SAE TECHNICAL PAPER SERIES 2007-01-0382

Effect of Primary Intake Runner Tapers and Bellmouths on the Performance of a Single Engine

V. Mariucci and A. Selamet The Ohio State University K. D. Miazgowicz Ford Motor Company

Reprinted From: Modeling of SI and Diesel Engines, 2007 (SP-2079)

2007 World Congress Detroit, Michigan April 16-19, 2007

400 Commonwealth Drive, Warrendale, PA 15096-0001 U.S.A. Tel: (724) 776-4841 Fax: (724) 776-0790 Web: www.sae.org Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 By mandate of the Engineering Meetings Board, this paper has been approved for SAE publication upon completion of a peer review process by a minimum of three (3) industry experts under the supervision of the session organizer.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE.

For permission and licensing requests contact:

SAE Permissions 400 Commonwealth Drive Warrendale, PA 15096-0001-USA Email: [email protected] Fax: 724-776-3036 Tel: 724-772-4028

For multiple print copies contact:

SAE Customer Service Tel: 877-606-7323 (inside USA and Canada) Tel: 724-776-4970 (outside USA) Fax: 724-776-0790 Email: [email protected]

ISSN 0148-7191 Copyright © 2007 SAE International Positions and opinions advanced in this paper are those of the author(s) and not necessarily those of SAE. The author is solely responsible for the content of the paper. A process is available by which discussions will be printed with the paper if it is published in SAE Transactions.

Persons wishing to submit papers to be considered for presentation or publication by SAE should send the manuscript or a 300 word abstract of a proposed manuscript to: Secretary, Engineering Meetings Board, SAE.

Printed in USA

Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 2007-01-0382 Effect of Primary Intake Runner Tapers and Bellmouths on the Performance of a Single Cylinder Engine

V. Mariucci and A. Selamet The Ohio State University

K. D. Miazgowicz Ford Motor Company

Copyright © 2007 SAE International

ABSTRACT the draws in fresh charge, an expansion wave propagates toward the inlet of the intake pipe. When The present experimental study investigates this expansion wave reaches the inlet of the pipe, it is systematically the effects of primary intake runner partially reflected as a compression wave. If this configurations on a firing single cylinder research compression wave arrives back at the intake valve while engine. Twelve different intake configurations were it is still open, it can aid the breathing process of the fabricated to investigate runners with tapers and engine, thus increasing the volumetric efficiency (Șv). bellmouths. For each configuration, the length from the This charging effect will happen near certain engine base of the test configuration to the start of the inlet speeds, governed by the intake geometry, giving rise to radius and pressure measurement locations were “tuning peaks.” retained in an effort to isolate the effect of runner geometry only. Each configuration is presented against Numerous early works have investigated the a baseline case with constant cross-sectional area and a effect of varying primary intake runner geometry on ratio of inlet radius, Ri, to internal diameter, D, of 0.35. intake tuning. In 1924, Matthews and Gardiner, working For the seven tapered runners, the length of the taper with a one-cylinder diesel engine, found that the engine varied from 25% to 100% of the overall length of the test would not start unless the compression pressure was piece, and the taper area ratios (TAR) varied from 1.5 - increased. They found that a long suction pipe attached 3; all tapers retained the inlet radius of the baseline. to the inlet port gave them the pressure increase they The four bellmouth runners had a constant cross- needed, and proceeded to test several lengths of pipe sectional area, and varying inlet radii from Ri/D = 0.05 to up to 177.8 cm in addition to the case where no pipe 1.0. The time-averaged quantities such as volumetric was present from 500 to 1800 RPM; they noticed an efficiency and brake power, and time-resolved intake increase in compression pressure of 17% with the 62 in. pressures are presented for each configuration. For the pipe compared to no inlet pipe. The study was purely bellmouth runners, Ri/D > 0.20 was found to be most experimental and did not attempt to explain the reason beneficial to volumetric efficiency, and a TAR larger than for the increase in pressure. 1.5 was detrimental to the intake tuning for higher-speed tuning peaks. Morse et al. (1938) developed a simple method of intake tuning prediction, assuming linear, acoustic INTRODUCTION behavior in the intake duct. This method suggests that the quasi-standing wave (QSW) developed in the intake Whether the motivating factor is more power or pipe when the intake valves are closed dictates intake better fuel economy, several techniques have been tuning. It was found that when the ratio q of pipe employed throughout the years to improve the IC frequency ( = c/4Ɛ) to engine frequency ( = RPM/120) engine. One of the most well-known and used methods was equal to 3, 4, or 5, beneficial resonance would of enhancing an engine’s performance is by intake occur since the pressure fluctuations at the valves will tuning. Due to the unsteady nature of the gas exchange be large, leading to process in the IC engine, pressure fluctuations are generated when the engine takes in air. During the 30 c intake stroke of a single cylinder engine with intake pipe N , (1) leading from atmosphere to the cylinder, for example, as " q

1 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 where N [RPM] is the engine speed of beneficial 2/1 955 § A · resonance, c [m/s] is the speed of sound, Ɛ [m] is the RPMN ][ c¨ ¸ , (3) length of the intake pipe, and q = 3, 4, or 5. They t ¨ ¸ K " Veff verified their method by testing several lengths of intake © ¹ pipe of 6.5 cm diameter at a constant engine speed of 1220 RPM. where c [m/s] is the speed of sound, A [cm²] is the effective cross-sectional area of the intake, Ɛ [cm] is the Kastner (1945) studied the intake tuning effect effective length of the inlet system, K is a factor usually of various lengths and diameters of primary runners on equal to 2.1, but that varies from 2.0 to 2.5 depending two separate single-cylinder engines. Standard air upon valve timing and other factors, and, consumption, intake port and in-cylinder pressures were analyzed. One engine, a Rolls-Royce, was motored rV  )1( cmV 3 ][ cd , (4) between 1000 and 2000 RPM with several lengths of eff r )1(2 intake pipes of diameters 3.5, 5.3, and 7.3 cm. The c  other engine, a J.A.P. motorcycle engine, was fired from 3 1200 to 4000 RPM with various runner lengths of where Vd [cm ] is the displacement volume and rc is the diameters 3.2, 3.5, and 5.3 cm. Kastner found that, for . This method provides a simple tool an engine running over a wide range of speeds, the for calculating the location of a single intake tuning peak, intake pipe “natural period” should be between 160-180 yet gives no insight into its magnitude. Even though it crank angle degrees (CAD) at the engine speed of incorporates extra parameters over the organ pipe desired tuning, where natural period, p, is defined as method such as the pipe area and cylinder volume, which Thompson (1968) showed gives a predictive 24N " advantage over the organ pipe method, the value for K p [deg] , (2) is slightly ambiguous without experimental data. Intakes V p of non-constant cross-section, such as tapers, can be handled by using (Engelman, 1973)

