1 Introduction In order to consider the presence of cracks in an abandoned gypsum pillar in numerical simulations, a hybrid method FEM/DEM, which allows the transition from continuum to discontinuum, was assumed. By means of a specific numerical code (ELFEN), this approach has been calibrated involving both physical quantities introduced by fracture mechanics and numerical aspects in order to support this hybrid method. Furthermore, the approaches FEM and FEM/DEM have been compared, showing advantages and disadvantages through experimental tests carried out to characterize geomechanical response of the pillar. The interaction domain has been calculated thanks to the implementation of both methods. The meaning of determining this domain is related to the evaluation of failure limit when a coupled system of loads (normal and tangential force and momentum) is acting on pillars. An application to a case study of an abandoned gypsum mine interacting with building in San Lazzaro di Savena (Bologna) is shown. 2 Abandoned Gypsum Mine – specific pillar (P7) chosen The choice of this pillar is related to its complex and redundant joint's system, its reduced section and its location. In fact, it was possible to obtain its topography by TLS to furnish a detailed geometrical for the simulation. In the following figure the mine system and the selected pillar is shown. 3 Laboratory tests (UCS, BT, Triax) A series of laboratory tests permitted to obtain failure parameters to be inserted both in constitutive model for the continuum approach (FEM) than for the crack propagation model for the discontinuum approach (FEM/DEM). 4 A numerical comparison between the two approaches A coupled system of loads acting on pillar P7 has been simulated. In particular, 3D numerical simulations have been run to calculate an interaction domain represented by normal and tangential force and momentum. Also for the 2D FEM/DEM approach an interaction domain has been built thanks to a simplified solution by choosing two representative sections of pillar P7. In the following the numerical calibration of the continuum and hybrid approach is shown.

1 Introduction Natural temperature variations are thought to play a role in rock slope instabilities. Thermal and and geotechnical in situ measurements have been carried out in sites were the succession of thermomechanical effects are suspected to have led progressively to failure. Parallel laboratory tests show the damage of specimens subjected to thermal fatigue. Site and laboratory observations are briefly described below and elements of interpretation are proposed. 2 In situ observations In situ observations accredited the role of temperature variations in the progressive destabilization of rock slopes. We refer below to 2 French sites, but Vargas et al. (2004, 2009) made similar observation in Brazil. 2.1 A gneissic rock mass For several years, thermal and geotechnical in situ measurements have been collected at the "Rochers de Valabres" site (Southern French Alps), a gneissic rockmass, to examine surface thermomechanical phenomena and their potential contribution to rock fall triggering. In fact, in May 2000 (Fig. 1) the site experiments a 2000 m3 rock fall during a period with any particular rainfall event, seismic activity, or freeze-thaw period and we suspect it could be linked to the repetition of thermal variation (Gunzburger et al. 2005). The site was monitored with different equipment to record the temperature variation in the rock-mass and the potential induced effect. The measured data were analyzed based on analytical solutions and numerical modelling. 2.2 A limestone cliff At "La Roque Gageac", a village located near Bergerac at the right bank of the Dordogne, a cliff composed of sandy limestone was subjected to rupture accredited to temperature fluctuations. The site experienced 4 rock falls since 1920, which resulted in three deaths. Among the potential causes for these rock falls, temperature variations were considered as one of the possible preparatory factor and the site is monitored with temperature gauges and strain measurement and under study.

1 Introduction Since linear elastic fracture mechanics (LEFM) was derived from the work of Griffith (1921), it has been assumed that the stored elastic strain energy that is involved in extending a crack in a brittle material becomes potential surface energy associated with the new surface as it is created. Thus the area of new surface that can be created in a fracture episode should be limited by the amount elastic strain energy input before the fragmentation episode. We conducted laboratory and field (rock avalanche) measurements to test this assumption and consistently found more fracture surface area than the assumption would predict. Nevertheless, this result does not affect conventional LEFM because it is unaffected by the fate of the elastic strain energy in brittle failure. We see evidence that the elastic strain energy that causes cracks to propagate in individual grains becomes transient elastic strain, and influences further breakage. This result has broad and growing significance as developing technologies allow ever finer fragments to be detected and measured in brittle fracture. 2 Borosilicate glas fragmentation Kolzenberg et al. (2013) report dry confined compression tests on borosilicate glass (Pyrex), presenting the elastic strain energy released at failure and total surface area of the fragmented samples. We have carried out unconfined compression tests on dry 10mm diameter and 20 mm long borosilicate glass (Pyrex) cylinders; four tests yielded fragments with estimated total surface area up to 91.8 m2. Since the specific surface energy of borosilicate glass is established as 4.5 Jm −2 (Wiederhorn 1969), this would require about 413 J, but the elastic strain energy available at failure was about 133 J.

