- Email: [email protected]
An experimental investigation into the starting process of free-piston engine generator
Applied Energy xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Applied Energy journal homepage: www.elsevier.com/locate/apenergy
An experimental investigation into the starting process of free-piston engine generator q Boru Jia a,b, Guohong Tian b, Huihua Feng a,⇑, Zhengxing Zuo a, A.P. Roskilly b a b
School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China Sir Joseph Swan Centre for Energy Research, Newcastle University, Newcastle upon Tyne NE1 7RU, UK
h i g h l i g h t s A prototype free-piston engine generator is developed and tested. The in-cylinder mixture was ignited successfully with the proposed control strategy. Two practical input parameters are selected for the future motor switching control. Three possible reasons of misfire are analysed.
a r t i c l e
i n f o
Article history: Received 2 September 2014 Received in revised form 3 February 2015 Accepted 18 February 2015 Available online xxxx Keywords: Free-piston generator Linear electric machine Experiment Starting process Combustion
a b s t r a c t This paper presents an experimental investigation of the starting process of a prototype free piston engine generator (FPEG). Experimental test results show that during the motoring stage, the peak in-cylinder pressure and compression ratio increase in a non-linear manner and trend to reach a stable state after a number of cycles. The motoring force is suggested to be within a reasonable range. With a fixed starting force of 125 N, the in-cylinder air fuel mixture was successfully ignited at the fourth cycle with a compression ratio of over 9:1. The peak in-cylinder pressure for the first combustion cycle reached over 40 bar. The piston ran at high and relatively constant speed at the middle portion of the stroke. The peak piston velocity increases significantly to around 4.0 m/s. Cycle-to-cycle variation of the piston movement was significant and the engine misfired frequently. During the misfire cycles, the peak piston velocity decreased to nearly 2.5 m/s; and the piston dynamics were similar to the motoring process. Based on these, discussion on misfire and further stable running control, as well as the linear electric machine mode switch were presented. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Free-piston engine generator (FPEG) is a novel type of energy conversion device that integrates a linear combustion engine and a linear electric machine [1]. Combustion in the chamber drives the piston reciprocates in a nearly resonant way and the linear electric machine converts part of the mover’s kinetic energy to electrical energy. The effective efficiency is estimated to be up to 46% (including friction and compressor losses) at a power level of 23 kW and shows promising results with respect to engine performance and emissions [2]. Since the FPEG was first proposed, despite the interests it has attracted from all over the world
q This paper is included in the Special Issue of Clean Transport edited by Prof. Anthony Roskilly, Dr. Roberto Palacin and Prof. Yan. ⇑ Corresponding author. Tel.: +86 10 68911062. E-mail address: [email protected] (H. Feng).
[3–6], there has not been any stably operating prototype reported by now. The FPEs were first proposed around 1930, and during 1930 to 1960 they were mainly used as air compressors and gas generators as they provided advantages over conventional combustion engines and gas turbines at that time [7]. In recent years, modern applications of the free-piston concept have been proposed for the generation of electric and hydraulic power, typically in hybrid electric vehicles [8–9]. Different prototype designs have been reported using the FPEG concept. The majority of these were, however, not commercially successful. This section gives an overview of known FPEG development, with an emphasis on reports where prototype performance data have been reported. Researchers at West Virginia University described the development of a spark-ignited (SI) dual piston FPEG [10]. This 36.5 mm bore size prototype was reported to have achieved 316 W power output at 23.1 Hz, with a 50 mm maximum stroke. High
http://dx.doi.org/10.1016/j.apenergy.2015.02.065 0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
2
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
cycle-to-cycle variations were observed particularly at low loads [11], which indicated an unstable operation. A two stroke compression ignition (CI) version linear engine prototype was also developed [11]. This CI engine had a similar mechanical arrangement to the SI prototype. The linear alternator interposed with the linear engine also operated as a starting device. In order to aid the cold start of the engine, each cylinder was equipped with a glow plug [12]. The research team in Sandia National Laboratories presented the design of a dual cylinder FPEG. The engine employed a homogeneous charge compression ignition (HCCI) and was aimed to operate on a variety of hydrogen-containing fuels. Test results from a compression–expansion machine showed nearly constant volume combustion with hydrogen, bio-gas, and ammonia at the equivalence ratio of approximately 0.3. The target efficiency was 50% overall considering 56% engine thermal efficiency and 96% generator efficiency [13]. In 2008, the dual cylinder free-piston engine configuration was changed into an opposed piston type, which was adopted to utilize the self-balance effect [14]. A European Commission-funded Free Piston Energy Converter (FPEC) project researched the subject of FPEG aimed at the development of an efficient new technology suitable for vehicle propulsion, auxiliary power units and distributed power generation since 2002 [2]. The prototype ran on diesel fuel, and was used primarily for testing in a test cell for validation of the specific FPEG issues. The converter was equipped with fuel injectors, pneumatic operated valves, cylinder pressure sensors, and translator displacement sensors. The combustion system is a Direct Injection, 2-stroke, two cylinder systems with electrical controlled valves and scavenging ports operating in HCCI mode on diesel fuel [15]. However, there has not been many test data reported by now. The approach by the Beijing Institute of Technology used a loop scavenged, carburetted free-piston, double-ended cylinder arrangement with a linear alternator; the bore of the engine was 34 mm, and the effective stroke was 20 mm [16]. Experiments results showed that the engine misfired every one or two cycles [16]. For the compression ignition FPEG prototype developed by the same group, the cylinder bore and stroke were enlarged to 62 mm and 86 mm respectively. If a compression ratio of 15 was required at the end of the compression stroke to enable selfignition, the corresponding in-cylinder pressure force was over 11 kN, which made it practically impossible to start the engine at one compression stroke without an largely oversized electric linear machine [17]. Pempek Systems Pty. Ltd., an Australian company, was one of the research leaders in this area. Their conceptual free piston engine generator design aims the power source of a high performance hybrid vehicle which has a top speed of 160 km/h and 0–100 km/ h acceleration of only 5.4 s. This vehicle will also be equipped with a brake recovery system. They conceptually designed a 25 kW free piston engine generator which has 50% engine thermal efficiency and over 93% of the generator efficiency [18]. The German Aerospace Centre (DLR) developed an FPEG prototype [19–22]. It consisted of three main subsystems: an internal combustion engine (ICE) converting chemical energy into kinetic energy, a linear generator (LG) converting kinetic energy into electric energy and a gas spring (GS) storing energy and inverting the piston movement. At 21 Hz, a power output of roughly 10 kW had been measured. Increasing the frequency up to 50 Hz should lead to a power output of 25 kW of a single piston FPEG system [19,20]. The FPEG prototype developed by Toyota Central R&D Labs Inc. was thin and compact [23,24]. The developed FPEG consisted of a two-stroke combustion chamber, a linear generator, and a gas spring chamber. A power generation experiment was carried out, and the results demonstrated that the prototype operated stably