L N where N [RPM] is the engine speed, Ɛ [cm] is the " dx " j # # . (5) effective intake length, and Vp [cm/s] is the velocity of ¦ A effective ³ AA propagation of the waves. Note that when p = 180 CAD, 0 j 1 j the expression is equal to Eq. (1) with q = 4. Kastner also noted that Vp increased as the intake diameter In 1968, Thompson investigated the effects of increased. not only different lengths and areas of primary runners, but also of curved runners, pipes with and without well- Downing (1958) performed a study on the rounded inlets, non-circular tubes, and runners of effects of primary runner length and diameter using one varying cross-sectional area. His experiments were cylinder of a motored six-cylinder engine. He tested performed on individual cylinders of a motored Cummins runner lengths of 13.3, 27.9, 38.1, 54.6, and 82.6 cm for V6 diesel engine and compression pressure was diameters of 4.0 and 5.0 cm from 1500 to 6000 RPM for acquired from 500 to 3000 RPM. The main purpose of the purposes of designing a complete fuel-injection his thesis was to validate Engelman’s approach to intake system for a racing engine, including intake manifold. tuning prediction. He found that, for both intake diameters, as the length increased, the tuning peak moved toward lower speeds. In a precursor to the present study, Howard Also, for same-length runners, the tuning peak of the 5.0 (2003) studied the effects of tapered runners on a cm diameter pipe was at a higher speed than the 4.0 cm motored single cylinder research engine. His tapers pipe except for the shortest pipe where the tuning peaks varied from TARs of 1.5 to 3 and the taper length was for both diameters were nearly identical. varied from 25% to 100% of the overall length of the test piece. He examined the tapers’ effects on Șv as well as Later, Engelman and his coworkers developed intake and exhaust pressure. He found that, as TAR an intake tuning prediction tool based on a lumped increased, the tuning peaks moved to higher speeds, Helmholtz resonator; they modeled the intake and which correlated well with the predicted speeds given by cylinder by treating the air in the cylinder as a spring with the lumped Helmholtz intake approximation. He also no inertia and the air in the runner as a mass with no observed that, in general, the peak intake pressure near compressibility. The tuning peak is predicted to occur the valve before intake valve closing (IVC) was when the resonant frequency of the Helmholtz system is indicative of Șv trends, but the correlation was sensitive around twice the piston frequency. For a single cylinder to the timing of the peak pressure. with an intake runner open to atmosphere, Engelman (1973) suggested for the speed of the tuning peak Although the literature is extensive on the effects of primary intake runner geometry, most of it has focused on changing lengths of constant-area tubes. Little is available on the effects that taper and inlet 2 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 radius of the primary runner have on the Șv, brake The single cylinder engine was connected to a power, and intake pressure of an IC engine. Thus, the General Electric DC engine dynamometer (dyno) objective of the present study is to systematically study capable of both motoring the engine and absorbing the the effect of these geometries on engine performance. load while firing at a specific speed. It is capable of Seven tapered runners with TAR varying from 1.5 to 3 delivering 134 kW as a motor and absorbing 142 kW as and taper length varying from 25% to 100% of the a dyno; the maximum speed is 6300 RPM. The dyno overall length of the test piece and four bellmouth was computer-controlled by Horiba Systems EDTCS- runners varying from Ri/D of 0.05 to 1 were tested on a 1000. Each intake was tested from 1000 to 5500 RPM single cylinder engine at speeds from 1000 to 5500 in 500 RPM increments from 1000 to 2000 RPM, then in RPM. 250 RPM increments from 2000 to 5500 RPM. Additional points were tested around the Șv peaks for EXPERIMENTAL SETUP each intake configuration. All experiments were performed at wide-open throttle (WOT), and the oil and ENGINE - All experiments for this work were cooling water were held to 93°C. performed on a single cylinder research engine designed by Ford Motor Company. This is a 4-valve INTAKE AND EXHAUST - The intake of the single spark-ignition engine designed to mimic one cylinder of cylinder engine, shown schematically in Fig. 1, consists the Jaguar 3.0L V6 X200 engine, including the of the intake port, a fuel rail block, an adapter section, a combustion chamber, and stroke, piston geometry, barrel-style throttle, and the test piece. The intake port valve timing, and intake and exhaust ports. Although the is a split-port design with a diameter of 3.02 cm at each Jaguar engine has variable cam phasing, the single valve; the port stays separated for 6.48 cm, then the two cylinder engine does not. The only engine accessory is branches come together in an oval of 14.32 cm² for 3.52 a dry sump oil pump which is run off the ; the cm until the head face. Although all experiments are water pump is electric and not powered by the engine. done at WOT, a barrel throttle is used in case the engine The follower is a direct-acting mechanical needed to be throttled due to unexpected knocking or bucket. The overall engine specifications are given in other engine-damaging phenomena. The throttle is Table 1. made such that at WOT, there would be no obstruction in the flow to the pressure waves. The only component of the intake that varied for each experiment is the test piece, which is described in detail in the next section. Bore 8.90 cm Stroke 7.95 cm The exhaust, shown schematically in Fig. 2, is Rod Length 13.81 cm designed to lead the exhaust gas outside as efficiently Compression Ratio 10.5:1 as possible while keeping external noise to an 3 acceptable level. It is made of 6.03 cm diameter straight Clearance Volume 47.10 cm pipe and mandrel-bent 90° elbows with bend radii of Maximum Valve Lift 19.4 cm. The exhaust port is a split-port design with a Intake 0.914 cm diameter of 2.54 cm at each valve; the port stays Exhaust 0.937 cm separated for 4.86 cm, and then comes together into a Valve Timing 10.18 cm² oval for a length of 3.48 cm until the head Intake Open 308.0 CAD face. A 16.5 cm long adapter, shown in Fig. 3, provides Intake Duration 286.0 CAD a smooth transition from the oval exhaust port of 10.18 Exhaust Open 86.5 CAD cm² to the rest of the exhaust with a diameter of 6.03 Exhaust Duration 326.0 CAD cm. A silencer is placed between the exit of the 6.03 cm diameter section of the exhaust and the large tube that routes the exhaust gas outside. Table 1. Single Cylinder Research Engine Specifications. TEST PIECES – Twelve configurations consist of a baseline case and two groups of geometries: tapers and bellmouths. One characteristic common to all test pieces is the large flange at the entrance to the piece, The fuel and spark delivery to the engine was which ensures a hemispherical propagation of the controlled by the Haltech E6K programmable engine pressure waves leaving the duct. The baseline, control unit. It was connected to the engine through a designated intake #1 and shown in Fig. 4a, is a straight custom-built wiring harness and programmed through a duct 26.45 cm long with an inner diameter (D) of 4.20 PC. The spark timing was set at each speed to a cm. At the inlet of the baseline piece is a 1.45 cm radius conservative amount close to MBT of the baseline and bellmouth and the large flange. remained constant for each speed during all experiments. The air-fuel ratio was set to 12.5:1 for all experiments and monitored with a Horiba Mexa-110Ȝ AFR Analyzer. 3 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 Fuel Rail Adapter Barrel Block Section Throttle Test Piece

