S Arungalai Vendan Liang Gao Akhil Garg P Kavitha G Dhivy 1

S.
Arungalai Vendan · Liang Gao
Akhil Garg · P. Kavitha · G. Dhivyasri
Rahul SG
Interdisciplinary
Treatment to Arc
Welding Power
Sources
Interdisciplinary Treatment to Arc Welding Power
Sources
S. Arungalai Vendan Liang Gao
•
Akhil Garg P. Kavitha G. Dhivyasri

• •
Rahul SG
Interdisciplinary Treatment
to Arc Welding Power
Sources
123
S. Arungalai Vendan P. Kavitha
School of Electrical Engineering School of Electrical Engineering
VIT University VIT University
Vellore, Tamil Nadu Vellore, Tamil Nadu
India India
Liang Gao G. Dhivyasri

State Key Lab of Digital Manufacturing School of Electrical Engineering
Equipment and Technology VIT University
Huazhong University of Science and Vellore, Tamil Nadu
Technology India
Wuhan, Hubei
China Rahul SG
School of Electrical Engineering
Akhil Garg VIT University
Intelligent Manufacturing Key Laboratory of Vellore, Tamil Nadu
Ministry of Education India
Shantou University
Shantou, Guangdong
China
ISBN 978-981-13-0805-5 ISBN 978-981-13-0806-2 (eBook)

https://doi.org/10.1007/978-981-13-0806-2
Library of Congress Control Number: 2018943712
© Springer Nature Singapore Pte Ltd. 2019

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part
of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission
or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, express or implied, with respect to the material contained herein or
for any errors or omissions that may have been made. The publisher remains neutral with regard to
jurisdictional claims in published maps and institutional affiliations.
Printed on acid-free paper
This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Contents
1 Welding an Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Basics of Arc Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.1.1 Electric Charge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.2 Electric Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.3 Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.1.4 Electrical Resistance . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.5 Heat Input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.1.6 Power . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.2 Equivalent Electrical Analogous Representation
of Arc Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3 Arc Welding as a Confluence of Several Disciplines . . . . . . . . . . 6
1.4 Applications of Arc Welding . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2 Insight into Arc Welding Power Source Terminologies . . . . . . . . . . . 15
2.1 Critical Arc Power Source Terminologies for Welding . . . . . . . . 15
2.1.1 Arc Plasma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.1.2 Arc Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.1.3 Arc Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.1.4 Arc Blow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.5 Pinch Effect . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.1.6 Arc Shielding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.2 Influence of Power Source Parameters on Weldment . . . . . . . . . 18
2.2.1 Open-Circuit Voltage (OCV) . . . . . . . . . . . . . . . . . . . . 18
2.2.2 Arc Voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
2.2.3 Welding Current . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
2.2.4 Electrode Polarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2.5 Power Factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.2.6 Duty Cycle and Current Rating . . . . . . . . . . . . . . . . . . . 22
2.2.7 Class of Insulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
v
vi Contents
2.3 Impact of Power Source Characteristics on Weldments . . . . . . . . 23

2.3.1 Static Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . 23
2.3.2 Dynamic Characteristics . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4 Classification of Arc Welding Power Sources . . . . . . . . . . . . . . . 25
2.4.1 Static Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
2.4.2 Rotating Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
2.5 Power Sources Components Briefing . . . . . . . . . . . . . . . . . . . . . 31
2.5.1 Diode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
2.5.2 BJT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.5.3 MOSFET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
2.5.4 Insulated Gate Bipolar Transistor (IGBT) . . . . . . . . . . . 36
2.5.5 Silicon-Controlled Rectifier (SCR) . . . . . . . . . . . . . . . . 36
2.5.6 Pulse Width Modulators (PWM) . . . . . . . . . . . . . . . . . . 38
2.5.7 Microprocessor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.5.8 Microcontroller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
2.5.9 Field-Programmable Gate Arrays (FPGAs) . . . . . . . . . . 39
2.6 Evolution of Arc Welding Power Sources . . . . . . . . . . . . . . . . . 41
2.7 Switch-Based Techniques Adopted for Welding Power
Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... 43
2.8 Literature Addressing Power Source Parameters . . . . . . ....... 61
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ....... 67
3 Control Terminologies and Schemes for Arc Welding Processes . . . . 71
3.1 Control System Terminologies . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.1.1 Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
3.1.2 System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.1.3 Control System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.1.4 Parameters/Variables . . . . . . . . . . . . . . . . . . . . . . . . . . 72
3.1.5 Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.1.6 Disturbances . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.1.7 Setpoint . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.1.8 Feedback . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
3.1.9 Error . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.1.10 Transfer Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.1.11 Open Loop System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
3.1.12 Closed-Loop System . . . . . . . . . . . . . . . . . . . . . . . . . . 75
3.2 Control System Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.2.1 Order of the System . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.2.2 Zeroth Order System . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.2.3 First-Order System . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.2.4 Second-Order System . . . . . . . . . . . . . . . . . . . . . . . . . . 77
3.2.5 Linearity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
3.2.6 Sensitivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Contents vii
3.3 Introduction to Fundamental Controllers . . . . . . . . . . . . . . . . . . . 78

3.4 Stability Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
3.5 Significance of Control System . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.6 Control System for Arc Welding . . . . . . . . . . . . . . . . . . . . . . . . 80
3.6.1 Sensing System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.6.2 Control Strategy and Algorithms . . . . . . . . . . . . . . . . . . 81
3.6.3 Desired Gating Signals . . . . . . . . . . . . . . . . . . . . . . . . . 82
3.7 Controller Schemes Adopted for Welding Power Sources . . . . . . 82
3.8 Process Parametric Influences on Weld Quality . . . . . . . . . . . . . 84
3.9 Real-Time Sample Reports on Formulating Adaptive Control
Scheme for Cold Metal Transfer for JoiningAA6061 . . . . . . . . . 119
3.9.1 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.9.2 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
3.9.3 Controller Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
3.9.4 MRAC Controller Response . . . . . . . . . . . . . . . . . . . . . 122
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
4 Power Sources and Challenges for Different Arc Welding
Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
4.1 Power Sources in Manual Metal Arc Welding (MMA) . . . . . . . . 127
4.2 Power Sources in Shielded Metal Arc Welding (SMAC) . . . . . . . 127
4.3 Power Sources in Gas Tungsten Arc Welding
(GTAW)/Tungsten Inert Gas Arc Welding (TIG) . . . . . . . . . . . . 128
4.4 Power Sources in Gas Metal Arc Welding/Metal Inert
Gas Welding (GMAW/MIG) . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
4.5 Power Sources in Submerged Arc Welding (SAW) . . . . . . . . . . . 130
4.6 Major Challenges in Power Sources . . . . . . . . . . . . . . . . . . . . . . 130
4.6.1 Harmonics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
4.6.2 Effects of Magnetic Field in Arc Welding . . . . . . . . . . . 132
4.6.3 Protection of Power Sources . . . . . . . . . . . . . . . . . . . . . 134
4.6.4 Cooling System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
5 Sensors for Welding Data Acquisition . . . . . . . . . . . . . . . . . . . . . . . 139
5.1 Data Acquisition System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139
5.1.1 What Are Sensors and Transducers? . . . . . . . . . . . . . . . 140
5.1.2 Signals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141
5.1.3 What Is a DAQ Hardware? . . . . . . . . . . . . . . . . . . . . . . 144
5.2 Physical Principles of Sensing . . . . . . . . . . . . . . . . . . . . . . . . . . 144
5.2.1 Characteristics of Different Sensor Types . . . . . . . . . . . . 145
5.2.2 Basic Terminologies . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
5.2.3 Choosing a Sensor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 146
5.3 Key Measurement Components of a DAQ Device . . . . . . . . . . . 147
5.3.1 Signal Conditioning . . . . . . . . . . . . . . . . . . . . . . . . . . . 147
viii Contents
5.3.2 Analog-to-Digital Converter (ADC) . . . . . . . . . . . . . . . . 148

5.3.3 Computer Bus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 148
5.4 Role of Computer in a DAQ System . . . . . . . . . . . . . . . . . . . . . 148
5.4.1 Application Software . . . . . . . . . . . . . . . . . . . . . . . . . . 148
5.4.2 Driver Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151
5.5 Data Acquisition in Arc Welding Processes . . . . . . . . . . . . . . . . 151
5.5.1 Measuring Current and Voltage . . . . . . . . . . . . . . . . . . . 151
5.5.2 Wire Feed Speed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
5.5.3 Shielding Gas Flow . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
5.5.4 Temperature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155
5.5.5 Sensors for Geometrical Parameters . . . . . . . . . . . . . . . . 157
5.5.6 Arc Sensors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 160
5.5.7 Typical Sensors and Their Outputs . . . . . . . . . . . . . . . . 161
5.6 Parameters of Arc Welding Sensors for Various
Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161
5.7 Data Acquisition Using LabVIEW . . . . . . . . . . . . . . . . . . . . . . . 161
5.7.1 Physical Input/Output Signals . . . . . . . . . . . . . . . . . . . . 163
5.7.2 DAQ Device/Hardware . . . . . . . . . . . . . . . . . . . . . . . . . 163
5.7.3 Driver Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
5.7.4 Application Software . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5.7.5 Measurement and Automation Explorer . . . . . . . . . . . . . 164
5.7.6 DAQ Assistant . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
5.8 Case Study 1: Measurement of Temperature During Joining
of 316L Stainless Steel by CMT Process . . . . . . . . . . . . . . . . . . 166
5.8.1 Process Details . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166
5.8.2 Description of DAQ Unit . . . . . . . . . . . . . . . . . . . . . . . 167
5.8.3 Experimental Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
5.8.4 Temperature Plots . . . . . . . . . . . . . . . . . . . . . . . . . . . . 168
5.9 Case Study 2: Characterization of Gas Metal Arc Welding
System Using DAQ . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.9.1 Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169
5.9.2 Welding Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . 171
5.10 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179
6 Optimization in Arc Welding Process . . . . . . . . . . . . . . . . . . . . . . . . 181
6.1 Introduction to Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . 181
6.1.1 Constructing a Model . . . . . . . . . . . . . . . . . . . . . . . . . . 181
6.1.2 System Identification in Arc Welding . . . . . . . . . . . . . . 182
6.2 Significance of Optimization in Welding . . . . . . . . . . . . . . . . . . 182
Contents ix
6.3 ANN-Based Optimization Techniques to Arc Welding

Processes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183
6.3.1 Introduction to ANN . . . . . . . . . . . . . . . . . . . . . . . . . . 183
6.3.2 Backpropagation Neural Network (BP-NN) . . . . . . . . . . 186
6.4 Development of PSO-Based Backpropagation Neural
Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
6.4.1 Particle Swarm Optimization . . . . . . . . . . . . . . . . . . . . . 188
6.4.2 Development of BP-NN Using PSO Algorithm . . . . . . . 189
6.5 Development of Levenberg–Marquardt (LM) Algorithm-Based
Backpropagation Neural Network . . . . . . . . . . . . . . . . . . . . . . . 192
6.5.1 Introduction to LM Algorithm . . . . . . . . . . . . . . . . . . . . 192
6.5.2 Computing the Jacobian Matrix . . . . . . . . . . . . . . . . . . 192
6.5.3 Steps in Levenberg–Marquardt Algorithm . . . . . . . . . . . 193
6.6 Genetic Algorithm for Tuning the Neural Network . . . . . . . . . . . 194
6.7 Case Study 1: Optimization of Flux Cored Arc Welding
Parameters Using GA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
6.7.1 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
6.7.2 Experimentation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
6.7.3 Optimization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
6.8 Case Study 2: Optimization and Prediction of Hardness
and Shear Strength Using PSO Based ANN in FSW
of AA6061 Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
6.8.1 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 200
6.8.3 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 201
6.9 Case Study 3: LM Algorithm-Based ANN Model to Predict
Strength and Joint Resistance of Al-Cu Alloys Joined
by Ultrasonic Welding Process . . . . . . . . . . . . . . . . . . . . . . . . . 203
6.9.1 Objective . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 203
6.9.3 Implementation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 204
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
7 Codes and Safety Standards During Welding . . . . . . . . . . . . . . . . . . 209
7.1 Risk Management Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
7.1.1 Identifying the Potential Hazards . . . . . . . . . . . . . . . . . . 209
7.1.2 Assessment of Risk . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
7.1.3 Risk Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
7.2 Specific Hazards and Control Measures . . . . . . . . . . . . . . . . . . . 211
7.2.1 Airborne Contaminants . . . . . . . . . . . . . . . . . . . . . . . . . 211
7.2.2 Radiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
7.2.3 Electrical Risks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
7.2.4 Risks Due to Electromagnetic Fields . . . . . . . . . . . . . . . 215
x Contents
7.2.5 Exposure to Heat and Burns . . . . . . . . . . . . . . . . . . . . . 215

7.2.6 Compressed and Liquefied Gases . . . . . . . . . . . . . . . . . 216
7.2.7 Personal Protective Equipment (PPE) . . . . . . . . . . . . . . 216
7.2.8 Health Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . 216
7.3 Standard Operating Procedures During Arc Welding . . . . . . . . . . 220
7.3.1 Engine Power Equipment . . . . . . . . . . . . . . . . . . . . . . . 220
7.3.2 In Presence of Electric and Magnetic Fields . . . . . . . . . . 221
7.3.3 During Handling Cylinders . . . . . . . . . . . . . . . . . . . . . . 222
7.3.4 While Handling Shielding Gases . . . . . . . . . . . . . . . . . . 222
7.4 Welding Codes: American Welding Society (AWS) . . . . . . . . . . 222
7.5 Quality Assurance and Quality Management . . . . . . . . . . . . . . . 224
7.5.1 En ISO 15609 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
7.5.2 En ISO 15614-1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 224
7.5.3 EN ISO 15614-2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
7.5.4 EN ISO 15610 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
7.5.5 EN ISO 5817 and ISO 10042 . . . . . . . . . . . . . . . . . . . . 227
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Chapter 1
Welding an Overview
This book categorically presents the mechanical, metallurgical and electrical termi-
nologies and the associated science for welding power sources. Key focus is laid
on arc welding processes owing to its growing demand in versatile applications.
Figure 1.1 illustrates the types of welding processes.
Arc welding is a fusion process wherein joining of materials is accomplished by
the application of sufficient electrical potential to an electrode and creating an arc
between the electrode and the work piece. The electric arc serves as a conductive
path for the current flow via the electrode. Consumable electrode conducts the current
while melting to provide filler material for the joint [1]. If the electrode is of a non-
consumable type, it simply conducts current between electrode and workpiece and
a separate filler material is used. The arc welding process is classified under the
following categories based on the work setup configuration (Fig 1.1):
(i) Shielded metal arc welding (SMAC) or manual metal arc welding (MMAW)
or stick Welding,
(ii) Gas Metal Arc Welding (GMAW) or Metal Inert Gas Welding (MIG),
(iii) Flux-Cored Arc Welding (FCAW),
(iv) Gas Tungsten Arc Welding (GTAW) or tungsten Inert Gas Welding (TIG),
(v) Submerged Arc Welding (SAW),
(vi) Cold Metal Transfer (CMT).
Power source is establishing the electric arc. It is a vital component in a welding
process that is responsible for delivering specific voltage–current characteristics with
a well-defined control to achieve preciseness while regulating the static and dynamic
characteristics. In general, the power source is designed to operate with 230 V single
phase or 440 V three phase while delivering a current of 0–300 A and a voltage
between 0 and 90 V. The operating modes vary based on AC/DC current, which
is identified and finalized depending on the material to be welded. The welding
process requires low voltage and high current to minimize the heat generated, which
is a challenge for the researchers. The dynamic behaviour of power sources and their
modernization through automation adds to the existing complications [2].
© Springer Nature Singapore Pte Ltd. 2019 1
S. A. Vendan et al., Interdisciplinary Treatment to Arc Welding Power Sources,
https://doi.org/10.1007/978-981-13-0806-2_1
2 1 Welding an Overview
Welding Process
Resistance High Beam Solid State

Gas Welding Arc Welding
Welding Energy Welding Welding
• Friction
• Carbon arc • Ultrasonic
• Metal arc • Diffusion
• Butt
• Oxy-acetylene • Plasma arc • Explosive
• Spot • Electron beam
• Air,-Acetylene • Gas Metal Arc • Magnetically
• Seam • Laser
• Oxy-Hydrogen (MIG) impelled Arc Butt
• Projection
• Gas Tungsten Arc Welding (MIAB)
(TIG) • Induction Pressure
• Cold Metal Welding (IPW)
Transfer (CMT) • Flash Butt welding
(FBW)
• Magnetic Pulse
Welding (MPW) –
cold welding process
Fig. 1.1 Types of welding processes
Despite its key role, the literature based on the behaviour of welding
power source and its implication on weld efficiency that accounts for mechani-
cal/metallurgical/electrical phenomena or behaviour are limited. Moreover, most of
the power source studies are presented by the authors in their native language which
poses more complexity and raises ambiguous pointers.
This book attempts to explore the inter- and intra-dependent parameters of the
joining process as a whole and the physics underlying. The physics of welding met-
allurgy involves the material being subjected to heat and pressure. This necessitates
the inclusion of welding power source, control module, electrode setup and the spec-
imens to be welded. Eventually, the weld specimens are subjected to non-destructive
analysis, mechanical and metallurgical characterization as illustrated in Fig. 1.2 to
establish the feasibility of that particular process for the chosen material.
1.1 Basics of Arc Welding
Governing to the available literature of arc welding power sources, the following
terminologies are to be understood from the electrical perspective of welding. All
matter is made up of atoms and each atom contains fundamental particles; electrons,
protons and neutrons. Electrons, the negatively charged particles have the capability
to move from one place to another; and protons, the positively charged particles
remains stationary at normal conditions.
1.1 Basics of Arc Welding 3
Mechanical Behaviour
Input Parameters
(Current, Voltage, Metallurgical Characteristics
Properties
Power factor...) Tensile strength
WELDING SYSTEM Micro and Macro Images
Bead Width SEM analysis
Impact TEM analysis
Hardness
Power
Source
Convertors Non Destructive Analysis
Radiography
Intermediate
Ultrasonic
System
Liquid penetrant test(LPT)
Dye penetrant Test(DPT)
Controllers Visual Inspection
Optical Inspection
Eddy Current
Magnetic particle test
Thermal Analysis
Finite Element Analysis (FEA)

Finite Element Modeling (FEM)
Thermogravimetric Analysis
(TGA)
Thermal Stress Analysis
Analysis on Heat Affected Zone
(HAZ)
Fig. 1.2 Intra- and inter-dependent electrometallurgy of arc welding process
1.1.1 Electric Charge
Electric charge is the physical property of matter that causes it to experience

a force when placed in an electromagnetic field. There are two types of electric
charges; positive and negative. Like charges repel and unlike attract. An object with
an absence of net charge is referred to as neutral. The unit of electric charge is
the coulomb (C).
1.1.2 Electric Current
Electrical current is a measure of the amount of electrical charge transferred per unit
of time. It represents the flow of electrons through a conductive material, such as a
metal wire. It is measured in amperes (A).
1.1.3 Voltage
A material is said to be in an electrically uncharged state when the number of protons

and electrons in a material are equal in number. This balance is disturbed when an
external energy forces the electrons to move from one atom to another atom or
away from that atom. This energy is termed as the electromotive force or voltage.
Without voltage, there is no current flow. The transfer of electrons from a negative
to a positive charge throughout the length of a conductor in presence of a voltage
potential constitutes the electrical current. Materials vary in their ability to transfer
electrons. Insulators have a strong electron bonding and their atoms resist the free
movement of electrons. Hence, they are poor conductors of electricity. However,
conductors, on the other side have more number of free electrons due to their weak
electron bonding, thereby exhibiting superior electron flow leading to conduction.
The unit of voltage is the volts (V).
1.1.4 Electrical Resistance
Electrical resistance can be said as the resistance to the flow of current. Resistance is
best understood through Ohm’s Law which emphasizes that even though the voltage
and current changes, the ratio of their relationship remains as per Eq. (1.4)
V I∗R (1.1)
where
V Voltage (V),
I Current (A),
R Resistance ().
The theory of electrical resistance is of immense importance in the arc welding
process. And it is this resistance in the air gap between the electrode and the base
material that contributes to the conversion of electrical energy into heat energy. In arc
welding process, this energy conversion causes the workpiece and the electrode/filler
material to melt resulting in the formation of weld pool (crater).
1.1.5 Heat Input
Study of welding fundamentals involves heat-flow concepts. The heat energy is

defined as the quantity of energy introduced per unit length of weld from a trav-
elling heat source. The heat density is expressed in joules per meter2 or joules per
millimetre2 [3]. This important measure is calculated as the ratio of total input power
in Watts to its velocity which is represented in Eq. (1.4),
f ∗v∗ I
H (1.3)
V
1.1 Basics of Arc Welding 5
where
H Heat density (KJ/cm2 ),
ƒ Heat transfer efficiency,
V Voltage (V),
I Current (A),
v Travel velocity of heat source (mm/s).
In arc welding, as arc voltage forces the electrons to move faster, the energy
generated by them is used partially to overcome the resistance created by the arc
gap. Thus, this energy is converted into heat energy and the temperature increases to
the point where it brings the materials to a molten state.
1.1.6 Power
Watt is the term defined as the amount of power required to maintain a current of 1 A
at a pressure of 1 V. The circuit voltage that comes is typically a constant factor, but
the current drawn from the utility company depends on the number of watts required
to run the electrical appliance. Power is denoted by Eq. (1.4) as follows:
W V∗I (1.3)
where
W Power (W),
V Voltage (V),
I Current (A).
The current required by an electrical equipment can be calculated by dividing the
watts rating of the device by the primary voltage for which it is designed is shown
in Eq. (1.4),
I W/V (1.4)
1.2 Equivalent Electrical Analogous Representation of Arc

Welding
The arc generated causes a heating effect which makes the workpiece and consumable
electrode or the filler material to melt and fuse. To understand the principle of energy
conversion, a fundamental knowledge on electricity and the electrical aspects of
welding are important. A typical arc welding process is depicted along with the
equivalent electrical analogous circuit which is shown in Fig. 1.3.
The equivalent electrical analogous notations are presented in Table 1.1.
Mains Electrode
Electric Welding Holder
Supply Cable
Arc Welding Welding current Welding Arc
Power Electrode
Source
Return Cable
Return Workpeice
Clamp
I2R Power or
diode in reverse
bias
Current I
AC/DC
Voltage Load
Source Resistor
(Or)
Ground
Fig. 1.3 Equivalent electrical circuit of arc welding process
Table 1.1 Equivalent electrical analogous notations

Mechanical component Electrical equivalent
Welding cable Welding current
Arc Welding power source AC/DC Voltage source
Electrode and workpiece Resistive load or diode in reverse bias
Welding arc Power I2 R (W)
1.3 Arc Welding as a Confluence of Several Disciplines
The flow diagram shown presents a detailed overview on the arc welding power
source and its interdependent terminologies. All the subsections in the flow diagram
are explained briefly in the subsequent chapters (Fig. 1.4).
1.4 Applications of Arc Welding
Although most of the welding processes at the time of their developments were used
for repairing in production, gradually they started gaining relevance to actual compo-
nent manufacturing. Presently welding is widely being used in fabrication of pressure
vessels, bridges, building structures, aircraft and space crafts, railway coaches and
Fig. 1.4 Power source and interdependent terminologies

7
8
Table 1.2 Application of arc welding. (Arc Welding in Ships, Aerospace and Automobile Applications - Google Search [Online]. Avail-
able: https://www.google.com/search?biw=1536&bih=710&tbm=isch&sa=1&ei=1eotW-z7I8nrvATroI4o&q=arc+welding+in+ships%2C+aerospace+and+aut
omobile+applications&oq=arc+welding+in+ships%2C+aerospace+and+automobile+applications&gs_l=img.3...15329.22442.0.22806.33.19.0.)
Application industry Type of arc welding Components (name with images)
Aerospace GTAW/TIG, CMT, MIG/GTAW Fuel pipes
Bicycle GTAW/TIG, MMA Cycle joints

1 Welding an Overview
(continued)
Table 1.2 (continued)
Automotive GMAW/MIG Automobile components
Metal Pipes MMA, TIG, MIG, MIAB Pipe joining
(continued)
9
10
Ship Underwater welding, FCAW,

GMAE/MIG, SMAW
(continued)
Railways FCAW, GMAW, SAW, SMAW Railway tracks
(continued)
11
12

Defence industry GTAW/TIG, GMAW/MIG, SMAC, SAW Military vehicle components
(continued)
Construction industry MMA, TIG Bridges
Metal structures
13
general applications. It is also being used in shipbuilding, automobile, electrical,

electronic and defence industries, laying of pipe lines and nuclear plant structures.
Some critical application areas of arc welding are presented in the Table 1.2.
References
1. C. TLE, Arc Welding fundamentals (1999). Available: http://www.lincolnelectric.com/en-us/su

pport/process-and-theory/Pages/arc-welding-detail.aspx. Accessed 2017
2. B. Ogunbiyi, The influence of power source type on welding performance and weld quality.
Exploiting Adv. Arc Weld. Technol. 55 (1999)
3. M. A. Gaodi and . D. I. Sangotra, A review paper on effect of varying welding heat inputs on
microstructure, mechanical properties and corrosion behaviours of ferritic stainless steel & mild
steel. Int. J. Mod. Eng. Res. 4(1), 105–109 (2014)
Chapter 2
Insight into Arc Welding Power Source
Terminologies
2.1 Critical Arc Power Source Terminologies for Welding
Welding arc is an electric current flowing between two electrodes through an ionized
gas column. Arc stability is a critical problem which influences the metal transfer
during welding. When the arc is stable, metal transfer is uniform with minimum spat-
ter [1]. The terminologies of major concerns during the joining process are presented
below.
2.1.1 Arc Plasma
Arc plasma as shown in Fig. 2.1 is the ionized state of welding gas, which is a mixture
of equal amounts of ions and electrons [2]. To study the effects of the arc plasma
over the weld pool, four main factors to be considered are heat flux, current density,
shear stress and arc pressure. Heat flux and current density have a direct relationship
towards the weld pool. While, increase in shear stress (shear stress is the amount of
force per unit area perpendicular to the axle of the member) results in outward flow
at the top of weld pool and increase in arc pressure (force exerted by arc per unit
area) results in a more concave surface of weld pool.
Where,
1. Gas plasma,
2. Protection nozzle,
3. Shielding gas,
4. Electrode,
5. Nozzle constriction,
6. Electric arc.

https://doi.org/10.1007/978-981-13-0806-2_2
16 2 Insight into Arc Welding Power Source Terminologies
Fig. 2.1 Arc plasma
2.1.2 Arc Efficiency
Arc efficiency is defined as the heat input to the metal divided by the total heat
energy of the arc [2]. Or simply, it is defined as the amount of arc energy delivered
to the workpiece [3]. It is essential to know the arc efficiency to measure the melting
efficiency using heat-flow models and experimentation. The arc efficiency of non-
consumable electrodes is lower than that of the consumable electrodes [4]. Heat input
can be calculated by the Eq. (2.1) [3]
Arc voltage*Arc current
Heat input KJ/cm2 ∗ Arc efficiency (2.1)
Welding speed*1000
2.1.3 Arc Stability
Arc stability is influenced by arc power and metal transfer modes. Spattering is the
emission of metals during welding process, which is a negative effect of poor arc
stability. The properties of ideal and stable welding arc are
(i) Constant shape of all transferred materials,
(ii) Constant length of arc and
(iii) Low or no spattering.
Another factor affecting arc stability is the shielding gas mixture of CO2. The
stability is low when the shielding gas mixture has high CO2 .
2.1 Critical Arc Power Source Terminologies for Welding 17
2.1.4 Arc Blow
Arc blow is a phenomenon where the arc tends to deviate from the welding point [5].
This is because of the imbalance in the magnetic field surrounding the workpiece.
This phenomenon occurs in the following three situations:
(i) Presence of magnetic materials around welding arc,
(ii) Changes in the direction of current,
(iii) Presence of magnetic materials at the end of the plate.
Arc blow mainly occurs in DC welding currents and can be reduced using AC
current. Arc with low voltage is stiff and resists deflection than the arc with a high
voltage. The melting rate is influenced by the heating effects of current which is
given by the Eq. (2.2)
β
Melting Rate α I + ∗ ls ∗ I 2 (2.2)
αω
where
α and β are constants
l s Electrode resistivity (m),
α Cross-sectional area of wire (m2 ),
l s Welding current (A).
2.1.5 Pinch Effect
The arc carrying the current is always surrounded by a magnetic field [5]. In arc
welding process the direction of electromagnetic force varies, and it depends on the
direction of flow of welding current. The magnetic field has a force directed towards
the centre of arc, called as Lorentz force. With increase in current, the arc radial
constriction increases due to increased magnetic force. This process is termed as
pinch effect (as shown in Fig. 2.2) [6]. The pinch force is directly related with the
welding current and electrode wire diameter and influences drop detachment to the
weld pool [7].
2.1.6 Arc Shielding
At elevated temperatures, metals tend to react with oxygen and nitrogen in the atmo-
sphere. When the molten metal pool is exposed to air it results in formation of oxides
and nitrides on metals which deteriorates the toughness and strength of the weld
joint. Arc welding processes require a protective shield of gas, vapour or slag which
is termed as arc shielding [2].
Fig. 2.2 Pinch effect during

short-circuit transfer
2.2 Influence of Power Source Parameters on Weldment
Critical electrical power source parameters influencing the final weldment are
explained briefly in the following context.
2.2.1 Open-Circuit Voltage (OCV)
The open-circuit voltage refers to the potential difference between the two terminals
of a welding power source at no load condition. The selection of an optimal value
of OCV depends on type of workpiece, type of welding current, polarity, type of
arc welding, type of electrode and its composition, etc. The base metal with a lower
ionization potential such as potassium (K), sodium (Na), calcium (Ca) requires a
low OCV than the base metal with a higher ionization potential such as oxygen (O),
fluorine (F) and nitrogen (N). Among all arc welding processes, GTAW requires
less OCV since it uses tungsten electrode which has higher free electron emitting
capability [8].
• When an AC source is used then the setting up of correct OCV is very important
for arc stability.
• AC power source requires higher OCV than DC because AC is prone to arc insta-
bility, since the welding current continuously changes its magnitude and direction
for every half cycle. But, in DC-type power sources the current remains constant.
• Very high value of OCV also results in electric shock.
• OCV for transformers ranges from 70 to 90 V; and for rectifiers it is 20–60 V.
The OCV of a power source is always 5–10 V greater than the arc voltage.
2.2 Influence of Power Source Parameters on Weldment 19
Electrode Electrode
Tip Tip
Cathode (-) Anode (+)
Welding Welding
Arc Arc
Arc Voltage Arc Voltage
Base Metal Base Metal

Anode (+) Cathode (-)
Fig. 2.3 Arc column
Cathode drop
Arc length (cm)
Arc Column
Anode drop
Arc Voltage
(Volts)
Fig. 2.4 Plot for arc voltage and arc length
2.2.2 Arc Voltage
Arc voltage shown in Fig. 2.3 is the potential difference between the electrode tip
and the workpiece when there is current flow [9].
Any fluctuation in the arc length affects the resistance to the flow of current, which
in turn affects the arc voltage. Increase in arc length increases arc voltage. Increase
of electrode extension increases the arc voltage and increases the electrode heating
due to increase in resistance to current flow. The arc voltage can be categorized into
three zones as shown in Fig. 2.4;
(i) Cathode drop,