1 Introduction The stability of fractured rock slopes is strongly related to the geometrical and mechanical characteristics of natural discontinuities, which concentrate stresses at their tips giving rise to progressive failure of rock bridges. Fracture mechanics makes it possible to take this phenomenon into account assuming discontinuities as cracks and studying their triggering and propagation inside rock bridges. As an example, a Linear Elastic Fracture Mechanics approach to the back analysis of a rockfall triggering phenomenon is presented in this paper, carried out through the code GEOF (Scavia 1995; Castelli 1998), based on the BEM technique of the Displacement Discontinuity Method (DDM, Crouch and Starfield 1983). 2 Description of the rockfall and definition of the geometrical model The simulated instability occurred in 2004 from an approximately vertical slope made of Urgonian limestone and located in the Vercors Sub-alpine Chain (South-western France). After the failure a detailed geological and structural survey has been carried out by Frayssines (2005), leading to a complete description of the detaching niche (scar) and the surrounding slope. The main part of the scar is limited by (Figs. 1a and 1b): a mean surface (A) roughly parallel to the front, a bedding joint at the top (B), a lateral sub-vertical surface on the western side (C). The existence of fresh rock zones was observed and measured during the survey in the lower part of surface (A). This zones (52 m 2 ) correspond to the 5% of the total surface and don't show any sign of alteration, in contrast with the rest of the scar, which is coated with a calcite crust. A detailed digital model shows that the cliff surface presented a prominent overhang before the failure. The failure mechanism can then be a topple, due to tensile fracturing of intact rock zones. The unstable volume is characterized by a prismatic shape with average height (H) around 20 m, width (w) around 2.5 m and length (L) around 50 m, as shown in Fig. 1c.

1 Abstract Our understanding of the nature of progressive failure of rockfall is in part limited by the temporal and spatial resolution, and coverage of monitoring data. We present a uniquely high-resolution 4D dataset that captures over 180,000 rockfall, largely < 1m 3 . Of these rockfall, only a small portion show evidence of pre-failure deformation. We use our data to look in detail at the nature of deformation during the week prior to failure. We see multiple phases of movement that relate to weather conditions, in addition to pulses of movement immediately prior to collapse, which we hypothesize relate to the underlying progressive failure mechanism controlling the release of the rockfall. 2 Introduction A series of studies using repeat surveys of actively failing rock slopes have revealed that rockfalls rarely occur as instantaneous events, and are often preceded by deformation or sequences of precursors (e.g. Rosser et al. 2007). It has been suggested that the behavior of a slope prior to failure reflects, at least in part, the underlying mechanisms, including progressive failure. For example, Royán et al. (2015) identify an accelerating pattern of small-scale movements prior to a larger rockfall, which the authors attribute to a progressive mode of failure. Similarly, Rosser et al. (2013) identified an increasing rate of increasingly large rockfalls preceding subsequent rock cliff collapse. Whilst these studies offer insight, they are limited by the frequency of data collection which is normally collected over intervals longer than the duration of changes, or they rely on fortuitously capturing precursors. To address this, here we present a uniquely long-term, high-resolution (space and time) monitoring dataset, captured using repeat laser scans over 9-months. 3 Methods Our dataset comprises c . 9,880 scans captured at < 1 hr intervals. Each scan contains c. 1.8m points at point spacing ≈ 0.05 m. The scanning system is a Riegl vz1000 with a controls system that enables sequential scans of the monitored rockface (area = 9,592 m 2 ). Data is combined with contemporaneous weather conditions to assess controls on deformation. Data is processed using a set of algorithms developed specifically for optimizing the detection of fine-scale change in large archives of point cloud data.

