**3. Engine design factors that influence quality of exhaust emissions**

At the core of diesel engine performance analysis is the efficiency of combustion and quality of exhaust emissions. These are related to the air utilization and fuel economy. Many factors affect engine combustion efficiency, which can individually and collectively affect the quality of exhaust emissions. Of interest in diesel engine emissions are the factors that promote generation of regulated variables which include soot, which along with other adsorbed particles like condensates of unburnt hydrocarbon (UHC) and its intermediates, is generally called particulate matter (PM); oxides of nitrogen (NOx) and carbon monoxide (CO). Their emissions into the atmospheric environment have impact on human health and are therefore regulated in the concentrations they could be emitted from modern engines. The factors that influence their generation include fuel injection pressure, injection timing, ignition delay, fuel atomization, spray-configuration, technique of supplying air, either quiescent or swirl and level of turbulence. These are related to the way fuel is introduced, mixed and ignited in the chamber to achieve good air utilization in the combustion process. Another inherent factor is the very nature and properties of diesel fuel. Every factor simply helps to achieve the underlying chemical balance in combustion equation which can be technically challenging

in the complicated engine environment. To obtain a chemically correct fuel-air combustion in diesel engine is complicated due to the presence of partly premixed fuel and air, and mainly due to unsteady-turbulent diffusion burning encountered in rapid vaporization of liquid fuel, with very short time for the fuel and air mixing. Diesel fuels contain different fractions of alkyl and aromatic components and the combustion mechanisms depend on the molecular structures of different species. These structures are based on the intra-molecular carbon bonds that may be single, double or triple bonds; and straight, branched or ringed chains. They determine the combustion characteristics associated with diesel fuels [5].

The consideration of these factors with the aim of improving exhaust gas emissions through better air utilization to achieve better combustion has led to optimal designs of key operating parameters [6].

#### **3.1 Air induction**

The air induction system consists of air filter, ducts, intake manifold, intake port and intake valves. The air charge induced into the cylinder is related to power output, quality of combustion and exhaust emission during a cycle. Cylinder displacement is fixed based on engine geometry; therefore naturally aspirated capacity is constrained. Design improvement to increase induction capacity is by compressing air to higher density prior to introduction into the cylinder. This is referred to as boosting or supercharging. Typical method of supercharging diesel engines is by allowing a portion of the escaping exhaust gases at high thermal and kinetic energy levels to flow and expand through turbine blades. The power imparted on the shaft is used to drive a centrifugal compressor, axially connected on the same turbine shaft to compress more air into the cylinder. This is known as turbocharging and the increased air charge in the cylinder boosts the power output from the fixed cylinder volume. Optimal design of turbochargers has increased power-to-weight ratio, improved engine efficiency, lowered exhaust emission and in general engine downsizing [1, 3, 4, 7].

#### **3.2 Fuel injection equipment**

Fuel injection equipment systems are used to supply finely atomized fuel sprays into the combustion chamber at precisely controlled time. Types of injection systems include pump-line-nozzle (P-L-N) systems, unit injector and common rail injection systems. These are general categories of injection systems, each having distinct designs. Common in the P-L-N system are in-line (**Figure 2**), the distributor or rotary pumps and the unit pump designs which are either mechanically or electronically controlled. The unit injectors are known to deliver very high pressure because of active mechanical intensifiers like plungers or large cams as in the pressure-time (PT) controlled systems. They are similarly controlled either mechanically or electronically. The common rail injection system (**Figure 3**) in conjunction with electronic control has been developed to overcome some of the draw backs in P-L-N and unit injection pumps. Essentially, it has overcome the dependency of fuel injection pressure and quantity on engine speed, enabled flexibility in matching the fuel injection timing, and smoothening fluctuations due to cycle-to-cycle variability. In all versions, filtered diesel fuel is first lifted from the tank to main gallery of a high pressure pump. The high pressure pump then delivers fuel to the injectors, (which are of different nozzles tip profiles) to spray fuel into the combustion chamber. Modern engine fuel injection systems are controlled by dedicated electronics giving rise to precise fuel spray at high pressures; and the fuel spray profiles are matched to optimized chamber to improve combustion and level

#### **Figure 2.**

*Pump-line-nozzle system (adapted from DieselNet. Courtesy of Robert Bosch GmbH).*

#### **Figure 3.**

*Typical common rail lay-out of a typical Cummins PT fuel injection system (adapted from DieselNet. courtesy of Cummins Inc.).*

exhaust emissions. Many manufacturers especially for heavy duty diesel engines have developed proprietary electronic fuel injection systems relevant to their total engine management system. Good discussions about different injection systems and their impact on exhaust emissions can be found on reference [3].

