VIETNAM NATIONAL UNIVERSITY, HANOI VIETNAM JAPAN UNIVERSITY CHO THU THU NAING ROCKFALL MECHANISM AND COUNTERMEASURE FOR HOANG SA ROAD IN DA NANG CITY MASTER’S THESIS VIETNAM NATIONAL UNIVERSITY, HANOI VIETNAM JAPAN UNIVERSITY CHO THU THU NAING ROCKFALL MECHANISM AND COUNTERMEASURE FOR HOANG SA MAJOR: INFRASTRUCTURE ENGINEERING CODE: 8900201.04 QTD RESEARCH SUPERVISOR: Dr NGUYEN CHAU LAN ROAD IN DA NANG CITY Hanoi, 2023 ABSTRACT The phenomenon of rockfall is a critical and dangerous case along the highway, especially in mountainous areas, that affects the traveler, vehicle, building, and other structures As the infrastructure develops, more rockfall cases happen on the mountainous road For this reason, modeling programs have become necessary for the decision-making process, such as protection and prevention against rockfalls, by simulating the behavior of rock failures, the stability of the slope, and remedial measures The thesis presents a study on rockfall behavior and mitigating the impacts of rockfall on the Hoang Sa road to the Son Tra Peninsula in Da Nang city This study area has recorded rock collapse failure problems several times in the past Especially significant rockfall events occurred three or four times along this highway last year The retaining wall was damaged, and the tourism industry was delayed on the Hoang Sa road in Da Nang City due to the rockfall events The aim of the study is to solve the rockfall mitigation problem by selecting suitable remedial measures for preventing highway rockfalls on the Hoang Sa road The study uses two types of numerical software: RocFall 2D and Plaxis 2D RocFall software calculates rock trajectories, rock end-points, kinetic energy, velocity, and bounce height to examine the behavior of the rockfall process Additionally, it determines a reliable remedial design in the passive method for the mitigation of rock slope failures On the other hand, Plaxis software calculates the prediction of the failure surface and factor of safety for slope stability According to the field results, unstable blocks on the steep slope, highly fractured and cracked rock masses, as well as weathered rocks on the surface of granite rock, are factors that caused rockfall in the study area The numerical model simulated the rockfall mechanism and rock slope stability analysis to find the suitable method for rockfall mitigation based on the m3 of rock volume (8100 kg of rock mass) that fell from the steep granite slope surface Based on each simulation result, the study concludes that the flexible barrier method and the anchored mesh method are the most feasible solutions for preventing future rockfall events on the Hoang road ACKNOWLEDGEMENTS My most sincere appreciation goes to my supervisor, Dr Nguyen Chau Lan from the University of Transport and Communication I am greatly indebted to him for his valuable advice and incredible support throughout my thesis period I would like to say thank you to Prof Nguyen Dinh Duc (MCE Director) Moreover, my special thanks go to Prof Hironori Kato (MCE co-director), Dr Nguyen Tien Dung (MCE coordinator), Assoc Prof Takeda Shinichi (MCE JICA expert), and Dr Nguyen Ngoc Vinh (MCE lecturer) for their kind support, guidance, and recommendations during the lecture time and research period I am grateful to the members of the JAIF scholarship organization as well as the rector of Vietnam-Japan University for providing me with the opportunity to gain valuable knowledge I am thankful to the staff of the Master's Program in Civil Engineering for their support not only with course work via online learning but also with my thesis work Additionally, a very special thanks to the staff from (UTCGeo) for data collection and guidance on the thesis preparation I am also thankful to my colleagues from MCE for their ideas and encouragement to successfully complete this work I would like to extend further thanks to all individuals for their direct and indirect help Finally, I am grateful for my family’s continued support for my graduate studies TABLE OF CONTENTS LIST OF TABLES i LIST OF FIGURES ii LIST OF ABBREVIATIONS iv CHAPTER INTRODUCTION 1.1 Background of the study 1.2 Location of the study area 1.3 Statement of the problem 1.4 Objectives 1.5 Scope of the study 1.6 The layout of the thesis 1.6.1 Outline of the thesis 1.6.2 Structure of the thesis 1.7 Thesis contribution CHAPTER LITERATURE REVIEW 2.1 Definition of rockfall 2.2 Rockfall process 2.3 Causal mechanism of rockfall 10 2.3.1 Rock mass characteristics 11 2.3.2 Triggering mechanisms 12 2.4 Computer programs to simulate rockfall 12 2.4.1 RocFall software 14 2.4.2 RocFall modelling 15 2.5 Rockfall mitigation methods and selection 19 2.5.1 Classification of countermeasures 20 2.5.2 Rock slope stabilization method 21 2.5.3 Rockfall protection method 24 2.6 Case studies of rockfall events and mitigation measure 28 2.7 Overview of the study area 33 2.7.1 Past rockfall events in the study area 36 CHAPTER METHODOLOGY 40 3.1 Introduction 40 3.2 Data Collection 42 3.2.1 Collection with field investigation 42 3.2.2 Collection with UAV, DEM model and GIS 43 3.3 Rockfall mechanisms analysis based on RocFall software 45 3.3.1 Rockfall mechanisms analysis without proposed method 46 3.3.2 Rockfall mechanisms analysis with proposed method 49 3.4 Rock slope stability analysis based on Plaxis software 51 3.4.1 Slope stability analysis without proposed method 54 3.4.2 Slope stability analysis with proposed methods 55 CHAPTER DATA ANALYSIS RESULTS AND DISCUSSIONS 59 4.1 Field data results 59 4.1.1 Field investigation results 59 4.1.2 UAV, DEM and GIS results 61 4.2 Rockfall simulation results based on RocFall software 63 4.2.1 Simulation results of rockfall mechanisms without proposed method 63 4.2.2 Simulation results of rockfall mechanisms with proposed method 68 4.3 Rock slope stability simulation results based on Plaxis software 73 4.3.1 Simulation results of slope stability without proposed method 73 4.3.2 Simulation results of slope stability with proposed methods 75 4.4 Discussions 77 CHAPTER CONCLUSIONS AND RECOMMENDATIONS 81 5.1 Conclusions 82 5.2 Limitations 84 5.3 Recommendations 84 RFERENCES 85 APPENDIX 87 LIST OF TABLES Table 2.1 Illustration of motion mechanism of falling rocks 10 Table 2.2 Classification of countermeasures for rock falls 20 Table 2.3 Overview of stabilization procedures 21 Table 2.4 Overview of protection measures 24 Table 2.5 Geologic classification .34 Table 2.6 Annual and maximum rainfall data of Da Nang City 35 Table 3.1 Input materials for RocFall 48 Table 3.2 Input materials for initial condition of stone .48 Table 3.3 Input parameters for granite .52 Table 3.4 Input parameters for granite calculated in the RocLab program 53 Table 3.5 Input parameters for rainfall 54 Table 3.6 Input parameters for retaining wall .54 Table 3.7 Parameters of retaining components 55 Table 3.8 Input parameters for shotcrete .57 Table 3.9 Input parameters for rock bolt .57 Table 4.1 Rockfall mechanisms analysis result with varied rock mass .66 Table 4.2 Proposed barrier design (option 1) 69 Table 4.3 Proposed barrier design (option 2) 69 Table 4.4 Maximum total kinetic energy results without and with barrier 79 i LIST OF FIGURES Figure 1.1 Rockfall on the road connecting with the Ho Sau area in 2017 (Tuong Quan, 2021) .2 Figure 1.2 Location map of study area Figure 1.3 Rockfall happened on the Hoang Sa road (a) April 2022; (b) May 2022; (c) August 2022 .4 Figure 1.4 Past countermeasure design and current condition Figure 2.1 A schematic slope profile of the rockfall process (Dorren, 2003) .9 Figure 2.2 Motion pattern of falling rocks (N C Koei, 2007) 10 Figure 2.3 Schematic of rock slope collapse (N Koei, 2007) .12 Figure 2.4 Installation of structural shotcrete (Richard Andrew et al., 2011) .23 Figure 2.5 Installation of an anchored Tecco mesh system in Northern California (Richard Andrew et al., 2011) .27 Figure 2.6 Flexible barrier system (Qi et al., 2018) 28 Figure 2.7 Rockfall event in Badouzih, Keelung (Wei et al., 2014) 29 Figure 2.8 Rockfall event in Badouzih, Keelung (Hancock, 2019) 29 Figure 2.9 Rockfall on Sea to Sky highway (B.C.) (Volkwein et al., 2011) .30 Figure 2.10 Rockfall case in Lai Chau province (News, 2018) 30 Figure 2.11 Rockfall case in Binh Dinh province (News, 2021) 31 Figure 2.12 Hoang Sa road, Son Tra Peninsula (taken by using UAV investigation) 34 Figure 2.13 Rockfalls alongside a section of the road connecting with the Ho Sau area (NGUYEN DUC NAM, 2018) 36 Figure 2.14 Rockfalls along a section of a road leading to the Ho Sau (Deep Hole) area (Landslide Warning Along Roads Leading to Son Tra Peninsula, 2020) .37 Figure 2.15 Rock rolled down onto the Hoang Sa road after a heavy rain in 2022 (Early Handling of Rockfalls on Son Tra Peninsula Taken, 2022.) 38 Figure 2.16 After rockfall conditions (VnExpress, 2022) 39 Figure 3.1 Flow chart of overview of the study 40 Figure 3.2 Flow chart of the analysis study (1) 41 Figure 3.3 Flow chart of the analysis study (2) 42 Figure 3.4 (a) UAV equipment; (b) Example of workflow in Agisoft Metashape to generate a DEM (González-Quiñones et al., 2022) .44 Figure 3.5 Slope profile taken from google earth 45 Figure 3.6 Simulation model procedures for rockfall mechanisms analysis .46 Figure 3.7 Model geometry without barrier method in RocFall for Son Tra area 47 Figure 3.8 The procedure for choosing the fallen stone location 47 Figure 3.9 Range of energy capacities for a variety of passive protection methods 49 Figure 3.10 Slope geometry with two proposed barrier options in RocFall software for Son Tra area .50 Figure 3.11 The procedure for specifying the height and capacity of the barrier .50 ii Figure 3.12 Simulation model procedures for rock slope stability analysis .51 Figure 3.13 Monthly rainfall data in 2022 54 Figure 3.14 The model without proposed method for Plaxis (a) Ground water table; (b) Ground water and rainfall; (c) High water table; (d) High water table and rainfall 54 Figure 3.15 The model with shotcrete method for Plaxis (a) Ground water table; (b) Ground water and rainfall; (c) High water table; (d) High water table and rainfall .56 Figure 3.16 The model with anchored mesh method for Plaxis (a) Ground water table; (b) Ground water and rainfall; (c) High water table; (d) High water table and rainfall 58 Figure 4.1 Survey result photos (May, 2022) 60 Figure 4.2 Survey result photos (October, 2022) 60 Figure 4.3 (a) DEM model from UAV pictures; (b) Imaged of rocks with a high-risk of rockfall taken by UAV; (c) Serious rockfall location point taken from google earth .62 Figure 4.4 Overview of contour map in Son Tra Penninsula 63 Figure 4.5 Rock trajectories and end point results (a) Position of stone (1); (b) Position of stone (2); (c) Position of stone (3); (d) Position of stone (4); (e) Position of stone (5) 64 Figure 4.6 Bounce height results with different fallen rock location 65 Figure 4.7 Rockfall mechanisms analysis result without barrier method 66 Figure 4.8 Maximum total kinetic energy with varied rock mass .67 Figure 4.9 Bounce height results 68 Figure 4.10 Rock mechanism analysis results of barrier option with 2000 kJ (a) m barrier height; (b) m barrier height; (c) m barrier height; (d) m barrier height 70 Figure 4.11 Rock mechanism analysis results of barrier option with 3000 kJ (a) m barrier height; (b) m barrier height; (c) m barrier height; (d) m barrier height 71 Figure 4.12 Rock mechanism analysis results of barrier option with 2000 kJ (a) m barrier height; (b) m barrier height; (c) m barrier height; (d) m barrier height 72 Figure 4.13 Rock mechanism analysis results of barrier option with 3000 kJ (a) m barrier height; (b) m barrier height; (c) m barrier height; (d) m barrier height 73 Figure 4.14 Failure surfaces corresponding to FOS values for existing condition .74 Figure 4.15 Failure surfaces corresponding to FOS values under used shotcrete method 76 Figure 4.16 Failure surfaces corresponding to FOS values under used anchored mesh method 77 Figure 4.17 Comparison results (a) Barrier option 1; (b) Barrier option 79 Figure 4.18 Factor of safety results without or with reinforcement methods .81 iii LIST OF ABBREVIATIONS COR Coefficient of restitution D Disturbance factor DSM Digital Surface Model DEM Digital Elevation Model E Young’s modulus E Kinetic energy of falling rocks Ev Linear velocity energy of falling rocks Er Rolling energy of falling rocks EA1 Elastic axial stiffness EI Elastic bending stiffness FOS Factor of safety GSI Geological strength index GIS Geographic Information System H Falling height LMA Lumped mass approach m Mass unit of falling rocks mi Intact rock parameter Msf Strength reduction factor RBA Rigid body approach Rn Normal coefficient of restitution Rt Tangential coefficient of restitution RMR Rock mass rating Pmax The maximum impact force UAV Unmanned aerial vehicle w The unit weigh of the plate iv CHAPTER INTRODUCTION 1.1 Background of the study Rockfall is one of the most common geohazard problems and the fastest type of landslide that happens mainly in mountainous regions Compared to other slope instability failures, rockfall events happen more frequently It is very dangerous for people who travel on the road at the toe of the steep terrain Rockfalls with high energies and velocities can cause loss of life and significant damage to vehicles and infrastructure that are near or within the area of rockfall movement Rock slope failure cases occur not only in developed countries but also in developing countries Most countries around the world have experienced rockfall hazards along the road in hilly regions For example, rock boulders fell on a vehicle, killed a woman and disabled her father, and delayed traffic on British Columbia Highway 99 in 1982 (MM95013, 1994) A major rockfall (5000 m3) killed two tourists on the main highway crossing the Alps through the Gotthard Tunnel in Switzerland on May 31, 2006 (Liniger & Bieri, 2006) Another example is the rockfall that took place in northern Taiwan due to Typhoon Kong-Rey on August 31, 2013, which damaged infrastructure within this area (N Koei, 2007) In 2007, there were many rock slope collapse problems on the road, such as on the Kennon Road, Marcos Highway, Lagawe-Banaue Road, Dalton Pass, Cebu-Balamban Transcentral Highway, and Wright-Taft Road They occurred mainly on high-cut slopes of highly fractured and jointed hard rocks (N Koei, 2007) The above cases are a few examples of serious rockfall hazard problems around the world Moreover, with the development of infrastructure in Vietnam, rock slope failures are also frequently encountered along roads No 1A, 12, 4C, 4D, 4E, 16, etc in Vietnam (Duong-Ngo et al., 2020) In 2018, rockfalls occurred along the Ha Long-Van Don highway in the northern area of Vietnam, and concrete barriers were damaged (DuongNgo et al., 2020) Last year, on October 25, 2021, a large stone block from Ba Hoa mountain suddenly fell on Nguyen Tat Thanh street in area 5, Le Hong Phong ward, Quy Nhon city, Binh Dinh province, as illustrated in Figure 1.1 Three people who were walking on the road were killed due to the unexpected rock detached from the slope surface, including one person who was seriously injured (Tuong Quan, 2021) In generally, the causes of rock slope failures depend on the following factors: rainfall, natural hazards, seepage, weak rock properties, the free-thaw process, and soil erosion The falling of large blocks of rock, which injure people and delay highway traffic flow, is the biggest issue, especially in mountainous terrain It is crucial to perform protection or mitigation measures in rockfall-prone areas Thus, understanding the rockfall mechanisms and predicting the rockfall behavior is first required to examine the suitable remedial measures and design for preventing the rockfall problems along the highway Many researchers have carried out rockfall studies in various fields, such as remediation and mitigation strategies, rock failure characterization, and the development of a rockfall hazard rating system (Jaccard et al., 2020) Figure 1.1 Rockfall on the road connecting with the Ho Sau area in 2017 (Tuong Quan, 2021) 1.2 Location of the study area Da Nang has a 92-kilometer coastline and is one of the gateways to the Eastern Sea of the East-West Economic Corridor, linking Myanmar, Thailand, Laos, and Vietnam Da Nang is the biggest city in the central region and the midpoint between Hanoi and Ho Chi Minh City It has six urban districts (Hai Chau, Thanh Khe, Lien Chieu, Cam Le, Son Tra, and Ngu Hanh Son), one rural district (Hoa Vang), and the island district of Hoang Sa The Son Tra Peninsula is located to the east of Da Nang City, in the middle of central Vietnam It is a well-known place in Vietnam for tourists, with beautiful panoramas of majestic mountains and seas in downtown Da Nang The study area is located on the Hoang Sa road on the Son Tra Peninsula, which is the main route leading to Linh Ung pagoda Hoang Sa is the famous tourist highway of the Son Tra Peninsula, near Da Nang City, which passes through complex mountainous terrain Figure 1.2 illustrates the location of the study area Hoang Sa road Figure 1.2 Location map of study area 1.3 Statement of the problem There are many rockfall events along the Hoang Sa road almost every year In 2012, due to the rock-falling process, damaged infrastructure, threatened travelers, moving vehicles on this road, and delayed traffic were more serious cases than in other years Hence, Da Nang's government solved the rock collapse case along the Hoang Sa road by building a reinforced retaining wall at the toe of the slope as a prevention countermeasure Last year, the most significant and serious rockfall event occurred again in Da Nang City During last year 2022, slope failures total cases were 33 cases that included rockfalls, landslide and debris flow happened three or four times along the Hoang Sa road The phenomenon of rolling stones and areas at risk of landslides were observed on the Hoang Sa route in April, May, and August 2022, as shown in Figure 1.3 (a) (b) (c) Figure 1.3 Rockfall happened