Opposed Piston Design
Opposed Piston Design
Opposed Piston Design
David Johnson, Michael Wahl, Fabien Redon, Eric Dion, Shauna McIntyre, Gerhard Regner
and Randy Herold
Achates Power, Inc., 4060 Sorrento Valley Boulevard, San Diego, CA 92121, USA
even for single-cylinder configurations, and thereby reduce accurately model the fuel spray, in-cylinder gas motion,
the loading of the crankshaft bearings. and combustion using computational fluid dynamic (CFD)
software packages (e.g. [4,5]) has allowed the engineering of
The two-stroke cycle and its double firing frequency gives
the combustion chamber geometry and nozzle configuration
engine designers the choice of decreasing brake mean
to achieve clean and efficient combustion. The ability to
effective pressure (BMEP) levels and increasing power density
shape two combustion chamber surfaces (the two pistons
compared to four-stroke engines of equivalent power output.
crowns) and incorporate multiple fuel injection locations on
The lower BMEP levels can be accomplished with lower
the liner has provided a larger design space than is available
peak cylinder pressures and therefore lower peak cylinder
in common four-stroke engines.
temperatures, both of which lead to design advantages. The
lower cylinder pressures result in lower mechanical stress The higher thermal loading of mechanical components
on engine components and therefore can be designed to be compared to four-stroke engines that results from the lack
of lighter weight. The lower cylinder temperatures result of the thermally relieving exhaust and intake cycles is also
in decreased NOx formation during combustion, lowering addressed with modern tools and materials. As discussed
the requirements for exhaust gas recirculation (EGR) and/or previously, the combustion is designed to occur away
NOx aftertreatment devices. The increased power density from the piston and liner surfaces, which reduces cooling
leads directly to smaller engine package size and weight, requirements of these components. Additionally, conjugate
both of which are beneficial to increasing overall vehicle fuel heat transfer (CHT) simulations are used to analyze the
economy and to decreasing manufacturing costs. cooling circuits of the engine and engineer an effective
cooling system that protects components and prevents oil
Modern Solutions to Opposed-Piston Two- degradation.
Stroke Engine Challenges Cylinder scavenging, which is a primary difference between
two- and four-stroke engines, is a technical challenge for
The challenge posed by the emissions regulations in
all two-stroke engines. In order to achieve a once-per-
the latter half of the twentieth century was difficult to
revolution firing frequency, the two-stroke engine must
overcome for two-stroke engines of any architecture and led
accomplish cylinder scavenging in roughly one-third of an
engine manufacturers to generally favor four-stroke engine
engine revolution, as opposed to a full revolution in four-
development. As demonstrated by the results reported in this
stroke engines, and must do so without the aid of a direct
paper, the emissions challenge - when revisited with modern
mechanical displacement pump (i.e. the piston). Instead, the
analytical tools, materials, and engineering methods - is no
two-stroke engine requires an external pressure differential
longer limiting the successful design of a clean and efficient
between the intake and exhaust ports to induce flow through
OP two-stroke engine.
the cylinder that allows the fresh charge to replace the
Since the OP arrangement has no cylinder head, the fuel exhaust products. Compared to other two-stroke architectures
injector must be installed in the cylinder liner. Historically, that use loop-scavenging, the opposed-piston engine with
this has created a technical challenge compared to the intake and exhaust ports being located at opposite ends of
common crank-slider arrangement with the fuel injector the cylinder employs the more efficient uniflow scavenging
located in the center of the cylinder head. The large [6]. The development goal for scavenging is to minimize the
distances in the fuel-spray direction (i.e. across the diameter external pumping required to purge the exhaust residual from
of the cylinder) combined with low fuel injection pressures the cylinder while creating the charge motion for subsequent
made accessing all of the available air in the combustion fuel/air mixing during injection. Complete characterization of
chamber difficult, resulting in inefficient combustion with the scavenging process is possible with the use of efficient
relatively high NOx and soot formation. Additionally, and accurate CFD software packages available today (e.g. [5],
the interaction between the fuel spray and the in-cylinder [7]). Optimization of port, cylinder, and piston geometries
fresh-charge motion with traditionally high swirl resulted in and their effects on the developed in-cylinder flow field can
combustion occurring near the combustion-chamber surfaces be accomplished in software without time-consuming and
causing increased thermal loading of the piston, piston rings, costly hardware fabrication and testing.
