Effect Injection Timing

8/12/2019 Effect Injection Timing

1/18

An ASAE/CSAE Meeting Presentation Paper Number: 046081

Fuel Property Effects on Injection Timing, Ignition

Timing and Oxides of Nitrogen Emissions from

Biodiesel-Fueled Engines

Mustafa Ertunc TatOsmangazi UniversityMechanical Engineering DepartmentBati Meselik/Eskisehir26480 TurkeyEmail: [email protected]

Jon H. Van Gerpen*University of IdahoDepartment of Biological and Agricultural EngineeringP.O. Box 440904Moscow, Idaho 83844-0904

Telephone Number: (208) 885-7891Fax Number: (208) 885-7908Email: [email protected]

Paul S. WangUniversity of IdahoDepartment of Biological and Agricultural EngineeringP.O. Box 440904Moscow, Idaho 83844-0904Telephone Number: (208) 885-7891Fax Number: (208) 885-7908Email: [email protected]

*corresponding author

Written for presentation at the2004 ASAE/CSAE Annual International Meeting

Sponsored by ASAE/CSAEFairmont Chateau Laurier, The Westin, Government Centre

Ottawa, Ontario, Canada1 - 4 August 2004


2/18

2

Abstract . Biodiesel is an environmentally friendly alternative diesel fuel consisting of the alkylmonoesters of fatty acids. It is obtained from triglycerides through the transesterfication

process. Biodiesel has been observed to reduce most engine exhaust pollutants with theexception of oxides of nitrogen (NO x ), which generally increase by 5 to 15%. The objective ofthis research was to determine the reason for the higher levels of NO x emissions. A conceptmap was developed to show the interrelationships between the fuel and engine variables thataffect NO x production. It was determined that a change in combustion timing caused bychanges in the fuel properties between diesel fuel and biodiesel is an important source of theNO x increase. The properties investigated in this research included the lower heating value,volatility, density, speed of sound, bulk modulus, and cetane number of biodiesel.

It was found that half of the start of combustion advance associated with biodiesel originatedfrom a start of injection advance that was split approximately evenly between the automatictiming advance of the pump as it injects more fuel to compensate for the lower heating value ofbiodiesel and the effect of the bulk modulus, viscosity, and density of the fuel. At the sametemperature, the fuel delivery of biodiesel was higher than for petroleum-based diesel fuelbecause of the higher viscosity of biodiesel. At the same viscosity level, it was found that thefuel delivery of petroleum-based diesel fuel was higher than for biodiesel. This was attributed tothe metering orifices in the fuel injection pumps restricting the amount of fuel flow for moredense fuels. The other half of the start of combustion timing advanc e was due to the highercetane number of the biodiesel.

Keywords. biodiesel, diesel fuel, alkyl esters, fuel injection, diesel engine, diesel combustion,diesel emission, NO x emission.

Introduction

Biodiesel is an environmentally friendly alternative diesel fuel consisting of the alkyl monoestersof fatty acids. It is obtained from triglycerides through the transesterfication process. Biodieseluse in diesel engines reduces diesel engine exhaust emissions with the exception of nitrogen

oxides (NO x) (Sharp et al., 2000). Most biodiesel emissions data has been collected fromengine dynamometer testing using the Federal Test Procedure (FTP-75) and these testsgenerally give 5% to 15% increases in NO x. In contrast to these engine dynamometer tests, invehicle testing with chassis dynamometers, biodiesel has been observed to produce less NO x emission than petroleum-based No. 2 diesel fuel (Peterson and Reece, 1996; Peterson et al.,1999). The objective of this project was to determine the reasons for the higher NO x emissionsof diesel engines fueled by biodiesel and why these emissions would be affected by testprocedure.

Monyem et al. (2001) performed steady state engine tests investigating the effect of injectiontiming on biodiesel emissions. They used a John Deere 4276T turbocharged diesel enginefueled with oxidized and nonoxidized biodiesel and No. 2 diesel fuel. They reported that NO x emissions increased from 0.5% to 18% for the two neat biodiesel fuels and for biodiesel blendsat all injection timings, relative to the base diesel fuel. They found a linear relationship betweenthe NO x emission and injection timing. The most significant finding was that when biodiesel anddiesel fuel were compared at the same start of combustion timing, the biodiesel produced lessNO x emission than diesel fuel. Monyem et al. (2001) also reported that the actual injectiontiming was advanced about 2.3 for the neat biodiesel fuels compared to diesel fuel at the samefuel injection pump setting. The injection timing advance was attributed to the physical propertydifferences between biodiesel and diesel fuel.


