Rockfall Characterisation and Structural Protection - a Review A

Home , Geogrid, Load testing

Rockfall characterisation and structural protection - a review A. Volkwein, K. Schellenberg, V. Labiouse, F. Agliardi, F. Berger, F. Bourrier, L.K.A. Dorren, W. Gerber, M. Jaboyedoff

To cite this version:

A. Volkwein, K. Schellenberg, V. Labiouse, F. Agliardi, F. Berger, et al.. Rockfall characterisation and structural protection - a review. Natural Hazards and Earth System Sciences, Copernicus Publ. / European Geosciences Union, 2011, 11, p. 2617 - p. 2651. hal-00653458

HAL Id: hal-00653458 https://hal.archives-ouvertes.fr/hal-00653458 Submitted on 19 Dec 2011

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Natural Hazards doi:10.5194/nhess-11-2617-2011 and Earth © Author(s) 2011. CC Attribution 3.0 License. System Sciences

Rockfall characterisation and structural protection – a review

A. Volkwein1, K. Schellenberg2, V. Labiouse3, F. Agliardi4, F. Berger5, F. Bourrier6, L. K. A. Dorren7, W. Gerber1, and M. Jaboyedoff8 1WSL Swiss Federal Institute for Forest, Snow and Landscape Research, Zurcherstrasse¨ 111, 8903 Birmensdorf, Switzerland 2Gruner+Wepf Ingenieure AG, Thurgauerstr. 56, 8050 Zurich,¨ Switzerland 3Swiss Federal Institute of Technology Lausanne EPFL, Rock Mechanics Laboratory LMR, GC C1-413 Station 18, 1015 Lausanne, Switzerland 4Universita` degli Studi di Milano-Bicocca, Dip. Scienze Geologiche e Geotecnologie, Piazza della Scienza 4, 20126 Milano, Italy 5Cemagref, Mountain Ecosystems and Landscapes Research, 38402 Saint Martin d’Heres` Cedex, France 6Cemagref, UR EMGR, 2, rue de la Papeterie, BP 76, 38402 Saint Martin d’Heres` Cedex, France 7Landslides, Avalanches and Protection Forest Section, Federal Ofﬁce for the Environment FOEN, Bern, Switzerland 8University of Lausanne, Institute of Geomatics and Analysis of Risk, Amphipole 338, 1015 Lausanne, Switzerland Received: 25 March 2011 – Revised: 26 July 2011 – Accepted: 7 August 2011 – Published: 27 September 2011

Abstract. Rockfall is an extremely rapid process involving a rockfall event from its initiation to suitable protective mea- long travel distances. Due to these features, when an event sures. This includes a presentation of typical applications as occurs, the ability to take evasive action is practically zero well as an extensive literature survey for the relevant topics and, thus, the risk of injury or loss of life is high. Damage that are evaluated and discussed with regard to their perfor- to buildings and infrastructure is quite likely. In many cases, mance, reliability, validation, extreme loads, etc. Contribu- therefore, suitable protection measures are necessary. This tions include contribution provides an overview of previous and current research on the main topics related to rockfall. It covers the – Rockfall susceptibility together with hazard assessment onset of rockfall and runout modelling approaches, as well as and zoning. hazard zoning and protection measures. It is the aim of this article to provide an in-depth knowledge base for researchers – Rockfall initiation and runout modelling and practitioners involved in projects dealing with the rock- – Design and performance evaluation of rockfall protec- fall protection of infrastructures, who may work in the ﬁelds tion systems, with particular attention paid to structural of civil or environmental engineering, risk and safety, the countermeasures such as fences, walls, galleries, em- earth and natural sciences. bankments, ditches or forests

Rockfall hazard (or risk) can be assessed using different 1 Introduction approaches (Einstein, 1988), depending on the characteristics of the investigated areas. Often the hazard must be as- Rockfall is a natural hazard that – compared to other haz- sessed along a communication (transport) route; in this case, ards – usually impacts only small areas. However, the dam- ﬁeld records and lists of past rockfall events (inventories) are age to the infrastructure or persons directly affected may be often used (Luckman, 1976; Bunce et al., 1997; Hungr et al., high with serious consequences. It is often experienced as a 1999), but have proved to be limited. For example, on 31 harmful event. Therefore, it is important to provide the best May 2006 a major rockfall (5000 m3) killed two tourists on possible protection based on rigorous hazard and risk man- the main highway crossing the Alps through the Gotthard agement methods. This contribution gives an overview of Tunnel in Switzerland (Liniger and Bieri, 2006). The event the assessment on parameters needed to deal effectively with caused global headlines and led to somewhat emotional me- dia reporting of major rockfall incidents in the Alps in the following weeks, including rockfall on the Eiger mountain Correspondence to: A. Volkwein (Hopkins, 2006; Oppikofer et al., 2008). Another recent ([email protected]) event shows the difﬁculties of forecasting rockfall events.

Published by Copernicus Publications on behalf of the European Geosciences Union. 2618 A. Volkwein et al.: Review on rockfall characterisation and structural protection

has to be clarified why and where rocks are released and the total volume or extent. The rockfall initiation also depends on different factors, mostly not yet quantified, such as weathering, freezing/melting cycles or heavy rainfall (see Sect. 3). Subsequent trajectory analyses determine the areas that have to be protected by measures. To account for their high sensitivity to just small changes in the landscape, such as bedrock, dead wood, small dips, etc., stochastic analyses are usually performed, preferably including an evaluation of the accuracy of the results. This is described in more detail in Sect. 4. However, for a quick preliminary analysis and estimation of the rockfall hazard, simpler and manual calculation methods might also be useful as described in Sect. 4.4.1. There is a large variety of structural protection measures against rockfall. These include natural protection by means Fig. 1. Rockfall on Sea to Sky highway (B.C.). Note the jointed of forests, semi-natural structures such as embankments and structure of the source area (Canadian Press photos). ditches and fully artificial structures such as fences, galleries or walls. The structural part of this contribution focuses mainly on fences and galleries. A short summary for em- During the night of 29 July 2008, a rockfall blocked the high- bankments is also given. Natural protection by means of way Sea to Sky joining Vancouver to the ski resort Whistler forests is mentioned in Sect. 5.5. (Fig. 1). This road is the cover picture of the well-known rock mechanics book by Hoek and Bray (1981). The area has been extensively investigated for risk analysis in the past 2 Rockfall hazard: definition, assessment and zonation (Bunce et al., 1997) and still is, because of an increase in population density (Blais-Stevens, 2008) and the Olympics Rockfall is a major cause of landslide fatality, even when el- Games in 2010. ements at risk with a low degree of exposure are involved, Further difficulties exist when the goal is to assess risk (or such as traffic along highways (Bunce et al., 1997). Al- hazard) on a regional scale for a limited area or over an entire though generally involving smaller rock volumes compared territory. Generally, inventories exist only in inhabited areas. to other landslide types (e.g., rock slides/rock avalanches), Moreover, some studies suggest that the number of events in- rockfall events also cause severe damage to buildings, increases in proportion to urbanization (Baillifard et al., 2004). frastructures and lifelines due to their spatial and temporal As a consequence, it is necessary to find ways that allow frequency, ability to easily release and kinetic energy (Ro- one to detect rockfall hazard source areas in the absence of chet, 1987b). The problem is even more relevant in large any inventory or clear morphological evidence, such as scree alpine valleys and coastal areas, with a high population den- slopes or isolated blocks. sity, transportation corridors and tourist resorts. Rockfall This article is structured following the typical work- protection is, therefore, of major interest to stakeholders, ad- flow when dealing with rockfall in practice (Vogel et al., ministrators and civil protection officers (Hungr et al., 2005). 2009), covering rockfall occurrence and runout modelling Prioritization of mitigation actions, countermeasure selection approaches, hazard zoning and protection measures. and land planning should be supported by rockfall hazard as- When a rockfall hazard or risk analysis (including the pro- sessment (Raetzo et al., 2002; Fell et al., 2005, 2008). On the tective effect of forests) reveals a threat to people, buildings other hand, risk analysis is needed to assess the consequences or infrastructures (see Sect. 2), suitable structural protection of expected rockfall events and evaluate both the technical measures have to be selected according to the expected event suitability and the cost-effectiveness of different mitigation frequency and impact energies. For proper design and di- options (Corominas et al., 2005; Straub and Schubert, 2008). mensioning of the measures, it is essential to know the magnitude of the impact loads and the performance of the struc- 2.1 Rockfall hazard: a definition tures. This knowledge can be obtained from rockfall onset susceptibility/ hazard analysis, numerical simulations, exper- Landslide hazard has been defined as the probability that a iments, models or existing guidelines, and provides guidance landslide of given magnitude occurs in a given area over on the design of roof galleries, fences, embankments and a specified time interval (Varnes, 1984; Einstein, 1988). forests as a natural protection system. This definition envisages the concepts of spatial location, However, rockfall protection considerations involve not temporal frequency and intensity. Nevertheless, for long- only structural protection measures but also the avoidance runout landslides, such as rockfall or rock avalanches, the of infrastructure or buildings in endangered areas. Firstly, it definition of the occurrence probability needs to account for

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2619 A. Volkwein et al.: Rockfall review 3 the concept of landslide propagation. This means the trans- 2.2 Hazard assessment fer of landslide mass and energy from the source to the maximum runout distance of up to tens of kilometres for rock In principle, rockfall hazard assessment would require eval- avalanches and debris flows or several hundred metres for uating: fragmental rockfall, characterised by poor interaction between falling blocks with volumes up to 105 m3 (Evans and a. the temporal probability (annual frequency or return pe- Hungr, 1993). Thus, rockfall hazard depends on (Jaboyed- riod) and the spatial susceptibility of rockfall events; off et al., 2001; Crosta and Agliardi, 2003; Jaboyedoff et al., b. the 3D trajectory and maximum runout of falling 2005b, Fig. 2) blocks; – the probability that a rockfall of given magnitude occurs c. the distribution of rockfall intensity at each location and at a given source location resulting in an onset probabil- along each fall path. ity Exposed elements at risk are not considered in the definition of hazard. Nevertheless, hazard assessment ap- – the probability that falling blocks reach a specific loca- proaches should be able to deal with problems charac- tion on a slope (i.e., reach probability), and on terized by different spatial distributions of potentially exposed targets, point-like (houses), linear (roads, rail- – rockfall intensity. ways) or areal (villages). Moreover, targets of different shape and size are likely to involve a different number of The latter is a complex function of block mass, velocity, rota- trajectories running out from different rockfall sources tion and jump height, significantly varying both along single (Jaboyedoff et al., 2005b, Fig. 2), influencing the local fall paths and laterally, depending on slope morphology and reach probability. Thus, assessment methods should be rockfall dynamics (Broili, 1973; Bozzolo et al., 1988; Azzoni able to account for the spatially distributed nature of the et al., 1995; Agliardi and Crosta, 2003; Crosta and Agliardi, hazard (Crosta and Agliardi, 2003). Although several 2004). Rockfall hazard can, thus, be better defined as the Fig.Fig. 2. 2.DefinitionDefinition of of rockfall rockfall hazard and related parameters parameters (modi- (modi- hazard assessment methods have been proposed, very probability that a specific location on a slope is reached by fied,fied, after afterJaboyedoff Jaboyedoff et et al. al.,, 2001 2001).). few satisfy all these requirements. They differ from one a rockfall of given intensity (Jaboyedoff et al., 2001), and another in how they account for rockfall onset frequency expressed as: or susceptibility, estimated reach probability, and com- Exposed– the probability elements that at risk falling are blocksnot considered reach a specific in the defini- loca- tion of hazard. Nevertheless, hazard assessment approaches bine them to obtain quantitative or qualitative hazard H = P (L) ·P (T |L) (1) tion on a slope (i.e. reach probability), and on ijk j ijk should be able to deal with problems characterised by differ- ratings. ent– spatialrockfall distributions intensity. of potentially exposed targets, point- where P (L)j is the onset probability of a rockfall event in the 2.2.1 Onset probability and susceptibility | like (houses), linear (roads, railways) or areal (villages). magnitude (e.g., volume) class j, and P (T L)ijk is the reach The latter is a complex function of block mass, velocity, ro- Moreover, targets of different shape and size are likely to in- The frequency of events of given magnitude (volume) should probability. This is the probability that blocks triggered in tation and jump height, significantly varying both along sin- volve a different number of trajectories running out from dif- be evaluated using a statistical analysis of inventories of the same event reach the location i with an intensity (i.e., ki- gle fall paths and laterally, depending on slope morphology ferent rockfall sources (Jaboyedoff et al., 2005b, Fig. 2), in- rockfall events, taking into account the definition of suitable netic energy) value in the class k. Since both probability and and rockfall dynamics (Broili, 1973; Bozzolo et al., 1988; fluencing the local reach probability. Thus, assessment meth- magnitude-frequency relationships (Dussauge-Peisser et al., intensity strongly depend on the initial magnitude (i.e., mass) Azzoni et al., 1995; Agliardi and Crosta, 2003; Crosta and ods should be able to account for the spatially distributed 2003; Malamud et al., 2004). They are also called magni- of rockfall events, rockfall hazard must be assessed for dif- Agliardi, 2004). Rockfall hazard can thus be better defined nature of the hazard (Crosta and Agliardi, 2003). Although tude - cumulative frequency distributions (MCF; Hungr et al., ferent magnitude scenarios, explicitly or implicitly associ- as the probability that a specific location on a slope is reached several hazard assessment methods have been proposed, very 1999). Although this approach is well established in the field ated to different annual frequencies or return periods (Hungr by a rockfall of given intensity (Jaboyedoff et al., 2001), and few satisfy all these requirements. They differ from one an- of natural hazards (e.g. earthquakes), its application to land- et al., 1999; Dussauge-Peisser et al., 2003; Jaboyedoff et al., expressed as: slide hazards is limited by the scarce availability of data and 2005b). other in how they account for rockfall onset frequency or sus- by the intrinsic statistical properties of landslide inventories ceptibility,H = P ( estimatedL) · P (T reach|L) probability, and combine them(1) to ijk j ijk (Malamud et al., 2004). The frequency distribution of rock- 2.2 Hazard assessment obtain quantitative or qualitative hazard ratings. fall volumes has been shown to be well fitted by the power where P (L)j is the onset probability of a rockfall event in law: In principle, rockfall hazard assessment would require the 2.2.1the magnitude Onset probability (e.g. volume) and class susceptibilityj, and P (T |L)ijk is the reach probability. This is the probability that blocks trig- evaluating of: log N(V ) = N0 − b · log V (2) Thegered frequency in the same of events event reachof given the magnitude location i with (volume) an intensity should (a) the temporal probability (annual frequency or return pe- be(i.e. evaluated kinetic energy) using valuea statistical in the class analysisk. Since of both inventories probabil- of where N(V ) is the annual frequency of rockfall with a vol- riod) and the spatial susceptibility of rockfall events; rockfallity and intensity events, stronglytaking into depend account on the the initial definition magnitude of (i.e. suit- ume exceeding V , N0 is the total annual frequency of rock- ablemass) magnitude-frequency of rockfall events, rockfall relationships hazard must (Dussauge-Peisser be assessed for fall, and b is the power law exponent, ranging between 0.4 (b) the 3-D trajectory and maximum runout of falling etdifferent al., 2003 magnitude; Malamud scenarios, et al., explicitly2004). They or implicitly are also associ- called and 0.7 (Dussauge-Peisser et al., 2003). According to Hungr blocks; magnitude-cumulativeated to different annual frequency frequencies distributions or return periods (MCF; (HungrHungr et al. (1999), magnitude-cumulative frequency curves (MCF) etet al. al.,, 1999 1999;). Dussauge-Peisser Although this approach et al., 2003; is well Jaboyedoff established et al., in derived from rockfall inventories allow estimating the annual (c) the distribution of rockfall intensity at each location and the2005b). field of natural hazards (e.g., earthquakes), its applica- frequency of rockfall events in specified volume classes, thus along each fall path. tion to landslide hazards is limited by the scarce availability www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2620 A. Volkwein et al.: Review on rockfall characterisation and structural protection of data and by the intrinsic statistical properties of landslide the maximum extent of rockfall runout areas is estimated inventories (Malamud et al., 2004). The frequency distribu- (Fig. 3a). However, this approach has been implemented in tion of rockfall volumes has been shown to be well fitted by a GIS tool (CONEFALL, Jaboyedoff and Labiouse, 2003) the power law: allowing a preliminary estimation of rockfall reach susceptibility and kinetic energy (Fig. 3b), according to the energy = − · logN(V ) N0 b logV (2) height approach (Evans and Hungr, 1993). Many existing where N(V ) is the annual frequency of rockfall with a vol- hazard assessment methodologies estimate reach probability and intensity using 2-D rockfall numerical modelling (Matte- ume exceeding V , N0 is the total annual frequency of rockfall and b is the power law exponent, ranging between 0.4 rock, Rouiller and Marro,1997; Rockfall Hazard Assessment and 0.7 (Dussauge-Peisser et al., 2003). According to Hungr Procedure RHAP, Mazzoccola and Sciesa, 2000; Cadanav, et al. (1999), magnitude-cumulative frequency curves (MCF) Jaboyedoff et al., 2005b). This provides a more accurate derived from rockfall inventories allows for the estimating description of rockfall physics and allows for a better eval- of the annual frequency of rockfall events in specified vol- uation of rockfall reach probability (i.e., relative frequency ume classes, thus, defining hazard scenarios. Major limita- of blocks reaching specific target locations) and of the spa- tions to this approach include the lack of rockfall inventories tial distribution of kinetic energy). However, 2-D modelling for most sites and the spatial and temporal heterogeneity of neglects the geometrical and dynamic effects of a 3-D to- available inventories. These are possibly affected by cen- pography on rockfall, leading to a subjective extension of soring, hampering a reliable prediction of the frequency of simulation results between adjoining 2-D fall paths (Fig. 3c). either very small and very large events (Hungr et al., 1999; Although this limitation has, in part, been overcome by intro- Dussauge-Peisser et al., 2003; Malamud et al., 2004). The ducing pseudo 3-D assumptions (Jaboyedoff et al., 2005b), hazard has been completely assessed using this approach by full 3-D numerical modelling has been shown to be required Hungr et al. (1999) in the case of a section of highway. On to account for the lateral dispersion of 3-D trajectories and a regional scale, Wieczorek et al. (1999) and Guzzetti et al. the related effects on reach probability and intensity. Nev- (2003) partially included the MCF within the method; while ertheless, a few hazard assessment methodologies based on Dussauge-Peisser et al. (2002, 2003) and Vangeon et al. 3-D numerical modelling are available (Crosta and Agliardi, (2001) formalized the use of the MCF on a regional scale 2003, Fig. 3d). merging it with susceptibility mapping. Where site-specific rockfall inventories are either unavail- 2.3 Hazard zoning: current practice and unresolved able or unreliable, the analysis of rockfall hazard can only questions be carried out in terms of susceptibility. This is the relative Rockfall hazard or susceptibility mapping/zoning is the final probability that any slope unit is affected by rockfall occur- step of hazard assessment, leading to the drafting of a doc- rence, given a set of environmental conditions (Brabb, 1984). ument useful for land planning, funding prioritization or the Onset susceptibility (see Sect. 3) can be assessed preliminary assessment of suitable protective measures. The – in a spatially distributed way by heuristic ranking of se- major issue in hazard zoning is to find consistent criteria to lected instability indicators (Pierson et al., 1990; Can- combine onset probability or susceptibility, reach probabil- celli and Crosta, 1993; Rouiller and Marro, 1997; Maz- ity and intensity in a map document, especially when formal zoccola and Sciesa, 2000; Budetta, 2004), probabilities cannot be evaluated. Swiss guidelines (Raetzo et al., 2002, see Fig. 4) require – by deterministic methods (Jaboyedoff et al., 2004a; that rockfall hazard are zoned according to the onset proba- Guenther et al., 2004; Derron et al., 2005) or bility (i.e., return period) and intensity (i.e., kinetic energy), – by statistical methods (Frattini et al., 2008). thus, defining three hazard zones, namely red, blue and yel- low. Nevertheless, these do not explicitly account for the 2.2.2 Reach probability and intensity reach probability and the spatial variability of kinetic energy. Thus, Jaboyedoff et al. (2005b) proposed a method- The reach probability and intensity for rockfall of given mag- ology (Cadanav) based on 2-D numerical modelling to map nitude (volume) depends on the physics of rockfall processes hazard according to the probability where blocks involved in and on topography (see Sect. 4). The simplest methods de- events with a specified return period reach a specific location scribing rockfall propagation are based on the shadow an- along a 2-D profile with a given kinetic energy. gle approach, according to which the maximum travel dis- When only onset susceptibility can be evaluated, hazard tance of blocks is defined by the intersection of the topog- zoning is based on the combination of hazard indicators or raphy with an energy line having an empirically-estimated reclassified values of the parameters contributing to the haz- inclination (Evans and Hungr, 1993, Fig. 2). Unfortunately, ard to obtain suitable hazard indices. Some authors (Rouiller with this approach there is no physical process model for and Marro, 1997; Jaboyedoff et al., 2001; Derron et al., rockfall and its interaction with the ground behind and only 2005; Copons and Vilaplana, 2008) used simple methods for

Intensity Intensity/ Energy / Energy

High > 300 kJ Highhazard > 300 kJ hazard

Medium 30 – 300 kJ Medium 30 – 300 kJ hazard hazard

<30kJ< 30 kJ Low <30kJ< 30 kJ Lowhazard hazard

1 – 30 y 30 – 100 y 100 – 300 y Onset probability 1 – 30 y 30 – 100 y 100 – 300 y Onset/ return probability period / return period

