Simulation Modeling Method and Experimental Investigation on the Uniﬂow Scavenging System of an Opposed-Piston Folded-Cranktrain Diesel Engine

energies

Article Simulation Modeling Method and Experimental Investigation on the Uniﬂow Scavenging System of an Opposed-Piston Folded-Cranktrain Diesel Engine

Fukang Ma 1,*, Zhenfeng Zhao 2, Yangang Zhang 1, Jun Wang 1, Yaonan Feng 1, Tiexiong Su 1, Yi Zhang 1 and Yuhang Liu 2

1 School of Mechanical and Power Engineering, North University of China, Taiyuan 030051, China; [email protected] (Y.Z.); [email protected] (J.W.); [email protected] (Y.F.); [email protected] (T.S.); [email protected] (Y.Z.) 2 School of Mechanical and Vehicle Engineering, Beijing Institute of Technology, Beijing 100081, China; [email protected] (Z.Z.); [email protected] (Y.L.) * Correspondence: [email protected]; Tel.: +86-351-3922020

Academic Editor: Chang Sik Lee Received: 20 December 2016; Accepted: 18 May 2017; Published: 20 May 2017

Abstract: The scavenging process for opposed-piston folded-cranktrain (OPFC) diesel engines can be described by the time evolution of the in-cylinder and exhaust chamber residual gas rates. The relation curve of in-cylinder and exhaust chamber residual gas rate is called scavenging profile, which is calculated through the changes of in-cylinder and exhaust chamber gas compositions determined by computational fluid dynamics (CFD) simulation. The scavenging profile is used to calculate the scavenging process by mono-dimensional (1D) simulation. The tracer gas method (TGM) is employed to validate the accuracy of the scavenging profile. At the same time, the gas exchange performance under different intake and exhaust state parameters was examined based on the TGM. The results show that the scavenging process from 1D simulation and experiment match well, which means the scavenging model obtained by CFD simulation performs well and validation of its effectiveness by TGM is possible. The difference between intake and exhaust pressure has a significant positive effect on the gas exchange performance and trapped gas mass, but the pressure difference has little effect on the scavenging efficiency and the trapped air mass if the delivery ratio exceeds 1.4.

Keywords: opposed-piston folded-cranktrain (OPFC) diesel engine; uniﬂow scavenging system; validation method; tracer gas method; computational ﬂuid dynamics (CFD)

1. Introduction The opposed piston two stroke (OP2S) engine concept can be traced back to late 1800s. Since then, many novel applications have been used on aircraft, ships and vehicles. In the first half of the 20th century, OP2S engines were developed in multiple countries for a wide variety of applications, however, modern emissions regulations stopped the widespread development of most two-stroke engines in the latter half of the 20th century [1]. In recent years, with the application of advanced design technology, modern analytical tools, materials and engineering methods, the emission problem no longer limits the successful design of a clean and efficient OP2S [2], so OP2S engines have again attracted intensive attention to improve engine efficiency and emission performance [3–7]. Compared with conventional engines, OP2S engines have many fundamental advantages [8]. The opposed-piston structure is characterized by two pistons reciprocating opposite each other in a common cylinder, which cancels the need for the cylinder head and valve mechanism, which leads to lower heat loss for higher wall temperatures of the two piston crowns compared to a cylinder head. The nearly

Energies 2017, 10, 727; doi:10.3390/en10050727 www.mdpi.com/journal/energies Energies 2017, 10, 727 2 of 18 symmetrical movement of opposed pistons also leads to excellent engine balance even for single cylinder configurations. In the uniflow scavenging process of an OP2S engine, two ideal models must be firstly introduced: the perfect displacement model and the perfect mixing model. The perfect displacement model is used to describe if the burned gases are pushed out by the fresh gases without any mixing. The perfect mixing model applies if the entering fresh mixture mixes instantaneously and uniformly with the cylinder contents. Since 1914, when Hopkinson proposed a perfect mixing scavenging model, lots of different models have been offered to draw conclusions about the quality of the scavenging process [9]. A model with good ability to reproduce the real process as well as low mathematical burden is always needed. Merker and Gerstle classified these models into single-zone, multi zone and fluid mechanical models [10]. In single-zone models, Benson and Brandham assumed that the cylinder content can be divided into two zones, a perfect displacement and a perfect mixing zone [11]. In the first phase, the scavenging process works as a perfect displacement while the second phase takes place as a perfect mixing process. Dang and Wallace combined perfect displacement with perfect mixing and short circuiting [12]. Multi-zone models divide the cylinder volume into two or more sub-volumes or systems. All sub-volumes or systems share a constant pressure but different zones can have different temperatures. The temperature of each sub-volume is constant and no heat transfer between zones is assumed. Several such models have been proposed by Streit, Borman and Wallace [13–15]. In recent years, a synthetic scavenging model for the overall characterization of the scavenging process of two-stroke engines was widely used. It gives the mass fraction of burned gases exhausted as a function of the mass fraction of burned gases in the cylinder. The model is usually used as input data in many mono-dimensional (1D) softwares [16]. Ferrara, Balduzzi and Vichi presented an innovative solution to reduce the short-circuit by designing a rotating valve positioned in correspondence with the exhaust port [17]. The scavenging model is used as input data for the second step of the 1D simulation. The new solution has been analyzed both in terms of global performance using 1D code, and by the fluid dynamics behavior of the system through computational fluid dynamics (CFD) simulations. Laget, Ternel and Thiriot used the synthetic scavenging model to study the influence of intake and exhaust pressure on the scavenging process in their preliminary design of a two-stroke uniflow diesel engine for passenger cars [18]. Tribotte, Ravet and Dugue compared three kinds of cylinder heads following the guideline of the synthetic scavenging model [19]. The results after redesign proved satisfactory. Pohorelsky, Brynych and Macek also used the synthetic scavenging model as a part of their 1D engine model to determine the air management requirements for a down-sized two-stroke engine [20]. For the improvement of scavenging efficiency the transient gas exchange simulation was carried out for multiple cases, including two intake port configurations at various back pressures in the exhaust system and two port timings [21]. The effects of exhaust back pressure, port timing and intake port layout on scavenging and trapped air mass in the cylinder all were investigated by transient CFD simulation including blow-down and scavenging. By three dimensional computational fluid dynamics under different intake pressures and engine speeds, Wang et al., evaluated the scavenging process by analyzing delivery ratio, trapping efficiency, scavenging efficiency and charging efficiency [22]. In addition, the in-cylinder flow motions, which play important roles in controlling the charge mixing and combustion process, were studied for different scavenging port designs. In order to achieve aggressive engine downsizing, a boosted uniflow scavenged direct injection gasoline engine concept has been proposed and researched by means of CFD simulation and demonstration in a single cylinder engine [23]. In summary, the scavenging model is always used in pre-design and some optimization work. From the literatures above, we can see most scavenging profiles are calculated by 1D code. The cylinder is divided into different zones and a combination of the perfect displacement model, perfect mixing model and short circuiting model is used to simulate the scavenging process. The mass change of fresh air and burned gas are used as the calculation method to obtain the profile, which is a model of the flow conditions in the cylinder closer to the reality. Energies 2017, 10, 727 3 of 18

In addition, trapping efficiency has been a difficult parameter to measure because trapping efficiency varies with the level of in-cylinder mixing, gas velocities, exhaust tuning and other physical phenomena that are difficult to predict. Numerous techniques involving in-cylinder sampling have been successfully implemented to measure trapping efficiency [24]. Although these techniques have proven to be reliable given that a significant portion of the cylinder contents is removed, or represented with each sample, it generally requires major engine modifications and a laboratory environment. In-cylinder sampling can be intrusive to the combustion process and is costly to carry out. The engine cylinder typically must be physically modified to accommodate a sampling valve. The biggest advantage of the tracer gas method (TGM) is that it does not require major engine modifications to execute it. Additionally, it is adaptable to field applications. The experimental setup is external to the engine, eliminating the need for any major engine modification [25–27]. In this paper, the scavenging profile is calculated through the changes of in-cylinder and exhaust chamber gas compositions by CFD simulation, which can reflect the flow conditions in the cylinder closer to the reality [28]. On the one hand the TGM validates the scavenging model, on the other hand it shows the rules influencing the scavenging process.

