$&7,9(12,6(&21752/7(&+1,48( 72,03529((1*,1(()),&,(1&<

Transcription

1 $&7,9(12,6(&21752/7(&+1,48( 72,03529((1*,1(()),&,(1&< )HGHULFR5RVVL Università degli Studi di Perugia Dipartimento di Ingegneria Industriale Via G. Duranti, 1/A Perugia 6200$5,2 Nel presente lavoro si propone un sistema originale per la riduzione del rumore allo scarico degli autoveicoli: marmitta elettronica (ME). Il funzionamento della ME si basa sulla tecnica del controllo attivo del rumore: un segnale della stessa ampiezza ma di fase opposta al rumore di scarico interferisce distruttivamente con esso producendo, come risultato, la riduzione del rumore globale. Un vantaggio aggiuntivo dell impiego della ME, rispetto ai sistemi tradizionali di scarico, è la diminuzione delle perdite di carico che determina un incremento delle prestazioni del motore. Mediante una campagna sperimentale condotta con la ME applicata ad una Fiat Marea 1900 JTD sono stati riscontrati: 1) riduzione del rumore di circa 3 db; la riduzione è dovuta alle componenti del rumore con frequenza inferiore a 200 Hz; 2) incremento del rendimento del motore di circa il 2 %.,1752'8&7,21 Exhaust noise abatement is usually obtained by passive mufflers. Traditional mufflers noise reduction is accompanied with transmission losses due to the exhaust duct (Alfredson HWDO, 1971) (Munjal, 1987); a great pressure drop between the inlet and outlet sections of the exhaust pipe rises up. Thus, engine efficiency decreases (Adams, 1980). An active control noise muffler (electronic muffler - EM) is proposed: engine noise and canceling signal destructively interfere at the end of exhaust pipe. The active control system is more efficient than the passive systems for low frequency noise components. Thus, the electronic muffler may substitute the exhaust duct final stage (low frequency stage LFS). The electronic muffler introduces very low transmission losses since gases flow through an empty low roughness pipe. An electronic muffler prototype for variable spin internal combustion engine exhaust noise reduction has been designed and realized at Perugia University Acoustic Laboratory. The prototype has been installed on a Fiat Marea SW 1900 JTD car, equipped with a Diesel engine: low frequency noise reduction, engine efficiency increase and fuel consumption reduction has been proved by different experimental methods. 7+((/(&7521,&08))/(5 A four-stroke engine active control muffler is proposed. Noise reduction is obtained thank to the interference between primary and canceling signals (Cotana HW DO, 1997). Engine noise signal propagates inside the exhaust pipe (Sujbert, 1999) (Lan HWDO, 1999). Canceling noise is

2 generated by an acoustical source. The canceling signal is generated by a loudspeaker which induces vibration on a pipe the final section of which is external (coaxial) to the exhaust gas one. The first part of canceling noise duct is characterized by a parallelepiped shape with triangular base. Primary and canceling noises destructively interfere at the electronic muffler outlet (see fig. 1). The acoustical source is insulated by a 3mm thickness steel box, which is characterized by a truncated pyramid shape. The acoustical box is internally filled by absorbing material. In order to achieve the best acoustical match between canceling noise source and electronic muffler, final section, the following peculiarities have been adopted: Bass reflex: two 15 millimeters diameter holes connect acoustic box to canceling noise duct; thus, loudspeaker low frequency efficiency increases. Impedance adaptation stage: canceling noise pipe section have been gradually varied in order to obtain a circular shape final section which matches external acoustical impedance. The acoustical emitter is fed by a DSP control unit. The control unit implements a Filtered-x LMS control algorithm. Two small ducts (40 millimeters length, 15 millimeters diameter) have been realized upstream and downstream the electronic muffler. Two microphones are put into the two ducts. The upstream microphone picks up the reference signal; the reference signal is synchronized with exhaust noise. The downstream microphone picks up the error signal. Control unit algorithm elaborates reference and error signals to generate canceling signal. The electronic muffler scheme is shown in fig. 1. In fig. 2 two pictures of electronic muffler are shown. Exhaust Gas Bass Reflex Impedence Adapter Stage A Final Exhaust Gas Section B B BB View Insulation Cover Loudspeaker A AA View Upstream Mic Duct Downstream Mic Duct )LJHOHFWURQLFPXIIOHUVFKHPH