Fig. 4. Hazard classification for rockfall in Switzerland Fig.Fig. 4. 4.HazardHazard classification classification for for rockfall rockfall in in Switzerland Switzerland

scale assessment capabilities. Major uncertainties in rockfall scaleity)hazard of assessment hazard mapping maps capabilities. are also depends related Major on to a the uncertainties number uncertainty of factors. in of rockfall rockfall Dif- hazardferentonset mapping frequencydescriptions are when also of requiredrockfall related to (e.g.dynamics the Swiss uncertainty can Code). be of adoptedThis rockfall is of- to onsetmodelten unknown, frequency rockfall thus trajectorieswhen requiring required (e.g., that (e.g. 2-D a set Swiss or of 3-D, scenario-based Code). empirical, This iskine- haz-of- Fig. 3. Comparison of hazard maps derived for the area of Mt. tenmaticalard unknown, maps or rather dynamic). thus than requiring a Moreover, single that map a set complex are of produced scenario-based phenomena, (Jaboyedoff haz- such Fig. 3. Comparison of hazard maps derived for the area of Mt. Fig.S.Martino 3. Comparison (Lecco, Italy; of hazard Jaboyedoff maps derived et al., 2001;for the Crosta area andof ardaset blockmaps al., 2005b). rather fragmentation From than a this single or perspective, the map effects are theproduced of choice vegetation, (Jaboyedoff of the maydesign be S.Martino (Lecco, Italy; Jaboyedoff et al., 2001; Crosta and Mt.Agliardi, S. Martino 2003) (Lecco, using Italy; differentJaboyedoff modelling et approachesal., 2001; Crosta and zoning and etaccountedblock al., 2005b). volume for From scenarioin different this perspective, is critical ways to (Crosta the avoid choice either et al. of,risky the2004designunder-; Dor- Agliardi, 2003) using different modelling approaches and zoning Agliardimethods., 2003 a)) Maximumusing different runout modelling area estimated approaches by a shadow and zoning angle blockren et volume al., 2004 scenario) and isgreatly critical influence to avoid all either therisky hazardunder- com- methods. a) Maximum runout area estimated by a shadow angle estimation or cost-ineffective overestimation of a hazard. Fi- methods.approach(a) usingMaximum the code runout CONEFALL area estimated (Jaboyedoff by a shadow and Labiouse, angle estimationponents related or cost-ineffective to rockfall propagation overestimation and, of athus, hazard. the Fi- final approach using the code CONEFALL (Jaboyedoff and Labiouse, nally, the extent of mapped hazard zones is greatly influenced approach2003); using b) hazard the mapcode obtained CONEFALL by applying (Jaboyedoff the RHV and methodology Labiouse, nally, the extent of mapped hazard zones is greatly influenced 2003); b) hazard map obtained by applying the RHV methodology hazardby subjectivity map. The in spatial establishing resolution class boundariesof the adopted for parame- descrip- 2003(Crosta); (b) hazard and Agliardi, map obtained 2003) to by the applying reach probability the RHVmethodology and kinetic en- by subjectivity in establishing class boundaries for parame- (Crosta and Agliardi, 2003) to the reach probability and kinetic en- tionters of contributing topography, to the especially hazard. when These 3-D should models be constrained are used, (Crostaergy and estimated Agliardi by, CONEFALL;2003) to the reach c) rockfall probability hazard andmap kinetic obtained en- by ergy estimated by CONEFALL; c) rockfall hazard map obtained by terscontrolsby contributing physically-based primarily to the criteria hazard. lateral depending dispersion These should on of the rockfall be envisaged constrained trajecto- use ergy2D estimated numerical by modellingCONEFALL; using(c) therockfall RHAP hazard methodology map obtained (modified by 2D numerical modelling using the RHAP methodology (modified byriesof physically-based the and maps the computed (e.g. landcriteria dynamic planning depending quantities, or countermeasure on the thus, envisaged affecting design; use the 2-Dafter numerical Mazzoccola modelling and using Sciesa, the 2000); RHAP d) methodology rockfall hazard (modified map ob- after Mazzoccola and Sciesa, 2000); d) rockfall hazard map ob- oflocalCrosta the reachmaps and (e.g.Agliardi, probability land 2003; planning and Jaboyedoff intensity or countermeasure ( etCrosta al., 2005b). and design;Agliardi, aftertainedMazzoccola by 3D numerical and Sciesa modelling, 2000); using(d) rockfall the code hazard HY-STONE map ob- and tainedthe RHV by 3D methodology numerical modelling (modified using after Crostathe code and HY-STONE Agliardi, 2003). and Crosta and Agliardi, 2003; Jaboyedoff et al., 2005b). tained by 3-D numerical modelling using the code HY-STONE and 2004). The applicability of hazard models on different scales the RHV methodology (modified after Crosta and Agliardi, 2003). 2.4 From hazard to quantitative risk assessment the RHV methodology (modified after Crosta and Agliardi, 2003). and with different aims also depends on model resolution, 2.4thus, From requiring hazard tools to quantitative with multi-scale risk assessment assessment capabili- evaluation of onset susceptibility by means of multivariate ties.Although Major hazard uncertainties mapping in is rockfall a useful hazard tool for zoning land planning, are also evaluation of onset susceptibility by means of multivariate Although hazard mapping is a useful tool for land planning, statistical techniques. relatedrisk analysis to the should uncertainty be carried of rockfall out to support onset frequency the design when and statistical techniques. risk analysis should be carried out to support the design and largeWhen scale susceptibility drafting hazard mapping, maps for basedpractical on purposes, the use of it muston- requiredoptimization (e.g., of Swiss both Code). structural This and is non-structural often unknown, protection thus, re- When drafting hazard maps for practical purposes, it must optimization of both structural and non-structural protection setbe susceptibility kept in mind indicators that the reliability and the (and shadow practical angle applicabil- method quiringactions that (Fell a et set al., of 2005; scenario-based Straub and hazardSchubert, maps 2008). rather Never- than be kept in mind that the reliability (and practical applicabil- actions (Fell et al., 2005; Straub and Schubert, 2008). Never- (Fig.ity)3a). ofMazzoccola hazard maps and depends Sciesa on(2000 a number) proposed of factors. a method- The atheless, single map a standard are produced risk analysis (Jaboyedoff approach et for al. rockfall, 2005b is). yet From to ity) of hazard maps depends on a number of factors. The theless, a standard risk analysis approach for rockfall is yet to ologydescription (RHAP) of in rockfall which dynamics 2-D numerical are adopted simulation to model is used rock- thisbeproposed perspective, because the choice of the of still the difficultdesign assessmentblock volume of haz- sce- description of rockfall dynamics are adopted to model rock- be proposed because of the still difficult assessment of haz- to zonefall trajectories reach probability (e.g. 2D along or 3D, profiles, empirical, later weightedkinematical ac- or narioards. is In critical fact, when to avoid a hazard either isrisky expressedunderestimation as susceptibility, or cost- fall trajectories (e.g. 2D or 3D, empirical, kinematical or ards. In fact, when a hazard is expressed as susceptibility, cordingdynamic). to indicators This way, of cliff complex activity phenomena (Fig. 3c). suchCrosta as block and ineffectiverisk can only overestimation be assessed through of a hazard. relative Finally, scales or the matrices extent dynamic).fragmentation This or way, the complexeffects of phenomena vegetation are such accounted as block for risk(Guzzetti can only et be al., assessed 2004; Fell through et al., relative 2005). scales The simplest or matrices form Agliardi (2003) combined reclassified values of reach sus- of mapped hazard zones is greatly influenced by subjectivity fragmentation(Crosta et al., or 2004; the effects Dorren of etvegetation al., 2004) are and accounted greatly influ- for (Guzzettiof rockfall et al.,risk 2004; analysis Fell consists et al., of2005). analysing The simplestthe distribution form ceptibility and intensity values such as kinetic energy or in establishing class boundaries for parameters contributing (Crostaence all et al.,the 2004;hazard Dorren components et al., related 2004) to and rockfall greatly propaga- influ- ofof rockfall elements risk at analysis risk with consists different of analysing postulated the vulnerability distribution in jump height derived by distributed 3-D rockfall modelling to the hazard. These should be constrained by physically- encetion, all and the thushazard the components final hazard related map. toThe rockfall spatial propaga- resolution ofdifferent elements hazard at risk zones with different (Acosta et postulated al., 2003; vulnerability Guzzetti et inal., to obtain a physically-based index (Rockfall Hazard Vector, based criteria depending on the envisaged use of the maps RHV).tion,of andthe This adopted thus allows the description final for a hazard quantitative of map. topography, ranking The spatial ofespecially hazards, resolution when ac- different2003, 2004). hazard However, zones (Acosta this approach et al., 2003; does not Guzzetti fully account et al., of3D the models adopted are description used, controls of topography, primarily the especially lateral dispersion when 2003,(e.g.for the 2004). land probability planning However, of or thisrockfall countermeasure approach impact, does the design; not vulnerability fullyCrosta account and and counting for the effects of 3-D topography (Fig. 3d) while Agliardi, 2003; Jaboyedoff et al., 2005b). keeping3Dof models rockfall information are trajectories used, about controls and the primarily of contributing the computed the lateral parameters. dynamic dispersion Thisquan- forvalue the probability of exposed of targets. rockfall Guidelines impact, the for vulnerability Quantitative and Risk oftities, rockfall thus trajectories affecting and theof local the reach computed probability dynamic and quan- inten- valueAnalysis of exposed (QRA) targets. based on Guidelines Hong Kong for rockfall Quantitative inventories Risk approach was implemented by Frattini et al. (2008) to include 2.4 From hazard to quantitative risk assessment atities, quantitativesity thus(Crosta affecting evaluation and Agliardi, the of local onset 2004). reach susceptibility The probability applicability by and means of inten- hazard of Analysis(Chau et (QRA) al., 2003) based were on proposed Hong Kong by GEO rockfall (1998), inventories whereas multivariatesitymodels (Crosta on statisticaland different Agliardi, techniques. scales 2004). and The with applicability different aims of hazard also de- (ChauStraub et and al., Schubert2003) were (2008) proposed combined by GEO probability (1998), theory whereas and modelspends on on different model resolution, scales and thus with requiring different tools aims with also multi- de- StraubAlthough2D numerical and Schuberthazard modelling zoning (2008) in is combined order a useful to improve probability tool for risk land theory analysis planning, and for pendsWhen on drafting model resolution, hazard maps thus for requiring practical tools purposes, with it multi- must 2Drisk numerical analysis shouldmodelling be carriedin order out to improve to support risk the analysis design for and be kept in mind that the reliability (and practical applicabil- optimization of both structural and non-structural protective www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2622 A. Volkwein et al.: Review on rockfall characterisation and structural protection actions (Fell et al., 2005; Straub and Schubert, 2008). Never- < 100 000 m3) methods of rock slope stability analysis are theless, a standard risk analysis approach for rockfall is yet to well established and their application is relatively easy when be proposed because of the still difficult assessment of haz- the slope and the source area are well characterised (Hoek ards. In fact, when a hazard is expressed as susceptibility, and Bray, 1981; Norrish and Wyllie, 1996; Wyllie and Mah, risk can only be assessed through relative scales or matrices 2004). However, this procedure does not give any informa- (Guzzetti et al., 2004; Fell et al., 2005). The simplest form tion about time-dependence and is difficult to apply on a re- of rockfall risk analysis consists of analysing the distribution gional scale (Guenther et al., 2004). of elements at risk with different postulated vulnerability in Most rockfall source area assessment methods are based different hazard zones (Acosta et al., 2003; Guzzetti et al., on stability assessment or on rockfall activity quantification. 2003, 2004). However, this approach does not fully account In order to get an estimate of rockfall activity, either inven- for the probability of rockfall impact, the vulnerability and tories or indirect methods, such as dendrochronology, are value of exposed targets. Guidelines for Quantitative Risk needed (Perret et al., 2006; Corominas et al., 2005). Several Analysis (QRA) based on Hong Kong rockfall inventories parameters can be used to create a hazard map for rockfall (Chau et al., 2003) were proposed by GEO (1998), whereas source areas, which, most of the time, involves susceptibility Straub and Schubert (2008) combined probability theory and mapping (Guzzetti et al., 1999). The parameters used de- 2-D numerical modelling in order to improve risk analysis for pend mainly on the availability of existing documents or the single countermeasure structural design. Bunce et al. (1997) budget available to collect field information (Jaboyedoff and and Hungr et al. (1999) quantitatively estimated rockfall risk Derron, 2005). along highways in British Columbia, based on inventories Source area susceptibility analysis has often used multi- of rockfall events. Nevertheless, major efforts are still re- parameter rating systems derived from tunnelling and mining quired to perform a quantitative evaluation of rockfall risk in engineering, such as Rock Mass Rating (Bieniawski, 1973, spatially distributed situations (e.g., urban areas; Corominas 1993, RMR;). Its evolution to the Slope Mass Rating SMR et al., 2005), where long runout and complex interactions be- (Romana, 1988, 1993) led to more suitable results by adding tween rockfall and single elements at risk occur, requiring a an explicit dependence on the joint-slope orientation rela- quantitative assessment of vulnerability. tionship. Recently, Hoek (1994) introduced the Geological In this perspective, Agliardi et al. (2009) proposed a quan- Strength Index (GSI) as a simplified rating of rock quality. titative risk assessment framework exploiting the advantages In recent years, it has been applied successfully to slope sta- of 3-D numerical modelling to integrate the evaluation of the bility analysis (Brideau et al., 2007). A similar approach was temporal probability of rockfall occurrence, the spatial prob- proposed by Selby (1980, 1982) for geomorphological appli- ability and intensity of impacts on structures, their vulnera- cations. Later, with the increasing availability of digital ele- bility, and the related expected costs for different protection vation models (DEM; Wentworth et al., 1987; Wagner et al., scenarios. In order to obtain vulnerability curves based on 1988) and of geographic information systems (GIS), several physical models for reinforced concrete buildings, Mavrouli other techniques (heuristic and probabilistic) have been ex- and Corominas (2010) proposed the use of Finite Element plored (Van Westen, 2004). However, this can be refined con- (FE)-based progressive collapse modelling. ceptually because a slope system can be described in terms of internal parameters (IP) and external factors (EF), which provide a conceptual framework to describe the instability po- 3 Rockfall source areas tential using the available data (Fig. 5). Therefore, instability detection requires locating (1) the pre-failure processes and 3.1 Influencing factors (2) the areas sensitive to rapid strength degradation leading to slope failure (Jaboyedoff et al., 2005a; Leroueil and Locat, As pointed out in Sect. 2, the rockfall hazard H at a given 1998). IP are the intrinsic features of the slopes. Some exam- location and for a given intensity and scenario depends on ples are summarized below (Jaboyedoff and Derron, 2005): two terms, namely: the onset probability (i.e., temporal fre- (a) Morphology: slope types (slope angle, height of slope, quency of rockfall occurrence) of a rockfall instability event profile, etc.), exposure, type of relief (depends on the and the probability of propagation to a given location (see controlling erosive processes), etc. Eq. 1)(Jaboyedoff et al., 2001). The latter, P (T |L)ijk, can be evaluated by propagation modelling or by observation. In (b) Geology: rock types and weathering, variability of the order to evaluate P (L), it is first necessary to identify poten- geological structure, bedding, type of deposit, folded tial rockfall sources, whereas their susceptibility is mainly zone, etc. based on rock slope stability analysis or estimates and can (c) Fracturing: joint sets, trace lengths, spacing, fracturing be evaluated by field observations or modelling. Anyway, intensity, etc. it must be kept in mind that inventories are the only direct way to derive the true hazard in small areas. For rockfall (d) Mechanical properties of rocks and soil: cohesion, fric- involving limited volumes (i.e., fragmental rockfall, usually tion angle, etc.

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2623 6 A. Volkwein et al.: Rockfall review following different methods that have been proposed to as- single countermeasure structural design. Bunce et al. (1997) sess the value of failure frequency P (L) in general by using and Hungr et al. (1999) quantitatively estimated rockfall risk susceptibility mapping. GIS and related software allow one along highways in British Columbia, based on inventories to manage most of these parameters regionally. For example, of rockfall events. Nevertheless, major efforts are still rein Switzerland the 1 : 250000 topographic vectorized maps quired to perform a quantitative evaluation of rockfall risk in spatially distributed situations (e.g. urban areas; Corominas include the cliff area as polygons (Jaboyedoff and Labiouse, et al., 2005), where long runout and complex interactions be- 2003; Loye et al., 2009). tween rockfall and single elements at risk occur, requiring a quantitative assessment of vulnerability. 3.2 Methods of identification and description Thus, Agliardi et al. (2009) proposed a quantitative risk assessment framework exploiting the advantages of 3D nu- 3.2.1 Methods using regional geomechanical merical modelling to integrate the evaluation of the temporal approaches probability of rockfall occurrence, the spatial probability and intensity of impacts on structures, their vulnerability, and the Basically, methods such as the Rock Fall Hazard Rating Sys- tem (RFHRS, Pierson et al., 1990) or the Missouri Rockfall related expected costs for different protection scenarios. In Fig.Fig. 5. 5. EFEF and IP for rockfall (modified (modified from from JaboyedoffJaboyedoff and and Labi- Labi- order to obtain vulnerability curves based on physical mod- ouseouse,, 2003 2003;; Jaboyedoff and Derron, Derron, 2005).2005). Hazard Rating System (MRFHRS, Maerz et al., 2005) mix els for reinforced concrete buildings, Mavrouli and Coromi- both P (L) and P (T | L) estimates at the same level, as well nas (2010) proposed the use of Finite Element (FE)-based as risk. Both methods are designed for talus slopes close to progressive collapse modelling. budget(e) Activity: available movements to collect fieldor rockfall, information etc. (Jaboyedoff and roads and have been refined in two ways, i.e., simplifying Derron, 2005). the number of parameters from 12 (or 18) to 4 for the RHRS (f)SourceHydrogeology: area susceptibility permeability, analysis joint has permeability, often used multi- etc. (Santi et al., 2008) or by mixing them with the RMS param- 3 Rockfall source areas parameter rating systems derived from tunnelling and mining eters (Budetta, 2004). These methods mix IP and EF at the Noteengineering, that within such a asgiven Rock framework, Mass Rating the (Bieniawski,joint sets or discon- 1973, same levels. 3.1 Influencing factors tinuities1993, RMR;). are the Its anisotropies evolution to that the mainly Slope controlMass Rating the stability SMR In addition to the classical rock mass characterisation (Bi- ((Romana,Hoek and 1988, Bray 1993), 1981 led); points to more b suitable to d are results related by to adding these eniawski, 1973; Romana, 1988), some methods are proposed an explicit dependence on the joint-slope orientation rela- As pointed out in Section 2, the rockfall hazard H at a given properties. The link between rockfall activity and the inten- to regionalise susceptibility parameters. Using mixed IP and tionship. Recently, Hoek (1994) introduced the Geological location and for a given intensity and scenario depends on sity of pre-existing fracturing, as in fold hinges with a steep EF Mazzoccola and Hudson (1996) developed a rating sys- Strength Index (GSI) as a simplified rating of rock quality. two terms, namely: the onset probability (i.e. temporal fre- limb, has been demonstrated by Coe and Harp (2007). tem based on the matrix interaction approach of the Rock En- In recent years, it has been applied successfully to slope sta- quency of rockfall occurrence) of a rockfall instability event The IP can evolve with time due to the effects of the EF, gineering System (RES) methodology (Hudson, 1992). This bility analysis (Brideau et al., 2007). A similar approach was and the probability of propagation to a given location (see which are (Jaboyedoff and Derron, 2005): allows one to create a modular rock mass characterisation Eq. 1) (Jaboyedoff et al., 2001). The latter, P (T |L) , can proposed by Selby (1980, 1982) for geomorphological appli- ijk method of slope susceptibility ranking. Based on a similar be evaluated by propagation modelling or by observation. In cations.– gravitational Later, with effects; the increasing availability of digital ele- approach, Vangeon et al. (2001) proposed to calibrate a sus- order to evaluate P (L) it is first necessary to identify po- vation models (DEM; Wentworth et al., 1987; Wagner et al., ceptibility scale using a geotechnical rating with a regional tential rockfall sources whereas their susceptibility is mainly 1988)– water and ofcirculation: geographic hydrology information or hydrogeology, systems (GIS) severalclimate, inventory, designed for a linear cliff area (Carere et al., 2001). based on rock slope stability analysis or estimates and can otherprecipitation techniques (heuristic in the form and of probabilistic) rainfall or snow, have infiltration been ex- Rouiller et al. (1998) developed a susceptibility rating system be evaluated by field observations or modelling. Anyway, ploredrates, (Van groundwater; Westen, 2004). However, this can be refined con- it must be kept in mind that inventories are the only direct ceptually because a slope system can be described in terms of based on 7 criteria mixing IP and EF. – weathering; way to derive the true hazard in small areas. For rockfall internal parameters (IP) and external factors (EF), which provide a conceptual framework to describe the instability po- 3.2.2 GIS and DEM analysis-based methods involving limited volumes (i.e. fragmental rockfall, usually – erosion; < 100, 000 m3) methods of rock slope stability analysis are tential using the available data (Fig. 5). Therefore, instability The first studies on rockfall using DEM or GIS were per- well established and their application is relatively easy when detection– seismicity; requires locating (1) the pre-failure processes and the slope and the source area are well characterized (Hoek (2) the areas sensitive to rapid strength degradation leading formed by Toppe (1987a), using simply the slope angle cri- and Bray, 1981; Norrish and Wyllie, 1996; Wyllie and Mah, to– slopeactive failure tectonics; (Jaboyedoff et al., 2005a; Leroueil and Locat, terion, and by Wagner et al. (1988) and Wentworth et al. 2004). However, this procedure does not give any informa- 1998). IP are the intrinsic features of the slopes. Some exam- (1987); Wu et al. (1996); Soeters and Van Westen (1996), tion about time-dependence and is difficult to apply on a re- ples– aremicroclimate summarized including below (Jaboyedoff freezing and and thawing, Derron, 2005): sun ex- using structural data for slope modelling. Of course, the gional scale (Guenther et al., 2004). posure, permafrost, which are increasingly invoked to simplest way to detect a source area is to use a slope angle Most rockfall source area assessment methods are based a.explain Morphology: rockfall slope activities types (slope (Frayssines angle,, height2005; Matsuoka of slope, threshold (Guzzetti et al., 2003), or to add some other crite- on stability assessment or on rockfall activity quantification. andprofile, Sakai etc.),, 1999 exposure,; Matsuoka type, 2008 of relief; Gruner (depends, 2008 on); the ria such as the presence of cliff areas (Jaboyedoff and Labi- In order to get an estimate of rockfall activity either in- controlling erosive processes), etc. ouse, 2003). The slope threshold can be deduced from a de- – nearby instabilities; tailed slope angle statistical analysis permitting one to iden- ventories or indirect methods such as dendrochronology are b. Geology: rock types and weathering, variability of the needed (Perret et al., 2006; Corominas et al., 2005). Several tify cliff areas (Strahler, 1954; Baillifard et al., 2003, 2004; – humangeological activities structure, (anthropogenic bedding, type factors); of deposit, folded parameters can be used to create a hazard map for rockfall zone, etc. Loye et al., 2009). In addition, some other approaches can source areas, which, most of the time, involves susceptibility – etc. be used for assessing the susceptibility of source areas, such mapping (Guzzetti et al., 1999). The parameters used de- c. Fracturing: joint sets, trace lengths, spacing, fracturing as using an index obtained by the back-analysis of rockfall pend mainly on the availability of existing documents or the Theseintensity, lists of etc. internal parameters and external factors are propagation. This index links the source area to the deposit, not exhaustive, but allow one to introduce key points for the by counting the number of intersections of the trajectories

www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2624 A. Volkwein et al.: Review on rockfall characterisation and structural protection with the scree slopes. This can be performed either using the quality data from DEM that – regarding some points – is bet- shadow angle method (Baillifard, 2005) or the HY STONE ter than that from standard fieldwork, especially for geologi- programme by intersecting the trajectory simulation with the cal structures (joint sets, fractures). However, for a local fully scree slopes (Frattini et al., 2008). detailed analysis, on-site inspection using Alpine techniques Along one particular road in Switzerland, five parame- is unavoidable in order to correctly asses the amount of open- ters: proximity to faults, nearness of a scree slope, cliff ings, fillings or roughness of joints or to verify automatically height, steep slope and proximity to road, were used to obtain determined rock face properties. good results using a simple classical GIS approach (Bailli- At the present time, the attempt to extract information such fard et al., 2003). as GSI from LiDAR DEM is still utopian (Sturzenegger et al., The major improvement related to GIS or/and the use of 2007b), but we can expect future generations of terrestrial Li- DEM is the automatic kinematical analysis (Wagner et al., DAR to allow the extraction of such information. The anal- 1988; Rouiller et al., 1998; Gokceoglu et al., 2000; Dorren ysis of geological structures in high resolution DEM and the et al., 2004; Gunther¨ , 2003; Guenther et al., 2004), which al- simulation of all possible instabilities in a slope have already lows one to determine whether the discontinuity sets are able been performed at the outcrop level (Grenon and Hadjigeor- to create instabilities. Using the standard stability criterion giou, 2008). We can expect that such methods will be ap- (Norrish and Wyllie, 1996) and a statistical analysis of the plicable on a regional scale within the next 10 yr by using kinematical tests, Gokceoglu et al. (2000) were able to pro- remote-sensing techniques associated with limited field ac- duce maps of probability of sliding, toppling or wedge type quisition that will provide rock parameters, structures and failures. Gunther¨ (2003) and Guenther et al. (2004) used a include stability simulations. However, the goal of hazard partial stability analysis using a Mohr-Coulomb criterion and assessment will not be reached as long as this analysis does an estimate of the stress state at a given depth of about 20 m not account for temporal dependencies. That can only be at each pixel of the DEM, also integrating in the analysis the achieved if we understand the failure mechanisms, i.e., the regionalisation of discontinuities such as folded bedding and degradation of the IP under the action of EF, such as weath- geology. The number of slope failures linked to joint sets ering (Jaboyedoff et al., 2007). Expected climate changes depends on the apparent discontinuity density at the ground will affect the frequency and magnitude of the EF. There is a surface, which can also be used as an input for the rock slope need to understand their impact on rock slope stability, other- hazard assessment and to identify the most probable failure wise we will either miss or overestimate a significant amount zone (Jaboyedoff et al., 2004b). In addition to structural tests, of potential rockfall activity. it may also be possible to combine several of the EF and IP, such as water flow, erodible material volume, etc., to obtain a rating index (Baillifard et al., 2004; Oppikofer et al., 2007). 4 Trajectory modelling Rock failure is mainly controlled by discontinuities. The It is important to describe the movement of a falling rock main joint sets can be extracted from the orientation of the to- along a slope, i.e., its trajectory. This allows the description pography (DEM) using different methods and software (Der- of existing hazard susceptibility or hazard assessment for a ron et al., 2005; Jaboyedoff et al., 2007; Kemeny et al., 2006; certain area. In addition, the information on boulder velocity, Voyat et al., 2006). Extracting the discontinuity sets from jump heights and spatial distribution is the basis for correct DEM allows one to perform a kinematic test on a regional design and the verification of protective measures. area (Oppikofer et al., 2007). New techniques such as ground A description of rockfall trajectories can be roughly ob- based LiDAR DEM allow one to extract the full structures, tained by analytical methods (see Sect. 4.4.1). If more de- even in the case of inaccessible rock cliffs (Lato et al., 2009; tailed analyses are needed and stochastic information has to Sturzenegger et al., 2007a; Voyat et al., 2006). be considered, numerical approaches are recommended. In landslide hazard assessment, many statistical or other This section, therefore, attempts to summarize the numer- modern techniques are now used (Van Westen, 2004); ous currently available rockfall trajectory simulation mod- e.g., Aksoy and Ercanoglu (2006) classified the susceptibility els. To do this, existing models are grouped firstly accord- of source areas using a fuzzy logic-based evaluation. ing to their spatial dimensions: (1) two-dimensional (2-D) trajectory models, (2) 2.5-D or quasi-3-D trajectory models 3.3 Concluding remarks on source detection and (3) 3-D trajectory models, and secondly according to the underlying calculation principles. Whether a rockfall trajec- Until now, most rock slope systems have been described by tory model is 2-D or 3-D, irrespective of its underlying cal- considering the EFs and IPs that control stability. This pro- culation procedure, the experience in applying the model and cedure only gives approximate results, mainly because field a knowledge of its sensitivity to parameter settings, as well access is usually limited. Moreover, to assess the hazard as how to determine model parameter values in the field, is a from susceptibility maps remains very difficult. Neverthe- prerequisite to obtaining acceptable results. Berger and Dor- less, recently developed technologies like photogrammetry ren (2006) defined the latter as results with an error of 20 %. or LiDAR (Kemeny et al., 2006) permit one to extract high