Pressure Engine Transducer (i3) Head Face 12.0

5.14 9.51 8.67 20.0

10.0

Intake 4.31 equivalent 4.20 DIA 4.20 DIA Pressure Valves DIA Transducer (i1) Pressure Transducer (i2) Dimensions are in cm

Figure 1. Single Cylinder Intake.

Dimensions are in cm to outside

Muffler 55.8

6.35 R 19.4 263

R 19.4

Pressure In-Cylinder Transducer (e1) Pressure Transducer (c1) 90.2 12.7

16.5

R 19.4 Engine

Figure 2. Single Cylinder Exhaust.

4 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 Pressure Head Face Transducer (e1)

Oxygen Sensor 27.3 Emissions 9.53 Tap

3.18 3.60 equivalent DIA

8.34 8.57 Exhaust 6.03 DIA Thermocouple Valves Dimensions are in cm 16.5

Figure 3. Adapter Section of Single Cylinder Exhaust.

Pressure Transducer (i1)

20.0 5.00

4.20 DIA 4.20 DIA

Ri All Dimensions are in cm

(a)

Pressure Transducer (i1) 6.45

20.0

4.20 DIA Dt

Ri=1.45 All Dimensions are in cm

Lt (b)

Figure 4. Intake Test Pieces: (a) baseline and bellmouth group, (b) taper group.

5 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 The tapers, referred to as intakes #2 – 8, are the test piece. An aluminum bushing is made to push shown in Fig. 4b along with the key dimensions given in into each test piece and fish-mouthed to fit flush with the Table 2. They have overall lengths and inlet radii (Ri) inner wall of the duct. The transducer threads tightly into equal to the baseline. The TAR varies from 1.5 to 3. this bushing, also is flush with the inner wall of the duct For TAR = 1.5, the length of the tapered section (Lt) and is perpendicular to and points toward its centerline. varies from 25% to 100% of the overall length and for For each piece, the i1 transducer is located 20.00 cm TAR = 2, Lt varies from 50% to 100%. The others with upstream from the surface that mates with the barrel TAR = 2.5 and 3 are tapered over their entire length. throttle. The pressures acquired at locations i1, i2, and i3 have a resolution of 1 CAD and are averaged for 64 cycles. Above the inlet of the intake duct is a thermocouple for measuring ambient air temperature.

Intake # Lt (cm) Dt (cm) TAR 2 6.613 The exhaust of the single cylinder engine is 3 13.23 5.14 1.5 fitted with a pressure transducer, an emissions tap, a 4 26.45 thermocouple, and a wide-band oxygen sensor (recall 5 13.23 Fig. 3). The pressure transducer, e1, is a Kistler 4045A5 5.94 2 6 26.45 piezoresistive pressure transducer with a 0-5 bar range. 7 6.62 2.5 This transducer threads into a water jacket, which is 26.45 8 7.26 3 supplied with cooling water. The water jacket is threaded into its bung in the adapter section such that it is flush with the inner wall of the duct. The e1 transducer is located 9.53 cm from the head face. As Table 2. Taper Group Dimensions. with the intake pressure traces, the exhaust pressure traces at e1 have a 1 CAD resolution and are averaged for 64 cycles. The thermocouple is located 8.57 cm from the head face. The wide-band oxygen sensor is located The bellmouth group, labeled as intakes #9 – 27.3 cm from the head face and connected to a Horiba 12, is shown in Fig. 4a along with the key dimensions MEXA-110Ȝ AFR Analyzer. The emissions tap is given in Table 3. Each bellmouth piece retains the connected first through stainless steel tube and then length from the start of the test piece to R and D of the i macroline to a Horiba MEXA-7100 Motor Exhaust Gas baseline, and R /D varies from 0.05 to 1.0. i Analyzer. The engine brake and motoring are measured with a Revere Transducers load cell (model 606555-24). The fuel flow is metered with the Pierburg Instruments PII401*11863 Fuel Measurement System. Intake # Ri (cm) Ri/D Overall Length Emissions, , and fuel flow data are acquired by (cm) the Horiba dyno controller computer and are averaged 1 (baseline) 1.45 0.35 26.45 over a minute at each engine speed. 9 0.21 0.05 25.21 10 0.84 0.20 25.84 A PCB 145A07 Piezotronics in-cylinder pressure 11 2.10 0.50 27.10 transducer, labeled c1 in Fig. 2, is located in the top 12 4.20 1.00 29.20 center section of the combustion chamber, fitting flush with its inner surface, between the and the rearward-most intake and exhaust valves. The output, Table 3. Bellmouth Group Dimensions. amplified via a Kistler 5010 dual-mode amplifier, is recorded by the Concurrent data-acquisition system.

