Research and application of over-expansion cycle (Atkinson and Miller) engines – A review

Applied Energy 185 (2017) 300–319

Contents lists available at ScienceDirect

Applied Energy journal homepage: www.elsevier.com/locate/apenergy

Research and application of over-expansion cycle (Atkinson and Miller) engines – A review Jinxing Zhao School of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai, China

h i g h l i g h t s
A review of study and application of over-expansion cycle engine is provided.
Mechanical realizations and real applications of over-expansion cycle are studied.
Some novel strategies for applying ‘‘Atkinson cycle effect” are provided.
Challenges and prospective solutions for using Atkinson cycle engine are discussed.
Primary problems and suggestions for future R&D in potential fields are given.

a r t i c l e

i n f o

Article history: Received 15 June 2016 Received in revised form 19 October 2016 Accepted 20 October 2016

Keywords: Atkinson and Miller cycles Thermal efficiency Load control NOx Knock Hybrid electric vehicle

a b s t r a c t Vehicle electrification has to be addressed to reduce dependence on fossil fuels and meet future emission regulations. Pure electric vehicles still have many limitations, but hybrid vehicles are the optimum transference scheme. An over-expansion cycle (Atkinson or Miller) engine is the most suitable for hybrid vehicles. Compared with a conventional Otto cycle engine, an over-expansion cycle engine can realize a larger expansion ratio and thus, a higher thermal efficiency while maintaining a normal effective compression ratio to avoid the knock. Basics for the Atkinson and Miller cycles are introduced first. An in-depth survey on mechanical realizations for the over-expansion cycle is conducted. Challenges and general recommendations for real applications of these mechanical realizations are presented. After a comprehensive review of the advantages and applications of the ‘‘Atkinson cycle effect” in load control, reducing NOx formation and suppressing the knock, primary problems are discussed and some novel strategies are provided. Prospective technical solutions that handle the reduced effective compression ratio and power density for over-expansion cycle engines are studied and discussed. Finally, in the potential application fields of range-extended electric vehicles and cogeneration plants, a significant problem is presented; the efficiency optimum working points for the engine and generator do not match. A multi-disciplinary design and optimization methodology is provided to resolve the problem. The main objective of this paper is to explore the critical problems, challenges and prospective solutions that push forward broader applications for over-expansion cycle engines. This paper can be used as a critical review of the current state-of-the-art research for over-expansion cycle engines and also as guidance towards future research directions in this domain. Ó 2016 Elsevier Ltd. All rights reserved.

Contents 1.

2.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.1. Background and significance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2. Aim of this paper. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Basics for over-expansion cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Introduction to Atkinson cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2. Atkinson cycle vs. Miller cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

E-mail address: [email protected] http://dx.doi.org/10.1016/j.apenergy.2016.10.063 0306-2619/Ó 2016 Elsevier Ltd. All rights reserved.

301 301 302 303 303 303

J. Zhao / Applied Energy 185 (2017) 300–319

301

Nomenclature ICE TWC EGR PM EV REEV GCR WOT VVT BDC VCR PMEP IVC k TIVC OA GDI ACE PFI AFR

3.

4.

5.

6.

7.

internal combustion engine three-way catalyst converter exhaust gas recirculation particulate matters electric vehicle range extended electric vehicle geometrical compression ratio widely open throttle variable valve timing bottom dead centre variable compression ratio pumping mean effective pressure intake valve closure specific heat ratio in-cylinder temperature at the IVC timing Otto-Atkinson gasoline direct injection Atkinson cycle engine port fuel injection air-to-fuel ratio

GHG CR HCCI NO BEV PHEV ECR LIVC EIVC IMEP MBT Tsoi CReff CA BSFC TDC SOC MCE ETC CI

greenhouse gases compression ratio homogeneous charge compression ignition nitrogen oxide battery electric vehicle plug-in hybrid electric vehicle effective compression ratio late intake valve closure early intake valve closure indicated mean effective pressure minimum spark angle for best torque in-cylinder air temperature at the start of fuel injection effective compression ratio crank angle brake specific fuel consumption top dead centre state of charge Miller cycle engine electric throttling control compression ignition

2.3. Thermodynamic analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods and challenges for realizing the Atkinson cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Atkinson cycle type . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Mechanical realizations and challenges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.1. Survey on the existing mechanisms for realizing the Atkinson cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2.2. Challenges and general recommendations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Applications and problems for the ‘‘Atkinson cycle effect” . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Aid engine load control. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.2. Reduce NOx formation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3. Suppress the knock . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Primary problems and discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Challenges and solutions for using an Atkinson cycle engine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Technical measures and challenges for reduced ECR . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Technical measures and challenges for improving power density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. Challenges and recommendations on transient AFR control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Critical issues and suggestions for future R&D. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.1. Current research direction for ACE and MCE. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.2. Suggestions for future R&D. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1. Introduction In 1882, an English engineer named James Atkinson invented the first ACE [1]. The original ACE, with a long expansion stroke and short intake and compression strokes, was realized with a complex linkage mechanism. High thermal efficiency of the original ACE was achieved at the expense of reduced power density and increased complexity. Minimal attention from the automotive industry was focused on the ACE for several years. In recent years, ACEs realized via VVT technology have been widely applied in hybrid vehicles. Increasing studies involving ACEs have been published. 1.1. Background and significance A shortage of energy sources and climate warming have attracted global attention and become a severe problem that impacts the sustainable development of humans [2]. GHGs are a

303 304 304 305 305 308 309 310 310 311 311 312 312 312 314 315 315 315 316 317 317

main factor leading to global climate warming [3,4]. Reducing the GHGs and harmful emissions (NOx, HC, etc.) from the transport sector, especially from vehicles, is a key factor that eliminates the climate change risk. In Europe, approximately 22% of CO2 emissions are caused by transportation systems [5] while in China, almost 85% of transportation system emissions are caused by vehicles [6]. Vehicle emissions are also the main contributor leading to haze and photochemical smog. Fossil fuels are the main energy source, but these have finite reserves. The U.S. Energy administration estimates that approximately 2/3 of the total petroleum demand is from the transportation sector [3,7]. Assuming daily production remains steady at 63.5 million barrels, the total global petroleum reserves will be consumed in 50 years [7]. It is particularly crucial to decrease the degree of vehicle dependence on fossil fuels. Developing new energy resources and significantly enhancing the energy efficiency of the conventional ICE are both effective paths and are also necessary to control GHGs and harmful emissions.

302

J. Zhao / Applied Energy 185 (2017) 300–319

Since the first ICE was invented, ICE has achieved great success especially in the automotive industry. Demand for fuel savings and environmental protection is imposing a huge challenge on conventional ICEs. More and more academic and government departments have transferred their focus to alternative automotive technology and fuels (such as vehicle electrification, fuel cell, and hydrogen fuel). The conventional ICE technology is thought to be approaching its development limit [8,9]. However, the ICE technology is far from its development plateau. Much effort is being made to improve the fuel efficiency and reduce harmful emissions. The thermal efficiency of diesel engines is excellent, but they have a high noise level and an expensive post treatment system to handle a great deal of soot and PM production. The TWC of gasoline engines is easy to resolve but gasoline engines have a lower mean fuel efficiency level. At present, popular gasoline engine technology mainly focuses on the turbocharged downsized GDI [10–12] and cooled EGR [13–15]. A problem in automotive ICE is that there is huge gap between the maximal power and the most frequent operating loads. One feasible method is a turbocharged downsized engine that is more efficient with partial loads [16]. Increased power output could be achieved through a turbocharger with GDI technology depressing the knock. Turbocharged downsizing combined with GDI could achieve a 10–15% improvement in fuel consumption [9,17]. Further emissions reductions can be achieved by combining with a cooled EGR [9,15]. However, a deeply downsized engine with extremely high intake boost is impeded by the increased exhaust temperature before turbine and knock tendency, especially in low speed ranges [16,18,19]. Greatly enhancing the fuel efficiency of a conventional gasoline engine is restricted. In the long term, the keys for conventional ICE sustainable development are: (1) accordance with emission regulations; (2) new fuels; (3) optimization of the combustion process; and (4) advanced energy saving concepts [20]. Two conflict problems exist when optimizing the combustion process: minimizing fuel consumption and reducing harmful emissions [9]. HCCI is a novel combustion concept that can realize both high fuel efficiency and low in-cylinder NOx and PM formation [9,20–23]. For diesel engines, PM filter and NOx post-treatment equipment are not necessary. For gasoline engines, a much higher thermal efficiency level (almost 50%) may be realized. However, the HCCI’s major problems are [20,22,23]: (1) difficulty controlling the ignition timing especially during transient operating conditions; (2) high HC and CO emissions; and (3) narrow load range. These challenges prevent the commercial development of an HCCI engine. As the global amount of automobiles increases, the EPA’s analysis shows that an annual reduction of 5–6% of GHGs is necessary to achieve the targets: 95 g/km for 2020, 70–75 g/km for 2025, and 50–55 g/km for 2030. The 2020 target can be realized through advanced ICE technology while further targets must depend on vehicle electrification (pure electric and hybrid vehicles) [9]. EV can draw electricity from the electric grid. In regions where electricity is generated with low GHG intensity, EVs can contribute to a significant reduction of energy consumption and GHG emissions [24–26]. The EV’s primary problems are: low battery energy density, thus a short range; long charging time; short battery cycle life; and high cost [25–31]. A joint report by the EPA, CARB and NHTSA [32] shows that the conventional vehicle technologies are more cost effective than any of the other technology options. Moreover, if the upstream electricity generation and distribution are considered in the overall life cycle, the advantages of an EV, including energy savings and emissions reductions, will decrease [33–35]. In this situation, the ICE-based vehicles (including a substantial portion of hybrid electric vehicles) tend to be superior in terms of fuel consumption and GHG emissions. Research in Ref. [36] shows that in countries with low GHG intensity of electricity generation, EVs have a significant contribution to GHG reduction.

In countries such as China and India where electricity is produced dominantly from coal, total well-to-wheel emissions of the most efficient diesel engine are even lower than in EVs. Therefore, the ICE still has significant potential for energy savings and emissions reduction. PHEV integrates nearly all the advantages of conventional ICE vehicles and EVs, which are the best transference scheme from conventional ICE vehicles to EVs [9,36–38]. In 1995, Toyota Company issued the first globally mass produced hybrid electric car, Prius, at the Tokyo Motor Show and brought it to market in 1997 [39]. The EPA and California Air Resources Board named it the cleanest car of the year. In 1999, Honda Company launched its hybrid car Honda Insight [40]. The 2000 Insight was rated as the most efficient gasoline car by the EPA. Ford Company also issued its own hybrid car called the Fusion Hybrid [41]. An ICE in a hybrid vehicle is so crucial that it considerably determines the vehicle fuel consumption and emissions [42–44]. As an example, study in Ref. [45] shows that propulsion electrification of a medium duty truck does not necessarily lead to lower well-to-wheel CO2 emissions such as in a passenger car because higher efficiency diesel engines are generally used in this type of vehicle. However, a diesel engine is not the best for a hybrid vehicle compared to a gasoline engine because of the increased complexity, higher cost and weight. The Otto cycle SI engine cannot use a higher GCR because of the knocking limit; thus, further thermal efficiency improvement is restricted. If the hybrid vehicle such as the Toyota Prius adopts an over-expansion Atkinson cycle SI engine as one of primary power sources, significant reduction in fuel consumption can be achieved [46–48]. Yoshiharu Yamamoto stated that, ‘‘If we hit the 45% efficiency mark, the ICE will long remain a worthy being.” One possible way to increase the efficiency of ICEs is the usage of the Atkinson cycle [49]. The studies in Ref. [50,51] show that in most speed/load ranges, an over-expansion cycle engine combined with optimum valve timing and ECR may achieve higher fuel efficiency than diesel engines. The Toyota Prius is the most typical hybrid car; it is a full hybrid type [48,52–54] that primarily consists of an ACE, a generator, a driven motor and a power split device. A vehicle controller regulates the energy flow from engine combustion and makes all components work in their own high efficiency area, thus maximizing the system efficiency. For example, as shown in Fig. 1, the ACE can achieve a lower BSFC value than the baseline Otto cycle engine. Moreover, the area of the lower BSFC such as 250 g/kW h for the ACE is obviously larger. The ACE is controlled to cross over the low BSFC region. Under a low load, the engine is cut off while the driven motor is working to propel the vehicle. As the load increases, the engine begins to work. If the engine output power is excessive, mechanical energy from the engine is split into two parts: a part used to propel the wheel and another part used to propel the generator to produce electricity that is stored in battery pack. When running a high load or with high speed, if the required power exceeds the engine efficient power or the maximum value, the driven motor can provide the leaving power. The reduced power density for ACE can be compensated by the driven motor. Therefore, ACE is the most suitable for hybrid vehicles. 1.2. Aim of this paper In addition to the ACE described above, MCE is another type of over-expansion cycle engine. In 1947, an American engineer named Ralph Miller patented a MCE [55,56]. This paper was written because no comprehensive review on over-expansion engines has been published to date. The main objective of this paper is to explore the critical problems and challenges and prospective solutions that push forward broader applications for over-expansion cycle engines. First, ACE and MCE basics are introduced. The similarities and differences

