Mo­tors & drives – test­ing is­sues

DEMM Engineering & Manufacturing - - CONTENTS - by Amin Al­masi Abridged – read the full pa­per on the Demm web­site: http://bit.ly/1tMwkIv

The de­sign of a large elec­tri­cal mo­tor is an op­ti­miza­tion in­volv­ing var­i­ous en­gi­neer­ing dis­ci­plines. The main fo­cus is the elec­tri­cal de­sign of the ma­chine, which has a fun­da­men­tal im­pact on the func­tion­al­ity and the per­for­mance.

With in­creases in the power, size, and speed of the elec­tric mo­tors, the me­chan­i­cal and tor­sional de­signs be­come more im­por­tant for the op­er­a­tion to re­main within the tight­ened noise, re­li­a­bil­ity and vi­bra­tion lim­its.

The usual tech­nol­ogy for large elec­tric mo­tor driv­ers is to use an LCI con­verter as­so­ci­ated with a syn­chro­nous two-pole elec­tric mo­tor.

How­ever, VSI (volt­age source in­verter) tech­nolo­gies are be­com­ing popular. Be­cause of the re­ac­tive power con­sump­tion of its thyris­tor bridge, an LCI con­verter can­not prop­erly power an in­duc­tion mo­tor. A VSI can power both in­duc­tion mo­tors and syn­chro­nous mo­tors.

The LCI tech­nol­ogy gen­er­ates torque pul­sa­tions and a har­monic fil­ter is usu­ally re­quired. LCIs have been used for decades and there have been many suc­cess­ful ref­er­ences.

There is a very low har­monic con­tent when us­ing a VSI sys­tem (no har­monic fil­ter re­quired) with a bet­ter ‘power fac­tor’ (PF). A VSI so­lu­tion could also of­fer a bet­ter cost, but ref­er­ences are limited. The se­lec­tion be­tween LCI and VSI de­pends on an ap­pli­ca­tion.

For large elec­tric mo­tors, the f lex­i­ble ro­tor con­cept is used (the first crit­i­cal speed usu­ally lies be­low the op­er­at­ing speed range). The ro­tor should be dy­nam­i­cally bal­anced. The field bal­anc­ing would not be re­quired (whereas it is of­ten pos­si­ble).

When pass­ing the first crit­i­cal speed, the lo­cal ro­ta­tional cen­tre changes from geo­met­ric to lo­cal mass cen­tre, which means, the lo­cal un­bal­ance in an elas­tic ro­tor varies with the speed.

There­fore, modal sets of un­bal­ance weights should be used to bal­ance each mode in­di­vid­u­ally. As a min­i­mum, “n+2” bal­anc­ing planes (n= the num­ber of modes to bal­ance) are nec­es­sary for bal­anc­ing.

A large elec­tric mo­tor should han­dle prop­erly the ther­mal un­bal­ances. Be­cause of the in­evitable use of var­i­ous ma­te­ri­als with very dif­fer­ent ther­mal ex­pan­sion co­ef­fi­cients, com­bined with non-uni­form tem­per­a­ture dis­tri­bu­tion and large sizes, a sym­met­ri­cal me­chan­i­cal and ther­mally in­sen­si­tive de­sign should be achieved. A small asym­me­try can cause an un­ac­cept­able dy­namic load.

To re­duce the risk of hav­ing non­per­form­ing drive sys­tems shipped, full-load, full-speed per­for­mance tests of en­tire drive sys­tem are manda­tory. Usu­ally fol­low­ing tests should be con­ducted: The test­ing a mo­tor alone. The back-to-back test to ver­ify the elec­tric mo­tor and the VSD per­for­mance. The string test for a com­plete com­pres­sor train sys­tem. The open-cir­cuit and short-cir­cuit

To re­duce the risk of hav­ing non-per­form­ing drive sys­tems shipped, full-load, full-speed per­for­mance tests of en­tire drive sys­tem are manda­tory.

tests could de­ter­mine the con­ven­tional mo­tor losses. The no-load test, con­ducted at the rated speed can give the open cir­cuit curve, which could in­di­cate var­i­ous loss con­tri­bu­tions.

The majority of the losses come from the fric­tion and the windage. Strong cool­ing air f lows pro­duced by the cool­ing fans (an in­ter­nal cool­ing) is as­so­ci­ated with some losses. Typ­i­cally ef­fi­cien­cies in range of 97-99 per cent could be ex­pected.

When at least two sim­i­lar VSD and mo­tor sys­tems are be­ing sup­plied, the VSD-mo­tor back-to-back test can be done (one in the mo­tor­ing mode and another in the gen­er­at­ing mode). It is the­o­ret­i­cally pos­si­ble just to sup­ply the losses and the re­ac­tive power de­mands.

The bear­ing prob­lems, ex­ces­sive vi­bra­tions and oil sys­tems is­sues are re­spon­si­ble for a con­sid­er­able por­tion of failed per­for­mance tests of large elec­tric mo­tors. Dur­ing a back-to-back test, ob­ser­va­tions are:

1) As­sess­ment of the mo­tor ther­mal per­for­mance: Heat-run tests should be per­formed to as­sess the mo­tor full-load ther­mal be­hav­iour at dif­fer­ent op­er­at­ing and emer­gency modes.

2) As­sess­ment of mo­tor vi­bra­tion per­for­mance. 3) Torque rip­ple mea­sure­ment. 4) As­sess­ment of mo­tor torque over­load ca­pa­bil­ity: The torque re­quired for the startup of the train (usu­ally a pres­sur­ized com­pres­sor), which could be 130-145 per­cent of nor­mal torque for around 90-130 seconds.

5) Mo­tor volt­age and cur­rent wave­forms.

In a case study, dur­ing an elec­tric mo­tor test, sus­pi­cious noises and smokes were iden­ti­fied on the elec­tric mo­tor. The first ob­ser­va­tion (after the trip) were: The mo­tor shaft drop by 1.6mm. The max­i­mum tem­per­a­ture on the bear­ing reached 135oC. The sleeve bear­ings were dam­aged. The root cause was the lack of lu­bri­ca­tion oil, be­cause of the main oil pump fail­ure (and also the fail­ure of standby oil pump to start).

Newspapers in English

Newspapers from New Zealand

© PressReader. All rights reserved.