on the Hoang Sa road (a) April 2022; (b) May 2022; (c) August 2022 Figure 1.4 Past countermeasure design and current condition Even though the study area had already prevented rock slope failures with the retaining wall method, large blocks of rocks had detached from the surface of the cliff and damaged road infrastructure as well as impacted traffic flow on the entire route in 2022 Until now, this countermeasure has not prevented rock slope failure problems The photos of damaged existing retaining structures from the site survey are illustrated in Figure 1.4 As can be clearly seen in Figure 1.4, this countermeasure was an ineffective method for protecting against highway rockfall because the existing retaining wall height on the actual site was not enough to prevent fragmentation of rocks detached from the rock surface Beside this, many places had slid strongly due to the strong weathering of rock blocks on a high slope Thus, these locations need to be handled with an appropriate protection method The local government and organizations of Da Nang face big challenges in the decision-making process, including choosing the most effective method of handling rockfall events along the Hoang Sa road In addition, this region has suffered many rockfall events in the past few decades, but only a limited rockfall study has been carried out in this region The main issue for the study area is how to mitigate rockfall events and how to choose the most effective method to protect and prevent future rockfall cases 1.4 Objectives The main objective of this study is to determine a suitable prevention method for rockfall mitigation on the Hoang Sa road in Da Nang City The specific objectives have been listed below: - To study rockfall mechanism in high-risk rockfall areas on the Hoang Sa road - To propose a suitable mitigation method for rockfall failures on the Hoang Sa road 1.5 Scope of the study - Conduct geology, geomorphology, and hydrogeological survey research in the study area - Evaluate the rockfall mechanism and slope stability under the existing conditions - Analyze the rockfall mechanism with the proposed method by using RocFall and slope stability with the proposed countermeasure by using Plaxis software - Recommend suitable countermeasures for rockfall mitigation based on each numerical simulation result 1.6 The layout of the thesis 1.6.1 Outline of the thesis This thesis study mainly deals with rockfall mechanisms and the selection of suitable countermeasures for rockfall mitigation in the research area Firstly, the site assessment is performed through desk studies and field investigations The study analyzes the rock failure mechanism with RocFall 2D (v.4.0) software and rock slope stability with Plaxis 2D (V21.01) software The software simulates rock end-point, bounce height, energy, and velocity by using the RocFall 2D (v.4.0) software under existing condition and under the proposed method condition Plaxis 2D (V21.01) software is utilized to evaluate the slope stability with and without countermeasures Finally, the research recommends suitable countermeasures for rockfall mitigation based on the numerical analysis results 1.6.2 Structure of the thesis There are five chapters in the thesis format, along with an appendix included at the end of the thesis The structure of the thesis was organized as follows: Chapter consists of an introduction, which includes background, the study area, the problems, and the objective and scope of the research Chapter is a review of the literature on causes of rockfall, mitigation measuresrelated topics, modeling software, and rockfall case studies inside and outside Vietnam Chapter explicitly describes the methodology and process regarding the thesis progress Chapter describes rockfall mechanisms and slope stability analysis, as well as mitigation analysis and their results Chapter provides conclusions, limitations, and recommendations 1.7 Thesis contribution The study focuses on finding a mitigation solution to reduce losses and damages from rock falls along the highway The aim of the study is to solve the rockfall mitigation problem by selecting suitable remedial measures for preventing highway rockfalls on the Hoang Sa road Hence, the study can contribute to provide an in-depth knowledge base for researchers involved in the rockfall remediation and prevention of infrastructure projects dealing with the fields of geotechnical or environmental engineering, and risk disaster Therefore, this thesis would be more valuable for generating mitigation measures for rockfalls as a result of the development of infrastructure on the road along the mountain Moreover, it also gives guidelines for selecting proper mitigation measures with a proper safety factor to prevent rockfalls CHAPTER LITERATURE REVIEW This chapter presents the definition and process of rockfalls and the causal factors that cause rockfall problems This chapter explains the computer programs used to simulate the rockfall mechanisms Moreover, this chapter expresses the research papers around the world corresponding to the rockfall failure mechanism and remedial measures for rockfall mitigation works Finally, this chapter describes the background of the study area, including geology, topography, geomorphology, and past rockfall events in the study area 2.1 Definition of rockfall Rockfall means the movement of a single rock or groups of rocks detached from an unstable slope surface, and it is downslope movement in a free fall, bouncing, rolling, sliding, and finally stopping on the flat area The slope process of the rockfall is a complicated one, and the process is based on gravity, slope material properties, and the laws of motion (Botha, 2017) It divides into various steps of slope movement, starting from the source of the detachment, initial impacts, ballistic trajectory, impact, launching, rolling, sliding, and stopping, as shown in Figure 2.1 A is detachment B is initial impact C is ballistic trajectory D is impact E is ground contact and interaction F is launching G is rolling H is sliding I is stopping Figure 2.1 A schematic slope profile of the rockfall process (Dorren, 2003) 2.2 Rockfall process A block detached from the rockfall source area is the first step in the rockfall process The detachment of the blocks is based on the source materials susceptibility and the triggering mechanisms (Dorren, 2003) After the block detachment from the source area is completed, it may start a period of free fall It can be seen in free-fall motion when the angle of the slope is greater than 700 (Dorren, 2003) The blocks will move downward along the path of the slope as the motion bounces, rolls, and slides if the slope angle is lower than 700 The initial impact before a block's trajectory is a fundamental one in the motions of the slope process While the block is sourced from high above the initial impact zone, the potential energy is converted to kinetic energy (Botha, 2017) The lost amount of kinetic energy upon the impacts and the corresponding impacts with the slope surface are based on the ground conditions A block's loss range is from 75% to 86% in the initial free fall, depending on the initial impact (Dorren, 2003) The impacts of the hard surface of the rock permit the boulder to retain more energy as a result of the stiffness of the surface For soft surfaces, in other words, soil, it changes deformation under the pressure of an impact and, following that, forms impact scars (Botha, 2017) It is observed that the impact process absorbs some of the energy of the block, and this process slows down and reduces the run-out length of the block After the initial interaction with the slope, whether a block is bouncing, rolling, sliding, or even moving at all Table 2.1 illustrates the general motion of falling rocks on a steep slope, which has three types, namely, sliding, rolling, and bouncing motions (N C Koei, 2007) The flow chart of the motion pattern of the blocks is shown in Figure 2.2 Table 2.1 Illustration of motion mechanism of falling rocks Motion pattern Sliding Rolling Bouncing Slides down Rolls down a slope Bounces in the air and Diagram slopes moves downwards Slow Average Fast Bounce height Zero Small Great Occurrence of Rock Falls Falling speed Linear motion Sliding Rolling Parabolic motion Bouncing Collision Characteristics Free fall Figure 2.2 Motion pattern of falling rocks (N C Koei, 2007) 2.3 Causal mechanism of rockfall A rockfall hazard is a rapid movement that causes the deaths of people and facilities within the rockfall area The main external factors are the slope characteristics, which are slope height, irregularities, slope materials, slope inclination, and blocks of rock size and shape to control the motion of the rock along the terrain (Botha, 2017) There are two types of causal mechanisms for rockfall: rock mass characteristics and 10 triggering mechanisms It is defined as triggering mechanisms such as jointing, weathering, erosion of surrounding material during heavy rain storms, freeze-thaw cycles, pore pressure increase due to rainfall infiltration in the rock joints, earthquakes, and root growth or leverage by roots moving in high winds 2.3.1 Rock mass characteristics The orientation of the discontinuities or sets of discontinuities indicates the kinematic feasibility and whether a rock mass may fail or not, which can be determined using the dip and slip directions of the discontinuity sets relative to the open rock face When the upper and lower bounding joints of a rock block are dipping out of the face, gravitational forces can act upon the block Planar and wedge failures in a rock mass are controlled by these joints If these forces overcome the shear strength of the basal plane of the block, they will allow the block to slide Planar failures appear in a rock mass when the slope angle is greater than the dip angle of the joint In contrast, the slope angle is less than the potential slope angle for sliding failure, and the slope is likely to remain stable When the discontinuities dip directly into the face of the slope, toppling failures occur in a rock mass, and it is possible for this to happen due to triggering mechanisms such as freeze-thaw and so on (Wyllie, 2014) The nature of the discontinuities and how they are formed, as this has a large impact on the susceptibility of the rock mass, should be considered a crucial factor for a rock mass's characteristics (Rock Slope Engineering, 2005) The rock mass discontinuities are called the nature of their formation and involve faulting, bedding, foliation, jointing, and cleavage A joint with a low roughness coefficient, such as a fault plane, needs less shear stress to overcome the joint's friction component On the other hand, a discontinuity with a rough joint surface should have more shear stress to overcome the friction component Hence, it makes the block more stable and less likely to fail (Rock Slope Engineering, 2005) The size of the blocks that are detached from the surface of the rock can be determined by the spacing of the discontinuities (Palma et al., 2012) The probability of the smaller blocks being released, which are highly fractured rock masses with closely spaced discontinuities, will be higher than the larger blocks with widely spaced discontinuities The shape of the block that is released according to the axis length can be determined by the orientation of the discontinuities in the rock mass The 11 orientation of the discontinuities in the rock mass specifies the shape of the block that is released with regards to the axis lengths The rock slope collapse types based on the failure conditions are described in Figure 2.3 Figure 2.3 Schematic of rock slope collapse (N Koei, 2007) 2.3.2 Triggering mechanisms Triggered factors, whether naturally occurring or induced by human-induced processes, are crucial in causing dangerous rockfall hazards, especially in mountainous areas The natural occurrence facts include rainfall, weathering, freeze-thaw, earthquakes, root wedging, volcanism, snow melt, wind, and differential erosion (Wyllie, 2014) Human-made processes are involved in constructing the infrastructure, for example, buildings, roads, or tunnels, with vibrations from machinery and blasting, deforestation, slope morphology, and undercutting of slopes (Dorren, 2003) 2.4 Computer programs to simulate rockfall There are many numerical programs to simulate slope failures, help understand rockfall behavior, and investigate structural remedial measures for the solution of rockfall problems Moreover, numerical applications are an easy way to simulate and take a short time to find the most effective method for rock slope failures The rockfall run-out distance and kinetic energies, jump heights, and impact loads for each point along their fall paths are required to be evaluated by using the computer program It can predict the rockfall threat and provide suitable design and dimensioning for the rockfall protection method According to the simulation results, the design process includes selecting appropriate mitigation measures and determining the suitable 12 physical parameters of a rockfall prevention structure The mitigation method consists of ditches, barriers, rock catch fences and attenuators, draped mesh, rock sheds, and embankments The prevention structures include position, type, strength, length, height, and the magnitude and direction of the impact loads Some literature also investigates a holistic approach to designing the rockfall protective structure based on the numerical modeling results Nowadays, many numerical software programs have been developed corresponding to quantitative and modeling rockfall behavior, relying on different mechanical frameworks Generally, most of the numerical programs for rockfall hazard analysis are classified into four methods: (1) Lumped mass approach (LMA); (2) Rigid body approach (RBA); (3) Discrete element method (DEM); (4) Discontinuous deformation analysis (DDA) Above, these methods provide information to the user about the energy, trajectory, moving velocity, and jumping height of dangerous rockfall problems As a result, the analysis of the rockfall process can be calculated in the field of advanced engineering practices and protective measures (Agliardi & Crosta, 2003) The Colorado Department of Transportation developed the Colorado Rockfall Simulation Program to help solve the rockfall problem and determine the mitigation design for rock slope failures It can predict the flow of debris along the slope path Maerz & Youssef (2009) simulated the rock cuts with different geometry after dividing the slope into five sections by using CRSP Moreover, the Washington State Department of Transportation (WSDOT) chose CRSP software to specify the rock cut design for mitigation of the rockfall hazard (M., & Al E, 2017) I’nan (2011) performed rockfall risk assessments on a settlement near the intersection of the North Anatolian fault line (NAF) and East Anatolian fault line (EAF), Turkey, with the Rockfall Program (M., & Al E, 2017) Ansari and Singh (2013) also used the Rockfall program to simulate the jointed basaltic rocks of the area (M., & Al E, 2017) Another software name is the STONE computer program, which can simulate slopes and describes the movement of free-falling rock along the slope Guzzetti (2003) calculated the rockfall runout in Yosemite National Park with the aid of the STONE program (M., & Al E, 2017) Budetta (2004) analyzed the rockfall cases along the 13 section of the Sorrentine Road in Southern Italy by using Hoek’s rockfall program (Budetta, 2004) The RocFall software of the Rocscience product and the Colorado Rocfall Simulation Program (CRSP) are the common, specific methods to investigate the rock collapse problems (M., & Al E, 2017) Recently, RocFall software has become a well-known program related to rockfall events, and it is the most widely used for identifying slopes at risk for rockfalls Even though RocFall software simulates rockfall problems, this modeling differs in the way it defines the problem, processes the parameters, and displays the results 2.4.1 RocFall software In 1996, the Rocscience company started to create the RocFall software, which can assist with rockfall hazards and remedial measures Nowadays, this software has become popular and advanced among many geomechanics software programs A group of rock engineers at the University of Toronto, Canada, developed this program The engineers from the geological, civil, and mining sectors mostly use Rocfall in their related fields The program can simulate both in two dimensions and in three dimensions, and it is easy to define parameters and analyze the results The main task of the RocFall software is the design of remedial measures, and this software can test their effectiveness It uses particle analysis for the evaluation of the motion of the rock RocFall can be performed with a slope and a rock for the simplest simulation The highlight feature of the RocFall program is the barrier feature, and more advanced simulations involve incorporating random variation in the mass, velocity, and position of the rock and random variation in the location and material properties of each segment of the slope Furthermore, the RocFall program has more advantages compared with the CRSP software Both of them are based on statistical analysis to assess the rockfall risk Input parameters for RocFall software are the profile of the slope geometry, the properties of rock mass, and the mass of detachable lumped rock to carry out the calculation of energy, velocity, and bounce envelopes for the entire slope Additionally, it simulates 14 the distributions of energy, velocity, and bounce height along the slope profile (M., & Al E, 2017) 2.4.2 RocFall modelling A kind of statistical analysis program, RocFall simulates the rockfall problems corresponding with the rock falling along the slope, involving the energy, velocity, and bounce height of the rocks at the location of the endpoints of their paths Moreover, it indicates the condition of the rocks along the slope paths