Fuel Economy Requirements

In addition to emission measurement, each manufacturer must determine the fuel consumption in MPG for each type of vehicle and must compute the corporate average fuel economy (CAFE) for all vehicles of all types produced in a year. Fuel consumption is measured during both an urban and a highway test, and the composite fuel economy is calculated.

Table 5.1 is a summary of the exhaust emission requirements and CAFE standards for a few representative years. It shows the emission requirements and increased fuel economy required, demonstrating that these regulations have become and will continue to become more stringent with passing time. Not shown in Table 5.1 is a separate regulation on nonmethane hydrocarbon (NMHC). Because of these requirements, each manufacturer has a strong incentive to minimize exhaust emissions and maximize fuel economy for each vehicle produced.

Table 5.1 Emission and MPG requirements.

New regulations for emissions have continued to evolve and encompass more and more vehicle classes. Present-day regulations affect not only passenger cars but also light utility vehicles and both heavy- and light-duty trucks. Furthermore, regulations apply to a variety of fuels, including gasoline, diesel, natural gas, and alcohol-based fuels involving mixtures of gasoline with methanol or ethanol.

As an example, we present below the standards that were written for the vehicle half-life (5 years or 50,000 miles–whichever comes first) and full life cycle (10 years or 100,000 miles) as of 1990. The standards are:

These regulations were phased in according to the following schedules:

There are many details to these regulations that are not relevant to the present discussion. However, the regulations themselves are important in that they provided motivation for expanded electronic controls.

Meeting the Requirements

Unfortunately, as seen later in this chapter, meeting the government regulations causes some sacrifice in performance. Moreover, attempts to meet the standards exemplified by Table 5.1 using mechanical, electromechanical, hydraulic, or pneumatic controls like those used in pre-emission control vehicles have not been cost-effective. In addition, such controls cannot operate with sufficient accuracy across a range of production vehicles, over all operating conditions, and over the life of the vehicle to stay within the tolerance required by the EPA regulations. Each automaker must verify that each model produced will still meet emission requirements after traveling 100,000 miles. As in any physical system, the parameters of automotive engines and associated peripheral control devices can change with time. An electronic control system has the ability to automatically compensate for such changes and to adapt to any new set of operating conditions and make electronic controls a desirable option in the early stages of emission control.

The Role of Electronics

The use of digital electronic control has enabled automakers to meet the government regulations by controlling the system accurately with excellent tolerance. In addition, the system has long-term calibration stability. As an added advantage, this type of system is very flexible. Because it uses microcomputers, it can be modified through programming changes to meet a variety of different vehicle/engine combinations. Critical quantities that describe an engine can be changed relatively easily by changing data stored in the system’s computer memory.

Inputs to Controller

Figure 5.6 identifies the major physical quantities that are sensed and provided to the electronic controller as inputs. They are as follows:

Output from Controller

Figure 5.7 identifies the major physical quantities that are outputs from the controller. These outputs are

This chapter discusses the various electronic engine control functions separately and explains how each function is implemented by a separate control system. Chapter 7 shows how these separate control systems are being integrated into one system and are implemented with digital electronics.

For certain readers of this book, a brief review of engine configuration and operation may be helpful. Although several types of engines have found application as the prime mover in automobiles, the one most commonly used continues to be the multicylinder, four-stroke IC engine as explained earlier in the chapter. The configuration and operation of electric propulsion (e.g., in hybrid vehicle) are discussed in Chapter 6.

The configuration of a single cylinder of an IC engine is depicted in Figure 5.8a. Mechanical power is produced by the engine in the form of torque acting on the rotating crank shaft. There are four basic engine processes that occur during the two complete revolutions of the crankshaft that occur during any single cycle of operation. This engine configuration includes a component called the piston, which fits within a cylinder and is mechanically linked to the crankshaft by the connecting rod. Airflow into and out of the cylinder is controlled by poppet valves (simply called the valves here). One of these is termed the intake valve and the other the exhaust valve. Additional components of the engine include a so-called “intake port system” consisting of a system of passageways (e.g., tubes) that direct fuel/air mixture into the engine and a so-called “exhaust port system” that directs the products of combustion out of the engine.

During any single cycle of engine operation (involving two complete rotations of the crankshaft), there are four portions of the crankshaft rotation called strokes. Each stroke corresponds to piston (reciprocating) motion from its highest point (called “top-dead-center” or TDC) to the lowest point (called “bottom-dead-center” or BDC). These four strokes of any given engine cycle are termed intake, compression, power, and exhaust strokes. During the intake stroke, the piston moves from TDC to BDC. During most of this stroke, the intake valve is open and the exhaust valve is closed. During this stroke, air mixed with fuel is pumped into the cylinder by the positive differential pressure between the intake port and the cylinder internal pressure. During the compression stroke, the piston moves from BDC to TDC. Both intake and exhaust valves are closed. For an ideal IC engine, this compression is adiabatic and modeled by the following expression:

(1)

where V_c is the cylinder contained volume (between the piston upper surface and the top of the combustion chamber), p_c the combustion chamber pressure, and γ the ratio of specific heat at constant pressure to the specific heat at constant volume. For the intake air/fuel mixture for air/gasoline mixture.

