Pinnacle is developing a multicylinder 1.2 L gasoline engine for automotive applications using high-performance computing (HPC) and analysis methods. Pinnacle and Oak Ridge National Laboratory executed large-scale multidimensional combustion analyses at the Oak Ridge Leadership Computing Facility to thoroughly explore the design space. These HPC-led investigations show high fuel efficiency (∼46% gross indicated efficiency) may be achieved by operating with extremely high charge dilution levels of exhaust gas recirculation (EGR) at a light load key drive cycle condition (2000 RPM, 3 bar brake mean effective pressure (BMEP)), while simultaneously attaining high levels of fuel conversion efficiency and low NOx emissions. In this extremely dilute environment, the flame propagation event is supported by turbulence and bulk in-cylinder charge motion brought about by modulation of inlet port flow. This arrangement produces a load and speed adjustable amalgamation of swirl and counter-rotating tumble which provides the turbulence required to support stable low-temperature combustion. At higher load conditions, the engine may operate at more traditional combustion modes to generate competitive power. In this paper, the numerical results from these HPC simulations are presented. Further HPC simulations and test validations are underway and will be reported in future publications.
Introduction
Increasingly stringent fuel-economy and emissions standards are compelling significant changes in gasoline-fueled internal combustion engines for automotive applications. Some of the strategies to achieve compliance with regulations are to run the engine with excess air or with recirculated exhaust gases as part of the combustion charge. The Pinnacle-opposed reciprocating piston with a four-stroke combustion cycle is a novel engine architecture which allows stable low-temperature combustion in a highly dilute combusting environment. Compared to conventional engines, Pinnacle's unique sleeve valve engine construction generates much higher levels of in-cylinder turbulence due to the collapse of counter-rotating tumble during the flame propagation event. Additionally, by equipping one crankshaft with a crankshaft phaser, Pinnacle's engine can manipulate the geometric compression ratio (CR) with piston phasing in a continuous fashion. This allows the engine to operate at a much higher CR at low- and midload operations and gradually lower CRs as load increases to avoid engine knock and lower peak cylinder pressures. The combination of higher turbulence and higher CR at a wide range of engine operating conditions enables this engine to operate at very highly dilute conditions compared to conventional piston engines, resulting in higher fuel economy and lower emission levels.
Overview of Pinnacle's Engine Technology
Pinnacle's second generation-opposed piston gasoline scooter engine utilizes a lean-burn combustion system and has shown in vehicle drive cycle bag emissions tests that uro 4 emissions can be easily met without NOx after treatment with acceptable combustion stability (<3% co-efficient of variation of indicated mean effective pressure (IMEP)) [1]. In this work, the authors are focused on developing a four-stroke opposed piston gasoline engine with a goal to meet the latest emissions standards such as Tier 3 Bin 160 and Euro 6 at fuel efficiency comparable to light duty diesel engines and at much lower costs. The work is focused on developing a high efficiency low-emission combustion recipe for a four-stroke, opposed piston gasoline engine for automotive applications (Fig. 1).
A highly dilute combustion recipe is chosen to achieve these goals. Spark-ignited (SI) gasoline engines are approaching fuel-economy levels typically achieved by compression ignition diesel counterparts by employing high compression ratio(s) (CR) with cooled exhaust gas recirculation (EGR) and direct-injection. A well-tuned combustion recipe that leverages higher CR to enhance low-temperature combustion burn rates can deliver a high effective expansion ratio (EER) together with low heat losses. High EER is essential to extract piston work most effectively and reduce energy losses to the exhaust stream. However, in SI engines, CR is often limited by the occurrence of knock.
