CORE Metadata, citation and similar papers at core.ac.uk
Provided by Okayama University Scientific Achievement Repository
MEMOIRS OF THE SCHOOL OF ENGINEERINGOKAYAMA UNIVERSITY Vol. 9, No.2, DECEMBER 1974
Reduction of Engine Exhaust Noise by Throttling in an Exhaust Manifold
Kazuie NISHIWAKI* and Yuzuru SHIMAMOTO*
(Received December, 10, 1974)
Synopsis
This paper covers the work done in an attempt to reduce exhaust noise level without the decay of brake mean effective pressure of a two-cylinder two-stroke cycle engine, by means of an exhaust manifold having a throttle plate in its junction. The data are shown for various dimensions and configurations of the manifold and the discussion is given on the effect of exhaust throttling on exhaust noise level and engine performance. The principal results obtained by this study are as follows: (1) The lower limit of the throttle area was about one third of a cross-sectional area of the manifold from the view point of brake mean effective pressure. (2) Within this limit brake mean effective pressure was scarcely influenced by throttling under the condition where a number of pressure oscillation related to the pulsation effect during an interval between discharges was less than 1.7. (3) A reduction in exhaust noise level attained was 8 dB(A) at higher engine speeds. (4) The junction angle of the manifold gave little significance to exhaust noise level and brake mean effective pressure.
* Department of Mechanical Engineering
35 36 Kazuie NISHIWAKI and Yuzuru SHIMAMOTO
1. Introduction
It may be inevitable in most cases that flow resistance due to silencers and throttled sections in an exhaust system reduces engine output. ~herefore, designers endevour to make a compromice between engine output and noise reduction through silencers. Th~ utili~ation of dynamic effects arising in an exhaust system may be one possible way to make better noise suppression without the decay of brake mean effective pressure, in particular for a two-stroke cycle engine, the performance of which is sensitive to pressure oscillation in the exhaust system. From this view point, an attempt was made on a two-cylinder two stroke cycle engine to set an throttle plate at the junction of the exhaust manifold so that it might reduce the amplitude of pressure oscillation transmitted to silencers. As to the influence of the throttle plate on engine performance, there is a possibility that the oscillating component of flow velocity is absorbed through the junction into the adjacent pipe and thus the throttle gives little influence on engine performance for appropriate dimensions of the manifold in respect to the dynamic effects. This principle is deduced from the experimental work by the authors(l) which indicates that a throttle plate at the end of an exhaust pipe having a resonance pipe gives little influence on inlet charge.
2. Lxperimental Apparatus
The engine used for the investi Gation was a two-cylinder crankcase scavenged otto engine with lead valves ( Mitsubishi 28-10 ), having a displacement volume of 180 cm 3 and an,'exhaust port opening period oflLl2deg.( symmetrical timing ). Fig.l shows the configurations and dimensions of the manifold used in the experiment. As is indicated in the figure, the junction angles of the manifolds tested are 180 deg., 90 deg. , and 45 deg. The lengths Fig.l Configurations and dimen of the manifolds L are shown on sions of the manifolds Reduction of Engine Exhaust Noise by Throttling in an Exhaust Manifold 37
Table 1 Length of the 850 manifolds !!!-. 2035 ~. ,.r. ~ Junction e-,- Length ~ dl angle cA ~_.~~....Jrl ~~fler =====::: 180 deg. 588 700 1400 = mm Exhaust tank : Exhaust system A 90 746 1446 { / Throttle Pre-muffler : Exhaust system B 45 / 798 / Fig.2 Exhaust systems behind the manifold