The term adiabatic refers to a process with zero heat loss. The compression process for an actual engine is not adiabatic since heat is lost (e.g., through the cylinder sidewalls). The actual function p_c(V_c) for a practical engine is shown graphically later in this chapter.

The difference between combustion chamber maximum volume (piston at BDC) and its minimum volume (at TDC) (which is often called the “clearance volume”) is called the cylinder displacement V_D. The ratio of cylinder pressure at TDC to that at BDC is called the compression ratio r. At some point before the piston reaches TDC, the spark is generated and combustion of the fuel/air mixture is initiated and cylinder pressure rises rapidly.

During the next stroke, the power stroke, the cylinder pressure, acting on the piston via the connecting rod, applies a torque to the crankshaft. This expansion ideally would be adiabatic, but in fact is not adiabatic due to heat losses (as is demonstrated later from measurements made on an actual engine).

During the final stroke, the exhaust stroke, the piston again moves from BDC to TDC. The exhaust valve is open during most of this stroke, and the products of combustion (mostly CO₂ and H₂O) are pumped out of the cylinder into the exhaust system and released through this system to the atmosphere.

The actual point in the 720° crankshaft rotation angle at which the valves open and close (called valve timing) is determined by a mechanism that includes the camshaft and mechanical linkage connecting it to the valves. The camshaft, which is illustrated in Figure 5.8b, has lobes that force the valves open against the restoring forces of valve springs that otherwise hold the valves closed. The reader should imagine that the valves depicted in Figure 5.8a extend to the end of the valve stem depicted in Figure 5.8b. The camshaft is coupled via a gear system to the crankshaft such that it rotates at half the speed of the latter. This mechanism assures that the valves operate synchronously within each engine cycle. During the development period of any new engine design, the optimal valve timing is determined. In Chapter 6, the drive mechanism for the camshaft and the means for rotating it at half the crankshaft angular speed are explained with respect to a system known as variable valve phasing. For the present, however, the discussion is focused on basic engine processes.

Energy is produced by a four-stroke/cycle internal combustion engine only during the power stroke. The energy produced during this stroke must be greater than the energy required for the other strokes as well as by internal friction losses. Normally, in any well-designed engine the power stroke energy far exceeds the magnitude of all mechanical losses, thereby yielding net output energy.

A basic method of evaluating the output mechanical energy involves the so-called indicator diagram, which is also a plot of the p_c vs. V_c for the entire cycle. Figure 5.9 represents an indicator diagram for an ideal engine cycle (in the sense of no heat loss to the engine) by the dashed curve and the p_c(V_c) plot for an actual engine by the solid curve.

For the ideal engine cycle, the valves are assumed to open or close at exactly TDC or BDC. Any time delays associated with the gas dynamics of intake and exhaust are taken to be negligible. In Figure 5.9, point a is at BDC with the cylinder filled with fuel/air mixture. The segment from point a to point b corresponds to the compression stroke. Ignition occurs at point b and the combustion chamber pressure increases instantaneously to point c, which is the beginning of the power stroke. The power stroke is represented by the segment from point c to point d. At point d, the exhaust valve opens and pressure drops to the exhaust system pressure (p_e) at point a. The segment from point a to point e corresponds to the exhaust stroke. At point e, the exhaust valve closes and the intake valve opens. The intake stroke corresponds to the segment from point e to point a where the cycle began and where the next engine cycle commences. For this ideal engine cycle, both intake and exhaust gas pressures are taken to be at atmospheric pressure.

The indicator diagram for an actual engine is depicted by the solid curve for which the compression stroke is the segment of the solid curve from point 1 to point 2. Notice that the pressure at point 1 is slightly below atmosphere pressure, which occurs because of pressure losses in the intake system. At point 2, ignition occurs and the pressure rises to its maximum value at point 3. The expansion from point 3 to point 4 is somewhat different in shape than the ideal adiabatic expansion due to heat losses. Pressure continues to drop after the exhaust valve opens (near point 4), but remains slightly above atmospheric pressure due to “back pressure” in the exhaust system. The exhaust stroke occurs between point 4 and point 0. The intake stroke occurs from point 0 to point 1 at pressure somewhat below atmospheric due to pressure drop across the throttle plate as well as some pressure losses in the intake system. From point 1, the cycle begins again.

The net energy/cycle (called the indicated energy, W_i) is given by the contour integral around the curve from points 1–6 below:

(2)

The only positive contribution to this integral comes from the portion from point 3 to point 4 (i.e., the power stroke). The energy/cycle is influenced markedly by the timing of the valve openings and closing as will be explained in Chapter 7. As clearly shown from Figure 5.9, in any practical engine the indicator diagram deviates from the ideal as indicated by the continuous curve of Figure 5.9.

Torque

Engine torque is produced on the crankshaft by the cylinder pressure pushing on the piston during the power stroke. In an IC, engine torque is produced at the crankshaft as explained below. The torque that is applied to the crankshaft is called “indicated torque T_i.” The output torque from the engine at the transmission end of the crankshaft differs from T_i due to friction and pumping losses and is called the brake torque (denoted T_b).