3 a) )LJ D ORXGVSHDNHUV RI HOHFWURQLF PXIIOHU DQG FRQVWUXFWLRQ SHFXOLDULWLHV E HOHFWURQLF PXIIOHU SURWRW\SHLQVWDOOHGRQ)LDW0DUHD6:-7' (/(&7521,&08))/(5$&2867,&$/3(5)250$1&(6 $FRXVWLFDOPHDVXUHPHQWFRQGLWLRQV EM performances and traditional passive muffler ones have been compared: the exhaust gas final stage of a Fiat Marea 1900 JTD was employed. Noise level measurements have been carried out for different engine rpm regimes under the following conditions: A- final passive stage is substituted by the EM which is ON; B- final passive stage is substituted with EM which is OFF; C- exhaust duct is equipped with final passive muffler. b) N Control Loudspeaker 50 cm 1 m 2 m F C A D I 45 G L Reference Microphone Error Microphone B E 45 H 5 m Control Unit )LJ±PHDVXUHPHQWSRLQWVSRVLWLRQV M

4 Instrumentation is constituted by the model Investigator 2260 phonometer, by Bruel & Kjaer, equipped with the model 4189 microphone, by Bruel & Kjaer. Measurement points, shown in fig. 3, lies on a plane which is 35 cm height from ground. Measurements have been carried out for a 10 seconds time interval; equivalent noise level (L eq ) and A-weighted equivalent noise level (L Aeq ) have been measured. 0HDVXUHPHQWVUHVXOWV The highest low frequency (lower than 200 Hz) noise emissions occur when engine rotation regime is below 3000 rpm (Burrahm, 1995); noise level comparisons are reported just for two of the most important engine regimes in terms of low frequency noise emissions: 2000 and 3000 rpm. L eq and L Aeq values and the main 1/3 octave band noise component level (MNC) are reported in tab. 1 for A, B and C conditions. Noise level comparisons among the three different conditions are reported in Tab. 2. 7DE±(0DQG/)6PXIIOHU/ DQG/ YDOXHVIRUDQGUSPHQJLQHUHJLPHV Measurement points A B C D E F G H I L M N Conditions Engine rpm Equivalent noise level L eq [db] A-weighted noise level L Aeq [dba] MNC (80 Hz for 2000 rpm, 100 Hz for 3000 rpm) [db] A B C A B C A B C

5 7DE±QRLVHOHYHOFRPSDULVRQ Measurement Points A B C D E F G H I L M N Conditions Engine A - B A C C B rpm L eq L Aeq MNC L eq L Aeq MNC L eq L Aeq MNC Global noise average reductions are 9.5 db for 2000 rpm and 9.3 db for 3000 rpm when the active control system is ON with respect to the control system OFF. The A-weighted noise reductions are 4.2 dba for 2000 rpm and 4.7 db for 3000 rpm, while the average MNC reductions are 16.1 db for 2000 rpm and 17 db for 3000 rpm. The L eq average global noise reductions are 3.1 db for 2000 rpm and 3.5 db for 3000 rpm when the active control system is ON with respect to LFS. The A-weighted noise reductions are 1.5 db for 2000 rpm and 1.8 db for 3000 rpm, while the average MNC reductions are 3.7 db for 2000 rpm and 4.1 db for 3000 rpm. The maximum reduction is measured in L,M and N points and the minimum one in B point; it is expected that as the muffler-evaluation point distance grows, noise reduction grows too; in fact, noise reduction tends to reach a limit value which is the MNC reduction (see tab. 2) because low frequency components better propagate than the other ones. The A- weighted reductions are much lower than linear ones because control system performs the highest noise reductions at low frequencies.

6 (/(&7521,&08))/(5)/8,'2'<1$0,&$/3(5)250$1&(6 A measurement campaign has been led in order to verify if EM introduces also an engine efficiency increase or a fuel consumption decrease. Three measurements methods have been carried out: - pressure drop measurement; - engine test bench measurement; - car traveling fuel consumption test. The EM tests have been carried out with control system OFF; it has been verified that control system status does not interfere on measurement results. 3UHVVXUHGURSPHDVXUHPHQWV The pressure drop measurement method has been applied to evaluate transmission losses due to LFS and to EM. The knowledge of transmission losses allows to calculate the engine efficiency increase. The present test was carried out at Perugia University Fluidodynamic Laboratories. The following instrumentation has been configured as shown in fig. 4: - ELVE KA 451 industrial fan, built by F.lli Ferrari Spa. - Aluminium extensible tube (1 m extensible to 3 m length, 120 mm diameter). - AS85 anemometer, built by Salmoiraghi; it measures the air instantaneous velocity. - Inox steel duct (700 mm length, 60 mm diameter). - Plastic rigid duct (220 mm length, 75 mm diameter); - DPA-100P differential manometer, built by Kytola. Anemometer Differential Manometer Fan Aluminium Extensible Inox Steel Duct Section Change Plastic Rigid Connector Duct )LJ±SUHVVXUHGURSPHDVXUHPHQWVFKHPH Muffler EM and LFS pressure drop versus air flow rate are shown in fig. 5. Volumetric air flow rate G s (m 3 /s) may be related to engine angular velocity (USP) by mean the following equation (Caputo, 1987): USP = Power losses : are calculated by G s and pressure drop: * : (1) = * 3 (2) Transmission losses 7/ (%) are calculated as follows: : 7/ = : 100 (3)

