5th Internatonal Conference on Measurement, Instrumentaton and Automaton (ICMIA 06) Methods for Measurng and Compensatng Ball Screw Error on Mult-mode Industral CT Scannng Platform Yuje Zhang, a, Shangfeng Pan, b and Jan Deng, c Department of Mechancal Engneerng, Tsnghua Unversty, Bejng 00084, Chna a zhang-yj4@mals.tsnghua.edu.cn, bpansf@mal.tsnghua.edu.cn, c deng-j6@mals.tsnghua.edu.cn Keywords: Postonng accuracy, Ball screw, Lead error, Backlash, Laser nterferometer Abstract. Precson of ball screw had a great mpact on the postonng accuracy of NC postonng platform. The lead error and backlash of ball screw, whch could be accurately measured by laser nterferometer, were two key factors affectng the postonng accuracy of NC postonng platform. By settng the compensaton data n the numercal control system, sgnfcant mprovement of the postonng accuracy of NC postonng platform could be acheved. A method for measurng lead error and backlash of ball screw by laser nterferometer was ntroduced and another method for calculatng compensaton data basng on the results of measurement was putted forward. Experments were carred out on a mult-mode ndustral CT scannng platform to verfy these two methods. And the postonng accuracy of mult-mode ndustral CT scannng platform was advanced obvously by compensatng ball screw lead error and backlash wth the compensaton data measurng by laser nterferometer. Introducton In the sem-closed loop NC postonng platform, the postonng accuracy depends on the accuracy of ball screw. Screw error could be dvded nto load error, lead error and thermal error whch affect the postonng accuracy of NC postonng platform obvously []. Load error was caused by the external force that appled to the screw, lead error was brought about by manufacturng precson and long-term frcton wear, and thermal error was ncted by the hgh temperature durng operaton. To heavy work load NC postonng platform, such as CNC lathes and CNC mllng machne, thermal error and lead error play the major role n affectng the postonng accuracy, whle to lght work load NC postonng platform, ncludng measurng nstruments and laboratory apparatus, only lead error affected the postonng accuracy sgnfcantly. Meanwhle backlash led sgnfcant effects on the postonng accuracy n postonng platform whch needs to recprocate. Although the backlash could be reduced by mechancal structure, t could not be totally elmnated. A method for measurng lead error and backlash of ball screw by laser nterferometer was ntroduced and another method for calculatng compensaton data was carred out n ths paper. The methods were verfed on a mult-mode ndustral CT platform. A Method for Measurng Postonng Error of Mult-mode Industral CT Scannng Platform The laser beam entered n the spectroscope as shown n Fg., and was splt nto two beams. A laser beam called the reference beam was guded to a mrror mounted on the nterferometer, whle the other laser beam called measurng beam passed through the spectroscope and reached the second mrror. Then, two laser beams were reflected back to the beam nterferometer, where they were re-combned and back nto the laser transmtter. The detector n the laser transmtter montored nterference frnges changes whle the second mrror movng []. Durng the lnear measurement process, a mrror remaned statonary, whle the other mrror moved along the lnear drecton. By montorng the changes of nterference frnges, the precse movng dstance could be calculated. 06. The authors - Publshed by Atlants Press 795
Fg. The measurng prncple of laser nterferometer Mult-mode ndustral CT scannng platform was composed by the three movement axes, called X-axes, Y-axes and Z-axes, as shown n Fg.. X-axes and Y-axes can drve the platform n horzontal drecton, Z-axes can drve the platform n vertcal drecton. Renshaw XL30 laser nterferometer measurng system was appled n the measurng experment. The measurng system conssted of XL laser transmtter, nterferometer, mrror and envronmental compensaton unt. The measurement precson of movng dstance could be less than um and the resoluton of movng dstance could acheve less than 0.0 um [3]. The mrror was fxed on the X table by a magnetc seat, the laser transmtter was mounted on a trpod head and the nterferometer was fxed on the base platform. The mrror, the laser transmtter and the nterferometer should be n