For an understanding of the various torques at different points in the powertrain, it is helpful to refer to Figure 5.10, which illustrates the geometry of a single cylinder in a four-stroke IC engine. Figure 5.10 shows the centerline of the cylinder which is a line along the cylinder axis through the crankshaft rotational center. The piston is connected via the connecting rod to the crankshaft. The connecting rod is fastened to the piston via the piston pin about which this rod can rotate. During the power stroke, a torque is applied to the crankshaft resulting from the force acting on the piston due to combustion chamber pressure acting through a lever arm which is proportional to the crank throw R and which varies with crankshaft angular position (θ_e). This torque is known as the indicated torque to distinguish it from other torque acting on the crankshaft and can be computed as explained below.

Figure 5.10 presents the geometry of the piston, connecting rod, and crankshaft in a way which permits a model for the indicated torque T_i(θ_e) as a function of crankshaft angle (θ_e) to be developed. In Figure 5.10 the connecting rod length is denoted L_r and the radius from the crankshaft axis of rotation to the center of the connecting rod journal is denoted R. It is this radius of the crankshaft rotation which provides the lever arm for the production of indicated torque due to the force on the top of the piston due to combustion chamber pressure (P_c(θ_e)). In many engines, the piston pin is located slightly off the cylinder centerline (C_L) in a plane that is orthogonal to the crankshaft axes of rotation, which benefits torque production. The piston pin offset from the cylinder C_L is denoted δ in Figure 5.10. The angle between the connecting rod plane of symmetry and the cylinder axis is denoted β. Owing to the piston offset, the indicated torque for a piston on the down stroke is given by

(3)

On the up stroke T_i is given by

(4)

where A is the piston cross-sectional area. The factors represent the lever arm through which torque is applied to the crankshaft.

The combustion chamber pressure for a representative four-stroke reciprocating IC engine is shown in Figure 5.11 for a complete engine cycle (720° of crankshaft rotation) beginning at –180° (BDC) for the start of compression and ending at 540° (BDC) at the end of intake stroke. Note that following ignition (point x), the pressure rises abruptly due to combustion reaching a maximum at a point (y) slightly beyond TDC.

The region of positive work for each cycle is indicated in the drawing as the power stroke (i.e., 0 to 180°). The fluctuations in combustion chamber pressure along with the geometry factor relating p_c to T_i cause T_i to fluctuate with crankshaft angle and of course with time. However when the engine produces power, the time-average value for T_i (i.e., ) is positive:

There are other contributors to the total dynamic indicated torque at the crankshaft, including torques due to the reciprocating forces of the piston and connecting rod. The details of the reciprocating torque (T_r) are beyond the scope of this book and are not relevant to the present discussion, but in general increase quadratically with rotational speed. In addition, there are contributors to the torque at the crankshaft due to internal friction of the rotating and reciprocating components as well as due to pumping of intake and exhaust gases.

The indicated torque is the maximum available torque which is applied at each crankshaft segment for the corresponding cylinder. Typically, between each crankshaft “throw” are sleeve bearings that have friction. In addition to friction, there are negative torques applied to the crankshaft owing to the nonzero cylinder pressures—p_e during exhaust and p_i during intake – and a relatively large negative torque associated with compression. The time-average torque averaged over an engine cycle at the crankshaft output end is called the brake torque and is given by

(5)

where is the average torques associated with friction and pumping losses. It is this brake torque acting through the drivetrain that provides the torque to drive the vehicle. The drivetrain includes the transmission and other gear systems (e.g., differential) as explained in Chapter 7.

Power

One of the most important metrics for engine performance is output power. This power is related to the indicated torque applied to the crankshaft (as explained above). The instantaneous power applied to the crankshaft by the indicated torque is known as the indicated power (P_i[θ_e(t)]), given by

where

(6)

The units for P_i(t) are Nm/sec (metric) or ft lb/sec (English units). Normally it is the average indicated power averaged over N engine cycles that is useful as a metric for engine available power (with θ_e in radians) is given by:

(7)

The appropriate unit for P_i is kW, although in the USA the popular unit (with the driving public) remains horsepower (Hp), where 1Hp = 0.75 kW and 550 ft lb/s.

The engine output power at the crankshaft is known as the brake power (P_b) since historically engine power was measured using a Prony brake. This brake power (in kW or Hp) is the difference between indicated power and the power associated with internal power losses due, e.g., to friction and pumping of the intake mixture and exhaust gases. Generally, the cycle-averaged friction and pumping power are combined and denoted . The brake power P_b is given by

(8)

Measurements are readily made of P_fp by driving the engine from an external power source such as an electric dynamometer. The latter is an instrumented electric motor/generator having the capacity to absorb all brake power produced by the running engine under test. Normally, instrumentation permits measurements to be made of output torque, angular speed ω_e, and P_b. It is also common practice to evaluate engine performance via the averaged torque at the engine output, which is called “brake torque” and is denoted T_b and which is related to P_b by the expression

(9)

Another metric of performance for an engine is the so-called mean-effective pressure (mep). It is defined as the indicated work done on the piston (W_i) (given in Eqn (2)) divided by displacement volume V_D. As in the case of torques, it is convenient to consider the indicated mep (imep) which is defined as

(10)

where V_D = V₁ − V₂ is the displacement, V₁ the cylinder maximum volume (at BDC), and V₂ the cylinder minimum volume at TDC; (i.e., clearance volume).