3/18

3

Peterson et al. (1999) and Peterson and Reece (1996) investigated the effect of biodieselfeedstock on the regulated emissions for chassis dynamometer tests of a pickup truck. Theypointed out that even though most studies showed NO x increases with biodiesel fuel, theirresearch with a chassis dynamometer showed decreases in the oxides of nitrogen withbiodiesel even when the same engine was tested. A Dodge 2500 pickup truck with a CumminsB5.9 turbocharged and intercooled direct injection diesel engine was used. They found thatlower iodine numbers, which corresponds to fewer carbon-carbon double bonds and a moresaturated feedstock, could be correlated with lower NO x. They showed that when the iodinenumber was increased from 8 to 129.5, the NO x increased by 29.3%. They were able to use alinear relationship to correlate iodine number and NO x emission with an R 2 of 0.091. The mostsignificant difference between engine dynamometer and chassis dynamometer tests is that theaverage engine torque level is much less for the chassis dynamometer tests than for the FTPtransient engine tests.

McCormick et al. (2001) examined the effect of biodiesel source material and chemical structureon the regulated emissions from a heavy-duty diesel engine. They specifically focused on theimpact of biodiesel's fatty acid chain length and the number of double bonds on the emissions ofNO x and particulates (PM). A heavy-duty truck engine was tested and the U.S. E.P.A. heavy-duty Federal Test Procedure was used. McCormick et al. found that the molecular structure ofbiodiesel can have a significant effect on diesel engine emissions. They showed that density,cetane number, and iodine number are highly correlated with each other and increasing densityand decreasing cetane number increased NO x emission. They found that increasing thenumber of double bounds can be correlated with increased NO x emission. They observed thatNO x emission increased with decreasing chain length for fully saturated fatty acid esters. Theyreported no significant difference in NO x and PM emission of methyl and ethyl esters of bodieselfuels with identical fatty acid distributions.

Theory of nitric oxide formation

The combustion process of hydrocarbon fuels can be considered to be a sequence ofelementary reaction processes that break the fuel into smaller molecules. Highly reactiveradical species are produced and consumed by the reactions. Most of the reactions areexothermic, so the temperature rises rapidly as the reactions proceed and the reactantsapproach an equilibrium mixture consisting mostly of CO, CO 2, H 2, H 2O, O 2, and N 2, as well asthe radicals O, N, H, HO 2 and OH. Because the nitric oxide (NO) formation reactions requiretime and high temperatures, most of the NO is formed in high temperature, post-flame regionsof the cylinder where near-equilibrium conditions exist. Nitric oxide is the predominant speciesin NO x.

The state of the equilibrium combustion products depends only on the temperature andpressure and the relative amounts of C, H, O, and N atoms. Further, for the same amount offuel energy delivered to the cylinder, the amounts of C, H, O, and N present in the products aresimilar for biodiesel and No. 2 diesel fuel. Since the in-cylinder mass will be the same for thetwo fuels, the pressure is mostly dependent on the temperature. Therefore, the temperature is

the key variable that determines NO production (Heywood, 1988 and Lavoie, 1970). Changesto the engine variables that increase the temperature in the post-flame gases can be expectedto increase NO emissions. Note that this simple model for NO production does not incorporatemany of the complexities of the heterogeneous fuel-air mixture in the diesel engine. However, itis consistent with the observation that the primary variable that affects NO levels in dieselengines is injection timing and it accurately predicts the effects of higher cetane number and theresulting decreased level of premixed burning, since these tend to decrease gas temperatureand NO.


4/18

4

A concept map, shown in Figure 1, was developed to organize the information about thevariables that affect NO formation and their interrelationships. In the concept map, it wasassumed that the combustion timing and the combustion rate will determine the temperature inthe post-flame gases and thus the NO emission. Early timing causes combustion to occurcloser to TDC and perhaps during the compression process, increasing the pressure,temperature, and NO emission (Heywood, 1988; Lyn and Valdmanis, 1968; and Wong andSteere, 1982). Combustion timing in a diesel engine is determined by the injection timing, or thestart of injection, and the ignition delay, which is the time between the start of injection and thestart of combustion (Heywood, 1988; Oven and Coley, 1995; Lavoie et al., 1970; Lyn andValdmanis, 1968; and Wong and Steere, 1982). Shorter ignition delay advances the start ofcombustion and the ignition delay time is mostly affected by the fuel's cetane number. If thecetane number is higher, the ignition delay time will be shorter and the start of combustion willcome earlier, which tends to increase NO. However, shorter ignition delay also tends todecrease premixed combustion, which usually decreases NO. Premixed combustion is the termused for the period of rapid combustion that follows autoignition and involves the fuel that hasbeen prepared for combustion during the ignition delay period. Which effect is dominantdepends on the specific engine design and operating conditions.

Biodiesel has a 12% lower energy content than diesel fuel by weight and when a greater volumeof fuel is injected to correct for this, some fuel injection pumps will advance the start of injectiontiming, causing an additional increase in NO x emission. This will be identified as the "pumpeffect" in later discussions. In addition, the faster propagation of pressure waves caused bybiodiesel's higher speed of sound and the more rapid pressure rise that results from biodiesel'sgreater bulk modulus may shift the injection timing settings from their optimized factory settings,leading to earlier combustion (Tat and Van Gerpen, 1999; Tat and Van Gerpen, 2000; and Tatet al., 2000).

