A Study on the Optimal Actuation Structure Design of a Direct Needle-Driven Piezo Injector for a CRDi Engine

Sangik Han; Juhwan Kim; Jinwook Lee

doi:10.3390/app7040320

A Study on the Optimal Actuation Structure Design of a Direct Needle-Driven Piezo Injector for a CRDi Engine

Han, Sangik;Kim, Juhwan;Lee, Jinwook 2017-03-24 00:00:00 applied sciences Article A Study on the Optimal Actuation Structure Design of a Direct Needle-Driven Piezo Injector for a CRDi Engine 1 1 2 , Sangik Han , Juhwan Kim and Jinwook Lee * Department of Mechanical Engineering, Graduate School, Soongsil University, Seoul 06978, Korea; shdow88@naver.com (S.H.); jwan0502@gmail.com (J.K.) Department of Mechanical Engineering, Soongsil University, Seoul 06978, Korea * Correspondence: immanuel@ssu.ac.kr; Tel.: +82-2-820-0929 Academic Editor: Jose Ramon Serrano Received: 13 January 2017; Accepted: 14 March 2017; Published: 24 March 2017 Abstract: Recently, the high-pressure fuel injection performance of common-rail direct injection (CRDi) engines has become more important, due to the need to improve the multi-injection strategy. A multiple injection strategy provides better emission and fuel economy characteristics than a normal single injection scheme. The CRDi engine performance changes with the type of high-pressure electro-mechanical injector that is used and its injection response in a multi-injection scheme. In this study, a direct needle-driven piezo injector (DPI) was investigated, to optimize its actuation components, including the plate length, number of springs, and the elasticity of the spring between the injector needle and the piezo stack. Three prototype DPIs were proposed by this research. They were classiﬁed as Type 1, 2, and 3, depending on whether the injector needle was hydraulic or mechanical. Then, the optimal prototype was determined by conducting four evaluation experiments analyzing the maximum injection pressure, injection rate, spray visualization, and real engine combustion application. As a result, it was found that the Type 3 DPI prototype, with several pan-springs and plates, had the highest injection pressure, a steady injection rate, and the fastest spray speed. It also demonstrated the most effective emission reduction for a two-stage rapid spray injection in a single-cylinder CRDi engine. The Type 3 DPI displays an increased elasticity from its hydraulic needle that provides a synergy effect for improving DPI actuation. Keywords: direct needle-driven piezo injector; high-pressure injection pressure; injection rate; spray visualization; CRDi engine 1. Introduction The combustion process in an automotive engine is never perfect and small amounts of more harmful emissions are also produced in internal combustion engines. In order to reduce engine emissions, modern automotive engines carefully control the amount of fuel that they burn. The thermal efﬁciency of a combustion engine is deﬁned as the relationship between the total energy contained in the fuel, and the amount of energy used to perform useful work or generate kinetic energy. Carbon dioxide (CO ) emissions, as a product of combustion by bonding the carbon in the fuel with the oxygen in the air, are by far the largest amount of emissions produced by internal combustion engine vehicles. Until recently, they were thought to pose no immediate threat to the environment and the health of human beings. Therefore, CO has not been regulated in the same way as HC, CO, and NO . However, 2 x due to recent concerns about the increasing production of greenhouse gases and the increasing use of fossil fuels, regulators have begun attempts to limit the release of CO and other greenhouse gases into the atmosphere. The ICCT (International Council on Clean Transportation) [1] says that most Appl. Sci. 2017, 7, 320; doi:10.3390/app7040320 www.mdpi.com/journal/applsci Appl. Sci. 2017, 7, 320 2 of 16 developed countries have set CO emissions targets for new vehicles. Compared to the EU’s target of 95 g CO /km by 2020/2021, the US’s 97 g CO /km by 2025 for passenger cars, Canada’s 97 g 2 2 CO /km by 2025, China’s 117 g CO /km by 2020, and South Korea’s 97 g CO /km by 2020, illustrate 2 2 2 similar targets. Recently, combustion engine vehicles with a high fuel economy have also been included as eco-friendly vehicles that have less harmful impacts on the environment. In general, modern practical engines are always compromised by trade-offs between different properties, such as efﬁciency, weight, power, exhaust emissions, or NVH (noise, vibration, harshness). Sometimes, the economy not only plays a role in the engine manufacturing cost, but also in the manufacturing and distributing of the fuel. Increasing the engine’s efﬁciency produces a better fuel economy. Reduced exhaust emissions have been strengthened in passenger cars and light duty trucks after EURO 1 was initiated in 1992. Europe adopted EURO 6 in 2013, and it was applied in South Korea in 2014. Nitrogen oxides (NO ) of 80% and particulate matters (PM) of 60% are now more tightened than EURO 5. Some countermeasure technologies have been developed for satisfying this regulation. Among them, clean diesel vehicles and combustion engine-based hybrid electric vehicles have been studied, requiring a great deal of high-tech concentration. Speciﬁcally, a common-rail direct injection system is at the core of clean diesel technology. This makes it possible to pressurize above 200 MPa and then spray fuel into the combustion chamber through an electro-hydraulic injector. This presents accurate and precise fuel spraying using electronic control and is a great development for clean diesel engines. Typical solenoid-driven and piezo-driven fuel injectors can generate a favorable response and provide an effective injection in a common-rail direct injection system. These effects reduce exhaust emissions and improve the combustion performance in a clean diesel engine. Therefore, a rapid injector needle response and multiple injections are needed to obtain a high degree of freedom spray strategy, in order to improve diffusion combustion with a higher air to fuel (A/F) mixing ratio, reduce exhaust emissions, and increase engine performance. Research [2,3] has shown that ﬁve to seven multiple injections and rapid On/Off switching of the injector hole by the needle can freely control the shape of the injection rate. Daiji Ueda, etc., announced G4P (4th generation diesel piezo) injector concepts, structures, and hydraulic performances, using a rectangular injection rate and shorter multiple injection intervals. Also, they showed engine performance improvement results with G4P [2]. Arpaia, A., etc., reported that a numerical model for simulating a piezo indirect acting fuel injection system under a steady state and transient operating conditions was conducted through a commercial code [3]. Other papers [4], referring to the above result, showed that an injection mechanism with a high-pressure performance reduces not only harmful PM exhaust emissions, but also the auto-ignition combustion performance. However, the design of the piezo injector needs further improvement to increase the response of the injector needle. In this study, a direct needle-driven piezo injector (DPI) based on the Delphi product [5] was considered to improve injection characteristics with a maximum injection pressure, injection response, and manufacturing simplicity. For investigation, three steps were followed in this study: spray visualization based on Mie-scattering, an injection rate measurement, and a compression ignition (CI) combustion experiment based on a single-cylinder CRDi engine. DPI prototype design concepts focused on an internal DPI constitution, depending on whether or not a hydrodynamic needle was used. This study investigates the injection characteristics and combustion performance of three prototype DPIs, to obtain practical results. The effects of the internal design structure of the direct needle-driven piezo-injector and its optimal design parameters on the spray formation and CI combustion are discussed experimentally. 2. Experimental Apparatus and Procedure 2.1. Design Concept of Prototype DPIs By using a solenoid coil or a piezo stack as an actuator, diesel injectors have recently been designed in response to stringent emission regulations. A solenoid-driven injector, operated by peak and hold Appl. Sci. 2017, 7, 320 3 of 16 Appl. Sci. 2017, 7, 320 3 of 16 size. However, this type of injector is likely to exhibit an injection delay. Therefore, a number of current control, is still commonly used for its cost effectiveness, reliability, and smaller unit size. studies [6,7] have focused on the characteristics and injection flexibility of a piezo-driven injector. However, this type of injector is likely to exhibit an injection delay. Therefore, a number of studies [6,7] A DPI directly utilizes a longer piezo stack to lift the needle for injections. The longer piezo stack have focused on the characteristics and injection ﬂexibility of a piezo-driven injector. generates enough displacement for the lifting using an adapting amplifier located between the needle A DPI directly utilizes a longer piezo stack to lift the needle for injections. The longer piezo stack and the piezo stack. The elongation generated by the piezo stack is used to directly control the needle. generates enough displacement for the lifting using an adapting ampliﬁer located between the needle Appl. Sci. 2017, 7, 320 3 of 16 During DPI operation, the piezo stack shrinks when its electric energy is discharged. This causes the and the piezo stack. The elongation generated by the piezo stack is used to directly control the needle. size. However, this type of injector is likely to exhibit an injection delay. Therefore, a number of needle to move upward and reveal previously blocked injector holes. Then, high-pressure fuel is During DPI operation, the piezo stack shrinks when its electric energy is discharged. This causes studies [6,7] have focused on the characteristics and injection flexibility of a piezo-driven injector. injected through the injector holes. This allows the DPI to have a low fuel consumption by controlling the needle to move A DPI d upwar irectly utilizes a d and reveal longer pie previou zo stack to li sly blocked ft the needle injector for injectholes. ions. The Then, longer pie high-pr zo stackessur e fuel is various injection rate generates eno s. The ugh displace refore ment for , as sho the lifting wn in Fi using gur an e ad 1, the prototype DPI design concepts in this apting amplifier located between the needle injected through the injector holes. This allows the DPI to have a low fuel consumption by controlling and the piezo stack. The elongation generated by the piezo stack is used to directly control the needle. study focused on the internal DPI construction with three types. Type 1 uses a needle body, various injection rates. Therefore, as shown in Figure 1, the prototype DPI design concepts in this study During DPI operation, the piezo stack shrinks when its electric energy is discharged. This causes the a cylindrical shell, and a spring. This type is simple to make, produces a good response, and reduces needle to move upward and reveal previously blocked injector holes. Then, high-pressure fuel is focused on the internal DPI construction with three types. Type 1 uses a needle body, a cylindrical shell, the unit cost. Type 2 consists of pan-springs, plates, a needle body, a needle cylindrical shell, injected through the injector holes. This allows the DPI to have a low fuel consumption by controlling and a spring. This type is simple to make, produces a good response, and reduces the unit cost. Type 2 various injection rates. Therefore, as shown in Figure 1, the prototype DPI design concepts in this and a small spring. There is no cap to decrease the hydraulic influence and increase the needle consists of pan-springs, plates, a needle body, a needle cylindrical shell, and a small spring. There is study focused on the internal DPI construction with three types. Type 1 uses a needle body, response. Type 3 is constructed without limiting the number of pan-springs and plates. It is designed no cap to decr a cy ease lindric the al sh hydraulic ell, and a spring. inﬂuence This typeand is simincr ple to ease make, the produc needle es a good response. response, and r Type educe 3 is s constructed using a needle that has a needle body, a cylindrical shell, a spring, and cap. This design allows the unit cost. Type 2 consists of pan-springs, plates, a needle body, a needle cylindrical shell, without limiting the number of pan-springs and plates. It is designed using a needle that has a needle and a small spring. There is no cap to decrease the hydraulic influence and increase the needle hydraulics to be used. This Type 3 design is used to analyze and study the injector influence by body, a cylindrical shell, a spring, and cap. This design allows hydraulics to be used. This Type 3 response. Type 3 is constructed without limiting the number of pan-springs and plates. It is designed varying the number of plate springs and pressure plates. using a needle that has a needle body, a cylindrical shell, a spring, and cap. This design allows design is used to analyze and study the injector inﬂuence by varying the number of plate springs and hydraulics to be used. This Type 3 design is used to analyze and study the injector influence by pressure plates. varying the number of plate springs and pressure plates. + - Electrode A + - Electrode A F + - - - - Electrode B F Electrode B F Piez Pi oe e le z ct o rie c lectric ceram ics F ceram ics - + - + (a) (b) (a) (b) Figure 1. DPI proposed by this study; (a) Three prototype DPIs, (b) Principle of the inverse piezo- Figure 1. DPI proposed by this study; (a) Three prototype DPIs, (b) Principle of the inverse electric effect. Figure 1. DPI proposed by this study; (a) Three prototype DPIs, (b) Principle of the inverse piezo- piezo-electric effect. electric effect. 2.2. Experimental Apparatus and Methods 2.2.1. Maximum Injection Pressure Measurement 2.2. Experimental Apparatus and Methods 2.2. Experimental Apparatus and Methods A 1.5 kW class 3-phase motor (Hyosung Co., Seoul, Korea) was used as a high-pressure injection 2.2.1. Maximum Injection Pressure Measurement pump for a common-rail system. It can pressurize up to 200 MPa, in order to inject at the maximum 2.2.1. Maximum Injection Pressure Measurement possible injection pressure, as listed in Table 1. A temperature sensor is placed inside the fuel tank to A 1.5 kW class 3-phase motor (Hyosung Co., Seoul, Korea) was used as a high-pressure injection automatically turn on and off the radiator and a cooling fan is used to keep the temperature constant. A 1.5 kW class 3-phase motor (Hyosung Co., Seoul, Korea) was used as a high-pressure injection As shown in Figure 2, the injection period and the injection pressure were controlled through the pump for a common-rail system. It can pressurize up to 200 MPa, in order to inject at the maximum pump for a common-rail system. It can pressurize up to 200 MPa, in order to inject at the maximum common rail controller (Zenobalti Co., Model:ZB-9013, Daejeon, Korea) and the universal piezo possible injection pressure, as listed in Table 1. A temperature sensor is placed inside the fuel tank to possible injection pressure, as listed in Table 1. A temperature sensor is placed inside the fuel tank to injector driver (Zenobalti Co., Model:ZB-6200, Korea). automatically turn on and off the radiator and a cooling fan is used to keep the temperature constant. automatically turn on and off the radiator and a cooling fan is used to keep the temperature constant. As shown in Figure 2, the injection period and the injection pressure were controlled through the As shown in Figure 2, the injection period and the injection pressure were controlled through the common rail controller (Zenobalti Co., Model:ZB-9013, Daejeon, Korea) and the universal piezo injector common rail controller (Zenobalti Co., Model:ZB-9013, Daejeon, Korea) and the universal piezo driver (Zenobalti Co., Model:ZB-6200, Korea). injector driver (Zenobalti Co., Model:ZB-6200, Korea). (a) (b) Figure 2. Fuel injection control system; (a) Common-rail controller, (b) Injector driver. (a) (b) Figure 2. Fuel injection control system; (a) Common-rail controller, (b) Injector driver. Figure 2. Fuel injection control system; (a) Common-rail controller, (b) Injector driver. Appl. Sci. 2017, 7, 320 4 of 16 Appl. Sci. 2017, 7, 320 4 of 16 Appl. Sci. 2017, 7, 320 4 of 16 Table 1. Experimental conditions for maximum injection pressure measurement. Table 1. Experimental conditions for maximum injection pressure measurement. Table 1. Experimental conditions for maximum injection pressure measurement. Item Speciﬁcation Item Specification Item Fuel Spec ULSD(Ultra-Low-Sulfurification Diesel) Fuel ULSD(Ultra-Low-Sulfur Diesel) Injection duration (s) 1000 Fuel ULSD(Ultra-Low-Sulfur Diesel) Injection duration (µs) 1000 Injection pressure (MPa) 30~180 Injection duration (µs) 1000 Injection pressure (MPa) 30~180 Injection pressure (MPa) 30~180 2.2.2. Injection Rate Measurement 2.2.2. Injection Rate Measurement 2.2.2. Injection Rate Measurement There are two methods which can be used to measure the injection rate. The Zeuch method There are two methods which can be used to measure the injection rate. The Zeuch method obtains the injection rate after injecting fuel into the pressure vessel. With the Bosch-Tube method, fuel There are two methods which can be used to measure the injection rate. The Zeuch method obtains the injection rate after injecting fuel into the pressure vessel. With the Bosch-Tube method, is injected into the pipe to detect the pressure change in the injection rate meter. Then, the injection obtains the injection rate after injecting fuel into the pressure vessel. With the Bosch-Tube method, fuel is injected into the pipe to detect the pressure change in the injection rate meter. Then, the rate is measured. In this study, the Bosch-Tube method was used [5]. The Bosch-Tube measurement fuel is injected into the pipe to detect the pressure change in the injection rate meter. Then, the injection rate is measured. In this study, the Bosch-Tube method was used [5]. The Bosch-Tube principle involves feeding the fuel into the pipe with a cross-sectional area A and selecting the fuel injection rate is measured. In this study, the Bosch-Tube method was used [5]. The Bosch-Tube measurement principle involves feeding the fuel into the pipe with a cross-sectional area A and control volume in the pipe ﬂowing at the speed of sound as c, when the fuel in the pipe moves at measurement principle involves feeding the fuel into the pipe with a cross-sectional area A and selecting the fuel control volume in the pipe flowing at the speed of sound as c, when the fuel in the a speed of u. The characteristics of the fuel moving inside can be expressed as shown in Figure 3. selecting the fuel control volume in the pipe flowing at the speed of sound as c, when the fuel in the pipe moves at a speed of u. The characteristics of the fuel moving inside can be expressed as shown By knowing the sonic velocity and fuel density in the pipe and ascertaining the pressure change in the pipe moves at a speed of u. The characteristics of the fuel moving inside can be expressed as shown in Figure 3. By knowing the sonic velocity and fuel density in the pipe and ascertaining the pressure chamber, the injection rate can be calculated. in Figure 3. By knowing the sonic velocity and fuel density in the pipe and ascertaining the pressure change in the chamber, the injection rate can be calculated. change in the chamber, the injection rate can be calculated. Figure 3. Control volume in the pipe to measure the injection rate. Figure 3. Control volume in the pipe to measure the injection rate. Figure 3. Control volume in the pipe to measure the injection rate. (a) (b) (a) (b) Figure 4. Injection rate measurement; (a) Injection rate measuring device, (b) Schematic diagram. Figure 4. Injection rate measurement; (a) Injection rate measuring device, (b) Schematic diagram. Figure 4. Injection rate measurement; (a) Injection rate measuring device, (b) Schematic diagram. Table 2. Experimental conditions for measuring injection rate. Table 2. Experimental conditions for measuring injection rate. Table 2. Experimental conditions for measuring injection rate. Properties Specification Properties Specification Fuel ULSD Properties Speciﬁcation Fuel ULSD Tube length (m) 12.607 Fuel ULSD Tube length (m) 12.607 Tube area (m ) 1.6409×10 Tube length (m) 12.607 Tube area (m ) 1.6409×10 Injection pressure (MPa) 30~180 2 5 Tube area (m ) 1.6409 10 Injection pressure (MPa) 30~180 Duration (µs) 1000 Injection pressure (MPa) 30~180 Duration (µs) 1000 Duration (s) 1000 As shown in Figure 4, the injection rate measurement system consists of a measuring tube with As shown in Figure 4, the injection rate measurement system consists of a measuring tube with a length of about 12.6 m, needle valve, accumulator, pressure gauge, and regulator. The injector is a length of about 12.6 m, needle valve, accumulator, pressure gauge, and regulator. The injector is installed on the adapter in the injection rate measuring device and is connected to the common-rail installed on the adapter in the injection rate measuring device and is connected to the common-rail to supply high-pressure fuel. When fuel injection starts, the injection rate is measured by passing the to supply high-pressure fuel. When fuel injection starts, the injection rate is measured by passing the Appl. Sci. 2017, 7, 320 5 of 16 As shown in Figure 4, the injection rate measurement system consists of a measuring tube with a length of about 12.6 m, needle valve, accumulator, pressure gauge, and regulator. The injector is installed on the adapter in the injection rate measuring device and is connected to the common-rail to supply high-pressure fuel. When fuel injection starts, the injection rate is measured by passing the ﬂuid through the pressure sensor (Kistler, Model:6052c, Winterthur, Switzerland) that is mounted on Appl. Sci. 2017, 7, 320 5 of 16 the adapter. The experimental conditions for the injection rate are expressed in Table 2. fluid through the pressure sensor (Kistler, Model:6052c, Winterthur, Switzerland) that is mounted on 2.2.3. Spray Visualization Measurement the adapter. The experimental conditions for the injection rate are expressed in Table 2. Spray visualization experiments were carried out in order to directly conﬁrm the injection speed and behavior 2.2.3. Sprafor y Vieach sualizat prion Mea ototype. surem In this ent study, a static chamber with a volume of 885 cc was used, and images were taken with a high speed camera through a quartz glass window with a diameter Spray visualization experiments were carried out in order to directly confirm the injection speed of 108 mm and a thickness of 50 mm. This experiment was conducted at a normal temperature and behavior for each prototype. In this study, a static chamber with a volume of 885 cc was used, and pressure, and used a Bosch common-rail system. Since the prototype injector used in the and images were taken with a high speed camera through a quartz glass window with a diameter of experiment 108 mmwas and a a tpiezo-driven hickness of 50 m type, m. Thi it swas expecontr riment olled was conduct using aeuniversal d at a norm piezo al temp actuation erature andriver d pressure, and used a Bosch common-rail system. Since the prototype injector used in the experiment (Zenobalti Co., Model:ZB-6200, Korea). Detailed speciﬁcations of the high-speed camera (Vision was a piezo-driven type, it was controlled using a universal piezo actuation driver (Zenobalti Co., Research, Model:Phantom V7.3, Wayne, NJ, USA) used in this study are shown in Table 3. Direct Model:ZB-6200, Korea). Detailed specifications of the high-speed camera (Vision Research, lighting spray images were obtained using an 80 W LED light source. The experimental conditions Model:Phantom V7.3, Wayne, NJ, USA) used in this study are shown in Table 3. Direct lighting spray of the spray visualization experiment are shown in Table 4. Figure 5 shows a schematic and actual images were obtained using an 80 W LED light source. The experimental conditions of the spray picture of the injection test apparatus used in this experiment. visualization experiment are shown in Table 4. Figure 5 shows a schematic and actual picture of the injection test apparatus used in this experiment. Table 3. Speciﬁcations of the high-speed camera. Table 3. Specifications of the high-speed camera. Properties Speciﬁcation Properties Specification Model Phantom V7.3 Model Phantom V7.3 Max. resolution 800 600 Max. resolution 800 × 600 Sensor type 16 bit sensor Sensor type 16 bit sensor Max. PPS 500,000 Max. PPS 500,000 Trigger source TTL signal Trigger source TTL signal Memory 8 G DDR RAM Memory 8 G DDR RAM Table 4. Experimental conditions for spray visualization. Table 4. Experimental conditions for spray visualization. Properties Properties Spec Speciﬁcation ification Injection duration (µs) 1000 Injection duration (s) 1000 Injection Pressure (MPa) 120 Injection Pressure (MPa) 120 Resolution (pixels) 320 × 320 Resolution (pixels) 320 320 Sample Sample ra ratete (f (frame rame per per second) second) 25,000 25,000 Exposure (s) 40 Exposure (µs) 40 Ambient condition 1 atm. 300 K Ambient condition 1 atm. 300 K Spray Spray Co Cone ne An Angle gle Sp Spr ra ay y T Ti ip p P Pe en ne et tr ra at ti io on n 5m 5mm m (a) (b) Figure 5. Spray visualization measurement; (a) Visualization apparatus, (b) Definition of spray Figure 5. Spray visualization measurement; (a) Visualization apparatus, (b) Definition of spray parameter. parameter. Appl. Sci. 2017, 7, 320 6 of 16 Appl. Sci. 2017, 7, 320 6 of 16 2.2.4. Real Engine Combustion System 2.2.4. Real Engine Combustion System The engine used in this study is a direct-injection type single-cylinder CI engine (Daedong Co., The engine used in this study is a direct-injection type single-cylinder CI engine (Daedong Co., Model:ND10DE, Kyeongsan, Korea). Figure 6 shows the single-cylinder diesel engine system and its Model:ND10DE, Kyeongsan, Korea). Figure 6 shows the single-cylinder diesel engine system and its speciﬁcations are given in Table 5. Table 6 shows the combustion test conditions for the single-cylinder specifications are given in Table 5. Table 6 shows the combustion test conditions for the single- diesel engine. The emission performance of the most optimized DPI type and the conventional type cylinder diesel engine. The emission performance of the most optimized DPI type and the was analyzed by measuring the exhaust gas through an exhaust gas measurement system (Horiba, conventional type was analyzed by measuring the exhaust gas through an exhaust gas measurement Model:MEXA-554JKNO , Kyoto, Japan). The main speciﬁcations of the exhaust gas measurement system (Horiba, Model:MEXA-554JKNOx, Kyoto, Japan). The main specifications of the exhaust gas system are shown in Table 7. measurement system are shown in Table 7. Figure 6. Experiment system of single-cylinder diesel engine. Figure 6. Experiment system of single-cylinder diesel engine. Table 5. Main specifications of the single-cylinder diesel engine. Table 5. Main speciﬁcations of the single-cylinder diesel engine. Properties Specification Properties Speciﬁcation Engine type Horizontal DI diesel engine Bore Engine × Stroke type(mm) Horizontal 95DI ×diesel 95 engine Bore Stroke (mm) 95 95 Displacement (cc) 673 Displacement (cc) 673 Compression ratio 18.0 Compression ratio 18.0 Valve type SOHC (Single OverHead Camshaft) 2 valve Valve type SOHC (Single OverHead Camshaft) 2 valve Combustion chamber type Reentrant bowl-in-piston Combustion chamber type Reentrant bowl-in-piston Table 6. Engine combustion test conditions. Table 6. Engine combustion test conditions. Item Specification Item Speciﬁcation Engine speed (rpm) 1200 Engine speed (rpm) 1200 Injection quantity (mg/stroke) 3, 7 Injection quantity (mg/stroke) 3, 7 Injection timing (BTDC) 5, 15, 20 Injection timing (BTDC) 5, 15, 20 Table 7. Specifications of the exhaust gas measuring device. Table 7. Speciﬁcations of the exhaust gas measuring device. Item CO HC CO2 NOx O Item CO HC CO NO O Measuring rage 0.00~10.00% 0~10,000 ppm 0.00~20. 