Development and Evaluation of Setup Strategies in Printed Circuit Board Assembly (eBook)

Dr. Ihsan Onur Yilmaz war wissenschaftlicher Mitarbeiter im Fachgebiet Produktionsmanagement an der Technischen Universität Berlin.

Foreword 6
Acknowledgement 7
Table of Contents 8
List of Figures 11
List of Tables 15
List of Abbreviations 17
1. Introduction 21
2. Technological Background 27
2.1 Printed Circuit Boards 27
2.2 Components 28
2.3 SMT Assembly Process 30
3. SMT Placement Machines 35
3.1 Structure of SMT Placement Machines 36
3.2 Pick-and-Place Machines 40
3.3 Collect-and-Place Machines 42
3.4 Chip Shooter 44
3.5 Other Forms of SMT Machines 45
3.6 Selection of the Appropriate SMT Machine 48
4. Production Planning in Electronics Assembly 51
4.1 Overview of Planning Problems 51
4.2 Hierarchical Decomposition of Planning Problems 53
4.3 Setup Strategies 59
5. Development of Group Setup Strategies 95
5.1 Hierarchical Clustering Based on Conventional Approaches 96
5.2 Hierarchical Clustering Based on Inclusion Trees 106
5.3 Improving Group Setup Solutions 112
5.4 Machine-Specific Operations Scheduling 113
6. Numerical Investigation 143
6.1 Experimental Design 143
6.2 Comparison of Group Setup Approaches 149
6.3 Comparison of the Proposed Group Setup Approach Against Other Approaches 158
6.4 Numerical Tests on Industrial PCBs 167
6.5 Concluding Remarks 178
7. Conclusions 179
References 183

7. Conclusions (S. 159-160)

The rapidly developing market for electronic products requires the use of flexible and highly automated assembly systems. Placement machines constitute the heaviest investment of an electronics assembly system which usually defines its throughput. Several analyses demonstrate that the real productive capacity of a placement machine remains much below the theoretical capacity given by the machine vendors. Hence, development and implementation of advanced planning and control systems are essential for exploiting the potentials of highly automated assembly lines. In order to conceive the problems arising in the electronics assembly, one should understand the SMT technology and its future trends. The categorization of placement machines given in the technological background section assists the reader to apprehend how machine-specific optimization problems depend on the kinematics of the placement machine. Thus, an appropriate solution approach has to be tailored for each placement machine individually.

The socalled X-Y gantry systems appear to be promising in responding to the requirements of the future electronics assembly because they provide a high degree of flexibility and modularity. Depending on the requirements of the production environment, these machines can be integrated into an assembly line in different modular configurations. An overview of the planning problems is presented using a novel hierarchical decomposition based on the characteristics of the production environment. The previous studies which are discussed in the literature review have focused mainly on the setup strategies without tackling the intertwined problems of other hierarchy levels. On the contrary, the novel group setup strategy developed in this study integrates the machine-specific optimization problems into the solution approach and covers the new aspects of modern placement machines, e.g. utilizes the advantages of an offline setup. The term group setup is considered as a synonym for creation of setup families in order to reduce the changeover effort. However, several analyses presented in this study have shown that the placement time of an individual PCB increases if different board types are assembled using a joint group setup.

This is due to the new group feeder assignment which cannot be optimized for each single PCB as in the case of a unique setup strategy. This study focuses on reducing the global makespan by observing both the savings in setup time and the increase of individual placement times by integrating detailed machine-specific algorithms into the solution methodology. Two different approaches based on hierarchical agglomerative clustering schemes are presented. The hierarchical clustering methodology enables observing different criteria at each decision stage for merging groups of PCBs. Machine-specific algorithms are used to evaluate the quality of the new solution at each agglomerative clustering step in terms of makespan improvement and magazine capacity constraints. Using a hierarchical scheme also reduces the number of times machine-specific algorithms are called, and thus the computational burden.

The experimental results reveal that fast solutions can be generated in short computational times. In the first group setup approach presented in this study, grouping is performed by use of well-known similarity measures (Jaccard’s and simple matching) and linkage methods (single, complete, and average). The second group setup approach employs inclusion measure as a similarity coefficient and generates setup families using a novel hierarchical clustering technique which is based on the inclusion tree representation scheme. Conventional clustering techniques are modified in order to comply with the characteristics of the job grouping problem in PCB assembly. Hence, proposed group setup approaches allow grouping only if both conditions for makespan improvement and magazine capacity constraint are satisfied. Because of the hierarchical structure of the grouping process, two improvement heuristics are implemented and applied to the initial group setup solution.