2. Uniﬂow Scavenging System of an OPFC Diesel Engine

2.1. OPFC Diesel Engine Concept As shown in Figure1, the cylinder of the opposed-piston folded-cranktrain (OPFC) diesel engine was horizontal and the injector was placed on the cylinder liner. Meanwhile, at the end of the liner there were gas ports, intake ports on one side and exhaust ports on the other side. Intake ports were used to deliver fresh air into the cylinder, and exhaust ports were used to deliver burnt gases out from the cylinder. There were two pistons placed in the same cylinder liner, and a combustion chamber was formed when two pistons moved to the most closed position, which is inner dead center (IDC). Rockers were placed either end of the cylinder body, and the upper end of the rocker was connected to piston by a small connecting rod, and the lower end of the rocker was connected to crankshaft by a big connecting rod. The reciprocating movements of pistons were driven by the crankshaft through the rockers and connecting rods. The pistons can be defined by the differences between control gas ports. The piston controlling the opening and closing of intake ports was defined as the intake piston and the piston controlling the opening and closing of exhaust ports was defined as the exhaust piston. Usually a phase difference between the intake piston and exhaust piston is an effective way to improve gas exchange performance in the OPFC engine by changing the scavenging duration and timing. At the same time, due to the opposed-piston motion phase difference, opposed-pistons on both sides can not arrive at each top dead center (TDC) simultaneously. When opposed-pistons are around IDC, the opposed-pistons’ relative velocity affects the isochoric combustion duration. When the opposed-pistons are approaching IDC, the greater the motion phase difference is, the smaller the relative velocity is. This is caused by the opposed-piston motion phase difference, which results in the exhaust piston moving away from the IDC, while the intake piston will be close to the IDC. There is a catch-up process between two pistons, and when the phase difference is bigger, the catch-up process will be longer [29,30]. The engine specifications are listed in Table1. Energies 2017, 10, 727 4 of 18 Energies 2016, 9, 727 4 of 18

FigureFigure 1. 1.OPFC OPFC dieseldiesel engineengine concept.concept.

TableTable 1. 1.Engine Engine speciﬁcations. specifications.

ParametersParameters (unit) (unit) Value NumberNumber of cylinders of cylinder (-)s (-) 22 BoreBore (mm) (mm) 100100 StrokeStroke (intake/exhaust) (intake/exhaust) (mm) (mm) 110/110110/110 NominalNominal compression compression ratio (-)ratio (-) 2222 Trapping compression ratio (-) 15.8 TrappingDisplacement compression (L) ratio (-) 15.8 3.4 Intake portDisplacement height (mm) (L) 243.4 ExhaustIntake port port height height (mm) (mm) 2824 Intake portExhaust circumference port height ratio (mm) (-) 0.7528 Exhaust port circumference ratio (-) 0.6 Intake port circumference ratio (-) 0.75 Intake port radial angle (◦) 15 ExhaustExhaust portport radial circumference angle (◦) ratio (-) 0.6 0 Opposed-pistonIntake motion port phase radial difference angle ( (°◦)CA) 1715 MaximumExhaust engine port power radial (kW) angle (°) 80 @ 24000 rpm · OpposedMaximum-piston engine motion moment phase (N differencem) (°CA) 420 @ 160017 rpm Maximum engine power (kW) 80 @ 2400 rpm 2.2. Scavenging Process CharacterizationMaximum engine moment (N·m) 420 @ 1600 rpm Figure2 shows the principle of the uniflow scavenging process of the OPFC diesel engine gas exchange.2.2. Scavenging The exhaustProcess C portharacterization open first before outer dead center (ODC) and a blow down discharge processFigure commences. 2 shows Thethe principle discharge of period the uniflow up to thescavenging time of theprocess scavenging of the OPFC port openingdiesel engine is called gas theexchange. free exhaust The exhaust period. port The intakeopen first port before also opens outer beforedead center ODC when(ODC) the and cylinder a blow pressure down discharge slightly exceedsprocess thecommences. scavenging The pressure. discharge Because period up the to burned the time gas of flowthe scavenging is toward port the exhaustopening port, is called which the nowfree has exhaust a large period. open area, The intake the exhaust port alsoflow open continuess before and ODC little backflowwhen the occurs.cylinder When pressure the cylinder slightly pressureexceeds isthe less scaven thanging the scavengingpressure. Because pressure the fresh burned air entersgas flow the is cylinder toward the and exhaust the scavenging port, which process now starts.has a This large flow open continues area, the as exhaust long as the flow intake continues port is and open little and thebackflow scavenging occurs. total When pressure the exceedscylinder thepressure pressure is less in the than cylinder. the scavenging As the pressure pressure rises fresh above air enters the exhaust the cylinder pressure, and the the fresh scavenging charge flowing process intostarts. the This cylinder flow continues displaces theas long burned as the gas: intake a part port of is the open fresh and air the mixes scavenging with the total burned pressure gases exceeds and is expelledthe pressure with them. in theThe cylinder. intake portAs the closes pressure after the rises exhaust above port the closes, exhaust and pressure, since the the flow fresh toward charge the intakeflowing port into is the continuous, cylinder displaces additional the fresh burned air is gas: obtained. a part of The the additional fresh air mixes air inflow with periodthe burned up to gas thees timeand ofis intakeexpelled port with close them. is called The theintake post port intake close period.s after the exhaust port closes, and since the flow toward the intake port is continuous, additional fresh air is obtained. The additional air inflow period up to the time of intake port close is called the post intake period.

Energies 2017, 10, 727 5 of 18 Energies 2016, 9, 727 5 of 18

FigureFigure 2. Gas2. Gas exchange exchange inin an OPFC OPFC diesel diesel engine. engine.

Delivery ratio, trapping efficiency and scavenging efficiency are usually employed as evaluation Deliveryindexes ratio, in the trappingtwo-stroke efficiency scavenging and system scavenging: efficiency are usually employed as evaluation indexes in theThe two-stroke delivery ratio: scavenging system: The delivery ratio: mass of delivered air (or mixture) per cycle l0  (1) reference mass The reference mass is massdefined of as delivered displaced airvolume(or mixture× ambient) airper (or cycle mixture). Ambient air (or l0 = (1) mixture) density is determined at atmosphericreference conditions mass or at intake conditions. The trapping efficiency: The reference mass is defined as displaced volume × ambient air (or mixture). Ambient air (or mass of delivered air (or mixture) retained mixture) density is determined at atmospheric  conditions or at intake conditions. tr mass of delivered air (or mixture) (2) The trapping efficiency: The trapping efficiency indicates what fraction of the air (or mixture) supplied to the cylinder is retained in the cylinder. mass of delivered air (or mixture) retained η = (2) The scavenging efficiency:tr mass of delivered air (or mixture) mass of delivered air (or mixture) retained sc  (3) The trapping efficiency indicates whatmass fraction of trapped of cylinder the air charge (or mixture) supplied to the cylinder is retained in the cylinder. The scavenging efficiency indicates to what extent the residual gases in the cylinder have been Thereplaced scavenging with fresh efficiency: air. When the reference mass in the definition of delivery ratio is the trapped cylinder mass (or mass of delivered air (or mixture) retained closely approximatedη by =it) then: (3) sc mass of trapped cylinder charge sc l0 tr (4)

The scavengingFor the perfect efﬁciency displacement indicates model, to trapping what extent and scavenging the residual efficiency gases vary in the with cylinder the delivery have been replacedratio with as freshfollows air. [31]: When the reference mass in the deﬁnition of delivery ratio is the trapped cylinder mass (or closely tr1 sc  ll00 for  1 (5) approximated by it) then: tr1ll00 sc  1 for  1 ηsc = l0 · ηtr (4)

For the perfect displacement model, trapping and scavenging efﬁciency vary with the delivery ratio as follows [31]: η = 1 η = l for l ≤ 1 tr sc 0 0 (5) ηtr = 1/l0 ηsc = 1 for l0 > 1 For the perfect mixing model, trapping and scavenging efﬁciency vary with the delivery ratio as follows: Energies 2017, 10, 727 6 of 18

1 −l ηtr = · 1 − e 0 l0 (6) −l ηsc = 1 − e 0

3. Uniﬂow Scavenging System Modeling

3.1. Working Process 1D Model Based on the hypothesis of one-dimensional isentropic flow, the fluid flow condition of free exhaust process can be written as a supercritical condition [31]:

s 1 dms Cv Fs 2gk 2 k−1 = · pz · ( ) (7) dϕ 6n (k − 1)RT pz + 1

The ﬂuid ﬂow condition of scavenging process can be written as a subcritical condition:

r s 2 k+1 dm C F 2gk p p k p k s = v s · √ s · ( z ) − ( z ) (8) dϕ 6n k − 1 RT ps ps where Cv is the intake or exhaust port flow coefficient, n is the engine speed, ps is the inlet pressure, pz is the outlet pressure. For the exhaust ports, ps is the in-cylinder pressure and pz is the exhaust chamber pressure; for the intake ports, ps is intake chamber pressure and pz is in-cylinder pressure. Fs is area of intake or exhaust port in different crank angle, g is gravitational acceleration, k is adiabatic exponent, R is gas constant, T is gas temperature. The uniflow scavenging process is assumed to be complete in the three models: perfect displacement model, perfect mixing model and short circuit model. In practice, the scavenging process includes multiple scavenging models, giving a relation for scavenging efficiency: ( l0 l0 ≤ l0c η = (9) sc −i·l 1 − e 0 l0 > l0c where the term i is the scavenging model index, and l0c is the demarcation point between perfect scavenging and rich exhaust scavenging. Section 3.3 shows the scavenging profile which was calculated by the three-dimensional (3D) CFD simulation as the input boundary conditions of 1D simulation scavenging model [30]. A simulation model based on GT-Power were established, Wiebe mode was used to describe the combustion process in the cylinder, and the Woschni mode was used to calculate the heat transfer in the cylinder [32]. The relationship between the residual gas coefficient in the exhaust and the residual gas coefficient in the cylinder is employed to describe the two-stroke scavenging process.