(ii) Anode drop, and
(iii) Arc column.
Fig. 2.5 Depth of

penetration for different
welding currents [44]
The cathode drop is near the electrode tip where the potential drop occurs swiftly
in a narrow zone. Near the anode drop zone, the voltage drop is fast on the anode
surface and is uniform. Sharp drop in arc potential near cathode and near anode is
attributed to the cooling effect of electrode and workpieces.
2.2.3 Welding Current
There are two types of electric current; Alternating Current (AC) and Direct Current
(DC). DC is a stream of negatively charged electrons flowing through the cable,
moving from the negative pole to the positive pole, whose flow is always unidirec-
tional. While, AC is achieved when the power supply switches between positive and
negative polarity at a frequency of 50 or 60 Hz [10].
As welding current increases, the weld penetration also increases and vice versa,
which is illustrated in Fig. 2.5. Among the several types of arc welding, only TIG
welding of aluminium utilizes AC because the continuous change in polarity of
AC helps in tearing the layer of oxide on the surface of weld pool. Most welding
applications are implemented with negative DC polarity to get the required depth of
penetration.
2.2 Influence of Power Source Parameters on Weldment 21
2.2.4 Electrode Polarity
When DC power source is used polarity of the electrode has a significant role. Polarity
in arc welding designates the direction of current flow. The types of connection are
Direct Current Electrode Positive and Direct Current Electrode Negative [10].
i) DCEP (Direct Current Electrode Positive)—The continuous flow of avalanche
of electrons in small passage produces the arc. After the electrons are emitted
from the base material (negative polarity), they are accelerated due by the poten-
tial difference and strike the electrode (positive polarity) with high velocity. The
kinetic energy of the electrons is partially converted partially to thermal energy
upon striking, which results in heat dissipation at the electrode [11]. Two-thirds
of arc heat is generated at electrode; while, only one-third heat is generated
at base material. Eventually, the electrode melts swiftly and weld penetration
increases.
(ii) DCEN (Direct Current Electrode Negative)—The electrons flow from electrode
to the base plates. Subsequently, more heat is generated at the base material as
compared to electrode. So, depth of penetration reduces.
2.2.5 Power Factor
Power factor is defined as the ratio of actual power (KW) used to produce the rated
load to the apparent power drawn from the supply (KVA) as shown in Eq. (2.3)
Real Power (KW)
Power factor (2.3)
Apparant Power (KVA)
Real power (W) is the one that does a work. For example, torque produced in
rotating a machine. It is a real value. Apparent power (VAR) is the power that bounces
back and forth between source and load. It is an imaginary value. Power factor is an
important parameter because it gives a measure of effectiveness of the real power
utilization in the system [10]. It also denotes the measure of distortion of the line
voltage and line current and the presence phase shift between them. The power source
should have high power factor for utilizing the energy effectively. When power source
has higher power factor it reduces reactive power in system and there by reduces
power consumption and cost. The power factor is high in welding transformers than
the other power sources. The advantages of having high power factor are improved
voltage quality, occasional voltages drop, less transmission losses, more economic
operation, reduced reactive power thereby reduced power consumption, reduced
cable cross-sectional area, etc. [12].
2.2.6 Duty Cycle and Current Rating
Duty cycle refers to the percentage of welding time to the total welding cycle (welding
type + rest time). Total welding time is taken as approximately 5 min. Duty cycle and
associated current are important for the safety of power sources. The welding current
drawn at a duty cycle can be expressed by the Eq. (2.4)
DR ∗ IR2 I100
2
∗ D100
2
(2.4)
where
I Current (A) at 100% duty cycle,
D100 100% duty cycle,
IR Current (A) at required duty cycle,
DR Required duty cycle.
At 100% duty cycle, minimum current will be drawn to avoid excessive heating of
winding and power source. Reduction of duty cycle results in withdrawal of higher
magnitude of welding current [12]. When high current is drawn from power source,
the flow of current causes heating of coil and cables due to electrical resistance of
heating. During such time welding, must be stopped to protect the welding power
source (resting time). Maximum current that can be drawn from a power source
at a duty cycle depends on size of wire and type of insulation. Effective cooling
systems reduce the withdrawal of higher current from the power source. Duty cycle
is inversely proportional to the current drawn.
2.2.7 Class of Insulation
The insulation is classified as A, E, B, F and G in the increasing order of maximum

allowable temperature 60, 75, 80, 100 and 120 °C, respectively. Duty cycle of power
source is governed by the class of insulation. The duty cycle of a power source at
given current also is governed by the maximum allowable temperature of the various
components such as primary and secondary coils, cables, connectors, etc. This, in
turn, depends on the material quality, type of materials used in the power source.
2.3 Impact of Power Source Characteristics on Weldments 23
2.3 Impact of Power Source Characteristics on Weldments
2.3.1 Static Characteristics
Static characteristic of a welding power source tells the variations in voltage with
respect to current when a pure resistive load is connected to the power source. They
are also called as constant characteristics.
2.3.1.1 Constant Current Characteristics (CC)
The CC power sources are designed to maintain the welding current approximately
constant, during changes in the arc length and arc voltage. The V–I curve shown in
Fig. 2.6 is called as droop characteristics because there exists a negative slope [8].
When consumable electrodes are used then the electrode melting rate remains con-
stant without making them to get stuck with the workpiece. The CC power sources are
suitable for thicker consumable electrodes and for non-consumable electrodes such
as tungsten electrode since the welding current is limited and maintained constant.
They are widely used in MMA, TIG welding, etc. [12].
2.3.1.2 Constant Voltage Characteristics (CV)
In the CV power source, the output voltage is set on the source which then estab-
lishes the required arc voltage. The power source supplies the welding current that
is necessary to melt the electrode at a rate to maintain the arc voltage limit that is
pre-set. These types of power sources are used only for MIG welding and GMAW
[13]. The speed of electrode wire feeder drive becomes an important parameter to
control the amount of welding current, since the wire feed rate is directly proportional
to the welding current. When the power source is used in combination with a con-
stant electrode wire feed system, it then results in a system with a self-regulating arc
Fig. 2.6 CC Characteristics
Constant Current
Characteristics
Voltage (Volts)
Current (Amperes)
Fig. 2.7 CV characteristics
Constant Voltage
Characteristics
Voltage (Volts)
Current (Amperes)
Fig. 2.8 Rising

characteristics
75 Constant Current
Voltage (Volts)
Constant
Arc length
Voltage
Long
35
Medium
25 Small
100 200 300

Welding Current
(Amperes)
length. This power source maintains a constant arc voltage by increasing or decreas-
ing the welding current since the arc length can sometimes vary due to external or
internal fluctuations such as operator errors, puddle movement, plate irregularities,
etc., thereby the V-I characteristic curve shown in Fig. 2.7, which is called as flat
characteristics. But practically, it has a slight negative slope due to internal resistance
and inductance of the power source.
2.3.1.3 Rising Characteristics
Rising characteristics is mainly observed in automated welding process, where con-

stant voltage is required for the success of welding. A slight positive slope can be
observed in the arc characteristics and welding current is increased little. There are
significant changes in voltage with slight changes in arc length while there is low
variation in current. The arc length in Fig. 2.8 is divided into three parts namely,
small, medium and large which are called as voltage source curves. The junction of
constant current and constant voltage curve is called as operating point of the power
source, which can be changed during the welding process [13].
2.3 Impact of Power Source Characteristics on Weldments 25
2.3.2 Dynamic Characteristics
Dynamic characteristics is essential for success of welding because during arc weld-
ing, the arc is subjected to severe transients and fluctuations in arc voltage due to
minor changes in arc length. Thereby, the arc current rapidly fluctuates for a very
short time. Dynamic characteristics depicts the instantaneous variations in welding
voltage and current relationships caused by the metal transfer across arc, welding
parameters, electrical parameters of power source such as resistance, inductance
and capacitance, and determines the process stability. The load is subjected to rapid
changes from arc condition like open circuit and with load state to short circuit and
vice versa. During short circuit, there is rapid rise in welding current and electrode
wire tip melts and falls on weld pool resulting in a spatter. With advancement in tech-
nology, automated welding is available that necessitates stabilizing of the welding
parameters and controlling the process remotely using a computer [13].
2.4 Classification of Arc Welding Power Sources
A variety of welding power sources are available for electric arc welding and some of
the major types are discussed briefly. Arc welding power sources are majorly divided
into two main categories: static type and rotating type as shown in Fig. 2.9.
2.4.1 Static Types
Static-type power sources utilize commercial electrical supply to energize a trans-

former that steps down the input voltage to a useable voltage for the welding process.
The two subcategories of static power sources are the transformer type and rectifier
type.
2.4.1.1 Transformer-Type Power Sources
Transformer-type power source as shown in Figs. 2.10 and 2.11 produce only Alter-
nating Current (AC). They are termed as ‘Welding Transformers’. All the AC-type
sources are utilized with single-phase primary power and are constant current type
[14].
ARC WELDING
POWER
SOURCES
STATIC TYPE
ROTATING TYPE
Uses commercially
Uses electromechanical
generated power to energize
components during
a transformer that steps
absence of commercial
down the line voltage to
power
usable welding voltage
Transformer Rectifier Motor Engine

Type type Generators Driven
Welding Rectifiers
Only Alternating Produce DC
Current (welding Current Or DC & Motor coupled to
Gasoline Or
Transformer) AC Welding generator or
Diesel Engine
Currentt alternator
All AC types uses
single-phase Constant current or
Constant current Constant voltage or
type combination of
both Both have AC or DC
welding power sources or
combinations of both.
They also use Constant
voltage and Constant
current source
Fig. 2.9 Classification of arc welding power sources
Ac Power
50/60Hz
Primary of Step Secondary of Welding

Switch AC down AC step down AC Filter output
Transformer Transformer Power
Fig. 2.10 Block diagram of Transformer-type power source
2.4.1.2 Rectifier-Type Power Sources
Rectifier-type power sources also termed as welding rectifiers are introduced in mid-
1970s. They produce DC or, both AC and DC welding current. They are being utilized
with either single-phase or three-phase input supply. They consist of a transformer
and perform rectification of AC to DC using power semiconductor devices such
as selenium rectifiers, silicon diodes or Silicon-Controlled Rectifiers (SCR). These
2.4 Classification of Arc Welding Power Sources 27
Fig. 2.11 Transformer-type power source
Ac Power
50/60Hz
Voltage Welding
Control
transformation AC AC Rectifier DC Filter output
circuit
and isolation Power
Fig. 2.12 Block diagram of Rectifier-type power source
power sources are available in both the CC and CV type by incorporating as a

single unit. They find their application in coated electrode welding, non-consumable
electrode welding and for welding with solid or flux cored wires. Figures 2.12 and
2.13 show a rectifier-type power source.
The rectifier- and transformer-type power sources have features such as low noise,
high efficiency and minimal maintenance compared to that of generator-type power
sources [14].
2.4.1.3 Inverter-Type Power Sources
Inverter-type power sources are introduced in the 1940s and became successful in
the 1970s. They boost the frequency as much as 1000 times of the input frequency.
They are beneficial in reducing the coil turns of transformers and core area resulting
in smaller and lighter machine than the conventional type transformers. They sup-
port both single- and 3-phase supplies. Their applications include GTAW, GMAW,
SMAW, FCAW processes. Figures 2.14 and 2.15 depict the inverter-type power
source. First, the high-voltage and low-current AC is rectified into DC which is then
filtered and fed to the switching components. Switching occurs at very high rates
Rectifier type
DC
Fig. 2.13 Rectifier-type DC power source
Ac Power Inverter section

50/60Hz Welding
output
Power
Input Power Output
AC Filter DC Transformer AC DC Filter
rectifier switches rectifier
Control Sensing
Circuit circuit
Fig. 2.14 Block diagram of Inverter-type power source
up to 50,000 times per second. This high-voltage and low-current fast-switching DC

appears like an AC to the transformer. The transformer then steps down the voltage
and increases the current. The resultant low voltage and high current are again filtered
for the improvement of welding performance [15].
Inverter-type power sources provide features such as overload protection and
power control for better controllability of welding parameter efficiency than the other
Fig. 2.15 Inverter-type power source
type of power sources. They are also used in combination with microcontrollers to
access and modify the electrical characteristics of the power source in real time
through software [14].
2.4.2 Rotating Types
Rotating-type power sources may be divided into two classifications.
2.4.2.1 Motor Generators
Motor generator power source as shown in Figs. 2.16 and 2.17 consist of an electric
motor coupled to a generator or alternator that produces the desired welding power.
These machines produce superior weld qualities. However, due to the moving parts
they require frequent maintenance. Welding generators can be of both AC and DC
type [15].
Shaft
coupling
Power In Motor Generator Power Out
Fig. 2.16 Block diagram of motor generator-type power source
Fig. 2.17 Motor

generator-type power source
2.4.2.2 Engine Driven
Earlier, arc welding power sources used motor generator types to produce welding
current, which followed the principle of current generation by the movement of
conductor through a magnetic field. These engine-driven electromechanical devices
find its applications, where there is no availability of commercial power. They are
generally called as rotating-type power sources and are classified further as alternators
that produce AC and generators that produce DC [14]. A gasoline or diesel engine
is coupled to a generator or alternator that produces the desired welding power as
shown in Figs. 2.18 and 2.19. They are used extensively on applications beyond
commercial power lines and as mobile units. Both rotating types can deliver either
AC or DC welding power, or a combination of both. Both types are available as
constant current or constant voltage types.
Shaft
coupling
Gasoline or
Generator /
Diesel Power Out
Alternator
Engine
Fig. 2.18 Motor generator-type power source

Fig. 2.19 Engine driven

power sources
2.5 Power Sources Components Briefing
Some of the fundamental terminologies addressing the literature are briefly explained
below.
2.5.1 Diode
A diode is a semiconductor device with two semiconductor materials in physical

contact; one with excessive electrons (N-type) and the other with excessive holes
(P-type). The function of a diode is to control the direction of current flow. The two
terminals of the diode are anode and cathode [15]. The symbolic representation of a
PN junction diode and the diode component is given in Fig. 2.20.
Fig. 2.20 Symbol of diode (left) and diode component (right)

Depending on the voltage applied across a diode, it operates in one of the three
regions as shown in Fig. 2.21;
Fig. 2.21 Forward and reverse characteristics of PN junction diode
1. Forward bias: The diode is ‘ON’ and current conduction takes place when the
voltage across the diode is greater than the forward voltage.
2. Reverse bias: This is the ‘OFF’ state of a diode, where the voltage is less than
forward voltage but greater than reverse breakdown voltage. A very small amount
of current in the order of nA flows which is called as reverse saturation current.
3. Breakdown: When the voltage applied across the diode becomes very large and
negative, lots of current flows in the reverse direction from cathode to anode.
2.5.2 BJT
Bipolar Junction Transistor (BJT) is a semiconductor device which consists of three

terminals (emitter, base and collector) separated by two P–N junctions as shown in
Figs. 2.22 and 2.23. The P–N junction between the ease and emitter has a barrier
voltage of about 0.6 V, which is an important parameter of a BJT [15]. In BJT, the
current flow takes place due to both types of charge carriers; electrons and holes,
hence they have the name ‘bipolar’.
2.5 Power Sources Components Briefing 33
Fig. 2.22 Symbol of BJT:

NPN transistor (left) and
PNP transistor (right)
Fig. 2.23 BJT component
The bipolar transistors can operate within three different regions:

1. Active region—The BJT operates as an amplifier
2. Saturation region—The transistor is in ‘ON’ state and operates as a switch
3. Cut-off region—The transistor is in ‘OFF’ operating as a switch and the collector
current is zero.
BJT can be connected in three configurations as shown in Fig. 2.24;
1. Common Base Configuration (CB): It is also called as grounded base configu-
ration. The emitter is input terminal and collector is the output terminal. Base is
the common terminal.
Fig. 2.24 Different BJT configurations
2. Common Emitter Configuration (CE): It is called as grounded emitter configu-

ration. The base is input terminal and collector is the output terminal. Emitter is
the common terminal.
3. Common Collector Configuration (CC): It is called as grounded collector config-
uration. The base is input terminal and emitter is the output terminal. Collector
is the common terminal.
2.5.3 MOSFET
MOSFET is abbreviated as for Metal–Oxide Semiconductor Field Effect Transistor.

It has 4 terminals; gate, drain, source and bulk as shown in Fig. 2.25. A pulse at the
gate controls the current that flows between the drain and source. The voltage in the
bulk plays a minor role and usually not shown in the symbol [15]. Since it has a
controlling terminal gate, it is used to control signals in electronic circuits.
The transistor has a metal region, which is conductive and is where the gate
terminal is connected, a part of oxide region which is an insulator, and a region of
semiconductor as shown in Fig. 2.25.
Fig. 2.25 Schematic of MOSFET (left) and MOSFET component (right)

There are two ways in which a MOSFET works. The first is depletion mode;
when there is no gate pulse, the channel exhibits its maximum conductance. As
the gate voltage increases either towards positive or negative direction, the channel
conductivity decreases. The second is called enhancement mode; when there is no
voltage on the gate, there is no channel, and the device does not conduct (Fig. 2.26).
A channel is only produced by the application of a voltage to the gate. The greater
the gate voltage, better the conduction. Since the oxide layer is so thin MOSFET is
susceptible to permanent damage due to electrostatic charges.
MOSFET operating regions as shown in Fig. 2.27 are as follows:
1. Cut-off region: The MOSFET is off and there is no current flow. In this region,
MOSFET behaves like an open switch.
2. Ohmic or linear region: The drain-to-source current increases with an increase
in the value of drain-to-source voltage. They act as amplifiers in this region.
Fig. 2.26 Symbol of

MOSFET: NMOS (left) and
PMOS (right)
Pinch-off voltage
Ohmic region Saturation region

Drain to source Current (A)
Cut-off region
Drain to source Voltage (V)
Fig. 2.27 Characteristics of MOSFET

Fig. 2.28 IGBT symbol

(left) and component (right)
3. Saturation region: In this region, the MOSFETs have constant drain-to-source

current in spite of an increase in drain-to-source voltage. Once VDS exceeds the
value of pinch-off voltage, the device acts like a closed switch through which a
saturated value of drain-to-source current flows. Thereby, this region is chosen
whenever MOSFETs are required to perform switching operations.
2.5.4 Insulated Gate Bipolar Transistor (IGBT)
The insulated gate bipolar transistor is a cross between a conventional BJT and
a MOSFET. It combines the insulated gate technology of the MOSFET with the
output performance characteristics of a conventional BJT [15]. The three terminals
of IGBT are collector, emitter and gate as shown in Fig. 2.28.
Collector and emitter are associated with the conductance path of current, while
its gate terminal controls the device.
Since IGBT is a voltage-controlled device, it requires only a small voltage on
the gate to maintain conduction unlike BJTs which require that the base current
is continuously supplied in a sufficient quantity to maintain the saturation. Also,
IGBT is a unidirectional device it switches the current in forward direction only
from collector to emitter. An IGBT is simply turned on or off by activating and
deactivating its gate terminal. Applying a positive voltage signal in the gate and the
emitter keeps device in its ‘ON’ state, while making the input gate signal zero results
in ‘OFF’ state. The IGBT characteristics are shown in Fig. 2.29.
The IGBT is presently one of the most popular devices due to its wide ratings,
switching speed of about 100 kHz, easy voltage drive and a square safe operating
area devoid of a second breakdown region.
2.5.5 Silicon-Controlled Rectifier (SCR)
Silicon-Controlled Rectifier is the most popular of the thyristor family of four-layer

regenerative devices [15]. It has three terminals; anode, cathode and a third control
terminal, Gate. SCRs are unidirectional devices as shown in Fig. 2.30.
Fully “on”
Collector current (A)
Active region
Fully “off”
Collector to emitter voltage (V)
Fig. 2.29 Characteristics of IGBT
Fig. 2.30 SCR symbol (left) and component (right)
SCRs are used in power-switching applications. In the, OFF state, SCR restricts
current flow. The SCR characteristics are shown in Fig. 2.31. When the gate-to-
cathode current exceeds a threshold value, the device turns ON and conducts. The
SCR remains the ON state even after gate current is removed so long as the current
through the device exceeds the holding current. If the gate is pulsed and the current
through the device is below latching current, the device will remain in the OFF state.
Holding current is the minimum current which must pass through from anode
to cathode for it to remain in the ‘ON’ state [15]. The minimum anode current is
required to maintain a thyristor in ‘ON’ state immediately after a thyristor has be
triggered ON. SCR can be turned ON by the application of a gate pulse when a
forward bias voltage is present at the main terminals. However, being regenerative
or ‘latching’, it cannot be turned off via the gate terminals, at the extremely high
amplification factor of the gate. The two main types of SCRs are: Converter grade
or Phase Control thyristor and Inverter grade thyristor.
Fig. 2.31 SCR characteristics
Fig. 2.32 Pulse width modulation
2.5.6 Pulse Width Modulators (PWM)
Pulse width modulators are powerful tools for controlling analog circuits with a
microprocessor’s digital outputs. They widen the pulse width or narrow the pulse
width without changing the time as shown in Fig. 2.32. PWM is employed in a wide
variety of applications, ranging from measurement and communications to power
control and conversion [15]. The main advantage of using PWM is the power loss
becomes low in the switching devices.
Fig. 2.33 Microprocessor chip
2.5.7 Microprocessor
Microprocessor is an Integrated Circuit (IC) (shown in Fig. 2.33), which incor-

porates core functions of a computer’s Central Processing Unit (CPU) [15]. It is a
programmable multipurpose silicon chip, clock driven, register based, accepts binary
data as input and provides output after processing it as per the instructions stored in
the memory.
2.5.8 Microcontroller
Microcontroller is a computer present in a single integrated circuit (shown in

Fig. 2.34) which is dedicated to perform one task and execute one specific applica-
tion. It contains memory, programmable input/output peripherals as well a processor
[15].
2.5.9 Field-Programmable Gate Arrays (FPGAs)
FPGAs are semiconductor devices that are based around a matrix of Config-
urable Logic Blocks (CLBs) connected via programmable interconnects as shown
in Fig. 2.35. FPGAs can be reprogrammed to desired application or functionality
requirements after manufacturing [15].
Fig. 2.34 Microcontroller board
Fig. 2.35 FPGA

2.6 Evolution of Arc Welding Power Sources 41
2.6 Evolution of Arc Welding Power Sources
A literature review, based on evolution of welding power source is conducted to build

a framework of this book. The electrical components such as resistors, inductors,
transformers and capacitors influence the efficiency, product cost, weight, power
density, size and volume of the finished power source.
The solid-state devices such as thyristor and bipolar transistor replaced the vacuum
tubes during the 1950s by increasing operating frequency and power rating demands
of semiconductors. In 1969, based on the static and dynamic characteristics, the per-
formance of the welding parameter is established [16]. E.H Dagget designed a power
source in which dynamic characteristics are measured for obtaining good arc stabil-
ity [17]. Bipolar transistors came into use as switching devices in the power circuit
design. However, in those periods, importance was given to snubbing technology,
switch technology and design optimization of weld transformer and protection [18].
The conventional power sources used 50 Hz operating transformer for converting
low current and high voltage of the main power supply to high-current and low-
voltage welding power. With the development of power-switching devices such as
transistors, MOSFETs and IGBTs [5].They find prominence in inverter-based weld-
ing power sources. In 1980, power sources were introduced with feedback control
systems and supported welding process automation [19].
In 1987, J.F. Lancaster et al., introduced a power source with feedback control
systems which allows constant current or constant voltage operation. Pan and Zhang
considered the use of transistorized power sources to improve weld quality. Dorn
and his colleagues, in 1982, developed a technique for plotting instantaneous values
of voltage or power against current for a test weld, thus obtaining a direct picture
of the variability of the operation [20]. Analysis on electrode behaviour by using
pure AC Supply is done [21] which reveals that the welding power source at short-
circuit period has strong correlation with arc stability [22]. Then condenser voltage
multiplier is introduced in power sources [23]. In 1998, M.J.M. Hermans and G.
den Ouden, derived a model using transistorized power source and observed that
electrode melting rate values and heat flow to the work piece are consistent [24].
With the advancements in digital control, control techniques are implemented using
high-performance PCs.
In the beginning of the twenty-first century, more work was published on power
source with linear and nonlinear regression models [25]. Guo-rong et al. presented the
dynamic characteristics of power sources [26]. The high operating speed and univer-
sal nature of static compensators had wider applications [27]. After performing com-
parison with the existing power sources of that period, it is observed that transformer
thyristor rectifier source maintains constant voltage stabilization of welding condi-
tions in the absence of main power supply. In case of short circuiting time, the current
form is effectively reproduced [28]. Controlled resonant inductor technology is used
in power source. Low-voltage and high-current arc welding inverter power supply
achieved the Zero-Voltage Source (ZVS) and reduced the loss of duty cycle of the
phase-shift full-bridge converter. Thus, the efficiency of the power supply is increased
and switching losses are decreased [29]. The power source is enhanced with single
phase welding current based on new energy conversion methods by replacing tradi-
tional transformer-rectifier systems [30]. Then fuzzy-based power sources are sug-
gested. The designed power source is implemented with the phase-shift full-bridge
Zero-Voltage Zero-Current Source (ZVZCS) [31]. Phase-shift full-bridge topology
is used in the high-power arc welding/cutting inverter power source that consisted of
three control methods; Sliding mode control, PID control and fuzzy control [32]. The
control technology widely used in inverter welding consists of full-bridge-isolated
pulse width modulation providing high current, low voltage and high power. The
Zero-Voltage Switching (ZVS) is realized easily by leading leg however the lag-
ging is difficult, that increases switching loss and Electromagnetic Inference (EMI)
[29]. Lebedev, V.A., et al., analysed the single-phase welding current source using
transformers and condenser [33].
To improve the blocking voltage capability of semiconductor devices, two meth-
ods are defined. The first method is to combine MOS and bipolar physics and the
second one is Wide-Bandgap (WBG) semiconductor material that replaces the Si
material. IGBT came out from the first method which is widely employed in power
electronics [34]. Second method with WBG material, where the silicon is replaced
by the silicon carbide in the power devices to reduce the resistance in drift regions.
The benefits of silicon carbide devices are demonstrated in [35] and have realized
the desired features such as high-power density and lightweight in medium voltage
application [36]. The optimal selection of welding current and arc voltage deter-
mines the quality of weld joints [37], which contributes to study of the power source
characteristics of power source under dynamic and static operating conditions. By
arriving at optimal parameter ranges for welding, the parameter settings are done in
the power sources. A reliable method based on Artificial Neural Networks (ANN) is
developed for detection of magnetic core saturation in welding transformers, with no
additional sensor. The proposed ANN is trained by the resilient backpropagation rule
using samples, then the dynamic model of the welding system and its measurements
are obtained [38]. The invention of high-speed solid-state devices leads to greater
saving of energy in the welding processes.
Later an arc welding Switched Mode Power Supply (SMPS)-type power source
is developed using simulation software PSIM, the proposed converter is built and
verified by experimental prototype measurements [1]. In the IGBT-type power
source used in electron beam welding, transient behaviour of the system is enhanced
by altering the characterization circuit and stability of the system [39]. Inverter
power supply for welding achieves the Zero-Voltage Source (ZVS). The controlled
resonant inductor technology is employed which reduced the loss of duty cycle
of the phase-shift full-bridge converter. The efficiency of the power supply is
increased, and the switching losses are decreased [29]. The current trends in various
welding methods are reported by JFE [40]. Recently, the performance of welding
parameters: Probability Density Distributions (PDDs), artificial neural network
comprising of Self-Organizing Maps (SOM), voltage and current is analysed using
inverter and generator power source [41]. The list of papers on evolution of welding
2.6 Evolution of Arc Welding Power Sources 43
1400
Min Max
1200
1000
800
Amps
600
400
200
0
1980 1985 1990 1995 2000 2004 2010 2014 2016
Year
Fig. 2.36 Year-wise variation of average welding current in the welding processes
power source is presented in the following table. The year wise variation of average
welding current is plotted in Fig. 2.36. Table 2.1 presents various papers pertaining
to the evolution of arc welding power sources.
2.7 Switch-Based Techniques Adopted for Welding Power

Sources
In 1998 Du Dong et al., considered the inverter power source for the robotic arc
welding. The inverter power source shall meet the following conditions:
(i) The power source should have a control property, constant output current reg-
ulation at a wide range and improvise the welding quality.
(ii) The power source should have high duty cycle and rated output current to
ensure a long life time.
(iii) The power source should have electromagnetic-compatible interfaces.
With the implementation of modern electronic power techniques, the IGBT (Isolated-
Gate Bipolar Transistor) inverter has acceptable control properties and high switch-
ing frequency (more than 20 kHz). The IGBT power sources are electromagnetically
compatible, therefore, each unit must have the same phase, frequency (25 kHz) and
synchronized switching. The inverter-type power source system consists of substan-
tial number of independent, parallel IGBT inverter units combined with a controlled
single-chip microcomputer unit. This system has good static and dynamic indices,
100% duty cycle and high-power capacity. They not only meet the welding require-
ment, but in addition its lower cost leads to batch production of inverter power sources
[42]. Instantaneous output current control method is used in these power sources.
The major characteristics of this method are as follows;
Table 2.1 Papers on evolution of power source

Author(s) Developments/remarks Technologies/methodologies
used
Kolasa et al. (1985) [19] Power supplies are suitable for No protection on dynamic
automation welding process behaviour
Lucas and Murch (1987) [21] Arc sustained at 40 V power Square wave power supplies
supply
Lancaster (1987) [20] Attained constant current or Feedback control systems and
constant voltage operation transistorized power sources
with improved weld quality
Mita et al. (1988) [22] Arc stability of the waveform Multiple regression analysis
factors had strong correlation
with arc stability
Schiedermayer (1997) [5] Converted low current, high Transistors, MOSFETs and
voltage to high current, low IGBTs are used in
voltage inverter-based welding power
source
Pentegov et al. (1998) [23] Power source resulted with Condenser voltage multiplier
higher degree of stability, and single-phase rectifiers
welding current, better arc
ignition and a lower power
consumption
Hermans and Den Ouden Electrode melting rate values Transistorized power source
(1998) [24] and heat flow to the workpiece
is consistent and heating of the
electrode and workpiece is
analysed
Kang and Rhee (2001) [25] Reduced spatter at low voltage Linear and nonlinear
regression model
Agunov et al. (2002) [27] Power coefficient is corrected Compensators is used
Lebedev et al. (2007) [33] Optimum control of the Power source with
electrode metal transfer transformers and condenser
process is achieved
Baliga (2008) [34] From MOS and bipolar Combined MOS and bipolar
physics, IGBT came into physics and Wide-Bandgap
existence and using WBG Si is
replaced with SiC to reduce
resistance in drift region
Guo-rong et al. (2008) [32] Digital control logic Phase-shift full-bridge
technology is exact, reliable topology consists of sliding
and flexible mode control, PID and Fuzzy
control
Shklovski and Janson (2012) Experimental Prototype of SMPS-type power source
[1] stabilized power at weld bead using Page Sequence Master
is verified. Interleave (PSIM) software
(continued)
2.7 Switch-Based Techniques Adopted for Welding Power Sources 45

Author(s) Developments/remarks Technologies/methodologies
used
Parthipan et al. (2013) [39] System stability and H bridge IGBT-type power source
inverter topology flashover
detection is done. Fault
recovery control circuit
provided superior performance
and achieved defect-free
welding
Ji et al. (2014) [29] Provided high current, low Full-bridge-isolated pulse
voltage and high power width modulation
Konovalov (2014) [28] Presented different types of Transformer thyristor rectifier
comparison of universal source
inverter-type power sources,
weld joints is analysed
Ji et al. (2014) [29] Low-voltage and high-current Controlled resonant inductor
arc welding inverter power technology
supply achieved the
Zero-Voltage Source (ZVS),
reduced the loss of duty cycle
of the phase-shift full-bridge
converter
and efficiency of the power
supply is increased and the
switching losses are decreased
OI et al. (2015) [40] Recent trends in various Automation technologies
welding methods are
developed independently
Kumar et al. (2016) [41] Performance of welding Using inverter and generator
parameters; Probability power source
Density Distributions (PDDs),
artificial neural network
comprising of Self-Organizing
Maps (SOM), voltage and
current is analysed
(i) When compared to traditional inverter arc welding machine it reduces spatter
formation.
(ii) Short-circuit metal transfer process is stabilized and enhanced the bead state.
The authors proposed a power supply based on the voltage drop caused by leakage
inductance in the transformer. This leakage inductance is proportional to the excita-
tion frequency. The study also reveals that, the inverter drive for a transformer causes
effects in feedback mode. The equivalent load impedance of the inverter drive varies
approximately ±20% more than the variation in sine wave operation. Further testing
is required to improve the controlling methods [43]. Few authors analysed the weld-
ing power source equipped with digital filtering process. Welding power sources
with accurate control grasps instantaneous weld phenomenon, processes the data
ultrafast and performs control in real time. Research and developments are ongoing
in this domain and there are several reports available on weld instability where mul-
tiple abnormal voltage events’ occurrences are introduced as a side effect of digital
filtering [44].
In welding, low-voltage and high-current sources are used with either AC or
DC welding currents as an effective approach. To implement them in low-power
applications, the use of zero-voltage and zero-current switching of the inverter is
introduced [45]. The switching frequency can be increased up to 40 kHz due to the
applied zero-current, zero-voltage switching mode, however the switching losses
remain low. The iron core size of the transformer can be considerably reduced when
the transformer is supplied with increased frequency. Some authors modified the
topologies of the basic DC–DC converter structure and it is observed from the studies
the sensitivity of the system reduces [46]. With the evolution of fast converter and
inverter grade switches, waveform-controlled power sources are developed. The list
of waveform-controlled welding power sources and weldability for different arc
welding processes are illustrated in Tables 2.2 and 2.3 respectively.
In 2009, an inverter power source is designed in which, the increase of volt-
age in the individual elements is restricted. Thus, process of Weld with Modulated
Current (WMC) by double modulation of welding current provides an efficient
control [7]. Later, an inverter-type power source equipped with a microcontroller
(MCF5213CAF8) and FPGA is experimented and presented to attain high reliabil-
ity, fast dynamic response, stable arc length control and good welding quality [47].
Also, the compatibility of the arc welding power source is analysed.
The equipment of the power source are gradually replaced from transformer
devices to rectifier and inverter devices which had many advantages such as pre-
defined sets of parameters, smooth adjustment of welding parameters, smaller and
lighter size. They have also introduced other welding methods such as Cold Metal
Transfer (CMT) and Surface Tension Transfer (STT). The Electromagnetic Com-
patibility (EMC) in power sources show some disadvantages and relatively resistant
to originate interferences from the mains. Using semiconductors devices, the inter-
ference problems are eliminated. Also, the power supply operation is a nonlinear
character due to core saturation, which is due to the harmonic occurrences.
The FPGA digital control strategy is adopted to analyse the voltage, current and
switching frequency. Thus, switching losses occurring are reduced using the phase of
the current in synchronized gate signal by incorporating predictive digital algorithm
[48]. In the following Table 2.4, a review of few switch-based techniques in welding
power source-related papers are presented.
2.7 Switch-Based Techniques Adopted for Welding Power Sources 47
Table 2.2 Waveforms and its applications for different processes of welding
Welding Waveforms Materials used
AC-TIG Sine wave Aluminium, Al alloys,
Soft square wave Stainless steel, Copper
Rectangular wave
Sawtooth wave
Asymmetric (wave shape is
not balanced either positive or
negative)
DC-TIG Square wave Stainless, Mild, Copper Alloy,
Nickel alloy, Titanium alloy,
Low alloy
AC/DC MMA Square wave Steel, Iron, Aluminium,
Nickel, Copper alloy
AC SAW Sine wave Carbon steels, Low-alloy
Square Wave steels, Stainless steels,
Nickel-based alloys
DC SAW Square wave Steel, Nonferrous metals,
Aluminium
AC-GTAW Square wave Aluminium, Copper alloys,
Magnesium alloys
DC-MIG Triangular wave Carbon steel, Stainless steel,
Aluminium, Thin sheet metals
DC-RSW Square wave Aluminium, Magnesium,
Sine wave Inconel, Nickel, Nickel silver,
Monel, Low-carbon steel,
Low-alloy steel, Stainless steel
AC/DC CMT Sine wave Cold Rolled Steel, Galvanneal,
Square wave Stainless steel, Aluminium,
and Copper
Table 2.3 Weldability of ferrous and nonferrous alloys