Table 1. The throttle plate was located just behind the junction. The experiment was done for three different area ratios of the opening area of the throttle to the cross-sectional area of the manifold, including 1/1, 1/3, and 1/6. In Fig.2, two exhaust systems to be connected behind the manifold are shown. The exhaust system A was used for the investigation of basic characteristics of each manifold concerning engine performance, and the exhaust system B for the measurement of exhaust noise level. The tests were carried out at full engine load. 1.0 . ---'--'---T 0(' L om I --180 700 --- 90 746 ,g O.9I-----A=-...... /-'o.--j!----.=-·- 45 798 3. Engine Performance Influenced e Throttle ratio by Throttling o III ~ 0.81-----+-----'>~T_-~0 1/3 ~ 6 I~ 3.1 Delivery Ratio ~ Delivery ratio,based on atmos pheric conditions and the displace ment volume, was obtained from flow rate through orifices in pipes. Q6r--t--~~~~·_~--~ The pipes were connected to an inlet surging tank the volume of which 05 was 1320 times as large as the displacement volume. 'Yi 0.4 Fig.3 presents delivery ratio 2000 3000 4000 5000 Engine speed rpm versus engine speed for the various I 2:5 20 q 1.5 throttle ratios and junction angles ! ! I 1.0 08 m 0.6 of the manifolds followed by the exhaust system A. To clarify the Fig.3 Delivery ratio versus engine dynamic behaviour, the manifolds speed, q, and m for various are reduced to an acoustic model throttle ratios and Junction as shown in Fig.4. For simplicity, angles 38 Kazuie NISHIWAKI and Yuzuru SHIMAMOTO
the exhaust tank is discarded, for the volume of the tank was so large .~-<-~----f-~- that a pressure amplitude produced -J there may be negligible compared l, (.1 with the one in the manifold. The Fig.4 Acoustic model model shown in the figure is related to the dynamic behaviour of the inertia effect. For the model related to the pulsation effect, the boundary between the cylinder and the pipe is replaced by a closed end boundary. Using those symbols which are indicated in the figure, and applying the acoustic impedance theory (2) to this model, a natural frequency (Hz) for the inertia effect Vm may be obtained from the following relation. 1 f/Vzk tan kl + 1 l tan kL -f/Vzk + tan kL tan kl 2 l l where k 211: V I a, a: sound velocity m The sound velocity was given by the indicator diagram measured just behind the exhaust port. A number of pressure oscillation m during the exhaust port opening period is then given by m JJm 801 6n where n engine speed (rpm), eO: exhaust port opening period (degree) A natural frequency (Hz) for the pulsation effect may be given Vq by
Vq = a I 4 II For the analysis of the pulsation effect in an exhaust system of a two-cylinder engine, any characteristic value has not been shown yet. For present purposes, a number of pressure oscillation related to the pulsation effect for an interval of the discharges ( 180 deg. in this case ) is then taken as this characteristic value, considering that the pressure wave transmitted from the other cylinder might be a predominant factor in exhaust and scavenging processes rather than the from preced ing cycle. Thus let q be this number, we have q 180 I 6n = JJq The characteristic values m and q calculated for the manifold without throttling the length of which was 700 mm, are shown below the abcissa in Fig.3. Examination of these values reveals that the inertia effect predominates over the pulsation effect in the lower speed range extended from 2000 rpm to around 3700 rpm, where delivery ratio is greatly influenced by throttling. One important relation is observable from the figure, uamely, that delivery ratio for a throttle ratio of Reduction of Engine Exhaust Noise by Throttling in an Exhaust Manifold 39