1 Introduction Cycles of glaciation drive in-situ stress changes in underlying bedrock as glaciers advance, erode, and retreat, generating damage in adjacent rock slopes and influencing paraglacial slope stability. Glacial debuttressing is frequently implicated as a trigger for paraglacial rock slope failures, despite commonly observed large lag-times between deglaciation and the timing of failure and often without clear mechanical reasoning. Rock slope damage generated during glacial cycles is hypothesized to have a strong role in preparing rock slope failures, however, the mechanics of paraglacial rock slope damage remain poorly characterized. Glacial cycles mechanically load and unload proximal rock slopes by the changing weight of ice, and in addition produce strongly varying thermal and hydraulic rock-surface boundary conditions tied to the fluctuating glacier (Fig. 1). Bedrock beneath temperate glacier ice maintains near isothermal surface temperatures at ~0 °C. Glacier retreat exposes rock walls to new thermal boundary conditions with strongly varying daily and seasonal cycles, a transition we term paraglacial thermal shock . Temperature changes generate thermal strain, inducing thermo-mechanical stresses capable of generating rock mass damage. In addition, high subglacial water pressures near the ice overburden level prevail at the base of temperate glaciers, and affect groundwater conditions in proximal valley flanks. Groundwater recharge by precipitation and snowmelt raises the water table seasonally, which is superposed on changes in hillslope groundwater tied to varying glacial ice elevations. Changing cleft water pressures control effective stresses and the strength of rock mass discontinuities. Together, these thermo-hydro-mechanical stresses act in concert with glacial loading cycles to generate rock slope damage, preparing slopes for future failure.

1. Introduction 1.1 Main challenge and purpose of the abstract After finalizing the excavation and support installation in the main caverns of the Uma Oya MPDP (Multi-Purpose Development Project), cracks in the shotcrete lining of the caverns and in the connecting galleries occurred. This happened during the ongoing work on the inner concrete structure of the caverns. The designers had to give proof to the client that these events will not endanger the integrity of the structures during construction and service life. As one of the main hazard scenarios potential buckling failure along pre-defined rock slabs in the cavern's side walls was recognized. The potential slabs were formed by the specific discontinuity pattern with sub-vertical joints almost in parallel to the caverns. In case of buckling failure, additional load would have been transferred into the rock pillar between the caverns. The decisions taken on site during construction were based on long-term monitoring data of which actual examples can be presented. The focus is on readings from load cells on anchors, extensometers and crack meters. 1.2. General outline of the project The Uma Oya MPDP in Uva province of Sri Lanka is a classical hydro power plant with the additional benefit conveying water from the central mountain region to the dry south part of the island. The scheme consists of two catchment areas confined by each a dam. These areas are linked via a low pressure conveyance tunnel of 3.9 km length. A 15.3 km long headrace tunnel starting from the Dyraaba dam guides the design discharge of 19.5 m 3 /s past a surge shaft, a valve chamber, a vertical pressure shaft with a length of 630 m, an underground powerhouse with an adjacent transformer cavern both bearing an overburden of almost 700 m and a tailrace tunnel of 3.7 km length.

1 Introduction Rockslides are one of the most hazardous natural events, because of their velocity and rarely detectable precursors. Possible strategies to prevent the risk associated to these events consist in: a) studying the gradual creep-driven deformation of rock mass by setting a proper monitoring system (Loew et al. 2016); b) realizing numerical models also considering progressive failure of rock slopes (Eberhardt et al. 2004); c) coupling microseismic monitoring systems and numerical models, taking in account rock mass damage during time (Xu et al. 2014; Tang et al. 2015); d) analyzing variations of ambient noise as indicators of imminent failure (Got et al. 2010; Lévy et al. 2010). To investigate long term rock mass deformations due to temperature, wind and rainfalls, the abandoned Acuto quarry rock wall (Frosinone – Central Italy) was selected in Autumn 2015 as test site for the installation of a multi-sensor monitoring system on a rock block prone to failure. The experimental site is included in the carbonatic Monti Ernici ridge. More in particular, Mesozoic wackestone with rudists crops out on the sub-vertical quarry wall with a height ranging from 15 m up to 50 m (Accordi et al. 1986). A geomechanical characterization of the rock mass led to the recognition of four joint sets, here indicated according to dip direction/dip convention: S0 (130/13) corresponding to the limestone strata, S1 (270/74), S2 (355/62) and S3 (190/64) (Fantini et al. 2016). The monitored sector is located in the NW portion of the 500 m long quarry front, which is characterized by the presence of a 64 m 3 densely cracked protruding block, separated from the back quarry wall by a main fracture. The multi-sensor monitoring system consists in two weather stations with air-thermometer, hygrometer, pluviometer and anemometer for wind speed and direction, installed at foot and top of the slope wall, one thermometer for the rock mass temperature, six strain-gauges installed on micro-fractures of the rock mass, four extensimeter installed on open fractures and one optical device. Moreover, a railway track was reproduced near the rock mass, to simulate a real rockfall hazard scenario (Fig. 1).