for a long period of time, despite of the abnormal combustion during the test [24]. The unique piston motion and its effect on combustion and power generation in the FPEG prototype were experimentally analysed [23]. FPEG can be divided into three categories according to piston/cylinder configuration: single piston, dual piston and opposed piston [25]. The basic operation principles are equal for each concept; differences between the concepts are the number of combustion chambers and compression stroke realization [25]. There has been successful implementation of the single piston type, coupled with a gas spring chamber [20,24]. However, despite the problems being reported, the dual piston configuration remains the most popular layout due to the following advantages over single piston and opposed piston configurations: 1. The only moving part is a linear magnet mover coupled with pistons at each end and placed between two opposing combustion chambers. This allows a simple and more compact device with higher power to weight ratio. 2. It eliminates the need for a rebound device, as the combustion force drives the piston assembly to overcome the compression pressure in the other cylinder. Because of the potential advantages above, the dual piston type FPEG is adopted in the prototype reported in this paper. A newly designed FPEG prototype has been built to validate the feasibility of the technical scheme of dual piston FPEG. This paper describes the configuration of the newly designed system as well as the fundamental test results. Based on these, discussion on misfire and further stable running control, as well as the linear electric machine mode switch were presented. 2. FPEG prototype 2.1. Prototype specification The FPEG prototype is demonstrated in Fig. 1. This prototype is a dual piston, two-stroke, spark-ignited, uniflow scavenging engine. It is integrated with a dedicated bench of steel plates, constructed to reduce the vibrations during testing. The specifications of the prototype are summarised in Table 1. More details about the prototype have been presented in our previous paper [26,27], and the carburettor has been replaced by a port fuel injection (PFI) system for a more precise air fuel ratio control. The PFI system with integrated electric fuel pump and fuel filter conveys the required amount of fuel from the tank to the injectors at a constant pressure. The fuel is injected into the intake manifold
Fig. 1. The prototype of free-piston engine generator.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx Table 1 Prototype specifications. Parameters
Value
Number of cylinder Stroke Combustion mode Bore Maximum total stroke Maximum effective stroke (compression stroke) Moving mass
2 2 SI 52.5 mm 70.0 mm 35.0 mm 5.0 kg
3
( X0) displacement from centre and falling edge is at X1 ( X1), as illustrated in Fig. 3. These two signals were used to trigger both spark ignition and fuel injection. The function is summarised in Table 2. The injection is triggered by the falling edge, and the duration of the injection is determined by the injected fuel quantity after calibration to deliver a stoichiometric fuel/air mixture. For the test carried out at this stage, the ignition timing and the injection duration were fixed and ignition timing has not been optimised. 2.3. Test facilities
to form a stoichiometric air/fuel mixture. The ignition system mainly consisted of a 12 V battery, ignition coil, and spark plug, and they are all off-the-shelf products. The ignition and injection systems are activated simultaneously when the engine is motored to achieve a good compression ratio.
2.2. Control strategy The general working process of the FPEG consists of three stages, i.e. the starting process, linear electric machine mode switching process, and generating process. The starting process discussed by this paper can be primarily divided into two stages. In stage one, the linear electric machine runs as a motor that provides motoring force to start the engine by mechanical resonance (motoring process). When piston speed and compression ratio achieve the desired values, fuel is delivered and ignited which is regarded as the second stage (ignition process). From the control point of view, each process is supposed to be coupled with proper control strategies to ensure stable running. Then the control system of the prototype can be divided by four parts, which are the starting system, ignition control system, linear electric machine mode control system, and the external load control system. All these systems are integrated in an electronic controller, which will generate the control signals for each sub system according to the feedback from the sensors. The structure of the control system is illustrated in Fig. 2. This paper focuses on the control strategy of the motoring process and the engine ignition control system, and presents suggestions for the further electric machine mode switching control as well as the stable running control for the generating process according to the fundamental test results. During motoring process, the linear electric machine is operated as a motor and generates constant motor force in the direction of the piston velocity. The displacement amplitude as well as the peak in-cylinder gas pressure is expected to grow and finally reach the required values for ignition. Detailed investigation of the starting strategy and implementation can be found in our previous paper [26]. The control system for the motoring process was coupled with PID compensators to reduce the influence from back electromagnetic voltage. It took the mover’s displacement and velocity, as well as the actual current in the coil as feedbacks. As a result, the electric machine was operated as closed-loop control system and generated constant motor force in direction of the mover’s velocity [26]. At the end of the motoring process, the ignition system and the PFI system are both activated simultaneously. Ignition is usually triggered at the end of the compression stroke, while injection starts at the end of the expansion stroke. The ignition and injection signals are both triggered using piston position signal as reference. This signal is provided by a linear encoded integrated in the linear electric machine and can be adjusted. The triggers are two TTL signals symmetric to the central position. The rising edge is at X0
The data of piston displacement and velocity is measured through the linear encoder integrated in the linear electric machine. The resolution of the encoder is 400 micron. The combustion cylinder volume is calculated from the piston displacement data. Two Kistler pressure transducers (6052 with charge amplifier 5064) are installed in the cylinder heads on both sides to measure the in-cylinder gas pressure. The high speed data is acquired by a National Instrument PXI QAQ card (PXIe-6358) and an in-house made LabView Real Time program for further analysis. 3. Test results for the motoring process As compression ratio and in-cylinder gas pressure are regarded as two important factors, the targets for a successful motoring process are: a. To reach an effective compression ratio of 8:1. b. To reach a maximum in-cylinder pressure of 10 bar. During the experiment of the motoring process, the ignition system and the fuel supply system were deactivated. Different motoring forces were implemented to start the engine; the data of piston dynamics and engine operating characteristics were collected for further analysis. During the testing, the mover could not be dragged with a motor force less than 60 N, indicating that the static friction force was nearly 60 N. Fig. 4 shows the in-cylinder pressure with a starting motoring force of 80 N and 95 N respectively. It can be seen that the amplitude of the in-cylinder pressure grows as expected, and the peak pressure is higher with a greater motoring force. With an insufficient motoring force, after a few cycles the peak in cylinder pressure tends to reach a stable state which is lower than the requested value. For a motoring force of 80 N, the maximum in-cylinder pressure increases to nearly 5 bar after 4 cycles and remain stable afterwards. When the starting motor force increases to 95 N, the in-cylinder pressure grows to over 7 bar in 6 cycles and then stay stable. The running frequency is 8.5 Hz with a motor force of 80 N, and 10 Hz with 95 N, so the equivalent speed of the engine is approximately 510 cycle per minute and 600 cycle per minute respectively. From the velocity versus displacement profile, shown in Fig. 5, it can be observed that, if a fixed motoring force acts on the mover in the direction of its velocity, the amplitude of displacement and the peak piston velocity grow from cycle to cycle. The motoring force has significant impact on the piston velocity and displacement. With higher motoring force, the achieved peak piston velocity and compression ratio are higher. Moreover, the piston is seen to run at high and relative constant speed at the middle portion of the stroke, and then slow down at bottom dead centre (BDC) and top dead centre (TDC). Without combustion process, the velocity profile is nearly symmetrical as the difference of acceleration
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
4
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
Fig. 2. Control system.
before and after BDC/TDC is small. The maximum velocity occurs at the middle stroke, where the motor force equals to the sum of compression gas force and friction force.
avoid oversizing and at the same time, ensure sufficient motor force can be applied to successfully start the engine. The peak in-cylinder pressure and the equivalent engine speed at stable motoring are shown in Fig. 7. From the experimental results, it can be observed that both of them are nearly in a linear relationship with motoring force. As a result, in order to smoothly start this engine prototype, the minimum starting motor force should be above 103 N. When the motoring force increased to over 103 N, the engine is ready for ignition after several working cycle. The running frequency of the engine is equivalent to approximately 650 cycles per minute, which is feasible for ignition compared with the required starting speed for conventional crankshaft engines. Within a reasonable range, higher motoring force is preferred in order to reach higher in-cylinder pressure and faster running speed, which will make it easier for ignition.