the same lne meanwhle the parallel degree requrements of laser nterferometer system should be meted. The expermental measurement ste was showed n Fg. 3. Fg. 3D model of ndustral CT Fg. 3 Measurement system The Measurement Results of Postonng Accuracy wthout Compensaton Data Precson-grade ball screws were selected to make up the mult-mode ndustral CT scannng platform. In the ntal state, there was no compensaton data n the NC system. The X-axs measurement results were descrbed as an example. The travel of X-axs was 687.87 mm and the standard value of lead was 5 mm. ponts were selected to measure the postonng accuracy of X-axs, and each pont measured 5 tmes n both postve and negatve drectons. The average postonng error were shown n Fg. 4. Accordng to ISO 30-04 [4], the postonng accuracy was calculated by below gven equatons from Eq. to Eq. 5, the fnal calculaton of the postonng accuracy was 0.0436 mm. Mean undrectonal postonng devaton at a poston X = n X j n j = () and 796
X n Xj ( n j = = ) Estmator for the undrectonal axs postonng repeatablty at a poston n S = ( Xj X ) n j= (3) and n S = ( Xj X ) n j = (4) B-drectonal postonng accuracy of an axs A= max X S ; X S mn X S ; X S + + (5 ) The value n remnds of the number of measurements, the value remnds of the number of pont. negatve drecton postve drecton mean postonng devaton/um 50 40 30 0 0 0 00 00 300 400 500 600 700 movng dstance/mm Fg. 4 Mean postonng devaton wthout compensaton data Solvng the Compensaton Data Based on Optmzaton Method Consderng the ball screw manufacturng errors and drve characterstcs, the vsual analyss for measurement results showed that the lead should be lnear corrected and the backlash should be set to a fxed value. By settng the compensaton data n the numercal control system, the postonng accuracy of mult-mode ndustral CT scannng platform could be mproved sgnfcantly. In accordance wth the calculaton method of postonng accuracy stpulated n ISO 30-04, the b-drectonal postonng accuracy of X-axs was set as the objectve functon and exhaustve method was used n a certan range to solve optmum compensaton data. The solvng process was shown n Fg. 5, and the relatons of compensaton data and postonng accuracy was shown n Fg. 6. The postonng error was mnmal when the lead compensaton value was set to -0.005 µ m and the backlash compensaton value was set to 0.007 µ m. Theoretcally, the postonng accuracy of X-axs could be less than 0.0079 mm. 797
Fg. 5 The solvng process of compensaton data Fg. 6 The relatons of compensaton data and postonng error The Measurement Result of Postonng Accuracy wth Compensaton Data The compensaton data mentoned above was set to numercal control system. Then, the postonng accuracy of X-axs was measured usng a laser nterferometer. ponts were selected to measure the postonng accuracy of X-axs, and each pont measured 5 tmes n both postve and negatve drectons. The average postonng error wth compensaton data were shown n Fg. 7. Accordng to the calculaton method of postonng accuracy stpulated n ISO 30-04, the postonng accuracy of X-axs was 0.04 mm. 798
negatve drecton postve drecton mean postonng devaton/um 3 0 - Conclusons - 0 00 00 300 400 500 600 700 movng dstance/mm Fg. 7 Mean postonng devaton wth compensaton data In summary, the postonng accuracy of mult-mode ndustral CT scannng platform was advanced obvously by compensatng screw lead and backlash wth the compensaton data measurng by laser nterferometer. The screw lead and backlash would change because of screw abrason n use, so the screw lead and backlash should be measured and compensated regularly. Due to the affecton of temperature and other envronmental factors, f much hgher postonng accuracy was requred, only usng closed loop control system could meet the requrement. Acknowledgements Ths work was fnancally supported by the natonal key foundaton for explorng scentfc nstrument of Chna (04YQ40445). References [] T. Satom and S. Toba: Journal of Japan Socety for Desgn Engneerng Vol. 3-7 (997), p. 3 [] K. Zhang, Z.X. Cu, L.Y. Ga, P. Ge and D.G. Ca: Appled Mechancs & Materals Vol. 30-34 (0), p. 36 [3] Informaton on http://www.renshaw.com [4] ISO 30-: 04, Test Code for Machne Tools - Part : Determnaton of Accuracy and Repeatablty of Postonng of Numercally Controlled Axes, an Internatonal Organzaton for Standardzaton (04) 799