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2625

4.1 Types of rockfall model 2008) or as shown in Masuya et al. (1999). The major advan- tage of 3-D models is that diverging and converging effects of 4.1.1 2-D rockfall trajectory models the topography, as well as exceptional or surprising trajectories, i.e., those that are less expected at first sight in the field, We define a 2-D trajectory model as a model that simulates are clearly reflected in the resulting maps. A disadvantage of the rockfall trajectory in a spatial domain defined by two 3-D models is the need for spatially explicit parameter maps, axes. This can be a model that calculates along a user-defined which require much more time in the field than parameter slope profile (Azzoni et al., 1995) that is defined by a dis- value determination for slope profile-based trajectory simu- tance axis (x or y) and an altitude axis (z). Such a profile lations. often follows the line of the steepest descent. Table 1 shows that the majority of the rockfall trajectory models belongs 4.2 Calculation approaches to this group. In the second type of 2-D model rockfall trajectories are calculated in a spatial domain defined by two A second main characteristic that allows one to distinguish distance axes x and y, e.g., a raster with elevation values or a between different rockfall trajectory models, which is closely map with contour lines. Such models generally calculate the related to the calculation of the rebound, is the representation rockfall path using topographic-hydrologic approaches and of the simulated rock in the model. As shown in Table 1, this velocity and runout distance with a sliding block approach can be done firstly by means of a lumped mass, i.e., the rock (cf. Van Dijke and van Westen, 1990; Meissl, 1998). As such is represented by a single, dimensionless point. The second these models do not provide information on rebound heights. approach is the rigid body, i.e., the rock is represented by a real geometrical shape, which is often a sphere, cube, cylin- 4.1.2 2.5-D rockfall trajectory models der or ellipsoid. In general, this approach is used in the deterministic models mentioned above. The last approach is the The second group of trajectory models defined here are 2.5- hybrid approach, i.e., a lumped mass approach for simulat- D models, also called quasi-3-D models. These are simply ing free fall and a rigid body approach for simulating rolling, 2-D models assisted by GIS to derive pre-defined fall paths. impact and rebound (Crosta et al., 2004; Frattini et al., 2008; The key characteristic of such models is that the direction of Agliardi et al., 2009). the rockfall trajectory in the x,y domain is independent of the Most of the rockfall trajectory models use a normal and kinematics of the falling rock and its trajectory in the vertical a tangential coefficient of restitution for calculating the re- plane. In fact, in these models the calculation of the hori- bound of simulated rock on the slope surface and a fric- zontal fall direction (in the x,y domain) could be separated tion coefficient for rolling. Details on these coefficients completely from the calculation of the rockfall kinematics are, among others, presented in Guzzetti et al. (2002). An and the rebound positions and heights. This means that these overview of typical values of the coefficients of restitu- models actually carry out two separate 2-D calculations. The tion can be found in Scioldo (2006). The models that use first one determines the position of a slope profile in an x,y these coefficients generally apply a probabilistic approach domain and the second one is a 2-D rockfall simulation along for choosing the parameter values used for the actual re- the previously defined slope profile. Examples of such mod- bound calculation (see Table 1). This is to account for els are those that calculate rockfall kinematics along a slope the large variability in the real values of these parameters, profile that follows the steepest descent as defined using dig- due to the terrain, the rock shape and the kinematics of ital terrain data, as in the model Rocky3 (Dorren and Seij- the rock during the rebound. Bourrier et al. (2009b) pre- monsbergen, 2003). sented a new rebound model that linked the impact angle, the translational and the rotational velocity before and after 4.1.3 3-D rockfall trajectory models the rebound based on multidimensional, stochastic functions, These models are defined as trajectory models that calcu- which gave promising results for rocky slopes. There are late the rockfall trajectory in a 3-dimensional plane (x, y, also models that use deterministic approaches for calculat- z) during each calculation step. As such, there is an in- ing the rockfall rebound. These models use mostly a discrete terdependence between the direction of the rockfall trajec- element method (Cundall, 1971), such as the Discontinuous tory in the x,y domain, the kinematics of the falling rock, Deformation Analysis (Yang et al., 2004) or percussion the- its rebound positions and heights and if included, impacts ory (Dimnet, 2002). on trees. Examples of such models are EBOUL-LMR (De- The parabolic free falls are calculated with standard algo- scoeudres and Zimmermann, 1987), STONE (Guzzetti et al., rithms for a uniformly accelerated parabolic movement, ex- 2002), Rotomap (Scioldo, 2006), DDA (Yang et al., 2004), cept for those models that use the sliding block theory for STAR3-D (Dimnet, 2002), HY-STONE (Crosta et al., 2004) calculating the rockfall velocity over its complete trajectory. and Rockyfor3-D (Dorren et al., 2004), RAMMS:Rockfall (Christen et al., 2007); Rockfall-Analyst (Lan et al., 2007), PICUS-ROCKnROLL (Rammer et al., 2007; Woltjer et al., www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2626 A. Volkwein et al.: Review on rockfall characterisation and structural protection

Table 1. Main characteristics of a selection of existing rockfall trajectory models (modiﬁed from Guzzetti et al., 2002).

Model/programme name Reference/Year Spatial Dimensions Approach Probabilistic Forest* N.N. (Ritchie, 1963) 2-D (slope profile) Lumped-mass No No Discrete Element Method (Cundall, 1971) 2-D (slope profile) Rigid body No No Computer Rockfall Model (Piteau and Clayton, 1976) 2-D (slope profile) Lumped-mass Partly No N.N. (Azimi et al., 1982) 2-D (slope profile) Lumped-mass Yes No N.N. (Falcetta, 1985) 2-D (slope profile) Rigid body No No ROCKSIM (Wu, 1985) 2-D (slope profile) Lumped-mass Yes No SASS (Bozzolo and Pamini, 1986) 2-D (slope profile) Hybrid Yes No EBOUL-LMR (Descoeudres and Zimmermann, 1987) 3-D (x,y,z) Rigid body No No (Labiouse et al., 2001) PROPAG/CETE Lyon (Rochet, 1987a) 2-D (slope profile) Lumped-mass No No N.N. (Hungr and Evans, 1988) 2-D (slope profile) Lumped-mass No No CRSP (4.0) (Pfeiffer and Bowen, 1989) 2-D (slope profile) Hybrid Yes No (Jones et al., 2000) N.N. (Van Dijke and van Westen, 1990) 2-D (x,y) Lumped-mass No No N.N. (Kobayashi et al., 1990) 2-D (slope profile) Rigid body No No Rotomap (Scioldo, 1991) 3-D (x,y,z) Lumped-mass Yes No CADMA (Azzoni et al., 1995) 2-D (slope profile) Hybrid Yes No Rockfall (Dr. Spang) (Spang and Sonser¨ , 1995) 2-D (slope profile) Rigid body Yes Yes ROFMOD 4.1 (Zinggeler et al., 1990) 2-D (slope profile) Hybrid Yes Yes (Krummenacher and Keusen, 1996) 3-D-GEOTEST-Zinggeler (Krummenacher et al., 2008) 3-D (x,y,z) Hybrid Yes Yes RocFall (Stevens, 1998) 2-D (slope profile) Lumped-mass Yes No Sturzgeschwindigkeit (Meissl, 1998) 2-D (x,y) Lumped-mass No No STONE (Guzzetti et al., 2002) 3-D (x,y,z) Lumped-mass Yes No STAR3-D (Dimnet, 2002) 3-D (x,y,z) Rigid body No Yes (Le Hir et al., 2006) Rocky3 (Dorren and Seijmonsbergen, 2003) 2.5-D (x.y coupled Hybrid Yes Yes with slope profile) HY-STONE (Crosta et al., 2004) 3-D (x,y,z) Hybrid Yes Yes (Frattini et al., 2008) (Agliardi et al., 2009) RockyFor (Dorren et al., 2004) 3-D (x,y,z) Hybrid Yes Yes (Dorren et al., 2006) (Bourrier et al., 2009a) DDA (Yang et al., 2004) RAMMS::Rockfall (Christen et al., 2007) 3-D (x,y,z) Rigid body Yes Yes RockFall Analyst (Lan et al., 2007) 3-D (x,y,z) Lumped-mass Partly No PICUS-ROCKnROLL (Woltjer et al., 2008) 3-D (x,y,z) Lumped-mass Yes Yes (Rammer et al., 2007)

∗ Forest characteristics such as tree density and corresponding diameters can be taken into account explicitly

4.3 Block-slope interaction between the falling block and the slope’s surface. Models are usually classified into two main categories, the rigid- The trajectories of falling rocks can be described as com- body and the lumped-mass methods (Giani, 1992; Hungr binations of four types of motion: free fall, rolling, sliding and Evans, 1988). Rigid-body methods consider the block and bouncing of a falling block (Ritchie, 1963; Lied, 1977; as a body with its own shape and volume, solve the fun- Descoeudres, 1997). The occurrence of each of these types damental equations of dynamics and account for all types strongly depends on the slope angle (Ritchie, 1963). For of block movement, including rotation (Azzoni et al., 1995; steep slopes, free fall is most commonly observed, whereas Cundall, 1971; Descoeudres and Zimmermann, 1987; Fal- for intermediate slopes, rockfall propagation is a succession cetta, 1985). Lumped-mass methods consider the block to of free falls and rebounds. For gentle slopes, the prevalent have either no mass or a mass concentrated into one point motion types are rolling or sliding. and do not take into account either the shape of the blocks A significant number of rockfall simulation programmes or rotational movement (Guzzetti et al., 2002; Hoek, 1987; exist to perform trajectory analyses. The challenge is not in Hungr and Evans, 1988; Piteau and Clayton, 1977; Ritchie, the free flight simulation, but in modelling the interactions 1963; Stevens, 1998).

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A.A. Volkwein Volkwein et et al.: al.: ReviewRockfall on review rockfall characterisation and structural protection 262711

andthethe slope slope surface’s (type and irregularity size of debris). and the rock shape, the rolling motionThe transition is more a condition succession between of small the bounces. bouncing and the rollingTherefore, mode is most discussed rockfall in Piteau models (1977), simulate Hungr trajectories and Evans as (1988)successions and Giani of free (1992). fall and bouncing The transition phases. from Only sliding a few con- to rollingsider sliding is defined and in rolling Bozzolo motions et al. (1988). (e.g., Azzoni et al., 1995; BozzoloThe whole and rockfall Pamini, trajectory1986; Statham is sometimes, 1979). modelled In these as models the slidinga tangential or rolling damping of a mass coefficient on a sloping related surface to the with rolling an aver- and/or agesliding friction friction angle between assumed block to be and representative slope is introduced. of the mean The energysliding losses friction along is thedefined block’s by path means (Evans of the and normal Hungr, 1993;compo- Govi,nent with 1977; respect Hungr to and the Evans, soil surface 1988; of Japan the block’s Road Associa-weight action,cording 1983; to Coulomb’sLied, 1977; law. Rapp, For 1960; rolling Toppe, motion, 1987b). according This to methodStatham (called(1979), the a fairly Fahrb accurateoschung,¨ description the shadow is angle also given or the by coneusing method) Coulomb’s provides law awith quick a rolling and low-cost friction preliminary coefficient de- that lineationdepends of on areas the characteristics endangered by of rockfall, the block either (size on and a local shape) orand a regional the slope scale (type (Jaboyedoff and size of and debris). Labiouse, 2003; Meissl, 2001).The transition condition between the bouncing and the rolling mode is discussed in Piteau and Clayton (1977), 4.3.2 Rebound models Hungr and Evans (1988) and Giani (1992). The transition Fig. 6. Definition of the block velocity before and after rebound. Fig. 6. Definition of the block velocity before and after rebound. Bouncingfrom sliding occurs to rolling when is the defined falling in blockBozzolo collides et al. with(1988 the). slopeThe surface. wholerockfall The height trajectory of the bounce is sometimes and the modelled rebound as di- the sliding or rolling of a mass on a sloping surface with an aver- V + components of the velocity after rebound also allow the rection depend on several parameters characterizing the im- n age friction angle assumed to be representative of the mean definitionThere are of other a plane programmes called the reflected that could plane. be considered The angle asδ pact conditions. Of the four types of movement that occur energy losses along the block’s path (Evans and Hungr, 1993; hybrid,between taking these advantagetwo planes of is thecalled fast the and deviation easy simulation angle. The of during rockfall, the bouncing phenomenon is the least well freenormal, flight tangential for lumped and rotational masses whileω+ velocities considering after geometri- rebound understoodGovi, 1977 and; Hungr the most and difficult Evans, to1988 predict.; Japan Road Associa- calare and computed mechanical from characteristics the normal, tangential of the slope and rotational and the blockω− tionA, number1983; Lied of rockfall, 1977; modelsRapp, 1960 represent; Toppe the, 1987b rebound). This in tovelocities model the before impact rebound (Azimi using and Desvarreux a rebound, model,1977; Bozzolo and the amethod simplified (called way the by Fahrbone oroschung,¨ two overall the coefficients,shadow angle which or the anddeviation Pamini angle, 1986δ is; Dorren determined, et al. leading, 2004; toJones the complete et al., 2000 def-; arecone called method) restitution provides coefficients. a quick and Some low-cost models preliminary use only de- Kobayashiinition of the et rock al., 1990 velocity; Pfeiffer after rebound. and Bowen, 1989; Rochet, onelineation restitution of areas coefficient, endangered quantifying by rockfall, the either dissipation on a local in 1987b; Crosta et al., 2004). termsor a regional of either scale velocity (Jaboyedoff magnitude and loss Labiouse (Kamijo, 2003 et al.,; 2000;Meissl, Paronuzzi,2001). 1989; Spang and Rautenstrauch, 1988; Spang and 4.3.1If 3-D Sliding rockfall and simulations rolling models are based on a “pseudo-2-D” Sonser,¨ 1995) or kinetic energy loss (e.g., Azzoni et al., approach (see Sect. 4) the block’s tangential V − and nor- 4.3.2 Rebound models t 1995; Bozzolo and Pamini, 1986; Chau et al., 1999a; Ur- malSlidingV − mainlyvelocity occurs components at small (before velocities, rebound) when a with block respect starts n ciuoli,Bouncing 1988). occurs In this when case, the an falling assumption block regarding collides thewith re- the toto the move slope or surface comes allow to rest. definition It is not of accounted a plane called for in the many inci- slope surface. The height of the bounce and the rebound di- rockfall models because it does not entail large+ propagations bound direction is necessary to fully determine the velocity dent plane (Fig. 6). Similarly, the tangential Vt and normal rection depend on several parameters characterising the im- of+ the blocks. Pure rolling is quite a rare motion mode, except vector after impact (i.e. the α+ angle in Figure 6). The Rv V components of the velocity after rebound also allow the pact conditions. Of the four types of movement that occur onn soft soils when the boulder penetrates the soil (Bozzolo coefficient is considered for the formulation in terms of ve- definition of a plane called the reflected plane. The angle δ during rockfall, the bouncing phenomenon is the least under- and Pamini, 1986; Ritchie, 1963). The distinction between locity loss and the RE coefficient is used for the formula- between these two planes is called the deviation angle. The stood and the most difficult to predict. the rolling and sliding modes is sometimes+ difficult since a tion in terms of kinetic energy (neglecting in general the ro- normal, tangential and rotational ω velocities after rebound A number of rockfall models represent the rebound in combination of the two movements can occur (Descoeudres,− tational part): are computed from the normal, tangential and rotational ω a simplified way by one or two overall coefficients, which 1997; Giani, 1992). On stiffer outcropping materials, due to + + 2 + 2 velocities before rebound using a rebound model, and the are calledV restitution coefficients.1/2[I(ω ) Some+ m( modelsV ) ] use only the slope surface’s irregularity and the rock shape, the rolling RV = and RE = (3) deviation angle δ is determined, leading to the complete def- one restitutionV − coefficient,1/2[ quantifyingI(ω−)2 + m the(V − dissipation)2] in motion is more a succession of small bounces. inition of the rock velocity after rebound. terms of either velocity magnitude loss (Kamijo et al., 2000; Therefore, most rockfall models simulate trajectories as However, the most common definition of block rebound Paronuzzi, 1989; Spang and Rautenstrauch, 1988; Spang and successions of free fall and bouncing phases. Only a few con- involves differentiation into tangential Rt and normal Rn Sonser¨ , 1995) or kinetic energy loss (e.g., Azzoni et al., 4.3.1sider sliding Sliding and and rolling rolling motions models (e.g., Azzoni et al., 1995; restitution coefficients (Budetta and Santo, 1994; Evans and 1995; Bozzolo and Pamini, 1986; Chau et al., 1999a; Ur- Bozzolo and Pamini, 1986; Statham, 1979). In these models Hungr, 1993; Fornaro et al., 1990; Giani, 1992; Guzzetti Slidinga tangential mainly damping occurs coefficient at small velocities, related to when the rolling a block and/or starts etciuoli al.,, 2002;1988). Hoek, In this 1987; case, Kobayashi an assumption et al., regarding 1990; Pfeiffer the re- andbound Bowen, direction 1989; is Piteau necessary and Clayton, to fully 1976; determine Urciuoli, the 1988; veloc- tosliding move friction or comes between to rest. block It is and not slope accounted is introduced. for in many The + rockfallsliding frictionmodels becauseis defined it does by means not entail of the large normal propagations compo- Ushiroity vector et al., after 2000; impact Wu, 1985): (i.e., the α angle in Fig. 6). The R coefficient is considered for the formulation in terms of ofnent the with blocks. respect Pure to rolling the soil is surfacequite a rare of the motion block’s mode, weight except ac- v + + velocityVt loss and the R Vncoefficient is used for the formu- oncording soft soils to Coulomb’s when the law. boulder For penetratesrolling motion, the soil according (Bozzolo to Rt = and Rn = E (4) V − V − andStatham Pamini (1979),, 1986 a; fairlyRitchie accurate, 1963). description The distinction is also given between by lation int terms of kineticn energy (neglecting in general the theusing rolling Coulomb’s and sliding law with modes a rolling is sometimes friction difficult coefficient since that a Theserotational coefficients part): are used conjointly and characterize the combinationdepends on the of the characteristics two movements of the can block occur (size (Descoeudres and shape), decreaseV in+ the tangential and1/2 the[I (ω normal+)2 +m(V components+)2] of the RV = and RE = (3) 1997; Giani, 1992). On stiffer outcropping materials, due to V − 1/2[I (ω−)2 +m(V −)2] www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2628 A. Volkwein et al.: Review on rockfall characterisation and structural protection

However, the most common definition of block rebound 4.3.3 Barrier effect of trees involves differentiation into tangential Rt and normal Rn restitution coefficients (Budetta and Santo, 1994; Evans and There are only a few spatial rockfall trajectory models that Hungr, 1993; Fornaro et al., 1990; Giani, 1992; Guzzetti explicitly (i.e., spatial distribution of different forest stands, et al., 2002; Hoek, 1987; Kobayashi et al., 1990; Pfeiffer stand densities, distribution of diameters at breast height and Bowen, 1989; Piteau and Clayton, 1976; Urciuoli, 1988; DBH and species) take into account the mitigating effect of Ushiro et al., 2000; Wu, 1985): existing forest cover (e.g., Dorren et al., 2006; Crosta et al., 2004; Krummenacher et al., 2008; Woltjer et al., 2008; Ma- + + Vt Vn suya et al., 2009). These models would allow determining Rt = − and Rn = − (4) Vt Vn optimal combinations and locations of technical and silvi- cultural measures at a given site. Furthermore, they enable These coefficients are used conjointly and characterise the rockfall hazard zoning with and without the mitigation ef- decrease in the tangential and the normal components of the fect of forests. Recent data describing the energy dissipa- block velocity, respectively. This definition fully determines tive effect of trees is published in Dorren and Berger (2006) the rebound direction (α+ angle in Fig. 6) and no further as- and Jonsson (2007). Older data seriously underestimated the sumption is needed to characterise it. energy dissipative capacity of trees, i.e., mature coniferous An alternative approach is based on impulse theory trees were thought to dissipate up to 15 kJ instead of 200– (Fremond´ , 1995; Goldsmith, 1960; Stronge, 2000) and con- 500 kJ (cf. the review on the interaction between trees and siders the change in the momentum of the block during the falling rocks by Dorren et al., 2007). compression and restitution phases of impact (Bozzolo et al., 1988; Descoeudres and Zimmermann, 1987; Dimnet, 2002; 4.3.4 Modelling variability Dimnet and Fremond´ , 2000). According to Newton’s theory of shocks, the restitution A deterministic prediction of the interaction between a block coefficients should have a constant value irrespective of the and the slope’s surface is not relevant because our under- impact energy (“elastic” collision) and of the impact direc- standing of the phenomena is insufficient and many param- tion. However, since this assumption does not match obser- eters are not completely characterised. Uncertainties are re- vations, several models have been developed to account for lated to the block (shape, dimensions), the topography (in- the dependency of the block velocity after rebound on the clination, roughness) and the outcropping material (strength kinematical conditions before impact (Bourrier et al., 2009b; and stiffness). As a consequence, even with a thorough field Chau et al., 2002; Dorren et al., 2004; Heidenreich, 2004; survey, data collection cannot be exhaustive and the rebound Pfeiffer and Bowen, 1989). These models can be considered prediction should take into consideration a certain variability. as extensions to classical models based on constant restitu- Stochastic rebound models have, therefore, been pro- tion coefficients. posed (Agliardi and Crosta, 2003; Azzoni et al., 1995; Bour- In addition, some very detailed models have been elab- rier et al., 2009b; Dudt and Heidenreich, 2001; Guzzetti orated for the interaction between the block and the slope et al., 2002; Paronuzzi, 1989; Pfeiffer and Bowen, 1989; (Azimi et al., 1982; Falcetta, 1985; Ushiro et al., 2000). They Wu, 1985). A model correctly assessing rebound variabil- differentiate between impact on hard and soft ground materi- ity should separate the different sources of uncertainty (due als, considering for the latter the penetration of the block into to randomness of characteristics or lack of data) and quan- the soil modelled with a perfectly plastic or elasto-plastic be- tify the variability associated with each of them separately. haviour. As for the fragmentation of blocks that can occur The variability of the bouncing phenomenon is quantified by with impact on hard ground, it is rarely accounted for (Az- several statistical laws that need to be calibrated based on the imi et al., 1982; Chau et al., 1998a; Fornaro et al., 1990) as statistical analysis of impact results. modellers generally assume that unbreakable blocks propa- Back-analysis of observed events or field experiments is gate further than breakable ones. not feasible for this purpose because either the dataset is in- Finally, apart from the rigid-body models which integrate complete or reproducible impact conditions are difficult to the fundamental equations of motion, only a few models ac- achieve. On the other hand, extensive laboratory experi- count for the rotational velocity along the block path. In this ments, or thoroughly calibrated numerical simulations, can case, a relationship between translation and rotation is usu- be used. These approaches have already been used for coarse ally established, assuming that blocks leave the ground after soils (Bourrier et al., 2009b). The challenge for such an ap- impact in a rolling mode. Either sticking or slipping condi- proach is the generation of appropriate datasets composed of tions are considered at the contact surface (Chau et al., 2002; results for different ground properties and kinematical con- Kawahara and Muro, 1999; Ushiro et al., 2000). ditions before rebound.