#### **3.3 Diesel fuel injection**

The scheme of injecting diesel fuel into the combustion chamber is multipurpose apart from mere delivery. The purposes are intricate functions that ensure efficient combustion and reduction of hazardous emissions. These include: timely control of fuel injection, control of right quantity of fuel delivered to meet engine duty or load and at high pressure to enhance good mixing with air. Good air-fuel mixture is an important parameter for efficient combustion through the injection of high pressure sprays that effectively entrains the compressed air in the combustion chamber. These functions are generally enhanced through different injection strategies. Studies have shown that splitting fuel injection into divided doses per

combustion cycle reduces emissions [8–11]. This has enabled injection requirements to be met for different engine operating conditions since engines encounter varying demands, from low-speed, low-load; to medium and high-speed, high-load applications. Good fuel injection is synonymous with the choice of injection equipment which achieve the engine schedule of time, quantity and pressure. Injection pressures can range from 200 to 2000 bars or higher, with nozzle exit velocity of about 100–300 m/s. The in-cylinder air pressure is about 40–100 bars during the time of injection and about 15–25 kg/m3 in density.

#### **3.4 In-cylinder and gas exchange processes**

The processes that occur inside the cylinder are usually described using the corresponding pressure—crank angle diagram. The sequence of valve actuations are numbered 1–4 in **Figure 4** for a four stroke naturally aspirated diesel engine.

A momentary period of valve overlap expressed in degrees of crank angle occurs between points 1 and 4 when both valves are open. This period is significant to the engine performance and emission characteristics because the volumetric efficiency, particularly in the absence of boosting depends on the in-cylinder filling during this period. As valve openings and closures are not instantaneous, the timing is designed to ensure that both are fully open when the benefit of maximum pressure difference will cause the working fluid to flow across them. In current designs, the poppet valves are located in overhead position in the cylinder head. The geometric profiles of intake manifold, intake port and the poppet valves are designed to enhance air breathing capacity of the engine and also to impact swirl and/or squish motion to the bulk flow into the cylinder. The air utilization during combustion process is dependent on how well it is premixed with fuel sprays prior to combustion or accessible during diffusion burning. Engine combustion chamber is designed to complement the effect of fuel spray profile while the air motion enhances mixing/ burning process. The optimization of these engine elements are very significant to the level of exhaust emissions.

#### **3.5 Heat transfer**

The heat transfer considerations in the engine environment are complicated as all modes are involved within the combustion chamber. Preservation of the structural integrity and geometric profiles of the mating surfaces necessitates removal of heat generated during combustion, particularly from the hottest regions to avoid

#### **Figure 4.**

*In-cylinder pressure-crank angle diagram with sequence of valve actuations and processes in four stroke CI engine (IVO, intake valve open; IVC, intake valve closure; EVO, exhaust valve open; EVC, exhaust valve closure; TDC, top dead center; BDC, bottom dead center) [1, 3].*

*Diesel Exhaust Emissions and Mitigations DOI: http://dx.doi.org/10.5772/intechopen.85248*

distortion of these surfaces. Engine heat transfer involves use of cooling water jackets around the regions with highest heat fluxes; the hot water is then pumped to the radiator where heat is lost to ambient environment by forced convection. Some low duty designs adopt the use of air cooling as it flows across an increased surface area created by fins around the cylinder head and walls. The engine efficiency, performance and emissions are greatly affected by heat transfer as such; it is optimized to retain enough heat required to maintain normal engine working temperature and the excess transferred to the cooling medium. This explains the inefficiency usually encountered during cold starting when the work transfer to the piston is reduced as high heat transfer occurs to the chamber walls and other working parts for quick warm up. Similarly, during this transient period, heat transfer affects the process of forming gaseous and particulate matter emissions both within the cylinder and exhaust line where post combustion reactions occur.

#### **4. Thermo-chemistry**

The stoichiometric air-fuel ratio (A/F)s is the chemically correct air-fuel proportion that will completely burn to yield only water, carbon dioxide, and nitrogen. It is dependent on the chemical composition of the fuel. Engine combustion makes use of atmospheric air which contains nitrogen; therefore a general expression for complete combustion of a molecule of hydrocarbon fuel CaHb with air is given as [1]:

$$C\_aH\_b + (a+b/4)\left(O\_2 + \text{3.773}N\_2\right) = aCO\_2 + (b/2)\_2O + \text{3.773}(a+b/4)N\_2 \quad \text{(1)}$$

