An experimental study of a single-piston free piston linear generator

Abstract

Free piston linear generator (FPLG) is a promising range extender for the electrical vehicle with unparallel advantages, such as compact structure, higher system efficiency, and reduced maintenance cost. However, due to the lack of the mechanic crankshaft, the related piston motion control is a challenge for the FPLG which causes problems such as misfire and crash and limits its widespread commercialization. Aimed at resolving the problems as misfire, a single-piston FPLG prototype has been designed and manufactured at Shanghai Jiao Tong University (SJTU). In this paper, the development process and experimental validation of the related control strategies were detailed. From the experimental studies, significant misfires were observed at first, while the FPLG operated in natural-aspiration conditions. The root cause of this misfire was then identified as the poor scavenging process, and a compressed air source was leveraged to enhance the related scavenging pressure. Afterward, optimal control parameters, in terms of scavenging pressure, air-fuel equivalence ratio, and ignition position, were then calibrated in this charged-scavenging condition. Eventually, the FPLG prototype has achieved a continuous stable operation of over 1000 cycles with an ignition rate of 100% and a cycle-to-cycle variation of less than 0.8%, produced an indicated power of 2.8 kW with an indicated thermal efficiency of 26% and an electrical power of 2.5 kW with an overall efficiency of 23.2%.

Graphical abstract

Keywords

free piston linear generator (FPLG) / charged scavenging / engine control / misfire / stable operation

1 Introduction

1.1 Background

With the rapid development of the world’s economy, the population of passenger cars has been skyrocketed nowadays. The huge vehicle population inevitably generates serious pollution problems, including greenhouse gas (GHG), CO, unburned hydrocarbon (HC), and NO_x, etc. Those pollutants not only affect the environment and climate adversely but also significantly harm human health [1‒4]. To mitigate the environmental and health problems caused by vehicle emissions, a series of policies and emission regulations have been issued in China, and the development of battery electrical vehicles (BEV) is remarkably encouraged in parallel. However, the BEV still faces numerous technical challenges, including the so-called “mileage anxiety” and lack of existing infrastructure [5‒7]. To solve these challenges, significant efforts have to be committed to advanced battery technology and related infrastructure investment. Consequently, a range extender for electrical vehicle receives increasing attention from the public and is considered as an alternative solution which can be realized in the near future.

Though the range extender usually refers to a conventional internal combustion engine (ICE), a novel engine architecture, i.e., free piston engine (FPE), is also investigated nowadays. In general, the FPE is an ICE without the mechanical crankshaft and thus its piston motion is free [8]. Compared to the conventional ICE, the unique architecture of the FPE has the following unparalleled advantages [9‒14]. First, due to the freedom of its piston motion, FPE is able to adjust the compression ratio on the fly and thus optimally controls the combustion of various fuels. In addition, linear piston motion in the FPE significantly reduces the friction loss caused by the lateral force and then substantially enhances the related mechanical efficiency. Moreover, the overall structure of the FPE is simple and compact, which reduces both the manufacturing cost and the maintenance cost.

1.2 Literature review

The history of the FPE can be dated back to 1928 when Pateras published the first related patent [15]. From the 1930s to the 1960s, due to its higher thermal efficiency (> 40%) compared to crankshaft ICE and gas turbine, the FPE was extensively used as free piston air compressors and free piston gas generators [16‒18]. However, limited by the technologies at that time, FPE was criticized for its inaccurate piston motion control and unstable operation. In the past 20 years, the R&D efforts of the FPE have grown again, mainly due to the substantial development of the mechatronic system and related dynamics control. Extensive studies related to the hydraulic FPE [19‒23] and the electrical FPE, also named as free piston linear generator (FPLG), have been conducted, the latter of which becomes even more attractive as a range extender for electrical vehicles (EV) since it combines an FPE with a linear generator directly and possesses a higher energy efficiency and power density [24].

West Virginia University (WVU) is one of the early institutions that started FPLG research. Clark et al. [25‒29] first manufactured an ignited dual-piston prototype FPLG with a cylinder bore of 36.5 mm and a maximum piston stroke of 50 mm. The related piston motion was fully regulated through fuel injection amount and timing control. The prototype could produce an output power of 316 W at 23.1 Hz, but with a huge cycle-to-cycle variation. The main problems, the researchers believed, stem from the insufficient fuel injection pressure and the limited scavenging efficiency. Leveraging this knowledge, a second-generation FPLG was manufactured in WVU and the corresponding startup test was conducted as well [30].

Since the 1990s, the Sandia National Laboratory has also conducted research on FPLG for efficient hybrid vehicle power plants. Blarigan et al. [31‒33] built up a dual-piston prototype FPLG in 1998, with a bore of 70 mm and a maximum piston stroke of 164 mm. The prototype makes the homogeneous charge compression ignition (HCCI) combustion mode possible. A related experimental study proved that the FPLG could employ various fuels, including hydrogen, natural gas, and ammonia, and delivered an overall system efficiency of more than 50% of (more than 56% thermal efficiency and 96% power generation efficiency). In 2008, this prototype was modified into an opposed-piston FPLG to enhance its stable operation under lean-burn conditions [34].