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2629

Table 2. Parameters assumed to influence the bouncing phe- and Statham, 1975; Statham and Francis, 1986). Indeed, nomenon (Labiouse and Descoeudres, 1999). when the falling block size is greater than the average debris particle size, rolling is the prevailing movement and the Slope Rock Kinematics block propagates further (Bozzolo and Pamini, 1986; Evans characteristics characteristics and Hungr, 1993; Giani, 1992; Kirkby and Statham, 1975; Ritchie, 1963; Statham and Francis, 1986). However, on strength strength velocity (translational loose soils, increasing block weight induces greater plastic stiffness stiffness ... and rotational) deformation of the soil (formation of a bigger crater), which roughness weight incidence angle somewhat reduces the previous influence. As for the shape inclination size configuration of... shape ...the rock at impact of blocks, tests carried out with cubic blocks have shown that the impact configuration (e.g., impact on face, edge or cor- ner) has a very significant influence on the block’s movement during and after impact (Giani, 1992; Heidenreich, 2004). Bouncing is found to depend significantly on the transfer 4.3.5 Relevance of impact parameters of energy between the block and the slope. The initial kinetic energy of the block is converted into kinetic energy after re- As emphasized by the number of different definitions of the bound, together with diffused and dissipated energies inside restitution coefficients used in computer codes, the rebound the slope material. Elastic deformation of the slope material of rock blocks on a slope’s surface is still a poorly understood also occurs, but, in general, can be neglected. Energy diffu- phenomenon. In particular, modelling by means of constant sion is due to wave propagation from the impact point (Bour- restitution coefficients only as a function of the slope mate- rier et al., 2008; Giani, 1992), while energy dissipation is re- rial is not very satisfactory, at least from a scientific point-of- lated to frictional (plastic) processes inside the slope material view. Indeed, as mentioned above, the rebound also depends during impact (Bourrier et al., 2008; Bozzolo and Pamini, on several parameters related to the boulder and its kinemat- 1986; Giani, 1992; Heidenreich, 2004) and is also due to ics before impact (Table 2). Experimental investigations of block and/or soil particle fragmentation (Azimi et al., 1982; the influence of these parameters are, therefore, worthwhile Fornaro et al., 1990; Giani, 1992). The magnitude of energy for reaching a deeper understanding of the mechanisms oc- dissipation is mainly governed by the ratio between the block curring during impact and to put forward mathematical ex- and the slope particles (Bourrier et al., 2008; Statham, 1979), pressions between the restitution coefficients and those pa- the soil properties (Azzoni et al., 1995, 1992) and the block rameters. These studies also attempt to determine reliable shape and incident orientation (Chau et al., 1999a; Falcetta, values for the parameters used in the rebound models. 1985; Heidenreich, 2004). Energy diffusion and dissipation Experimental investigations were carried out both in the processes are also strongly dependent on the kinetic energy field (e.g., Azzoni and De Freitas, 1995; Azzoni et al., of the block before impact, which is related to its mass m and − − 2 1992; Berger and Dorren, 2006; Bozzolo et al., 1988; Broili, its velocity before rebound V , i.e., Ec = 1/2×m×(V ) . 1977; Evans and Hungr, 1993; Fornaro et al., 1990; Gia- The effects of variations in block mass (Jones et al., 2000; comini et al., 2009; Giani, 1992; Japanese highway public Pfeiffer and Bowen, 1989; Ushiro et al., 2000) and in block corporation, 1973; Kirkby and Statham, 1975; Kobayashi velocity before rebound (Urciuoli, 1988; Ushiro et al., 2000) et al., 1990; Lied, 1977; Pfeiffer and Bowen, 1989; Ritchie, are different due to the linear and square dependencies. 1963; Statham, 1979; Statham and Francis, 1986; Teraoka Another very important feature observed in many exper- et al., 2000; Urciuoli, 1996; Wu, 1985; Yoshida, 1998) iments is the strong influence of the kinematical conditions and in the laboratory (Azimi and Desvarreux, 1977; Az- before rebound. In particular, experiments show that small imi et al., 1982; Bourrier, 2008; Camponuovo, 1977; Chau impact angles result in greater energy conservation by the et al., 1998a, 1999a, 2002, 1999b, 1998b; Heidenreich, 2004; block (Bozzolo and Pamini, 1986; Chau et al., 2002; Hei- Kamijo et al., 2000; Kawahara and Muro, 1999; Murata and denreich, 2004; Ushiro et al., 2000; Wu, 1985). Indeed, only Shibuya, 1997; Statham, 1979; Ujihira et al., 1993; Ushiro a small part of the kinetic energy before impact is associ- et al., 2000; Wong et al., 2000, 1999; Masuya et al., 2001). ated with normal to soil surface velocity and consequently These experiments contributed to determining the most im- less energy is dissipated into the soil. On the other hand, a portant impact parameters and to quantifying their influence significant part of the kinetic energy related to the tangential on block rebound. component of velocity is retained by the block after impact Experimental investigations have shown the dependence and a part of it (up to 30 %) is transformed into rotational en- of block bouncing on geometrical parameters and, in par- ergy (Kawahara and Muro, 1999; Ushiro et al., 2000). The ticular, on the roughness of the slope (usually characterised reflected rotational velocity depends, to a large extent, on the by the ratio of block size to average debris particle size). incidence angle and on the soil type. It is governed by the The influence of slope roughness on rebound is generally re- interaction conditions at the contact surface, either sticking ported as an explanation for size sorting along slopes (Kirkby or slipping (Chau et al., 2002). www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2630 A. Volkwein et al.: Review on rockfall characterisation and structural protection

Given the limited amount of results, most of the above- 2008; Chau et al., 2002; Heidenreich, 2004). From a prac- mentioned experimental investigations were insufficient for tical point-of-view, the implementation in computer codes a thorough understanding of the phenomenon or for statis- of the mathematical relationships deduced from the labora- tical and parametric analyses. Therefore, some systematic tory tests should lead to better predictions of rebound. This experimental investigations were carried out in laboratories can improve the determination of areas at risk, particularly on small- and medium-scale models (Bourrier, 2008; Chau for sites where no rockfall events have been experienced and et al., 2002; Heidenreich, 2004). These experiments were monitored. dedicated to analyse the influence on the rebound of param- However, from a scientific point-of-view, the relevance of eters related to the ground, the block and the kinematics. restitution coefficients expressed for the mass centre of the Blocks (mainly spherical) were released on different soil ma- blocks (Eqs. 3–4) is challenged (Labiouse and Heidenreich, terials with different degrees of compaction either normally 2009). Indeed, from a thorough analysis of impact films, or with different incidences using specific throwing devices. the movement of blocks during impact is found to consist All experiments were filmed using high-speed cameras. Con- of three main interdependent mechanisms: a normal transla- trary to field experiments, controlled laboratory experiments tion (penetration), a tangential translation (sliding) and a ro- provide precisely measured and reproducible results that are tation. It is illusory to model this complexity by means of two valid over larger domains. The trends obtained can, there- overall restitution coefficients expressed for the mass centre fore, be used with confidence to improve rebound models. of the block, as adopted by most existing rockfall trajectory The results from laboratory experiments also provide a lot of codes. Only rigid-body methods that take into consideration information, much of it relevant in the calibration of numer- the shape of the blocks and fully consider the interaction be- ical models of the impact that can, in turn, be used to study tween boulder and ground material at the contact surface (in- energy transfer during impact (Bourrier et al., 2008). How- cluding the creation of a crater) would be able to model the ever, the quantitative interpretation of laboratory experiments impact phenomenon. is not straightforward, because matching the similitude requirements for all the parameters involved in the dynamic 4.3.6 Concluding remarks on block-slope interaction process can be difficult (Bourrier, 2008; Camponuovo, 1977; Heidenreich, 2004). The number of different rebound models used in rockfall The main results gathered from these experimental investi- simulations emphasizes that block-slope interaction is still gations confirm the general trends obtained in previous stud- poorly understood. This complex phenomenon depends not ies. Regarding the influence of the slope material charac- only on the ground conditions (stiffness, strength, roughness, teristics, the motion of the block during and after impact is inclination), but also on the block’s characteristics (weight, found to be significantly influenced by the degree of com- size, shape, strength) and the kinematics before impact (ve- paction of the soil material and somewhat less by its friction locities, collision angle, configuration of the block at impact). angle (Bourrier, 2008; Heidenreich, 2004). As for the influ- One should, therefore, keep in mind that if common re- ence of the kinematics before impact, experiments confirm a bound models are used, the predictive ability of rockfall sim- clear dependency of the restitution coefficients on the block ulation is conditioned by a good calibration of its parameters velocity and the impact angle on the slope surface. The in- on already experienced or monitored rockfall at the site of fluence of the latter seems to prevail (Bourrier, 2008; Chau interest. In cases where data on natural or artificial events is et al., 2002; Heidenreich, 2004). Additionally, the depen- lacking for the specific site, one should be aware that calcu- dency on block mass and size is more marked for normal lations of rock trajectories can be very misleading when per- than for smaller impact angles because energy transfer to the formed with the restitution coefficients stated in the literature soil is greater for normal impact (Bourrier, 2008; Heiden- or assessed from in situ rockfall events or back-analyses of reich, 2004). The shape of the block and its configuration events on other slopes. at impact were also shown to have a clear influence on the To achieve better reliability in trajectory simulations, sev- motion of the block after impact and especially on the rota- eral studies have been carried out, or are still in progress, tional rate. Finally, the large amount of experimental results to develop rebound models that account for the influence of allowed, for coarse soils in particular, quantifying the high the most important impact parameters. The parameters can variability of the kinematics of the block after rebound de- then be calibrated by a more objective field data collection. pending on both the surface shape and the geometrical con- To achieve this goal, many experimental investigations were figuration of soil particles near the point of impact (Bourrier conducted, either in the field or in the laboratory, to reach et al., 2009b, 2008). a deeper understanding of the mechanisms involved during The results from the above-mentioned laboratory exper- impact and to quantify the influence of the most important iments allowed determining the most important geometri- geometrical and geotechnical parameters. After a thorough cal and geotechnical parameters that influence rebound and calibration using experimental data, numerical modelling can proposing mathematical expressions for the restitution coef- contribute to studying energy transfer during impact and to ficients as a function of the impact characteristics (Bourrier, assess the influence of parameters outside the range of tested

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. VolkweinA. Volkwein et al.: et al.: Review Rockfall on rockfallreview characterisation and structural protection 263115 values.Azzoni From and these De Freitas,studies, 1995;mathematical Falcetta, expressions 1985; Giani, for 1992; the reboundHungr models’ and Evans, parameters 1988), the can characteristics be derived as of a functionmotion after of theimpact impact are characteristics. conditioned by several factors other than the slope Implementationmaterial properties, of the such rebound as the weight, models size in rockfall and shape simula- of the tionblocks, codes should as well provide as their more velocity, accurate collision predictions angle and of config- rock- falluration trajectories at impact. and energies Consequently, and consequently the restitution improve coefficients the delineationthat characterize of areas at the risk rebound and the of design blocks of during protection rockfall struc- are not constant parameters but simply a function of the slope tures. material. Owing to our incomplete knowledge both of and in mod- 4.4 Rebound model calibration elling the bouncing phenomenon and to the rather subjective description of the slope material, the reliability of the simu- In general, the rebound parameters used for trajectory calculation results could be improved. This is evident when com- lations are estimated on the basis of a rough description of paring the results provided by different models on a specific the slope material (rock, scree deposits, loose soil), some- site, or even by the same program used by different users times(Berger complemented and Dorren, by 2006; information Labiouse, regarding 2004; Labiouse its roughness, et al., its degree2001). of The compaction limits of predictions and the vegetation are also clear cover. when Now, values as mentionedof model by parameters several authors taken from who the have literature experienced or obtained natu- by ral and/orin situ tests artificial or back-analyses in situ rockfall of natural(e.g., Azimi events et on al. particular, 1982; Azzonislopes and doDe not Freitasprovide, satisfactory1995; Falcetta results, 1985 when; Giani used, on1992 other; Hungrslopes. and Evans, 1988), the characteristics of motion after impactTo are achieve conditioned goodreliability by several of factors trajectory other predictions, than the slope the materialprogram properties, parameters such must as the be thoroughly weight, size calibrated and shape at theof the site blocks,of interest. as well For as this their purpose, velocity, during collision the field angle data and collection, config- urationparticular at impact. attention Consequently, should be paid the restitution to gain information coefficients on thatthe characterise rockfall paths the of rebound previous of events, blocks such during as scars rockfall on cliffs, are notimpacts only a function on slopes, ofthe damage slope to material. vegetation and accumulation Owingzones. to Provided our incomplete the numerical knowledge model isboth well of calibrated and in mod- with ellingthese the field bouncing observations, phenomenon confidence and to in the the rather trajectory subjective results descriptionwill be greatly of the enhanced. slope material, the reliability of the simu- Fig. 7. Schematic illustration of rockfall traces on the ground and lation results could be improved. This is evident when com- Fig. 7. Schematic illustration of rockfall traces on the ground and tree branches paring4.4.1 the Fieldresults data provided collection by differentand analysis models on a specific tree branches. site, or even by the same programme used by different users For a complete back-analysis of the rock’s trajectory the alti- (Berger and Dorren, 2006; Labiouse, 2004; Labiouse et al., tudes of the release and deposition positions must be known. f/s = 1/12 for shallow jumps 2001). The limits of predictions are also clear when values sible should be detected with their (inclined) distance s and In addition, all traces should be recorded on a map in or- of model parameters taken from the literature or obtained by the slope inclination. Additional traces above ground allow- der to obtain the horizontally projected length of the trajec- in situ tests or back-analyses of natural events on particular ingIf for the a tracesderivation on the of ground the jump cannot height be should assigned also to the be logged.single tory. Along this, as many follow-up impact craters as pos- However,jumps because these of traces several usually overlapping belong rockfall to the centre trajectories of gravity the slopessible do should not provide be detected satisfactory with their results (inclined) when used distance on others and slopes. ofterrain the block, profile whereas of the potential the traces trajectory on the shouldground be belong recorded. to its the slope inclination. Additional traces above ground allow- This may allow a later modelling of the rock’s movements. To achieve good reliability of trajectory predictions, the lower boundary. This has to be considered dealing with small ing for a derivation of the jump height should also be logged. From the field data the ”air parabolas” of the single jumps programme parameters must be thoroughly calibrated at the jump heights in combination with large blocks. In rare cases, However, these traces usually belong to the centre of grav- can be derived with the corresponding velocities. The upper site of interest. For this purpose, during the field data col- even the (vertically measured) maximum jump height f in ity of the block whereas the traces on the ground belong to impact crater O is the starting point of a parabola, the other lection, particular attention should be paid to gain informa- the middle of the jump (s/2 if the inclination of the slope its lower boundary. This has to be considered dealing with end is defined by the lower crater E. The start velocity is tionsmall on the jump rockfall heights paths in combination of previous events, with large such blocks. as scars In rare on doesn’t change significantly) can be measured (Fig. 7). In mostcalled cases,vO and however,vE defines the thejump next height impactf velocitymust be split estimated into cliffs,cases impacts even the on slopes,(vertically damage measured) to vegetation maximum and jump accumu- height horizontal and vertical components x and z: lationf in zones. the middle Provided of the the numerical jump (s/2 modelif the is inclination well calibrated of the based on the inclined jump length s. Observations show the withslope these doesn’t field observations, change significantly) confidence can bein themeasured trajectory (Fig. re- 7). following relations to be valid for characteristic jumps: vOx = lift-off velocity in horizontal direction sultsIn will most be cases, greatly however, enhanced. the jump height f must be estimated based on the inclined jump length s. Observations show the f/s = 1/6 for high jumps v = following relations to be valid for characteristic jumps: Oz lift-off velocity in vertical direction 4.4.1 Field data collection and analysis f/s = 1/8 for normal jumps v = impact velocity in horizontal direction For a completef/s = 1back-analysis/6 for high jumps of the rock’s trajectory, the alti- f/sEx= 1/12 for shallow jumps tudes of the release and deposition positions must be known. In addition,f/s = all 1 traces/8 for normal should jumps be recorded on a map in or- If thevEz traces= impact on the velocity ground incannot vertical be direction assigned to the sin- der to obtain the horizontally projected length of the trajec- gle jumps because of several overlapping rockfall trajecto- tory. Along this, as many follow-up impact craters as pos- ries, the terrain profile of the potential trajectory should be www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 16 A. Volkwein et al.: Rockfall review 263216 A. Volkwein et al.: Review on rockfall characterisationA. Volkwein and et structuralal.: Rockfall protection review Table 3. Start and end velocities of a parabolic trajectory for differ- Tableent values 3. Start of jump and end height velocities of of a a parabolic parabolic trajectory trajectory for for differ- differ- ent values of jump height height Jump height f 3.50 m 3.75 m 4.0 m JumpJumpJump height heightlengthfs 3.50303.50.0 m m 3.7530 3.75.0m m30 4.0 4.0.0m m ◦ ◦ ◦ JumpInclinationJump length lengthβs 3030.0.400 m30 30.0.400 m30 30.0.400m m x 22.98m◦ 22.98m◦ 22.98m◦ InclinationJumpInclination lengthβ 4040◦ 40 ◦ 40◦ JumpJump length xz 1922.98.28 m19 22.98.28m m19 22.98.28m m Jump length x 22.98m 22.98m 22.98m Jump length z 19.28 m 19.28 m 19.28 m Jump length z 19.28m 19.28m 19.28m Lift-off velocity vOx 13.60m/s 13.14m/s 12.72m/s −1 −1 −1 Lift-offLift-off velocity vvOxOz 13.603.13 mm/s s 13.142.45 mm/s s 12.721.82 mm/s s Lift-off velocity vOx 13.60m/s−1 13.14m/s−1 12.72m/s−1 Lift-offLift-off velocity vvOzO 3.1314. m0m/s s 2.4513.4 mm/s s 1.8212.9 mm/s s Lift-off velocity vOz 3.13m/s−1 2.45m/s−1 1.82m/s−1 Lift-off velocity vO 14.0 m s 13.4 m s 12.9 m s Lift-off velocity v 14.0m/s 13.4m/s 12.9m/s Impact velocity vO 13.60m/s 13.14m/s 12.72m/s Ex −1 −1 −1 ImpactImpact velocity vvEx 13.6019.70 mm/s s 13.1419.60 mm/s s 12.7219.54 mm/s s Impact velocity v Ez 13.60m/s−1 13.14m/s−1 12.72m/s−1 ImpactImpact velocity vvEzEx 19.7023. m9m/s s 19.6023.6 mm/s s 19.5423.3 mm/s s E −1 −1 −1 ImpactImpact velocity velocity vvEEz 1923.9.70 mm/s s 1923.6.60 mm/s s 1923.3.54 mm/s s Impact velocity vE 23.9m/s 23.6m/s 23.3m/s

Fig. 8. Details of air parabola with velocity vectors

Fig. 8. Details of air parabola with velocity vectorsvectors. The jump height f is defined in the middle of the jump length s (Fig. 8). The horizontal and vertical fractions of the recorded. This mayf allow a later modelling of the rock’s jumpThejump length heights with a slopeis defined inclination in theβ middleare : of the jump movements.length s (Fig. 8). The horizontal and vertical fractions of the jumpFrom length the fields with data, a slope the “air inclination parabolas”β are of the: single jumps canx = bes derived cosβ withand thez = correspondings sinβ velocities. The upper(5) impactx = s cosβ crater Oandis thez = startings sinβ point of a parabola, the other(5) endThe is defined coordinate by the components lower crater of theE. lift-off The start velocity velocityvO are is called vO and vE defines the next impact velocity split into The coordinater g components of the lift-offr g velocity vO are horizontalvOx = x and verticaland componentsvOz = (z −x4andf) z: (6) r 8f r 8f vOx = lift-offg velocity in horizontal directiong vOx = x and vOz = (z − 4f) (6) resulting in8f a total lift-off velocity of 8f vOz = lift-off velocity in vertical direction resultingp= in a total lift-offr velocityg of v v=Ex x2impact+ (z − velocity4f)2 in horizontal. direction (7) O 8f = r vEzp 2 impact velocity2 ing vertical direction Fig. 9. Lift-off and impact velocity for an assumed jump height of vO = x + (z − 4f) . (7) Fig. 9. Lift-off and impact velocity for an assumed jump height of 8f 2 f/s = 1/8 as a tool for rapid trajectory analyses in the field Herein, g stands forf the gravitational constant g = 9.81m/s f/s = 1/8 as a tool for rapid trajectory analyses in the field. The jump height is defined in the middle of the jump Fig. 9. Lift-off and impact velocity for an assumed jump height of lengthand thes (Fig. vertical8). The direction horizontal is used and with vertical a positive fractions sign of if the di- Herein, g stands for the gravitational constant g = 9.81m/s2 f/s = 1/8 as a tool for rapid trajectory analyses in the field jumprected length upwards.s with Accordingly, a slope inclination the impactβ are: velocity vE is and the vertical direction is used with a positive sign if di- As an example, the series of measured values (see Fig. 7) r xrected= s cosβ upwards.and Accordingly,z =ps sinβ the impact velocityg vE is (5) would5 Structural result in thecountermeasures velocities shown in Table 3. The different v = v + v = x2 + (z + 4f)2 . (8) E Ex Ez 8f assumed jump heights of 3.5−4.0 m result in similar lift-off The coordinate components of the lift-offr velocity vO are p 2 2 g 5and Structural impact velocities. countermeasures vE = vExr + vEz = x + (z + 4f)r . (8) As an example,g the series of measuredg8f values (see Fig. 7) InThe thedetermination case of infrastructure of the start or buildings and end velocitiessituated withinvO and a vOx = x and vOz = (z−4f ) (6) would result8f in the velocities shown in Table8f 3. The different vrockfallE can be hazard simplified zone either and suitablespeeded newly up by planned/built making use pro- of a assumedAs an example, jump heights the series of 3.5 of− measured4.0 m result values in similar (see Fig. lift-off 7) Intection the case measures of infrastructure are needed or such buildings as are situated necessitated within by a resulting in a total lift-off velocity of diagram that depends on the jump length s and slope incli- wouldand impact result in velocities. the velocities shown in Table 3. The different rockfallnationchangedβ hazard boundariespaired zone with either of an rockfall assumed suitable occurrence.jump newly height planned/built This relationship section pro- assumedTheq determination jump heightsr of of3 the.5 − start4.0 andm result endvelocities in similar lift-offv and gives an overview of modern protection systems and provides 2 2 g O tectionof f/s = measures1/8. Such are graphics needed or can such be easily as are prepared necessitated for any by vO = x +(z−4f ) . (7) andvE impactcan be velocities. simplified and8f speeded up by making use of a di- changedothera short relation summary boundaries of f/s for. dams, of rockfall embankments occurrence. and ditches This sectionin sec- agramThe determination that depends onof the the startjump and length ends velocitiesand slopev inclina-O and givestion 5.2. an overview A more comprehensive of modern protection state-of-the-art systems and report provides deals g g = . −2 Herein,vEtioncanβ bepairedstands simplified with for the an and gravitational assumed speeded jump up constant by height making relationship use9 81 of m a s di- of awith short fences summary and galleries for dams, (sections embankments 5.3 and and 5.4). ditches For forests, in sec- and the vertical direction is used with a positive sign if di- agramf/s = that 1/ depends8. Such on graphics the jump can length be easilys and prepared slope inclina- for any tion5reference Structural 5.2. A should more countermeasures becomprehensive made to a recent state-of-the-art review of the report protection deals rectedtionotherβ upwards.paired relation with of Accordingly,f/s an. assumed the jump impact height velocity relationshipvE is of withof forests fences in and section galleries 5.5. (sections 5.3 and 5.4). For forests, f/s = 1/8. Suchq graphics can ber easily prepared for any referenceIn the case should of infrastructure be made to a recent or buildings review of situated the protection within 2 2 g votherE = v relationEx +vEz of=f/sx. +(z+4f ) . (8) of forests in section 5.5. 8f a rockfall hazard zone, either suitable newly planned/built