The imep (which has the dimensions of pressure) is the value of constant pressure, which, if acting during an engine cycle, would produce the work done on the crankshaft. There is also a friction mep (fmep):

(11)

where W_f is the work done by the friction torque. The most commonly used mep is the brake mep (bmep), which is defined as

It has the units of pressure (e.g., N/m² or lb/in²) and is the value of constant pressure acting over a full engine cycle to produce the output mechanical work/cycle.

Fuel Consumption

Fuel economy can be measured while the engine delivers power to the dynamometer. The engine is typically operated at a fixed RPM and a fixed brake power (fixed dynamometer load), and the fuel flow rate (in kg/hr or lb/hr) is measured. The fuel consumption is then given as the ratio of the fuel flow rate to the brake power output (P_b). This fuel consumption is known as the brake-specific fuel consumption, or BSFC. BSFC is a measurement of the fuel economy of the engine alone and is given by

(12)

The unit for BSFC is kg/(kW·hr) or (lb/CHp·hr) in British units. By improving the BSFC of the engine, the fuel economy of the vehicle in which it is installed is also improved. It is shown later in this chapter that electronic controls can optimize BSFC.

In gasoline-fueled engines, airflow into the engine at any operating angular speed (RPM) is determined by the throttle angular position. In fact, the throttle is the control by the driver that determines the engine output power.

As explained above, any internal combustion must pump air/fuel into its combustion chamber. If an IC engine were a perfect air pump, then at wide open throttle the air volume pumped into the engine for each complete engine cycle (i.e., two complete revolutions) would be its displacement volume V_d:

(13)

where A_p is the piston cross-sectional area and S_c the cylinder stroke which is the distance traveled by the piston from TDC to BDC.

Formally, the volumetric efficiency is defined as the ratio of the mass of fresh mixture (i.e., air and fuel) that is actually pumped into the cylinder during an intake stroke at inlet air density to the mass of this mixture, which would fill the cylinder at the inlet air density. The volumetric efficiency for any given engine is determined empirically and varies with throttle angle, RPM, inlet pressure and temperature as well as exhaust pressure (p_e). Assuming initially that all cylinders receive mixture at identical density the definition of can be expressed by

(14)

where is the air intake mixture mass flow rate (slugs/sec) (or kg/sec), N the number of revolutions/sec, and ρ_i the inlet mixture density (slugs/ft³) (or kg/m³).

The variable ρ_i is the density of the mixture in the intake system downstream from the throttle plate in or near a cylinder inlet port. When inlet air density is defined at this point in the intake system, it provides a measurement of the air pumping efficiency of the cylinder and valves alone. It is this definition which is used for the present discussion. However, it should be noted that volumetric efficiency could be based on the air density at the input to the intake system (i.e., upstream of the throttle plate). With air density taken at this point, the volumetric efficiency is termed the overall volumetric efficiency. Unfortunately, it is not always convenient to measure the density of the inlet mixture which consists of air, fuel, and atmospheric water vapor. On the other hand, since fuel, airflow, and water vapor occupy the same volume and have the same intake volume flow rate , the following relationship is valid:

(15)

where is the mass flow rate of dry air and ρ_a the inlet density of dry air.

For mixtures of air, water vapor, and gaseous or evaporated fuel, Dalton’s law of partial pressures states:

(16)

where p_i is the total inlet pressure, p_a the partial pressure of air, p_f the partial pressure of fuel, and p_w the partial pressure of water vapor.

Each constituent (denoted k) of the inlet air mixture behaves as a perfect gas such that

(17)

where R is the perfect gas law constant. Also, it can be shown that

(18)

where M_k= mass of constituent k and m_k = molecular mass of constituent k:

The air density in the mixture ρ_a is given by

(19)

where is the fuel/air mass ratio, h the ratio of mass water vapor to the mass of air, m_a = 29, and m_w = 18.

In this form it is possible to compute inlet air density from measurements of total inlet air pressure (p_i) and inlet absolute air temperature T_i as well as standard environmental variable measurements (e.g., relative humidity). As presently shown, F_i is determined by the fuel-control system to achieve certain engine performance requirements.

As explained earlier in this chapter, the engine power is regulated by the driver via an air valve in the form of a movable throttle plate in the intake system. Linkage connects the accelerator pedal to the throttle plate such that it partially restricts airflow into the engine. Typically, the throttle plate is in the form of a circular disk that pivots about a diametric axis in a cylindrical portion of the intake manifold (e.g., see Figure 5.3). Effectively, the airflow into the engine at any given engine angular speed (RPM) varies in proportion to the opening of this plate as represented by the throttle angle θ_T. This empirically determined volumetric efficiency is a convenient variable that can be used to characterize engine pumping efficiency during the development of a new engine control system and can also be used in fuel control of a production engine as explained later.