by producing graphs of the energy, velocity, and bounce height of the rocks and determining the impact on the mountain roads The users define the slope geometry, the properties of the rock, and the barrier Besides, the density, mass, quantity, and shape of rock are defined The users built the slope profile by using a number of segments and then assigning material properties to each segment It is easy to define and change the parameters of the rock and the profile of the slope This software allows the users to use the barrier features, which include predefined barriers or creating new barriers The required capacity, size, and location of the barrier can be determined based on the energy information and the input on the barrier from the program Barriers are defined as line segments standing vertically anywhere along the slope path They can be used to stop falling rocks or absorb some of their energy during their fall along their paths The users can use or define a new barrier with special properties and the height for a suitable slope, referred to as Macafferri barrier specifications In addition, the highest velocity impact on the barrier will display on the screen if the barrier selects only the paths The output results in the RocFall software are displayed as clear graphs and histograms, and these results can be exported to an Excel file to help the users with analysis and reports There are two output programs to aid in the design of the remedial measures with statistical analyses: the envelope program and the data collector program An envelope program describes the maximum velocity, kinetic energy, and bounce height of the rocks along the entire length of the slope profile The result of the envelope program is very important for deciding where the remedial measures should be placed "Data collectors" in the program give the distribution of the velocity, kinetic energy, and bounce height of the rocks at any location along the 15 slope profile as displayed in the histograms The data collector in the program is also useful when designing remedial measures, in particular when deciding the capacity of a barrier It is a vertical line segment utilized to pinpoint the location on the slope The users can create a segment anywhere on the slope and collect data about the movement of the rocks through the segment, which means recording data involving the kinetic energy, velocity, and vertical and horizontal location of all rocks that pass through when they fall down from the slope surface Particle analysis of RocFall software The movement of the rock can be calculated by using particle analysis in the RocFall software It is also the model of the physical process of a rockfall The particle analysis does not consider the effects of the size, shape, or angular momentum of the particle, whereas it has the advantage of being easy to calculate, getting a fast result, and allowing the sensitivity analysis to be performed When the input of the rockfall is poorly defined, it needs to be analyzed to determine the sensitivity of the results (Stevens, 1998) It is classified into three distinct sections for the particle analysis, which are the particle algorithm, the projectile algorithm, and the sliding algorithm The particle algorithms set up all of the initial conditions to prepare for the projectile and sliding, and the following starts with the projectile algorithms The projectile algorithm calculates the rock’s movement while the rock is traveling through the air, bouncing from one point of the slope to another The sliding algorithm simulates the motion of the block when it is in contact with the slope The simulation process time of the projectile algorithm is longer than that of the other algorithm because the velocity of the block is very low before it leaves the projectile algorithm (Stevens, 1998) Projectile Algorithm In the projectile algorithm, it is assumed that the rock has some velocity, which means that it will move through the air from its present location to a new location because of the force of gravity and a parabola The rock will strike another object, which can be further along the same object The projectile algorithm tends to find the location of the intersection both in the path of the rock (also known as a parabola) and in a slope segment (called a line segment) When the intersection point is found, the impact of 16 the rock can be calculated, corresponding to the coefficients of restitution After that, the rock is moving fast enough that the process begins again with the search for the next intersection point (Stevens, 1998) The minimum velocity (VMIN) is defined as fast enough in this program, and the users specify it at the beginning of the simulation For the minimum velocity of a projectile, it is the transition point between the projectile state and the state where the block is moving too slowly and should instead be considered rolling, sliding, or stopped After the determination of the correct intersection, the velocities can be evaluated, and all data collectors are checked for intersection with the parabola The data collector can record the location, velocity, and kinetic energy of the rock when it passes it After the velocity of the rock is calculated, it can be compared to VMIN If the value of the velocity is greater than VMIN, the process starts over again with the search for the next intersection point On the other hand, if it is less than VMIN, the rock can no longer be considered a particle and is changed into the sliding algorithm Sliding algorithm The purpose of the sliding algorithm is to calculate the movement of the rocks when they leave the projectile algorithm The sliding algorithm works on any segment of the slope and on any barrier It consists of a single straight-line segment that has the properties of slope angle (  ) and friction angle (  ) A constant value or sample from a random distribution is classified as the friction angle The rock starts sliding at any location, and it has an initial velocity according to the upslope or downslope For the equation in the program, it considers only the velocity component tangential to the slope After the sliding is initiated, the algorithm selected depends on whether the initial velocity is upslope or downslope (Stevens, 1998) Sliding downslope The initial velocity of the rock is downslope (or zero) The behavior of the rock depends on the relative magnitudes of the friction angle (  ) and the slope angle (  ) (    ) If the slope angle is equal to the friction angle, the driving force (gravity) is equal to the resisting force (friction) and the rock will slide off the downslope end of the segment, with a velocity equal to the initial velocity (i.e VEXIT = V0) There is a 17 special case when V0 = 0; in this case, the rock does not move, and the simulation ends (    ) If the slope angle is greater than the friction angle, the driving force is greater than the resisting force and the rock will slide off the downslope endpoint with an increased velocity The speed with which the rock leaves the slope segment is calculated by: VEXIT  V02  2sgk (2.1) where: VEXIT is the velocity of the rock at the end of the segment , V is the initial velocity of the rock, tangential to the segment, s is the distance from the initial location to the endpoint of the segment, g is the acceleration due to gravity (-9.81m/s/s) k is ± sin(  ) - cos(  ) tan(  ) where:  is the slope of the segment,  is the friction angle of the segment, ± is + if the initial velocity of the rock is downslope or zero ± is - if the initial velocity of the rock is upslope (    ) If the slope angle is less than the friction angle, the resisting force is greater than the driving force and the rock will decrease in speed The rock may come to a stop on the segment, depending on the length of the segment and the initial velocity of the rock Assuming that the segment is infinitely long, a stopping distance is calculated The distance is found by setting the exit velocity (VEXIT) to zero in equation 2.1 and rearranging: s V02 gk (2.2) It calculates the distance from the initial location of the rock to the end of the segment While the stopping distance is larger than the distance to the end of the segment, the blocks will side off at the end of the segment In contrast, when the stopping distance is smaller than the distance to the end of the segment, the block will stop on the segment, and the simulation will end Equation 2.1 is used to calculate the exit 18 velocity The location where the rock stops is a distance of s downslope from the initial location (Stevens, 1998) Sliding upslope For the sliding upslope, both the frictional force and the gravitational force reduce the velocity of the particle The segment assumed as the length is infinite, and the particle will finally come to rest The stopping distance and the distance from the starting location of the rock to the upslope end of the segment can be evaluated with equation 2.2 While the stopping distance is larger than the distance to the end of the segment, the blocks will side off at the end of the segment In contrast, when the stopping distance is smaller than the distance to the end of the segment, the block comes to rest, and the simulation will end Equation 2.1 is used to calculate the exit velocity When the rock slides up and resets, it is then added to the downslope sliding algorithm If the segment is steep for sliding (i.e    ) later the rock will slide off the toe end of the segment Alternatively, if the segment is not steep, then the location where the block stopped moving is taken as the final location, and the simulation will stop (Stevens, 1998) 2.5 Rockfall mitigation methods and selection The effect of the rockfalls is damage to vehicles, death or injury to drivers and passengers, and economic loss due to the road closures because of the high speed of the rock slope collapse Consequently, a large number of road sections require countermeasures against rock slope collapse, especially on long, large slopes and steep cliffs in mountainous zones Stabilization methods and protective measures are the countermeasures to protect against rock slope failures, which have a tremendous impact on roadway safety The aim of the stabilization method is to reduce the frequency of rockfall cases by removing the source of the rockfall, increasing the stability of the rock face, increasing the resisting forces, and decreasing the driving forces This method is generally installed within the rock mass and is much less visible than other protection measures Protection methods include the ability to act to stop, divert, or control when the rockfall will occur It is installed externally to the rock mass, and thus these are much more visible 19 2.5.1 Classification of countermeasures Rockfall countermeasures are defined as two types: rockfall prevention works and rockfall protection works Prevention works correspond with the rock fall source, for instance, removal of the rocks and crib work, whereas protection methods involve protecting works from the damage of rock fall, such as barriers and catchment works Table 2.2 describes the most common countermeasures for rock slope failures (N.C Koei, 2007) Table 2.2 Classification of countermeasures for rock falls Countermeasures Classification Type of works Hand/ Mechanical Scaling Slope Geometry Modification Trim Blasting Rock Bolts Rockfall Prevention Method/ Slope Stabilization Internal Rock Dowels Stabilization Shear Pins Reinforcement Injectable Resin/ Epoxy Method External Shotcrete Stabilization Drainage Weep Drain Draped Mesh/Nets Mesh/ Cable Nets Anchored Mesh/Nets Rockfall Earthen Barriers Protection Method Concrete Barriers Barriers Structural Walls Flexible Barriers 20 Attenuators Catchments Ditches/hybrid Ditches 2.5.2 Rock slope stabilization method In many rock collapse failures, engineers mostly consider slope stabilization methods to reduce localized slope failures in both rockfall and erosion Commonly, the slope stabilization method is the most effective strategy for preventing failure at the source through stabilization without the installation of structures to protect against failure problems in the future A stabilization method is classified as altering the slope geometry, installing drainage, adding reinforcement, or using combinations of these methods An overview of common stabilization procedures is provided in Table 2.3 (Richard Andrew et al., 2011) Table 2.3 Overview of stabilization procedures Mitigation Measure Description/ Purpose Slope Geometry Modification Hand/ Mechanical Remove loose rock from slope by hand tools and/or Scaling mechanical equipment Commonly used in conjunction with other stabilization methods Trim Blasting Remove overhanging faces and protruding knobs and to modify the slope angle to improve rockfall trajectory and slope stability Reinforcement Internal Stabilization Rock Bolts Increase the normal-force friction and shear resistance along discontinuities and potential failure surfaces Can 21 apply in a pattern or in a specific block Rock Dowels Increase shear resistance and reinforce a block To increase normal-force friction once block movement occurs Less visible than rock bolts Shear Pins Provide shear support at the leading edge of a dipping rock block or slab using grouted steel bars Can easily be blended with surrounding rock by colored concrete Injectable Resin/ Epoxy Resin/epoxy injected into the rock mass through a borehole; travels along joints to add cohesion to discontinuities Decreases the number of rock bolts or dowels needed in a rock slope External Stabilization Shotcrete Pneumatically applied concrete requiring high velocity and proper application to consolidate Primarily used to halt the ongoing loss of support caused by erosion and raveling Adds small amount of structural support for small blocks Sculpted and/or colored shotcrete can be used for improved aesthetics and to cover rock bolts and dowels Drainage must be installed Drainage Weep Drain Reduce water pressures within a slope using horizontal drains Commonly used in conjunction with other design elements Good for aesthetics because drains are rarely visible Shotcrete The shotcrete method is sprayed directly onto a slope using compressed air, either manually or by machine, and involves a wet or dry-mix mortar with a fine aggregate 22 The required thickness of the shotcrete should be considered based on the actual site requirements This method is classified as structural shotcrete and unreinforced shotcrete as a kind of slope stabilization method Unreinforced shotcrete is applied to protect a well-defined strip of rocks with a higher rate of erosion than the surrounding rock, such as faults or shale lenses in sandstone Furthermore, this method should be used on an entire slope composed of highly erodible material It is obviously true that unreinforced shotcrete is applicable when laying the slope back to avoid a structured solution and the differential erosion of the rock slope, which will create stability problems that become worse with time The purpose of structural shotcrete work is to protect the surface of the rock slopes, which, left untreated, would erode into a fault zone or clay seam, as well as provide structural support for rock slope failures Otherwise hard rock that is either undermined by erosion or is unstable as a result of unfavorable orientations or a degree of fracturing This type can be used not only for the original construction condition or as part of the remediation of an existing unstable rock slope problem, but also for a retaining wall supporting the rock slope Welded wire mesh, rock bolts, and dowels are the structural components of the shotcrete system The advantage of using dowels in the shotcrete is that it improves its tensile capacity Particularly, the shotcrete will be needed to retain or transfer loads, and it also includes an essential surface protection function in conjunction with its structural function Figure 2.4 Installation of structural shotcrete (Richard Andrew et al., 2011) In the form of weep holes or wick drains, a drainage system requires drawing out the water from behind the shotcrete to prevent increased water pressure that causes 23 cracking and instability for any type of shotcrete work The face of the rock should be scaled and cleaned to remove any loose material such as stone, dirt, ice, and vegetation before shotcrete application Additionally, the removal of highly fractured or weak rock should be done to expose more competent material It is noted that the rock face should be free of flowing water but damp enough to facilitate proper curing, and any wire mesh should be attached securely to the area of application It will serve as a frame to hold the shotcrete in place For the mixing process, the wet method should be mixed with water before entering the application nozzle; however, dry work is mixed with the water at the nozzle The thickness of the shotcrete varies between 50 mm (2 in) and 0.6 m (2 ft) Shotcrete allows for multiple layers of about 50 to 100 mm (2 to in) each, and it needs to cure between applications for the thicker thickness of shotcrete (Richard Andrew et al., 2011) An air compressor, application nozzle, and cement mixer are needed to install shotcrete on the surface of the slope 2.5.3 Rockfall protection method Mesh or cable nets, barriers and fences, catchment areas, and ditches at the toe of a slope are structure designs to prevent rockfall from reaching the highway This method allows the blocks to fall, but it prevents them from causing any damage to a structure, the public, or road users The aim of the rockfall protection method is to stop a rock, control its trajectory, reduce its energy, or provide a catchment, which is used in conjunction with an inclined deceleration zone Due to the fact that the structure of the protection is constructed externally on the slope, the protection structure is more difficult to hide and fit within the context of the surrounding area than the structures of slope stabilization Building a tunnel, realigning the roadway, or