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2633 protection measures are needed or are necessitated by its kinematics (velocity and impact angle) and to the layer changed boundaries of rockfall occurrence. This section of absorbing material (thickness, compaction degree). For gives an overview of modern protection systems and pro- rockfall protection galleries, the action on the structure is also vides a short summary for dams, embankments and ditches in found to depend on the structure’s stiffness. Sect. 5.2. A more comprehensive state-of-the-art report deals Most of the above-mentioned studies provided quantitative with fences and galleries (Sects. 5.3 and 5.4). For forests, ref- data on the temporal evolution of the impact force induced by erence should be made to a recent review of the protection of the block (measured accelerations by means of accelerome- forests in Sect. 5.5. ters on the boulder and/or using image processing of high- speed camera films to obtain the evolution of velocity over 5.1 Action of rocks on protection structures time), on the penetration of the block into the absorbing material and, for some of them, on the earth pressures acting For a long time, estimations of the impact load caused by a at the base of the cushion layer (i.e., on the structure). The rockfall were only drawn from empirical relationships based data gathered provide information on the transfer of energy on experimental observations. Then several other formu- during the impact and on the force exerted on the structure. lations were developed from theoretical considerations as- Formulas were worked out to assess the magnitude of the suming the ground behaviour to be elastic, plastic or elasto- forces, with the aim of improving the design of protection plastic. The first family of relationships, derived from structures (e.g., SBB, 1998). However, these results and for- Hertz’s elastic contact theory, assumes that a rigid ball im- mulas must be interpreted with caution because the thick- pacts an elastic medium (Goldsmith, 1960; Japan Road As- ness of the absorbing cushion and the boundary conditions sociation, 1983; Lang, 1974; Tonello, 1988). Other formula- strongly influence the dynamics of the interaction (Calvetti, tions are based on a plastic or elasto-plastic behaviour of the 1998; Montani-Stoffel, 1998). ground material (Azimi and Desvarreux, 1988; Habib, 1976; When carefully calibrated on the experimental data, nu- Heierli, 1984; Lang, 1974; Tonello, 1988). Recently, for- merically modelling the impacts can help to better under- mulas were derived from the penetration of nondeformable stand and quantify the energy diffusion and dissipation inside ogive-nose projectiles onto concrete and soil targets (Pichler the absorbing cushion. It can also contribute to assessing the et al., 2005). For roughly the last decade, many efforts are influence of various parameters that could not be studied, or devoted to the numerical modelling of the impact on rock- only in a limited range of values, during the experimental fall protection structures, using finite element (FE) and dis- campaigns, and to improving the design of protection struc- crete element (DE) methods (Bertrand et al., 2006; Calvetti, tures. 1998; Calvetti et al., 2005; Magnier and Donze´, 1998; Ma- suya and Kajikawa, 1991; Nakata et al., 1997; Nicot et al., 5.2 Embankments and ditches 2007; Peila et al., 2002, 2007; Plassiard et al., 2004). The DE method seems quite promising for studying impact prob- Embankments and ditches belong to the quasi-natural class lems, provided that a careful calibration of the parameters is of protection measures against rockfall. Their construction first achieved. along the side of the infrastructure is efficient and they are To gather data on the action of rocks on protection struc- one of the most reliable protection measures. Therefore, they tures and then to calibrate numerical codes, experimental are more likely to be used to protect permanent buildings. campaigns are essential. Several half-scale and full-scale Embankments are able to withstand high impact energies experimental studies have been conducted to determine the of e.g., 20 MJ (personal communication with practitioners). damping abilities of the cushion covering rockfall protection However, the cross sections of embankments and ditches re- galleries (often called rock sheds) for design purposes, by quire a rather large area in front of the protected object. dropping blocks of different weights and shapes from var- For structural measures, like fences or galleries, the perfor- ious heights on concrete slabs covered with different ab- mance of the protective system is quite well known and the sorbing materials (Calvetti et al., 2005; Chikatamarla, 2006; planning of protection measures does not have to take into Labiouse et al., 1996; Montani-Stoffel, 1998; Murata and account the deceleration process. However, this has to be Shibuya, 1997; Sato et al., 1996; Schellenberg et al., 2008; clarified for the structural safety of earth embankments. This Yoshida et al., 1988). Other testing campaigns were car- includes the questions: What is the impact load as a func- ried out on gravel layers (Pichler et al., 2005), embankments tion of the impact energy? What is the effect of changing (Blovsky, 2002; Burroughs et al., 1993; Lepert and Corte´, mass or impact velocity? What is the limit state of the em- 1988; Peila et al., 2002; Yoshida, 1999) and composite struc- bankment? What is the influence of soil properties such as tures (Lambert et al., 2009; Lorentz et al., 2006). Paramet- density, strength, angle of internal friction? What is the pene- rical analyses performed in the framework of these experi- tration depth? How does the cross section of an embankment mental campaigns allowed for the determining of the most or ditch affect the interaction with the block? important factors and quantifying their influence on the im- For example and theoretically, the front face taking the impact force. They are related to the block (mass, shape) and pact could be (at least partially) vertical. This might deviate www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2634 A. Volkwein et al.: Review on rockfall characterisation and structural protection 18 A. Volkwein et al.: Rockfall review thebya block damping into alayer vertical to dissipate path and energy its rotation and reduce does not bouncing cause itheight). to roll Furthermore, over the embankment rather low or inclined roll out hillside of a ditch. slopes Inof practice,embankments several being impacts covered on rockfall by a damping embankments layer (being are docu- built mentedwith its where friction the angle) construction will prevent fulfilled a rolling its task block for inclinedto over- hillslidecome the slopes construction even with as the angles material that will represent react with the friction ground anglesfailure of as thesoon construction as the block material. will induce The shear geometry forces of to the embankmentslope. Therefore should, it should therefore, be noted, reflect that more for the the design local of geo- the metricalgeometry boundaries of the embankment and can also (especially be strongly the influencedinclination by of thethe existencehillside slope) and widthshould of be a done hillside with catchment respect to the zone geome- (e.g., beingtry of coveredthe slope by where a damping the construction layer todissipate will be done. energy Ideally and reducethe slope bouncing of the embankment height). Furthermore, will be rectangular rather low to the inclined hills- hillsidelope. slopes of embankments covered by a damping layer (builtThe with deceleration its friction process angle) into will soil prevent has been a rolling investigated block on to Fig. 10. Penetration and deceleration of impacting rocks rocks onto onto con- con- overcomedifferent scales, the construction i.e. small as (Heidenreich, the material 2004), reacts with large ground (Labi- solidated soil of thickness thickness 0.5 0.5 m and and 1.3 1.3 m for different different impact impact ve- ve- failureouse et as al., soon 1996; as theMontani-Stoffel, block induces 1998) shear and forces full to scale the slope. (Ger- locities. Therefore,ber, 2008). it The should main be results noted, are that the for maximum the design deceleration of the ge- ometryand penetration of the embankment of blocks. Both (especially results arethe important inclination for of gal- the 2 hillsideleries (see slope) section should 5.3) be to done design with the respect strength to of the the geometry underly- afeature= 0.8v elasto-plastic/(gt) deformation in the direction of a free(9) ofing the structures slope where and the the construction thickness of will the soilbe done. layer Ideally (Labiouse the surface (valley-side slope of the embankment). Furthermore, 2 slopeet al., of 1996; the embankment ASTRA, 2008; will Schellenberg be rectangular et to al., the 2008). hillslope. The pthe= measured0.8v /a parameters p and a are difficult to be obtained(10) dynamicThe deceleration decelerating process force into is then soil usually has been transformed investigated into on in the field without having appropriate data on the behaviour differenta statically-equivalent scales, i.e., small force. (Heidenreich, 2004), large (Labi- Thus,of the theblock relationship at the impact between on the penetration surface of an depth embankment. and maxi- ouseMost et al. experiments, 1996; Montani-Stoffel presented in, 1998 Montani-Stoffel) and full scale (1998); (Ger- mumThe data deceleration from vertical can be falling formulated tests on as damping a function layers of the above soil berGerber, 2008 (2008);). The Pichler main results et al. are(2005) the maximumdeal with experimental deceleration layera stiff thickness layer do not(see necessarily Fig. 10). However, reflect the the load-case formulas experi- result anddata penetration gained in an of effort blocks. to quantify Both results forces are acting important on a horizon- for gal- fromenced experiments on rockfall embankments and the parameters but might measured be used after as long the im- as leriestal and (see stiff Sect. concrete5.3) to slab design being the covered strength by of various the underlying damping pactsno better of rigid results bodies are available. on cushion layers after a vertical fall. structureslayers. The and impact the thickness in these experiments of the soil layer is done (Labiouse by free fallet al. in, TheTo cushion optimize layer embankment overlies a stiff dimensions construction further and, therefore,full-scale 1996a vertical; ASTRA direction., 2008 Opposed; Schellenberg to these et experiments al., 2008). the The impact dy- cannottests on easily earth embankment be transferred structures to earth are embankments, necessary. In which Peila namicacting deceleratingon rockfall embankments force is then (being usually usually transformed constructions into a featureet al. (2002) elasto-plastic and Peila deformation et al. (2007) in the the direction performance of a free of statically-equivalentbuilt with compacted force. soils and not featuring stiff layers) will surfacereinforced (valley-side embankments slope of is the described embankment). showing Furthermore, penetration mostMost probably experiments react differently presented to in theMontani-Stoffel behaviour of the(1998 tested); thedepths measured of 0.6 parameters− 1.1 m forp embankmentsand a are difficult with to a obtain base width in the Gerberstructures.(2008 The); Pichler few projects et al. dealing(2005) with deal embankments with experimental built fieldof 5 withoutm and a having height appropriate of around data4.5 m onand the behaviour rockfall impact of the datafrom gained soil exclusively in an effort deal to quantify with real forces scale acting experiments on a horizon- (Peila blockenergies at betweenthe impact2, 400 on theand surface4, 200 ofkJ an. An embankment. overview on The the talet al., and 2002, stiff 2007)concrete or model slab covered tests (Blovsky, by various 2002) damping made from lay- datadesign from methods vertical for falling embankments tests on damping is given layers by Lambert above a andstiff ers.geogrid The reinforcedimpact in these soil embankments. experiments is Thisdone reveals by free that fall infur- a layerBourrier do not(submitted) necessarily and reflect an example the load-case of the design experienced of a rock- on verticalther tests direction. to characterize Opposed theto behaviour these experiments, of earth embankments the impact rockfallfall protection embankments, embankment but might is given be usedin Baumann as long (2008).as no better actingwith and on rockfallwithout geogridembankments reinforcements (being usually are necessary. constructions results are available. builtGerber with compacted (2008) measured soils and the not impact featuring on stiff soil layers) of varying will 5.3To Rockfall optimize protection embankment galleries dimensions, further full-scale mostthickness probably of free react falling differently blocks to of the800 behaviourand 4, of000 thekg testedwith tests on earth embankment structures are necessary. In Peila structures.falling heights The few varying projects from dealing2 ... 15 withm resulting embankments in impact built etThere al. are(2002 many) and differentPeila typeset al. of(2007 rockfall) the protection performance gallery of fromenergies soil in exclusively the range deal20 to with600 realkJ. scale Based experiments on these experi- (Peila reinforcedin regard to embankments structural design is described (Fig. 11). showing The most penetration common etments al., 2002 the following, 2007) orformulas model tests for (Blovsky the maximum, 2002) deceleration made from depthstype in Switzerlandof 0.6 − 1.1 is m a for monolithic embankments reinforced with concrete a base struc- width geogrida and penetration reinforced depth soil embankments.p due to an impact This reveals velocity thatv have fur- ofture 5 m covered and a height by a cushion of around layer 4.5 (Schellenbergm and rockfall and impact Vogel, en- therbeen tests proposed: to characterise the behaviour of earth embankments ergies2005). between 2400 and 4200 kJ. An overview on the design with and without geogrid reinforcements are necessary. methodsRockfall for galleries embankments are appropriate is given by protectiveLambert measures and Bourrier for a = 0.8v2/(gt) (9) Gerber (2008) measured the impact on soil of varying (small2011) and and well-defined an example endangered of the design zones of a withrockfall a high protection rate of thicknessp = 0.8v2 of/a free falling blocks of 800 and 4000 kg with(10) embankmentmedium magnitude is given events in Baumann (Jacquemoud,(2008). 1999). While pro- falling heights varying from 2...15 m resulting in impact en- viding protection against high energy impacts, galleries can ergiesThus the in the relationship range 20 to between 600 kJ. penetration Based on these depth experiments and maxi- 5.3provide Rockfall a low maintenance protection galleries solution for frequent low energy themum following deceleration formulas can be for formulated the maximum as a functiondeceleration of thea and soil events, for which the rocks accumulating on the gallery are penetrationlayer thickness depth (seep due Fig. to 10). an However, impact velocity the formulasv have result been Thereremoved are at many given different time intervals. types of rockfall protection galleries proposed:from experiments and the parameters measured after the im- withThe regard working to structural range of galleries design (Fig. has been11). estimatedThe most to com- be pacts of rigid bodies on cushion layers after a vertical fall. monfor impact type in energies Switzerland up to is about a monolithic 3000 kJ reinforced (ASTRA, concrete 2003). The cushion layer overlies a stiff construction and therefore structureBased on covered recent research, by a cushion which layer focuses (Schellenberg on either improv- and Vo- cannot easily be transferred to earth embankments, which geling, the2005 damping). properties of the cushion layer, increasing

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A.A.A. Volkwein Volkwein Volkwein et et et al.: al.: al.: Rockfall Rockfall Review on review review rockfall characterisation and structural protection 263519 19 A. Volkwein et al.: Rockfall review 19

Fig.Fig. 11. 11.DifferentDifferent types types of of shed shed structures structures (fltr): (fltr): reinforced-concrete reinforced-concrete slab, shell type, in situ reinforced concrete, and steel-concrete-composite Fig.typetypeFig. 11. (from (from 11.DifferentDifferent VogelVogel types et et types al., al. of, 2009).2009 of shed shed). structures structures (fltr): (fltr): reinforced-concrete slab, slab, shell shell type, type, in situ in situ reinforced reinforced concrete, concrete, and steel-concrete-composite and steel-concrete-composite typetype (from (from Vogel Vogel et etal., al., 2009). 2009).

Fig. 13. Steel-concrete composite structure for a rockfall protection Fig. 13. Steel-concrete composite structure for a rockfall protection Fig. 12. Full-scale steel-concrete composite rock shed subjected Fig.gallery 13. (Steel-concreteKonno et al., 2008 composite) structure for a rockfall protection Fig. 12. Full-scale steel-concrete composite rock shed subjected gallery (Konno et al., 2008) Fig. 12. Full-scale steel-concrete composite rock shed subjected gallery (Konno et al., 2008) toto a a falling falling weight weight (left; (left; MaegawaMaegawa et et al., al., 2003),2003), gallery gallery with with PSD PSD Fig. 13. Steel-concrete composite structure for a rockfall protection dissipationto a falling system weight in (left; Val d’Arly, Maegawa France et al., (right; 2003), taken gallery from withMasuya PSD, Fig.dissipation 12. Full-scale system in steel-concrete Val d’Arly, France composite (right; takenrock fromshed Masuya, subjected gallery (Konno et al., 2008) dissipation system in Val d’Arly, France (right; taken from Masuya, 5.3.1 Cushion layer to a2007 falling). weight (left; Maegawa et al., 2003), gallery with PSD 2007))2007)) material is often used as a cushion material, whereas in Japan . material is often used as a cushion material, whereas in Japan dissipation. system in Val d’Arly, France (right; taken from Masuya, sandThe mainis generally function used of (Ishikawa, a cushion 1999). layer is to act as a shock 2007)) sand is generally used (Ishikawa, 1999). Rockfall galleries are appropriate protective measures for absorbersmaterialThe dynamic (Jacquemoud is often force used applied, 1999 as a cushionto). the Shock top material, wavesof the cushion in whereas reinforced layer in Japan . The dynamic force applied to the top of the cushion layer thesmall structural and well-defined capacity or endangered adding energy-dissipating zones with a high supports, rate of dueconcretesand to a is structures falling generally block could used is is cause empirically empirically (Ishikawa, the separation given 1999).given by by of equation equationthe concrete 11 11 the structural capacity or adding energy-dissipating supports, cover on the soffit, so called scabbing, even for impacts with themedium galleries magnitude can provide events protection (Jacquemoud for up, 1999 to 5000). While kJ (Vogel pro- (Montani-Stoffel,The dynamic 1998). 1998). force appliedThe The impact impact to the force force top depends depends of the cushion on on the the layer vidingthe galleries protection can provide against protectionhigh energy for impacts, up to 5000 galleries kJ (Vogel can E-Moduliless intensity of the than cushion the structural layers M capacityas well (Herrmann as on the, 2002 block). theet structural al., 2009). capacity or adding energy-dissipating supports, E-Modulidue to aof falling the cushion block layers is empiricallyMEE as well given as on theby blockequation 11 provideet al., 2009). a low maintenance solution for frequent low energy radiusTher cushionand the layer rock’s also kinematic dissipates energy, some expressed of the impact in terms en- Steel-concrete-composite galleries (Fig. 12 Maegawa radius(Montani-Stoffel,r and the rock’s 1998). kinematic The energy, impact expressed force depends in terms on the theevents, galleriesSteel-concrete-composite for can which provide the rocks protection accumulating galleries for up (Fig. on to 5000the 12 gallery Maegawa kJ (Vogel are ofergy, mass distributesm and impact the contact velocity stresses,v. decreases the peak load- et al., 2003) or composite sandwich structures with high- ofE-Moduli mass m and of impact the cushion velocity layersv. ME as well as on the block et al.,removedet 2009).al., 2003) at given or composite time intervals. sandwich structures with high- ing on the impacted structure and also increases the duration tensile bolt connections (Fig. 13 Konno et al., 2008) have radius r and the rock’s kinematic energy, expressed in terms Steel-concrete-compositetensile bolt connections (Fig. galleries 13 Konno (Fig. et al., 12 2008) Maegawa have of impact. For economic reasons, locally available0.6 granular The working range of galleries has been estimated to be 22 0.6 been evaluated in Japan and could provide future solutions of mass m and impact0.2 velocity0.4 mmv.××vv etfor al.,been impact 2003) evaluated energiesor composite in Japan up to and sandwich about could 3000 provide structures kJ (ASTRA future with solutions, 2003 high-). Pmaterial= is1.765 often× usedr0.2 as× aM cushion0.4 × material, whereas in Japan(11) for specific applications. Pmax = 1.765 × r × MEE × 2 (11) tensileBasedfor specific bolt on connectionsrecent applications. research (Fig. which 13 focuses Konno onet al., either 2008) improv- have sand is generally used (Ishikawa, 19992). The following section gives a summary of research related 0.6 beening evaluatedThe the following damping in Japan propertiessection and gives of could a the summary cushion provide of layer, researchfuture increasing solutions related The dynamic force applied to the topm of× thev cushion2 layer toto protection protection galleries galleries with with emphasis emphasis on on the the cushion cushion layer layer PFor structural= 1.765 design× r0. purposes, purposes,2 × M 0.4 however, however,× the the forces forces trans- trans- (11) forthe specific structural applications. capacity or adding energy-dissipating supports, duemax to a falling block is empiricallyE given2 by Eq. (11) andand the the structural structural evaluation evaluation of of the the galleries. galleries. mitted across the interface between between the the cushion cushion layer layer and and Thethe galleries following can section provide gives protection a summary for up of to research 5000 kJ (Vogel related (Montani-Stoffel, 1998). The impact force depends on the et al., 2009). structureE-Moduli are of required.the cushion Of Of layers interest interestM are areas the the well definition definition as on the of of block load load to protection galleries with emphasis on the cushion layer For structural design purposes,E however, the forces trans- 5.3.15.3.1Steel-concrete-composite Cushion Cushion layer layer galleries (Fig. 12 Maegawa magnituderadius r and and the loading rock’s area.kinematic area. Both, Both, energy, of of course, course, expressed vary vary with with in terms time time and the structural evaluation of the galleries. mitted across the interface between the cushion layer and et al., 2003) or composite sandwich structures with high- duringof mass them impactand impact process velocity and and depend dependv. on on the the material material prop- prop- ThetensileThe main main bolt function function connections of of a a cushion (Fig. cushion13 layer layerKonno is is etto to al. act act, 2008 as as a a) shock shock have ertiesstructure of the cushionare required. layer. layer. Of interest are the definition of load 0.6 5.3.1absorberabsorber Cushion (Jacquemoud, (Jacquemoud, layer 1999). 1999). Shock Shock waves waves in in reinforced reinforced magnitudeIn experimental and loading research research area. (Kishi (Kishi Both, et et2 al.,! al., of 1993), 1993),course, the the vary trans- trans- with time been evaluated in Japan and could provide future solutions 0.2 0.4 m×v concreteconcrete structures structures could could cause cause the the separation separation of of the the concrete concrete mittedPmax = force1.765 was×r found×M to to× be be about about 1.8 1.8 times times the the impact impact(11) for specific applications. during the impact processE and2 depend on the material prop- Thecovercover mainThe on onfollowing functionthe the soffit, soffit, section ofso so called a called cushion gives scabbing, scabbing, a summary layer even even is of to for for research act impacts impacts as arelated withshock with forceerties in the of the case cushion of a a sand sand layer. cushion cushion layer layer or or only only half half the the im- im- lessless intensity intensity than than the the structural structural capacity capacity (Herrmann, (Herrmann, 2002). 2002). absorberto protection (Jacquemoud, galleries 1999).with emphasis Shock on waves the cushion in reinforced layer pactFor forceIn structural experimental for a special design three three research purposes, layer layer (Kishi cushion cushion however, et system system theal., forces 1993), (Ishikawa, (Ishikawa, trans- the trans- concreteandTheThe the cushionstructures cushionstructural layer layer could evaluation also also cause dissipates dissipates of the the separation galleries. some some of of the the of theimpact impact concrete en- en- 1999).mittedmitted across The force transmitted the was interface found force, force, between to which which be about the is is the the cushion 1.8 load load times acting layeracting the and on on impact ergy, distributes the contact stresses, decreases the peak load- the structures, can also be determined numerically. A sim- coverergy, on distributes the soffit, the so contact called stresses, scabbing, decreases even for the impacts peak loadwith thestructureforce structures, in are the required. cancase also of Of a be sand interest determined cushion are the numerically. layer definitions or only ofA half sim-load the im- inging on on the the impacted impacted structure structure and and also also increases increases the the duration duration plified method using an an ordinary ordinary FE FE code, code, assuming assuming one- one- less intensity than the structural capacity (Herrmann, 2002). magnitudepact force and for loading a special area. three Both, layer of course, cushion vary system with (Ishikawa,time ofof impact. impact. For For economic economic reasons, reasons, locally locally available available granular granular dimensional stress wave propagation propagation and and elastic-plastic elastic-plastic soil soil The cushion layer also dissipates some of the impact en- 1999). The transmitted force, which is the load acting on ergy,www.nat-hazards-earth-syst-sci.net/11/2617/2011/ distributes the contact stresses, decreases the peak load- the structures, Nat. Hazards can Earth also beSyst. determined Sci., 11, 2617– numerically.2651, 2011 A sim- ing on the impacted structure and also increases the duration plified method using an ordinary FE code, assuming one- of impact. For economic reasons, locally available granular dimensional stress wave propagation and elastic-plastic soil 20 A. Volkwein et al.: Rockfall review