The combustion rate also has a significant effect on NO production. More premixed combustionmeans a high initial rate of combustion which causes the fuel to burn earlier, resulting in highergas temperatures and increased NO production (Heywood, 1988). Cetane number and fuel

volatility are the two most important fuel properties that determine the amount of premixedcombustion and thus the combustion rate (Tat and Van Gerpen, 1999; 2000). Biodiesel's highcetane number is expected to shorten the ignition delay period and thus lower the amount offuel that is involved with the premixed combustion portion of the biodiesel combustion, loweringNO emission. Biodiesel's lower volatility decreases the amount of fuel vaporized during theignition delay and, therefore, also decreases premixed combustion. Properties that could affectatomization and spray penetration were not considered to be as important. Other factors thatare unrelated to fuel properties, such as air swirl, can affect the combustion rate by causingmore rapid fuel-air mixing. However, since these factors were the same for both fuels, theywere not considered likely to be the source of the higher NO levels for biodiesel.

Materials and Methods

A four-stroke, four-cylinder, turbocharged, direct injected John Deere 4045T diesel engine wasused for this research. The engines operating condition was held constant at 352.5 N-m (260 ft-lbf) of torque and 1400 rpm. The engine had a Stanadyne model D8DB4429-5415 distributor-type fuel pump. The basic specifications of the engine and the exhaust emission instrumentswere described by Tat (2003). Regular No. 2 diesel fuel was purchased from a local supplier foruse as a baseline fuel. Soybean oil methyl ester and yellow grease methyl ester were preparedat the Biomass Energy Conversion Center (BECON) facilities of the Iowa Energy Center inNevada, Iowa. The physical and chemical properties of the fuels are given in the Appendix.


5/18

Figure 1: Concept map of NO emission and combustion characteristics .

NO

Combustion Timing

Combus

Injection Timing

Ignition Delay (ID)

Cetane Number

Fuel Property EffectPump Effect

advance d combustion timing: higher temperature increase d NO

lower combustiolowers temperatu

and NO

early fuel injection timingleads to early combustiontiming

higher cetane numberearlier combustiontiming

more fuel early fuelinjection

higher cetane number lower premixed combustion rate

higher density, viscosity, speed ofsound, and bulk modulus leads toearly fuel injection

higher cetane number shorter ignition delay

longer ID giveslate combustiontiming longer ID gives m

premixed combus

early injectiontiming longer ID


6/18

6

In this research, the impact of biodiesel's 8% lower heating value (by volume) and the 11% lowercompressibility of soybean oil biodiesel fuel than No. 2 diesel fuel were investigated. The testmatrix was completed with two different fuel injection pumps, designated as pump #1 and pump#2. The effect of the lower heating value and higher bulk modulus of biodiesel can be observedon the power and the start of injection timing. The lower heating value of biodiesel fuels isexpected to cause 8% less power, or for the same power level, a correspondingly greater volumeof fuel to be consumed. Because the Stanadyne injection pump has a fixed end of injection, thisextra amount of fuel sent to the engine by the fuel injection pump will cause an advance in thestart of injection timing. Besides this, the higher isentropic bulk modulus of the biodiesel is alsoexpected to affect the start of injection timing. It was mentioned earlier in the concept mapdiscussion that early injection and combustion timing have a significant effect on the NOproduction. Therefore, in this test matrix, soybean oil biodiesel and No. 2 diesel fuel's start ofinjection were compared both at the same power and at the same volumetric fuel consumptionlevel. The engine emissions, start of combustion, and ignition delays were also compared at thesame power level. The engine was run at load conditions of 100, 95, 90, 80, 70, 60, 50, 40, 30,20% of the maximum torque that could be obtained from each fuel.

Results and Discussion

Figure 2 presents the start of injection timing comparison for soybean-based biodiesel and No. 2diesel fuel for the range of brake mean effective pressures (BMEP) tested. It can be seen thatwhen the engine load decreases, the start of injection is retarded until the light load advancesystem in the pump is activated and the timing is advanced again. It is clearly shown that, forpump #1, the biodiesel fuel injection timing was more advanced than for regular diesel fuel at thesame BMEP, and the advance was about 1.34 for the range of loads between the release of thepumps light load advance mechanism and full load. This is the total timing advance due to thecombined effect of the injection pump load compensation advance due to the lower energycontent of biodiesel and the different physical properties of biodiesel, such as bulk modulus andviscosity.

01

2

3

4

5

6

7

8

9

10

11

2 3 4 5 6 7 8 9 10 11BMEP (Bar)

S t a r t o f

I n j e c t

i o n

B T D C ( )

Biodiesel Start of Injection

No. 2 Diesel Fuel Start of Injection

Figure 2. Start of injection comparison of soybean oil biodiesel and No. 2 diesel fuel at varying theload conditions from 100% to 20%, at 1400 rpm and with pump #1.


7/18

7

It can also be seen that the light load advance system engaged at a higher load level with regulardiesel (about 5 bar) than with biodiesel fuel (about 3 bar). This should have a significant effecton the light load emissions of the engine and may explain the lower NO x emissions observed bysome researchers using light load chassis dynamometer tests (Peterson and Reece, 1996;Peterson et al., 1999). From these data, it is not possible to discriminate the relative significanceof the pump effect and the bulk modulus effect on the timing. However, when the comparisonsare made based on the volumetric fuel delivery per injection, as presented in Figure 3, it is easierto separate the effects.