In 2006, Mikalsen [35] of Newcastle University designed a single-piston FPLG with a turbocharger. The prototype consisted of a combustion cylinder with a bore of 131 mm, a rebound chamber with a bore of 150 mm, and a linear alternator. Other components included an exhaust gas turbocharger, a common rail fuel supply, and a direct fuel injector. A corresponding dynamic model was also developed. The simulation results showed that at a compression ratio of 15: 1 and a charged intake air of 1.68 bar, the prototype could produce an output of 44.4 kW with an overall efficiency of up to 42%.

The German Aerospace Center (DLR) has been conducting FPLG-related research since 2012. At first, Ferrari et al. [36] and Rinderknecht & Kock [37] developed a single-piston FPLG and realized a “quasi-independent operation” with the assistance of a hydraulic device. The prototype consisted of a single combustion cylinder, a linear alternator, and an air spring. The bore of the combustion cylinder and air spring were both 82.5 mm and the piston stroke was 40‒95 mm. The prototype was equipped with two overhead intake valves and exhaust valves and a direct fuel injection system, which could generate a power of 10 kW at 21 Hz. Kock et al. [38‒40] conducted further research on the prototype and achieved a completely independent operation in 2013, increasing the power to 12 kW. They further designed a power module that could be applied to hybrid vehicles. The module was composed of two single-piston FPLGs in parallel, with a targeted output power of 30 kW at 50 Hz.

Toyota also revealed its first-generation FPLG prototype in 2014 [41,42], which adopted a single-piston architecture. An innovative W-shaped piston was developed in the prototype, which made the air spring bore larger than the combustion cylinder bore, thereby reducing heat transfer loss and ensuring sufficient cooling and lubrication. With an effective piston motion control system, the prototype could operate stably for more than 4 h. In the simulation study, the prototype could reach an output power of 10.4 kW and an overall efficiency of 36.2% in spark-ignition mode, and an output power of 12.7 kW and an overall efficiency of 42.0% in premixed-charged compression ignition (PCCI) mode. In 2016, Toyota manufactured the second-generation prototype [43] and proposed a new simple harmonic control algorithm that kept the top and bottom dead center (TDC and BDC) error of the piston within 1 mm. Such a control algorithm remarkably enhanced the stability of the FPLG.

Beijing Institute of Technology also conducted research related to the FPLG. Wang [44] designed an ignited double-piston prototype FPLG with a bore of 34 mm. The test results showed that the FPLG could only operate for a few cycles without the assistance of the linear alternator. Afterward, a piston motion detector and a state switch control were developed accordingly. With the assistance of the above control methods, a second generation of the ignited double-piston FPLG [45] as well as a diesel FPLG [46] was manufactured and tested in start mode.

In Norwegian University of Science and Technology, the design and experimental testing of a free-piston engine with two stroke and a single piston was reported [47]. A dynamic model without differential equation and controller was established, and a scavenging pump for a two-stroke FPE was designed.

In Shanghai Jiao Tong University (SJTU), extensive FPLG research was also conducted through simulation and experiments. For example, Li [48] built up a dual-piston FPLG with a cylinder bore of 40.8 mm and experimentally investigated the potential root cause for the misfire. Enlightened by this study, more studies were performed to design an active fuel injection control strategy aimed to realize the stable operation of the FPLG [49,50].

1.3 Aims and methodology

In this paper, a recent study of the FPLG in SJTU was presented. This study leverages the knowledge learned from previous experiences and finally achieves the stable operation of a single-piston FPLG with calibrated fuel injection control. This paper is contributive because a single-piston FPLG, including an air-spring and a scavenging pump, was manufactured and integrated with an active fuel injection control system. In addition, the root cause of misfire in the FPLG was identified through extensive experimental tests. Moreover, an appropriate charged pressure and an optimal fuel injection strategy, in terms of injection amount and timing, were determined to significantly reduce the cycle-to-cycle variation. Furthermore, more than 1000 cycles of stable operation were achieved in the prototype FPLG and the corresponding performance was characterized.

2 Test setup

The FPLG under investigation in this study is within a single-piston architecture. Unlike the dual-piston architecture [44,46,48,49], the single-piston FPLG adopts a recovery system (air spring chamber in this study) to replace a combustion cylinder on one side, which reduces the cycle-to-cycle variation significantly and ensures stable operation with simple control [41,42].

Fig.1 and Fig.2 are the schematic and picture of the FPLG prototype, respectively. As can be seen, the FPLG consists of three systems, namely a combustion chamber, an air spring chamber, and a linear generator. In the combustion chamber, a direct fuel injector and a spark plug are mounted on the cylinder head, which realize gas direct-injection and spark-ignition combustion. When the combustion occurs, the abruptly increased air pressure in the combustion chamber will push the piston to move to the BDC. During this left-forward piston motion, the linear generator will produce the electrical output power and the air spring will absorb the rest of the kinetic energy of the piston and convert it to the stored potential energy. After the piston reaches the BDC, the compressed air in the air spring chamber will push the piston back to the TDC and form the next combustion. The above cycle will be repeated during the FPLG continuous operation.