1 Introduction The Westwood mine is located approximately 80 km west of the town of Val-d'Or in Quebec, Canada. Production in the 104-mining block in Zone 2 of the Westwood mine was halted by three largemagnitude seismic events over two days in May 2015 (Fig. 1). These seismic events are technically interesting because mining activities leading up to these events were reviewed and no specific blast (production or development) could be identified as a trigger. Further, these pillar bursting events occurred in the Central Infrastructure Corridor (CIC) well away from ore zone production, where adverse seismicity in mining operations would more typically be anticipated. This paper provides an overview of the seismic events and uses numerical models to demonstrate how geological conditions and mine geometry both contributed to what was considered, by a review panel of experts, to be an unforeseen series of events. 2 Numerical back analysis The main objective of the numerical analyses performed in this study was to improve the understanding of overall rock mass behaviour in Zone 2 at the Westwood mine and to better define the factors that contributed to the May 2015 seismicity. To achieve this, global mine-scale models and local drift-scale models were developed. Model calibration utilized a combination of the available seismic data, underground observations, the condition of boreholes, and inspection of core drill after the events. The global model was developed using FLAC3D (V. 5.0 by Itasca Consulting Group 2012) and included all development and stoping geometries in the relevant mining blocks. Drift-scale modelling utilized Phase2 (V.8.0 by RocScience 2015). Fig. 2 provides a summary of material properties and boundary conditions. The first iteration of numerical simulations focused on the global mine model in order to narrow the range of possible stress tensors and material property combinations. The first iteration of global-scale modelling produced non-unique results. Multiple combinations of boundary conditions and material properties could reproduce the observed ground response. However, these preliminary numerical solutions required either excessively high k-ratios, or un-realistically low material strengths to induce sufficient rock mass yield to drive the pillar failure.

1 Background Recent studies have demonstrated important links between the strength and failure mode of brittle rock materials, how they respond to varying principal stress conditions and stress paths, and resultant field-scale responses, manifest as differences in the pattern and nature of slope failures and landform and landscape evolution (Diederichs 2003; Leith et al. 2014). However, the influence of strain rate in driving differences in intact rock strength and failure mode is less well constrained (cf. Amann et al. 2011). Furthermore, how differences in stress path and strain rate influence future failures by controlling if, where and when progressive fracture can develop has not previously been considered in sufficient detail. 2 Approach In contrast to most conventional confined compression tests that employ an increase in axial load to cause failure, we focus on the effects of both slow and rapid (near-instantaneous, quasi-dynamic) reductions in confining pressure on the compressive strength and failure mode of rock samples. We consider the former (slow loss of confining pressure) to be representative of a longer-term, gradual loss of hillslope material, driven by small-scale but continuous rockfall activity. The latter (rapid loss of confining pressure) mimics stress conditions in remnant hillslope materials during and following large slope failures that cause a sudden reduction in lateral confinement of remaining, unfailed materials. 3 Results We present results of laboratory tests undertaken on a medium- to coarse-grained sandstone (unconfined compressive strength = 51.0 ± 10.6 MPa). We undertook a series of confined compression tests over a confining pressure range of 1 to 10 MPa. Our results indicate that differences in strength and failure mode occur where strain rates differ, despite (un-)loading along the same stress path (constant axial load, reducing confining pressure). A gradual, controlled reduction in confining pressure causes minor, but not insignificant, strengthening relative to standard axial loading tests, though we observed standard shear failure typical of confined compression (Fig. 1a). In contrast, rapid reduction in confining pressure causes an increase in compressive strength relative to baseline axial loading tests, but results in greater rock fragmentation upon failure, indicative of hybrid shear and tensile axial splitting modes (Fig. 1b,c).

