-fMPHA FRACTION OF ALCOHOL 1.0 Fig Brake Pwer at CR=7:1

84 76 BRAKE POWER AT KLCR INDO.-METH. & INDO.-MPHA BLENDS LEGEND INDO.+METH. INDO.+MPHA m m O Q ^ 'III] I < tr CD FRACTION OF ALCOHOL 1.0 Fig Brake Pwer at CR=KLCR

85 77 FUEL CONSUMPTION AT C.R =5.0 INDO.-METH. & INDO.-MPHA BLENDS = LEGEND INDO.+METH. INDO.+MPHA :-:: ^ ^<:^ FRACTION OF ALCOHOL Fig BSFC at CR=5:1