Baixe o app para aproveitar ainda mais
Prévia do material em texto
Developments in Geotechnical Engineering A. Murali Krishna Takeshi Katsumi Editors Geotechnics for Natural Disaster Mitigation and Management Developments in Geotechnical Engineering Series Editors Braja M. Das, California State University, Henderson, CA, USA Nagaratnam Sivakugan, James Cook University, Townsville, QLD, Australia More information about this series at http://www.springer.com/series/13410 http://www.springer.com/series/13410 A. Murali Krishna • Takeshi Katsumi Editors Geotechnics for Natural Disaster Mitigation and Management 123 Editors A. Murali Krishna Department of Civil Engineering Indian Institute of Technology Tirupati Tirupati, AP, India Takeshi Katsumi Graduate School of Global Environmental Studies Kyoto University Kyoto, Japan ISSN 2364-5156 ISSN 2364-5164 (electronic) Developments in Geotechnical Engineering ISBN 978-981-13-8827-9 ISBN 978-981-13-8828-6 (eBook) https://doi.org/10.1007/978-981-13-8828-6 © Springer Nature Singapore Pte Ltd. 2020 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, expressed 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. 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 https://doi.org/10.1007/978-981-13-8828-6 Preface Indian Geotechnical Society (IGS) and Japanese Geotechnical Society (JGS) made an agreement of cooperation to promote and enlarge the exchange of technical, scientific and professional knowledge related to geotechnical engineering. In this context, a series of workshops are being conducted with collaboration between the two societies (IGS and JGS) for the growth of geotechnical engineering and fruitful interaction between researchers in both the countries. The first workshop was held in 2011 at Kochi, India, on ‘Earthquake Geotechnical Engineering’ theme, and the second workshop was held in 2015 at Fukuoka, Japan, on ‘Geotechnics for Resilient Infrastructure’ theme. The third workshop was held in conjunction with ‘Indian Geotechnical Conference 2017’ in December 2017, Guwahati, India, on ‘Geotechnics for Natural Disaster Mitigation and Management’ theme. The workshop focussed on the recent advances and the developments that are taking place in geotechnical aspects of natural disaster mitigation and management. Speakers from Japan and India presented some of the salient aspects of natural disasters and their mitigation strategies. A total of about 30 eminent researchers and practitioners from both the countries participated in the workshop deliberations. This book titled Geotechnics for Natural Disaster Mitigation and Management is the compilation of the some of the expert deliberations made at the third Indo-Japan workshop held on 13 December 2017 at IIT Guwahati, India. This book is organized into 13 chapters covering the landslides and earthquake natural disasters for effective mitigation and management. Chapter 1 discusses the geotechnical and geological perspectives of The 2017 July Northern Kyushu Torrential Rainfall Disaster. Chapter 2 outlines disaster management strategies that are being adopted in India and typical geotechnical characterization aspects for earthquake disaster. Chapter 3 presents details of shear strength characterization gravel–tyre chips mixtures for sustainable geotechnical engineering through the usage and recycled waste tyres for various disaster mitigation techniques. At times, back analysis-based approaches are to be adopted to the characterization of soils through numerical simulations and analyses. Chapter 4 explains an example of the application of such an approach for determination of elastic modulus of soil. It is also essential to account for a special variation of the properties of soils for realistic analyses. Chapter 5 provides the v details of the evaluation of the spatial distribution of strength of few embankments in Japan and India through field investigation. Various aspects of landslides like integrating rainfall load in landslide warning system (Chap. 6); application of physically based models for landslide hazard evaluation (Chap. 7); and debris flow-related aspects in the case of rock fall (Chap. 8) are also included in this book. Chapter 9 describes the significance of surface and sub-surface drainage measures for efficient landslide mitigation strategies. Waterfront retaining structures including breakwater systems are being affected by earthquake and tsunami. Design aspects (Chap. 10) and countermeasures (Chap. 11) against disaster-induced instabilities are also included herein. For efficient disaster mitigation strategies, ground improvement methods play a vital role. Ground modification techniques using vibro methods (Chap. 12) and cement-based grouting are briefly discussed. We sincerely thank and appreciate the efforts of all the expert contributors in formulating the book chapter contributions. We also thank Springer team for giving inputs for finalizing the manuscripts and publishing the contributions to spread the conglomerated ideas through this book. It is believed that this book will be a good resource for academicians, researchers, practising professionals and, especially, students of the geotechnical fraternity related to the natural disaster mitigation and management. Tirupati, India A. Murali Krishna Kyoto, Japan Takeshi Katsumi vi Preface Contents 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster—Geotechnical and Geological Perspectives . . . . . . . . . . . . 1 H. Hazarika, S. Yamamoto, T. Ishizawa, T. Danjo, Y. Kochi, T. Fujishiro, K. Okamoto, D. Matsumoto and S. Ishibashi 2 Disaster Management in India and Characterization for Geohazards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 B. K. Maheshwari 3 Shear Strength Behaviour of Gravel–Tire Chips Mixture . . . . . . . . 33 S. M. K. Pasha, H. Hazarika and N. Yoshimoto 4 Elastic Modulus Estimation Using a Scaled State Parameter in the Extended Kalman Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43 M. C. Koch, A. Murakami and K. Fujisawa 5 Spatial Distribution of Strength—Comparison Between Indian and Japanese Embankments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 S. Nishimura, K. Imaide, T. Ueta, T. Hayashi, K. Inoue, T. Shibata and B. Chaudhary 6 Integrating Rainfall Load into Remedial Design of Slopes Affected by Landslides . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 G. L. Sivakumar Babu and Pinom Ering 7 Investigation of Rainfall-Induced Landslides at the Hillslopes of Guwahati Region, Assam . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75 Chiranjib Prasad Sarma, Arindam Dey and A. Murali Krishna 8 Evaluation of the Risk Distribution of the Debris Flow Occurred Using Numerical SimulationSubjected to Rockfall . . . . . . . . . . . . . 89 Y. Isobe, H. Inagaki and H. Ohno vii 9 Significance of Drainage Measures on Landslide Mitigation Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Minimol Korulla 10 Design of Waterfront-Retaining Walls Subjected to Waves and Earthquakes: A Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 109 D. Choudhury and B. G. Rajesh 11 Instability of Composite Breakwater Subjected to Earthquake and Tsunami and Its Countermeasures . . . . . . . . . . . . . . . . . . . . . . 119 B. Chaudhary, H. Hazarika, A. Murakami and K. Fujisawa 12 Ground Modification Techniques to Improve Liquefaction Resistance in Indo-Gangetic Soils . . . . . . . . . . . . . . . . . . . . . . . . . . 127 J. T. Shahu and Mamata Mohanty 13 Extended Application of Cement-Based Grouting to Gravel/Boulder Ground Improvement . . . . . . . . . . . . . . . . . . . . 135 H. Ishii viii Contents About the Editors Dr. A. Murali Krishna joined the Department of Civil Engineering, Indian Institute of Technology (IIT) Tirupati as an Associate Professor (Geotechnical Engineering) inMay 2019.He has been a facultymember in Department of Civil Engineering at IIT Guwahati, since 2008. He obtained doctoral degree from Indian Institute of Science Bangalore, M.Tech degree from IIT Kanpur and B.Tech degree from Sri Venkateswara University, Tirupati. His research interests include: Earthquake Geotechnics, Geosynthetics and Ground Improvement, Site characterization and Numerical and Physical modelling of geotechnical structures. Dr. Murali Krishna supervised 7 Doctoral students and 24 Masters students. He co-authored nearly 170 publi- cations of technical papers in international/national Journals and conference/seminar proceedings, includ- ing book chapters. He is a recipient of BRNS Young Scientist Research award, BOYSCAST fellowship and HERTAGE fellowship. Dr. Murali Krishna is an exec- utive member of ISRM (India) and ISET. He is also a Member of TC 203 of ISSMGE, since 2011. He served Indian Geotechnical Society as an ‘Executive Member’ for four terms (2011–2018). Dr. Murali Krishna organ- ised several national and international workshops and short courses. He is a reviewer for several national and international journals. ix Dr. Takeshi Katsumi is a Professor at the Graduate School of Global Environmental Studies (GSGES), Kyoto University, Japan. He served as Assistant to the Executive Vice-President of Kyoto University for two years from 2012 October, and is currently Vice Dean of GSGES. He graduated from the Department of Civil Engineering, Kyoto University, and obtained his doc- toral degree from the same university in 1997. He has research interests in a variety of topics of environmental geotechnics, including waste landfills, remediation of contaminated sites, and re-use of by-products in geotechnical applications. He has received several awards including the “JSPS Prize” by the Japan Society for the Promotion of Science. He has been a member of ISSMGE Technical Committee No. 215 on