In order to alleviate knock in high CR SI engines, a spark retarding strategy is often employed, which in turn reduces EER and increases exhaust gas energy losses. In previous studies, elevated levels of cooled EGR have proven effective to reduce the propensity of knock occurrence [2] in addition to increasing the fuel octane level. Higher dilution levels reduce flame temperature, which in turn reduces NOx emission. However, increased dilution levels also decrease laminar flame speed. To mitigate this or rather extend the dilution tolerance of this engine, it is important to manage charge motion to generate adequate in-cylinder turbulence to enhance flame speeds. In-cylinder turbulence in a four-stroke opposed piston engine is generated in part by in-cylinder shear planes that occur between cylinder halves during the induction stroke and enhanced as the pistons come together and cause the collapse of a counter-rotating dual tumble during the compression stroke (Fig. 2). Higher in-cylinder turbulence with a highly dilute combustion environment creates a novel combustion design uniquely positioned to deliver high fuel economy benefit at low emissions.

Presence of counter-rotating dual tumble in Pinnacle's 1.2 L engine. Streamline of in-cylinder charge motion at intake valve close.
Figure 2 shows the presence of counter-rotating dual tumble as visualized by streamlines of velocity in the tumble plane. Pinnacle's opposed piston architecture incorporates sleeve valves for controlling intake and exhaust flow. The sleeve valves each open a port near the center of the engine. Flow entering through the intake valve flows into both the “primary side” and the “secondary side” representing two halves of the combustion chamber corresponding to the opposed pistons' stroke-path. Two large scale, counter-rotating tumble eddies are produced from the induction stroke. During the compression stroke, the piston compresses these tumble motions resulting in higher small-scale turbulence if compared to collapse of single tumble motion of traditional single-piston engine architecture. This mechanism is responsible for creating a highly turbulent environment that can be taken advantage of in highly dilute combustion system to maintain or enhance burn rates relative to a less-dilute low turbulence system.
In-cylinder flow developed from the intake charge has vortex formation perpendicular to the piston motion. Figure 3 shows the flow velocity vector of in-cylinder flow motion in the swirling plane, generated during the intake stroke. It may be noticed that counter-rotating vertical swirl is generated in this case. By regulating the flow from one of the inlet port legs, it is possible to manipulate in-cylinder charge to generate net swirl.
Pinnacle's 1.2 L engine is designed with Variable Valve timing (VVT) as well as variable compression ratio (VCR) system. The VVT is capable of continuously adjusting intake and exhaust valve timing independently, whereas VCR is capable of continuously adjusting the phasing of the secondary side piston relative to primary side piston, giving ability to adjust compression ratio. Because of the ability to independently control the intake and exhaust valve phasing, the internal residual charge (internal EGR) can be controlled with VVT. Combining VCR with VVT makes it possible to couple a load and speed-dependent high geometric compression ratio (and therefore expansion ratio) with delayed exhaust valve open (EVO) event. The naturally aspirated performance goals for this engine in a nonhybrid vehicle motivated the selection of a late (Atkinson) inlet lift profile to reduce pumping work and effective compression ratio, but an early closing inlet profile (Miller) is also feasible to leverage depending on the performance requirements [3]. Figure 4 shows example intake and exhaust valve profiles and timing ranges.
The VCR system is used to increase CR at low-load operation with high dilution levels. At high-load operation, the engine operates at lower charge dilution levels and to avoid knock, compression ratio can be lowered by lagging the secondary piston's arrival to its highest position with respect to the primary piston. The specifications of the 1.2 L engine are given in Table 1. In traditional single-piston engine architectures, VCR capability is often limited to a limited discrete number of piston adjustment mechanisms [4]. This crank phasing method in an opposed-piston architecture enables continuously adjustable CR by leveraging already proven cam phaser technologies.