Ferrous and Weldability
nonferrous
alloys
MMA TIG MIG FCAW SMAW SAW
Carbon Yes Yes Yes Yes Yes Yes
steel
Stainless Yes Yes (DC) Yes Yes Yes Yes
steel
Cast iron Yes
Nickel alloy Yes Yes Yes
Aluminium Yes Yes (AC) Yes Yes Yes Yes
Copper Yes Yes (DC) Yes Yes Yes
Magnesium Yes (AC)
Titanium Yes (DC)
Table 2.4 Switch-based techniques in welding power source
48
Author(s) Switch-based methodologies Block diagrams and inference
Dong et al. (1998) [42] Power source with several IGBTs

connected in parallel
Single-chip microcomputer is regulated in individual power unit consisting of parallel connected IGBTs to make the power source system to work over
a wide range of continuous output current with good dynamic properties
Low cost, good dynamic and static indices, high-power capacity and 100% duty cycle is produced
Chae et al. (1999) [58] Output current control using
single-chip microcontroller
Inverter circuit topology is implemented to improve welding performance, reduced spatter and improved bead state
2 Insight into Arc Welding Power Source Terminologies
(continued)
Takasaki and Sonoda (2005) Current controllability

[59]
Inverter frequency is varied to easily control load current. The use of transformer leakage inductance for current control for obtaining low-voltage and
high-current power sources
Uezono & Hongjun (2008) Digital filtering process with digital
[44] calculation technology
2.7 Switch-Based Techniques Adopted for Welding Power Sources
For abnormal voltage events, the optimal value is different depending on the material. Digital filtering removes the noise in real time and proved its
advantages in manufacturing efforts
(continued)
49
50
Beno Klopčič et al. (2008) Advanced hysteresis control (AHC)

[60]
AHC adjusts itself automatically to any change in iron core size and in the value of applied DC bus voltage. Proposed AHC assures a very short rise
time of the welding current and best possible utilization of transformer iron core
(continued)
Knyaz’kov et al. (2009) [7] Modulated current technique

Control algorithms of the power section in the IPS enables efficient control
(continued)
51
52
Pang and Zhang (2010) [47] Microcontroller

(MCF5213CAF80) and
FPGA-based control system
The software design of FPGA and software design of operating system µC/OS-II on MCF5213CAF80 are proposed which overcomes the
disadvantages of analogous control technology and single-chip control. Microcontroller and FPGA-based control system for MIG welding has faster
response, better reliability and more stable arc length
(continued)
Skrzyniecki et al. (2013) [45] Monitoring THD in inverter power

source
The detailed analysis on impact of sub-harmonics and inter-harmonics on a power system is done. Efficient detection of system parameters is performed
by monitoring the distortion introduced into the system
(continued)
53
54
Navarro-Crespin and Zero-block crossing control and

Azcondo (2015) [48] predictive digital algorithm
The algorithm generates two complementary signals to drive the 4QSWs, required to synchronize the switches and the output current of the transformer.
Switching losses on 4QSW are minimized by the proper synchronization of the gate signals with the phase of the input current using a predictive digital
algorithm. Phase lags are measured by zero-block crossing
(continued)
Polloclk and Flower [61] Convertor using IGBTs
Delivered output voltage of 25 V and output current of 200A. Their operation at high frequencies has minimized the size, weight and reduced current
ripples. The power source is controlled using the control logic developed using Xilinx logic cell array
Janson and JWik [62] Series-to-parallel switching of
inductor and capacitor in
oscillatory circuits
AC-DC resonant is developed. The reactive power of the inductor and capacitor compensates each other in both parallel and series connection so the
power factor is maintained unity. There are no requirement for filters and reactive power compensative circuits
55
(continued)
56
Oshaben designed [57] IDDD DC–DC welding convertor
Low EMI. Most overlooked features of this design are minimal engineering time and cost-effectiveness. As a future work, the author recommended
building and testing of a prototype with further investigations on size and economic design
Shklovski and Janson [63] Switch-mode load-resonant
converter using PSIM
The authors proposed that inherent short-circuit current limitation and a fast-parametrical response to the load would result in good weldment. It has
wide operating range from no-load to short-circuit without overloading, high power factor, THD of 0.1 PF between 0.99 and 1, and low transient
overshoots
(continued)
Narula et al. [64] (PFC) two stage AC-DC converter
The proposed circuit consists of a bridgeless (BL) zeta converter in the front and a full bridge (FB) buck converter for high-frequency isolation. This
converter eliminated conduction losses and provided power factor correction
Narula et al. [65] Modified power factor
(PF)-corrected single stage and two
switches-isolated zeta converter
The proposed converter operates in a discontinuous inductor current mode (DICM) for achieving good PF correction and reduced control complexity
(continued)
57
58
Mecke et al. [66] ZVZCS-PS-FB-Convertor
A closed-loop adaptive control strategy to minimize switching losses for a lightweight ZVZCS-PS-FB-Convertor built with IGBTs and MOSFETs
Snubber capacitors are used to reduce switch turn-off loss and saturable inductors are used to prevent discharge of the blocking capacitor and to limit
the current rise
Chae et al. [67] Arc welding inverter
Proposed a new control algorithm for instantaneous current control using a single-chip microprocessor for CO2 arc welding inverter to achieve proper
metal transfer, reduced spattering and good digital control
(continued)
Ji-qiang et al. [68] Inverter power supply made of

IGBTs
A three-phase welding inverter power supply made of IGBTs with power factor correction and validated for high quality and unity power factor. The
proposed design aims to minimize inactive power flow problems when the power source works under open-load condition. The author suggested further
research to improve the performance characteristics when the power source is used for arc GTAW and GMAW
Zhu et al. [69] Constant current IGBT-based
inverter-type power source
A constant current IGBT-based inverter-type power source for CO2 arc welding using LPC213-embedded ARM unit to realize output waveform control
digitally. The developed power source utilizes PWM control to adjust the waveform parameters, which maintained a stable welding condition and
improves the welding performance. The author suggested to further investigate and add devices in the main circuit and develop high-frequency power
source to improve the welding property
(continued)
59
60
De Li et al. [70] IGBT-based switching-inverter

power source
A fuzzy adaptive PID control and PWM control algorithm for IGBT-based switching-inverter power source for obtaining smooth, constant DC output.
The proposal aimed to overcome nonlinearities and time variations and improve the steady-state precision and dynamic performance of the system
2.8 Literature Addressing Power Source Parameters 61
Fig. 2.37 Probability distribution of short-circuit period with CO2 , Ar/CO2 , and TIME shielding
gases. I 90 A, U 21 V (as illustrated in [57])
2.8 Literature Addressing Power Source Parameters
Based on the review of parameters associated with power sources, the following
points are discussed.
Suban and Tušek [6] have emphasized that poor arc stability causes spattering
which results in material losses, production time wastage, anaesthetic appearance
on workpiece and quality deterioration. The experimental study revealed a stable,
spatter-free weldment when carbon dioxide shielding gas (82% Ar/18% CO2 gas
mixture) is used for short-circuit metal transfer. TIME mixture (gas mixture of 65%
Ar/26.5% He/8% CO2 /0.5% O2 ) for spray transfer. (Variation of the curves during
the short-circuit periods are smaller than in the two other cases as in Fig. 2.37).
Palani and Murugan [49] have performed a study on pulsed current GMAW. The
study aimed at selection of optimal parameters such as average current, peak current,
peak time, base current, feed rate, frequency and shielding gas based on trial and error
experimentation. The review reveals that only a few researchers have used Design of
Experiments (DOE) to perform experiments for selecting the pulse parameters and
to study their effect on weldment.
A. K. Paul [50] has analysed the power electronic techniques and technologies for
the optimal performance of arc welding process. The analysis concluded that power
electronic technology is beneficiary in minimizing the energy usage, maximized
productivity, enhanced quality and minimal environmental impact.
Skrzyniecki et al. [51] performed an experimental study on power source stability
for MIG/MAG welding processes using LabVIEW software and a computer. The
computer-controlled process resulted in good weldment. The author emphasizes on
the choice of suggesting advanced methods for recording and online monitoring of
data.
Kolasa et al. [52] conducted an experimental study on dynamic characteristics
of TIG arc welding that is supplied with square and triangular current waveforms
as shown in Figs. 2.38 and 2.39. Two anodes; steel molten anode and water-cooled
copper anode are utilized. The outcomes revealed that for very low frequencies arc
voltage coincides with static characteristics, and at high frequencies, the arc dynamic
Fig. 2.38 Difference in

voltage response waveforms
in steel molten anode and
water-cooled cathode for
triangular waveform (as
illustrated in [52])
characteristics becomes like that of a constant resistance. It is observed that the square
waveform of current is optimal.
Lucas and Murch [21] conducted an experimental study on reduction of OCV by
developing a two new square wave power sources; squared AC and switched DC.
The experiments indicated that welding arc is sustained at 62 V on squared sine wave
and 40 V on square wave as shown in Figs. 2.40 and 2.41 respectively. The developed
power source had a greater current change rate during polarity reversal with reduced
OCV requirements.
Jinhong et al. [53] developed an optimal nonlinear model of control system for
arc welding inverter using MATLAB and proposed a Computer-aided analysis of
dynamic process. The dynamic process of open circuit to short circuit is simulated
and it is proven as an effective approach.
Yamane et al. [54] proposed a new power source characteristic to improve the arc
stability to avoid spattering. Experiments are carried out to compare the nonlinear
power source characteristic with the conventional characteristic under 100% CO2
shielding gas. The conventional power source characteristic revealed the occurrence
of unregulated metal transfer and spattering as shown in Fig. 2.42. While using
of nonlinear power source characteristic, the spatter is reduced, and arc stability is
improved as shown in Fig. 2.43.
Fig. 2.39 Difference in

voltage response waveforms
in steel molten anode and
water-cooled cathode for
square waveform (as
illustrated in [52])
Devakumaran et al. [55] preformed a study on the power source characteristics

involving the response to voltage and current variation under different modes of
the power source operations such as conventional GMAW, Pulsed gas Metal Arc
Welding (P-GMAW), Dual Pulse Gas Metal arc Welding (DP-GMAW) and Modified
Short-Circuit Gas Metal Arc Welding (MS-GMAW) processes. All the investigations
are carried out under static and dynamic operating conditions. It is observed that
MS-GMAW process provided better weld quality and less spatter generation when
compared with other processes as shown in Fig. 2.44.
Skrzyniecki et al. [56] conducted a study on static and dynamic characteristics of
welding power sources to analyse the welding circuit in terms of its electrical param-
eters such as inductance, resistance and capacitance which dynamically changes
during welding. Under static characteristics as shown in Fig. 2.45, the power source
is loaded with a reducing resistance and the welding parameters are measured and
recorded using computer. Under dynamic characteristics as shown in Fig. 2.46, alter-
ations to load are performed to provide information about the dynamics of the tested
object and current and voltage are recorded.
Fig. 2.40 Voltage and

current waveforms of
positive and negative
re-ignitions when welding at
40 V OCV on a square wave
supply [21]
Fig. 2.41 Voltage and

current waveforms of
positive and negative
re-ignitions when welding at
62 V OCV on a sine wave
supply [21]
Fig. 2.42 Waveforms of

voltage and current obtained
using conventional power
source [54]
Fig. 2.43 Waveforms of

voltage and current obtained
using nonlinear power
source [54]
Fig. 2.44 Typical photographs of surface appearance of weld bead with respect to spatter generation
under different welding processes of a MS-GMAW, b GMAW, c P-GMAW, and d DP-GMAW [55]
Fig. 2.45 Static characteristics recorded for power source [56]

References 67
Fig. 2.46 Dynamic characteristic recorded for power source during welding with controlled load
[56]
References
1. J. Shklovski, K. Janson, Development of constant-power source for arc welding, in 13th Bien-
nial Baltic Electronics Conference, pp. 255–258 (2012)
2. H. Latifi, R. Suoranta, J. Martikainen, M. Pirinen, P. Kah, Usability of arc types in industrial
welding. Int. J. Mat. Eng. 9(1), 1–12 (2014)
3. T.W. Eagar, The Physics of Arc Welding Processes (1990), pp. 17–19
4. J.N. DuPont, A.R. Marder, Thermal efficiency of arc welding processes. Weld. Res. Suppl.
406–416(1995)
5. M. Schiedermayer, Improving reliability of inverter based welding machines. Weld. J. 76(2)
(1997)
6. M. Suban, J. Tušek, Methods for determination of arc stability. J. Mat. Process. Technol.
143–144, 430–437 (2003)
7. A.F. Knyaz’kov, S.A. Knyaz’kov, K.I. Dementsev, An inverter power source for welding with
modulated current. Weld. Int. 23(12), 957–962 (2009)
8. V. Agelidis, O. Anaya-Lara, T. Miller, E. Acha, Power Electronic Control in Electrical Systems
(Elsevier, India, 2002)
9. Arc Welding Power Sources:NPTEL online course (2017). Available: http://nptel.ac.in/course
s/112107090/module3/lecture1/lecture1.pdf
10. T.E. GROUP, Basics of Arc Welding. Available: http://www.esabna.com/euweb/awtclesson1_
21.htm. Accessed 2017
11. Pintu, Arc Welding Polarity—Types, Effects, Selection and Pros & Cons. Avail-
able: http://www.minaprem.com/physics-of-welding/arc-welding-polarity-types-effects-selec
tion-pros-and-cons/. Accessed 2017
12. T.L.E. Company, Weld Penetration Variables: Lincon Electric (1999). Accessed 2017
13. W.S. Service, Maritime Welding Handbook (2010)
14. M. Electric, Guidelines to GTAW process (2003). Available: http://manuals.chudov.com/Mill
er/Miller-TIG-Handbook.pdf. Accessed 2017
15. M.H. Rashid, Power Electronics Handbook (Elsevier)
16. A.F. Manz, Inductance vs. Slope Control for Gas Metal Arc Power, vol. 9 (1969), pp. 707–7012
17. E.H. Dagget, A power supply for pulsed power welding. IEEE Spectr. 1, 67–70 (1968)
18. J.D. Van Wyk, J.A. Ferreira, Transistor invertor design optimization in frequency range above
5KHz up to 50 KVA. IEEE Trans. Ind. Appl. 2, 296–302 (1983)
19. A. Kolasa. A. Matsunawa, Y. Arata, Experimental study on dynamic properties of power sources
for MIG/MAG Welding. Trans. JWRI, 2, 255–265 (1985)
20. J.F. Lancaster, The physics of fusion welding Part-1: the electric arc welding. IEEE
Proc.—Electric Power Appl. 134(5), 233–254 (1987)
21. W. Lucas, M.G. Murch, Arc reignition characteristics when welding with sine-wave and square-
wave power supplies. IEEE Proc.—Electr. Power Appl. 134(86), 348–354 (1987)
22. T. Mita. A. Sakabe, T. Yokoo, Quantitative estimates of arc stability for CO2 gas shielded arc
welding. Weld. Int. 2(2), 152–159 (1988)
23. I.V. Pentegov, S.V. Rymar, V.P. Latanskij, Optimisation of welding parameters in power sources
with condenser voltage multipliers. Weld. Int. 12(2), 89–91 (1988)
24. M.J.M. Hermans, G. Den Ouden, Modelling of heat transfer in short circuiting gas metal arc
welding. Sci. Technol. Weld. Joining 3(3), 135–138 (1998)
25. M.J. Kang, S. Rhee, The statistical models for estimating the amount of spatter in the short
circuit transfer mode of GMAW. Weld. J. (USA). 80(1) (2001)
26. Z. Jinhong, L. Wenlin, S. Yaowu, Study on the dynamic process of arc welding inverter. in
Proceedings of the Power Electronics and Motion Control Conference, 2000. IPEMC 2000,
vol. 1, pp. 308–322 (2000)
27. A.V. Agunov, M.V. Agunov, G.M. Korotkova, V.I. Stolbov, A.A. Shevtsov, Energy character-
istics of the power source-welding arc system. Weld. Int. 16(12), 966–969 (2002)
28. Y.N. Konovalov, Comparison of the properties of the MAGMA-350 universal inverter power
source for the welding arc with traditional rectifiers for mechanized welding. Weld. Int. 28(4),
317–319 (2014)
29. J. Ji, X. Hu, Z. Hua, G. Zeng, L. Guo, Research of soft switching arc welding inverter power
supply with high-frequency and high-power. Int. Power Electron. Appl. Conf. Exposition, 924
(2014)
30. V.A. Lebedev, V.S. Romanyuk, Single-phase welding current power sources for mechanised
carbon dioxide welding. Weld. Int. 18(6), 489–493 (2004)
31. G.R. Zhu, Z. Liu, X. Li, B.Y. Liu, S.X. Duan, Y. Kang, Research on digital soft-switch weld-
ing/cutting inverter power source, in IEEE International Conference on Power Electronics and
Drive Systems, vol. 7 (2007), pp. 325–329
32. Z. Guo-rong, L. Zhao, Z. Ai-yun, Y. Mi, S.X. Duan, Y. Kang, Sliding mode control and PI control
for arc welding/cutting inverter, in IEEE International Conference In Industrial Technology
(2008), pp. 1–4
33. V.A. Lebedev, A.V. Motrii, A.D. Glushchenko, N.I. Postolatii, Experimental examination of
welding current sources for semiautomatic welding equipment poared from 220 V single-phase
mains. Weld. Int. 21(6), 454–457 (2007)
34. B.J. Baliga, Power Mosfets—In Fundamentals of Power Semiconductor Devices (2008)
35. J.S. Glaser, J.J. Nasadoski, P.A. Losee, A.S. Kashyap, K.S. Matocha, J.L. Garrett, L.D. Ste-
vanovic, Direct comparison of silicon and silicon carbide power transistors in high-frequency
hard-switched applications, in Applied Power Electronics Conference and Exposition (APEC),
pp. 1049–1056 (2011)
36. R.C.S. Machado, J.C. Braz Filho, Next generation arc welding machines based on Silicon
Carbide MOSFETS and high frequency planar magnetics, in IEEE 13th Brazilian Power Elec-
tronics Conference and 1st Southern Power Electronics Conference (2015)
37. K. Devakumaran, P.K. Ghosh, Thermal characteristics of weld and HAZ during pulse current
gas metal arc weld bead deposition on HSLA steel plate. Mat. Manuf. Process 25(7), 616–630
(2010)
38. K. Dezelak, J. Pihler, G. Stumberger, B. Klopcic, D. Dolinar, Artificial neural network applied
for detection of magnetization level in the magnetic core of a welding transformer. IEEE Trans
Mag. 2, 46 (2010)
References 69
39. T. Parthipan, C. Ribton, P. Mudge, R. Nilavalan, W. Balachandran, Enhancement of high voltage

electron beam welding power supply: Rapid recovery after flashover detection for void-free
welding, in IEEE Proceedings (2013)
40. K. Oi, M. Murayama, Recent trend of welding technology development and applications. JFE
Technical Report (2015)
41. V. Kumar, N. Chandrasekhar, S.K. Albert, J. Jayapandian, Performance analysis of arc welding
parameters using self-organizing maps and probability density distributions, in IEEE Confer-
ence on Control Systems (2016)
42. D. Dong, H. Zandong, X. Ping, Z. Qian, Structure and control of an inverter type power source
for robot arc welding, Tsinghua Sci. Technol. (1998) pp. 1026–1028
43. Y. Takasaki, T. Sonoda, Current controllability of the low-voltage 10 kA inverter power sources.
IEEE Trans. Mag. 10, 4054–4056 (2005)
44. T. Uezono, T. Hongjun, Application to MIG welding using welding power source equipped
with digital filtering process. Weld. Int. 299–303 (2008)
45. K. Skrzyniecki, P. Cegielski, A. Kolasa, A. Krajewski, Electromagnetic compatibility of power
supplies for arc welding. Weld Int. 27, 623–628 (2013)
46. B. Klopcic, D. Dolinar, G. Stumberger, Advanced control of a resistance spot welding system.
IEEE Trans. Power Electron. 23(1), 144–152 (2008)
47. Q. Pang, M. Zhang, Design of digital control system for pulsed MIG welding power source, in
Intelligent Control and Automation (WCICA) 8th World Congress (2010), pp. 2492–2495
48. A. Navarro-Crespin, R. Casanueva, F.J. Azcondo, Alternating current welding using four quad-
rant switches, in IEEE Applied Power Electronics Conference and Exposition (APEC) (2015),
pp. 3330–3334
49. P.K. Palani, N. Murugan, Selection of parameters of pulsed current gas metal arc welding, J.
Mater. Process. Technol. 172, 1–10 (2006)
50. A.K. Paul, Power electronics help reduce diversity of arc welding process for optimal per-
formance, in 2010 Joint International Conference on Power Electronics, Drives and Energy
Systems & 2010 Power India (2010), pp. 4–10
51. K. Skrzyniecki, P. Kołodziejczak, P. Cegielski, Experimental studies on stability of power
source—ARC, in IEEE conference (2013) pp. 359–362
52. A. Kolasa, A. Matsunawa, Y. Arata, Dynamic characteristics of variable frequency pulsed TIG
arc. Trans. JWRI (1986)
53. J. Bo, Study on the dynamic process of arc welding inverter, Polymer (Guildf) (1992)
pp. 783–786
54. S. Yamane, S. Xiang, Y. Kaneko, K. Oshima, Effect of power source characteristic on CO2
short circuiting arc welding_TF. Sci. Technol. Weld. Join 10, 281–286 (2005)
55. K. Devakumaran, N. Rajasekaran, P.K. Ghosh, Process characteristics of inverter type GMAW
power source under static and dynamic operating conditions. Mater. Manuf. Process 27,
1450–1456 (2012)
56. A. Kolasa, P. Cegielski , K. Skrzyniecki, Study of static and dynamic characteristics of welding
power source-arc systems. Weld. Int. 29, 865–867 (2015)
57. E.J. Oshaben, DC-DC Power Converter Design For Application In Welding Power Source For
The Retail Market (2005)
58. Y.M. Chae, J.S. Gho, H.S. Mok, G.H. Choe, W.S. Shin, A new instantaneous output current
control method for inverter arc welding machine, in Power Electronics Specialists Conference
(1999), pp. 521–526
59. Y. Takasaki, T. Sonoda, Current controllability of the low-voltage 10 kA inverter power sources.
IEEE Trans. Mag. 4054–4056 (2005)
60. B. Klopcic, D. Dolinar, G. Stumberger, Advanced control of a resistance spot welding system.
IEEE Trans. Power Electron. 144–152 (2008)
61. H. Pollock, O. Flower, Series-parallel load-resonant converter for controlled-current arc weld-
ing power supply. IEEE Proc. 3 (1996)
62. K. Janson, J. Jarvik, Load adapting mains frequency resonant converters for supplying electrical
arc—a new way in power electronics. EEE Annu. Power Electron. Spec. Conf. 2, 2090–2096
(1998)
63. J. Shklovski, K. Janson, Development of constant-power source for arc welding. Proc. Bienn.
Balt. Electron. Conf. BEC, 225–258 (2012)
64. B. Singh, G. Bhuvaneswari, S. Narula, PFC bridgeless converter for welding power supply
with improved power quality, in IEEE International Conference on Power Electronic Drives
Energy System PEDES (2014)
65. B. Singh, G. Bhuvaneswari, S. Narula, Power factor corrected welding power supply using
modified zeta converter. in IEEE Conference (2016) pp. 617–625
66. W. Fischer, F. Werther, H. Mecke, Soft switching inverter power source for arc welding, in
EPE’79, Trondheim, vol. 1 (1998) pp. 333–337
67. Y.M. Chae, J.S. Gho, H.S. Mok, G.H. Choe, W.S. Shin, A New Instantaneous Output Current
Control Method for Inverter Arc Welding Machine (1999), pp. 0–5
68. C. Shu-Jun, Y. Shu-Yan, W. Dong-Ping, Z. Hua, H. Ji-qiang, Novel three-phase welding inverter
power supply with high power factor, pp. 1113–1118 (2003)
69. J. Zhu, H. Shi, K. Lu, K.Y. Lee, Study on IGBT inverter power supply for CO2 arc welding
and embedded arm based waveform control, in International Conference on Mechatronics and
Automation, 2007. ICMA 2007, pp. 2634–2638
70. S.Z. Wei, W. Qiong, X. Peng, J. De Li, A Switching-inverter power controller based on fuzzy
adaptive PID, in Proceedings of the 6th International Forum on Strategic Technology, IFOST
2011 (2011), vol. 2, pp. 695–699
Chapter 3
Control Terminologies and Schemes
for Arc Welding Processes
A control system interconnects various components of a system to provide a desired

response. The control theory provides a foundation for analysing a system, which
indicates cause–effect (input–output) relationship for various elements involved in a
system. A control system performs the following three key tasks:
(i) Measurement of original output;

(ii) Comparison of original output to reference value;
(iii) Adjustment to minimize error (difference between original output and reference
value).
3.1 Control System Terminologies
Some of the fundamental concepts involved in control system are presented in this
chapter.
3.1.1 Process
A process is a continuing operation by a series of gradual changes that succeed

one another in a relatively fixed way and lead towards certain result.For example,
arc welding shown in Fig. 3.1 is a process that creates an electric arc between an
electrode and the base material to melt the metals at the welding point.

https://doi.org/10.1007/978-981-13-0806-2_3
72 3 Control Terminologies and Schemes for Arc Welding Processes
Electrode holder
Electrode cable
Electrode
Filler metal
Welding machine
AC or DC Power
Arc source
Work
Molten weld Solidified weld

metal metal
Fig. 3.1 Arc Welding process
3.1.2 System
When several elements are connected in a sequence to perform a specific function, it

is called as system.For example, an arc welding process is a system comprising var-
ious elements like power source, connecting cables, electrode, workpieces, clamps,
holders and sensors, deployed for material joining.
3.1.3 Control System
In a system, the regulation of output parameter by varying its input parameter, in

a system to accomplish the desired result is called as control system. Example: In
Constant Current (CC) type power source for a Gas Metal Arc Welding Process
(GMAW) welding process, the control system plays a vital role in maintaining a
constant welding current.
3.1.4 Parameters/Variables
Parameters/variables are entities that are measured and controlled. Example: In arc
welding, welding current, arc length, arc voltage, depth of penetration and bead width
are few key parameters.
(i) Controlled Variable: It is the quantity or variable that is measured and controlled.
Example: Maintaining constant arc voltage in a Constant Voltage (CV) type
power source. Here, arc voltage is the controller variable.
3.1 Control System Terminologies 73
(ii) Manipulated Variable: It is the quantity or variable that is varied by the controller
to alter the value of the controlled variable. Example: To maintain a constant
voltage (controlled variable) in CV type power source, the gating signal applied
to a thyristor for modulating the pulse width is the manipulated variable.
3.1.5 Control
Control refers to measurement of controlled variable of the system and applying the
manipulated variable to the system to correct or limit deviation of the measured value
from a desired value.
3.1.6 Disturbances
A disturbance is an external signal that tends to adversely affect the value of the output
of a system. Example: Sudden change in arc length due to environmental effects
(operator mishandling, mechanical vibrations, etc.) are considered as disturbance.
3.1.7 Setpoint
Fixed value of process parameter as desired is called as setpoint. Example: Before

performing welding, the operator sets a value of welding current, say 70A in a CC
type power source which is called as the set point.
3.1.8 Feedback
Feedback is an integral part of control system that permits the output dynamic
response to be altered without modifying the system components. The process param-
eter is measured using measuring devices called sensors or transducers. Example: In
arc welding process, the welding current is controlled after measuring it with a hall
effect sensor. The current is measured using a hall effect sensor and its feedback to
the input and the error is calculated.
3.1.9 Error
The deviation of the measured value from setpoint is called as error. Example: If
the operator sets a current value of 70 A in a CC type power source and if the hall
effect sensor measurement indicates 68 A of current, then the error is the difference
between both which is 2 A. This error signal is fed to the controller.
3.1.10 Transfer Function
Any physical system can be represented in the form of a differential equation denoted
by
dn c(t) dn−1 c(t) dm r (t) dm−1 r (t)

an + a n−1 + · · · + a 0 c(t) bm + bm−1 + · · · + b0 r (t)
dt n dt n−1 dt m dt m−1
(3.1)
After taking Laplace transform, the above time domain equation, we can obtain the
transfer function. The transfer function of a linear, time-invariant system is defined
as the ratio of the Laplace transform of the output (response function) to the Laplace
transform of the input (driving function) with zero initial conditions. The general
representation of transfer function is shown

C(s) bm s m + bm−1 s m−1 + · · · + b0
G(s) n (3.2)
R(s) an s + an−1 s n−1 + · · · + a0
C(s) Output polynomial of the system.

R(s) Input polynomial of the system.
Roots of the numerator and denominator polynomials are called as zeroes and
poles, respectively.
3.1.11 Open Loop System
A process in which the output parameter has no effect on the input parameter is called
as an open loop system. It is a system without feedback and does not automatically
correct the variation in output as shown in Fig. 3.2.
Example: A current limit sensor is an open loop system. The sensor gets activated
when the current exceeds a prefixed limit and it starts activating buzzer. The system
needs human interaction to intervene and stop the buzzer (as shown in Fig. 3.3).
3.1 Control System Terminologies 75
Setpoint Controller Process Output
Fig. 3.2 Open loop systems
Power Arc welding Current

Controller Buzzer
Source process limit sensor ON/OFF Switch
Fig. 3.3 Example of open loop system
error
Setpoint +
Controller Process Output
-
Feedback
Fig. 3.4 Closed-loop system
Setpoint
error
current +
Arc welding
Controller Output
process
-
Hall effect sensor

(measures current)
Feedback
Fig. 3.5 Example of closed-loop system
3.1.12 Closed-Loop System
A process in which the output parameter alters the input parameter to achieve the
set point is called as closed-loop system as shown in Fig. 3.4. The output signal is
measured and feedback to the input. Hence, the closed-loop control system is often
referred as feedback control system.
A feedback control system tends to maintain a relationship between one parameter
and another parameter by comparing both the values. Their difference is fed as error
signal to the controller. Example: In a CC type power source for a GMAW process,
after a hall effect sensor measures the welding current, the difference between mea-
sured current and setpoint current is calculated and the corresponding error signal is
fed back to the controller (as shown in Fig. 3.5).
3.2 Control System Analysis
There are two methods to analyse the functioning of a control system; time domain
analysis and frequency domain analysis.
(i) In time domain analysis: The system response is a function of time. The two
components of time domain response are transient response and steady-state
response. The time domain analysis can be applied only when the mathematical
model and nature of input is known. Time domain analysis is the preferred one
for welding applications.
Example: Transient analysis can be performed on welding current measured
from any arc welding process.
(ii) In frequency domain analysis: The system response is a function of frequency.
The two components of frequency domain response are magnitude response and
phase response. The frequency response of a system is defined as the steady-
state response to a sinusoidal signal. One major advantage of frequency response
is that we can use the data obtained from measurements on physical systems
without deriving its mathematical mode.
Both the analysis can be applied only when the mathematical model and the nature
of the input are known.
3.2.1 Order of the System
Order of the system refers to the highest derivative of the controlled quantity in the
equation describing the dynamics of the system, i.e. the highest power of ‘s’ in the
denominator polynomial of the transfer function without cancelling the common
terms in the numerator and denominator polynomial.
3.2.2 Zeroth Order System
A zero-order system is one in which output changes instantaneously as the input

changes. It is a memoryless system. Example: A potentiometer is a zeroth order
system. When potentiometer is connected in a circuit, changing the resistance value
of the potentiometer instantaneously changes the output voltage drop.
3.2 Control System Analysis 77
3.2.3 First-Order System
If the input–output equation of a system is a first order differential equation, it is

called as first-order system. For example, a series RC circuit is a first order system
which contains only one energy storing element (capacitor). The standard form of
first order system is given by
C(s) 1
(3.3)
R(s) Ts + 1
where
T Time constant.
Example: Thermocouple is an example of first-order system. Measurement of tem-
perature distribution in the Heat Affected Zones (HAZ) of the weldment is performed
using a thermocouple. As the temperature of the thermocouple rises, the differential
temperature between surroundings and HAZ decreases, the rate of heat flow and the
rate of change of temperature of thermocouple decreases proportionately.
3.2.4 Second-Order System
The second-order system is represented by the differential equation as shown in.