0

1

2

3

4

5

6

7

8

9

10

11

12

0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09Fuel Delivery cc/inj

S t a r t o f

I n j e c t

i o n

B T D C ( ) Biodiesel Start of Injection


1400 rpm80-400 Nt-mPump #1

Figure 3. Volumetric comparison of the fuel delivery versus start of injection of biodiesel and No. 2diesel fuels at varying load conditions.

Figure 3 presents the start of injection timing versus fuel delivery (cm 3 /injection) data for bothfuels. From this comparison it can be seen that the start of injection of biodiesel from soybean oilwas advanced about 0.68 relative to No. 2 diesel fuel when the same volume of fuel wasinjected. This comparison is made at a single value of intermediate load (7.67 bar) but it can be

seen on the figure that the offset is relatively constant. The injection pump is designed to retardthe start of injection as the volume of fuel is reduced. For the same volume of fuel, the differenceof 0.68 was initially assumed to be purely due to the effect of the speed of sound and isentropicbulk modulus. Therefore, 1.34 - 0.68 = 0.66 was proposed as the effect of the lower energy

content of biodiesel on the start of injection timing due to fuel injection pump advancement atconstant torque.

There is a second way to estimate the pump effect. When the slope of the fuel delivery wascalculated from Figure 2, it was found that an increase of 8% in the fuel delivery led to about0.54 degrees of injection advance for the biodiesel fuel. Other inferences to be made are thatthe start of injection timing curves are parallel to each other until the light load advancementsystem in the pump engages, which occurs earlier with diesel fuel than biodiesel. It should alsobe noted that later work to be presented indicates that the 0.54 - 0.68 difference may not besolely due to speed of sound and isentropic bulk modulus effects. Viscosity and densitydifferences may also be responsible for the variation in timing.


8/18

8

After an accident that was unrelated to the test, pump #1 was was replaced with a second pumpdesignated pump #2. This pump matched the characteristics of the engine but was a differentpart number (Stanadyne DB4429-5776). After minor calibration of the pump, the same testmatrix was repeated with pump #2.

It was observed that with pump #2 the start of injection timing of biodiesel was advanced about0.6 compared with diesel fuel at a BMEP of 7.67 bar. It should be noted that this is the samecondition that gave a difference of 1.34 in Figure 2 with pump #1. In Figure 4, the start ofinjection timing versus volumetric fuel delivery is presented for the pump #2. As shown in thisfigure, when the same volume of both fuels was injected, the start of injection timing was thesame, unlike the results obtained with pump #1. This means that either the higher speed ofsound and isentropic bulk modulus of biodiesel have no effect on the start of injection (in contrastto what was observed with pump #1) or there are other effects that cancel the speed of soundand the bulk modulus effects. In Figure 4, it also appears that there was some kind ofmechanical problem with the light load regulation system that was more severe with the dieselfuel.

2.00

3.00

4.00

5.00

6.00

7.00

8.00

9.0010.00

11.00

12.00

0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09Fuel Delivery, cc/inj

S t a r t o

f I n j e c

t i o n

( ) Biodiesel Start of Injection


Figure 4. Volumetric comparison of the fuel delivery versus start of injection of biodiesel and No. 2diesel fuels at varying load conditions from 100% to 20% at 1400 rpm with pump #2

Up to this point, the effect of fuel viscosity on timing was assumed to be negligible. However, asa result of the significant differences between the two pumps, it was speculated that the fuelviscosity might have an effect on the fuel injection timing. Fuel is used as a lubricant and coolantin the pump, and the amount of fuel delivered can be significantly affected by the leakage of fuelpast the plungers and the metering valve. It was also thought that the physical property effect onthe fuel injection timing could be a function of the pump speed. Therefore, viscosity and pumpspeed effects on the injection timing were investigated and are described elsewhere (Tat, 2003).

Since the light load advance system for pump #2 was clearly not working properly, a new pumpwas purchased, designated pump #3 (Stanadyne, DB4429-5415), and was tested for fuelinjection timing and fuel delivery values for the same fuel and speed conditions. To investigatethis, start of injection timing and fuel delivery comparisons were made at engine speeds of 1000,1400, 1800, and 2100 rpm with pump #2 and pump #3. The start of injection timing was thesame for both fuels at the same fuel delivery at all engine speed conditions, even though theinjection timing varied at different speed levels. While the differences between the fuels do notseem to be affected by speed, it was still believed that viscosity effects were a potential source of


9/18

9

the differences between pump #1 and the subsequent pumps. The control systems in the dieselfuel injection system can be affected by the high viscosity of biodiesel.