constructing an elevated portion are the applicable methods to mitigate rockfall problems Table 2.4 describes an overview of the most common protective method for rockfall (Richard Andrew et al., 2011) Before the evaluation of effective rockfall protection, a site investigation should be conducted to determine the factors affecting rockfall, which consist of slope height, topographic profile, variable slope angles, potential launch points, rock type variations, soil cover, vegetative cover, potential runout areas, and impact zones Table 2.4 Overview of protection measures 24 Mitigation Measure Description/ Purpose Mesh/ Cable Nets Draped Mesh/ Nets Hexagonal wire mesh, cable nets, or high-tensilestrength steel mesh draped over a slope face to slow erosion, control the descent of falling rocks, and restrict them to the catchment area Anchored Mesh/ Nets Free-draining, pinned/anchored-in-place nets or mesh Used to apply active retention force to retain rocks and soil on a slope Barriers Earthen Barriers Barriers constructed of natural soil and rocks (berms) or mechanically stabilized earth (MSE), placed at the distal end of a catchment area to improve its effectiveness MSE walls, in particular, can withstand large kinetic energies and repeated impacts Easily repaired Earthen material blends well with surrounding landscape Concrete Barriers Rigid barriers provide protection from low- energy impacts Relatively cheap, easy to obtain, and fast to install Normally used as temporary barrier in context sensitive areas Structural Walls Rigid barriers used to intercept falling rocks and restrict them to a prescribed catchment area Can withstand significant kinetic energies and repeated impacts Facing can be installed on road side of walls to improve aesthetics Flexible Barriers Flexible barriers made of wire ring or high- strength wire mesh with high energy-absorption capacity, 25 Mitigation Measure Description/ Purpose supported by steel posts and anchor ropes with a braking system Fence is fixed at the bottom to hold rocks Effective on high- to low-energy events Attenuators Flexible barriers is similar to fencing (above) but not attached at bottom (an extra length of fence lies on the slope face); allow rocks to move beneath the two sections of fence and direct them into a catchment area Require less maintenance than standard fencing Catchments Ditches / hybrid ditches Shaped catchment areas normally placed along the roadside or slope base and used to contain rockfall Hybrid ditches are a combination of a barrier and a ditch Anchored mesh/ nets Anchored mesh work is defined as being anchored to the rock face in either a grid pattern across the face or via a row, or anchoring the entire upper and side perimeters of the wire mesh to inhibit erosion and rockfall Anchored mesh is a different system compared with the secured drapery method The mesh or cable net is used in an anchored system because it is vastly stronger than that used for drapery Moreover, the mesh or cable net is a fastening method that includes closely spaced rock bolts with plates that anchor the mesh to the slope The advantage of these anchors is that they provide a more robust and permanent approach to stabilizing the slope 26 Figure 2.5 Installation of an anchored Tecco mesh system in Northern California (Richard Andrew et al., 2011) Rock bolts are the main way to attach the mesh or net to the slope in an anchored mesh system The mesh can stretch between the bolts to increase its contact with the slope, which adds a normal force to protruding blocks Thus, the slope is more stable due to the application of an anchored mesh system, but these systems are more expensive than the unsecured draped system In that case, the designed and constructed works should be carefully checked, and this method requires periodic maintenance to remove accumulated material from behind the wire mesh (Richard Andrew et al., 2011) Flexible barrier A flexible barrier is a kind of rockfall fence system that absorbs energy through deformation of the fence material and braking elements The components of a flexible barrier are illustrated in Figure 2.6 (Qi et al., 2018) Deformable cables and mesh are the most commonly used fencing materials, the most common types of which are woven wire-rope mesh nets or interlocking ring nets Interlocking rings can be tested to determine which are providing the greatest deformation and energy absorption The function of the mesh is to support a series of steel beam posts anchored to a foundation with grouted bolts to prevent rock from falling to the road Commonly, the foundation is composed of a concrete cap secured to rock bolts or a large concrete mass The braking elements consist of upslope anchor ropes It gives additional support to the fence and applies friction brakes that are activated during the high-energy process to absorb energy Some of the largest fences can withstand impact energies up to 5000 kJ (1,844 ft-tons) Some of the largest fences can protect the energies of the rockfall above 5000 kJ (1,844 ft-tons) (Richard Andrew et al., 2011) 27 Figure 2.6 Flexible barrier system (Qi et al., 2018) These systems need a variety of machines or equipment to install the fences on the cliff For example, when fence posts and brake cables are installed into the ground, they require grouted anchors Handheld drills are for drilling the anchor holes, and pumps are needed to place the grout For supporting the fence, steel H-beams are essential to use as fence posts A woven steel cable is used for constructing the braking system, attaching the net to the fence, and attaching sections of fence to one another It is remarkable that helicopter support may be required for the difficult access areas Removing the accumulated rocks and debris should be done as a periodic cleaning for the maintenance of fences 2.6 Case studies of rockfall events and mitigation measure Case studies of rockfall events Rockfall disaster caused by typhoon Kong-Rey with the highest intensity rainfall of 94.5 mm/h on August 31, 2013 With an average weight of 150 T and dimensions of 4.5 m x 3.5 m x 3.5 m, the rock fell to the road, and the fallen rock struck the passing vehicle on Provincial Highway No 2, in Badouzih, Keelung Figure 2.7 describes the source area of the rockfall, the rockfall events, and the damage to the car 28 Figure 2.7 Rockfall event in Badouzih, Keelung (Wei et al., 2014) A rockfall happened on the Norwegian motorway near the entrance to the Larvik tunnel on Friday, December 13, as described in Figure 2.8 The cause of the major rockfall was the large wedge failure developed according to the failure mechanism Figure 2.8 Rockfall event in Badouzih, Keelung (Hancock, 2019) A rockfall-type geohazard occurred on the highway Sea to Sky joining Vancouver to the ski resort Whistler on July 29, 2008, as shown in Figure 2.9 The highway road was blocked due to the rock-falling process Moreover, the road was seriously damaged due to the accumulation of high-speed fallen rock, as can be seen clearly in Figure 2.9 29 Figure 2.9 Rockfall on Sea to Sky highway (B.C.) (Volkwein et al., 2011) In the northern mountainous province of Lai Chau, a huge rock suddenly dropped onto a car, and this dangerous event killed the driver and damaged the vehicle, as shown in Figure 2.10 This accident happened on June on the National Highway 12 in Le Loi Commune, Nam Nhun District Figure 2.10 Rockfall case in Lai Chau province (News, 2018) A very large rock from the slope surface of Ba Hoa mountain that suddenly fell on Nguyen Tat Thanh street, area 5, Le Hong Phong ward, Quy Nhon city, Binh Dinh province, on October 25, 2021 Ba Hoa mountain is mostly a steep mountain Nguyen Tat Thanh road is closed at the toe of the mountain Even with the heavy rain in the 30 past years, rock slope failure cases were very rare, but last year rockfall occurred on this road, and last year's event threatened the lives of people traveling on this road Figure 2.11 Rockfall case in Binh Dinh province (News, 2021) Case studies of rockfall mitigation measure The rockfall cases have become more serious and have attracted more attention, resulting in the construction of highways in mountains and the excavation of high slopes Some researchers analyze the degree of influence of different factors on the movement distance of falling rocks and determine the main factors affecting the movement trajectory of the blocks by carrying out probabilistic analysis in their research progress Cha et al and Xiao-hui et al calculated the rockfall motion parameters with the theoretical formula method (Jiang et al., 2021) Other researchers also predicted the movement distance and movement path of falling rocks by carrying out a model test or numerical simulation On the other hand, many researchers have investigated the selection of remedial measures and countermeasures for rockfall mitigation around the world Below are a few examples of research papers dealing with rockfall analysis and the design of rockfall mitigation measures De Graaf et al (2004) illustrated the design and implementation of rockfall mitigation measures for the DeCew Falls hydroelectric station An overview of the study included the stability of the rock slope back of the powerhouse and the proposed remediation method, which was designed to protect the powerhouse and related infrastructure The author considered technical effectiveness and costs in order to provide a relative ranking of the options for the determination of the design Based on detailed stability 31 analyses involving a number of remedial options that were evaluated with rockfall simulations, the author chose draped ring nets with friction brakes near the toe without permanent rockfall fences for protection of the hydropower station (Beers, 2004) Maerz et al (2015) studied the determination of hazardous zones on steep mountain roads and specified the solutions for remediation and mitigation measures to improve safety along the roads of Fayfa Mountain in the Kingdom of Saudi Arabia The author considered suitable remediation methods by minimizing the cost and making use of local expertise and available equipment The remediation solutions in this research suggested scaling loose rock, reshaping the slope with machines, making the ditch more capacity by widening and deepening it, installing jersey barriers on the down gradient side of the ditch, constructing retaining walls to stabilize the lower section of slopes and gabion, and using large block barriers to stabilize small debris flow channels Furthermore, the research recommended a small number of cuts, trim blasting, and creating benches For this reason, anchoring systems, advanced walls, for instance, sheet pile and anchored retaining walls, draped mesh, and sacrificial fences, were not selected in this research because of their high cost and because local contractors could not use these methods effectively (Maerz et al., 2015) Volkwein et al (2011) conducted rockfall hazards, rockfall source areas, trajectory modeling, and structural countermeasures in the research paper "Rockfall characterization and structural protection: a review." This research described an overview of previous and current studies corresponding to the rockfall The researcher indicated that there was a specific need to improve the prediction of probabilities in hazard and risk assessment for a better quality of the risk of the rockfall and to improve the resolution of hazard and risk maps (Volkwein et al., 2011) Budhbhatti et al (2016) studied the research topic related to the analysis and design of advanced rockfall mitigation measures implemented at Saptshrungi Gad Temple, Nashik This research described in detail the study of site and engineering investigations performed to understand the problem and contribution factors for proposed suitable mitigation methods In addition, this paper contributed to the discussion of the design principles and basic equations that tend to be applied in the analyses of the software as well as in the analyses of the proposed solutions The 32 research analyzed probabilistic analysis to determine the suitable location for the positioning of a rockfall barrier against rockfall events on the temple and Parikrama paths Finally, the author suggested covering the slope with a secured drapery system along with a dynamic rockfall barrier above the temple and Parikrama paths in accordance with all the investigation and analysis results From a technical and cost perspective, this method was the most reasonable and reliable solution for this area (Budhbhatti et al., 2016) Duong-Ngo et al (2020) investigated the problem of the rockfall from km 18+980 to km 19+120 near the Halong-Vandon highway in Vietnam The estimated source area of the rockfall was 140 m long and 50 m high The excavated part of the mountain to construct the highway caused an imbalance of the remaining slope following the rockfall event Manual measurement and an unmanned aerial vehicle (UAV) were used to investigate the topography of the location The researcher found the solution to prevent the rockfall event for the Halong-Vandon highway, which was the retaining system composed of rock bolts, wire mesh, and an anchored wall, based on the analysis results from Plaxis software While the rock was bolted together with wire mesh constructed at the top of the mountain to protect against rockfall, the anchored wall was installed at the bottom of the mountain to stabilize the slope (Duong-Ngo et al., 2020) 2.7 Overview of the study area Son Tra Peninsula is situated in Tho Quang Ward, Son Tra District, in Da Nang City This peninsula is on average 10 kilometers northeast of the Da Nang city center, and it has an elevation of 695 meters above sea level Additionally, it covers an area of 60 square kilometers, with km of width and 13 km of length Topography features The area's topographical features are characterized by high mountains, deep abysses, long mountain slopes, and a wide water collection area The landform covers mountains and hilly, upland areas based on the micro-geomorphological map of Da Nang city The Hoang Sa road, which is in the study area, is constructed in a mountainous area that passes through many road cuts and high-slope sections Hoang Sa is the famous tourist highway of the Son Tra Peninsula, near Da Nang City, which 33 passes through complex mountainous terrain One side is the sea, and on the other side is the mountain, as shown in Figure 2.12 Figure 2.12 Hoang Sa road, Son Tra Peninsula (taken by using UAV investigation) Furthermore, there are many large slopes cut without reinforced protection The average elevation of the study area terrain is 695 m above sea level The slope within the study area is steep, with a range of 35% to 50% based on the slope map of Da Nang according to international standards The Hoang Sa Road is a type of provincial road that connects mainly urban districts and mountainous areas This road has lanes for motor vehicles and lanes for rudimentary vehicles, and the roadway width is 10.5 m Geological characteristics The geological characteristics of the rocks in the Son Tra area are mainly granite, and the surface of the granite is sandy soil to a depth of to 10 m The surface-cover soil is mixed with loose and porous particles When the humidity increases, the cohesive force and internal friction angle decrease sharply, resulting in sliding failure problems Beside this, many places along the Hoang Sa road have had sliding failures in recent years due to the strong weather and rocks with a high sliding risk Table 2.5 Geologic classification Geology Name Content 34 γaT3hv1 Hai Van Complex: Phase Biotite Granite, Two-mica Granite The geological feature of high-risk, rock-prone areas (Points 28, 29, and 30) is a steep mountain with granite rock that belongs to the Hai Van complex (γaT3hv1) The geologic classification of the Hai-Van Complex is described in Table 2.5 (The Study on Integrated Development Strategy for Danang City and Its Neighboring Area in the Socialist Republic of Vietnam (DaCRISS), 2010) Climate characteristics The study area is located in a tropical monsoon climate with a typical climate regime for the central climate region The climate here is divided into two distinct seasons: the rainy season and the dry season The rainy season here usually starts from August to December, and the dry season lasts from January to July Table 2.6 describes an overview of the rainfall data in the Son Tra Peninsula and Da Nang city Table 2.6 Annual and maximum rainfall data of Da Nang City The biggest annual rainfall (1964) 3307 mm Average number of rainy days per year 147 days Average annual rainfall 2066 mm Month with the most average number of rainy 22 days (in October every year) days Maximum daily rainfall 332mm Maximum rainfall in October, 2022 800 mm Floods, landslides, and rockfall cases mostly happen around the Hoang Sa road during heavy rain Especially in the rainy season, slope failure cases, in particular rockfall problems, have been experienced along the Hoang Sa road, whereas rock slope failures occurred on the Hoang Sa road even during the sunny season Compared with past 35 rockfalls recorded, the times of the rockfall cases were higher in the rainy season, but the rockfall problems in 2022 also happened in the sunny season Hydrological characteristics The location of the study area is adjacent to the sea, so the flood cases occurred within the Son Tra Penisular Small basins all have large vertical and horizontal slopes and short main bed lengths Thus, the flow is concentrated quickly and high intensity occurs in a short time, and this is the main factor leading to rock rolling and subsidence A hydrogeological survey of the route section showed that water occurred almost all over the body of the slope of the talus When there were heavy rains lasting into the rainy season, surface water flowed along the slopes, causing erosion of the slope surface A part of the water seeped into the rock, changing the state of the slope rock, and then rock falls happened 2.7.1 Past rockfall events