Experimental analysis and hybrid predictive modeling of microhole drilled with high-energy electron beam on industrial materials

목차 (Table of Contents)

• ABSTRACT I
• TABLE OF CONTENTS III
• LIST OF FIGURES V
• LIST OF TABLES VIII
• NOMENCLATURE IX
• 1. Introduction 1
• 1.1. Research background and motivation 1
• 1.2. Research objectives 4
• 1.3. Outline of this work 5
• 2. Literature reviews 7
• 2.1. Trends in microhole drilling technology 7
• 2.1.1. Microhole drilling on metallic substrate 8
• 2.1.2. Microhole drilling on Carbon Fiber-Reinforced Plastic 9
• 2.2. Methodologies of Through Silicon/Glass Via 10
• 2.2.1. Progress in nonconvetional drilling for brittle material 10
• 2.2.2. Trends in Through Silicon Via technology 10
• 2.2.3. Advancement in Through Silicon Glass technology 13
• 2.3. Research field of electron beam drilling 16
• 3. Experimental drilling with a high energy electron beam 17
• 3.1. Experimental setup 17
• 3.1.1. Equipment of electron beam drilling and measurement 17
• 3.1.2. Major process prameters of electron beam drilling 22
• 3.1.3. Industrial materials as target substrate 24
• 3.2. Experimental results of drilling process 26
• 3.2.1. Metallic substrate 26
• 3.2.2. Thin silicon wafer and glass 31
• 3.2.3 Carbon Fiber-Reinforced Plastics 43
• 3.3. Subconclusion and summary 48
• 4. Enhanced numerical predictive modeling 50
• 4.1. Original predictive model of electron beam drilling 50
• 4.1.1. Governing equation and gaussian model-based profiling 51
• 4.1.2. Numerical prediction for temperature distribution 53
• 4.2. Modified approach for enhancing predictive model 56
• 4.2.1. Adjusted application of absorptance mechanism 58
• 4.2.2. Modeling of AIbased electromagnetic lens 60
• 4.2.3. Machine-learned magnetic field and beam spot generation 64
• 4.3. Validation of predictive accuracy in enhanced predictive model 71
• 4.3.1. Limitation of original predictive results 71
• 4.3.2. Enhanced accuracy and visualization. 73
• 4.4. Subconclusion and summary 78
• 5. Deep learning-based prediction of microhole profile 79
• 5.1. Prediction in post-drilling step with generative adversarial network 79
• 5.1.1. Pix2Pix model for prediction of cross-sectional-hole 80
• 5.1.2. Modification of U-Net architecture by embedding Multi-head Attention 82
• 5.1.3. Preprocessing of training/test data for deep learning model 87
• 5.2. Performance validations of deep learning-based model 90
• 5.2.1. Performance evaluation of training and test for predictive model 90
• 5.2.2. Validation of predictive accuracy and precision for generative cross-sectional hole 96
• 5.3. Prediction in pre-drilling step using hybrid modeling 103
• 5.4. Subconclusion and summary 109
• 6. Conclusion 110
• 6.1. Conclusion and recommandation 110
• 6.2. Path forward 112
• REFERENCES 113
• ACKNOWLEDGEMENT 124