The viscosity of biodiesel is about twice that of diesel fuel. Viscosity is a strong function oftemperature with correlations for biodiesel available in the literature (Tat and Van Gerpen, 1999).During fuel compression a small amount of fuel is permitted to leak past the plunger forlubrication and the amount of this leakage fuel is directly correlated with the fuel viscosity and theclearance between the cylinder and the plunger. The more viscous the fuel, the less the amountof fuel that will leak. The amount of the leakage fuel is sufficiently important that the temperatureof the fuel has a significant impact on the engine power. The start of injection timing can also beaffected by the amount of fuel leakage. During the fuel compression process, the amount ofleakage can impact the pressure rise and the start of injection. The engine was run at fueltemperatures of 25, 30, 40, 50, and 55 C using soybean-based biodiesel and No. 2 diesel fuel.The engine was run at wide open throttle (wide open metering) at 1400 rpm so that the fuelinjection pump was delivering its maximum flow and this corresponds to an equal volume forboth fuels. Since the viscosities of the fuels were different, the leakage was different and thiscaused differences in the fuel delivery. Both pump #2 and pump #3 were tested. Due todamage from the accident mentioned earlier, pump #1 could not be tested.

Temperature versus fuel delivery of both fuels with both pumps is shown in Figure 5. At all

temperatures, a greater volume of biodiesel was injected compared with diesel fuel. This isassumed to be a result of biodiesels greater viscosity reducing leakage within the fuel pump. At40 C, the temperature at which fuel is usually supplied to the engine during our tests, about 1.2and 3.2% more biodiesel was injected than No. 2 diesel fuel with pump #2 and pump #3,respectively. At 40 C, the No. 2 diesel fuel delivery is about 0.087 cc/inj for both pumps.However, the biodiesel fuel delivery is 0.0885 cc/inj with pump #2 and 0.0900 cc/inj with pump#3. Therefore, 0.0015 cc/inj and 0.003 cc/inj more biodiesel was injected with pump #2 andpump #3, respectively, compared to diesel fuel. Using the data presented in Figure 2, anincrease in the volumetric fuel delivery of 0.0015 cc/inj and 0.003 cc/inj is expected to cause 0.2 and 0.5 advance in the start of injection timing with pump #2 and pump #3, respectively.

Figure 5: Fuel viscosity and fuel delivery comparison for biodiesel and diesel fuel with pump #2and pump #3.

0.083

0.084

0.085

0.086

0.087

0.088

0.089

0.09

0.091

0.092

0.093

25 30 35 40 45 50 55 Temperature, C

No. 2 Diesel Fuel Delivery cc/inj, pump #2Biodiesel Fuel Delivery cc/inj, pump #2No. 2 Diesel Fuel Delivery cc/inj, pump #3Biodiesel Fuel Delivery cc/inj, pump #3

F u e

l D e

l i v e r y ,

c c

/ i n

j


10/18

10

The kinematic viscosities of the fuels at each temperature were calculated using data from Tatand Van Gerpen (1999) and a plot of viscosity versus fuel delivery was prepared and is shownas Figure 6. The viscosity data in Figure 6 is based on the fuel inlet temperature and is assumedto be representative of the fuel temperature within the pump. It should be noted that even at thesame fuel viscosity and metering valve position (wide open), the quantity of fuel delivered is notthe same for the two fuels. At the same fuel viscosity, the quantity of diesel fuel delivered ishigher than for biodiesel. When the fuel temperature and the start of injection timings of fuels

were compared, it was found that, at the same fuel temperature, both the fuels had the samestart of injection timings. This confirms our observations from Figure 4. It can also be seen thatthe fuel delivery of No. 2 diesel fuel was twice as sensitive to viscosity change as the fueldelivery of biodiesel. This is based on the observation that the slope of No. 2 diesel isapproximately twice the slope of the biodiesel.

0.083

0.084

0.085

0.086

0.087

0.088

0.089

0.09

0.091

0.092

1.5 2 2.5 3 3.5 4 4.5 5 5.5 6 6.5Kinematic Viscosity of No. 2 Diesel and Biodiesel Fuel Consumption, cSt

F u e l

D e l

i v e r y c c

/ i n j

No. 2 Diesel Fuel Delivery cc/inj, pump #2Biodiesel Fuel Delivery cc/inj, pump #2No. 2 Diesel Fuel Delivery cc/inj, pump #3Biodiesel Fuel Delivery cc/inj, pump #3

Figure 6. Fuel viscosity and fuel delivery comparison for biodiesel and diesel fuel with pump #2and pump # 3.

The leakage flow between the cylinder and the plunger under high pressure can be modeled asa laminar flow between two flat plates and the governing equation would be Equation 1 (Munsonet al., 1998).

l

phq

**3

**2 3

= (1)

where q is the flow rate, h is the half distance between the plates, p is the pressure difference, l is the length of the plates and is the viscosity. For this case, all of the variables are the samefor both fuels except the viscosity. Equations 2 and 3 show how the relative amount of fuelleakage as a function of viscosity can be estimated.


