Abstract
Spray investigations are critical for understanding internal combustion engine combustion. Optimised spray atomisation helps improve engine output/performance and reduce tailpipe emissions. The spray from the injector nozzle depends on nozzle hole diameter, fuel injection pressure, ambient density, pressure and temperature in the spray chamber, and test fuel properties. This study evaluated macroscopic and microscopic spray characteristics of dimethyl ether (DME) and baseline diesel under atmospheric conditions (1.013 bar pressure at 298 K temperature). It correlated the spray parameters with distinctive physicochemical properties of diesel and DME using dimensionless numbers, namely Reynolds number, Weber number, and Ohnesorge number. The fuel injection system consisted of a high-pressure mechanical injection pump and mechanical fuel injectors having an original equipment manufacturer fixed nozzle opening pressure in the constant volume spray chamber. The microscopic spray investigations were performed using a phase Doppler interferometer along the spray direction at three axial distances (50, 70, and 90 mm) from the nozzle. The three orthogonal spray droplet velocities of diesel and DME were compared. The droplet number-size distributions for baseline diesel and DME were compared. Macroscopic spray characteristics were evaluated using high-speed imaging. Reynolds number was higher for DME, leading to more turbulence in the spray and accelerating the spray breakup phenomenon. Weber number of DME was also much higher than baseline diesel due to its lower surface tension. The higher Weber and lower Ohnesorge numbers justified the finer droplets of DME sprays. DME showed superior spray atomization characteristics than baseline diesel, leading to superior fuel–air mixing and efficient and sootless combustion.
1 Introduction
Dimethyl ether (DME; CH3OCH3) is drawing attention as the next-generation transport fuel, with the potential to ensure energy security by reducing fossil fuel dependence worldwide. The primary feedstocks for DME production are biomass, municipal solid waste, high-ash coal, and natural gas. DME has a higher cetane number (>55), lower boiling point (−25 °C), and lower auto-ignition temperature (235 °C) than baseline diesel, and it burns smoke-free due to the absence of C–C bonds. The research to utilize DME as fuel in internal combustion (IC) engines started in the early 1990s and included characterizing combustion, performance, and emissions [1–6]. Later, dedicated DME-fueled vehicles were developed [7–9]. After that, on-road testing and validation of DME-fueled vehicles were also undertaken [10,11].
The oxygen content (34.78% w/w), absence of C–C bonds, lower C/H ratio, lower auto-ignition temperature, and higher cetane number (55–70) of DME leads to its superior combustion than baseline diesel in compression ignition (CI) engines [12]. However, DME’s energy density is ∼65% of diesel, necessitating changes in the fuel injection system design, such as enlarging injector nozzle diameter and fuel injection pumps of higher capacity. These are required to inject ∼1.5 times DME mass than baseline diesel to generate diesel equivalent power and torque [13,14]. Higher compressibility (lower bulk modulus) and lower viscosity of DME than baseline diesel lead to a reduction in fuel injection pressure (FIP) while increasing the compression work of the high-pressure pump (HPP). DME’s higher vapor pressure (0.53 MPa at atmospheric conditions of 1 bar and 298 K) promotes rapid fuel spray vaporization after its injection in the engine combustion chamber [4]. DME properties such as lower boiling point (−24.8 °C), lower surface tension (11.37 g/s2), and lower kinematic viscosity (0.197 cSt) at normal temperature and pressure (NTP) aid faster spray atomization and vaporization [15].
The spray characteristics of a fuel govern the fuel–air mixing in the combustion chamber, which determines the engine’s combustion, performance, and emission characteristics [16]. Therefore, a detailed study of macroscopic spray characteristics (namely spray penetration length (SPL), spray area, and spray cone angle) and microscopic spray characteristics (namely spray droplet size and velocity distributions) is a fundamental step for analyzing the fuel’s atomization and vaporization processes.
Significant differences exist in the physical properties, such as the bulk modulus, vapor pressure, density, viscosity, and surface tension of DME and baseline diesel. These properties play a crucial role in their distinctive spray atomization characteristics. DME’s very low bulk modulus (one-third of baseline diesel) results in higher compressibility, causing an injection delay and lower FIP for the same compression work by the HPP. Higher vapor pressure of the DME is responsible for cavitation. Cavitation in the nozzle was reportedly more in DME than in diesel due to its higher vapor pressure, resulting in superior spray atomization [36,37].
![Nozzle spray breakup of fuel [39]](https://asmedc.silverchair-cdn.com/asmedc/content_public/journal/openengineering/2/10.1115_1.4063202/1/m_aoje_2_021045_f002.png?Expires=1704379627&Signature=4OdIMTKnJIxS9~WvsnPNeIkWy141lRBYQzkQPS0u4fCCkkWIiC-R-UceKtllpfTSLxWmqeaFBGOvQvITGZ7r--2MG-2bY5UJNydKJTqziUeanHhJPZwUfgMiL8HSDw-kOFYvlCU-UM6JdDZz8dYq6KIiunbufFV75ukfX7er1b8QeiVrxDUBZec29VBAYQ9UTgy45-qk0pkp7WzdCL353ZpcE3aWWJ4~-sSgSQ7pdDK8S8S1mt8dY0h330SPQgzkYPOlbcNCpCribbR3opbTFPIEP3tOkJClotzpxirY-NU3RDgV~GkXvzPpBwg4IWDtLBUKnhJvzYEAVTeDGL4s8Q__&Key-Pair-Id=APKAIE5G5CRDK6RD3PGA)
The negative CN of DME is nearly 4.21 times lower than baseline diesel for the same NOP.