3.1. Comparison of varied motor forces
4. Test results for the ignition process
The motoring process experiments were carried out with different motoring forces from 80 N to 125 N in 15 N interval. The compression ratios achieved for each cycle are demonstrated in Fig. 6. With a fixed motoring force, the compression ratio achieved at each running cycle shows a non-linear index increase and tends to reach a stable state after a number of cycles. The increase rate of the compression ratio and the maximum compression ratio are higher with a higher motoring force. If the motor force is insufficient, the stable compression ratio cannot meet the target for ignition. However, for a commercial linear electric motor, the peak output motor force and the maximum speed of the mover are determined by the physical structure of the machine. The linear electric motor used in the prototype, for example, has a maximum continuous force output of 232 N, and maximum allowed speed of 3.1 m/s. As a result, the motoring force needs to be controlled within a reasonable range. The selection of linear motor should also
During this part of experiment, a fixed motoring force of 125 N was implemented to start the engine. The ignition and fuel delivery system were activated and the ignition timing is set at 27 mm from the central position. The engine was operated at stoichiometric air–fuel ratio (k = 1.0). In this paper, the ignition timing and injected fuel were not optimised to the best performance and the electric linear machine was not switched to generation mode, i.e. the motoring force was constantly provided.
Fig. 3. Ignition and injection trigger signals.
Table 2 Description of the trigger signals. Signal
Usage
Rising edge at X0 Falling edge at X1
Start charging the ignition coil of the left cylinder Trigger the injection signal of the right cylinder Trigger the spark plug of the left cylinder Start charging the ignition coil of the right cylinder Trigger the injection signal of the left cylinder Trigger the spark plug of the right cylinder
Rising edge at X0 Falling edge at X1
Fig. 4. In-cylinder gas pressure.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
5
Fig. 8. Comparison with motoring process and ignition process. Fig. 5. Displacement and velocity.
Fig. 6. Compression ratio at each running cycle.
Fig. 9. Comparison of the fourth cycle between the two modes.
Fig. 9 shows the detailed comparison of the fourth cycle between the two modes. For the ignition mode, the in-cylinder pressure was much lower than without combustion. One possible reason to explain this is because the variations and error of the frequency for different tests, and the piston positions on the x axis are not exactly the same. But after the mixture was ignited, the pressure rose very rapidly and eventually achieved nearly 40 bar. The heat release seems violent due to the high pressure rise rate, but no knocking was observed. For a very dynamic starting process, this can be considered successful. 4.2. Piston dynamics
Fig. 7. Peak in-cylinder pressure and equivalent running speed.
4.1. Engine performance The comparison of the measured in-cylinder pressure traces between motoring process and ignition process is demonstrated in Fig. 8. For the first three cycles, the peak in-cylinder pressure gradually grew to over 10 bar as the piston assembly oscillated. At the fourth cycle, there was a sharp increase of the in-cylinder pressure with a compression ratio of over 9:1, indicating that the in-cylinder air fuel mixture was successfully ignited and combustion happened. Moreover, the successful combustion proves that the starting strategy proposed before is feasible for the FPEG and the target set for the starting process is reasonable.
Fig. 10 demonstrates the piston displacement and velocity profile for 30 consecutive cycles during the starting process. The engine was motored to a desirable compression ratio and ignition started from cycle 4. It can be observed that the amplitude of piston displacement and velocity grew as expected. Before ignition, the piston velocity archived 2 m/s at the middle range of the stroke. After ignition, the piston ran at a higher but still relatively constant speed and the maximum velocity was nearly 4 m/s at the middle position. The shape of the velocity-displacement oval changed as well. It became less symmetric with a longer stroke due to the greater acceleration in expansion stroke caused by combustion. 4.3. Discussion on engine misfire The successful ignition notwithstanding, cycle-to-cycle variation is deemed severe and the engine misfired every one or two
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
6
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
engine started to get warmer for more stable ignition, despite the influence of the inside brake from the motor. To meet these challenges, optimal working conditions for the engine should be analysed. The mode of the linear electric machine is supposed to be switched properly, and a robust control system is required to ensure stable running. Mikalsen and Roskilly have presented parametric analysis and detailed control strategies in their papers on the control of a diesel free-piston engines [28–30]. For a spark-ignited free-piston engine, potential control variables can be the ignition timing, the throttle opening, the injection timing, and the resistance force from the generator as well as the external load. Further research relevant on the optimisation and control of the FPEG is undertaken and will be published in our following papers. Fig. 10. Displacement and velocity in 30 running cycles.
4.4. Discussion on linear electric machine mode switch cycles, which is demonstrated in Fig. 11. During the misfire cycles, the peak piston velocity decreased to nearly 2.5 m/s. The piston assembly was driven by the linear electric motor, and the piston dynamic was similar with the motoring process. The possible reasons of the variations and unstable operation could be a combination of a number of factors. The air/fuel mixture formation might vary from cycle to cycle in cold engine conditions, and the spark and initial flame propagation have cyclic variation as normal SI engines. These factors will result in an unstable combustion process during starting. Unlike normal engines, the piston dynamics of the FPE are coupled with the combustion process; therefore the unstable combustion could lead to an undesired piston motion trajectory and then affect the combustion process in the next cycle. Moreover, in this preliminary test, the ignition timing was fixed without careful optimisation. More experiments are being undertaken now to optimise the ignition timing and other control factors. Furthermore, according to the introduction, the motor is suggested to be operated below 3.1 m/s, otherwise might cause overheat in the stator. An inside brake would be triggered if the mover’s speed reaches its maximum allowed value and slow down the mover. During the test, the piston velocity could reach 4 m/s after ignition, and then the inside brake was supposed to be engaged to slow down the piston, which would affect the gas flow rate through the piston ring, heat transfer through cylinder wall, as well as the intake and scavenging processes. Then misfire happened for the cylinder on the other side. However, the linear electric machine was still operating as a motor, and the piston was kept driving and oscillating to reach the expected conditions for ignition. For the last few cycles shown in Fig. 11, continuous ignition and combustion was achieved. One of the reason to explain this was because the
Fig. 11. Tested piston velocity vs time after ignition.
