TURBOCHARGING A LOW BMEP PUMP SCAVENGED ENGINE

Proceedings of Gas Machinery Research Council Gas Machinery Conference 2005 October 3-5, 2005 Covington Kentucky TURBOCHARGING A LOW BMEP PUMP SCAVENGED ENGINE David Stickler DigiCon Incorporated Randy Potter Enginuity Incorporated Damian Kuiper DigiCon Incorporated Jonathan Goss El Paso Corporation ABSTRACT Increasingly stringent emissions standards are continually driving the natural gas industry to seek cost-effective methods of reducing environmentally harmful pollutants. The two primary pollutants of concern are NO x and CO. NO x formation is exponentially dependent on temperature and time. Combustion temperatures which provide the environment where oxides of nitrogen form are directly linked to the cylinder temperature and, therefore, air/fuel ratio. Also related to air/fuel ratio is CO, with rich fuel-air mixtures there is insufficient oxygen to fully burn all of the carbon present within the fuel to CO 2 [1]. Engine operation under a leaner air/fuel ratio effectively reducing combustion temperatures can decrease NO x production. To avoid the typical CO increase with lean air/fuel ratio combustion, the poor air/fuel mixing inherent in the Worthington design had to be improved. To meet these desired operating conditions, a turbocharger in combination with enhanced mixing nozzles were installed directly to seven Worthington LTC 2-stroke, spark ignition, direct injection engines. Turbocharger installation in combination with modified port and piston geometry provided higher air and exhaust manifold pressures, thus increasing the trapped air/fuel ratio as well as improving cylinder scavenging by increasing the engine air through flow. The fuel injection valves were modified to improve the air/fuel mixing by injecting the fuel at speeds greater than Mach. However, these modifications were not undertaken without a few continuing technical challenges. Low BMEP internal combustion engines such as the Worthington LTC have less energy available within the exhaust stream due to their low operating efficiency. This reduces the available energy within the exhaust stream and therefore reduces the exhaust energy available to the turbocharger turbine and compressor. To account for and overcome this challenge, the design team implemented modifications to the piston, intake port and exhaust port geometry. During later stages of design parametric modeling and computational fluid dynamic (CFD) software were utilized. The intake, exhaust and mixing processes are visualized ensuring that desired flow characteristics are employed by the geometric modifications made to the cylinder, piston, and liner. INTRODUCTION The Worthington LTC-8 natural gas engine is by OEM design a two-stroke pump scavenged engine with direct injection and spark ignition. The primary driver for these modifications is emissions reductions of NO x from a baseline near 18.00 g / bhp h to a target below 10.00 g / bhp h NO x while maintaining CO below 1.97 g / bhp h. To meet the desired operating condition of reduced NO x and CO on all seven engines three primary objectives were targeted. First, a higher air/exhaust manifold pressure to increase the trapped mass air/fuel ratio. Second, a higher air mass flow rate across the cylinder to improve scavenging efficiency. Third, along with the increased air delivery, fuel injection modifications were implemented via supersonic fuel injection nozzles achieving a non-stratified, thoroughly mixed charge. In order to obtain a leaner air/fuel ratio and a higher scavenging efficiency the initially pump scavenged engine was converted to turbocharged aspiration. The conversion reduces parasitic losses of a reciprocating scavenge cylinder and provides the ability to dictate a greater charge density and air flow rate through the engine within the turbocharger operating range. Though a turbocharger can be built for a specific design point the turbocharger/engine interaction plays a significant role in whether the turbocharger compressor can operate at a desired design point. More specifically when high engine through flow is a desirable trait the only significant restriction within the airexhaust loop should be the turbine nozzle ring. This was not the case with the stock LTC-8 power cylinder / piston design. Figure 1: Surface Model of the Worthington LTC-8 Combustion Cylinder with Power Piston 1