In addition, a compressor, integrated with a buffer air tank, is used as the air supply for the FPLG. This device is connected to the air spring chamber with an intake port and the combustion chamber with an intake port and an exhaust port. Two pressure regulators are also employed to these two air paths. The regulator to the air spring chamber can change the inside air pressure and adjust the elastic coefficient of the air spring. The other one is used to adjust the charged pressure of the fresh air flowing into the combustion chamber. Theoretically, the burned gas inside the combustion chamber can be swept out completely by applying sufficient pressure difference between the intake pressure and the exhaust backpressure. The FPLG prototype is also connected to an external fuel supply system, which utilizes a 2-L pressured accumulator to provide an adjustable fuel injection pressure (up to 207 bar). Other specifications of the FPLG prototype are listed in Tab.1.

3 Control system and strategy

3.1 Control system

Fig.3 shows the general schematic of the control system that is utilized in this study. As can be seen, the entire system consists of multiple sensors, a controller, and related actuators.

The CompactRIO (cRIO)-9039 embedded controller developed by National Instruments (NI) was selected as the main controller in this study due to its high performance and modularity. The cRIO controller can be integrated with different modular according to various functional requirements. Specific to this study, the NI-9205 data acquisition module is used to collect all the signals of sensors, the NI-9751 direct injector driver module controls the fuel injector in the FPLG, and the NI-9401 digital signal module is able to output 5 V transistor-transistor logic (TTL) signals to control the spark plug as well as the solid-state relays. The specifications of the main components in the control system are listed in Tab.2.

3.2 Control strategy

3.2.1 State-switching control

From the start of moving the piston to continuous generation of electrical power, the FPLG undergoes three operational states, i.e., the motoring state, the transition state, and the generation state. To coordinate these states, a state-switching control is developed, as shown in Fig.4.

During the motoring state, the linear electric machine works as an electrical motor to produce reciprocated electromagnetic force which drives the piston move back and forth. When the piston moves to the TDC, the in-cylinder air will be compressed but the fuel injector and spark plug are not triggered. After several cycles, the air pressure in the combustion chamber will reach a prescribed threshold (12 bar in this study) and then the FPLG enters the transition state. In the transition state, the linear electric machine still works as an electric motor, but the fuel injector and the spark plug will function as designed. The transition state is completed if the air-fuel mixture in the combustion chamber is successfully ignited. Given this condition, the FPLG enters the stable combustion state or the so-called generation state, while the linear electric machine is disconnected from the motor controller but linked with a three-phase rectifier bridge. In this way, the FPLG produces the electrical output power from the combustion of the fuel. It is worth noting that a misfire monitor (through the in-cylinder pressure in the combustion chamber) is also integrated with the state-switching control. If a misfire is captured in the generation state, the state-switching control will regulate the FPLG back to the transition state until the air-fuel mixture is ignited successfully again.

Specifically, the state-switching control is realized through the two solid-state relays. As shown in Fig.5, in the moving and the transition state, Relay 1 is closed and Relay 2 is disconnected; while in the stable combustion state, Relay 1 is then disconnected and Relay 2 is closed.

3.2.2 Control strategy in each state

As can be seen in Fig.6, the detailed control strategies in the motoring state and the generation state are different. It is worth noting that in the transition state, both control strategies will be employed simultaneously. In addition, the following definitions are kept for all the control strategies: the midpoint of the maximum displacement of the piston is set as the origin of the coordination, and the rightward direction (piston moves toward the TDC) is set as the positive direction.

As shown in Fig.6, in the motoring state, the electromagnetic thrust force applied to the piston is switched between the positive and the negative directions, and the switch points are determined by predefined displacements S− and S+. On top of that, the absolute value of the negative thrust force is designed to be greater than the positive counterpart. This difference is used to coordinate with the air-spring force, which also provides the thrust force while the piston moves to the right, but the resistant force while the piston moves in the opposite direction. The value of the electromagnetic thrust force is adjusted by the input current to the electrical motor, with the thrust coefficient as 93 N/A.

In the generation state, the air-fuel mixture in the combustion chamber can be ignited and the subsequent combustion generates sufficient force to push the piston to the BDC. In this process, no control is applied, and the free-piston motion triggers the linear generator to produce the electrical power and compress the air-spring. After the piston reaches the BDC, the compressed air-spring pushes the piston to move back to the right, and then specific controls are employed to adjust the fuel injection timing and the spark timing. Both timings are activated according to the prescribed piston displacements.

4 Experimental results and discussion

4.1 Data processing

4.1.1 Ignition rate

Fig.7 shows a test sample including both fire and misfire occurrences. Apparently, the in-cylinder pressure and the piston displacement profiles in these two conditions are distinct. As shown in Fig.7, the in-cylinder pressure is generally higher than 20 bar in the fire events, while the pressure is less than 16 bar within the misfire. Beyond, the piston can only expand around −30 mm in the misfire condition, while the piston movement can reach farther than −40 mm in the normal operation. An appropriate approach to determine the misfire event is to check the difference between the compress process and the expansion process within one cycle through the P-V diagram. In the misfire cycle, the expansion line is lower than the compression line due to air leakage and heat transfer loss. However, in the normal cycle, the expansion line is significantly higher than the compression line due to the combustion. Using this approach, the ignition rate can be easily calculated as

(1)R_{\text{igni}}=\frac{N_{\text{fire}}}{N_{\text{cycle}}},

where R_igni is the ignition rate, N_fire is the number of cycles that catch the fire after the first successful ignition, and N_cycle is the total number of the cycles after the first successful ignition.