1/3 is scarecely different from that for the manifold without throttl ing in the speed range higher than around 3700 rpm, where the pulsation effect is influential judging from the value of q. These trends are the same as those which were observed on the experimental work by the authors(l) on a sigle-cylinder two-stroke cycle engine with an exhaust pipe having a resonance pipe related to the pulsation effect. Based on the analysis of the previous work, it is deduced that the trend observed at the higher speeds is owing to the resonance in the manifold for the pressure oscillation of the pulsation effect. A considerable decrease in delivery ratio for a throttle ratio of 1/6 is seen in the figure within the speed range by this test. This may be attributed to the fact that the mean back pressure increases with a decreasing throttle area. Thus, the lower limit of the throttle ratio may be around 1/3, as far as delivery ratio is concerned. Experimental results for different lengths of the manifold with a junction angle of 180 deg. are shown in Figs.5 and 6. The former is the data for the speed range where the inertia effect is predominant, and the latter is the result for the speed range where delivery ratio is chiefly influenced by the pulsation effect. It is seen from Fig.5
0.8·r------:Jr------, Q9'~~------'---'---=:--::----' Exhaust system A Throttle system A Throttle 01= 1800 ratio ()(=1800 ratio {=588mm 0 III o (~1400mm 0 III o o 1/3 Q8:I----+~,...__----!----t=0~__:_I_':_/_':'3___1 0.71--1-----+-"''---'"''<- :s i! 6. 1/6 ~ 6 1/6 > cu' > ~ 0.6l--+----e1-1-----+---:~J---_i o
a*--t_-~--+----t_---___i 0!FJ--f--4'..L---+---t--+-----\:------i
O.4I---+-----I-----''fr---t------i Q51--+-----+-----r-'Yr-----j
0.3 2000 3000 4000 5000 0.4 Engine IS peed rpm I I Engine speeD rpm I ! I 3P 2.5 2.0 1.75 q L25 1.0 q 075 I I I , 1.25 1.0 m 0.75 ! 075 0.5 m
Fig.5 Delivery ratio versus engine Fig.6 Delivery ratio versus engine speed, q, and m in the range speed, q, and m in the range where the inertia effect is where the pulsation effect predominant is predominant 40 Kazuie NISHIWAKI and Yuzuru SHIMAMOTO
that delivery ratio is greatly 09'~------:=-----:-' influenced by throttling. On the Exhaust system A Throttle ()I = 900 ratio other hand, it appears in Fig.6 '- = 1446 mm 0 III o ~3 that a difference between delivery '~08I---DI-C>-----I--- o ~ /:, Y6 ratios at throttle ratios of 1/1 and 1/3 is comparatively small. As ;.,... ~o.71----+------:I---~*----l stated above, this trend is due to a; the resonance in the manifold which o causes little difference to delivery 061--+----+-..B.;,------1--'=t-\-----1 ratio for the varying throttle ratio. Fig.7 shows the data for a junction 051--1----+---+:.__-----1 angle of 90 deg. and a manifold length of 1440 mm. Comparing the results indicated in Figs.6 and 7, 20'00 3000 4000 5000 little significance is seen concern , , Engine speed, rpm 0.75 ing the junction angle. 1.25 , 1.0 ,q 0.75 05m 3.2 Brake Mean Effective Pressure A series of tests were carried Fig.7 Delivery ratio versus out for the manifolds with a junc engine speed, q, Gnd m for tion angle of 180 deg. and throttle a junction angle of 90 deg. ratios of 1/1, 1/3, and 1/6. Figs. 8,9, and 10 show the brake mean
6.------.-----, .. 6r------~ .. Exhaust system A Thr; ttle ra1io E Exhaust system A Throttle ratio 0 0 fX , ~ «=180 , l.=700mm 0 1/ I a, = 180 l =588 0 1/ 1 ~ 0 1/3 ~ D 1/3 f::,. 1/6 6 1/6 CIl 5 ~1_--~_+-7~.£::,.De:.~!t_":::.....:'-'--'~_t ~!l ... :::l ::3 rn VI rn VI ~ ~ a. a. .~ 4\-+-----+-----+-~~-___j ~ 4 1-l-----+-----+-~.c::.'----___1 "-5 "0 2 2 'Q) CIl c 6 :63 ~~~--~.l-----..L..----.....J ~ 3 2000 3000 4000 5000 20~0 Engi~e 4!P~ed , rpm i I 3000 r qR,0o I Enlline speed e 3.0 2.5 2.0 q 1.75 2.5 2.0 q 1.5, I I CD L'!l Lb 0.6 0.75 1.0 0.8 m m Fig.8 Brake mean effective Fig.9 Brake mean effective pressure versus engine speed, pressure versus en~ine q, and m in the range where speed, q, and m in the range both the inertia and~ulsation where the inertia effect is effects are influential predominant Reduction of Engine Exhaust Noise by Throttling in an Exhaust Manifold 41