Cavitation disappears immediately when the hydraulic flip starts. A hydraulic flip is the formation of vapors around the flow through the inner part of the nozzle sac because of cavitation. Due to the lower viscosity of DME, its droplet velocity and Reynolds number are higher than baseline diesel. Therefore, DME’s energy loss is much lower than baseline diesel due to its higher discharge coefficient during a hydraulic flip [18]. However, there is a hydraulic flip for DME spray, which increases the spray droplet velocity distribution. The other factors affecting the spray breakup process are fuel injection velocity and test fuel properties. These properties are density, viscosity, and surface tension. As the test fuel is injected into the combustion chamber by the nozzle, instabilities initiate the breakup of the fuel jet. The fuel breakups are divided into two stages: primary breakup near the nozzle exit, where the fuel jet is detached into fuel ligaments and droplets, and then the secondary breakup of larger droplets into smaller droplets downstream of the spray, as shown in Fig. 2. Figure 2 shows cavitation and consequential turbulence in the fuel flow inside the nozzle hole. The turbulence and cavitation cause the velocity of spray droplets to increase in all directions in the CVSC. There are primary and secondary fuel spray breakups in the ambient. The primary breakup occurrs adjacent to the nozzle, where the size of the droplets is much smaller than the nozzle hole diameter. The primary breakup mainly occurrs in the liquid fuel region. However, the secondary breakup occurs further downstream in the vapor region. There have been studies on the macroscopic spray characteristics of DME. However, the microscopic spray studies of 100% DME are rather limited in the open literature. Therefore in the present study, the macroscopic and microscopic spray characteristics of DME and diesel under atmospheric conditions (1.013 bar pressure at 298 K temperature) are experimentally evaluated and are correlated with distinct physicochemical properties of diesel and DME using dimensionless numbers, namely Reynolds number, Weber number, and Ohnesorge number. The DME spray morphology, the 3D distribution of spray droplet velocities, droplet number-size distribution, and different mean diameters of the spray droplets are assessed in this study. The DME and baseline diesel microscopic spray characteristics are compared at three axial distances (50 mm, 70 mm, and 90 mm) from the nozzle exit, along the direction of the spray.
2 Experimental Setup and Methodology
The spray characterization for this study was done using a mechanical fuel injection system in a CVSC, made of glass with a 5-hole mechanical injector mounted on the top (Fig. 3).
DME was drawn from the fuel tank and then compressed by a pneumatic pump (Maximator; HTPU-MSF1-150) up to a pressure of 70 bar to avoid vapor lock formation in the fuel injection system. The compressed DME was supplied to a mechanical fuel injection HPP for further increasing the FIP and then eventually supplying it to the mechanical injector via a high-pressure fuel line. The HPP was coupled to an electric motor via a belt drive to operate at 1400 rpm speed. The start of injection (SoI) timing was measured using an optical sensor fitted onto the HPP. The injector tip was covered by a 12 mm outer diameter cap such that only one plume is allowed to emerge for investigations from the five nozzles of the injector. The macroscopic spray characterization was done using a high-speed CCD camera (Photron; SA-1) and a white light source (Thorlabs) using the Mie scattering technique. The white light source illuminated the spray plume. The camera was synchronized using an optical sensor signal. The spray images were captured using Mie scattering technique at 5400 frames per second (fps). matlab and imagej software were used for spray image processing and calculating the macroscopic spray parameters such as SPL, spray area, and spray cone angle. The spray images were converted to binary images, and the background was subtracted from the images so that only the spray area was obtained to calculate the macroscopic spray parameters.
The microscopic spray characteristics of DME and baseline diesel were investigated using phase Doppler interferometer (PDI) (Artium Technologies Inc.; PDI-300 MD; Fig. 4). The PDI system consisted of two transmitters and a receiver, attached with three advanced signal analysers. Solid-state lasers of three different wavelengths (532, 491, and 561 nm) were used for generating interference patterns. Transmitter 1 emitted four laser beams (two green of 532 nm and two blue of 491 nm) such that pairs of the two laser beams converged at a single point to create the “probe volume.” Similarly, transmitter 2 emitted two yellow laser beams of 561 nm wavelength, converging into the same probe volume. Interference fringe patterns develop when two coherent beams with a phase difference intersect at a point. Therefore, when light is scattered by any droplet passing through the bright fringe in the proble volume, the PDI receiver detects frequency and phase difference changes. This information is passed to the photomultiplier tube of the detector, which detects the Doppler burst signal and frequency changes to provide information about the velocity of individual droplet, and the phase difference gives information about the individual droplet size.