Engine specification of Pinnacle's 1.2 L engine
Bore/stroke (single piston) ratio | 0.937 | Fuel delivery | Port fuel injection |
---|---|---|---|
Peak surface to volume ratio (1/cm) | 4.76 | In-cylinder flow system | Dual tumble and adjustable Swirl (Swumble) |
Compression ratio system | VCR—continuous adjustment between 10:1 and 20:1 CRs | Valve timing | VVT—continuous adjustment of Cam phasing |
Ignition system | Synchronously fired dual spark plugs | Combustion system | Spark ignition with cooled EGR |
Bore/stroke (single piston) ratio | 0.937 | Fuel delivery | Port fuel injection |
---|---|---|---|
Peak surface to volume ratio (1/cm) | 4.76 | In-cylinder flow system | Dual tumble and adjustable Swirl (Swumble) |
Compression ratio system | VCR—continuous adjustment between 10:1 and 20:1 CRs | Valve timing | VVT—continuous adjustment of Cam phasing |
Ignition system | Synchronously fired dual spark plugs | Combustion system | Spark ignition with cooled EGR |
Schematic of the 1.2 L engine system is given in Fig. 5. The primary throttle regulates the amount of fresh inlet charge mixed with cooled EGR into the intake manifold. A secondary throttle placed in one of the legs of each inlet port regulating the type of in-cylinder flow. When the secondary throttle is closed, the flow from one of the inlet port legs is shut off, generating a high swirl motion inside the cylinder enhancing dilution tolerance. On the other hand, when the secondary throttle is open, flow from both inlet port legs occurs resulting in no net in-cylinder swirl. The in-cylinder swirl is continuously modulated by the secondary throttle over different loads and speeds. Regardless of the secondary throttle control, the ensuing dual tumble in-cylinder charge motion persists (Fig. 2). When the swirl is present along with the dual tumble, a new mode of in-cylinder charge motion is generated. This co-existing swirl and dual tumble in-cylinder charge will be referred to as “swumble” flow in this paper. There are two spark plugs located inside each of the cylinder of combustion chamber. The spark plugs are located on the side of the cylinder due to absence of a cylinder head in the opposed-piston engine architecture.
Analysis-Led Development of Combustion System
In this work, multidimensional engine computational fluid dynamics (CFD) simulations are performed using the commercially available CONVERGE software package [5]. The converge software can perform numerical solutions for a chemically reactive, two-phase flow problems with moving boundary to emulate piston and valve motions. converge uses a structured, orthogonal, finite volume computational grid with automatic mesh refinement and mesh-embedding methodologies to reduce numerical viscosity of the multidimensional flow solution. It uses a time-splitting numerical scheme for the flow solver with a first-order Euler in time and a second-order accurate central difference scheme in space.
In this paper, the multidimensional numerical investigation is conducted using the renormalization group k–ε turbulence model [6] and detailed chemistry solver for combustion chemistry [7]. The law of wall model is used for near-wall physics [8] and zero-dimensional (0D) crevice model to represent gas dynamics between cylinder charge and piston crevice regions [9]. For ignition, an energy dump model is chosen with prescribed inductive coil energy dump profile based on prior experimental measurements. The NOx formation and reduction is captured using extended Zel'dovich reaction kinetics [10].
A best-practice method of an analysis-led design and development process has been developed for the 1.2 L development program. The engine design is carried out in parallel with results from high-fidelity numerical models. In order to simulate thermal and fluid phenomena, a one-way coupled approach is adopted using a one-dimensional (1D)/0D system level simulation (performed using the GT-POWER simulation package, Chicago, IL), and the multidimensional numerical tool where spatial resolution is needed. This approach, portrayed in Fig. 6, shows that boundary conditions are generated by the 1D/0D system models. The turbulent reactive combustion simulation with higher fidelity physical submodels is performed to predict combustion, emissions, and combustion knock characteristics using CFD simulations. Results from CFD results are used to improve the accuracy of the system level models.
Using the analysis-led design and development process, massively parallel simulations of many designs and control degrees-of-freedom (DOFs) have been launched at the Oak Ridge Leadership Computing Facility (OLCF). In this work, meta-models (model of models) are constructed based on the results from the large number of parallel CFD results [11]. The meta-models are constructed using statistical model construction principles of deriving response functions from large set of sample results.
For example, a large number of numerical design of experiments cases are conducted with design and control degrees-of-freedom such as piston crown shapes, port design, EGR level, spark timing, swirl level, CR, valve timing, and engine speed and load. Based on the CFD results, a set of response models for individual parameters of interest (such as gross-indicated efficiency (GIE), fuel-specific nitrogen oxides (fsNOx), fuel-specific HC (fsHC), knock, combustion noise, co-efficient of variation of IMEP) are constructed relative to other parameters of interest (such as speed/load, 50% mass fraction burned (CA50), swirl, tumble, etc.).