effective pressure related to engine 6.------. ~ speed, q, and m. Here the brake Exhaust system A TtYottle ratio l3 (I( = 1800 l= 1400 mm 0 1/ I mean effective pressure is seen to ~ 0 1/3 "'" £::, 1/6 decrease with decreasing throttle ~5 ::l ratio in the range where the inertia U> ~ effect is predominant. The amount a of the decrease, however, is not so ~ 4t-t-----!------="",....---:='"'r--""-i large compared with the decrease in n ~.. delivery ratio as previously shown. w ~ Presumably this is owing to a short E3 2000 circuited fresh charge during the 1.25 scavenging period. One interesting 0.75 trend is observable from these figures, namely, that brake mean Fig. 10 Brake mean effective effective pressure for a throttle pressure versus engine ratio of 1/3 increases slightly speed, q, and m in the comparing with that for a throttle range where the pulsation ratio of 1/1 in part of the range effect is predominant where the pulsation effect is influential. The tendency may not be well explained by the resonance in the manifold. To clarify this point, the indicator diagramp Were taken just behind the exhaust port. The results are shown in Fig.ll. The high pressure pulse observed at 3750 rpm just before the time of exhaust port closing is the reflected pressure pulse from the other cylinder whose exhaust port is closed. This pUlse may be effective to prevent fresh charge from flowing out after scavenging port closing. It will be noted on this figure that an amplitude of the reflected
/\ .4500rpm~h w~'\;V
~j I I ! _.L-.l_--L _._-- EO SOC EC TOC EO SOC EC TOC ~ ~lh
_-1 II EO SOC EC TOC
Fig.ll Indicator diagram just behind the exhaust port 42 Kazuie NISHIWAKI and Yuzuru SHIMAMOTO
pulse is greater at a 1/3 throttle '"E Exhaust system B ratio than that at a 1/1 throttle ~ ~ 0'= 1800 l =700mm ratio. As will be expected, higher .. reflected pulse results in improved .,~ 5t-+---=---+?'~P""on:J-.<.>....-IL.JoL..---l charging, and therefore in increased ..~ brake mean effective pressure. ~ It is seen from the figure that ~ 4t-t--h--I::I""--+-----+-"'*-+.,....---! :t.. reflected pulse at 2250 rpm reaches <: o too early and that at 4500 rpm ..E .. 3...':'::!:=:----=!=----:-::!'==:---"'::":~ comes back too late to improve -a 2000 ~ 4000 5000 charging. Thus the favourable eli Engine speed rpm effect of throttling on brake mean Fig. 12 Brake mean effective effective pressure may be mainly pressure versus engine due to the reflected pulse which speed for the exhaust increases in amplitude with a system with a silencer
decreasing throttle ratio, and !g 15 r-,------,------,------, small influence on delivery ratio Exhaust system B Q) > Qnaccount of resonance in the ~ manifold. As a considerable decrease ~ ::;] in brake: mean effective pressure is ~ IOI---',...... L::~'--*=----_+....:....,,____--_1 Q)... " seen at a throttle ratio of 1/6, Q. '0 Q) the lower limit of the throttle > -- in front of .... throttle ratio may be about 1/3 also in o --- beI*1d throttle Q) a:: 5 3X)() respect to brake mean effective 3000 4000 5000 pressure. Engine speed rpm Fig. 13 Relative pressure levels in front of and behind 4. Exhaust Noise the throttle
For the study of exhaust noise emitted from the exhaust system with the manifold under consideration, the exhaust system B .shown in Fig.l was attached behind the throttle plate. The measuring and rec~rding·equipment consists of a condenser micIJophone{ Bion DC-ll ), which is used in conjunction with a precision ,sound level meter (Bion NA-51 ), a spectrum analyzer ( SA 35 ), and a level·:recorder ( Rion LR-OIE ). For spectrum analyze, the o~tput o~the ~recision sound'.lev~l meter was recorded cin a data recorder ( TEAC R-~IO ), and supplied to the spectrum analyzer thereafter. As noise leve is affected by cylinder pressure at the beginning of Reduction of Engine Exhaust Noise by Throttling in an Exhaust Manifold 43