Schematic of experimental setup for microscopic spray characterization of diesel and DME sprays using PDI
Schematic of experimental setup for microscopic spray characterization of diesel and DME sprays using PDI
The PDI was synchronized with the optical sensor signals using a function generator and an oscilloscope. The SoI was measured when the pump started compressing the test fuel. The pulse time is defined as the time after the static injection. The microscopic spray parameters were measured at three axial distances (50, 70, and 90 mm) from the injector nozzle exit with a fixed pulse time. Static injection timing was ∼30 crank angle degrees (CAD), equivalent to ∼7 ms at 1400 rpm speed.
Microscopic spray characteristics, namely spray droplet size-velocity distributions in the three directions, were measured and compared for DME and baseline diesel. Figure 5 shows the orthogonal and non-orthogonal coordinate system of spray droplet velocities. The spray droplet velocity distributions were measured as V1, V2, and V3 in non-orthogonal coordinates. These components were transformed into orthogonal coordinates (Vx, Vy, and Vz) to explain the experimental results.
Schematic showing spray parameters and 3D velocity directions of the spray droplets
Schematic showing spray parameters and 3D velocity directions of the spray droplets
3 Results and Discussion
The macroscopic and microscopic spray investigations of DME and baseline diesel were performed under ambient conditions in the spray chamber (1.013 bar pressure at 298 K temperature). Since DME existed as a gas in ambient conditions (1 bar and 298 K), it was pressurized up to 70 bar using a pneumatic feed pump to avoid vapor lock and cavitation in the fuel line. The feed pressures for DME and diesel were 70 bar and 5 bar, respectively.
3.1 Macroscopic Spray Characterization.
Macroscopic spray characterization was done for the two test fuels (diesel and DME) to understand the effect of test fuel properties on the spray evolution under ambient conditions in a CVSC. The macroscopic spray characteristics include spray cone angle, spray area, and SPL.
DME and diesel spray investigations were performed using the Mie scattering technique in the CVSC under atmospheric conditions (1.013 bar pressure at 298 K temperature). The spray evolution after the SoI of fuel from the injector nozzle for both test fuels is shown in Fig. 6. The nozzle opening pressures for both test fuels were identical.
The images of the liquid and vapor spray phases with time were observed. The spray images were obtained by high-speed imaging using Mie scattering. The outer spray boundary near the nozzle was a straight line, representing a liquid spray jet, which changed to an irregular vortex, away from the nozzle. Overall, the spray shape was conical. However, the actual shape of the DME spray was not completely visible due to very fast evaporation of DME spray droplets around the outer periphery of the spray due to flash boiling effect. The spray characteristics of test fuel were affected by fuel properties (namely density, viscosity, surface tension, boiling point, and latent heat of vaporization), ambient conditions, and fuel injection parameters. At the same ambient conditions, diesel spray evolution was faster than the DME spray. Since the test fuel density affects the spray droplet momentum, and diesel has a higher density (830 kg/m3) than DME (665.7 kg/m3), diesel spray evolution was also faster than DME. The atomization and vaporization of test fuels affected the spray evolution with time.
Figure 7 shows the SPL, spray area, and spray cone angle variations with time after the SoI. SPL is the distance between the injector nozzle and the spray tip. The spray area is the total area of spray projected in a 3D coordinate system. The spray cone angle is the angle between the two straight lines drawn from the nozzle tip to the outermost periphery of the fuel spray on both sides. Figure 7(a) shows that DME’s SPL was lower than diesel due to a faster evaporation rate, lower density, and lower kinematic viscosity of DME. DME’s lower density caused considerable deceleration of spray droplets, resulting in a slower SPL of DME than diesel. The lower kinetic viscosity of DME also reduced the SPL due to a faster DME spray atomization. Other researchers also reported similar results [24,27,42]. DME’s shorter spray penetration length prevented the chances of washing away the lubricating oil layer on the cylinder walls (wall-wetting effect). This minimized the chances of oil burning and prevented soot and unburnt hydrocarbons emissions, otherwise a major problem in diesel engines. Figure 7(b) shows the spray area of diesel and DME at different times after the SoI from the injector nozzle. The spray area of diesel was greater than DME. The DME’s lower spray area was due to DME’s shorter SPL than diesel, which was affected by the fuel properties namely surface tension, density, and viscosity.