There are several factors that make second-order systems important. For example,
a series RLC circuit is a second order system which contains two energy storing
elements, inductor and capacitor.
The standard form of second order system is given by
C(s) ωn2
2 (3.4)
R(s) s + 2ξωn s + ωn2
where
ωn Natural frequency (rad/s)
ξ Damping ratio.
Example: An RLC circuit an example of a second order system. The two energy
storing elements are inductor and capacitor. Inductor stores energy in terms of mag-
netic field. Capacitor stores energy in terms of charge.
Table 3.1 Basic control actions

Control action Mathematical notation Effects
Proportional control u(t) k p e(t) Adjustable gain (amplifier)
t
Integral control u(t) ki 0 e(t)dt Eliminates offset (deviation in
the response of the system
from the desired settling value)
error. Also causes oscillations
Derivative control u(t) kd de
dt Provides faster response.
Never used alone
where,
u(t) control signal
e(t) control error (e feedback signal—setpoint value)
k p , ki , kd are the proportional gain, integral gain and derivative gain, respectively
3.2.5 Linearity
Linearity is the behaviour of a system in which the output signal varies in direct
proportion to the input signal. It is attributed to homogeneity and superposition.
Example: An amplifier amplifies the voltage signal proportional to the gain.
3.2.6 Sensitivity
Sensitivity is the relative change in the total transfer function of a system for a given
relative change in some parameter in the system.
3.3 Introduction to Fundamental Controllers
The fundamental controllers are based on a fixed structure controller family, the
so-called PID controller family. These controllers have proven to be robust and
extremely beneficial in the control of important applications. PID stands for Pro-
portional, Integral and Derivative controller. The basic control actions are shown in
Table 3.1.
In general, PID controller is used in most closed-loop industrial processes despite
the abundance of sophisticated and advanced controllers. The general equation of a
PID control is represented by
t de
u(t) k p e(t) + ki ∫ e(t)dt + kd (3.5)
0 dt
3.4 Stability Analysis 79
Fig. 3.6 Stability concepts using a cone
3.4 Stability Analysis
Stability is defined as the state of a system that remains in a constant state unless
affected by an external action and which returns to a constant state when the external
action is removed. If a system is stable for all variations of its parameters, then the
system is called stable system. If the system is stable for limited range of variations
of its parameters, then the system is called conditionally stable system. If the system
does not adapt to the variations in parameter, then it becomes unstable. The concept
of stability is explained using an example shown in Fig. 3.6.
Force is represented as
F m∗a (3.6)
where
F Force (N)
m mass (Kg)
a acceleration (m/s2 )
G centre of gravity
From the figure, it is seen that the cone in position (a) experiences unstable equi-
librium as it topples over left and right directions. In the position (b) the cone falls
back. The cone remains in neutral equilibrium at position (c).
The stability of a feedback system is directly related to the location of the roots of
the characteristic equation (denominator polynomial) of the system transfer function.
When the poles of the closed loop transfer function of a given system are in the right-
half of the S-plane, the system becomes unstable. When the poles of the system are in
the left-half plane the system is stable. When the poles are located on the imaginary
axis ( jω-axis) then the system is said to be marginally or conditionally stable.
Some of the basic stability test criteria are given as follows:
(i) Routh-Hurwitz Criteria,

(ii) Root-Locus,
(iii) Bode plot,

(iv) Polar plot.
3.5 Significance of Control System
The function of a feedback control system is to ensure that the closed-loop system
has desirable dynamic and steady-state response characteristics. Ideally, we would
like the closed loop system to satisfy the following performance criteria:
(i) Stability and robustness;
(ii) Minimal disturbance providing good disturbance rejection;
(iii) Good setpoint tracking (i.e. rapid smooth responses to set point changes);
(iv) Steady-state error (offset) is eliminated;
(v) Zero steady-state error;
(vi) Avoiding excessive control action.
3.6 Control System for Arc Welding
Welding controller improves the efficient welding and the essential features are given
as follows:
(i) Provides provision for programming;

(ii) Ability to create electrical isolation for the interface components;
(iii) Feedback signals should be noise free and responsive and
(iv) Hosts a data acquisition system that captures and analyses the feedback signals
obtained from the power source [1].
Earlier, solid-state switch mode rectification converters were developed for

improving the power quality in terms of power factor correction, reducing total
harmonic distortion, regulating the dc output in buck, boost or buck/boost modes.
The conventional AC-DC converters (rectifiers) are developed using thyristors and
diodes which provide controlled DC power flow. They have disadvantages such as
induced current harmonics which results in voltage distortion and poor power factor
at AC input mains and rippled DC output. With technological advancement, new set
of rectifiers are developed using self-commutating switches such as IGBT, MOS-
FET, GTO, etc. called as convertors, power factor correctors, PWM rectifiers, etc.
Further to sort out issues of power quality, active filters, passive filters and hybrid
filters are incorporated in power source units. Owing to the bulkiness of filters, the
efficiency reduces as a consequence of various losses [2]. The essential stages of
welding control are shown in Fig. 3.7.
3.6 Control System for Arc Welding 81
Mains Supply
230V or 440V
Output
Input to Sensors/ Parameters (Welding
(Weld
Power Source Transducers current, arc voltage)
Efficiency)
Parameter Correction
Control Unit
(gating signals)
Fig. 3.7 Welding control system
3.6.1 Sensing System
The variables used in control are sensed, scaled and fed to the processor where the
control schemes are applied. The signals include AC input voltage, welding cur-
rent, OCV, arc voltage, arc length and sometimes additional voltage like capacitor
or inductor voltage are used. The AC voltage is sensed using potential transformers.
For sensing DC voltages, hall effect sensors, optocouplers, isolation amplifiers are
used. The current signals are sensed using current transformers, shunt resistors, hall
effect current sensors, etc. The parameters are scaled and fed to the controller via
ADC channels which can be then used as a feedback signal in the control schemes
of the voltage and current controllers. These sensed parameters can be used to anal-
yse performance indices such as power factor, Total Harmonic Distortion (THD),
distortion factor, ripple factor, etc. [2].
3.6.2 Control Strategy and Algorithms
Analog controllers are implemented using microcontrollers and Application Specific

Integrated Circuits (ASIC), etc. Some of the digital control schemes includes Propor-
tional Integral (PI) controller, Proportional Integral and Derivative (PID) controller,
Fuzzy logic controller, Neural network-based controllers, Sliding Mode Controllers
(SMC) and adaptive control techniques, etc., which provide fast dynamic responses
and maintains the power source parameters in optimal range [2].
3.6.3 Desired Gating Signals
The desired gating signal involves deriving the gating signals for the solid-state
switches used in power sources. Some of the controllers used in deriving gating sig-
nals are PID controller, pulse width modulator (PWM) controller, Neural network-
based controllers, SMC, etc. They are implemented through hardware using Digital
Signal Processor (DSP) or microcontrollers or digitally build using computer soft-
ware. The derived gating signals are measured and obtained from dedicated hardware
units and fed to optocoupler for isolation, and then given to the power switches. The
full integration of the control unit, interfacing unit and power source unit are in devel-
opment such that they are cost-effective, compact, highly efficient, reduced weight
and reliable.
3.7 Controller Schemes Adopted for Welding Power

Sources
The control methods and algorithms adopted for welding power source are repre-
sented as the backbone for improving the efficiency in welding. Different control
techniques have been proposed in research reports to achieve better weld quality.
Some published literature reviewed consist of various implementation to overcome
the shortcomings in welding. To obtain a uniform weld quality a microprocessor-
based online feedback control is suggested in 1985. In reference to the experimental
results, they obtained almost closer to the desired weld strength with the help of pro-
portional control algorithm. The amount of energy utilized for performing welding
action uses thyristor controls. Welding current, voltage and energy are the parameters
which depend upon different control methods. The desired welding current are con-
trolled using SCR controller. Based on this control system, input voltage to the SCR
controller is determined. To produce the controlling input voltage the authors partic-
ularly used P algorithm [3]. In 1995, robotic arc welding thyristor type power source
is investigated with the control of microprocessor measuring registering facility. The
investigations found the direct, functional relationship between the dynamic and
technological properties of power sources of the electric arc welding. The dynamic
properties affect only the magnitude of the spatter during welding whereas the weld
quality is based on welding parameters [4]. During 2005, the experimental investiga-
tions reveal that consistent weld penetration and fusion area is achieved by adopting
constant voltage control technique. Although welding setup, base material and torch
angle inconsistency are unavoidable to a certain extent. Using constant current tech-
nique coated surface damage and weld spatter are considerably reduced [5]. In 2006,
pulsed current GMAW power source is investigated based on selection parameters.
Peak duration, peak current and pulse parameters, play a major role in finding the
properties of weld bead [6]. In 2008, the following evolutions happened. The digital
control technology became reliable, exact and flexible. The phase-shift full-bridge
3.7 Controller Schemes Adopted for Welding Power Sources 83
topology is used in the high-power arc welding/cutting inverter power source that
consisted of three control methods; Sliding mode control, PID control and fuzzy
control [7].
Later in 2011, TIG welding power sources incorporated fuzzy controller to per-
form rapid current control [8]. The power source with conventional voltage con-
trolled short circuit mechanisms resulted in accomplishment of stable and efficient
weld quality [9]. Silva et al. determined the following factors of MIG/MAG weld-
ing process: globular transfer mode, metal transfer dynamics, drop formation, forces
acting on it, shielding gases, which developed the controlled short-circuiting [9]. In
2012, authors realized that Fuzzy PI control reduced current error than traditional
PI control in burning arc stage. Using STT, no splash transition is observed, and
the appearance of welding is improved [10] and also investigated the occurrence of
breakdown in electron beam gun welding. The power sources reactive parameters
on the time characteristics of voltage and current in different parts of the circuit
are presented [11]. In 2013, Dual-input/Single-output fuzzy control welding power
source is analysed. Using 89S51 microcontroller, change of current is checked that
showed that welding current remains stable, control effect is appreciable. This con-
trol system is low cost, flexible control and high reliability [12]. Recently in 2014,
authors presented a single-phase ac RSW power source which utilized two SCRs
connected in parallel. Current passing through the coil is dependent on the trigger
time and the system properties [13, 14]. In 2014, 1985 some papers stated that in
RSW system the most important input variable is SCR trigger time, which controls
the energy used for welding action. For each controlling phase, the SCR trigger time
is related to the real-time key factor to produce the accurate control of welding. Weld-
ing voltage, energy and current are the selected variables that depend upon various
control strategies [3, 14]. To maintain the weld quality, different controlling methods
in the welding current have been reported by many authors. Different control meth-
ods have been proposed, such as learning control in the frequency domain lookup
table method and fuzzy logical control to deal with the non-linear and time-varying
properties of RSW. The regulation is obtained using PD controller. The experimental
result showed the effectiveness of the controlled output which is compared with that
of the PID controller [13]. During 2014, the analysis of arc welding inverter type
power source and control of droplet transfer is studied at closed-circuit welding. The
control of the current in short circuit phase was used to optimize the welding process
and parameter control in welded joint. This has reduced the electrode metal losses.
Using digital controlling methods in each phase of the droplet formation and droplet
transfer of the electrode in short circuit arc welding produced different possibility of
joining with minimum energy consumption and electrode material losses to achieve
high weld quality [15].
In 2015, authors designed a new hysteresis current controller technique for steady-
state and dynamic performance. A non-linear wide-bandwidth digital current con-
troller for DC–DC and DC–AC converters has been presented. Controller imple-
mented in FPGA kit examined the oversampling of current error by using ADC. This
method is only used for current error signal acquisition. Then it is operated at a con-
stant switching frequency, wide small signal bandwidth and minimum large signal
response delay of nonlinear controller [16]. In 2016, power supply system is devel-
oped with adaptive control method which can adjust to momentous change of system
parameters by using input–output variables for design instead of state variable [17].
Swati Narula also been reported in the literature of a full-bridge buck converter and
a half-bridge buck converter for the AWPS. In conventional isolated zeta AC–DC
converter, many limitations are identified across the switch. The proposed welding
converter is introduced with single stage power conversion, ease of control and high-
frequency isolation. The zeta converter-based AWPS has been designed, modelled
and implemented in hardware using closed-loop PI controller. Constant output volt-
age is maintained with respect to load and supply voltage variations. Due to the
reduced voltage stress in the devices, high reliability is obtained. The improved per-
formance during overload conditions is proportional to weld quality. Thus, the system
attained major advantages ,i.e. one gating signal is used to drive both the switches and
AWPS is simple in control [18]. The following Table 3.2 presents scholarly papers
based on the control strategy in welding as the main application areas.
3.8 Process Parametric Influences on Weld Quality
Configuration of input parameters for optimized welding is important. The effects of

input parameters like welding current, open circuit voltage, welding speed influences
the weld stability, weld penetration, bead width, depth of penetration, strength and
hardness of the final weldment [26].
Welding current is the most influential parameter because it affects bead shape
and controls the rate at which electrode is melted and that eventually controls the
deposition rate, heat affected zone, the depth of penetration and the amount of base
metal melted. Penetration and reinforcement increase with the increase in welding
current. If the current is too high at a given welding speed, the depth of fusion or
penetration will also be too high so that the resulting weld may tend to melt through
the metal being joined. High current also leads to the wastage of electrodes in the
form of excessive reinforcement and produces digging arc and undercut. This over
welding increases weld shrinkage and causes greater distortion. If the current is too
low, inadequate penetration or incomplete fusion may result. Too low current also
leads to unstable arc, inadequate penetration and overlapping.
Arc voltage also varies the efficiency of welding, i.e. increase in arc length causes
increase in arc voltage between the electrode and molten weld metal. The voltage
is the key factor that determines the shape of the weld bead cross section and its
external appearance. Increasing the welding voltage with constant current and weld-
ing speed produces flatter, wider, less penetrated weld beads and tends to reduce the
porosity caused by rust or scale on weldments material [6]. Increase in voltage also
increases the size of droplets. Further increase in voltage increases the possibility of
breaking the arc and disrupting the normal welding process. Increase in voltage also
enhances flux consumption which increases pick up or loss of the alloying elements
and therefore affects the mechanical and metallurgical properties of the weld metal.
Table 3.2 Control strategies in welding power sources
Author(s) Control circuits/inferences
Cho et al. (1985) [3]
• Microprocessor-based online feedback control

• P control Algorithm is used to obtain a uniform weld quality in changing environment
Dean et al. (2005) [5]

• Constant voltage control technique and constant current technique is used to achieve uniform weld penetration, fusion area and reduced weld spatter
(continued)
85
86
Zhu Guo-rong et al.

(2008) [7]
• Sliding Mode control (Voltage loop) and PI control (Current loop) were implemented to solve voltage overshoots and faster current response is realized
• Proposed sliding mode control has good dynamic response for the control system and easy to apply the algorithms
(continued)
3 Control Terminologies and Schemes for Arc Welding Processes
Yu, Bai et al. (2011) [8]
• Fuzzy control algorithm has been implemented

• Information of fuzzy controller in power source realized the control of molten pool width
(continued)
87
88
Shcherbakov, A. V.
(2012) [11]
• Electrical circuit of the power source of accelerating voltage is used to investigate the breakdowns
(continued)
Peng Tang & Xiao Jun

Jing (2013) [12]
• Dual-input/Single-output fuzzy control welding power source is implemented

• Using 89S51 microcontroller, change of current is checked and showed the welding current remains stable
• Control effects are also found to be good
• Control system is economical, flexible and reliability
(continued)
89
90
Kang Zhou & Lilong

Cai (2014) [14]
• The SCRs is used to approximate the value of the trigger time

• Proportional Derivative (PD) Controller is implemented to find the non-linear relationship between input and output variable
• Results in better performance than non-linear, time-varying and conventional PID control schemes
(continued)
Simone Buso &

Tommaso Caldognetto
(2015) [16]
• New digital non-linear wideband current controller using ADC and FPGA is implemented
• Current error processing algorithm causes power converters to attain constant switching frequency
• This exploited the oversampling and signal conversion delays
Y. Liu et al. (2016) [17]

• Adaptive control method is adjusted to the system parameters using input–output variables
(continued)
91
92
Swati Narula (2016)

[18]
• Zeta Converter-based power source is implemented

• PI controller is also utilized in this model
• Results obtains is that reduced the voltage stress and provides reliability
(continued)
J Schupp, W Fischer
and H Mecke [19]
• Various possibilities of arc control for GMAW process using power electronics circuits and external magnetic field is emphasized
• To provide optimal control of welding power source, improves in weld quality and reduce spattering, modern power sources consists of inverter on primary side and rectifier with smoothening
choke on secondary side

• The authors suggested that the movement and position of the arc can be controlled by applying external magnetic field and it revealed that influence of additional magnetic fields has a more
direct influence on the arc properties
(continued)
93
94
Young Min Chae et al.

[20]
• A mixed current and voltage control scheme for an inverter-controlled arc welding machine is proposed
• This scheme incorporates a closed loop current controller and an open-loop voltage controller to optimize the output voltage and current waveform
• An experimental comparison is made between the conventional and newly proposed control scheme and it revealed that the control schemes reduced the spatter generation by 30–50% by
stabilizing the metal transfer and improved the welding performance
(continued)
Dominic Cuiuri, John

Norrish and
Christopher David
Cook [21]
• Developed control strategies improves the control of highly unstable metal transfer in the GMAW process
• The research reviewed the issues involved in arc stability with CO2 shielding
• The authors proposed a solution by developing a new power source control technique to reduce the spatter levels when the wire feed rate increases
(continued)
95
96
Dominic Cuiuri and

John Norrish [1]
• Authors designed a modular programmable GTAW welding controller with graphical user interface (GUI)
• The control algorithms are developed and implemented using Digital Signal Processor DSP
• The weld performance is evaluated readily by comprehensive monitoring capabilities
(continued)
Zhu Guo-rong [7]
• Developed a 20KW arc welding inverter power source with Sliding mode control (SMC) and PI control, based on average state space model
• PI controller is applied for current loop to improve the stability. SMC is applied for voltage loop to decrease the overshoot of voltage loop without affecting the current loop
• And phase-shift PWM generation is applied additionally for making a simple control and attain superior performance
(continued)
97
98
Pang Qingle [22]
• The proposed fuzzy and PI control system for MIG welding inverter power source overcomes the demerits of the analogous control schemes with improved reliability, faster response and stable
arc length
• The fuzzy control and PI control are implemented to control arc voltage and welding current, respectively
• The proportional controller is used to obtain good dynamic performance and integral control is used to reduce the static error, increase the precision
• And the fuzzy control system is an adjustor of the closed-loop control system to determine the parameters of PI controller for different welding conditions
(continued)
Liu Zhao et al. [23]

• The proposed sliding mode control (SMC) and PWM control based on average state space model for a 20 KW Arc Welding
• Inverter power source overcomes voltage overshoot and obtain fast response time of PI control
Qingle Pang and Min

Zhang [24]
• The proposed Microcontroller(MCF5213CAF80)and FPGA based control system for pulsed MIG welding power source improves the flexibility, control, precision and reliability
• The overall hardware circuit, sampling, ADC circuits and PWM drive circuits and software design of FPGA and real-time embedded operating system μC/OS-II on MCF5213CAF80 are
designed. The experimental results showed that the microcontroller and FPGA based control system provided quicker response, better reliability and more stable arc length
(continued)
99
100
Tomoyuki Ueyama [25]
• Digitizing the arc welding power source using microprocessors and computer controls the spatter
• Digitisation of feed wire servo motor control is also implemented using feedback control system
• Recent improvements in digital computation capabilities paved a way to research on complex phenomena
(continued)
Peng Tang and XiaoJun

Jing [12]
• Authors developed a cheap, flexible arc welding power source using fuzzy control and PWM control schemes
• The traditional welding machine is applied to a simple systems and found to have demerits such as high energy consumption, poor welding performance, etc.
• When it comes to complex systems with too many variables the dynamics of the system is difficult to describe which emphasizes the significance of fuzzy controller which queries the fuzzy
control table and performs reasoning by saving time
(continued)
101
102
Y. Liu, X. Miao, and C.

Zhang [17]
• Proposed Model reference adaptive control (MRAC) for Submerged Arc welding (SAW) power source is used to control arc voltage, feed wire speed, arc stability and to achieve, constant current
characteristics
• The MRAC is compared with PID control algorithm.
• The limitation of PID is that it can be used for simple systems and single variable linear systems. Welding process being more complex with multiple input, multiple output and non-linearities;
the MRAC is opted to monitor the welding process and eliminate error by modifying the controller parameters
3.8 Process Parametric Influences on Weld Quality 103
Excessively high voltage lead to cracking, increases undercut and creates difficulty
in removing slag. Lowering the voltage produces stiffer arc, which improves penetra-
tion in a deep weld groove and resists arc blow. An excessively low voltage produces
a narrow bead and causes difficult slag removal along the bead edges.
Welding speed is the linear rate at which the electrode feed is moved along the
weld joint. With any combination of welding voltage and welding current, the effect
of changing the welding speed confirms to a general pattern. If the welding speed is
increased, power or heat input per unit length of weld is decreased and less filler metal
is applied per unit length of the weld, resulting in less weld reinforcement. Thus, the
weld bead becomes smaller. Weld penetration is affected more by welding speed than
any variable other than current. For excessively slow speeds the molten weld pool
is beneath the welding electrode. Excessive speed may cause undercutting, porosity,
arc blow, uneven bead shape, cracking and higher slag inclusion in the weld metal.
Higher welding speed results in less heat affected zone. Within limits, welding speed
can be adjusted to control weld size and penetration. Relatively slow welding speed
provides time for gases to escape from the molten metal, thus reducing porosity. If
the welding speed is too low, it may also result in a large molten pool that flows
around the arc, resulting in rough bead, slag inclusions and burn through of the weld
plate [26].
In 2002, authors studied the relationship between the process variable and bead
penetration, the effect of welding current and arc voltage on bead penetration on
increasing the voltage. The welding current and depth of penetration of bead also
increases correspondingly [27]. In welding process, deformation is unavoidable.
Inherent deformation of weld bead is analysed using elastic finite element modelling
(FEM). Major influence of the weld length to transverse shrinkage is observed [28].
In 2008, digital computation began to gain attention. The information obtained from
the device needs to be controlled. For the digital data used by the computer, numerical
calculations are performed on the data. Thus, controlling is done for the MIG welding.
The voltage and current control enables the overall stability of the molten pool after
solidification [29]. Dynamic response is the feature of transistor characteristics in
a transistorized technology, which determines the variation in the characteristics of
the electric circuit [30].
In traditional method, the process stability is determined based on the arc stabil-
ity by considering both dynamic and static characteristics assessment [31], which
resulted in better power source properties. In 2010, a thermal model for single seam
per layer multipass narrow gap P-GMA is published. Here, appropriate pulse param-
eter is decided by considering heat and current. Pulse parameters influence at a given
heat input. Then, the higher mean current assured the preparation of sound weld. This
also avoided the lack of groove wall fusion. Single seam per layer multipass narrow
gap P-GMA welding is operated by changing the pulse parameters, with the constant
heat input to produce a sound weld [32]. To reduce the bead widening AC frequency
control technique is done, such that by increasing the input frequency parameter,
widening of the bead during welding is reduced accordingly [25]. Deva Kumaran,
K and his team [33] has investigated the static and dynamic characteristics of an
inverter type synergic GMAW power source. The operational mode study concluded
Table 3.3 Welding current Type of welding Output voltage calculation

versus voltage relations in
welding CO2 0.04·I AC + 15
TIG 0.04·I AC + 10
DC ARC 0.04·I DC + 20
MIG 0.04·I AC + 13
that there exists a direct correlation between welding wire feed rate, arc voltage
and mean current. The dynamic setting voltage of operating conditions is raised to
higher value than that of the static operating condition. The authors concluded that
using modified short circuit MS-GMAW mode of power source produced less spatter
when compared with that of the GMAW, Pulsed P-GMAW and double pulsation DP-
GMAW. The quality of the weld joint is largely determined by the optimal selection
of arc voltage and welding current. The relationship between welding voltage and
welding current in diverse types of welding processes are shown in Table 3.3.
It is required to study the power source characteristics under static and dynamic
operating conditions to decide about the optimal range of welding parameters [33]. In
2013, the stability of the power source is experimentally investigated. The arc stability
which determines the process stability is was assessed by implementing conventional
P, PID controller. The static (constant voltage) and dynamic (duty cycle) characteris-
tics assessment showed good power source properties. Using LabVIEW (Laboratory
Virtual Instrument Engineering Workbench), stability of the power source is anal-
ysed. The disturbances which affect the arc stability, welding parameters and metal
transfer irregularities are identified. Based on the voltage and current values, weld-
ing process stability and output current are related to each other [34]. Yamane et al.
investigated the effects of welding power source by analysing the current and volt-
age behaviour and identified the cause of spatter generation. Then, characteristics of
the power source are proposed to enhance the stability and self-regulation of the arc.
The developed power source with non-linear characteristics verified the performance
[35]. The analysis of arc welding inverter type power source and control of droplet
transfer during closed circuit welding is analysed. The control of the current in short
circuit phase was used to optimize the welding process and parameter control in
welded joint. This has reduced the electrode metal losses [15]. Overall study reveals
the influence of input parameters such as current, voltage, frequency and weld speed
that directly contributes to the mechanical behaviour and output characteristics of
welding. In the following Table 3.4, some of the comparative literature are illustrated.
Table 3.4 Process parametric effects on welding
Author(s) Welding process/input parameters Outputs and inferences
I. S. Kim et al. 2002 [27] CO2 arc welding

I 180–260 A
V 18–26 V
When
voltage increases, current and bead penetration also increases
(continued)
105
106
LIANG et al. 2004 [28] I 100–200 A

V 18–20 V
Weld Speed 10 mm/s,
15 mm/s and 20 mm/s
Inherent deformation of weld bead is analysed using elastic FEM

T. Uezono and T. Hongjun 2008 MIG Welding
[29] 100A
17.1 V
• Digital filtering process

• Removes abnormal voltage
(continued)
T. Uezono and T. Hongjun 2008 MIG Welding

[29] 140 A
23 V
Bead appearance is uniform on stainless steel material using digital filtering process
107
(continued)
108
T. Uezono and T. Hongjun 2008 MIG welding

[29] 250A
33 V
Mild steel subjected to Inconel deposition welding. Large grain sputter on the surface is observed
Dutra, Jair Carlos 2009 [30] MIG/MAG Welding Fillet weld carried out manually with CO2 at low current and voltage observed more splashes
20 A
21 V
Dutra, Jair Carlos 2009 [30] MIG/MAG Welding Fillet weld carried out manually with CO2 at high current and voltage observed minimum splashes
44 A
21 V
(continued)
Dutra, Jair Carlos 2009 [30] MIG/MAG Welding

24 A
20 V
Low value of current increases stability

109
(continued)
110
Ueyama, Tomoyuki 2010 [25] GMA Welding

200 A
50 Hz/200 Hz
Ueyama, Tomoyuki 2010 [25] GMA Welding

IavL: 340 A VavL: 26.5 V
IavT: 140 A VavT: 22.5 V
AC frequency control (synchronized control method) technique in high-speed fillet lap welding by tandem pulse GMA welding method
(continued)
Agrawal et al. 2010 [32] GMA Welding

V 8V 220 A
0.2 240 A
Appearance of interrupted intimate contact of single seam per layer weld deposit with the groove wall in narrow gap welding
(continued)
111
112

V 28 V 265 A and 0.08
13.4 265 A and 0.15
265 A and 0.25
240 A and 0.25
Appearance of interrupted intimate contact of single seam per layer weld deposit with the groove wall in narrow gap welding
(continued)

V 28 V 220 A and 0.25
10.2 240 A and 0.25
240 A and 0.15
Lack of fusion at groove wall is typically observed in transverse section of weld joint
(continued)
113
114

V 28 V 240 A and 0.25
13.4 265 A and 0.07
265 A and 0.15
265 A and 0.25
Lack of fusion at groove wall typically observed in transverse section of weld joint
(continued)
Devakumaran, K et al. 2012 [33] (a) MS-GMAW

(b) GMAW
(c) P-GMAW
(d) DP-GMAW
Spatter generation for different process is analysed

MS-GMAW
produce less spatter than that of GMAW, P-GMAW, and DP-GMAW
(continued)
115
116
Skrzyniecki et al. 2013 [34] MAG Welding

110 A
11.1 V
Each droplet caused dynamic change of current and voltage
(continued)
Yamane. S et al. 2013 [35] Arc Welding

120 A
• Using non-linear power source

• Smooth metal transfer is achieved
• During short-circuiting, the current increases according to the metal transfer
• Before the droplet is released, the current decreases to a low value to reduce the spatter
117
(continued)
118
Lebedev, V. A 2014 [15] CO2 short circuit welding

200 A
18–20 V
(1) Resistance of the welding circuit; (2) controlling effect; (3) welding voltage; (4) welding current
Observed that splashing increases with the reduction of the rate of increase of current
3.9 Real-Time Sample Reports on Formulating Adaptive Control … 119
3.9 Real-Time Sample Reports on Formulating Adaptive

Control Scheme for Cold Metal Transfer
for JoiningAA6061
3.9.1 Objective
In this chapter, an adaptive control scheme is employed for joining Aluminium 6061
alloy sheets by Cold Metal Transfer process. The transfer function model of the
CMT welding system is derived using empirical equations. The CMT plant transfer
function is estimated using system identification technique. For the estimated plant
model, a conventional PID controller is initially designed by tuning the controller
parameters k p , ki , and kd . The designed control system is tested for its ability to
control the welding current when short circuit phase and arcing phase are detected.
Following the conventional PID controller, a Model Reference Adaptive Controller
(MRAC) is implemented to maintain the welding current at desired range, during
melting and electrode wire short-circuiting. The performance analysis for the pro-
posed adaptive control scheme and the conventional PID controller is compared.
3.9.2 Implementation
The experimental trails for the CMT process is performed on 6061 Aluminium alloy
sheets (2 mm thickness) using the Fronius advanced CMT 7000 VR machine as
shown in Fig. 3.8. Argon shielding gas is used with a constant flow rate of 18 lit/min.
The CMT welding is performed in pulsed mode using a linear six-dimensional robotic
displacement machine. The welding trails were designed by varying the wire feed
speed and welding current. The remaining welding parameters are maintained con-
stant. The parameters evaluated in the study are weld bead width, Depth of Penetration
(DOP) and reinforcement height.
PID controller is the most preferred conventional controller. It is a feedback mech-
anism widely used in industrial control systems and a variety of other applications
requiring continuously modulated control. A PID controller continuously calculates
an error value and applies corrective action depending on the proportional, integral
or derivative terms. Its ideal transfer function is represented by

ki
G c (s) k p 1 + + kd s (3.7)
s
where k p , ki , kd denote the proportional, integral and derivative controller gains,

respectively.
Fig. 3.8 CMT Experimental setup
Fig. 3.9 PID control loop
After obtaining the plant model from the system identification toolbox, the con-
ventional PID controller is implemented and tested as shown in Fig. 3.9. The plant
model obtained is a third-order system which consists two-zeroes and three-poles as
shown. The optimal controller gains were obtained using the auto tuning feature in
MATLAB. The optimal k p , ki and kd values provide the desired performance and
achieve appropriate time domain specifications. The obtained controller gains are
Reference
Model
RM
Adaptive
Mechanism
AM e
ym
r Controller u Actual plant yp y

C P
- -
Fig. 3.10 MRAC control scheme
k p 26.6; ki 8.67e04; kd 0.0008202

In CMT process, the electrode feed is retracted when a short circuit is detected.
At that instant, the welding current decreases and droplet detachment takes place.
Once the droplet detaches, the electrode feed inches forward and there is an increase
in welding current. When the short circuit is detected, the MRAC controller adjusts
the wire feed rate and controls the welding current.
The feedback element in MRAC measures the plant output and generates an
error signal which is fed to the controller block (as shown in Fig. 3.10). A standard
second-order reference model is used in the reference loop for which the adaptation
mechanism works to result in a plant output response like that of the reference model.
The adaptation mechanism block uses suitable adaptation law to achieve desired plant
response. The theory of the developed MRAC is presented as follows, the idea behind
MRAC is to create a closed-loop controller whose parameters are updated to change
the system response to match with a desired reference model
3.9.3 Controller Results
The PID controller is implemented to control the welding current when short circuit
of the electrode wire feed is detected. Initially, for the setpoint of 50A, the perfor-
mance of the auto-tuned PID controller is evaluated. The short circuit is fed as a
disturbance to the control system. The PID controller response is shown in Fig. 3.11.
The experimental weld images are presented.
From the observed results, the PID controller is unable to adapt to the disturbance
while detecting a short circuit. Initially, the welding current increases to a value of
Fig. 3.11 Response of PID Controller
Fig. 3.12 Response of MRAC
50A. After being subjected to a disturbance (short circuit phase), the welding current
reduces which thereby decreases the bead width and DOP. The experimental images
are shown in Table 3.5.
3.9.4 MRAC Controller Response
The response of the implemented MRAC is shown in Fig. 3.12. It is observed that
the current decreases when a short circuit is detected, and the electrode feed retracts.
The MRAC ensures that the current is maintained at 50 A with uniform bead width
and DOP.
The experimental weld images are shown in Table 3.6.
Table 3.5 Weld images of PID Controller

Image location Weld image Bead width (mm) Remarks
(at 50 A welding current)
3.37 Good DOP is
observed
1
3.22 Due to short

circuit phase, the
DOP decreases
2 with decrease in
welding current
and heat input
2.32 After the short

circuit phase, the
welding current
3 and heat input
decreases further
to a low value.
DOP is very less
Table 3.6 Weld images of MRAC