7 # # The comparison between EM and LFS power losses versus the engine angular velocity is shown in fig. 6. The comparison between EM and LFS transmission is shown in fig. 7. LFS transmission losses are greater than the EM ones; the difference rises up when engine angular velocity increases. Fiat Marea SW 1900 JTD engine maximum torque corresponds to rpm=4000. Flow rates * value is: 3 P * = (4) V Thus LFS and EM pressure drops are: 3 3 = = [ 3D] [ 3D] (5) According to eq. (2), power losses values are: : : $ % && $ % && '( )!" = 250 = 112 [.: ] [.: ] (6) By eq. (3), transmission losses values are: 7/ 7/,-. *+ = 324 = 145 [ ] [ ] (7) EM equipped engine performances are approximately 2% better than the LFS equipped engine ones. Pressure drop (Pa) LFS EM Flow Rate (m 3 /h) )LJ±(0DQG/)6SUHVVXUHGURSLQUHVSHFWWRDLUIORZUDWH

8 12 10 Power losses (KW) EM LFS 2 0 Engine angular velocity (rpm) )LJ±(0DQG/)6SRZHUORVVHVZLWKUHVSHFWWRHQJLQHDQJXODUYHORFLW\ LFS EM TL (%) Engine angular velocity (rpm) )LJ±(0DQG/)6WUDQVPLVVLRQORVVHVZLWKUHVSHFWWRHQJLQHDQJXODUYHORFLW\ (QJLQHWHVWEHQFK Fuel consumption measurements have been carried out by mean of an engine test bench. EM and LFS inlets have been connected to the gas exhaust system. Fig. 8 shows the electronic muffler connected to the engine test bench. Measurements have been led for two engine angular velocities: 2500 rpm and 4000 rpm (maximum engine torque). EM and LFS measured

9 fuel consumptions are reported in Table 3. Measurement result is the average value of a 15 single measures set; each single measure has been carried out for a different value of environmental atmospheric pressure (see tab. 3): EM equipped engine fuel consumption is about 2% lower than the LFS equipped engine one. 7DE(QJLQHWHVWEHQFKPHDVXUHPHQWVUHVXOWV Measures Atmosphere pressure [hpa] Average 995 Engine Angular Velocity [rpm] LFS Fuel Consumption [g/kwh] EM Fuel Consumption [g/kwh] EM Fuel Consumption reduction [%]

10 Electronic Muffler )LJ±HOHFWURQLFPXIIOHUFRQQHFWHGWRWKHHQJLQHWHVWEHQFK &DUWUDYHOOLQJWHVW Fuel consumption tests have been carried out in Umbria Autodrome (Magione) under the following conditions: - wind absence, - dry road, - external air temperature equal to 7 C, - vehicle windows closed, - air conditioning system off, - two people on board. 7DE±FDUWUDYHOOLQJWHVWPHDVXUHPHQWVUHVXOWV Fuel Consumption [lit.] Fuel consumption Measures reduction LFS EM [%] Average Evaluation of engine fuel consumption was made by repeating the following procedure 10 times. Thus engine consumption is the average value of the 10 single consumption measures. Fiat Marea engine has been equipped with electronic muffler. The vehicle has been positioned on the track. Engine was warmed up. A graduate tank has been connected to the vehicle embedded tank. The vehicle tank has been filled until the graduate tank level was 5 litres. Test