J. Zhao / Applied Energy 185 (2017) 300–319

Atkinson

303

operating line for the Atkinson cycle engine

Otto

Fig. 1. Experimental BSFC comparison between Otto and Atkinson cycle engines.

between Atkinson and Miller cycles are clarified. Then, an in-depth survey on the mechanical realizations for the over-expansion cycle is conducted. Challenges and general recommendations on real applications of these mechanical realizations are presented. After a comprehensive review of the advantages and applications of the ‘‘Atkinson cycle effect” in load control, reducing the NOx formation and suppressing the knock, the primary problems are discussed and a novel load control strategy is provided. To consider the problems of reduced ECR and power density for using overexpansion cycle engines, some practical and prospective technical solutions are studied and discussed. Finally, the paper demonstrates a significant problem for the potential applications of over-expansion cycle engines including range-extended electric vehicles and cogeneration plants. A parallel design and optimization methodology based on a multi-disciplinary optimization theory is provided to resolve the problem and draw the thermal efficiency potential of the over-expansion cycle. 2. Basics for over-expansion cycles

barometric pressure. There is no doubt that the Atkinson cycle is more efficient than the Miller cycle because of the increased work output (5–4M–4A–5). However, in a practical ACE, the piston is not necessary for expansion to the barometric pressure. First, a long expansion stroke lengthens the cylinder body, increasing piston friction loss and engine weight. Then, low in-cylinder pressure at the end of the expansion stroke will lead to large exhaust pumping loss because there is not a free exhaust process when the exhaust valves open [49,58]. This conversely discounts the thermodynamic merit of the full-expansion Atkinson cycle. Therefore, the cycle process in a practical ACE is closer to the ideal Miller cycle that is shown in Fig. 2. Despite the fact that in-cylinder gas pressure at 4M is still higher than the barometric pressure, more thermal energy (1–4O–4M–5–1) is drawn from the working charges compared with the Otto cycle. A real MCE generally uses LIVC (or EIVC) to realize the overexpansion cycle such as in many modern ACEs. However, ACE is naturally aspirated while MCE is generally equipped with a super-/turbocharger [55,56]. Thus, the Miller cycle is essentially the same as the Atkinson cycle [59].

2.1. Introduction to Atkinson cycle The original ACE is a type of full-expansion cycle. Fig. 2 presents the P-V and T-S diagrams comparison among the ideal Otto, Miller and Atkinson cycles. 1–2–3–4O–1 is the Otto cycle process, and 1– 2–3–4A–1 is the Atkinson cycle. For the Otto cycle, the expansion stroke equals the compression stroke. At the end of the power stroke, the in-cylinder pressure is obviously higher than the barometric pressure, meaning that working charges still have considerable potential to make useful work. While the Atkinson cycle fully expands to 4A with the in-cylinder pressure equal to the barometric pressure, more thermal energy (1–4O–4A–1) can be converted to useful mechanical work, resulting in higher thermal efficiency [57]. 2.2. Atkinson cycle vs. Miller cycle Fig. 2 demonstrates the Miller cycle process (1–2–3–4M–5–1), which is an over-expansion cycle but does not expand to the

2.3. Thermodynamic analysis Thermodynamic cycle analysis was performed to confirm the thermodynamic merits of Atkinson and Miller cycles using the finite-time thermodynamics method [50,60–68]. Some assumptions were introduced in the thermodynamic model and some practical impact factors, such as the fuel vaporization and combustion process, were ignored. Effects of some important parameters, such as compression ratio and equivalence ratio, on cycle efficiency have been computed. Fig. 3 qualitatively demonstrates the effects of compression ratio rc, equivalence ratio £ and expansion-to-compression ratio r on cycle efficiency. Within the entire effective range of rc, the cycle efficiency of the Atkinson cycle is always higher than the Otto and Miller cycles. However, as the rc increases, the efficiency differences decrease. There is an optimum rc approximately 20 that maximizes cycle efficiency of the Atkinson or Miller cycles. It is

304

J. Zhao / Applied Energy 185 (2017) 300–319

(a) P-V diagram

(b) T-S diagram

Fig. 2. P-V and T-S diagrams for ideal Otto, Atkinson and Miller cycles.

worth mentioning that the equivalence ratio also has a considerable effect on cycle efficiency. As the £ increases, cycle efficiency of the Otto cycle obviously decreases. For a practical Otto cycle engine, a rich mixture is generally carried out to avoid overheating the TWC at full loads. According to formula

T4a ¼ T3 re1k ;

ð1Þ

for the Atkinson or Miller cycle engine, a higher expansion ratio contributes to lower in-cylinder temperature at the end of the expansion stroke, thus resulting in a lower exhaust temperature, as shown in Fig. 2(b). In addition to a slighter heat load in the exhaust pipelines [69], an equivalence ratio closer or equal to the stoichiometric value can be used to further improve the fuel efficiency without the risk of TWC overheating under high or full loads. According to the writer’s studies and a literature review, some important thermodynamic characteristics for the over-expansion cycles have been summarized: (1) An over-expansion cycle has more work output and higher thermal efficiency than an Otto cycle at the same operating conditions [57]. (2) There is an optimum compression ratio that maximizes the performances of Atkinson or Miller cycles. The effective power, power density and effective efficiency increase up to an optimum value and then start to decrease with increasing compression ratio [68]. The effective power and

effective efficiency increase with increasing cycle temperature ratio, cycle pressure ratio and inlet pressure while decrease with increasing friction coefficient, inlet temperature and retarding angle of intake valve closing [67]. (3) The equivalence ratio has a significant impact on engine performance [62]. Under high or full loads, an Atkinson or Miller cycle engine can implement an equivalence ratio equal or closer to 1, compared with an Otto cycle engine. First, the combustion efficiency increases, and the fuel economy is greatly improved. In addition, HC and CO emissions decrease because of increased O2 concentration and a higher transformation efficiency of the TWC under the stoichiometric AFR. (4) The Miller cycle can decrease the maximum cylinder temperature of a turbocharged engine, but variations of the exhaust temperature are also determined by the turbocharger efficiency [66]. (5) Partial load efficiency of the over-expansion cycle with the largest feasible expansion ratio is the highest. Diesel and dual cycles, with higher partial load efficiency than the Otto cycle, must apply a compression ratio over 25:1 to achieve the same efficiency as the over-expansion cycle [50]. 3. Methods and challenges for realizing the Atkinson cycle As described above, the original Atkinson engine concept is a full expansion cycle; a later Atkinson cycle realized via LIVC (or EIVC) is essentially the same as the Miller cycle. This section mainly studies the types and mechanical realizations of the Atkinson cycle including the Miller cycle. 3.1. Atkinson cycle type Fig. 4 shows three types of Atkinson cycles that are described as:

Fig. 3. Effects of rc, r and / on cycle efficiency.

① The original Atkinson cycle is realized by fully expanding to the barometric pressure to draw the full working potential of in-cylinder working charges while the intake and compression strokes are the same as a normal Otto cycle. The engine, with less geometrical displacement VD, has to be designed as a larger volume VA for a long expansion stroke. ② The Atkinson cycle is realized by retarding or advancing intake valve closure timing based on a normal Otto cycle (or diesel cycle): LIVC or EIVC. The expansion stroke is the same as the Otto cycle, while the effective compression stroke shortens. In this simple way, an over-expansion cycle is realized. ③ The expansion stroke length is the same as a normal Otto cycle, but a higher expansion ratio is realized by reducing dead volume Vcc, as shown in Fig. 4(b). LIVC or EIVC is carried out to

305

J. Zhao / Applied Energy 185 (2017) 300–319

(a)

P

(b) 2

Atkinson cycle obtained Advancing or Retarding IVC

3

1

Original Atkinson cycle

IVC Vcc

IVC

v VD

VA

Vcc

VD

Fig. 4. Three types of Atkinson cycle [70,71].

realize the over-expansion cycle. The reduced ECR resulting from LIVC or EIVC can be compensated by the increased GCR. In this way, a greater efficiency improvement can be achieved.

Table 1 Survey and comparison for mechanical realizations of Atkinson cycle. Type

Mechanisms

Advantages

Disadvantages

①

1. Original opposed piston and mechanical linkage structure [1,46,72]

Stationary compression to expansion ratio without the need for control

2. Multi-link mechanism [73– 77]

1. Simple, compact, suitable for mass production; 2. Acceptable friction loss

3. Planetary gear mechanism [49]

1. Simple, compact; 2. Great potential for mass production

1. Most complex, not compact; considerable friction loss; 2. Invariable inlet stroke length at different loads; 3. Not suitable for mass production 1. More kinetic components and complexity; not suitable for highspeed engines; 2. Difficult to adjust intake stroke length 1. Higher complexity and cost; 2. Cannot adjust intake stroke length at different engine loads

1. Manually delay or advance inlet camshaft [78–80]

Simple, low cost, achieved based on Otto (or diesel) cycle engine

2. VVT [81–86]

1. Simple; 2. Easy to control; 3. Mass production

1. Manually adjust camshaft + increased GCR [51,87–90]

1. Simple, low cost; 2. Easy to realize via increasing piston top height

2. VVT + increased GCR [18,48,91,92]

1. Simple, low cost; 2. Better performance at different loads; 3. Most frequently used Optimum performance

3.2. Mechanical realizations and challenges 3.2.1. Survey on the existing mechanisms for realizing the Atkinson cycle Table 1 summarizes the existing measures and mechanisms for realizing the Atkinson cycle that was shown in Fig. 4. Mechanism 1 for type ① is the first practical ACE arranged as an opposed piston engine, the Atkinson differential engine [1]. As shown in Fig. 5(a), a single crankshaft was connected to two opposed pistons through a toggle jointed linkage that had a non-linearity. For half a revolution, one piston remained almost stationary while the other approached it and returned. Then, for the next half revolution, the pistons changed over which piston was almost stationary and which piston approached and returned [46]. Thus, in each revolution, one piston provided a compression stroke and a power stroke, and the other piston provided an exhaust stroke and a charging stroke. A later version, also designed by James Atkinson in 1887, is one of the most familiar ACE structures, as shown in Fig. 5(b) [46,72]. The piston is not directly connected to the crankshaft but is connected through a unique mechanical linkage. In this way, the engine could implement longer expansion and exhaust strokes than its intake and compression strokes. The expansion ratio is approximately 1.78 times the compression ratio. Higher thermal efficiency is achieved while a shorter compression stroke reduces the compression ratio and prevents the knock. Since then, some other novel multi-link mechanical systems have been researched and designed for realizing the extended expansion Atkinson cycle [73–77]. The mechanism shown in Fig. 6 is comprised of a four bar mechanism that connects the piston to the crankshaft via an oscillating member positioned on the opposite side of the cylinder relative to the crankshaft. The oscillating member is supported on an eccentric shaft that rotates at half of the crankshaft speed in the same direction and is driven by the crankshaft. The oscillating member, eccentric shaft and crankshaft jointly impact the piston movement and give rise to a shorter intake and compression stroke than a long expansion and exhaust stroke. A phasing mechanism between the crankshaft and the eccentric shaft allows relative angular motion between them. This relative motion changes the piston position at TDC and thus changes the combustion chamber volume, allowing the compression ratio to be altered continuously. The mechanism can be used to improve partial load fuel efficiency as