in the study area Small landslides and rockfalls happened due to the unfavorable weather along some sections of concrete road leading to the Ho Sau area in 2017 Thus, a volume of soil and stones blocked some drainage systems along the route At that time, this road was facing the high risk of having quantities of rocks detached freely from the cracked surface of the cliffs, which consequently caused danger for road users Figure 2.13 Rockfalls alongside a section of the road connecting with the Ho Sau area (NGUYEN DUC NAM, 2018) On December 28, 2018, landslides and rockfalls occurred along several roads to the Son Tra Peninsula due to the heavy downpours dampening Da Nang during December 36 Moreover, a concrete route adjacent to Yet Kieu Street leading to the Bai Cat Vang (Yellow Sand Beach) tourist area was faced with traffic movement as a result of small landslides and rockfalls Roads leading to the Ho Sau area, the Vong Canh hills, the Dong Dinh Museum, the Linh Ung Pagoda, the ‘Ho Xanh’ also known as Green Lake, and the InterContinental Danang Sun Peninsula Resort, all located on the Son Tra Peninsula, had landslides and rockfall cases due to the heavy rains in parts of Da Nang on October 18, 2020 A volume of soil and stones fell onto the road to the Son Tra Peninsula, threatening road users' safety, and it was blocked by fallen trees Figure 2.14 Rockfalls along a section of a road leading to the Ho Sau (Deep Hole) area (Landslide Warning Along Roads Leading to Son Tra Peninsula, 2020) A rockfall occurred on a section of Hoang Sa Road, Son Tra Peninsula, Da Nang City, on April 15, 2022 Due to the detachment of the fallen block process, the retaining wall structure to prevent slope failure problems was damaged, and road users faced traffic blocking in this portion Generally, the months of April and May are in the middle of the sunny season in Da Nang City Even though the weather was not rainy, a volume of blocks spilled out onto the section of the Hoang Sa route While the stone fell, it hit the concrete retaining wall and damaged the embankment system, which 37 followed onto the surface of the road In 2022, there were six locations at risk of rockfalls along a section of Hoang Sa road through the Son Tra Peninsula After the heavy rainfall and the storm called Son Ca happened, within the Peninsula area, there were 41 locations of slope erosion on October 14th and 15th Roads were destroyed, and soil, rubble, trees, and other objects blocked many paths Both sides of the Hoang Sa road section above the Linh Ung Pagoda were seriously eroded due to the landslide and rocks Figure 2.15 Rock rolled down onto the Hoang Sa road after a heavy rain in 2022 (Early Handling of Rockfalls on Son Tra Peninsula Taken, 2022) Rockfall events along the Hoang Sa road occur almost every year However, the damage and hazard conditions caused by rockfall vary from year to year depending on the size of the fallen rock Almost all rockfall cases happen during the rainy season, but last year there were three or four rock slope failures on the Hoang Sa road, not only during the rainy season but also in sunny weather Due to the rock falling process, people who live within this area or who used the Hoang Sa road and moved vehicles on this road face damaged and delayed traffic Hence, the study area is required to investigate the causes of rock slope failures and find remedial measures for rockfall problems along the Hoang Sa road leading to the Son Tra Peninsula 38 Figure 2.16 After rockfall conditions (VnExpress, 2022) 39 CHAPTER METHODOLOGY This chapter explains the field survey data, the requirements for the simulation software, and the numerical procedures performed for rock mechanisms and slope stability analysis in this research Section 3.2 consists of a field investigation and a DEM model with a UAV Section 3.3 provides an overview of procedures and necessary data to calculate the rockfall mechanisms based on two conditions: the normal condition and the rock protection method Section 3.4 describes the necessary data and the procedures for the simulation model to calculate the slope stability based on two conditions: the existing condition and the slope stabilization methods 3.1 Introduction The thesis belongs to a numerical research study that focuses on the rock fall behavior and stability of the rock slope as well as the selection of suitable methods for rockfall prevention on the Hoang Sa road The general methodology of the study is described in the flowchart shown in Figure 3.1 Figure 3.1 Flow chart of overview of the study 40 In this analysis of the study, there are two parts: rockfall mechanisms and rock slope stability analysis under normal condition (without proposed method) and the rockfall mechanisms analysis using passive method and rock slope stability using active method The analysis of the procedures for the rockfall mechanisms and rock slope stability is illustrated in Figures 3.2 and 3.3 The flowchart of the analysis study (1) shows the progress made to simulate the mechanisms of the rock fall and the stability of the rock slope based on the normal condition (no prevention method was applied) Under normal conditions, there are two numerical software packages to simulate the rockfall mechanisms using the RocFall 2D software (v.4.0) and the rock slope stability using the Plaxis software (V21.01) Additionally, the considerations for the rock slope stability analysis in the Plaxis model consist of ground water, ground water and rainfall, high water, and high water and rainfall Figure 3.2 Flow chart of the analysis study (1) The flowchart of the analysis study (2) indicates the procedure for the simulation of the rockfall mechanisms and the rock slope stability based on the countermeasures Three countermeasures were proposed in this thesis: a flexible barrier, the shotcrete method, and the anchored mesh method, based on the consideration of design, software, construction feasibility to determine the solution for rockfall problems RocFall software was chosen to simulate the rockfall mechanisms, and the Plaxis program was applied to calculate the factor of safety for the rock slope as well as the stability condition when shotcrete and anchored mesh were applied to the rock slope 41 Also, ground water, ground water and rainfall, high water, and high water and rainfall were considered in the calculation of the rock slope stability analysis stage without and with the proposed method condition Figure 3.3 Flow chart of the analysis study (2) 3.2 Data Collection 3.2.1 Collection with field investigation Desktop studies and field surveys were done in the data collection section for the site assessment of the study area The desktop study is to collect information dealing with the site and the history of past rock slope collapses The purpose of the field survey is to classify the geomorphology characteristics of the site and the nature of the geological features, especially rock run-out zones In addition, a geological and rock mechanical study was performed in the field survey, finding that the factors affecting the rockfall include slope height, slope angles, the structure of the rock, joint patterns, and types of rocks It is very important to select the most effective rockfall protection method with the aid of numerical software The information to be collected during the site assessment is a critical role in the selection of countermeasure designs for rock slope failures The data was collected using the following procedures: pre-field work and field work Collection and processing of the geological documents related to the 42 study area, interpreting satellite images and aircraft images, etc., to help the field work achieve the best results, was carried out in pre-field work Field work was conducted to investigate the survey routes according to the geological survey process 3.2.2 Collection with UAV, DEM model and GIS The second part of the data collection is to specify the high-risk rock fall source area among the other risk areas on the Hoang Sa road The location of potential rock falls can be determined based on UAV surveys The abbreviation "UAV" refers to an unmanned aerial vehicle, also known as a drone, that does not include a human pilot, crew, or passengers on the aircraft The functions of UAV technology are numerous, such as mapping, agriculture, and surveillance Moreover, it has been used recently for terrain surveying, 3D modeling, land damage assessment, and geological hazards It is one of the most effective 3D terrain survey equipment nowadays Combining UAV with cameras for high-resolution, sharp images, this technology is a quick and safe tool for collecting detailed topographic surface data that is used as a reference for the design of barriers The primary surface, classified as Digital Surface Model (DSM), Digital Elevation Model (DEM), and orthomosaic images, can be exported using UAV function equipment with digital cameras, for example, Agisoft Photoscan and Pix4D Mapper Agisoft Metashape is commonly known as Agisoft PhotoScan UAV companies and archaeologists widely use Agisoft Metashape software Many researchers underlined the necessity of DEM in landslide and rockfall investigations for prediction at either small or large scales (Kakavas & Nikolakopoulos, 2021) Geographical Information Systems (GIS) can be used for conducting risk assessments and contributing to risk zonation mapping Mason and Rosenbaum (2002), Mancini et al (2010), and Calvello et al (2013) revealed that civil engineering and protection purposes utilized GIS for creating spatial models of potential hazard zonation maps in the last decade Additionally, Carrara et al (1991), Barredo et al (2000), Fernandez et al (2003), Kolat et al (2006), Yilmaz and Yildirim (2006), Nandi and Shakoor (2009), and Paulin et al (2014) described the important role of GIS in hazard assessment and mitigation Besides, Chau et al (2003) and Chau et al (2004) performed a rockfall susceptibility map with the aid of GIS-based techniques, referred to as a rockfall 43 inventory for Hong Kong From the past decade until now, many researchers and young engineers from geotechnical, land use, and disaster work have mostly used GIS to investigate hazard assessments, including landslides, rockfalls, debris, and other related disasters, using simplified geotechnical models (Burrough and McDonnel, 1998) The GIS technique can be used in rockfall hazard simulation models and supports the zonation of large-scale areas, such as an entire province or a river basin (Sakellariou and Ferentinou, 2001, Ferentinou, 2004, Sakellariou et al., 2006) In this thesis, the study was carried out to determine the high-risk rock-prone area using UAV, DEM models, and Agisoft software The DEM model was generated from the Agisoft software after it was combined with the collected data from taking pictures using the UAV equipment at the site, as shown in Figure 3.4 Moreover, the rockfall risk location can be determined using GIS and Google Earth The rock slope profile of the high-risk rockfall area in this thesis, including slope elevation, slope angle, and distance, was taken from Google Earth, as shown in Figure 3.5 (a) (b) (b) Figure 3.4 (a) UAV equipment; (b) Example of workflow in Agisoft Metashape to generate a DEM (González-Quiñones et al., 2022) 44 308 ft (93.88 m) 128 ft (39.14 m) Figure 3.5 Slope profile taken from google earth 3.3 Rockfall mechanisms analysis based on RocFall software The first analysis of the study was the mechanism of the rockfall based on the numerical simulation using RocFall 2D (v.4.0) software Before starting the calculation of rockfall mechanisms, the collected data from the site assessment was required to run the program It consists of three main categories: source characterization, slope characterization, and previous rockfall Source characterization is related to location, rock mass characteristics, block size, and block shape Slope characterization is defined as slope topography, slope morphology, and slope materials After that, define the slope, define the slope material properties, choose the fallen rock location, and select block size and velocity in the software These procedures were first carried out to simulate rockfall behavior with this program And then, the software calculated the rock end point, rock run-out paths, kinetic energy, bounce height, and velocity The results of rockfall energy and bounce height in the software are important factors in selecting the structure type and location Furthermore, structure type and location were considered depending on the design, technical issues, construction time, and construction cost According to the rockfall 45 energy and bounce height results from the program, the structure type and location were chosen for barrier design If a designed barrier can save the blocks and the paths of the fallen stones stop at the barrier location, the simulation of the program will end The below-mentioned procedures are the simulation model for rockfall mechanism analysis, as well as those described in Figure 3.6 Figure 3.6 Simulation model procedures for rockfall mechanisms analysis 3.3.1 Rockfall mechanisms analysis without proposed method In the study, the RocFall 2D (v.4.0) program simulated the rockfall mechanism analysis under two conditions: without the barrier method and with the barrier method The geometry of the slope profile was taken, as referenced in the photo and the video, from the site investigation, and also taken used Google Earth The model geometry and slope geometry with the assumed location of the fallen stone to analyze the rockfall mechanisms are illustrated in Figure 3.7 46 Figure 3.7 Model geometry without barrier method in RocFall for Son Tra area Figure 3.8 The procedure for choosing the fallen stone location The phenomenon of rockfall was considered before determining suitable mitigation measures for rock slope failures Thus, rock trajectory, rock end-points, bounce height, kinetic energy, and velocity were simulated under existing condition with the aid of RocFall 2D (v.4.0) software It was obvious that the software could not determine the location where the stone would fall, which is why the study determined the location of the fallen stone The procedure for determining the location of the stone is shown in Figure 3.8 47 Table 3.1 Input materials for RocFall Material type Normal coefficient Tangential Friction angle of restitution (Rn) coefficient of (degree) restitution (Rt) Mean Standard Mean Deviation Standard Mean Deviation Standard Deviation Bedrock 0.35 0.04 0.85 0.04 30 Retaining 0.5 0.03 0.5 0.03 30 0.4 0.04 0.9 0.04 30 wall Asphalt road Physical properties of rock material, including size and shape of blocks, slope profile, roughness, velocity, and hardness of the material along the profile, are common input parameters for the RocFall program Among the parameters in this software, it mainly requires normal and tangential restitution coefficients, friction angle, velocity, and rock mass density Additionally, the coefficient of restitution (COR) is an important parameter for the analysis of rockfall problems Some researchers used the in-situ experiment or a table provided by Rocscience for determining the value of the coefficient of restitution (Verma et al., 2019) The study was taken from the in-situ condition of the material on the surface and the recommended value in Rocscience; the values of the input parameters for this program are shown in Table 3.1 Table 3.2 Input materials for initial condition of stone Parameter Horizontal velocity (falling speed of fallen stone) (m/s) Value The mass of fallen stone (kg) 2700 Density (kg/m3) 2700 48 In the geological profile of the slope, the result of the field study was classified as three material types: bedrock, retaining walls, and asphalt roads The initial condition of stone dealing with horizontal velocity, the mass of fallen stone, and density require parameters to define the properties of falling blocks in RocFall software for simulation of the rockfall phenomenon These values in this study, taken from the field study and site video recorded, are described in Table 3.2 3.3.2 Rockfall mechanisms analysis with proposed method The selection of the proposed countermeasure among passive protection methods was considered based on maximum total kinetic energy simulation results as referenced in Design considerations for passive protection structures, October 2016 (Rori Green, 2016) and the European Guideline ETAG 027 (EOTA 2008) (Peila & Ronco, 2009) Passive protection methods include rigid barriers, flexible barriers, attenuators, and rock sheds The types of passive protection methods are classified based on their energy capacities as described in Figure 3.9 After the proposed method is chosen, the model examines where a barrier should be installed The model geometry and slope geometry with two proposed barrier location points are illustrated in Figure 3.10 Figure 3.9 Range of energy capacities for a variety of passive protection methods 49 Figure 3.10 Slope geometry with two proposed barrier options in RocFall software for Son Tra area The consideration of the location where a barrier should be placed on the slope surface and the type of barrier design consisting of barrier height and barrier capacity are very important to determine the effective mitigation method for rockfall The motion of blocks from the detachment of the rock surface still moving or stopping at the designed barrier is critical based on the above factors in the simulation of the RocFall 2D (v.4.0) software Figure 3.11 shows the process of defining the barrier location and the reliable barrier design Two types of barrier designs were chosen after analyzing the results of bounce height and kinetic energy using RocFall 2D (v.4.0) software Figure 3.11 The procedure for specifying the height and capacity of the barrier 50 3.4 Rock slope stability analysis based on Plaxis software In the slope stability analysis, it also collected source characterization, slope characterization, and recent historic events Additionally, hydrogeological