2 00% 0~5000 ppm x 0.00~25.00% Repeatability ± 0.06 ± 12 ± 0.5 ± 7 - Measuring rage 0.00~10.00% 0~10,000 ppm 0.00~20.00% 0~5000 ppm 0.00~25.00% Repeatability 0.06 12 0.5 7 - Measurement methods NDIR NDIR NDIR CLD CLD Measurement methods NDIR NDIR NDIR CLD CLD Responsibility up to 10 s, 90% response Responsibility up to 10 s, 90% response Appl. Sci. 2017, 7, 320 7 of 16 Appl. Sci. 2017, 7, 320 7 of 16 3. Results and Discussion 3. Results and Discussion 3.1. Design Complexity and Specialized Knowledge in Needed for Each DPI Prototype 3.1. Design Complexity and Specialized Knowledge in Needed for Each DPI Prototype 3.1.1. Type 1 DPI 3.1.1. Type 1 DPI In Type 1, the spring (70C hardened SW-C steel wire) was fabricated according to the wire In Type 1, the spring (70C hardened SW-C steel wire) was fabricated according to the wire thickness, outer diameter, number of spring revolutions, and length. It was inserted into the injector thickness, outer diameter, number of spring revolutions, and length. It was inserted into the injector inner-body, and then the needle body and needle cylinder were also inserted. Experimental results inner-body, and then the needle body and needle cylinder were also inserted. Experimental results showed that there is no problem in spraying up to the available injection pressure of 120 MPa. showed that there is no problem in spraying up to the available injection pressure of 120 MPa. However, However, at 130 MPa, normal injection becomes intermittent. For this reason, the injector normally at 130 MPa, normal injection becomes intermittent. For this reason, the injector normally injected only injected only 20 of the total injection signals. 20 of the total injection signals. As shown in Figure 7, each graph is divided into steel wire thicknesses of 0.7 mm and 0.8 mm. As shown in Figure 7, each graph is divided into steel wire thicknesses of 0.7 mm and 0.8 mm. It is It is shown that the production spring constant (K) increases toward the right of the X-axis, and the shown that the production spring constant (K) increases toward the right of the X-axis, and the actual actual number of injections continually increases. The reason for this is that the number of actual number of injections continually increases. The reason for this is that the number of actual injections is injections is expected to increase as the spring stiffness and K increases. The spring with high K expected to increase as the spring stiffness and K increases. The spring with high K becomes harder becomes harder than the lower K springs, such that the pressure inside the injector provides the force than the lower K springs, such that the pressure inside the injector provides the force to lift the injector to lift the injector needle. It was found that the best Type 1 performance was achieved by the spring needle. It was found that the best Type 1 performance was achieved by the spring with a steel wire with a steel wire thickness of 0.8 mm, outer diameter of 4.8 mm, five spring revolutions, and a length thickness of 0.8 mm, outer diameter of 4.8 mm, ﬁve spring revolutions, and a length of 6 mm. of 6 mm. (a) (b) Figure 7. Actual number of injections according to the thickness of the steel wire; (a) 0.7 mm, Figure 7. Actual number of injections according to the thickness of the steel wire; (a) 0.7 mm, (b) 0.8 mm. (b) 0.8 mm. 3.1.2. Type 2 DPI 3.1.2. Type 2 DPI As shown in Figure 8, the total number of leaf springs and pressure plates is speciﬁed as a As shown in Figure 8, the total number of leaf springs and pressure plates is specified as a design design parameter, and the needle part is assembled by inserting the needle body, cylinder, and parameter, and the needle part is assembled by inserting the needle body, cylinder, and spring. The spring. The number of leaf springs was reduced by one and three cases were analyzed. As shown in number of leaf springs was reduced by one and three cases were analyzed. As shown in Figure 9, the Figure 9, the Type 2 needle conﬁguration allows ﬂuid to move more freely than the conventional type. Type 2 needle configuration allows fluid to move more freely than the conventional type. When When hydraulic pressure is used in the conventional type, the needle moving distance is ampliﬁed hydraulic pressure is used in the conventional type, the needle moving distance is amplified and the and the needle can move with a small pressure force. However, the injection responsiveness is needle can move with a small pressure force. However, the injection responsiveness is considered to considered to be slow. Therefore, the design purpose of Type 2 is to increase the injection responsiveness be slow. Therefore, the design purpose of Type 2 is to increase the injection responsiveness by by allowing pressured ﬂuid to ﬂow freely without hydraulic pressure, by taking advantage of the allowing pressured fluid to flow freely without hydraulic pressure, by taking advantage of the mechanical characteristics. mechanical characteristics. Case 2-1 in Figure 8 has an available maximum injection pressure of 130 MPa, as shown in Case 2-1 in Figure 8 has an available maximum injection pressure of 130 MPa, as shown in Figure 10. Figure 10. The injection pressures of Case 2-2 and Case 2-3 were about 100 bar lower than Case 2-1. The injection pressures of Case 2-2 and Case 2-3 were about 100 bar lower than Case 2-1. It was found It was found that the available maximum injection pressure can be obtained, depending on the number that the available maximum injection pressure can be obtained, depending on the number of leaf of leaf springs. If one leaf spring in Type 2 is absent, the force for lifting the needle under CR high springs. If one leaf spring in Type 2 is absent, the force for lifting the needle under CR high pressure pressure is insufﬁcient. However, when two leaf springs are provided, it becomes possible to lift the is insufficient. However, when two leaf springs are provided, it becomes possible to lift the needle Appl. Sci. 2017, 7, 320 8 of 16 Appl. Appl. Sci. Sci. 2017 2017 , , 77 , 320 , 320 8 of 8 of 16 16 Appl. Sci. 2017, 7, 320 8 of 16 with a with a gre gre aa ter ter force, due to force, due to the incre the incre aa se se in the elasti in the elasti c c force. It is con force. It is con ss ider ider ed that t ed that t h h e elastic e elastic forc forc e in the e in the needle with a greater force, due to the increase in the elastic force. It is considered that the elastic force with a greater force, due to the increase in the elastic force. It is considered that the elastic force in the design of Type 2 is determined by the number of leaf springs. design of Type 2 is determined by the number of leaf springs. in design of Typ the design of e 2 is determined by Type 2 is determined the n byu the mber of number leaf of springs. leaf springs. Figure 8. Detailed designs of Type 2 DPIs. Figure 8. Detailed designs of Type 2 DPIs. Figure 8. Detailed designs of Type 2 DPIs. Figure 8. Detailed designs of Type 2 DPIs. (( aa) ( ) ( b b )) (a) (b) Figure 9. Figure 9. Com Com p p arison of nee arison of nee d d le le configu configu rr ation; ( ation; ( aa ) Type ) Type 2, ( 2, ( b b ) Convention ) Convention al Type DPI al Type DPI . . Figure 9. Comparison of needle configuration; (a) Type 2, (b) Conventional Type DPI. Figure 9. Comparison of needle conﬁguration; (a) Type 2, (b) Conventional Type DPI. Figure 10. Comparison of the maximum injection pressure of Type 2 DPIs. Figure 10. Comparison of the maximum injection pressure of Type 2 DPIs. Figure 10. Comparison of the maximum injection pressure of Type 2 DPIs. Figure 10. Comparison of the maximum injection pressure of Type 2 DPIs. 3. 3. 1. 1. 3. 3. T T y y p p ee 3 DPI 3 DPI 3.1.3. Type 3 DPI 3.1.3. Type 3 DPI A A ss s s h h own i own i n n F F ii gu gu rr ee 11 11 , the Ty , the Ty pe pe 3 3 i i n n te te rr n n aa ll s s tt rr u u cc tt ur ur ee wa wa ss a a n n aa ll y y zz ed by ed by ei ei g g h h t ca t ca ss ee ss wi wi th di th di ff ff ee rr ent ent As shown in Figure 11, the Type 3 internal structure was analyzed by eight cases with different As shown in Figure 11, the Type 3 internal structure was analyzed by eight cases with different combinat combinations ions of pr of press essuure re plat plates es and and l leeaf af sprin springgs,s, in includ cluding t ing thhee s saame no me nozzzzle le c coonfig nfigur urat ation ion as as aa combinations of pressure plates and leaf springs, including the same nozzle configuration as a combinations of pressure plates and leaf springs, including the same nozzle conﬁguration as a convent conventiional onal ttyype. In Type pe. In Type 3 3 DPIs DPIs, , it it wa was fo s found und t th hat at t t h he av e av ai aillaable ble maximum maximum inj injeect ction pres ion pressu sure w re w aass conventional type. In Type 3 DPIs, it was found that the available maximum injection pressure was conventional type. In Type 3 DPIs, it was found that the available maximum injection pressure 11 80 MPa i 80 MPa i n n all all ca ca ses, excep ses, excep tt Ca Ca ses ses 33 -- 3 a 3 a n n d 3-6 d 3-6 i i n n Fi Fi gure 12 gure 12 . In Ca . In Ca se se 33 -- 33 wi wi th th 80 80 0 0 ba ba r r aa n n d Case d Case 33 -- 66 , the , the 180 MPa in all cases, except Cases 3-3 and 3-6 in Figure 12. In Case 3-3 with 800 bar and Case 3-6, the was 180 MPa in all cases, except Cases 3-3 and 3-6 in Figure 12. In Case 3-3 with 800 bar and Case pr pr ee ss ss u u rr ee d fu d fu el el l l ee aa k k ee d d a a n n d the d the fu fu el el s s p p rr aa y y cc o o u u ll d not be d not be me me aa ss u u rr ee d d . It . It wa wa s thou s thou g g h h t tha t tha tt Ca Ca ss ee 3 3 -- 3 3 d d ii d not d not pressured fuel leaked and the fuel spray could not be measured. It was thought that Case 3-3 did not 3-6, the pressured fuel leaked and the fuel spray could not be measured. It was thought that Case inject inject m m o o re re t t h h an an 88 00 00 b b aa rr , ,d d u u e t e t o o t t h h ee ab ab se se nce of nce of a p a p ll at at e e sp sp ring ring for for am am p p lif lif ying ying t t h h e p e p ii ezo ezo st st ack ack for for cc e, e, as as inject more than 800 bar, due to the absence of a plate spring for amplifying the piezo stack force, as 3-3 did not inject more than 800 bar, due to the absence of a plate spring for amplifying the piezo seen seen in t in t h h e T e T y y pe 2 pe 2 DPI. DPI. U U n n like like Ca Ca se se 22 -- 3 o 3 o ff Type Type 2 2 , t , t h h e re e re ason ason for for not not inject inject ing ing under under 11 22 00 M M P P a a is is t t h h at at seen in the Type 2 DPI. Unlike Case 2-3 of Type 2, the reason for not injecting under 120 MPa is that stack force, as seen in the Type 2 DPI. Unlike Case 2-3 of Type 2, the reason for not injecting under there is not enough elast there is not enough elast ii cc ii ty to use the needle’s hy ty to use the needle’s hy dr dr aulic pre aulic pre ssur ssur ee . This in . This in dic dic aa tes that when there is tes that when there is there is not enough elasticity to use the needle’s hydraulic pressure. This indicates that when there is 120 MPa is that there is not enough elasticity to use the needle’s hydraulic pressure. This indicates that not a not a large large en en ough e ough e ll astic astic force, force, it is it is adv adv a a ntageo ntageo us us to to spray spray only w only w h h en using en using the the needle needle mechanically, mechanically, not a large enough elastic force, it is advantageous to spray only when using the needle mechanically, when there is not a large enough elastic force, it is advantageous to spray only when using the needle wi without a thout anny y hydrauli hydraulic pressure. In c pressure. In Case 3-6, the Case 3-6, the difference betwe difference betweeen n C Caase 3- se 3-6 a 6 annd d C Caase 3- se 3-4 i 4 iss tthhe e without any hydraulic pressure. In Case 3-6, the difference between Case 3-6 and Case 3-4 is the mechanically, without any hydraulic pressure. In Case 3-6, the difference between Case 3-6 and Case thickness o thickness o ff the pressure plate. The thic the pressure plate. The thic kness o kness o ff th th e thick pressure e thick pressure plate is 0.65 plate is 0.65 mm, and the thickness mm, and the thickness thickness of the pressure plate. The thickness of the thick pressure plate is 0.65 mm, and the thickness 3-4 is the thickness of the pressure plate. The thickness of the thick pressure plate is 0.65 mm, and the of of the thi the thinn pressure pla pressure platte is 0 e is 0.3 .3 mm. The tota mm. The totall leng length of th of the va the varia riabbllees, s, excludi excludinng the l g the leength ngth of of the the of the thin pressure plate is 0.3 mm. The total length of the variables, excluding the length of the thickness of the thin pressure plate is 0.3 mm. The total length of the variables, excluding the length of needle needle an an d p d p ii ezo ezo stack stack , ,is is 1.95 mm 1.95 mm for for C C aa se se 3-4 3-4 and and 1.6 mm 1.6 mm for for Case 3-6. Therefo Case 3-6. Therefo rr e, e, it w it w aa s fo s fo und that und that needle and piezo stack, is 1.95 mm for Case 3-4 and 1.6 mm for Case 3-6. Therefore, it was found that the minimum length inside the injector must exceed 1.6 mm. the minimum length inside the injector must exceed 1.6 mm. the minimum length inside the injector must exceed 1.6 mm. Appl. Sci. 2017, 7, 320 9 of 16 Appl. Sci. 2017, 7, 320 9 of 16 the needle and piezo stack, is 1.95 mm for Case 3-4 and 1.6 mm for Case 3-6. Therefore, it was found For Type 3 DPI Cases that have an injection performance of up to 180 MPa, spray images were that the Appl. minimum Sci. 2017, 7, 320 length inside the injector must exceed 1.6 mm. 9 of 16 acquired by a high-speed camera and the spray speed was compared, as shown in Figure 13. For Type 3 DPI Cases that have an injection performance of up to 180 MPa, spray images were As known, the spray speed was derived by the definition of the spray tip penetration and spray cone For Type 3 DPI Cases that have an injection performance of up to 180 MPa, spray images were acquired by a high-speed camera and the spray speed was compared, as shown in Figure 13. As known, angle, with the edge defined as a line of 80% transmittance of back-light in the raw spray images. acquired by a high-speed camera and the spray speed was compared, as shown in Figure 13. the spray speed was derived by the deﬁnition of the spray tip penetration and spray cone angle, with It was found that Case 3-8 had the best spray performance with an average speed of 107.3 m/s, an As known, the spray speed was derived by the definition of the spray tip penetration and spray cone the edge deﬁned as a line of 80% transmittance of back-light in the raw spray images. It was found that observed-maximum speed of 144 m/s, and an SOI (start of injection) value of 80 µs. The observed- angle, with the edge defined as a line of 80% transmittance of back-light in the raw spray images. Case 3-8 had the best spray performance with an average speed of 107.3 m/s, an observed-maximum maximum It was found that speed in this study mean Case 3-8 had the s that the spra best spray perfor y hmance with as not reache an aver d the real m age speed of 107.3 aximum speed b m/s, an e cause speed of 144 m/s, and an SOI (start of injection) value of 80 s. The observed-maximum speed in observed-maximum speed of 144 m/s, and an SOI (start of injection) value of 80 µs. The observed- the bore size of the test engine used in this study is 95 mm, as shown in Table 5, which is about the this study means that the spray has not reached the real maximum speed because the bore size of maximum speed in this study means that the spray has not reached the real maximum speed because same as the 100 mm diameter disk shown in Figure 5b. In other words, the high speed camera shoots the test the bore size engine used of the test engine in this studyused is 95in this study mm, as shown is 95 mm, in Table as shown i 5, which n Ta isble 5, whi about the ch is same about the as the 100 the fuel stream until it reaches the end of the disk and identifies the highest speed. When the fuel is same as the 100 mm diameter disk shown in Figure 5b. In other words, the high speed camera shoots mm diameter disk shown in Figure 5b. In other words, the high speed camera shoots the fuel stream injected into the cylinder, it is blocked by the inner wall of the cylinder and the maximum speed is the fuel stream until it reaches the end of the disk and identifies the highest speed. When the fuel is until it reaches the end of the disk and identiﬁes the highest speed. When the fuel is injected into the not fully achieved in the real spray chamber. injected into the cylinder, it is blocked by the inner wall of the cylinder and the maximum speed is cylinder, it is blocked by the inner wall of the cylinder and the maximum speed is not fully achieved in In some Type 3 DPIs, as the elasticity increases, the needle movement is amplified and its not fully achieved in the real spray chamber. the real spray chamber. moving speed increases. Therefore, it can be expected to have a good effect on two- or multi-injection In some Type 3 DPIs, as the elasticity increases, the needle movement is amplified and its In some Type 3 DPIs, as the elasticity increases, the needle movement is ampliﬁed and its moving schemes that are favorable for enhancing the air-fuel mixing rate of the diffusion combustion in the moving speed increases. Therefore, it can be expected to have a good effect on two- or multi-injection speed increases. Therefore, it can be expected to have a good effect on two- or multi-injection schemes CI engine. schemes that are favorable for enhancing the air-fuel mixing rate of the diffusion combustion in the that are favorable for enhancing the air-fuel mixing rate of the diffusion combustion in the CI engine. CI engine. Case 3-1 (leaf spring 2, thin plate 1, thick plate 1) Case 3-2 (leaf spring 1, thick plate 2) Case 3-1 (leaf spring 2, thin plate 1, thick plate 1) Case 3-2 (leaf spring 1, thick plate 2) Case 3-3 (thick plate 3) Case 3-4 (leaf spring 2, thick plate 1) Case 3-3 (thick plate 3) Case 3-4 (leaf spring 2, thick plate 1) Case 3-5 (leaf spring 2, thin plate 2) Case 3-6 (leaf spring 2, thin plate 1) Case 3-5 (leaf spring 2, thin plate 2) Case 3-6 (leaf spring 2, thin plate 1) Case 3-7 (leaf spring 2, thin plate 3) Case 3-8 (leaf spring 3, thin plate 1, thick plate 1) Case 3-7 (leaf spring 2, thin plate 3) Case 3-8 (leaf spring 3, thin plate 1, thick plate 1) Figure 11. Detailed designs of Type 3 DPIs. Figure Figure 11. 11. Detailed Detaileddesigns designs of of T Ty ype pe 3 DPI 3 DPIs.s. Figure 12. Comparison of the maximum injection pressures of Type 3 DPIs. Figure 12. Comparison of the maximum injection pressures of Type 3 DPIs. Figure 12. Comparison of the maximum injection pressures of Type 3 DPIs. Appl. Sci. 2017, 7, 320 10 of 16 Appl. Sci. 2017, 7, 320 10 of 16 Appl. Sci. 2017, 7, 320 10 of 16 Figure 13. Comparison of the spray speed results in Type 3 DPIs. Figure 13. Comparison of the spray speed results in Type 3 DPIs. Figure 13. Comparison of the spray speed results in Type 3 DPIs. 3.2. Comparison and Evaluation of Prototype DPIs 3.2. 3.2. C Comparison omparison an and d E Evaluation valuation of Pro of Prototype totype DP DPIs Is The following prototype DPIs were selected to compare the injection performance, spray, and The following prototype DPIs were selected to compare the injection performance, spray, and The following prototype DPIs were selected to compare the injection performance, spray, and combustion characteristics. (1) The Type 1 DPI has a 0.8 mm thick steel wire spring with an outer combustion combustion c characteristics. haracteristics. ((1) 1) Th The e Typ Type e 1 D 1 DPI PI h has as a a 0.8 0.8 m mm m t thick hick st steel eel wire wire sp spring ring w with ith an an outer outer diameter of 4.8 mm, spring rate of five, and a length of 6 mm; (2) For the Type 2 DPI, Case 2-1 diameter of 4.8 mm, spring rate of ﬁve, and a length of 6 mm; (2) For the Type 2 DPI, Case 2-1 (two leaf diameter of 4.8 mm, spring rate of five, and a length of 6 mm; (2) For the Type 2 DPI, Case 2-1 (two leaf springs + one thick pressure plate + one thin pressure plate) was selected; (3) For the Type springs (two leaf +springs one thick + one thic pressurk pressure p e plate + one late + one th thin pressur in pressure p e plate) was latselected; e) was selecte (3) For d; (3) For the the Type 3Type DPI, 3 DPI, Case 3-8 (three plate springs + one thick pressure plate + one thin pressure plate) was selected. Case 3-8 (three plate springs + one thick pressure plate + one thin pressure plate) was selected. 3 DPI, Case 3-8 (three plate springs + one thick pressure plate + one thin pressure plate) was selected. 3.2.1. Comparison of Maximum Injection Pressure Results 3.2.1. Comparison of Maximum Injection Pressure Results 3.2.1. Comparison of Maximum Injection Pressure Results The maximum injection pressure with the representative Types 1, 2, and 3, is shown in Figure 14. The maximum injection pressure with the representative Types 1, 2, and 3, is shown in Figure 14. The maximum injection pressure with the representative Types 1, 2, and 3, is shown in Figure 14. The Type 3 DPI showed the highest injection pressure at 180 MPa. It was found that the hydraulic The Type 3 DPI showed the highest injection pressure at 180 MPa. It was found that the hydraulic The Type 3 DPI showed the highest injection pressure at 180 MPa. It was found that the hydraulic application required to build up needle movement is an essential element to obtain an injection application required to build up needle movement is an essential element to obtain an injection application required to build up needle movement is an essential element to obtain an injection performance at high pressures, over 130 MPa. As a result, it can be said that the Type 3 DPI prototype performance at high pressures, over 130 MPa. As a result, it can be said that the Type 3 DPI prototype performance at high pressures, over 130 MPa. As a result, it can be said that the Type 3 DPI prototype has a similar injection performance to the conventional type DPI, and it has a better flexibility for has a similar injection performance to the conventional type DPI, and it has a better ﬂexibility for has a similar injection performance to the conventional type DPI, and it has a better flexibility for designing the internal structure of a direct needle-driven piezo-injector. designing the internal structure of a direct needle-driven piezo-injector. designing the internal structure of a direct needle-driven piezo-injector. Figure 14. Comparison of the maximum injection pressure for representative Type 1, 2 and 3 DPI. Figure 14. Comparison of the maximum injection pressure for representative Type 1, 2 and 3 DPI. Figure 14. Comparison of the maximum injection pressure for representative Type 1, 2 and 3 DPI. 3.2.2. Comparison of Mie-Scattering Spray Image Results 3.2.2. Comparison of Mie-Scattering Spray Image Results 3.2.2. Comparison of Mie-Scattering Spray Image Results In this study, LED back illumination for Mie-scattering was applied to make the initial liquid In this study, LED back illumination for Mie-scattering was applied to make the initial liquid spray In this study, LED back illumination for Mie-scattering was applied to make the initial liquid spray to full development spray visible. This enabled an accurate SOI determination and to full development spray visible. This enabled an accurate SOI determination and simultaneously spray to full development spray visible. This enabled an accurate SOI determination and simultaneously measured the initial liquid portion of the spray. Additionally, this Mie scattering measured the initial liquid portion of the spray. Additionally, this Mie scattering technique is suited to simultaneously measured the initial liquid portion of the spray. Additionally, this Mie scattering technique is suited to the spray characteristics in cold conditions, where there is insignificant the spray characteristics in cold conditions, where there is insigniﬁcant vaporization [8,9]. The spray technique is suited to the spray characteristics in cold conditions, where there is insignificant vaporization [8,9]. The spray image was taken with an exposure time of 38 µs at a resolution of 320 × image was taken with an exposure time of 38 s at a resolution of 320 320 pixels in 25,000 pps by vaporization [8,9]. The spray image was taken with an exposure time of 38 µs at a resolution of 320 × 320 pixels in 25,000 pps by a high-speed camera. a high-speed camera. 