Using the air-fuel ratio obtained from the MEXA- EXPERIMENTAL MEASUREMENTS - The intake of 7100 and the fuel flow rate, Șv is calculated for each the single cylinder engine is outfitted with three Kistler engine speed by 4045A2 piezoresistive pressure transducers with a 0-2 bar range. Each of these transducers is connected through a Kistler 4603A Amplifier to a Concurrent high- 2m a K , (6) speed data acquisition system (model MC68040). This v U NV system is capable of acquiring data at a sampling rate of , dia 2 MHz from 32 channels simultaneously. The pressure transducers labeled i2 and i3 are both located in the wherem a [g/s] is the mass flow rate of air (determined adapter section of the intake, 14.65 cm from the head from the measured fuel flow rate and the air-fuel ratio) face as shown in Fig. 1. Both transducers are flush with 3 into the cylinder, Vd [cm ] is the displacement volume, N the inner wall of the adapter piece, and both are 3 [RPS] is the engine speed, and ȡ [g/cm ] is the inlet air perpendicular to and point toward the centerline of the a,i density. The inlet air density is determined from the duct. The pressure transducer labeled i1 is located in 6 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 ideal gas relationship by using the barometer and 117 21 temperature readings. solid = run 1 Power dashed = run 2 110 18 RESULTS

103 15 This section presents Șv, brake power, and intake pressures for each test piece. For brevity, only intake pressures at i2 are presented since they are more 96 12 indicative of Ș trends than i1, which is further upstream. v Volumetric Efficiency Locations i2 and i3 were the same distance from the 89 9 back of the intake valves, and pressures measured there Volumetric Efficiency (%) Corrected Brake PowerCorrected (kW) proved to be identical. For further information on 82 6 pressure measurements from the i1, e1, and c1 locations, the reader is referred to Mariucci (2006). 75 3 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 Engine Speed (RPM) BASELINE – The Șv and brake power for repeated runs of the baseline are presented in Fig. 5. The level of repeatability observed in this figure is similar in other test Figure 5. Experimental Șv and Brake Power for the Baseline. pieces. There are three distinct Șv peaks, at 3000, 3750, and 4750 RPM. With c = 348 m/s and Ɛ = 0.6226 m (the total geometric length of the intake), the intake tuning prediction method developed by Morse et al. [recall Eq.

(1)] gives the tuning peak speeds shown in Table 4. 115 This method over-predicts the tuning peak speeds for all intake # 1 three values of q. Using Engelman’s lumped-parameter 110 intake # 2 Helmholtz model [recall Eq. (3)] with c = 348 m/s, Ɛ = intake # 3 intake # 4 62.26 cm, A = 13.85 cm2, and a preliminary guess of K = 105

2.1, a tuning peak location was predicted at 4314 RPM, 100 a 15% difference when compared to the largest experimental tuning peak at 3750 RPM. To accurately 95 predict this tuning peak, K = 2.42, within the range prescribed by Engelman. 90

Volumetric Efficiency (%) 85

80 q Predicted Experimental % Difference (a) 75 Peak (RPM) Peak (RPM) 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 3 5590 4750 17.7 Engine Speed (RPM) 4 4192 3750 11.8 5 3353 3000 11.8 21 intake # 1 intake # 2 18 Table 4. Morse et al. Method for Intake Tuning intake # 3 intake # 4 Prediction. 15

12 TAPERS - The measured Șv and brake power of the engine for each taper (intakes #2-8) are compared to the baseline. Figure 6 shows all intakes with TAR = 1.5 9 (intakes #2-4) and Fig. 7 with TAR = 2.0 (intakes #5 and Corrected Brake Power Brake (kW) Corrected 6). All intakes that are tapered over their full length 6 (intakes #4, 6, 7, and 8) are compared in Fig. 8 to show (b) the overall trend of increasing TAR. The location of the 3 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 peak Șv and its magnitude are given in Table 5 for each taper. Engine Speed (RPM)

As predicted by Eq. (3), the tuning peaks move Figure 6. Experimental (a) Șv and (b) Brake Power for toward higher engine speeds as Ɛ/Aeffective decreases. Intakes with TAR = 1.5 Compared to the Baseline. The Ɛ/Aeffective for each taper, calculated using Eq. (5), along with the predicted tuning peaks using the lumped- 7 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 115 21 intake # 1 intake # 1 110 intake # 5 intake # 5 18 intake # 6 intake # 6 105 15 100

95 12

90 9

Volumetric Efficiency (%) 85 Corrected Brake Power (kW) 6 80 (a) (b) 75 3 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 Engine Speed (RPM) Engine Speed (RPM)

Figure 7. Experimental (a) Șv and (b) Brake Power for Intakes with TAR = 2.0 Compared to the Baseline.

115 21 intake # 1 intake # 1 110 intake # 4 intake # 4 intake # 6 18 intake # 6 105 intake # 7 intake # 7 intake # 8 intake # 8 15 100

95 12

90 9

Volumetric Efficiency (%) 85 Corrected Brake Power(kW) 6 80

(a) (b) 75 3 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 Engine Speed (RPM) Engine Speed (RPM)

Figure 8. Experimental (a) Șv and (b) Brake Power for Intakes Tapered over Entire Length Compared to the Baseline.