3.2. Scavenging Process 3D CFD Model AVL-Fire software is used to build the CFD model in the working process simulation [33]. Fame Engine plus is used to generate the moving meshes of the cylinder by defining movement selection, buffer selection, interpolation selection and the relative motion rule of the opposed-piston. The intake and exhaust chamber are generated as non-moving meshes which is refined near the intake and exhaust ports, in order to capture the significant flow gradients accurately, as shown in Figure3. The full-scale three-dimensional CFD model consists of 472,895 cells for the scavenging process and 76,419 cells for the compression process after rezoning. The dynamic mesh of piston motion in the intake and exhaust stroke has been treated according to the realistic motion rules of opposed pistons. The scavenging calculation is from exhaust port opening (EPO) to intake ports closing (IPC), while the in-cylinder working process is from IPC to EPO. Mesh movement includes three parts: intake and exhaust piston and cylinder to simulate the gas motion during the entire working process model which was used in the calculation of turbulence. Energies 2016, 9, 727 7 of 18 Energies 2017, 10, 727 7 of 18 Energies 2016, 9, 727 7 of 18

Figure 3. CFD calculation model. Figure 3 3.. CFD calculation model model.. The boundary conditions were chosen to reflect the physical conditions in the validation model Theand theboundary prototype conditions engine. Because were chosen the pressure to reflect reflect at the the intake physical ports conditions fluctuates, the in intakethe va validation pressurelidation ismodel and the obtained prototype prototype by an engine. engine.experimental Because Because method. the the pressure pressureTwo hundred at at the the intakecycles intake ofports portsintake fluctuates,fluctuates, pressure the measurement the intake intake pressure pressure are is obtainedis obtainedcarried by out. by an an Forexperimental experimental each cycle point, method. method. the average Two Two hundredvalue hundred is calculated cycles cycles of and of intake intake the result pressure pressure is shown measurement measurement as Figure 4. are carriedThe out. k-epsilon For For each each-xi-f cycle cyclemodel point, point, is employed the the average average to capture value value is is turbulence calculated calculated [3 and4 and]. The the the kresult result-epsilon is is- xishown shown-f model as as Fig Figurecanure 44.. The ksimulate-epsilon-xi--epsilon the-xi -intakef model vortex is and employed flow separation to to capture capture with turbulence turbulencehigher accuracy. [[3434]. ]With. The the k-epsilon-xi- -kepsilon-epsilon--xixi--ff model,model can the scavenging process with high pressure drop is calculated. The time step for the calculation is set simulate the intake vortex and flow flow separation with higher accuracy. With the k-epsilon-xi--epsilon-xi-f model,model, at about 0.5°. A constant pressure boundary condition is used for exhaust ports and a mean exhaust the s scavengingcavenging process process with with high high pressure pressure drop drop is is calculated. calculated. The The time time step step for forthe thecalculation calculation is set is receiver pressure◦ of 1 bar is taken. Frictional effects at the walls are not taken into account, i.e., the atset about atsmooth about 0.5°. wall 0.5A constant option. A constant for pressure turbulent pressure boundary flow boundary boundary condition condition condition is used is used. is for used Theexhaust forinitial exhaust ports conditions and ports a in mean cylinder and exhaust a mean receiverexhaustare receiverextractedpressure pressurefrom of 1 thebar ofGT is 1-taken.Power bar is softwareFrictional taken. Frictional simulation effects effectsat for the every atwalls the case. wallsare The not are flow taken not field taken into is initializedaccount, into account, byi.e., the i.e., smooththe smoothspecif wallying wall option the option temperature, for forturbulent turbulent pressure flow flow andboundary boundary turbulence condition condition intensity. is is Byused. used. performance The The initial initial prediction, conditions conditions the in ininitial cylinder cylinder are extractedpressure andfrom temperature the GT GT-Power-Power in the software cylinder are simulation computed for in everya scheme case. of 80The kW flow flow at an field field engine is initializedspeed of by specifspecifying2400ying rpm, the whichtemperature, are the initial pressure conditions and turbulence for CFD. Initial intensity. temperatures By performance of cylinder, prediction, intake chamber the initial pressureand exhaust and temperature chamber are in given the cylindera value of are 788 computed K, 322 K, and in 634a scheme K, respectively. of 80 kW at an engine speed of 2400 rpm, which are the initial conditions for CFD. Initial tem temperaturesperatures of cylinder, intake chamber and exhaust chamber are given a value of 788 K, 322 K, and 634 K,K, respectively.respectively.

Figure 4. Intake pressure curve.

In order to investigate mesh independence, two additional meshes are tested. One with approximately 150,000 cells denoted “coarse” and one with approximately 237 ,000 cells denoted Figure 4. Intake pressure curve. Figure 4. Intake pressure curve. In order to investigate mesh independence, two additional meshes are tested. One with approximately 150,000 cells denoted “coarse” and one with approximately 237,000 cells denoted

Energies 2017, 10, 727 8 of 18

EnergiesIn 2016 order, 9, to727 investigate mesh independence, two additional meshes are tested. One with approximately8 of 18 150,000 cells denoted “coarse” and one with approximately 237,000 cells denoted “medium”. The reference “medium”.mesh of 304,000 The cellsreference is referred mesh toof as304 “fine”,000 cells [35]. isThe referred intake to port as radial“fine” angle[35]. The (obliquity intake angle)port radial was anglethe between (obliquity of flow angle) direction was the and between radius of direction, flow direction which and could radius be used direction, to form which swirl. could For uniflowbe used toscavenging form swirl. system, For uniflow the intake scavenging swirl can system, increase the scavenging intake swirl flow can velocityincrease andscavenging improve flow scavenging velocity process.and improve Because scavenging of the intakeprocess. port Because radial of angle, the intake in-cylinder port radial flow angle, was swirl in-cylinder mainly flow in scavenging was swirl process.mainly in The scavenging tangential process. velocity The was tangential selected for velocity comparison. was selected The effect for comparison. of mesh resolution The effect is presented of mesh resolutionby comparing is presented radial profiles by comparing of tangential radial velocity profiles as of shown tangential in Figure velocity5. The as profiles shown arein Fig sampledure 5. The the profilescross section are sampled of cylinder the cross center section when theof cylinder opposed center piston when is at the the opposed ODC. The piston comparison is at the shows ODC. thatThe comparisonthe velocity profilesshows that are inthe good velocity agreement, profiles the are maximum in good agreement error is 3.3%, the and maximum the medium error meshis 3.3% can and be the medium considered mesh mesh. can be the considered mesh.

Figure 5 5.. ComparisonComparison between different mesh resolutions resolutions..

33.3..3. Scavenging P Profilerofile FigFigureure 66 showsshows thethe realreal scavengingscavenging processprocess fromfrom thethe CFDCFD simulation.simulation. TheThe redred colorcolor representsrepresents the burned gas while the blue color represents the fresh air. The The number numberss from from 0 0 to 1 for the color legend represents the percentage of burned gas. Before 120 °◦CA, the scavengi scavengingng ports are sealed by the piston. The cylinder and exhaust chamber are are filled filled with burned gas and the intake chamber is full of fresh charge. With the separation of two pistons, the intake port opens, and the fresh charge begins toto flowflow intointo cylinder. cylinder. The The portion portion of of fresh fresh charge charge increases increases until until 160 ◦160CA ° and,CA and, meanwhile, meanwhile, fresh freshcharge charge and burned and burned gas mix gas at mix the contactat the contact regions. regions. At 170 At◦CA, 170 a ° mixtureCA, a mixture of fresh of charge fresh andcharge burned and burnedgas exists gas in exists exhaust in exhaust chamber. chamber. The burned The burned gas still gas accounts still accounts for a great for proportiona great proportion in the exhaust in the exhaustchamber chamber and this alsoand meansthis also that means short that circuiting short happens.circuiting From happens. 180 ◦ CAFrom to 250180 ◦°CA,CA to the 250 fresh °CA charge, the freshconstantly charge replaces constantly the burned replaces gas the in burned the cylinder gas in and the exhaust cylinder chamber. and exhaust When chamber. the two pistonsWhen the seal two the pistonsscavenging seal portsthe scavenging once again, ports almost once all again, the burned almost gas all isthe displaced burned gas by freshis displaced charge. by We fresh can observecharge. Wethe variationcan observe of burnedthe variation gas and of freshburned charge gas and in cylinder fresh charge from in Figure cylinder6. from Figure 6. Since the scavenging process of two stroke engines occurs in a very short time, it is hard to know know the proportion proportion of of fresh fresh air andair and burned burned gas in-cylinder gas in-cylinder during oneduring cycle one under cycle experimental under experimental conditions. Wecondition assumes. We the assume fresh air the is comprisedfresh air is ofcomprised oxygen andof oxygen nitrogen and and nitrogen the fuel and in the cylinder fuel in is cylinder burned tois burnedcarbon dioxideto carbon and dioxide water. and Under water. full loadUnder conditions, full load condition through trackings, through the tracking variation the of varia oxygention and of oxygennitrogen, and the nitrogen, in-cylinder the residualin-cylinder ratio residual and exhaust ratio and chamber exhaust residual chamber ratio residual are calculated. ratio are calculated.