The lower surface tension of the DME caused the smaller droplets to form, which evaporated quickly, resulting in a smaller spray area and shorter SPL. Also, the lower density of DME led to deceleration of the DME spray. On the other hand, diesel has higher viscosity, density, and surface tension (Table 1), leading to formation of coarser spray droplets. The larger droplets of diesel experienced lower drag and had higher momentum, leading to longer SPL than DME. The diesel and DME spray areas increased with time after the SoI from the injector. Figure 7(c) shows the variations of the spray cone angle of diesel and DME with time after the SoI. DME showed a higher spray cone angle than baseline diesel due to the flash boiling effect upon injection of DME into the CVSC, which caused spray atomization and dispersion in the radial direction, increasing the spray cone angle [24]. Since DME was injected at atmospheric pressure (1 bar), much lower than its saturation vapor pressure (∼5 bar at NTP), the flash boiling effect led to a dense vapor cloud formation around the spray periphery. Hence, a more extensive spray spread (greater spray volume and larger spray cone angle) of DME than baseline diesel was observed. The macroscopic spray characteristics of DME showed its superior atomization and vaporization characteristics than baseline diesel in atmospheric conditions.
| Fuel properties | Diesel | DME |
|---|---|---|
| Oxygen (% w/w) | 0 | 38.4 |
| Cetane number | 46 | 68 |
| Lower heating value (MJ/kg) | 42.5 | 28.4 |
| Flammability limits (%, v/v) | 0.6–7.5 | 3.4–18 |
| Ignition temperature (°C) | 135 | 235 |
| Flashpoint (°C) | 55 | −42 |
| Boiling point (°C) | 280–338 | −24.8 |
| Latent heat of vaporization at NTP (kJ/kg) | 250 | 460 |
| Vapor pressure at 25 °C (bar) | ≪0.10 | 5.17 |
| Liquid density at 25 °C (kg/m3) | 831 | 656.62 |
| Specific gravity (gas) (vs air) | – | 1.59 |
| Kinematic viscosity at 25 °C (cSt) | 3 | 0.15 |
| Bulk modulus (MN/m2) | 1486 | 637 |
| Surface tension at 25 °C (mN/m) | 27.88 | 11.71 |
| Fuel properties | Diesel | DME |
|---|---|---|
| Oxygen (% w/w) | 0 | 38.4 |
| Cetane number | 46 | 68 |
| Lower heating value (MJ/kg) | 42.5 | 28.4 |
| Flammability limits (%, v/v) | 0.6–7.5 | 3.4–18 |
| Ignition temperature (°C) | 135 | 235 |
| Flashpoint (°C) | 55 | −42 |
| Boiling point (°C) | 280–338 | −24.8 |
| Latent heat of vaporization at NTP (kJ/kg) | 250 | 460 |
| Vapor pressure at 25 °C (bar) | ≪0.10 | 5.17 |
| Liquid density at 25 °C (kg/m3) | 831 | 656.62 |
| Specific gravity (gas) (vs air) | – | 1.59 |
| Kinematic viscosity at 25 °C (cSt) | 3 | 0.15 |
| Bulk modulus (MN/m2) | 1486 | 637 |
| Surface tension at 25 °C (mN/m) | 27.88 | 11.71 |
3.2 Microscopic Spray Characterization.
Microscopic spray investigations of baseline diesel and DME were performed using PDI at three axial distances (50, 60, and 70 mm) from the injector nozzle exit. This section uses spray droplet velocity distribution with pulse time in the three orthogonal directions (radial, axial, and tangential), droplet number-size distribution, droplet velocity-size distribution, and different mean diameters of spray droplets to understand the spray atomization, fuel vaporization, and fuel–air mixing. Three-dimensional spray droplet size-velocity variations with pulse time were investigated. Due to injection delay, the droplets’ velocity was detected at the probe volume a few milliseconds after the SoI. The time difference between the first spray droplet detection in the probe volume and the SoI is “injection delay.” The components of the velocity of spray droplets in all three dimensions were generated for the overall spray volume. These three-dimensional velocities were in axial, radial, and tangential directions. The three velocity components were measured by the three channels of the PDI system (Ch1, Ch2, and Ch3), which were converted to X, Y, and Z velocity components, signifying radial, axial, and tangential velocity components, respectively (Fig. 5).
3.2.1 Radial Droplet Velocity Component (X-Axis).
The velocity measured by Ch1 was radial spray droplet velocity (in X-direction), represented by V1 in the spray coordinate system. The pattern of radial spray droplet velocity was along positive and negative X-directions and symmetrical to the axis of the spray direction.
Figure 8 shows the radial spray droplet velocity component for baseline diesel and DME, measured at different axial distances of 50 mm, 70 mm, and 90 mm from the injector nozzle under ambient conditions. The maximum radial velocity of spray droplets was visibly much higher for DME than baseline diesel in all axial distances from the injector nozzle due to flash boiling of DME. Flash boiling of DME is promoted by its significantly lower boiling point and higher vapor pressure [43]. The density factor of the spray droplets radial velocity of higher magnitude was significantly more for DME than baseline diesel. However, the density factor of spray droplets showing nearly zero velocity was higher for baseline diesel than DME. The spray droplet velocity increased for DME as the axial distance from the injector nozzle increased. However, the spray droplet velocity for diesel got suppressed with increasing axial distance from the injector nozzle. The increase in the radial spray droplet velocity of DME was due to its superior fuel properties, namely boiling point, surface tension, and viscosity, which promoted smaller droplet formation and fuel atomization [24]. DME’s very low surface tension may cause a secondary breakup of larger droplets, resulting in higher radial spray droplet velocity than baseline diesel [44]. Another reason for the higher spray droplet velocity could be DME’s higher vapor pressure, which caused higher flow velocity in the nozzle due to hydraulic flip [45]. The hydraulic flip in DME was due to its very low cavitation number. Also, the radial spray droplet velocity of DME may be due to turbulent flow in the injector nozzle hole.