②

③

3. VVT + VCR [70,77,93–95]

1. Invariable valve timing; 2. Less partial load efficiency improvement; 3. WOT power loss 1. Additional component and cost; 2. Additional calibration time 1. Stationary valve timing; 2. Reduced power density; 3. Not optimum for different engine operating conditions 1. Reduced power density; 2. Additional cost and time 1. Complex, expensive; 2. Bad dynamic response for VCR; 3. Many unresolved challenges

well as full load output. This mechanism realizes the overexpansion cycle as in type ① but also alters the CR by adjusting the combustion chamber volume as in type ③. The multi-link mechanical system shown in Fig. 7 has been developed for Honda’s EXlink (Extended Expansion linkage engine), as the world’s first mass-produced multiple linkage ACE.

306

J. Zhao / Applied Energy 185 (2017) 300–319

Fig. 5. Patent drawing of the original Atkinson engine [1,46,72].

Oscillating member

Phasing mechanism (a)

(b)

Eccentric shaft (control shaft)

(c)

(d) Fig. 6. Mechanism for variable compression and stroke ratios [77].

A multi-link system is primarily comprised of a trigonal link, a swing rod, an eccentric shaft, a connecting rod and a crankshaft. In the EXlink engine, the trigonal link is positioned between the connecting rod and the crankshaft. The trigonal link is connected to the eccentric shaft through the swing rod to accomplish the extended expansion linkage mechanism. The eccentric shaft, driven by the crankshaft, turns at half of the crankshaft speed in the same direction. This multi-link mechanism allows the piston’s stroke to lengthen for the expansion/exhaust and shorten for the intake/compression per cycle. In this way, the type ① Atkinson cycle can be realized in a simpler and more compact structure than the original one in Fig. 5. The expansion ratio is approximately 1.5 times the compression ratio. It is worth noting that the two multilink mechanisms in Figs. 6 and 7 are similar in realizing the Atkinson cycle except that the phasing mechanism in Fig. 6 allows the CR adjustment.

Shown in Fig. 8, a crank train concept realized via a planetary gear mechanism has been researched as an effective measure to realize the Atkinson cycle (type ①) [49,96]. Because of the defined gear ratios of the sun wheel (crankshaft), planetary wheel (extender) and annulus gear, piston motion for the Atkinson cycle could be generated. The Atkinson cycle piston motion is based on the superposition of the two movements of the extender and crankshaft. As a result, the dead volume remains invariable while an almost two times longer expansion stroke than the compression stroke is achieved. In this way, cylinder pressure at the end of the expansion stroke as low as the barometric pressure can be realized to fully extract the working potential of in-cylinder gases. The negative effect of this Atkinson cycle with high gas exchange losses has been reported [49]. The cause may be the lower gas pressure at the end of expansion that eliminates the free exhaust process and leads to increased exhaust pumping losses.

J. Zhao / Applied Energy 185 (2017) 300–319

307

Multiple linkage system Piston motion and strokes

Fig. 7. EXlink multiple linkage system and piston movement [73–75].

Atkinson crank train concept

Fig. 8. Piston motion and schematic diagram of the IVT Atkinson cycle concept [49].

Atkinson or Miller cycles for type ② are much easier to implement than type ①. It can be realized via LIVC (or EIVC) based on a baseline Otto (or Diesel) cycle engine. A series of discrete LIVC timings can be performed by manually retarding the camshaft or fabricating some camshafts with different cam profiles [80], as shown in Fig. 9(a). Continuously variable LIVC timings can be implemented by a VVT mechanism. VVT technology has been widely used in gasoline engines [81–85,97,98], making it particularly easy to realize the over-expansion cycle. This type of ACE or MCE is actually a modified Otto (or diesel) cycle engine. With a VVT mechanism, LIVC can be performed by keeping intake valves open longer than in the normal condition in the compression stroke shown in Fig. 10. In this way, the expansion ratio, close or equal to the

GCR, is larger than the ECR. The disadvantage is that a reverse flow process decreases the engine power density [48,78,79,91]. For the EIVC shown in Fig. 9(a), the intake valves close extremely early in the intake stroke. Then, pistons continue to run towards BDC, inducing an internal cooling process. After the BDC, the pistons run upward, compressing in-cylinder charges. The expansion and subsequent compression back to the volume at which the intake valves are closed is relatively energy-free, because both processes are nearly isentropic at such low temperatures [99]. In this way, the effective expansion stroke is longer than the effective intake and compression strokes and enables equivalently realizing the over-expansion cycle. The EIVC strategy was originally used in diesel engines to realize the Miller cycle [56].

308

J. Zhao / Applied Energy 185 (2017) 300–319

(a)

LIVC or EIVC (

Increase GCR

(b)

)

Fig. 9. Atkinson or Miller cycle realized by LIVC (EIVC) + increased GCR [51].

Otto cycle

Atkinson cycle

intake stroke

intake stroke

compression stroke

power stroke

exhaust stroke

reverse flow process

compression stroke

power stroke

exhaust stroke

Fig. 10. Comparison between the Otto cycle and over-expansion cycle realized via VVT [78].

However, the studies and applications for LIVC are obviously more than EIVC. One primary cause is that EIVC will lead to considerable WOT power loss, especially in the high speed range [25,51]. Higher intake boost pressure is needed to compensate the more power loss for EIVC [100]. The efficiency is higher for EIVC because of the lower compression work compared to LIVC, when backflow to the intake manifold occurs [100,101]. With a partial load, type ② can reduce pumping loss and improve fuel economy. However, the improvement is less as a result of the reduced ECR and the same expansion ratio as the baseline Otto cycle. The initial purpose of the Atkinson cycle SI engine is an increased expansion ratio to enhance thermal efficiency while maintaining a normal ECR to avoid the knock [46]. The expansion ratio can be increased by lengthening the expansion stroke (type ①) or reducing the dead volume that is type ③. For type ③, mechanical realization by VVT + increased GCR is the most frequently adopted technical measure [18,48,91– 93,102]. The combustion chamber volume is reduced via changing the geometry of the piston top to increase the expansion ratio while a normal ECR is maintained based on LIVC. As shown in Fig. 9(b), a series of pistons with different GCR were manufactured from an original diesel engine piston. These pistons have been used to test the performances of different Miller cycle SI engines.

However, these pistons are not ideal for SI engines. Fig. 11 shows how the piston of a gasoline ACE with PFI is manufactured from that of an Otto cycle engine. The GCR increased from 10.6:1 to 12.5:1 by increasing the height of the piston top. For an ACE with GDI, in addition to increasing the height of the piston top, the shape of the piston crown must simultaneously be optimized to accelerate combustion and reduce heat transfer loss [92]. The method using VVT + VCR can achieve optimum performance over the entire speed and load range. As shown in Fig. 12, the combustion chamber volume can be continuously adjusted by changing the position of the outer piston relative to the inner piston, which is realized by controlling the oil pressure in the upper and lower oil chambers. In this way, ECR can remain invariable even though significant LIVC is performed and the expansion ratio can be the optimum value under any operating condition. 3.2.2. Challenges and general recommendations Mass production for type ① is generally realized via the multilink mechanism. Compared to type ③, under the same operating condition and expansion ratio, the shorter intake and compression strokes for type ① contribute to less intake pumping loss and compression work, resulting in slightly higher thermal efficiency [73]. However, this type of mechanical realization increases the kinetic

J. Zhao / Applied Energy 185 (2017) 300–319

Otto cycle

309

Atkinson cycle

Fig. 11. Piston crowns of the baseline Otto cycle and ACE.

(LIVC)

VCR

(a)

(b) Fig. 12. Atkinson cycle realized by VVT + VCR [90].

mass and is generally more complex than conventional engines. The NVH problem is also more severe. It is not suitable for highspeed operation because of the quickly increased inertia force and friction loss. Moreover, the intake and compression strokes are generally fixed and cannot respond to varying operating conditions. Although the design in Fig. 6 can adjust the compression to expansion stroke ratio, no mass production is reported and further investigation is required. A general recommendation is that this type of mechanical realization is particularly suitable for small general-purpose low-speed engine with simple operating conditions. For the multi-link mechanism, the friction loss between the piston and cylinder wall can be reduced by minimizing the angle between the connecting rod and the cylinder wall. As a result, the total friction loss of the multi-link Atkinson engine can be maintained as much as that of a conventional engine [75]. The Atkinson cycle realized via LIVC + increased GCR has been widely used in automotive engines and is better for use in hybrid

electric vehicles. The primary problem for LIVC + increased GCR is the mixture back-flow reducing WOT power/torque and power density. LIVC + VCR is the most promising technical measure for achieving optimum performance within the entire engine speed and load range. A VCR mechanism is crucial but it increases engine complexity and cost. The engine friction loss and weight also accordingly increases discounting the fuel economy. The dynamic response rate of current VCR mechanisms are always slower than required. More technical challenges are needed to resolve VCR mechanisms [103,104].

4. Applications and problems for the ‘‘Atkinson cycle effect LIVC or EIVC in types ② and ③ are generally called ‘‘Atkinson cycle effect” or ‘‘Miller cycle effect” and have special applications. According to previous publications, the ‘‘Atkinson cycle effect”

310

J. Zhao / Applied Energy 185 (2017) 300–319

can be carried out to aid load control, decrease NOx formation and suppress the knock. 4.1. Aid engine load control During most of the operating time (almost more than 50%), conventional automotive engines work at partial load conditions [99]. It is particularly convenient and flexible to use a throttle valve to regulate fresh charge amounts. However, the throttle valve causes a large vacuum in the intake manifold, leading to considerable intake pumping loss. If the throttling losses can be cancelled out, 15–20% fuel consumption can be saved [105]. The ‘‘Atkinson cycle effect” has been proved to be an effective method to assist in engine load control and improve partial load fuel economy by reducing pumping loss [81–88,98,99,105–108]. LIVC or EIVC can be implemented by a VVT mechanism to aid engine load control. As shown in Fig. 13, the exhaust-intake pumping loop area (=PMEP) decreases because of LIVC or EIVC operation. At the same engine load IMEPnet (=IMEPgross  PMEP), the compression-expansion loop area IMEPgross decreases, meaning that a lesser fuel-air mixture is required. In this way, the fuel economy is improved because of the PMEP reduction. A principle of the other research that uses LIVC or EIVC to reduce pumping loss is essentially the same as that in Fig. 13. At a constant engine load, a larger throttling width can be implemented because of the ‘‘Atkinson cycle effect”. As a result, the vacuum degree in the intake manifold decreases, thus reducing the pumping loss [81–85,98,99]. For example, the studies in Ref. [81] confirmed that the highest reduction of pumping loss of 47% and approximately 6% BSFC improvement with LIVC to aid load control at 2000 rpm could be achieved.