4.1.2 Cycle-to-cycle variation

Cycle-to-cycle combustion variation is an important feature possessed by the spark-ignition ICE. Even when the engine works in a stable condition, the combustion process of this cycle and the next cycle is still changing, in terms of the pressure curve, flame propagation, and engine output power [2].

In general, the cycle-to-cycle variation is used as an indicator to measure the consistency of each cycle. For parameter x, the related cycle-to-cycle variation is calculated as

(2)\text{CO}{\text{V}}_x=({\sigma }_x/\overline{x})\times 100\mathrm{\%},

where COV_x is the coefficient of variation of the parameter x, \overline{x} is the average of x and {\sigma }_x is the standard deviation of x. Note that in this paper, the parameter x is the measured data, such as in-cylinder pressure and operational frequency, from the file cycles. In other words, the signals from the misfire cycle are not included in the calculation.

4.2 Motoring test

In the motoring tests, the fuel injector and the spark plug were not activated, and the air pressure in the air-spring chamber was set at 1.5 bar.

Since the intake air is already charged in the air-spring chamber, the piston will always receive a positive force during the operation. To start the FPLG, the electrical motor needs to apply a negative force on the piston to counter the air-spring force and compresses the air-spring to a certain level. In this way, sufficient potential energy can be stored in the air-spring chamber. After the negative thrust force is removed, the compressed air in the air-spring chamber will push the piston back to the TDC and trigger the subsequent combustion.

The maximal thrust force that the electrical motor can achieve is 250 N in this study. For the sake of safety, apply the negative 200 N force at first and the test results show that the peak in-cylinder pressure in the combustion chamber can only reach 6 bar. Enhancing the force value to negative 240 N, the related peak pressure is around 7 bar. This value is still lower than the required peak value, which makes successful spark ignition possible according to Ref. [49].

Consequently, a positive thrust force from the linear electric machine is added to the piston while it moves to the TDC. As a result, both air-spring force and this positive thrust force will work together to push the piston further and compress the in-cylinder air more aggressively. Fig.8 shows the FPLG motoring test while negative 200 N and positive 50 N thrust forces are applied alternatively as the start forces. It is observed clearly from Fig.8 that after adding the positive thrust forces, the piston displacement can reach 42 mm and the corresponding in-cylinder pressure in the combustion chamber remains at 14 bar after 6 motoring cycles.

4.3 Natural-aspirated combustion tests

Preliminary test parameters of the combustion test were first designed based on previous experiences and literature, as listed in Tab.3. For the sake of safety, the fuel-air equivalence ratio is set as 0.7 in the test. The actual fuel injection amount is derived by multiplying the above air-fuel equivalence ratio to the mass of intake air calculated through the in-cylinder pressure. The fuel injection timing is set as the time when the piston reaches 13 mm from the midpoint of the stroke, while the exhaust port is just closed. The spark timing is first set as the time when the piston reaches 33 mm according to Ref. [49]. The fuel under investigation is commercial No. 92 gasoline.

Fig.9 shows the corresponding experimental result of the preliminary natural-aspirated combustion test in the FPLG prototype. Significant cycle-to-cycle variation is observed. In addition, if the air-fuel mixture is ignited, the in-cylinder pressure in the combustion chamber can reach 24‒30 bar, as shown in Fig.9. However, after each firing cycle, there is always a misfire cycle followed, where the peak in-cylinder pressure is only about 13 bar, close to the pressure rise during the pure compression. Further investigation on the entire test reveals that one or two misfires always occur between two successful ignitions, and the total ignition rate of the test is only 47.5%.

It is widely believed that misfires in the FPLG may stem from inappropriate fuel-air ratio, early or later spark timing, malfunction of the air-spring system, and inefficient scavenging process [46, 48‒50]. To identify the corresponding root cause, the fuel-air equivalence ratio and the sparking timing are varied respectively during the following tests, and the related operations of FPLG are characterized.

4.3.1 Fuel-air equivalence ratio

Since the FPLG experiences significant misfire at a fuel-air equivalence ratio of 0.7, a range of the fuel-air equivalence ratios were then tested to enhance the related ignition rate.

Fig.10 shows the test results at different fuel-air (F/A) equivalence ratios. Note that the max peak pressure indicates the maximal in-cylinder pressure obtained in the test of a specific equivalence ratio, and the average peak pressure refers to the mean value of the peak in-cylinder pressures from all the fire cycles. Intuitively, the in-cylinder pressure (from both peak and average values) is indeed increased with a higher fuel-air equivalence ratio. However, the highest ignition rate is achieved when the equivalence ratio is 0.7, and when the ratio is over 0.7, the ignition rate is reduced.

Another interesting observation is that the maximal peak pressure usually exists in the first combustion cycle. Fig.11 shows that this phenomenon happens in both the 0.75 ratio test and the 0.9 ratio test. Enlightened by this observation, it is figured out that the misfire is mainly caused by insufficient gas scavenging in the prototype FPLG. With poor scavenging, the exhaust cannot be discharged in a short time and a large amount of the residual burned gas is left in the chamber, which causes the misfire in the next cycle. On the contrary, since the combustion chamber is filled with fresh air in the first cycle, the fuel can be completely burned and thus generates aggressive combustion.