NOMENCLATURE ν φ η sc Power Specific, Volume Equivalence Ratio Scavenging Efficiency A / A Ratio of Exhaust to Inlet Port Area e i EMISSIONS The two main pollutants targeted for reduction by this project are NO x and CO. The majority of NO x will be nitrogen oxide (NO), with a small amount of nitrogen dioxide (NO 2 ) formation within the exhaust system, downstream of the combustion chamber [2]. Common NO generating reactions that occur during the combustion process are [2]: O+ N NO+ N 2 (1) N+ O NO+ O 2 (2) N+ OH NO+ H (3) As expressed by equations (1) through (3), during combustion only NO is formed. Further reactions immediately following the combustion process at high temperatures are [2]: NO + H O NO + H 2 2 2 (4) NO + O NO + O 2 2 (5) Reactions (4) and (5) take place after the combustion process. Internal combustion engines within the natural gas industry consistently operate at equivalence ratios less than one (excess oxygen). This operating condition ensures sufficient amounts of oxygen (O 2 ) for combustion; however, this excess oxygen provides an environment for the production of NO x to be directly related to combustion temperature. A measure of the direct impact that equivalence ratio and therefore temperature has on NO x production is represented by Figure 2. Figure 2: Emissions as a Function of φ [3] As Figure 2 identifies, while running rich there is not enough oxygen to fully react with all of the fuel and higher levels of HC and CO are present. Substantial leaning of the air/fuel ratio, beyond stoichiometric, provides a reduction in all three major pollutants. ABOUT THE WORTHINGTON LTC - 8 The Worthington LTC engine is considered a low BMEP engine. A low BMEP leads to less energy available within the exhaust stream and therefore a reduced amount of work is delivered to the turbocharger turbine. This chain of events leads to a reduced flow rate of air available for scavenging the cylinder. To put this in perspective, characteristics of the Worthington LTC-8 are directly compared to a Worthington UTC-8T of similar displacement volume. Within Table 1 specific volume is defined as Equation 6. engine volume v = (6) rated power Worthington Engines LTC - 8 UTC 8T Rated Horsepower [hp] 1100 1850 Number of Power Cylinders 8 8 Bore [inches] 15 16 Stroke [inches] 15 15 Displacement [ft 3 / rev] 12.3 13.9 Specific Volume (ν ) [ft 3 /hp] 0.0112 0.0075 BMEP [psig] 68 77 Scavenging Method Cross Uniflow Table 1: Comparison of Similar Displacement Engines (The UTC-8T is a field turbocharged UTC-8) As noted by Heywood [1], specific volume indicates the effectiveness with which the engine designer has used the engine materials, where a lower value indicates more efficient use. The LTC-8 specific volume is 49% greater than that of the UTC-8T. Table 1 identifies the LTC volumetric displacement is only 11.5% less than that of the UTC-8T yet the UTC-8T provides 68% more power. As a result a very similar displacement volume must be scavenged with a much lower available brake horsepower and brake mean effective pressure. WORTHINGTON LTC - 8 SCAVENGING Scavenging is the process of clearing exhaust gas from the cylinder of a two-stroke engine. There are two primary concepts of this, perfect mixing and perfect displacement though neither of these actually occur. Perfect displacement assumes that all exhaust gases within the cylinder are displaced by the fresh air. Perfect mixing states that though some exhaust gas will be left in the cylinder after scavenging the mixture will be perfectly homogeneous. In actuality scavenging is done partially by 2

