Key-Words: - Power quality, voltage fluctuation, induction motors, electrical machines testing.

Transcription

1 On the Behavior of Induction Motor in Preence of Voltage Aplitude GIOVANNI BUCCI, EDOARDO FIORUCCI, ANTONIO OMETTO, NICOLA ROTONDALE Departent of Electrical and Inforation Engineering Univerity of L Aquila, Italy Loc. Monteluco di Roio, L Aquila ITALY {bucci, fiorucci, oetto, roto}@ing.univaq.it Abtract: The voltage fluctuation of the electrical power upply and their effect on both indutrial and doetic apparatue are today a central topic of dicuion. The ai of thi paper i the evaluation of the effect of the low-frequency voltage aplitude odulation on induction otor, with both nuerical and experiental analyi. For the nuerical analyi different kind of odulation have been conidered, a uggeted by the international coittee; an experiental analyi ha been applied to a three phae, 7. kw induction otor with the adoption of an ad hoc yte for the EMC lowfrequency iunity teting, with the adoption of rectangular aplitude odulation. The nuerical and the experiental reult how that a few percent aplitude odulation can involve a critical behavior in ter of torque. A for the light flicker effect eaureent, a tatitical approach can be adopted for the evaluation of the torque ripple, and a iplified relation can be ued to relate the torque ripple to the eaureent of the intantaneou voltage wavefor, without direct echanical eaureent. Key-Word: - quality, voltage fluctuation, induction otor, electrical achine teting. Introduction The evaluation of Quality in low voltage yte require the eaureent of a wide et of paraeter and indexe, for the characterization of different phenoena; high order haronic, interharonic, voltage fluctuation and high frequency diturbance with lew-rate of oe icroecond or le can be entioned a exaple []. On detail, voltage fluctuation can involve a phyiological irritating phenoenon, due to luinance fluctuation of the lighting (light flicker effect) [2]. The ae voltage fluctuation that generate the flicker effect can produce undeired effect on other apparatue, a otor, tranfor and variable peed drive. In thi paper we conider the effect of the voltage aplitude odulation on three-phae induction otor, tarting fro the reult obtained in the previou reearch activitie about light flicker [3]. Specifically, our goal are: i) the nuerical analyi of the behavior of induction otor in preence of different kind of voltage fluctuation; ii) the ipleentation of a eaureent and teting yte for the experiental characterization of the induction otor perforance in the decribed condition; iii) the analyi of the nuerical and experiental data, coparing the electrical and echanical wavefor to evaluate the effect of the applied voltage odulation; iv) the ynthei and teting of a iplified odel for the tatitical evaluation of the torque ripple tarting fro eaured voltage and current wavefor. 2 The Voltage Fluctuation A wide et of voltage fluctuation can be adopted for the teting of the behavior of induction otor: i) inuoidal aplitude odulation; ii) rectangular aplitude odulation; iii) over ipoed interharonic coponent; iv) real world fluctuaction. In the Quality eaureent, the voltage aplitude odulation (Fig. ) have been firtly conidered, both with rectangular and inuoidal hape: the IEC tandard define the teting procedure for the perforance evaluation of the flicker eter applying both inuoidal and rectangular aplitude odulation, in the frequency range 0.-2 Hz. For thi kind of odulation, the IEC require an aplitude odulation eaureent range of at leat 0 (20 i uggeted). The preence of interhaonic coponent on the ain i a well known phenoenon[4]; the effect of thee coponent in ter of light flicker on fluorecent lap and on the increaing of loe in electrical achine are known []. The real world fluctuation have been recently introduced by [6] for the evaluation of the perforance of the light flicker eter in realitic condition; the propoed aplitude odulation are not periodical and can be obtained a a equence of different path. Fig. how the firt inute of the pattern ; it conit of a equence of rectangular aplitude odulation with variable duty cycle (0-30), rate Hz, with a ret period between each tep; the other inute are a iilar repetition, with increaing rate (Table. I) of the decribed equence.