Eq. (1) defines the stoichiometric ratio of fuel and air reaction that could be utilized for gravimetric analysis of complete fuel combustion in air. It depends on the fuel molecular structure. In engine combustion analysis, it is usual to also express (A/F)s as a reciprocal, meaning fuel-air ratio (F/A)s −1 . Either way is useful in expressing engine operating condition because during engine testing, both air and fuel flow rates are measured to determine if the mixture is rich or lean. The degree to which fuel is rich or lean, is central to engine performance analysis especially with regards to emission considerations. The products of combustion emitted in either of the mixture situation differ significantly due to fuel chemical composition and structure. In this regard, a parameter: equivalence ratio (ϕ) that expresses the actual fuel/air ratio to the stoichiometric value is normally adopted for expressing the composition of mixture.

$$\phi = \left( F \!\!/ A \right)\_{\text{actual}} / \left( F \!\!/ A \right)\_{\text{s}} \tag{2}$$

Alternatively, the inverse of ϕ could be used as relative air/fuel ratio (λ), similarly defined as:

$$
\lambda = \phi^{-1} = \left< A/F \right>\_{\text{actual}} \left< A/F \right>\_{s} \tag{3}
$$

Comparatively, mixture strengths are expressed as follows:

Fuel rich mixture: ˃ 1, ˂ 1

Fuel lean mixture: ˂ 1, ˃ 1

Stoichiometric mixture: = = 1

From Eq. (1), exhaust emissions associated with engine combustion on stoichiometric condition are supposed to be carbon dioxide (CO2), water vapor (H2O) and nitrogen (N2). However this is not practically possible as other emissions like carbon monoxide (CO), unburnt hydrocarbons (UHC), oxides of nitrogen (NOx), oxides of sulfur (Sox) and soot are inevitably present. These emissions arise due to unsteady combustion, post combustion and engine cycle-to-cycle fluctuations. During combustion and about some degrees after the top dead center (TDC), the composition of combustion products corresponds to local equilibrium with respect to pressure, temperature and equivalence ratio. But as expansion stroke advances, recombination reactions of active species occur, and their compositions are changed. Due to fast changing cylinder condition during this period, recombination reaction is not able to maintain the species in chemical equilibrium. As the exhaust valve opens, exhaust flow process forces species compositions to freeze as they are expelled from the cylinder. By implication, even if the fuel-air mixture was lean, not all fuel species had opportunity to be fully oxidized. Also, it is not possible to equally match all the in-cylinder flow conditions in a multi-cylinder engine; therefore it is difficult to have the same exhaust composition from all cylinders. During testing, engines are calibrated to meet transient and stable operations of which fuel-air mixtures vary from rich to lean conditions, thereby giving various emission levels. Therefore evaluation of engine-out emissions is cumbersome and is best approached through experimental measurements [1].

### **5. Diesel spray structure, atomization and spray combustion**

Prior to development of optical accessible engines with laser diagnostics, direct measurement of combustion events inside the engine cylinder was difficult; high speed movies were used to record the spread of fuel jets and flame propagations. Efforts were also made to calculate fuel injection rates from measurements of fuel injection parameters; and the associated heat release rates from engine pressure data in order to gain insights about the combustion process but, details could not be obtained about the true in-cylinder reactions. The development of laser diagnostics enabled in-situ measurements to be made as in-cylinder diesel-spray combustion reactions occur. It is through this approach that planar images have been recoded with good temporal and spatial resolutions, and generates data that offers better interpretations. The fuel spray trajectory, combustion and extinction is the conceptual basis for describing combustion in diesel engine. Currently, there are good insights into the complexity of fuel spray, lift-off length, air-fuel mixture formation, combustion and resultant emission formation through modelling and experiments. The conceptual model presented by Dec in 1997 based on laser sheet imaging [12] is the hallmark of several efforts made with his co-workers at Sandia National Laboratory and many other authors toward contributing to the current understanding. The laser sheet imaging diagnostics has revealed the true nature of in-cylinder processes like: entrainment of air as liquid fuel ejects from nozzle and mixes with hot swirling air to form combustible mixture, images of poly-aromatic hydrocarbons (PAHs), the auto-ignition event, soot formation which starts well from the upstream and central regions of the jet toward the head vortex, as well as the relative particle size distributions and concentrations of soot emission. Separate investigations made for fuel-air mixing, combustion and post-combustion reactions were used to establish the conceptual model (**Figure 5**) as a framework for describing spray combustion.

The original work of Dec used large bore diesel engine with nearly quiescent flow-field, and represented 'quasi-steady' free jet combustion without collision or wall impingement. Subsequent studies that gave information on fuel jet ignition and penetration, have dwelt on this model to elucidate the combustion phenomenon. Detailed information about these developments through modeling and experimental examinations can be checked in the following references: [13–21].