EnergiesEnergies 20162017, ,910, 727, 727 99 of of 18 18 Energies 2016, 9, 727 9 of 18

FigureFigure 6 6.. InIn-cylinder-cylinder scavenging scavenging process process of the of OPFC the OPFC engine engine.. Figure 6. In-cylinder scavenging process of the OPFC engine. Before exhaust port opening, the in-cylinder burnt gas is defined as a gas mixture which is made Before exhaust port opening, the in-cylinder burnt gas is deﬁned as a gas mixture which is made upBefore of H2O, exhaust CO2 and port N2. Asopening, shown inthe Figure in-cylinder 7a, the ratio burnt of gasin-cylinder is defined burnt as gas a gas composition mixture which remains is made up of H O, CO and N . As shown in Figure7a, the ratio of in-cylinder burnt gas composition up ofunchanged H2O,2 CO during2 and2 N the2. As free2 sho exhaustwn in Figurephase. During7a, the ratiothe early of in stages-cylinder of the burnt scavenging gas composition process, the remains unchangedremainscomposition unchanged during of H2 Othe during and free CO the2exhaust decrease free exhaustphase.s but the During phase. composition Duringthe earlyof O the2 isstages earlyincreased. stagesof the Because ofscavenging the the scavenging proportion process, process, the compositiontheof composition N2 in the of fresh H2O of charge and H2O CO is and 2greater decrease CO 2thandecreasess butin in the the composition but burnt the gas, composition the of N O2 composition2 is increased. of O2 isis increased.increasedBecause the slightly. Because proportion the ofproportion NDuring2 in the the offresh middle N2 incharge theand freshlater is greater stages charge ofthan is the greater inscavenging in the than burnt inprocess, in gas, the the burntthe in N-cylinder gas,2 composition the N N2 and2 composition O is2 compositionsincreased is increasedslightly. remain unchanged because the CO2 and H2O are completely expelled from the cylinder. As shown in Duringslightly. the During middle the and middle later stages and later of the stages scavenging of the process, scavenging the in process,-cylinder the N in-cylinder2 and O2 compositions N2 and O2 Figure 7b, the ratio of gas compositions in the exhaust chamber has the same change trend as the in- remaincompositions unchanged remain because unchanged the CO because2 and H the2O are CO 2completelyand H2O areexpelled completely from the expelled cylinder. from As the show cylinder.n in cylinder gas compositions. The primary difference is that the variation of gas compositions in the FigureAs shown 7b, the in Figure ratio 7ofb, gas the compositions ratio of gas compositions in the exhaust in thechamber exhaust has chamber the same has change the same trend change as the trend in- as theexhaust in-cylinder chamber gaslags compositions. behind that of the The gas primary compositions difference in the is cylinder that the by variation 40 °CA. When of gas the compositions ratio cylinderof gas gas compositions compositions. in exhaust The prima chamberry difference is changed, is thethat fresh the variation charge is of short gas-circuited compositions in the in the in the exhaust chamber lags behind that of the gas compositions in the cylinder by 40 ◦CA. When exhaustscavenging chamber process. lags behind that of the gas compositions in the cylinder by 40 °CA. When the ratio ofthe gas ratio compositions of gas compositions in exhaust in exhaust chamber chamber is changed, is changed, the fresh the fresh charge charge is is short short-circuited-circuited in in the the scavengingscavenging process. process.

(a) (b)

Figure 7. The variation of gas components in the scavenging process: (a) in-cylinder gas components; (b) exhaust chamber gas components. (a) (b) Through the abovementioned analysis, the state parameter of scavenging process can be Figure 7. The variation of gas components in the scavenging process: (a) in-cylinder gas components; obtainedFigure 7.byThe calculating variation the of gaschange components of fresh charge in the scavenging (O2 and N2 process:). The residual (a) in-cylinder gas coefficient gas components; in the (b) exhaust chamber gas components. cylinder(b) exhaust is described chamber by gas the components. O2 and N2 in the cylinder as follows:

Through the abovementioned analysis, the state parameter of scavenging process can be Through the abovementioned analysis, the state parameter of scavenging process can be obtained obtained by calculating the change of fresh charge (O2 and N2). The residual gas coefficient in the by calculating the change of fresh charge (O and N ). The residual gas coefﬁcient in the cylinder is cylinder is described by the O2 and N2 in the cylinder2 2 as follows: described by the O2 and N2 in the cylinder as follows:

Energies 2016, 9, 727 10 of 18

Energies 2017, 10, 727 10 of 18 76.8% O%O%2,cyl 2, cyl  1 23.2% (10) R, cyl + 76.8% O2,cyl %O2, Ncyl 2, cyl% % 23.2% CO 2, cylO %2,cyl  H% 2 O cyl % ηR,cyl = 1 − (10) O2,cyl% + N2,cyl% + CO2,cyl% + H2Ocyl% where O2,cyl%, N2,cyl%, CO2,cyl% and H2Ocyl% are the instantaneous mass percentages of O2, N2, CO2

2 andwhere H OO2, incyl the%,N cylinder.2,cyl%, CO 2,cyl% and H2Ocyl% are the instantaneous mass percentages of O2,N2, CO2 and HThe2O residual in the cylinder. gas coefficient in exhaust is described by the O2 and N2 in the exhaust as follows: The residual gas coefficient in exhaust is described76.8% by the O2 and N2 in the exhaust as follows: O%O%2,exh 2, exh  1 23.2%76.8% (11) R, exh O2,exh% + 23.2% O2,exh% ηR,exh = 1 − O2,exh % N 2, exh %  CO 2, exh %  H 2 O exh % (11) O2,exh% + N2,exh% + CO2,exh% + H2Oexh% where O2,exh%, N2,exh%, CO2,exh% and H2Oexh% are the instantaneous mass percentages of O2, N2, CO2 where O2,exh%,N2,exh%, CO2,exh% and H2Oexh% are the instantaneous mass percentages of O2,N2, and H2O in the exhaust. CO and H O in the exhaust. 2 As shown2 in Figure 8, the profile should be analyzed from 1.0 to zero. Before the intake port opens, As shown in Figure8, the profile should be analyzed from 1.0 to zero. Before the intake port no fresh charge passes into the cylinder. Both the cylinder residual ratio and exhaust chamber residual opens, no fresh charge passes into the cylinder. Both the cylinder residual ratio and exhaust chamber ratio remain at 1. With the outward movement of the pistons, the intake and exhaust ports are unsealed residual ratio remain at 1. With the outward movement of the pistons, the intake and exhaust ports are and a fresh charge flows into the cylinder. The burned gas is constantly replaced by fresh charge. When unsealed and a fresh charge flows into the cylinder. The burned gas is constantly replaced by fresh the cylinder residual ratio is 0.35, the synthetic scavenging profile begins to decline obviously. Through charge. When the cylinder residual ratio is 0.35, the synthetic scavenging profile begins to decline the whole process, the OPFC diesel engine scavenging profile remains above the perfect mixing curve obviously. Through the whole process, the OPFC diesel engine scavenging profile remains above which means the scavenging process of OPFC diesel engine is satisfactory. This profile will be used as the the perfect mixing curve which means the scavenging process of OPFC diesel engine is satisfactory. boundary condition of the 1D simulation to calculate the OPFC diesel engine scavenging efficiency and This profile will be used as the boundary condition of the 1D simulation to calculate the OPFC diesel trapping efficiency. engine scavenging efficiency and trapping efficiency.

Figure 8. Figure 8. TheThe uniflow uniﬂow scavenging scavenging profile proﬁle of the OPFC diesel engine engine..