Intake Peak Șv Peak Șv % difference Intake Ɛ/A|effective Predicted Experimental % -1 # location magnitude from baseline # (cm ) Peak Peak (RPM) Difference (RPM) (%) (RPM) 1 3750 111.5 -- 2 4.407 3787 3850 1.6 2 3850 112.1 0.53 3 4.319 3825 3875 1.3 3 3875 110.5 0.89 4 4.145 3905 4000 2.3 4 4000 111.6 0.09 5 4.214 3873 4000 3.1 5 4000 110.7 0.72 6 3.935 4008 4125 2.8 6 4125 109.6 1.70 7 3.800 4078 4250 4.0 7 4250 108.5 2.69 8 3.689 4139 4375 5.4 8 4375 107.5 3.59 Table 6. Taper Intake Tuning Predictions using the Lumped-Parameter Helmholtz Approximation. Table 5. Șv Characteristics of the Taper Group.

8 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 parameter Helmholtz approximation are presented in approximate time in CADs required for the compression Table 6. Experimentally, the location of peak Șv wave to travel from i2 to the back of the intake valves increases by 125 RPM as TAR increases by 0.5 for each may be estimated from intake tapered over its entire length; Eq. (3) did not predict the increase to be so large, thus the prediction 6 '" N error associated with this method generally increases 'CAD , (7) with TAR. The overall magnitudes of peak Șv remain c similar to the baseline for intakes #2 – 5, while the peak Șv begins to diminish for intakes #6 – 8. This may be where ǻƐ is the distance from i2 to the back of the due to the decrease in Ɛ/Aeffective, which decreases the valves (= 0.2465 m for this study), N [RPM] is the engine inertial effect of the air due to weaker reflected speed, and c [m/s] is the speed of sound. Equation (7) compression waves. Another contributor may be the gives ǻCAD = 16° at 3750 RPM, 17° at 4000 RPM, and flow losses having a more pronounced effect as the 18° at 4250 RPM. Each pressure trace is presented in engine speed of the peaks increases (Heywood, 1988). the crank-angle resolved time domain, as well as the engine-order resolved frequency domain. The brake power for each taper generally follows Șv, with spikes in power occurring near Șv peaks. Figure 9 shows Pi2 at 3750 RPM for intakes #1 Peak power and its location are given in Table 7 for and 4. For intake #4, the compression wave returning to each taper. For intakes #2 – 5, the peak brake power i2 near IVC is reduced in magnitude by 0.166 bar occurs near the highest-speed Șv peak and is of similar (12.6%) at its peak, which decreases the effectiveness magnitude to the baseline. The peak power for intakes of intake tuning; at the same speed, Fig. 6 shows a #6 – 8 are near the maximum Șv speeds, and substantial reduction in Șv of 9.4% from intake #1 to intake #4. The increases in peak power are seen w.r.t. the baseline. dominant frequency of the “quasi standing wave” High-speed brake power generally increases for the (QSW), defined as the pressure wave in the intake tapered intakes compared to the baseline, with the during the intake valve (IV) closed period, has increased largest belonging to intake #6 (30.0% at 5500 RPM), for intake #4, from about 136 Hz for the baseline to 150 intake #7 (35.8% at 5500 RPM), and intake #8 (33.6% at Hz. Note that any dominant frequency of the QSW 5500 RPM). presented in this paper is calculated by measuring the CAD between peaks of the QSW. For the straight baseline, the dominant frequency of the QSW is similar to the first resonance frequency of a quarter-wave Intake Peak power Peak power % difference silencer, # location magnitude from (RPM) (kW) baseline c f , (8) 1 4750 18.66 -- 4 2 4750 18.87 1.13 " 3 4750 18.61 0.27 where c [m/s] is the speed of sound and Ɛ [m] is the 4 5000 18.56 0.54 effective length of the intake duct. With c = 348 m/s and 5 5000 18.31 1.88 Ɛ = 0.6226 m, ƒ becomes 140 Hz, a 3% difference with 6 4250 19.06 2.14 respect to the measured frequency. The increase in 7 4375 19.35 3.70 average intake area for the tapered case has a similar 8 4500 19.77 5.95 effect on the frequency of the QSW as shortening the effective length of the intake. For intake #4, there is a 5.5 dB reduction in sound pressure level (SPL) at order Table 7. Brake Power Characteristics of the Taper 2, the dominant order. Figure 10 shows Pi2 at the main Group. tuning peak speed for intake #4, 4000 RPM. At this speed, the peak magnitude of the compression wave near IVC is larger for the baseline than for intake #4; however, it occurs 18 CAD later for intake #1 and right at For brevity, only the intake pressure at the i2 IVC. Thus, it seems that this compression wave is location (Pi2) are compared for a select set that includes slightly late in getting from i2 to the back of the valves, intakes #1, 4, 6, 7, and 8. These intakes were chosen while the compression wave for intake #4 is early because of their significant differences in Șv with respect enough to affect positive intake tuning, giving a rise of to one another. Comparisons are presented for speeds 11% in Șv over the baseline case. At 4000 RPM, the corresponding to the main tuning peak of the baseline dominant frequency of the QSW has increased for the case, 3750 RPM, and the main tuning peak of each of taper case just as it had for 3750 RPM. The peak SPL the other intakes. While examining the impact of of intake #4 has decreased by 0.5 dB compared to pressure waves at the intake valve in the following intake #1, while those of neighboring orders have comparisons, the distance between the transducer increased. location i2 and the valve should be kept in mind. The 9 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 1.4 180 intake #1 intake #1 intake #4 intake #4 1.3 170

160 1.2

150 1.1 140

1 (dB) SPL

Pressure (bar) Pressure 130

0.9 120

0.8 110 (a) c/4L (baseline) (b) IVO EVC IVC 0.7 100 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 9. Experimental Pi2 at 3750 RPM for Intakes #1 and 4 in the (a) Time Domain and (b) Frequency Domain.

1.4 180 intake #1 intake #1 intake #4 1.3 intake #4 170

1.2 160 1.1

1 150 SPL (dB) SPL

Pressure (bar) 0.9 140

0.8

130 0.7 c/4L (baseline) (a) (b) IVO EVC IVC 120 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 10. Experimental Pi2 at 4000 RPM for Intakes #1 and 4 in the (a) Time Domain and (b) Frequency Domain.