11 has been executed with standing start. Ten Autodrome laps (corresponding to 25.2 Km) have been covered. The vehicle speed has been maintained constant to 50 Km/h. The engine has been maintained in third gear, corresponding to 2500 rpm angular velocity. When ten laps were covered, the engine has been turned off. The fuel consumption amount has been measured. Therefore Fiat Marea engine has been equipped with low frequency stage muffler. The previous test has been executed in the EM same conditions. In tab. 4 are reported the fuel consumption amount with EM and LFS muffler for each consumption measurement. The EM fuel consumption reduction is approximately 2 %. &21&/86,216 An active control noise muffler (EM) has been proposed. EM improves noise reduction and engine performances by substituting the vehicle exhaust duct low frequency stage (LFS). An acoustical measurement campaign has been led to evaluate the EM noise abatement with respect to the LFS one. The L eq average global reduction is 3.1 db for 2000 rpm and 3.5 db for 3000 rpm. EM and LFS transmission losses and fuel consumptions have been evaluated through some measurement methods: 1) Pressure drops measurements; 2) Fuel consumption measurement by an engine test bench; 3) Fuel consumption measurement on vehicle travelling conditions. Pressure drop measurement results show that LFS transmission losses are greater than the EM ones; the difference between EM and LFS performances rises up when engine angular velocity increases; engine average performances improvement is about 2%. The same average value has been found by mean of engine test bench. Fuel consumption tests confirmed pressure drop results; EM fuel consumption reduction is approximately 2%. Noise measurements results and three different methods for engine efficiency evaluation have verified that EM improves passive muffler system low frequency abatement and meanwhile increase engine efficiency. /,672)6<0%2/6 6\PERO 8QLW0HDQLQJ 'HVFULSWLRQ EM Acronym Electronic Muffler LFS Acronym Low Frequency Stage MNC Acronym Main Noise Component rpm Acronym Revolutions per minute L eq db Equivalent noise level L Aeq dba A-weighted noise level G S m 3 /s Volumetric air flow rate P 1 Pa Cylinder internal pressure just before pumping cycle V 1 m 3 Cylinder internal volume just before pumping cycle T 1 K Cylinder internal temperature just before pumping cycle P 2 Pa Average gas pressure upstream mufflers set T 2 K Average gas temperature upstream mufflers set W loss W Power losses P Pa Pressure drop TL Adimensional Transmission losses

12 %LEOLRJUDSK\ Alfredson R.J. et al. (1971) : The Radiation of Sound from the Engine Exhaust, J. Sound and Vibrations, pp Munjal M.L. (1987): Acoustics of Ducts and Mufflers, John Wiley & Sons, New York, Adams T.G. (1980): Effect of Exhaust System Design on Engine Performance, SAE Paper Cotana F., Felli M., Rossi F. (1997): An Active Control Device for Electric Power Plant, Inter-Noise 97, Aug. 1997, Budapest, Hungary. Sujbert L. (1999): A new filtered LMS algorithm for active noise control, ACTIVE 99, Dec. 99, Fort Lauderdale, Florida, USA. Lan H., Zhang M., Ser W. (1999): A noise reduced car communication system, ACTIVE 99, Dec. 99, Fort Lauderdale, Florida, USA. Burrahm R. (1995): Engine and component evaluation, SwRI Publications, Sep. 95, San Antonio, Texas, USA. Caputo C. (1987): Motori termici volumetrici, UTET, Turin, Italy, Tabaczinsky R.J. (1980): Effects of Inlet and Exhaust System Design on Engine Performance, SAE Paper $%675$&7 An Active Noise Control system was employed to increase the performances of a diesel engine which equips a commercial car (FIAT Marea 1900 JTD). The system is an electroacoustical device (electronic muffler EM) which substitutes the exhaust duct final stage (low frequency stage - LFS). EM is constituted by an inner roughless tube wrapped by an external acoustical box which contains a loudspeaker. Inner duct and external box show a coaxial shape at the outlet section. The loudspeaker is driven by a control unit which generates the canceling noise. A destructive interference occurs at the EM final section; thus a strong reduction of global noise is achieved. A comparison between EM and LFS showed what follows: í EM noise reduction is higher than LFS thank to active system high performances at low frequency range; í EM pressure drop is much lower than LFS because gas are exhausted through the roughless tube instead through the labyrinth shape LFS duct. Previous statements were proved by means of three different methods: 1) Engine test bench; 2) Fuel consumption measurement on traveling condition, 3) Laboratory pressure drop measurement facility. Each method showed that an EM equipped engine fuel consumption is 2% (average value) lower than a LFS equipped engine: lower pressure drop produces a lower engine pumping work and thus a consumption reduction. EM may be conveniently employed to meanwhile reduce exhaust gas noise and fuel consumption.