in diesel engines, NO2 is generally 10–30% of the total exhaust NOx emissions [109]. The principal NO source is the oxidation of atmospheric nitrogen; the principal reactions that govern the NO formation in the combustion of near-stoichiometric fuel-air mixture are:

O þ N2 NO þ N

ð2Þ

N þ O2 NO þ O N þ OH NO þ H

ð3Þ ð4Þ

The temperature ranges for forward reactions in Eqs. (2)–(4) are 2000–5000 K, 300–3000 K, and 300–2500 K, respectively. The fresh air only contains molecular nitrogen (N2), the required atomic nitrogen for the forward reactions in Eqs. (3) and (4) must be from the productions of the forward reaction in Eq. (2). Although the initial temperature for forward reactions in Eqs. (3) and (4) is particularly low (300 K), the reactions hardly occur if the flametemperature is lower than 2000 K. Another crucial factor that impacts the forward reactions is the AFR [80,110]. When the flame-temperature is high enough, more N2 and O2 in the fresh mixture accelerates the forward reactions and leads to more NO formation. An excessively rich/lean mixture can also greatly reduce the NO formation because of the lower flame-temperature. However, the CO or HC will sharply increase, and the transformation efficiency of TWC will also decrease. Therefore, lowering the flame-temperature inside the engine cylinder is the most effective measure to reduce NOx formation at the source. For SI engines, the reduced ECR by LIVC or EIVC decreases the mixture pressure and temperature at the end of the compression stroke. Accordingly, the peak flame-temperature declines [78]. For diesel engines, according to formula [71]: ðk1Þ

Tsoi ¼ TIVC CReff

4.2. Reduce NOx formation The harmful emission NOx mainly includes NO and NO2; NO is the predominant oxide of nitrogen that is produced inside the engine cylinder. In SI engines, NO2/NO ratios are negligibly small;

;

ð5Þ

the reduced ECR contributes to reduced air temperature at the start of injection and decreases the peak flame-temperature. Thus, the Atkinson or Miller cycle, as an effective way to reduce the peak flame-temperature, is an ‘‘internal cooling cycle”. The NOx forma-

Fig. 13. Comparisons of load control methods: EIVC or LIVC vs. Throttle [107].

J. Zhao / Applied Energy 185 (2017) 300–319

tion is particularly temperature-dependent; this is the original idea for reducing the NOx emissions based on the ‘‘Atkinson cycle effect”. Numerous researches on the application of LIVC or EIVC in reducing NOx emissions have been published [58,71,78–80,86,90, 100,101,110–124]. The original MCE with EIVC was aimed at reducing the NOx and PM emissions for diesel engines [56,57]. A 48.27% reduction of NOx formation by delaying IVC to 83CA after BDC has been reported [111]. Despite a significant NOx reduction, the fuel consumption merits for ACE are not obviously discounted [71,111–113,115]. 4.3. Suppress the knock The knock is an abnormal combustion phenomenon that is the result of auto-ignition of the end-gas before the spark-flame front arrives. Its occurrence mainly depends on the end-gas temperature and pressure as well as the spark-flame development rate [125]. The ‘‘Atkinson cycle effect” is an effective measure for suppressing the knock because it can reduce the end-gas pressure and temperature. The principle for suppressing the knock is the same as that of reducing NOx formation. The advantage of using LIVC to suppress the knock can be confirmed by Fig. 14. The results were computed based on the induced time formula by Douaud and Eyzat [91,126]. The KI (knock index) value equal to 200 represents the knocking occurrence. It is seen that the effectiveness of LIVC in suppressing the knock is close to the spark angle. Thus, the ‘‘Atkinson cycle effect” is a fine alternative instead of the spark angle in impeding the knock since retarding the spark timing has an intense negative effect on the indicated thermal efficiency. Suppressing the knock is the original purpose of introducing the Atkinson cycle in a SI engine. A normal ECR could be maintained through LIVC or EIVC [46,48,91]. Another significant application of the ‘‘Miller cycle effect” is knock suppression in modern super-/turbocharged SI engines. The intake boost in a turbocharged SI engine increases the fresh charge temperature at the start of compression, thus intensifying the knocking tendency especially in the low speed range. If the knock is suppressed via reducing the GCR or retarding the spark timing, the thermal efficiency improvement will be discounted [18]. LIVC or EIVC can effectively impede the pressure and temperature increase resulting from the usage of the intake boost. Some research has been addressed concerning the application of LIVC or EIVC in turbocharged SI engines for knock suppression

Fig. 14. Effect of IVC timings and SAs (spark angles) on the KI.

311

[18,77,89,93,127]. In a gas SI engine with an expansion ratio of 15, using an LIVC to realize an ECR 11 for knock suppression, combined with an optimized combustion system and a high efficiency turbocharger, an efficiency of 42.2% has been achieved [89]. In another publication, a knock limited LIVC operation contributing to 4.7% fuel economy improvement at a high load of a highly boosted GDI engine with GCR 12 has been reported [18]. 4.4. Primary problems and discussion In an SI engine, the most applicable load ranges for reducing pumping loss and NOx formation, as well as suppressing the knock, are different. In the medium to low load range, the effect of pumping loss is more significant while it is more crucial for NOx and the knock in the high load range. The primary problems are: (1) LIVC has two contrary effects on thermal efficiency [89,99]. On one side, it can reduce pumping loss; on another side, it reduces the ECR and thus reduces the indicated thermal efficiency. (2) The effectiveness of the ‘‘Atkinson cycle effect” on reducing pumping loss in the high speed range declines because of the shortened time for a fresh charge to flow back to the intake manifold [99]. Moreover, the effective LIVC range in reducing the charge amount is limited. As shown in Fig. 15, when the load is low enough, the LIVC 115CA has nearly no effect on the intake airflow rate. Further retardation of LIVC timing would sharply reduce the ECR, counteractively discounting the throttling-less merit. Although EIVC has a larger load regulation range [51], its disadvantages are also significant, as described previously. For the primary problems of applying the ‘‘Atkinson cycle effect” in Atkinson or Miller cycle SI engines, a novel load control strategy over the entire load range is provided. (1) Under extremely low loads, advance IVC timing for high ECR and only use the throttle valve to govern the engine load; (2) Around a medium load, perform LIVC operation since pumping loss is considerable holding a substantial proportion of energy loss in this load range; (3) At high loads, the NOx and knock are crucial and the pumping loss is not of importance. In this situation, if highly efficient cooled EGR is adopted to suppress the NOx and knock [11,13,14], advance the IVC timings towards BDC for higher indicated thermal efficiency; otherwise, perform LIVC operation for reduced pumping loss [128]. LIVC reduces the mixture temperature at the end of the compression stroke and accordingly prolongs the 0–5% combustion duration. More advanced MBT timing can be performed to enhance the degree of constant volume combustion and

Fig. 15. Effect of ETC and LIVC on the net mass airflow rate. Speed: 2000 rpm, WOT.

312

J. Zhao / Applied Energy 185 (2017) 300–319

improve the indicated thermal efficiency [88]. In this way, the LIVC disadvantage can be moderately compensated by the reduced pumping loss and advanced MBT. For SI engines, the NOx formation monotonically decreases as increasing the Atkinson or Miller cycle degrees [100,91,111,117]. Some investigations reported different trends for CI engines with diffusion combustions. A study demonstrated that at low loads increasing LIVC or EIVC increases NOx close to the reference value with standard valve timing [118]. A longer ignition delay allowed more fuel being mixed prior to ignition, leading to a greater magnitude of premixed combustion. Moreover, the long ignition delay leads to combustion instability; and lets the fuel spray penetrate further, enhancing air entrainment favoring NOx formation [119]. The premixed combustion generally leads to faster combustion and thus higher flame temperature. Split injection strategy is expected to reduce the magnitude of the premixed combustion and NOx [120]. However, a recent study [115] demonstrated that at high loads higher Miller degree with single injection always brings a further reduction of NOx. Under the same Miller degree, split injection generally reduces BSFC but leads to increased NOx, when compared with the single injection cases. The reasons for these results are: (1) In the medium to high load range, although high Miller cycle degree enhances the magnitude of the premixed combustion, the heat release rate of the main diffusion combustion (dominant), with longer duration, decreases because of lower reactants temperature. Eventually, the NOx increase because of the premixed combustion is less than the NOx reduction because of lower reactants temperature; the net NOx decreases. At low loads with less fuel mass injected, when high Miller cycle degree is carried out the NOx increase because of the high proportion of premixed combustion (dominant) exceeds that of diffusion combustion; the net NOx increases. (2) In the cases of split injection, pilot injection increases the magnitude of the premixed combustion and flametemperature. Although modest pilot injection indeed can reduce the peak of the premixed combustion, the main diffusion combustion is simultaneously enhanced due to compression heating from the premixed combustion. Eventually the split injection doesn’t produce further NOx reduction than the single injection. Considering the problems when applying the ‘‘Miller cycle effect” in CI engines with diffusion combustions, the VVT mechanism can be applied to adjust the Miller timings according to the engine’s operating conditions. At low loads, eliminate Miller cycle for improving thermal efficiency and combustion instability; at higher loads, increase Miller degree for great NOx reduction. The Miller timings together with fuel injection strategy should be simultaneously optimized under different operating conditions in order to achieve an optimum trade-off among NOx, fuel efficiency and combustion stability.

and increases the NOx formation. Within the medium to low load range, improvement in fuel economy does not match the reduction of pumping loss. In this situation, the mixture heating method to compensate the negative effect of the ECR reduction can be carried out [99], as shown in Fig. 16. The mixture temperature in an ACE with mixture heating (LIVC = 105CA) is even higher than that in an Otto cycle engine without mixture heating. Thus, the thermal efficiency can be greatly improved because of the mixture heating and LIVC. To not discount the fuel economy merit of mixture heating, engine cooled water or exhaust gas can be used to heat intake charges [99]. (2) Higher GCR. Research through experimentation and calculation has confirmed that the expansion ratio was 10 times as effective as the ECR in improving the thermal efficiency [87]. It is thought that the thermal efficiency is mainly determined by the expansion ratio [89]:

gt v ¼ 1 

1

eek1

;

ðee is the expansion ratio; k is the specific heat ratioÞ

ð6Þ

For a modern ACE or MCE, a higher expansion ratio is generally realized by way of increasing GCR. Therefore, the negative effects of the ECR reduction can be compensated by the increased GCR. On the one hand, as shown in Fig. 17, for a constant ECR, more LIVC can be performed [92]; on the other hand, the indicated efficiency reduction can be compensated by the increased expansion ratio. However, greatly increasing GCR is limited by reduced power density; this will be discussed in detail in the next section. (3) VCR. To maintain an invariable ECR, the clearance volume at TDC can be accordingly reduced as the load becomes light. A VCR mechanism can realize continuous compensation for the reduced ECR and can achieve the highest thermal efficiency improvement by optimizing the expansion ratio and extending the advantageous less throttling LIVC range [70,94,95]. 5.2. Technical measures and challenges for improving power density The power density reduction for type ① in Table 1 is unavoidable because the engine volume has to be increased to realize the long expansion stroke. For SI engines with type ③, LIVC (or EIVC) is carried out to suppress the knock at WOT operating conditions. For CI engines with type ②, LIVC (or EIVC) is generally carried out to reduce the NOx formation. The mixture backflow (or shortened intake stroke) reduces the power density; this is a signif-

5. Challenges and solutions for using an Atkinson cycle engine Three primary challenges for modern ACE and MCE have been clarified: (1) LIVC reduces the ECR, which decreases the indicated thermal efficiency and limits the throttling-less range; (2) reduced power density; and (3) transient AFR control. 5.1. Technical measures and challenges for reduced ECR Three technical measures can be adopted to address the reduced ECR and extend the throttling-less load control range: (1) mixture heating; (2) higher GCR; and (3) VCR mechanism. (1) Mixture heating. Notably, mixture heating is only suitable for medium to low loads. At high loads, mixture heating is not advantageous because it intensifies the knock tendency

Fig. 16. Changes of in-cylinder charge temperature with and without mixture heating [99].