In this work, quadratic functions are used to construct the response models. The collection of all such response functions is referred to here as a meta-model. In this work, meta-models are used to rapidly explore continuously a much wider range of design options and control authority in order to arrive at an optimal engine design within a set of constraints (such as emission, noise, and/or knock). This model-based prediction and optimization workflow is shown in Fig. 7. The robustness of the optimal solution is also calculated using Monte Carlo simulation based on meta-models.
An example of the influence of two parameters, namely combustion duration and phasing, on heat release is shown in Fig. 8. As the resultant combustion phasing and many other parameters are unknown at design of experiment generation, the spark timing is leveraged as a degree-of-freedom to modulate the resultant combustion phasing across bounds where it is expected to perform optimally. Subsequently, the meta-model is utilized to precisely identify the combination of conditions where a constrained optimum is found. The influence of EGR dilution level on both lowering the peak heat release rate and extending combustion phasing is also visible from these results.
Results and Discussion
This section highlights the results from the numerical simulation of Pinnacle's 1.2 L development engine. In this section, two port designs and two piston crown design will be discussed. The effect of charge motion on dilution tolerance, fuel efficiency, and emissions is also discussed. Finally, the capability of the VCR to increase fuel efficiency will be presented.
Intake Port and Charge Motion.
The design of the intake port is critical to the development of in-cylinder charge motion. In this study, two families of inlet ports were used, one symmetric in design and the other which used different design concepts of each flow leg (Fig. 9). Each port housed a secondary throttle in one of the legs and port fuel injector on the other (Fig. 5).

Inlet port designs for 1.2 L engine: (a) asymmetric inlet port legs and (b) symmetric inlet port legs
The asymmetric inlet port design uses dissimilar hydraulic flow areas between its port legs, while the symmetric port design maintains equivalent areas. Figure 10 shows a top view of the in-cylinder flow at spark timing for each port with fully open and fully closed throttle at a plane coinciding with the location of the spark plugs. The results show the asymmetric port design is able to generate some level of in-cylinder swirl even without the restriction of the secondary throttle due to preferred flow between the port legs. With a restriction added by the secondary throttle closing, the swirl generated by the asymmetric inlet port is higher than the symmetric port design. Figure 11 shows tradeoff between in-cylinder turbulence (quantified by modeled turbulent kinetic energy at minimum cylinder volume) and swirl. The asymmetric port can provide higher turbulence than the symmetric port at low swirl levels and deliver a higher swirl magnitude when the secondary throttle is closed. When the secondary throttle is fully open and inlet port legs deliver similar flow (Fig. 3) and pure dual tumble flow is generated inside the combustion chamber. In both flow configurations, the bulk in-cylinder turbulence is generated by the collapse of the dual tumble (Fig. 2). When the swirl is present, it acts to supplement the in-cylinder turbulence to enhance flame kernel development. In asymmetric inlet port leg design, the dynamics of the dual tumble collapse is undisturbed by the presence of net swirl. Because of its higher swirl capacity, the asymmetric inlet port provides an in-cylinder flow mode that transports the flame propagation tangentially around the combustion chamber. It also provides the dual-tumble collapse turbulent energy necessary to enhance flame speed in tumble mode. Hence, this port provides multiple modes with which to extend dilution tolerance.
It is also important to confirm the valve flow coefficient curve is sufficient to achieve the target full throttle power and torque requirements. Figure 12 shows the normalized flow coefficient against valve lift position for the inlet port designs with secondary throttle wide open. Though the asymmetric port design has a lower flow coefficient, both port designs provide the necessary flow to achieve performance targets. Overall, as the asymmetric port satisfies the flow requirement and offers advantages in both turbulence and high swirl (which both enable extended highly dilute low-temperature combustion), the asymmetric inlet port design is selected for further combustion studies.
Charge Motion—Mixing Field and Flame Propagation Behavior.