The scavenging issue also explains why the ignition rate first rises and then reduces as the fuel-air equivalence ratio increases. During the misfire cycles, the control system still activates the fuel injection and part of the injected fuel will remain in the chamber due to the poor scavenging. Therefore, the next successful ignition has ignited the fuel injected in the current cycle and the remaining fuel from previous misfire cycles. Only when the above fuel combination has the correct equivalence ratio with the intake of fresh air, can the combustion occur in the FPLG. Therefore, when the equivalence ratio is too low (less than 0.65), abundant fresh air is captured in the combustion chamber and the correct fuel amount needs to be accumulated through multiple misfire cycles to trigger the combustion; when the equivalence ratio is too high (over 0.7), more fuel is injected and remain during the misfire cycles while substantial fresh air is required to form the next combustion. Both conditions above will inevitably reduce the ignition rate.

4.3.2 Spark timing control

When the fuel-air equivalence ratio is determined as 0.7, which can achieve the maximal ignition rate, various sparking timings are also tested to further mitigate the cycle-to-cycle variation. Note that in the tests all the sparking timings are before the TDC to give sufficient time to complete the entire combustion.

Fig.12 represents the experimental results of the FPLG natural-aspiration combustion with different sparking ignition positions. All the other test parameters are fixed as listed in Tab.3. Similar to the influence of equivalence ratios, the ignition rate also increases first and then decreases with increased sparking ignition positions. 50% of ignition rate is achieved while the ignition position is around 30 mm. Advancing this position will lead to a low in-cylinder temperature and pressure when the spark plug is activated and cause a subsequent misfire; delaying this position will make the majority of the combustion occur during the expansion stroke and result in an unstable combustion or even misfire. In addition, no matter what ignition position is employed, the ignition rate of the FPLG cannot exceed 50% in the natural-aspiration combustion mode. In other words, even the ignition position can affect the ignition rate, but it is not the root cause for the misfire in the prototype FPLG.

Another observation from Fig.12 is that the average peak in-cylinder pressure is reduced with delayed ignition position. The main reason for this is that more fuel will be burned during the expansion stroke if the ignition position is closer to the TDC. In this case, the pressure rise from combustion is partially offset by the increasing combustion chamber volume. Further delay in the ignition timing (for example, 38 mm) may lead to a misfire.

Based on the analysis in Sections 4.3.1 and 4.3.2, the optimal control parameters for the prototype FPLG under natural-aspiration combustion conditions are similar to those listed in Tab.3, while the only exception is that the spark ignition position is changed from 33 mm to 30 mm. However, even in this best scenario, the ignition rate of the FPLG is only 50% and significant cycle-to-cycle variation is still observed. Consequently, the FPLG is then investigated under super-charged conditions.

4.4 Charged-scavenging combustion tests

4.4.1 Effect of charged-scavenging pressure on the ignition rate

Generally, the discharge of exhaust and the intake of fresh air is overlapped in a two-stroke ICE, which is named as the scavenging process. To discharge the exhaust completely, the two-stroke ICE usually compresses the intake air through a scavenging pump or a crankcase [2]. Such a compression process cannot be achieved easily in the FPLG due to the lack of the crankshaft mechanism. Without such a scavenging pump, the air exchange in the FPLG only depends on the free exhaust and the natural suction of fresh air. There is no doubt that the scavenging efficiency is limited (< 45% according to Ref. [44]) and a huge amount of residual exhaust remains in the chamber.

To resolve this challenge, a compressed air source (see Fig.1) is connected directly to the combustion chamber and a regulator is used to adjust the charged air pressure. Intuitively, the compressed fresh air is expected to discharge the most of exhaust, substantially enhance the scavenging efficiency, and ensure the stable operation of the FPLG.

To experimentally prove the above capability, multiple tests were conducted at different scavenging pressures and fuel-air equivalence ratios. Three scavenging pressures, i.e., 1 bar (natural aspiration), 1.2 bar, and 1.4 bar, were tested with a range of equivalence ratios from 0.7 to 1.1, all the other test parameters are identical to those listed in Tab.3, except that the ignition position is set to 30 mm again. The corresponding ignition rates achieved from these tests are shown in Fig.13. Charged intake air will significantly enhance the ignition rate. At a pressure of 1.2 bar, the ignition rates at a ratio of 0.9 and 1.0 are around 50%, while the corresponding ignition rates in natural aspiration conditions are both under 30%. If the scavenging pressure is further increased to 1.4 bar, the ignition rate for 0.9, 1.0, and 1.1 ratios are even 86%. 88%, and 69%, respectively. With such a high ignition rate, most of the misfires can be considered as nonexistent, and the exhaust in the combustion chamber is effectively removed during scavenging.

Encouraged by the above results, experiments at a wide range of scavenging pressure ware also conducted at the stoichiometric equivalence ratio, as shown in Fig.14. An obvious enhancement in the ignition rate can be seen while the scavenging pressure increases to 1.4 bar. At the same time, the cycle-to-cycle variation is also remarkably reduced. However, if the scavenging pressure is further raised, neither the ignition rate nor the cycle-to-cycle variation is improved. This phenomenon reveals that a scavenging pressure of 1.4 bar is sufficient for the scavenging purpose and other causes should be identified for the misfire and unstable operation of the FPLG.