320 pixels in 25,000 pps by a high-speed camera. Figure 15 and Table 8 show the comparative results of the spray speeds with four different DPIs. Figure 15 and Table 8 show the comparative results of the spray speeds with four different DPIs. The Type 1 and 2 DPI had observed-maximum speeds of 91 m/s and 104 m/s, respectively, with an The Type 1 and 2 DPI had observed-maximum speeds of 91 m/s and 104 m/s, respectively, with an Appl. Sci. 2017, 7, 320 11 of 16 Appl. Sci. 2017, 7, 320 11 of 16 Figure 15 and Table 8 show the comparative results of the spray speeds with four different DPIs. The Type 1 and 2 DPI had observed-maximum speeds of 91 m/s and 104 m/s, respectively, with an average speed of 59 m/s and 67 m/s. In comparison to the conventional type, Types 1 and 2 have average speed of 59 m/s and 67 m/s. In comparison to the conventional type, Types 1 and 2 have slower spray speeds and a slow SOI response of 200 µs. It is known that the motion condition of a slower spray speeds and a slow SOI response of 200 s. It is known that the motion condition of a Appl. Sci. 2017, 7, 320 11 of 16 DPI needle is highly affected by the response parameter [9]. Generally, this shows that the use of DPI needle is highly affected by the response parameter [9]. Generally, this shows that the use of hydraulic pressure for needle movement enhances the response performance. In this study, Type 1 average speed of 59 m/s and 67 m/s. In comparison to the conventional type, Types 1 and 2 have hydraulic pressure for needle movement enhances the response performance. In this study, Type 1 and and 2 DPIs with the needle only being controlled by a mechanical mechanism, without the use of slower spray speeds and a slow SOI response of 200 µs. It is known that the motion condition of a 2 DPIs with the needle only being controlled by a mechanical mechanism, without the use of hydraulic hydraulic pre DPI needle ssur is highly e, at abo affect ut 80 µ ed by the response par s, would be slower ame tha ter [9]. Gene n the conv rally, this sh entional type i ows that the use of n terms of the SOI pressure, at about 80 s, would be slower than the conventional type in terms of the SOI response. hydraulic pressure for needle movement enhances the response performance. In this study, Type 1 response. In particular, by considering the internal actuation structure of Type 2 and the conventional In particular, by considering the internal actuation structure of Type 2 and the conventional type DPI, and 2 DPIs with the needle only being controlled by a mechanical mechanism, without the use of type DPI, it can be seen that the plate spring and the pressure plate are key factors influencing the it can be seen that the plate spring and the pressure plate are key factors inﬂuencing the determination hydraulic pressure, at about 80 µs, would be slower than the conventional type in terms of the SOI determination of the spray characteristics. of the spray characteristics. response. In particular, by considering the internal actuation structure of Type 2 and the conventional type DPI, it can be seen that the plate spring and the pressure plate are key factors influencing the determination of the spray characteristics. Figure 15. Comparison of spray speed with various D PIs. Figure 15. Comparison of spray speed with various DPIs. Figure 15. Comparison of spray speed with various DPIs. Table 8. Comparison of spray characteristics with various DPIs. Table 8. Comparison of spray characteristics with various DPIs. Table 8. Comparison of spray characteristics with various DPIs. Item Type1 Type2 Type3 Conventional Type Item Type 1 Type 2 Type 3 Conventional Type Item Type1 Type2 Type3 Conventional Type Max. speed (m/s) 91 104 113 103 Max. speed (m/s) 91 104 113 103 Max. speed (m/s) 91 104 113 103 Ave. speed (m/s) 59 67 80 69 Ave.Ave. speedsp(m/s)eed (m/s) 5959 67 67 80 80 69 69 SOI (µs) 200 200 80 120 SOI (s) 200 200 80 120 SOI (µs) 200 200 80 120 Conventional Type DPI Conventional Type DPI 80 μs 120 μs280 μs320 μs 80 μs 120 μs280 μs320 μs Type 3 DPI (as a representative prototype) Type 3 DPI (as a representative prototype) 80 μs 120 μs280 μs320 μs Figure 16. Comparison of Mie-scattering spray image between conventional type and Type 3 DPI. 80 μs 120 μs280 μs320 μs Figure 16. Comparison of Mie-scattering spray image between conventional type and Type 3 DPI. Figure 16. In the c Comparison of M ase of the Type 3 D ie-scat PI, the tering sp observed-ma ray image be ximum tween conventional ty spray speed is 113 m/s, the pe and aver Type 3 DPI. age speed is 80 m/s, and the SOI response is 80 µs. This value means that the Type 3 DPI has a fast needle In the case of the Type 3 DPI, the observed-maximum spray speed is 113 m/s, the average speed is 80 m/s, and the SOI response is 80 µs. This value means that the Type 3 DPI has a fast needle Appl. Sci. 2017, 7, 320 12 of 16 Appl. Sci. 2017, 7, 320 12 of 16 In the case of the Type 3 DPI, the observed-maximum spray speed is 113 m/s, the average movement, since about 40 µs would be faster than the conventional type DPI. In other words, the speed is 80 m/s, and the SOI response is 80 s. This value means that the Type 3 DPI has a fast Type 3 DPI needle is able to start fuel injection at 80 µs, but the conventional type DPI is still closed, needle movement, since about 40 s would be faster than the conventional type DPI. In other words, maintaining a needle state until the SOI timing is 120 µs. the Type 3 DPI needle is able to start fuel injection at 80 s, but the conventional type DPI is still closed, As shown in Figure 16, there is an obvious difference in the spray characteristics with time after maintaining a needle state until the SOI timing is 120 s. the start of the injection. As stated previously, the Type 3 DPI had a faster needle opening than the As shown in Figure 16, there is an obvious difference in the spray characteristics with time after other three DPIs, so it displayed a rapid spray behavior until obtaining a full development spray. It the start of the injection. As stated previously, the Type 3 DPI had a faster needle opening than the is clear that the actuation structure of a direct needle-driven injector significantly affects the initial other three DPIs, so it displayed a rapid spray behavior until obtaining a full development spray. spray shape. This is due to hydraulic internal flow dynamics within the DPI nozzle. Therefore, when It is clear that the actuation structure of a direct needle-driven injector signiﬁcantly affects the initial the needle response is relatively high, the spray tip fuel penetration increases in the liquid phase spray shape. This is due to hydraulic internal ﬂow dynamics within the DPI nozzle. Therefore, when spray characteristics. the needle response is relatively high, the spray tip fuel penetration increases in the liquid phase spray characteristics. 3.2.3. Comparison of Injection Rate Results 3.2.3. Comparison of Injection Rate Results Figures 17 and 18 compare the injection rate results between Types 1, 2, 3, and the conventional type DPI. To confirm the reliability of the injection rate meter, the injection rate obtained from the Figures 17 and 18 compare the injection rate results between Types 1, 2, 3, and the conventional meter used in this study was first validated with reference data from Delphi [10]. It was observed type DPI. To conﬁrm the reliability of the injection rate meter, the injection rate obtained from the that the injection quantity of the Type 1 DPI is 3.14 mg, which is relatively smaller than Type 2 or meter used in this study was ﬁrst validated with reference data from Delphi [10]. It was observed that Type 3, as shown in Table 9. Since the spring constant K is relatively higher than the leaf spring, the the injection quantity of the Type 1 DPI is 3.14 mg, which is relatively smaller than Type 2 or Type 3, length of the compressed spring is shortened when the needle is lifted, so the distance of travel of the as shown in Table 9. Since the spring constant K is relatively higher than the leaf spring, the length of needle in the Type 1 DPI is also shortened, and the injection quantity is expected to be reduced. In the compressed spring is shortened when the needle is lifted, so the distance of travel of the needle in the case of Type 2, the injection amount is high after the start of injection, but then falls sharply. the Type 1 DPI is also shortened, and the injection quantity is expected to be reduced. In the case of Compared to Type 3, the end of the injection (EOI) of the Type 2 DPI occurs early and pressured fuel Type 2, the injection amount is high after the start of injection, but then falls sharply. Compared to moves freely in the needle. So, the Type 2 DPI has a higher injection rate than other DPI types, until Type 3, the end of the injection (EOI) of the Type 2 DPI occurs early and pressured fuel moves freely in it reaches the EOI. This is a key cause of decreasing injection quantity due to the quick closing operation. the needle. So, the Type 2 DPI has a higher injection rate than other DPI types, until it reaches the EOI. This is a key cause of decreasing injection quantity due to the quick closing operation. Figure 17. Comparison of the injection rate with various DPIs. Figure 17. Comparison of the injection rate with various DPIs. Table 9. Comparison of the injection quantity with various DPIs. Table 9. Comparison of the injection quantity with various DPIs. Type 1 Type 2 Type 3 Type 1 Type 2 Type 3 Quantity (mg) 3.14 3.53 3.56 Quantity (mg) 3.14 3.53 3.56 Appl. Sci. 2017, 7, 320 13 of 16 Appl. Sci. 2017, 7, 320 13 of 16 Appl. Sci. 2017, 7, 320 13 of 16 Figure 18. Comparison of the injection rate between conventional type and Type 3 DPI. However, the Type 3 DPI showed a more stable and sufficient injection rate than Type 1 and 2 Figure 18. Comparison of the injection rate between conventional type and Type 3 DPI. DPIs. The shape of the injection rate of the Type 3 DPI is relatively horizontal and the fuel injection Figure 18. Comparison of the injection rate between conventional type and Type 3 DPI. lasts for the longest time. This means that the Type 3 DPI has a favorable degree of freedom for However, the Type 3 DPI showed a more stable and sufficient injection rate than Type 1 and 2 controlling the injection amount and the injection shape in real injection mapping processes. However, the Type 3 DPI showed a more stable and sufﬁcient injection rate than Type 1 and 2 DPIs. The shape of the injection rate of the Type 3 DPI is relatively horizontal and the fuel injection DPIs. The shape of the injection rate of the Type 3 DPI is relatively horizontal and the fuel injection lasts Figure 18 shows a comparison of the injection rates between the conventional type and the Type lasts for the longest time. This means that the Type 3 DPI has a favorable degree of freedom for for the longest time. This means that the Type 3 DPI has a favorable degree of freedom for controlling 3 DPI. As can be seen from this graph, the conventional type shows a shorter start of injection than controlling the injection amount and the injection shape in real injection mapping processes. the injection amount and the injection shape in real injection mapping processes. Type 3. However, as compared with the results of Table 10, the injection rate and injection quantity Figure 18 shows a comparison of the injection rates between the conventional type and the Type Figure 18 shows a comparison of the injection rates between the conventional type and the Type 3 do not show much deviation. This opinion is based on using a different observational method for the 3 DPI. As can be seen from this graph, the conventional type shows a shorter start of injection than DPI. As can be seen from this graph, the conventional type shows a shorter start of injection than injection rate measurement system and the spray visualization measurement system. Moreover, the Type 3. However, as compared with the results of Table 10, the injection rate and injection quantity Type 3. However, as compared with the results of Table 10, the injection rate and injection quantity do not show much deviation. This opinion is based on using a different observational method for the first “injection signature” marked with one circle in Figure 18 was detected for the Type 3 DPI. They do not show much deviation. This opinion is based on using a different observational method for injection rate measurement system and the spray visualization measurement system. Moreover, the show an almost equal injection performance, even if the Type 3 DPI has more simple structures in the injection rate measurement system and the spray visualization measurement system. Moreover, first “injection signature” marked with one circle in Figure 18 was detected for the Type 3 DPI. They the design processes of the direct needle-driven piezo injector. the ﬁrst “injection signature” marked with one circle in Figure 18 was detected for the Type 3 DPI. show an almost equal injection performance, even if the Type 3 DPI has more simple structures in They show an almost equal injection performance, even if the Type 3 DPI has more simple structures the design processes of the direct needle-driven piezo injector. Table 10. Comparison of injection quantity between conventional type and Type 3 DPI. in the design processes of the direct needle-driven piezo injector. Table 10. Comparison of injection quantity between conventional type and Type 3 DPI. Conventional Type Type 3 Table 10. Comparison of injection quantity between conventional type and Type 3 DPI. Quantity (mg) 3.51 3.56 Conventional Type Type 3 Conventional Type Type 3 Quantity (mg) 3.51 3.56 Quantity (mg) 3.51 3.56 (a) (b) (a) (b) Figure 19. Comparison of HC and NOx emissions between conventional type and Type 3 DPI; Figure 19. Comparison of HC and NOx emissions between conventional type and Type 3 DPI; Figure 19. Comparison of HC and NO emissions between conventional type and Type 3 DPI; (a) 7 mg/stroke (injection quantity) and BTDC 10 CA (injection timing); (b) 3, 7 mg/stroke (pilot/main (a) 7 mg/stroke (injection quantity) and BTDC 10 CA (injection timing); (b) 3, 7 mg/stroke (pilot/main (a) 7 mg/stroke (injection quantity) and BTDC 10 CA (injection timing); (b) 3, 7 mg/stroke (pilot/main injection quantity) and BTDC 15, 5 CA (pilot/main injection timing). injection quantity) and BTDC 15, 5 CA (pilot/main injection timing). injection quantity) and BTDC 15, 5 CA (pilot/main injection timing). The effect of the injection rate on combustion is discussed in Section 3.2.4. It predicts that the The effect of the injection rate on combustion is discussed in Section 3.2.4. It predicts that the Type 2 DPI has a higher NOx emission due to its higher injection quantity at the beginning of the Type 2 DPI has a higher NOx emission due to its higher injection quantity at the beginning of the combustion. It can be inferred that the Type 3 DPI might have the advantage of multiple injections, combustion. It can be inferred that the Type 3 DPI might have the advantage of multiple injections, Appl. Sci. 2017, 7, 320 14 of 16 The effect of the injection rate on combustion is discussed in Section 3.2.4. It predicts that the Type 2 DPI has a higher NO emission due to its higher injection quantity at the beginning of the combustion. It can be inferred that the Type 3 DPI might have the advantage of multiple injections, because the test results indicated that it had a lower injection delay and faster injection start, with a higher initial injection quantity. Therefore, a pilot injection strategy was applied and described, as shown in Figure 19b. 3.2.4. Comparison of Engine HC and NO Emission Results Figure 19 shows the diesel engine experimental results using the same conditions applied in previous chapters. The speed of the engine was maintained at 1200 rpm, and the injection timing was set to BTDC 10 CA in a single injection case and BTDC 15 CA and 5 CA in a two-injection stage (pilot/main) injection case. The fuel injected was 3 mg and 7 mg with the pilot and main injections, respectively. This was done to maintain the total fuel mass with the single injection experiment. By ﬁxing the pilot injection timing at BTDC 15 CA, it was observed that the combustion characteristics and performance were inﬂuenced by the injection timing of the main injection. First, it was observed that HC and NO emissions for the conventional type and the Type 3 DPI were different, regardless of single and two-injection stages. As shown in Figure 19a, in the case of the Type 3 DPI at the injection quantity of 7 mg/stroke, the HC emission increased by about 8 ppm, but the NO value decreased by about 23 ppm, which is more than 50% lower than the conventional type DPI. However when the injection timing of BTDC 10 CA differs for the two-injection stage, as shown in Figure 19b. The NO emission value is about 109 ppm, which is about 4.5 times more than the difference. The reason for this is that when the injection angle of the conventional type DPI is small, the injector response is relatively slow and any special distinction between the pilot injection stage and the main injection stage is not clear. Therefore, it is expected to have a similar effect to that of injecting at the injection stroke. Conversely, Type 3 can exhibit low NO emission values with a relatively fast injection response and spray speed. This shows that HC and NO emissions are signiﬁcantly reduced in a two-stage injection scheme. The premixed combustion phase and the mixing controlled combustion phase were mostly affected by the injection rate, as stated in Section 3.2.3. It was observed that most of the fuel was injected rapidly following the injection signal at BTDC 15 CA, due to the characteristics of the DPI, a fast SOI, and a high injection rate. Therefore, the premixed combustion phase was promoted. Also, the rest of the remaining fuel was injected after the ignition. Therefore, the combustion was continuous, until the controlled combustion phase. As a result, the air-fuel mixture was saturated, and some fuel in the cylinder was not able to participate in combustion, remaining in an unburned condition. Later, the unburned fuel was used in the late combustion phase. Therefore, a multiple injection strategy could be advantageous in combustion performance assessments. In this sense, a Type 3 DPI speciﬁed as Case 3-8 (three plate springs + one thick pressure plate + one thin pressure plate) is a very useful measure for a FIE (fuel injection equipment) system in a newly clean CRDi diesel engine. 4. Conclusions In this study, a direct needle-driven piezo injector (DPI) based on the Delphi product was considered to improve injection characteristics, including: a maximum injection pressure and injection response, and manufacturing simplicity. The effects of the injection response and spray characteristics on CI combustion emissions were investigated with three concept designs, in order to ascertain the optimal DPI actuation structure for a clean CRDi diesel engine. First, a maximum injection pressure test and a Mie-scattering spray visualization with a high-speed camera were carried out, to compare the realizable criteria for the prototype DPIs. Second, injection rates based on Bosch tube method were applied for each prototype DPI and compared with a conventional type DPI. Finally, diesel engine experiments were carried out to investigate HC and NO emissions, and to investigate the engine adaptability and combustion performance of the Type 3 DPI prototype. This approach led to the following conclusions: Appl. Sci. 2017, 7, 320 15 of 16 (1) In internal design structure for a useful DPI needle response, it is very difﬁcult to apply a high injection pressure over 130 MPa if the hydraulic pressure required to excite the needle’s initial state is not applied. From the maximum injection pressure test, it was concluded that a high DPI injection pressure depends on the number of leaf springs. If the leaf spring is not present, the DPI can’t spray more than 180 MPa, due to a lack of elasticity. As the number of leaf springs increases, the elasticity increases. (2) Type 1 and 2 DPI prototypes were conceptually designed with non-hydraulic needles. Therefore, they were able to spray up to a maximum injection pressure of 130 MPa. Type 1, with an internal coil spring, had a high spring constant K, which a shortened needle distance of travel and a reduced injection value. Type 2 had the highest injection rate due to free needle movement. However, the Type 3 DPI, with an optimal number of pan-springs, plates, and needle structure for available hydraulics, had a maximum injection pressure of 180 MPa and advantageous degrees of freedom for controlling the injection value and injection rate with a simpler structure in the design process for a direct needle-driven piezo injector. (3) In the spray visualization experiment, the Type 3 DPI showed the fastest spraying speed and responsiveness at an average speed, the observed-maximum speed, and the maximum SOI response. As a result, the key factor inﬂuencing the spray characteristics is whether the DPI plate spring and the pressure plate of the needle are controlled by a mechanical mechanism only, or by hydraulic pressure. (4) For the injection rate, the Type 3 DPI had a faster injection rate, which corresponded to the visualization results. Furthermore, by looking at the slope of the injection rate at the beginning and end, the Type 3 DPI had a relatively horizontal slope, and fuel injection lasted the longest. (5) In diesel engine experiments, relatively high fuel injection rates for the Type 3 DPI prototype generated un-burned emission gases such as HC. However, it had a lower injection delay and faster injection start, with a much higher initial injection quantity. Therefore, it is expected to demonstrate a great advantage for multi-stage injections, where a rapid response and the spraying speed are subdivided. Acknowledgments: This research was supported by the CEFV (Center for Environmentally Friendly Vehicle) as Global-Top Project of KMOE (Ministry of Environment, Korea). Author Contributions: Sangik Han, Juhwan Kim and Jinwook Lee conceived and designed the experiments; Sangik Han and Juhwan Kim produced prototype DPI; Sangik Han performed the experiments; Sangik Han and Jinwook Lee analyzed the data and wrote the paper. Jinwook Lee supervised and advised all parts. Conﬂicts of Interest: The authors declare no conﬂict of interest. References 1. Jeong, Y.I.; Lee, J.W.; Cho, G.B.; Kim, H.S. Automotive and Environment; Publishing Department of Soongsil University: Seoul, Korea, 2010. 2. Ueda, D.; Tanada, H.; Utsunomiya, A.; Kawamura, J.; Weber, J. 4th Generation Diesel Piezo Injector (Realizing Enhanced High Response Injector); SAE Technical Paper 2016-01-0846; SAE International: Warrendale, PA, USA, 2016. 3. Arpaia, A.; Catania, A.; Ferrari, A.; Spessa, E. Development and Application of an Advanced Numerical Model for CR Piezo Indirect Acting Injection Systems; SAE Technical Paper 2010-01-1503; SAE International: Warrendale, PA, USA, 2010. 4. Husted, H.; Piock, W.; Ramsay, G. Fuel Efﬁciency Improvements from Lean, Stratiﬁed Combustion with a Solenoid Injector. SAE Int. J. Engines 2009, 2, 1359–1366. [CrossRef] 5. Kim, J.S. Effects of Injection Rate and Spray Characteristics of Piezo Injectors on Diesel Combustion. Master ’s Thesis, Soongsil University, Seoul, Korea, 2015. 6. Jo, I.; Jeong, M.; Kim, S.; Sung, G.; Lee, J. Experimental Investigation and Hydraulic Simulation of Dynamic Effects on Diesel Injection Characteristics in Indirect Acting Piezo-driven Injector with Bypass-Circuit System. Int. J. Autom. Technol. 2015, 16, 173–182. [CrossRef] Appl. Sci. 2017, 7, 320 16 of 16 7. Hardy, M.; Tolliday, A.; Delphi Technologies, Inc. Improvements Relating to Fuel Injector Control. Patent EP 2136062 A1, 2009. 8. Karimi, K. Characterisation of Multiple-Injection Diesel Sprays at Elevated Pressures and Temperatures. Ph.D. Thesis, University of Brighton, Brighton, UK, 2007. 9. Raul, P.; Jaime, G.; Michele, B.; Plazas, A.H. Study liquid length penetration results obtained with a direct acting piezo electric injector. Appl. Energy 2013, 106, 152–162. 10. Detlev, S.; Stefan, Z.; Martin, H.; Derk, G.; Rainer, W.; Nigel, B. Delphi common rail system with direct acting injector. MTZ Worldw. 2008, 69, 32–36. © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/). http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Applied Sciences Multidisciplinary Digital Publishing Institute http://www.deepdyve.com/lp/multidisciplinary-digital-publishing-institute/a-study-on-the-optimal-actuation-structure-design-of-a-direct-needle-WwCvSdaCBZ