5.3.2 Structural Evaluation 2636 A. Volkwein et al.: Review on rockfall characterisation and structural protection during the impact process and depend on the material properties of the cushion layer. In experimental research (Kishi et al., 1993), the transmitted force was found to be about 1.8 times the impact To date guidelines for the design of rockfall galleries have force in the case of a sand cushion layer or only half the impact force for a special three layer cushion system (Ishikawa, been published in Switzerland and in Japan (ASTRA, 2008; 1999). The transmitted force, which is the load acting on the structures, can also be determined numerically. A sim- Japan Road Association, 2000). In both cases, a static- plified20 method using an ordinary FE code, assuming one- A. Volkwein et al.: Rockfall review dimensional stress wave propagation and elastic-plastic soil equivalent force is applied, which apart from the rock mass properties was used to estimate the stress distributions for relatively small impact loads (Sonoda, 1999). and velocity depends mostly on the geotechnical conditions Today, advanced FE models (e.g., LS-DYNA code) are 5.3.2 Structural Evaluation used to model entire galleries including the cushion layer and are able to match results from large scale tests (Kishi of the cushion layer. This approach is simple to use by prac- et al., 2009). In the latest simulations for the cushion layer, a cap-hardening model is used, in which parameters are de- ticing engineers, but presents difficulties in accounting for termined by curve fitting using experimental data (Ghadimi- Fig. 14. Alternative cushion layers: (left) Fence box structure Khasraghy et al., 2009). To date guidelines for the designthe of complex rockfall galleries dynamic have processes during the impact. A sum- Numerical simulations, by means of the DE method, have with cellular glass material (Schellenberg, 2008), (right) Multilayer been applied for rockfall impact on embankments (Plassiard sandwich structure (Lorentz et al., 2008).been published in Switzerland andmary in Japan of older (ASTRA, formulations 2008; for the impact force is given in and Donze´, 2009) and could potentially lead to future improvements in the design of rockfall protection galleries. It Japan Road Association, 2000).Montani-Stoffel In both cases, (1998) a static- and a comparison of the different cal- has also been proposed to simulate the processes taking place within the cushion layer by a rheological model (Calvetti and equivalent force is applied, whichculation apart from methods the rock can mass be found in Casanovas (2006). Di Prisco, 2009) or by a simplified nonlinear spring describing the overall relationship between force and rock penetra- and velocity depends mostly on the geotechnical conditions tion into the cushion layer (Schellenberg, 2009). properties was used to estimate the stress distributions for The selection of the cushion material can significantly im- Fig. 14. Alternative cushion layers: (top) Fence box structure with Based on a system of multiple degrees of freedom for cellular glass material (Schellenberg, 2008), (bottom) Multi-layer of the cushion layer. This approach is simple to use by prac- prove the capacity of the gallery. The energy dissipation relatively small impact loads (Sonoda, 1999). sandwich structure (Lorentz et al., 2008). impact loads (Comite-Euro-International´ du Beton,´ 1988), a for different materials and mixtures has been studied in cen- ticing engineers, but presents difficulties in accounting for trifugeFig. tests, with 14. the resultAlternative that a mixture of cushion sand-rubber layers:Today, (left) advanced Fence box FE structure models (e.g. LS-DYNA code) are (70 %–30 %) with clay lumps seems to be an efficient cush- 5.3.2 Structural evaluation the complex dynamic processesnew during analytical the impact. model A sum- has been proposed for the design of ion materialwith (Chikatamarla cellular, 2006 glass). material (Schellenberg,used to model 2008), (right) entire Multilayer galleries including the cushion layer Full scale tests in Japan showed that the impact forces can To date guidelines for the design of rockfall galleries have mary of older formulations for therockfall impact galleries, force is given which in allows predicting both shear and also besandwich substantially reduced structure by the above-mentioned (Lorentz three- et al.,andbeen 2008). published are in able Switzerland to and match in Japan (ASTRA results, 2008; from large scale tests (Kishi layered absorbing system (TLAS), which is composed of an Japan Road Association, 2000). In both cases, a static- Montani-Stoffel (1998) and a comparisonbending of failure the different (Schellenberg cal- et al., 2008, Fig. 15). EPS (expanded polystyrol) layer, a reinforced concrete core etequivalent al., force 2009). is applied, In which the apart from latest the rock simulations mass for the cushion layer, slab and a sand layer (Nakano et al., 1995). A large-scale and velocity depends mostly on the geotechnical conditions culation methods can be found in Casanovas (2006). test in Switzerland with foam glass as cushion layer mate- aof cap-hardening the cushion layer. This approach model is simple to use is by used, prac- in which parameters are de- The time histories of the spring forces are derived from rial alsoproperties showed promising was results (Schellenberg used to et al., estimate2007, ticing engineers, the stress but presents distributions difficulties in accounting for for Fig. 14top). Lorentz et al. (2008) investigated the perfor- terminedthe complex dynamic by processes curve during fitting the impact. using A sum- experimentalBased on data a system (Ghadimi- of multiplethe degrees equations of freedomof motion for with the given masses and spring mancerelatively of sandwich structures small composed impact of two or threeloads re- (Sonoda,mary of older formulations 1999). for the impact force is given in inforced concrete layers separated by tyres (Fig. 14bottom). KhasraghyMontani-Stoffel (1998) et and al.,a comparison 2009). of the different cal- impact loads (Comite-Euro-International´ properties du described Beton,´ 1988), above. a The peak loads are performance- A differentToday, approach to advanced dissipate energy without FE a cushion modelsculation (e.g. methods can LS-DYNA be found in Casanovas code)(2006). are layer is the PSD system (Pare-blocs Structurellement Dissi- BasedNumerical on a system of multiple simulations degrees of freedom by for meansnew of analytical the DE method model have has beenbased proposed results for the and design can be of compared with the resistance in the pantes)used proposed to in France model and shown entire in Fig. 12 right. galleries The impact including loads (Comite-Euro-International´ the cushion du Beton´ , layer1988), a slab is subjected to direct impact and energy absorbing de- new analytical model has been proposed for the design of rockfall galleries, which allowscritical predicting sections both of shear the slab. and vicesand are placed are at the able slab supports to match(Tonello, 2001 results). Test beenrockfall from galleries, applied large which scale allows for predicting rockfall tests both (Kishi shear impact and on embankments (Plassiard results on a scale of 1/3 are presented in (Berthet-Rambaud, bending failure (Schellenberg et al., 2008, Fig. 15). bending failure (Schellenberg et al., 2008, Fig. 15). 2004). andThe time Donz historiese,´ of 2009)the spring forces and are derived could from potentially lead to future im- et al., 2009). In the latest simulationsthe equations of motion for the with the cushion given masses and layer, spring With this model relative values between the maximum provementsproperties described above. in The the peak loads design are performance- of rockfallThe protection time histories galleries. of the It spring forces are derived from a cap-hardening model is used,based in results which and can be parameters compared with the resistance are inde- the forces and the load bearing capacities for punching (η2) and termined by curve fitting usinghascritical experimental also sectionsbeen of the slab. proposed data (Ghadimi- to simulatethe the equationsprocesses of taking motion place with the given masses and spring bending failure (η3) are obtained, leading to an iterative pro- Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011within www.nat-hazards-earth-syst-sci.net/11/2617/2011/ the cushion layer by a rheological model (Calvetti and Khasraghy et al., 2009). properties described above. The peakcess loads for the are structural performance- design. Numerical simulations byDi means Prisco, of the 2009) DE or method by a simplified have based non-linear results spring and can describ- be compared with the resistance in the been applied for rockfall impacting on the embankments overall relationship (Plassiard betweencritical force sections and rock of penetra-the slab. This procedure is particularly suitable for the evaluation of and Donze,´ 2009) and couldtion potentially into the lead cushion to future layer im- (Schellenberg, 2009). existing galleries. Fig. 16 shows the ratio values reached for With this model relative values between the maximum provements in the design of rockfall protection galleries. It The selection of the cushion materialforces can and significantly the load bearing im- capacitiesrocks for with punching different (η masses) and falling from different heights for has also been proposed to simulate the processes taking place 2 prove the capacity of the gallery.bending The energy failure ( dissipationη ) are obtained,the leading gallery to an Axen-S iterativeud¨ pro- in Switzerland. Future evaluations of within the cushion layer by a rheological model (Calvetti and 3 for different materials and mixturescess has for been the studied structural in design. cen- the force penetration relationship of the rock into the cushion Di Prisco, 2009) or by a simplifiedtrifuge non-linear tests, with spring the describ- result that a mixture of sand-rubber layer would improve this model. ing the overall relationship between(70%-30%) force with and rock clay penetra- lumps seems toThis be an procedure efficient is cushion particularly suitable for the evaluation of In recent years, significant advances have been made re- tion into the cushion layer (Schellenberg,material (Chikatamarla, 2009). 2006). existing galleries. Fig. 16 shows the ratio values reached for garding numerical simulations to aid structural design (Kishi The selection of the cushion materialFull scale can tests significantly in Japan im- showedrocks that the with impact different forces masses can falling from different heights for prove the capacity of the gallery. The energy dissipation the gallery Axen-Sud¨ in Switzerland.et al., Future 2009; Masuya evaluations and of Nakata, 2001). The simulations al- also be substantially reduced by the above mentioned three for different materials and mixtures has been studied in cen- the force penetration relationshiplow of the a detailed rock into evaluation the cushion of the structure and its response to layered absorbing system (TLAS), which is composed of an trifuge tests, with the result that a mixture of sand-rubber layer would improve this model.rockfall impact (Fig. 17). This approach, however, requires (70%-30%) with clay lumps seemsEPS (expanded to be an efficient polystyrol) cushion layer, a reinforced concrete core In recent years, significant advancesexperimental have been data made for calibration re- and significant resources, material (Chikatamarla, 2006).slab and a sand layer (Nakano et al., 1995). A large-scale garding numerical simulations tolimiting aid structural its application design (Kishi in practice. Such efforts, though, are Full scale tests in Japan showedtest in that Switzerland the impact withforces foam can glass as cushion layer mate- et al., 2009; Masuya and Nakata,useful 2001). for The the simulations development al- of design guidelines and for eval- also be substantially reducedrial by also the above showed mentioned promising three results (Schellenberg et al., 2007, low a detailed evaluation of the structureuating critical and its sections response and to parametric influences. layered absorbing system (TLAS),Fig. 14left). which is Lorentz composed et of al. an (2008) investigated the perfor- rockfall impact (Fig. 17). This approach, however, requires EPS (expanded polystyrol) layer,mance a reinforced of sandwich concrete structures core composed of two or three re- Despite advances in understanding the structural perfor- experimental data for calibration and significant resources, slab and a sand layer (Nakanoinforced et al., concrete 1995). A layers large-scale separated by tyres (Fig. 14right). mance of rockfall galleries, there are still large uncertain- limiting its application in practice. Such efforts, though, are test in Switzerland with foam glassA different as cushion approach layer to mate- dissipate energy without a cushion ties regarding the definition of design situations. Therefore, useful for the development of design guidelines and for eval- rial also showed promising resultslayer (Schellenberg is the PSD system et al., 2007, (Pare-blocs Structurellement Dissi- probabilistic methods are attractive tools because the uncer- uating critical sections and parametric influences. Fig. 14left). Lorentz et al. (2008)pantes) investigated proposed in the France perfor- and shown in Fig. 12right. The tainties can be better quantified. In addition, future develop- mance of sandwich structuresslab composed is subjected of two to or direct three impact re- andDespite energy advances absorbing in understandingde- ments the in the structural design perfor- of new protection galleries or the eval- inforced concrete layers separatedvices by are tyres placed (Fig. at 14right). the slab supportsmance (Tonello, of rockfall 2001). galleries, Test thereuation are still of existing large uncertain- sheds might involve evaluating the failure A different approach to dissipateresults energy on a scale without of 1/3a cushion are presentedties regarding in (Berthet-Rambaud, the definition of designprobability situations. for different Therefore, design situations and select the de- layer is the PSD system (Pare-blocs2004). Structurellement Dissi- probabilistic methods are attractivesign tools situations because based the uncer- on overall risk acceptance criteria. pantes) proposed in France and shown in Fig. 12right. The tainties can be better quantified. In addition, future develop- slab is subjected to direct impact and energy absorbing de- ments in the design of new protection galleries or the eval- vices are placed at the slab supports (Tonello, 2001). Test uation of existing sheds might involve evaluating the failure results on a scale of 1/3 are presented in (Berthet-Rambaud, probability for different design situations and select the de- 2004). sign situations based on overall risk acceptance criteria. A. Volkwein et al.: Rockfall review 21 A. Volkwein et al.: Rockfall review 21 A. Volkwein et al.: Review on rockfall characterisation and structural protection 2637

a) b) c) d) e) a) b) c) d) e)

Fig. 15. System with multiple degrees of freedom (SMDF) a) and b), from the section of a gallery to the model definition together with the Fig.Fig. 15. 15.SystemSystem with with multiple multiple degrees degrees of freedom (SMDF) (SMDF)(a) a) and(b) b),, from from the the section of a a gallery gallery to to the the model model definition definition together together with with the the force-displacement relationship of the springs for c) cushion layer, d) shear behaviour and e) global bending stiffness (from Schellenberg and force-displacementforce-displacement relationship relationship of of the the springs springs for for c)(c) cushioncushion layer, layer, d)(d) shearshear behaviour behaviour and and e)(e) globalglobal bending bending stiffness stiffness (from (from SchellenbergSchellenberg and Vogel, 2009). Vogel,and Vogel 2009)., 2009).

5.4 Flexible protection systems the latter being concatenated like a historical byrnie and orig- the latter being concatenated like a historical byrnie and originate from the torpedo protection nets used in front of har- inate from the torpedo protection nets used in front of har- Today,bours one and of ships the most in the common 2nd World protection War. measures Only few against knowl- bours and ships in the 2nd World War. Only few knowl- rockfalledge exists is the on use the of flexible use of protectionalternative systems. net materials Such barri- (Tajima edge exists on the use of alternative net materials (Tajima erset areal., usually 2003). Theinstalled support like ropes fences(ropesectiondiameter along the boundary of an12 − et al., 2003). The support ropes (ropesectiondiameter12 − infrastructure22 mm) are orspanned in front between of buildings steel acting posts with as a passive typical pro-lengths 22 mm) are spanned between steel posts with typical lengths tectionbetween system, 2 and i.e., 7 mthey and are field meant spacings to stop avarying moving between block. 5 Muchbetween research 2 and has 7 m already and field been spacings performed varying on such between barri- 5 andand 12 12 m. m. The The posts posts are are fixated fixated by by ground ground plates plates either either by by ersclamped in recent support years. or At hinged first, the support research with work additional concentrated upslope onclamped the general support ability or hinged of flexible support systems with to additional reliably retainupslope Fig. 16. Loading capacity of protection gallery Axen-Sud¨ for dif- ropes at the post head. Details regarding the state-of-the-art Fig. 16. Loading capacity of protection gallery Axen-Sud¨ for dif- fallingropes at rocks the post (Sect. head.5.4.1 Details). Later, regarding the emphasis the state-of-the-art was on how Fig.ferent 16. impactLoading masses capacity (from ofSchellenberg protection gallery, 2009). Axen-Sud¨ for dif- of post foundations including suggestions for load measure- ferent impact masses (from Schellenberg, 2009). toof improve post foundations our knowledge including of such suggestions barriers, e.g., for loadby means measure- of ferent impact masses (from Schellenberg, 2009). ments can be found in Turner et al. (2009). Additional ropes systematicments can and be found extensive in Turner testing et (Grassl al. (2009)., 2002 Additional), overall eval- ropes may be placed depending on the individual systems. Connec- With this model relative values between the maximum uationsmay be ( placedSpang anddepending Bolliger on, 2001 the individual) or numerical systems. simulations Connec- tions to the ground are usually achieved by drilled anchors. 5.4forces Flexible and the protection load bearing systems capacities for punching (η2) and (seetions Sect. to the5.4.5 ground). The are knowledge usually achieved gained thereby by drilled formed anchors. the 5.4 Flexible protection systems For higher impact energies most systems have additional en- bending failure (η3) are obtained, leading to an iterative pro- basisFor higher for standardization impact energies as described most systems in Sect. have5.4.2 additional. Because en- cess for the structural design. ergyergy absorbing absorbing elements elements attached attached to to the the ropes. ropes. Such Such elements elements Today,Today, one one of of the the most most common common protection protection measures measures against against the research is usually rather application-oriented and carried This procedure is particularly suitable for the evaluation of outdeformdeform in close plastically plastically cooperation with with with large large the displacements displacements manufacturers, (up (up typically to to22mm the)) inin- rockfallrockfall is is the the use use of of flexible flexible protection protection systems. systems. Such Such barri- barri- existing galleries. Figure 16 shows the ratio values reached publishedcreasingcreasing the results the flexibility flexibility consider of of just the the one supporting supporting barrier type. structure. structure. However, Fig. Fig. it 18 18 ersers are are usually usually installed installed like like fences fences along along the the boundary boundary of of an an for rocks with different masses falling from different heights stillshowsshows would some some be typical typicalpossible braking braking to compare elements. elements. the different The The barriers barriers systems are are re- usu- usu- infrastructureinfrastructure or or in in front front of of buildings buildings acting acting as as a a passive passive pro- pro- for the gallery Axen-Sud¨ in Switzerland. Future evaluations gardingallyally erected erected their by performance, by local local mounting mounting braking teams teams distance, according according energy to to balance, the the manu- manu- tectiontection system, system, i.e. i.e. they they are are meant meant to to stop stop a a moving moving block. block. of the force penetration relationship of the rock into the cush- etc.,facturer’sfacturer’s as done installation installation by Gerber manual andmanual Volkwein that that comes comes(2007 with with). the the barrier. barrier. MuchMuch research research has has already already been been performed performed on on such such barri- barri- ion layer would improve this model. Today,There after are severalvarious decades advantages of development favouring flexibleand improve- nets for ersers in in recent recent years. years. At At first, first, the the research research work work concentrated concentrated There are various advantages favouring flexible nets for In recent years, significant advances have been made re- ment,an increasingly a typical flexible wide rockfall distribution. protection They system are cheaper consists ofcom- onon the the general general ability ability of of flexible flexible systems systems to to reliably reliably retain retain an increasingly wide distribution. They are cheaper com- garding numerical simulations to aid structural design (Kishi apared steel net with attached other protection longitudinally systems, to so-called e.g. about support one ropes. tenth of fallingfalling rocks rocks (section (section 5.4.1). 5.4.1). Later, Later, the the emphasis emphasis was was on on how how pared with other protection systems, e.g. about one tenth of et al., 2009; Masuya and Nakata, 2001). The simulations al- Thea gallery nets with structure. mesh openings They are ranging quickly from installed 5–35 cm requiring are made little toto improve improve our our knowledge knowledge of of such such barriers, barriers, e.g. e.g. by by means means of of a gallery structure. They are quickly installed requiring little low a detailed evaluation of the structure and its response to fromequipment. chain-link Their meshes, performance wire-rope is nets effective, or steel efficient rings, the and lat- reli- systematicsystematic and and extensive extensive testing testing (Grassl, (Grassl, 2002), 2002), overall overall eval- eval- equipment. Their performance is effective, efficient and reli- rockfall impact (Fig. 17). This approach, however, requires terable. being The concatenated impact on the like landscape a historical during byrnie construction and originate is low uationsuations (Spang (Spang and and Bolliger, Bolliger, 2001) 2001) or or numerical numerical simulations simulations able. The impact on the landscape during construction is low (seeexperimental section 5.4.5). data for The calibration knowledge and gained significant thereby resources, formed fromand the a certain torpedo transparency protection nets afterwards used in front is guaranteed. of harbours Dueand to (seelimiting section its 5.4.5). application The in knowledge practice. Such gained efforts, thereby though, formed are and a certain transparency afterwards is guaranteed. Due to the basis for standardization as described in section 5.4.2. shipstheir in wide the2nd range World of energy War. Only retention limited capacity, knowledge flexible exists fence theuseful basis for for the standardization development of as design described guidelines in section and for 5.4.2. eval- ontheir the wide use of range alternative of energy net retentionmaterials capacity,(Tajima et flexible al., 2003 fence). Because the research is usually rather application-oriented systems can be used for most applications. And, finally, an Becauseuating critical the research sections is and usually parametric rather influences. application-oriented Thesystems support can ropes be used (rope for section most applications. diameter 12 − And,22 mm) finally, are an and carried out in close cooperation with the manufacturers, increasing number of manufacturers results in healthy com- and carriedDespite out advances in close in cooperation understanding with the the structural manufacturers, perfor- spannedincreasing between number steel of manufacturersposts with typical results lengths in healthy between com- 2 typically the published results consider just one barrier type. petition, guaranteeing continuous development and improve- typicallymanceof the rockfall published galleries, resultsthere consider are just still one large barrier uncertain- type. andpetition, 7 m and guaranteeing field spacings continuous varying betweendevelopment 5 and and 12 m. improve- The However, it still would be possible to compare the different ments with a parallel reduction in prices. However,ties regarding it still the would definition be possible of design to compare situations. the Therefore, different postsments are with fixated a parallel by ground reduction plates in either prices. by clamped support systemssystemsprobabilistic regarding regarding methods their their are performance, performance, attractive tools braking braking because distance, distance, the uncer- en- en- or hingedHowever,However, support there there with are are additionalsome some limiting limiting upslope factors factors ropes in in at the the the case casepost of of ergyergytainties balance balance canetc. be etc. better as as done done quantified. by by Gerber Gerber In addition,and and Volkwein Volkwein future (2007). (2007). develop- head.flexibleflexible Details barriers. barriers. regarding Long-term Long-term the state-of-the-art protection protection post against against foundations corrosion corrosion mentsToday,Today, in after after the designseveral several of decades decades new protection of of development development galleries and and or improve- the improve- eval- includingmustmust be be guaranteed; suggestionsguaranteed;working for working load measurements life life is is defined defined can in in EOTA be EOTA found (2008) (2008) in ment,ment,uation a a typical typicalof existing flexible flexible sheds rockfall rockfall might protection involve protection evaluating system system consists the consists failure of of Turnerwithwith 25 25 et years al. years(2009 (or (or). eveneven Additional shorter shorter ropes if if installed installed may be in placed in aggressive aggressive depend- en- en- aa steelprobability steel net net attached attached for different longitudinally longitudinally design situations to to so-called so-called and support supportselect the ropes. ropes. de- ingvironmentalvironmental on the individual conditions). conditions). systems. If If a aConnections barrier barrier has has to experienced experienced the ground at at TheThesign nets nets situations with with mesh mesh based openings openings on overall ranging ranging risk acceptance from from 5 5 -criteria. - 35 35 cm cm are are areleastleast usually one one medium-sized medium-sized achieved by drilled rockfall rockfall anchors. event, event, it it For is is usually usually higher deformed impact deformed mademade from from chain-link chain-link meshes, meshes, wire-rope wire-rope nets nets or or steel steel rings, rings, resultingresulting in in a a reduced reduced barrier barrier height height after after a a successfully successfully www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 22 2638 A. Volkwein et al.: Review on rockfall characterisationA.A. Volkwein Volkwein and structural et et al.: al.: Rockfall Rockfall protection review review

