Quality Assurance and Control for Robotic GMA Welding
Total Page:16
File Type:pdf, Size:1020Kb
StPrSIGOObl)ft LUTMDN/(TMMV-1013)/l-171/(1992 Quality Assurance and Control for Robotic GMA Welding Max Xie Department of Production and Materials Engineering Lund University, Sweden January 1992 Organization Document name LUND UNIVERSITY Doctoral Dissertation Dept. of Production and Materials Date of issue Engineering 2 January 1992 Box 118 CODEN:LUTMDN/(TMMV-1013)/1-171/(1992) S-221 00 Lund, Sweden Authors) Sponsoring organization Max X. Xie Title and subtitle Quality Assurance and Control for Robotic GMA Welding Abstract A quality assurance (QA) model has been developed. This model systematically considers the relevant activities before, during and after the welding operations with respect to quality. Efficient quality assurance requires that the functionality of the present robotic welding systems needs to be increased ak. j that the knowledge of the personnel involved in the design and production needs to be improved. The collaboration between different departments and personnel needs also to be improved. The procedure specification aspects have been studied and a method for the determination of optimal welding parameters is presented with regards to process stability, quality requirements and productivity. A main productivity problem of robotic welding systems for small series production is due to the time spent on the specification of welding procedures. In order to improve the efficiency, expert systems technology has been studied and applied to automatically generate optimal welding procedures. Many industries have shortages of skilled and experienced welding engineers or welders. This implies problems in the selection of optimal welding parameters. An objective method for the assessment of process stability has been developed, based upon the analysis of the electrical signals of welding arcs. Furthermore, a method has been developed to monitor the process stability. It is found that it is possible to identify the causes of the disturbance of process stability and to predict the weld quality characteristics based on the analysis of the electrical signals. Though quality is formed during the welding operation, the diagnosis of the causes of quality disturbances is important for the prevention of quaL y problems of subsequent welds and has been discussed. To assist the operators, expert systems technology is also applied. system so that both quality and productivity aspects of the system can be further improved. Key words Welding, Robotics, Quality Assurance, Quality Control Classification system and/or index terms (if any) Supplementary bibliographical information Un«ua«e English ISSN and kev title ISBN Recipient's notes Number of pages j-i Pric« SEK 300 Security classification Distribution by (name and address) I, the undersigned being (he copyright owner of the abstract of the above-mentioned dissertation, hereby grant to all reference sources permission to publish and disseminate the abstract of the above-mentioned dissertation. Max Xie 2 January 1992 Signature / Date Quality Assurance and Control for Roboiic GMA Welding av Max Xie Tekn Lie Akademisk avhandling so . ör avläggande av teknologie doktorsexamen vd Tekniska fakulteten v i Universitetet i Lund kommer att offentligen försvaras i sal M:I3, Tek ska Högskolan i Lund, torsdagen den 20 febru- ari 1992, kl 10:15. fal i, otsopponent är Dr-Ing Osama Al-Erhayem, In- geni0rh0j?kolen Helsingv, Teknikurn, Danmark. Quality Assurance and Control for Robotic GMA Welding Max X. Xie Department of Production and Materials Engineering Lund University, Sweden January 1992 Department of Production and Materials Engineering Lund University CODEN: LUTMDN/(TMMV-1013)/l-171/(1902) ©1992 by Max X. Xie. All rights reserved Published 1992 Printed in Sweden KF-Sigma Lund To my wife Sofi and son Henry Foreword This thesis is a continuing work of a licentiate's thesis [90], with the main objectives of increasing the understanding of quality assurance problems for robotic GMA welding systems and to develop methods for quality control. This work has been carried out at the Department of Production and Materials Engineering, Lund University, and has been financially sponsored by the National Swedish Board for Technical Development within the national program, under the section for adaptive manufacturing systems. I wish to express my greatest gratitude to my supervisor Professor Gunnar Bolmsjö for his invaluable support and helpful discussions. I am very grateful for Professor Jan-Eric Ståhl for his supervision during the initial phase of this study. I also want to thank all the staff at the department who have helped me during the course of this study. Max X Xie Lund January 1992 1 U Abstract A main objective of any manufacturing