The linear electric motor is expected to be switched to generator mode and output electricity after successful combustion being achieved at the end of the starting process. Appropriate brake power is required to allow a stable operation and the best energy conversion efficiency. Moreover, proper switching timing is necessary as it may change the piston dynamics significantly. Three possible parameters to identify a sufficient brake power or a misfire for each running cycle are: TDC position (or equivalently effective stroke), peak in-cylinder pressure, maximum piston velocity at half stroke (middle position). From the experimental results above, the variations of the TDC position for each running cycle is small compared to the other two parameters. As a result, the peak in-cylinder pressure and maximum piston velocity are considered as two practical input parameters as criteria for the switching control of the linear electric motor. The motor then needs to be powered off and connected to an external load. Detailed investigation on the switching and generation is still under investigation, the simulation results and test data will be presented in our further publications. 5. Conclusion This paper presents an experimental investigation of the starting process of an FPEG prototype, including motoring and ignition processes. Test results show that for the mechanical resonance starting strategy, the in-cylinder properties trends to reach a stable state after several cycles and the compression ratio shows a nonlinear increase. The motoring force is required to be controlled within a reasonable range to achieve the desired compression ratio. The peak in-cylinder pressure and the equivalent engine speed are nearly in a linear relationship with the motoring force. With a fixed motoring force of 125 N, the in-cylinder mixture was ignited successfully at the fourth cycle. The peak combustion pressure was over 40 bar. This suggests that the proposed starting strategy is feasible. After ignition, the piston ran at high and relatively constant speed at the middle portion of the stroke. The peak piston velocity increases significantly to around 4.0 m/s. Cycle-to-cycle variation of the piston movement was significant and the engine misfired frequently. During the misfire cycles, the peak piston velocity decreased to nearly 2.5 m/s; and the piston dynamics were similar to the motoring process. Possible reasons of the variations and unstable operation could be a combination of several factors such as the variation of the air/fuel mixture formation, the non-optimised ignition and injection timing, and the influence of the inside brake from the motor. The peak in-cylinder pressure and maximum piston velocity are considered as two practical input parameters as criteria for the future switching control of the linear electric motor.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
The future work will focus on the optimisation of the ignition timing and other control parameters, and an adaptive control system is being developed to improve the engine performance and to ensure a stable operation. The mode switch of the linear electric machine will be undertaken and the overall generating efficiency of the FPEG system will be calculated by testing the electricity power output. Acknowledgement This project is supported by the Programme of Introducing Talents of Discipline to Universities of China (B12022) and National Nature Science Foundation of China (51006010). We would like to thank the sponsors. References [1] Mikalsen R, Roskilly AP. The design and simulation of a two-stroke free-piston compression ignition engine for electrical power generation. Appl Therm Eng 2008;28(5):589–600. [2] Max Erland. FPEC, Free piston energy converter. In: Proceedings of the 21st electric vehicle symposium & exhibition, EVS, vol. 21; 2005. [3] Atkinson C, Petreanu S, Clark N, Atkinson R et al. Numerical simulation of a two-stroke linear engine-alternator combination, SAE Technical Paper 199901-0921; 1999. [4] Bergman Miriam, Fredriksson Jakob, Golovitchev Valeri I. CFD-based optimization of a diesel-fueled free piston engine prototype for conventional and HCCI combustion. SAE Technical Paper 2008-01-2423; 2008. [5] Zulkifli Saiful A, Karsiti Mohd N, Aziz Rashid A. Starting of a free-piston linear engine-generator by mechanical resonance and rectangular current commutation. In: Vehicle Power and Propulsion Conference, IEEE; 2008. [6] Goldsborough S. Scott, Van Blarigan Peter. A numerical study of a free piston IC engine operating on homogeneous charge compression ignition combustion. SAE Technical Paper 1999-01-0619; 1999. [7] Mikalsen Rikard, Roskilly AP. A review of free-piston engine history and applications. Appl Therm Eng 2007;27(14):2339–52. [8] Zhao Z, Zhang F, Huang Y, Zhao C, Guo F. An experimental study of the hydraulic free piston engine. Appl Energy 2012;99:226–33. [9] Xiao Jin, Li Qingfeng, Huang Zhen. Motion characteristic of a free piston linear engine. Appl Energy 2010;87(4):1288–94. [10] Nandkumar Subhash. Two-stroke linear engine. Dissertation, West Virginia University; 1998. [11] Petreanu Sorin. Conceptual analysis of a four-stroke linear engine. Dissertation, West Virginia University; 2000. [12] Houdyschell David. A diesel two-stroke linear engine. Dissertation, West Virginia University; 2000.
7
[13] Van Blarigan P, Paradiso N, Goldsborough S. Homogeneous charge compression ignition with a free piston: a new approach to ideal otto cycle performance. SAE Paper 1998;982484. [14] http://www4.eere.energy.gov/vehiclesandfuels/resources/meritreview/sites/ default/files/ace008_johnson_2013_o.pdf. [15] Hansson Jörgen. Analysis and control of a hybrid vehicle powered by a freepiston energy converter, Dissertation, Royal Institute of Technology; 2006. [16] Mao JL, Zuo ZX, Lin W, Feng H. Multi-dimensional scavenging analysis of a free-piston linear alternator based on numerical simulation. Appl Energy 2011;88. [17] Mao Jinlong, Zuo Zhengxing, Feng Huihua. Parameters coupling designation of diesel free-piston linear alternator. Appl Energy 2011;88:4577–89. [18] Carter D, Wechner E. The free piston power pack: sustainable power for hybrid electric vehicles. SAE Technical Paper 2003-01-3277; 2003. [19] Ferraril Cornelius, Friedrich Horst E. Development of a free-piston linear generator for use in an extended-range electric vehicle. EVS26 international battery, hybrid and fuel cell electric vehicle symposium. California, May 6–9; 2012. [20] Haag J, Ferrari C, Starcke J, Stöhr M et al. Numerical and experimental investigation of in-cylinder flow in a loop-scavenged two-stroke free piston engine, SAE Technical Paper 2012-32-0114; 2012. [21] Kock F, Haag J, Friedrich H. The free piston linear generator – development of an innovative, compact, highly efficient range-extender module, SAE Technical Paper 2013-01-1727; 2013. [22] Schneider Stephan, Frank Rinderknecht, Friedrich Horst E. Design of future concepts and variants of the free piston linear generator. inecological vehicles and renewable energies (EVER). In: 2014 ninth international conference on, IEEE; 2014. p. 1–8. [23] Kosaka H, Akita T, Moriya K, Goto S et al. Development of free piston engine linear generator system part 1 – investigation of fundamental characteristics. SAE Technical Paper 2014-01-1203; 2014. [24] Goto S, Moriya K, Kosaka H, Akita T et al. Development of free piston engine linear generator system part 2 – investigation of control system for generator. SAE Technical Paper 2014-01-1193; 2014. [25] Aichlmayr HT. Design considerations, modeling, and analysis of microhomogeneous charge compression ignition combustion free-piston engines. Dissertation, University of Minnesota; 2002. [26] Jia B, Zuo Z, Feng H, Tian G, Roskilly AP. Development approach of a sparkignited free-piston engine generator. SAE Technical Paper 2014-01-2894; 2014. [27] Jia B, Zuo Z, Tian G, Feng H, Roskilly AP. Development and validation of a freepiston engine generator numerical model. Energy Convers Manage 2015;91:333–41. [28] Mikalsen R, Roskilly AP. The control of a free-piston engine generator. Part 1: fundamental analyses. Appl Energy 2010;87(4):1273–80. [29] Mikalsen R, Roskilly AP. The control of a free-piston engine generator. Part 2: engine dynamics and piston motion control. Appl Energy 2010;87(4):1281–7. [30] Mikalsen R, Jones E, Roskilly AP. Predictive piston motion control in a freepiston internal combustion engine. Appl Energy 2010;87(5):1722–8.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
Contents lists available at ScienceDirect
Applied Energy journal homepage: www.elsevier.com/locate/apenergy
An experimental investigation into the starting process of free-piston engine generator q Boru Jia a,b, Guohong Tian b, Huihua Feng a,⇑, Zhengxing Zuo a, A.P. Roskilly b a b
School of Mechanical Engineering, Beijing Institute of Technology, Beijing 100081, China Sir Joseph Swan Centre for Energy Research, Newcastle University, Newcastle upon Tyne NE1 7RU, UK
h i g h l i g h t s A prototype free-piston engine generator is developed and tested. The in-cylinder mixture was ignited successfully with the proposed control strategy. Two practical input parameters are selected for the future motor switching control. Three possible reasons of misfire are analysed.