44.. Experimental Investigation Investigationss

4.1. Tracer Gas Method 4.1. Tracer Gas Method The TGM was selected as the simplest method to implementedimplemented to measure the scavenging performance of a running diesel engine. As As described described in in the the [ [3366],], the the engine may may be idealized as as a steady ﬂowflow system.system. ThisThis methodmethod isis basedbased on on the the use use of of small small quantities quantities of of a gasa gas which which are are assumed assumed to toburn burn completely completely in the in combustion the combustion temperature temperature but not but to burn not toat temperatures burn at temperatures below those below prevailing those prevailingin the cylinder in the when cylinder the exhaust when the ports exhaust open. ports Figure open.9 shows Figure the 9 scavengingshows the scavengi efﬁciencyng test efficiency schematic test schematicusing the tracerusing gasthe method.tracer gas It consistsmethod. of It threeconsists parts: of tracerthree parts: gas injection tracer gas system, injection tracer system, gas sampling tracer gasand sampling analysis systemand analysis and OPFC system principle and OPFC prototype. principle prototype.

Energies 2017, 10, 727 11 of 18 EnergiesEnergies 20162016,, 99,, 727727 1111 ofof 1818

FigureFigure 99.. TGMTGM ttestest schematicschematic.. Figure 9. TGM test schematic.

AsAs shownshown inin FigureFigure 10a10a,, inin thethe tracertracer gasgas injectioninjection system,system, thethe tracertracer gasgas ((methanemethane)) isis injectedinjected intointo theAsthe shown intakeintake inmanifoldmanifold Figure 10 forfora, homogeneous inhomogeneous the tracer gas mixing.mixing. injection TheThe system, experimentexperiment the tracer utilizesutilizes gas (methane) aa reducingreducing is injectedvalve,valve, flowflow into-- meterthemeter intake and and manifold pressure pressure for stabilizing stabilizing homogeneous device device mixing. for for monitoring monitoring The experiment and and controlling controlling utilizes a thethe reducing gasgas flow. flow. valve, The The flow-meter methane methane injectionandinjection pressure pointpoint stabilizing isis locatedlocated deviceatat thethe entranceentrance for monitoring ofof thethe compressor.compressor. and controlling AfterAfter the thethe gas methanemethane flow. The isis injected,injected, methane itit injectionisis mixedmixed pointwithwith highhigh is located speedspeed at inin thetaketake entrance air.air. AsAs of shownshown the compressor. inin FigureFigure 10b10b After,, thethe the distancedistance methane betweenbetween is injected, thethe inlet itinlet ismixed collectingcollecting with pointpoint high andspeedand injection injection intake air. point point As shownisis approximateapproximate in Figurelyly 10 4.24.2b, the m.m. distance Thus, Thus, it it between can can be be concluded concluded the inlet collecting that that homogeneous homogeneous point and injection mixing mixing pointbetweenbetween is approximately tracertracer gasgas andand 4.2 intakeintake m. Thus, airair isis complete itcomplete can be concluded andand thethe exhaustexhaust that homogeneous collecticollectionon isis aboutabout mixing 0.80.8 between mm apartapart tracer fromfrom gasthethe intakeandintake intake.. TheThe air effecteffect is complete ofof intakeintake and collecticollecti theonon exhaust pointpoint positionposition collection onon is thethe about inletinlet 0.8 molarmolar m apart fractionfraction from isis the mainlymainly intake. embodieembodie The effectdd inin theofthe intake degreedegree collection ofof mixingmixing point ofof methanemethane position andand on intake theintake inlet,, ww molarhilehile thethe fraction exhaustexhaust is collecticollecti mainlyonon embodied positionposition influence ininfluence the degree onon thethe of exhaustmixingexhaust of molarmolar methane fractionfraction and isis intake, mainlymainly while reflectedreflected the exhaust inin thethe consumptionconsumption collection position andand combustioncombustion influence on ofof the tracertracer exhaust gasgas causedcaused molar byfractionby thethe dischargedischarge is mainly temperature.temperature. reflected in the InIn particular,particular, consumption thethe andexhauexhau combustionstst temperaturetemperature of tracer effecteffect gas isis negligiblenegligible caused by asas the thethe discharge exhaustexhaust temperaturetemperature.temperature isis In lowerlower particular, duringduring the thethe exhaust OPFCOPFC temperature dieseldiesel engineengine effect underloadunderload is negligible performance.performance. as the exhaust AfterAfter temperature filtfiltrationration andand is coollowercoolinging during,, gasgas samplsampl the OPFCeses cancan diesel bebe delivereddelivered engine underload toto aa FourierFourier performance. infraredinfrared analysisanalysis After filtration systemsystem and byby an cooling,an airair pumppump gas samples forfor onon-- linecanline berealreal delivered titimeme analysisanalysis to a Fourier.. infrared analysis system by an air pump for on-line real time analysis.

((aa)) ((bb))

FigureFigure 10.1010.. TGTGMTGMM t testtestest bed bed:bed:: ( (aa)) i injection,injection,njection, co collectioncollectionllection and and analysis analysis system system;system;; ( (bb)) OPFC OPFC diesel diesel engine.engineengine..

4.2. Scavenging Process Parameters 4.2.4.2. Scavenging Process ParametersParameters According to Equation (1), the delivery ratio can be calculated by Equation (12): AccordingAccording toto EquationEquation (1),(1), thethe deliverydelivery ratioratio cancan bebe calculatedcalculated byby EquationEquation (12):(12): GG G G/_/_ t t cycle cycle g g ll SS   S S   s s G00S GG VS/t_cycle  N  n V   Ngs  n ((1212)) l0 = =GRR V h h s s  N cyl cyl  n= V h h  s s  N cyl cyl  n (12) GR Vh · ρs · Ncyl · n Vh · ρs · Ncyl · n wherewhere ll00 isis thethe deliverydelivery ratioratio,, GGSS isis thethe massmass ofof delivereddelivered airair (or(or mixture)mixture) perper cyclecycle,, GGRR isis thethe referencereference massmass,, t_cyclet_cycle isis thethe timetime perper cycle,cycle, VVhh isis thethe displacementdisplacement perper cylindercylinder,, ρρss isis thethe intakeintake density,density, NNcylcyl isis thethe engineengine cylindercylinder number,number, nn isis thethe engineengine speedspeed,, andand ggss isis thethe intakeintake flowflow rate.rate.

Energies 2017, 10, 727 12 of 18

where l0 is the delivery ratio, GS is the mass of delivered air (or mixture) per cycle, GR is the reference mass, t_cycle is the time per cycle, Vh is the displacement per cylinder, ρs is the intake density, Ncyl is the engine cylinder number, n is the engine speed, and gs is the intake ﬂow rate. Intake density can be calculated by Equation (13):

ps M ρs = (13) RTs

where ps is the intake pressure, M is the molar mass, R is the molar gas constant, Ts is the intake temperature. As shown in Figure 11, Equations (14) and (15) can be derived. Energies 2016, 9, 727 12 of 18

mair = mtrap + mscav (14) Intake density can be calculated by Equation (13):

m W = mtrap,tracerpM+ mscavWtracer,scav (15) air tracer,air s s  (13) RT where mair is the mass of scavenging gas, mtrap is the masss of trapped gas, mscav is the mass of the short W m wherecircuit ps gas,is the tracer,airintake pressureis the mass, M fraction is the molar of tracer mass gas, inR theis the total molar scavenging gas constant air, trap,tracer, Ts is theis intake the mass temperatureof tracer gas. A ins shown the total in scavengingFigure 11, Eq air,uationWtracer,scavs (14) andis the (15 mass) can fractionbe derive ofd tracer. gas that is circulated.

FigureFigure 11.11. PrinciplePrinciple of TGM of TGM..