Variations of spray droplets’ radial velocity component with pulse time for diesel and DME sprays at different axial distances from the nozzle for: (a) diesel and (b) DME
Variations of spray droplets’ radial velocity component with pulse time for diesel and DME sprays at different axial distances from the nozzle for: (a) diesel and (b) DME
The injection delay for DME was lower than baseline diesel, caused by hydraulic flip inside the nozzle, which indicated that it had higher radial spray droplet velocity than baseline diesel, as shown in Fig. 8. The injection delay for baseline diesel were 1.6, 1.9, and 2.3 ms at axial distances of 50, 70, and 90 mm, respectively. The corresponding delay for DME were 1.1, 1.7, and 2.2 ms. DME’s maximum radial spray droplet velocities were 18.5, 24, and 25.4 m/s at axial distances of 50, 70, and 90 mm from the injector nozzle. Diesel’s corresponding maximum radial spray droplet velocities were 17.8, 8.7, and 7.8 m/s. Thus, radial spray droplet velocity reductions were very sharp in the case of baseline diesel than DME because diesel droplet sizes were much larger and hence experienced higher drag force. Hence, DME’s radial spray droplet velocity distribution was affected mainly by its thermophysical properties [24,41,46].
3.2.2 Axial Droplet Velocity Component (Y-Axis).
The axial velocity component of the spray droplets was along the spray direction, denoted by droplet velocity in the Y-direction in the spray coordinate system. The axial velocity of the spray droplets depends on the test fuel properties and FIP differences. DME has higher vapor pressure, lower boiling point, viscosity, and surface tension than baseline diesel. However, thermophysical test fuel properties dominate the FIP when the spray velocities of diesel and DME are compared.
Figure 9 shows the spray droplet velocity variations with pulse time for diesel and DME at three axial distances (50, 70, and 90 mm) from the injector nozzle along the axial direction (Y-axis). DME spray showed a higher maximum droplet axial velocity than baseline diesel. This trend was consistent at all axial distances from the injector nozzle. This was attributed to the higher vapor pressure of DME than diesel in the ambient conditions, which caused the hydraulic flip inside the nozzle. It could be one reason for higher spray droplet axial velocity [47]. DME properties namely lower viscosity, surface tension, and boiling point promoted the higher axial velocity of spray droplets but formation of smaller droplets. Tinier droplets experience lower drag, resulting in higher droplet velocity. Lower fuel density and smaller size of DME droplets contributed to lower mass, which showed higher droplet velocities having similar momentum. DME showed higher axial velocity than baseline diesel due to a lower droplet density factor and viscosity than baseline diesel. Lower DME viscosity promoted a faster fuel mass flow through the nozzle. As the axial distance from the injector nozzle increased, the axial velocity droplet density decreased for diesel and DME (Fig. 9). This trend was due to a faster spray atomization rate and vaporization of DME. The axial droplet velocity at 90 mm axial distance was very low due to DME’s high evaporation rate. The axial velocity of droplets decreased with increasing pulse time for both test fuels (Fig. 9). Axial droplet velocity decreased with increasing pulse time. However, for DME, at 90 mm axial distance, the axial droplet velocity dropped to the lowest level with increasing pulse time due to higher evaporation. The maximum spray droplet axial velocities of diesel and DME droplets at an axial distance of 50 mm from the nozzle were 48.2 and 52 m/s, respectively, which decreased to 33.6 and 49.7 m/s, respectively, at 90 mm axial distance from the injector nozzle. Thus, the DME spray droplet’s more scattered velocity distribution promoted superior fuel–air mixing. A lower number of DME droplets having higher axial velocity and higher evaporation rate suppressed the piston top surface impingement of sprays.
Variations of spray droplets’ axial velocity component with pulse time for diesel and DME sprays at different axial distances from the nozzle for: (a) diesel and (b) DME
Variations of spray droplets’ axial velocity component with pulse time for diesel and DME sprays at different axial distances from the nozzle for: (a) diesel and (b) DME
3.2.3 Tangential Droplet Velocity Component (Z-Axis).
Tangential velocity is the velocity component measured by Ch3. It is perpendicular to the axial direction of the spray and in the tangential direction of the spray cone formed. This velocity component is responsible for the spray cone angle. Figure 10 shows the Z-direction (tangential) velocity component variations of spray droplets with pulse time for diesel and DME at three axial distances (50, 70, and 90 mm). The tangential velocity of diesel and DME droplets decreased with pulse time for both test fuels.