and cylinder liner, and leading to reduced thermal efficiency
With regards to emissions, one of the opportunities afforded
and increased cooling requirements.
by the two-stroke scavenging process is the ability to retain
Thanks to modern development tools and advanced fuel some portion of the charge in the cylinder after combustion
systems, the OP arrangement with a liner-mounted injector (“internal” EGR) as a means to control NOx by simply
has turned from a technical challenge into a unique reducing the pumping work applied by the aircharge system.
opportunity. The availability of fuel systems with high This helps improve fuel efficiency at part load. For high
injection pressures and the greater ease of manufacturing rates of EGR, the use of cooled external EGR is still
asymmetric injector nozzle hole directions have enabled required.
the fuel spray of liner-mounted injectors to better utilize
High oil consumption is a traditional challenge for two-stroke
the air within the combustion chamber with little-to-no
engines and is problematic for two reasons. First, free oil
wall impingement. Additionally, the ability to quickly and
3
Symposium on International Automotive Technology 2011
is a significant source of particulate emissions, and second, model is based on engine friction measurements, and the
additives in the oil create high ash residue that tends to heat release rate is measured by a single-cylinder research
contaminate after-treatment devices. The engineering goal engine. In turn, the 1 D engine model supplies intake and
is to develop a piston-liner and ring-liner interface that uses exhaust boundary conditions for the single-cylinder research
as little oil as possible without compromising durability. engine and performance predictions for the multi-cylinder
Fortunately, many of the technologies developed to reduce target engine. Furthermore, emission results from the single-
oil consumption in four-stroke engines apply directly here: cylinder research engine and measured fuel spray data are
improvements in cylinder bore materials, cylinder bore used for correlation with the 3 D combustion simulations that
finishing, piston ring technologies, crankcase breathing are aimed at optimizing the combustion chamber geometry.
systems, management of cylinder bore oil impingement,
synthetic oils, and low ash and phosphorus oils are all
technologies that have been applied to four-stroke engines
and are equally applicable to two-stroke engines. For the
work reported here, an advanced oil consumption analyzer
has been used [8], making research and development of these
oil-control technologies quicker and much more focused.
The lubrication of wrist pin bearings is another well-
known two-stroke challenge as the oil replenishing of the
continuously loaded wrist pin bearing is difficult. Several
techniques have evolved utilizing bearings with substantially
greater surface area and ladder grooves as well as special
coatings that aid boundary lubrication.
mixed, the intake gas flows through a second heat exchanger In-cylinder pressure was measured at 0.5 crank-angle
followed by a heater to precisely control the intake manifold resolution with a Kistler 6052C piezoelectric pressure
temperature. The exhaust manifold pressure is set with a back transducer coupled with a Kistler 5064 charge amplifier. The
pressure valve in the exhaust system. cylinder pressure signal is pegged to the intake and exhaust
manifold pressures, measured with Kistler 4005B and 4049A
high-speed pressure transducers, respectively. Custom in-
house software is used to acquire and process the data. A
California Analytical Instruments (CAI) emissions analyzer is
used to measure the steady-state concentration of five exhaust
species (CO2, CO, O2, HC, NOx) and intake CO2. A Dekati
DMM-230A Mass Monitor provided real-time particulate
matter values, and an AVL 415s Smoke Meter provides a
measure of exhaust soot content.