The tracer gas mass that is circulated equals m m the tracer m gas mass in the exhaust gas, so Equation (16) air trap scav (14) can be derived from Equation (15): m W m m W air tracer,air trap,tracer scav tracer,scav (15) mtrap,tracer = mairWtracer,air − mexhWtracer,exh (16) where mair is the mass of scavenging gas, mtrap is the mass of trapped gas, mscav is the mass of the short circuitwhere gasW, tracer,airWtracer,airis is the the mass mass fraction fraction of of tracer tracer gas gas that in isthe circulated, total scavengingmexh is the air mass, mtrap,tracer of tracer is the gas mass in the of exhausttracer gas gas, in Wthetracer,exh total scavengingis the mass air fraction, Wtracer,scav of tracer is the gas mass in fraction the exhaust of tracer gas. gas that is circulated. TheAccording tracer gas to mass the definitionthat is circu ofla trappingted equals efficiency, the tracer Equation gas mass (17) incan the beexhaust derived: gas, so Equation (16) can be derived from Equation (15): mtracer,exh m m W mexh m = G0 = trap,tracer = − exh tracer,exh = − exh ηtr G mmW  m1 Wm W m W 1 mtracer,air S airtrap,tracertracer,air air tracer,airair tracer,air exh tracer,exh mair (16) mair Mtracer,exhntracer,exh ! Mtracer,exh ! (17) + M n t,e M where Wtracer,air is= the1 −massm airfractionmfuel of tracerexh gasexh that is =circu1 −la(ted1 +, (mFexh/A is) the) mass ofexh tracer gas in the mair Mtracer,airntracer,air ov Mtracer,air t,i exhaust gas, Wtracer,exh is the mass fraction ofMair tracernair gas in the exhaust gas. Mair According to the definition of trapping efficiency, Equation (17) can be derived: where ηtr is the trapping efficiency, G0 is the mass of delivered (or mixture) air retained, mtracer,exh is m the mass of tracer gas in the exhaust gas, mfuel is the mass ofm fuel,tracer,exhMtracer,exh is the molar mass of tracer Gmm mW exh gas in the exhaust gas, n 0trap,traceris the molarityexh of tracer,exh tracer gas in the exh exhaust gas, M is the molar mass tr tracer,exh  11    exh G m W m W mtracer,air of exhaust air, nexh is theS molarityair tracer,air of exhaust air gas, tracer,airMtracer,airmis the molar mass of tracer gas in the total air m n air M scavenging air, tracer,air is the molarity of tracer gas in the total scavenging air, air is the molar(17 mass) Mn M of the intake air, nair is the molarity oftracer,exh the tracer,exhintake air, F is the fuel, A is thetracer,exh intake air, (F/A)ov is the m m M n  te, M  global fuel-air ratio,1Xt,e isair the fuel molar concentration exh exh 1 of  tracer 1  (FA / gas ) in  the intakeexh pipe,  Xt,i is the molar Mn ov M concentration of tracer gasm inair the exhaust tracer,air pipe. tracer,air  tracer,air    ti,   Mnair air  M air  where ηtr is the trapping efficiency, G0 is the mass of delivered (or mixture) air retained, mtracer,exh is the mass of tracer gas in the exhaust gas, mfuel is the mass of fuel, Mtracer,exh is the molar mass of tracer gas in the exhaust gas, ntracer,exh is the molarity of tracer gas in the exhaust gas, Mexh is the molar mass of exhaust air, nexh is the molarity of exhaust gas, Mtracer,air is the molar mass of tracer gas in the total scavenging air, ntracer,air is the molarity of tracer gas in the total scavenging air, Mair is the molar mass of the intake air, nair is the molarity of the intake air, F is the fuel, A is the intake air, (F/A)ov is the global fuel-air ratio, Xt,e is the molar concentration of tracer gas in the intake pipe, Xt,i is the molar concentration of tracer gas in the exhaust pipe. Since the molar masses of tracer gas in the intake gas and exhaust gas are equal. The trapping efficiency can be calculated by Equation (18):

XM  1  1  (FA / ) te, air tr ov  (18) XMti, exh

Energies 2017, 10, 727 13 of 18

Since the molar masses of tracer gas in the intake gas and exhaust gas are equal. The trapping efﬁciency can be calculated by Equation (18): Xt,e Mair Energies 2016, 9, 727 ηtr = 1 − (1 + (F/A)ov) 13 of(18) 18 Xt,i Mexh

ScavengingScavenging efficiency efﬁciency can can be be calculated calculated by by Eq Equationuation ( (4)4) after after the delivery ratio and trapping efficiencyefﬁciency are calculated.

44.3..3. Validation Validation of of S Scavengingcavenging M Modelodel ByBy comparing comparing the the experimental experimental data data and simulation and simulation results, results, it can be it found can that be thefound experimental that the experimentalresults approximate results approximate the perfect scavenging the perfect model scavenging and perfect model mixing and perfect model, mixing as shown model, in as Figure shown 12 . inIt shouldFigure 12. be It noted should that be the noted experimental that the experiment data are closeral data to are the closer perfect to the scavenging perfect scavenging model, indicating model, indicatingthe diesel enginethe diesel is superiorengine is insuperior gas transfer in gas quality. transfer The quality. simulation The simulation results are results shown are to shown be in to close be inagreement close agreement with the experimental with the experimental data and the data error and is less the thanerror 5%. is less Therefore, than 5%. it can Therefore, be concluded it can that be concludedin-cylinder that and exhaust in-cylinder coefficients and exhaust of residual coefficients gas are moreof residual accurate gas in theare representation more accurate of in OPFC the representationdiesel engine scavenging of OPFC diesel simulation. engine Figurescavenging 12b shows simulation. a gradual Figure increase 12b shows in scavenging a gradual efficiency increase in as scavengingthe delivery efficiency ratio grows. as the When delivery the delivery ratio grows. ratio exceeds When 1.4,the thedelivery effect ofratio delivery exceeds ratio 1.4, on the scavenging effect of deliveryefficiency ratio is weakened on scavenging and anyefficiency further is increaseweakened in and delivery any further ratio would increase only in worsendelivery the ratio pumping would onlylosses, worsen therefore, the pumping it is suggested losses, thattherefore, the delivery it is suggested ratio should that the be delivery controlled ratio within should 1.4. be Meanwhile, controlled withinwhen the 1.4. delivery Meanwhile, ratio when is less the than delivery 0.5, the ratio number is less of misfiringthan 0.5, cyclesthe number in the of experiments misfiring cycle growss in while the experimentthe torques declines, grows while resulting the torque in a worsening declines, resulting engine performance. in a worsening In engine view of performance. this, a delivery In view ratio ofranging this, a delivery from 0.5 ratio to 1.4 ranging is proposed from 0.5 for to an 1.4 optimization is proposed goalfor an taking optimization both stable goal operation taking both and stable less operationpumping lossesand less into pumping consideration. losses into consideration.

(a) (b)

FigureFigure 1 12.2. ResultResult compar comparisonison between between simulation simulation and and experimental experimental data data:: ( (aa)) t trappingrapping efficiency efﬁciency varivariationation with delivery ratio ratio;; and ( b) s scavengingcavenging efficiency efﬁciency variationvariation with delivery ratio.ratio.

4.4. Experimental Results and Analysis 4.4. Experimental Results and Analysis

44.4.1..4.1. E Effectffect of of P Pressureressure D Differenceifference on Scavenging P Processrocess CoolingCooling water temperature in the experiment is controlled atat 8080 ±± 22 °C.◦C. The The principle principle prototype prototype runsruns respectively atat 800 800 rpm, rpm 1200, 1200 rpm rpm and and 1500 1500 rpm rp withm with a stable a stable load ofload 20 ±of 12 N0 ·±m. 1 FigureN·m. Fi 13gure shows 13 showsthe effect the of effect pressure of pressure difference difference between between the intake the and intake exhaust and exhaust on delivery on delivery ratio at differentratio at different running runningspeeds. Delivery speeds. ratio Delivery increases ratio when increases the pressure when differencethe pressure between difference the intake between and exhaust the intake grows and for exhausta certain grows engine for running a certain speed. engine Delivery running ratio speed. declines Deliver wheny ratio the enginedeclines running when speedthe engine is accelerated running speedwith a is certain accelerated pressure with difference a certain pressure between difference the intake between and exhaust. the intake The following and exhaust. can explainThe following such a candifference: explain insuch the a case difference: of a certain in the engine case of running a certain speed engine and running scavenging speed duration, and scavenging a larger intake duration, and a larger intake and exhaust pressure difference is much more beneficial to fresh charge. Consequently, the delivery ratio is bigger; with a certain pressure difference between the intake and exhaust, the higher the engine running speed is, the shorter the scavenging duration is. As fresh charge is weakened consequently, the delivery ratio decreases. Therefore, scavenging pressure difference and engine running speed are the main factors affecting the delivery ratio.

Energies 2017, 10, 727 14 of 18 exhaust pressure difference is much more beneﬁcial to fresh charge. Consequently, the delivery ratio is bigger; with a certain pressure difference between the intake and exhaust, the higher the engine running speed is, the shorter the scavenging duration is. As fresh charge is weakened consequently, the delivery ratio decreases. Therefore, scavenging pressure difference and engine running speed are Energiesthe main 2016 factors,, 9,, 727 affecting the delivery ratio. 14 of 18

FigureFigure 13 13... EffectEffect of of pressure pressure difference difference on deliver deliveryy ratio ratio...

FigureFigure 1144 showshowss the the effect of pressure difference between the intake and exhaust on trapping efficiency.efficiency. With With a a certain certain engine engine running running speed, speed, trapping trapping efficiency efficiency decreases decreases when when the the pressure differencedifference between the intake and exhaust increases. This This is is due due to to an an increas increasee in in the the total total air air supply supply with the increase of pressure difference. Consequently, Consequently, the amount of gas captured in the cylinder is is relativelyrelatively smaller smaller than than that that of of intake intake a air.ir. Besides, Besides, with with the the certain certain pressure pressure difference, difference, the the lower lower the the engineengine running speedspeed is,is, thethe smaller smaller the the trapping trapping efficiency efficiency is is since since a lowera lower engine engine speed speed prolongs prolongs the thegas gas exchange exchange duration, duration, thus thus leading leading to a to larger a larger scavenging scavenging amount. amount.