313

J. Zhao / Applied Energy 185 (2017) 300–319

EIVC

(a)

LIVC

(b)

Fig. 17. Effective compression ratios [94].

EIVC

icant challenge for using ACE [46–48,129]. Table 2 summarizes the technical measures for improving the power density. (1) Turbocharged Miller cycle: Unlike ACE, MCE can improve engine efficiency without any reduction in power output because of the application of supercharger or turbocharger. With Miller cycle to reduce the ECR and highly boosted pressure to compensate the air mass losses, the in-cylinder compression temperature can be reduced [122,132]; and the power lost because of mixture backflow (or shortened intake stroke) can be recovered [101]. A turbocharger is superior to a supercharger because that a supercharger is directly driven by the crankshaft, consuming partial engine power, and reduces the fuel economy. Fig. 18 demonstrates the effects of IVC timings on the turbineinlet pressure, the compressor-outlet pressure, PMEP and efficiencies. As shown in Fig. 18(b), the efficiencies decrease for both extreme cases of LIVC and EIVC, which is because of the increased exhaust PMEP imposed by the turbocharger. Mixture backflow (or shortened intake stroke) results in less total mixture mass trapped. As the Miller degree increases the boost pressure has to accordingly increase in order to maintain the fresh mixture mass trapped. Then the exhaust backpressure increases leading to the increase of exhaust PMEP; and the brake efficiency declines. It is shown better efficiency for extreme LIVC. The reason is the much increased exhaust PMEP for EIVC and the high boost pressure requirement. Main problem for EIVC and very large LIVC is the need for a high intake boost pressure, which results in a higher

LIVC

Fig. 18. Effects of IVC timings on boost pressure and engine performances [101] (PMEP = intake PMEP  exhaust PMEP).

exhaust backpressure and a need for a bigger turbocharger. However, higher intake boost pressure requirement (higher than 5 bar) is very challenging in current engines. For this, a dual stage turbocharger configuration may be very promising but it largely increases the engine complexity. The R&D for high pressure ratio turbocharger is strongly recommended to address in the future. Crucial technology would be required to achieve up to 6 bar boost pressure over the whole engine speed range. High turbine efficiency would be necessary to avoid BSFC penalty. Almost all the studies point out the need for specially designed high efficiency turbochargers, able to provide high pressure ratios at low flow rates [101,133,134]. On the other hand, it is particularly of importance that the high intake boosting can’t lead to additional or significant increase to intake temperature, which will discount the temperature decrease by the Miller cycle. A highly efficient intake intercooler is always needed to cool down the intake air after boosted by a turbocharger.

Table 2 Technical measures for improving power density. Measures

Description

Intake boost (Miller cycle engine)

Partial load: Miller cycle; Full load: Turbocharged Miller cycle

Advantages 1. No power loss; 2. Simple and suitable for production

Drawbacks

VCR (OA cycle engine)

Partial load: Atkinson cycle; Full load: Otto cycle

1. Constant ECR at partial load; 2. Optimum performance over the entire operating range; 3. Permit high GCR

Knock resistance (special OA cycle engine)

Partial load: Atkinson cycle Full load: improve knock resistance for approximate even full Otto cycle

1. Less power loss even increased power output; 2. Simple and suitable for mass production

mass

1. Big turbocharger; 2. Need high efficiency intercooler

VCR is too complex with bad dynamic response

1. Additional cost and time; 2. More measures for knock resistance

Examples [66] [122] [100] [101] [115] [70] [77] [90] [94] [95] Has power loss: [48] No power loss: [92,102,130,131]

314

J. Zhao / Applied Energy 185 (2017) 300–319

In addition, water [135,136] or ethanol [101] injection into the engine intake ports may also be feasible methods since the latent heat of vaporization could effectively cool down the intake air. With current technical levels, a reduction in power output seems inevitable when applying Miller cycle [19,100]. So, a high efficiency turbocharger and an efficient intake intercooler are always needed to realize the low temperature cycle and guarantee the fuel economy at the same time [66]. (2) VCR: realize OA cycle engine [70,77,90,94,95,105,107,112]. The OA cycle engine operates on the Otto cycle at full load for a high power density while it operates on the Atkinson cycle at partial load for reduced fuel consumption. In an OA cycle engine, the Otto and Atkinson cycles have the same displacement but different clearance volume at TDC. At WOT operating conditions, the engine increases the clearance volume and implements a normal GCR (Otto cycle) to eliminate LIVC operation and increase the power density. This can be achieved by a VCR mechanism, which is necessary for an OA cycle engine. Many VCR mechanisms have been investigated and designed by some researchers for OA cycle engine realization [77,90,94,95]. As described previously, the VCR mechanisms have some technical challenges that need to be resolved [103,104], leading to difficulty for the mass production of an OA cycle engine. (3) Improve the knock resistance: realize special OA cycle engine. Improving the power density for Atkinson SI engines is essentially a problem of suppressing the knock. LIVC operation at a full load can be reduced and even fully eliminated by adopting some technical measures to improve the knock resistance. Approximately even full Otto cycle operation at a full load can be realized, thus enhancing engine power density. At partial load, the Atkinson cycle is operated for improved fuel economy [92,130]. Modern ACEs are mainly used for hybrid vehicles, the power/torque loss at full loads can be compensated by a driven motor [52–54,129]. As a result, the research on knock suppression at full loads for ACE is less [47,92,102]. It is particularly important to enhance the power density and reduce the engine displacement since it is advantageous for reducing vehicle fuel consumption. The knock resistance of an SI engine can be improved via the following aspects: accelerate the combustion process; reduce the flame-front propagation distance; lower the end-gas temperature and change the fuel/mixture properties. Practical feasible technical measures for ACE have been summarized in Table 3. Of course, knock suppression must be under the premise of not worsening engine fuel economy and emissions. Measures such as rich mixtures and delaying spark timing are not considered [14]. A new 1.8 L ACE with a GCR of 13:1 has been developed by Toyota Company [48], aiming at replacing its 1.5 L predecessor [54]. A highly efficient water-cooled EGR and electric water pump were introduced to decrease the end-gas temperature and thus improve the anti-knock performance. More advanced spark timing and less LIVC can be performed to improve the thermal efficiency and prevent considerable WOT power loss. Combining some anti-knock measures can ensure an approximate or full Otto cycle at WOT operating conditions. In 2014, Toyota Company issued two small 1.0 L and 1.3 L new ACE with GCRs of 11.5 and 13.5, respectively [102,137]. Up to 14 technical solutions were adopted to resolve the problem of significant torque reduction in the medium to low engine speed range. Technical highlights for the knock resistance are: special design of intake port (induce strong vertical swirl), cooled EGR, combustion cham-

Table 3 Practical techniques for suppressing the knock. Methods

Technical measures

Accelerate combustion

Increase turbulence intensity: vertical intake port [102]; optimize entry angle of intake port and valve fillet angle [92]; optimize piston crown (hemispherical cavity) [92]

Reduce flame distance

Place the spark-plug in the centre of combustion chamber; Reduce cylinder bore thus flame transferring distance [125]

Decrease end-gas temperature

Decrease intake temperature: highly efficient watercooled EGR [13–15,48,102,134]; water injection [14,135,136]; Reduce heating effect of cylinder wall: optimize engine cooled system [125,137]; Reduce hot residual gas: 4–2–1 exhaust pipeline [92,137]; proper exhaust timing and valve overlap [92]; GDI: latent heat of fuel vaporization [18,92]

Change fuel/mixture properties

Alternative fuel: natural gas, ethanol [101], LPG. Blended fuel: gasoline blended with other fuel for higher octane number; Use antiknock additives; Residual gas: suppress the end-gas activity impeding auto-ignition [14,125]; Lean end-mixture: multiple injections for stratified mixture (GDI engine) [138]

ber scavenging by 4–2–1 exhaust manifold, and water jacket spacers around the cylinders to control wall temperature [137]. These technical measures realize an approximate Otto cycle operation at full loads and ensure sufficient power/torque output. It is thought that high WOT torque could be ensured even in a high-GCR SI engine as long as the knocking problems are resolved [92]. As shown in Fig. 19, Mazda’s gasoline SI engine, with the world’s highest GCR of 14:1, achieves a 15% fuel consumption improvement and an astonishing 15% WOT torque rise compared with the baseline 2.0 L PFI engine with a GCR of 10:1. The knocking problems are resolved from two sides: accelerate combustion and reduce initial in-cylinder temperature. A hemispherical cavity is created on the piston top to enable the initial flame kernel to grow quickly and reduce cooling loss. The entry angle of the intake port and valve fillet angle are optimized to intensify the in-cylinder flow to accelerate combustion. The cylinder bore is made smaller to shorten the flame propagation distance. Furthermore, the high-pressure GDI and special 4–2–1 exhaust manifold are employed to reduce the initial in-cylinder temperature. The latent heat of fuel vaporization cools the in-cylinder gas. Through proper design of the 4–2–1 exhaust manifold and proper valve overlap, the exhaust interference between adjacent cylinders is removed and an efficient scavenging process is realized, largely reducing in-cylinder hot residual gas. As a result, the knock resistance and charging efficiency are considerably improved. Because of the knocking suppression measures, the fuel efficiency over the entire speed range of the high-GCR engine is greatly improved without any WOT torque loss. With a partial load, the Atkinson cycle is operated for improved fuel economy [92,130,131]. Moreover, more LIVC can be performed to further reduce pumping loss at partial load since the GCR is high [92]. The engines from Toyota [137] and Mazda [92] are essentially a type of special OA cycle engine without using the VCR mechanism.

5.3. Challenges and recommendations on transient AFR control The primary challenge for transient AFR control [59] is that the mixture containing fresh air and fuel flows back to the intake man-

315

Torque improvement ratio [%]

J. Zhao / Applied Energy 185 (2017) 300–319

Engine speed 1500rpm Combustion speed increase DI DI DI CR14 CR14 CR14 Flat piston Cavity piston Cavity piston + high tumble intake port combustion lead to torque reduction Base Current 2.0L PFI CR10

DI CR11.2

Air cooling DI DI CR14 CR14 + MHI + Scavenging + Split + 4-2-1 injection exhaust manifold

Fig. 19. Measures on torque loss resulting from knocking [92].

ifold because of LIVC operation. In the subsequent intake stroke of another cylinder, the backflow mixture together with fresh air flows into the cylinder again. Under stationary operating conditions, the injected fuel mass can be experimentally calibrated to aim for the stoichiometric AFR value. Under transient operating conditions, the mechanism based on feedback compensation makes quickly correcting the injected fuel mass difficult. As a result, AFR departs from the stoichiometric value, which worsens fuel economy and emissions. It is worth noting that the challenge on transient AFR control only exists in the ACE or MCE realized via LIVC. Type ① ACE does not have this problem because there is no mixture backflow occurring. Research recommendations: (1) The operating conditions for ACE or MCE can be as easy as possible even with only a single working point. (2) Build a prediction model for backflow fuel mass: a prediction model for backflow fuel mass resulting from LIVC can be built through experiments and simulations; and integrated to the AFR controller as a feed forward predicting module. As a result, the AFR controller can correct the injected fuel mass in advance rather than feedback compensation. 6. Critical issues and suggestions for future R&D 6.1. Current research direction for ACE and MCE As described above, the original ACE, with a complex structure and reduced power density, was not widely applied. Although the multi-link ACE is simpler than the original one, it still has the problem of reduced power density. There is no report on the application of the multi-link ACE in the automotive industry. The first mass produced Miller cycle engine, used as an automotive power unit, was released by Mazda in 1993 [139–141], showing up to 15% reduction in BSFC. The higher expansion ratio was increased by combining LIVC and bigger GCR, while the power density issue was solved using a specially designed high efficiency Lysholm supercharger [142]. At present, the turbocharged Miller cycle engines are more popular and frequently studied. The high efficiency turbocharger, with low-inertia turbine blades and small response delay, and high efficiency intercooler are being developed in order to realize higher Miller cycle degree.

