Material property evaluation by analysis of indentation imprint image
- 주제어 (키워드) Indentation testing , Optical flow , Displacement field , Mechanical properties , Residual stress
- 발행기관 서강대학교 일반대학원
- 지도교수 이형일
- 발행년도 2025
- 학위수여년월 2025. 2
- 학위명 박사
- 학과 및 전공 일반대학원 기계공학과
- 실제 URI http://www.dcollection.net/handler/sogang/000000079448
- UCI I804:11029-000000079448
- 본문언어 영어
- 저작권 서강대학교 논문은 저작권 보호를 받습니다.
초록 (요약문)
This study proposes a non-destructive method for material property evaluation based on the analysis of surface in-plane displacement fields induced by indentation, and develops the necessary experimental equipment and algorithms to implement it. The primary goal is to establish a methodology capable of evaluating various mechanical properties, including plastic properties, residual stress, and viscoelastic behavior. While conventional instrumented indentation testing (IIT) has been used for material property evaluation, its reliance on load-displacement curves limits the accuracy. This study addresses the limitations by introducing an approach that utilizes surface displacement field analysis to extract more comprehensive information. To achieve this, an inhouse-made testing system integrating an optical imaging system with a precision indentation module was developed. The system captures pre- and post- indentation images to calculate surface displacement fields using an optical flow (OF) algorithm. Furthermore, to correct errors caused by micro-vibrations and positional restoration inaccuracies, a correction algorithm based on the origin symmetry condition of in-plane displacement field generated by axisymmetric indenters was proposed. This approach reduced measurement errors in the displacement field. The displacement fields obtained from indentation tests were validated by comparison with indentation finite element analysis (FEA) solutions, showing over 99 % agreement, thus confirming the reliability of the proposed system and algorithm. Subsequently, a neural network-based material property evaluation method was developed using data from indentation tests and FEA. The artificial neural network (ANN) model was trained on FEA data, and the input dataset was optimized by analyzing the identifiability index (I-index) derived from the input dataset configuration, which enhanced the model's performance. Random noise was added to the training data to incorporate potential noise from experimental data into the model, improving its robustness. To verify the accuracy of residual stress and plastic property evaluations, indentation tests were conducted on specimens with induced surface stresses using the custom-built equipment. The measured radial and vertical displacement fields were input into the ANN model to predict residual stress and plastic properties, which were then compared with actual values to validate the model's reliability. For analyzing viscoelastic properties, this study proposed a method to measure the creep characteristics through in-plane displacement fields after indentation tests. In particular, time-dependent recovery of displacement fields was observed for viscoelastic materials like PMMA and PC under various loading conditions. The analysis revealed that the recovery of displacement fields is directly attributable to the viscoelastic characteristics of the materials, indicating the feasibility of deriving viscoelastic parameters through indentation tests. By applying the embedded center of dilatation (ECD) model, the initial residual stress state on the material surface was determined without complex inverse analysis. Using strain calculated from the displacement field and the initial stress, creep model parameters were estimated. This approach demonstrated the ability to evaluate the initial residual stress state and time-dependent creep deformation behavior from simple analysis of indentation imprint images. FEA and experimental results for PMMA and PC confirmed that the proposed method achieves accuracy comparable to conventional tensile creep tests and can serve as a practical tool for evaluating viscoelastic properties, even with small-scale specimens. In conclusion, this study presents a novel non-destructive method for material property evaluation that overcomes the limitations of conventional indentation tests. By incorporating high-precision data analysis and noise correction, this method provides a powerful tool for evaluating various mechanical properties through indentation testing. The results of this research have high applicability not only in material science but also in diverse industrial fields such as aerospace, automotive, and biomedical engineering, contributing significantly to the advancement of non-destructive material evaluation technologies.
more목차
1. Introduction 1
1.1 Motivation of the research 1
1.2 Research objectives 4
1.3 Research outline 5
2. Literature review 7
2.1 Instrumented indentation test (IIT) 8
2.2 Material evaluation based on the theoretical analysis 10
2.2.1 Elastic modulus and hardness evaluation 10
2.2.2 Plastic properties evaluation 11
2.2.3 Viscoelastic properties evaluation 14
2.3 Material evaluation through indentation inverse analysis 16
2.3.1 Plastic properties evaluation 16
2.3.2 Residual stress evaluation 18
2.3.3 Viscoelastic properties evaluation 20
2.4 Material evaluation through indentation imprint image 22
2.5 Summary 24
3. Measuring in-plane indentation displacement 25
3.1 Objectives 25
3.2 Theoretical and numerical approaches 26
3.2.1 Noise reduction algorithm 26
3.2.2 Repositioning error correction 28
3.2.3 Indentation simulations 32
3.2.4 Determination of element size in FEA 33
3.2.5 Determination of friction coefficient in FEA 34
3.3 Experiments 35
3.3.1 Equipment design 35
3.3.2 Materials 37
3.3.3 Microindentation test 39
3.4 Results and discussion 40
3.4.1 Effect of image averaging on the displacement field 40
3.4.2 Measurement of in-plane displacement field 42
3.4.3 Grain size and surface roughness effects 44
3.4.4 Effect of indenter stiffness on in-plane displacement field 48
3.4.5 Verification of corrected ui field 49
3.5 Summary 52
4. Evaluation of residual stress and plastic properties 53
4.1 Objectives 53
4.2 Numerical approaches 54
4.2.1 Indentation simulations 54
4.2.2 Evaluation parameters for residual stress and plastic properties 55
4.2.3 Artificial neural network for material evaluation 57
4.2.3 Sensitivity of indentation parameters 60
4.3 Experiments 62
4.3.1 Materials 62
4.3.2 Stress induced indentation test 63
4.4 Results and discussion 66
4.4.1 Identifiability of inverse parameters 66
4.4.2 Construction of ANN model 67
4.4.3 Numerical verification of ANN model 70
4.4.4 Experimental verification of ANN model 74
4.5 Summary 77
5. Evaluation of creep properties 78
5.1 Objectives 78
5.2 Theoretical and numerical approaches 79
5.2.1 Embedded center of dilatation model 79
5.2.2 Time hardening creep model 81
5.2.3 Indentation simulation 83
5.3 Experiments 84
5.3.1 Uniaxial creep test 84
5.3.3 Indentation relaxation test 85
5.4 Results and discussion 86
5.4.1 Recovery of indentation in-plane displacement 86
5.4.2 Estimation of indentation-induced residual stress via ECD model 90
5.4.3 Comparison of indentation and uniaxial creep tests 93
5.5 Summary 95
6. Conclusions and future aspects 96
6.1 Measurement of indentation included in-plane displacement 96
6.2 Application in-plane displacement to material evaluation 97
6.3 Future aspects 98
6.3.1. Enhancement of measurement resolution 99
6.3.2. Application to non-homogeneous materials 99
6.3.3. Real-time indentation deformation measurement 100
6.3.4. Standardization for industrial applications 101
Appendix 102
A.1 Gunnar Farnebäk algorithm for optical flow 102
A.2 Effect of optical flow parameters on displacement field 104
References 106