a r t i c l e
i n f o
Article history: Received 2 September 2014 Received in revised form 3 February 2015 Accepted 18 February 2015 Available online xxxx Keywords: Free-piston generator Linear electric machine Experiment Starting process Combustion
a b s t r a c t This paper presents an experimental investigation of the starting process of a prototype free piston engine generator (FPEG). Experimental test results show that during the motoring stage, the peak in-cylinder pressure and compression ratio increase in a non-linear manner and trend to reach a stable state after a number of cycles. The motoring force is suggested to be within a reasonable range. With a fixed starting force of 125 N, the in-cylinder air fuel mixture was successfully ignited at the fourth cycle with a compression ratio of over 9:1. The peak in-cylinder pressure for the first combustion cycle reached over 40 bar. The piston ran at high and relatively constant speed at the middle portion of the stroke. The peak piston velocity increases significantly to around 4.0 m/s. Cycle-to-cycle variation of the piston movement was significant and the engine misfired frequently. During the misfire cycles, the peak piston velocity decreased to nearly 2.5 m/s; and the piston dynamics were similar to the motoring process. Based on these, discussion on misfire and further stable running control, as well as the linear electric machine mode switch were presented. Ó 2015 Elsevier Ltd. All rights reserved.
1. Introduction Free-piston engine generator (FPEG) is a novel type of energy conversion device that integrates a linear combustion engine and a linear electric machine [1]. Combustion in the chamber drives the piston reciprocates in a nearly resonant way and the linear electric machine converts part of the mover’s kinetic energy to electrical energy. The effective efficiency is estimated to be up to 46% (including friction and compressor losses) at a power level of 23 kW and shows promising results with respect to engine performance and emissions [2]. Since the FPEG was first proposed, despite the interests it has attracted from all over the world
q This paper is included in the Special Issue of Clean Transport edited by Prof. Anthony Roskilly, Dr. Roberto Palacin and Prof. Yan. ⇑ Corresponding author. Tel.: +86 10 68911062. E-mail address: [email protected] (H. Feng).
[3–6], there has not been any stably operating prototype reported by now. The FPEs were first proposed around 1930, and during 1930 to 1960 they were mainly used as air compressors and gas generators as they provided advantages over conventional combustion engines and gas turbines at that time [7]. In recent years, modern applications of the free-piston concept have been proposed for the generation of electric and hydraulic power, typically in hybrid electric vehicles [8–9]. Different prototype designs have been reported using the FPEG concept. The majority of these were, however, not commercially successful. This section gives an overview of known FPEG development, with an emphasis on reports where prototype performance data have been reported. Researchers at West Virginia University described the development of a spark-ignited (SI) dual piston FPEG [10]. This 36.5 mm bore size prototype was reported to have achieved 316 W power output at 23.1 Hz, with a 50 mm maximum stroke. High
http://dx.doi.org/10.1016/j.apenergy.2015.02.065 0306-2619/Ó 2015 Elsevier Ltd. All rights reserved.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
2
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
cycle-to-cycle variations were observed particularly at low loads [11], which indicated an unstable operation. A two stroke compression ignition (CI) version linear engine prototype was also developed [11]. This CI engine had a similar mechanical arrangement to the SI prototype. The linear alternator interposed with the linear engine also operated as a starting device. In order to aid the cold start of the engine, each cylinder was equipped with a glow plug [12]. The research team in Sandia National Laboratories presented the design of a dual cylinder FPEG. The engine employed a homogeneous charge compression ignition (HCCI) and was aimed to operate on a variety of hydrogen-containing fuels. Test results from a compression–expansion machine showed nearly constant volume combustion with hydrogen, bio-gas, and ammonia at the equivalence ratio of approximately 0.3. The target efficiency was 50% overall considering 56% engine thermal efficiency and 96% generator efficiency [13]. In 2008, the dual cylinder free-piston engine configuration was changed into an opposed piston type, which was adopted to utilize the self-balance effect [14]. A European Commission-funded Free Piston Energy Converter (FPEC) project researched the subject of FPEG aimed at the development of an efficient new technology suitable for vehicle propulsion, auxiliary power units and distributed power generation since 2002 [2]. The prototype ran on diesel fuel, and was used primarily for testing in a test cell for validation of the specific FPEG issues. The converter was equipped with fuel injectors, pneumatic operated valves, cylinder pressure sensors, and translator displacement sensors. The combustion system is a Direct Injection, 2-stroke, two cylinder systems with electrical controlled valves and scavenging ports operating in HCCI mode on diesel fuel [15]. However, there has not been many test data reported by now. The approach by the Beijing Institute of Technology used a loop scavenged, carburetted free-piston, double-ended cylinder arrangement with a linear alternator; the bore of the engine was 34 mm, and the effective stroke was 20 mm [16]. Experiments results showed that the engine misfired every one or two cycles [16]. For the compression ignition FPEG prototype developed by the same group, the cylinder bore and stroke were enlarged to 62 mm and 86 mm respectively. If a compression ratio of 15 was required at the end of the compression stroke to enable selfignition, the corresponding in-cylinder pressure force was over 11 kN, which made it practically impossible to start the engine at one compression stroke without an largely oversized electric linear machine [17]. Pempek Systems Pty. Ltd., an Australian company, was one of the research leaders in this area. Their conceptual free piston engine generator design aims the power source of a high performance hybrid vehicle which has a top speed of 160 km/h and 0–100 km/ h acceleration of only 5.4 s. This vehicle will also be equipped with a brake recovery system. They conceptually designed a 25 kW free piston engine generator which has 50% engine thermal efficiency and over 93% of the generator efficiency [18]. The German Aerospace Centre (DLR) developed an FPEG prototype [19–22]. It consisted of three main subsystems: an internal combustion engine (ICE) converting chemical energy into kinetic energy, a linear generator (LG) converting kinetic energy into electric energy and a gas spring (GS) storing energy and inverting the piston movement. At 21 Hz, a power output of roughly 10 kW had been measured. Increasing the frequency up to 50 Hz should lead to a power output of 25 kW of a single piston FPEG system [19,20]. The FPEG prototype developed by Toyota Central R&D Labs Inc. was thin and compact [23,24]. The developed FPEG consisted of a two-stroke combustion chamber, a linear generator, and a gas spring chamber. A power generation experiment was carried out, and the results demonstrated that the prototype operated stably
for a long period of time, despite of the abnormal combustion during the test [24]. The unique piston motion and its effect on combustion and power generation in the FPEG prototype were experimentally analysed [23]. FPEG can be divided into three categories according to piston/cylinder configuration: single piston, dual piston and opposed piston [25]. The basic operation principles are equal for each concept; differences between the concepts are the number of combustion chambers and compression stroke realization [25]. There has been successful implementation of the single piston type, coupled with a gas spring chamber [20,24]. However, despite the problems being reported, the dual piston configuration remains the most popular layout due to the following advantages over single piston and opposed piston configurations: 1. The only moving part is a linear magnet mover coupled with pistons at each end and placed between two opposing combustion chambers. This allows a simple and more compact device with higher power to weight ratio. 2. It eliminates the need for a rebound device, as the combustion force drives the piston assembly to overcome the compression pressure in the other cylinder. Because of the potential advantages above, the dual piston type FPEG is adopted in the prototype reported in this paper. A newly designed FPEG prototype has been built to validate the feasibility of the technical scheme of dual piston FPEG. This paper describes the configuration of the newly designed system as well as the fundamental test results. Based on these, discussion on misfire and further stable running control, as well as the linear electric machine mode switch were presented. 2. FPEG prototype 2.1. Prototype specification The FPEG prototype is demonstrated in Fig. 1. This prototype is a dual piston, two-stroke, spark-ignited, uniflow scavenging engine. It is integrated with a dedicated bench of steel plates, constructed to reduce the vibrations during testing. The specifications of the prototype are summarised in Table 1. More details about the prototype have been presented in our previous paper [26,27], and the carburettor has been replaced by a port fuel injection (PFI) system for a more precise air fuel ratio control. The PFI system with integrated electric fuel pump and fuel filter conveys the required amount of fuel from the tank to the injectors at a constant pressure. The fuel is injected into the intake manifold