Modern ACEs used in automobiles are generally realized via LIVC. The engine’s high-speed power and low-speed torque generally decrease. At present, the ACE is primarily used in the hybrid vehicle. However, because of technical progress, the ACE is beginning to be used as stand-alone automotive power unit [92,102,130,137,143]. Typical ACEs used as stand-alone power units are listed in Table 4. The current research focus for ACE is to enhance WOT power/torque under high-GCR, the special OA cycle engine. The crucial problem is improving the knock resistance under high-GCR. The practical technical measures have been discussed in Section 5.

6.2. Suggestions for future R&D An ACE in a hybrid vehicle has a large number of operating points with varying loads and speeds [54]. Many engine structure parameters have to be determined to compromise the performance from low to high speeds. The efficiency potential of an ACE in a hybrid vehicle is not fully developed. When an ACE is applied in a REEV [38,144,145], the situation is different. The REEV is a type of electric vehicle with functions of on-board electricity generation and plug-in charging [38]. Its power system contains a rangeextender (a genset), a suit of battery package and a propulsion motor. When the battery SOC is lower than a preset value, a

Table 4 Typical ACEs used as stand-alone power units. SKYACTIV-G 2.0 L [92,130]

Toyota 1KR-FE [102,137]

Toyota 1NRFKE [143]

Displacement Bore/stroke GCR Pmax

1.998 cc 83.5/91.2 14 120 kW/6000 rpm

0.996 cc 71/84 11.5 51 kW/6000 rpm

Tmax

210 N m/ 4000 rpm No report

95 N m/ 4300 rpm 37%

1.329 cc 72.5/80.5 13.5 73 kW/ 6000 rpm 121 N m/ 4400 rpm 38%

Mazda CX-5, Mazda MX-5

Toyota Aygo

Toyota Vitz

Highest thermal efficiency Application

J. Zhao / Applied Energy 185 (2017) 300–319

/Nm

/Nm

316

Rated torque for generator

Engine

Generator

N/rpm N/rpm Fig. 20. Performance maps of the generator and engine of a range-extender system.

range-extender engine starts to work under the global efficiency optimum point (the red1 point in Fig. 1) to propel a generator to produce electricity. The engine does not directly provide mechanical power to the wheels, and its maximal power requirement is greatly lower than a conventional automotive engine [146]. A smalldisplacement ACE or MCE may be safely adopted as a rangeextender engine. As a result, all of the engine structure parameters can be optimized under the global efficiency optimum point under the premise of meeting the power requirement. A survey has been made to identify a total of 35 REEVs [147], most of which use Otto cycle engines. Although some adopt over-expansion cycle engines, the engines are operated over a broad speed-load range instead of working at the efficiency optimum point. For example, the range-extender engine of Opel Ampera works under variable speeds [148,149]. The fuel consumption and CO2 emissions of a REEV based on different rangeextender engines have been tested [25]. The ACE in this study was allowed to work at two fixed points: one was a 15 kW high efficient point with thermal efficiency of almost 37%; another was a 40 kW boost mode but with a low efficiency of approximately 27%. The experimental tests show that the vehicle with an ACE is the most efficient in global annual combined operation cycles (urban-highway driving cycles). The REEV with an ACE achieves 6% fuel consumption improvement and 9% CO2 reduction compared to a REEV with a conventional Otto cycle engine. The advantages when an ACE or a MCE is used in a REEV and works at one normal operating point are: (1) The maximal power requirement is lower, thus higher GCR and LIVC can be realized for higher thermal efficiency; (2) Only one fixed normal working point: engine structure may be simpler, all relevant engine structure parameters can be optimized to greatly improve antiknocking ability and fuel economy, and VVT and VCR mechanisms are not necessary; (3) No transient AFR control problem because of mixture backflow [59]. Another similar application for an over-expansion cycle engine is in the stationary cogeneration plant [89,150,151]. An ICE based cogeneration plant is gaining popularity because of its high efficiency and low exhaust emissions. It can meet the needs of customers with comparatively low household demand for electricity and heat. In a cogeneration unit, an ACE or MCE may be applied as the power source of a generator for further improvement of fuel economy and emissions. For example, development of a 10 kW micro-cogeneration unit based on an Atkinson cycle methane SI engine has been reported [150]. A global electric efficiency of 31.5% has been achieved. 1 For interpretation of color in Fig. 1, the reader is referred to the web version of this article.

Significant problems have been clarified by investigating some domestic REEV’s range-extenders. Fig. 20 shows the engine fuel consumption map and generator efficiency map. The engine efficiency optimum point is obviously higher than the generator rated torque line. The generator’s input torque and power must be lower than the rated values to safely and stably work for a long time. As a result, the engine has to work in the fuel economy non-optimum area for a long time. The efficiency optimum points between the engine and generator do not match. Moreover, the torque of the generator’s extreme working point (red circle point) is considerably lower than the engine WOT torque under the same speed. The engine’s maximum power/torque is excessive. According to previous analysis, the crucial problem of applying an ACE or a MCE in a range-extender or cogeneration plant is efficiency optimum point matching. The coupling effect of the generator must be considered when designing and optimizing the ACE. For a specific requirement of electrical power, the power requirement for the engine will increase if the working point efficiency of the generator is lower. Then, under a specific engine displacement, less LIVC has to be carried out to achieve enough power output, and the GCR has to be reduced to avoid the knock. As a result, the ACE’s thermal efficiency drops. Investigations on the matching and co-optimization for the engine and generator have not been reported. It is suggested to study the parallel design and optimization for the ACE and generator based on multi-disciplinary design and optimization theory [152]. First, precise models must be built to predict the working processes and final performances for the ACE and generator, respectively. Then, a multi-disciplinary design and optimization methodology based on the models may be carried out to perform the matching of working points and co-optimization for the ACE and generator. In this way, the interactive coupling effect of a matched generator can be considered during the design stage of the ACE. Under the premise of meeting the required electrical power, the efficiency at normal working points for the ACE and generator can be optimized to fully draw the ACE’s fuel-saving potential. 7. Conclusions Vehicle electrification is crucial so that it can reduce the dependence on fossil fuels and meet future emissions regulations. ACE, with higher thermal efficiency than the conventional Otto cycle engine, is the most suitable for hybrid electric vehicles. MCE, similar to the ACE, is another type of over-expansion cycle engine. This paper provides a critical review of current state-of-the-art research on ACE and MCE. Their similarities and differences are clarified. An in-depth survey on the mechanical realizations for

J. Zhao / Applied Energy 185 (2017) 300–319

the over-expansion cycle is conducted; crucial challenges are clarified and general recommendations on real applications are given. After a comprehensive review of the applications of the ‘‘Atkinson cycle effect”, some novel strategies are provided. Promising technical solutions are studied and discussed to handle reduced ECR and power density. For potential applications, including REEV and cogeneration plants, a significant problem is presented and a parallel design and optimization methodology is provided. Based on the research performed for this paper, the following conclusions can be given: (1) The over-expansion cycle contributes to more useful mechanical work, lower exhaust energy and temperature. AFRs close or equal to the stoichiometric value can be used at high to full loads for further improvement of fuel economy and emissions. (2) The original ACE realized via a complex mechanical linkage is not practical. The mass-produced multi-link ACEs are particularly suitable for small general purpose low-speed engines. The planetary gear mechanism based Atkinson cycle is particularly prospective. Most of the modern ACEs and MCEs are realized via LIVC + increased GCR. The overexpansion cycle realized via LIVC + VCR can achieve optimum performance but many challenges for the VCR mechanisms must be resolved. (3) The ‘‘Atkinson cycle effect” is particularly helpful in load control, suppressing NOx formation and the knock. Primary problems are: (1) the most applicable load ranges for these applications are different; (2) LIVC has two contrary effects on the thermal efficiency; (3) the LIVC’s effectiveness in reducing the fresh charging amount is limited; (4) for CI engines with diffusion combustions, applying Atkinson cycle doesn’t always lead to further reduction of NOx. For these problems, the strategies are provided. For SI engines, at low loads, advance IVC timing and use the throttle valve to regulate the load; around a medium load, LIVC is used to aid load control; at high loads, LIVC is used to aid load control if no EGR used while advancing IVC timing if using EGR. For CI engines, at low loads, eliminate Miller cycle for improving thermal efficiency and combustion instability; at higher loads, increase Miller degree for great NOx reduction. (4) Three primary challenges for ACE and MCE have been clarified: reduced ECR, reduced power density, and transient AFR control. Reduced ECR can be compensated by a higher expansion ratio and heating the inlet mixture. Reduced power density can be solved by applying high efficiency turbocharger with efficient intake cooling, and improving the knocking resistance. The VCR mechanism is the most advantageous for dealing with the reduced ECR and power density. (5) Once the knocking problem under high-GCR is resolved, a special OA cycle engine can be realized and used as a stand-alone automotive power plant. When an ACE or a MCE is applied in a REEV or cogeneration unit, it can work at the global efficiency optimum point to propel a generator. Although the energy transforming times increase, the system electrical efficiency can still be greatly improved because of the maximum possible engine working point efficiency. Investigation on the optimization and matching methodology for the ACE and generator is suggested for future research.

Acknowledgments Funding: This work was supported by the National Natural Science Foundation of China (Grant number 51506130).