In fact, when the intake air pressure raises to 1.4 bar, the piston movement and corresponding in-cylinder pressure dynamics of the FPLG are completely different from their counterpart under natural-aspiration conditions, as can be seen in Fig.15.

The piston speed is slower, and the operational frequency is much lower in the natural aspiration test, compared to the case of charged scavenging. From the identical ignition position (30 mm) to the corresponding TDC, it takes 4.6 ms for the piston to move in the case of natural aspiration, while in the case of 1.4 bar pressure, it takes only 2.0 ms. Such a time difference substantially affects the subsequent combustion process, as can be seen in Fig.15(b). The 4.6 ms duration is sufficient for the air-fuel mixture in the natural-aspiration condition to complete the majority of the combustion process before the TDC. Therefore, most of the combustion energy has been released and the corresponding peak pressure (around 26 bar) comes just slightly later than the TDC. However, in the case of 1.4 bar scavenging pressure, the 2.0 ms duration is too short to ignite most of the fuel. Without the combustion before the TDC, the in-cylinder pressure cannot push the piston back, and instead, the piston collides with the safety buffer in the FPLG. After the collision, the piston moves to the BDC, and the main combustion is then produced during the piston expansion stroke, which also leads to a limited peak pressure (only 22 bar) occurring much later than the TDC. On top of that, due to the unique characteristics of the piston motion of the FPLG, the piston will abruptly accelerate away from the TDC and cause the combustion chamber to rapidly increase in volume, which is unfavorable for flame propagation. All in all, in the case of a scavenging pressure of 1.4 bar, an ignition position of 30 mm is too late to form the stable operation of the FPLG.

4.4.2 Multiple tests with different ignition positions

Since an ignition position of 30 mm is too late for the case of a scavenging pressure of 1.4 bar, a new appropriate ignition position needs to be calibrated. Consequently, a set of tests with different ignition positions were conducted. Note that in these tests, the scavenging pressure is kept as 1.4 bar and the fuel-air equivalence ratio is set to 1.0. Other parameters remain the same as those listed in Tab.3.

Five ignition positions, from 26 mm to 30 mm with 1 mm increments, were tested with continuous 300 cycles, respectively. The corresponding ignition rate and the average peak in-cylinder pressure are depicted in Fig.16. It can be seen from the Fig.16 that the FPLG prototype achieves a 100% ignition rate while the ignition position is within 27‒29 mm, realizing continuous robust operation (cycle-to-cycle variation is around 0.8%). When the ignition positions are 26 mm and 30 mm, the related ignition rates are 93.77% and 88.68%, respectively. The 26 mm ignition position places a remarkable difficulty to cold start the FPLG, while the 30 mm ignition position leads to a piston collision and abnormal combustion due to the late ignition.

Detailed investigations, in terms of P-V diagram and piston velocity, were then proceeded with the 27 mm, 28 mm, and 29 mm test scenarios, as shown in Fig.17.

From Fig.17(a), it is clearly observed that the 27 mm ignition position produces the maximal peak pressure as well as the most indicated output work among these three cases. The main reason for this is that the 27 mm ignition position triggers the combustion early and ensures that the majority of the combustion occurs before the TDC. Later ignition timing inevitably delays the combustion energy release to the expansion stroke and diminishes the indicated output work. It is also obviously seen from Fig.17(b) that the 27 mm ignition position leads to a fast piston motion and a higher operational frequency. These characteristics also benefit the in-cylinder spark-ignition combustion through enhanced turbulence. To summarize, ignition position is a critical parameter for the stable operation of the FPLG. However, the range of available ignition positions is usually narrow [41], and therefore, extensive efforts are required to design an appropriate ignition system and the related control strategy.

4.4.3 Continuous stable operation of the FPLG prototype

Enlightened by the above analyses on the scavenging pressure and ignition position, the final test parameters to achieve continuous stable operation of the FPLG prototype are listed in Tab.4.

With the test parameters in Tab.4, the FPLG prototype has achieved more than 1000 cycles of continuous and stable operation after the motoring and transition states, as shown in Fig.19. The prototype produced an indicated power of 2.8 kW with an indicated thermal efficiency of 26% and an electrical power of 2.5 kW with an overall efficiency of 23.2%.

The operational frequency curve of the first 30 cycles (start from motoring) of the FPLG during the test is also illustrated in Fig.19. As can be seen, during the motoring state, the operational frequency of the FPLG is gradually increased and reaches a plateau after six cycles, indicating the stable condition of pure compression (around 22 Hz). Then, the FPLG enters the transition state till the ninth cycle, while the air-fuel mixture in the combustion chamber has been first ignited successfully. The combustion force enhances the frequency from 22 Hz to 43 Hz within one cycle. With the further operation of the FPLG, the frequency is gradually stable and after the 20th cycle, the operational frequency of the FPLG is stabilized around 43.2 Hz, with a cycle-to-cycle variation of only 0.8%.