2 Table I: Rate adopted in the pattern The Fig.2 illutrate a equence of the pattern 2; the aplitude of the odulation and the interediate value are not defined yet. Differently fro the other odulating hape and fro interharonic, it i difficult to analyze thee realitic odulation with a fft-baed approach. All the decribed fluctuation can be adopted for the iulation of the voltage fluctuation when heavy load a wood chipper, rolling ill and arc furnace are turned on, and to evaluate their effect on other electrical load. Fig. Realitic fluctuation: pattern, Firt 30 equence. Fig.2 Realitic fluctuation: pattern 2, inute -2 equence. 3 The Effect of the Voltage Fluctuation on Induction Motor On the ain, the voltage fluctuation uually happen iultaneouly on each of the three phae voltage wavefor. Each faily of electrical apparatue i characterized by different behavior in thee condition; we focued our interet on the induction otor, that in our country repreent the ajority of the low-voltage load. Generally, the rapid variation of the voltage level can be conidered a the caue of tranitory phenoena, with fluctuation of electrical and echanical quantitie, and increaing of loe [7,8].On detail, the experiental application of three phae voltage wavefor affected by aplitude odulation to an induction otor ha hown the following iportant effect: i) torque ripple, ii) angular peed ripple, iii) vibration tranferred to the load and to the upport; iv) acoutic noie. Thee phenoena can reduce the life of echanical coponent a the bearing and the joint. Moreover, in indutrial application, the decribed vibration and torque ripple are not uually related to the probable preence of a odulation in the upply voltage wavefor, and the evaluation of the perforance of induction otor in the decribed condition i yet not conidered and defined in the IEC tandard. 4 A Nuerical Analyi Baed on a Dynaic el The evaluation of the effect of the voltage fluctuation can be perfored both with nuerical and experiental analye; the choice of the nuerical odel hould guarantee the correct evaluation of the tranitory phenoena both on electrical and echanical quantitie. A atheatical dynaic odel of the achine, in ter of pace vector, aking the uual iplifying auption (iron loe neglected, agnetic aturation neglected and o on), can be expreed a: di dir v + k = αi () dt dt Lt di dir kr + = jkr pωri ( αr jpωr ) i (2) r dt dt T () t = 3 pl I i i r (3) 2 where v, i and i r are the pace vector of the upply voltage, of the tator current and of the rotor current referred to the tator repectively, T i the electroagnetic torque, ω r i the echanical rotating peed, the caret ^ denote the coplex conjugate operator and where: α k r = l + L L α r rr = l + L r k = l L + L r = L t = l + L (4) lr + L with: l r rotor leakage inductance referred to the tator; l tator leakage inductance; p pole pair; r r rotor reitance referred to the tator; r tator reitance; agnetizing inductance. L The odel of the whole electroechanical yte i copleted conidering the echanical equation:

3 dωr T + TL = J + ( D r + F ) dt ω () where T L i load torque, J i the oent of inertia of the whole echanical yte and (Dω r + F) i the friction torque. All the adopted paraeter have been obtained with the experiental characterization (Table II) of the induction otor under tet. Table II Electrical and echanical paraeter adopted in the dynaic odel D [N ] 0.00 r [Ω] 0. r r [Ω] 0.6 F [N ] 0.3 l [H] 4 l r [H] 4 J [kg 2 ] l [H] Nuerical reult obtained with the pattern The pattern [6] i a equence of rectangular pule train, with increaing rate (Fig. ); even if the applied ax aplitude odulation i 2, the effect on the torque i high. The reult obtained applying the all pattern are preented in Table III-VI. Table III Pattern input power tatitical indexe Table V Pattern torque tatitical indexe [N] [N] [N] Table VI Pattern peed tatitical indexe [rp] [rp] [rp] Table IV Pattern output power tatitical indexe Nuerical reult obtained with the pattern 2 The pattern 2 [6] i a equence of 4 different odulating hape. In thi cae, an aplitude odulation of 4 generate a torque increaing up to 3. The tatitical indexe obtained applying the inute -2 are reported in Table VII-X. The reult obtained applying the odulation in the inute 3-4 of the pattern 2 are very iilar to reult in the previou table; the effect of the odulation in the inute -6 and inute 7-8 are negligible, due to the voltage low lew rate. aplitude Table VII Patten 2, inute -2 input power tatitical indexe