displacement of exhaust gas with fresh air and partially by entraining exhaust gas with fresh air as the mixture then discharges through the exhaust ports, it is never perfectly displaced or perfectly mixed. Cross, loop, and uniflow are the three main methods for cylinder scavenging. Figure 3 identifies the general direction of flow for each [1]. Figure 3: Three Primary Methods of Scavenging [1] Top View Side View Per OEM design the LTC-8 is a cross-scavenged engine which, as with all three methods, relies heavily on a pressure differential between the intake and exhaust passages to flow air through the cylinder. It can be stated that the pressure differential across the cylinder represents the potential for air flow through the cylinder volume during the scavenging period hence a larger differential is preferred. A very practical means of measuring how well the exhaust gases are replaced with fresh air is scavenging efficiency ( η ) [1]. sc sc mass of delivered air retained mass of trapped cylinder charge η = (7) Under OEM design conditions the LTC-8 power cylinders are supplied with air from a scavenging piston at 6 inch Hg while the exhaust is free flowing through a silencer generating 5 inches of H 2 O backpressure. This provides the power cylinder with a net 5.92 inches of Hg pressure differential. Due to the backpressure generated with a turbocharger turbine, when turbocharged, the LTC-8, with stock cylinder and piston configuration, receives at best a 1.5 inch Hg pressure differential across the power cylinder, this represents a 75% reduction. INTAKE AND EXHAUST PORT DESIGN Successful turbocharging of a two-stroke engine requires minimal air and exhaust flow restriction. For this reason high flow and low pressure loss were design requirements while specifying the air inlet filter, intercooler, air manifold and exhaust manifold. The OEM design of the intake and exhaust ports is identified by Figure 4. Figure 4: OEM Intake and Exhaust PortDesign By calculating the area of intake port exposed to the cylinder volume at every 1 degree of crank angle and summing the values for all eight cylinders the net exposure area is generated from 0 to 360 of revolution. This cross-sectional area represents the total area available for the air flow to pass through at each degree of revolution. The minimum, maximum and average were found to be 23.7 in 2, 32.3 in 2, and 28.1 in 2 respectively. This concept was also applied to the stock exhaust ports where the minimum, maximum and average were found to be 78.29 in 2, 87.36 in 2, 82.8 in 2. These calculations identified an approximate 3:1 ratio between average exhaust port and average inlet port exposure areas. At a 5.92 inch Hg cylinder pressure differential this ratio was more acceptable for scavenging the cylinder but at a 1.5 inch Hg differential this ratio identified the intake port area as a flow restriction. In conjunction with flow restriction it has been found that the ratio of exhaust port area to inlet port area (A e /A i ) plays a key role in the ability for some two-stroke engines to scavenge its cylinders. For a loop scavenged engine scavenging efficiency has been found to increase markedly as the ratio of exhaust to inlet port area (A e /A i ) was decreased from 1.2 to 1.0, and even to as low as 0.6 [3]. The largest increase in scavenging efficiency 3