5 Conclusions

In this paper, a prototype of single-piston FPLG was designed and manufactured. Specific control strategies for the three operational states of the FPLG, i.e., the motoring, the transition, and the combustion state, were developed and experimentally validated. Specifically, alternative electrical thrust forces, with negative 200 N and position 50 N, were designed to guarantee the start of the FPLG through motoring operation. The FPLG prototype was first tested in natural-aspiration conditions, while significant cycle-to-cycle variation and a large portion of misfire cycles (with only 50% ignition rate) were observed even after the optimal fuel-air equivalence ratio and ignition position were determined. The root cause of this misfire was identified through extensive experiments, which was the poor scavenging process with limited scavenging efficiency. Enlightened by this analysis, a compressed air source was directly connected to the combustion chamber of the FPLG to provide a charged scavenging process. After calibrating the related control parameters, e.g., equivalence ratio (1.0), scavenging pressure (1.4 bar), and ignition position (27 mm), in this charged scavenging environment, a continuous stable operation of the FPLG was achieved with more than 1000 cycles at an operational frequency of 43.2 Hz. The corresponding ignition rate was 100% (no misfire events) and the cycle-to-cycle variation was less than 0.8%. Detailed investigations were also conducted on the piston motion characteristics and the subsequent combustion process in this stable operation of the FPLG.

In the future, an integrated scavenging pump and related air path design will be conducted in the next-generation FPLG to eliminate the use of the external compressed air source. In addition, an upgrade on the control system is also planned to realize active accurate control on the piston motion of the FPLG to further enhance the system performance.

References

[1] Huang Z, Zhu L, Li A. . Renewable synthetic fuel: turning carbon dioxide back into fuel. Frontiers in Energy, 2022, 16(2): 145–149

[2] Kizu R, Okamura K, Toriba A. . Antiandrogenic activities of diesel exhaust particle extracts in PC3/AR human prostate carcinoma cells. Toxicological Sciences, 2003, 76(2): 299–309

[3] TheUnited States Environmental Protection Agency. Health and environmental effects of particulate matter. 2018–04–20, available at website of epa-gov

[4] Bahrampour H, Beheshti Marnani A K, Askari M B. . Evaluation of renewable energies production potential in the Middle East: confronting the world’s energy crisis. Frontiers in Energy, 2020, 14(1): 42–56

[5] Wang J. The number of public charging stations in China ranks the first in the world. Machinery and Electronics Business (Chinese version), 2018, 0(1): 28–28

[6] Sun Y, Liu X. Impact of mileage anxiety on intention to use pure electric vehicle. Journal of Wuhan University of Technology (Transportation Science and Engineering), 2017, 41(1): 87–91

[7] Habib S, Khan M M, Abbas F. . A framework for stochastic estimation of electric vehicle charging behavior for risk assessment of distribution networks. Frontiers in Energy, 2020, 14(2): 298–317

[8] Mikalsen R, Roskilly A P. The design and simulation of a two-stroke free-piston compression ignition engine for electrical power generation. Applied Thermal Engineering, 2008, 28(5–6): 589–600

[9] Jia B, Smallbone A, Zuo Z. . Design and simulation of a two- or four-stroke free-piston engine generator for range extender applications. Energy Conversion and Management, 2016, 111: 289–298

[10] Jia B, Zuo Z, Feng H. . Effect of closed-loop controlled resonance based mechanism to start free piston engine generator: simulation and test results. Applied Energy, 2016, 164: 532–539

[11] Zhang C, Sun Z. Trajectory-based combustion control for renewable fuels in free piston engines. Applied Energy, 2017, 187: 72–83

[12] Zhang C, Sun Z. Using variable piston trajectory to reduce engine-out emissions. Applied Energy, 2016, 170: 403–414

[13] Zhang C, Li K, Sun Z. Modeling of piston trajectory-based HCCI combustion enabled by a free piston engine. Applied Energy, 2015, 139: 313–326

[14] Lin M, Mou J, Chi C. . A space power system of free piston Stirling generator based on potassium heat pipe. Frontiers in Energy, 2020, 14(1): 1–10

[15] PaterasP R. US Patent, 1 657 641, 1928–01–31

[16] Mikalsen R, Roskilly A P. A review of free-piston engine history and applications. Applied Thermal Engineering, 2007, 27(14–15): 2339–2352

[17] Farmer H O. Free-piston compressor-engines. Proceedings–Institution of Mechanical Engineers, 1947, 156(1): 253–271

[18] NorenO BErwinR L. The future of the FREE-PISTON ENGINE in commercial vehicles. SAE Technical Paper: 580032, 1958 oi:10.4271/580032

[19] Peter A J, Achten J. A review of free piston engine concepts. SAE Transactions, 1994, 103: 1836–1847

[20] TikkanenSLammilaMHerranenM, . First cycles of the dual hydraulic free piston engine. SAE Technical Paper: 2000-01-2546, 2000

[21] Hibi A, Ito T. Fundamental test results of a hydraulic free piston internal combustion engine. Proceedings of the Institution of Mechanical Engineers. Part D, Journal of Automobile Engineering, 2004, 218(10): 1149–1157

[22] Zhao Z, Zhang F, Huang Y. . An experimental study of the cycle stability of hydraulic free-piston engines. Applied Thermal Engineering, 2013, 54(2): 365–371

[23] Li K, Zhang C, Sun Z. Precise piston trajectory control for a free piston engine. Control Engineering Practice, 2015, 34: 30–38

