Wednesday seminar Assessment of Shear Band Characteristics in Cohesive Soils using Digital Image Analysis Technique for Plane Strain Tests - Concept & Evaluation of accuracy for DIC- Kwak, Tae young Master course Seoul National University Contents 1. Introduction 2. Digital Image Correlation 3. Evaluation of Accuracy for DIC 4. Summary Page 1 Introduction Assessment of Shear Band Expansion of shear band by loading is related to failure mode of soil Traditional methods measure overall soil behavior - Difficult to assessment of shear band Evaluation of internal soil behavior near shear band - Importance of the digital image analysis Page 2 Introduction Object of This Research – Internal deformation of specimen during compression shearing • Internal behavior with stress-strain response • Spatial distribution for deformation – Assessing the shear band characteristics for cohesive soil • Generation of shear band, corresponding to the stress state and strain level • Characteristics of shear band for soil type (NC vs. OC) Page 3 Introduction Methodology – Plane strain test for the observation of failure surface • Internal behavior of the soil follows the surface behavior at plane strain condition. • Occurrence of shear band characteristics is clear at plane strain test. – Digital image analysis for the internal deformation of specimen • Quantitative measurement for internal deformation value of specimen • Available for high accuracy and continuous evaluation during shearing Page 4 Introduction Flow Chart of the Research – Specimen: EPK Kaolinite – Test Procedure Specimen preparation Plane strain test Page 5 Transducers in test apparatus digital image acquisition Overall stress-strain behavior Internal deformation using digital image analysis Internal deformation and Shear band characteristics with shearing level Digital Image Correlation Concept of Digital Image Correlation – Evaluation of displacement between two digital images • Determine displacement vector for selected subset Before After ? Page 6 Digital Image Correlation Digital Image Correlation Procedure – Select the subset pixel from the first picture. And then consider expected deformation of the selected subset pixel – Perform cross correlation for each pixel by moving 1pixel, so correlation between two subset pixels can be found out – Determine the location of highest correlation peak – Interpolate the surrounding of highest correlation peak Page 7 Digital Image Correlation Influence Factor for DIC Method Image resolution Size of pixel subset Deformation condition Determined by pixel of digital camera Image pattern Page 8 Higher pixel of digital camera, higher image resolution Digital Image Correlation Influence Factor for DIC Method Image resolution Size of pixel subset Determined by size of pixel subset Deformation condition Image pattern Page 9 Larger size of pixel subset, higher uniqueness Difficulty in observing the behavior for each location of the soil at too large size of pixel subset Digital Image Correlation Influence Factor for DIC Method Image resolution Size of pixel subset Rigid body movement / Deformation Deformation condition Image pattern Page 10 Higher accuracy for rigid body movement Lower accuracy for high deformation - Set-up DIC analysis interval shorter! Digital Image Correlation Influence Factor for DIC Method Image resolution Size of pixel subset Image pattern classification Deformation condition Grid pattern (Alshibli et al.) - Same pattern makes difficult to classify Image pattern Color of soil particle (Jang et al.) Arbitrary pattern by spray (Kim et al.) Page 11 Evaluation of Accuracy for DIC Evaluation of Accuracy for DIC Method (Jang et al, 2009) – Comparison between criterion displacement and analyzed displacement • Criterion displacement : Exact particle movement › Make a particle movement by DEM program • Analyzed displacement : Measured movement by DIC method › Use initial digital image and deformed digial image by DEM program N [( u Std _ E rr u D E M ) ( v D IA v D E M ) ] 2 D IA i N (u D E M ,v D E M ):D isp la ce m e n t b y D E M p ro g ra m (u D IA ,v D IA ):D isp la ce m e n t b y D IC m e th o d Page 12 2 Evaluation of Accuracy for DIC Evaluation of Accuracy for DIC Method (Jang et al. 2009) – Image resolution • Higher image resolution, lower standard error ! – Deformation condition • Larger deformation, higher standard error ! – SizeImage of pixel subset Volume change Distortion resolution •Limitation Larger sizefor of pixel subset, lower standard error ! with DEM program DIC method for granular soils 0.5% 1.5% 5.0% 0.5% 1.5% 5.0% (mm/pixel) – Image pattern DEM program is not available for cohesive soil However, 150×150 0.1477 0.2335 • Distribution of 0.0852 soil particles with different color0.0978 300×300 0.0283 0.0521 0.1171 0.0258 0.1531 0.1212 0.0557 0.0941 Evaluation for cohesive soil in plane strain test is needed With alternative methodology instead of DEM program 600×600 0.0147 0.0408 0.1067 0.0116 0.0101 0.0585 Page 13 Evaluation of Accuracy for DIC Evaluation of DIC method in cohesive soil for PS test Image OK resolution Size of pixel subset Deformation condition Image pattern OK Digital camera : 12-million pixels (4288 x 2848 Pixels) digital camera Nikkon D90 Arbitrary pattern for cohesive soil by spray (Kim et al.,2011) Adequate pixel subset size Uniqueness ↔ Spatial distribution Small deformation interval Page 14 Accuracy & Precision for Analysis ↔ Analysis efficiency Evaluation of Accuracy for DIC Evaluation of DIC method in cohesive soil for PS test 1) Evaluation by void-ratio Almost rigid body movement Comparison between void ratio change and digital image analysis(△e ↔△εv) Exclude upper & near shear band section - Consideration of soil disturbence Page 15 Evaluation of Accuracy for DIC Evaluation of DIC method in cohesive soil for PS test 2) Evaluation by LVDT Volume change during consolidation Comparison between LVDT and digital image analysis(△εLVDT ↔△εDIA) Page 16 Summary Assessment of shear band localization – Importance of digital image analysis – Investigate overall stress-strain behavior and observe deformation of specimen during shearing Introduction of digital image correlation – Concept of the digital image correlation – Influence factors for accuracy of the DIC method Evalation of accuracy for DIC – Literature review about Evaluation of accuracy in granular soil – Suggestion about method for evaluation of accuracy in cohesive soil Page 17 Thank You!