317

References [1] US 336505. Atkinson J. Gas engine [issued 1886-02-16]. [2] Climate change: basic information. United States Environmental Protection Agency, 2006-12-14. In common usage, ‘global warming’ often refers to the warming that can occur as a result of increased emissions of greenhouse gases from human activities. 2007-02-09. [3] Varnhagen S, Same A, Remillard J, Park JW. A numerical investigation on the efficiency of range extending systems using advanced vehicle simulator. J Power Sources 2011;196:3360–70. [4] U.S. Environmental Protection Agency. Climate change— science. U.S. Environmental Protection Agency; 2009 [online, cited: March 07, 2010]. http://www.epa.gov /climate change /science /stateofknowledge.html. [5] European Commission. EU energy in figures Retrieved from http://ec.europa. eu /energy/observatory/statistics/statistics de.htm; 2013. [6] http://auto.163.com/13/0122/14/8LR2UOOA000851PI.html and http:// i.daozhou.net/G996nbaie1x60119n435012784.shtml. [7] Energy Information Administration. U.S. consumption by sector, independent statistics and analysis; 2010 [online, cited: March 11, 2010]. http://www. eia.doe.gov/pub/oilgas/petroleum/analysispublications/oilmarketbasics/ demandtext.htm#U.S.%20Consumption%20by%20sector. [8] Fontaras G, Samaras Z. On the way to 130 g CO2/km—estimating the future characteristics of the average European passenger car. Energy Policy 2010;38 (4):1826–33. [9] Berggren C, Magnusson T. Reducing automotive emissions—the potentials of combustion engine technologies and the power of policy. Energy Policy 2012;41:636–43. [10] Lake T, Stokes J, Murphy R, Osborne R, Schamel A. Turbo-charging concepts for downsized DI gasoline engines. SAE paper 2004-01-0036; 2004. [11] Galloni E, Fontana G, Palmaccio R. Effects of exhaust gas recycle in a downsized gasoline engine. Appl Energy 2013;105(2):99–107. [12] Zamboni G, Moggia S, Capobianco M. Hybrid EGR and turbocharging systems control for low NOX and fuel consumption in an automotive diesel engine. Appl Energy 2016;165:839–48. [13] Fontana G, Galloni E. Experimental analysis of a spark-ignition engine using exhaust gas recycle at WOT operation. Appl Energy 2010;87(7):2187–93. [14] Bozza F, Bellis VD, Teodosio L. Potentials of cooled EGR and water injection for knock resistance and fuel consumption improvements of gasoline engines. Appl Energy 2016;169:112–25. [15] Wei H, Zhu T, Shu G, et al. Gasoline engine exhaust gas recirculation – a review. Appl Energy 2012;99(2):534–44. [16] Bellis VD. Performance optimization of a spark-ignition turbocharged VVA engine under knock limited operation. Appl Energy 2016;164:162–74. [17] King J. A review of low-carbon cars. HM Tresaruy; 2007. [18] Li T, Gao Y, et al. The Miller cycle effects on improvement of fuel economy in a highly boosted, high compression ratio, direct-injection gasoline engine: EIVC vs LIVC. Energy Convers Manage 2014;79:59–65. [19] Li T, Wang B, Zheng B. A comparison between Miller and five-stroke cycles for enabling deeply downsized, highly boosted, spark-ignition engines with ultra expansion. Energy Convers Manage 2016;123:140–52. [20] Taylor AM. Science review of internal combustion engines. Energy Policy 2008;36(12):4657–67. [21] Applied energy editorial. Novel technologies and strategies for clean transport systems. Appl Energy 2015;157:563–6. [22] Sherazi H Iqbal, Li Y. Homogeneous charge compression ignition engine: a technical review. In: Proceedings of the 17th international conference on automation & computing, 10 September 2010. Huddersfield (UK): University of Huddersfield; 2010. [23] Sharma TK, Rao GAP, Murthy KM. Homogeneous charge compression ignition (HCCI) engines. A review. Arch Comput Methods Eng 2015:1–35. [24] Dincer I, Rosen MA, Zamfirescu C. Economic and environmental comparison of conventional and alternative vehicle options. Electric and hybrid vehicles. Elsevier; 2010. p. 1–17. [25] Ribau J, Silva C, Brito FP, Martins J. Analysis of four-stroke, Wankel, and microturbine based range extenders for electric vehicles. Energy Convers Manage 2012;58(6):120–33. [26] Donateo T, Licci F, D’Elia A, Colangelo G, Laforgia D, Ciancarelli F. Evaluation of emissions of CO2 and air pollutants from electric vehicles in Italian cities. Appl Energy 2015;157:675–87. [27] Schneidereit T, Franke T, Günther M, Krems JF. Does range matter? Exploring perceptions of electric vehicles with and without a range extender among potential early adopters in Germany. Energy Res Soc Sci 2015;8:198–206. [28] Ziegler A. Individual characteristics and stated preferences for alternative energy sources and propulsion technologies in vehicles: a discrete choice analysis for Germany. Transp Res Part A: Policy Pract 2012;46:1372–85. [29] Ruan J, Walker P, Zhang N. A comparative study energy consumption and costs of battery electric vehicle transmissions. Appl Energy 2016;165:119–34. [30] Hidrue MK, Parsons GR. Is there a near-term market for vehicle-to-grid electric vehicles? Appl Energy 2015;151:67–76. [31] Zhao X, Doering OC, Tyner WE. The economic competitiveness and emissions of battery electric vehicles in China. Appl Energy 2015;156:666–75. [32] EPA, Environmental Protection Agency. Light-duty vehicle greenhouse gas emission standards and corporate average fuel economy standards for model years 2017–2025. Interim joint technical assessment report. Office of Transportation and Air Quality, September; 2010.

318

J. Zhao / Applied Energy 185 (2017) 300–319

[33] Li Y, Davis C, Lukszo Z, Weijnen M. Electric vehicle charging in China’s power system: energy, economic and environmental trade-offs and policy implications. Appl Energy 2016;173:535–54. [34] Ferrero E, Alessandrini S, Balanzino A, Yan J. Impact of the electric vehicles on the air pollution from a highway. Appl Energy 2016;169:450–9. [35] Onat NC, Kucukvar M, Tatari O. Conventional, hybrid, plug-in hybrid or electric vehicles? State-based comparative carbon and energy footprint analysis in the United States. Appl Energy 2015;150:36–49. [36] Doucette RT, McCulloc MD. Modeling the CO2 emissions from battery electric vehicles given the power generation mixes of different countries. Energy Policy 2011;39(2):803–11. [37] Tomaz K. Hybridization of powertrain and downsizing of IC engine – a way to reduce fuel consumption and pollutant emissions – Part 1. Energy Convers Manage 2007;48(5):1411–23. [38] Tate ED, Michael OH and Peter JS. The electrification of the automobile: from conventional hybrid, to plug-in hybrids, to extended-range electric vehicles. SAE paper 2008-01-0458; 2008. [39] http://en.wikipedia.org/wiki/Toyota_Prius. [40] http://en.wikipedia.org/wiki/Honda_Insight. [41] http://en.wikipedia.org/wiki/Ford_Fusion_Hybrid. [42] Wang Y, Zu B, Xu Y, et al. Performance analysis of a Miller cycle engine by an indirect analysis method with sparking and knock in consideration. Energy Convers Manage 2016;119:316–26. [43] Millo F, Rolando L, Fuso R, Mallamo F. Real CO2 emissions benefits and end user’s operating costs of a plug-in hybrid electric vehicle. Appl Energy 2014;114(114):563–71. [44] Guo J, Ge Y, Hao L, Tan J, Peng Z, Zhang C. Comparison of real-world fuel economy and emissions from parallel hybrid and conventional diesel buses fitted with selective catalytic reduction systems. Appl Energy 2015;159:433–41. [45] Katrašnik T. Impact of vehicle propulsion electrification on well-to-wheel CO2 emissions of a medium duty truck. Appl Energy 2013;108(8):236–47. [46] http://en.wikipedia.org/wiki/Atkinson_cycle. [47] Auto Tech: Atkinson cycle engines and hybrids. Autos.ca. 2010-07-14 [retrieved 2013-02-23]. [48] Kawamoto N, Naiki K, Kawai T, Shikida T, Tomatsuri M. Development of new 1.8-L engine for hybrid vehicles. SAE paper 2009-01-1061; 2009. [49] Pertl P, Trattner A, Abis A, Schmidt S, et al. Expansion to higher efficiency investigations of the Atkinson cycle in small combustion engines. SAE paper 2012-32-0059; 2012. [50] Martins JJG, Uzuneanu K, Sousa B, Jasansky O. Thermodynamic analysis of an over-expanded engine. SAE paper 2004-01-0617; 2004. [51] Ribeiro B, Martins J. Direct comparison of an engine working under Otto, Miller and Diesel cycles: Thermodynamic analysis and real engine performance. SAE paper 2007-01-0261; 2007. [52] Abe S. Development of the hybrid vehicle and its future expectation. SAE paper 2000-01-C042; 2000. [53] Inoue T, Kusada M, Kanai H, Hino S, Hyodo Y. Improvement of a highly efficient hybrid vehicle and integrating super low emissions. SAE paper 200001-2930; 2000. [54] Muta K, Yamazaki M and Tokieda J. Development of new-generation hybrid system THS II – drastic improvement of power performance and fuel economy. SAE paper 2004-01-0064; 2004. [55] https://en.m.wikipedia.org/wiki/Miller_cycle. [56] Miller R. Supercharge and internally cooling for high output. ASME Trans 1947;69:453–64. [57] Hou SS. Comparison of performances of air standard Atkinson and Otto cycles with heat transfer considerations. Energy Convers Manage 2007;48:1683–90. [58] Wang C, Daniel R, Xu H. Research of the Atkinson cycle in the spark ignition engine. SAE paper 2012-01-0390; 2012. [59] Li T, Zheng B, Yin T. Fuel conversion efficiency improvements in a highly boosted spark-ignition engine with ultra-expansion cycle. Energy Convers Manage 2015;103:448–58. [60] Al-Sarkhi A, Jaber JO, Probert SD. Efficiency of a Miller engine. Appl Energy 2006;83:343–51. [61] Zhao Y, Chen J. Performance analysis and parametric optimum criteria of an irreversible Atkinson heat-engine. Appl Energy 2006;83:789–800. [62] Ebrahimi R. Effects of mean piston speed, equivalence ratio and cylinder wall temperature on performance of an Atkinson engine. Math Comput Model 2011;53:1289–97. [63] Ge Y, Chen L, et al. Performance of an Atkinson cycle with heat transfer, friction and variable specific-heats of the working fluid. Appl Energy 2006;83:1210–21. [64] Ge Y, Chen L, Sun F. Finite time thermodynamic modeling and analysis for an irreversible Atkinson cycle. Therm Sci 2010;14(4):887–96. [65] Ge Y, Chen L, Sun F. Progress in finite time thermodynamic studies for internal combustion engine cycles. Entropy 2016;18(4):139. [66] Zhu S, Deng K, Liu S, Qu S. Comparative analysis and evaluation of turbocharged Dual and Miller cycles under different operating conditions. Energy 2015;93:75–87. [67] Gonca G, Sahin B. The influences of the engine design and operating parameters on the performance of a turbocharged and steam injected diesel engine running with the Miller cycle. Appl Math Model 2015;40(5– 6):3764–82. [68] Gonca G. Thermodynamic analysis and performance maps for the irreversible Dual-Atkinson cycle engine (DACE) with considerations of

[69]

[70] [71]

[72] [73]

[74]

[75] [76] [77] [78] [79] [80]

[81] [82]

[83]

[84] [85]

[86] [87]

[88]

[89] [90]

[91] [92] [93]

[94] [95] [96] [97] [98]

[99]

[100] [101] [102] [103]