Interface Model
Friction and pumping energy losses, which represent the
difference between indicated work and brake work, are
specific for each engine configuration and do not translate
from a single-cylinder to a multi-cylinder engine by
simple multiplication. In order to predict the brake-specific
performance of a multi-cylinder engine based on single-
cylinder combustion results, an “interface model” has
been created in Ricardo WAVE. This model is exercised
(as described below) to match the experimental boundary
conditions and measured in-cylinder pressure trace so as
to provide multi-cylinder-based predictions of the friction
and pumping work required at the operating point measured
on the dynamometer. The results from the interface model
therefore provide predictions of multi-cylinder brake specific
performance and emissions parameters based on measured
Figure 2. Single Cylinder Research Engine Installed in single-cylinder results.
Test Cell
Figure 4 shows the schematic of the input data and
assumptions of the interface model. The combustion chamber
geometry, the piston motions, and the porting profiles
are identical to what exists in the single-cylinder engine,
while the number of cylinders and associated manifold
configurations are application specific. Engine speed,
fuel flow rate, air flow rate, EGR percentage, and intake
pressures and temperatures match the measured values. The
rate of heat release is derived from the measured cylinder
pressure and is input directly into the combustion sub-
model. Assumptions for the air-handling equipment, charge
cooling components, and aftertreatment system are used in
the pumping loss prediction. The Chen-Flynn mechanical
friction model is based on the mechanism design and analysis
and is correlated to experimental friction results. The work
needed to drive all accessories, including the supercharger,
is also taken into account.
The interface model air-handling system (Fig. 5) consists
mainly of a supercharger, a turbocharger, and a charge
air cooler after each compression stage. The size and
characteristics of the air-handling system components are
application specific. The supercharger model uses a full map
obtained from a supplier and a dual-drive mechanism was
Figure 3. Schematic of the Air and EGR Conditioning assumed to provide two drive ratios for the supercharger,
System which is useful for maintaining high thermal efficiency
5
Symposium on International Automotive Technology 2011
RESULTS
Medium-Duty Engine Performance
The process of measuring single-cylinder combustion results
and then using the interface model to predict multi-cylinder
engine performance has been applied to an engine size
typically found in a medium-duty commercial vehicle. The
specifications of this medium-duty engine are provided in
Table 1. The engine is operated at the corner points (C100,
C25, A100, A25) and center point (B50) of the steady-state
supplemental certification cycle adopted by the US and
Europe [10]. Only 5 of the 13 engine modes are considered
to reduce total testing time, although the weighting factors
Figure 5. Air Handling System Configuration
6
Symposium on International Automotive Technology 2011
specified by the legislation are used to calculate the cycle- performance values based on the multi-cylinder interface
average fuel consumption and emissions values. model that includes predictions of friction and pumping
losses. As shown, the brake thermal efficiencies at C100 and
Table 1. Medium-Duty Engine Specifications B50 exceed 42%, which is equivalent to a brake-specific fuel
consumption of less than 200 g/kWh. The cycle-averaged fuel
consumption of 203.2 g/kWh is achieved at a cycle-averaged
NOx level lower than the EURO IV NOx emissions standard
of 3.5 g/kWh. The cycle-averaged brake-specific PM, CO,
and HC levels are higher than the EURO IV standards, but
state-of-the-art DOC/DPF aftertreatment devices provide
sufficient reduction of these species to bring them within
acceptable limits. It should be noted that the interface model
assumed a turbine outlet pressure based on a complete DOC,
DPF and SCR aftertreatment model. Eliminating the SCR
would reduce the engine back pressure and would further
decrease the fuel consumption.