11/18

11

58.047.467.2

40@

40@

40@

40@ ==

=

cSt cSt

q

q

C biodiesel

C diesel

C diesel

C biodiesel

(2)

90.155@

25@

25@

55@ =

=

C biodiesel

C biodiesel

C biodiesel

C biodiesel

q

q

(3)

The biodiesel fuel leakage at 40 C is estimated to be 58% of the value for #2 diesel fuel. It isalso estimated that the fuel temperature change in biodiesel from 25 to 55 C will increase theamount of leakage fuel by 90%. These calculations show that the viscosity of the fuels has avery significant effect on the amount of the leakage fuel. This is probably the primary reason forthe change in fuel delivery as the fuel temperature changes that was shown in Figure 6. It isunderstood that higher viscosity will increase the amount of fuel delivery per injection andincreased fuel delivery advances the start of injection timing.

However, Figure 6 also shows that, at the same viscosity level, the pump delivers more No. 2diesel fuel than biodiesel fuel at wide-open throttle. In Figure 7, the fuel delivery and the start ofinjection are compared at wide open throttle for different fuel temperatures. For the same

metering valve position and equal fuel delivery, The start of injection for No. 2 diesel fuel wasadvanced 0.25 to 0.35 relative to the biodiesel fuel.

7.90

8.00

8.10

8.20

8.30

8.40

8.50

8.60

8.70

0.083 0.084 0.085 0.086 0.087 0.088 0.089 0.09 0.091 0.092Fuel Delivery, cc/inj

S t a r t o f

I n j e c t

i o n , d e g

B T D C

No.2 Diesel Start of Injection, pump #3

Soybean Biodiesel Start of Injection,

Wide Open Throttle

55 C

30C

40 C

50 C

26 C

30 C

40 C

50 C

55 C

Figure 7: Fuel delivery and start of injection comparison for biodiesel and diesel fuel with pump #3

Figures 6 and 7 indicate that the flow is not only affected by viscosity, but is also affected bysome other property, probably the fuel density. Equation 4 shows the equation that describesthe flow of liquids through an orifice. Because the density is in the denominator, it is clear thatless dense fuels will have more flow through orifices at the same pressure drop. Equation 4 can


12/18

12

be reduced to Equation 5 when diesel fuel at 26 C and biodiesel fuel at 52 C are compared atthe same pressure difference.

[ ]21221

2 )/(1*)(*2

A A P P

AC q D

=

(4)

The density of biodiesel at 52 C is 0.863 g/cc and the density of the diesel fuel at 26 C is 0.836g/cc.

016.1836.0

836.0

26@

52@

52@

26@ ===C diesel

C biodiesel

C biodiesel

C diesel

q

q

(5)

It is found that the diesel flow rate is about 1.6% more than biodiesel fuel. This can explain thehigher fuel delivery at the same viscosity and also explains the start of injection advance thatoccurs with diesel fuel presented in Figure 7. Higher fuel delivery due to the lower density ofpetroleum-based diesel fuel also tends to advance the start of injection timing. Biodiesels higherdensity tends to oppose the effect of viscosity, speed of sound, and isentropic bulk modulus.

It should be noted that the 0.2 and 0.5 of advance is a significant portion of the 0.63 increasein the start of injection due to the higher speed of sound and bulk modulus of biodiesel that wasidentified with pump #1. These tests have tried to separate the effects of pump advance withload from the property effect by comparing the timing at the same volume of fuel injected. Nowthat the effect of viscosity has been identified as significant, it cannot be separated as easily.The effect of leakage in the pump will be such that the pump will be trying to inject a largerquantity of fuel than is actually injected, and will have the earlier start of injection timing thatcorresponds to this greater quantity of fuel, but will actually have a lower volume of fuel delivereddue to leakage during the pumping strokes. So, the technique proposed earlier of separating thetiming advance into its two major components by comparing at the same volume of fuel deliveredis not necessarily valid because the viscosity effect cannot be separated from the pump advance

that corresponds to the greater fuel delivery needed to compensate for biodiesel's lower energycontent. The viscosity effect is also expected to be highly variable between different pumps dueto variations in factory tolerances, which are different for each pump, and any accumulated wear.

Effect of Fuel Properties on Combustion Timing

The timing for the start of combustion as a function of engine load for biodiesel from soybean oiland No. 2 diesel fuel is shown in Figure 8. It is possible to see the light load advancementdifference between the two fuels at about 2-4 bar of BMEP. Start of injection, start ofcombustion, and ignition delay comparisons were made at the intermediate load condition of7.67 bar. This load condition was chosen for the comparison because it was a representative

engine condition and it is a point where the measured values were relatively consistent. At thisload, biodiesel's start of combustion timing was advanced by 2.4 crank angle degrees relative todiesel fuel. When the ignition delay periods of biodiesel and diesel fuel were compared it wasfound that biodiesel had a shorter ignition delay period than diesel fuel at all load conditions. Itwas also found that the ignition delay periods were longer as the load decreased and thedifference between biodiesel and diesel fuel increased until the light load advancement wasengaged. The ignition delay difference between the two fuels at 7.67 bar BMEP was 1.06.When added together, the lower heating value effect of 0.66, the higher isentropic bulk modulusand other fuel property effects of 0.68 , and the shorter ignition delay period of 1.06 due to


13/18

13

higher cetane number, the total start of combustion advances 2.4 which matches what weobserve in Figure 8.