Figure 11 compares Pi2 at 3750 RPM for intakes average pressure between intake valve opening (IVO) #1 and 6. For intake #6, the compression wave arriving and exhaust valve closing (EVC) (during the valve at i2 near IVC is drastically reduced, by 0.182 bar overlap period) is higher for intake #6. This may mean (13.8%), at its peak. At the same speed, Fig. 8 shows the difference between pulling fresh charge into the that Șv is reduced by 11% compared to the baseline. As cylinder (increasing Șv) and back flow of exhaust gas with intake #4, an increase in the dominant frequency of (decreasing Șv) during the overlap, and may be another the QSW is observed for intake #6 compared to the reason for the 13% increase in Șv for intake #6 over baseline: about 150 Hz for intake #6 vs. 136 Hz for the intake #1 at 4250 RPM. Compared to the baseline, the baseline. The peak SPL for intake #6 has moved from frequency spectrum shows a decrease in SPL from order 2 to 2.5 and reduced by 5.9 dB compared to intake order 0.5 to 1.5 and an increase from orders 2 to 3 for #1. A direct comparison of pressure at the intake #6 the taper, with the largest increase (15.5 dB) at order tuning peak speed of 4125 RPM is not made since that 2.5. speed is not acquired for the baseline case; instead

4250 RPM is presented in Fig. 12. The peak Pi2 at 3750 RPM for intake #7 is compared to the magnitudes of the compression waves occurring during baseline in Fig. 13. The peak pressure of intake #1 is the IV open period are similar for both intakes #6 and 1, 0.15 bar (11.4%) higher than intake #7. Also, the but the peak occurs 10° sooner for intake #6, while the average Pi2 during valve overlap is higher for intake #1. peak for the baseline case arrives at i2 right at IVC and Both of these phenomena likely contribute to the 11% thus its effect on intake tuning has decreased. Also, the reduction in Șv for intake #7. The frequency spectra

10 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 1.4 180 intake #1 intake #1 intake #6 intake #6 1.3 170

1.2 160

1.1 150

1 (dB) SPL 140 Pressure (bar)

0.9 130

0.8 120 c/4L (baseline) (a) (b) IVO EVC IVC 0.7 110 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 11. Experimental Pi2 at 3750 RPM for Intakes #1 and 6 in the (a) Time Domain and (b) Frequency Domain.

1.4 180 intake #1 intake #1 1.3 intake #6 intake #6 170

1.2

160 1.1

1 150 SPL (dB) SPL

Pressure (bar) 0.9 140

0.8

130 0.7 c/4L (baseline) (a) (b) IVO EVC IVC 120 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 12. Experimental Pi2 at 4250 RPM for Intakes #1 and 6 in the (a) Time Domain and (b) Frequency Domain.

1.4 180 intake #1 intake #1 intake #7 intake #7 1.3 170

1.2 160

1.1 150

1 (dB) SPL 140 Pressure (bar)

0.9 130

0.8 120 c/4L (baseline) (a) (b) IVO EVC IVC 0.7 110 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 13. Experimental Pi2 at 3750 RPM for Intakes #1 and 7 in the (a) Time Domain and (b) Frequency Domain.

11 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 show that the peak SPL has shifted from order 2 to 2.5 For intake #9, Șv at each tuning peak has and reduced by 4.6 dB for intake #7. Figure 14 decreased, with the magnitudes at 3000, 3750, and compares the same intakes at the main tuning peak 4750 RPM, lowered by 3.1%, 3.8%, and 3.5%, speed of 4250 RPM for intake #7. The peak magnitude respectively. Ri/D = 0.05 results in an inlet loss of the compression wave near IVC is lower for intake #7 coefficient of about 0.2 (Miller, 1990), which is than for intake #1; however, the compression wave for measurably detrimental to the overall engine the baseline arrives at i2 later and right at IVC, thus performance. Compared to the baseline, the peak reducing its effect on Șv and allowing higher Șv for intake power for intake #9 has reduced slightly, and an #7 at this speed. Compared to the baseline, the SPL for increase in power at 5250 to 5500 RPM is noticeable intake #7 is 15.5 dB higher at order 2.5 and slightly corresponding to the increase in Șv at those speeds. lower from orders 0.5 to 2. The Șv and brake power of intakes #1 and 10 (Ri/D = 0.20) are nearly identical. For an Ri/D > 0.15, the inlet Figure 15shows Pi2 of the largest tapered intake loss is almost negligible (Miller, 1990), which appears to (Dt = 7.26 cm), intake #8, compared to intake #1 at the be the case for the bellmouth group, as the Șv main tuning peak of intake #1, 3750 RPM. A reduction magnitudes are similar for intakes #10 - 12. of 0.14 bar (10.6%) can be seen in the peak pressure for intake #8, which decreases its effectiveness on intake It appears that Șv of intake #11 has shifted tuning at this speed, contributing to a 9% reduction in Șv toward lower speeds slightly, increasing at 3650 RPM compared to baseline. As with the other tapered cases, and decreasing at 3850, 4000, and 4750-5500 RPM. the dominant frequency of the QSW has increased for Selamet et al. (2001) have shown that for a duct with intake #8, from about 136 Hz for the baseline to about flanged bellmouth, as Ri/D increases, the length of the 161 Hz, an increase of 18.4% over the baseline and end correction, a hypothetical length of pipe attached to 7.3% over intake #6. When compared to intake #1, the the inlet of the duct to account for inertial effects of the frequency spectrum for intake #8 shows a decrease in air, increases for frequencies low enough to affect SPL from orders 0.5 to 2 followed by an increase from engine tuning. This gives a longer effective length for orders 2.5 to 4, with the peak SPL moving from order 2 intake #11 compared to intake #1 despite both intakes to 2.5. Data is not acquired for the baseline case for the having equal distance to the bellmouth, causing the shift exact Șv peak speed for intake #8 of 4375 RPM, thus Pi2 in Șv toward lower speeds for the former. The slight shift is presented at 4250 RPM instead in Fig. 16. The peak in Șv between the two intakes does not have much pressure of intake #8 is 0.132 bar (9.9%) lower than that influence on the brake power. of the baseline, but it occurs 25° earlier, which is early enough to give the cylinder a supercharging effect For intake #12, since the bellmouth is larger, the before IVC, aiding in the 8% increase in Șv for intake #8. end correction is larger than that for the baseline and The SPL from orders 0.5 to 2, have decreased for intake intake #11; therefore the overall length of the intake is #8, while that of order 2.5 has increased by 16.0 dB longer, and Șv continues to shift toward lower speeds for compared to intake #1. intake #12 compared to the baseline or intake #11, with tuning peaks moving from 3750 and 4750 RPM of the BELLMOUTHS – First, the measured Șv and corrected baseline to 3650 and 4650 RPM for intake #12. The brake power of the engine for each bellmouth (intakes power, following the trend of Șv, shifts toward lower #9-12) are compared to the baseline. Figure 17 shows engine speeds for intake #12, while the peak power intakes #9 and 10 (Ri/D = 0.05 and 0.20, respectively) magnitude remains similar. and Fig. 18 shows intakes #11 and 12 (Ri/D = 0.50 and 1.0, respectively). Table 8 gives the location and Figure 19 shows the Pi2 for intakes # 1 and 9 at magnitude of peak Șv for each bellmouth. 3750 RPM. The peak pressure near IVC of intake #9 is 0.063 bar (4.8%) lower than that of the baseline, and the pressures have also decreased during valve overlap. These reductions in pressure contribute to a drop in Șv of Intake Peak Șv Peak Șv % difference 4% for intake #9. The overall behavior of the frequency # location magnitude from baseline spectra is similar for both intakes, with similar or lower (RPM) (%) SPL for intake #9 for most orders. 1 3750 111.5 -- 9 3850 107.9 3.23 Figure 20 shows Pi2 for intakes #1 and 10 at 10 3750 111.9 0.36 3750 RPM. Since Șv for these two intakes is nearly 11 3750 110.7 0.72 identical, it is expected that the intake pressures should 12 3650 111.5 0 also be nearly identical. This proves to be the case, as only the subtlest of differences can be seen in both the time and frequency domain. The pressure shift seen in the time domain may be attributed partly to the change Table 8. Șv Characteristics of the Bellmouth Group. in barometric pressure between experiments, with that of intake #1 being 0.013 bar higher.