Fig. 17. General view of an FE analysis model of an impacted rock shed and the resulting crack patterns for different loading cases (from Fig. 17. GeneralGeneral view view of of an an FE FE analysis analysis model model of of an an impacted impacted rock rock shed shed and and the the resulting resulting crack crack patterns patterns for for different different loading loading cases cases (from (from Kishi et al., 2009). Kishi et et al., al., 2009) 2009)

Fig. 18. Different types of energy absorbing barrier components (friction of tensioned rope between friction plates, friction between rope Fig. 18. Fig.clamps, 18. DifferentDifferent bent steel types types pipe of ofcircle energy energy narrowing absorbing absorbing under barrier barrier tension components components and elongating (friction (friction spiral structures) of of tensioned tensioned and roperope mesh between between types (original friction friction anti-submarine plates, plates, friction friction net, between between hexagon rope rope clamps,mesh bent and spliced steel pipe rope circle net, ring narrowing net, rope under net with tension, clamps). and elongating spiral structures) and mesh types (originally anti-submarine net, clamps, bent steel pipe circle narrowing under tension, and elongating spiral structures) and mesh types (originally anti-submarine net, hexagon mesh and spliced rope net, ring net, rope net with clamps). hexagon mesh and spliced rope net, ring net, rope net with clamps). energies most systems have additional energy absorbing el- a gallery structure. They are quickly installed requiring lit- ements attached to the ropes. Such elements deform plas- tle equipment. Their performance is effective, efficient and resistedresistedtically rockfall rockfall with large event. event. displacements Further, Further, after(up after to large-sized large-sized 2 m) increasing rockfall rockfall the reliable.temtem has has not Thenot been been impact capable capable on the of of landscape withstanding withstanding during the the construction dynamic dynamic snow snow events,flexibility the the remaining remaining of the supporting retention retention structure. capacity capacity Figure might might18 shows be be reduced reduced some isloadload low (Margreth, (Margreth, and a certain 1995; 1995; transparency Nicot Nicot et et al., al., afterwards 2002b,a). 2002b,a). is In guaranteed. In such such a a case, case, requiringrequiringtypical immediate immediate braking elements. maintenance. maintenance. The barriers Therefore, Therefore, are usually regular regular erected inspec- inspec- by Duethethe alternatives alternatives to their wide would would range be be of a a energy partial partialretention removal removal and capacity, and re-installation re-installation flexi- tionlocal is necessary mounting for teams all according installed barriers to the manufacturer’s to prevent reduced instal- bleevery fence year systems or an alternative can be used protection for most applications. measure such And, as gal- tion is necessary for all installed barriers to prevent reduced every year or an alternative protection measure such as gal- performancelation manual as a that result comes of, e.g., with barriers the barrier. being partially filled finally,leries. an increasing number of manufacturers results in performance as a result of, e.g., barriers being partially filled leries. by smallThere rocks, are variouswood etc. advantages Flexible favouringbarriers cannot flexible be nets used for if healthy competition, guaranteeing continuous development by small rocks, wood etc. Flexible barriers cannot be used if thean expected increasingly impact wide energies distribution. are too high They or are if the cheaper calculated com- andIn improvements the recent years with new a parallel rockfall reduction mitigation in prices. measures have the expected impact energies are too high or if the calculated In the recent years new rockfall mitigation measures have pared with other protection systems, e.g., about one tenth of gained increasing attention. So-called attenuating systems block trajectories would overtop the barriers reaching the ob- gained increasing attention. So-called attenuating systems block trajectories would overtop the barriers reaching the ob- do not try to stop a falling rock but to catch it and to guide it ject to be protected. If the place of installation is also subject do not try to stop a falling rock but to catch it and to guide it ject to be protected. If the place of installation is also subject downhill in a controlled manner (see Fig. 19). Such barriers to avalanchesNat. Hazards in Earth winter, Syst. up tillSci., now 11, a2617– rockfall2651 protection, 2011 sys- downhill www.nat-hazards-earth-syst-sci.net/11/2617/2011/ in a controlled manner (see Fig. 19). Such barriers to avalanches in winter, up till now a rockfall protection sys- are also called Hybrid Barriers or Hanger Nets (Glover et al., are also called Hybrid Barriers or Hanger Nets (Glover et al., A. Volkwein et al.: Review on rockfall characterisation and structural protection 2639 A. Volkwein et al.: Rockfall review 23 to around 5000 kJ. However, it must be stated that research relatedThe first to flexible guideline fence world-wide systems was generally initiated involves in Switzerland coopera- intion 2000 between (Gerber, a research 2001a). institute This guideline and a particular defines the fence testing man- procedures that allow an a posteriori evaluation of the barri- ufacturer focusing only on its own products (Grassl, 2002; ers with respect to the maximum energy retention capacity, Volkwein, 2004; Nicot, 1999; Wienberg et al., 2008; Peila the actual rope forces, the braking distance, the remaining et al., 1998). There are only few studies which compare barrier height, the performance for small and medium-sized different net systems. For instance, Gerber and Volkwein rockfall events and the corresponding maintenance work. (2007) analysed the performance of different systems for ei- therIn soft 2008 or the hard European dynamic Guideline decelerating ETAG processes. 027 was published The grow- (EOTA,ing understanding 2008; Peila of and fence Ronco, systems 2009.). and By their letter dynamic of the Eu- be- ropean Commission to the Member States, the 1st of Febru- Fig. 19. Principle mode of operation for rockfall attenuating system haviour also allows the use of various net-type systems to Fig. 19. Principle mode of operation for rockfall attenuating system ary 2008 has been considered as the date of its availabil- (left,(left,Glover Glover et et al. al.,, 2010 2010)) and and system system sketch sketch for for typical typical hanger net resist impact forces caused by other natural hazards such ity and applicability. ETAG 027 defines a testing procedure systemsystem (right). (right). as avalanches (Margreth, 1995), falling sliding trees (Volk- similarwein et to al. the, 2009 Swiss; guidelineHamberger and and - after Stelzer successful, 2007), system debris testingflows (Wendeler and identification, 2008) or testing shallow of landslides the main components (Bugnion et as al., well as after initial factory production inspection by the in- 2010;However, Dhakal there et al., are 2011a). some limiting factors in the case of 2008). flexible barriers. Long-term protection against corrosion volved approval body - allows the producers to attach the CE marking for the barrier on basis of relevant EC certifi- must5.4.1 be Historicalguaranteed; development working life and is currentdefined research in EOTA (2008) 5.4.2 Standardization cate of a notified certification body and EC declaration of with 25 yr (or even shorter if installed in aggressive environ- conformity by the manufacturer. The basis for issuing the mentalMostly, conditions). the old-type If fences a barrier were has able experienced to withstand at just least small one It is important for the planning and design of effective protec- EC certificate is the European technical approval as the con- medium-sizedrockfall events. rockfall Only event, in theit early is usually 1990s withdeformed research result- on tion systems that their behaviour is well understood and thor- cerned harmonized technical specification, issued by an ap- inghow in to a reduced stop falling barrier rocks height efficiently after a was successfully the dynamics resisted of oughly verified. This also ensures an efficient use of public proval body entitled for these tasks and the implementation rockfallthe decelerating event. Further, process after considered large-sized and rockfallused to design events, new the investment. Due to the complex, dynamic and difficult to de- of a factory production control system on basis of the con- remainingretention systemsretention (Hearn capacity et might al., 1992). be reduced This included requiring also im- scribe decelerating process a typical barrier design is based trol plans, accompanying the European technical approval. mediatethe development maintenance. of fences Therefore, with regular retention inspection capacities is neces- of up on prototype testing. This procedure has also been adapted It is typical for such a broad guideline that many different 50kJ based on dynamic design approaches (Duffy, 1992; to produce standardization guidelines defining the minimum sary for all installed barriers to prevent reduced performance interests have to be combined and formulated. This usually Duffy and Haller, 1993). Since then continuous research performance limits of solid barriers. as a result of, e.g., barriers being partially filled by small becomes a quasi-minimum standard requiring National Ap- and engineering development has increased their retention The first guideline world-wide was initiated in Switzerland rocks, wood, etc. Flexible barriers cannot be used if the ex- plication Documents for the single member states. pectedcapacities impact to energiesaround 5000 are tookJ. high However, or if the it calculated must be stated block in 2000 (Gerber, 2001a). This guideline defines the testing trajectoriesthat research would related overtop to flexible the barriers fence reaching systems generally the object in- to proceduresIt must also that be allow borne a posteriori in mind that evaluation there will of alwaysthe barriers be bevolves protected. cooperation If the between place of a researchinstallation institute is also and subject a partic- to loadwith cases respect outside to the the maximum scope of energy the guidelines, retention suchcapacity, as ec- the avalanchesular fence inmanufacturer winter, up focusingtill now a only rockfall on its protection own products sys- centricactual rope impact forces, forces, the post braking or rope distance, strikes, the high remaining or low speed barrier tem(Grassl, has not 2002; been Volkwein, capable of2004; withstanding Nicot, 1999; the Wienberg dynamic et snow al., rockfallheight, the events performance with the same for small impact and energy, medium-sized etc (Wienberg rockfall load2008; (Margreth Peila et, al.,1995 1998).; Nicot There et al. are, 2002b only,a few). In studies such a which case, etevents al., 2008; and the Volkwein corresponding et al., 2009). maintenance work. thecompare alternatives different would net be systems. a partial removal For instance, and re-installation Gerber and In 2008, the European Guideline ETAG 027 was published Volkwein (2007) analysed the performance of different sys- every year or an alternative protection measure such as gal- 5.4.3(EOTA Dimensioning, 2008; Peila and Ronco, 2009.). By letter of the Euro- leries.tems for either soft or hard dynamic decelerating processes. pean Commission to the Member States, the 1st of February TheIn the growing recent understanding years new rockfall of fence mitigation systems measures and their have dy- 2008 was considered the date of its availability and appli- gainednamic increasingbehaviour also attention. allows the So-called use of various attenuating net-type systems sys- Ifcability. a flexible ETAG protection 027 defines fence is a suitable testing forprocedure a specific similar site it to dotems not to try resist to stop impact a falling forces rock, caused but to by catch other it natural and to hazardsguide it hastheto Swiss be located guideline in the and field – after in such successful a way that system it covers testing most and such as avalanches (Margreth, 1995), falling sliding trees downhill in a controlled manner (see Fig. 19). Such barriers trajectoriesidentification and testing that the of falling the main rock components does not come as well to rest,as af- (Volkwein et al., 2009; Hamberger and Stelzer, 2007), de- are also called Hybrid Barriers or Hanger Nets (Glover et al., e.g.ter initial on the factory road to beproduction protected, inspection or reaches by the the clearance involved sec- ap- bris flows (Wendeler, 2008) or shallow landslides (Bugnion 2010; Dhakal et al., 2011a). tionproval of roadbody or – railwayallows the during producers deceleration to attach process. the CE A suit-mark- et al., 2008). ableing for fence the system barrier is on selected the basis according of relevant to the EC expected certificate max- of a 5.4.1 Historical development and current research imumnotified impact certification energy obtainedbody and with EC declarationthe aid of geological of conformity ex- 5.4.2 Standardization pertise.by the manufacturer.The arrangement The of the basis barrier for issuing in the field the has EC to certifi- fol- Mostly, the old-type fences were able to withstand just small lowcate the is the installation European instructions technical approval given in as the the accompanying concerned har- rockfallIt is important events. for Only the in planning the early and 1990s, design with of effective research pro- on manual.monized A technical ready-made specification, design load issued for the by anchors an approval according body tection systems that their behaviour is well understood and to the measured rope forces during prototype tests (see sec- how to stop falling rocks efficiently, was the dynamics of the entitled for these tasks and the implementation of a factory thoroughly verified. This also ensures an efficient use of pub- tion 5.4.4) is sometimes available online (BAFU Bundesamt decelerating process considered and used to design new re- production control system on the basis of the control plans, lic investment. Due to the complex, dynamic, and difficult to fur¨ Umwelt, 2011). In Switzerland, a partial safety factor of tention systems (Hearn et al., 1992). This also included the accompanying the European technical approval. It is typi- describe decelerating process a typical barrier design is based 1.3 has to be applied in compliance with (SIA261, 2003) on development of fences with retention capacities of up 50 kJ cal for such a broad guideline that many different interests on prototype testing. This procedure has also been adapted the load side. The safety of anchorage (e.g. micro-piles, bolts based on dynamic design approaches (Duffy, 1992; Duffy have to be combined and formulated. This usually becomes to produce standardization guidelines defining the minimum and anchors) has to be guaranteed according to CEN (2010). and Haller, 1993). Since then continuous research and engi- a quasi-minimum standard requiring National Application performance limits of solid barriers. Shu et al. (2005) describe results from anchorage testing. neering development has increased their retention capacities Documents for the single member states. www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 24264024 A. Volkwein et al.: Review on rockfall characterisationA.A. Volkwein Volkwein and et et al.:structural al.: Rockfall Rockfall protection review review

Fig. 20. Different testing methods for rockfall protection systems: Fig.Fig. 20. 20.DifferentDifferent testing testing methods methods for for rockfall rockfall protection protection systems: systems: free trajectory (left) with impact including rotation, but imprecise Fig.Fig.Fig. 21. 21. 21.StandardizedStandardizedStandardized test test test blocks blocks blocks for for for flexible flexible flexible rockfall rockfall rockfall protection protection protection freefree trajectory trajectory (left) (left) with with impact impact including including rotation rotation but but imprecise imprecise impact location; cable car guided oblique (middle) and vertical systemssystemssystems related related related to to to a a aregular regular regular cube cube cube with with with edge edge edge length length lengthLLLaccordingaccordingaccording to to to impactimpact location; location; cable cable car car guided guided oblique oblique (middle) (middle) and and vertical vertical (right) impact with precise impact location. thethethe approval approval approval guidelines guidelines guidelines of of of Switzerland Switzerland Switzerland (left, (left, (left, Gerber,Gerber Gerber,, 2001a,2001a 2001a,, until until until (right)(right) impact impact with with precise precise impact impact location. location. 2008)2008)2008) and and and the the the European European European Union Union Union (right, (right, (right, EOTA,EOTA EOTA,, 2008).2008 2008).).

5.4.45.4.4It must Field Field also testing testing be borne in mind that there will always be load cases outside the scope of the guidelines, such as eccentric ingingFor impossible impossible the tests, due duemainly to to the the two aim aim different not not to to stop stop setups the the falling are falling possible block block but de- but impact forces, post or rope strikes, high or low speed rockfall topendingto deviate deviate on it it and how and simply simply the falling to to control control rock its its is trajectory. trajectory. accelerated: inclined InIn order order to to verify verify and and validate validate the the setup setup for for newly-developed newly-developed guidance of test blocks along a track cable or their vertical rockfalleventsrockfall with protection protection the same fences fences impact full-scale full-scale energy, field field etc., tests tests (Wienberg are are necessary. necessary. et al., 2008; Volkwein et al., 2009). 5.4.5drops5.4.5 (see Numerical Numerical Fig. 20 Modelling, Gerber Modelling, 2001b). The barrier is then usu- FieldField testing testing has has been been performed performed from from the the beginning beginning (Hearn (Hearn ally installed with an inclination so that an impact angle be- etet al., al., 1992; 1992; Duffy, Duffy, 1992) 1992) and and continues continues to to the the present present day day FlexibletweenFlexible barrier rockfall rockfall and protectionprotection rockfall trajectory barriers barriers have ofhave 60 reached◦ reached(Gerber a adevelop-, develop-2001a) 5.4.3 Dimensioning (Zaitsev(Zaitsev et et al., al., 2010). 2010). A A summary summary of of flexible flexible barrier barrier testing testing mentorment±20 stage stage◦ between where where barrier considerable considerable and reference effort effort wouldslope would (EOTA be be required required, 2008) to is to toto withstand withstand rockfall rockfall up up to to 2008 2008 can can be be found found in in Thommen Thommen extendobtained.extend their their This rockfall rockfall represents retention retention a typical capacity. capacity. situation A A corresponding corresponding for free rockfall nu- nu- If a flexible protection fence is suitable for a specific site it (2008).(2008). Since Since then, then, the the testing testing methods methods have have not not changed changed mericalwhenmerical impacting simulation simulation a barrier enables enables in a the a more more field. efficient efficient development development or or has to be located in the field in such a way that it covers significantly.significantly. But But due due to to better better measurement measurement methods methods more more optimizationoptimizationThe test results of of new new are types retrievedtypes due due tousing to a a reduced differentreduced number number measurement of of ex- ex- most trajectories and that the falling rock does not come to detaileddetailed results results can can be be obtained, obtained, as as shown shown for for example example in in pensivesystems.pensive prototype prototype The geometry field field tests. oftests. the In barrierIn addition, addition, before the the and use use after of of soft- soft- the rest, e.g., on the road to be protected, or reaches the clear- GottardiGottardi and and Govoni Govoni (2010). (2010). waretestware is allows allows surveyed the the simulation usingsimulation leveling of of designed designed instruments barriers barriers or by tachymeters by consider- consider- ance section of road or railway during deceleration process. ForFor the the tests, tests, mainly mainly two two different different setups setups are are possible possible de- de- ingwithing special specialadditional load load manualcases cases that that measurements cannot cannot be be reproduced reproduced of brake element in in field field elon-tests tests A suitable fence system is selected according to the expected pendingpending on on how how the the falling falling rock rock is is accelerated: accelerated: inclined inclined guid- guid- (high-speedgations,(high-speed post rockfall, rockfall, inclinations, post/rope post/rope etc. strikes strikesThe braking etc.), etc.), as asprocess well well as as special for special the maximum impact energy obtained with the aid of geologi- anceance of of test test blocks blocks along along a atrack track cable cable or or their their vertical vertical drops drops geometricalfallinggeometrical rock boundary can boundary be obtained conditions conditions either for for individual from individual frame-per-frame topographi- topographi- cal expertise. The arrangement of the barrier in the field has (see(see Fig. Fig. 20, 20, Gerber Gerber (2001b)). (2001b)). The The barrier barrier is is then then usually usually calanalysiscal situations situations of high-speed or or the the influence influence video of recordings of structural structural (min. changes changes 100 on frames on bar- bar- to follow the installation instructions given in the accompa- installedinstalled with with an an inclination inclination such such that that an an impact impact angle angle be- be- rierperrier performance second performance recommended) (Fornaro (Fornaro et or et al., from al., 1990; 1990; numerical Mustoe Mustoe integration and and Huttel- Huttel- of nying manual. A ready-made design load◦ ◦ for the anchors tweentween barrier barrier and and rockfall rockfall trajectory trajectory of of6060(Gerber,(Gerber, 2001a) 2001a) maier,themaier, block’s 1993; 1993; internal Akkaraju, Akkaraju, acceleration 1994; 1994; Nicot Nicot measurements et et al., al., 1999, 1999, (sample 2001; 2001; Caz- rateCaz- according◦ ◦ to the measured rope forces during prototype tests oror±±2020betweenbetween barrier barrier and and reference reference slope slope (EOTA, (EOTA, 2008) 2008) zani>zani1− et et2 al., kHz al., 2002; 2002; recommended). Anderheggen Anderheggen et et al., al., 2002; 2002; Volkwein, Volkwein, 2004; 2004; (see Sect. 5.4.4) is sometimes available online (BAFU Bun- isis obtained. obtained. This This represent represent a a typical typical situations situations for for free free rock- rock- SasiharanSasiharanThe typical et et al., al., test 2006). 2006). boulders Apart Apart are from from specially the the numerical numerical manufactured modelling modelling con- desamt fur¨ Umwelt, 2011). In Switzerland, a partial safety fallfall when when impacting impacting a abarrier barrier in in the the field. field. ofcreteof full full elements protection protection (see systems systems Fig. 21 also) alsowith justjust different single single masses components components according can can factor of 1.3 has to be applied in compliance with (SIA261, TheThe test test results results are are retrieved retrieved using using different different measurement measurement betobe evaluatedguideline evaluated energynumerically. numerically. classes Related with Related an work impact work has has velocity been been done of done mini- for for 2003) on the load side. The safety of anchorage (e.g., micro- systems.systems. The The geometry geometry of of the the barrier barrier before before and and after after the the e.g.mume.g. energy 25energy m s dissipating− dissipating1. This velocity elements elements is considered (del (del Coz Coz D being D´ıaz´ıaz et in et al., the al., 2010; upper 2010; piles, bolts and anchors) has to be guaranteed according to testtest is is surveyed surveyed using using levelling levelling instruments instruments or or tachymeters tachymeters Studer,rangeStuder, of 2001; 2001;rockfall Dhakal Dhakal events. et et al., al., 2011b) 2011b) or or net net rings rings (Nicot (Nicot et et al., al., CEN (2010). Shu et al. (2005) describe results from anchor- withwith additional additional manual manual measurements measurements of of brake brake element element elon- elon- 1999;1999;In Volkwein, recentVolkwein, years 2004). 2004). the investigations have concen- age testing. gations,gations, post post inclinations inclinations etc.. etc.. The The braking braking process process for for the the tratedLargeLarge more deformations deformations on the causing causing testing geometrical geometrical of attenuating non-linearity, non-linearity, systems, the the fallingfalling rock rock can can be be obtained obtained either either from from frame-per-frame frame-per-frame short-timee.g.,short-timeGlover simulation et simulation al. (2010 period). period Here, and oblique and nonlinear nonlinear impact material is material mandatory be- be- 5.4.4 Field testing analysisanalysis of of high-speed high-speed video video recordings recordings (min. (min. 100 100 frames frames haviourandhaviour vertical requires requires testing explicit explicit impossible FE FE analysis analysis due to strategies strategies the aim such not such to as as stop the the perper second second recommended) recommended) or or from from numerical numerical integration integration of of CentraltheCentral falling Differences Differences block, but Method Method to deviate used used it e.g.e.g. and by simplyby Bathe Bathe to (2001); (2001); control An- An- its theInthe order block’s block’s to internalverify internal and acceleration acceleration validate the measurements measurementssetup for newly-developed (sample (sample rate rate derheggentrajectory.derheggen et et al. al. (1986). (1986). This This provides provides a adetailed detailed view view of of the the >rockfall>1 1−−2 2kHz protectionkHzrecommended).recommended). fences, full-scale field tests are necessary. system’ssystem’s dynamic dynamic response. response. It It can can also also deliver deliver information information FieldTheThe testing typical typical was test test performed boulders boulders are from are specially specially the beginning manufactured manufactured (Hearn et con- con- al., on5.4.5on the the loading Numericalloading and and degree modelling degree of of utilization utilization of of any any modelled modelled sys- sys- crete1992crete elements; Duffy elements, 1992 (see (see) Fig. and Fig. 21) continues 21) with with different to different the present masses masses day according according (Zaitsev temtem configuration. configuration. The The simulation simulation of of the the falling falling rock rock should should toetto guideline al. guideline, 2010). energy energy A summary classes classes of with with flexible an an impact impact barrier velocity velocitytesting of to of min-with- min- takeFlexibletake into into account rockfall account large protection large three-dimensional three-dimensional barriers have displacements displacements reached a devel- and and imumstandimum rockfall2525m/sm/s. up. Thisto This 2008 velocity velocity can be is is considered found considered in Thommen of of being being( in2008 in the the). rotations.opmentrotations. stage When When where impacting impacting considerable a a steel steel neteffort net at at wouldany any location, location, be required spe- spe- upperSinceupper range then, range of the of rockfall rockfall testing events. events.methods have not changed signifi- cialtocial extend contact contact their algorithms algorithms rockfall prevent prevent retention the the capacity.net net nodes nodes from A from correspond- penetrat- penetrat- cantly.InIn recent recent But, years yearsdue the to the better investigations investigations measurement have have concentratedmethods, concentrated more more more de- inginging the numerical the rock rock permitting permitting simulation only only enables tangential tangential a more movements. movements. efficient All develop- All slid- slid- ontailedon the the results testing testing can of of be attenuating attenuating obtained, systems,as systems, showne.g. for e.g. example Glover Glover in etGot- et al. al. ingmenting effects effects or optimization taking taking place place of in innew the the model types model usually due usually to occura occur reduced over overnum- long long (2010).tardi(2010). and Here, Here, Govoni oblique oblique(2010 impact). impact is is mandatory mandatory and and vertical vertical testtest- distancesberdistances of expensive and and also also prototype cause cause friction friction field between tests. between In the addition, the various various the compo- compo- use of