Fig. 1. The prototype of free-piston engine generator.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx Table 1 Prototype specifications. Parameters
Value
Number of cylinder Stroke Combustion mode Bore Maximum total stroke Maximum effective stroke (compression stroke) Moving mass
2 2 SI 52.5 mm 70.0 mm 35.0 mm 5.0 kg
3
( X0) displacement from centre and falling edge is at X1 ( X1), as illustrated in Fig. 3. These two signals were used to trigger both spark ignition and fuel injection. The function is summarised in Table 2. The injection is triggered by the falling edge, and the duration of the injection is determined by the injected fuel quantity after calibration to deliver a stoichiometric fuel/air mixture. For the test carried out at this stage, the ignition timing and the injection duration were fixed and ignition timing has not been optimised. 2.3. Test facilities
to form a stoichiometric air/fuel mixture. The ignition system mainly consisted of a 12 V battery, ignition coil, and spark plug, and they are all off-the-shelf products. The ignition and injection systems are activated simultaneously when the engine is motored to achieve a good compression ratio.
2.2. Control strategy The general working process of the FPEG consists of three stages, i.e. the starting process, linear electric machine mode switching process, and generating process. The starting process discussed by this paper can be primarily divided into two stages. In stage one, the linear electric machine runs as a motor that provides motoring force to start the engine by mechanical resonance (motoring process). When piston speed and compression ratio achieve the desired values, fuel is delivered and ignited which is regarded as the second stage (ignition process). From the control point of view, each process is supposed to be coupled with proper control strategies to ensure stable running. Then the control system of the prototype can be divided by four parts, which are the starting system, ignition control system, linear electric machine mode control system, and the external load control system. All these systems are integrated in an electronic controller, which will generate the control signals for each sub system according to the feedback from the sensors. The structure of the control system is illustrated in Fig. 2. This paper focuses on the control strategy of the motoring process and the engine ignition control system, and presents suggestions for the further electric machine mode switching control as well as the stable running control for the generating process according to the fundamental test results. During motoring process, the linear electric machine is operated as a motor and generates constant motor force in the direction of the piston velocity. The displacement amplitude as well as the peak in-cylinder gas pressure is expected to grow and finally reach the required values for ignition. Detailed investigation of the starting strategy and implementation can be found in our previous paper [26]. The control system for the motoring process was coupled with PID compensators to reduce the influence from back electromagnetic voltage. It took the mover’s displacement and velocity, as well as the actual current in the coil as feedbacks. As a result, the electric machine was operated as closed-loop control system and generated constant motor force in direction of the mover’s velocity [26]. At the end of the motoring process, the ignition system and the PFI system are both activated simultaneously. Ignition is usually triggered at the end of the compression stroke, while injection starts at the end of the expansion stroke. The ignition and injection signals are both triggered using piston position signal as reference. This signal is provided by a linear encoded integrated in the linear electric machine and can be adjusted. The triggers are two TTL signals symmetric to the central position. The rising edge is at X0
The data of piston displacement and velocity is measured through the linear encoder integrated in the linear electric machine. The resolution of the encoder is 400 micron. The combustion cylinder volume is calculated from the piston displacement data. Two Kistler pressure transducers (6052 with charge amplifier 5064) are installed in the cylinder heads on both sides to measure the in-cylinder gas pressure. The high speed data is acquired by a National Instrument PXI QAQ card (PXIe-6358) and an in-house made LabView Real Time program for further analysis. 3. Test results for the motoring process As compression ratio and in-cylinder gas pressure are regarded as two important factors, the targets for a successful motoring process are: a. To reach an effective compression ratio of 8:1. b. To reach a maximum in-cylinder pressure of 10 bar. During the experiment of the motoring process, the ignition system and the fuel supply system were deactivated. Different motoring forces were implemented to start the engine; the data of piston dynamics and engine operating characteristics were collected for further analysis. During the testing, the mover could not be dragged with a motor force less than 60 N, indicating that the static friction force was nearly 60 N. Fig. 4 shows the in-cylinder pressure with a starting motoring force of 80 N and 95 N respectively. It can be seen that the amplitude of the in-cylinder pressure grows as expected, and the peak pressure is higher with a greater motoring force. With an insufficient motoring force, after a few cycles the peak in cylinder pressure tends to reach a stable state which is lower than the requested value. For a motoring force of 80 N, the maximum in-cylinder pressure increases to nearly 5 bar after 4 cycles and remain stable afterwards. When the starting motor force increases to 95 N, the in-cylinder pressure grows to over 7 bar in 6 cycles and then stay stable. The running frequency is 8.5 Hz with a motor force of 80 N, and 10 Hz with 95 N, so the equivalent speed of the engine is approximately 510 cycle per minute and 600 cycle per minute respectively. From the velocity versus displacement profile, shown in Fig. 5, it can be observed that, if a fixed motoring force acts on the mover in the direction of its velocity, the amplitude of displacement and the peak piston velocity grow from cycle to cycle. The motoring force has significant impact on the piston velocity and displacement. With higher motoring force, the achieved peak piston velocity and compression ratio are higher. Moreover, the piston is seen to run at high and relative constant speed at the middle portion of the stroke, and then slow down at bottom dead centre (BDC) and top dead centre (TDC). Without combustion process, the velocity profile is nearly symmetrical as the difference of acceleration
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
4
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
Fig. 2. Control system.
before and after BDC/TDC is small. The maximum velocity occurs at the middle stroke, where the motor force equals to the sum of compression gas force and friction force.
avoid oversizing and at the same time, ensure sufficient motor force can be applied to successfully start the engine. The peak in-cylinder pressure and the equivalent engine speed at stable motoring are shown in Fig. 7. From the experimental results, it can be observed that both of them are nearly in a linear relationship with motoring force. As a result, in order to smoothly start this engine prototype, the minimum starting motor force should be above 103 N. When the motoring force increased to over 103 N, the engine is ready for ignition after several working cycle. The running frequency of the engine is equivalent to approximately 650 cycles per minute, which is feasible for ignition compared with the required starting speed for conventional crankshaft engines. Within a reasonable range, higher motoring force is preferred in order to reach higher in-cylinder pressure and faster running speed, which will make it easier for ignition.