[24] JiaB. Analysis and control of a spark ignition free-piston engine generator. Dissertation for the Doctoral Degree. Newcastle: Newcastle University, 2016

[25] Clark N, Nandkumar S, Atkinson C. . Operation of a smallbore two-stroke linear engine. In: Proceeding of the Fall Technical Conference of the ASME Internal Combustion Engine Division, New York, USA: American Society of Mechanical Engineers, 1998, 31: 33–40

[26] FamouriPCawthorneW RClarkN, . Design and testing of a novel linear alternator and engine system for remote electrical power generation. In: IEEE Power Engineering Society 1999 Winter Meeting (Cat. No.99CH36233), New York, NY, USA: IEEE: 108–112

[27] Cawthorne W R, Famouri P, Chen J. . Development of a linear alternator-engine for hybrid electric vehicle applications. IEEE Transactions on Vehicular Technology, 1999, 48(6): 1797–1802

[28] CawthorneWFamouriPClarkN. Integrated design of linear alternator/engine system for HEV auxiliary power unit. In: IEEE International Electric Machines and Drives Conference (Cat. No.01EX485). Cambridge, MA, USA. IEEE: 267–274

[29] NandkumarS. Two-stroke linear engine. Dissertation for the Doctoral Degree. West Virginia: West Virginia University, 1998

[30] HoudyschellD. A diesel two-stroke linear engine. Dissertation for the Doctoral Degree. West Virginia: West Virginia University, 2000

[31] van BlariganP. Advanced internal combustion electrical generator. In: Proceedings of the 2001 DOE Hydrogen Program Review, Maryland, USA, 2002

[32] van BlariganPParadisoNGoldsboroughS. Homogeneous charge compression ignition with a free piston: a new approach to ideal Otto cycle performance. SAE Technical Paper: 982484, 1998

[33] WuJParadisoNvan BlariganP, . Homogeneous charge compression ignition free piston linear alternator. Office of Scientific and Technical Information (OSTI), 1998

[34] Woo Y, Lee Y J. Free piston engine generator: Technology review and an experimental evaluation with hydrogen fuel. International Journal of Automotive Technology, 2014, 15(2): 229–235

[35] MikalsenR. An investigation into the free-piston engine concept and its potential for high efficiency and low emissions power generation. Dissertation for the Doctoral Degree. Newcastle: Newcastle University, 2008

[36] FerrariCFriedrichH E. Development of a free-piston linear generator for use in an extended-range electric vehicle. In: Proceeding to EVS26 International Battery, Hybrid and Fuel Cell Electrical Vehicle Symposium, Los Angeles, USA, 2012

[37] Rinderknecht F, Kock F. A high efficient energy converter for a hybrid vehicle concept. World Electric Vehicle Journal, 2012, 5(2): 475–481

[38] KockFHaagJFriedrichH E. The free piston linear generator - development of an innovative, compact, highly efficient range-extender module. SAE Technical Paper: 2013-01-1727, 2013

[39] Kock F, Heron A, Rinderknecht F. . The free-piston linear generator potentials and challenges. MTZ Worldwide, 2013, 74(10): 38–43

[40] HaagJKockFChiodiM, . Development approach for the investigation of homogeneous charge compression ignition in a free-piston engine. SAE Technical Paper: 2013-24-0047, 2013

[41] KosakaHAkitaTMoriyaK, . Development of free piston engine linear generator system part 1 — investigation of fundamental characteristics. SAE Technical Paper: 2014-01-1203, 2014

[42] GotoSMoriyaKKosakaH, . Development of free piston engine linear generator system part 2–investigation of control system for generator. SAE Technical Paper: 2014-01-1193, 2014

[43] MoriyaKGotoSAkitaT, . Development of free piston engine linear generator system part 3–novel control method of linear generator for to improve efficiency and stability. SAE Technical Paper: 2016-01-0685, 2016

[44] WangD. Study of free piston linear generator’s stable operation. Dissertation for the Doctoral Degree. Beijing: Beijing Institute of Technology, 2015 (in Chinese)

[45] MaoJ. Numerical simulation and experimental study of free piston linear generator operation. Dissertation for the Doctoral Degree. Beijing: Beijing Institute of Technology, 2011 (in Chinese)

[46] YuanC. Design and characterization of a diesel free piston linear generator. Dissertation for the Doctoral Degree. Beijing: Beijing Institute of Technology, 2015 (in Chinese)

[47] Johansen T A, Egeland O, Aa Johannessen E. . Free-piston diesel engine timing and control — toward electronic cam- and crankshaft. IEEE Transactions on Control Systems Technology, 2002, 10(2): 177–190

[48] LiQ. Study of free piston linear generator. Dissertation for the Doctoral Degree. Shanghai: Shanghai Jiao Tong University, 2011 (in Chinese)

[49] Yang X, Wang Z, Zhu C. . Control system design of a free piston linear generator through LabVIEW. Internal Combustion Engine and Power System (Chinese Version), 2018, 35(01): 7–12 (in Chinese)

[50] Lin B, Xiao J, Zhang L. Numerical simulation model of a free piston engine through Simulink and GT-Power. Internal Combustion Engine and Power System (Chinese Version), 2017, 34(1): 17–23 (in Chinese)

RIGHTS & PERMISSIONS

Higher Education Press 2022