Engine Overall Efficiency

There are numerous ways to characterize the performance of an engine as indicated above. One of the most meaningful of these is the efficiency with which the engine converts the energy available in the fuel (in chemical form) to mechanical work. This efficiency, which we denote η_m, can be evaluated on an engine cycle by engine cycle basis. However, it is more convenient to express η_m as the ratio of the instantaneous mechanical power delivered to the load to the rate of change of available energy in the fuel being delivered:

where P_m is the mechanical power delivered to load (kW), the instantaneous mass flow rate of fuel (kg/sec), and Q_f the energy content of fuel (joule/kg).

Calibration

The definition of engine calibration is the setting of the air/fuel ratio and ignition timing for the engine for any given operating condition. With the new electronic control systems, calibration is determined by the electronic engine control system.

As will be shown later in this chapter, electronic engine control systems are based upon microprocessors or microcontrollers. Under program control, the engine control system determines the correct fuel delivery amount and the ignition timing as a function of driver command (via throttle setting) and other operating variables and parameters. Typically, these correct values are found from a table look-up process with interpolation. The calibration tables for any given engine configuration are found empirically as described below. As will also be shown below, an additional component of fuel delivery is determined from a closed-loop portion of the control.

The majority of present-day engines deliver fuel by means of an individual fuel injector associated with each cylinder. A fuel injector is essentially an electromechanical valve to which fuel, under pressure, is supplied. Chapter 6 explains the configuration and operation of fuel injectors. As will be explained later, each fuel injector delivers fuel in a pulse mode in which fuel quantity is determined primarily by the duration of fuel delivery at a nominally constant delivery rate. Another important engine calibration variable for such systems is the time of fuel delivery relative to cycle for the associated cylinder.

Engine Mapping

The development of any control system comes from knowledge of the plant, or system to be controlled. In the case of the automobile engine, this knowledge of the plant (the engine) comes primarily from a process called engine mapping.

For engine mapping, the engine is connected to a dynamometer and operated throughout its entire speed and load range. Measurements are made of the important engine variables while quantities, such as the air/fuel ratio and the spark control, are varied in a known and systematic manner. Such engine mapping is done in engine test cells that have engine dynamometers and complex instrumentation that collects data under computer control. At each operating point, calibration is varied and performance is measured. An optimum calibration can be found that is a compromise between performance and allowable exhaust gas emission rates under federal regulations.

From the engine mapping a calibration table can be created of optimum values for later incorporation into the engine control system ROM as explained in Chapter 7. Also, from this mapping, a mathematical model can be developed empirically that explains the influence of every measurable variable and parameter on engine performance. The control system designer can, if desired, select a control configuration, control variables, and control strategy that will satisfy all performance requirements (including stability) as computed from this model and that are within the other design limits such as cost, quality, and reliability. To understand a representative engine control system, it is instructive to consider the influence of control variables on engine performance and exhaust emissions.

Effect of Air/Fuel Ratio on Performance

Figure 5.12 illustrates the variation in the performance variables of indicated torque (T_i) BSFC as well as engine emissions with variations in the air/fuel ratio with fixed spark timing and a constant engine speed.

In this figure, the exhaust gases are represented in brake-specific form. This is a standard way to characterize exhaust gases whose absolute emission levels are proportional to power. The definitions for the brake-specific emissions rates are

BSHC = brake-specific HC concentration

(20)

BSCO = brake-specific CO concentration

(21)

BSNOx = brake-specific NOx concentration

where r_HC is the HC rate of flow, r_CO the CO rate of flow, r_NOx the NOx rate of flow, and P_b the brake power. The indicated torque is denoted T_i in Figure 5.12.

One specific air/fuel ratio is highly significant in electronic fuel-control systems, namely, the stoichiometric mixture. The stoichiometric (i.e., chemically correct) mixture corresponds to an air and fuel combination such that if combustion was perfect, all the hydrogen and carbon in the fuel would be converted by the burning process to H₂O and CO₂. For gasoline, the stoichiometric mixture ratio is 14.7:1.

Stoichiometry is sufficiently important that the fuel and air mixture is often represented by a ratio called the equivalence ratio, which is given the specific designation λ. The equivalence ratio is defined as follows:

A relatively low air/fuel ratio, below 14.7 (corresponding to λ < 1), is called a rich mixture and an air/fuel ratio above 14.7 (corresponding to λ > 1) is called a lean mixture. Emission control is strongly affected by air/fuel ratio, or by λ.

Note from Figure 5.12 that torque (T_i) reaches a maximum in the air/fuel ratio range of 12–14. The exact air/fuel ratio for which torque is maximum depends on the engine configuration, engine speed, and ignition timing. Also note that the CO and unburned hydrocarbons tend to decrease with increasing air/fuel ratios, as one might expect because there is relatively more oxygen available for combustion with lean mixtures than with rich mixtures.

Unfortunately, for the purposes of controlling exhaust emissions, the NOx exhaust concentration increases with increasing air/fuel ratios. That is, there is no air/fuel ratio that simultaneously minimizes all regulated exhaust gases.