Environmental Geotechnics for more than 15 years, and has been the International Secretary of the Japanese Geotechnical Society (JGS) since 2014. He has deliv- ered keynote lectures at several international confer- ences such. He has been involved in several projects regarding the recovery works from the 2011 East Japan earthquake and tsunami, and has been a contributing member to the Central Environment Council of Japan for the last two years. x About the Editors Chapter 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster—Geotechnical and Geological Perspectives H. Hazarika, S. Yamamoto, T. Ishizawa, T. Danjo, Y. Kochi, T. Fujishiro, K. Okamoto, D. Matsumoto and S. Ishibashi 1.1 Introduction Record-breaking localized and torrential rainfall on July 5, 2017, caused many land- slides, debris flows, and flooding of rivers in several areas of the Fukuoka and Oita prefectures of Kyushu, Japan. The disaster, officially known as “The 2017 July Northern Kyushu Torrential Rainfall Disaster”, is one of the unprecedented in disaster histories of Japan in many decades. The record-breaking magnitude of the heavy and localized rains triggered landslides and debris flow that swept massive volumes of driftwood and soils into cities, towns, and villages located at the bases of mountains, leaving 34 dead including 4 still missing. It also resulted in 1,428 H. Hazarika (B) Department of Civil Engineering, Kyushu University, Fukuoka, Japan e-mail: hazarika@civil.kyushu-u.ac.jp S. Yamamoto Chuo Kaihatsu Corporation, Fukuoka, Japan T. Ishizawa · T. Danjo NIED, Tsukuba, Japan Y. Kochi K’s Lab Inc., Yamaguchi, Japan T. Fujishiro · K. Okamoto Fukuyama Consultants Co. Ltd., Kitakyushu, Japan D. Matsumoto Japan Foundation Engineering Co. Ltd., Fukuoka, Japan S. Ishibashi Nihon Chiken Co. Ltd., Fukuoka, Japan © Springer Nature Singapore Pte Ltd. 2020 A. M Krishna and T. Katsumi (eds.), Geotechnics for Natural Disaster Mitigation and Management, Developments in Geotechnical Engineering, https://doi.org/10.1007/978-981-13-8828-6_1 1 http://crossmark.crossref.org/dialog/?doi=10.1007/978-981-13-8828-6_1&domain=pdf mailto:hazarika@civil.kyushu-u.ac.jp https://doi.org/10.1007/978-981-13-8828-6_1 2 H. Hazarika et al. completely, half and partially destroyed houses. The large amount of driftwood that got caught on bridges and other structures blocked the flow of the rivers, thereby spreading the damage even further. According to data analysis by the Japan Meteorological Agency (JMA), in Asakura city, Fukuoka prefecture recorded some 1,000 mm of rain in the 24-h period ending onmidnight of July 5, 2017. The city of Hita, Oita Prefecture and Toho village of Fukuoka Prefecture each had some 600 mm of rain; and the city of Nakatsu, Oita Prefecture, and Soeda town of Fukuoka Prefecture each had around 500 mm of rain, according to an analysis by the JMA’s Fukuoka Regional Headquarters [1]. Figure 1.1 shows the extent of the flooded area and the landslide disaster areas. In the figure, blue areas represent the flooded rivers and their tributaries, and the red areas represent the extent of damage due to landslides, debris flows, and flood- related damage. The authors conducted an extensive investigation of the damage covering several cities, towns, and villages of the Fukuoka and Oita Prefectures. The investigation team consisted of people from the industry, government, and academia. Figure 1.2 shows the survey routes and locations of the surveyed areas (Location 1 to Location 6). Flooding, numerous slope failures and debris flows caused serious damage to many infrastructures, and they can be broadly classified into the following categories: (1) Landslides and debris flow related damage, (2) Formation of landslide dam, (3) River erosion, (4) Agricultural dam damage, (5) Roads and railways damage, and 6) Environmental damage due to disaster debris (including driftwood). A preliminary investigation report by the Japanese Geotechnical Society has covered many aspects of the abovementioned damage [2]. Asakura City Toho Village Hita City Haki IC Asakura IC Chikugo River Fig. 1.1 Distribution of the flooded and damaged area [6] 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster … 3 Loc.1 Hita-City Ono-District Nagio Loc.3 Asakura-City Hakiakatani Loc.2 Kazurahae Toho-Village Asakura- Gun Loc.4 Choan-temple Asakura-City Sugawa Loc.5 Asakura-City Miyano Loc.6 Pond around Asakura-City Sugawa Survey area of slope failure along Otoishi River Fig. 1.2 Location map of the investigation area (modified from the [6]) Many professional societies and research institutions of Japan have carried out disaster survey and submitted their reports. However, in most of the reports, they focused only on the shallow landslides, and none provide an in-depth analysis of the failure form. The authors have felt that the damage causedby this disaster cannot be analyzed solely from the geotechnical engineering perspectives, as the areas covered by the disaster have many complicated geologies and geomorphologies, which made the damage devastating with far-reaching economic implications [3]. Therefore, the authors conducted a special investigation of the slope failure areas in and around Otoishi River basin, located approximately 3 km upstream from Matsusue Elemen- tary School (see Fig. 1.2). It was found that the slope failures that occurred in this basin were not only shallow landslides of strongly weathered granite (decomposed granite) but also included failures that reflected the geological structure (fault fracture zone) and large-scale (deep-seated) landslides in the areas with crystalline schist. In this paper, the characteristics of the slope failures that occurred in the Otoishi River basin are summarized, and the classification of failure focusing on the geological structure are described based on the slope failure mechanism. 1.2 Topographical and Geological Features of the Otoishi River Basin As shown in Fig. 1.3, the Otoishi River is located in the southern area of the Sangun Mountains where themain peaks extend in the northwest–southeast direction. It joins the Akatani River, the right branch of the Chikugo River, near Matsusue Elementary School, which was badly affected due to accumulation of disaster debris in the school compound leading to the closure of the school for months. 4 H. Hazarika et al. Loc. 3: Landslide of crystalline schistLoc. 4: Failure of crystalline schist Loc. 2: Failure around the fault fracture zone Loc. 1: Failure of Strongly weathered granite Matsusue Elementary Fig. 1.3 Location map of the detailed investigation (modified from [6]) The geological map of the investigated areas (Fig. 1.4) suggests that a valley plain lies in the Otoishi River basin in the northwest–southeast direction. The geological formations of the entire basin are indicated in the figure. On the left side of the bank, valleys lay vertically in the northeast–southwest direction, while on the right bank side, valleys lay horizontally in the northeast–southwest direction. In addition, several topographic elements such as saddles, knick lines, and straight valleys can also be observed in the same direction. These directions are in line with the direction of the fault fracture zone in northern Kyushu. This suggests that the topography of this basin is vulnerable to discriminate erosion due to the geological structure (faults). Moreover, Asakura granodiorite is distributed in the middle-to-downstream basin of the Otoishi River, and Sangun metamorphic rocks are distributed in the upstream area. Mount Hiko volcanic rocks from the Pliocene epoch in the neo-tertiary period are distributed in the ridge, which is the drainage basin of the uppermost basin. 1.3 Characteristics of the Slope Failures 1.3.1 Granite Failure The most common failure form found along the basin occurred in the so-called strongly weathered “decomposed granite,” which transformed into soil (Location 1 of Fig. 1.3), as shown in Fig. 1.5a. The failure was roughly 1 m deep, and many traces of springs are thought to be due to piping. Gully erosions had formed on the failure surface. Since there was little colluvial deposit at the toe of the slope, it had likely been eroded by the debris flow and been washed downstream. The probable failure mechanism of the slope is described schematically in Fig. 1.5b. 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster … 5 Akatani River Akatani River Otoishi River Chikugo River Fig. 1.4 Geologicalmap of the investigated area (tl: Low terrace deposit, A4:Aso 4 pyroclastic flow deposit, tm: Medium terrace deposit, H: Mount Hiko volcanic rocks, K: Kita-Sakamoto formation, E: Paleogene period formations (Hoshuyama, Kawamagari, and Haji), Ku: Asakura granodiorite, m2: Muddy schist, s2: Sandy schist, g2: Hard-iron schist) (a) Slope failure with many gully erosions (b) Schematic profile Fig. 1.5 Failure in highly weathered granite InLocation 2,which is in the riverbank side slope in themiddle basin of theOtoishi River (Refer toFig. 1.3), a 10mwide fault fracture zonewas found (Fig. 1.6).Multiple such formations were observed in the northwest–southeast direction, as well as the north-northeast–south-southwest and northeast–southwest directions. In some, fault fracture zones strongly crushed by fault gouge could be observed. In this location, at the end of the slope failure zone, a 1 m-wide gouge zone and a fault fracture zone with a crack zone with a large opening of around 5 m wide were verified (Fig. 1.7). 