attained at lower piston speeds (600 ft/min) occurred while shifting A e /A i from 1.25 to 0.75 [3]. Before the stock LTC intake ports could be modified the cylinder water jackets were located and dimensioned. It was found that little room was available between the intake ports but the ports could be raised 1 inch without risk of mechanical failure. This modification increased to total intake port area per cylinder from 31.28in 2 to 46.92in 2. This increase in area shifted the A e /A i ratio from 1.46 to 0.97. It was also found that at bottom dead center the piston covered ½ inch of the intake port height. In response to this a ½ inch, 45 chamfer was placed on the intake edge of the piston providing full exposure of the intake ports while positioned at bottom dead center. The piston chamfer provided a twofold benefit. The piston edge was no longer a flow obstruction and the time allotted for the intake event was lengthened increasing the available time for cylinder scavenging. By raising the intake ports 1 inch and applying a ½ inch chamfer to the piston intake edge the start of the intake event was shifted from 136 ATDC to 122 ATDC. While the increased time for scavenging is a benefit, initiating the intake event 14 earlier eliminated the blow-down period. A blow-down period is typically between 10 and 20 of duration providing time for a substantial portion of the higher pressure exhaust gas to discharge through the exhaust ports prior to intake port exposure. As a necessity the exhaust ports as well as the exhaust edge of the piston received a ½ inch, 45 chamfer. This shifted the start of exhaust and blow-down events 12 earlier from 120 to 108 ATDC. Though the majority of the OEM blow-down period has been restored both the intake and exhaust events are being initiated earlier in the piston-cylinder expansion process where higher in-cylinder pressures are prevalent. With the current port modifications providing a 14 blowdown period the engine was tested over an array of operating conditions identifying the combustion mechanism as fairly unstable. The condition is best described as engine operation at low load where sporadic misfires followed by detonation were common. Pressure versus crank angle traces obtained from field testing coincided with this thought process showing cylinder pressures over 35 psig during the start of the intake event. CFD INITIATIVE With less than desired performance and a minimal amount of cylinder pressure differential to work with for cylinder scavenging an in-depth Computational Fluid Dynamic (CFD) analysis was conducted. Though a complete CFD analysis was desired prior to initiation of any engine modifications the rigid timeline of this project demanded progress prior to completion of the computational analysis. The CFD effort within DigiCon was initiated in late January 2005. As a brief background, the entire cylinder, cylinder head, piston top contour, intake, and exhaust ports were parametrically surface modeled as displayed within Figure 1. These surfaces were then filled with over 1.3 million volume cells ensuring sufficient cell density to capture highly transient flow structures. This process was initially completed and validated by cylinder pressure vs. crank angle plots for OEM geometry then applied to the modifications on the LTC as well as proposed future modifications which have been computationally found to significantly improve flow through the cylinder. The blow-down and scavenging periods were the main focus of this CFD analysis in search of a primary reason for poor air flow across the cylinder. The principle results assessed are cylinder pressure at the intake opening event and time duration of flow from cylinder into the intake ports (pulsation). CFD ANALYSIS OF OEM CONDITIONS The first simulation conducted was a replication of OEM conditions. This means no modifications to the piston, intake port, or exhaust port geometry as well as boundary conditions representing the original pump scavenged engine. The event timing for intake and exhaust are those of the original timing diagram. The computational model was initially positioned at 100 crank angle (CA) from top dead center (TDC), where both intake and exhaust ports are closed and run to 180 CA (BDC). The intake, exhaust, and cylinder were all provided respective pressures as found during field testing. The cylinder is only provided an initial condition pressure at 100 CA all other results of the simulation are functions of the model geometry and boundary conditions provided to the intake and exhaust porting. Exhaust Ports Start to Open 120 ATDC Intake Ports Start to Open 136 ATDC Cylinder Pressure at 100 ATDC 55.92 psig Boundary Pressure for Exhaust Volume 5.00 inch H 2 0 Boundary Pressure for Intake Volume 6.00 inch Hg To ensure the accuracy of the model the average pressure of the cylinder volume was recorded at every 2 degrees of crank angle and compared to an actual pressure trace obtained from field testing. The results show minimal overall deviation from the experimental trace as identified by Figure 5 (next page). The maximum difference between experimental and computational results was 3.91 psig the minimum difference was 0.08 psig the average difference was 1.75 psig. With the overall pressure trend being virtually duplicated this validation instills confidence within the methodology utilized to generate the model. CFD OF MODIFIED CYLINDER AND PISTON The next step was to parametrically model the modified piston and cylinder geometry currently in the field. This includes modifications to the intake ports, exhaust ports, and the piston top on the intake and exhaust side. As with the OEM model the modified design surface model was then meshed with CFD software and a simulation was run from 100 CA to 180 CA allowing an assessment of the blow-down period and clear detection of the pulsation magnitude. The initial and boundary conditions provided were all identified from field testing of the actual modified engine. 4