FigureFigure 14 14... EffectEffect of of pressure pressure difference difference on on trapping trapping efficiency efﬁciency...

Figure 15 shows the effect of pressure difference between the intake and exhaust on scavenging efficiency. With a certain engine running speed, scavenging efficiency increases with a rising pressure difference. The scavenging efficiency remains almost unchanged when the pressure difference reaches a certain value beyond which it will only increase pumping losses rather than improve the scavenging efficiency. The experimental result shows that engines running at different speeds require different pressure differences to achieve the same scavenging efficiency and the higher the engine running speed is, the more pressure difference it requires. This is mainly due to a shorter scavenging duration caused by an increasing engine running speed, which hinders residual gas scavenging.

Energies 2017, 10, 727 15 of 18

Figure 15 shows the effect of pressure difference between the intake and exhaust on scavenging efficiency. With a certain engine running speed, scavenging efficiency increases with a rising pressure difference. The scavenging efficiency remains almost unchanged when the pressure difference reaches a certain value beyond which it will only increase pumping losses rather than improve the scavenging efficiency. The experimental result shows that engines running at different speeds require different pressure differences to achieve the same scavenging efficiency and the higher the engine running speed is, the more pressure difference it requires. This is mainly due to a shorter scavenging duration Energiescaused 2016 by,an 9, 727 increasing engine running speed, which hinders residual gas scavenging. 15 of 18

FigureFigure 15 15.. EffectEffect of of pressure pressure difference difference on on scavenging scavenging efficiency efﬁciency..

44.4.2..4.2. E Effectffect of P Pressureressure D Differenceifference on TrappingTrapping AirAir Figure 16 shows the effect of pressure difference and delivery ratio on trapping air mass. With With a certain engineengine runningrunning speed, speed, the the pressure pressure difference difference has has a greater a greater influence influence on trapping on trapping air mass air whenmass whenthe pressure the pressure difference difference between between the intake the and intake exhaust and isexhaust smaller. is As smaller. pressure As difference pressure difference increases, increases,the trapping the air trapping mass remains air mass almost remain unchangeds almost unchanged at different at engine different running engine speeds running once speed the pressures once thedifference pressure reaches difference a certain reaches value. a certain The result value. also The shows result that also the shows prototype that the running prototype at a running lower speed at a lowerrequires speed a relatively requires smaller a relatively pressure smaller difference pressure for difference a stable trapping for a stable air mass.trapping Such air a mass. result Such has an a resultaffinity has for an total affinity scavenging for total consumptionscavenging consumption during the scavenging during the scavenging process. The process. definition The of definition delivery ofratio delivery shows ratio that theshows larger that the the total larger scavenging the total mass scavenging is, the higher mass is, the the delivery higher ratio the delivery is. Delivery ratio ratio is. Deliveryis completely ratio determinedis completely by determined pressure difference by pressure during difference scavenging. during Thus, scavenging. the relation Thus, of delivery the relation ratio ofand delivery trapping ratio air and mass trapping can be used air mass to explain can be the used effect to explain of pressure the differenceeffect of pressure on trapping difference air mass. on trappingAssuming air exhaust mass. As backsuming pressure exhaust is consistent back pressure and theis consistent delivery ratioand the is “1” delivery in an idealratio is scavenging “1” in an idealsituation, scavenging a perfect situation scavenging, a perfect can be completedscavenging so can that be in-cylinder completed fresh so that charge in-cylinder reaches itsfresh maximum. charge reachesAs the delivery its maximum. ratio continues As the delivery to increase ratio while continues in-cylinder to increase fresh charge while remains in-cylinder stable, fresh extra charge gas remainsin the cylinder stable, extra is short-circuited. gas in the cylinder Since is the short OPFC-circuited. diesel Since engine the scavenging OPFC diesel approximates engine scavenging perfect approximatesscavenging and perfect perfect scavenging mixing scavenging, and perfect the mixing delivery scavenging, ratio should the delivery exceed “1”ratio consequently should exceed when “1” consequentlythe fresh charge when reaches the fresh the charge inflection reaches point. the Figure inflection 10 indicates point. Figure that 10 fresh indicates charger that under fresh different charger underrunning different speeds reaches running the speeds inflection reaches point the when inflection the delivery point when ratio isthe 1.4. delivery The pressure ratio differencesis 1.4. The pressurerequired difference for a deliverys require ratiod of for 1.4 a aredelivery respectively ratio of 0.075, 1.4 are 0.117 respectively and 0.158 0.075, MPa. 0.117 and 0.158 MPa.

(a) (b)

Figure 16. Trapping air mass effect in the scavenging process: (a) effect of pressure difference on trapping air mass; and (b) effect of delivery ratio on trapping air mass.

Energies 2016, 9, 727 15 of 18

Figure 15. Effect of pressure difference on scavenging efficiency.

4.4.2. Effect of Pressure Difference on Trapping Air Figure 16 shows the effect of pressure difference and delivery ratio on trapping air mass. With a certain engine running speed, the pressure difference has a greater influence on trapping air mass when the pressure difference between the intake and exhaust is smaller. As pressure difference increases, the trapping air mass remains almost unchanged at different engine running speeds once the pressure difference reaches a certain value. The result also shows that the prototype running at a lower speed requires a relatively smaller pressure difference for a stable trapping air mass. Such a result has an affinity for total scavenging consumption during the scavenging process. The definition of delivery ratio shows that the larger the total scavenging mass is, the higher the delivery ratio is. Delivery ratio is completely determined by pressure difference during scavenging. Thus, the relation of delivery ratio and trapping air mass can be used to explain the effect of pressure difference on trapping air mass. Assuming exhaust back pressure is consistent and the delivery ratio is “1” in an ideal scavenging situation, a perfect scavenging can be completed so that in-cylinder fresh charge reaches its maximum. As the delivery ratio continues to increase while in-cylinder fresh charge remains stable, extra gas in the cylinder is short-circuited. Since the OPFC diesel engine scavenging approximates perfect scavenging and perfect mixing scavenging, the delivery ratio should exceed “1” consequently when the fresh charge reaches the inflection point. Figure 10 indicates that fresh charger underEnergies 2017 different, 10, 727 running speeds reaches the inflection point when the delivery ratio is 1.4.16 The of 18 pressure differences required for a delivery ratio of 1.4 are respectively 0.075, 0.117 and 0.158 MPa.

(a) (b)

FigureFigure 16 16.. TrappingTrapping air air mass mass effect effect in in the the scavenging process process:: ( (aa)) e effectffect of of pressure pressure difference difference on on trappingtrapping air air mass; mass; and and ( (bb)) e effectffect of delivery ratio on trapping air mass mass..

5. Conclusions (1) The scavenging profile can be calculated by 3D CFD simulation through the changes of gas compositions. As a vital boundary condition, results from 1D simulation with correct scavenging profiles show good agreement with the experimental data. (2) The scavenging profile obtained by 3D CFD simulation can simulate the real scavenging process. Moreover, the relationship of residual gas coefficient in the exhaust and residual gas coefficient in the cylinder in a 1D scavenging model can characterize the scavenging process of OPFC engines with reasonable accuracy. (3) With the increase of pressure difference, the delivery ratio increases and the trapping efficiency reduces. To achieve the same delivery ratio under different running speeds, a higher running speed requires a greater pressure difference; when the pressure difference reaches a certain value (the delivery ratio is 1.4), the scavenging efficiency will gradually stabilize. At this time, any further increase of pressure difference cannot effectively improve the scavenging efficiency. (4) Pressure difference also has a major effect on trapping air mass. When the delivery ratio is less than 1.4, the trapping air mass will increase to a larger extent with the increase of pressure difference. However, when the delivery ratio exceeds 1.4, the effect of increasing pressure difference is not obvious or even decreases the trapping air mass. The scavenging process of an OPFC diesel engine is satisfactory and the delivery ratio range is 0.5–1.4.

Acknowledgments: The authors gratefully acknowledge the financial support by the National Natural Science Foundation of China (Grant No. 51605447), the Applied Basic Research Programs of Shanxi Province in China (Grant No. 201601D021085) and the National Ministry Fundamental Research Foundation of China (Grant No. B2220110005). Author Contributions: Fukang Ma and Zhenfeng Zhao designed the experiment set-up; Yaonan Feng and Jun Wang performed the simulation; Tiexiong Su and Yuhang Liu analyzed the data; Yi Zhang and Yangang Zhang contributed to the editing and reviewing of the document. Conflicts of Interest: The authors declare no conflict of interest.