6 H. Hazarika et al. The direction of the fault was N54°W, 70°SE, and it worked as a receiving board to the slope. Furthermore, a substantial volume of groundwater was flowing from the crack zone on the slope side (around 50 L/min). An analysis of this failuremechanism (Fig. 1.7b) based on the characteristics of the crush zone and flowing water suggests that the gouge zone, which normally indicates low permeability, became a cut-off wall, and the groundwater flowed from the crack zone. However, during the torrential rainfall event, a large volume of groundwater that accumulated in a short period of time likely raised the groundwater level inside the slope resulting in the increase of pore water pressure, and eventually led to a bedrock failure that included the crack zone. Fig. 1.6 Whole view of the failure (a) Enlarged view (b) Schematic profile Gouge zone Crack zone (Large opening) Sound rock (Decomposed granite) Crack zone Topsoil Fig. 1.7 Fault fracture zone and flowing water 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster … 7 Fig. 1.8 Large-scale landslide in the crystalline schist-distributed area (Source GIJ) 1.3.2 Failure of Crystalline Schist 1.3.2.1 Landslide on the Left Side of the Upstream Bank In the upstream area of the Otoishi River with crystalline schist (Location 3 in Fig. 1.3), a landslide with a length of approximately 300 m (landslide scarp to clod end), width of 110 m, and an average depth of 15–20 m occurred (Fig. 1.8). This landslide formed two landslide scarps, and a landslide clod that still remains in the slope. The frontal part of the landslide clod traversed the road below, reached the Otoishi riverbed and stopped, blocking most of the river channel (Fig. 1.9). However, there was no evidence that this moving clod turned into a debris flow. The slope was inclined to the south-southwest, and the landslide occurred near the ridge. The topography of this area before the landslide had typical characteristics, such as a steep slope in the end and a gentle slope at the halfway point (Fig. 1.10). There was a layer of debris with rock fragments over 5 m thick near the left end (Fig. 1.11), indicating that the same activity may have occurred several times in the past. 8 H. Hazarika et al. Fig. 1.9 Frontal section of the landslide clod blocking most of the Otoishi River Fig. 1.10 Landslide scarp seen from halfway up the slope (gentle gradient) Fig. 1.11 Layer of debris with rock fragments on the left end 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster … 9 The geology of this slope was crystalline schists consisting of muddy and basic schist penetratedwith fine-grain granite. Xenolith, crystalline schist, was found in the fine-grain granite near the landslide scarp (Fig. 1.12). The crystalline schists, which dominated the landslide scarp and side, were more evident on the upper areas. Also, this area was thought to be the bedrock, which showed many cracks, with evidence of fractioning and loosening. Since there was no regularity in the direction of cracks (Fig. 1.13), the landslide may have caused high deformation and crushing. This was because the strike and inclination measured in the area were N48°E and 36°NW, which did not match the measurements outside the landslide range of N86°W and 32°N. Furthermore,the entire geological structure of the original slope presumably formed an opposite slope. In addition to the fine-grain granite (aplite), a small-scale fault fracture zone (a fault breccia zone less than 1 m wide) from the north-northeast to south-southwest direction was found in some places with the intense formation of clay. Heavily transformed fine-grain granite boulders were also observed. The rela- tionship between these observations and the mechanism of the landslide is unknown, but there may be a geological predisposition involving faults and alterations. Except for the clod immediately below the landslide scarp, the landslide mass of this slope did not preserve its original form. The status of the landslide scarp and the side indicated that the landslide mass was composed of a sediment layer with rock fragments and crystalline schists that had turned into soil and crumbled (rock classification: D ~ CL). The thickness of this landslide mass was estimated to be 15–20 m on average, with a maximum of 25 m. The relatively firm crystalline schist (rock classification: mainly CM) with few cracks on the lowermost part of the mass strongly suggests Fig. 1.12 Fine-grain granite and crystalline schist 10 H. Hazarika et al. Fig. 1.13 Left landslide scarp that it was composed ofD ~ CL class crystalline schist (Fig. 1.14). However, because landslide clods remained on the failure surface, there was no geological weak line that could act as a sliding surface. Multiple points of outflowing water were observed in the area from the land- slide scarp to its end, where the flow increased (Fig. 1.15). However, since most of these gushed out after the landslide clod yielded, the detailed status of groundwa- ter is unknown. Since it has been estimated that the landslide occurred at a depth of approximately 20 m, the deep groundwater likely contributed to the landslide movement. Schematic diagram of the mechanism of failure is shown in Fig. 1.16. 1.3.2.2 Group of Failures on the Right Side of the Upstream Bank In Location 4, multiple slope failures occurred (Fig. 1.17) opposite to the landslide in Location 3 of the Otoishi Riverbank. The failures were smaller than the landslide of the opposite bank (left side of the bank) and around 5 m deep, but different types of failure forms were found, including failure of colluvial deposit that filled the valley, failure of schist that weathered and transformed into soil, and failure dominated by a fault fracture zone. Since there were fault fracture zones of almost the same formation as in the granite-distributed areas, as well as schists full of cracks and aplite formations in some parts of the failure surface, the primary cause of the failure was likely controlled by these complicated geological structures. The failure mechanism in this location can be classified into two groups. The first failuremechanism (Fig. 1.18) is caused by fault fracture zones, as shown in Figs. 1.19 and 1.20, where the bedrock around the fault fracture zone weakened, likely resulting in the failure in this part. The second failure mechanism (Fig. 1.21) is caused by a highly weathered rock, as seen in Figs. 1.22 and 1.23. Soft sedimentation of bedrock 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster … 11 Fig. 1.14 Geologic stratigraphy found at the landslide scarp from aging due to weathering likely caused the failure of highly weathered rock due to the heavy rainfalls. Regardless of the failure mechanism, both bedrock failures undoubtedly involved the weathered rock. 1.3.3 Classification of Failure Patterns in Otoishi River Basin Table 1.1 shows the various failure forms found in this investigation. The classifica- tion in the Table summarizes the failure forms according to the geology and geolog- ical structure. The granite group (Group I) can be subdivided as shallow landslides of decomposed granite (Ia) and bedrock failures (Ib) dominated by the presence of several fa fracture zones. On the other hand, the crystalline schist group (Group II) 12 H. Hazarika et al. Fig. 1.15 Alteration and flowing water Fig. 1.16 Schematic profile (failure and landslide of crystalline schist) can be subdivided as failures of strongly weathered rocks distributed at the surface (surface failure, IIa), bedrock failures (IIb) controlled by the geological structure, and failures caused by large-scale (deep-seated) landslides (IIc). 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster … 13 Colluvial deposit Fault fracture zone Highly weathered rock Schist and granite Fault fracture zone Highly weathered rock Fig. 1.17 Group of slope failures on the right side of the upstream bank (Source GIJ) Fig. 1.18 Illustrative diagram of the failure form involving the crush zone 14 H. Hazarika et al. Fig. 1.19 Low-angle fault distributed on the failure surface Fig. 1.20 Failure controlled by the fault fracture zone (whole view) Fig. 1.21 Illustrative diagram of the failure form due to highly weathered rock 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster … 15 Fig. 1.22 Whole view of the landslide scarp Fig. 1.23 State of the collapsed surface 1.4 Conclusions and Future Challenges Below are the few important findings from this study: (1) Fault fracture zones of multiple formations, such as NW, NS, and EW types, are distributed along the Otoishi River. (2) These zones are the primary cause of the slope failures and landslides that occurred in this basin. (3) The slope failures in the granite-distributed areas can be classified into two groups: shallow landslides of highly weathered granite (decomposed granite) 16 H. Hazarika et al. Ta bl e 1. 1 C la ss ifi ca tio n of fa ilu re pa tte rn s C la ss ifi ca tio n sy m bo l G eo lo gi ca lf ea tu re s/ G eo lo gi ca l st ru ct ur e Sl op e fa ilu re ch ar ac te ri st ic s Sc he m at ic pr ofi le I (G ra ni te ) Ia St ro ng ly w ea th er ed gr an ite so il Sh al lo w la nd sl id es Ib Fa ul tf ra ct ur e zo ne B ed ro ck fa ilu re (f au lt fr ac tu re ty pe ) II (C ry st al sc hi st ) II a So ft se di m en ta tio n of be dr oc k fr om ag in g du e to w ea th er in g Su rf ac e la ye r fa ilu re (c on tin ue d) 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster … 17 Ta bl e 1. 