The use of geothermal energy in Bavaria increased considerably in recent years, and practical experience in geothermal plant operation for heat and power generation has shown a primary need for research. To answer questions related to deep geothermal energy and to concentrate and link operating experience of the operators, the Geothermie-Allianz Bayern (GAB) was founded. Bavaria has a unique geothermal potential in Germany. In particular, the South-German Molasse Basin south of Munich offers ideal conditions for the development of hydrothermal geothermal energy. The present project aims to reduce the risk of geothermal search in the South German Molasse Basin and to optimize reservoir engineering. For this purpose, the hydraulic effectiveness of disturbances will be investigated by a structure-stress analysis (Reinecker 2010). The disturbance zones near drilled holes will be characterized geologically, seismically and geomechanically. The results provide data for a reference model of the southern Molasse basin. In addition, the local stress situation around the borehole will be evaluated and a geomechanical reservoir model will be constructed using data from the geomechanical near-field analysis (Seithel et al. 2015). In order to complete the database of the mechanical properties of representative types of rock, extensive investigations will be carried out in the laboratory. This will be done with samples from geothermal drillings as well as with so-called analogue samples, which come from geologic or petro-graphic explorations, e.g. quarries. Furthermore, petrothermal potential in crystalline rocks is focused in the northern part of Bavaria (Riemer 2011). The development and use of this future technology initially requires extensive research, but then it has an outstanding future potential, since the use of geothermal energy is not limited to water in depth (Bauer 2000). Considering local geology, natural stress conditions and geomechanical laboratory parameters, numerical modeling approaches shall show in which part of the investigation area hydraulically division surfaces of defined orientation will appear with priority in case of stimulation measurements. The results will be validated using the above mentioned borehole and exploration analogue data and will serve to predict existing rock anisotropies after stimulation measurements. The slip-tendency analysis is used here to investigate the behavior of already existing interfaces / disturbance zones in a defined stress field. It is used to identify potentially re-activatable faults and to selectively prevent seismic events during fluid injections.

1 Introduction In a paper published in 2013, Yaméogo et al. showed how rock failure could affect stress measurements performed using the so-called modified doorstopper technique developed at Polytechnique Montréal. The modified doorstopper stress measurement technique (Corthésy et al. 2016) often categorized as an overcoring technique, consists in recording the recovered strains caused by stress relief at the bottom of a borehole drilled in a rock mass and converting the relieved strains into stresses by using the deformability parameters at the measuring point. The recovered strains are monitored using a specially designed data logger (Fig. 1) which is linked to a modified doorstopper cell. The recovered strains are converted to stresses assuming the rock behaviour is linear elastic and anisotropic. If rock failure occurs during the measurement process, the stress tensor calculated by assuming an elastic behaviour can be wrong. Core discing is probably the most visible manifestation of core failure when performing such measurements, however, even if no visible signs of damage are present, plastic strains can be induced at or near the borehole bottom, affecting the recovered strains and the calculated stresses (Corthésy & Leite 2008). When Yaméogo et al. (2013) published their work, no interpretation model that could account for the progressive failure occurring prior to or during the stress relief drilling process was available. Recently, Corthésy et al. (2016) proposed using the inverse problem approach for interpreting modified doorstopper stress measurements. In this abstract, the potential for using this approach when progressive failure occurs during the stress measurement operations is evaluated.

1 Introduction The well-documented 2014–2016 exfoliation of Twain Harte Dome, located in the western foothills of California's Sierra Nevada Mesozoic granitic batholith, provides an unprecedented opportunity to study progressive fracture of exfoliation domes. The dome seemingly spontaneously fractured on five occasions during the late summer of 2014, with subsequent fracturing in the summers of 2015 and 2016. The warm-weather timing of all events strongly suggests that thermally-induced environmental conditions are responsible for triggering exfoliation. Some of the exfoliation events consisted of rock slabs being thrust ~40 cm into the air in a matter of seconds, with both audible and observable explosive-like energy release. We initiated a research project to investigate the triggering conditions and fracture mechanics of dome exfoliation under this active setting to serve as a proxy for understanding precursor signals that might occur prior to rock falls on near-vertical cliffs. Here we present the portion of our data that captures the collapse of the slab that was deformed during the 2014 events. The data suggest that the collapse and other ongoing deformation is occurring as a result of progressive fracture driven by diurnal heating and cooling. 2 Exfoliation sheet monitoring Following the August 2014 fracture events, we installed three vibrating-wire-type strain gauge "crackmeters" (Geokon modified 4420; see Collins & Stock 2016) between a newly detached portion of the uppermost slab of the dome and the intact rock beneath (Fig. 1). Two crackmeters monitored open fractures and a third crackmeter acted as a control to differentiate between thermal strains in the instrumentation verses thermal strains in the granitic slab. We also deployed pendent-type temperature and light sensors (Onset Hobo UA-002–08) to act as a proxy for surface rock temperatures. Our instrumentation captured data at 5-minute intervals beginning on 22 August 2014 for 41 days until the upper slab detached and crushed the crackmeters.