Exhaust Ports Start to Open 108 ATDC Intake Ports Start to Open 122 ATDC Cylinder Pressure at 100 ATDC 67.66 psig Boundary Pressure for Exhaust Volume 6.28 psig Boundary Pressure for Intake Volume 7.02 psig Again this simulation has the ability to be validated against an experimental pressure trace obtained in the field. As with the OEM model, cylinder pressure was recorded every 2 degrees of crank angle for comparison. The maximum difference between experimental and computational results was 5.56 psig the minimum difference was 0.29 psig the average difference was 2.66psig. The pressure trends during the blow-down period, as identified by Figure 6, are again very similar though some variation in the pressure fluctuation is found near bottom dead center. The rate of pressure drop matches experimental results, further building the confidence level of the model s ability to accurately predict the blow-down mechanism. Pressure Vs Crank Angle 100 Cylinder Pressure [psig] 90 80 70 60 50 40 30 20 OEM - UNIT 6A - Cyl 3 ProStar Sim Press - OEM Cond. BLOW-DOWN PERIOD 10 0 90 100 110 120 130 140 150 160 170 180 190 Crank Angle [degrees] Figure 5: Pressure Trace Comparison of Experimental Vs. Computational for OEM Conditions 100 Pressure Vs Crank Angle Cylinder Pressure [psig] 90 80 70 60 50 40 30 20 MODIFIED - UNIT 9A - Cyl 8 - as of 9/1/05 ProStar Sim Press - Current Design Condition - as of 9/1/05 BLOW-DOWN PERIOD 10 0 80 90 100 110 120 130 140 150 160 170 180 190 200 Crank Angle [degrees] Figure 6: Pressure Trace Comparison of Experimental Vs. Computational for Modified Conditions 5

The pressure trends during the blow-down period, as identified by Figure 6, are again very similar though some variation in the pressure fluctuation is found near bottom dead center. With the rate of pressure drop matching experimental results this further builds the confidence level of the models ability to accurately predict the blow-down mechanism. COMPARISON OF MODIFIED AND OEM DESIGN The results show significantly different blow-down periods as indicated by Table 2. Under OEM and modified conditions there is initially flow into the intake ports from the cylinder (pulsation). Though the time period during which this occurs for OEM design is only 7.94% of the entire intake event (48.88 milliseconds), for the modified piston and cylinder this pulsation lasts much longer. The magnitude of flow disturbance can be viewed within Figures A, B, and C of the OEM Condition Appendix. Similar plots for the Modified Design, Figures D, E, and F within the appendix, show significantly elevated levels of flow disturbance over the OEM design. It was found that the modified design condition introduced a substantial pressure wave into the intake port volume severely reducing the effectiveness of the enlarged port area. ACTING ON CFD RESULTS In recognition of the cylinder pressure and pulsation indicating an insufficient blow down period, a series of iterative design simulations were conducted using the existing port modifications presently in the field while applying geometric changes to the piston exhaust side edge. This was done in effort to increase the blow-down period effectively reducing cylinder pressure at the point the intake event is initiated. The computational model shows a more reasonable level of flow disturbance will occur as the intake ports are uncovered. This increases the time available for positive flow into the cylinder and significantly reduces the initial presence of exhaust gas within the intake port volume. Figure 7 presents the OEM, Modified Design, and Proposed Design piston geometry on the intake and exhaust edges. Within Figure 7 the OEM design identifies the exhaust edge of the piston to be a rounded corner, Initial Modifications has a 0.5 inch (45 ) chamfer, and the Proposed Design profile was extended down along the pistons height to initiate the blowdown period 9 earlier. The proposed design has been simulated using the currently modified intake and exhaust ports with promising results. Results Units OEM Condition Modified Design Cylinder Pressure at 100 Crank Angle [psig] 55.92 67.66 Air Manifold Pressure [psig] 2.94 7.02 Exhaust Manifold Pressure [psig] 0.18 6.28 Crank Angle at Exhaust Open Event [degrees] 120 108 Crank Angle at Intake Open Event [degrees] 136 122 Avg. Cylinder Pressure at Intake Open Event [psig] 12.24 37.31 Pulse Differential at Intake Open Event (Cyl. Press. - Intake Press.) [psig] 9.3 30.29 Crank Angle When Fresh Air Begins Flowing into Cylinder (End of Pulsation Period) [deg] 143 137 Time from Intake Open Event to Intake Close Event [millisec] 48.88 64.44 Time from Intake Open Event to End of Pulsation Period [millisec] 3.88 8.33 Percentage of Time the Pulsation takes of the Entire Intake Event [%] 7.94% 12.93% Total Time of Positive Flow Into Cylinder [Intake Event - Pulse Time] [millisec] 45 56.11 Table 2: Computational Pressure Trace Comparison of OEM Vs. Modified Design Exhaust Edge - OEM Design Exhaust Edge - Initial Modifications Exhaust Edge - Proposed Piston Design Figure 7: Comparison of OEM, Initial Modification, and Proposed Piston Design 6