1 (c on tin ue d) C la ss ifi ca tio n sy m bo l G eo lo gi ca lf ea tu re s/ G eo lo gi ca l st ru ct ur e Sl op e fa ilu re ch ar ac te ri st ic s Sc he m at ic pr ofi le II b Fa ul tf ra ct ur e zo ne B ed ro ck fa ilu re (f au lt fr ac tu re ty pe ) II c O ld ta lu s ac cu m ul at io n an d st ro ng -t o- w ea k w ea th er ed sc hi st D ee p- se at ed la nd sl id es 18 H. Hazarika et al. and bedrock failures (a type of failure around the fault fracture zones driven by the geological structure and groundwater). (4) The failure form of the crystalline schist-distributed area on the left side of the upstream bank suggests that it was a landslide failure. Judging from the topographic form before the failure, this area had likely suffered repeatedly from gravity transformation. (5) In the crystalline schist-distributed area on the right side of the upstream bank, multiple failures of strongly weathered areas were found. In addition, a bedrock failure near the fault fracture zone was observed. Many aspects of the torrential rainfall disaster this time remain unclear. For exam- ple, it is still not clearly known how the geological structure and groundwater charac- teristics in the upstream,where crystalline schist is distributed, are related to the slope failures. It is necessary to understand the topography of a wider area and perform surface geological and hydrogeological surveys to examine the failure mechanism in greater detail. Hydrological phenomena are closely related to the geology that forms aquifers and the formation process of its structure [4, 5]. The five failure forms proposed in this paper can be validated from the hydrogeologicalperspective by measuring simple water quality parameters and the specific flow rate. To prepare for the risk of landslide disasters caused by torrential rains, it is first necessary to understand the evolution of the topography of the entire basin, as well as the geological composition and structure in addition to the geotechnical engineering knowledge. This enables a better understanding of the classification of failure forms and failure mechanism to adopt both the hard-type and soft-type countermeasures according to the types of disasters. Therefore, validating the classification of the slope failure forms proposed in this paper from the geotechnical engineering perspective is another important task of the future. Acknowledgements The authors would like to express their sincere gratitude to the following individuals for their assistance and cooperation during the field survey: Mr. Yokou (Chuo Kaihatsu Corporation, Kyushu branch), Mr. Kariya (K’s Lab Inc., Yamaguchi), Mr. Ide (Nihon Chiken Co. Ltd., Fukuoka). The tremendous help during the survey from the graduate students (Messrs Manafi, Ali, Ogo, Zawad and Makimoto) of Geodisaster Prevention Laboratory of Kyushu University, Fukuoka are also gratefully acknowledged. References 1. Japan Meteorological Agency (JMA). http://www.jma.go.jp/jma/menu/H29kyusyu_hokubu. html (2017) 2. Japanese Geotechnical Society (JGS): Report of the 2017 July Northern Kyushu Tor- rential Rainfall Disaster. https://www.jiban.or.jp/wp-content/uploads/2017/07/saigai_ kyusyuhokubugoukinkyuhoukoku20170713.pdf (2017) (In Japanese ) 3. Yamamoto, S., Yata, J., Yano, K.: Classification of slope failure along the Otoishi River: distri- bution and features. In: Proceedings of the Annual Conference of Japan Society of Engineering Geology, pp. 65–74 (2017) (In Japanese) http://www.jma.go.jp/jma/menu/H29kyusyu_hokubu.html https://www.jiban.or.jp/wp-content/uploads/2017/07/saigai_kyusyuhokubugoukinkyuhoukoku20170713.pdf 1 The 2017 July Northern Kyushu Torrential Rainfall Disaster … 19 4. Miyazaki, S.: Hydrogeological survey method. In: Booklet Hydrogeology Series 4. Senjōchi Suikankyō Kenkyū Kikō, Japan (2017) (In Japanese) 5. Kashiwagi, T.: Hydrology: research and measurement method (flow survey and hydrogeologi- cal properties of mountainous areas). In: Booklet Hydrogeology Series 3. Senjōchi Suikankyō Kenkyū Kikō, Japan (2017) (In Japanese) 6. Geospatial InformationAuthority of Japan (GIJ). https://saigai.gsi.go.jp/3/20170726handokuzu/ handokuzu.pdf (2017) https://saigai.gsi.go.jp/3/20170726handokuzu/handokuzu.pdf Chapter 2 Disaster Management in India and Characterization for Geohazards B. K. Maheshwari 2.1 Introduction Thedisastersmaydamage the public infrastructure of billions of dollars. Thedisasters have occurred from millions of years and may continue in the future. However, to mitigate the effects of disasters, it is necessary to understand how a hazard is converted into a disaster. The primary objective is to prevent from turning a hazard into a disaster. India, particularly the Himalayan region, is vulnerable to various disasters including earthquakes, landslides and floods. As a follow up to Disaster Management Act [1], in 2006, the Govt. of India established “National Disaster Management Authority (NDMA)” headed by the Prime Minister. Later guidelines on “National Policy on Disaster Management” was published [10]. Accordingly, there is a paradigm shift, from relief-centric response to a proactive prevention, mitigation and preparedness approach towards disasters. Disasters disrupt the progress of nations particularly, pushbackdevelopingnations by several decades. In recent times, efficient management of disasters, rather than a mere response to their occurrence, has received increased attention worldwide. This is the result of the recognition of the increasing frequency and intensity of disasters and acknowledgement by the government that good governance in a caring and civilized society, needs to deal effectively with the devastating impact of disasters. India is vulnerable to a large number of various types of natural and man-made disasters. The approximate proportions are 58.6% of the landmass is prone to earthquakes of moderate to very high intensity; over 12% of land (40 million hectare) is prone to floods and river erosion; about 76% of coastline (5,700 km out of 7,516 km) is prone to cyclones and tsunamis; 68% of the cultivable area is vulnerable to drought and hilly areas are at risk from landslides and avalanches. B. K. Maheshwari (B) Department of Earthquake Engineering and Former Head, Centre of Excellence in Disaster Mitigation & Management, IIT Roorkee, Roorkee, India e-mail: bkmahfeq@iitr.ac.in © Springer Nature Singapore Pte Ltd. 2020 A. M Krishna and T. Katsumi (eds.), Geotechnics for Natural Disaster Mitigation and Management, Developments in Geotechnical Engineering, https://doi.org/10.1007/978-981-13-8828-6_2 21 http://crossmark.crossref.org/dialog/?doi=10.1007/978-981-13-8828-6_2&domain=pdf mailto:bkmahfeq@iitr.ac.in https://doi.org/10.1007/978-981-13-8828-6_2 22 B. K. Maheshwari Vulnerability to Chemical, Biological, Radiological and Nuclear (CBRN) origin disasters/emergencies also exists. Heightened vulnerabilities to disaster risks can be related to population expansion, urbanization, and industrialization development within high-risk zones, environmental degradation and climate change. Disaster Management Act [1] led to the creation of the NDMA headed by the Prime Minister, State Disaster Management Authorities (SDMAs) headed by the Chief Ministers, and District Disaster Management Authorities (DDMAs) headed by the District Collector/Magistrate or Deputy Commissioner to adopt a holistic and integrated approach to disaster management (DM). The vision of NDMA is to build a safe and disaster resilient India by developing a holistic, proactive, multi-disaster oriented and technology-driven strategy through a culture of prevention, mitigation, preparedness and response. The process, objective and elements of disaster manage- ment as per the government’s policy on disaster management [10] are discussed. One of the strategies for disaster mitigation is “seismic microzonation”, which involves ground motion, liquefaction hazard and slope stability analysis (ISSMG Manual). For seismic microzonation, the geotechnical site characterization is must and the same is discussed here in detail. This characterization has been carried out based on the in situ and laboratory test data. The data of tests conducted by the author and his research group in Roorkee region are presented. The Roorkee is a town in north India and located at about 200 km north of New Delhi in Uttarakhand state. According to IS:1893 [2], Roorkee falls in seismic zone IV. The data indicate that how geotechnical engineering plays a crucial role in the characterization of sites which has been also indicated by Eurocode-8 and NEHRP. 2.2 Process and Objectives of Disaster Management A disaster refers to a catastrophe or mishap from natural or man-made causes, which is beyond the coping capacity of the affected community. Disaster manage- ment involves following integrated process of planning, organizing, coordinating and implementing measures [10]: (i) Prevention of danger or threat. (ii) Mitigation or reduction of risk. (iii) Preparedness to deal with. (iv) Prompt response to disaster threatening situation. (v) Assessing the severity or magnitude of effect. (vi) Evacuation, rescue and relief. (vii) Rehabilitation and reconstruction. (viii) Capacity building including research. Objectives of the National Policy on Disaster Management [10] are (i) Promoting a culture of prevention, preparedness and resilience at all levels through knowledge, innovation and education. 2 Disaster Management in India and Characterization for Geohazards 23 (ii) Encouraging mitigation measures based on technology, traditional wisdom and environmental sustainability. (iii)Mainstreaming disaster management into the developmental planning process. (iv) Establishing institutional and techno-legal frameworks to create an enabling regulatory environment and a compliance regime. (v) Ensuring efficient mechanism for identification, assessment and monitoring of disaster risks. (vi) Developing contemporary forecasting and early warning systems backed by responsive and fail-safe communicationwith information technology support. (vii) Ensuring efficient response and relief with a caring approach towards the needs of the vulnerable sections of the society. (viii) Undertaking construction as an opportunity to build disaster resilient struc- tures and habitat for ensuring safer living. (ix) Promoting a productive and proactive partnership with the media for disaster management. Institutional framework under DM act consists of NDMA, SDMAs, DDMAs and local authorities. Besides these, two important organizations are NIDM (National Institute of Disaster Management) and NDRF (National Disaster Response Force). 