Effect of Spark Timing on Performance

Spark advance is the time before top dead center (TDC) when the spark is initiated. It is usually expressed in number of degrees of crankshaft rotation relative to TDC. Figure 5.13 reveals the influence of spark timing on brake-specific exhaust emissions with constant speed and constant air/fuel ratio for a representative engine. Note that both NOx and HC generally increase with increased advance of spark timing. BSFC and torque are also strongly influenced by timing. Figure 5.13 shows that maximum torque occurs at a particular advanced timing (called advance for mean best torque) denoted MBT.

Operation at or near MBT is desirable since this spark timing tends to optimize performance. This optimal spark timing varies with RPM. As will be explained, engine control strategy involves regulating fuel delivery at a stoichiometric mixture and varying ignition timing for optimized performance. However, there is yet another variable to be controlled, which assists the engine control system in meeting exhaust gas emission regulations.

Effect of Exhaust Gas Recirculation on Performance

Up to this point in the discussion, only the traditional calibration parameters of the engine (air/fuel ratio and spark timing) have been considered. However, by adding another control variable, the undesirable exhaust gas emission of NOx can be significantly reduced while maintaining a relatively high level of torque. This new control variable, exhaust gas recirculation (EGR), consists of recirculating a precisely controlled amount of exhaust gas into the intake. The engine control configuration depicted in Figure 5.5 shows that exhaust gas recirculation is a major subsystem of the overall control system. Its influence on emissions is shown in Figures 5.14 and 5.15 as a function of the percentage of exhaust gas in the intake. Figure 5.14 shows the dramatic reduction in NOx emission when plotted against air/fuel ratio, and Figure 5.15 shows the effect on performance variables as the percentage of EGR is increased. Note that the emission rate of NOx is most strongly influenced by EGR and decreases as the percentage of EGR increases. The HC emission rate increases with increasing EGR; however, for relatively low EGR percentages, the HC rate changes only slightly. Thus, a compromise EGR rate between NOx reduction and HC increase is possible in which the benefits of EGR on NOx reduction far offset the adverse effect on HC emissions. This compromise amount of EGR varies with engine configuration.

The mechanism by which EGR affects NOx production is related to the peak combustion temperature. Roughly speaking, the NOx generation rate increases with increasing peak combustion temperature if all other variables remain fixed. Increasing EGR tends to lower this temperature; therefore, it tends to lower NOx generation. It should be noted that EGR, though relatively small, does influence the air partial pressure in the intake mixture. Compensation for this effect is required as explained later.

Oxidizing Catalytic Converter

The oxidizing catalytic converter (Figure 5.16) has been one of the more significant devices for controlling exhaust emissions since the era of emission control began. The purpose of the oxidizing catalyst (OC) is to increase the rate of chemical reaction, which initially takes place in the cylinder as the compressed air–fuel mixture burns, toward an exhaust gas that has complete oxidation of HC and CO to H₂O and CO₂.

The extra oxygen required for this oxidation is often supplied by adding air to the exhaust stream from an engine-driven air pump. This air, called secondary air, is normally introduced into the exhaust manifold.

The most significant measure of the performance of the OC is its conversion efficiency,

(22)

where M_o is the mass flow rate of gas that has been oxidized leaving the converter and M_ic is the mass airflow rate of gas into the converter

The conversion efficiency of the OC depends on its temperature. Figure 5.17 shows the conversion efficiency (expressed as a percent) of a typical OC for both HC and CO as functions of temperature. Above about 300 degrees C, the efficiency approaches 98–99% for CO and more than 95% for HC.

The Three-Way Catalyst

Another catalytic converter configuration that is extremely important for modern emission control systems is called the three-way catalyst (TWC). It uses a specific catalyst formulation containing platinum, palladium, and rhodium to reduce NOx and oxidize HC and CO all at the same time. It is called three way because it simultaneously reduces the concentration of all three major undesirable exhaust gases. The three-way catalyst uses a specific chemical design to reduce all three major emissions (HC, CO, and NOx) by approximately 90%.

The conversion efficiency of the TWC for the three exhaust gases depends mostly on the air/fuel ratio. Unfortunately, the air/fuel ratio for which NOx conversion efficiency is high, corresponds to a very low conversion efficiency for HC and CO and vice versa. However, as shown in Figure 5.18, there is a very narrow range of air/fuel ratio (called the window and shown as the lined region in Figure 5.18) in which an acceptable compromise exists between NOx and HC/CO conversion efficiencies. The conversion efficiencies within this window are sufficiently high to meet the very stringent EPA requirements established so far.

Note that this window is only about 0.1 air/fuel ratio wide (± 0.05 air fuel ratio) and is centered at stoichiometry. (Recall that stoichiometry is the air/fuel ratio that would result in complete oxidation of all carbon and hydrogen in the fuel if burning in the cylinder were perfect; for gasoline, stoichiometry corresponds to an air/fuel ratio of 14.7.) This ratio and the concept of stoichiometry is extremely important in an electronic fuel controller. In fact, the primary function of most modern electronic fuel-control systems is to maintain average air/fuel ratio at stoichiometry. The operation of the three-way catalytic converter is adversely affected by lead. Thus, in automobiles using any catalyst, it is necessary to use lead-free fuel.

Controlling the average air/fuel ratio to the tolerances of the TWC window (for the full life requirement) requires accurate measurement of mass airflow rate and precise fuel delivery and is the primary function of the electronic engine control system. A modern electronic fuel-control system can meet these precise fuel requirements. In addition, it can maintain the necessary tolerances for government regulations for over 100,000 miles.