**Figure 5.**

*Dec's conceptual model of quasi-steady spray combustion in conventional diesel engine [12].*

#### **5.1 Spray structure**

The structure of spray evolving from the injector nozzle is related to the injection pressure, size of nozzle holes and their orientation in the combustion chamber. If flow through the nozzle is considered to be of a quasi-steady, incompressible nature in one dimension, the model for mass flow ṁƒ rate of fuel injected is given as [1]: \_

$$\dot{\mathbf{m}}\_f = \mathbf{C}\_d A\_n \sqrt{\left(\mathbf{2} \,\mathbf{p}\_f \,\Delta P\right)} \tag{4}$$

Where the Cd is the discharge coefficient, An is the nozzle flow area, ƿƒ is fuel density, Δ<sup>Ρ</sup> <sup>=</sup> (Pƒ <sup>−</sup> <sup>Ρ</sup>cyl) the pressure drop across nozzle orifice, Δθ is the injection duration in crank angle degrees and N is the engine speed in RPM. If the pressure drop across the nozzle and nozzle open area are taken to be constant during the injection period, the mass of fuel injected is then \_

action period, the mass of fuel injected is then

$$
\dot{\mathbf{m}}\_f = C\_d A\_n \sqrt{\left(\mathbf{2p}\_f\right)} \Delta P \Big|\_{360N} \tag{5}
$$

This relation gives the dependency of overall fuel delivery characteristics of the injection system on injection parameters [2]. However in practice diesel injection is unsteady, the injection pressure as well as in-cylinder pressure vary during the injection period. Similarly, there are significant compressibility effects on the injection process at the very high unsteady pressures encountered in operation; but these are not considered here.

Information about the spray structure has been established though modeling and simulation as well as through laser diagnostics as in the references given. Volumes of publications that describe spray structure through simulations, used Reynolds Average Navier-Stokes (RANS) model and large eddy simulation (LES). The LES turbulence model capture large unsteady flow structures that are responsible for mixing. At a finite distance after nozzle exit called break-up length (L*bu*), the jet flow becomes fully turbulent and the liquid fuel starts to breaks into small droplets as in **Figure 6**. This development creates a large surface area for air to be entrained into the spray, making it to increase in diameter even as it trust axially away from the nozzle tip. As the bulk of the spray penetrates into the cylinder, the early droplets that encounter hot dense air especially at the periphery, loses momentum to the air and the droplets behind the wake penetrate deeper into the

**Figure 6.** *Schematic of diesel spray structure (adapted from AdvancedEngineeringCenter@brighton.ac.uk).*

combustion chamber. For large bore cylinders, the penetration is undisturbed and reaches full spray tip penetration length (L*tip*); but for small bores cylinders, spray impinges on the chamber walls. This make the droplets to spread on the hot walls and experience quick evaporation and mixing with air.

#### **5.2 Spray atomization**

Diesel fuel is injected as conical-shaped spray jet into hot air at conditions that are higher than the critical temperatures and pressures of the fuel, and the increase in velocity of the jet leads to large scale instability. Studies have shown that the jet breakup depends on characteristic factors like the nozzle hole length to diameter ratio, its shape, smoothness, and the rate of injection pressure rise. The break-up mechanisms are classified into primary and secondary regimes depending on prevailing conditions [1, 22]. Primary break-up occurs when the liquid jet near the injector nozzle disintegrates into very small droplets due to interaction with in-cylinder gas. Generally this regime is used to classify break-up at the outer surface of the plane of nozzle exit due to very high jet velocity. The average droplet diameter are much smaller than the nozzle diameter. This initial break-up are considered to be initiated by many factors like turbulence and collapse of cavitating bubbles causing unsteadiness in the nozzle flow and full interaction with swirling air, effects of surface tension and aerodynamic forces, leading to break-off of some filaments or droplets. The continued influence of aerodynamic forces induce further break-up and reduction of the droplets sizes as flow penetrates into the surrounding air. This is called secondary break-up and together with progressively strong evaporative effects, droplet size diminishes in axial flow direction. At the droplet level, the break-up is assumed to be controlled by droplet Weber number (We), defined as the ratio of inertia forces to surface tension forces:

$$\text{We} = \text{p}\_a D\_d \, \text{U}\_{rel}^2 / \sigma\_f \tag{6}$$

where Ƿa—ambient air density, Dd—droplet diameter, Urel—relative velocity between droplet and the ambient gases, σf—surface tension of fuel.

Weber number is typically in excess of 100 in modern diesel engines implying that stripping and catastrophic regimes are the important modes of droplet breakup regime (**Table 1**).