3.1. Comparison of varied motor forces
4. Test results for the ignition process
The motoring process experiments were carried out with different motoring forces from 80 N to 125 N in 15 N interval. The compression ratios achieved for each cycle are demonstrated in Fig. 6. With a fixed motoring force, the compression ratio achieved at each running cycle shows a non-linear index increase and tends to reach a stable state after a number of cycles. The increase rate of the compression ratio and the maximum compression ratio are higher with a higher motoring force. If the motor force is insufficient, the stable compression ratio cannot meet the target for ignition. However, for a commercial linear electric motor, the peak output motor force and the maximum speed of the mover are determined by the physical structure of the machine. The linear electric motor used in the prototype, for example, has a maximum continuous force output of 232 N, and maximum allowed speed of 3.1 m/s. As a result, the motoring force needs to be controlled within a reasonable range. The selection of linear motor should also
During this part of experiment, a fixed motoring force of 125 N was implemented to start the engine. The ignition and fuel delivery system were activated and the ignition timing is set at 27 mm from the central position. The engine was operated at stoichiometric air–fuel ratio (k = 1.0). In this paper, the ignition timing and injected fuel were not optimised to the best performance and the electric linear machine was not switched to generation mode, i.e. the motoring force was constantly provided.
Fig. 3. Ignition and injection trigger signals.
Table 2 Description of the trigger signals. Signal
Usage
Rising edge at X0 Falling edge at X1
Start charging the ignition coil of the left cylinder Trigger the injection signal of the right cylinder Trigger the spark plug of the left cylinder Start charging the ignition coil of the right cylinder Trigger the injection signal of the left cylinder Trigger the spark plug of the right cylinder
Rising edge at X0 Falling edge at X1
Fig. 4. In-cylinder gas pressure.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
5
Fig. 8. Comparison with motoring process and ignition process. Fig. 5. Displacement and velocity.
Fig. 6. Compression ratio at each running cycle.
Fig. 9. Comparison of the fourth cycle between the two modes.
Fig. 9 shows the detailed comparison of the fourth cycle between the two modes. For the ignition mode, the in-cylinder pressure was much lower than without combustion. One possible reason to explain this is because the variations and error of the frequency for different tests, and the piston positions on the x axis are not exactly the same. But after the mixture was ignited, the pressure rose very rapidly and eventually achieved nearly 40 bar. The heat release seems violent due to the high pressure rise rate, but no knocking was observed. For a very dynamic starting process, this can be considered successful. 4.2. Piston dynamics
Fig. 7. Peak in-cylinder pressure and equivalent running speed.
4.1. Engine performance The comparison of the measured in-cylinder pressure traces between motoring process and ignition process is demonstrated in Fig. 8. For the first three cycles, the peak in-cylinder pressure gradually grew to over 10 bar as the piston assembly oscillated. At the fourth cycle, there was a sharp increase of the in-cylinder pressure with a compression ratio of over 9:1, indicating that the in-cylinder air fuel mixture was successfully ignited and combustion happened. Moreover, the successful combustion proves that the starting strategy proposed before is feasible for the FPEG and the target set for the starting process is reasonable.
Fig. 10 demonstrates the piston displacement and velocity profile for 30 consecutive cycles during the starting process. The engine was motored to a desirable compression ratio and ignition started from cycle 4. It can be observed that the amplitude of piston displacement and velocity grew as expected. Before ignition, the piston velocity archived 2 m/s at the middle range of the stroke. After ignition, the piston ran at a higher but still relatively constant speed and the maximum velocity was nearly 4 m/s at the middle position. The shape of the velocity-displacement oval changed as well. It became less symmetric with a longer stroke due to the greater acceleration in expansion stroke caused by combustion. 4.3. Discussion on engine misfire The successful ignition notwithstanding, cycle-to-cycle variation is deemed severe and the engine misfired every one or two
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
6
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
engine started to get warmer for more stable ignition, despite the influence of the inside brake from the motor. To meet these challenges, optimal working conditions for the engine should be analysed. The mode of the linear electric machine is supposed to be switched properly, and a robust control system is required to ensure stable running. Mikalsen and Roskilly have presented parametric analysis and detailed control strategies in their papers on the control of a diesel free-piston engines [28–30]. For a spark-ignited free-piston engine, potential control variables can be the ignition timing, the throttle opening, the injection timing, and the resistance force from the generator as well as the external load. Further research relevant on the optimisation and control of the FPEG is undertaken and will be published in our following papers. Fig. 10. Displacement and velocity in 30 running cycles.
4.4. Discussion on linear electric machine mode switch cycles, which is demonstrated in Fig. 11. During the misfire cycles, the peak piston velocity decreased to nearly 2.5 m/s. The piston assembly was driven by the linear electric motor, and the piston dynamic was similar with the motoring process. The possible reasons of the variations and unstable operation could be a combination of a number of factors. The air/fuel mixture formation might vary from cycle to cycle in cold engine conditions, and the spark and initial flame propagation have cyclic variation as normal SI engines. These factors will result in an unstable combustion process during starting. Unlike normal engines, the piston dynamics of the FPE are coupled with the combustion process; therefore the unstable combustion could lead to an undesired piston motion trajectory and then affect the combustion process in the next cycle. Moreover, in this preliminary test, the ignition timing was fixed without careful optimisation. More experiments are being undertaken now to optimise the ignition timing and other control factors. Furthermore, according to the introduction, the motor is suggested to be operated below 3.1 m/s, otherwise might cause overheat in the stator. An inside brake would be triggered if the mover’s speed reaches its maximum allowed value and slow down the mover. During the test, the piston velocity could reach 4 m/s after ignition, and then the inside brake was supposed to be engaged to slow down the piston, which would affect the gas flow rate through the piston ring, heat transfer through cylinder wall, as well as the intake and scavenging processes. Then misfire happened for the cylinder on the other side. However, the linear electric machine was still operating as a motor, and the piston was kept driving and oscillating to reach the expected conditions for ignition. For the last few cycles shown in Fig. 11, continuous ignition and combustion was achieved. One of the reason to explain this was because the
Fig. 11. Tested piston velocity vs time after ignition.