Loading next page...

References (14)

(2009)
Delphi Technologies, Inc. Improvements Relating to Fuel Injector Control
(2009)
Improvements Relating to Fuel Injector Control
Anthony Arpaia, A. Catania, A. Ferrari, E. Spessa (2010)
Development and Application of an Advanced Numerical Model for CR Piezo Indirect Acting Injection Systems
(2010)
Development and Application of an Advanced Numerical Model for CR Piezo Indirect Acting Injection Systems; SAE Technical Paper 2010-01-1503
D. Ueda, H. Tanada, Atsushi Utsunomiya, J. Kawamura, J. Weber (2016)
4th Generation Diesel Piezo Injector (Realizing Enhanced High Response Injector)
R. Payri, J. Gimeno, Michele Bardi, A. Plazas (2013)
STUDY LIQUID LENGTH PENETRATION RESULTS OBTAINED WITH A DIRECT ACTING PIEZO ELECTRIC INJECTOR
Applied Energy, 106
K. Karimi (2007)
Characterisation of multiple-injection diesel sprays at elevated pressures and temperatures
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons
(2015)
Effects of Injection Rate and Spray Characteristics of Piezo Injectors on Diesel Combustion
Detlev Schöppe, S. Zülch, Martin Hardy, Derk Geurts, R. Jorach, Nigel Baker (2008)
Delphi Common Rail system with direct acting injector
MTZ worldwide, 69
(2010)
Automotive and Environment
I. Jo, M. Chung, Si-Jin Kim, G. Sung, Jin-Yeon Lee (2015)
Experimental investigation and hydraulic simulation of dynamic effects on diesel injection characteristics in indirect acting piezo-driven injector with bypass-circuit system
International Journal of Automotive Technology, 16
(2010)
Automotive and Environment; Publishing Department of Soongsil University
H. Husted, W. Piock, G. Ramsay (2009)
Fuel Efficiency Improvements from Lean, Stratified Combustion with a Solenoid Injector
SAE International journal of engines, 2