To simulate the proposed piston design the boundary conditions of the cylinder and intake/exhaust ports were collected from field testing of the actual modified engine. The exhaust event is initiated 9 earlier than Modified Design conditions. Exhaust Ports Start to Open 99 ATDC Intake Ports Start to Open 122 ATDC Cylinder Pressure at 100 ATDC 67.66 psig Boundary Pressure for Exhaust Volume 6.28 psig Boundary Pressure for Intake Volume 7.02 psig Figure 8 identifies the computational pressure trace of the proposed piston design which is overlaid with the experimental pressure trace of the modified design. The results of the proposed exhaust profile were positive. The cylinder pressure at the intake open event was reduced 62% from the current 37.31psig to 13.89psig. The time during which pulsation occurs was reduced 60% from 8.33 milliseconds to 3.33 milliseconds. The total time available for positive flow into the cylinder was increased 8.9% from 56.11 milliseconds to 61.11 milliseconds. Figures G, H, and I of the proposed condition appendix identify the magnitude of disturbance as compared to Figures D, E, and F of the modified condition appendix. The proposed design results are compared to the modified and OEM condition with Table 3. Pressure Vs Crank Angle 100 90 MODIFIED - UNIT 9A - Cyl 8 - as of 9/1/05 Cylinder Pressure [psig] 80 70 60 50 40 30 20 10 0 ProStar Sim Press - Proposed Design Condition BLOW-DOWN PERIOD FOR MODIFIED DESIGN BLOW-DOWN PERIOD FOR PROPOSED DESIGN 80 90 100 110 120 130 140 150 160 170 180 190 200 Crank Angle [degrees] Figure 8: Pressure Trace Comparison of 14 and 23 Blow-Down Periods Results Units OEM Condition Modified Design Proposed Design Cylinder Pressure at 100 Crank Angle [95 CA for Proposed] [psig] 55.92 67.66 72.914 Air Manifold Pressure [psig] 2.94 7.02 7.02 Exhaust Manifold Pressure [psig] 0.18 6.28 6.28 Crank Angle at Exhaust Open Event [degrees] 120 108 99 Crank Angle at Intake Open Event [degrees] 136 122 122 Avg. Cylinder Pressure at Intake Open Event [psig] 12.24 37.31 13.89 Pulse Differential at Intake Open Event (Cyl. Press. - Intake Press.) [psig] 9.3 30.29 6.87 Crank Angle When Fresh Air Begins Flowing into Cylinder (End of Pulsation Period) [deg] 143 137 128 Time from Intake Open Event to Intake Close Event [millisec] 48.88 64.44 64.44 Time from Intake Open Event to End of Pulsation Period [millisec] 3.88 8.33 3.33 Percentage of Time the Pulsation takes of the Entire Intake Event [%] 7.94% 12.93% 5.17% Total Time of Positive Flow Into Cylinder [Intake Evt - Puls Time] [millisec] 45 56.11 61.11 Table 3: Computational Pressure Trace Comparison of OEM, Modified Design, and Proposed Piston Design 7