The linear electric motor is expected to be switched to generator mode and output electricity after successful combustion being achieved at the end of the starting process. Appropriate brake power is required to allow a stable operation and the best energy conversion efficiency. Moreover, proper switching timing is necessary as it may change the piston dynamics significantly. Three possible parameters to identify a sufficient brake power or a misfire for each running cycle are: TDC position (or equivalently effective stroke), peak in-cylinder pressure, maximum piston velocity at half stroke (middle position). From the experimental results above, the variations of the TDC position for each running cycle is small compared to the other two parameters. As a result, the peak in-cylinder pressure and maximum piston velocity are considered as two practical input parameters as criteria for the switching control of the linear electric motor. The motor then needs to be powered off and connected to an external load. Detailed investigation on the switching and generation is still under investigation, the simulation results and test data will be presented in our further publications. 5. Conclusion This paper presents an experimental investigation of the starting process of an FPEG prototype, including motoring and ignition processes. Test results show that for the mechanical resonance starting strategy, the in-cylinder properties trends to reach a stable state after several cycles and the compression ratio shows a nonlinear increase. The motoring force is required to be controlled within a reasonable range to achieve the desired compression ratio. The peak in-cylinder pressure and the equivalent engine speed are nearly in a linear relationship with the motoring force. With a fixed motoring force of 125 N, the in-cylinder mixture was ignited successfully at the fourth cycle. The peak combustion pressure was over 40 bar. This suggests that the proposed starting strategy is feasible. After ignition, the piston ran at high and relatively constant speed at the middle portion of the stroke. The peak piston velocity increases significantly to around 4.0 m/s. Cycle-to-cycle variation of the piston movement was significant and the engine misfired frequently. During the misfire cycles, the peak piston velocity decreased to nearly 2.5 m/s; and the piston dynamics were similar to the motoring process. Possible reasons of the variations and unstable operation could be a combination of several factors such as the variation of the air/fuel mixture formation, the non-optimised ignition and injection timing, and the influence of the inside brake from the motor. The peak in-cylinder pressure and maximum piston velocity are considered as two practical input parameters as criteria for the future switching control of the linear electric motor.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065
B. Jia et al. / Applied Energy xxx (2015) xxx–xxx
The future work will focus on the optimisation of the ignition timing and other control parameters, and an adaptive control system is being developed to improve the engine performance and to ensure a stable operation. The mode switch of the linear electric machine will be undertaken and the overall generating efficiency of the FPEG system will be calculated by testing the electricity power output. Acknowledgement This project is supported by the Programme of Introducing Talents of Discipline to Universities of China (B12022) and National Nature Science Foundation of China (51006010). We would like to thank the sponsors. References [1] Mikalsen R, Roskilly AP. The design and simulation of a two-stroke free-piston compression ignition engine for electrical power generation. Appl Therm Eng 2008;28(5):589–600. [2] Max Erland. FPEC, Free piston energy converter. In: Proceedings of the 21st electric vehicle symposium & exhibition, EVS, vol. 21; 2005. [3] Atkinson C, Petreanu S, Clark N, Atkinson R et al. Numerical simulation of a two-stroke linear engine-alternator combination, SAE Technical Paper 199901-0921; 1999. [4] Bergman Miriam, Fredriksson Jakob, Golovitchev Valeri I. CFD-based optimization of a diesel-fueled free piston engine prototype for conventional and HCCI combustion. SAE Technical Paper 2008-01-2423; 2008. [5] Zulkifli Saiful A, Karsiti Mohd N, Aziz Rashid A. Starting of a free-piston linear engine-generator by mechanical resonance and rectangular current commutation. In: Vehicle Power and Propulsion Conference, IEEE; 2008. [6] Goldsborough S. Scott, Van Blarigan Peter. A numerical study of a free piston IC engine operating on homogeneous charge compression ignition combustion. SAE Technical Paper 1999-01-0619; 1999. [7] Mikalsen Rikard, Roskilly AP. A review of free-piston engine history and applications. Appl Therm Eng 2007;27(14):2339–52. [8] Zhao Z, Zhang F, Huang Y, Zhao C, Guo F. An experimental study of the hydraulic free piston engine. Appl Energy 2012;99:226–33. [9] Xiao Jin, Li Qingfeng, Huang Zhen. Motion characteristic of a free piston linear engine. Appl Energy 2010;87(4):1288–94. [10] Nandkumar Subhash. Two-stroke linear engine. Dissertation, West Virginia University; 1998. [11] Petreanu Sorin. Conceptual analysis of a four-stroke linear engine. Dissertation, West Virginia University; 2000. [12] Houdyschell David. A diesel two-stroke linear engine. Dissertation, West Virginia University; 2000.
7
[13] Van Blarigan P, Paradiso N, Goldsborough S. Homogeneous charge compression ignition with a free piston: a new approach to ideal otto cycle performance. SAE Paper 1998;982484. [14] http://www4.eere.energy.gov/vehiclesandfuels/resources/meritreview/sites/ default/files/ace008_johnson_2013_o.pdf. [15] Hansson Jörgen. Analysis and control of a hybrid vehicle powered by a freepiston energy converter, Dissertation, Royal Institute of Technology; 2006. [16] Mao JL, Zuo ZX, Lin W, Feng H. Multi-dimensional scavenging analysis of a free-piston linear alternator based on numerical simulation. Appl Energy 2011;88. [17] Mao Jinlong, Zuo Zhengxing, Feng Huihua. Parameters coupling designation of diesel free-piston linear alternator. Appl Energy 2011;88:4577–89. [18] Carter D, Wechner E. The free piston power pack: sustainable power for hybrid electric vehicles. SAE Technical Paper 2003-01-3277; 2003. [19] Ferraril Cornelius, Friedrich Horst E. Development of a free-piston linear generator for use in an extended-range electric vehicle. EVS26 international battery, hybrid and fuel cell electric vehicle symposium. California, May 6–9; 2012. [20] Haag J, Ferrari C, Starcke J, Stöhr M et al. Numerical and experimental investigation of in-cylinder flow in a loop-scavenged two-stroke free piston engine, SAE Technical Paper 2012-32-0114; 2012. [21] Kock F, Haag J, Friedrich H. The free piston linear generator – development of an innovative, compact, highly efficient range-extender module, SAE Technical Paper 2013-01-1727; 2013. [22] Schneider Stephan, Frank Rinderknecht, Friedrich Horst E. Design of future concepts and variants of the free piston linear generator. inecological vehicles and renewable energies (EVER). In: 2014 ninth international conference on, IEEE; 2014. p. 1–8. [23] Kosaka H, Akita T, Moriya K, Goto S et al. Development of free piston engine linear generator system part 1 – investigation of fundamental characteristics. SAE Technical Paper 2014-01-1203; 2014. [24] Goto S, Moriya K, Kosaka H, Akita T et al. Development of free piston engine linear generator system part 2 – investigation of control system for generator. SAE Technical Paper 2014-01-1193; 2014. [25] Aichlmayr HT. Design considerations, modeling, and analysis of microhomogeneous charge compression ignition combustion free-piston engines. Dissertation, University of Minnesota; 2002. [26] Jia B, Zuo Z, Feng H, Tian G, Roskilly AP. Development approach of a sparkignited free-piston engine generator. SAE Technical Paper 2014-01-2894; 2014. [27] Jia B, Zuo Z, Tian G, Feng H, Roskilly AP. Development and validation of a freepiston engine generator numerical model. Energy Convers Manage 2015;91:333–41. [28] Mikalsen R, Roskilly AP. The control of a free-piston engine generator. Part 1: fundamental analyses. Appl Energy 2010;87(4):1273–80. [29] Mikalsen R, Roskilly AP. The control of a free-piston engine generator. Part 2: engine dynamics and piston motion control. Appl Energy 2010;87(4):1281–7. [30] Mikalsen R, Jones E, Roskilly AP. Predictive piston motion control in a freepiston internal combustion engine. Appl Energy 2010;87(5):1722–8.
Please cite this article in press as: Jia B et al. An experimental investigation into the starting process of free-piston engine generator. Appl Energy (2015), http://dx.doi.org/10.1016/j.apenergy.2015.02.065