Publisher: Multidisciplinary Digital Publishing Institute
Copyright: © 1996-2019 MDPI (Basel, Switzerland) unless otherwise stated
ISSN: 2076-3417
DOI: 10.3390/app7040320
Publisher site: See Article on Publisher Site

Abstract

applied sciences Article A Study on the Optimal Actuation Structure Design of a Direct Needle-Driven Piezo Injector for a CRDi Engine 1 1 2 , Sangik Han , Juhwan Kim and Jinwook Lee * Department of Mechanical Engineering, Graduate School, Soongsil University, Seoul 06978, Korea; shdow88@naver.com (S.H.); jwan0502@gmail.com (J.K.) Department of Mechanical Engineering, Soongsil University, Seoul 06978, Korea * Correspondence: immanuel@ssu.ac.kr; Tel.: +82-2-820-0929 Academic Editor: Jose Ramon Serrano Received: 13 January 2017; Accepted: 14 March 2017; Published: 24 March 2017 Abstract: Recently, the high-pressure fuel injection performance of common-rail direct injection (CRDi) engines has become more important, due to the need to improve the multi-injection strategy. A multiple injection strategy provides better emission and fuel economy characteristics than a normal single injection scheme. The CRDi engine performance changes with the type of high-pressure electro-mechanical injector that is used and its injection response in a multi-injection scheme. In this study, a direct needle-driven piezo injector (DPI) was investigated, to optimize its actuation components, including the plate length, number of springs, and the elasticity of the spring between the injector needle and the piezo stack. Three prototype DPIs were proposed by this research. They were classiﬁed as Type 1, 2, and 3, depending on whether the injector needle was hydraulic or mechanical. Then, the optimal prototype was determined by conducting four evaluation experiments analyzing the maximum injection pressure, injection rate, spray visualization, and real engine combustion application. As a result, it was found that the Type 3 DPI prototype, with several pan-springs and plates, had the highest injection pressure, a steady injection rate, and the fastest spray speed. It also demonstrated the most effective emission reduction for a two-stage rapid spray injection in a single-cylinder CRDi engine. The Type 3 DPI displays an increased elasticity from its hydraulic needle that provides a synergy effect for improving DPI actuation. Keywords: direct needle-driven piezo injector; high-pressure injection pressure; injection rate; spray visualization; CRDi engine 1. Introduction The combustion process in an automotive engine is never perfect and small amounts of more harmful emissions are also produced in internal combustion engines. In order to reduce engine emissions, modern automotive engines carefully control the amount of fuel that they burn. The thermal efﬁciency of a combustion engine is deﬁned as the relationship between the total energy contained in the fuel, and the amount of energy used to perform useful work or generate kinetic energy. Carbon dioxide (CO ) emissions, as a product of combustion by bonding the carbon in the fuel with the oxygen in the air, are by far the largest amount of emissions produced by internal combustion engine vehicles. Until recently, they were thought to pose no immediate threat to the environment and the health of human beings. Therefore, CO has not been regulated in the same way as HC, CO, and NO . However, 2 x due to recent concerns about the increasing production of greenhouse gases and the increasing use of fossil fuels, regulators have begun attempts to limit the release of CO and other greenhouse gases into the atmosphere. The ICCT (International Council on Clean Transportation) [1] says that most Appl. Sci. 2017, 7, 320; doi:10.3390/app7040320 www.mdpi.com/journal/applsci Appl. Sci. 2017, 7, 320 2 of 16 developed countries have set CO emissions targets for new vehicles. Compared to the EU’s target of 95 g CO /km by 2020/2021, the US’s 97 g CO /km by 2025 for passenger cars, Canada’s 97 g 2 2 CO /km by 2025, China’s 117 g CO /km by 2020, and South Korea’s 97 g CO /km by 2020, illustrate 2 2 2 similar targets. Recently, combustion engine vehicles with a high fuel economy have also been included as eco-friendly vehicles that have less harmful impacts on the environment. In general, modern practical engines are always compromised by trade-offs between different properties, such as efﬁciency, weight, power, exhaust emissions, or NVH (noise, vibration, harshness). Sometimes, the economy not only plays a role in the engine manufacturing cost, but also in the manufacturing and distributing of the fuel. Increasing the engine’s efﬁciency produces a better fuel economy. Reduced exhaust emissions have been strengthened in passenger cars and light duty trucks after EURO 1 was initiated in 1992. Europe adopted EURO 6 in 2013, and it was applied in South Korea in 2014. Nitrogen oxides (NO ) of 80% and particulate matters (PM) of 60% are now more tightened than EURO 5. Some countermeasure technologies have been developed for satisfying this regulation. Among them, clean diesel vehicles and combustion engine-based hybrid electric vehicles have been studied, requiring a great deal of high-tech concentration. Speciﬁcally, a common-rail direct injection system is at the core of clean diesel technology. This makes it possible to pressurize above 200 MPa and then spray fuel into the combustion chamber through an electro-hydraulic injector. This presents accurate and precise fuel spraying using electronic control and is a great development for clean diesel engines. Typical solenoid-driven and piezo-driven fuel injectors can generate a favorable response and provide an effective injection in a common-rail direct injection system. These effects reduce exhaust emissions and improve the combustion performance in a clean diesel engine. Therefore, a rapid injector needle response and multiple injections are needed to obtain a high degree of freedom spray strategy, in order to improve diffusion combustion with a higher air to fuel (A/F) mixing ratio, reduce exhaust emissions, and increase engine performance. Research [2,3] has shown that ﬁve to seven multiple injections and rapid On/Off switching of the injector hole by the needle can freely control the shape of the injection rate. Daiji Ueda, etc., announced G4P (4th generation diesel piezo) injector concepts, structures, and hydraulic performances, using a rectangular injection rate and shorter multiple injection intervals. Also, they showed engine performance improvement results with G4P [2]. Arpaia, A., etc., reported that a numerical model for simulating a piezo indirect acting fuel injection system under a steady state and transient operating conditions was conducted through a commercial code [3]. Other papers [4], referring to the above result, showed that an injection mechanism with a high-pressure performance reduces not only harmful PM exhaust emissions, but also the auto-ignition combustion performance. However, the design of the piezo injector needs further improvement to increase the response of the injector needle. In this study, a direct needle-driven piezo injector (DPI) based on the Delphi product [5] was considered to improve injection characteristics with a maximum injection pressure, injection response, and manufacturing simplicity. For investigation, three steps were followed in this study: spray visualization based on Mie-scattering, an injection rate measurement, and a compression ignition (CI) combustion experiment based on a single-cylinder CRDi engine. DPI prototype design concepts focused on an internal DPI constitution, depending on whether or not a hydrodynamic needle was used. This study investigates the injection characteristics and combustion performance of three prototype DPIs, to obtain practical results. The effects of the internal design structure of the direct needle-driven piezo-injector and its optimal design parameters on the spray formation and CI combustion are discussed experimentally. 2. Experimental Apparatus and Procedure 2.1. Design Concept of Prototype DPIs By using a solenoid coil or a piezo stack as an actuator, diesel injectors have recently been designed in response to stringent emission regulations. A solenoid-driven injector, operated by peak and hold Appl. Sci. 2017, 7, 320 3 of 16 Appl. Sci. 2017, 7, 320 3 of 16 size. However, this type of injector is likely to exhibit an injection delay. Therefore, a number of current control, is still commonly used for its cost effectiveness, reliability, and smaller unit size. studies [6,7] have focused on the characteristics and injection flexibility of a piezo-driven injector. However, this type of injector is likely to exhibit an injection delay. Therefore, a number of studies [6,7] A DPI directly utilizes a longer piezo stack to lift the needle for injections. The longer piezo stack have focused on the characteristics and injection ﬂexibility of a piezo-driven injector. generates enough displacement for the lifting using an adapting amplifier located between the needle A DPI directly utilizes a longer piezo stack to lift the needle for injections. The longer piezo stack and the piezo stack. The elongation generated by the piezo stack is used to directly control the needle. generates enough displacement for the lifting using an adapting ampliﬁer located between the needle Appl. Sci. 2017, 7, 320 3 of 16 During DPI operation, the piezo stack shrinks when its electric energy is discharged. This causes the and the piezo stack. The elongation generated by the piezo stack is used to directly control the needle. size. However, this type of injector is likely to exhibit an injection delay. Therefore, a number of needle to move upward and reveal previously blocked injector holes. Then, high-pressure fuel is During DPI operation, the piezo stack shrinks when its electric energy is discharged. This causes studies [6,7] have focused on the characteristics and injection flexibility of a piezo-driven injector. injected through the injector holes. This allows the DPI to have a low fuel consumption by controlling the needle to move A DPI d upwar irectly utilizes a d and reveal longer pie previou zo stack to li sly blocked ft the needle injector for injectholes. ions. The Then, longer pie high-pr zo stackessur e fuel is various injection rate generates eno s. The ugh displace refore ment for , as sho the lifting wn in Fi using gur an e ad 1, the prototype DPI design concepts in this apting amplifier located between the needle injected through the injector holes. This allows the DPI to have a low fuel consumption by controlling and the piezo stack. The elongation generated by the piezo stack is used to directly control the needle. study focused on the internal DPI construction with three types. Type 1 uses a needle body, various injection rates. Therefore, as shown in Figure 1, the prototype DPI design concepts in this study During DPI operation, the piezo stack shrinks when its electric energy is discharged. This causes the a cylindrical shell, and a spring. This type is simple to make, produces a good response, and reduces needle to move upward and reveal previously blocked injector holes. Then, high-pressure fuel is focused on the internal DPI construction with three types. Type 1 uses a needle body, a cylindrical shell, the unit cost. Type 2 consists of pan-springs, plates, a needle body, a needle cylindrical shell, injected through the injector holes. This allows the DPI to have a low fuel consumption by controlling and a spring. This type is simple to make, produces a good response, and reduces the unit cost. Type 2 various injection rates. Therefore, as shown in Figure 1, the prototype DPI design concepts in this and a small spring. There is no cap to decrease the hydraulic influence and increase the needle consists of pan-springs, plates, a needle body, a needle cylindrical shell, and a small spring. There is study focused on the internal DPI construction with three types. Type 1 uses a needle body, response. Type 3 is constructed without limiting the number of pan-springs and plates. It is designed no cap to decr a cy ease lindric the al sh hydraulic ell, and a spring. inﬂuence This typeand is simincr ple to ease make, the produc needle es a good response. response, and r Type educe 3 is s constructed using a needle that has a needle body, a cylindrical shell, a spring, and cap. This design allows the unit cost. Type 2 consists of pan-springs, plates, a needle body, a needle cylindrical shell, without limiting the number of pan-springs and plates. It is designed using a needle that has a needle and a small spring. There is no cap to decrease the hydraulic influence and increase the needle hydraulics to be used. This Type 3 design is used to analyze and study the injector influence by body, a cylindrical shell, a spring, and cap. This design allows hydraulics to be used. This Type 3 response. Type 3 is constructed without limiting the number of pan-springs and plates. It is designed varying the number of plate springs and pressure plates. using a needle that has a needle body, a cylindrical shell, a spring, and cap. This design allows design is used to analyze and study the injector inﬂuence by varying the number of plate springs and hydraulics to be used. This Type 3 design is used to analyze and study the injector influence by pressure plates. varying the number of plate springs and pressure plates. + - Electrode A + - Electrode A F + - - - - Electrode B F Electrode B F Piez Pi oe e le z ct o rie c lectric ceram ics F ceram ics - + - + (a) (b) (a) (b) Figure 1. DPI proposed by this study; (a) Three prototype DPIs, (b) Principle of the inverse piezo- Figure 1. DPI proposed by this study; (a) Three prototype DPIs, (b) Principle of the inverse electric effect. Figure 1. DPI proposed by this study; (a) Three prototype DPIs, (b) Principle of the inverse piezo- piezo-electric effect. electric effect. 2.2. Experimental Apparatus and Methods 2.2.1. Maximum Injection Pressure Measurement 2.2. Experimental Apparatus and Methods 2.2. Experimental Apparatus and Methods A 1.5 kW class 3-phase motor (Hyosung Co., Seoul, Korea) was used as a high-pressure injection 2.2.1. Maximum Injection Pressure Measurement pump for a common-rail system. It can pressurize up to 200 MPa, in order to inject at the maximum 2.2.1. Maximum Injection Pressure Measurement possible injection pressure, as listed in Table 1. A temperature sensor is placed inside the fuel tank to A 1.5 kW class 3-phase motor (Hyosung Co., Seoul, Korea) was used as a high-pressure injection automatically turn on and off the radiator and a cooling fan is used to keep the temperature constant. A 1.5 kW class 3-phase motor (Hyosung Co., Seoul, Korea) was used as a high-pressure injection As shown in Figure 2, the injection period and the injection pressure were controlled through the pump for a common-rail system. It can pressurize up to 200 MPa, in order to inject at the maximum pump for a common-rail system. It can pressurize up to 200 MPa, in order to inject at the maximum common rail controller (Zenobalti Co., Model:ZB-9013, Daejeon, Korea) and the universal piezo possible injection pressure, as listed in Table 1. A temperature sensor is placed inside the fuel tank to possible injection pressure, as listed in Table 1. A temperature sensor is placed inside the fuel tank to injector driver (Zenobalti Co., Model:ZB-6200, Korea). automatically turn on and off the radiator and a cooling fan is used to keep the temperature constant. automatically turn on and off the radiator and a cooling fan is used to keep the temperature constant. As shown in Figure 2, the injection period and the injection pressure were controlled through the As shown in Figure 2, the injection period and the injection pressure were controlled through the common rail controller (Zenobalti Co., Model:ZB-9013, Daejeon, Korea) and the universal piezo injector common rail controller (Zenobalti Co., Model:ZB-9013, Daejeon, Korea) and the universal piezo driver (Zenobalti Co., Model:ZB-6200, Korea). injector driver (Zenobalti Co., Model:ZB-6200, Korea). (a) (b) Figure 2. Fuel injection control system; (a) Common-rail controller, (b) Injector driver. (a) (b) Figure 2. Fuel injection control system; (a) Common-rail controller, (b) Injector driver. Figure 2. Fuel injection control system; (a) Common-rail controller, (b) Injector driver. Appl. Sci. 2017, 7, 320 4 of 16 Appl. Sci. 2017, 7, 320 4 of 16 Appl. Sci. 2017, 7, 320 4 of 16 Table 1. Experimental conditions for maximum injection pressure measurement. Table 1. Experimental conditions for maximum injection pressure measurement. Table 1. Experimental conditions for maximum injection pressure measurement. Item Speciﬁcation Item Specification Item Fuel Spec ULSD(Ultra-Low-Sulfurification Diesel) Fuel ULSD(Ultra-Low-Sulfur Diesel) Injection duration (s) 1000 Fuel ULSD(Ultra-Low-Sulfur Diesel) Injection duration (µs) 1000 Injection pressure (MPa) 30~180 Injection duration (µs) 1000 Injection pressure (MPa) 30~180 Injection pressure (MPa) 30~180 2.2.2. Injection Rate Measurement 2.2.2. Injection Rate Measurement 2.2.2. Injection Rate Measurement There are two methods which can be used to measure the injection rate. The Zeuch method There are two methods which can be used to measure the injection rate. The Zeuch method obtains the injection rate after injecting fuel into the pressure vessel. With the Bosch-Tube method, fuel There are two methods which can be used to measure the injection rate. The Zeuch method obtains the injection rate after injecting fuel into the pressure vessel. With the Bosch-Tube method, is injected into the pipe to detect the pressure change in the injection rate meter. Then, the injection obtains the injection rate after injecting fuel into the pressure vessel. With the Bosch-Tube method, fuel is injected into the pipe to detect the pressure change in the injection rate meter. Then, the rate is measured. In this study, the Bosch-Tube method was used [5]. The Bosch-Tube measurement fuel is injected into the pipe to detect the pressure change in the injection rate meter. Then, the injection rate is measured. In this study, the Bosch-Tube method was used [5]. The Bosch-Tube principle involves feeding the fuel into the pipe with a cross-sectional area A and selecting the fuel injection rate is measured. In this study, the Bosch-Tube method was used [5]. The Bosch-Tube measurement principle involves feeding the fuel into the pipe with a cross-sectional area A and control volume in the pipe ﬂowing at the speed of sound as c, when the fuel in the pipe moves at measurement principle involves feeding the fuel into the pipe with a cross-sectional area A and selecting the fuel control volume in the pipe flowing at the speed of sound as c, when the fuel in the a speed of u. The characteristics of the fuel moving inside can be expressed as shown in Figure 3. selecting the fuel control volume in the pipe flowing at the speed of sound as c, when the fuel in the pipe moves at a speed of u. The characteristics of the fuel moving inside can be expressed as shown By knowing the sonic velocity and fuel density in the pipe and ascertaining the pressure change in the pipe moves at a speed of u. The characteristics of the fuel moving inside can be expressed as shown in Figure 3. By knowing the sonic velocity and fuel density in the pipe and ascertaining the pressure chamber, the injection rate can be calculated. in Figure 3. By knowing the sonic velocity and fuel density in the pipe and ascertaining the pressure change