12 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 1.4 180 intake #1 intake #1 intake #7 intake #7 1.3 170

1.2 160

1.1 150

1 (dB) SPL Pressure (bar) 140 0.9

130 0.8 c/4L (baseline) (a) (b) IVO EVC IVC 0.7 120 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 14. Experimental Pi2 at 4250 RPM for Intakes #1 and 7 in the (a) Time Domain and (b) Frequency Domain.

1.4 180 intake #1 intake #1 intake #8 intake #8 1.3 170

160 1.2

150 1.1 140

1 (dB) SPL

Pressure (bar) 130

0.9 120

0.8 110 (a) c/4L (baseline) (b) IVO EVC IVC 0.7 100 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 15. Experimental Pi2 at 3750 RPM for Intakes #1 and 8 in the (a) Time Domain and (b) Frequency Domain.

1.4 180 intake #1 intake #1 intake #8 intake #8 1.3 170

1.2 160

1.1 150

1 (dB) SPL Pressure (bar) 140 0.9

130 0.8 c/4L (baseline) (a) (b) IVO EVC IVC 0.7 120 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 16. Experimental Pi2 at 4250 RPM for Intakes #1 and 8 in the (a) Time Domain and (b) Frequency Domain.

13 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 115 21 intake # 1 intake # 1 intake # 9 110 intake # 9 intake # 10 18 intake # 10 105

15 100

95 12

90 9

Volumetric Efficiency (%) Efficiency Volumetric 85 Corrected Brake Power (kW) 6 80

(a) (b) 75 3 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 Engine Speed (RPM) Engine Speed (RPM)

Figure 17. Experimental (a) Șv and (b) Brake Power for Intakes #9 and 10 compared to the Baseline.

115 21 intake # 1 intake # 1 110 intake # 11 intake # 11 18 intake # 12 intake # 12 105 15 100

95 12

90 9

Volumetric Efficiency (%) Efficiency Volumetric 85 Corrected Brake Power (kW) 6 80

(a) (b) 75 3 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 Engine Speed (RPM) Engine Speed (RPM)

Figure 18. Experimental (a) Șv and (b) Brake Power for Intakes #11 and 12 compared to the Baseline.

1.4 180 intake #1 intake #1 intake #9 intake #9 1.3 170

1.2 160

1.1 150

1 (dB) SPL 140 Pressure (bar)

0.9 130

0.8 120 c/4L (baseline) (a) (b) IVO EVC IVC 0.7 110 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 19. Experimental Pi2 at 3750 RPM for Intakes #1 and 9 in the (a) Time Domain and (b) Frequency Domain.

14 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 1.4 180 intake #1 intake #1 intake #10 intake #10 1.3 170

1.2 160

1.1 150

1 (dB) SPL Pressure (bar) Pressure 140 0.9

130 0.8 c/4L (baseline) (a) (b) IVO EVC IVC 0.7 120 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 20. Experimental Pi2 at 3750 RPM for Intakes #1 and 10 in the (a) Time Domain and (b) Frequency Domain.