2.3 Elements of Disaster Management In general, disaster management can be broadly categorized into two groups, i.e. actions before the disaster which include prevention, mitigation and preparedness and actions after the disaster which includes response, relief, rehabilitation, recovery and reconstruction. These are further discussed in detail [10]. 2.3.1 Prevention, Mitigation and Preparedness As the natural hazards like floods, earthquakes and cyclones cannot be avoided, with mitigation measures along with proper planning of developmental work in the risk- prone area, these hazards can be prevented from turning into disasters. To undertake mitigation measures a multi-pronged approach needs to be adopted: • Building mitigation measures into all development projects. • Initiating of national level mitigation projects, in high priority areas. • Encouraging and assisting state level mitigation projects. • Indigenous knowledge of disaster and coping mechanisms. 24 B. K. Maheshwari Further, mitigation measures include risk assessment and vulnerability mapping, climate change adaption. Preparedness for disasters includes forecasting and early warning systems, strengthening emergency centres, medical preparedness, corporate social partnerships, mock drills, communication, IT support and media partnerships. 2.3.2 Response, Relief and Rehabilitation and Reconstruction and Recovery The existing and new institutional arrangements need to ensure proactive and inte- grated approaches in dealing with any disaster. Prompt and effective response min- imizes loss of life and property. This is possible through contemporary forecasting and early warning systems, fail-safe communication and anticipatory deployment of specialised response forces. A well-informed and prepared community can mitigate the impact of disasters. The NEC (National Executive Committee) coordinates the overall response in the disaster. Further state governments and SDMAare responsible to monitor the situation and informNDMA. All stakeholders need to follow standard operating procedures (SOPs). Incident Command System andMedical response need to be strengthened. Relief is not a provision of emergency relief supplies in time but also facilitating an overarching system of assistance to the disaster victims for their rehabilitation and ensuring their social safety and security. The relief should be prompt, adequate and of approved standards. This includes setting up of temporary relief camps, management of relief supplies, temporary rehabilitation and socio-economic rehabilitation. Incorporating disaster resilient features to “build back better” is the guiding princi- ple of reconstruction. Essential services and intermediate shelters need to be provided at earliest. The capacity development can be addressed effectively with the active and enthusiastic participation of the stakeholders. This process comprises awareness generation, education, training, Research and Development (R&D), etc. It further addresses putting in place appropriate institutional framework, management sys- tems and allocation of resources for efficient prevention and handling of disasters. Thus, a roadmap for disaster management is laid by NDMA. 2.4 Characterization for Earthquake Geohazard Assessment of damage during past earthquakes indicated that the degree of damage due to an earthquake is controlled mainly by three important factors [10]: (a) Earthquake source and path characteristics. (b) Local geological and geotechnical characteristics. (c) Structural design and construction features. 2 Disaster Management in India and Characterization for Geohazards 25 Thus, the seismic ground response at a site is strongly influenced by local soil and geological conditions.Hence, the details of the local geological and geotechnical data along with the background of regional seismotectonic and seismicity are needed for effective evaluation of ground response and site effects. The damage patterns during an earthquake depend on the soil characteristics at a site which will have a major effect on the level of ground shaking. This high- lights the importance of site characterization, particularly, in microzonation studies. The regional tectonic maps as well as surface geology maps and vertical geologi- cal profiles are the essential ingredients for the seismic microzonation study. The geological, geomorphological and geotechnical databases including the thickness of site’s soil conditions are needed for assessing the local site effects for site amplifi- cation as well as for liquefaction and landslide susceptibility. To obtain these data either in situ geophysical and geotechnical explorations and/or laboratory tests are conducted. The results of some of the tests conducted by the author and his research group are discussed. 2.5 In Situ Field Tests for Site Characterization The most common field tests are the Standard Penetration Test (SPT), Cone Pene- tration Test (CPT) and Multichannel Analysis of Surface Wave (MASW) which are discussed in following subsections. 2.5.1 Standard Penetration Test (SPT) The standard penetration test (SPT) is done in a borehole using a split-spoon sampler. The sampler consists of a driving shoe, a split-barrel of circular cross-section. The procedure is discussed in detail in IS: 2131 [3]. In this test, the soil sample is also obtained which can be further examined in the laboratory for index and dynamic properties. The measured N value from field is denoted as Nm and this is equal to the number of blows required to penetrate the sampler into the soil for 300 mm beyond seating drive of 150 mm. To find the liquefaction resistance of soil, this N value is corrected as follows [11]: (N1)60 = NmCnCeCbCrCs where values of various correction factors are given in Table 2.1. Figure 2.1 shows soil profiles obtained from SPT conducted at four sites near Roorkee [4, 6]. These SPT data are used either to evaluate liquefaction resistance of the site or shear wave velocity (using correlations as discussed later). For liquefaction 26 B. K. Maheshwari Table 2.1 Various correction factors applied to SPT values (after [11]) Factor Equipment variable Term Correction Overburden pressure Cn See Fig. 4.3 Energy ratio Safety hammer Ce 0.60–1.17 Donut hammer 0.45–1.00 Automatic trip 0.9–1.6 Hammer See Table 4.2 for details Borehole diameter 65–115 mm Cb 1.0 150 mm 1.05 200 mm 1.15 Rod length 3–4 m Cr 0.75 4–6 m 0.85 6–10 m 0.95 10–30 m 1.0 >30 m <1.0 Sampling method Standard sampler Cs 1.0 Sampler without liners 1.2 resistance, first (N1)60 is found and then corresponding Cyclic Resistance Ratio (CRR) is evaluated using empirical charts [11]. The liquefaction potential of these sites was found as presented by Muley [8] and Muley et al. [9]. 2.5.2 Cone Penetration Test (CPT) Cone Penetration Test (CPT) is an in situ test done to determine the soil properties and to get the soil stratigraphy. The cone havingan apex angle of 60° and an end contact area of 10 cm2 will be pushed through the ground at 2 cm/sec. However, like SPT no soil sample is collected in case of CPT. During the penetration of cone penetrometer through the ground surface, the forces on the cone tip (qc) and sleeve friction (f s) are measured. The Friction Ratio (f r = f s/qc), will vary with soil type and it is also an important parameter. CPT with Piezocone termed as CPTu can also measure the porewater pressure with the depth. Muley [8] examined the liquefaction potential of few sites in Roorkee region using both CPT and CPTu data. Also, the CPT and CPTu profiles can be used for determining soil behaviour type (Ic) and fines content. One such profile for Solani Riverbed site is shown in Fig. 2.2. 2 Disaster Management in India and Characterization for Geohazards 27 (a) Solani Riverbed Site (b) Bhagwanpur Site (c) Bahadrabad Site (d) Haridwar Site Fig. 2.1 Variation of penetration resistance (N) with depth for four sites near Roorkee (after [6]) 28 B. K. Maheshwari Fig. 2.2 Soil profile at Solani riverbed site, CPT; a normalized tip resistance (qc1N); b normalized friction ratio (F); c soil behaviour type (Ic) and d fines content (after [8]) 2.5.3 Multichannel Analysis of Surface Waves (MASW) The most widely used geophysical technique is MASW (Multichannel Analysis of Surface Waves) based on the seismic method that can be used for geotechnical characterization of near-surface materials. The MASW has been found to be an efficient method for unravelling the shallow subsurface properties. In particular, the MASW is used in geotechnical engineering for the measurement of shear wave velocity (V s) and dynamic properties, identification of subsurface material boundaries and spatial variations of V s. MASW generates a shear wave velocity (V s) profile (i.e. V s vs. depth) by analysing surface waves obtained from multichannel records for active or passive seismic sources. Figure 2.3 shows two-dimensional soil profiles at Bahadrabad and Haridwar site [9]. Based on MASW and SPT data in Roorkee region, Kirar et al. [5] proposed the following relationship. Vs = 99.5N 0.345 (2.1) For site classification, Eurocode-8 andNEHRP usedweighted average shear wave velocity for 30 m depth ( V 30s ) , for n number of layers of thickness di with shear wave velocity Vi. This is defined as: V 30s = 30 n∑ i=1 ( di / Vi ) (2.2) 2 Disaster Management in India and Characterization for Geohazards 29 Fig. 2.3 2D shear wave velocity (V s) profiles obtained from MASW tests (after [9]) 2.6 Laboratory Tests for Site Characterization Besides the index property tests, the most common laboratory tests for dynamic analysis are Cyclic Triaxial for High-Strain testing and Resonant Column for Low- Strain testing. Both are described in detail in the following subsections. 