in the chamber, the injection rate can be calculated. change in the chamber, the injection rate can be calculated. Figure 3. Control volume in the pipe to measure the injection rate. Figure 3. Control volume in the pipe to measure the injection rate. Figure 3. Control volume in the pipe to measure the injection rate. (a) (b) (a) (b) Figure 4. Injection rate measurement; (a) Injection rate measuring device, (b) Schematic diagram. Figure 4. Injection rate measurement; (a) Injection rate measuring device, (b) Schematic diagram. Figure 4. Injection rate measurement; (a) Injection rate measuring device, (b) Schematic diagram. Table 2. Experimental conditions for measuring injection rate. Table 2. Experimental conditions for measuring injection rate. Table 2. Experimental conditions for measuring injection rate. Properties Specification Properties Specification Fuel ULSD Properties Speciﬁcation Fuel ULSD Tube length (m) 12.607 Fuel ULSD Tube length (m) 12.607 Tube area (m ) 1.6409×10 Tube length (m) 12.607 Tube area (m ) 1.6409×10 Injection pressure (MPa) 30~180 2 5 Tube area (m ) 1.6409 10 Injection pressure (MPa) 30~180 Duration (µs) 1000 Injection pressure (MPa) 30~180 Duration (µs) 1000 Duration (s) 1000 As shown in Figure 4, the injection rate measurement system consists of a measuring tube with As shown in Figure 4, the injection rate measurement system consists of a measuring tube with a length of about 12.6 m, needle valve, accumulator, pressure gauge, and regulator. The injector is a length of about 12.6 m, needle valve, accumulator, pressure gauge, and regulator. The injector is installed on the adapter in the injection rate measuring device and is connected to the common-rail installed on the adapter in the injection rate measuring device and is connected to the common-rail to supply high-pressure fuel. When fuel injection starts, the injection rate is measured by passing the to supply high-pressure fuel. When fuel injection starts, the injection rate is measured by passing the Appl. Sci. 2017, 7, 320 5 of 16 As shown in Figure 4, the injection rate measurement system consists of a measuring tube with a length of about 12.6 m, needle valve, accumulator, pressure gauge, and regulator. The injector is installed on the adapter in the injection rate measuring device and is connected to the common-rail to supply high-pressure fuel. When fuel injection starts, the injection rate is measured by passing the ﬂuid through the pressure sensor (Kistler, Model:6052c, Winterthur, Switzerland) that is mounted on Appl. Sci. 2017, 7, 320 5 of 16 the adapter. The experimental conditions for the injection rate are expressed in Table 2. fluid through the pressure sensor (Kistler, Model:6052c, Winterthur, Switzerland) that is mounted on 2.2.3. Spray Visualization Measurement the adapter. The experimental conditions for the injection rate are expressed in Table 2. Spray visualization experiments were carried out in order to directly conﬁrm the injection speed and behavior 2.2.3. Sprafor y Vieach sualizat prion Mea ototype. surem In this ent study, a static chamber with a volume of 885 cc was used, and images were taken with a high speed camera through a quartz glass window with a diameter Spray visualization experiments were carried out in order to directly confirm the injection speed of 108 mm and a thickness of 50 mm. This experiment was conducted at a normal temperature and behavior for each prototype. In this study, a static chamber with a volume of 885 cc was used, and pressure, and used a Bosch common-rail system. Since the prototype injector used in the and images were taken with a high speed camera through a quartz glass window with a diameter of experiment 108 mmwas and a a tpiezo-driven hickness of 50 m type, m. Thi it swas expecontr riment olled was conduct using aeuniversal d at a norm piezo al temp actuation erature andriver d pressure, and used a Bosch common-rail system. Since the prototype injector used in the experiment (Zenobalti Co., Model:ZB-6200, Korea). Detailed speciﬁcations of the high-speed camera (Vision was a piezo-driven type, it was controlled using a universal piezo actuation driver (Zenobalti Co., Research, Model:Phantom V7.3, Wayne, NJ, USA) used in this study are shown in Table 3. Direct Model:ZB-6200, Korea). Detailed specifications of the high-speed camera (Vision Research, lighting spray images were obtained using an 80 W LED light source. The experimental conditions Model:Phantom V7.3, Wayne, NJ, USA) used in this study are shown in Table 3. Direct lighting spray of the spray visualization experiment are shown in Table 4. Figure 5 shows a schematic and actual images were obtained using an 80 W LED light source. The experimental conditions of the spray picture of the injection test apparatus used in this experiment. visualization experiment are shown in Table 4. Figure 5 shows a schematic and actual picture of the injection test apparatus used in this experiment. Table 3. Speciﬁcations of the high-speed camera. Table 3. Specifications of the high-speed camera. Properties Speciﬁcation Properties Specification Model Phantom V7.3 Model Phantom V7.3 Max. resolution 800 600 Max. resolution 800 × 600 Sensor type 16 bit sensor Sensor type 16 bit sensor Max. PPS 500,000 Max. PPS 500,000 Trigger source TTL signal Trigger source TTL signal Memory 8 G DDR RAM Memory 8 G DDR RAM Table 4. Experimental conditions for spray visualization. Table 4. Experimental conditions for spray visualization. Properties Properties Spec Speciﬁcation ification Injection duration (µs) 1000 Injection duration (s) 1000 Injection Pressure (MPa) 120 Injection Pressure (MPa) 120 Resolution (pixels) 320 × 320 Resolution (pixels) 320 320 Sample Sample ra ratete (f (frame rame per per second) second) 25,000 25,000 Exposure (s) 40 Exposure (µs) 40 Ambient condition 1 atm. 300 K Ambient condition 1 atm. 300 K Spray Spray Co Cone ne An Angle gle Sp Spr ra ay y T Ti ip p P Pe en ne et tr ra at ti io on n 5m 5mm m (a) (b) Figure 5. Spray visualization measurement; (a) Visualization apparatus, (b) Definition of spray Figure 5. Spray visualization measurement; (a) Visualization apparatus, (b) Definition of spray parameter. parameter. Appl. Sci. 2017, 7, 320 6 of 16 Appl. Sci. 2017, 7, 320 6 of 16 2.2.4. Real Engine Combustion System 2.2.4. Real Engine Combustion System The engine used in this study is a direct-injection type single-cylinder CI engine (Daedong Co., The engine used in this study is a direct-injection type single-cylinder CI engine (Daedong Co., Model:ND10DE, Kyeongsan, Korea). Figure 6 shows the single-cylinder diesel engine system and its Model:ND10DE, Kyeongsan, Korea). Figure 6 shows the single-cylinder diesel engine system and its speciﬁcations are given in Table 5. Table 6 shows the combustion test conditions for the single-cylinder specifications are given in Table 5. Table 6 shows the combustion test conditions for the single- diesel engine. The emission performance of the most optimized DPI type and the conventional type cylinder diesel engine. The emission performance of the most optimized DPI type and the was analyzed by measuring the exhaust gas through an exhaust gas measurement system (Horiba, conventional type was analyzed by measuring the exhaust gas through an exhaust gas measurement Model:MEXA-554JKNO , Kyoto, Japan). The main speciﬁcations of the exhaust gas measurement system (Horiba, Model:MEXA-554JKNOx, Kyoto, Japan). The main specifications of the exhaust gas system are shown in Table 7. measurement system are shown in Table 7. Figure 6. Experiment system of single-cylinder diesel engine. Figure 6. Experiment system of single-cylinder diesel engine. Table 5. Main specifications of the single-cylinder diesel engine. Table 5. Main speciﬁcations of the single-cylinder diesel engine. Properties Specification Properties Speciﬁcation Engine type Horizontal DI diesel engine Bore Engine × Stroke type(mm) Horizontal 95DI ×diesel 95 engine Bore Stroke (mm) 95 95 Displacement (cc) 673 Displacement (cc) 673 Compression ratio 18.0 Compression ratio 18.0 Valve type SOHC (Single OverHead Camshaft) 2 valve Valve type SOHC (Single OverHead Camshaft) 2 valve Combustion chamber type Reentrant bowl-in-piston Combustion chamber type Reentrant bowl-in-piston Table 6. Engine combustion test conditions. Table 6. Engine combustion test conditions. Item Specification Item Speciﬁcation Engine speed (rpm) 1200 Engine speed (rpm) 1200 Injection quantity (mg/stroke) 3, 7 Injection quantity (mg/stroke) 3, 7 Injection timing (BTDC) 5, 15, 20 Injection timing (BTDC) 5, 15, 20 Table 7. Specifications of the exhaust gas measuring device. Table 7. Speciﬁcations of the exhaust gas measuring device. Item CO HC CO2 NOx O Item CO HC CO NO O Measuring rage 0.00~10.00% 0~10,000 ppm 0.00~20. 2 00% 0~5000 ppm x 0.00~25.00% Repeatability ± 0.06 ± 12 ± 0.5 ± 7 - Measuring rage 0.00~10.00% 0~10,000 ppm 0.00~20.00% 0~5000 ppm 0.00~25.00% Repeatability 0.06 12 0.5 7 - Measurement methods NDIR NDIR NDIR CLD CLD Measurement methods NDIR NDIR NDIR CLD CLD Responsibility up to 10 s, 90% response Responsibility up to 10 s, 90% response Appl. Sci. 2017, 7, 320 7 of 16 Appl. Sci. 2017, 7, 320 7 of 16 3. Results and Discussion 3. Results and Discussion 3.1. Design Complexity and Specialized Knowledge in Needed for Each DPI Prototype 3.1. Design Complexity and Specialized Knowledge in Needed for Each DPI Prototype 3.1.1. Type 1 DPI 3.1.1. Type 1 DPI In Type 1, the spring (70C hardened SW-C steel wire) was fabricated according to the wire In Type 1, the spring (70C hardened SW-C steel wire) was fabricated according to the wire thickness, outer diameter, number of spring revolutions, and length. It was inserted into the injector thickness, outer diameter, number of spring revolutions, and length. It was inserted into the injector inner-body, and then the needle body and needle cylinder were also inserted. Experimental results inner-body, and then the needle body and needle cylinder were also inserted. Experimental results showed that there is no problem in spraying up to the available injection pressure of 120 MPa. showed that there is no problem in spraying up to the available injection pressure of 120 MPa. However, However, at 130 MPa, normal injection becomes intermittent. For this reason, the injector normally at 130 MPa, normal injection becomes intermittent. For this reason, the injector normally injected only injected only 20 of the total injection signals. 20 of the total injection signals. As shown in Figure 7, each graph is divided into steel wire thicknesses of 0.7 mm and 0.8 mm. As shown in Figure 7, each graph is divided into steel wire thicknesses of 0.7 mm and 0.8 mm. It is It is shown that the production spring constant (K) increases toward the right of the X-axis, and the shown that the production spring constant (K) increases toward the right of the X-axis, and the actual actual number of injections continually increases. The reason for this is that the number of actual number of injections continually increases. The reason for this is that the number of actual injections is injections is expected to increase as the spring stiffness and K increases. The spring with high K expected to increase as the spring stiffness and K increases. The spring with high K becomes harder becomes harder than the lower K springs, such that the pressure inside the injector provides the force than the lower K springs, such that the pressure inside the injector provides the force to lift the injector to lift the injector needle. It was found that the best Type 1 performance was achieved by the spring needle. It was found that the best Type 1 performance was achieved by the spring with a steel wire with a steel wire thickness of 0.8 mm, outer diameter of 4.8 mm, five spring revolutions, and a length thickness of 0.8 mm, outer diameter of 4.8 mm, ﬁve spring revolutions, and a length of 6 mm. of 6 mm. (a) (b) Figure 7. Actual number of injections according to the thickness of the steel wire; (a) 0.7 mm, Figure 7. Actual number of injections according to the thickness of the steel wire; (a) 0.7 mm, (b) 0.8 mm. (b) 0.8 mm. 3.1.2. Type 2 DPI 3.1.2. Type 2 DPI As shown in Figure 8, the total number of leaf springs and pressure plates is speciﬁed as a As shown in Figure 8, the total number of leaf springs and pressure plates is specified as a design design parameter, and the needle part is assembled by inserting the needle body, cylinder, and parameter, and the needle part is assembled by inserting the needle body, cylinder, and spring. The spring. The number of leaf springs was reduced by one and three cases were analyzed. As shown in number of leaf springs was reduced by one and three cases were analyzed. As shown in Figure 9, the Figure 9, the Type 2 needle conﬁguration allows ﬂuid to move more freely than the conventional type. Type 2 needle configuration allows fluid to move more freely than the conventional type. When When hydraulic pressure is used in the conventional type, the needle moving distance is ampliﬁed hydraulic pressure is used in the conventional type, the needle moving distance is amplified and the and the needle can move with a small pressure force. However, the injection responsiveness is needle can move with a small pressure force. However, the injection responsiveness is considered to considered to be slow. Therefore, the design purpose of Type 2 is to increase the injection responsiveness be slow. Therefore, the design purpose of Type 2 is to increase the injection responsiveness by by allowing pressured ﬂuid to ﬂow freely without hydraulic pressure, by taking advantage of the allowing pressured fluid to flow freely without hydraulic pressure, by taking advantage of the mechanical characteristics. mechanical characteristics. Case 2-1 in Figure 8 has an available maximum injection pressure of 130 MPa, as shown in Case 2-1 in Figure 8 has an available maximum injection pressure of 130 MPa, as shown in Figure 10. Figure 10. The injection pressures of Case 2-2 and Case 2-3 were about 100 bar lower than Case 2-1. The injection pressures of Case 2-2 and Case 2-3 were about 100 bar lower than Case 2-1. It was found It was found that the available maximum injection pressure can be obtained, depending on the number that the available maximum injection pressure can be obtained, depending on the number of leaf of leaf springs. If one leaf spring in Type 2 is absent, the force for lifting the needle under CR high springs. If one leaf spring in Type 2 is absent, the force for lifting the needle under CR high pressure pressure is insufﬁcient. However, when two leaf springs are provided, it becomes possible to lift the is insufficient. However, when two leaf springs are provided, it becomes possible to lift the needle Appl. Sci. 2017, 7, 320 8 of 16 Appl. Appl. Sci. Sci. 2017 2017 , , 77 , 320 , 320 8 of 8 of 16 16 Appl. Sci. 2017, 7, 320 8 of 16 with a with a gre gre aa ter ter force, due to force, due to the incre the incre aa se se in the elasti in the elasti c c force. It is con force. It is con ss ider ider ed that t ed that t h h e elastic e elastic forc forc e in the e in the needle with a greater force, due to the increase in the elastic force. It is considered that the elastic force with a greater force, due to the increase in the elastic force. It is considered that the elastic force in the design of Type 2 is determined by the number of leaf springs. design of Type 2 is determined by the number of leaf springs. in design of Typ the design of e 2 is determined by Type 2 is determined the n byu the mber of number leaf of springs. leaf springs. Figure 8. Detailed designs of Type 2 DPIs. Figure 8. Detailed designs of Type 2 DPIs. Figure 8. Detailed designs of Type 2 DPIs. Figure 8. Detailed designs of Type 2 DPIs. (( aa) ( ) ( b b )) (a) (b) Figure 9. Figure 9. Com Com p p arison of nee arison of nee d d le le configu configu rr ation; ( ation; ( aa ) Type ) Type 2, ( 2, ( b b ) Convention ) Convention al Type DPI al Type DPI . . Figure 9. Comparison of needle configuration; (a) Type 2, (b) Conventional Type DPI. Figure 9. Comparison of needle conﬁguration; (a) Type 2, (b) Conventional Type DPI. Figure 10. Comparison of the maximum injection pressure of Type 2 DPIs. Figure 10. Comparison of the maximum injection pressure of Type 2 DPIs. Figure 10. Comparison of the maximum injection pressure of Type 2 DPIs. Figure 10. Comparison of the maximum injection pressure of Type 2 DPIs. 3. 3. 1. 1. 3. 3. T T y y p p ee 3 DPI 3 DPI 3.1.3. Type 3 DPI 3.1.3. Type 3 DPI A A ss s s h h own i own i n n F F ii gu gu rr ee 11 11 , the Ty , the Ty pe pe 3 3 i i n n te te rr n n aa ll s s tt rr u u cc tt ur ur ee wa wa ss a a n n aa ll y y zz ed by ed by ei ei g g h h t ca t ca ss ee ss wi wi th di th di ff ff ee rr ent ent As shown in Figure 11, the Type 3 internal structure was analyzed by eight cases with different As shown in Figure 11, the Type 3 internal structure was analyzed by eight cases with different combinat combinations ions of pr of press essuure re plat plates es and and l leeaf af sprin springgs,s, in includ cluding t ing thhee s saame no me nozzzzle le c coonfig nfigur urat ation ion as as aa combinations of pressure plates and leaf springs, including the same nozzle configuration as a combinations of pressure plates and leaf springs, including the same nozzle conﬁguration as a convent conventiional onal ttyype. In Type pe. In Type 3 3 DPIs DPIs, , it it wa was fo s found und t th hat at t t h he av e av ai aillaable ble maximum maximum inj injeect ction pres ion pressu sure w re w aass conventional type. In Type 3 DPIs, it was found that the available maximum injection pressure was conventional type. In Type 3 DPIs, it was found that the available maximum injection pressure 11 80 MPa i 80 MPa i n n all all ca ca ses, excep ses, excep tt Ca Ca ses ses 33 -- 3 a 3 a n n d 3-6 d 3-6 i i n n Fi Fi gure 12 gure 12 . In Ca . In Ca se se 33 -- 33 wi wi th th 80 80 0 0 ba ba r r aa n n d Case d Case 33 -- 66 , the , the 180 MPa in all cases, except Cases 3-3 and 3-6 in Figure 12. In Case 3-3 with 800 bar and Case 3-6, the was 180 MPa in all cases, except Cases 3-3 and 3-6 in Figure 12. In Case 3-3 with 800 bar and Case pr pr ee ss ss u u rr ee d fu d fu el el l l ee aa k k ee d d a a n n d the d the fu fu el el s s p p rr aa y y cc o o u u ll d not be d not be me me aa ss u u rr ee d d . It . It wa wa s thou s thou g g h h t tha t tha tt Ca Ca ss ee 3 3 -- 3 3 d d ii d not d not pressured fuel leaked and the fuel spray could not be measured. It was thought that Case 3-3 did not 3-6, the pressured fuel leaked and the fuel spray could not be measured. It was thought that Case inject inject m m o o re re t t h h an an 88 00 00 b b aa rr , ,d d u u e t e t o o t t h h ee ab ab se se nce of nce of a p a p ll at at e e sp sp ring ring for for am am p p lif lif ying ying t t h h e p e p ii ezo ezo st st ack ack for for cc e, e, as as inject more than 800 bar, due to the absence of a plate spring for amplifying the piezo stack force, as 3-3 did not inject more than 800 bar, due to the absence of a plate spring for amplifying the piezo seen seen in t in t h h e T e T y y pe 2 pe 2 DPI. DPI. U U n n like like Ca Ca se se 22 -- 3 o 3 o ff Type Type 2 2 , t , t h h e re e re ason ason for for not not inject inject ing ing under under 11 22 00 M M P P a a is is t t h h at at seen in the Type 2 DPI. Unlike Case 2-3 of Type 2, the reason for not injecting under 120 MPa is that stack force, as seen in the Type 2 DPI. Unlike Case 2-3 of Type 2, the reason for not injecting