characteristics data are considered to analyze the stability of the slope using Plaxis 2D (V21.01) software The surface water was found on the rock slope as a result of a field investigation; thus, the study considered the rainfall data not only for existing conditions but also for shotcrete and anchored mesh methods in the Plaxis model Moreover, Plaxis 2D (V21.01) software can calculate slope stability under rainfall conditions Figure 3.12 Simulation model procedures for rock slope stability analysis The numerical simulation for rock slope stability analysis is shown as a flow chart in Figure 3.12 Primary checked the stability of the slope before choosing which methods could prevent rock collapse failures This step is very important to consider the remediation measure for rockfall mitigation After that, the selection of 51 countermeasures for the mitigation works was done in accordance with the design, technology, and construction issues The geometry of the slope, the properties of the material for rock, and the existing retaining wall were defined in Plaxis 2D (V21.01) software to simulate the stability of the slope under normal condition Additionally, the material properties for each countermeasure, including rock bolt properties and shotcrete properties, were defined to calculate whether the slope of the rock is stable or unstable when applied for rockfall mitigation After remedial measures were installed on the slope of the rock surface, the software simulated the slope stability analysis The material model for this study was the Hoek-Brown model to analyze the stability of the rock slope in the investigated area This model is an isotropic, elastic, perfectly plastic model for weathered rock in accordance with the Hoek-Brown failure criterion It consists of the elastic parameters: Young’s modulus (E), Poisson’s ratio ( ), practical rock parameters such as uni-axial compressive strength ( ci), intact rock parameter (mi), geological strength index (GSI), and disturbance factor (D) The required input parameters for the Plaxis program were taken as referenced using the ROCKLAB program and from the 2012 laboratory test data, as summarized in Tables 3.3 and 3.4 The model considered the seepage for groundwater tables and high-water tables in the Plaxis Also, the model used infiltration in the boundary condition for considering the rainfall Figure 3.13 shows the monthly rainfall data in 2022 for the Son Tra Peninsula The maximum rainfall data is described in Table 3.5, and it is recorded from the metrological station (Suối Đá, Da Nang City) The model utilized a plate element for an existing retaining wall in Plaxis 2D (V21.01) software The values of the input parameter for the existing retaining wall in Plaxis are given in Table 3.6 Table 3.3 Input parameters for granite Parameter Material model Drained type Unsaturated unit weight, unsat Layer Layer Layer Hoek-Brown Hoek-Brown Hoek-Brown Drained Drained Drained 26 26 26 (kN/m ) 52 Parameter Layer Layer Layer 27 27 27 Poisson's ratio,  0.0232 0.0232 0.0232 Uni-axial compressive strength of 80*103 22*103 15*103 Saturated unit weight, sat (kN/m ) the intact rock, ci (kN/m2) Table 3.4 Input parameters for granite calculated in the RocLab program The rock mass Young's modulus , 6 16.4*10 2.7*10 1.4*10 Intact rock parameter, msi 32 31 30 Geological strength index, GSI 68 48 40 Disturbance factor, D 0.7 0.7 0.7 Rock mass parameter, mb 5.515 1.780 1.110 Rock mass parameter, s 0.0097 0.0005 0.0002 Rock mass parameter, a 0.502 0.507 0.511 Rock mass tensile strength, t 140.4 6.591 2.261 Erm (kN/m ) (kN/m2) 53 Figure 3.13 Monthly rainfall data in 2022 Table 3.5 Input parameters for rainfall Item Rainfall, q Model condition Value Unit Boundary 0.8 m/day Infiltration Table 3.6 Input parameters for retaining wall Parameter Value Elastic axial stiffness, EA1=EA2 (kN/m) 556*103 Elastic bending stiffness, EI (kN/m2/m) 465.3 The unit weigh of the plate, w (kN/m/m) 2.4 Poisson's ratio, ν 0.15 3.4.1 Slope stability analysis without proposed method (a) (b) (c) (d) Figure 3.14 The model without proposed method for Plaxis (a) Ground water table; (b) Ground water and rainfall; (c) High water table; (d) High water table and rainfall 54 Among the material models in the Plaxis 2D (V21.01) software, the study used the Hoek-Brown model to analyze the stability of the slope under without and with reinforcement conditions The Hoek-Brown failure criterion is a better non-linear approximation of the strength of the rocks (Material Models Manual 2D, 2023) The slope geometry with ground water level and high water level, rock material layer, and existing retaining wall are expressed as model Figure 3.14 The ground water and high water table in the model were taken as assumption table, as illustrated in Figure 3.14 3.4.2 Slope stability analysis with proposed methods When the countermeasures are selected to prevent rockfall problems, they are regarded as a requirement of the site condition, which means the space is sufficient or not sufficient to accommodate the collapsed mass The prevention method is to increase the stability of the rockface by increasing the resisting forces or decreasing the driving forces The slope stabilization method is one of the prevention methods, and it does not need to consider the space to construct the prevention structure for rock collapse cases In the slope stabilization method, there is an external stabilization method and an internal stabilization method According to the design and technology related to the construction case, the study chose the shotcrete method for external stabilization and the anchored mesh method for internal stabilization Moreover, the purpose of the study is to find an effective countermeasure for rockfall mitigation on the Hoang Sa road, thus shotcrete and anchored mesh were proposed to protect the highway against rockfall The rock slope stability was analyzed with two countermeasures: under the shotcrete method and under anchored mesh method Table 3.7 describes the parameters for retaining components in the Plaxis Table 3.7 Parameters of retaining components Rock bolt Diameter 25mm Material Steel SD400 55 Yielding strength > 400 N/mm2 Ultimate tensile strength > 560 N/mm2 Wire mesh Type 4T×50×50 Core diameter 3.16 mm Gross diameter 3.98 mm > 480 N/mm2 Ultimate tensile strength Shotcrete method (a) (c) (b) (d) Figure 3.15 The model with shotcrete method for Plaxis (a) Ground water table; (b) Ground water and rainfall; (c) High water table; (d) High water table and rainfall The structural shotcrete method was used to prevent the rock collapse failure problems in this study, and this method involves covering the rock with wire mesh and then spraying the concrete on the surface of the high-risk rockfall area Figure 3.15 illustrates the model with the slope geometry, rock material properties, existing 56 retaining structure, and shotcrete for the program The model considered the water table and rainfall to analyze the slope stability under the shotcrete method, as illustrated in Figure 3.15 In the Plaxis model, the ground water and high-water levels were taken as assumptions, as illustrated in Figure 3.15 The required data about shotcrete parameters for the RocFall model is shown in Table 3.8 Table 3.8 Input parameters for shotcrete Parameter Value Elastic axial stiffness, EA1=EA2 (kN/m) 2.1*106 Elastic bending stiffness, EI (kN/m2/m) 1750 The unit weigh of the plate, w (kN/m/m) 2.4 Poisson's ratio, ν 0.15 Anchored mesh method The required data about rock bolt parameters for the RocFall model is given in Table 3.9 For analysis of rock slope stability using the anchored mesh method, this method involves high-strength wire mesh and rock bolts, and it covers the rock surface as well as controls the stability of the slope The anchored mesh method is classified as the spacing and length of a rock bolt, with a m spacing and a m length of rock bolt based on the rock mass characteristics, such as the RMR value referenced in Intraratna & Kaiser, 1980 (Einstein, 2003) It can be seen that the slope profile, rock properties, retaining wall, and rock bolt properties are as shown in Figure 3.16 The model considered the water table and rainfall to analyze the slope stability under the anchored mesh method The ground water table and high-water table in the model were taken as assumptions, as described in Figure 3.16 Table 3.9 Input parameters for rock bolt Parameter Value Stiffness, E (kN/m2) 200*106 The unit weight of the material,  (kN/m3) 78 57 Parameter Value Diameter, d (m) 0.025 Axial skin resistance, Tskin(start) (kN/m2) 63 Axial skin resistance, Tskin(end) (kN/m2) 63 (a) (c) (b) (d) Figure 3.16 The model with anchored mesh method for Plaxis (a) Ground water table; (b) Ground water and rainfall; (c) High water table; (d) High water table and rainfall 58 CHAPTER DATA ANALYSIS RESULTS AND DISCUSSIONS This chapter presents the survey results from the field investigation and the location of the high-risk rockfall hazard area from UAV and DEM model This chapter illustrates the simulation results related to the rockfall mechanism analysis under the existing condition and the proposed countermeasure method In addition, it describes the analysis results that deal with the stability of the rock slope under existing condition and determines the stability condition of the rock slope when the shotcrete method and anchored mesh method are applied In conclusion, this chapter discussed the results of rockfall mechanism analysis and rock slope stability analysis without and with proposed countermeasures 4.1 Field data results 4.1.1 Field investigation results The field survey was performed to study the rock slope failure conditions and find the causal factors for rockfall problems along the Hoang Sa road The survey was conducted twice, in May and October of 2022, in the rockfall area along the Hoang Sa road It was done by taking photos and measuring manually According to the first field results, m3 of orphaned stones rolled down onto the road, damaged the steel mesh, broke the cable, and warped the steel poles, as shown in Figure 4.1 (a) Consequently, gabion walls and stone retaining nets were rusted and damaged quite badly The retaining wall was on top of a rolling stone screen with a height of m Although a retaining wall exists in this section, this wall has not obstructed the process by which fallen blocks continue to roll or slide onto the road In Figure 4.1 (b), it can be seen that some blocks detached from the surface of the rock; therefore, the remaining part of the rock structure was unstable and would later fall from the surface The field study found that there are many cracks on the surface due to the weathering effect, unstable block structures, and unstable blocks on the steep slope, as indicated in Figure 4.1 (b), (d), (e), (f), and (g) Additionally, tectonic fractures appeared on the granite rock surface, corresponding to the site results as described in Figure 4.1 (c) As a result, these causes tend to detach from the rock surface and become unstable for the rock 59 (b) (a) (c) (e) (d) Tectonic fracture Unstable blocks on the steep slope Unstable block Weathered rocks Fractures, cracks & joints (f) Fractures (g) Unstable block on the steep surface Strongly weathered rock surface Figure 4.1 Survey result photos (May, 2022) (a) (b) Figure 4.2 Survey result photos (October, 2022) 60 For the second investigation’s results, a 1.5 m high retaining wall structure was already constructed, and the slope of the mountain was very high According to the site investigation, the high-risk rockfall area had slightly weathered rock and an intrusive large formation and was normally granite, quite homogeneous lithologically Additionally, surface water was found on the rock slope surface within the rockfall prone area even though the weather was sunny, as shown in Figure 4.2 (b) According to the survey result, it can be seen that blocks detached from the rock slope surface and then scattered on the retaining wall and jumped onto the road surface, as expressed in Figure 4.2 (a) Many rock locations within the Hoang Sa road were unstable due to collisions and weathering Rockfall problems occurred due to heavy rain on October 19, 2022, and damaged the retaining wall, as illustrated in Figure 4.2 (b) Thus, the study area was necessary to apply long-term durable and stable measures for rockfall prevention works, according to the field investigation 4.1.2 UAV, DEM and GIS results The field survey was carried out to investigate rock slope failure locations along the Hoang Sa road by taking photos using an unmanned aerial vehicle (UAV) The study specified three high-risk rockfall-prone points, which are points 28, 29, and 30, depending on the site investigation using UAV and DEM models, as illustrated in Figures 4.3 (a) and (b) The location of serious rock slope failures among three points can be clearly seen in Figures 4.3 (b) and (c) Figure 4.3 (c) was taken from Google Earth Pro (a) 61 (b) (c) Failure surface High-risk rockfall area Figure 4.3 (a) DEM model from UAV pictures; (b) Imaged of rocks with a high-risk of rockfall taken by UAV; (c) Serious rockfall location point taken from google earth The high-risk rockfall area is situated on the Hoang Sa road to the Son Tra Peninsula The total length of the high-risk rockfall is about 700 m There are three points in this prone area: points 28, 29, and 30 Among these three points, the rockfall case in point 29 is a serious and significant one on the Hoang Sa road The starting point is 16° 06' 21.79" north latitude and 108° 18' 27.82" east longitude, and the ending point is 16° 62 06' 35.96" north latitude and 108° 18' 30.01" east longitude The location of point 29 is 16° 06' 26.20" north latitude and 108° 18' 29.05" east longitude The contour map along the Hoang Sa road is described in Figure 4.4 The ASTER Global Digital Elevation Model was downloaded from NASA Earth data, and after that, a GIS result figure was generated using ArcGIS software Figure 4.4 Overview of contour map in Son Tra Penninsula 4.2 Rockfall simulation results based on RocFall software The procedure for determining the location of the fallen blocks is mentioned in the Methodology section Position refers to the initial location of the fallen block that detached the slope surface or the source area of fallen block rock Positions to represent the highest to lowest slope height of fallen boulders from the slope toe, respectively The result of the rock failure mechanism without the proposed method is explained in Section 4.2.1 For rockfall mechanism analysis with the barrier method, the numerical result is expressed in Section 4.2.2 4.2.1 Simulation results of rockfall mechanisms without proposed method Figure 4.5 presents the assumed five locations in this study to simulate the rock endpoint and specify the location The rock parameters used in this analysis were m3 63 of volume and 2700 kg of rock mass based on past rockfall data The program calculated rock trajectories, kinetic energy, bounce height, and velocity along the slope profile (a) (b) (0,0) Position of stone (1) (5,4.5) Falling trajectories of rock Position of stone (2) Falling trajectories of rock Rock end point Existing Retaining wall Asphalt road Rock end point (c) Existing Retaining wall Asphalt road (d) (7.5,11.5) Falling trajectories of rock Position of stone (3) (10,17.8) Falling trajectories of rock Position of stone (4) Rock end point Rock end point Existing Retaining wall Asphalt road Existing Retaining wall Asphalt road (e) (12.5,23) Falling trajectories of rock Position of stone (5) Rock end point Existing Retaining wall Asphalt road Figure 4.5 Rock trajectories and end point results (a) Position of stone (1); (b) Position of stone (2); (c) Position of stone (3); (d) Position of stone (4); (e) Position of stone (5) In Figure 4.5(a), it can be shown that when the source area of fallen blocks is on a higher slope from the toe, the sliding distance of the blocks is greater In contrast, 64 when the block falls from the lower slope, which means the slope location of the fallen block is not so high from the toe of the slope, the stopping distance of the block due to the impact of the block is short, as can be seen in Figure 4.5 (e) According to the overview of the figure results, when the blocks detach anywhere on the surface of the rock slope as well as the slope high or low from the slope toe, the blocks roll over past the existing retaining wall and then stop on the road Figure 4.6 Bounce height results with different fallen rock location Figure 4.6 illustrates the bounce height results with varied fallen block locations It is obvious that the higher the slope height of the fallen rock from the toe, the higher the bounce height On the other hand, the lower the slope height, the lower the bounce height Generally, the highest bounce height has the maximum kinetic energy as well as the maximum velocity The bounce height parameter in RocFall software is one of the main factors in choosing the source area of a fallen block In Figure 4.6, it can be clearly seen that the position of the fallen stone (2) has one of the highest values among the other positions of fallen blocks Based on numerical results, the boulders detached at position on the surface of the slope Moreover, according to the site video recorded (Vietnamese News, 2022), position assumed fallen source area Hence, the study chose position for the fallen block’s location to simulate the behavior of rocks 65 The energy of rock boulders is lost after the first impact, causing the fragmentation of blocks into smaller pieces on impact Consequently, rock blocks are bouncing and rolling along their trajectories before stopping on the asphalt road As a result of the rock-fall analysis under existing conditions, rock blocks reached the asphalt road, causing damage to the road infrastructure and harm to commuters on the Hoang Sa road, as illustrated in Figure 4.7 The analysis results concluded that this area needs to mitigate the rock collapse failures and prevent future