Figure 8. Comparison of the start of combustion timings of Biodiesel and No. 2 Diesel fuels atvarying load conditions from 100% to 20% at 1400 rpm and with pump #1

The brake specific oxides of nitrogen emission (BSNOx) of biodiesel from soybean oil and No. 2diesel fuels are compared in Figure 9. It is seen that at the maximum load conditioncorresponding to a BMEP of 10.5 bar, the biodiesel produced about 16% more NO x than thediesel fuel. When the light load advance system was engaged for No. 2 diesel fuel at between 4

and 5 bar, the No. 2 diesel fuel started to produce more NOx than the biodiesel and thedifference was increased until the light load advancement system engaged for biodiesel. Evenafter the system was engaged and the biodiesel start of injection was slightly more advanced,the BSNOx emission of biodiesel was less than for the No. 2 diesel fuel. This might explain thelower biodiesel NO x emission results obtained by Peterson et al. [44, 44, 45]. In chassisdynamometer tests, the engine load conditions are much lower than for engine dynamometertests. Therefore, the difference in the engagement conditions for the light load advancemechanism for both fuels may explain the lower NO x and higher particulate emissions of thosetests.

The brake specific unburned hydrocarbon (BSHC), the brake specific carbon monoxide (BSCO),and the Bosch Smoke Numbers (SN) emissions were measured and reported in detail in Tat(2003). BSHC emissions of both fuels increased as the load decreased. The BSHC emission ofbiodiesel was less than No. 2 diesel fuel at most of the load conditions, and at the maximum loadcondition, the BSHC emission was about 30% less than that of diesel fuel.


14/18

14

4

5

6

7

8

9

10

11

12

13

14

2 3 4 5 6 7 8 9 10 11BMEP (Bar)

B r a

k e S p e c i

f i c N O x E m i s s i o n

( g / k W - h r ) Biodiesel

No.2 Diesel

Figure 9. Comparison of the brake specific oxides of nitrogen (BSNOx) of biodiesel and No. 2diesel fuel at varying the load conditions from 100% to 20% at 1400 rpm and with pump #1

The BSCO emission of biodiesel was less than the BSCO emission of No. 2 diesel fuel at almostall load conditions in a manner that was similar to the BSHC results. The Bosch SmokeNumbers (SN) of the fuels were also compared. The SN of biodiesel was considerably less thanfor No. 2 diesel fuel at all load conditions. Biodiesel's SN is about 50% less than diesel fuel'sand the difference decreases as the load decreases to 2 bars, which is 20% of the maximumpower, where the smoke numbers for the two fuels are about equal.

Conclusion

The properties of biodiesel affect the fuel injection system and can cause an advance in the fuelinjection timing. It was found that about half to all of the advancement in injection timing was dueto the increase in the fuel delivery needed to overcome the power loss that results frombiodiesels lower energy content. As much as half of the advancement was due to the effects ofhigher speed of sound, isentropic bulk modulus, viscosity, and density of biodiesel fuel. Thehigher cetane number of soybean biodiesel shortens the ignition delay and advances the start ofcombustion, and this also contributes to the higher NOx emission of soybean biodiesel. This iscurrently the subject of on-going research.

At the same temperature, the fuel delivery of biodiesel was higher than for petroleum-baseddiesel fuel because of the higher viscosity of biodiesel. At the same viscosity level, it was foundthat the fuel delivery of petroleum-based diesel fuel was higher than for biodiesel fuel. This was

judged to be a result of the orifices in the fuel injection pumps restricting the amount of fuel flow

for more dense fuels and the reduction in the amount of fuel delivery affecting the fuel injectiontiming. It was also concluded that the effects of some properties are coupled and cannot beseparately identified by changing one without changing the other, such as density and viscosity.


15/18

15

References

Chang, Y.Z.D., Van Gerpen, J.H., Lee, I., Johnson, L.A., Hammond, E.G. and Marley, S.J. 1996.Fuel Properties and Emissions of Soybean Oil Esters as Diesel Fuel. Journal of

American Oil Chemists Society 73(11): 1549-1555.

Graboski, S.M., Ross, J.D. and McCormick, R.L. 1996. Transient Emissions from No. 2 Dieseland Biodiesel Blends in a DDC Series 60 Engine. Society of Automotive Engineers paper961166.

McCormick, R.L., Graboski, M.S., Alleman, T.L., and Herring, A.M. 2001. Impact of BiodieselSource Material and Chemical Structure on Emissions of Criteria Pollutants from aHeavy-Duty Engine. Environmental Science and Technology 35(9):1742-1747.

Heywood, J.B. 1988. Internal Combustion Engine Fundamentals. McGraw-Hill, New York.

Lavoie, G.A., Heywood, J.B. and Keck, J.C. 1970. Experimental and Theoretical Investigation ofNitric Oxide Formation in Internal Combustion Engines. Combustion ScienceTechnology. 1:313-326.