Image location Weld image Bead width (mm) Remarks
(at 50 A welding current)
3.37 Good DOP is
observed
1
3.37 Good DOP is

observed. After
short circuit and
2 arcing phase,
current and heat
input are
retrieved
3.37 Good DOP is

observed. After
short circuit and
3 arcing phase,
current and heat
input are
retrieved and
maintained
References 125
References
1. D. Cuiuri, J. Norrish, Verstatile Welding Power Source Controller for Research and Product
Development (2006)
2. I. Singh, Bhim Senior Member, B. N. Singh, I. Chandra, Ambrish Senior Member, I. Al-
haddad, Kamal Senior Member, A. Pandey, I. Kothari, Dwarka P Senior Member, A review of
three-phase improved Power quality AC–DC converters. IEEE Trans. Indus. Electron. 51(3),
641–660 (2004)
3. H.S. Cho, D.W. Chun, A microprocessor-based electrode movement controller for spot weld
quality assurance, IEEE Trans. Indus. Electron. IE-32(3), 234–238 (1985)
4. J. Dziubinski, J. Slania, As assessment of the welding properties which are characteristic of
the power source used in robotic arc welding. Weld. Int. 9(8), 601–606 (1995)
5. G. Dean, J. Norrish, C. Cook, Evaluation of control techniques for dip transfer gas metal arc
welding. Australas. Weld. J. 50, 34–41 (2005)
6. P.K. Palani, N. Murugan, Selection of parameters of pulsed current gas metal arc welding. J.
Mater. Process. Technol. 172(1), 1–10 (2006)
7. Z. Guo-rong, L. Zhao, Z. Ai-yun, Y. Mi, D. Shan-xu, K. Yong, Sliding Mode Control and PI
Control for Arc Welding/cutting Inverter (2008), pp. 4–7
8. X. Dl et al., The Fuzzy Control Algorithm in Copper-Coated Aluminium Wire TIG System
Application (2011), pp. 2214–2218
9. R.H.G. Silva, J.C. Dutra, R. Gohr, Scientific and technological fundamentals for the develop-
ment of the controlled short-circuiting MIG/MAG welding process: a review of the literature.
Part 2 of 3. Metal droplet formation, shield gases, penetration mechanisms, heat input and
economical asp. Weld. Int. 23(2), 141–149 (2009)
10. S. Liu, Y. Wang, Research of CO2 welding inverter power source under current waveform
control. IEEE International Conference on Automation and Logistics, August, pp. 116–121
(2012)
11. A.V. Shcherbakov, Switching processes in power sources for electron beam welding gas with
formation of high-voltage breakdown. Weld. Int. 26(3), 221–226 (2012)
12. P. Tang, X. Jing, A Novel Fuzzy Control of Welder Power Source Using PWM Chip. 2013
Sixth International Symposium on Computing Intelligence and Design, pp. 224–227 (2013)
13. L. Gong, C.L. Liu, X.F. Zha, Model-based real-time dynamic power factor measurement in AC
resistance spot welding with an embedded ANN. IEEE Trans. Ind. Electron. 54(3), 1442–1448
(2007)
14. K. Zhou, L. Cai, A Nonlinear Current Control Method for Resistance Spot Welding.
IEEE/ASME Trans. Mechatron. 19(2), 559–569 (2014)
15. V.A. Lebedev, M.S. Sorokin, A.A. Belov, Algorithms for controlling inverter sources of welding
current to optimize the electrode metal transfer parameters. Weld. Int. 28(12), 957–961 (2014)
16. S. Buso, T. Caldognetto, A nonlinear wide-bandwidth digital current controller for DC–DC
and DC–AC converters. IEEE Trans. Ind. Electron. 62(12), 7687–7695 (2015)
17. Y. Liu, X. Miao, C. Zhang, The Model Reference Adaptive Control Method of Submerged Arc
Welding Power Supply System (2016), pp. 3646–3649
18. S. Narula, B. Singh, G. Bhuvaneswari, Power factor corrected welding power supply using
modified zeta converter. IEEE J. Emerg. Sel. Top. Power Electron. 4(2), 617–625 (2016)
19. J. Schupp, W. Fischer, H. Mecke, Control with power electronics (475), 18–19 (2000)
20. Y.M. Chae, Y. Jang, M.M. Jovanovic, A Novel Mixed Current and Voltage Control Scheme for
Inverter Arc Welding Machines. APEC 2001. Sixth Annual IEEE Applied Power Electronics
Conference and Exposition (Cat. No.01CH37181), vol. 0, no. C, pp. 308–313 (2001)
21. D. Cuiuri, J. Norrish, C. Cook, New approaches to controlling unstable Gas Metal Arc Welding.
Australas. Weld. J. vol. 47, no. 3 (2002)
22. P. Qingle, Design of fuzzy control and expert system based MIG arc welding invert power
source. ICEMI 2009—Proceedings of the 9th International Conference on Electron. Measure-
ments and Instruments (2009), pp. 489–492
23. L. Zhao, X. Bai-lu, W. Shu-hui, Z. Guo-rong, The Sliding Mode Control for Arc Welding
Inverter Power Source. 2008 3rd IEEE (2008), pp. 1100–1104
24. Q. Pang, M. Zhang, Design of digital control system for pulsed MIG welding power source.
Proc. World Congr. Intell. Control Autom. 20090461204, 2492–2495 (2010)
25. T. Ueyama, Welding power sources_TF. Weld. Int. 24(9), 699–705 (2010)
26. I. The, Effect of welding parameters on bead. Flux 1 (1989)
27. I.S. Kim, J.S. Son, I.G. Kim, J.Y. Kim, O.S. Kim, A study on relationship between process vari-
ables and bead penetration for robotic CO2 arc welding. J. Mater. Process. Technol. 136(1–3),
139–145 (2003)
28. L. Wei, Measurement of inherent deformations in typical weld joints using inverse analysis
(part 2) prediction of welding distortion of large structures†. Trans. JWRI is (2005)
29. T. Uezono, T. Hongjun, Application to MIG welding using welding power source equipped
with digital filtering process. Weld. Int. 22(5), 299–303 (2008)
30. J.C. Dutra, MIG/MAG—Short circuit metal transfer—Welding power sources versus arc gases.
Weld. Int. 23(4), 231–236 (2009)
31. M. Suban, J. Tušek, Methods for the determination of arc stability (METAL TRANSFER). J.
Mater. Process. Technol. 143–144(1), 430–437 (2003)
32. B.P. Agrawal, P.K. Ghosh, Thermal modeling of multipass narrow gap pulse current GMA weld-
ing by single seam per layer deposition technique. Mater. Manuf. Process. 25(11), 1251–1268
(2010)
33. K. Devakumaran, N. Rajasekaran, P.K. Ghosh, Process characteristics of inverter type GMAW
power source under static and dynamic operating conditions. Mater. Manuf. Process. 27(12),
1450–1456 (2012)
34. K. Skrzyniecki, P. Kolodziejczak, P. Cegielski, A. Kolasa, Experimental Studies on Stability of
Power Source—Arc (2013), pp. 359–362
35. S. Yamane, S. Xiang, Y. Kaneko, K. Oshima, Effect of power source characteristic on CO 2
short circuiting arc welding_TF. Sci. Technol. Weld. Join. 10(3), 281–286 (2005)
Chapter 4
Power Sources and Challenges
for Different Arc Welding Processes
The choice of power source is based on the working principle of joining technique
and the material type to be welded. Power sources for various arc welding process
are discussed in the following sections.
4.1 Power Sources in Manual Metal Arc Welding (MMA)
In MMA welding, both AC and DC power sources are used. Mostly inverter type
power sources are preferred considering its adaptability for minimal maintenance.
The parameter range is presented in Table 4.1.
4.2 Power Sources in Shielded Metal Arc Welding (SMAC)
Welding Power Sources in SMAC process utilizes either AC or DC, but in either case,
the power source must be of the constant current (CC) type. Whether to use an AC,
DC, or AC/DC power source depends on the following factors presented in Table 4.2.
When DC power is used, polarity of electrode is a key factor. If electrode is connected
to negative polarity, the weld exhibits shallow penetration that contradicts to the deep
penetration exhibited when electrode is connected to positive polarity [1, 2].
Table 4.1 Parameter range in MMA power source

Approximate electrode length Weld time (s) Welding current (A)
(mm)
200–400 45–90 40–400

https://doi.org/10.1007/978-981-13-0806-2_4
128 4 Power Sources and Challenges for Different Arc Welding Processes
Table 4.2 Suitability of Factor Suitability and reasons

power source type for SMAC
Electrode selection DC power offers wider range of
electrode types
Thickness of metal DC power is suitable for both thick and
light metal workpieces
Distance from work AC power is suitable for layer distance
because the voltage drop through the
cables is less when AC is used instead of
DC.
Welding position DC power is suited for overhead and
vertical position welding since they can
be operated at low welding currents
Arc blow AC is suited since the rapid reversing
magnetic field produced does not affect
the arc and weld bead but in DC power
magnetic field is set up throughout the
weldment which would affect the
welding arc and the process
4.3 Power Sources in Gas Tungsten Arc Welding

(GTAW)/Tungsten Inert Gas Arc Welding (TIG)
In Gas Tungsten Arc Welding, also called as Tungsten Inert Gas Arc Welding, the
arc is established between a non-consumable tungsten electrode and the workpiece.
The electrode, the arc and the area around the weld puddle are protected from the
atmospheric reactions by inert gas shield of Helium or Argon. It is performed using
both DC and AC (30–300 Hz) for aluminium with pulsed current of 5–500A with
a frequency range of 2 Hz–20 kHz [3]. The choice between DC or AC depends on
the metal thickness and type of the metal to be welded. If DC power is chosen, then
the polarity is the key governing factor. If the electrode is connected to the negative
terminal of the power source, the electrons flow from the electrode to the workpiece,
and approximately 30% heat of the arc is concentrated at the electrode end and 70%
heat of the arc is concentrated at the workpiece [4]. This demands the use of small
tungsten electrodes to produce a narrow-concentrated arc. The resulting weld shape
indicates narrow and deep penetration. This polarity is suitable for welding most
metals.
Aluminium and magnesium have an oxide coating on their surface which must be
physically removed prior to welding if this polarity is to be used. If the electrode is
connected to the positive terminal of the welding power source, the flow of electron is
from the work to the electrode tip, concentrating approximately 30% at the workpiece
and 70% of the heat of the arc at the electrode. Such a higher heat at the electrode
side necessitates the use of larger diameter tungsten electrode to prevent its melting
and contamination of the workpiece. Since the electrode diameter is larger and the
concentration of heat is less at the work, the resulting weld penetration is wide and
4.3 Power Sources in Gas Tungsten Arc Welding (GTAW) … 129
Table 4.3 Different metal transfer modes in GMAW

Parameters Short-circuit transfer Globular transfer Spray transfer
Voltage Less than 22 V Greater than 22 V Greater than 22 V
Shielding gas 75 Argon/25 Carbon 75 Argon/25 Carbon Argon
dioxide dioxide
Welding position All positions Flat Flat
Material types Light structured Hard structured metal Hard structured metal
metals, ornamental with high deposition with high deposition
metals and gauge rates rates
metals
Material thickness 0.76–1.1 mm Greater than 1.1 mm Greater than 2.5 mm
Characteristics swift, causes mild More spattering and Efficient and high
spattering cleaning is essential speed, no spattering
shallow. For this polarity, the heavy oxide coating of Aluminium and magnesium must
be removed since it acts as an insulator for the electron flow [4] When DC power
sources are used, the arc should be initiated by touching the tungsten electrode on
the workpiece and quickly withdrawing that results in electrode contamination. So,
DC power sources for GTAW are designed with high-frequency stabilizer wherein,
the electrode when brought near 1/8th in. of workpiece, ionization of gas in the
occurs place thereby creating a conductive path for the arc to be established. When
AC power sources are used for GTAW it always incorporates high frequency which
maintains an ionized path for establishing, maintaining and reigniting the electric
arc each time, while the current changes the direction. The electrode size and the
resulting weld penetration is intermediate [1].
4.4 Power Sources in Gas Metal Arc Welding/Metal Inert

Gas Welding (GMAW/MIG)
GMAW process utilizes a consumable metal electrode. The electrode is a metal wire
which is transferred across the arc and into the molten weld puddle. All GMAW
is done using DCEP. This polarity provides deep penetration, arc stability with low
spatter. AC power is not used for gas metal arc welding [5]. The knowledge about
three types of metal transfer is essential in GMAW which are depicted in Table 4.3.
A DC constant voltage power source is more suited for GMAW because of
lower open circuit voltage (OCV) and self-correcting arc length feature. It can be a
transformer-rectifier or a rotary type. Constant voltage power sources are used for
spray transfer welding and globular transfer welding. If it is used for short-circuiting
arc welding, slope control must be included which aims to limit the high short-circuit
current [6]. The power sources of GMAW have an outlet for providing power to the
wire feeder. The function of the wire feeder is to deliver the welding wire to the arc at
Table 4.4 Welding process, current type and characteristics of power sources
Welding process Type of current Characteristics of power source
MMA DC/AC Constant current
TIG/GTAW DC Constant current
Plasma arc welding DC Constant current
SMAW DC/AC Constant current (if electrode thickness = 2.4 mm)
DC Constant voltage (if electrode thickness = 2.4 mm)
MIG/GMAW DC Constant voltage
constant speed [5]. Since the rate of wire feed determines the current, and ultimately
the amount of heat at arc region, wire feed speed influences the weld bead and pene-
tration. Inverter type power sources and microprocessor controls are popular features
in GMAW.
In Metal Inert Gas (MIG)/Metal Active Gas Welding (MAG) welding, the elec-
trode is of a continuous wire that is fed through a torch which would be attached to
a robotic arm. Welding continues for several minutes depending on electrode size.
MIG/MAG welding prefer DC sources predominantly while few AC sources are also
developed. The welding current is a pulsed type at 10–150 Hz frequency between
low level of 50A and high peak value of 400–600 A [3].
4.5 Power Sources in Submerged Arc Welding (SAW)
SAW process is mechanized and is like MIG/MAG welding process except that the
arc is submerged beneath the flux and along the joint instead of being protected by
shielding gas. AC or DC power sources are used with welding current Up to 1500 A
[3].
The following Table 4.4 presents the different welding processes, type of welding
current and the respective power source characteristics.
4.6 Major Challenges in Power Sources
4.6.1 Harmonics
In AC power systems, voltage and current are purely sinusoidal. When their wave-
forms are distorted using Fourier series they can be analysed into components at
fundamental frequency and their multiples as shown in Fig. 4.1 [7]. The frequency
components other than the fundamental ones are termed as harmonics. The main
source of origins of harmonics are given as follows:
4.6 Major Challenges in Power Sources 131
Fig. 4.1 Harmonics
1. Non-linear magnetic components like saturated transformer cores

2. Non-sinusoidal air gap flux distributions
3. Switched power devices such as inverters, thyristors, rectifiers and other power
converter devices.
The adverse effects of harmonics are given as follows:
1. Heating of cables and the power electronic devices
2. Electrical resonance resulting in undesired or hazardous voltage and current
levels
3. Interference to the communications in the electrical power circuits
4. Electromechanical resonance resulting in noise, vibration and failure of the elec-
trical and mechanical components.
Power quality includes all aspects of events in the system that deviates from normal
operation. Power sources act as non-linear loads, drawing distorted waveforms that
contain harmonics causing degradation problems of the power devices [8]. Total
harmonic distortion (THD) is the summation of all harmonic components of the
voltage or current waveform as compared against the fundamental component of
the voltage or current wave. The type of load (linear or non-linear) affect the power
quality of the system due to the current withdrawal of each type of load. Linear loads
draw current that is sinusoidal and does not distort the waveform. Non-linear loads
can draw current that is not perfectly sinusoidal. Since the current waveform deviates
from a sine wave, distortions are created in the voltage waveform. Harmonics have
frequencies that are integer multiples of the waveform’ s fundamental frequency of
the waveform. As the percentage of THD is more, the distortion on the mains is
more. Unwanted distortion creates adverse effects and maintaining low THD values
on a system will further ensure proper operation of equipment and a longer lifespan
[8].
Gopalakrishnan et al. [9] presented a survey on harmonic contents in MIG welding

power supplies and the issues of voltage and current harmonics based on the field
measurements in various plant locations. The field measurement is useful to design
shunt filter (tuned LC filters) to improve the power factor from 0.74 to 0.9 and
suppressing the total current harmonics up to 70%. Power flow analysis is performed
resulting in a power factor identical with measured value.
Wiseman and Wu [10] simulated and experimented on active damping to reduce
THD of line current for current source rectifiers using GCT (Gate Controlled Thyris-
tors). Two types of LC resonances such as series and parallel resonance and series
resonance, which gets excited by harmonics of the rectifier and source voltage,
respectively. The use of active damping control simplifies the design of input fil-
ter.
Vijaya Prasuna et al. [10] presented a dual boost PFC converter which provided
higher boost factor, reduced THD and good control capabilities using average current
control method. Simulations are implemented in MATLAB, for circuits incorporating
the PFC, and the results of current and voltage waveforms are observed. The authors
suggested further improvement by implementing Fuzzy and PI controllers.
4.6.2 Effects of Magnetic Field in Arc Welding
Magnetic arc blow (shown in Fig. 4.2) is one of the challenges that occur when the
magnetic field around the arc is not symmetrical completely. It can often result in a
poor and defective weld.
Arc blow results in the following problems:
(i) The weld pool, and thus the weld bead, is non-uniform as the arc gets deflected
towards one side, producing a defective weld.
(ii) If the arc is deflected along the joint, the width of the bead and the penetration
gets affected.
(iii) The protection provided by molten slag or gas can be affected, resulting in the
formation of pores.
The problem worsens when the welding current increases and becomes more notice-
able. This will further increase all the electromagnetic forces rapidly in and around
the electric arc.
The major causes of arc blow are given as follows:
(i) Asymmetric return current connection
(ii) Asymmetric position of the workpiece while performing welding
(iii) Close Electrode position during multiwire welding
(iv) Permanent magnetic fields from magnetic clamps, bead plates, workpiece, etc.
Fig. 4.2 Magnetic arc blow
The welding arc behaves as a fluid conductor and generates a magnetic field at
a certain radius from the axis [11]. Over the last two decades, developments are
made in inverter technology for providing lighter and smaller power sources with
improvement in performance and efficiency to reduce the magnetic arc blow. In the
inverter power source, a single or three-phase input is rectified, and then switched
at high frequency using a power transformer. The secondary is then rectified to
obtain DC output, The DC output consists of high-frequency ripples which produce
a magnetic field at that switching frequency. Today, the switching frequencies are
increased from few KHz to 100 kHz.
The arc welding equipment’s tested are given:
(i) Conventional arc welding transformer
(ii) Thyristor arc welding power source
(iii) Inverter arc welding power source
(iv) Square wave inverter arc welding power source
(v) Pulsed MIG inverter arc welding power source.
For inspecting the magnetic field, QinetiQ’s 3 m linear plotter, magnetometer, hall
sensor, coil sensor and a data acquisition card are used. The magnetic field is made
up of many harmonic components which must be considered when analysing it
with standard ICNIRP reference levels. The results in the investigation carried by
GB Melton et al., of TWI Limited [2] reveals that for conventional AC transformer
power source, the magnetic field is predominantly 50 Hz, with harmonics at 100 and
150 Hz. For MMA and TIG welding, square wave, inverter power source, the results
indicate that as output frequency increases, the magnetic field decreases. For square
wave outputs, the harmonic components are higher than that of a sine wave output.
In case of MIG/MAG welding, output current waveform depends on metal transfer

mode. At low current, the wire dips into the weld pool. At higher current, open arc
is maintained with small metal drops spraying at wire end resulting in spray transfer
with low magnetic field. For obtaining pulsed transfer, the welding current is pulsed
between high and low preset values at a frequency depending on mean current. For
pulsed transfer, the magnetic fields are higher and there are significant harmonic
components resulting in the summed magnetic field exceeding the reference level.
In most cases, the magnetic fields are higher than INCIRP reference levels which
necessitates detailed investigations.
4.6.3 Protection of Power Sources
The arc welding power sources are made of semiconductor materials and it is impor-
tant to determine semiconductor losses based on which adequate cooling means is
provided to keep the device temperature within prescribed limits. The semiconductor
losses are grouped as follows:
1. Conduction loss (on-state and dynamic saturation)
2. Switching (turn-on and turn-off)
3. Off state.
The conduction and switching losses depend on converter type, operating frequen-
cies, type of load and switching characteristics [7]. Off state losses are very negli-
gible. Snubber circuits reduce switching losses and associated stress due to voltage
fluctuations. The two types of snubber circuits are dissipative and non-dissipative.
1. Dissipative: The energy stored in reactive elements is dissipated in the resistors
and are converted into heat energy. This is not suited for higher operating powers
and frequencies.
2. Non-dissipative: The low losses are caused by nonideal device properties such
as transient switching loss and conduction losses of the switching devices in the
snubber circuits.
4.6.4 Cooling System
The power sources generate heat and they are not 100% efficient. There must be
a way to dissipate the excess heat generated to prevent power source damage and
maximize the reliability. This can be achieved by incorporating a cooling system.
A proper cooling system is necessary to increase the lifetime and reliability of the
power electronic devices in the power source unit [7]. The maximum allowable limit
of temperature for each component is prescribed in the datasheet and those limits
should be strictly maintained. The three mechanisms of heat transfer are shown in
Table 4.5.
Table 4.5 Heat transfer Heat transfer Significance

mechanism [7] mechanism
Conduction Occurs in fluids and solids that are in contact
with one another. The heat energy is transferred
from a warmer material to a cooler material
Pconduction dλ · A · (T 1 − T 2 )
Where
λ—Thermal conductivity of the material
(W/m °C)
T 1 , T 2 —Temperatures in °C
A—Area in the surface in m2
d—length in m
Convection Occurs between solid material and surrounding
air
Natural convection—occurs naturally when
colder air surrounds a warmer material
Forced convection—air flow around the
warmer material is forced by external means
such as fan or a blower
It is comparatively faster
Pconvection α · A · (T 1 − T 2 )
where,
α—convection coefficient (W/m2 °C)
T 1 , T 2— Temperatures in °C
Radiation This mode of heat transfer is due to
electromagnetic emissions when a transparent
medium surrounds a material
Pradiation S · E · A · (T 41 − T 42 )
Where
S—Stefan–Boltzmann constant (5.67 × 10−8
W/m2 K4)
E—Emissivity of the material
T 1, T 2—Temperatures in °C
There are few popular methods to minimize losses in the power sources using the
following cooling methods shown in Table 4.6.
Table 4.6 Cooling system and their significance [7]

Cooling mechanism Significance Image
Heatsink A type of conduction cooling,
that allows heat transfer via
direct contact with a cooler
component. Exists in two
primary forms—with fins and
without fins. A heat sink with fins
is beneficial as it allows a faster
heat dissipation because of
increased contact surface with the
air and it is usually coupled with
the forced air cooling mechanism
for a maximum dissipation
Cold plate A type of conduction cooling,

where a metal plate of high
thickness is used to transfer heat
away from a material. Cold
plates are very effective thermal
management solutions when
space and simplicity are of
greater concern. The power
source design also accounts for
enabling the heat to be
transferred from cold plate
through forced air flow or natural
air flow
Liquid cooling It consists of cold plate with an
attached pipe consisting a
running closed loop of a liquid.
The liquid is water and sometime
glycol is mixed to some
percentage. It is an optimal
solution in tight spaces where
fans or heat sinks cannot be
placed. This type of cooling is
more efficient but expensive and
complicated as it requires more
engineering and material efforts
(continued)
References 137

Cooling mechanism Significance Image
Natural cooling A type of convection cooling
technique. The natural air carries
away the heat surrounding the
power device components and fill
the surrounding with a colder air.
It is least expensive cooling
technique since there are no
requirements for any additional
components to set up the air
Forced cooling A type of convection cooling

technique. Small fans are placed
around the power source unit. It
is efficient method and improves
the lifetime of power devices
References
1. ESAB Welding &Cutting Products, Basic Welding Filler Metal Technology (2000). Available:
http://www.esabna.com/euweb/awtc/lesson1_1.htm
2. T.W. Eagar, The physics of arc welding processes. Adv. Join. Technol. 61–68
3. P. Kah, H. Latifi, R. Suoranta, J. Martikainen, M. Pirinen, Usability of arc types in industrial
welding. Int. J. Mech. Mater. Eng. 9(1), 1–12 (2014)
4. G.E. Cook, The Effect of High-Frequency Pulsing of Welding Arc,” vol. 5, no. 2 (1985), pp. 2–7
5. YAS HAM STD, Engineering, Costs, Quality, and Safety, vol. 5 (1997)
6. D. Cuiuri, J. Norrish, Verstatile Welding Power Source Controller for Research and Product
Development (2006)
7. E. Acha, V.G. Agelidis, O. Anaya-Lara, T.J.E. Miller, Power Electronic Control in Electrical
Systems—CH 2 (Elsevier, India, 2002)
8. C. Wang, Z. Wang, Q. Xu, Study on dynamic characteristics of inverter arc welding power
supply based on double-loop control, in 2009 IEEE 6th International Power Electronic Motion
Control Conference IPEMC ’09, vol. 3 (2009) pp. 1609–1612
9. C. Gopalakrishnan, K. Udayakumar, T.A. Raghavendiran, Survey of harmonic distortion from

power quality measurements and the application of standards including simulation, in Trans-
mission and Distribution Conference and Exhibition, vol. 2 (2002), pp. 1054–1058
10. J.C. Wiseman, B. Wu, Active damping control of a high-power PWM current-source rectifier
for line-current THD reduction. IEEE Trans. Ind. Electron. 52(3), 758–764 (2005)
11. TWI Limited, Measurement and Analysis of Magnetic Fields from Welding Processes (2005),
p. 51
Chapter 5
Sensors for Welding Data Acquisition
Sensors in arc welding process can obtain information about the geometry and posi-
tion of the workpiece, electrical parameters like voltage, current, wire feed rate, in
addition to providing suitable form of control over the welding process parameters.
The weld quality mainly depends on weld parameters like current, voltage, wire
speed and wire feed rate. The welding torch also influences the flow of the material.
The input heat for melting the component and constant heat flow is directly related
with the welding torch control and it further influences the weld quality. This chapter
presents the fundamental knowledge on data acquisition in arc welding process with
real-time case study illustrations.
5.1 Data Acquisition System
Data Acquisition (DAQ) (the system shown in Fig. 5.1) is the process of sampling
signals which involves the measurements of any physical or electrical quantities
such as current, voltage, pressure and temperature with a computer. A DAQ system
comprises of devices such as sensors, measurement hardware and a computer with
a programmable software. In comparison to traditional measurement systems, PC-
based DAQ systems consist of fast processing power, productivity and display, cost-
effective and powerful measurement solution [1].
Data acquisition begins with the physical parameter measurement. This physical
parameter could be the temperature of a room, the intensity of a light source, the
pressure inside a chamber, the force applied to an object or many other things. An
effective data acquisition system can measure all these different parameters.

https://doi.org/10.1007/978-981-13-0806-2_5
140 5 Sensors for Welding Data Acquisition
Sensor DAQ Device Computer

Bus
Analog to
Signal Digital Driver Application
Conditioning Converter Software Software
Fig. 5.1 Elements in DAQ system
5.1.1 What Are Sensors and Transducers?
A sensor is defined as a device that receives and responds to a signal or stimulus.

The purpose of a sensor is to respond to an input physical stimulus and to convert it
into an equivalent electrical signal which is compatible with electronic circuits. In
simpler terms, sensor is a translator of a nonelectrical parameter into an electrical
parameter. When it is said as ‘electrical’, it refers a signal that can be channelled,
amplified and modified using electronic devices. The output signal of sensor may
be in the form of charge, voltage or current. They may be further defined in terms
of amplitude, phase or frequency [2]. A sensor is also called ‘a transducer’, which
converts a physical parameter into a measurable electric signal. Some sensors may
require added components and circuitry to produce a signal that can accurately read
using a DAQ device.
A sensor may incorporate several transducers as shown in Fig. 5.2. The term
sensor is distinguished from transducer. The latter is a converter of one type of signal
into another, while the former converts any type of signal into electrical signal. An
example of a transducer is a loudspeaker which converts an electrical signal into a
variable magnetic field and, subsequently, into acoustic signal. A sensor does not
function by itself and is always a part of a larger system that incorporate other
detectors, signal conditioners, signal processors, memory devices, data recorders
and actuators. Sensors are of two kinds: passive and active. A passive sensor does
not need external energy source and directly generates an electric signal in response
to an external stimulus. Some examples are a thermocouple, a photodiode and a
piezoelectric sensor. The active sensors require external power for their operation,
which is called an excitation signal which is modified by the sensor to produce
the output signal. The active sensors sometimes are called parametric because their
own properties change in response to an external effect and these properties can be
subsequently converted into electric signals. The ability of a data acquisition system
to measure different phenomena depends on the transducers to convert the physical
parameter into signals measurable by the data acquisition hardware. Transducers are
synonymous with sensors in data acquisition systems. There are specific transducers
for many different applications, such as measuring temperature, pressure, or fluid
flow. Table 5.1 shows commonly used physical parameters and the transducers used
to measure them.
5.1 Data Acquisition System 141
Sensor
Electric
Stimulus
e2 e3 Signal
e1
Transducer 1 Transducer 2 Transducer 3
Fig. 5.2 Sensor in terms of transducer
Table 5.1 Phenomena and existing transducers

Phenomenon Transducer
Temperature Thermocouple, RTD, Thermistor
Light Photosensor
Sound Microphone
Force and Pressure Strain Gauge, Piezoelectric Transduce
Position and displacement Potentiometer, LVDT, Optical Encoder
Acceleration Accelerometer
pH pH Electrode
5.1.2 Signals
Different signals demand different measuring systems. For this reason, it is important
to understand the different types of signals and their corresponding attributes. Signals
can be categorized into two groups:
• Analog
• Digital.
5.1.2.1 Analog Signal
An analog signal can exist at any value with respect to time. Few examples of analog
signals include voltage, temperature, pressure, sound and load. The three primary
characteristics of an analog signal are level, shape and frequency (Fig. 5.3).
Level
Because analog signals can take on any value, the level gives vital information about
the measured analog signal. The intensity of a light source, the temperature in a room
and the pressure inside a chamber are all examples that demonstrate the importance
Fig. 5.3 Primary characteristics of analog signal
of the level of a signal. When we measure the level of a signal, the signal generally
does not change quickly with respect to time. The accuracy of the measurement,
however, is very important. We should choose a data acquisition system that yields
maximum accuracy to help with analog level measurements.
Shape
Some signals are named after their specific shapes—sine, square, sawtooth and tri-
angle. The shape of an analog signal can be as important as the level because by
measuring the shape of an analog signal, we can further analyse the signal, includ-
ing peak values, DC values and slope. Signals where shape is of interest, generally
change rapidly with respect to time, but system accuracy is still important. The anal-
ysis of heartbeats, video signals, sounds, vibrations and circuit responses are some
applications involving shape measurements.
Frequency
All analog signals can be categorized by their frequencies. Unlike the level or shape
of the signal, frequency cannot be directly measured. The signal is analysed using
software to determine the frequency information. This analysis is usually done using
an algorithm known as the Fourier transform. When frequency is the most important
piece of information, which considers both accuracy and acquisition speed. Although
the acquisition speed for acquiring the frequency of a signal is less than the speed
required for obtaining the shape of a signal, we still must acquire the signal at
the requisite speed to avoid the loos of the pertinent information while acquiring
5.1 Data Acquisition System 143
Fig. 5.4 Primary characteristics of digital signal
the analog signal. The condition that stipulates this speed is known as the Nyquist
Sampling Theorem. Speech, telecommunication and earthquake analysis are some
examples of common applications where the frequency of the signal must be known.
5.1.2.2 Digital Signal
A digital signal is discrete in nature. Instead, a digital signal has two possible levels:
high and low. Digital signals generally conform to certain specifications that define
the characteristics of the signal. They are commonly referred to as transistor-to-
transistor logic (TTL). TTL specifications indicate a digital signal to be low when
the level falls within 0–0.8 V, and the signal is high between 2 and 5 V. The useful
information that can be measured from a digital signal includes the state and the rate
Fig. 5.4.
State
The state of a digital signal is essentially the level of the signal—on or off, high
or low. Monitoring the state of a switch—open or closed—is a common application
showing the importance of knowing the state of a digital signal.
Rate
The rate of a digital signal defines how the digital signal changes state with respect to
time. An example of measuring the rate of a digital signal includes determining how
fast a motor shaft spins. Unlike frequency, the rate of a digital signal measures how
often a portion of a signal occurs. A software algorithm is not required to determine

the rate of a signal.
5.1.3 What Is a DAQ Hardware?
DAQ hardware performs as the interface between signals from the real world and
computer. It mainly functions as a device that digitizes incoming analog signals,
so that a computer can interpret them. The three key components of a DAQ device
used for measuring a signal are: the signal conditioning circuitry, Analog-to-Digital
Converter (ADC) and computer bus. Many DAQ devices are also utilized for other
performances like automating measurement systems and processes. It primarily func-
tions as a device that digitizes incoming analog signals so that the computer interprets
them. Other data acquisition functionality includes the following:
• Analog input/output
• Digital input/output
• Counter/timers
• Multifunction—a combination of analog, digital, and counter operations on a sin-
gle device.
The most readily available platform is the desktop computer. PCI DAQ boards
that plug into any desktop computer are available now. In addition, DAQ modules for
PXI/Compact PCI, a more rugged modular computer platform specifically for mea-
surement and automation applications are available. For distributed measurements,
compact field point platform delivers modular I/O, embedded operation and Ethernet
communication are available. For portable or handheld measurements, DAQ devices
for USB and PCMCIA work with laptops or Windows Mobile PDAs are used. In
addition, DAQ devices for PCI Express, the next-generation PC I/O bus and for PXI
Express, the high-performance PXI bus are introduced.
5.2 Physical Principles of Sensing
Some of the commonly detectable phenomena are as follows:

• Biological,
• Chemical,
• Electric,
• Electromagnetic,
• Heat/Temperature,
• Magnetic,
• Mechanical motion (displacement, velocity, acceleration, etc.),
• Optical and
5.2 Physical Principles of Sensing 145
• Radioactivity.
Some of the commonly measured quantities are tabulated in Table 5.2,
5.2.1 Characteristics of Different Sensor Types
The sensors are characterized based on the following criteria:

(i) Active versus Passive: Does sensor draw energy from the signal?
Active sensors require an external source of power such as excitation voltage
that provides most of the output power of the signal. While the passive sensors
output power is entirely provided by the measured signal without an excitation
voltage.
(ii) Digital versus Analog: Is the signal discrete or continuous?
The signal produced or reflected by digital sensor is binary. The signal produced
by the analog sensor is continuous and proportional to the measurand.
(iii) Null and deflection methods
Deflection: The signal produces some physical deflection effect closely related
to the measured quantity and transduced to be observable. Null: The signal
produced by the sensor is counteracted to minimize the deflection. That oppos-
ing effect necessary to maintain a zero deflection should be proportional to the
signal of the measurand.
(iv) Input–Output configuration
It depends on: Method of inherent insensitivity, Method of high gain feedback,
Method of calculated output corrections, Method of signal filtering and Method
of opposing inputs.
(v) Contact–Non-contact Type
Physical contact exists between contact type sensors.
Table 5.2 Commonly measured parameters

Stimulus Parameters
Acoustic Wave (amplitude, phase, polarization), Spectrum, Wave Velocity
Biological and Fluid Concentrations (Gas or Liquid)
Chemical
Electric Charge, Voltage, Current, Electric Field (amplitude, phase, polarization), Conductivity,
Permittivity
Magnetic Magnetic Field (amplitude, phase, polarization), Flux, Permeability
Optical Refractive Index, Reflectivity, Absorption
Thermal Temperature, Flux, Specific Heat, Thermal Conductivity
Mechanical Position, Velocity, Acceleration, Force, Strain, Stress, Pressure, Torque
Table 5.3 Sensor terminologies

Terminology Definition
Accuracy The agreement between the actual value and the measured value
Resolution The change in measured variable to which the sensor will respond
Repeatability Variation of sensor measurements when the same quantity is measured
several times
Sensitivity Upper and lower limits of the variable that can be measured
Linearity It is an expression of the extent to which the actual measured curve of a
sensor departs from the ideal curve.
Sensitivity It is the minimum input of physical parameter that will create a detectable
output change
Hysteresis A sensor should be capable of accommodating changes of the input parameter
regardless of the direction of change.
offset It is the difference between the actual output value and the specified output
value under some particular set of conditions
Table 5.4 Selection criteria for sensors

Environmental Temperature range, Humidity effects, Corrosion, Size, Range Protection,
Factors Susceptibility to Electromagnetic interference, Ruggedness, Power
consumption and Self-test capability
Economic Factors Cost, Availability and Lifetime
Sensor Sensitivity, Range, Stability, Repeatability, Linearity, Error, Response
Characteristics time and Frequency response
5.2.2 Basic Terminologies
Some of the basic sensor terminologies are tabulated in Table 5.3.
5.2.3 Choosing a Sensor
Selection of suitable sensor depends on the following techno-economic factors and

characteristics (Table 5.4).
According to [3] sensor technology in welding, following classification can be
made (Fig. 5.5):
(i) Wire touch sensing—after applying a sensing voltage to the weld wire, the robot
is programmed to move to a series of positions relative to the weld joints. The
tool point position is recorded when the wire touches the part and the voltage
drops to zero. After a series of touches, the original program is adjusted.
(ii) Through-arc sensing—the robot is programmed to weave the arc across the
weld joint which results in a current and voltage change in the weld power
5.2 Physical Principles of Sensing 147
Arc Welding
Sensors
Wire touch Through-arc Vision guided line Vision guided

sensing sensing scan system circular system
Fig. 5.5 Arc Welding Sensors Classification
Sensor Input
Analog to
Buffer Low pass Active Communication
Digital Microcontroller
Amplifier Filter Port
Converter
Signal Conditioning
Fig. 5.6 Block diagram representation of signal conditioning
supply. The robot controls offsets in the programmed trajectory to bring the
weld current back to a specified level.
(iii) Vision-guided line scan systems—a laser camera is mounted at a particular
distance ahead of the weld torch. An accurate position of the weld, down to
0.1 mm, and process variables such as gap and joint angles are measured.
(iv) Vision-guided circular systems—using a circular scan rather than a line scan,
three-dimensional data can be obtained from a single measurement. Usually,
this method is slower than the line scan system mentioned above.
5.3 Key Measurement Components of a DAQ Device
5.3.1 Signal Conditioning
Signals from sensors are generally noisy and direct measurements of signal are not
conductive. Signal conditioning circuitry manipulates a signal into a suitable form
of input into an ADC. This circuitry includes attenuation, amplification, isolation
and filtering. Some DAQ devices include built-in signal conditioning system which
is intended to measure the specific type of sensors. Figure 5.6 represents the general
block diagram of a signal conditioning process.
5.3.2 Analog-to-Digital Converter (ADC)
Analog signals from sensors are converted into digital before they are manipulated
using a computer. An ADC is a chip, providing a digital representation of an analog
signal at an instant. ADC takes periodic ‘samples’ of the signal at a predefined rate
and these samples are transferred to a computer bus, where the original signal is
reconstructed from the samples in software.
5.3.3 Computer Bus
A computer is connected to DAQ device using a slot. The computer bus serves as
the communication interface between the DAQ device and computer to measure
data’s and for sending information. DAQ devices are connected to computer buses
using USB (Universal Serial Bus), PCI, PCI Express, RS232, RS485, RS422, GPIB
(General Purpose Interface Bus) and Ethernet (as shown in Table 5.5). More recently,
DAQ devices have become available for 802.11 Wi-Fi for wireless communication.
There are some types of buses which offers different advantages for diverse types of
applications.
5.4 Role of Computer in a DAQ System
A programmable software in the computer controls the functions of the DAQ device
and is used for visualizing, processing and storing measured data. Many computers
are used for different types of applications. A desktop may be used in a lab for its
processing power, a laptop may be used in the field for its portability, or an industrial
computer may be used in a manufacturing plant for its ruggedness. LabVIEW is one
such graphical programming software which enables real-time data acquisition and
online monitoring of a process. The types of software’s used in computerized DAQ
system are given as follows:
5.4.1 Application Software
Application software enables the interaction between the user and computer for
analysing, acquiring and presenting measured data. It is also a prebuilt application
with predefined functionality and programming environment for building applica-
tions with custom functionality. Example: National Instruments LabVIEW is an
application software for real-time data acquiring in welding process.
5.4 Role of Computer in a DAQ System 149
Table 5.5 Commonly used computer Bus

Type of BUS Image
USB
Ethernet
RS232
RS485
(continued)

Type of BUS Image
GPIB
PCI
PCI Express
5.4 Role of Computer in a DAQ System 151
Sensors
Technological (Process Geometrical (Geometry

Oriented) Oriented)
• Voltage
• Optical
• Current
• Arc sensor
• Wire feed speed
• Primary process variables • Tactile

• Secondary process variables • Non tactile
Fig. 5.7 Sensors for welding data acquisition
5.4.2 Driver Software
Driver software provides application software the ability to interact with a DAQ
device. It communicates with the DAQ device easily by extracting low-level hard-
ware commands. In general, DAQ driver software comprises of an Application Pro-
gramming Interface (API) that is used within a programming environment to build
application software [1]. NI-LabVIEW provides different driver software for various
DAQ modules and chassis.
5.5 Data Acquisition in Arc Welding Processes
The information about the weld quality is identified using different sensors. Sensors
in welding applications are mainly used for checking the components position from
start to end, process parameter changes at joint/grooves and joint tracking. They are
used during the time of welding or at pre-welding and post-welding conditions [4]
(Fig. 5.7).
5.5.1 Measuring Current and Voltage
Welding voltage is measured at the terminals of power source as shown in Figs. 5.8
and 5.9. However, this introduces small error due to the voltage drop across the
Current
Power source V
Fig. 5.8 Measurement of welding current
Wire feeder
A
Power source V
Fig. 5.9 Measuring the welding voltage between the wire feeder and the workpiece
welding cables. For perfect measurement of actual welding voltage, it is important

to measure the voltage close to the welding position [4].
In most arc welding processes, the voltage between the wire at the feed rollers
and the workpiece are measured. The wire itself serves as an ideal conductor thereby
avoiding the voltage drop across the welding cables. The wire and the wire feeder
must be electrically insulated from welding power source up to the welding gun. The
welding current is the same everywhere in the welding current circuit and is very
high and cannot pass directly through the instrument (as shown in Fig. 5.10). Some
other methods aregiven as follows:
• The volt drops across a small calibrated shunt resistor fitted somewhere in the weld-
ing circuit can be measured. To measure rapidly changing currents, it is important
to have low inductance for the shunt, otherwise, it would result in spurious addition
to the measured value.
• Instead of a shunt resistor, Hall effect sensors (shown in Fig. 5.10) are used which
are based on measurement of the magnetic field around a conductor.
Hall-effect sensors consist of a thin rectangular p-type semiconductor material such
as gallium arsenide (GaAs), or indium arsenide (InAs) passing a continuous current
through itself. When the sensor is placed in a magnetic field, the magnetic flux lines
exert a force on the semiconductor material which deflects the charge carriers to
either side of the semiconductor. This movement of charge carriers is a result of the
magnetic force experienced while passing through the semiconductor material. As
the electrons and holes move sideward, a potential drop is produced between the two
sides of the semiconductor material. Then, the movement of electrons through the
semiconductor material is influenced by the presence of an external magnetic field
5.5 Data Acquisition in Arc Welding Processes 153
Fig. 5.10 Hall-effect sensor
which is at right angles to it. The effect of generating a voltage using a magnetic field
is called the Hall Effect.
• Clip-on ammeters are simple for occasional measurements (Fig. 5.11).

• Current transformers are used to measure high alternating currents (Fig. 5.12).
Current Transformer is an instrument transformer used to convert a primary current

into a secondary current through a magnetic medium. Its secondary winding then
provides a reduced current which is used for detecting overcurrent, undercurrent,
peak current, or average current conditions. Its primary coil is always connected in
series with the main conductor, and hence it is also referred as a series transformer.
5.5.2 Wire Feed Speed
The wire feed speed is a major parameter which has to be controlled in order to
achieve a stable welding process. Mostly, the welding power source is controlled to
produce a constant voltage and the preset parameters (voltage and current). However,
a current will represent a certain wire feed speed and the common method is to apply
a constant voltage and wire feed speed while adjusting the current. By keeping the
wire feed speed at a nominal preset value, a stable welding process can be achieved.
Fig. 5.11 Clip-on ammeter
Fig. 5.12 Current transformer

In normal robot systems, the wire feed unit is mounted on the robot arm. However,
in some cases, longer conduits must be used. This results in wire feed speed variations
when the conduit is bent and twisted. In practice, a push-pull wire feed system should
be used to counteract this problem. Measuring the wire feed speed is a major issue
and for laboratory purposes, custom-made solutions can be built that measure the
speed at the contact tube. A more realistic approach in a production system is to
measure the controlled speed of the drive wheel of the feeder unit. However, this
must be complemented with securing the functionality and reliability of the feeder
system as its robustness is important for the resulting quality of the welds. Special
sensors with idler rollers that generate a signal proportional to the wire feed speed
are available [4]. For occasional check measurements, it is sufficient to use a tape to
measure the length of wire delivered by the feeder in a certain length of time, from
which the speed can be calculated by the formula
Wire length
Wire feed speed ((1))
Measured time
Digital wire feed rate meters are also available as shown in Fig. 5.13.
5.5.3 Shielding Gas Flow
The common method to measure the gas flow is to use a marked tapered glass
tube with a float inside. When using this to check performance, it should be held
vertically against the welding gun nozzle. Sensors that provide an electrical output
signal are also available. Digital shielding gas flow meters are also available as shown
in Fig. 5.14.
5.5.4 Temperature
It is important to measure the temperature of workpiece. A simple method is to

use special chalks that can change colour to indicate the temperatures as shown in
Fig. 5.15. A contact thermometer is another instrument with a measurement tip that
is pressed against the workpiece.
A thermocouple (as shown in Fig. 5.15) is appropriate for use when continuously
monitoring a temperature which works on the principle of Seebeck effect. It consists
of two dissimilar wires that are soldered together at one end. The contact between the
two metals produces a thermoelectric voltage that is proportional to the temperature.
If the other end of the wires is held at room temperature, the voltage reading is
directly related to the temperature of the soldered junction. It is important that the
soldered tip is in good thermal contact with the part to be measured by placing it in
a drilled hole.
Fig. 5.13 Digital wire feed

rate meter
Fig. 5.14 Shielding gas flow meter
Fig. 5.15 A temperature chalk

Fig. 5.16 Schematic of thermocouple
A pyrometer (as shown in Fig. 5.16) is an instrument that measures temperature

by measuring radiation from the object without direct contact. Pyrometers are also
available with electrical signal outputs for recording other purposes. If the temper-
ature distribution over a particular area has to be measured, a thermal imaging IR
camera can be used, which displays the temperatures in the form of assorted colours
on a screen (Fig. 5.17).
5.5.5 Sensors for Geometrical Parameters
Sensors for geometrical parameters must be able to obtain information about the
weld that relates to the geometry of the weld joint. This information is of immense
importance for the seam tracking to be done and use this information for quality
control of the weld. The challenge, however, is to use sensors during welding. Due to
the harsh environment with elevated temperature, liquid metal spattering, intensive
light and high currents, purpose-built sensors must be applied. The most commonly
used sensors are optical and through-arc sensors.
Optical sensors use the basic principle for detecting the weld joint during arc
welding. A laser beam that is projected in a scanning motion across the seam and
a CCD-array is used to measure the features of weld joint in combination with a
laser stripe. Variations of this method are in use, for example, the laser stripe may
not be a linear line on the weld joint but circular instead. In such case, the sensor is
more flexible to detect weld joints in corners. To measure the distance, the method
of triangulation is used as shown in Fig. 5.18
Red filter Absorption screen
Objective lens
Reference
temprature
Temprature source
lamp
Observer
Eye-piece
Outer tube
Multimeter
Rheostat
Battery
Fig. 5.17 Schematic of optical pyrometer
Fig. 5.18 Triangulation

method
Laser
Detector array
Laser beam
Lens
Scanning axis
A laser beam is focused on an object, and then the reflection from the object as
seen from a lens in the laser sensor is determined by the distance between the sensor
and the object. If the object is close to the sensor, then the angle between the outgoing
beam and the reflection through the focusing lens of the detector is large, while it is
small if the object is farther away.
Depending on the weld joint preparation and geometrical shape, the laser beam can
produce reflections like mirrors. For example, consider a V-groove weld joint where
the laser light will produce several reflecting positions but with different intensities
depending on the surfaces of the weld joint. Therefore, these sensors must have real-
time image processing capabilities to filter out reflections that do not belong to the
point of interest. It should be noted in this context that highly reflective materials
may cause problems during welding and a real test may be needed to verify the
functionality.
The widespread use of triangulation in welding is for seam tracking and this
requires measuring the weld joint geometry. This is achieved through a scanning
technique of the beam across the weld joint (Fig. 5.19). During the scanning, the
sensor acquires a two-dimensional picture of the joint profile as an array of 2D
coordinates. When the robot is moving, a weld joint geometrical model can be made
that contains a full 3D description of the joint.
These optical sensors have a capability for more than seam tracking. Information
that they acquire includes joint volume, gap size, misalignment, track welds, etc. This
information is useful for adaptive feedback control of both the welding power source
and the robot. From a control point of view, seam tracking is usually done with full
compensation of the position error. The nominal path is the assumed trajectory of
the weld joint and during tracking, the robot controller receives new target positions
from the sensor. The robot controller then overrides the nominal path by changing
the position of the Tool Centre Point (TCP) while keeping a constant orientation.
The benefits of this method lies in the fact that for a given nominal path, it is rather
straightforward to verify the ability of the robot to follow the path with some minor
Fig. 5.19 Scanning Laser source

principle of seam tracking
Motor Focussing
lens
CCD line
camera
Angle sensor Objective
Scanning
motion Viewing
mirror
Object
changes while keeping the orientation constant. The drawback is that the user has to
define and program a nominal path.
Instead, if the robot is only instructed where to start and where to end, it must be
able to follow the path measured by the sensor on the fly. This puts some additional
requirements on the robot system as it must be able to calculate the trajectory includ-
ing both target positions and orientations of the weld torch. Typical operating data of
a laser scanner is a scan sweep frequency of 10–50 Hz. If welding speed of 1.2 m/min
is assumed, it indicates one sweep per mm during welding. This is considered sat-
isfactory in most cases. However, new welding processes such as laser welding will
increase the welding speed considerably and for high requirements careful analysis
and trials are essential.
Acoustic signals can also be used to determine the metal transfer mode during
welding process. The main characteristic being used for determination of the short
arc metal transfer mode is the sound of reignition of the arc.
Ultrasonic testing of welds is a proven and efficient method for detection of
defects, irregularities, cracks, inclusions and other welding defects. The major advan-
tages of this method is its adaptability to versatile materials and show possibility of
detecting minute irregularities.
5.5.6 Arc Sensors
The first basic thing to achieve with a fully automatic robotic welding system is the
capability to follow precisely the joint to be welded. This is because the welding
quality depends very much on the welding pool position apart from its geometry. A
preprogrammed path cannot be obtained with the desired precision since deviations
from the programmed path are likely to occur due to deficient path definition. This
is also due to material plate deficiencies and the effect of heating the plates. Con-
sequently, an online joint recognition and seam tracking system must be available.
Several techniques have been used for joint detection and seam tracking, namely,
for welding robotic systems. Using the arc characteristics and exploiting the propor-
tional relationship between the welding current and the distance from the electrode
to workpiece was the pivotal approach. Seam tracking using a weaving motion and
the arc itself as the sensor, sometimes referred to as through-arc sensing, was intro-
duced in the 1980s. The principle behind the method is to make use of the change
in current when the distance between the contact tube and the workpiece varies. The
underlying principle is relatively easy and cost-effective. Therefore, it is the com-
monly used sensor for tracking methods in robotic welding based on gas metal arc
welding and related processes, like flux-cored arc welding, submerged arc welding,
etc. The approximate relationship between arc voltage (U), arc current (I) and the
contact tube to workpiece distance (l) is expressed by,
β3
U β1 I + β2 + + β4 l ((2))
I
where β 1, β 2, β 3 and β 4 are dependent on factors like wire, gas and the characteristics
of the welding power source. As can be seen from the equation above, when the value
‘l’ varies, the arc current will also change proportionally with opposite sign.
For implementing arc sensor technology, no additional sensors are needed which
makes them a very simple and cost-efficient technology. Furthermore, it is possible
to implement them into closed loop systems, and in comparison, to vision systems,
they are fast [4]. Also, arc sensor technology is a non-contact method and, therefore,
it does not intervene with the welding process.
5.5.7 Typical Sensors and Their Outputs
See Table 5.6.
5.6 Parameters of Arc Welding Sensors for Various

Applications
See Table 5.7.
5.7 Data Acquisition Using LabVIEW
LabVIEW (short for Laboratory Virtual Instrumentation Engineering Workbench) is

a platform and development environment for a visual programming language from
National Instruments. The graphical language is named ‘G’. Originally released for
the Apple Macintosh in 1986, LabVIEW is commonly used for data acquisition,
Table 5.6 Different sensor Parameter Sensor Output

outputs
Temperature Thermocouple Voltage
RTD Resistance
Thermistor Resistance
Force/pressure Strain gauge Resistance
Piezoelectric Voltage
Acceleration Accelerometer Capacitance
Position LVDT AC voltage
Light intensity Photodiode Current
Phototransistor Current
Table 5.7 Advantages and disadvantages of different arc welding sensors

Operation time Type of Advantages Drawbacks
sensors
Sensing Touch Can recognize three-dimensional Can defect elastically, using tactile
independent sensing offset of the workpiece. The wire tip probes it is difficult, if not
or the gas nozzle can serve as a impossible, to provide information
sensor. Can be used for accurate on the joint fit up
learning of the path before welding Poor weld joint repeatability
Preview Contact Relatively low cost. The mechanical Not adaptable to suit a variety of
sensing sensing probes lead the welding spots joint geometries
Inductive Largely used in industry, Different sensor is needed for each
sensing configurations with one pick-up coil type of joint, should stay very close
can provide a cross-seam or vertical to the joint
path correction signal
Capacitive Offer the opportunity to measure It is hard to extract a correction
sensing the distance between the workpiece signal in two directions from the
and an electrical conduction plate of capacity variations
small dimension
Acoustical Apart from seam tracking Line of sight must not deviate from
sensing application, an acoustical sensing the surface normal; another
system can be used to explore the limitation is the temperature
workpiece for obstacle and maybe dependence of the speed of the
to inspect a produced weld sound
Optical Can be used for seam tracking as To prevent accessibility limitation,
sensing well as for geometrical recognition it may require additional axes for
of the weld pool, to adapt process seam tracking, tremendous effort to
parameters in the case of possible introduce technical integration,
deviations regularly check the lens protection
On-the-spot Weld pool Dedicated to welding pool geometry There should be a clear
sensing observation and properties. The obtained image interpretation of the image by the
is processed, and pattern system, to give torch corrective
recognition algorithms are used to changes accordingly
extract the dimensions and form of
the weld pool. Different sensors can
be applied: optic sensing, thermal
sensing, real-time radiography,
weld pool oscillation sensing
Through the- No additional voluminous sensor The torch must be weaved during
arc needs to be fixed to the weld torch. welding. The dimension of the joint
sensing Its simple operation and must exceed some critical
implementation have made arc dimension, e.g. it is not applicable
sensing a commonly accepted for sheet metal. In addition, a signal
off-the-shelf technique can be obtained only after the arc
has been established. Therefore, it
cannot be used for finding starting
point of the weld
5.7 Data Acquisition Using LabVIEW 163
instrument control and industrial automation on a variety of platforms including

Microsoft Windows, various flavors of UNIX, Linux and Mac OS X.
LabVIEW is very powerful in creating DAQ applications. LabVIEW includes a
set of VIs that let us configure, acquire data from and send data to DAQ devices.
Often, one device can perform a variety of functions, such as analog-to-digital (A/D)
conversion, digital-to-analog (D/A) conversion, digital I/O and counter/timer opera-
tions. Each device supports different DAQ and signal generation speeds. Also, each
DAQ device is designed for specific hardware, platforms and operating systems.
National Instruments, the inventor of LabVIEW, also make DAQ devices, so the
integration with the DAQ devices from NI and the LabVIEW software is seamless
and makes it easy to perform I/O operations from the LabVIEW environment.
The parts are given as follows:
• Physical input/output signals,
• DAQ device/hardware,
• Driver software,
• Software application (Application software).
5.7.1 Physical Input/Output Signals
A physical input/output signal is typically a voltage or current signal. A voltage signal

can typically be a 0–5 V signal, while a current signal can typically be a 4–20 mA
signal.
5.7.2 DAQ Device/Hardware
DAQ hardware acts as an interface between the computer and the outside world.
It primarily functions as a device that digitizes incoming analog signals so that the
computer can interpret them as A DAQ device (Data Acquisition Hardware) which
usually has these following functions:
• Analog input,
• Analog output,
• Digital I/O,
• Counter/timers.
5.7.3 Driver Software
Driver software is the layer of software for easily communicating with the hard-
ware. It forms the middle layer between the application software and the hardware.
Driver software also prevents a programmer from doing register-level programming

or complicated commands to access the hardware functions. Driver software from
National Instruments are the following:
• NI-DAQmx
• NI-DAQmx Base.
The DAQ Assistant, included with NI-DAQmx, is a graphical, interactive guide

for configuring, testing, and acquiring measurement data. With a single click, you
can even generate code based on your configuration, making it easier and faster to
develop complex operations. Since DAQ Assistant is completely menu-driven, you
will make fewer programming errors and drastically decrease the time from setting
up your DAQ system to taking your first measurement.
5.7.4 Application Software
Application software adds analysis and presentation capabilities to the driver soft-
ware. Our software application normally does such tasks as follows:
• Real-time monitoring,
• Data analysis,
• Data logging,
• Control algorithms,
• Human machine interface (HMI).
5.7.5 Measurement and Automation Explorer
Measurement and Automation Explorer (MAX) provides access to your National

Instruments devices and systems. With MAX, we can
• Configure your National Instruments hardware and software;
• and edit channels, tasks, interfaces, scales, and virtual instruments;
• Execute system diagnostics;
• View devices and instruments connected to your system;
• Update National Instruments software.
In addition to the standard tools, MAX can expose item-specific tools which can
be used to configure, diagnose or test our system depending on which NI products
are installed. As we navigate through MAX, the contents of the application menu
and toolbar change to reflect these new tools.
5.7 Data Acquisition Using LabVIEW 165
5.7.6 DAQ Assistant
The DAQ Assistant, included with NI-DAQmx, is a graphical, interactive guide

for configuring, testing, and acquiring measurement data. With a single click, we
can even generate code based on our configuration, making it easier and faster to
develop complex operations. Because DAQ Assistant is completely menu-driven,
the programming errors and decreases the time of setting up your DAQ system to
record the first are minimized measurement. One such example is presented.
1. When DAQ Assistant icon is dragged on our Block Diagram, the following
window appears as shown in Fig. 5.20. In this window, we need to select either
‘Acquire Signals’ (i.e. Input Signals) or ‘Generate Signals’ (i.e. Output Signals).
2. Select Acquire Signals → Analog Input → Voltage.
3. In the next window as shown in Figs. 5.21 and 5.22 select which Analog Input
is to be used. Select ai0 (Analog Input channel 0) and click Finish.
4. The following window appears (Fig. 5.23).
5. In the Timing Settings select ‘1 Sample (On Demand)’ as shown in Fig. 5.24.
The input range can be selected.
6. We may also rename the name of the channel (right-click on the name) (Fig. 5.25).
7. You are now finished with the configuration. Click OK in the DAQ Assistant
Window. The DAQ Assistant icon appears on the Block Diagram (Fig. 5.26).
Example: Wire the data output to a numeric indicator like this (and hit the Run
button). Then, numeric indicator will show such as Fig. 5.27
Fig. 5.20 NI Measurement and automation explorer

Fig. 5.21 DAQ assistant window
If you want a ‘continuous’ acquisition, put a While loop around the DAQ Assistant
like this (Fig. 5.28).
5.8 Case Study 1: Measurement of Temperature During

Joining of 316L Stainless Steel by CMT Process
5.8.1 Process Details
Base Material: Stainless steel 316L sheet (2 mm thickness);

Filler Wire: Stainless steel 308 (1.2 mm diameter);
Dimension: 10 cm * 10 cm;
Welding process: Cold Metal Transfer.
5.8 Case Study 1: Measurement of Temperature During Joining of … 167
Fig. 5.22 Channel selection in DAQ assistant
5.8.2 Description of DAQ Unit
NI-9219 (100 S/s/channel, 4-Channel C Series Universal Analog Input Module) is

used to measure the signals from the K-type thermocouples (pin type) with built-in
voltage and current excitation. The K-type thermocouple (chromel–alumel) is the
most commonly used general purpose thermocouple. It is inexpensive, and vari-
ety of probes are available for the range of −200 to +1350°C. It has a sensitivity of
approximately 41 µV/°C. Each channel is selectable individually to perform a differ-
ent measurement. One junction (hot) of each of the three thermocouples are placed
1 mm away from the weld zone and are fixed at equidistant positions in the base
material. The cold junction is inserted in one channel of the NI-9219 module. Using
the DAQ assistant block in LabVIEW, the temperature of the three thermocouples
T1, T2 and T3 are recorded and saved in the computer (Fig. 5.29).
Fig. 5.23 Configuration in DAQ assistant
5.8.3 Experimental Data
See Tables 5.8 and 5.9.
5.8.4 Temperature Plots
The temperature plots for three different parameter settings are shown in Figs. 5.30,
5.31, 5.32, 5.33, 5.34 and 5.35.
5.9 Case Study 2: Characterization of Gas Metal Arc Welding … 169
Fig. 5.24 Timing settings in DAQ assistant
5.9 Case Study 2: Characterization of Gas Metal Arc

Welding System Using DAQ
The data acquisition system employed in this study is an Arc Tracker, which is self-
contained data acquisition (DAQ) system manufactured by Lincoln Electric Com-
pany. The purpose of the DAQ unit is to obtain the welding data using a computer
(Fig. 5.36).
5.9.1 Description
The Arc Tracker system shown in Fig. 5.35, acquires voltage and current data at a
rate of 120 kHz. The rate of the controller in Lincoln Electric Power Waves 455R
and the main power supply model studied is about 10 kHz. This ensures that the data
observed at any change of state in the power supply is observed in the DAQ system.
This study utilized Lincoln Electric’s Power Wave Manager Software which is used
to connect the power supply directly and observes the operation of power supply.
This software can send the current pulse to the system and changes the calibration
set point of the power supply and represents the output in graphs. This results in
providing false offset voltage of the parameter setting and troubleshoots the entire
system.
Fig. 5.25 Renaming in DAQ assistant

5.9 Case Study 2: Characterization of Gas Metal Arc Welding … 171
Fig. 5.26 DAQ assistant icon
Fig. 5.27 Illustration 1
Fig. 5.28 Illustration 2
5.9.2 Welding Procedure
The study is undertaken to monitor the system functioning. The control architectures
and effects of system circuitry were relatively unexplored. The system model of the
welding circuit is shown in Fig. 5.37 and 5.38.
Temperature Acquisition System
CDAQ with
LabVIEW
NI-9219
Software Module
Three K-type thermocouples

placed equidistantly
SS316L Base
Material
Fig. 5.29 Temperature acquisition system schematic
Table 5.8 Experimental Parameters

Trial no Weld current Voltage (V) Travel speed Wire feed Weld time (S) Heat input
(A) (mm/min) rate (m/min) (J/mm)
1 90 15.9 400 3.1 14 228.15
2 90 15.9 450 3.1 13.3 202.8
3 90 15.9 500 3.1 12 182.52
4 85 15.8 400 2.9 14 214.2
5 85 15.8 450 2.9 13.3 190.4
6 85 15.8 500 2.9 12 171.36
7 80 15.5 400 2.6 14 198
8 80 15.5 450 2.6 13.3 176
9 80 15.5 500 2.6 12 158.4
The main goal of the study is to determine the differences in welding parameters
across various power supplies. The weld system is established and bead on plate welds
are performed on the sample for different power sources. The weld parameters are
maintained at a constant, and testing is done to find the intrinsic difference of the
system. All data is collected using Arc tracker and at 120 kHz of sampling rate in
software. Voltage and current are recorded, then the acquisition is triggered at 10 V,
and data is stored. The stored signals are investigated for irregularities such as high
or low voltage power drops and spikes. Voltage and current are recorded along with
the arc stability and process or plate changes.
5.10 Results
This case study provides a knowledge of how to standardize welding systems.

The initial step is to characterize the welding system which requires understand-
ing the differences between systems and the relationship between input and output
5.10 Results 173
Table 5.9 Measured peak temperature and time

Trial no Current Travel Thermocouple T1 Thermocouple T2 Thermocouple T2
(A) speed
(mm/min)
Peak tem- At time Peak tem- At time Peak tem- At time
perature (s) perature (s) perature (s)
(°C) (°C) (°C)
1 90 400 79.214 9.6 562.638 15.6 135.464 33.6
2 90 450 60.83 8.4 194.95 24.6 72.65 38.4
3 90 500 52.46 50.4 254.83 33.6 105.34 29.4
4 85 400 60.14 44.4 297.65 24 108.67 40.2
5 85 450 62.88 24 310.56 36 127.1 33.6
6 85 500 47.78 59.4 143.07 24.6 73.83 40.8
7 80 400 45.22 65.4 237.09 31.8 91.71 45.2
8 80 450 40.79 24 262.99 31.8 64.67 23.4
9 80 500 65.8 37.2 222.28 25.2 157.75 33.6
Fig. 5.30 Temperature plot 1
parameters. Testing results include simple voltage and current measurements based
on different power sources from the same variations. Figure 5.39 shows the graph of
voltages recorded from six different power supplies in three different systems.
The error is calculated as V error V set − V weld such that a negative error caused
by the weld voltage is above the set voltage and a positive error is caused by the weld
voltage being below the set voltage.
Figure 5.40 shows the same power supplies and their current error related to the
voltage setting. The power supplies which has higher voltage output than the others

5.10 Results 175

Fig. 5.36 Arc Tracker system

5.10 Results 177
Power
Controller
Supply
Wire feed Robot

Wire
Torch
feeder
Gas
Voltage Voltage
DAQ
Ground Welding System
Voltage Process
Current
Weld
Fig. 5.37 System Block diagram
+ Vposcable - + Vtorch -
Welding Wire Feeder
Power Supply +
Positive sense lead
Vweld
AC Work -
V
Work sense lead +
A Vfixturing
-
Fixturing Fixturing
- VGndcable +
Fig. 5.38 Circuit diagram of welding system

Fig. 5.39 Set voltage versus measured voltage
Fig. 5.40 Error in voltage for six power supplies
have a higher current output in this graph, and vice versa for the power supplies with
lower voltages (Fig. 5.41).
References 179
Fig. 5.41 Current readings versus voltage
References
1. National Instruments (2013). Available: http://www.ni.com/data-acquisition/

2. J. Fraden, Handbook of Modern Sensors, 4th ed. Springer (2010)
3. I. Garašić, Sensors and their classification in the fusion weldingtechnology. Teh. Vjesn. Gaz.
22(4), 1069–1074 (2015)
4. P. Kah, M. Shrestha, E. Hiltunen, J. Martikainen, Robotic arc welding sensors and programming
in industrial applications. Int. J. Mech. Mater. Eng. 10(1), 13 (2015)
Chapter 6
Optimization in Arc Welding Process
6.1 Introduction to Optimization
Optimization is a valuable tool in making decisions and in analysing physical systems.