Operating results for the proposed medium-duty engine This engine was also operated at an engine-out NOx level
are provided in Table 2. Results are given for each of the capable of achieving the more stringent US 2010 emissions
five operating conditions considered as well as the cycle- standards assuming use of an SCR aftertreatment device.
averaged fuel consumption and emissions values. Also The brake-specific fuel consumption and NOx emissions
provided in Table 3 are EURO IV emissions standards results are provided in Table 3. For comparison, the fuel
for the European Steady-State Cycle (ESC) and European consumption for this engine calibration is 13% lower than a
Transient Cycle (ETC), which are the relevant emissions state-of-the-art 2010 medium-duty diesel engine at the same
levels for commercial vehicles in India [11]. The engine engine-out NOx levels [12].
operation has been optimized to achieve the EURO IV NOx
standards in-cylinder, thereby eliminating the need for NOx Oil Consumption
aftertreatment devices.
As previously discussed, opposed-piston two-stroke engines
The results shown in Table 2 provide indicated performance breathe through ports on both ends of the cylinder liner.
values from the single-cylinder engine and brake-specific This implies that the compression rings must traverse the
7
Symposium on International Automotive Technology 2011
CONCLUSIONS
ENGINE COST ANALYSIS The results reported in this work have shown that the
As discussed previously and consistent with observations opposed-piston two-stroke engine architecture is a suitable
by Flint and Pirault [1], the opposed-piston architecture platform for a highly efficient and clean internal combustion
benefits from a low part count that results in a low engine engine featuring low oil consumption. EURO IV NOx
cost. This is particularly advantageous in high volume emissions are achievable in-cylinder obviating the need for
applications in which a $500 savings per engine can yield NOx aftertreatment in certain markets. A SCR after-treatment
significant savings to the manufacturer. To estimate this cost can be fitted with further fuel efficiency benefits and even
advantage, a study by a cost analysis expert was performed lower NOx levels at the tailpipe. The architectural benefit
that compared the component cost of the Achates Power of a reduced parts count (no cylinder head) reduces cost,
A40 engine to a conventional diesel engine meeting similar weight and complexity compared to conventional four-stroke
emissions standards and equivalent manufacturing volumes. engine designs.
8
Symposium on International Automotive Technology 2011
3. Junkers, H., “Engine”, U.S. Patent 2 031 318, Feb. 18, CONTACT
1936. David Johnson, [email protected]
4. Amsden, A.A., “KIVA-3V: A Block-Structured KIVA DEFINITIONS/ABBREVIATIONS
Program for Engines with Vertical or Canted Valves”,
Los Alamos National Laboratory report LA-13313-MS, OP Engine Opposed Piston Engine
July 1997. IC Engine Internal Combustion Engine
5. Richards, K. J., Senecal, P. K., and Pomraning, E., DOC Diesel Oxidation Catalyst
CONVERGE™ (Version 1.3), Convergent Science, Inc., DPF Diesel Particulate Filter
Middleton, WI, 2008.
SCR Selective Catalytic Reduction
6. Schweitzer, P.H., Scavenging of Two-Stroke Cycle Diesel
NOx Nitric Oxides
Engines, The Macmillan Co., New York, NY, 1949.
IMEP Indicated Mean Effective Pressure
7. ANSYS CFX, Retrieved from http://www.ansys.com/
BMEP Brake Mean Effective Pressure
products/fluid-dynamics/cfx/, 2010.
ISFC Indicated Specific Fuel Consumption
8. DALOC, Retrieved from http://www.davinci-limited.com/
DALOC.asp, 2010. BSFC Brake Specific Fuel Consumption
BSPM Brake Specific Particulate Matter
9. Teng, H. and Regner, G., “Characteristics of Soot
Deposits in EGR Coolers”, SAE International Journal of BSHC Brake Specific Hydrocarbons
Fuels and Lubricants, Vol. 2, No. 2, pp. 81-90, 2010. BSCO Brake Specific Carbon Monoxide
Also published as SAE Technical Paper 2009-01-2671,
2009. ESC European Steady-State Cycle
ETC European Transient Cycle
10. Heavy-Duty Supplemental Emissions Test (SET),
Retrieved from http://www.dieselnet.com/standards/
cycles/set.php, 2010.