The fundamentals of any electronic engine control system are that it regulates the fuel/air mixture and ignition timing in response to an arbitrary driver input (via the throttle plate). The driver input includes the accelerator pedal position which ultimately determines the throttle position. The electronic engine control system directly determines a corresponding fuel quantity delivered to the engine which optimizes performance subject to a somewhat complex set of constraints. The constrained optimization involves compromises between the conflicting constraints of emission regulations and required fuel economy. As explained above, there are other control variables (e.g., spark timing EGR) that are part of the constrained optimization process.

Engine Control Sequence

The step-by-step process of events in fuel control begins with engine start. During engine cranking the mixture is set rich by an amount depending on the engine temperature (measured via the engine coolant sensor), as explained in detail in Chapter 7. Generally speaking, the mixture is relatively rich for starting and operating a cold engine as compared with a warm engine. However, the discussion of this requirement is deferred to Chapter 7. Once the engine starts and until a specific set of conditions is satisfied, the engine control operates in the open-loop mode.

After combustion, the exhaust gases flow past the EGO sensor, through the TWC, and out the tailpipe. Once the EGO sensor has reached its operating temperature (typically a few seconds to about two minutes depending upon ambient conditions and the type of sensor used [see Chapter 6]), the EGO sensor signal is input to the controller and the system begins closed-loop operation.

Open-Loop Control

Fuel control for an electronically controlled engine operates open-loop any time the conditions are not met for closed-loop operation. Among many conditions (which are discussed in detail in Chapter 7) for closed-loop operations, there are some temperature requirements. After operating for a sufficiently long period after starting, a liquid-cooled automotive engine operates at a steady temperature.

However, an engine that is started cold initially operates in open-loop mode. This operating mode requires, at minimum, measurement of the mass airflow into the engine, and a measurement of RPM as well as measurement of coolant temperature. The mass airflow rate measurement in combination with RPM permits computation (by the engine controller) of the mass of air (M_a) drawn into each cylinder during intake for each engine cycle. The correct fuel mass (M_f) that is injected with the intake air is computed by the electronic controller:

(24)

where r_fa is the desired ratio of fuel to air.

For a fully warmed-up engine, this ratio is 1/14.7, which is about .068. That is, 1 lb of fuel is injected for each 14.7 lb of air, making the air/fuel ratio 14.7 (i.e., stoichiometry). The desired fuel/air ratio varies with temperature in a known way such that the correct value can be found from the measurement of coolant temperature. For a very cold engine, the mixture ratio can go as low as about 2 (i.e., ).

Theoretically, if there were no changes to the engine, the sensors, or the fuel injector, an engine control system could operate open-loop at all times. In practice, owing to errors in the calculation of M_a, variations in manufactured components, as well as to factors such as wear, the open-loop control would not be able to maintain the mixture at the desired air/fuel ratio if it were used alone. In order to maintain the very precise air/fuel mixture ratio required for emission control over the full life of the vehicle, the engine controller is operated in closed-loop mode for as much of the time as possible. Compensation for the open-loop mode variations above is possible via adaptive closed-loop control as explained in Chapter 7.

Closed-Loop Control

Referring to Figure 5.20, the control system in closed-loop mode operates as follows. For any given set of operating conditions, the fuel metering actuator provides fuel flow to produce an air/fuel ratio set by the controller output. This mixture is burned in the cylinder and the combustion products leave the engine through the exhaust pipe. The EGO sensor generates a feedback signal for the controller input that depends on the exhaust gas oxygen concentration. This concentration is a function of the intake air/fuel during the intake portion of the same cycle which is 1.5 crankshaft revolutions earlier than the time at which the EGO sensor output is measured.

One closed-loop control scheme that has been used in practice (i.e., limit-cycle control) results in the air/fuel ratio cycling around the desired set point of stoichiometry. The important parameters for this type of control include the amplitude and frequency of excursion away from the desired stoichiometric set point. Fortunately, the three-way catalytic converter’s characteristics are such that only the short-term time-average air/fuel ratio determines its performance. The variation in air/fuel ratio during the limit-cycle operation is so rapid that it has no effect on engine performance or emissions, provided that the average air/fuel ratio remains at stoichiometry.

Closed-Loop Operation

In the present example, the closed-loop portion of the pulse duration τ_F(k) for each cycle τ_Fc(k) is a function of the equivalence ratio at the EGO sensor:

(28)

whenever the mixture is lean of stoichiometry (i.e., λ > 1), the pulse duration increases by δτ from the previous cycle, thereby richening the mixture (and causing λ to decrease toward 1). Correspondingly, whenever λ < 1 (rich of stoichiometry), the pulse duration is decreased from cycle k to cycle k + 1. The open-loop portion of τ_F is a pulse whose duration τ_o(k) (called base pulse duration) is calculated to yield M_F corresponding to stoichiometric mixture based upon mass airflow rate measurements. We illustrate the operation of this example fuel-control system in Figure 5.20.