under there is not enough elast there is not enough elast ii cc ii ty to use the needle’s hy ty to use the needle’s hy dr dr aulic pre aulic pre ssur ssur ee . This in . This in dic dic aa tes that when there is tes that when there is there is not enough elasticity to use the needle’s hydraulic pressure. This indicates that when there is 120 MPa is that there is not enough elasticity to use the needle’s hydraulic pressure. This indicates that not a not a large large en en ough e ough e ll astic astic force, force, it is it is adv adv a a ntageo ntageo us us to to spray spray only w only w h h en using en using the the needle needle mechanically, mechanically, not a large enough elastic force, it is advantageous to spray only when using the needle mechanically, when there is not a large enough elastic force, it is advantageous to spray only when using the needle wi without a thout anny y hydrauli hydraulic pressure. In c pressure. In Case 3-6, the Case 3-6, the difference betwe difference betweeen n C Caase 3- se 3-6 a 6 annd d C Caase 3- se 3-4 i 4 iss tthhe e without any hydraulic pressure. In Case 3-6, the difference between Case 3-6 and Case 3-4 is the mechanically, without any hydraulic pressure. In Case 3-6, the difference between Case 3-6 and Case thickness o thickness o ff the pressure plate. The thic the pressure plate. The thic kness o kness o ff th th e thick pressure e thick pressure plate is 0.65 plate is 0.65 mm, and the thickness mm, and the thickness thickness of the pressure plate. The thickness of the thick pressure plate is 0.65 mm, and the thickness 3-4 is the thickness of the pressure plate. The thickness of the thick pressure plate is 0.65 mm, and the of of the thi the thinn pressure pla pressure platte is 0 e is 0.3 .3 mm. The tota mm. The totall leng length of th of the va the varia riabbllees, s, excludi excludinng the l g the leength ngth of of the the of the thin pressure plate is 0.3 mm. The total length of the variables, excluding the length of the thickness of the thin pressure plate is 0.3 mm. The total length of the variables, excluding the length of needle needle an an d p d p ii ezo ezo stack stack , ,is is 1.95 mm 1.95 mm for for C C aa se se 3-4 3-4 and and 1.6 mm 1.6 mm for for Case 3-6. Therefo Case 3-6. Therefo rr e, e, it w it w aa s fo s fo und that und that needle and piezo stack, is 1.95 mm for Case 3-4 and 1.6 mm for Case 3-6. Therefore, it was found that the minimum length inside the injector must exceed 1.6 mm. the minimum length inside the injector must exceed 1.6 mm. the minimum length inside the injector must exceed 1.6 mm. Appl. Sci. 2017, 7, 320 9 of 16 Appl. Sci. 2017, 7, 320 9 of 16 the needle and piezo stack, is 1.95 mm for Case 3-4 and 1.6 mm for Case 3-6. Therefore, it was found For Type 3 DPI Cases that have an injection performance of up to 180 MPa, spray images were that the Appl. minimum Sci. 2017, 7, 320 length inside the injector must exceed 1.6 mm. 9 of 16 acquired by a high-speed camera and the spray speed was compared, as shown in Figure 13. For Type 3 DPI Cases that have an injection performance of up to 180 MPa, spray images were As known, the spray speed was derived by the definition of the spray tip penetration and spray cone For Type 3 DPI Cases that have an injection performance of up to 180 MPa, spray images were acquired by a high-speed camera and the spray speed was compared, as shown in Figure 13. As known, angle, with the edge defined as a line of 80% transmittance of back-light in the raw spray images. acquired by a high-speed camera and the spray speed was compared, as shown in Figure 13. the spray speed was derived by the deﬁnition of the spray tip penetration and spray cone angle, with It was found that Case 3-8 had the best spray performance with an average speed of 107.3 m/s, an As known, the spray speed was derived by the definition of the spray tip penetration and spray cone the edge deﬁned as a line of 80% transmittance of back-light in the raw spray images. It was found that observed-maximum speed of 144 m/s, and an SOI (start of injection) value of 80 µs. The observed- angle, with the edge defined as a line of 80% transmittance of back-light in the raw spray images. Case 3-8 had the best spray performance with an average speed of 107.3 m/s, an observed-maximum maximum It was found that speed in this study mean Case 3-8 had the s that the spra best spray perfor y hmance with as not reache an aver d the real m age speed of 107.3 aximum speed b m/s, an e cause speed of 144 m/s, and an SOI (start of injection) value of 80 s. The observed-maximum speed in observed-maximum speed of 144 m/s, and an SOI (start of injection) value of 80 µs. The observed- the bore size of the test engine used in this study is 95 mm, as shown in Table 5, which is about the this study means that the spray has not reached the real maximum speed because the bore size of maximum speed in this study means that the spray has not reached the real maximum speed because same as the 100 mm diameter disk shown in Figure 5b. In other words, the high speed camera shoots the test the bore size engine used of the test engine in this studyused is 95in this study mm, as shown is 95 mm, in Table as shown i 5, which n Ta isble 5, whi about the ch is same about the as the 100 the fuel stream until it reaches the end of the disk and identifies the highest speed. When the fuel is same as the 100 mm diameter disk shown in Figure 5b. In other words, the high speed camera shoots mm diameter disk shown in Figure 5b. In other words, the high speed camera shoots the fuel stream injected into the cylinder, it is blocked by the inner wall of the cylinder and the maximum speed is the fuel stream until it reaches the end of the disk and identifies the highest speed. When the fuel is until it reaches the end of the disk and identiﬁes the highest speed. When the fuel is injected into the not fully achieved in the real spray chamber. injected into the cylinder, it is blocked by the inner wall of the cylinder and the maximum speed is cylinder, it is blocked by the inner wall of the cylinder and the maximum speed is not fully achieved in In some Type 3 DPIs, as the elasticity increases, the needle movement is amplified and its not fully achieved in the real spray chamber. the real spray chamber. moving speed increases. Therefore, it can be expected to have a good effect on two- or multi-injection In some Type 3 DPIs, as the elasticity increases, the needle movement is amplified and its In some Type 3 DPIs, as the elasticity increases, the needle movement is ampliﬁed and its moving schemes that are favorable for enhancing the air-fuel mixing rate of the diffusion combustion in the moving speed increases. Therefore, it can be expected to have a good effect on two- or multi-injection speed increases. Therefore, it can be expected to have a good effect on two- or multi-injection schemes CI engine. schemes that are favorable for enhancing the air-fuel mixing rate of the diffusion combustion in the that are favorable for enhancing the air-fuel mixing rate of the diffusion combustion in the CI engine. CI engine. Case 3-1 (leaf spring 2, thin plate 1, thick plate 1) Case 3-2 (leaf spring 1, thick plate 2) Case 3-1 (leaf spring 2, thin plate 1, thick plate 1) Case 3-2 (leaf spring 1, thick plate 2) Case 3-3 (thick plate 3) Case 3-4 (leaf spring 2, thick plate 1) Case 3-3 (thick plate 3) Case 3-4 (leaf spring 2, thick plate 1) Case 3-5 (leaf spring 2, thin plate 2) Case 3-6 (leaf spring 2, thin plate 1) Case 3-5 (leaf spring 2, thin plate 2) Case 3-6 (leaf spring 2, thin plate 1) Case 3-7 (leaf spring 2, thin plate 3) Case 3-8 (leaf spring 3, thin plate 1, thick plate 1) Case 3-7 (leaf spring 2, thin plate 3) Case 3-8 (leaf spring 3, thin plate 1, thick plate 1) Figure 11. Detailed designs of Type 3 DPIs. Figure Figure 11. 11. Detailed Detaileddesigns designs of of T Ty ype pe 3 DPI 3 DPIs.s. Figure 12. Comparison of the maximum injection pressures of Type 3 DPIs. Figure 12. Comparison of the maximum injection pressures of Type 3 DPIs. Figure 12. Comparison of the maximum injection pressures of Type 3 DPIs. Appl. Sci. 2017, 7, 320 10 of 16 Appl. Sci. 2017, 7, 320 10 of 16 Appl. Sci. 2017, 7, 320 10 of 16 Figure 13. Comparison of the spray speed results in Type 3 DPIs. Figure 13. Comparison of the spray speed results in Type 3 DPIs. Figure 13. Comparison of the spray speed results in Type 3 DPIs. 3.2. Comparison and Evaluation of Prototype DPIs 3.2. 3.2. C Comparison omparison an and d E Evaluation valuation of Pro of Prototype totype DP DPIs Is The following prototype DPIs were selected to compare the injection performance, spray, and The following prototype DPIs were selected to compare the injection performance, spray, and The following prototype DPIs were selected to compare the injection performance, spray, and combustion characteristics. (1) The Type 1 DPI has a 0.8 mm thick steel wire spring with an outer combustion combustion c characteristics. haracteristics. ((1) 1) Th The e Typ Type e 1 D 1 DPI PI h has as a a 0.8 0.8 m mm m t thick hick st steel eel wire wire sp spring ring w with ith an an outer outer diameter of 4.8 mm, spring rate of five, and a length of 6 mm; (2) For the Type 2 DPI, Case 2-1 diameter of 4.8 mm, spring rate of ﬁve, and a length of 6 mm; (2) For the Type 2 DPI, Case 2-1 (two leaf diameter of 4.8 mm, spring rate of five, and a length of 6 mm; (2) For the Type 2 DPI, Case 2-1 (two leaf springs + one thick pressure plate + one thin pressure plate) was selected; (3) For the Type springs (two leaf +springs one thick + one thic pressurk pressure p e plate + one late + one th thin pressur in pressure p e plate) was latselected; e) was selecte (3) For d; (3) For the the Type 3Type DPI, 3 DPI, Case 3-8 (three plate springs + one thick pressure plate + one thin pressure plate) was selected. Case 3-8 (three plate springs + one thick pressure plate + one thin pressure plate) was selected. 3 DPI, Case 3-8 (three plate springs + one thick pressure plate + one thin pressure plate) was selected. 3.2.1. Comparison of Maximum Injection Pressure Results 3.2.1. Comparison of Maximum Injection Pressure Results 3.2.1. Comparison of Maximum Injection Pressure Results The maximum injection pressure with the representative Types 1, 2, and 3, is shown in Figure 14. The maximum injection pressure with the representative Types 1, 2, and 3, is shown in Figure 14. The maximum injection pressure with the representative Types 1, 2, and 3, is shown in Figure 14. The Type 3 DPI showed the highest injection pressure at 180 MPa. It was found that the hydraulic The Type 3 DPI showed the highest injection pressure at 180 MPa. It was found that the hydraulic The Type 3 DPI showed the highest injection pressure at 180 MPa. It was found that the hydraulic application required to build up needle movement is an essential element to obtain an injection application required to build up needle movement is an essential element to obtain an injection application required to build up needle movement is an essential element to obtain an injection performance at high pressures, over 130 MPa. As a result, it can be said that the Type 3 DPI prototype performance at high pressures, over 130 MPa. As a result, it can be said that the Type 3 DPI prototype performance at high pressures, over 130 MPa. As a result, it can be said that the Type 3 DPI prototype has a similar injection performance to the conventional type DPI, and it has a better flexibility for has a similar injection performance to the conventional type DPI, and it has a better ﬂexibility for has a similar injection performance to the conventional type DPI, and it has a better flexibility for designing the internal structure of a direct needle-driven piezo-injector. designing the internal structure of a direct needle-driven piezo-injector. designing the internal structure of a direct needle-driven piezo-injector. Figure 14. Comparison of the maximum injection pressure for representative Type 1, 2 and 3 DPI. Figure 14. Comparison of the maximum injection pressure for representative Type 1, 2 and 3 DPI. Figure 14. Comparison of the maximum injection pressure for representative Type 1, 2 and 3 DPI. 3.2.2. Comparison of Mie-Scattering Spray Image Results 3.2.2. Comparison of Mie-Scattering Spray Image Results 3.2.2. Comparison of Mie-Scattering Spray Image Results In this study, LED back illumination for Mie-scattering was applied to make the initial liquid In this study, LED back illumination for Mie-scattering was applied to make the initial liquid spray In this study, LED back illumination for Mie-scattering was applied to make the initial liquid spray to full development spray visible. This enabled an accurate SOI determination and to full development spray visible. This enabled an accurate SOI determination and simultaneously spray to full development spray visible. This enabled an accurate SOI determination and simultaneously measured the initial liquid portion of the spray. Additionally, this Mie scattering measured the initial liquid portion of the spray. Additionally, this Mie scattering technique is suited to simultaneously measured the initial liquid portion of the spray. Additionally, this Mie scattering technique is suited to the spray characteristics in cold conditions, where there is insignificant the spray characteristics in cold conditions, where there is insigniﬁcant vaporization [8,9]. The spray technique is suited to the spray characteristics in cold conditions, where there is insignificant vaporization [8,9]. The spray image was taken with an exposure time of 38 µs at a resolution of 320 × image was taken with an exposure time of 38 s at a resolution of 320 320 pixels in 25,000 pps by vaporization [8,9]. The spray image was taken with an exposure time of 38 µs at a resolution of 320 × 320 pixels in 25,000 pps by a high-speed camera. a high-speed camera. 320 pixels in 25,000 pps by a high-speed camera. Figure 15 and Table 8 show the comparative results of the spray speeds with four different DPIs. Figure 15 and Table 8 show the comparative results of the spray speeds with four different DPIs. The Type 1 and 2 DPI had observed-maximum speeds of 91 m/s and 104 m/s, respectively, with an The Type 1 and 2 DPI had observed-maximum speeds of 91 m/s and 104 m/s, respectively, with an Appl. Sci. 2017, 7, 320 11 of 16 Appl. Sci. 2017, 7, 320 11 of 16 Figure 15 and Table 8 show the comparative results of the spray speeds with four different DPIs. The Type 1 and 2 DPI had observed-maximum speeds of 91 m/s and 104 m/s, respectively, with an average speed of 59 m/s and 67 m/s. In comparison to the conventional type, Types 1 and 2 have average speed of 59 m/s and 67 m/s. In comparison to the conventional type, Types 1 and 2 have slower spray speeds and a slow SOI response of 200 µs. It is known that the motion condition of a slower spray speeds and a slow SOI response of 200 s. It is known that the motion condition of a Appl. Sci. 2017, 7, 320 11 of 16 DPI needle is highly affected by the response parameter [9]. Generally, this shows that the use of DPI needle is highly affected by the response parameter [9]. Generally, this shows that the use of hydraulic pressure for needle movement enhances the response performance. In this study, Type 1 average speed of 59 m/s and 67 m/s. In comparison to the conventional type, Types 1 and 2 have hydraulic pressure for needle movement enhances the response performance. In this study, Type 1 and and 2 DPIs with the needle only being controlled by a mechanical mechanism, without the use of slower spray speeds and a slow SOI response of 200 µs. It is known that the motion condition of a 2 DPIs with the needle only being controlled by a mechanical mechanism, without the use of hydraulic hydraulic pre DPI needle ssur is highly e, at abo affect ut 80 µ ed by the response par s, would be slower ame tha ter [9]. Gene n the conv rally, this sh entional type i ows that the use of n terms of the SOI pressure, at about 80 s, would be slower than the conventional type in terms of the SOI response. hydraulic pressure for needle movement enhances the response performance. In this study, Type 1 response. In particular, by considering the internal actuation structure of Type 2 and the conventional In particular, by considering the internal actuation structure of Type 2 and the conventional type DPI, and 2 DPIs with the needle only being controlled by a mechanical mechanism, without the use of type DPI, it can be seen that the plate spring and the pressure plate are key factors influencing the it can be seen that the plate spring and the pressure plate are key factors inﬂuencing the determination hydraulic pressure, at about 80 µs, would be slower than the conventional type in terms of the SOI determination of the spray characteristics. of the spray characteristics. response. In particular, by considering the internal actuation structure of Type 2 and the conventional type DPI, it can be seen that the plate spring and the pressure plate are key factors influencing the determination of the spray characteristics. Figure 15. Comparison of spray speed with various D PIs. Figure 15. Comparison of spray speed with various DPIs. Figure 15. Comparison of spray speed with various DPIs. Table 8. Comparison of spray characteristics with various DPIs. Table 8. Comparison of spray characteristics with various DPIs. Table 8. Comparison of spray characteristics with various DPIs. Item Type1 Type2 Type3 Conventional Type Item Type 1 Type 2 Type 3 Conventional Type Item Type1 Type2 Type3 Conventional Type Max. speed (m/s) 91 104 113 103 Max. speed (m/s) 91 104 113 103 Max. speed (m/s) 91 104 113 103 Ave. speed (m/s) 59 67 80 69 Ave.Ave. speedsp(m/s)eed (m/s) 5959 67 67 80 80 69 69 SOI (µs) 200 200 80 120 SOI (s) 200 200 80 120 SOI (µs) 200 200 80 120 Conventional Type DPI Conventional Type DPI 80 μs 120 μs280 μs320 μs 80 μs 120 μs280 μs320 μs Type 3 DPI (as a representative prototype) Type 3 DPI (as a representative prototype) 80 μs 120 μs280 μs320 μs Figure 16. Comparison of Mie-scattering spray image between conventional type and Type 3 DPI. 80 μs 120 μs280 μs320 μs Figure 16. Comparison of Mie-scattering spray image between conventional type and Type 3 DPI. Figure 16. In the c Comparison of M ase of the Type 3 D ie-scat PI, the tering sp observed-ma ray image be ximum tween conventional ty spray speed is 113 m/s, the pe and aver Type 3 DPI. age speed is 80 m/s, and the SOI response is 80 µs. This value means that the Type 3 DPI has a fast needle In the case of the Type 3 DPI, the observed-maximum spray speed is 113 m/s, the average speed is 80 m/s, and the SOI response is 80 µs. This value means that the Type 3 DPI has a fast needle Appl. Sci. 2017, 7, 320 12 of 16 Appl. Sci. 2017, 7, 320 12 of 16 In the case of the Type 3 DPI, the observed-maximum spray speed is 113 m/s, the average movement, since about 40 µs would be faster than the conventional type DPI. In other words, the speed is 80 m/s, and the SOI response is 80 s. This value means that the Type 3 DPI has a fast Type 3 DPI needle is able to start fuel injection at 80 µs, but the conventional type DPI is still closed, needle movement, since about 40 s would be faster than the conventional type DPI. In other words, maintaining a needle state until the SOI timing is 120 µs. the Type 3 DPI needle is able to start fuel injection at 80 s, but the conventional type DPI is still closed, As shown in Figure 16, there is an obvious difference in the spray characteristics with time after maintaining a needle state until the SOI timing is 120 s. the start of the injection. As stated previously, the Type 3 DPI had a faster needle opening than the As shown in Figure 16, there is an obvious difference in the spray characteristics with time after other three DPIs, so it displayed a rapid spray behavior until obtaining a full development spray. It the start of the injection. As stated previously, the Type 3 DPI had a faster needle opening than the is clear that the actuation structure of a direct needle-driven injector significantly affects the initial other three DPIs, so it displayed a rapid spray behavior until obtaining a full development spray. spray shape. This is due to hydraulic internal flow dynamics within the DPI nozzle. Therefore, when It is clear that the actuation structure of a direct needle-driven injector signiﬁcantly affects the initial the needle response is relatively high, the spray tip fuel penetration increases in the liquid phase spray shape. This is due to hydraulic internal ﬂow dynamics within the DPI nozzle. Therefore, when spray characteristics. the needle response is relatively high, the spray tip fuel penetration increases in the liquid phase spray characteristics. 