1 Subcritical crack growth and stress corrosion cracking Progressive deformation of rocks, e.g. by creep (Rutter 1972; Itô & Sasajima 1980; Brantut et al. 2012) and fatigue (Attewell & Farmer 1973; Costin & Holcomb 1981; Ko & Kemeny 2013), at low magnitude near constant stresses is facilitated by different subcritical mechanisms causing gradual irreversible damage, commonly termed subcritical crack growth (Atkinson 1984; Atkinson & Meredith 1987). Subcritical crack growth is an effective means of facilitating time-dependent material degradation (Atkinson 1984; Brantut et al. 2013; Nara et al. 2013). These mechanisms include plastic (Atkinson & Meredith 1987), as well as brittle deformations (Atkinson & Meredith 1987; Nara et al. 2013). These mechanisms are enhanced in the presence of chemically active fluids, such as water (Atkinson & Meredith 1981; Lawn 1993). An interaction of chemical and mechanical processes, as postulated in stress corrosion cracking (SCC), reduces brittle fracture toughness, and therefore increases subcritical fracture propagation velocities. SCC involves a range of range processes, including stress enhanced dissolution (as described by the Charle's law), hydrolysis and embrittlement, and a reduction of surface energies which in turn reduces cohesive forces and promotes granular disintegration. To better understand the coupling of chemical and mechanical processes in stress corrosion cracking and subcritical crack growth, we set up single edge notch bending (SENB) creep tests, mimicking natural conditions. 2 Experimental methods Six Alta-Quartzite samples (AQ 1–6, 300x30x70 mm) were brought to failure in stepped SENB creep tests. The load was increased over months in steps of 5–10 % of the intact wet flexural strength determined in preliminary tests, starting at 10% for samples AQ1–3 and at 50 % for samples AQ4–6 (Tab. 1). Samples were pre-cracked to ~ 50 % (F= 4000 N) of the intact strength in a standard SENB testing frame with 10 N/s. Straining of the sample in the long-term tests was measured using electrical resistivity strain gages 2 mm below the notch, and four samples were supplied with distilled water in the notch (AQ 1–2, 4–5). To assess the influence of dissolution in the wet samples, this water was dripped in and captured for analysis of silica content by Atomic Absorption Spectrometry (AAS). Scanning Electron Microscopy images were taken of failed dry and wet samples to allow a surface texture analysis and fractography of fracture planes, to better define and discriminate chemo-mechanical mechanisms of stress corrosion and subcritical crack growth.

1 Introduction Compiled field and literature data show that landslides are geologically pre-disposed by the lithological, structural and morphological setting. Concerning the Alps, radiometric age dating data suggest that several active moving landslides represent "old" slope deformations that have been (re-) activated for some 1000 yrs. At several locations, densely foliated and fractured metamorphic rocks (phyllites, schists, gneisses) are widely spread and affected by slowly moving deep-seated rock slides. According to Cruden and Varnes (1996) rock slides are characterized by a downslope movement of a rock mass on surfaces of rupture i.e. relatively thin zones of intensive shear strain. The landslides presented herein may be classified as "rock compound slides" (according to Varne's classification updated by Hungr et al. (2014)). A rock compound slide is characterized by sliding of a mass of rock on a rupture surface consisting of several planes, or an uneven surface, so that motion is kinematically feasible only if accompanied by significant internal distortion of the moving mass. Distinct extensional features i.e. horst- and graben structures at the head as well secondary shear surfaces are characteristic features of this type of landslide. Furthermore, this type of landslides shows fully persistent basal shear zones, and in some cases also internal shear zones formed through progressive failure processes. It is commonly agreed that shear zones of rock slides originate from to the growth, propagation and coalescence of pre-existing and new fractures as well as rock cataclasis. Based on a variety of field surveys and monitoring data, several slopes show actual gravitational deformation that is determined by sliding processes resulting from shear deformation along discrete shear zones. Generally, there is little knowledge about the temporal and spatial evolution of shear zones. In addition, there is also a lack in understanding how shear zones form by progressive failure processes. As a consequence, the assessment if a fully persistent basal shear zone has formed is still difficult when direct subsurface investigations such as core drillings or investigation adits are not available. This contribution aims to highlight open questions for future research rather than to provide final outcomes and conclusions. Beside others, one aim is to initiate a fruitful discussion focusing on the formation and geometry of shear zones of deep-seated rock slides. Field observations focusing on rock mass failure processes of slides and conceptual considerations focusing on the formation of shear zones are presented. Emphasis is given on surface investigations based on mapping and monitoring methods and their ability to develop reliable 3D rock slide models for slope analyses and hazard assessments.