4 Table VIII Patten 2, inute -2 output power tatitical indexe Table XI Rectangular odulation input power tatitical indexe aplitude Apl aplitude Table IX Patten 2, inute -2 torque tatitical indexe [N] [N] [N] Table X Patten 2, inute -2 peed tatitical indexe Table XII Rectangular odulation output power tatitical indexe aplitude [rp] [rp] [rp] Nuerical reult obtained with inuoidal and rectangular odulation The inuoidal aplitude odulation [2] ee to be the le critical for the induction otor; the reult how a torque increaing up to 7.7 in the wort cae. Thi behavior i due to the typical low lew rate of the inuoidal hape. The ituation i quite different for rectangular odulation [2]; the reult are iilar to the one obtained for the pattern. The tatitical reult are in Table XI-XIV. 0 Apl The Teting Syte and the Experiental Reult The voltage wavefor are upplied by the EMC PARTNER HARCS 000 THREE PHASE SYSTEM. Thi apparatu ha been developed for three phae haronic iunity and eiion teting. Hi power ection i fed by three-phae power line. It filter the input voltage wavefor, and it can add teady-tate haronic coponent, interharonic and voltage fluctuation. It perfor the generation of both inuoidal and rectangular odulated inuoidal voltage wavefor in copliance with IEC tandard, in the range 00-2 Vr, Vr, with a three-phae power axiu value 2 kva. A voltage generator it i characterized by hort-cut ipedance lower than 3 Ω, and it ha a Line Ipedance Stabilization Network type 0.4 Ω + j 0.2 Ω at 0 Hz, according to IEC

5 0 0 Table XIII Rectangular odulation torque tatitical indexe Apl. [N] [N] [N] Table XIV Rectangular odulation peed tatitical indexe Apl. [rp] [rp] [rp] environent [9,0]. After the preliinary nuerical teting, an experiental characterization have been perfored. The ain goal ai of thi econd activity paper i the validation of the nuerical reult; oreover, an intereting propective i to obtain a procedure for the firt evaluation of the torque and peed ripple tarting fro the direct eaureent of three phae voltage and current wavefor in indutrial environent, during the noral operating condition. With thi ai we adopted the following tep: i) choice of the otor load in ter of echanical power; ii) election of the voltage odulation aplitude and frequencie; iii) acquiition of the three phae voltage and current, torque and peed experiental wavefor; iv) nuerical evaluation of the r voltage and current wavefor, in order to iply the viualization of the aplitude odulation; v) nuerical evaluation of the input and output power wavefor; vi) tatitical evaluation of the obtained data. The ain feature of the induction otor under tet are: i) 380 V, 6.7 A tar-connected; ii) 7. kw; iii) 443 rp. The electrical and echanical paraeter of thi otor have been eaured and adopted in the previou nuerical analyi. The tie doain diagra in Fig. 3-8 how the effect of rectangular voltage odulation obtained with: i) load ean power of 4.6 kw; ii) input voltage aplitude odulation frequency of Hz; iii) input voltage aplitude odulation of ; iv) rectangular aplitude odulation. In thi cae the torque and peed variation are repectively in the range N and rp. For the teted otor, the input power wavefor and the torque wavefor are very iilar. A for the analyi of the nuerical reult, a tatitical approach have been ued. It ha been preferred to the frequency doain analyi becaue the voltage aplitude fluctuation in realitic condition are not periodic [6] and they can introduce pectral leakage. In Table XV the ain input power tatitical indexe obtained with different odulation paraeter are reported. The ax eaured input power range reache the 77.6 of the ean value. The tatitical indexe of the output power are hown in Table XVI. In the wort condition, the output power range reache the 66. of the ean value. For the tranduction of voltage wavefor, we adopted LEM LV Hall effect cloed loop tranducer. A Magtrol HD hyterei brake dynaoeter ha been ued; thi apparatu ha a controller unit that i connected to the hot coputer via IEE 488 bu. The torque and peed are available a analog output voltage ignal. On the ae coputer, the NI 602E 6 bit DAQ board perfor the A/D converion of the voltage, current, torque and peed wavefor. For each teting tep, 4000 aple at a apling rate of 000 aple/ have been acquired. Both the eaureent oftware and the brake control oftware have been developed in the NI LabVIEW Fig.3 Voltage wavefor with Hz, rectangular aplitude odulation.