In mathematical terms, optimization is the process of determining the best solution
achievable close to desired value among the set of all feasible solutions.
6.1.1 Constructing a Model
The first step in optimization process is to construct an appropriate model; modelling

is the process of identifying and expressing the variables, objectives and constraints
of a problem in mathematical terms ([1] and [2])
• An objective is a quantitative measure of the performance of the system which
has to be minimized or maximized. In arc welding, to improve the bead width,
DOP and mechanical properties of the welded materials (such as shear strength,
tensile strength, etc.), and while fitting the experimental data to a model, the total
deviation of the observed parameters from the predicted parameters should be
minimized.
• The variables are the components of the system that is to be fine-tuned. In welding,
the variables may be welding current, wire feed speed, whereas, in data fitting, the
variables would be the parameters of the model.
• The constraints are the functions that describe the relationships among the vari-
ables and that define the allowable values for the variables. During the process
of welding, the amount of heat input is limited to a value beyond which in over
melting of material occurs.

https://doi.org/10.1007/978-981-13-0806-2_6
182 6 Optimization in Arc Welding Process
6.1.2 System Identification in Arc Welding
The suitable model for Arc welding process can be obtained using a limited number
of experimental measurements of the input and outputs parameters, with prior system
knowledge.
The three basic steps are given as follows:
1. Collecting useful data,
2. Choosing a convenient model set and
3. Computing the best model within the model, followed by a certain identification
criterion, i.e. parameter estimation.
System identification toolbox in MATLAB enables us to estimate the system model
by using the experimental data. This toolbox provides MATLAB functions, Simulink
blocks and an app for constructing mathematical models of dynamic systems from the
measured input–output data. The dynamic systems can be easily modelled using this
toolbox compared to the first principles modelling. The time-domain and frequency-
domain input–output data are used to identify continuous-time and discrete-time
transfer functions, process models and state-space models. The toolbox also provides
algorithms for embedded online parameter estimation.
6.2 Significance of Optimization in Welding
A frequent problem being faced by the manufacturer is the control of the process
input parameters to obtain a good joint. Traditionally, it has been necessary to deter-
mine the weld input parameters for every new welded product to obtain good joint
with the required specifications. To do so, it requires a time-consuming trial and
error approach, with weld input parameters chosen by the skill of the engineer or
machine operator [3]. Welds are then examined to determine whether they meet the
desired specification or not. Finally, the weld parameters can be chosen to produce a
welded joint that closely meets the joint requirements. Also, there are often a more
ideal welding parameters combination, which can be used only if the parameters are
determined. To overcome this problem, various optimization methods are applied to
define the desired output variables through the development of mathematical models
to specify the relationship between the input and output parameters. Presently, appli-
cations of Design of Experiment (DoE), evolutionary algorithms and soft computing
techniques are useful to develop an empirical relationship between the welding pro-
cess input and output parameters to determine optimal parametric window which
leads to the desired weld quality.
6.3 ANN-Based Optimization Techniques to Arc Welding Processes 183
6.3 ANN-Based Optimization Techniques to Arc Welding

Processes
Artificial Neural networks are advanced soft computing tools which find its promi-
nence in optimization of arc welding parameters. Due to the high-speed learning
capability of ANN, experimental data can be used to train the neural networks using
appropriate algorithms and they predict the characteristic parameters of the weld-
ments in showing good agreement with experimental values. This would help in
creating optimal parametric window.
6.3.1 Introduction to ANN
Artificial neural networks are simplified model of biological network structure. The
building block of ANN is the artificial neuron. These neurons interact with the real
world to receive and transmit data. Some of the neurons remain hidden in their hidden
layers. Neurons are connected to one another by synapse (shown in Fig. 6.1), and each
synapse is associated with a weight factor [4]. Artificial neural networks have their
profound role in real-time, complex and interactive applications like speech recog-
nition, pattern recognition, medicine, finance, weather forecasting and control of
manufacturing processes. The meritorious feature is that they can be modelled using
the experimental data without making any assumption. Several types of ANN algo-
rithms, like Back Propagation (BPN), Multilayer Perception (MLP), Self-Organizing
Map (SOM), Radial Basis Function (RBF), etc. are widely used models [5].
Input Layer
X1
Activation
W1 Summation Function
Yin
W2 Output
X2 ∑ f
Y
Wn
Xn
Fig. 6.1 Schematic representation of ANN

INPUT LAYERS OUTPUT LAYERS
X1
Y1
X2 Artificial
Neural
Network
Yn
(Hidden Layer
X3 with bias)
Xn
Fig. 6.2 ANN model
Each input neuron is weighted (multiplied by an adjustable weight value Wi ), and

their sum is passed through a non-linear function called as an activation function. The
transfer functions have a sigmoid shape, but they can also have non-linear functions
like piecewise linear functions or step functions. ANN uses an extra weight (W0 )
that enables to add an extra value to this sum before computing the output of the
transfer function. For a given ANN, let there be n + 1 input neurons with signals X 0
through X n and weights W 0 through W n . Usually, the X 0 input is assigned the value
+1, which acts as a bias input (threshold) with W 0 (as shown in Figs. 6.1 and 6.2)
For the above general model of ANN, the net input is calculated as
Yin x1 w1 + x2 w2 + · · · + xn wn (6.1)

n
Yin w0 + xi wi (6.2)
i0
The output is calculated by applying activation function
Y f (Yin ) (6.3)
The ANN can be created and trained using neural network toolbox in MATLAB
software. This toolbox provides easy access and implementation of ANN. The input
data set is divided into time steps based on the number of data samples. The time of
steps are classified into three categories, such as training, validation and testing;
(1) Training: The data set is fed into the neural network during training stage, and
the network adjusts itself according to the error.
Unsupervised
Input Output
Learning
Fig. 6.3 Unsupervised learning
(2) Validation: It is used to measure the network generalization. The training is

stopped when generalization stops improving.
(3) Testing: It does not have any effect on the training. It provides an independent
measure of the network performance at pre-and post-training.
Processing of ANN depends upon the following three building blocks:
6.3.1.1 Network Topology
A network topology is the arrangement of a network along its nodes and connecting
lines (Based on this arrangement the ANN is classified into feedforward ANN and
backpropagation ANN).
6.3.1.2 Adjustments of Weights or Learning
Learning is the method of modifying the weights of connections between the neurons
of a specified network. Learning in ANN can be classified into three categories,
namely, supervised learning, unsupervised learning, and reinforcement learning.
Unsupervised Learning
In this algorithm, there is no specific target or outcome variable to predict/estimate.
It is used for clustering population in diverse groups, which is widely used for seg-
menting customers in diverse groups for specific intervention (shown in Fig. 6.3).
Examples: Apriori algorithm, K-means.
Supervised Learning
This algorithm consists of a target/outcome variable (or dependent variable) which
is to be predicted from a given set of predictors (independent variables). Using these
set of variables, a function is generated that map inputs to desired outputs (a shown
in Fig. 6.4). The training process continues until the model achieves a desired level of
accuracy on the training data. Examples: Regression, Decision Tree, Random Forest,
KNN, Logistic Regression, etc.
Reinforcement Learning
Using this algorithm, the machine is trained to make specific decisions. When the
machine is exposed to an environment it trains itself continually using trial and error
Target
Error
Comparator
Supervised
Input Output
Learning
Fig. 6.4 Supervised learning
Reward Performance Evaluation
Reinforcement
Input Output
Learning
Fig. 6.5 Reinforced learning
and also it machine learns from past experience and tries to capture the best possible
knowledge to make accurate business decisions (as shown in Fig. 6.5). Examples:
Markov Decision Process.
6.3.1.3 Activation Functions
The additional force or effort is applied at the input to obtain an exact output. (Exam-
ple: Linear, sigmoid). An activation function is a decision-making function that deter-
mines the presence of particular neural feature. Non-linearity is needed in activation
functions because its aim in a neural network is to produce a non-linear decision
boundary via non-linear combinations of the weight and inputs.
6.3.2 Backpropagation Neural Network (BP-NN)
A Backpropagation (BP) Neural Network (NN) is a Multilayered Feedforward Neural

Network in which learning occurs by constantly modifying the connection weights
Inputs Weight
Adjustments
Output Desired
(Wij)
Responses Outputs
(Wjk)
Network
Errors
Back Propagating
Errors Output Layers (k)
Input Layers (i) Hidden Layers (j)
Fig. 6.6 Backpropagation neural network Schematic
between neurons in each layer and the neuron thresholds (as shown in Fig. 6.6). Due
to its robustness, BP-NN is used in many fields [6]. Backpropagation or ‘backward
propagation of errors’, is an algorithm for supervised learning of artificial neural
networks using gradient descent method.
1 2 1
E k(ek ) e2 (6.4)
2 2
where
ek error,
e a vector with element ek.
For a given artificial neural network and an error function, this method calculates
the gradient of the error function with respect to the weights of neural network. The
‘backwards’ part of the name means that the calculation of the gradient proceeds
backwards through the network, with the gradient of the final layer of weights being
calculated first and the gradient of the first layer of weights being calculated last.
Partial computations of the gradient from one layer are reused in the computation of
the gradient for the previous layer.
This backward flow of the error information allows for efficient computation of
the gradient at each layer versus the naive approach of calculating the gradient of
each layer separately. Backpropagation’s popularity has experienced a recent resur-
gence given the widespread adoption of deep neural networks for image recognition
and speech recognition. It is considered as an efficient algorithm, and modern imple-
mentations take advantage of specialized GPUs for performance.
Some of the best-suited algorithms to train ANN are Particle Swarm Optimiza-
tion (PSO) algorithm, Levenberg–Marquardt (LM) and Genetic Algorithm (GA) for
optimizing and predicting the weldment characteristics.
6.4 Development of PSO-Based Backpropagation Neural

Network
The above-mentioned algorithms and their applications are discussed in the following
context. The main advantage of PSO is its ease of implementation as there are only
few parameters to adjust [7].
6.4.1 Particle Swarm Optimization
The Particle swarm optimization (PSO) is a stochastic optimization technique which

is conceptualized by the social behaviour of flocking birds or schooling fishes (as
shown in Fig. 6.7). The system to be investigated is initialized with a population
with random solutions and searches for an optimal one by updating the generations.
To implement since there are only few parameters to adjust. In this study, PSO is
applied in training the artificial neural network [8].
PSO simulates the behaviors of bird flocking or fish schooling. Consider the
following scenario, a school of fishes searches randomly for food in an aquatic
area. There is only single piece of food in the search area. All fishes do not know
the location of food. But they know how far the food is in each iteration. So, the
effective strategy is to follow the fish which is nearest to the food. In PSO, each
single solution is a fish in search space and is called ‘particle’. All particles have
fitness values that are evaluated by optimized fitness function and have velocities
which direct the swimming of the fishes (particles). The particles swim through a
problem space by following the current optimum particles. PSO is first initialized
with a group of random particles called solutions, and then searches for optimal
Fig. 6.7 Representation of

PSO
6.4 Development of PSO-Based Backpropagation Neural Network 189
solution by updating the generations. In each iteration, each particle is updated by

following the two best values. The first one is the best solution called fitness. The
fitness value is stored which is called as pbest. Another best value that is tracked by
the particle swarm optimizer is the best value, obtained so far by any particle in the
population. This best value is a global best and called gbest. When a particle takes
part of the population as its topological neighbours, the best value is a local best and
is called lbest [5].
After identifying two best values, the particle updates its velocity and positions
based on the following two equations:
v[] v[] + c1 ∗ rand() ∗ (pbest[] − present[]) + c2 ∗ rand() ∗ (gbest[] − present[])

(6.5)
present[] present[] + v[] (6.6)
where
v[] the particle velocity
present[] current particle (solution)
pbest[] and gbest[] stated already
rand () random number between (0,1)
c1, c2 learning factors (usually c1 c2 2).
6.4.2 Development of BP-NN Using PSO Algorithm
The PSO algorithm is used to optimize the BP Neural Network. The algorithm
is combined with NN using initial weights between BP-NN layers and the initial
thresholds between the nodes and to optimize the distribution; execute searches
within solution space and to find optimal initial weights and thresholds for the BP-
NN at a faster convergence rate. Subsequently, the weights and thresholds obtained
by the BP-NN are used to train and test the sample data set. Figure 6.8 shows a
flowchart of this optimization model.
The steps involved in the algorithm are given as follows:
1. Initialize the parameters of the PSO; population size, position, velocity of parti-
cles and number of iterations.
2. Decide the structure of BP-NN and generate the population of particles.
Particles: X i (x i1 , x i2 … x iD )T , i 1, 2, … n
D RD1 + D1 D2 + D1 + D1 (6.7)
where
R number of nodes in input layer
Initialize the
speed and Determine the ANN
position of network topology
the particle
Calculate
Initialize the weights and
fitness value
threshold of BP-NN
of the particle
Search the
Obtain optimal values of
pbest and
weight and threshold
gbest
Update the
position and Calculate error
velocity
Update the Updating of weight and

fitness value threshold
Yes
of particle No
No
Update pbest
and gbest Terminal
conditions
satisfied?
Yes
Terminal
conditions
satisfied? Output
Fig. 6.8 PSO-BP Neural Network algorithm
D1 number of nodes in hidden layer

D2 number of nodes in output layer of the BP-NN.
3. Define the fitness function. The BP-NN is trained to set weights and thresholds
using the samples as per the Particle population, X i . The fitness function of the
X i , is denoted as

M−1
2
fiti yj − yj where, i 1, 2, . . . n (6.8)
j1
6.4 Development of PSO-Based Backpropagation Neural Network 191
4. Calculate the fitness value for each particle and evaluate the Population X. Cal-
culate the fitness value of each particle X i .
5. Update the optimal fitness value of the individual (Pbest) and the position of the
particles. For each of the particle X i , if the current fitness value is less than its
optimal fitness value, then use the current fitness value to update the optimal
fitness value and the position of particle X i .
6. Update the population fitness value (Gbest). For each of the particle X i , if the
fitness value is less than the optimal fitness value for current population, then use
the fitness value of the current particle to update the optimal fitness value.
7. Update the speed and position of particles iteratively to generate new populations
as per the below equations.
k
Vidk+1 ωVidk + c1r1 Pidk − X id
k
+ c2 r2 Pgd − X id
k
(6.9)
k+1
X id k
X id + Vidk+1 , d ∈ [1, D], i ∈ [1, n] (6.10)
where
ω inertia weight
k number of iterations
Vid velocity of particle Xi
C 1 and C 2 acceleration factors (not smaller than zero)
r 1 and r 2 are arbitrary numbers between 0 and 1.
8. The iteration is stopped based on the optimal value of initial weights and thresh-
olds of the BP-NN. The overall fitness variance of the particle population is used
to judge whether the algorithm converges or not. If the overall fitness variance
is less than given threshold, then the algorithm converges. The optimal solution
is the outputs; initial weights and thresholds. Else, the iterations continue. The
overall fitness variance is denoted by
n
f i − f avg 2
σ −
2
(6.11)
i1
f
where
n number of particles in the current population,
f fitness value,
fi fitness value of particle X i ,
f avg average fitness of the current population particle.
6.5 Development of Levenberg–Marquardt (LM)

Algorithm-Based Backpropagation Neural Network
6.5.1 Introduction to LM Algorithm
In this study, Levenberg–Marquardt (LM) Optimization is used to train the ANN. LM

is a virtual standard in non-linear optimization which outperforms gradient descent
and conjugate gradient methods. The function evaluations and gradient information
are done, and estimation of Hessian matrix is done [4, 5, 9]. Following equation is
to be solved by using Levenberg–Marquardt

J T J + λI δ J T E (6.12)
where
J Jacobian matrix for the system
λ Levenberg’s damping factor
δ weight update vector
E error vector.
The LM algorithm is designed to minimize the sum of squared error function of
the form
1 2 1
E k(ek ) e2 (6.13)
2 2
where
ek is the error,
e is a vector with element ek .
Based on weight update vector, the network weights are changed to achieve appropri-
ate solution. The cross product of the Jacobian matrix is known as the approximated
Hessian. The Levenberg damping factor is adjusted in each iteration.
6.5.2 Computing the Jacobian Matrix
The Jacobian matrix consists of all first-order partial derivatives of a vector function.
In the neural network, it is a N × W matrix, where N is the number of elements in
training set and W is the total number of parameters (weights + bias). It is created by
taking the partial derivatives of each output with respect to each weight denoted by
the form
6.5 Development of Levenberg–Marquardt (LM) Algorithm … 193
⎡ ⎤
∂ F(x1 ,w) ∂ F(x1 ,w)
∂w1
··· ∂wW
⎢ ⎥
⎢ ⎥
J ⎢
⎢
..
.
..
.
..
.
⎥
⎥ (6.14)
⎣ ⎦
∂ F(x N ,w) ∂ F(x N ,w)
∂w1
··· ∂wW
where F(x1 , w) is the network function evaluated at i-th input vector of training set
using the weight vector w and wW is the w-th element of the weight vector w.
6.5.3 Steps in Levenberg–Marquardt Algorithm
1. The LM involves solving the equation of J T E with different values of λ values

until the sum of squared error decreases. Each iteration involves following basic
steps.
2. Computation of the Jacobian matrix.
3. Computation of error gradient.
g JTE (6.15)
4. Approximation of Hessian using the cross-product Jacobian.
H JTJ (6.16)
5. Solving the equation
(H + λI )δ g (6.17)
find δ.
6. Updating of the network weights w using δ.
7. Recalculation of the sum of squared errors using updated weights.
8. If the sum of squared errors not decreased, discard the new weights, increase λ
using some adjustment factor and go to step 4.
9. Else decrease λ and stop.
The flow chart of the implemented ANN model is depicted in Fig. 6.9.
Collect data set
Select input variables and input them

into ANN Model
Load the Data set
Set number of hidden layers and

Validate
(Training, Validation, Testing)
Train the ANN Model with

Levenberg-Marquardt Algorithm Reaches the Maximum
Iteration
YES
Calculate the Network
Output
Evaluate result to the
requirement?
Estimate the Error

NO
Retrain the Network Model

Stop training
Network
Fig. 6.9 Flow chart of the LM based ANN model
6.6 Genetic Algorithm for Tuning the Neural Network
Genetic Algorithm (GA) is a search-based algorithm based on the concept of natural

selection and genetics. GA belongs to a branch of computation called Evolutionary
Computation. GA is developed by John Holland in University of Michigan along with
his students and colleagues which finds its successful results in various optimization
problems. In GA, there is a pool or a population of workable solutions for a given
problem. These solutions then undergo recombination and mutation like natural
genetics, resulting in new children. This process is repeated over various generations
[10]. Every individual (the solution) is assigned with a fitness value (depending on
the objective function) and the fit individuals are given a higher chance to mate and
result in more ‘fit’ individuals. This is in line with Darwinian Theory of ‘Survival
of the Fittest’. In this manner, better individuals or solutions over generations are
6.6 Genetic Algorithm for Tuning the Neural Network 195
Chromosome
0 1 0 1 0 1 0 1 0 0 1 0 1 0 1 0 1 0
0 Gene
Population
(Set of Chromosomes)
0 Allele
Fig. 6.10 Representation of chromosome and gene
evolved, until a stopping criterion is reached. Some of the key terminologies are
discussed below.
• Population—It is a subset of all the workable solutions for a given problem. The
population for a GA is analogous to the human population but instead of human
beings, GA has Candidate Solutions representing human beings.
• Chromosomes—A chromosome is one solution to the given problem (shown in
Fig. 6.10).
• Gene—A gene is one element position in a chromosome.
• Allele—It is a value a gene takes in a particular chromosome.
• Genotype—Genotype is the population in a computation space. In the computation
space, the solutions are represented such that it is manipulated by a computing
system.
• Phenotype—Phenotype is the population in the actual real-world solution space
in which solutions are represented in real-world situations.
• Decoding and Encoding—For simple problems, the phenotype and genotype
spaces are the same. In some cases, the phenotype and genotype spaces are differ-
ent. Decoding is the process of transforming a solution from the genotype to the
phenotype space and encoding is a process of transforming from the phenotype
to genotype space. Decoding is done fast as it is carried out repeatedly in a GA
during the fitness value calculation (shown in Fig. 6.11).
• Fitness Function—A fitness function is one which takes the solution as input
and produces the suitability of the solution as the output. Sometimes, the fitness
function and the objective function may be same.
• Genetic Operators—They alter the genetic composition of the offspring. These
include crossover, mutation, selection, etc.
GA is used to optimize the arc welding parameters like bead width, depth of
penetration, reinforcement height, front height to front width ratio, back height to
back width ratio, etc. The basic flow structure of GA is given in Fig. 6.12.
Encoding
Genotype Space
(Computation Space)
Phenotype Space
(Actual Solution Space)
0 1 0 1 0 1 0 1
Decoding
Fig. 6.11 Encoding and decoding in GA
6.7 Case Study 1: Optimization of Flux Cored Arc Welding

Parameters Using GA
6.7.1 Objective
B. Senthil Kumar and T. Kannan and R. Madesh investigated the effect of FCAW pro-
cess parameters on the quality of the super duplex stainless steel (SDSS) claddings
using Taguchi L9 design of experiments. In this investigation, deposits are made with
30% bead overlap. Establishing the optimum combination of process parameters is
essential to ensure better bead geometry and desired properties. The above objectives
are achieved by identifying the significant input process parameters as input to the
mathematical models like welding voltage (X 1 ), wire feed rate (X 2 ), welding speed
(X 3 ) and nozzle to plate distance (X 4 ). The identified responses governing the bead
geometry are bead width (W ) and height of the reinforcement (H). The mathemati-
cal models are constructed using the data collected from the experiments based on
Taguchi L9 orthogonal array. Then, the responses are optimized using non-traditional
nature-inspired technique like genetic algorithm (GA) [11].
6.7.2 Experimentation
The experimental setup consists of multiprocess welding machine Lincoln Electric

Invertec V350 pro coupled with wire feeder LF-74. The electrode is 1.2 mm-diameter
Metrode™ Supercore™ FC2507 (AWS A5.22E2594T0-4) and base material IS:2602
low carbon structural steel. The shielding gas mixture consists of 80% Argon and
20% CO2 supplied at a rate of 24 l/min (as shown in Figs. 6.13 and 6.14; Table 6.1).
6.7 Case Study 1: Optimization of Flux Cored Arc Welding … 197
Start
Generate Number of
Individuals
Assign Fitness to each

Individuals
Select two individuals

(Parent 1 Parent 2)
Use crossover operation to

produce off-springs NO
Crossover
Assign fitness to off-springs
Finished?
YES
Select one off -spring
Apply Mutation operation

NO
Mutation
NO Assign fitness to off-spring
Finished?
YES
Apply replacement operation to
incorporate new induvidual
Terminate?
YES
Finish
Fig. 6.12 Flow diagram of GA

Fig. 6.13 Experimental setup
Fig. 6.14 Welded specimen
6.7.3 Optimization
The electrode melting and heat input rate regulates the bead dimensions. The deposi-
tion rate has greater impact on the bead dimensions and dilution but does not affect the
welding current. The main objective of weld cladding is to produce the desired bead
dimension with low percentage dilution without affecting the bead integrity. To meet
the above requirements, the objective function is composed of both reinforcement
height and bead width.
6.7 Case Study 1: Optimization of Flux Cored Arc Welding … 199
Table 6.1 Design of experiments

S. no Parameters Symbols Units Factors level
1 2 3
1 Welding X1 V 22 26 30
voltage
2 Wire feed X2 m/min 5.08 6.35 6.62
rate
3 Welding X3 m/min 0.12 0.16 0.2
speed
4 Nozzle to X4 m 0.015 0.019 0.023
plate
distance
Table 6.2 Optimized parameters

S. no Process parameters (un-coded) Responses
Welding Wire feed Weld speed Nozzle to Bead width Reinforcement
voltage (V) rate (m/min) X 3 plate (mm) height (m)
X1 (m/min) X 2 distance
(m) X 4
1 22.18 5.28 0.198 0.022 13.26 5.81
2 23.27 5.25 0.198 0.02 14.84 5.37
3 24.8 5.23 0.198 0.016 16.48 4.67
The table presents the parameter combinations for non-dominated Pareto optimal
points (as shown in Table 6.2). The details of the optimization environment are given
as follows:
GA settings,
Population 60,
Selection function tournament,
Crossover fraction 0.8,
Crossover ratio 1.0,
Distance measure function distance crowding,
Number of variables 4,
Upper limit [30, 300, 200, 23],
Lower limit [22, 200, 120, 15].
6.8 Case Study 2: Optimization and Prediction of Hardness

and Shear Strength Using PSO Based ANN in FSW
of AA6061 Alloys
6.8.1 Objective
In this study, Friction stir welding is applied to Aluminium alloy 6061 sheets of 3 mm
thickness. The influence of the process parameters; spindle speed, plunge speed and
plunge depth on the Shear Strength and Brinell’s Hardness of the joints are inves-
tigated by performing a single lap joint. Finally, a parallel design method based on
Particle Swarm Optimized—Backpropagation Neural Network is developed to pre-
dict the shear strength and hardness of the aluminium alloy joints. The PSO algorithm
improves the accuracy and runtime efficiency of the BP Neural Network and thereby
optimizes the initial weights and thresholds of the same. Based on the experimental
outcomes, validations are carried out. The results reveal good coherence between
experimental data and predicted data. Considering the vitality, this soft computing
algorithm can be incorporated to replicate the experimental results thereby saving
resources, energy consumption, manpower and material wastage during investiga-
tions in manufacturing processes.
Friction stir welding is performed on the aluminium alloy 6061 of 3 mm thickness

as shown in Fig. 6.15.
Welding parameters that control the process are spindle speed, welding speed
and plunge depth. Each factor, or independent variable, is placed for one of five
Fig. 6.15 Experimental

setup
6.8 Case Study 2: Optimization and Prediction of Hardness and … 201
Table 6.3 Parameter levels

Parameters Level −1.682 Level −1 Level 0 Level 1 Level 1.682
Spindle speed 563 700 900 1100 1236
(rpm)
Welding 0.46 0.8 1.3 1.8 2.14
speed (mm/s)
Plunge depth 0.02 0.05 0.1 0.15 0.18
(mm)
Input Layer Output Layer

Spindle Speed
(rpm)
Brinell Hardness
Welding Speed
Hidden Layer
(mm/sec)
with bias
Shear Strength
(Mpa)
Plunge Depth
(mm)
Fig. 6.16 Developed PSO-NN model
equally spaced values, usually coded as −1.682, −1, 0, +1, +1.682. The range of
the input parameters spindle speed, welding speed and plunge depth are selected as
563–1236 rpm, 0.46–2.14 mm/s, 0.02–0.18 mm, respectively as shown in Table 6.3.
The development of BP-NN and its training are performed using neural network
toolbox in MATLAB. The hidden network structure contains five layers with a bias
node. The neural network deployed in this study is a feedforward backpropagation
network with ‘PSO’ as learning algorithm. The input layer has three neurons (Spin-
dle speed, Welding speed and Plunge depth) and the output layer has two neurons
(Sher strength and Brinell Hardness) as shown in Fig. 6.16. Initially, various BP-NN
structures with one to five hidden layers (each with varying neuron numbers) in each
layer are evaluated. Out of all the evaluated NN, a NN with five hidden layers with
five hidden neurons in each layer resulted in superior performance.
Fig. 6.17 Comparison of

actual and predicted Brinell
hardness
Fig. 6.18 Comparison of

actual and predicted shear
strength
The results for the Brinell hardness and Shear strength obtained by developed BP-
NN model is in good agreement with the experimental results as shown in Figs. 6.17
and 6.18.
Three test samples are taken for validating the obtained results. The test results
reveal that the experimental and predicted values are in close coherence with each
other as shown in Table 6.4. The percentage difference of error while measuring
the spread of prediction by the BP-NN model (in predicting the hardness and shear
strength) is estimated. It is observed that the developed BP-NN model predicts param-
eters under study within the acceptable error limits.
6.9 Case Study 3: LM Algorithm-Based ANN Model to Predict Strength … 203
Table 6.4 Testing and validation of friction stir welded parameters

Test trial Input Actual values Predicted values % Error
parame-
ters
Brinell Shear Brinell Shear Brinell Shear
hardness strength hardness strength hardness strength
(MPa) (MPa) (MPa)
1 Spindle 71 100 71.23 99.7 −0.32% +0.3%
speed:
1236 rpm
Welding
speed:
1.8 mm/s
Plunge
depth:
0.05 mm
2 Spindle 72 98 71.87 96.98 +0.18% +0.02%
speed:
563 rpm
Welding
speed:
1.8 mm/s
Plunge
depth:
0.15 mm
3 Spindle 75 85 75.01 85.2 −0.01% −0.23%
speed:
1100 rpm
Welding
speed:
0.8 mm/s
Plunge
Depth:
0.15 mm
6.9 Case Study 3: LM Algorithm-Based ANN Model

to Predict Strength and Joint Resistance of Al-Cu
Alloys Joined by Ultrasonic Welding Process
6.9.1 Objective
This chapter presents the investigations undertaken to analyse the microstructural

behaviour and mechanical properties of ultrasonic welding of Al–Cu dissimilar wires
typically employed for electrical appliances and automobile parts. Levenberg–Mar-
quardt algorithm-based Artificial Neural Network Optimization is implemented to
Fig. 6.19 Ultrasonic welding machine used for joining Al–Cu wires
predict and validate the strength and joint resistance of the welded samples. It is
observed that there is a good agreement between the simulation and experimental
results.
The ultrasonic welding is performed for Al6061 and Cu alloys by varying parameters
based on L9 Taguchi Method. Based on the recent demands in industrial sector,
dissimilar materials aluminium and copper are chosen in this study. The power rating
of the ultrasonic machine is 10 KW and the process model used is UWMM 1000 as
shown in Fig. 6.19. This machine provides the option to vary the process parameters;
pressure, amplitude and weld time.
The parameters that have insignificant influence on the weldment as observed
from trials are maintained constant. For both Al and Cu, the thickness of the weld
samples is 0.5 mm and welding is performed for 30 trails to analyse the weldments
and to frame an optimum parameter window. The parameters levels in this study are
provided in Table 6.5.
The ANN model is developed with four set of input parameters: Pressure (bar), weld
time (sec), amplitude (microns) and holding time (s). The hidden layers are chosen in
the implemented LM algorithm. The output layer consists of two parameters: Strength
(N/m2 ) and Joint resistance () (as shown in Fig. 6.20). Thirty data samples are used
for training the ANN model and three samples are used for prediction and validation.
Based on the experimental data, the ANN is trained using LM algorithm in the
MATLAB tool platform to predict the strength and joint resistance of ultrasonically
6.9 Case Study 3: LM Algorithm-Based ANN Model to Predict Strength … 205
Table 6.5 Parameter levels

S.no Parameters Level
−2 −1 0 1 2
1 Pressure 2.5 3 3.5 4 4.5
(bar)
2 welding 1 1.5 2 2.5 3
time (s)
3 Amplitude 35 40 45 50 55
(µm)
4 Holding 1.5 2 2.5 3 3.5
time (s)
INPUT LAYER OUTPUT LAYER

Pressure
(bar)
Strength
(N/m2)
Time
(sec) Artificial
Neural
Network
Joint Resistance
(Hidden Layer (Ω)
Amplitude
(microns) with bias)
Holding time
(sec)
Fig. 6.20 Schematic of the ANN model
joined Al-Cu alloy wires. Figure 6.21 shows that the strength and joint resistance
predicted by the developed LM based ANN model is in good coherence with the
experimental validations.
Three test samples are taken for validating the obtained results. The test results
reveal that the experimental and predicted values are in close agreement with each
other as shown in Table 6.6.
The percentage difference of error while measuring the spread of prediction by
the ANN model (in predicting the strength and joint resistance) is estimated. It is
observed that the developed LM-ANN predicts within the acceptable error limits.
The error plots are shown in Fig. 6.22.
The optimization methods are appropriate for modelling, control and optimiz-
ing the different welding process. Most works indicate the adaptation of ANNs to
predict responses and optimize the welding process. Nevertheless, there is a lack of
comparative study regarding the performance of the optimization methods based on
Fig. 6.21 Prediction of strength (left) and Joint resistance (right)
Table 6.6 Testing and validation of evaluated parameters

Test no. Actual Predicted Error (%) Actual Joint Predicted Error for
strength strength for strength resistance joint joint
(N/m2 ) (N/m2 ) () resistance resistance
() (%)
1 0.9 0.91 −1.11 2.05 2.083 −1.60
2 1.34 1.33 0.74 2.28 2.281 −0.04
3 1.36 1.36 0 2.62 2.62 0
Fig. 6.22 Error plot in prediction of strength (left) and Joint resistance (right)
its suitability. Future work should focus on the application of these modelling and
optimization techniques to identify the optimal welding combinations for a certain
welding process for which the process is considered safe, environment friendly and
economical.
References 207
References
1. H. Guan, Gradient-type method for unconstrained optimization (2016)

2. K.Y. Benyounis, A.G. Olabi, Optimization of different welding processes using statistical and
numerical approaches—A reference guide. Adv. Eng. Softw. 39(6), 483–496 (2008)
3. A.J. Maren, C.T. Harston, R.M. Pap, Handbook of Neural Computing Applications. Elsevier
(1990)
4. S. Mandal, P.V. Sivaprasad, S. Venugopal, K.P.N. Murthy, Artificial neural network modeling
to evaluate and predict the deformation behavior of stainless steel type AISI 304L during hot
torsion. Appl. Soft Comput. J. 9(1), 237–244 (2009)
5. M. Vasudevan, Soft computing techniques in stainless steel welding. Mater. Manuf. Process.
24(2), 209–218 (2009)
6. Y. Hu, The Parallelism and Application in Data Mining of BP Algorithm, Chongqing University
(2003)
7. J. Cao, H. Cui, H. Shi, L. Jiao, Big Data: A Parallel Particle Swarm Optimization-Back-
Propagation Neural Network Algorithm Based on MapReduce (2016)
8. R.C. Eberhart, Y. Shi, Particle swarm optimization: developments, applications and resources,
in Proceedings of the 2001 Congress on Evolutionary Computation (IEEE Cat. No.01TH8546),
vol. 1, February 2001 (2015) pp. 81–86
9. H. Yu, B. Wilamowski, Levenberg–Marquardt Training (2011) pp. 1–16
10. D.S. Nagesh, G.L. Datta, Genetic algorithm for optimization of welding variables for height to
width ratio and application of ANN for prediction of bead geometry for TIG welding process.
Appl. Soft Comput. J. 10(3), 897–907 (2010)
11. B. Senthilkumar, T. Kannan, R. Madesh, Optimization of flux-cored arc welding process param-
eters by using genetic algorithm. Int. J. Adv. Manuf. Technol. 35–41 (2015)
Chapter 7
Codes and Safety Standards During
Welding
Arc welding becomes a safe procedure when adequate safety precautions are con-
sidered and followed to protect the workers from harmful hazards. When such safety
measures are neglected, the workers come across potential hazards such as exposure
to gases and fumes, arc radiation, electric shock and fire accidents, which could be
serious and fatal.
A person in charge of the work (welding) has the main responsibility to ensure
that welders are not exposed to safety and health hazards. A person directing a
business on welding activities must eradicate risks due to welding or minimize the
risks that are reasonably practicable [1].
The WHS (Workplace Health and Safety) Regulations possess specific neces-
sities for managing the hazards due to airborne contaminants, chemicals and other
associated risks. This necessitates the analysis about the substance and providing
adequate information.
Officers, such as directors of an industry, have the responsibility to ensure that
the undertaking business fulfils the regulations provided by WHS.
Workers hold the responsibility of taking care of their own safety and must
not affect the safety of other workers. Workers must cooperate with procedures and
policies relating to safety and health aspects. Workers must use the personal protective
equipment (PPE) as per the training instructions given to them.
7.1 Risk Management Process
7.1.1 Identifying the Potential Hazards
The preliminary step to manage the risks associated with arc welding is to find all
potential risks and hazards. Some welding results in the production of infrared (IR)
and ultraviolet (UV) radiation which cause blindness, burns and cancer.

https://doi.org/10.1007/978-981-13-0806-2_7
210 7 Codes and Safety Standards During Welding
Hazards can be identified by the following:
• Conducting a walk-through assessment in the work environment by observation

and interaction with workers.
• Inspection of materials and equipment used in welding.
• Reading the product labels, standards and manuals of the equipment.
• Discussing with manufacturers, associated suppliers and safety specialists.
• Review of incident reports.
7.1.2 Assessment of Risk
The risk assessment process involves considering the consequences when a worker
is exposed to a hazard. As per the regulations of WHS, a risk assessment is not
compulsory for welding, but it is essential while working in a confined space.
In certain circumstances, a risk assessment will assist to:
• Determination of sources and processes causing risk.