Variations of spray droplets’ tangential velocity component with pulse time for diesel and DME sprays at different axial distances from the nozzle for: (a) diesel and (b) DME
Variations of spray droplets’ tangential velocity component with pulse time for diesel and DME sprays at different axial distances from the nozzle for: (a) diesel and (b) DME
The tangential velocity component of diesel spray droplets decreased with increasing axial distance from 50 mm to 70 mm and then increased when moving away from 70 mm to 90 mm axial distance. However, DME showed a random trend in tangential velocity with increasing axial distance from the injector nozzle. The droplet density factor at a more considerable axial distance was lower for both test fuels. The droplets with tangential velocity were much fewer for DME than for diesel. The maximum tangential velocity was 28.96 m/s for the diesel droplets at 50 mm axial distance and 29.12 m/s for DME at 70 mm axial distance. The velocity distribution of droplets is vital in explaining the spray behavior. Figure 11 shows the droplet velocity distribution w.r.t. droplet diameter under atmospheric conditions for diesel and DME at 50, 70, and 90 mm axial distances from the injector nozzle. The density factor represented these velocity distributions. Results showed that droplet velocity distribution was more dispersed for DME than baseline diesel.
Spray droplet velocity versus droplet diameter distribution for diesel and DME sprays at different axial distances from the nozzle for: (a) diesel and (b) DME
Spray droplet velocity versus droplet diameter distribution for diesel and DME sprays at different axial distances from the nozzle for: (a) diesel and (b) DME
DME’s velocity distribution was dominated by smaller droplets (Fig. 11). It was observed that the velocity distribution for DME was lesser from 20-µm diameter droplets. However, the diesel droplet mobility was seen for up to 35-µm diameter droplets. The maximum velocity of DME spray droplets was higher than baseline diesel, and droplets were more scattered along the velocity axis. However, diesel showed more dispersed droplets along the droplet size axis. DME’s smaller droplet generation was due to its much lower surface tension, viscosity, and boiling point than baseline diesel. DME’s lower boiling point caused flash boiling of the spray droplets in ambient conditions. The higher velocity of DME droplets was due to much higher vapor pressure, resulting in a hydraulic flip inside the injector nozzle. Most diesel and DME spray droplets showed nearly zero velocity, which was also observed in the droplet density factor (Fig. 11). The very low boiling point and higher volatility of DME resulted in faster DME atomization than baseline diesel, resulting in smaller droplets. As the axial distance increased, the droplet density factor of all droplet sizes increased for diesel and DME due to incomplete atomization. Spray breakup enhanced the number of droplets along the increasing axial distance. The number of droplets at a 90 mm axial distance from the injector nozzle was much higher than the 50 mm axial distance for both test fuels. Droplet size distribution was smaller at a larger axial distance for baseline diesel and DME. The droplet count was higher at a longer axial distance, though. Diesel exhibited a higher possibility of droplet coalescence in the secondary spray breakup zone, wherein smaller droplets combined to form larger droplets. Hence, diesel droplet size varied from 1 to 55 µm at all axial distances. DME showed a negligible density factor for larger droplets at 70 and 90 mm axial distances. Most larger droplets had nearly zero velocity, and the number of such droplets was very low.
The spray droplet count and size distributions are essential observations to understand the evolution of the fuel–air mixture in a combustion chamber. Figure 12 shows the spray droplet count variations with the droplet diameter for baseline diesel and DME at the three axial distances.
Variations of spray droplet count with diameter for baseline diesel and DME sprays at different axial distances
Variations of spray droplet count with diameter for baseline diesel and DME sprays at different axial distances
The droplets count for the DME spray was more than double the baseline diesel droplet count at all axial distances. The higher droplet count of DME was due to its lower viscosity and surface tension. A larger number of DME droplets were produced due to flash boiling. Most DME droplets were smaller than baseline diesel due to its lower viscosity, surface tension, and higher flash boiling. The droplet count for any particular size was higher for DME than baseline diesel. The maximum number of DME droplets was observed for 3 µm diameter at all axial distances. However, the maximum number of baseline diesel droplets were of 5 µm diameter. Most DME droplets were lower than 15 µm diameter; however, in baseline diesel, most were less than 30 µm diameter. The physical properties of DME, namely lower surface tension, viscosity, and boiling point, promoted the formation of smaller droplets more than baseline diesel. DME’s thermophysical properties improved smaller droplet spray atomization [48]. The droplet count increased with increasing axial distance from the nozzle due to atomization and breakup of larger droplets into smaller ones. The droplets count for DME at 90 mm axial distance was nearly double that of 50 mm axial distance. The droplet count peaked for DME at different axial distances of 50, 70, and 90 mm, and were 6894, 8660, and 15,518, respectively, and the corresponding numbers for diesel droplets were 2247, 3683, and 5349, respectively. These observations showed that DME exhibited superior spray atomisation than baseline diesel, promoting superior fuel–air mixing.
Figure 13 shows different mean diameters of the spray droplets for baseline diesel and DME at three axial distances (50, 70, and 90 mm) from the nozzle exit. All mean diameters of DME spray were lower than baseline diesel at all axial distances (along the spray direction), except SMD at 50 mm axial distance. DME’s lower surface tension led to droplet breakup and, consequently, a lower SMD. At 50 mm axial distance, the SMD was slightly higher for DME than baseline diesel. This can be attributed to the droplet collision and coalescence, which might have led to slightly larger droplet formation. The fuel’s boiling point, kinematic viscosity, and surface tension affect mean droplet diameters of a spray. The boiling point of DME (−24.8 °C) was lower than diesel (280–338 °C), which promoted its faster atomization and vaporization [50]. DME also had much lower surface tension (0.012 N/m) than baseline diesel (0.028 N/m), promoting formation of smaller-diameter spray droplets. The lower kinematic viscosity and lower surface tension of DME also assisted in formation of smaller droplets. Superior DME evaporation promoted homogeneous fuel–air mixture formation in the combustion chamber, which resulted in more efficient combustion [51].