rockfall events Position of fallen stone Falling trajectories of rock Rock end point Existing Retaining wall Asphalt road Figure 4.7 Rockfall mechanisms analysis result without barrier method The rockfall mechanisms analysis evaluated bounce height, total kinetic energy, translational velocity, and run-out distance depending on the three rock mass types: 2700 kg (1 m3 of volume), 5400 kg (2 m3 of volume), and 8100 kg (3 m3 of volume), using RocFall software The program generated an output result about the phenomenon of rockfall mechanisms without the proposed method condition, as described in Table 4.1 It shows bounce height results and maximum translational velocity result values that change in accordance with each rock mass simulation in this program But the run-out distance value does not change even with varied rock masses In Table 4.1, the maximum total kinetic energy is a significant result compared with the other output results from the RocFall program Thus, the rock mass parameter affects the kinetic energy Table 4.1 Rockfall mechanisms analysis result with varied rock mass 66 Fallen rock mass (weight) Rock mass Rock mass Rock mass (2700 kg) (5400 kg) (8100 kg) Bounce height (m) 9.18 9.02 9.03 Max total kinetic energy (kJ) 736 1473 2210 Max translational velocity (m/s) 22.79 22.75 22.76 Run-out distance (m) 64.13 64.13 64.13 2500 Maximum Total Kinetic Energy (kJ) 2210 2000 1473 1500 1000 736 500 (2700kg) (5400kg) (8100kg) Rock mass Figure 4.8 Maximum total kinetic energy with varied rock mass Additionally, the study analyzed the behavior of the rockfall phenomenon based on three types of rock masses: 2700 kg, 5400 kg, and 8100 kg, respectively, in RocFall software According to Table 4.1 and Figure 4.8 results, it can be observed that the maximum kinetic energy is 736 kJ for 2700 kg of rock mass The maximum kinetic energy for 8100 kg of rock mass is 2210 kJ As a result, the behavior of the rockfall phenomenon demonstrates that the greater the weight of the fallen rock block, the greater the kinetic energy of the fallen block during the falling process along the slope profile Based on the total kinetic energy simulation, the results are 736 kJ, 1473 kJ, and 2210 kJ as shown in Table 4.1 These result values are within 100 kJ to 8500 kJ; therefore, the study selects the flexible barrier method in accordance with the calculation result of total kinetic energy as referenced in Design considerations for 67 passive protection structures, October 2016 (Rori Green, 2016) and the European Guideline ETAG 027 (EOTA 2008) (Peila & Ronco, 2009) Range of energy capacities for a variety of passive protection methods are described in section 3.3.2 4.2.2 Simulation results of rockfall mechanisms with proposed method Before the numerical simulation of rockfall mechanisms under the barrier method, the suitable place where a barrier should be installed on the surface of the slope was determined by analyzing the bounce height results According to the analysis result, the study chose two locations to install a barrier: one place had the highest bounce height among the other bounce height result values, and the second place had a higher bounce height than the other result values The proposed location of barrier option is 11 m in horizontal distance, as shown in Figure 4.9 The location for barrier option is proposed to be at 14 m of horizontal distance, as described in Figure 4.9 11 m 14 m Figure 4.9 Bounce height results The selection of barrier location was based on the bounce height analysis results of rockfall mechanisms without barrier condition The analysis results from RocFall were produced as a graph shown in Figure 4.9 The capacity and height of the barrier were specified as the design barriers to simulate the rockfall process, including rock run-out 68 paths, kinetic energy, bounce height, and velocity in the RocFall program, referring to the numerical simulation results The RocFall model was considered a flexible barrier type using high-strength steel wire mesh fencing Tables 4.2 and 4.3 present the proposed barrier designs (options and 2), which consist of the barrier height, installed location, and capacity of the barrier For proposed barrier option 1, a horizontal distance of 11 m and 19 m from slope toe was considered a barrier location in the RocFall model And then, the rock run-out path, rock end point, and sliding distance were calculated based on two barrier capacities and various barrier heights of m, m, m, and m The two barrier capacities were classified as 2000 kJ for 2700 kg of rock mass and 3000 kJ for 8100 kg of rock mass Barrier option proposed a design of 14 m of horizontal distance and 13 m from the toe of the slope, and barrier capacity and height were the same assumptions as option Table 4.2 Proposed barrier design (option 1) Designed barrier location Designed barrier height (m) 19 m (from slope toe) 3m 4m 5m 6m 2000 kJ (for rock mass= 2700 kg, volume=1 m3) Designed barrier capacity (kJ) 3000 kJ (for rock mass = 8100 kg, volume=3 m3) Table 4.3 Proposed barrier design (option 2) Designed barrier location Designed barrier height (m) 13 m (from slope toe) 3m 4m 5m 6m 2000 kJ (for rock mass= 2700 kg, volume=1 m3) Designed barrier capacity (kJ) 3000 kJ (for rock mass = 8100 kg, volume=3 m3) The rock trajectories and end points are based on the characteristics of the barrier height Moreover, barrier design parameters, particularly barrier height and barrier capacity parameters, in the RocFall program can be used to determine whether rock boulders are still moving or not at the proposed barrier They are critical factors when considering and examining suitable remedial measures against highway rockfall 69 Therefore, the study carried out the analysis by changing the values of each parameter, such as barrier height, location, and capacity, to recommend a suitable barrier design for Hoang Sa road rockfall problems (a) Barrier height=3 m Capacity=2000 kJ Barrier location=19 m (from toe slope) Barrier height=4 m Capacity=2000 kJ Barrier location=19 m (from toe slope) (b) Blocks rolled down Blocks rolled down (c) Barrier height=5 m Capacity=2000 kJ Barrier location=19 m (from toe slope) Blocks rolled down (d) Barrier height=6 m Capacity=2000 kJ Barrier location=19 m (from toe slope) Blocks stopped Figure 4.10 Rock mechanism analysis results of barrier option with 2000 kJ (a) m barrier height; (b) m barrier height; (c) m barrier height; (d) m barrier height The rockfall mechanisms analysis results of barrier option with 2000 kJ and 3000 kJ based on the different proposed barrier heights (3 m to m) are described in Figures 4.10 and 4.11 It can be seen that the proposed m, m, and m barrier height designs cannot prevent the movement of the blocks from the surface of the rock slope In contrast, it is obvious that the blocks cannot roll or move over the barrier when it chooses a m barrier height to install on the slope surface at 11 m of horizontal distance (19 m from the toe slope) To sum up the option result figure, only a m barrier height in both of the two proposed barrier capacities compared with other heights could be capable of preventing rock blocks from rolling on the terrain, as shown in Figures 4.10 and 4.11 70 (a) Barrier height=3 m Capacity=3000 kJ Barrier location=19 m (from toe slope) Blocks rolled down Blocks rolled down (c) Barrier height=5 m Capacity=3000 kJ Barrier location=19 m (from toe slope) Blocks rolled down Barrier height=4 m Capacity=3000 kJ Barrier location=19 m (from toe slope) (b) (d) Barrier height=6 m Capacity=3000 kJ Barrier location=19 m (from toe slope) Blocks stopped Figure 4.11 Rock mechanism analysis results of barrier option with 3000 kJ (a) m barrier height; (b) m barrier height; (c) m barrier height; (d) m barrier height Figure 4.12 illustrates the rock mechanism results for option with 2000 kJ and various barrier heights of m, m, m, and m Figure 4.13 shows the rock mechanism results for option with 3000 kJ and m, m, m, and m of barrier heights 71 (a) Barrier height=3 m Capacity=2000 kJ Barrier location=13 m (from toe slope) (b) Barrier height=4 m Capacity=2000 kJ Barrier location=13 m (from toe slope) Blocks rolled down Blocks rolled down (c) (d) Barrier height=6 m Capacity=2000 kJ Barrier location=13 m (from toe slope) Barrier height=5 m Capacity=2000 kJ Barrier location=13 m (from toe slope) Blocks stopped Blocks will Figure 4.12 Rock mechanism analysis results of barrier option with 2000 kJ (a) m barrier height; (b) m barrier height; (c) m barrier height; (d) m barrier height (a) Barrier height=3 m Capacity=3000 kJ Barrier location=13 m (from toe slope) Blocks rolled down (b) Barrier height=4 m Capacity=3000 kJ Barrier location=13 m (from toe slope) Blocks rolled down 72 (c) Barrier height=5 m Capacity=3000 kJ Barrier location=13 m (from toe slope) Blocks will roll (d) Barrier height=6 m Capacity=3000 kJ Barrier location=13 m (from toe slope) Blocks stopped Figure 4.13 Rock mechanism analysis results of barrier option with 3000 kJ (a) m barrier height; (b) m barrier height; (c) m barrier height; (d) m barrier height The proposed barrier option with a 2000 kJ result shows the barrier heights of m and m were not enough to protect blocks from rolling rocks The 3000 kJ barrier option result also indicates the m and m barrier heights are not safe designs for the prevention of rockfall In the two barrier capacities results of option 2, it is obvious that a m barrier height cannot fully catch the fallen rock blocks from the surface of the slope To conclude, a m barrier height is a reasonable height and can protect the blocks from rock bouncing or rolling, according to the barrier option result 4.3 Rock slope stability simulation results based on Plaxis software The stability or instability of the rock slope can be determined based on the factors of safety and failure surface in the Plaxis software, and the safety factor for slope stability is termed "Phi/c reduction’ The strength reduction method determines the stability of a slope, and the prediction of failure mechanisms depends on the finite element method In this study, the slope stability was simulated as a plane strain model with 15 node triangular elements and as a Hoek-Brown model for granite rock The boundary conditions were considered the default boundary in the Plaxis program There are two conditions, as follows: without the proposed method and with the proposed method (shotcrete and anchored mesh method) to examine the stability of the rock slope for the determination of suitable countermeasures in rock mitigation works 4.3.1 Simulation results of slope stability without proposed method The analysis was performed to investigate the slope stability under the existing condition The model was classified as the ground water table and the high-water table 73 in the Plaxis program Moreover, the model used rainfall infiltration (q) to simulate rock slope stability and the effect of rainfall Slope stability analysis results under existing condition (a) Ground water (b) Ground water and rainfall M =1.704 M =1.658 sf sf (c) High water (d) High water and rainfall M =1.391 sf M =1.340 sf Figure 4.14 Failure surfaces corresponding to FOS values for existing condition After the simulation ended, this program produced the output result for predicting the failure surface and the factor of safety, represented as an Msf value The failure surface and the safety factor value are expressed in the diagram shown in Figure 4.14 According to the simulation result of slope stability under existing condition, the value of the safety factor for ground water is the highest compared with the results of other conditions, which are ground water and rainfall, high water, and high water and rainfall But when starting with the rainfall data in the model, the factor of safety value is lower than when not considering the rainfall It can be seen that the high-water 74 rainfall condition has the lowest value among the other result values, as shown in Figure 4.14 4.3.2 Simulation results of slope stability with proposed methods Shotcrete method In this study, the proposed shotcrete method as a solution for rockfall prevention work on the Hoang Sa road is also taken into account for four conditions that deal with rainfall: ground water, ground water and rainfall, high water, and high water and rainfall, to construct the model in Plaxis software Figure 4.15 presents the failure surfaces in accordance with safety factor values under the shotcrete method According to the comparison of the factor of safety value for the without shotcrete method and the with shotcrete method, the factor of safety value for the without shotcrete method is less than the result value for the safety factor when the shotcrete method is applied on the slope surface The result from the numerical simulation concluded that the shotcrete method, among the slope stabilization methods, is one of the solutions for the Hoang Sa highway against rockfalls From the water level and rainfall point of view in the simulation model, the value of considering ground water is greater than that of considering high water levels, and ground water and rainfall also have greater value than high water and rainfall Slope stability analysis results under applied shotcrete method (a) Ground water (b) Ground water and rainfall M =1.793 M =1.747 sf (c) High water sf (d) High water and rainfall 75 M =1.484 M =1.501 sf sf Figure 4.15 Failure surfaces corresponding to FOS values under used shotcrete method Anchored mesh method The study proposed the anchored mesh method as a remedial measure, which is designed with m spacing and a m length of rock bolt The Plaxis program simulated the prediction of failure surfaces and factors of safety for stability The numerical simulation results of failure surface and safety factor values are illustrated and summarized in Figure 4.16 Slope stability analysis results under applied anchored mesh method (a) Ground water (b) Ground water and rainfall M =2.123 M =2.050 sf (c) High water sf (d) High water and rainfall M =1.647 sf 76 M =1.637 sf Figure 4.16 Failure surfaces corresponding to FOS values under used anchored mesh method The study proposed the anchored mesh method as a remedial measure, which is designed with m spacing and a m length of rock bolt The Plaxis program simulated the prediction of failure surfaces and factors of safety for stability The numerical simulation results of failure surface and safety factor values are illustrated and summarized in Figure 4.16 The numerical modeling was based on four conditions of water level and rainfall The consideration of rainfall and water table was similar to the conditions above mentioned in the existing condition and shotcrete method In accordance with the overall result in Figures 4.16, high water tables and rainfall infiltration are more affected than ground water tables and rainfall in the model for analyzing the slope stability of rock according to the factor of safety results The Msf value for groundwater, including surface runoff water, is significantly higher than the high water tables and rainfall It is concluded that the consideration of rainfall in the model has a significant impact on the calculation of the stability of rock slopes 4.4 Discussions The rockfall area in this study is 700 m long and 50 m high at 80 degrees of the overall face angle and is composed of granite rock that belongs to the Hai Van complex (γaT3hv1) Based on the field survey results, it is a highly active rockfall area due to geological and geomorphological factors Past rockfall events on the Hoang Sa road are due to highly fractured rocks, jointed hard rocks, weather rocks, cracks, and water flow appearing on the surface of the granite rock slope These factors cause the slope 77 of the terrain to be unstable, and rock falls would occur, leading to hazardous rockfall problems Several case studies investigated the factors contributing to the failure mechanism of rockfall, and they found that the geological condition, differential weathering, and high rainfall intensity were the main factors leading to rockfall events (Arbanas et al., 2012; Budhbhatti et al., 2016; Spang, 1996; Verma et al., 2019) In Croatia, rockfalls occurred due to unfavorable rock mass characteristics on the limestone slopes during the heavy rainy season (Arbanas et al., 2012) The causes of rockfall at Saptshrungi Gad Temple were weathering action, fractures, and joints on the surface of a basaltic steep slope (Budhbhatti et al., 2016) Due to opening joints and differential weathering during a rainy period, massive sandstone blocks lying on the shale bedding fell down on the highway NH-44A in India (Verma et al., 2019) For rockfall mechanisms analysis The result of the rockfall mechanism analysis without a barrier from the RocFall program expresses that an existing retaining wall is not an effective solution, and the study area needs a new structural countermeasure to prevent rock slope failures From the simulation result, the rock boulders detach from the surface of the cliff, bounce, and then stop on the roadway In its actual condition, a fallen rock boulder was found on the pavement road Thus, the numerical simulation result is similar to the field observation The kinetic energy value of fallen blocks in the output result of the RocFall software is 2210 kJ The result value is in the range of 100 kJ to 8500 kJ; therefore, the study selected a flexible barrier according to the European Guideline ETAG 027 (EOTA 2008) (Peila & Ronco, 2009) Budhbhatti et al., 2016 proposed flexible barrier based on the referenced the European Guideline ETAG 027 (EOTA 2008) when selecting the rockfall mitigation measures for the rockfall problem at Saptshrungi Gad temple According to the JICA and FHWA statements, a flexible barrier system is very easy and rapid to install and maintain, and the structure is deformable, lightweight, high quality, and long-lasting Furthermore, this method can be utilized on very steep slopes for rockfall prevention (N.C Koei, 2007, Richard Andrew et al., 2011) Thus, the study chose a flexible