2.6.1 Cyclic Triaxial Tests Cyclic triaxial test is the most commonly used laboratory test for the evaluation of cyclic strength and strain-dependent dynamic properties of soils at high strain levels. In this test [6], the deviator stress is applied cyclically on soil specimens under stress- controlled condition. An automated triaxial testing system with five sensors (a load cell to monitor the axial load; an LVDT to measure the vertical displacements; and three transducers to detect the cell pressure, pore water pressure and volume change) was used in this study.Kirar andMaheshwari [6] presented cyclic triaxial tests carried out on Solani riverbed sand specimens of 50 mm diameter and 100 mm height, at the in situ relative density and tested at effective confining pressure equivalent to overburden pressure. Figure 2.4 shows modulus reduction and damping ratio curves for Solani riverbed site. It can be observed that there is an effect of depth on these curves and good agreement with literature. 30 B. K. Maheshwari Fig. 2.4 a Modulus reduction curves. b Damping ratio curves of Solani riverbed sand (after [6]) 2.6.2 Resonant Column Tests The resonant column test is another laboratory test used for measuring the low-strain dynamic properties of soils. The basic principle of the resonant column test is to vibrate a cylindrical soil specimen in a fundamental mode of vibration, in torsion or flexure. Once the fundamental mode is established, measurements of resonant frequency and amplitude of vibration are obtained. In the present research, a com- puterized resonant column apparatus was used. To evaluate dynamic soil properties 2 Disaster Management in India and Characterization for Geohazards 31 Fig. 2.5 aModulus reduction curves. bDamping ratio curves in low-strain range of Solani riverbed sand (after [7]) in low-strain range (0.000–0.07%), resonant column tests were carried out on the soil samples obtained from SPT [7]. Each sample was prepared with 50mm diameter and 100 mm height. All the samples were prepared at the in situ relative density and tested at an effective confining pressure equivalent to overburden pressure at that depth. Figure 2.5 shows modulus reduction and damping ratio curves for the Solani riverbed site in low-strain range. It can be observed that there is an effect of depth on these curves and good agreement with literature. 32 B. K. Maheshwari 2.7 Summary and Conclusions In this article, the policy of the Government of India for disaster management is presented. The “National Disaster Management Authority (NDMA)” looks after the disaster management in the country. One of its wing “National Disaster Response Force (NDRF)” takes care of relief and rescue operations during the disasters. However, the government’s policy is to increase capacity for prevention, mitigation and preparedness before the occurrence of disaster rather than simply react after it. The paper also highlighted the important methods of geotechnical site character- ization for disaster mitigation. This includes both field and laboratory tests. Some of the results from the author and his research group are presented. References 1. DMAct: The DisasterManagement Act, the Gazette of India (Extraordinary), No. 53. Ministry of Law and Justice, Government of India, New Delhi (2005) 2. IS: 1893 (Part 1): Criteria for earthquake resistant design of structures: general provisions and buildings. Bureau of Indian Standards, New Delhi, India (2016) 3. IS: 2131: Method for standard penetration test for soils. Bureau of Indian Standards, New Delhi, India (1981) 4. Kirar, B.: Dynamic Stiffness Characteristics of Unreinforced and Reinforced Sands. Ph.D. thesis, Department of Earthquake Engineering, IIT Roorkee, India (2016) 5. Kirar, B., Maheshwari, B.K., Muley P.: Correlation between shear wave velocity (Vs) and SPT resistance (N) for Roorkee region. Int. J. Geosynth. Ground Eng. 2(1), Article 9 (2016) 6. Kirar, B., Maheshwari, B.K: Dynamic properties of soils at large strains in Roorkee region using field and laboratory tests. Indian Geotech. J. 48(1), 125–141 (2018) 7. Maheshwari, B.K., Kirar, B.: Dynamic properties of soils at low strains in Roorkee region using resonant column tests. Int. J. Geotech. Eng. 13(5), 399–410 (2019) 8. Muley, P.: Assessment of liquefaction potential using in-situ and laboratory tests. Ph.D. thesis, Department of Earthquake Engineering, IIT Roorkee, India (2016) 9. Muley, P., Maheshwari, B.K., Paul, D.K.: Liquefaction potential of Roorkee region using field and laboratory Tests. Int. J. Geosynth. Ground Eng. 1(4), Article 37 (2015) 10. NDMA: National Policy on Disaster Management. National Disaster Management Authority (NDMA), Ministry of Home Affairs, Government of India, New Delhi (2009) 11. Youd, L., Idriss, I.M.: Summary Report, Proceedings of the NCEER Workshop on Evaluation of Liquefaction Resistance of Soils. Technical Report NCEER-97-0022, pp. 1–40 (1997) Chapter 3 Shear Strength Behaviour of Gravel–Tire Chips Mixture S. M. K. Pasha, H. Hazarika and N. Yoshimoto 3.1 Introduction It has been estimated that more than 1.5 billionscrap tires are generated in all over the world annually. In 2016, over 95 million scrap tires in quantity and 1 million tones in weight are generated and disposed in Japan. One common disposal practice around the world is to dump scrap tires in very large mono-fill stockpiles. This can pose serious threat to environment and public health such as increased breeding ground for disease carrying pests, insidious animal’s activity and combustion. Recycling is a definite option to reduce the amount of stockpiled scrap tires in landfills. Over 63% of total scrap tires are being used as fuel for energy production purposes in industries, while only 16% are reused as Tire-Derived Materials (TDM). Due to unique physical and mechanical properties of Scrap Tire Derived Materials (STDM) like low unit weight, low bulk density, high hydraulic conductivity, and high elastic deformability, this has recently found it is way into the civil engineering application with an advance growing interest each year. Retaining wall and bridge abutment backfills, fills for lightweight embankment and landfill liners are just examples of using STDM in civil engineering applications [3, 5]. The use of STDM only may undermine the serviceability of civil and geotechnical structures, because they are highly compressible. Researches in recent years focused on the use of sand and STDM mixtures. Outcome of some researches have shown addition of STDM (mostly tire shreds) leads to slightly improvement in shear strength of sand [7], however referring to others, it was found to reduce the shear strength of sand [6]. On the other hand, most of studies unanimously have reported that sand and STDM S. M. K. Pasha (B) · H. Hazarika Department of Civil Engineering, Kyushu University, Fukuoka 819-0395, Japan e-mail: manafi.siavash@gmail.com N. Yoshimoto Department of Civil and Environmental Engineering, Yamaguchi University, Ube-shi, Yamaguchi 755-8611, Japan © Springer Nature Singapore Pte Ltd. 2020 A. M Krishna and T. Katsumi (eds.), Geotechnics for Natural Disaster Mitigation and Management, Developments in Geotechnical Engineering, https://doi.org/10.1007/978-981-13-8828-6_3 33 http://crossmark.crossref.org/dialog/?doi=10.1007/978-981-13-8828-6_3&domain=pdf mailto:manafi.siavash@gmail.com https://doi.org/10.1007/978-981-13-8828-6_3 34 S. M. Khajeh Pasha et al. Fig. 3.1 Particle size distribution of Gravel and tire chips Gravel Tire chips mixtures experienced high deformation resulted in non-explicit peak point or failure at stress–deformation curves. STDM and sand–STDM mixture are often being used as protective layer due to low liquefaction potential and interesting damping property [1]. However, high compressibility and low elastic modulus of tire chips and tire chips and STDM mixture could result in high differential settlement and inadequate bearing capacity of foundation. So in order to overcome aforementioned issues, gravel–tire chips mixture (GTCM) as an alternative geomaterial has been introduced by Hazarika et al. [4], Hazarika and Abdullah [2]. Unfortunately, there is a lack of study on the physical and mechanical properties of GTCM. Therefore, this study aims to identify influential parameters affecting the shear strength and dilatancy behaviour of gravel tire chips mixture. This paper presents results of the study on the effect of confining pressure, relative density, and gravel fraction (VGravel/VT) on the mechanical behaviour and deformation characteristic of gravel and GTCM. 3.2 Material Properties and Testing Program Stress-controlled drained triaxial compression testswere conducted on the specimens of 100 mm in diameter by 200 mm in height. The particle size distribution of the gravel and tire chips used in this research is plotted in Fig. 3.1. The maximum grains size of TC and gravel were limited to less than 1/6 of specimen diameter to avoid the effect of sample size on the results of experiments. According to JGS 0131 (JIS A1204), gravel is classified as poorly graded (SP). Regarding to shape andmaximum grain size of STDM, they are classified as tire chips (TC). Specific gravity (Gs) of gravel and TCwere obtained 2.81 and 1.17 respectively (Fig. 3.2). In order to