The effect of swirl on in-cylinder mixing field and flame propagation is briefly discussed in this subsection. Swirl with tumble (referred to as swumble) has a profound effect on in-cylinder mixing and flame kernel generation. It is observed from the spatial distribution of in-cylinder residual gases (Fig. 13) that pure tumble flow field generates more evenly distributed in-cylinder mixing than swumble, while swumble provides radial stratification in residual gas. Spatial stratification in charge dilution, and in-cylinder residuals in swumble flow configuration can thus be used to design the combustion recipe including pathways for flame development.
Figure 14 shows flame propagation behavior during early kernel growth. In this figure, the flame is depicted by an iso-temperature surface colored by local turbulent kinetic energy. It shows that the swumble system promotes kernel growth along the periphery of the cylinder chamber. In the swumble system, flame is assisted along the peripheral region due to presence of mean cylinder velocity (swirl), and followed by an inward flame propagation toward the center of the cylinder with chemical conversion. Conversely, in the tumble system, the flame propagates toward the center of the combustion chamber from the two spark plugs placed diametrically opposite in the cylinder and is promoted by turbulence enhancement.
Flame propagation to the periphery of the chamber serves an important role in combustion efficiency and HC emissions which are most prevalent in the geographic location in the cylinder that the flame reaches last. It is observed that the presence of swumble helps in mitigating HC emissions by facilitating kernel development in the regions where flame has the highest propensity to quench.
Results From Massively Parallel Multidimensional Combustion Simulations at 2000 RPM, 3 Bar Brake Mean Effective Pressure.
A computational campaign has been carried out to survey performance, emission, knock, and stability limits of the engine operation at the 2000 RPM and 3 bar BMEP (Table 2). At each of these cases, multiple simulations are carried out with spark timing sweeps to understand the effect of CA50 variations and ascertain the response of important combustion and emission parameters against combustion phasing. Between cases 1–12, the net in-cylinder EGR is varied between 36% and 42%. Both swumble and tumble combustion systems are investigated along with two different piston crown shapes (Fig. 15). The piston crown shapes chosen for this study are both axisymmetric about the bore axis and identical inlet to exhaust side of the opposed piston combustion chamber. The difference between these piston crowns is most prominent the central riser as the SIM0336 design offers a piston-to-piston squish zone and intended to minimize flame touch-down with a swumble/peripheral combustion flame propagation. An 18:1 CR is maintained for both pistons when cranks are phased in-sync. The CR can be reduced from 18:1 by the action of a VCR crank phaser.
Dilution level, flow configuration, and piston crown investigated at the 2000 RPM, 3 bar BMEP condition
Case ID | EGR (%) | In-cylinder flow configuration | CR | Piston crown ID |
---|---|---|---|---|
1 | 42 | Swumble | 18:1 | SIM0335 |
2 | 42 | Tumble | 18:1 | SIM0335 |
3 | 42 | Swumble | 18:1 | SIM0336 |
4 | 42 | Tumble | 18:1 | SIM0336 |
5 | 37 | Swumble | 18:1 | SIM0335 |
6 | 46 | Swumble | 18:1 | SIM0335 |
7 | 37 | Tumble | 18:1 | SIM0335 |
8 | 46 | Tumble | 18:1 | SIM0335 |
9 | 37 | Swumble | 18:1 | SIM0336 |
10 | 46 | Swumble | 18:1 | SIM0336 |
11 | 37 | Tumble | 18:1 | SIM0336 |
12 | 46 | Tumble | 18:1 | SIM0336 |
21 | 42 | Partial Swumble | 18:1 | SIM0336 |
22 | 42 | Partial Swumble | 17:1 | SIM0336 |
23 | 42 | Partial Swumble | 16:1 | SIM0336 |
Case ID | EGR (%) | In-cylinder flow configuration | CR | Piston crown ID |
---|---|---|---|---|
1 | 42 | Swumble | 18:1 | SIM0335 |
2 | 42 | Tumble | 18:1 | SIM0335 |
3 | 42 | Swumble | 18:1 | SIM0336 |
4 | 42 | Tumble | 18:1 | SIM0336 |
5 | 37 | Swumble | 18:1 | SIM0335 |
6 | 46 | Swumble | 18:1 | SIM0335 |
7 | 37 | Tumble | 18:1 | SIM0335 |
8 | 46 | Tumble | 18:1 | SIM0335 |
9 | 37 | Swumble | 18:1 | SIM0336 |
10 | 46 | Swumble | 18:1 | SIM0336 |
11 | 37 | Tumble | 18:1 | SIM0336 |
12 | 46 | Tumble | 18:1 | SIM0336 |
21 | 42 | Partial Swumble | 18:1 | SIM0336 |
22 | 42 | Partial Swumble | 17:1 | SIM0336 |
23 | 42 | Partial Swumble | 16:1 | SIM0336 |
Fuel Efficiency Maps.