#### **Table 1.**

*Secondary droplet break-up classification.*

#### **5.3 Spray penetration**

Spray penetration is of great benefit for overall fuel-air mixing rates and degree of air utilization. Fuel atomization and evaporation enable fuel-air mixing in the immediate vicinity while substantial quantity of air may remain unutilized for combustion if not quickly reached by fuel species further inside the cylinder. Faster fuel spray penetration rate into the combustion chamber quickens the mixing rate. Many studies have been carried out to get good correlations both experimentally and by turbulent jet mixing theory to propose penetration models [23–25]. While it is good to achieve good penetration, in multi-spray nozzles in direct injection combustion engine, over-penetration leads to fuel impingement on the chamber walls which might be cool and lowers mixing rates thereby causing emissions of unburnt and intermediate hydrocarbon species.

**Figure 7.** *Modified conceptual model [12, 19, 22].*

#### **5.4 Spray combustion**

The original work of Dec used to establish the framework that described spray combustion used large bore diesel engine with near-quiescent flow-field, and represented 'quasi-steady' free jet combustion without collision or wall impingement. The temporal sequence of the conceptual model that have been modified to include some recent contributions [19] is described and adapted here as in **Figure 7**. The crank angle degree after the start of injection (*ASI*) is indicated on each image, is used to trace the history of spray combustion in the chamber.

#### *5.4.1 Fuel jet and air entrainment (0.0–4.5° ASI)*

Fuel sprays exiting the nozzles penetrate into the chamber and expand to near conical shapes. Based on the investigations in [26], the images were captured from nozzle exit at intervals of 0.5° of crank angle degrees. At 1.0°, liquid jet core (dark brown) moves a maximum of 23 mm downstream corresponding to 3.0° *ASI*. The fuel is all vaporized by the entrained air and this limits liquid core penetration. The gas-phase jet penetrates downstream and develops vortex-head due to further entrainment. Correlations for entrainment in [27] and direct fuel vapor measurement in [28] show that as entrainment increased downstream, there is inverse variation of equivalence ratio along the jet axis. Here the description of equivalence ratio is given as the ratio of local fuel-ambient charge mass ratio to the stoichiometric fuel-ambient mass ratio. The liquid fuel is heated and vaporized by the thermal energy of the entrained air, therefore at some distance downstream referred as liquid length, all liquid fuel are vaporized. The entrainment correlations have been used to determine this distance and also establish that it could be shortened if the ambient charge density is increased.

#### *5.4.2 Auto-ignition (3.0–5.0° ASI)*

The beginning of flame initiation is not definite; however Chemiluminescence imaging offers opportunity to examine auto-ignition in diesel combustion. The arrows under the vortex head in jets 3.0–5.0° *ASI*, indicate the Chemiluminescence region of the jet which may even appear earlier. At onset around 3.0–3.5° *ASI*, the vapor fuel-air mixture is thin and around 4.5° ASI, most of the Chemiluminescence emerge from the leading edge portion of the vapor fuel-air mixture and indicates fuel breakdown and formation of PAH in the region 4.5–5.0° *ASI*.

#### *5.4.3 Pre-mixed combustion (4.0–6.5° ASI)*

From the curve of heat release rate (insert), the rate starts to rise after 3.0° and the sharply from 4.0–5.0° ASI. During the same time, emergence of Chemiluminescence is followed by rapid fuel breakdown and formation of PAHs increasingly in the leading edge of the jet where an equivalence ratio is about 2–4.

The initial burning of this premixed rich mixture is indicated by sharp rise in the heat release rate followed by rapid PAH and soot formation between 5.0 and 6.0° ASI downstream of the jet. Upstream, there is more fuel break-down due to increase in temperature, as typified by jet at 6.0° ASI.

#### *5.4.4 Early diffusion flame (5.5–6.5° ASI)*

As the apparent heat release peaks during pre-mixed burn period, there is corresponding appearance of thin diffusion flame on the periphery of the

#### *Diesel Exhaust Emissions and Mitigations DOI: http://dx.doi.org/10.5772/intechopen.85248*

vapor-fuel/air mixture. This is a consequence of fuel rich premixed combustion that burns comfortably in the presence of air. Quickly, the whole periphery of the jet downstream is covered by this thin diffusion flame burning at high temperature with consequent formation of nitrogen oxides. The flame is limited upstream to a point referred as 'Lift-off Length' from the nozzle tip which corresponds to the border region of the vapor-fuel/air mixture. Here the flame is lifted and due to increased local heating, the rate of fuel break-up increases as evidenced by a decrease in the liquid-fuel length by about 2–3 mm.