proceed with the preparation of specimens for CD triaxial tests at desired relative density, a series of vibratory test were conducted on Gravel and GTCM mixtures according to JGS-0161 standard . The following empirical expression has been proposed to estimate the maximum and minimum void ratio of GTCM: 3 Shear Strength Behaviour of Gravel–Tire Chips Mixture 35 Fig. 3.2 Specific gravity of GTCM particles Tire chips content ratio by weight (%) Sp ec ifi c Gr av ity o f G TC M Table 3.1 Fitting parameters for max. and min. void ratio Parameters A B C D Maximum void ratio 0.83 0.35 41.39 −0.05 Minimum void ratio 0.56 0.42 43.26 −0.03 Fig. 3.3 Maximum and minimum void ratio of GTCM Max Void ratio Min Void ratio emin,GTCM, emax,GTCM = A + B /( 1 + 10(C−(Gf%)×D ) , D50,Tc / D50,G ≈ 1.2, D50,Tc = 6mm (3.1) where emin,GTCM, emax,GTCM are minimum and maximum void ratio of GTCM for a given gravel fraction (Gf) in mixture. Where A, B, C,D are fitting parameters listed in Table 3.1. As it can be seen in Fig. 3.3, the value of void ratio decreases as the gravel fraction increases. Considering theory of packing of binary mixture, if large particles are of the same size as small particles, large particles will be replaced by small ones and gradual transition in void ratio without exhibiting any minimal value can be observed Reid et al. [8, 9]. Under-compaction method was used for the preparation of specimens. Gravel and tire chips were measured by mass corresponds to the desired volume of gravel in the mixture and afterward mixed carefully by hand and placed into the mould and sequentially compacted into 10 layers. The sample was saturated by allowing 36 S. M. Khajeh Pasha et al. de-aired water flow through from the bottom of the sample. 200 kPa back pressure was applied to the specimen in order to increase the degree of saturation (B > 0.95). An isotropic consolidation pressure was applied to the sample while maintain- ing initial backpressure constant. Specimen was undergone shearing with a constant axial strain rate of 0.1%/min until the axial strain of 20% was achieved. The effect of relative density, confining pressure and gravel fraction (Gf) in a mixture on shear strength and deformation characteristics of gravel and GTCM was investigated in this study. To this end, Gravel fraction by volume (Gf) in GTCM mixtures varies between 0 (pure tire chips) and 100% (pure gravel). Confining pressures are con- sidered to vary in the range of 50 and 200 kPa and three different relative density Dr = 35, 50 and 75% are taken into account for this study. 3.3 Results and Discussion 3.3.1 Effect of the Gravel Fraction in GTCM (Gf) As can be seen in Fig. 3.4 peak deviatoric stress decreases and the corresponding value of axial strain increases with a decrease in the percentage of gravel fraction (Gf) in GTCM. Three different behavioural zones of GTCM are clearly distinguishable in the figure. Gravel-like behaviour was observed for Gf > 83%, where essentially gravel particles forms GTCM soil matrix and stresses are mainly transmitted by gravel to gravel contact forces.Well defined peak deviatoric stress and highly dilative behaviour can be observed. Tire chips like behavior is evident for the specimens with Gf < 55%. Tire chips particles are dominant in GTCM, as a results force chains are mainly formed between tire chips particles. Therefore, mixture shows significant reduction in the shear strength and demonstrates properties which can be found in tire chips. For GTCM samples with 55% < Gf < 83%, very ductile behaviour was observed when the peak of the deviatoric stress–strain curve is reached. Slightly dilative behaviourwhich follows clear contractive behaviour can be seen in the figure. Except in the case of GTCM with Gf = 83%, no distinct peakshear strength has been observed for the GTCM samples with Gf < 83%. 3.3.2 Effect of the confining pressure ( σ́3 ) Test results demonstrated that the shear strength of GTCM with Gf = 83% has been remarkably enhanced by the effective confining pressure (see Fig. 3.5). GTCM exhibits completely contractive behaviour at high confining pressure (σ́3 = 200 kPa). At lower confining pressures, majority of gravel particles are still in contact and as a result, stresses transmitted mainly by gravel particle and is found to carry its properties that could be the reason for the observed phenomena. 3 Shear Strength Behaviour of Gravel–Tire Chips Mixture 37 Fig. 3.4 a Deviatoric stress–axial strain. b Volumetric strain–axial strain relation of GTCM at different GFs (Dr = 50%, σ́3 = 100 kPa) 3.3.3 Effect of the Relative Density (Dr) The rise in the degree of relative density (Dr) has almost no influence on the initial slope of stress–strain curve but slightly increases peak shear strength and intensifies dilatancy of GTCM specimens with Gf = 55% (see Fig. 3.6). Similar results have been reported by other researchers. 3.3.4 Tangent Modulus of GTCM The tangent modulus (Et) can be determined from stress–strain curve of the type shown in Fig. 3.7. As can be seen in Fig. 3.8, for the GTCM samples with GF > 83% tangent modulus (Et) decreases (to 25% of its initial value for axial strains over 6%) drastically with increasing axial strain and becomes negative beyond the peak deviator stress (softening) and eventually increases again to around zero as the residual strength is achieved (Gravel like behaviour). 38 S. M. Khajeh Pasha et al. Fig. 3.5 a Deviatoric stress–axial strain. b Volumetric strain–axial strain relation of GTCM at different confining pressures (Dr = 50%,Gf = 88%) 50 kN/ 100 kN/ 200 kN/ 50 kN/ 100 kN/ 200 kN/ GTCM like behavior (a) (b) For GTCM samples with 55% < Gf < 83%, Et decreases with axial strain until it reaches to an almost positive threshold value. However, for samples with Gf <=55%, no significant change in Et values has been observed (tire chips like behaviour). Another remark that can be concluded from the figure is that no softening was observed for samples with Gf < 70%. A slightly decreasing trend can be seen for the first few percentages of axial strain, in the case of pure tire chips (Gf = 0%). 3.4 Conclusion This paper presents the results of a series of triaxial test conducted on Gravel–Tire Chips Mixture (GTCM). A new relationship has been proposed to predict the maxi- mum andminimum void ratio of GTCM as a function of gravel fraction by volume in the mixture. The behaviour of GTCM has been found to be significantly influenced by the percentage of tire chips in mixture. 3 Shear Strength Behaviour of Gravel–Tire Chips Mixture 39 Fig. 3.6 a Deviatoric stress–axial strain. b Volumetric strain–axial strain relation of GTCM at different relative densities( Gf = 55%, σ́3 = 100 kPa ) Fig. 3.7 Definition of tangent modulus 40 S. M. Khajeh Pasha et al. Fig. 3.8 Variation of tangent modulus with axial strain. a 78% < Gf < 100% , b 0 < Gf < 70% ( ( Dr = 50%, σ́3 = 100 kPa ) Gf=100% Gf=87% Gf=83% Gf=78% Gf=70% Gf=55% Gf=44% Gf=0% (a) (b) Peak shear strength and tangent modulus of GTCM decreases as gravel fraction decrease in the mixture because tire chips particles have relatively low stiffness in comparison to that of gravel and do not contribute remarkably to improve shear stiffness ofmixture. Three different behavioural zones, gravel-like, gravel–tire chips- like and tire chips like behaviours have been observed. In the first zone, gravel particles form GTCM material matrix and in third zone tire chips form GTCM matrix and the second zone is established by a binary matrix where gravel and tire chips have contributed in forming GTCM material matrix. The effect of confining pressure on stress–strain and dilatancy behaviour of GTCM was also investigated. Shear strength increases with increasing confining pressure and contraction behaviour has been found to be intensified by the confining pressure. Shear strength ofGTCMhas been found to be enhanced by increasing relative density of GTCM up to a limited extent in comparison to that of the confining pressure. Acknowledgements The authors would like to express their sincere gratitude to the Bridgestone Corporation, Tokyo, Japan for their partial financial support. We would like to express our deep gratitude to Prof. Yukio Nakata of the geotechnical engineering laboratory of Yamaguchi University 3 Shear Strength Behaviour of Gravel–Tire Chips Mixture 41 for his support. A special thanks go to Mr. Kanta Matsui of the geotechnical engineering laboratory of Yamaguchi University for his help and support while conducting experiments. References 1. Bahadori, H., Manafi, S.: Effect of tyre chips on dynamic properties of saturated sands. Int. J. Phys. Model. Geotech. 15(3), 116–128 (2015) 2. Hazarika, H., Abdullah, A.: Improvement effects of two and three dimensional geosynthetics used in liquefaction countermeasures. Jpn. Geotech. Soc. Spec. Publ. 2(68), 2336–2341 (2016) 3. Hazarika, H., Kohama, E., Sugano, T.: Underwater shake table tests on waterfront structures protected with tire chips cushion. J. Geotech. Geoenviron. Eng. 134(12), 1706–1719 (2008) 4. Hazarika, H., Yasuhara, K., Kikuchi, Y., Karmokar, A.K., Mitarai, Y.: Multifaceted potentials of tire-derived three dimensional geosynthetics in geotechnical applications and their evaluation. Geotext. Geomembr. 28(3), 303–315 (2010) 5. Kaushik, M.K., Kumar, A., Bansal, A.: Performance of tire chips–gravel combinations with nonwoven geotextile and encapsulated tire chips layers used as filter/separator under incremental stress levels. Eur. J. Environ. Civ. Eng., 1–34 (2017) 6. Kawata, S., Hyodo, M., Orense, P., Yamada, S., Hazarika, H.: Undrained and drained shear behavior of sand and tire chips composite material. In: Proceedings of the International Work- shop on Scrap tire Derived Geomaterials—Opportunities and Challenges, Yokosuka, Japan, pp. 277–283 (2007) 7. Mashiri, M.S., Vinod, J.S., Sheikh, M.N., Tsang, H.