The results from meta-model correlations of gross-indicated efficiency are presented for the different engine designs studied (i.e., flow configuration and piston crown) across EGR and CA50 phasing degrees-of-freedom. These GIE maps are overlaid with contours of knock and unacceptable combustion stability regions. The combustion stability limits are derived based on correlation of spark to CA10 timing based on previous experimental analysis. The knock is correlated based on crank-angles resolved local pressure data from several virtual sensors placed across the combustion chamber at minimum volume. Knock is directly predicted from the multidimensional simulations using detailed chemistry and surrogate fuel formulations. The numerical model and methods used in this paper is further elaborated in a different publication by the authors in this technical conference. Figure 16 shows these GIE maps for tumble and swumble configurations, respectively, for the piston crown SIM0336 at one compression ratio.

Gross-indicated efficiency map overlaid with regions of unstable combustion (gray) and knock (black) with piston SIM0336 in tumble (a) and swumble (b) configurations
The key takeaway is that while the swumble system has significantly faster combustion and an improved dilution tolerance, there is an associated gross fuel efficiency penalty compared to tumble system because of higher heat losses from the higher wall flow velocities. Maximum gross-indicated efficiency occurs at CA50 between 5 and 10 deg after fired minimum volume with 42% EGR (tumble) and 44% (swumble).
Gross-indicated efficiency maps of the piston crown SIM0335 in tumble and swumble configurations (Fig. 17) are also presented to study the effect of piston crown shapes on efficiency, knock, and combustion stability. It is concluded that there is little difference in fuel efficiency between these two piston crown shapes. However, it is observed that knock propensity can be manipulated effectively with the piston crown shapes. This provides insight that future studies on piston crown shapes may be leveraged to reduce knock propensity.

Gross-indicated efficiency map overlaid with regions of unstable combustion (gray) and knock (black) with piston SIM0335 in tumble (a) and swumble (b) configurations
Effect of Swirl Ratio on Gross-Indicated Efficiency Map.
Since swirl can be manipulated by the secondary throttle restriction level, a follow-up investigation was conducted with partially open secondary throttle position. By modulating swirl at a fixed dilution level, the effect of swirl against GIE, knock, and combustion stability is presented (Fig. 18). It is seen that gross-indicated efficiency is improved up to a swirl level of approximately 3, after which GIE reduces due to increased wall heat transfer. It is also observed that at swirl ratio of ∼3, there is much wider region of control authority over combustion than at 0 swirl (pure tumble configuration). This leads to the conclusion that the secondary throttle plays an important role to improve engine performance over variability in cylinder-to-cylinder or primary engine condition control parameters.

Gross-indicated efficiency map overlaid with regions of unstable combustion (gray) and knock (black) with fixed dilution (42% EGR) with piston SIM0336
Emission Maps.
Fuel-specific NOx in a gram of NOx residual in the combustion chamber per kilogram of injected fuel is calculated at EVO. Contour maps of fsNOx from tumble and swumble configurations (Fig. 19) are compared. For most of the engine operation at 2000 RPM, 3 bar BMEP, the fsNOx is below 5 g/kg f. This low NOx emission is due to low-temperature combustion enabled by highly dilute cooled EGR. It is observed from these results that flow configurations of the swumble system enables stable operation in a region that can offer NOx emission reductions.