Nomenclature

1D mono-dimensional IPC intake ports closing 3D three-dimensional ODC outer dead center CA crank angle OP2S opposed-piston two-stroke CFD computational ﬂuid dynamics OPFC opposed-piston folded-cranktrain EPO exhaust port opening TDC top dead center IDC inner dead center TGM tracer gas method Energies 2017, 10, 727 17 of 18

References

1. Pirault, J.P.; Flint, M. Opposed Piston Engines: Evolution, Use, and Future Applications; SAE International: Warrendale, PA, USA, 2010. 2. Naik, S.; Johnson, D.; Koszewnik, J. Practical Applications of Opposed-Piston Engine Technology to Reduce Fuel Consumption and Emissions; SAE Technical Paper 2013-01-2754; SAE International: Warrendale, PA, USA, 2013. 3. Hofbauer, P. Opposed Piston Opposed Cylinder (OPOC) Engine for Military Ground Vehicles; SAE Technical Paper 2005-01-1548; SAE International: Warrendale, PA, USA, 2005. 4. Hirsch, N.R.; Schwarz, E.E.; McGough, M.G. Advanced Opposed-Piston Two-Stroke Diesel Demonstrator; SAE Technical Paper 2006-01-0926; SAE International: Warrendale, PA, USA, 2006. 5. Herold, R.E.; Wahl, M.H.; Regner, G. Thermodynamic Benefits of Opposed-Piston Two-Stroke Engines; SAE Technical Paper 2011-01-2216; SAE International: Warrendale, PA, USA, 2011. 6. Regner, G.; Herold, R.E.; Wahl, M.H. The Achates Power Opposed-Piston Two-Stroke Engine: Performance and Emissions Results in a Medium-Duty Application; SAE Technical Paper 2011-01-2221; SAE International: Warrendale, PA, USA, 2011. 7. Redon, F.; Kalebjian, C.; Kessler, J. Meeting Stringent 2025 Emissions and Fuel Efficiency Regulations with an Opposed-Piston, Light-Duty Diesel Engine; SAE Technical Paper 2014-01-1187; SAE International: Warrendale, PA, USA, 2014. 8. Regner, G.; Johnson, D.; Koszewnik, J. Modernizing the Opposed Piston, Two-Stroke Engine for Clean, Efficient Transportation; SAE Technical Paper 2013-26-0114; SAE International: Warrendale, PA, USA, 2013. 9. Hopkinson, B. The charging of two-cycle internal-combustion engines. J. Am. Soc. Naval Eng. 2010, 26, 974–985. [CrossRef] 10. Merker, G.; Gerstle, M. Evaluation on Two Stroke Engines Scavenging Models; SAE Paper 970358; SAE International: Warrendale, PA, USA, 1997. 11. Benson, R.S.; Brandham, P.T. A method for obtaining a quantitative assessment of the influence of charging efficiency on two-stroke engine performance. Int. J. Mech. Sci. 1969, 11, 303–312. [CrossRef] 12. Dang, D.; Wallace, F.J. Some single zone scavenging models for two-stroke engines. Int. J. Mech. Sci. 1992, 34, 595–604. [CrossRef] 13. Sher, E.; Harari, R. A simple and realistic model for the scavenging process in a crankcase-scavenged two-stroke cycle engine. Proc. Inst. Mech. Eng. Part A J. Power Energy 1991, 205, 129–137. [CrossRef] 14. Streit, E.E.; Borman, G.L. Mathematical Simulation of a Large Turbocharged Two-Stroke Diesel Engine; SAE International: Warrendale, PA, USA, 1971. 15. Wallace, F.; Cave, P. Experimental and Analytical Scavenging Studies on a Two-Cycle Opposed Piston Diesel Engine; SAE Technical Paper 710175; SAE International: Warrendale, PA, USA, 1971. 16. Bozza, F.; Gimelli, A. A Comprehensive 1D Model for the Simulation of a Small-Size Two-Stroke SI Engine; SAE Paper 2004-01-0999; SAE International: Warrendale, PA, USA, 2004. 17. Ferrara, G.; Balduzzi, F.; Vichi, G. An Innovative Solution for Two-Stroke Engines to Reduce the Short-Circuit Effects; SAE Technical Paper 2012-01-0180; SAE International: Warrendale, PA, USA, 2012. 18. Laget, O.; Ternel, C.; Thiriot, J.; Charmasson, S.; Tribotte, P.; Vidal, F. Preliminary design of a two-stroke uniflow diesel engine for passenger car. SAE Int. J. Engines 2013, 6, 596–613. [CrossRef] 19. Tribotte, P.; Ravet, F.; Dugue, V.; Obernesser, P.; Quechon, N.; Benajes, J.; Novella, R.; De Lima, D. Two strokes

diesel engine-promising solution to reduce CO2 emissions. Proced. Soc. Behav. Sci. 2012, 48, 2295–2314. [CrossRef] 20. Pohorelsky, L.; Brynych, P.; Macek, J.; Vallaude, P.-Y.; Ricaud, J.-C.; Obernesser, P.; Tribottre, P. Air System Conception for a Downsized Two-Stroke Diesel Engine; SAE Technical Paper 2012-01-0831; SAE International: Warrendale, PA, USA, 2012. 21. He, C.; Xu, S. Transient Gas Exchange Simulation and Uniflow Scavenging Analysis for a Unique Opposed Piston Diesel Engine; SAE Technical Paper 2016-01-1087; SAE International: Warrendale, PA, USA, 2016. 22. Wang, X.; Ma, J.; Zhao, H. Evaluations of Scavenge Port Designs for a Boosted Uniflow Scavenged Direct Injection Gasoline (BUSDIG) Engine by 3D CFD Simulations; SAE Technical Paper 2016-01-1049; SAE International: Warrendale, PA, USA, 2011. 23. Ma, J.; Zhao, H. The Modeling and Design of a Boosted Uniflow Scavenged Direct Injection Gasoline (BUSDIG) Engine; SAE Technical Paper 2015-01-1970; SAE International: Warrendale, PA, USA, 2015. Energies 2017, 10, 727 18 of 18

24. Nuti, M.; Martorano, L. Short-Circuit Ratio Evaluation in the Scavenging of Two-Stroke S. I. Engines; SAE 850177; SAE International: Warrendale, PA, USA, 1985. 25. Olsen, D.B.; Holden, J.C.; Hutcherson, G.C.; Willson, B.D. Formaldehyde characterization utilizing in-cylinder sampling in a large bore natural gas engine. ASME J. Eng. Gas Turbines Power 2001, 123, 669–676. [CrossRef] 26. Matthew, G.M.; Robert, F. Experimental Investigation of the Scavenging Performance of a Two-Stroke Opposed-Piston Diesel Tank Engine; SAE Paper, 2004-01-1591; SAE International: Warrendale, PA, USA, 2004. 27. Olsen, D.B.; Hutcherson, G.C.; Willson, B.D.; Mitchell, G.E. Development of the tracer gas method for large bore natural gas engines—Part I method validation. J. Eng. Gas Turbines Power 2002, 124, 678–685. [CrossRef] 28. Wu, Y.; Wang, Y.; Zhen, X.; Guan, S.; Wang, J. Three-dimensional CFD (computational ﬂuid dynamics) analysis of scavenging process in a two-stroke free-piston engine. Energy 2014, 68, 167–173. [CrossRef] 29. Ma, F.K.; Zhao, C.L.; Zhang, S.L. Scheme Design and Performance Simulation of Opposed-Piston Two-Stroke Gasoline Direct Injection Engine; SAE Technical Paper 2015-01-1276; SAE International: Warrendale, PA, USA, 2015. 30. Ma, F.K.; Zhao, C.L.; Zhang, F.J. An experimental investigation on combustion and heat release characteristics of an opposed-piston folded-cranktrain diesel engine. Energies 2015, 8, 6365–6381. [CrossRef] 31. Taylor, C.F. The Internal-Combustion Engine in Theory and Practice: Thermodynamics, Fluid Flow, Performance, 2nd ed.; MIT Press: Cambridge, UK, 1985. 32. GT-Powerv.7.1. Engine Performance Application Manual; Gamma Technologies: Detroit, MI, USA, 2010. 33. AVL-Fire vs. CFD User Guide; AVL Technologies: Graz, Austria, 2012. 34. Basara, B.; Krajnovic, S.; Girimaji, S.; Pavlovic, Z. Near-wall formulation of the partially averaged navier stokes turbulence model. AIAA J. 2011, 49, 2627–2636. [CrossRef] 35. Sigurdsson, E.; Ingvorsen, K.M.; Jensen, M.V. Numerical analysis of the scavenge ﬂow and convective heat transfer in large two-stroke marine diesel engines. Appl. Energy 2014, 123, 37–46. [CrossRef] 36. McGough, M.G.; Fanick, E.R. Experimental Investigation of the Scavenging Performance of a Two-Stroke Opposed-Piston Diesel Tank Engine; SAE International: Warrendale, PA, USA, 2004.

© 2017 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).