-H.: Shear strength and dilatancy behaviour of sand–tyre chip mixtures. Soils Found. 55(3), 517–528 (2015) 8. Reid, R.A., Soupir, S.P., Schaefer, V.R.: Mitigation of void development under bridge approach slabs using rubber tire chips. Recycl. Mater. Geotech. Appl. ASCE, 37–50 (1998) 9. Pasha, S.M.K., Hazarika, H., Yoshimoto, N.: Dynamic properties and liquefaction potential of gravel-tire chips mixture (GTCM), J. Jpn. Soc. Civ Eng, Ser. A1 Struct Eng & Earthq Eng (SE/EE), 74(4), I_649-I_655 (2018) Chapter 4 Elastic Modulus Estimation Using a Scaled State Parameter in the Extended Kalman Filter M. C. Koch, A. Murakami and K. Fujisawa 4.1 Introduction Kalman Filtering is a Bayesian approach and one of the most popular methods for state estimation in linear systems [1]. It is a powerful tool that combines the information in the presence of uncertainty. The Kalman Filter can be applied to the State-Space Model, that contains a system equation and an observation equation. The state covariance and observation noise (0, R) are input parameters to the Kalman Filter. Initially, the Kalman Filter makes an “apriori” estimate to the next step based on the system equation. Then the state is updated (“aposteriori” estimate) using the Gain that is calculated based on a weighting of the state covariance and observation noise variance. The Extended Kalman Filter is a natural extension of the Kalman Filter for nonlinear systems. It is based on a linearization along the state estimates resulting from the observations. TheKalman Filter and associated techniques like the Extended Kalman Filter (EKF) and Particle Filter have been vastly used to estimate a wide range of geotechnical parameters [2, 3]. In this paper, an attempt is made to estimate the Elastic Modulus (E) of soil in the presence of heterogeneities. The heterogeneous element in thesoil may be a cavity or a material of different Elastic Modulus embedded in the otherwise homogenous soil. The elastic wave propagation is simulated using the Finite ElementMethod. The unconditionally stable Newmark β method is used for evaluating the Finite Element equations in the time domain. The accelerations recorded at predetermined points are used as observation data. As the observation equation is nonlinear, hence, the EKF is used for state estimation. A simple case is considered wherein the location of the heterogeneity is known but the Elastic Modulus (E) is unknown. The EKF is then used to try and estimate the Elastic Modulus of the homogenous soil as well as the heterogeneity. The EKF is known to perform poorly in the caseswhere the initial state M. C. Koch (B) · A. Murakami · K. Fujisawa Graduate School of Agriculture, Kyoto University, Sakyo-ku, Kyoto, Japan e-mail: michaelkoch044@gmail.com © Springer Nature Singapore Pte Ltd. 2020 A. M Krishna and T. Katsumi (eds.), Geotechnics for Natural Disaster Mitigation and Management, Developments in Geotechnical Engineering, https://doi.org/10.1007/978-981-13-8828-6_4 43 http://crossmark.crossref.org/dialog/?doi=10.1007/978-981-13-8828-6_4&domain=pdf mailto:michaelkoch044@gmail.com https://doi.org/10.1007/978-981-13-8828-6_4 44 M. C. Koch et al. covariance is very large with respect to observation noise covariance [4, 5]. The large difference in covariance’s, makes the filter pay too much attention to the observation data which can cause numerical divergence. In practice the Elastic Modulus of soil could range between 106 and 109 Pa, hence the initial state covariance would be in this range. The accelerations recorded would typically be magnitudes lower and so would the associated noise. Citing the above problem, a simple scaled formulation is proposed that considers a scaled parameter as the state rather than the actual Elastic Modulus. The formulation is presented and the effectiveness of this formulation is investigated in relation to the given problem. 4.2 Extended Kalman Filter 4.2.1 Linearization of the System and Observation Equations A state-space description for a nonlinear system would be of the form shown below [6]. The system equation is given as xt+1 = ft(xt) + ξt (4.1) and the observation equation is zt = gt(xt) + ηt (4.2) where xt , zt are the true state vector and observation vector at time t, respectively. ft and gt are nonlinear functions of the state vector. ξt is the white noise contribution to the state vector while ηt is the error in the observation vector which is also assumed to be a Gaussian white noise. The covariance associated with ξt and ηt are given by E ( ξtξ T p ) = { Qt, t = p 0, t �= p (4.3) E ( ηtη T p ) = { Rt, t = p 0, t �= p (4.4) E ( ξkη T l ) = 0 for all t and p (4.5) Let the state estimate at time t be x ∧ t then the predicted state at t + 1 is given as x ∧ t+1|t = f t ( x ∧ t ) (4.6) The covariance associated with states x ∧ t|t−1 and x ∧ t|t are shown in Eq. (4.7) and Eq. (4.8), respectively: 4 Elastic Modulus Estimation Using a Scaled State Parameter … 45 Pt,t−1 = E [( xt − x ∧ t|t−1 )( xt − x ∧ t|t−1 )T] (4.7) Pt,t = E [( xt − x ∧ t )( xt − x ∧ t )T] (4.8) The state and observation functions ft and gt are linearized by considering a linear Taylor series approximation. While the function ft is linearized at time x ∧ t|t , gt is linearized at x ∧ t|t−1 . f t(xt) � f t ( x ∧ t ) + �t ( xt − x ∧ t ) (4.9) gt(xt) � gt ( x ∧ t|t−1 ) + Ht ( xt − x ∧ t|t−1 ) (4.10) where �t = ∂f t ∂xt ( x ∧ t ) and Ht = ∂gt ∂xt ( x ∧ t|t−1 ) (4.11) Equations (4.9) and (4.10) can be substituted back into Eqs. (4.1) and (4.2) to get xt+1 = �txt + ξ t + f t ( x ∧ t ) − �tx ∧ t (4.12) zt = Htxt + ηt + gt ( x ∧ t|t−1 ) − Htx ∧ t|t−1 (4.13) Setting ut = ft ( x ∧ t ) − Φtx∧t yt = zt − gt ( x ∧ t|t−1 ) + Htx ∧ t|t−1 (4.14) we get the linearized state and observation equations as shown in (4.15) and (4.16) xt+1 = �txt + ut + ξ t (4.15) yt = Htxt + ηt (4.16) Using the usual formula for the Kalman filter on this linearized system, to obtain the corrected state from the predicted state (x ∧ t|t−1 ), we get x ∧ t = x ∧ t|t−1 + Gt ( yt − Htx ∧ t|t−1 ) (4.17) where Gt is the Gain calculated at time t. On substituting wt from (4.14) in (4.17), we get x ∧ t = x ∧ t|t−1 + Gt ( zt − gt ( x ∧ t|t−1 )) (4.18) 46 M. C. Koch et al. 4.2.2 Algorithm The complete EKF algorithm is as follows: 1. Pt,t−1 = Var(x0), x ∧ 0 = E(x0) 2. For = 1, 2, . . . 3. Pt,t−1 = [�t−1 ( x ∧ t−1 )]Pt−1,t−1 [ �t−1 ( x ∧ t−1 )]T + Qt 4. x ∧ t|t−1 = ft−1 ( x ∧ t−1 ) 5. Gt = Pt,t−1 [ Ht ( x ∧ t|t−1 )]T[[ Ht ( x ∧ t|t−1 )] Pt,t−1 [ Ht ( x ∧ t|t−1 )]T + Rt ]−1 6. Pt,t = [ I − Gt [ Ht ( x ∧ t|t−1 )]] Pt,t−1 7. x ∧ t = x ∧ t|t−1 + Gt ( zt − gt ( x ∧ t|t−1 )) (4.19) 4.3 Formulation 4.3.1 Finite Element Method The Galerkin Finite Element Method was used to simulate the wave propagation based on the following discretized equation: Müt + K ( Ei ) ut = Ft (4.20) whereM is the Mass Matrix, K ( Ei ) is the stiffness matrix. Ei is the Elastic Modulus of all the elements in the domain and F(t) is the Force applied at time t. u(t) and ü(t) are the displacement and acceleration vector for all the nodes. No damping was considered. 4.3.2 Derivation of the H Matrix The dynamic analysis of the system was done using the Newmark β Method (γ = 0.5, β = 0.25) Newmark [7]. The equations of the Newmark Beta Method are as follows: ut+1 = ut + �tu̇t + [( 1 2 − β ) �t2 ] üt + β�t2üt+1 (4.21) u̇t+1 = u̇t + [( 1 2 − γ ) �t ] üt + γ�t2üt+1 (4.22) 4 Elastic Modulus Estimation Using a Scaled State Parameter … 47 Substituting Eqs. (4.21) and (4.22) in (4.20) and rearranging the terms, we get üt+1 ( Ei ) = S(Ei)−1[Ft+1 − T ( Ei ) üt − V ( Ei ) u̇t − K ( Ei ) ut ] (4.23) where, S ( Ei ) = M + K(Ei)�t 2 4 ,T ( Ei ) = K(Ei)�t 2 4 and V ( Ei ) = K(Ei)�t (4.24) As it can be clearly seen S, V, and T are all functions of the Elastic Modulus. More specifically, it can be seen that the dependence of üt+�t on E is nonlinear. In order to use the EKF, it is necessary to linearize this equation. This is done as follows [8]: Hit+1 = ∂üt+1 ∂Ei = −S−1 ∂S ∂Ei S−1Ft+1 − [ −S−1 ∂S ∂Ei S−1Tüt + S−1 ∂T ∂Ei üt ] − [ −S−1 ∂S ∂Ei S−1Vu̇t + S−1 ∂V ∂Ei u̇t ] − [ −S−1 ∂S ∂Ei S−1Kut + S−1 ∂K ∂Ei ut ] (4.25) Hit+1 = ∂üt+1 ∂Ei = −S−1 ∂S ∂Ei S−1 [ Ft+1 − Tüt − Vu̇t − Ku̇t ] − S−1 [ ∂T ∂Ei üt + ∂V ∂Ei u̇t + ∂K ∂Ei ut ] (4.26) H = Hit+1 = ∂üt+1 ∂Ei = −S−1 [ ∂S ∂Ei üt+1 + ∂T ∂Ei üt + ∂V ∂Ei u̇t + ∂K ∂Ei ut ] (4.27) 4.3.3 Setting Blocks and Scaling the State Vector The domain is set up in such a way that a finite number of elements are grouped together to form a block, i.e., all the elements in the same block are forced to have the same Elastic Modulus. If the domain discretized with n elements is divided into N blocks (N ≤ n) then the new state vector has only N dimensions. Let this vector be called W. A simple scaling operation is executed by simply considering the state vector E = [E1,E2,E3, · · · ,En]T = E′[w1,w1,w2, · · · ,wN]T i.e. Ei = E′wj ∀i ∈ block j (4.28) where wj is the scaled state vector representing the Elastic Modulus of block j to which element i belongs. The scaling parameter E′ is kept constant during the filtering process. The lin- earization of the observation equation now needs to be done with respect to the new variable. This is done as follows: 48 M. C. Koch et al. ht+1 = ∂üt+1 ∂W = ∂üt+1 ∂E ∂E ∂W = HJ (4.29) where, J = ∂E i ∂wj = { E′ if i ∈ blockj 0 otherwise (4.30) 4.3.4 The EKF Formulation The system equation is given as Wt+1 = IWt (4.31) while the Observation Equation is given as yt = htW t + ηt (4.32) where based on Eq. (4.14) yt = üt − üt|t−1 + htw ∧ t|t−1 (4.33) The Extended Kalman filtering process can then be started by substituted by placing
Compartilhar