IMPLEMENTATION OF PROPOSED PISTON DESIGN Based on the results presented within Table 3 the proposed piston design change was applied to all eight pistons of one engine. Though little testing time was available prior to this publication a cylinder pressure trace was obtained to compare with the computational values. Figure 9 shows this comparison which again provides confidence within the computational models ability to predict the blow-down period. coupled with specific investigation into the gas dynamics and turbocharger/engine interaction. Many of the same modifications would have been made though a few would have been done simultaneously. Pressure Vs Crank Angle 100 Cylinder Pressure [psig] 90 80 70 60 50 40 30 20 ProStar Sim Press - Mod006 Proposed Design - UNIT 9A - as of 9/20/05 BLOW-DOWN PERIOD FOR PROPOSED DESIGN 10 0 90 100 110 120 130 140 150 160 Crank Angle [degrees] Figure 9: Pressure Trace Comparison of Experimental Vs. Computational for Proposed Piston Design LESSONS LEARNED CONCLUDING STATEMENTS In summation, the high cylinder pressure (37.31psig) present while the intake ports are uncovered generates a pressure wave along with high magnitude reverse flow into the intake port volume. This flow structure delays the start of scavenging and reduces the overall ability of the turbocharger to flow air through the cylinder in the allotted time for the intake event. By modifying the exhaust side of the piston the blow-down period is initiated 9 earlier than current design conditions. This modification provided an overall blow-down period of 23 CA reducing cylinder pressure 62% from 37.3 psig to 13.9 psig at the intake open event. Computational results show the Proposed Design condition reduces the pulse differential 77% from 30.29 psig to 6.87 psig. It should be noted that 6.87psig is 24% lower than that found under OEM conditions (9.3 psig-oem). The change in area under the pressure vs. crank angle diagram of Figure 8 represents a minimal amount of lost horsepower which is offset by the non-use of the parasitic scavenging piston operating under OEM design conditions. The research and development phase conducted prior to the start of a new project such as this should never be dismissed as unnecessary or conducted as a rough approximation. Throughout this project many decisions could have been better guided through a more in-depth research and development phase The current road map for this project includes further computational research of the scavenging process and associated modifications to improve scavenging efficiency. More efficient use of the delivered scavenging air has potential to further reduce in-cylinder temperatures directly linked to emissions production. ACKNOWLEDGMENTS The authors extend their gratitude and thanks to the El Paso Pipeline Group employees of Station 32, the entire El Paso Clean Air Team, Thomas Burgett, and Gregg Beshouri. REFERENCES [1] Heywood J. B., Internal Combustion Engine Fundamentals, McGraw-Hill, Inc., New York, Apr 1988 [2] Pulkrabek, W.W., Engineering Fundamentals of the Internal Combustion Engine, Prentice-Hall Inc., Upper Saddle River, New Jersey, 1997 [3] Taylor, Charles F., The Internal Combustion Engine in Theory and Practice Vol. 2, The MIT Press, Cambridge, Massachusetts, 1985 8

APPENDIX OEM CONDITION 9

Figure A (OEM Condition): Exhaust Flow From Cylinder to Exhaust Port Figure B (OEM Condition): Exhaust Flow from Cylinder to Intake Port (Pulsation) 10

Figure C (OEM Condition): End of Pulsation Flow into Cylinder from Intake Port 11

APPENDIX MODIFIED DESIGN 12

FIGURE D (Modified Design): Exhaust Flow From Cylinder to Exhaust Port Figure E (Modified Design): Exhaust Flow from Cylinder to Intake Port (Pulsation) 13

Figure F (Modified Design): End of Pulsation Flow into Cylinder from Intake Port 14

APPENDIX PROPOSED PISTON DESIGN 15

FIGURE G (Proposed Design): Exhaust Flow From Cylinder to Exhaust Port Figure H (Proposed Design): Exhaust Flow from Cylinder to Intake Port (Pulsation) 16

Figure I (Proposed Design): End of Pulsation Flow into Cylinder from Intake Port 17