Pi2 are compared for intakes #1 and 11 in Fig. similar trend to the Șv, with intake #8 making the highest 21 for the baseline tuning peak speed of 3750 RPM. power of all tapers. The compression wave returning to The peaks and valleys of the wave for intake #11 trail i2 near IVC is known to cause intake tuning. However, those of the baseline by 2 to 4 CAD at this speed, the location of its peak must be taken into account as indicating that the effective length of this intake is slightly the wave takes about 17° to travel from i2 to the back of longer; the peak magnitudes are similar. At this speed, the valves at speeds typical of peak Șv for this group. If Șv for both intakes are also similar. The frequency the peak of the compression wave for the baseline was spectra for intakes #1 and 11 are almost identical for the closer to IVC than that of any taper and within 17° of dominant engine orders (from 0.5 to 2.5), while the IVC, then a significant decrease in Șv was observed for overall behavior of SPL for higher orders are similar. the baseline even if the taper had a lower peak pressure magnitude near IVC. For example, intake #4 had a Figure 22compares Pi2 for the largest bellmouth lower peak pressure near IVC at 4000 RPM, but it (Ri/D = 1.0), intake #12, to the baseline for its Șv peak occurred 18° earlier than that of the baseline, which speed of 3750 RPM. The dominant frequency of the contributed to an 11% increase in Șv for the taper. If the QSW is noticeably lower for intake #12 than it is for the compression wave peaks of the baseline and any taper baseline, by approximately 5 Hz, indicating that the were both an ample time ahead of IVC, then the effective length of intake #12 is longer than that of the difference in peak pressures was of the same order of baseline, despite both intakes having the same length to magnitude as the difference in Șv. For example, the the bellmouth. The phase lag from intake #1 to intake pressures for intakes #6 and 8 were 13.8% and 10.6% #12 is also considerably larger (5 to 15 CAD) than for lower than the baseline at 3750 RPM and Șv were 11.0% intake #11. Both these aspects suggest that, generally, and 9% lower. For the tapered intakes, the dominant a longer end correction is needed as Ri/D increases; frequency of the QSW at i2 increased compared to the Mariucci (2006) explores end corrections determined baseline, which is an effect similar to shortening the from the present study further. The peak pressure is length of a closed-end quarter-wave silencer. higher for intake #12 than for intake #1, but it occurs 7 CAD later for intake #12 (and only 10 CAD before IVC), The volumetric efficiency of the bellmouth group which appears to reduce its effect on intake tuning, as Șv showed slight movement of tuning peaks to lower for intake #12 is 5% lower than intake #1 at this speed. speeds as Ri/D increased. The lumped Helmholtz At this speed, the SPL of these two intakes are within 1 approach suggests that increased effective length dB from orders 0.5 to 2.5. causes this. The trend of increased length with Ri/D is echoed in the pressures at i2; the phase shifts from the CONCLUSIONS baseline indicates if the end correction for any particular intake configuration in this group increases or decreases The volumetric efficiency of the tapers exhibits with respect to that of the baseline. If the pressures led increasing tuning peak speeds with decreasing Ɛ/Aeffective. the baseline, a smaller end correction was needed, and The lumped Helmholtz approximation of the intake if the pressures trailed the baseline, a larger. A system accurately predicts this trend. As the taper area reduction in Șv was observed at the tuning peaks for ratio is increased above 1.5, the peak Șv begins to intake #9 (Ri/D = 0.05) due to presumably inlet flow deteriorate. The shape of the power curves follow a losses. When Ri/D > 0.20, no losses were evident, as the tuning peaks Șv of intakes #10 – 12 were similar.

15 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 1.4 180 intake #1 intake #1 intake #11 intake #11 1.3 170

1.2 160

1.1 150 1

SPL (dB) SPL 140

Pressure (bar) Pressure 0.9

130 0.8

0.7 120 c/4L (baseline) (a) (b) IVO EVC IVC 110 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 21. Experimental Pi2 at 3750 RPM for Intakes #1 and 11 in the (a) Time Domain and (b) Frequency Domain.

1.4 180 intake #1 1.3 intake #12 170

1.2 160

1.1 150 1

SPL (dB) SPL 140

Pressure (bar) Pressure 0.9

130 0.8

0.7 120 intake #1 c/4L (baseline) intake #12 (a) (b) IVO EVC IVC 110 0 90 180 270 360 450 540 630 720 0 1 2 3 4 5 6 7 8 9 10 CAD Engine Order

Figure 22. Experimental Pi2 at 3750 RPM for Intakes #1 and 12 in the (a) Time Domain and (b) Frequency Domain.

Brake power trends generally followed Șv. The pressure REFERENCES peak trend at i2 before IVC was generally indicative of the Șv behavior, although the timing of the peak was an Downing, E. W., 1957-58, “Petrol Injection: Some important factor for intake #12. Further Developments,” Institution of Mechanical Engineers: Proceedings of the Automobile Division, 6: Extensive engine simulation work has also been 161-173. performed using the experimental data presented in this paper. For more information, the interested reader is Engelman, H. W., 1973, “Design of a Tuned Intake referred to Mariucci (2006). Manifold,” ASME Paper 73-WA/DGP-2.

ACKNOWLEDGMENTS Heywood, J. B., 1988, Internal Combustion Engine Fundamentals. McGraw-Hill, Inc., New York. The authors would like to acknowledge Ford Motor Company for their generous contribution of the Howard, T. M., 2003, “Tapered Intakes on a Single single cylinder research engine and test pieces, Cylinder Engine,” MS Thesis, The Ohio State University. specifically Dr. Kevin Tallio, Zafar Shaikh, Mike Magnan, Graham Hoare, and Frank Fsadni. Kastner, L. J., 1945, “Induction Ramming Effects in Single-Cylinder Four-Stroke Engines,” Proceedings of the Institution of Mechanical Engineers, 153: 206-220.

16 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90 Mariucci, V. E., 2006, “An Experimental and Computational Investigation of the Effect of Primary Intake Runner Geometry on the Performance of a Single Cylinder Engine,” MS Thesis, The Ohio State University.

Matthews, R. and Gardiner, A. W., 1924, “Increasing the Compression Pressure in an Engine by Using a Long Intake Pipe,” NACA Technical Memorandum 180.

Miller, D. S., 1990, Internal Flow Systems, 2nd Edition. BHRA Information Services, Cranfield.

Morse, P. M., Boden, R. H., Schecter, H., 1938, “Acoustic Vibrations and Internal Combustion Engine Performance I: Standing Waves in the Intake Pipe System,” Journal of Applied Physics, 9: 16-23.

Selamet, A., Ji, Z. L. and Kach, R. A., 2001, “Wave Reflections from Duct Terminations,” Journal of the Acoustical Society of America, 109: 1304-1311.

Thompson, M. P., 1968, “Non-Mechanical Supercharging of a Four-Stroke Diesel Engine,” MS Thesis, The Ohio State University.

17 Author:Gilligan-SID:12324-GUID:50206991-164.107.10.90