activity is to efficiently manufacture prod- ucts that satisfy the customer's needs. In other words, the manufactured products must have the desired quality. Undeniably, quality has become a major concern of many industries and the importance of quality as a competitive strength is increasing world wide. This thesis contributes to increasing the understanding of quality assurance prob- lems for robotic GMA welding and to the development of methods and techniques for quality control. A quality assurance (QA) model has been developed. This model systematically considers the relevant activities before, during and after the welding operations with respect to quality. Efficient quality assurance requires that the functionality of the present robotic welding systems needs to be increased and that the knowledge of the personnel involved in the design and production needs to be improved. The collaboration between different departments and personnel needs also to be improved. Essential knowledge for QA is information about the causes of various weld de- fects. Quality disturbances can not be eliminated without the elimination of the quality disturbing sources. The causes of various weld defects are analysed, which forms the fundamental basis for the making of proper measures to prevent and to eliminate the sources of quality disturbances and for the diagnosis of the causes of the disturbances in production. A controlled welding process requires a satisfactory welding procedure. The pro- cedure specification aspects have thus been studied and a method for the de- termination of optimal welding parameters is presented with regards to process stability, quality requirements and productivity. A main productivity problem of robotic welding systems for small series production is due to the time spent on the specification of welding procedures. In order to improve the efficiency, expert sys- tems technology has been studied and applied to automatically generate optimal welding procedures. Many industries have shortages of skilled and experienced welding engineers or welders. This implies problems in the selection of optimal welding parameters. An objective method for the assessment of process stability has been developed, based upon the analysis of the electrical signals of welding arcs. Furthermore, a method has been developed to monitor the process stability. It is found that it is possible to identify the causes of the disturbance of process stability and to predict the weld quality characteristics based on the analysis of the electrical signals. Though quality is formed during the welding operation, the diagnosis of the causes of quality disturbances is important for the prevention of quality problems of subsequent welds and has been discussed. To assist the operators, expert systems technology is also applied. Further work should be directed to the integration of various QA functions in the robotic arc welding system so that both quality and productivity aspects of the system can be further improved. List of Symbols minimum background current [A] c 1 1 d arc anode heating coefficient [mm s" A" ] c? electrical resistance heating coefficient [s-1 A-2] D diameter of gas nozzle [mm] di diameter of molten metal bridge [mm] F current pulse frequency [Hz] FP push force [N] ha arcing energy [J] hi heat input [J/mm] h. short-circuiting energy [J] I,i current [A] »0 current during arcing period [A] h background current [A] h peak current [A] i, current during short-circuiting period [A] k constant [mm s"1] wire extension [mm] up pressure [N mm"2] p arc power [W] R Reynolds number S wire electrode cross area [mm2] Se contact area [mm2] t time W T current pulse cycle time H ta arcing time W Tt background current duration [s] Ti droplet detachment time H T peak current duration P W t. short-circuiting time W U,u welding voltage [V] ua voltage during arcing period [V] u« voltage during short-circuiting period [V] V, v welding speed [mm/s] 1 va velocity of shielding gas flow- [mm s" wire burn-off rate [mm s"1 wm 1 Wj wire feed rate [mm s" Greek Symbols H friction coefficient v droplet volume [mm3] p density of shielding gas [kg mm"3] a standard deviation 2 ffe elastic strength of wire electrode [N mm" ] Subscript min minimum value max maximum value opt optimal value Contents Foreword 1 Abstarct 3 List of Symbols 5 1 Introduction 11 1.1 Background 11 1.2 Objectives 12 1.3 Scope and Limitation 12 1.4 Disposition 12 2 Gas Metal Arc Welding 15 2.1 General Characteristics 15 2.2 Control of Metal Transfer 17 2.2.1 Steady DC Spray Metal Transfer 17 Transition Current 20 2.2.2 Pulse Controlled Spray Metal Transfer 21 Synergic Control 22 Wire Burn-off Criterion 23 Metal Transfer Criterion 24 Arc Stability Criterion 26 8 2.2.3 Short-Circuiting Metal Transfer 27 Effect of Welding Variables 27