1 Why do we study exfoliation joints? Exfoliation joints are natural fractures limited to near the ground surface, i.e., they occur at a depth range important for rock extraction and engineering works. Exfoliation joints (also known as sheeting joints) follow the landscape surface at the time of their formation; the age of the associated landscape feature then provides a maximum age of exfoliation joints. While landscape forms can change through time, exfoliation joints preserve elements of former landscape morphologies by their undisturbed orientations. Exfoliation fractures can play a key role in causing slope instabilities such as rockfalls and rockslides, increase the rates of rock mass weathering and erosion, but also, together with tectonic fractures, allow for economic extraction of dimension stones from a rock mass. Apart from this, we can use exfoliation joint properties (orientation, spacing, and fracture surface morphology) to constrain geologically recent and today's near-surface rock mass stresses, a rock mass state property that is essential for rock engineering. Furthermore, well-preserved fracture surface morphologies let us interpret fracture formation modes and processes. 2 What is special about the Grimsel region of the Central Alps? The Grimsel region in Switzerland is well suited for analyzing the impact of erosional episodes, and accompanying stress changes, on exfoliation joint formation in granitic-gneissic rocks (Ziegler et al. 2013, 2014, 2016). Mapping above and below ground revealed that exfoliation joints are widespread and occur between valley bottoms and mountain crests within glacial (inner and hanging U-shaped trough valleys, glacial cirques, and steep mountain crests) and predominantly fluvial landforms (gently inclined linear slopes above the inner trough valleys, narrow inner-valley gorges, and steep V-shaped side gullies). Relating the mapped distribution and characteristics of exfoliation joints to identified erosional episodes and landscape features of known and inferred ages enabled us to distinguish four exfoliation joint generations in the Grimsel area, ranging from the lower Pleistocene (~1.5–1 Ma; generation 1) to the Late Glacial/Holocene (<0.02 Ma; generation 4). The most prominent and deepest exfoliation joint generation is associated with erosion of the inner glacial troughs of the upper Aar valley, which likely occurred during the 'mid-Pleistocene Revolution' (generation 2), followed by a joint generation with a minimum age dating to about the Last Glacial Maximum (generation 3).

1 Introduction The growth of ice in porous rocks can generate large internal pressures and progressive crack growth. The resulting damage to rock is widely believed to arise from the 9% volumetric expansion of pore water during freezing. This belief is, in fact, at odds with much of the work on freezing of soils and porous solids; under natural conditions, freezing-induced damage in these solids is strongly related to water movement and often largely independent of the specific-volume change of the pore fluid during freezing (Walder & Hallet 1986; Hallet 2006). Herein, we outline the underlying processes, and highlight recent developments to provide an introduction to this rich subject. 2 Underlying processes The potential of the 9% water-to-ice expansion to cause high pressure in confined spaces is undeniable based on practical experience with, for example, broken bottles in freezers and broken pipes exposed to freezing weather, etc. This expansion may, however, rarely be significant for rocks under natural conditions, because it requires a tight orchestration of unusual conditions. Unless the rocks are essentially saturated with water (unlikely for many situations at the ground surface; Fig. 1 for instance) and frozen from all sides, the expansion can simply be accommodated by the flow of water into empty pores, or out of the rock through its unfrozen sides. The common notion that incipient cracks at the surface of rocks can be wedged open by freezing (as sketched in a number of textbooks) may also be rarely important in nature, because water can leak out of the cracks, and the ice capping the cracks can push out (Davidson & Nye 1985). Ice growth inside porous rocks has much in common with frost heave of soils, which has received considerable research attention. Less well known is that slow ice growth in soils hydraulically connected to an unpressurized water reservoir over long periods, hundreds of hours, can lead to pressures against a confining piston in a stiff apparatus exceeding 18 MPa (Radd & Oertle 1973). This is far in excess of the internal pressure required for crack propagation in most rocks. Such ice growth (known as segregation ice) is sustained by a supply of water driven thermodynamically along unfrozen films toward growing ice lenses (Taber 1929). Intermolecular forces acting between the mineral surfaces, ice, and water sustain these unfrozen films, and generate significant pressure between mineral and ice surfaces (Dash et al. 2006).