Lambda is used in the block diagram of Figure 5.20 to represent the equivalence ratio at the intake manifold. The exhaust gas oxygen concentration determines the EGO output voltage (V_o). The EGO output voltage abruptly switches between the lean and the rich levels as the air/fuel ratio crosses stoichiometry. The EGO sensor output voltage V_o is at its higher of two levels for a rich mixture and at its lower level for a lean mixture.

Reduced to its essential features, the engine control system operates as a limit-cycle controller in which the air/fuel ratio cycles up and down about the set point of stoichiometry, as shown by the idealized waveforms in Figure 5.21. The air/fuel ratio is either increasing or decreasing; it is never constant. The increase or decrease is determined by the EGO sensor output voltage. Whenever the EGO output voltage level indicates a lean mixture, the controller causes the air/fuel ratio to decrease, that is, to change in the direction of a rich mixture. On the other hand, whenever the EGO sensor output voltage indicates a rich mixture, the controller changes the air/fuel ratio in the direction of a lean mixture.

The electronic fuel controller changes the mixture by changing the duration of the actuating signal to each fuel injector. Increasing this duration causes more fuel to be delivered, thereby causing the mixture to become richer. Correspondingly, decreasing this duration causes the mixture to become leaner. Figure 5.21b shows the fuel injector signal duration.

In Figure 5.21a the EGO sensor output voltage is at the higher of two levels over several time intervals, including 0–1 and 1.7–2.2. This high voltage indicates that the mixture is rich. The controller causes the pulse duration (Figure 5.21b) to decrease during this interval. At time 1 second, the EGO sensor voltage switches low, indicating a lean mixture. At this point the controller begins increasing the actuating time interval to tend toward a rich mixture. This increasing actuator interval continues until the EGO sensor switches high, causing the controller to decrease the fuel injector actuating interval. The process continues this way, cycling back and forth between rich and lean around stoichiometry.

The engine controller continuously computes the desired fuel injector actuation duration and maintains the current value in memory. At the appropriate time in the intake cycle, the controller reads the value of the fuel injector duration and generates a pulse of the correct duration to activate the proper fuel injector for the computed time interval τ_F.

One point that needs to be stressed at this juncture is that the air/fuel ratio deviates from stoichiometry. However, the catalytic converter will function as desired as long as the time-average air/fuel ratio is at stoichiometry. The controller continuously computes the average of the EGO sensor voltage. Ideally, the air/fuel ratio should spend as much time rich of stoichiometry as it does lean of stoichiometry. In the simplest case, the average EGO sensor voltage should be halfway between the rich and the lean values:

(29)

Whenever this condition is not met, the controller adapts its computation of pulse duration (from EGO sensor voltage) to achieve the desired average stoichiometric mixture. Chapter 7 explains this adaptive control in more detail than is given here.

Measuring Air Mass

A critically important aspect of fuel control is the requirement to measure the mass of air that is drawn into the cylinder (i.e., the air charge). The amount of fuel delivered can then be calculated such as to maintain the desired air/fuel ratio. There is no practically feasible way of measuring the mass of air in the cylinder directly. However, the air charge can be determined from the mass flow rate of air into the engine intake since all of this air eventually is distributed to the cylinders (ideally uniformly).

There are two methods of determining the mass flow rate of air into the engine. One method uses a single sensor that directly measures mass airflow rate. The operation of this sensor is explained in Chapter 6. The other method uses a number of sensors that provide data from which mass flow rate can be computed. This method is known as the speed-density method.

Influence of Valve System on Volumetric Efficiency

For any given engine configuration, volumetric efficiency is determined by the intake manifold, the valve sizes, and locations, as well as the timing and profile of the cam lobe characteristics. The design of the cam lobe profile determines when the valves open and close and determines the maximum valve opening (lift). Any given cam profile is optimum only for a relatively narrow range of RPMs and throttle settings. Compromises are made between low-, high-, and mid-range RPMs as well as part throttle versus open or closed throttle.

Ideally, it would be desirable to vary valve timing and lift continuously as the engine operates so as to optimize volumetric efficiency. One technology exists for such variable valve timing (VVT) and is found in certain production vehicles. Variable valve timing is also called variable valve phasing (VVP) which is the preferred terminology in this book.

This technology involves separate camshafts for intake and exhaust valves. These two camshafts are driven via a mechanism that varies the relative timing for intake and exhaust. This mechanism (which includes an electromechanical actuator) and its operation are explained in Chapter 6. There it is shown that either (or both) intake and exhaust valve timing is varied relative to the engine cycle. The control strategy for regulating VVP is explained in Chapter 7.

In essence, the exhaust valve is open primarily during the exhaust stroke and that the intake valve is open primarily during the intake stroke. Typically in automotive engines, the exhaust valve remains open during the initial portion of the intake valve-opening period. The crankshaft angle over which the two valves are both open (or partially open) is called overlap. Valve overlap permits exhaust action to assist the intake and improve volumetric efficiency. It also permits some exhaust gas to be mixed with intake gases such the EGR system is at least partially implemented by engine pumping processes. In a variable cam phasing system, this overlap is minimum at idle and varies with operating conditions to optimize emissions and performance. This topic is discussed in detail in Chapters 6 and 7.