Fuel-specific HC emission in a gram of unburned HC in combustion chamber per kilogram of injected fuel is calculated at EVO. Contour maps of fsHC from tumble and swumble configurations are compared in Fig. 20.
It is observed that operating at optimal EGR and CA50 phasing, similar fsHC emissions are generated between tumble and swumble systems. However, with the tumble system at high EGR levels, it is noted that fsHC increases quicker with ignition delay (larger CA50 or combustion phasing retard) than the quicker-burning swumble system. This is because the swumble system promotes flame propagation into areas more prone to quenching than seen with the tumble configuration.
Both fsNOx and fsHC (Fig. 21) emission responses are presented against swirl and CA50 at a fixed dilution level. It is seen from this figure that both NOx and HC have a high degree of sensitivity to combustion phasing (CA50). fsHC emissions do not respond strongly to swirl level, but fsNOx shows the impact of swirl on peak NOx is highest at approximately 3. This peak in NOx production is proposed to be due to competing interactions of combustion speed and heat transfer, both of which increase with swirl. Faster heat release tends to reduce HC and improve GIE if unburned volumes exist, whereas increasing heat transfer tends to reduce GIE from thermal losses. This peak NOx highlights the swirl at which in-cylinder temperatures are highest and hence heat release is fastest, akin to a slightly lean condition in a conventional engine. The reduction of NOx with increasing swirl is interesting as it is likely the increasing rate of heat transfer is expected to mitigate the increase in chemical heat release.
It is proposed the reason for the HC not reducing measurably with increasing swirl (e.g., relative to zero swirl level) is that the tumble system (i.e., no swirl) already has a thorough conversion of fuel at this condition. Hence, negligible additional unburned fuel can be converted with the increased heat release rate offered by swirl. This insensitivity of HC against swirl is not expected to maintain at an engine operating condition where the tumble system leaves some volume of unburned fuel.
Effect of Compression Ratio on Fuel Efficiency.
With the VCR mechanism, CR can be manipulated as a control parameter to study its effect on fuel efficiency, knock, and combustion stability. Figure 22 shows a GIE map against CR and CA50 combustion phasing for 2000 RPM, 3 bar BMEP engine operating condition. Using second-order statistical meta-model (quadratic response function), an optimization of GIE against CR is conducted. Figure 23 shows that for every 0.5 increase in compression ratio, ∼0.5% gross-indicated fuel efficiency gain is realized between 16:1 and 18:1 compression ratio. The scope of this study was limited to 18:1 compression ratio.

Maximum GIE against CR with fixed EGR and swirl number while maintaining emission, knock, and combustion stability margins
Engine Operating Map.
A system level 1D/0D simulation is carried out following the multidimensional simulation results to evaluate drive cycle-averaged emission and fuel economy against development target (Table 3). The targets for emissions are derived from Tier3, Bin 160 (similar to Euro 6) emission for vehicle class N1 (<1350 kg), “positive ignition” engine [12].
Drive cycle emission and efficiency. System level simulation versus target for the 1.2 L development engine.
CO (g/km) | HC (g/km) | NOx (g/km) | Fuel economy (km/L) | |
---|---|---|---|---|
Target | 0.5 | 0.03 | 0.036 | 24 |
Noncatalyzed engine out | 0.11 | 1.34 | 0.028 | 24.2 |
CO (g/km) | HC (g/km) | NOx (g/km) | Fuel economy (km/L) | |
---|---|---|---|---|
Target | 0.5 | 0.03 | 0.036 | 24 |
Noncatalyzed engine out | 0.11 | 1.34 | 0.028 | 24.2 |
Based on the numerical studies, it is found that 1.2 L engine will be able to meet Tier 3, Bin 160 (similar to Euro 6) emission standards with only HC oxidation catalyst. For vehicle class N1, this 1.2 L development engine is on target to meet CO and NOx emissions without catalyst. From these simulations, it is expected that the vehicle fuel economy target of 24.2 km/L will be met.