Lyn, W.T. and Valdmanis, E. 1968. Effects of Physical Factors on Ignition Delay. Society of Automotive Engineers Paper No. 680101, SAE, Warrendale, Penn.

McClements J.D. and Povey, M.J.W. 1988. Ultrasonic Velocity Measurements in Some LiquidTriglycerides and Vegetable Oils, The Journal of the American Oil Chemists Society65(11):1787-1790.

McDonald, J.F., Purcell, D.L., McClure, B.T., and Kittelson, D.B. 1995. Emission Characteristicsof Soy Methyl Ester Fuels in an IDI Compression Ignition Engine. Society of AutomotiveEngineers Paper No. 950400, SAE, Warrendale, Penn.

Monyem, A., Van Gerpen, J.H., and Canakci, M. 2001. The Effect of Timing and Oxidation onEmissions. Transactions of the American Society of Agricultural Engineers, Vol 44 (1),2001, 35-42.

Monyem, A. 1998. The effect of biodiesel oxidation on engine performance and emissions.Ph.D. thesis, Iowa State University.

Munson, B.R., Young, D.F., and Okiishi, T.H. 1998. Fundamentals of Fluid Mechanics, ThirdEdition, John Wiley & Sons.

Owen, K. and Coley, T. 1995. Automotive Fuels Reference Book, Second Edition. Society of Automotive Engineers, Inc. Warrendale, PA.


16/18

16

Peterson C.L. and Reece, D.L. 1996. Emissions Characteristics of Ethyl and Methyl Ester ofRapeseed Oil Compared with Low Sulfur Diesel Control Fuel in a Chassis DynamometerTest of a Pickup Truck Transactions of the ASAE, 39(3):805-816.

Peterson, C.L., Taberski, J.S., and Thompson, J.C. 1999. The effect of biodiesel feedstock onRegulated Emissins in Chasis Dynamometer Tests of a Pickup Truck. Written forPresentation at the 1999 ASAE/CSAE-SGCR Annual Internatinal Meeting Paper No.996135.

Schmidt, K. and Van Gerpen, J.H. 1996. The Effect of Biodiesel Fuel Composition on DieselCombustion and Emission, Society of Automotive Engineers paper 961086.

Senatore, A., Cardone, M., Rocco, V., and Prati, M.V. 2000. A Comparative Analysis ofCombustion Process in D.I. Diesel Engine Fueled with Biodiesel and Diesel Fuel. Societyof Automotive Engineers Paper No. 2000-01-0691, SAE, Warrendale, Penn.

Sharp, C.A., S.A. Howell, J. Jobe. 2000. The Effect of Biodiesel Fuels on Transient Emissionsfrom Modern Diesel Engines, Part I Regulated Emissions and Performance Society of

Automotive Engineers Paper No. 2000-01-1967, SAE, Warrendale, Penn.

Tat, M.E. 2003. Investigation of oxides of nitrogen emissions from biodiesel-fueled engines.Ph.D. thesis, Iowa State University.

Tat, M.E. and Van Gerpen, J.H. 1999. The Kinematic Viscosity of Biodiesel and Its Blends withDiesel Fuel. Journal of American Oil Chemists Society, 76(12):1511-1513.

Tat, M.E. and Van Gerpen, J.H. 2000. The Specific Gravity of Biodiesel and Its Blends withDiesel Fuel. Journal of American Oil Chemists Society, Vol. 77(2):115-119.

Tat, M. E., Van Gerpen, J.H., Soylu, S., Canakci, M., Monyem, A. and Wormley, S. 2000. TheSpeed of Sound and Isentropic Bulk Modulus of Biodiesel at 21 C from AtmosphericPressure to 35 MPa. Journal of American Oil Chemists Society, 77(3): 285-289.

Tat, M.E. and Van Gerpen, J.H. 2003. Speed of Sound and Isentropic Bulk Modulus of AlkylMonoesters at Elevated Temperatures and Pressures, The. Journal of American OilChemists Society, 80(12):1-8.

Wong, C.L. and Steere, D.E. 1982. The Effects of Diesel Fuel Properties and Engine OperatingConditions on Ignition Delay. Society of Automotive Engineers Paper No. 821231, SAE,

Warrendale, Penn.


17/18

17

Appendix

Table A.1. The physical and chemical properties of No. 2 diesel fuel, soybean oil methyl ester,and yellow grease methyl ester

Test Property No 2 diesel fuel Soybean OilMethyl Ester

Yellow GreaseMethyl Ester

Carbon (% mass) d 86.66 a 77.00 76.66

Hydrogen (% mass) d 12.98 a 12.18 12.33

Ox en % mass d - 10.82 11.01

C/H Ratio 6.676 6.322 6.217

Sulfur (% mass) a 0.034


18/18

18

Table A.2. Fatty Acid Profiles for Biodiesel Fuels from Soybean Oil and Yellow Grease

Fatty Acid Profile*Soybean Oil

BiodieselYellow Grease

Biodiesel

C14:0 Tetradecanoic (Myristic)

Documents

Effect Injection Timing