1 Introduction A long period of exceptionally high temperatures in Finland in the summer of 2015 was associated with the formation of large ‘exfoliation’ or ‘sheeting’ fractures on Långören Island in the Finnish Archipelago. A video of the event shows sharp fractures forming along the edge of a thin bedrock sheet several meters across (Sarpaneva 2014), while total vertical strain of one section of the pop-up was up to 100 mm, and observations following the event indicate a slab of approximately 50 m2 was detached from the underlying bedrock. Long scratch marks (‘striations’) visible on the bedrock surface of the island are the result of boulders being dragged over the landscape during the last glacial period (>15 ka BP), hinting at the rarity of the recent events on the otherwise undamaged surfaces (Fig. 1). 2 Setting Deglaciation of the region is bracketed by two moraine complexes dated at 15.2–13 ka BP and 13–11.5/6 ka BP, while flow lineations based on a reconstructed ice sheet topology and regional field mapping of drumlins indicate the present-day island was located on the eastern margin of a southward- flowing ice stream sourced in the center of the Weichselian Scandinavian ice sheet (Boulton et al. 2004, and references therein). The ice sheet in this region was around 2500 m thick during the estimated maximum for the Last Glacial Maximum at approximately 18.5 ka BP (Näslund 2006). Present-day isostatic uplift rates relative to sea level in the region are approximately 3.5–4 mm/a (Ekman 1996). Tidal variations in the archipelago only cause a few centimeters of sea level change, and as a result, the slow emergence of the now 3.5 m – 4 m. a.s.l. bedrock surface can reasonably accurately be extrapolated back to 0.8–1 ka BP.

1 Introduction Stability and serviceability of brittle rocks or rock masses is mainly governed by load level and strength, whereby three processes can be distinguished: sub-critical crack and critical crack growth under constant or slowly changing load and dynamic triggered failure. Some selected recent research results with respect to (1.) continuum and discontinuum based numerical approaches considering subcritical and critical crack growth under constant load (Chen et al. 2014, 2015, 2016; Konietzky et al. 2009, 2015; Li et al. 2014, 2015, 2016; Tan et al. 2016) and (2.) first results of the stability of discontinuities under complex dynamic loading obtained by dynamic shear box tests (Dang et al. 2016a,b) are presented. 2 Damage evolution due to subcritical and critical crack growth Macroscopic fracturing is considered as the result of the growth and coalescence of microcracks. Based on fracture mechanics, (based on Charles or Charles-Hillig equations) and critical crack growth (linear elastic fracture mechanics) can be distinguished and used to determine the life time of rocks under loading using stress and crack-size dependent crack propagation velocity. Exemplary, Fig. 1 shows numerical simulation of damage evolution over time in terms of failed zone counts and axial deformation of a rock sample under uniaxial compression, which reveals primary, secondary and tertiary creep phases. This model assumes the existence of a huge number of microcracks with different length and orientation following certain stochastic distributions.

1 Introduction One of the main methods in hard rock quarry mining is the controlled fracture method that is carried out by the introduction of a slowly advancing crack by Non-Explosive Expansion Material (NEEM). The method of rock breakage is without noise and vibrations and its operation, compared to blasting method, is more controllable, very safe and easy and without extra undesirable cracks in the rock block (Hinze and Nelson, 1996; Gambatese 2003; Huynh, et al. 2009; Laefer, et al. 2010). In this method, some circular holes are drilled closely with equal length, diameter and spacing (center- to-center distance) in a rock block. Subsequently, the holes are filled by the NEEM, which it can be expanded and generates an incremental static loading into the holes after about two to four hours. If the spacing of the holes to be adequate, it will create a crack between two neighboring holes and the rock will fracture along the high-stress concentration path between the holes. In this paper will be try to introduce a new numerical modelling approach for showing location and length of the first cracks around the holes and predicting of the crack growing path with the geometry. 2 Experimental data and a case study A granite quarry mine was selected as a case study entitled "Ahrar Mine" which it has located near to Natanz in Iran (Fig. 1). Some laboratory tests have been done on the granite specimens for measuring of some rock mechanic's properties such as uniaxial compressive strength, Brazilian tensile strength, Young's modulus, Poisson's ratio, specific gravity and fracture toughness by ISRM standard methods (Fowell, 1995), (Tab. 1). Process of rock fracture in due time by the NEEM has been shown in Figs. 1 and 2. The drill holes were 38 mm with 1 to 3 meters depth and the average of the hole's spacing was 12 cm (10–14 cm) in the mine.