6 Fig.4 Three phae r current wavefor Fig. wavefor. Fig.8 wavefor. The ot intereting effect involve the torque; the torque range i fro 24.6 up to 63.9 of the eaured ean torque (Table XVII). Thi phenoenon can involve an unacceptable increaing of torional tre in the coupling and vibration. It can be een that the peed ripple obtained with the yte under tet are not very ignificant (Table XVIII). It i due to the value of the overall oent of inertia. The proble of undeirable peed ripple would be ore critical with lower power induction otor. Starting fro the previou diagra, it can be een that the input power and torque wavefor are proportional, becaue of the all peed ripple. TABLE XV. Input power tatitical indexe Apl. Fig.6 Total input power wavefor TABLE XVI. Output power tatitical indexe Apl. range Fig.7 Output power wavefor

7 0 TABLE XVII tatitical indexe Apl. [N] range [N] [N] intantaneou voltage wavefor, with a dynaic odel of the otor, it i poible to identify a iplified relation for the torque ripple evaluation, without direct echanical eaureent. 7 Acknowledgeent Thi work wa upported by Italian Minitry for Univerity and Scientific and Technologic Reearch, under the grant Cluter 3, Project 7, at the Departent of Electrical and Inforation Engineering, Univerity of L Aquila. Reference 0 TABLE XVIII. tatitical indexe Apl. [rp] range [rp] [rp] Thee reult confir that input power fluctuation ainly produce torque ripple becaue of the aller peed and torage agnetic energy fluctuation in the yte under tet. 6 Concluion The voltage fluctuation and odulation in low voltage feeder can involve undeirable effect on induction otor. The ot iportant are torque and peed ripple that can generate vibration, noie and echanical tre in the bearing, coupling and upport. Today, thee phenoena are not defined in IEC tandard, and it i difficult to relate the torque ripple to the voltage fluctuation in any indutrial environent. The ai of thi paper i the evaluation of thee effect with both nuerical and experiental analyi of the behavior of an induction otor, with the application of a three phae voltage wavefor affected by different kind of aplitude fluctuation. A tatitical approach ha been propoed and adopted for the analyi of the acquired data. The obtained diagra ugget the coparion between the input power and the torque wavefor. Thi approach can ugget intereting teting application, becaue fro the eaureent of the [] EN 060 Voltage characteritic of electricity upplied by public ditribution yte [2] EN , Flickereter Functional and deign pecification [3] G. Bucci, E. Fiorucci, D. Gallo, C. Landi, Coparion aong Traditional and New Digital Intruent for the Meaureent of the Light Flicker Effect 2003 IEEE Intruentation and Meaureent Technology Conference, Vail, CO,USA May 20-22, [4] Teta, A.; Gallo, D.; Langella, R. Interharonic eaureent in IEC tandard fraework Engineering Society Suer Meeting, 2002 IEEE Volue 2, 2002 Page(): vol.2 [] Eanuel, A.E. Etiating the Effect of Haronic Voltage Fluctuation on the Teperature Rie of Squirrel-Cage Motor Engineering Review, IEEE Volue, Iue 3, March 99 [6] CIGRE 36.0/CIRED WG 2/UIRPQ Joint Working Group CCU2 on Voltage Quality Tet Protocol of IEC Flickereter ue in power yte voltage onitoring, Draft, July [7] C. A. G. Medeiro, J. C. de Oliveira, Effect of Voltage Fluctuation Aociated to Flicker Liit on Equipent Perforance IEEE ICHQP Rio De Janeiro, October, [8] Ran Cao, E. R. Collin, The Effect of Load Type on the Behavior of AC Motor Drive during Voltage Sag. IEEE ICHQP Rio De Janeiro, October, [9] G. Bucci, E. Fiorucci, C. Landi State of the Art and Future Trend of Meaureent technique for Aynchronou Motor Characterization. IEEE SDEMPED 0 3rd IEEE International Sypoiu on Diagnotic for Electric Machine, Electronic and Drive. Gorizia, Septeber -3, 200. [0] G. Bucci, E. Fiorucci, A. Oetto N. Rotondale The Evaluation of the Effect of the voltage aplitude fluctuation on induction otor Proceeding of IEEE tech 200, June St. Peterburg, Ruia.