blow-down,brake mean effective pressure, which is approximately proportional to this pressure, ~"T77/~rl was obtained for the exhaust system B. The result is shown in Fig. 12. Comparison of Main brake mean effective pressures for the exhaust muffler system B and A reveals that little difference arises between them, in particular at throttle Height ratios of 1/3 and 1/6. One explanation f.or 740 this might be the fact that reflected pressure wave from the exhaust system following the throttle is damped in large volume o~ the pre Fig. 14 Location of muffler and through the throttle, and thus gives the microphone little influence to the pressure variation at the exhaust port, which was experimentally observed oni.indicator diagrams. From the above discussion, the exhaust system B and A in conjunction with the manifold are considered to be under approximately the same condition in respect to brake mean effective pressure. In Fig. 13, relative pressure levels measured just in front ·of and behind the throttle are shown for various throttle ratios. As would be expected, the pressure level is seen to decrease with a decreasing throttle ratio. Noise level and spectrum were measured outside of the test room as shown in Fig. 14. No significant influence of reflected sound was observable by the preliminary test. Figs. 15 and 16 show noise level in dB(C) and dB(A), respectively, obtained for the manifold with junction angle of 180 deg. and a length of 700 mm. It is seen from
100 I 100 Exhaust system B Exhaust sysfem B Th:ottle q 1)'=180·.1=700mm ci= 180 ratio I =700mm lD LI 1/1 "0 en 90 90 ~ Q; 1/3 >
7OL-":C2-:l000"."..--...",3.".OOO'="-----:4":Co'=o-=0-~5~O~0:-::O 70,L-:::"20'=f0:::"0=----=3-=0'=0-=0:----4:-:0:':0=-=0=----=5-=0-=00=' Engine speed rpm Engine speed rpm Fig. 15 Noise level in dB(C) versus Fig. 16 Noise level in dB(A) engine speed for various versus engine speed for throttle ratios various throttle ratios 44 Kazuie NISHIWAKI and Yuzuru SHIMAMOTO
these figures that a considerable 100 Exhaust system B I Throttl reduction in noise level is attain~ u "" ed by throttling, in particular at i I~g: ~~mm higher speeds, for instance noise Qi 901---1~-;::-A::l:°=----tL~S~kt--D--l levle at a throttle ratio of 1/3 ~ decreases by 8dB(A) at 3500 rpm as .~ o compared with that at a throttle Z . 80'L-.".."J",..,..---~-:-:----="=----::=-=-=-= ratlo of 1/1. Though a decrease 3000 4000 5000 in noise level at a throttle ratio Engine speed rpm of 1/6 is seen to be greater, a Fig. 17 Noise level in dB(C) for decay of engine output is signifi various junction angles cant as previously shown. In Fig. 17, Doise level in dB(C) at a throttle ratio of 1/3 is shown for various junction angles. It appears in Fig.17 that a difference in noise level for different Junction angles is less than 3 dB. Therefore the result indicates that a junction angle of the manifold has little influence on noise level. Fig. 18 shows the typical noise spectrums for the manifold with a junction angle of 180 deg. and a length of 700 mm for various throttle ratios. It is noticeable in Fig. 18 that a reduction in spectrum level by throttling is more rOOi pronounced at lower frequencies. !ExhaJsf system B 0 90 oL~ 180 I This change in noise spectrum /.. 700 rnrn 80 Throttl~ ~atl~ 1(.1 , results in smaller difference ±h~~-~rJ; !! in noise levelsi.in dB(C) and dB(A) at throttle ratio of 1/3 and 1/6 as shown in Figs. 15 and 16. m ."
5. Conclusion
By the use of the manifold having a throttle at its junction, a considerable reduc tion in noise level was attain~ ed without a significant decay , ; in engine output. The experi Frequency Hz ment suggests that the manifold acts as a resonance pipe for Fig. 18 Typical noise spectrums as pressure oscillation related to influenced by throttling Reduction of Engine Exhaust Noise by Throttling in an Exhaust Manifold 45
the pulsation effect and thus absorbs the variation of flow velocity to avoid excessive back pressure. The lower limit of the throttle ratio was around 1/3. Within this limit, brake mean effective pressure was scarecely influenced by throttling under the condition where q is less than 1.7. A reduction in,noise level attained at a throttle ratio of 1/3 was more pronounced at higher speeds and the maximum reduction obtained was 8 dB(A). 'I'he junction angle of the manifold gave little influence on engine output and noise level within the speed range in the test.
References
(1) K.Nishiwaki and Y.Shimamoto : Preprint of 50th Meeting, Japan Soc. Hech. Engrs., No.720-14(1972), 5. (2) M.Maegawa : Preprint of 54th Lecture Meeting, Japan Soc. Mech. Engrs., (1953) ,25.