3.2.3. Comparison of Injection Rate Results 3.2.3. Comparison of Injection Rate Results Figures 17 and 18 compare the injection rate results between Types 1, 2, 3, and the conventional type DPI. To confirm the reliability of the injection rate meter, the injection rate obtained from the Figures 17 and 18 compare the injection rate results between Types 1, 2, 3, and the conventional meter used in this study was first validated with reference data from Delphi [10]. It was observed type DPI. To conﬁrm the reliability of the injection rate meter, the injection rate obtained from the that the injection quantity of the Type 1 DPI is 3.14 mg, which is relatively smaller than Type 2 or meter used in this study was ﬁrst validated with reference data from Delphi [10]. It was observed that Type 3, as shown in Table 9. Since the spring constant K is relatively higher than the leaf spring, the the injection quantity of the Type 1 DPI is 3.14 mg, which is relatively smaller than Type 2 or Type 3, length of the compressed spring is shortened when the needle is lifted, so the distance of travel of the as shown in Table 9. Since the spring constant K is relatively higher than the leaf spring, the length of needle in the Type 1 DPI is also shortened, and the injection quantity is expected to be reduced. In the compressed spring is shortened when the needle is lifted, so the distance of travel of the needle in the case of Type 2, the injection amount is high after the start of injection, but then falls sharply. the Type 1 DPI is also shortened, and the injection quantity is expected to be reduced. In the case of Compared to Type 3, the end of the injection (EOI) of the Type 2 DPI occurs early and pressured fuel Type 2, the injection amount is high after the start of injection, but then falls sharply. Compared to moves freely in the needle. So, the Type 2 DPI has a higher injection rate than other DPI types, until Type 3, the end of the injection (EOI) of the Type 2 DPI occurs early and pressured fuel moves freely in it reaches the EOI. This is a key cause of decreasing injection quantity due to the quick closing operation. the needle. So, the Type 2 DPI has a higher injection rate than other DPI types, until it reaches the EOI. This is a key cause of decreasing injection quantity due to the quick closing operation. Figure 17. Comparison of the injection rate with various DPIs. Figure 17. Comparison of the injection rate with various DPIs. Table 9. Comparison of the injection quantity with various DPIs. Table 9. Comparison of the injection quantity with various DPIs. Type 1 Type 2 Type 3 Type 1 Type 2 Type 3 Quantity (mg) 3.14 3.53 3.56 Quantity (mg) 3.14 3.53 3.56 Appl. Sci. 2017, 7, 320 13 of 16 Appl. Sci. 2017, 7, 320 13 of 16 Appl. Sci. 2017, 7, 320 13 of 16 Figure 18. Comparison of the injection rate between conventional type and Type 3 DPI. However, the Type 3 DPI showed a more stable and sufficient injection rate than Type 1 and 2 Figure 18. Comparison of the injection rate between conventional type and Type 3 DPI. DPIs. The shape of the injection rate of the Type 3 DPI is relatively horizontal and the fuel injection Figure 18. Comparison of the injection rate between conventional type and Type 3 DPI. lasts for the longest time. This means that the Type 3 DPI has a favorable degree of freedom for However, the Type 3 DPI showed a more stable and sufficient injection rate than Type 1 and 2 controlling the injection amount and the injection shape in real injection mapping processes. However, the Type 3 DPI showed a more stable and sufﬁcient injection rate than Type 1 and 2 DPIs. The shape of the injection rate of the Type 3 DPI is relatively horizontal and the fuel injection DPIs. The shape of the injection rate of the Type 3 DPI is relatively horizontal and the fuel injection lasts Figure 18 shows a comparison of the injection rates between the conventional type and the Type lasts for the longest time. This means that the Type 3 DPI has a favorable degree of freedom for for the longest time. This means that the Type 3 DPI has a favorable degree of freedom for controlling 3 DPI. As can be seen from this graph, the conventional type shows a shorter start of injection than controlling the injection amount and the injection shape in real injection mapping processes. the injection amount and the injection shape in real injection mapping processes. Type 3. However, as compared with the results of Table 10, the injection rate and injection quantity Figure 18 shows a comparison of the injection rates between the conventional type and the Type Figure 18 shows a comparison of the injection rates between the conventional type and the Type 3 do not show much deviation. This opinion is based on using a different observational method for the 3 DPI. As can be seen from this graph, the conventional type shows a shorter start of injection than DPI. As can be seen from this graph, the conventional type shows a shorter start of injection than injection rate measurement system and the spray visualization measurement system. Moreover, the Type 3. However, as compared with the results of Table 10, the injection rate and injection quantity Type 3. However, as compared with the results of Table 10, the injection rate and injection quantity do not show much deviation. This opinion is based on using a different observational method for the first “injection signature” marked with one circle in Figure 18 was detected for the Type 3 DPI. They do not show much deviation. This opinion is based on using a different observational method for injection rate measurement system and the spray visualization measurement system. Moreover, the show an almost equal injection performance, even if the Type 3 DPI has more simple structures in the injection rate measurement system and the spray visualization measurement system. Moreover, first “injection signature” marked with one circle in Figure 18 was detected for the Type 3 DPI. They the design processes of the direct needle-driven piezo injector. the ﬁrst “injection signature” marked with one circle in Figure 18 was detected for the Type 3 DPI. show an almost equal injection performance, even if the Type 3 DPI has more simple structures in They show an almost equal injection performance, even if the Type 3 DPI has more simple structures the design processes of the direct needle-driven piezo injector. Table 10. Comparison of injection quantity between conventional type and Type 3 DPI. in the design processes of the direct needle-driven piezo injector. Table 10. Comparison of injection quantity between conventional type and Type 3 DPI. Conventional Type Type 3 Table 10. Comparison of injection quantity between conventional type and Type 3 DPI. Quantity (mg) 3.51 3.56 Conventional Type Type 3 Conventional Type Type 3 Quantity (mg) 3.51 3.56 Quantity (mg) 3.51 3.56 (a) (b) (a) (b) Figure 19. Comparison of HC and NOx emissions between conventional type and Type 3 DPI; Figure 19. Comparison of HC and NOx emissions between conventional type and Type 3 DPI; Figure 19. Comparison of HC and NO emissions between conventional type and Type 3 DPI; (a) 7 mg/stroke (injection quantity) and BTDC 10 CA (injection timing); (b) 3, 7 mg/stroke (pilot/main (a) 7 mg/stroke (injection quantity) and BTDC 10 CA (injection timing); (b) 3, 7 mg/stroke (pilot/main (a) 7 mg/stroke (injection quantity) and BTDC 10 CA (injection timing); (b) 3, 7 mg/stroke (pilot/main injection quantity) and BTDC 15, 5 CA (pilot/main injection timing). injection quantity) and BTDC 15, 5 CA (pilot/main injection timing). injection quantity) and BTDC 15, 5 CA (pilot/main injection timing). The effect of the injection rate on combustion is discussed in Section 3.2.4. It predicts that the The effect of the injection rate on combustion is discussed in Section 3.2.4. It predicts that the Type 2 DPI has a higher NOx emission due to its higher injection quantity at the beginning of the Type 2 DPI has a higher NOx emission due to its higher injection quantity at the beginning of the combustion. It can be inferred that the Type 3 DPI might have the advantage of multiple injections, combustion. It can be inferred that the Type 3 DPI might have the advantage of multiple injections, Appl. Sci. 2017, 7, 320 14 of 16 The effect of the injection rate on combustion is discussed in Section 3.2.4. It predicts that the Type 2 DPI has a higher NO emission due to its higher injection quantity at the beginning of the combustion. It can be inferred that the Type 3 DPI might have the advantage of multiple injections, because the test results indicated that it had a lower injection delay and faster injection start, with a higher initial injection quantity. Therefore, a pilot injection strategy was applied and described, as shown in Figure 19b. 3.2.4. Comparison of Engine HC and NO Emission Results Figure 19 shows the diesel engine experimental results using the same conditions applied in previous chapters. The speed of the engine was maintained at 1200 rpm, and the injection timing was set to BTDC 10 CA in a single injection case and BTDC 15 CA and 5 CA in a two-injection stage (pilot/main) injection case. The fuel injected was 3 mg and 7 mg with the pilot and main injections, respectively. This was done to maintain the total fuel mass with the single injection experiment. By ﬁxing the pilot injection timing at BTDC 15 CA, it was observed that the combustion characteristics and performance were inﬂuenced by the injection timing of the main injection. First, it was observed that HC and NO emissions for the conventional type and the Type 3 DPI were different, regardless of single and two-injection stages. As shown in Figure 19a, in the case of the Type 3 DPI at the injection quantity of 7 mg/stroke, the HC emission increased by about 8 ppm, but the NO value decreased by about 23 ppm, which is more than 50% lower than the conventional type DPI. However when the injection timing of BTDC 10 CA differs for the two-injection stage, as shown in Figure 19b. The NO emission value is about 109 ppm, which is about 4.5 times more than the difference. The reason for this is that when the injection angle of the conventional type DPI is small, the injector response is relatively slow and any special distinction between the pilot injection stage and the main injection stage is not clear. Therefore, it is expected to have a similar effect to that of injecting at the injection stroke. Conversely, Type 3 can exhibit low NO emission values with a relatively fast injection response and spray speed. This shows that HC and NO emissions are signiﬁcantly reduced in a two-stage injection scheme. The premixed combustion phase and the mixing controlled combustion phase were mostly affected by the injection rate, as stated in Section 3.2.3. It was observed that most of the fuel was injected rapidly following the injection signal at BTDC 15 CA, due to the characteristics of the DPI, a fast SOI, and a high injection rate. Therefore, the premixed combustion phase was promoted. Also, the rest of the remaining fuel was injected after the ignition. Therefore, the combustion was continuous, until the controlled combustion phase. As a result, the air-fuel mixture was saturated, and some fuel in the cylinder was not able to participate in combustion, remaining in an unburned condition. Later, the unburned fuel was used in the late combustion phase. Therefore, a multiple injection strategy could be advantageous in combustion performance assessments. In this sense, a Type 3 DPI speciﬁed as Case 3-8 (three plate springs + one thick pressure plate + one thin pressure plate) is a very useful measure for a FIE (fuel injection equipment) system in a newly clean CRDi diesel engine. 4. Conclusions In this study, a direct needle-driven piezo injector (DPI) based on the Delphi product was considered to improve injection characteristics, including: a maximum injection pressure and injection response, and manufacturing simplicity. The effects of the injection response and spray characteristics on CI combustion emissions were investigated with three concept designs, in order to ascertain the optimal DPI actuation structure for a clean CRDi diesel engine. First, a maximum injection pressure test and a Mie-scattering spray visualization with a high-speed camera were carried out, to compare the realizable criteria for the prototype DPIs. Second, injection rates based on Bosch tube method were applied for each prototype DPI and compared with a conventional type DPI. Finally, diesel engine experiments were carried out to investigate HC and NO emissions, and to investigate the engine adaptability and combustion performance of the Type 3 DPI prototype. This approach led to the following conclusions: Appl. Sci. 2017, 7, 320 15 of 16 (1) In internal design structure for a useful DPI needle response, it is very difﬁcult to apply a high injection pressure over 130 MPa if the hydraulic pressure required to excite the needle’s initial state is not applied. From the maximum injection pressure test, it was concluded that a high DPI injection pressure depends on the number of leaf springs. If the leaf spring is not present, the DPI can’t spray more than 180 MPa, due to a lack of elasticity. As the number of leaf springs increases, the elasticity increases. (2) Type 1 and 2 DPI prototypes were conceptually designed with non-hydraulic needles. Therefore, they were able to spray up to a maximum injection pressure of 130 MPa. Type 1, with an internal coil spring, had a high spring constant K, which a shortened needle distance of travel and a reduced injection value. Type 2 had the highest injection rate due to free needle movement. However, the Type 3 DPI, with an optimal number of pan-springs, plates, and needle structure for available hydraulics, had a maximum injection pressure of 180 MPa and advantageous degrees of freedom for controlling the injection value and injection rate with a simpler structure in the design process for a direct needle-driven piezo injector. (3) In the spray visualization experiment, the Type 3 DPI showed the fastest spraying speed and responsiveness at an average speed, the observed-maximum speed, and the maximum SOI response. As a result, the key factor inﬂuencing the spray characteristics is whether the DPI plate spring and the pressure plate of the needle are controlled by a mechanical mechanism only, or by hydraulic pressure. (4) For the injection rate, the Type 3 DPI had a faster injection rate, which corresponded to the visualization results. Furthermore, by looking at the slope of the injection rate at the beginning and end, the Type 3 DPI had a relatively horizontal slope, and fuel injection lasted the longest. (5) In diesel engine experiments, relatively high fuel injection rates for the Type 3 DPI prototype generated un-burned emission gases such as HC. However, it had a lower injection delay and faster injection start, with a much higher initial injection quantity. Therefore, it is expected to demonstrate a great advantage for multi-stage injections, where a rapid response and the spraying speed are subdivided. Acknowledgments: This research was supported by the CEFV (Center for Environmentally Friendly Vehicle) as Global-Top Project of KMOE (Ministry of Environment, Korea). Author Contributions: Sangik Han, Juhwan Kim and Jinwook Lee conceived and designed the experiments; Sangik Han and Juhwan Kim produced prototype DPI; Sangik Han performed the experiments; Sangik Han and Jinwook Lee analyzed the data and wrote the paper. Jinwook Lee supervised and advised all parts. Conﬂicts of Interest: The authors declare no conﬂict of interest. References 1. Jeong, Y.I.; Lee, J.W.; Cho, G.B.; Kim, H.S. Automotive and Environment; Publishing Department of Soongsil University: Seoul, Korea, 2010. 2. Ueda, D.; Tanada, H.; Utsunomiya, A.; Kawamura, J.; Weber, J. 4th Generation Diesel Piezo Injector (Realizing Enhanced High Response Injector); SAE Technical Paper 2016-01-0846; SAE International: Warrendale, PA, USA, 2016. 3. Arpaia, A.; Catania, A.; Ferrari, A.; Spessa, E. Development and Application of an Advanced Numerical Model for CR Piezo Indirect Acting Injection Systems; SAE Technical Paper 2010-01-1503; SAE International: Warrendale, PA, USA, 2010. 4. Husted, H.; Piock, W.; Ramsay, G. Fuel Efﬁciency Improvements from Lean, Stratiﬁed Combustion with a Solenoid Injector. SAE Int. J. Engines 2009, 2, 1359–1366. [CrossRef] 5. Kim, J.S. Effects of Injection Rate and Spray Characteristics of Piezo Injectors on Diesel Combustion. Master ’s Thesis, Soongsil University, Seoul, Korea, 2015. 6. Jo, I.; Jeong, M.; Kim, S.; Sung, G.; Lee, J. Experimental Investigation and Hydraulic Simulation of Dynamic Effects on Diesel Injection Characteristics in Indirect Acting Piezo-driven Injector with Bypass-Circuit System. Int. J. Autom. Technol. 2015, 16, 173–182. [CrossRef] Appl. Sci. 2017, 7, 320 16 of 16 7. Hardy, M.; Tolliday, A.; Delphi Technologies, Inc. Improvements Relating to Fuel Injector Control. Patent EP 2136062 A1, 2009. 8. Karimi, K. Characterisation of Multiple-Injection Diesel Sprays at Elevated Pressures and Temperatures. Ph.D. Thesis, University of Brighton, Brighton, UK, 2007. 9. Raul, P.; Jaime, G.; Michele, B.; Plazas, A.H. Study liquid length penetration results obtained with a direct acting piezo electric injector. Appl. Energy 2013, 106, 152–162. 10. Detlev, S.; Stefan, Z.; Martin, H.; Derk, G.; Rainer, W.; Nigel, B. Delphi common rail system with direct acting injector. MTZ Worldw. 2008, 69, 32–36. © 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Journal

Applied Sciences – Multidisciplinary Digital Publishing Institute

Published: Mar 24, 2017

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A Study on the Optimal Actuation Structure Design of a Direct Needle-Driven Piezo Injector for a CRDi Engine

A Study on the Optimal Actuation Structure Design of a Direct Needle-Driven Piezo Injector for a CRDi Engine

Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

A Study on the Optimal Actuation Structure Design of a Direct Needle-Driven Piezo Injector for a CRDi Engine

A Study on the Optimal Actuation Structure Design of a Direct Needle-Driven Piezo Injector for a CRDi Engine

References (14)

Abstract

Journal

Recommended Articles

There are no references for this article.

Our policy towards the use of cookies