Comparison of different mean diameters for baseline diesel and DME spray droplets
Comparison of different mean diameters for baseline diesel and DME spray droplets
3.3 Dimensionless Analysis of the Spray Parameters.
DME spray undergoes flash boiling; hence, DME spray’s primary breakup may also contain fuel vapors [52]. The cohesive and disruptive forces (aerodynamic, centrifugal, and surface shear) also dictate the spray deformation and breakup. The aerodynamic forces, namely viscous and inertia forces, disrupt the formation of fuel droplets, while surface tension force tries to maintain the spherical shape of the spray droplets. The effect of these forces on the spray droplets may be captured by nondimensional numbers namely Reynolds number (Re), Weber number (We), and Ohnesorge number (Oh) [53–56]. Ohnesorge and Reitz explained Rayleigh breakups and atomization regimes [57,58] in a fuel sprays. The spray atomization of diesel and DME can be explained by Oh, We, and Re. The liquid Weber number decides the primary spray breakup; however, the secondary breakup is dictated by the gaseous Weber number, which is a relation between the nozzle diameter, spray-droplet velocity, fuel's surface tension, and ambient gas density [59,60]. The secondary breakup of spray droplets depends on the aerodynamic forces on moving droplets and their ambient gas interactions. The gaseous Weber number is obtained by replacing the liquid density with ambient gas density and nozzle diameter with the droplet mean diameter.
Calculated Reynolds number (Re), Weber number (We), and Ohnesorge number (Oh) for baseline diesel and DME from the spray parameters
| Axial distance from nozzle tip | Maximum axial velocity (DME) | Maximum axial velocity (Diesel) | Droplet diameter (DME) | Droplet diameter (Diesel) | Re (DME) | Re (Diesel) | We (DME) | We (Diesel) | Oh (DME) | Oh (Diesel) |
|---|---|---|---|---|---|---|---|---|---|---|
| 50 mm | 52.00 m/s | 48.20 m/s | 27.10 µm | 23.60 µm | 9394.7 | 379.2 | 7.5 | 2.5 | 2.9 × 10−4 | 41.6 × 10−4 |
| 70 mm | 51.16 m/s | 36.49 m/s | 16.70 µm | 26.20 µm | 5695.8 | 318.7 | 4.5 | 1.6 | 3.7 × 10−4 | 39.5 × 10−4 |
| 90 mm | 49.70 m/s | 33.60 m/s | 11.30 µm | 25.30 µm | 3744.1 | 283.4 | 2.9 | 1.3 | 4.5 × 10−4 | 40.2 × 10−4 |
| Axial distance from nozzle tip | Maximum axial velocity (DME) | Maximum axial velocity (Diesel) | Droplet diameter (DME) | Droplet diameter (Diesel) | Re (DME) | Re (Diesel) | We (DME) | We (Diesel) | Oh (DME) | Oh (Diesel) |
|---|---|---|---|---|---|---|---|---|---|---|
| 50 mm | 52.00 m/s | 48.20 m/s | 27.10 µm | 23.60 µm | 9394.7 | 379.2 | 7.5 | 2.5 | 2.9 × 10−4 | 41.6 × 10−4 |
| 70 mm | 51.16 m/s | 36.49 m/s | 16.70 µm | 26.20 µm | 5695.8 | 318.7 | 4.5 | 1.6 | 3.7 × 10−4 | 39.5 × 10−4 |
| 90 mm | 49.70 m/s | 33.60 m/s | 11.30 µm | 25.30 µm | 3744.1 | 283.4 | 2.9 | 1.3 | 4.5 × 10−4 | 40.2 × 10−4 |
It can be seen from Table 2 that the low dynamic viscosity of DME leads to a much higher Reynolds number than baseline diesel. The higher Re of DME leads to greater turbulence in the spray, accelerating the breakup phenomenon. Weber number of DME was also much higher than baseline diesel due to its lower surface tension. In addition, the higher droplet velocity of DME increased the value of these dimensionless numbers for DME. On the other hand, the droplet diameter, despite being higher for baseline diesel at 70 mm and 90 mm axial distance from the nozzle tip (Table 2), did not increase the Weber number of baseline diesel compared to DME. Therefore, it can be said that the surface tension of the fuel and its droplet velocity mainly controlled its Weber number. The higher Weber number of the DME spray represented higher inertia forces relative to the surface tension forces. Higher inertia forces promoted the spray breakup and atomization processes. The Ohnesorge number of diesel was much higher than DME (Table 2) due to its higher dynamic viscosity. The higher Oh of diesel represents its higher internal viscosity dissipation than its surface tension energy, leading to poor spray breakup [63]. The higher We and lower Oh justify the finer droplets in the DME sprays. Higher viscosity fuel generated larger spray droplets, while higher surface tension promoted the disintegration of droplets into smaller ones. Since DME has a lower viscosity and surface tension, it atomized quickly into smaller spray droplets than baseline diesel [64]. Other reasons for smaller DME spray droplets include faster breakup and quicker evaporation [65]. Therefore, it can be concluded that DME exhibits superior spray atomization characteristics than baseline diesel.