• Identification of appropriate control measures.
• Ensuring the efficacy of existing control measures.
• Identification of workers prone to exposure.
Risks depends on:
• Properties of the workpiece materials being joined.

• Conditions of the welding equipment.
• Surface coating of the workpiece materials being welded.
• Skills and experience of the welders.
Different welding processes influence risk in a different manner. For example, the risk
of electric shock is less in Gas Metal-Arc Welding (GMAW) process than Manual
Metal-Arc (MMA) welding process since the open-circuit voltage is lower and only
direct current is used.
7.1.3 Risk Control
The aim of eliminating a potential hazard and its associated risk is significant. If it
is not possible, the risks can be minimized by the following methods:
Substitution—Replacement of hazardous material with a less hazardous material,
for example, preferring Submerged Arc Welding (SAW) rather than Flux-Cored Wire
Welding (FCAW) reduces the risk of exposure to fumes and radiations.
Isolation—Removal of the workers from hazardous location or isolating the haz-
ards from the workers. For example, in the processes such as plasma cutting, grinding
7.1 Risk Management Process 211
can be performed in away from the usual manufacturing location to reduce the expo-
sure to noise.
Use of engineering controls—Risks can be eliminated using engineering con-
trols, for example, ventilation to remove welding fumes.
If risk then persists, it must be minimized by implementing administrative con-
trols. For example, if a welding is performed in a hot environment, allowing the welder
to weld for a limited time by taking break causes reduction in the heat exhaustion.
Any remaining risks can have minimized using personal protective equipment
(PPE). For example, if the worker stands on metal surfaces which belongs to a part
of the electric circuit, rubber-soled or leather-based boots can be used to prevent the
risk of electric shock.
7.2 Specific Hazards and Control Measures
A person performing the welding business must accomplish enough control measures
to manage the associated risks in a workplace [1]. This comprises ensuring of correct
labels for the hazardous chemicals that the workers access.
7.2.1 Airborne Contaminants
Welding produces fumes, dust, smoke, gases and vapours (as shown in Fig. 7.2). The
type and quantity emitted varies depending on the material and the joining process.
The health effects of exposure to these fumes can include the irritation in the nasal
cavity, respiratory tract, throat, asthma and tightness in chest area (Fig. 7.1).
A few basic terms to be known before approaching the control measures are given
as folows:
Natural Ventilation: It is the air movement through the workplace, which is caused
due to natural forces. Outside the workplace, it is caused by the wind. Inside, it occurs
through open doors and windows.
Fig. 7.1 Symbolic notation:

fumes and gas emissions
Fig. 7.2 Fumes during

welding [2]
Mechanical Ventilation: It is the air movement in the workplace caused using an

electrical device such as fan or exhaust fans mounted in the ceilings or walls.
Source Extraction: It is an electromechanical device used to capture the welding
smoke and fumes near the arc and work area.
Some control measures are given as follows:
• Elimination of exposure to the airborne contaminants.

• Sometimes, it is not practicable to eradicate risks, so measures to minimize the
risk can be chosen, for example, substituting a hazardous chemical with a less
hazardous one.
• Reduction in the quantity of used hazardous components.
• Isolation of the source of exposure.
• Utilizing engineering control measures like ventilators.
• Implementation of administrative control measures.
• Provision of respiratory protection.
7.2.2 Radiation
Radiation is the energy that travels in the form of waves such as electromagnetic
radiation or subatomic particles (as shown in Fig. 7.3). Arc welding emits ultraviolet
radiation, infrared radiations and visible radiation. The effect of radiation on the
body depends on type, intensity and the distance of the person being exposed. It may
7.2 Specific Hazards and Control Measures 213
Fig. 7.3 Radiation during

arc welding [3]
result in eye disorders and skin burns when exposed to intense UV or IR radiation
during welding. Exposure of radiation to eyes results in inflammation and pain in the
cornea called as ‘welders flash’ or ‘arc eye’. The cornea can repair itself in one or
two days, and sometimes may lead to loss of vision if the exposure is severe. Workers
involved in the welding environment are at greatest risk; however, other workers in
the workplace could be exposed to radiation.
Some of the control measures are given as follows:
• Installation of non-flammable partitions and screens.

• Usage of symbolic indications to warn that welding is performed.
• If welding is performed, entry can be restricted to the work area unless safety
equipment is used.
• Provide PPE equipment.
7.2.3 Electrical Risks
Using welding equipment may result in the risk of electric shock. Exposure to elec-
tromagnetic fields is a potential hazard for workers with certain medical conditions
[1] (Fig. 7.4).
Electric shock results in burns or sometimes death by electrocution. Electrocution
occurs by direct contact with the electrode, live part, workpiece or contact with any
unearthed cable or equipment. Electric shock risk worsens with moisture and high
humidity.
Figure 7.5 illustrates melting of wires due to short circuit and excessive current.
Fig. 7.4 Symbolic notation: electric shock
Fig. 7.5 Melting of cables due to short circuit errors [4]
Some of the control measures are given as follows:
• Usage of electrode holders with insulations. The holder should not be wet for
cooling.
• Prevention of contacting welding wire or and electrodes with bare hands.
• Ensure that holders are not held under the armpits.
• Prevention of holders or electrodes to meet any other person.
• Checking the working area for the presence of any live components or wetness.
• Inspection of equipment like power switches, connections, cables and insulation
whether they work in good condition.
7.2.4 Risks Due to Electromagnetic Fields
Arc welding results in the production of strong electromagnetic fields close to the
current-carrying cables and power source. They disturb the functioning of permanent
defibrillators, pacemakers or other medical devices which can slow down the heart
or even stop it. A worker will not be aware of a magnetic field hazard unless a heart
pacemaker or other device behaves irregularly. Some of the control measures are
given as follows:
• Usage of barriers to isolate people not involved in the welding.

• Usage of signs to indicate the presence of strong electromagnetic fields to alert
people.
• Avid the presence of workers close to the power source.
• Workers with pacemakers or other medical devices should substitute work that
involves exposure to electromagnetic fields.
• Seeking medical advice before exposing a worker to welding related to electro-
magnetic fields.
7.2.5 Exposure to Heat and Burns
Burns are most common injuries during welding. The IR and UV radiation result in
harmful effects to the workers. Burns occur on hands and skin surface eyes due to
sparks and weld fragments (Fig. 7.6).
Some of the control measures are given below:
• Isolation of workers from contacting the workpieces. For example, carrying post-
weld heat treatment where workpieces cannot be touched.
• Mark or label as ‘hot’ equipment.
• Contact with heated surfaces can be evaded with thermal insulators.
• Ventilation of work areas.
Fig. 7.6 Symbolic notation:

burns and explosions
7.2.6 Compressed and Liquefied Gases
These gases are used as shielding gases in arc welding, which are stored in cylinders
as large volumes under high pressure. Precautions are to be taken when storing and
handling the cylinders.
The potential hazards associated with these gases are fire, toxicity, explosion
asphyxiation, oxidization and uncontrolled pressure release. Gas leakage is also
dangerous which can be recognized by its odour, however, leakage of oxygen is
harder to sense and therefore it is more hazardous. Some of the control measures are
given below:
• Store the cylinders appropriately and handle it safe.

• Keep the cylinders free from leaks.
• The cylinders should be kept in an upright position.
• Secure the cylinders to prevent dislodgement.
• Flashback arrestors can be fit at the blowpipe and regulator end of both the oxygen
and fuel gas lines.
7.2.7 Personal Protective Equipment (PPE)
In most cases, PPE (as shown in Fig. 7.7) must be worn by workers when welding
to supplement higher levels of controls such as ventilation systems or administrative
controls.
Some of the critical Personal Protective Equipment (PPE) are presented in
Table 7.1.
7.2.8 Health Monitoring
Health monitoring means monitoring a person to identify changes in their respec-

tive health status due to exposure to certain substances during welding. It involves
collecting their health data to evaluate the effects of exposure and then to determine
whether the exposure dose is within safe limit. This enables the conclusions about
for eliminating the risk of worker’s exposure. For example, reassignment of a worker
with other duties involves less exposure.
Tables 7.2 and 7.3 present some of the fumes and gases, their sources and their
adverse health effects.
Table 7.1 Common PPE’s

Potential hazards Type of PPE to be used Suggestions
Heat, ultraviolet light and Gloves/gauntlets Gloves should be fire resistant
burns from hot debris and and protect exposed skin on
sparks the hands and wrists
Hearing loss Hearing protection (e.g. earEar plugs or ear muffs may be
muffs and ear plugs) required to minimize the risks
of noise
Heat, ultraviolet light and Clothing (e.g. flame resistant Avoid clothing that has the
burns from hot debris and long-sleeved shirts, long potential to capture hot sparks
sparks trousers, aprons and leather and metals, for example, in
spats) pockets or other folds.
Clothing should be made of
natural fibres
Dusts, hazardous fumes, gases Respiratory protective devices Respirators should be fitted for
and chemicals and (face respirators and air each person individually and if
oxygen-depleted atmospheres supplied respirators) one is to be used by another
operator, it must be disinfected
and refitted before use. The
tightness of all connections
and the condition of the face
piece, headbands and valves
should be checked before each
use. Air supplied respirators
may be required in some
situations, e.g. confined spaces
Hot metal debris, other metal Foot protection (e.g. boots and Foot protection should be
debris and electric shock shoes) non-slip and be heat and fire
resistant. Avoid using foot
protection that has the
potential to capture hot sparks
and metal debris, for example
in laces or in open style shoes
Light, radiation, burns from Eyes, face and head protection Workers should always have
hot debris and sparks (e.g. goggles, helmets, hand their eyes, face and/or head
shields and protective filters) protected whenever they are
welding
Exposure to the rays of an arc Screens Opaque or appropriate
during electric welding translucent screens can be
operations used to protect the health and
safety of people within the
vicinity of welding
Table 7.2 Source of fumes and health effects

Type of fumes Source of fume Health effect
Aluminium Aluminium component of some Respiratory irritant
alloys, e.g. nickel-chromium, copper,
zinc, steel, magnesium, brass and
filler materials
Vanadium Some steel alloys, iron, stainless Acute effect is irritation of the eyes,
steel, nickel alloys skin and respiratory tract. Chronic
effects include bronchitis, retinitis,
fluid in the lungs and pneumonia
Cadmium Stainless steel containing cadmium or Irritation of respiratory system, sore
oxides plated materials, zinc alloy and dry throat, chest pain and
breathing difficulty. Chronic effects
include kidney damage and
emphysema. Suspected carcinogen
Chromium Most stainless steel and high-alloy Increased risk of lung cancer. Some
materials, and welding rods, and also individuals may develop skin
used as plating material irritation. Some forms are
carcinogens (hexavalent chromium)
Beryllium Hardening agent found in copper, ‘Metal fume fever’. A carcinogen.
magnesium, aluminium alloys and Other chronic effects include damage
electrical contacts to the respiratory tract
Copper Alloys such as nickel-copper, brass, Acute effects include irritation of the
bronze, and also some welding rods eyes, nose and throat, nausea and
‘metal fume fever’
Iron oxides The major contaminant in all iron or Siderosis—a benign form of lung
steel welding processes disease caused by particles deposited
in the lungs. Acute symptoms include
irritation of the nose and lungs. Tends
to clear up when exposure stops
Molybdenum Steel alloys, iron, stainless steel, Acute effects are eye, nose and throat
nickel alloys irritation and shortness of breath
Nickel Stainless steel, nickel-chromium, Acute effect is irritation of the eyes,
nickel-copper and other high-alloy nose and throat. Increased cancer risk
materials, welding rods and plated has been noted in occupations other
steel than welding, and also associated
with dermatitis and lung problems
Manganese Most welding processes, especially ‘Metal fume fever’. Chronic effects
high-tensile steels may include central nervous system
problems. Ototoxic and therefore risk
of hearing loss
Zinc oxides Galvanized and painted metal Metal fume fever
Fluorides Common electrode coating and flux Acute effect is irritation of the eyes,
material for both low- and high-alloy nose and throat. Long-term exposures
steels may result in bone and joint
problems. Chronic effects also
include excess fluid in the lungs
Table 7.3 Source of gases and health effects

Type of gas Source of gas Health effect
Carbon monoxide Formed in the arc Absorbed readily into the
bloodstream, causing
headaches, dizziness or
muscular weakness. High
concentrations may result in
unconsciousness and death.
Ototoxic and therefore risk of
hearing loss
Phosphine Metal coated with rust Irritant to eyes and respiratory
inhibitors. Phosphine is system
formed by reaction of the rust
inhibitor with welding
radiation
Nitrogen oxides Formed in the arc Eye, nose and throat irritation
in low concentrations.
Abnormal fluid in the lung and
other serious effects at higher
concentrations. Chronic
effects include lung problems
such as emphysema
Ozone Formed in the welding arc Acute effects include fluid in
during open arc welding the lungs. Very low
processes including Manual concentrations (e.g. one part
Metal-Arc Welding (MMAW), per million) cause headaches
Flux-Cored Arc Welding and dryness of the eyes.
(FCAW), especially during Chronic effects include
plasma arc, Metal Inert Gas significant changes in lung
(MIG) and Tungsten Inert Gas function
(TIG) processes
Oxygen deficiency Welding in confined spaces, Dizziness, mental confusion,
and air displacement by asphyxiation and death
shielding gas
Hydrogen fluoride Decomposition of rod coatings Irritating to the eyes and
respiratory tract.
Overexposure can cause lung,
kidney, bone and liver
damage. Chronic exposure can
result in chronic irritation of
the nose, throat and bronchi
Fig. 7.7 Personal protective equipment
7.3 Standard Operating Procedures During Arc Welding
7.3.1 Engine Power Equipment
• Turn off the engines while troubleshooting and maintenance work are carried out
and except when the maintenance work requires it to be running.
• Operating engines in open and in well-ventilated areas helps the engine to exhaust
fumes outdoors.
• Avoid adding the fuel when the engine is running or near an open flame welding
arc.
• Before refuelling, stop the engine and allow it to cool in order to prevent spilled
fuel from vaporizing on contact with hot engine parts and igniting.
• Do not spill fuel while filling the tank. If fuel is spilled, wipe it up and do not start
the engine until fumes have been eliminated.
• Keep all equipment covers, safety guards and devices in position and in good
repair. Keep hands, clothing, hair and tools away from gears, fans, V-belts and all
other moving parts when starting, operating or repairing equipment.
7.3 Standard Operating Procedures During Arc Welding 221
• In few cases, it can be necessary to remove safety guards to do the required main-
tenance. Remove guards only when necessary and replace them when the main-
tenance requiring their removal is complete. Always use the greatest care when
working near moving parts.
• Do not put hands near the engine fan.
• Do not attempt to override the governor or idler by pushing on the throttle control
rods while the engine is running.
• To prevent accidentally starting gasoline engines while turning the engine or weld-
ing generator during maintenance work, disconnect the spark plug wires, distrib-
utor cap or magneto wire as appropriate.
• To avoid scalding, do not remove the radiator pressure cap when the engine is hot.
7.3.2 In Presence of Electric and Magnetic Fields
• Route the electrode and work cables together—secure them with tape when pos-
sible.
• Never coil the electrode lead around the body.
• Do not place the body between the electrode and work cables. If the electrode
cable is on our right side, the work cable should also be on our right side.
• Connect the work cable to the workpiece as close as possible to the area being
welded.
• Do not work next to welding power source.
• The electrode and work (or ground) circuits are electrically ‘hot’ when the welder
is on. Do not touch these ‘hot’ parts with our bare skin or wet clothing. Wear dry,
hole-free gloves to insulate hands.
• Insulate our self from work and ground using dry insulation. Make certain the
insulation is large enough to cover our full area of physical contact with work and
ground.
• In semiautomatic or automatic wire welding, the electrode, electrode reel, welding
head, nozzle or semiautomatic welding gun are also electrically ‘hot’.
• Always be sure the work cable makes a good electrical connection with the metal
being welded. The connection should be as close as possible to the area being
welded.
• Ground the work or metal to be welded to a good electrical (earth) ground.
• Maintain the electrode holder, work clamp, welding cable and welding machine
in good, safe operating condition. Replace damaged insulation.
• Never dip the electrode in water for cooling.
• Never simultaneously touch electrically ‘hot’ parts of electrode holders connected
to two welders because voltage between the two can be the total of the open-circuit
voltage of both welders.
• When working above floor level, use a safety belt to protect our self from a fall
should we get a shock.
7.3.3 During Handling Cylinders
• Use only compressed gas cylinders containing the correct shielding gas for the
process used and properly operating regulators designed for the gas and pressure
used. All hoses, fittings, etc. should be suitable for the application and maintained
in good condition.
• Always keep cylinders in an upright position securely chained to an undercarriage
or fixed support.
• Cylinders should be located; away from areas where they may be struck or sub-
jected to physical damage; a safe distance from arc welding or cutting operations
and any other source of heat, sparks or flame.
• Never allow the electrode, electrode holder or any other electrically ‘hot’ parts to
touch a cylinder.
• Keep our head and face away from the cylinder valve outlet when opening the
cylinder valve.
• Valve protection caps should always be in place and hand-tight except when the
cylinder is in use or connected for use.
7.3.4 While Handling Shielding Gases
• The operation of welding fume control equipment is affected by numerous fac-

tors including proper use and positioning of the equipment, maintenance of the
equipment and the specific welding procedure and application involved.
• Worker exposure level should be checked upon installation and periodically there-
after to be certain it is within applicable limits.
• Do not weld in locations near chlorinated hydrocarbon vapours coming from
degreasing, cleaning or spraying operations. The heat and rays of the arc can react
with solvent vapours to form phosgene, a highly toxic gas and other irritating
products.
• Shielding gases used for arc welding can displace air and cause injury or death.
Always use enough ventilation, especially in confined areas, to ensure breathing
air is safe.
7.4 Welding Codes: American Welding Society (AWS)
Welding, fabrication, testing, repair and inspection requirements are covered under
three main governing organizations: American Welding Society (AWS), the Amer-
ican Petroleum Institute (API) and the American Society of Mechanical Engineers
(ASME). All these organizations have specific codes for several types of processes,
construction and/or materials [5].
7.4 Welding Codes: American Welding Society (AWS) 223
The American Welding Society was founded in with a global mission to advance
the science and technology in the application of welding, cutting, allied joining. It
is headquartered in Florida. The organization is popular for its code and standard
procedures, for the welding and joining of materials. Some of the critical codes
pertaining to arc welding are presented as follows.
AWS D1.1: It contains the requirements for fabrication and erection of welded steel
structures. It applies to steel materials with a thickness of 3.2 mm or more. Most of
the provisions are mandatory.
AWS D1.2: It consists of the Structural Welding Code for Aluminium. The require-
ments are applicable to all types of welded aluminium alloy. This code is applied for
fabrication of supporting structures. It is not intended to the codes developed for use
in applications specifically to ASME Boiler, aerospace, pressure vessels and military
codes.
AWS D1.3: It consists of the Structural Welding Code for Sheet-based Steel. It covers
the arc welding of steel sheet including the cold-formed members which are less than
or equal to 4.8 mm thickness.
AWS D1.4: It consists of the Structural Welding Code for Reinforcing Steel. It applies
to the welding of reinforcing, carbon and low-carbon structural steel.
AWS D1.5: It consists of the Bridge Welding Code. It covers welding fabrication
requirements applicable to welded highway bridges. This code is not intended to
be used for steels with a minimum specified yield strength greater than 690 MPa,
pressure vessels, base metals other than carbon or low-alloy steels.
AWS D1.6: It consists of the Structural Welding Code for Stainless Steel. It covers
welding requirements applicable to stainless steel weldments. It shall be used in
conjunction with any complementary code for the design of stainless steel weldments.
AWS D3.5-93R: It consists of the guide for Steel Hull Welding. It is referenced in
many contract specifications for building vessels from barges to tugboats.
AWS D3.6M: It consists of the specification for Underwater Welding.
AWS D3.7: It consists of the guide for Aluminium Hull Welding. Like the Steel Hull
Welding Guide, but with a special emphasis on the unique properties of aluminium.
AWS D8.8-97: It consists of the specification for Automotive and Light Truck Weld
Quality: Arc Welding.
AWS D14.1: It consists of specification for Welding Earth Moving and Construction
Equipment.
AWS D14.5: It consists of specification for Welding Presses and Press Components.
It establishes minimum acceptable requirements for weld joint design and the fabri-
cation by welding of presses and press components.
7.5 Quality Assurance and Quality Management
Quality requirements with respect to welded structures are essential in regulations,

standards or customer specifications. The manufacturer performing welding needs
to analyse these requirements at the beginning stage, to decide whether they can be
fulfilled which is assisted by a systematic working method [6]. Companies with ISO
9001 certification have documented procedures for the same. ISO 9001 is a system
standard concerned with the quality of systems. They define welding as a special
process that must be properly controlled to ensure that the quality requirements are
satisfied. Under ISO 9001, some critical arc welding procedure specifications are
presented below.
7.5.1 En ISO 15609
It specifies the technical contents of the Welding Procedure Specification (WPS)

for arc welding, gas welding and resistance welding methods. WPS must specify in
detail how the welding is to be performed. It must contain all valuable information
regarding the welding work. It must indicate the factors that can affect the metallurgy,
mechanical properties or geometry of the weldments. The nomenclature of welding
and allied processes is specified and numbered in ISO 4063.
7.5.2 En ISO 15614-1
It sets out the conditions for welding procedure tests on standardized test pieces for
the arc welding methods for welding steel-based alloys as shown in Table 7.4.
7.5.2.1 Test Pieces
The standard specifies the shape and minimum dimensions of standardized test pieces
to be used in connection with the welding procedure. The test pieces must be suf-
Table 7.4 Welding method Welding method ISO 4063 designation

and corresponding ISO
designation for welding Flux-cored metal-arc welding 114
steel-based alloys Submerged arc welding 12
Metal inert gas welding 131
Metal active gas welding 135
Tungsten inert gas welding 141
7.5 Quality Assurance and Quality Management 225
ficiently large to ensure that there is sufficient material to conduct away the heat.
When impact testing of the heat-affected zone is required, the test pieces must be
marked with the rolling direction. All welding of test pieces must be carried out
in accordance with the preliminary WPS, and under the same conditions as can be
expected in production.
Welding positions, angles of slope and rotation must be as specified in EN ISO
6947. Tack welding must be included in the test welds if it is to be used in production.
Welding and testing must be supervised by an examiner or examining body.
7.5.2.2 Examination and Testing
Testing consists of both non-destructive and destructive testing:
• Visual inspection,
• Radiographic or ultrasonic testing,
• Crack detection,
• Transverse tensile test,
• Transverse bend test,
• Impact testing,
• Hardness test and
• Macro- and micro-examination.
The standards specify how the test pieces shall be positioned.
7.5.2.3 Retesting
If the welding procedure test pieces do not meet all the test requirements, the results
cannot be qualified. It is permissible to perform a further procedure test. If any single
test piece fails to meet the requirements due to geometrical defects, two new test
pieces may be selected for retesting. If either of them fails, then the entire WPS also
fails.
7.5.2.4 Range of Qualification
A WPS that has been qualified by a manufacturer is valid for welding in workshops
and at sites under the same technical management. Welding procedure tests form
the basis for qualification of a WPS, of which the important variables lie within the
qualification range of the procedure test. Essential variables are as follows:
• Parent material,
• Material thicknesses,
• Welding method,
• Welding position,
• Type of joint,
• Consumables,
• Type of welding current,
• Heat input,
• Preheat temperature,
• Intermediate pass temperature and
• Post-heat treatment.
7.5.2.5 Documents of Qualified Welding Procedure Tests (WPQR)
Records from welding and testing shall include all the information needed for qual-
ification. Welding Procedure Qualification Records (WPQR) must be signed by the
examiner. Model forms of WPQR are included in the standard.
7.5.3 EN ISO 15614-2
EN ISO 15614-2 describes the conditions applicable to qualification of welding

procedures to be used for arc welding of aluminium and its weldable alloys in accor-
dance with ISO 2092 and 2107 [6]. These welding methods are MIG welding and
TIG welding. The standard, which follows the same principles as in EN ISO 15614-
1, specifies how welding is to be performed and what tests that are to be carried out.
Important variables for the procedure test are the same as for steel, but with lesser
differences in the validity area.
7.5.4 EN ISO 15610
It allows welding procedures to be qualified based on the use of tested consum-

ables, which applies for repetitive welding operations and for workpiece materials of
which the structures and properties in the heat-affected zone do not degrade during
operation. For steel, applicable welding methods are metal-arc welding, MIG/MAG
welding and TIG welding, while MIG welding and TIG welding are applicable for
aluminium [6]. The standard applies to carbon manganese steels and chrome-nickel
steels, as well as for pure aluminium and non-heat-treatable aluminium alloys. Parent
material thicknesses are 3–40 mm.
7.5 Quality Assurance and Quality Management 227
7.5.5 EN ISO 5817 and ISO 10042
Quality levels must be clearly specified if the quality of welds is to be determined by

non-destructive testing. This is done in accordance with ISO 5817 and ISO 10042.
These two standards have three different quality levels as shown in Table 7.5 [6].
The quality level is determined by visual testing and non-destructive testing as
shown in Table 7.6. The most widely used methods of non-destructive testing are:
Radiography (shown in Fig. 7.8) is used to reveal primarily internal volumetric
discontinuities. The films are evaluated against the required quality level. Use of this
method is limited by the thickness of the weld, which should not exceed 50 mm.
Ultrasonic testing (as shown in Fig. 7.9) is most suitable for detecting internal plane
discontinuities. It requires a good test surface. The material should not be less than
8 mm thick for reliable evaluation.
Magnetic particle testing (as shown in Fig. 7.10) is a surface testing method and is
used to reveal discontinuities in or immediately below the surface of ferromagnetic
materials.
Table 7.5 Quality levels Quality level Symbol

Moderate D
Average C
High B
Table 7.6 Non-destructive Testing method Symbol

testing
Radiography RT
Ultrasonic testing UT
Magnetic particle testing MT
Liquid penetrant testing PT
Eddy current testing ET
Fig. 7.8 Radiographic test on weldment

Fig. 7.9 Ultrasonic testing on weldments
Fig. 7.10 Magnetic particle testing on weldments
Liquid penetrant testing (as shown in Fig. 7.11) is also a surface testing method
and is used to reveal discontinuities in the surface of non-porous materials.
Eddy current testing, or inductive testing (as shown in Fig. 7.12) is used to reveal
discontinuities on or immediately below the surface of electrically conducting mate-
rials. A common feature of all these test methods is that the test personnel must hold
certificates in accordance with EN ISO 9712. Testing must also be carried out in
accordance with qualified test procedures.
References 229
Fig. 7.11 Liquid penetrant testing
Fig. 7.12 Eddy current testing on weldments
References
1. UK Spa Association, Code of Practice, April. (2016)

2. Welding fumes increase risk of hearing loss. [Online]. Available: https://www.mysafetysign.co
m/blog/welding-fumes-hearing-loss/
3. The Kingdom of random, Gasget Hacks, (2013). [Online]. Available: https://www.mysafetysig
n.com/blog/welding-fumes-hearing-loss/
4. Health & Safety International. [Online]. Available: https://www.hsimagazine.com/article/weld
ing-precautions-349%5D
5. A. W. Society, P. Vessel, Welding Codes and How They ’ re Used
6. Klas Weman, Welding Processes Handbook, 2nd ed. Woodhead Publishing

S Arungalai Vendan Liang Gao Akhil Garg P Kavitha G Dhivy 1

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

S Arungalai Vendan Liang Gao Akhil Garg P Kavitha G Dhivy 1

Uploaded by

Copyright:

Available Formats

S.

Akhil Garg P. Kavitha G. Dhivyasri

Liang Gao G. Dhivyasri

ISBN 978-981-13-0805-5 ISBN 978-981-13-0806-2 (eBook)

© Springer Nature Singapore Pte Ltd. 2019

Printed on acid-free paper

2.3 Impact of Power Source Characteristics on Weldments . . . . . . . . 23

3.3 Introduction to Fundamental Controllers . . . . . . . . . . . . . . . . . . . 78

5.3.2 Analog-to-Digital Converter (ADC) . . . . . . . . . . . . . . . . 148

6.3 ANN-Based Optimization Techniques to Arc Welding

7.2.5 Exposure to Heat and Burns . . . . . . . . . . . . . . . . . . . . . 215

Resistance High Beam Solid State

Fig. 1.1 Types of welding processes

1.1 Basics of Arc Welding

Finite Element Analysis (FEA)

Fig. 1.2 Intra- and inter-dependent electrometallurgy of arc welding process

1.1.1 Electric Charge

Electric charge is the physical property of matter that causes it to experience

1.1.2 Electric Current

A material is said to be in an electrically uncharged state when the number of protons

1.1.4 Electrical Resistance

1.1.5 Heat Input

Study of welding fundamentals involves heat-flow concepts. The heat energy is

1.2 Equivalent Electrical Analogous Representation of Arc

Fig. 1.3 Equivalent electrical circuit of arc welding process

Table 1.1 Equivalent electrical analogous notations

1.3 Arc Welding as a Confluence of Several Disciplines

1.4 Applications of Arc Welding

Fig. 1.4 Power source and interdependent terminologies

Bicycle GTAW/TIG, MMA Cycle joints

Metal Pipes MMA, TIG, MIG, MIAB Pipe joining

Application industry Type of arc welding Components (name with images)

Ship Underwater welding, FCAW,

Application industry Type of arc welding Components (name with images)

general applications. It is also being used in shipbuilding, automobile, electrical,

1. C. TLE, Arc Welding fundamentals (1999). Available: http://www.lincolnelectric.com/en-us/su

2.1 Critical Arc Power Source Terminologies for Welding

2.1.1 Arc Plasma

© Springer Nature Singapore Pte Ltd. 2019 15

Fig. 2.1 Arc plasma

2.1.2 Arc Efficiency

2.1.3 Arc Stability

2.1.4 Arc Blow

2.1.5 Pinch Effect

2.1.6 Arc Shielding

Fig. 2.2 Pinch effect during

2.2 Influence of Power Source Parameters on Weldment

2.2.1 Open-Circuit Voltage (OCV)

Base Metal Base Metal

Fig. 2.3 Arc column

Fig. 2.4 Plot for arc voltage and arc length

2.2.2 Arc Voltage

(i) Cathode drop,

Fig. 2.5 Depth of

2.2.3 Welding Current

2.2.4 Electrode Polarity

2.2.5 Power Factor

2.2.6 Duty Cycle and Current Rating

2.2.7 Class of Insulation

The insulation is classified as A, E, B, F and G in the increasing order of maximum

2.3 Impact of Power Source Characteristics on Weldments

2.3.1 Static Characteristics

2.3.1.1 Constant Current Characteristics (CC)

2.3.1.2 Constant Voltage Characteristics (CV)