temperature-dependent specific heats, heat transfer and friction losses. Energy Convers Manage 2016;111:205–16. Zhang F, Okamoto K, et al. Improving performance and reliability of natural gas engines for power generation – a concept of realizing lower temperature and higher pressure combustion. SAE paper 1999-01-3511; 1999. Luria D, Taitel Y, and Stotter A. The Otto-Atkinson engine – a new concept in automotive economy. SAE paper 1982-02-01, 1982. Benajes J, Serrano JR, Molina S, Novella R. Potential of Atkinson cycle combined with EGR for pollutant control in a HD diesel engine. Energy Convers Manage 2009;50:174–83. US 367496. Atkinson J. Gas engine [issued 1887-08-02]. Watanabe S, Koga H, Kono S. Research on extended expansion generalpurpose engine – theoretical analysis of multiple linkage system and improvement of thermal efficiency. SAE paper 2006-32-0101; 2006. Kono S, Koga H, Watanabe S. Research on extended expansion generalpurpose engine-efficiency enhancement by natural gas operation. SAE paper 2010-32-0007; 2010. http://world.honda.com/powerproducts-technology/exlink/. Yamada Y. Engine of compression-ratio variable stroke. U.S. patent 6820577 B2; Nov. 2004. Boretti A, Scalzo J. Exploring the advantages of Atkinson effects in variable compression ratio turbo GDI engines. SAE paper 2011-01-0367; 2011. Wang Y, Lin L, Zeng S, Huang J. Application of the Miller cycle to reduce NOx emissions from the petrol engines. Appl Energy 2008;85:463–74. Wang Y, Lin L, Anthony P. An analytic study of applying Miller cycle to reduce NOx emission from petrol engine. Appl Therm Eng 2007;27:1779–89. Gonca G, Sahin B, Parlak A, et al. Theoretical and experimental investigation of the Miller cycle diesel engine in terms of performance and emission parameters. Appl Energy 2015;138:11–20. Fontana G, Galloni E. Variable valve timing for fuel economy improvement in a small spark-ignition engine. Appl Energy 2009;86:96–105. Fontana G, Galloni E, Torela E. The influence of variable valve timing on the combustion process of a small spark-ignition engine. SAE paper 2006-010445; 2006. Fiorenza R, Pirelli M, Palloti P, Kapus E. VVT + Port deactivation on a small displacement 4 cylinder 16 V engine: an effective way to reduce vehicle fuel consumption. SAE paper 2003-01-0020; 2003. Sellnau M, Sinnamon KJ, Burkhard J. 2-Step variable valve actuation: system optimization and integration on an SI engine. SAE paper 2006-01-0040; 2006. Sher E, Bar-Kohany T. Optimization of variable valve timing for maximizing performance of an unthrottled SI engine—a theoretical study. Energy 2002;27:757–75. Söderberg F, Johansson B. Load control using late intake valve closing in a cross flow cylinder head. SAE paper 2001-01-3554; 2001. Shiga S, Hirooka Y, Yagi S. Effects of over-expansion cycle in a spark-ignition engine using late-closing of intake valve and its thermodynamic consideration of the mechanism. Int J Auto Tech-Korea 2001;2(1):1–7. Okamoto K, Zhang F, Shimogata S, Shoji F. Development of a late intake-valve closing (LIVC) Miller cycle for stationary natural gas engines-effects of EGR utilization. SAE paper 972948; 1997. Fukuzawa Y, Shimoda H, Kakuhma Y. Development of high efficiency miller cycle gas engine. Tech Rev 2001;38(3):146–50. Boggs DL, Hilbert HS, Schechter MM. The Otto-Atkinson cycle engine: fuel economy and emissions results and hardware design. SAE paper 950089; 1995. Zhao J, Xu M, et al. Design and optimization of an Atkinson cycle engine with the artificial neural network method. Appl Energy 2012;92:492–502. Goto T, Isobe R, Yamakawa M, Nishida M. The new MAZDA gasoline engine SKYACTIV-G. MTZ Worldwide 2011;72(6):40–7. Wang Y, Du A. Reducing part load pumping loss and improving thermal efficiency through high compression ratio over-expanded cycle. SAE paper 2013-01-1744; 2013. Saunders RJ, Abdul-Wahab EA. Variable valve closure timing for load control and the Otto-Atkinson cycle engine. SAE paper 890677; 1989. Blakey SC, Saunders RJ. A design and experimental study of an Otto-Atkinson cycle engine using late intake valve closing. SAE paper 910451; 1991. Austin, WM. Variable stroke internal combustion engine. U.S. patent 1278563; Sep. 1918. Parvate-Patil GB, Hong H, Gordon B. An assessment of intake and exhaust philosophies for variable valve timing. SAE paper 2003-32-0078; 2003. Caton JA. Results from an engine cycle simulation of compression ratio and expansion ratio effects on engine performance. J Eng Gas Turb Power 2008;130(5):819–25. Miklanek L, Vitek O, Gotfryd O, Klir V. Study of unconventional cycles (Atkinson and Miller) with mixture heating as a means for the fuel economy improvement of a throttled SI engine at part load. SAE Int J Engines 2012;5 (4):1624–36. Tavakoli S, Jazayeri SA, Fathi M, Jahanian O. Miller cycle application to improve lean burn gas engine performance. Energy 2016;109:190–200. Martins MES, Lanzanova TDM. Full-load Miller cycle with ethanol and EGR: potential benefits and challenges. Appl Therm Eng 2015;90:274–85. http://www.gizmag.com/toyota-atkinson-engines-improved-thermal-fuelefficiency/31615/. Roberts M. Benefits and challenges of variable compression ratio (VCR). SAE paper 2003-01-0398; 2003.

J. Zhao / Applied Energy 185 (2017) 300–319 [104] Amjad S, Nsv M, Rudramoorthy R. Variable compression ratio engine: a future power plant for automobiles – an overview. Proc IMechE Part D: J Autom Eng 2007;221:1159–68. [105] Yates A. The Atkinson cycle revisited for improved part-load fuel efficiency? N&O J 1991(April):1–6. [106] Miklánek L. Tools for optimization of SI engine at part load Dissertation theses. Prague: CTU; 2011 (in Czech). [107] Kutlar OA, Arslan H, Calik AT. Methods to improve efficiency of four stroke, spark ignition engines at part load. Energy Convers Manage 2005;46:3202–20. [108] Anderson MK, Assanis DN, Filipi ZS. First and second law analyses of a naturally-aspirated, Miller cycle, SI engine with late intake valve closure. SAE paper 980889; 1998. [109] Heywood JB. Internal combustion engine fundamentals. New York: McGrawHill; 1988. [110] Gonca G, Sahin B, Parlak A, Ust Y, Ayhan V. The effects of steam injection on the performance and emission parameters of a Miller cycle diesel engine. Energy 2014;78(C):266–75. [111] Manivannan A. Numerical and experimental study of extended expansion concept applied to lean burn spark ignition engine. Int J Eng 2011;2:37–42. [112] Raynes HS. An Atkinson cycle engine for low pollution. SAE paper 984064; 1998. [113] Imperato M, Antila E, Sarjovaara T, Kaario O, et al. NOx reduction in a medium-speed single-cylinder diesel engine using miller cycle with very advanced valve timing. SAE paper 2009-24-0112; 2009. [114] Rinaldini CA, Mattarelli E, Golovitchev VI. Potential of the Miller cycle on a HSDI diesel automotive engine. Appl Energy 2013;112:102–19. [115] Imperato M, Kaario O, Sarjovaara T, Larmi M. Split fuel injection and Miller cycle in a large-bore engine. Appl Energy 2016;162:289–97. [116] Edwards SP, Frankle GR, Wirbeleit F, Raab A. The potential of a combined Miller cycle and internal EGR engine for future heavy duty truck applications. SAE paper 980180; 1998. [117] Luo QH, Sun BG. Effect of the Miller cycle on the performance of turbocharged hydrogen internal combustion engines. Energy Convers Manage 2016;123:209–17. [118] Imperato M, Nurmiranta J, Sarjovaara T, Larmi M, Wik C. Multi-injection and advanced Miller timing in large-bore CI engine. In: 27th CIMAC congress, paper no 157, Shanghai, China; 2013. [119] Kyrtatos P, Hoyer K, Obrecht P, Boulouchos K. Recent developments in the understanding of the potential of in-cylinder NOx reduction through extreme Miller valve timing. In: 26th CIMAC congress, paper no 225, Shanghai, China; 2013. [120] Park SH, Yoon SH. Injection strategy for simultaneous reduction of NOx and soot emissions using two-stage injection in DME fueled engine. Appl Energy 2015;143:262–70. [121] Gonca G, Sahin B, Ust Y, Parlak A. A study on late intake valve closing Miller cycled diesel engine. Arab J Sci Eng 2012;38(2):383–93. [122] Gonca G, Sahin B, Parlak A, et al. Application of the Miller cycle and turbo charging into a diesel engine to improve performance and decrease NO emissions. Energy 2015;93:795–800. [123] Gonca G, Sahin B, Ust Y, Parlak A, Safa A. Comparison of steam injected diesel engine and Miller cycled diesel engine by using two zone combustion model. J Energy Inst 2015;88(1):43–52. [124] Wang Y, Huang J, Huang X, Li S, Zeng S. Experimental investigation of applying Miller cycle to reduce NOx emission from diesel engine. Proc IMechE Part A: J Power Energy 2005;219(8):631–8. [125] Zhen X, Wang Y, Xu S, et al. The engine knock analysis – an overview. Appl Energy 2012;92:628–36. [126] Gamma Technologies. GT-power user’s manual, version 7.0 2009 (September).

319

[127] Gottschalk W, Lezius U, Mathusall L. Investigations on the potential of a variable Miller cycle for SI knock control. SAE paper 2013-01-1122; 2013. [128] Zhao J, Xu M. Fuel economy optimization of an Atkinson cycle engine using genetic algorithm. Appl Energy 2013;105:335–48. [129] Hirose K, Ueda T, Takaoka T, Kobayashi Y. The high-expansion-ratio gasoline engine for the hybrid passenger car. JSAE Rev 1999;20(1):13–21. [130] https://en.wikipedia.org/wiki/SkyActiv. [131] Ellies B, Schenk C, Dekraker P. Benchmarking and hardware-in-the-loop operation of a 2014 MAZDA SkyActiv 2.0L 13:1 compression ratio engine. SAE paper 2016-01-1007; 2016. [132] Benajes J, Molina S, Novella R, Belarte E. Evaluation of massive exhaust gas recirculation and Miller cycle strategies for mixing-controlled low temperature combustion in a heavy duty diesel engine. Energy 2014;71 (c):1–12. [133] Zhang F, Okamoto K, Morimoto S, Shoji F. Methods of increasing the BMEP (power output) for natural gas spark-ignition engines. SAE paper 981385; 1998. [134] Watson N, Janota M. Turgocharging the internal combustion engine. New York: John Wiley and Sons; 1982. [135] http://www.autox.in/quattroruote/bmw-m4-with-water-injectiontechnology/. [136] http://www.foxnews.com/leisure/2015/03/12/bmw-water-injectionsystem-gives-turbo-engine-boost/. [137] http://www.motortrend.com/news/toyotas-future-engines-employatkinson-cycle-for-max-efficiency/. [138] Bai Y, Wang Z, Wang J. Experimental research on knocking suppression using stratified stoichiometric mixture in a GDI engine. Trans CSICE 2009;27:534–9. [139] Goto T, Hatamura K, Takizawa S, et al. Development of V6 Miller cycle gasoline engine. SAE paper 940198; 1994. [140] Choshi M, Asanomi K, Abe H, Okamoto S, Shoji M. Development of V6 Miller cycle engine. JSAE Rev 1994;15(3):195–200. [141] Shimizu R, Fujii M, Suzuki T, Inoue M, Niinai S. Miller cycle engine management system and its distinctive feature. JSAE Rev 1994;15(4):305–8. [142] Miyagi Y, Takabe S, Miyashita K, Ikeva N. Experimental study of new Lysholm supercharger with a simple unloading system. SAE paper 960952; 1996. [143] https://en.wikipedia.org/wiki/Toyota_NR_engine#1NR-FKE. [144] Chen B, Wu Y, Tsai H. Design and analysis of power management strategy for range extended electric vehicle using dynamic programming. Appl Energy 2014;113(1):1764–74. [145] Zhang Z, Zhang X, Chen W, et al. A high-efficiency energy regenerative shock absorber using supercapacitors for renewable energy applications in range extended electric vehicle. Appl Energy 2016;178:177–88. [146] Bassett M, Fraser N, Brooks T, et al. A study of fuel converter requirements for an extended-range electric vehicle. SAE paper 2010-01-0832; 2010. [147] Ribeiro B, Brito FP, Martins J. A survey on electric/hybrid vehicles. SAE paper 2010-01-0856; 2010. [148] Opel Ampera. http://www.opel-ampera.com/index.php/mas/home. [149] Brooke L. Development of the 2011 Chevrolet volt – a new role for the ice. SAE Veh Electrif 2010;1(1):48–51. [150] Capaldi P. A high efficiency 10 kWe microcogenerator based on an Atkinson cycle internal combustion engine. Appl Therm Eng 2014;71(2):913–20. [151] Thiem M, Takeishi M, Miwa K. Introduction of micro combined heat and power generation unit for residential use in Germany. In: International gas union research conference. [152] Perez RE, Liu HHT, Behdinan K. Evaluation of multidisciplinary optimization approaches for aircraft conceptual design. In: 10th AIAA/ISSMO multidisciplinary analysis and optimization conference, 30 August–1 September 2004, Albany, New York.