4 Conclusions
The macroscopic and microscopic fuel spray characterizations are important to understand a CI engine’s combustion, performance, and emission characteristics. The spray investigations evaluated various spray parameters in ambient conditions at three axial distances (50, 70, and 90 mm) from the injector nozzle. These parameters were SPL, spray area, spray cone angle, droplet velocity-size distributions with pulse time and different mean diameters. DME properties indicated that there might be possibility of hydraulic flip inside the injector nozzle due to negative cavitation number, which was supported by a little higher velocity of DME droplets observed in the microscopic spray characterizations. The high-speed imaging of sprays showed faster evaporation of DME than baseline diesel due to the very low boiling point of DME. The macroscopic spray characterization studies showed lower SPL and spray area for DME than baseline diesel due to its faster evaporation, lower density, and viscosity. The maximum droplet velocities for DME were higher than baseline diesel in all three directions at all axial distances from the injector nozzle exit. The maximum droplet velocities were 17.4, 13.7, and 51.1 m/s for baseline diesel, whereas for DME, these were 29.4, 15.2, and 64.1 m/s in radial, axial, and tangential directions, respectively, at 50 mm axial distance. DME spray had smaller droplets with nearly double the count than baseline diesel due to its lower viscosity and surface tension, at all axial distances from the nozzle. The maximum DME droplet count was 15,518 for 3 µm droplets at 90 mm axial distance from the injector nozzle exit. DME spray’s mean diameters (D10, D20, D30, and D32) were smaller than baseline diesel. The Reynolds number was higher for DME, leading to greater spray turbulence, which accelerated the spray breakup phenomenon. Weber number of DME was also much higher than baseline diesel due to its lower surface tension. The surface tension of the fuel and its droplet velocity dominantly controlled its Weber number. The higher Weber number and lower Ohnesorge number justified the finer droplets of DME sprays. Since DME has a lower viscosity and surface tension, it atomized quickly into smaller spray droplets than baseline diesel. Other reasons for smaller DME spray droplets included faster spray breakup and quicker evaporation. Therefore, it can be concluded that DME has superior spray atomization characteristics than baseline diesel, leading to superior fuel–air mixing, resulting in efficient and sootless combustion in the engine.
Acknowledgment
The authors acknowledge the Science and Engineering Research Board (SERB), Department of Science and Technology (DST), Government of India (Sanction Number IMP/2019/000245 dated 22-01-2020) for funding the project entitled “Ultra-clean emissions DME-fueled tractor engine prototype development for agricultural applications,” for collaboration between Engine Research Laboratory (ERL), Indian Institute of Technology, Kanpur (IITK), and TAFE Motors and Tractors Limited (TMTL). The authors acknowledge the J C Bose Fellowship by SERB, Government of India (Grant EMR/2019/000920) and SBI endowed Chair Professorship from State Bank of India to Prof. Avinash Kumar Agarwal, which enabled this work. Assistance by Ms Utkarsha Sonawane and Mr Ashutosh Jena in performing the experiments is gratefully acknowledged. The contributions of Mr Manojit Pal, Mr Roshan Lal, and Mr Hemant Kumar for experimental setup development are also gratefully acknowledged.
Conflict of Interest
There are no conflicts of interest. This article does not include research in which human participants were involved. Informed consent not applicable. This article does not include any research in which animal participants were involved.
Data Availability Statement
The datasets generated and supporting the findings of this article are obtainable from the corresponding author upon reasonable request.
Nomenclature
Abbreviations
- ASA =
advanced signal analyser
- CI =
compression ignition
- CN =
cavitation number
- CVSC =
constant volume spray chamber
- D10 =
arithmetic mean diameter
- D20 =
surface mean diameter
- D30 =
volume mean diameter
- D32 =
Sauter mean diameter
- DME =
dimethyl ether
- FIP =
fuel injection pressure
- FIS =
fuel injection system
- HPP =
high-pressure pump
- IC =
internal combustion
- LHV =
latent heat of vaporization
- MSW =
municipal solid waste
- NOP =
nozzle opening pressure
- NTP =
normal temperature and pressure
- OEM =
original equipment manufacturer
- Oh =
Ohnesorge number
- PDI =
phase Doppler interferometer
- PODE =
polyoxymethylene dimethyl ether
- Re =
Reynolds number
- SI =
spark ignition
- SMD =
Sauter mean diameter
- SoI =
start of injection
- SPL =
spray penetration length
- Ta =
ambient temperature
- We =
Weber number