barrier among the passive methods as a new proposed countermeasure to prevent rockfalls along the Hoang Sa road 78 (a) (b) Barrier height=6m Barrier Capacity = 3000 kJ Barrier location=19 m (from toe slope) Barrier inclination= 60 degree Barrier height=6m Barrier Capacity = 3000 kJ Barrier location=13 m (from toe slope) Barrier inclination = 60 degree Figure 4.17 Comparison results (a) Barrier option 1; (b) Barrier option The study was carried out using parametric analysis to specify where the fence should be placed and the required barrier height and capacity using the RocFall program By changing the input parameters, barrier capacities, barrier locations, and barrier height, a suitable barrier design was determined Many researchers also investigated the proposed suitable barrier design, which consists of barrier location, height, and capacity based on the rockfall simulation results of RocFall software (Spang, 1996, Arbanas et al., 2012, Budhbhatti et al., 2016, Erfen & Musta, 2022) Figure 4.17 illustrates the rockfall mechanism analysis results of both proposed barrier designs (option and option 2) Barrier option (1) is placed on the surface of the steep rock slope (19 m from the toe), while barrier option (2) is installed on the lower slope surface (13 m from the toe) As can be seen in Figure 4.17, the results show that the m barrier height and 3000 kJ barrier capacity of both options can prevent rock fall phenomena and future rock fall problems However, the location of barrier option is nearer the toe of the slope than barrier option Table 4.4 presents the maximum total kinetic energy results without and with barriers (including barrier design options and 2) The maximum total kinetic energy of a fallen rock block was calculated based on the consideration of 8100 kg of rock mass (3 m3 volume) Compared with the kinetic energy results with and without barrier conditions, the maximum total kinetic result value without barrier condition is higher than the results with barrier condition Among the two barrier design options, the kinetic energy result value of option is lower than that of option Table 4.4 Maximum total kinetic energy results without and with barrier 79 With barrier Maximum total kinetic energy Without (option-1) (option-2) barrier Barrier location Barrier location 19 m (from the slope toe) 13 m (from the slope toe) 554 kJ 381 kJ 2210 kJ In the model, the barrier structure location can be examined based on the maximum energy of fallen block Budhbhatti et al (2016), Gnyawali et al (2015), and Erfen & Musta (2022) found that the energy of fallen blocks is reduced more due to gravity as they get closer to the toe of the slope Due to the reduced energy impact of fallen rock on the barrier, the proposed barrier where it is near the toe slope is more sufficient to prevent rockfall and withstand the accumulation of fallen boulders than the barrier at the steep slope From a construction point of view, the geometry in the study area is steep, thus the barrier should be constructed near the toe slope to be installed more easily, have enough space to construct, and save on construction costs Hence, proposed barrier option is an appropriate installation location and design according to the numerical result and construction issue to prevent rock slope failures on the Hoang Sa road For rock slope stability analysis The study area proposed shotcrete and the anchored mesh method, as active methods against highway rockfalls based on the consideration of site condition, technical, service life, and construction cost and time The Plaxis model calculated the failure surface and safety factor of slope stability with both of the proposed methods DuongNgo et al (2020) and Fahimifar & Ghadami (2013) investigated the rock fall event calculating the stability of the rock slope by using Plaxis software Son Tra Peninsula is a monsoon area, and surface water flows along the slope as a result of field investigation Therefore, the rainfall and water table were included in the model simulation Figure 4.18 illustrates the factor of safety with or without reinforcement methods For the numerical analysis of rock slope stability, it is indicated that the 80 factor of safety value without reinforcement is less than that with reinforcement methods On the other hand, the results prove that the anchored mesh method can effectively prevent rock collapse failure by stabilizing the rock slope than shotcrete method 2.5 2.123 GL GL & Rainfall HL HL & Rainfall 2.05 2.0 1.793 1.747 Factor of safety 1.704 1.658 1.5 1.6471.637 1.5011.484 1.391 1.34 1.0 0.5 0.0 Normal condition Shotcrete Anchored mesh Figure 4.18 Factor of safety results without or with reinforcement methods As shown in Figure 4.18, the consideration of rainfall in the model affects rock slope stability both normal condition and with proposed methods because the factor of safety result value involving rainfall is less than the result value not involving rainfall in the model Duong-Ngo et al (2020) and Fahimifar & Ghadami (2013) found that rainfall intensity and rising groundwater tables during the rainy season have an impact on the stability of the slope Fahimifar & Ghadami (2013) suggested that to increase the factor of safety for stability, a drain system should be installed Duong-Ngo et al (2020) recommended a retaining system of rock bolts, wire mesh, and an anchored wall for rockfall events and Fahimifar & Ghadami (2013) mentioned that the rock reinforcement with geogrid box method is based on the numerical result of rock slope stability with Plaxis software for prevention of rockfalls For practical field, the anchored mesh method can protect and have a more durable effect under rainfall conditions, based on the JICA report Wyllie and Norrish (1996) 81 and Maerz (2015) revealed that the shotcrete method can provide little support against sliding of the overall slope Anchored mesh can provide more support for small to medium-sized loose rock blocks in terms of slope stability (Maerz et al., 2015; Gnyawali et al., 2015) In the study area (Hoang Sa Road), the slope geometry is steep, it faces a differential weathering effect and surface water, and there are many fractures and joints on the surface of the slope Hence, the anchored mesh method should be applied to prevent future rockfall events on the Hoang Sa road because the factor of safety for the anchored mesh method under rainfall conditions is greater than the shotcrete safety factor as a result of slope stability, and this method has a more durable effect under rainfall conditions Moreover, the safety factor of the anchored mesh system is higher than the standard stability value based on the classification of roads taken from the Vietnamese standard TCVN 13346:2021 CHAPTER CONCLUSIONS AND RECOMMENDATIONS 5.1 Conclusions The study focuses on the rockfall mechanisms and mitigation measures for preventing rockfall on the Hoang Sa road leading to the Son Tra peninsula in Da Nang city This study can contribute for the selection of remedial measures to prevent future rock fall events along the highway The study investigated the behavior of rockfall mechanisms and rock slope stability without the proposed countermeasure and with the proposed countermeasure 82 For first objective According to the field investigation, the study found that granite rock with the steep slope geometry combined with weathered rocks, a highly fractured and cracked surface of the rock slope, and surface water caused the past rockfall cases in the Son Tra area These factors are making the slope unstable and creating future rockfalls in this area Moreover, the numerical result of rockfall mechanism analysis without the proposed method using RocFall software indicates that the existing countermeasures are not effective in preventing the rockfall process and that suitable structural countermeasures are needed against highway rockfall in this area For second objective The numerical simulations of the proposed method and design are carried out based on 8100 kg of rock mass (3 m3 rock blocks) falling from a 30–50 m high granite rock slope surface Based on the rockfall mechanism analysis results from RocFall software, the study recommends a flexible barrier and barrier design option with m of barrier height, 3000 kJ of barrier capacity, 60 degree of barrier inclination, and 13 m of barrier location from the pavement level, which is an effective passive method and design for rockfall mitigation in the Son Tra area On the other hand, the flexible barrier method can prevent boulders from falling with high speed and velocity from the cliff and, consequently, reduce the risk of damage to the traveler and moving vehicles Furthermore, on the construction site, the barrier can be constructed easily and rapidly because the recommended barrier option location is closer to the toe slope Rock slope stability analysis was carried out using the shotcrete method and the anchored mesh method with the Plaxis program According to the rock slope stability analysis, the factor of safety value for anchored mesh under normal and rainfall conditions is higher than the standard safety factor Besides, in both conditions, it is greater than that of the shotcrete method The anchored mesh method can prevent rock slope failures in features with steep slopes, joints, and cracks in the rock mass, in accordance with numerical results Thus, the study recommends the anchored mesh method, which included m spacing and m length of rock bolt and high-strength wire mesh, as a reasonable solution among the active methods for the study area 83 5.2 Limitations There is limited study in the thesis, as follows:  The study does not take rainfall conditions into account to simulate the rockfall mechanism analysis in the RocFall software This software does not consider rainfall levels in the simulation of the rock fall process 5.3 Recommendations The followings are some recommendations from this study  The numerical result from Plaxis showed that the rainfall intensity affects the rock slope stability Thus, the study recommends surface drains be installed, and the drainage system should be maintained frequently on the construction site  The study suggests the flexible barrier method should be chosen if the space is large enough to accommodate the collapsed mass on the construction site In contrast, if space is limited in the field, the anchored mesh method should be used instead of the flexible barrier method  Due to the limited geological and geotechnical properties of soil and rock in this area, a deep field investigation and laboratory tests should be performed as a future study  The study should carry out the barrier sensitivity analysis related to barrier inclination using the RocFall 3D model for further study 84 RFERENCES Agliardi, F., & Crosta, G B (2003) High resolution three-dimensional numerical modelling of rockfalls International Journal of Rock Mechanics and Mining Sciences, 40(4), 455–471 Business, M (2016) Rockfall: Design considerations for passive protection structures Beers, D (2004) Design and Implementation of Rockfall Mitigation Measures for DeCew Falls Hydroelectric Station Mine Closure-Geotechnical Aspects View Project Geotechnical Data Management View project Phil de Graaf Botha, A (2017) Creating an Engineering Modelling Dorren, L K A (2003) A review of rockfall mechanics and modelling approaches Progress in Physical Geography, 27(1), 69–87 Duong-Ngo, H., Do-Tuan, N., & Nguyen-Chau, L (2020) Retaining System Against Rockfall Near a Highway in Vietnam—A Case Study In Lecture Notes in Civil Engineering vol 62, (Springer) pp 865–869 Early handling of rockfalls on Son Tra Peninsula taken (2022) DaNangToDay Einstein, A (2003) Imagination is more important than knowledge Hancock, M (2019, December 23) Major Rock Fall closes Norwegian motorway Ground Engineering (GE) Jaccard, C J., Abbruzzese, J M., & Howald, E P (2020) An evaluation of the performance of rock fall protection measures and their role in hazard zoning Natural Hazards, 104(1), 459–491 Jiang, G., Feng, Z., Zhao, R., Wang, F., Yu, X., Wu, M., & Zhang, Z (2021) Case Study on Safety Assessment of Rockfall and Splash Stone Protective Structures for Secondary Excavation of Highway Slope Advances in Civil Engineering, 2021 Kakavas, M P., & Nikolakopoulos, K G (2021) Digital elevation models of rockfalls and landslides: A review and meta-analysis In Geosciences (Switzerland) vol 11, Issue (MDPI AG) Koei, N (2007) The Study on Risk Management for Sediment-Related Disaster on Final Report Guide III Selected National Highways in the Republic of the Philippines Road Slope Protection Guide III Road Slope Protection Landslide warning along roads leading to Son Tra Peninsula (2020) DaNangToDay Liniger, M., & Bieri, D (2006) A2, Gotthardautobahn, Felssturz Gurtnellen vom 31 Mai 2006, Beurteilung und Massnahmen 153(2006), 81–86 Material Models Manual 2D (2023) M., & Al E, M S (2017) Scholars’ Mine Scholars’ Mine Assessment of rockfall rollout risk along varying slope Assessment of rockfall rollout risk along varying slope geometries using the Rocfall and CRSP software geometries using the Rocfall and CRSP software News, V (2018) Large rock fall kills driver in Lai Chau VietNamNet News 85 News, V (2021) Large rock fall kills driver in Lai Chau VietNamNet News NGUYEN DUC NAM (2018, January) Landsides risk warning for travellers alongside roads to Son Tra Peninsula DaNangToDay Palma, B., Parise, M., Reichenbach, P., & Guzzetti, F (2012) Rockfall hazard assessment along a road in the Sorrento Peninsula, Campania, southern Italy Natural Hazards, 61(1), 187–201 PLAXIS CONNECT Edition V21.01 PLAXIS 2D-Tutorial Manual Peila, D., & Ronco, C (2009) Natural Hazards and Earth System Sciences Technical Note: Design of rockfall net fences and the new ETAG 027 European guideline In Hazards Earth Syst Sci vol.9 Qi, X., Yu, Z X., Zhao, L., Xu, H., & Zhao, S C (2018) A new numerical modelling approach for flexible rockfall protection barriers based on failure modes Advanced Steel Construction, 14(3), 479–495 Richard Andrew, A D., Bartingale, R., Hume, H., Haramy, K., DeMarco, M., Surdahl, R., Mohamed, K., Barrows, R., Anderson, S., Siel, B., Greer, M., & Ortiz, T (2011) Context Sensitive Rock Slope Design Solutions 13 Type of Report and Period Covered Final Report Rock Slope Engineering (2005) Stevens, W D (1998) Rocfall: A Tool for Probabilistic Analysis, Design of Remedial Measures and Prediction of Rockfalls TCVN 13346:2021 The Landslide Prevention Engineering on Road-Requirements for Investigation and Design The Study on Integrated Development Strategyfor Danang City and Its Neighboring Area in the Socialist Republic of Vietnam (DaCRISS) (2010) Volkwein, A., Schellenberg, K., Labiouse, V., Agliardi, F., Berger, F., Bourrier, F., Dorren, L K A., Gerber, W., Jaboyedoff, M., & Bourrier, F (2011) Rockfall characterization and structural protection-a review Natural Hazards and Earth System Sciences Rockfall characterisation and structural protection-a review Hazards Earth Syst Sci, 11, 2617–2651 VnExpress (2022, October 30) Da Nang peninsula grapples with unprecedented erosion damage VnExpress International – Latest News, Business, Travel and Analysis from Vietnam Wei, L W., Chen, H., Lee, C F., Huang, W K., Lin, M L., Chi, C C., & Lin, H H (2014) The mechanism of rockfall disaster: A case study from Badouzih, Keelung, in northern Taiwan Engineering Geology, 183, 116–126 Wyllie, D C (2014) Rock Fall Engineering: Development and Calibration of an Improved Model for Analysis of Rock Fall Hazards on Highways and Railways 86 APPENDIX Appendix 1: Drawing and construction of passive method (flexible barrier) (a) 6m (b) 87 (c) Figure (a) Schematic of flexible barrier system; (b) Barrier side view; (c) Scheme of maximum post drop front view The following is the sequence of construction of a flexible barrier using a highstrength falling rock barrier system  Clean up the surface of rock slope and locate the anchor hole  Drill holes D42 for steel anchor row and D76 for border cable anchor row  Spreading and bonding of wire mesh rolls  Drilling hole D76 on the slope surface  Clean and inspect boreholes  Anchor bar installation  Grouting the drill hole, plugging the anchor steel bar and installing the metal base plate  Anchoring the boundary cable and tensioning the boundary cable to strengthen the rock barrier  Finishing 88 Appendix 2: Drawing and construction of active method (anchored mesh) (a) (b) (c) Figure (a) Schematic of anchored mesh system; (b) Detail of rock bolt; (c) Detail of high strength wire mesh 89 The following is the sequence of construction of anchored mesh method using rock bolt and high strength wire mesh  Drilling to create holes  Installing ground nails (anchor pins)  Installing wire mesh B40  Installing the spacer and bolt head Appendix 3: Table Criteria for slope stability mitigation measures, FHWA-CFL/TD-11-002 (Richard Andrew et al., 2011) Complexity Effectiveness Durability Constructability Cost Maintenance requirement CRITERIA Hand/ Mechanical scaling L-M L-H L-M M L-M L-M Trim Blasting L-H L-H M-H L-H L-H L-M Shear Pins M M M M M L Shotcrete M-H M M H M L M M-H M-H M L-H M-H L L-H M L L H MITIGATION MEASURE STABILIZATION METHODS Excavation Reinforcement Wire mesh (anchored) Drainage Weep Drains PROTECTION METHODS 90 Barrier and Fences Earthen Berms L M-H H L L M-H Concrete Barriers L M L-M L L M-H Structural Walls L-M M-H M M L-M M-H Flexible barrier M-H M-H M-H M M M-H Ditches L M-H H L L-H H L= low, M= medium, H= high, VH= very high, N= no, Y= yes, P= possibly Table Types of mitigation methods (JICA final report, 2007) (N.C Koei, 2007) Type of work Durability Maintenance Construction Construction Degree ease cost of safety Rockfall prevention work Rockfall protection work Removal ◎ ◎ △ ○ ◎ Cutting ○ ○ ○ ○ ○ Drainage ditches ○ ○ ○ ◎ ○ Shotcrete ○ ◎ ◎ ◎ ○ Anchored mesh ◎ ◎ △ ○ ◎ Ground anchors ◎ ◎ △ △ ◎ Catch fill and ditches ○ ○ ◎ ○ △ Catch net (flexible barrier) ◎ ◎ ◎ ○ ◎ Rock sheds ◎ ◎ △ △ ◎ ◎ = Very good or very easy, ○ = Good or easy, △ = Good or easy in some cases 91