Conclusions
This paper presents high-performance computer-assisted numerical studies of 1.2 L four-stroke opposed piston sleeve valve development engines. Massive computational campaign was carried out to evaluate design and control concepts of the development engines using multidimensional engine simulations. From these computational combustion results, reduced-order meta-models were constructed to accelerate analysis-led design evaluations. Later system level simulations were carried out to evaluate drive cycle emission and efficiency and compared with Tier 3, Bin 160 (similar to Euro 6) emissions requirement as well as fuel economy targets for engine program.
Design concepts were evaluated using multidimensional engine modeling, and it is observed that:
Swumble system concept enhances dilution tolerance of low-temperature combustion operation.
Collapse of dual tumble in opposed piston architecture produces enhanced in-cylinder turbulence to support dilute gasoline combustion.
In-cylinder design features do play important role to manipulate flame dynamics and to improve combustion stability.
Using crank phasers, the compression ratio can be adjusted to result in higher fuel efficiency at cruising condition. To limit knocking, compression ratio is lowered at higher load engine operating conditions.
Currently, Pinnacle is building a prototype engine to conduct engine testing based on design principles and guidance obtained from these massive computation campaign. After obtaining engine test data, these models will be validated further to guide further refinements for production ready engine.
Acknowledgment
This research used resources of the Oak Ridge Leadership Computing Facility, which is a DOE Office of Science User Facility supported under Contract DE-AC05-00OR22725. This paper has been authored by UT-Battelle, LLC under Contract No. DE-AC05-00OR22725 with the U.S. Department of Energy. The United States Government retains and the publisher, by accepting the paper for publication, acknowledges that the United States Government retains a nonexclusive, paid-up, irrevocable, worldwide license to publish or reproduce the published form of this paper, or allow others to do so, for United States Government purposes. The Department of Energy will provide public access to these results of federally sponsored research in accordance with the DOE Public Access Plan2. Authors would also like to acknowledge Convergent Science for their technical support in conducting this work.
Funding Data
This research work was funded by the U.S. Department of Energy (DOE) under Small Business Voucher Program Cooperative Research and Development Agreement (CRADA) No. NFE-16-06356.
Nomenclature
- BMEP =
brake mean effective pressure
- CA =
crank angle time
- CA50 =
crank angle corresponding to 50% heat release
- CFD =
computational fluid dynamics
- EER =
effective expansion ratio
- EGR =
exhaust gas recirculation
- fsHC =
fuel-specific hydro carbon
- fsNOx =
fuel-specific nitrogen oxides
- GIE =
gross-indicated efficiency
- HC =
hydrocarbon
- HPC =
high-performance computing
- IMEP =
indicated mean effective pressure
- NOx =
nitrogen oxides
- OLCF =
Oak Ridge Leadership Computing Facility
- RPM =
revolutions per minute
- SI =
spark ignition
Appendix: Priliminary Validation of Simulation Model
The multidimensional simulation submodels were developed in lieu with predicting combustion and emission characteristics of opposed-piston four-stroke engines. Test data from a lean-burn version of Pinnacle Engine was used to validate and test the multidimensional model. A more detailed validation method is discussed in ASME ICEF2017-2618. The summary of the model validation results is given later.
Here is a summary of model validation. A test engine operating condition (4000 RPM, 3 bar IMEP) was chosen to validate model. A nominal compression ratio of 15:1 and in-cylinder equivalence ratio of 0.626 (lean-burn) is chosen for the validation test. A new chemical kinetics mechanism was developed to match laminar flame speed data from various literature source representing wide ranging conditions ranging both lean and dilute conditions. The CFD simulation with modified mechanism at the test engine operating condition showed good match (Fig. 24). More details of the validation methods, new mechanism development is discussed in the ASME ICEF2017-2618 paper.

Apparent heat release rate from simulation (solid line) and test (dotted black): (a) with original chemical kinetics mechanism and (b) from newly developed chemical kinetics mechanism and validation of the simulation model