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2641 software allows the simulation of designed barriers by con- 5.5 Forests sidering special load cases that cannot be reproduced in field tests (high-speed rockfall, post/rope strikes, etc.), as well as The most natural type of protection is a forest. Its protective special geometrical boundary conditions for individual topo- effect is basically due to the barrier effect (energy dissipa- graphical situations or the influence of structural changes on tion) of standing and lying trees. Whether this barrier effect barrier performance (Fornaro et al., 1990; Mustoe and Hut- is effective or not is determined by the size and kinetic en- telmaier, 1993; Akkaraju, 1994; Nicot et al., 1999, 2001; ergy of the rock, the total basal area that is available to inter- Cazzani et al., 2002; Anderheggen et al., 2002; Volkwein, cept the falling rock, as well as the tree species (Berger and 2004; Sasiharan et al., 2006). Apart from the numerical mod- Dorren, 2007). In rockfall protection forests, the concept of elling of full protection systems, also single components can the basal area is important as it comprises both the density be evaluated numerically. Related work has been done, for of the forest (how many tree stems per hectare are present) example, energy dissipating elements (del Coz D´ıaz et al., and the diameter distribution of the trees. The definition of 2010; Studer, 2001; Dhakal et al., 2011b) or net rings (Nicot total basal area is the total area covered by all trunks in cross et al., 1999; Volkwein, 2004). section, usually measured at breast height, per hectare. Basal − Large deformations causing geometrical nonlinearity, the area is, therefore, expressed in m2 ha 1. The lower limit of − short-time simulation period and nonlinear material be- an effective protection forest is about 10 m2 ha 1, whereas a − haviour requires explicit FE analysis strategies such as the forest with 25 m2 ha 1 will be able to provide a significant Central Differences Method used e.g., by Bathe (2001); An- level of protection against rockfall. This, however, depends derheggen et al. (1986). This provides a detailed view of the on the previously mentioned factors (rock energy, species, system’s dynamic response. It can also deliver information and length of forested slope, etc.). An assessment of the pro- on the loading and degree of utilisation of any modelled sys- tective function of the forest can be carried out using rapid as- tem configuration. The simulation of the falling rock should sessment tools and protection forest guidelines (e.g., Frehner take into account large three-dimensional displacements and et al., 2005; Berger and Dorren, 2007) or with more complex rotations. When impacting a steel net at any location, spe- rockfall trajectory models that account for the barrier effect cial contact algorithms prevent the net nodes from penetrat- of single trees (e.g., Dorren, 2010; Rammer et al., 2010). ing the rock permitting only tangential movements. All slid- Various research investigations have been carried out to ing effects taking place in the model usually occur over long obtain a detailed knowledge of the capacity of a forest to distances and also cause friction between the various compo- stop falling rocks, as shown in the fundamental summary on nents. the state of the art of rockfall and forest interactions (Dor- Up till now, different strategies to model flexible rockfall ren et al., 2007). It is generally agreed that not only large fences have been pursued. The design of a special tailor- trees are required in a rockfall protection forest, but that made software allows one to focus on the relevant details and well-structured stands with a wide diameter distribution and neglect unwanted parts and, therefore, speeds up the compu- a mosaic of different forest development phases provide the tations (Nicot et al., 1999; Volkwein, 2004). Such an ap- best rockfall protection. Experiments have shown clearly that proach also facilitates the setup of different barrier models, small trees are capable of stopping large rocks, provided that because all software elements are already optimized for the a large part of the kinetic energy has already been dissipated simulated components. This method, however, needs a large during preceding impacts against large trees. amount of time until usable results are available. Therefore, The repartition of large and small trees, which usually also the use of common multi-purpose FE codes is also recom- corresponds to the height of the trees, is referred to as the ver- mendable because it saves the time-consuming development tical forest structure. Furthermore, the higher the stand den- of routine functions (Fornaro et al., 1990). This again is at the sity, the higher the contact probability, but this also depends risk of non-ideal element properties or performance. Finally, on the rock size since small rocks have a lower encounter more abstract models, e.g., with a numerically much simpli- probability than large rocks. A problem in protection forest fied net performance, allow the simulation with systems that management is that dense forest stands cannot be maintained have not yet been fully explored. over a long period of time by having thick trees and a high Regardless of the approach adopted to simulate a flexible stability. Therefore, a compromise has to be found between barrier, the results of the simulations should be validated by an optimal protective function while assuring forest stabil- full-scale rockfall field tests measuring the cable and support ity and renewal (Brang, 2001). The number of tree stems forces as well as accelerations and the trajectory of the falling and their spatial repartition is referred to as the horizontal rock. forest structure. An important characteristic with respect to the horizontal structure that determines the protection against rockfall is the length and number of gaps and couloirs in the forest.

www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2642 A. Volkwein et al.: Review on rockfall characterisation and structural protection

Over the last decade, research on the interaction between susceptibility vs. rockfall hazard should be discussed. rockfall and protection forest has intensified. Examples are It is also important to have a thorough knowledge of the Lundstrom¨ (2010) and Jonsson (2007), who studied the me- extreme variations of trajectories within a certain area. chanical stability and energy absorption of single trees. A They define the decisive fractiles relevant for the map- link between the protective capacity of a single tree and the ping process. However, all this is of no avail, if the efficacy of a forest stand has been made by Kalberer (2007). reliability of models with a proper physical basis is not Jancke et al. (2009) investigated the protective effect of dif- checked properly. ferent coppice stands. Le Hir et al. (2006), Rammer et al. (2010) and Dorren (2010) have proposed new approaches for – Secondly, a specific design level has to be uniformly de- integrating forest in rockfall trajectory models. Monnet et al. fined for protection measures. This can be achieved by (2010) showed, by way of an example, how laser-scanning quantifying the risk level, the vulnerability of the pro- data can be used for the automatic characterisation of rock- tection countermeasures and the involved costs for life- fall protection. Advances in dendro-geomorphology provide cycles of the mitigation measure and for overall risk re- an improved spatiotemporal analysis of the silent witnesses duction. Of course, standardized evaluation and veri- of rockfall (e.g., Schneuwly and Stoffel, 2008). Important fication procedures for the countermeasures need to be remaining subjects in this area are the effect of lying stems defined. on rockfall trajectories, decomposition of lying and standing – Further, more discussion on what is the best way to clas- dead wood and the optimal protection forest stand character- sify a single rockfall event is needed. It could be satis- istics for different rockfall settings (coppice stands, homoge- factorily described using either the energy in kJ or the neous beech forest, maximum gap length, etc). impulse in Ns. The first is more common and state- of-the-art, but the latter is sometimes more exact when considering impact and rebound effects. 6 Summary and outlook – Finally, it is becoming increasingly important for re- Todays rockfall hazard issues and estimation of the risk searchers from different disciplines to establish close of rockfall are considered essential. Research on rockfall- collaboration. Today’s demands on applicability and related topics is an important task and advances are clearly efficiency rule out isolated studies lacking interaction. visible. In addition, structural countermeasures also based Such collaboration could result in valuable products like on uncertainty models are also of practical interests. This ar- this paper or a book on rockfall (Lambert and Nicot, ticle, therefore, consists of four main chapters, namely rock- 2011). fall hazard, rockfall source areas, trajectory modelling and structural countermeasures. Acknowledgements. Without the work of a lot of researches world Numerical simulation nowadays allows for a calculation wide this summary article wouldn’t contain so much information. of trajectories at a very high level of precision (see Sect. 4). The authors further thank E. G. Prater for the harmonization and For example, the rockfall process can be simulated using the improvement of this article, Johanna Scheidegger for her work on DE method based on highly detailed laser scans as input, etc. the reference list and two reviewers who did an excellent job. However, such a detailed level would also require the consid- Edited by: T. Glade eration of the block’s shape, its exact position before the re- Reviewed by: M. Molk¨ and another anonymous referee. lease, etc. Therefore, an alternative approach also has its va- lidity: There is no essential need for sophisticated simulation models to estimate the velocities in rockfall events. A few References clearly visible impact locations and some basic mathematics are sufficient to calculate the trajectory (see Sect. 4.4.1). The Acosta, E., Agliardi, F., Crosta, G. B., and Rios Aragues,` S.: Re- positions of impact locations on the ground, the inclinations gional rockfall hazard assessment in the Benasque Valley (Centra Pyrenees) using a 3-D numerical approach, in: 4th EGS Plinius between them and – if available – above ground traces on Conference – Mediterranean Storms, 555–563, Universitat des tree branches permit the definition of the block’s lift-off and Illes Balears, Mallorca, Spain, 2003. impact velocities. This contribution includes the formulas Agliardi, F. and Crosta, G.: High resolution three-dimensionnal nu- necessary to calculate the velocities and with the possibility merical modelling of rockfalls, Int. J. Rock. Mech. Min., 40, of graphical presentation. 455–471, 2003. What are the questions needing attention in the immediate Agliardi, F., Crosta, G. B., and Frattini, P.: Integrating rock- future? Here are some suggestions: fall risk assessment and countermeasure design by 3D modelling techniques, Nat. Hazards Earth Syst. Sci., 9, 1059–1073, – Firstly, there is a definite need to improve the prediction doi:10.5194/nhess-9-1059-2009, 2009. of probabilities in hazard and risk assessment in order to Akkaraju, L.: Dynamic Analysis of Cable Structures, Master’s the- better quantify the risk of rockfall and to improve haz- sis, University of Colorado, Boulder, masterthesis, University of ard and risk maps. In this context, in addition rockfall Colorado at Boulder, 1994.

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2643

Aksoy, H. and Ercanoglu, M.: Determination of the rockfall Bathe, K.-J.: Finite Element Methoden, Springer-Verlag, Berlin, source in an urban settlement area by using a rule-based 2001. fuzzy evaluation, Nat. Hazards Earth Syst. Sci., 6, 941–954, Baumann, P.: Lastfalle¨ und Bemessungsansatz bei Sturzprozessen, doi:10.5194/nhess-6-941-2006, 2006. in: FAN – Herstkurs 2008, Fachleute Naturgefahren Schweiz, Anderheggen, E., Elmer, H., and Maag, H.: Nichtlineare Finite- Bellinzona, 2008. Element-Methoden: Eine Einfuhrung¨ fur¨ Ingenieure, Institut fur¨ Berger, F. and Dorren, L.: Objective comparison of rockfall mod- Informatik, Zurich,¨ 1986. els using real size experimental data, in: Disaster mitigation of Anderheggen, E., Volkwein, A., and Grassl, H.: Computational debris flows, slope failures and landslides, 245–252, Universal Simulation of Highly Flexible Rockfall Protection Systems, Academy Press, Inc, Tokyo, Japan, 2006. in: Proc. Fifth World Congress on Computational Mechanics Berger, F. and Dorren, L. K. A.: Principles of the tool Rockfor.NET (WCCM V), edited by: Mang, H., Rammerstorfer, F., and Eber- for quantifying the rockfall hazard below a protection forest, hardsteiner, J., Vienna University of Technology, 2002. Schweizerische Zeitschrift fur¨ Forstwesen, 158, 157–165, 2007. ASTRA: Steinschlag – Naturgefahr fur¨ die Nationalstrassen, Berthet-Rambaud, P.: Structures rigides soumises aux avalanches Schlussbericht der ASTRA Expertengruppe, Tech. rep., Bunde- et chutes de blocs: modelisation´ du comportement mecanique´ et samt fur¨ Strassen, 2003. caracterisation´ de l’interaction phenom´ ene-ouvrage,` Ph.D. the- ASTRA and SBB: Einwirkungen auf Steinschlagschutzgalerien, sis, Universite Grenoble, 2004. Richtlinie, Tech. rep., Bundesamt fur¨ Baudirektion SBB, 18 Bertrand, D., Nicot, F., Gotteland, P., and Lambert, S.: Modelling pages, Bern, 1998. a geo-composite cell using discrete analysis, Comput. Geotech., ASTRA: Einwirkungen infolge Steinschlags auf Schutzgalerien, 32, 564–577, 2006. Tech. rep., Bundesamt fur¨ Strassen, Baudirektion SBB, Eid- Bieniawski, Z. T.: Engineering classification of jointed rock masses, genossische¨ Drucksachen- und Materialzentrale, 2008. Trans. S. Afr. Inst. Civ. Engrs, 15, 335–344, 1973. Azimi, C. and Desvarreux, P.: Calcul de chutes de blocs Bieniawski, Z. T.: Classification of rock masses for engineering: et verification´ sur modele` reduit,´ Association pour le the RMR system and future trends, Comprehensive Rock Eng., developpement´ des recherches sur les glissements de terrain, 3, 553–573, 1993. Grenoble, 1977. Blais-Stevens, A.: Landslide Hazards and their mitigation along the Azimi, C. and Desvarreux, P.: Les chutes de pierres: Exemple No2 sea to sky corridor, British Columbia, in: 4th Canadian Con- (Galerie de protection), Stage paravalanches, E.N.P.C., Paris, ference on Geohazards: from causes to management, edited by 1988. Locat, J., Perret, D., Turmel, D., Demers, D., and Leroueil, S., Azimi, C., Desvarreux, P., Giraud, A., and Martin-Cocher, J.: Quebec, Canada, 2008. Methode´ de calcul de la dynamique des chutes de blocs – Appli- Blovsky, S.: Model tests on protective barriers against rockfall, in: cation a` l’etude´ du versant de la montagne de La Pale (Vercors), 15th EYGEC – European Young Geotechnical Engineers Con- Bulletin de liaison des laboratoires des ponts et chaussees,´ 122, ference, 2002. 93–102, 1982. Bourrier, F.: Modelisation´ de l’impact d’un bloc rocheux sur un Azzoni, A. and De Freitas, M.: Experimentally gained parameters, terrain naturel, application a` la trajectographie des chutes de decisive for rock fall analysis, Rock Mech. Rock Eng., 28, 111– blocs, Ph.D. thesis, Institut Polytechnique de Grenoble, Greno- 124, 1995. ble, 2008. Azzoni, A., La Barbera, G., and Zaninetti, A.: Analysis and predic- Bourrier, F., Nicot, F., and Darve, F.: Physical processes within a 2D tion of rock falls using a mathematical model, Int. J. Rock Mech. granular layer during an impact, Granular Matter, 10, 415–437, Min., 32, 709–724, 1995. 2008. Azzoni, A., Rossi, P. P., Drigo, E., Giani, G. P., and Zaninetti, A.: Bourrier, F., Dorren, L., Nicot, F., Berger, F., and Darve, F.: To- In situ observation of rockfall analysis parameters, in: Sixth In- wards objective rockfall trajectory simulation using a stochastic ternational Symposium of Landslides, 307–314, Rotterdam, The impact model, Geomorphology, 110, 68–79, 2009a. Netherlands, 1992. Bourrier, F., Eckert, N., Nicot, F., and Darve, F.: Bayesian stochas- BAFU Bundesamt fur¨ Umwelt: Zugelassene Steinschlagschutzsys- tic modeling of a spherical rock bouncing on a coarse soil, Nat. teme, Tech. rep., Swiss Federal Office for the Environment, Hazards Earth Syst. Sci., 9, 831–846, doi:10.5194/nhess-9-831- Berne, http://www.umwelt-schweiz.ch/typenpruefung, 2011. 2009, 2009b. Baillifard, F.: Detection par SIG des zones rocheuses a` fortessus- Bozzolo, D. and Pamini, R.: Simulation of rock falls down a valley ceptibilites´ d’eboulement,´ Ph.D. thesis, University of Lausanne, side, Acta Mech., 63, 113–130, 1986. 2005. Bozzolo, D., Pamini, R., and Hutter, K.: Rockfall analysis – A Baillifard, F., Jaboyedoff, M., and Sartori, M.: Rockfall hazard mathematical model and its test with field data, in: 5th Inter- mapping along a mountainous road in Switzerland using a GIS- national Symposium on Landslides, 555–563, Balkema, Rotter- based parameter rating approach, Nat. Hazards Earth Syst. Sci., damm, Lausanne, Switzerland, 1988. 3, 435–442, doi:10.5194/nhess-3-435-2003, 2003. Brabb, E.: Innovative approaches to landslide hazard and risk map- Baillifard, F., Jaboyedoff, M., Rouiller, J. D., Couture, R., Locat, ping, 4th International Symposium on Landslides, 1, 307–323, J., Robichaud, G., and Gamel, G.: Towards a GIS-based hazard 1984. assessment along the Quebec City Promontory, in: Landslides Brang, P.: Resistance and elasticity: promising concepts for Evaluation and stabilization, edited by: Lacerda, W., Ehrlich, A., the management of protection forests in the European Alps, Fontoura, M., and Sayao, A., 207–213, Taylor & Francis, Que- Forest Ecol. Manage., 145, 107–119, doi:10.1016/S0378- bec, Canada, 2004. 1127(00)00578-8, 2001.

www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2644 A. Volkwein et al.: Review on rockfall characterisation and structural protection

Brideau, M.-A., Stead, D., Roots, C., and Orwin, J.: Geomorphol- Chau, K. T., Wu, J., Wong, R., and Lee, C.: The coefficient of resti- ogy and engineering geology of a landslide in ultramafic rocks, tution for boulders falling onto soil slopes with various values of Dawson City, Yukon, Eng. Geol., 89, 171–194, 2007. dry density and water content, in: International Symposium on Broili, L.: In situ tests for the study of rockfall, Geologia Applicata Slope Stability Engineering: Geotechnical and Geoenvironmen- e Idrogeologia, 8, 105–111, 1973. tal Aspects, 1355–1360, Matsuyama, Japan, 1999b. Broili, L.: Relations between scree slope morphometry and dynam- Chau, K. T., Wong, R., and Wu, J.: Coefficient of restitution and ics of accumulation processes, in: Meeting on Rockfall dynamics rotational motions of rockfall impacts, Int. J. Rock Mech. Min., and protective works effectiveness, 11–23, Bergamo, Italy, 1977. 39, 69–77, 2002. Budetta, P.: Assessment of rockfall risk along roads, Nat. Hazards Chau, K. T., Wong, R., Liu, J., and Lee, C.: Rockfall hazard analysis Earth Syst. Sci., 4, 71–81, doi:10.5194/nhess-4-71-2004, 2004. for Hong Kong based on rockfall inventory, Rock Mech. Rock Budetta, P. and Santo, A.: Morphostructural evolution and related Eng., 36, 383–408, 2003. kinematics of rockfalls in Campania (southern Italy): A case Chau, K. T., Wong, R., and Lee, C. F.: Rockfall Problems in Hong study, Eng. Geol., 36, 197–210, 1994. Kong and some new Experimental Results for Coefficients of Bugnion, L., Denk, M., Shimojo, K., Roth, A., and Volkwein, Restitution, Int. J. Rock Mech. Min., 35, 662–663, 1998b. A.: Full-scale experiments on shallow landslides in combina- Chikatamarla, R.: Optimisation of cushion materials for rockfall tion with flexible protection barriers, in: First World Landslide protection galleries, Ph.D. thesis, Swiss Federal Institute of Tech- Forum, 99–102, United Nations University, Tokyo, 2008. nology ETHZ, Zurich, 2006. Bunce, C. M., Cruden, D., and Morgenstern, N.: Assessment of the Christen, M., Bartelt, P., and Gruber, U.: RAMMS – a modelling hazard from rock fall on a highway, Can. Geotech. J., 34, 344– system for snow avalanches, debris flows and rockfalls based 356, 1997. on IDL., PFG Photogrammetrie – Fernerkundung – Geoinfor- Burroughs, D., Henson, H. H., and Jiang, S.: Full scale geotextile mation, 4, 289–292, 2007. rock barrier wall testing, analysis and prediction, Geosynthetics’ Coe, J. A. and Harp, E. L.: Influence of tectonic folding on rockfall 93, 1993. susceptibility, American Fork Canyon, Utah, USA, Nat. Hazards Calvetti, F.: Distinct Element evaluation of the rock-fall design load Earth Syst. Sci., 7, 1–14, doi:10.5194/nhess-7-1-2007, 2007. for shelters, Rivista Italiana di Geotecnica, 3, 63–83, 1998. Comite-Euro-International´ du Beton,´ C.: Concrete Structures under Calvetti, F. and Di Prisco, C.: An uncoupled approach for the de- Impact and Impulsive Load, Lausanne, 1988. sign of rockfall protection tunnels, Struct. Eng. Int., 19, 342–347, Copons, R. and Vilaplana, J.: Rockfall susceptibility zoning at a 2009. large scale: From geomorphological inventory to preliminary Calvetti, F., Di Prisco, C., and Vecchiotti, M.: Experimental and land use planning, Eng. Geol., 102, 142–151, 2008. numerical study of rock-fall impacts on granular soils, Rivista Corominas, J., Copons, R. J. M., Vilaplana, J., Altimir, J., and Italiana di Geotecnica, 4, 95–109, 2005. Amigo, J.: Quantitative assessment of the residual risk in a rock- Camponuovo, G.: ISMES experience on the model of St. Martino, fall protected area, Landslides, 2, 343–357, 2005. in: Meeting on Rockfall dynamics and protective works effec- Crosta, G. B. and Agliardi, F.: A methodology for physically based tiveness, 25–38, Bergamo, Italy, 1977. rockfall hazard assessment, Nat. Hazards Earth Syst. Sci., 3, Cancelli, A. and Crosta, G.: Hazard and risk assessment in rock- 407–422, doi:10.5194/nhess-3-407-2003, 2003. fall prone areas, Risk Reliability in Ground Engineering, Thomas Crosta, G. B. and Agliardi, F.: Parametric evaluation of 3D dis- Telford, 1993. persion of rockfall trajectories, Nat. Hazards Earth Syst. Sci., 4, Carere, K., Ratto, S., and Zanolini, F., eds.: Prevention´ des mouve- 583–598, doi:10.5194/nhess-4-583-2004, 2004. ments de versants et des instabilites´ de falaises, Programme In- Crosta, G., Agliardi, F., Frattini, P., and Imposato, S.: A three- terreg IIC – Falaises, Mediterran´ ee´ occidentale et Alpes latines, dimensional hybrid numerical model for rockfall simulation, confrontation des methodes´ d’etude´ des eboulements´ dans l’arc Geophys. Res. Abstr., 6, 2004. alpin, 2001. Cundall, P.: A computer model for simulating progressive, large- Casanovas, M.: Dimensionamiento de galerias de proteccion scale movements in blocky rock systems, in: Symp. Int. Soc. frente a desprendimientos de rocas, Master’s thesis, Universitat Rock Mech., 1, Paper No. II–8, Nancy, France, 1971. Politecnica` de Catalunya, 2006. del Coz D´ıaz, J., Nieto, P. G., Castro-Fresno, D., and Rodr´ıguez- Cazzani, A., Mongiovi, L., and Frenez, T.: Dynamic Finite Element Hernandez,´ J.: Nonlinear explicit analysis and study of the be- Analysis of Interceptive Devices for Falling Rocks, Int. J. Rock haviour of a new ring-type brake energy dissipator by FEM Mech., 39, 303–321, 2002. and experimental comparison, Appl. Math. Comput., 216, 1571– CEN: EN 1997-1 – Eurocode 7 – Geotechnical Design, Tech. rep., 1582, 2010. European Committee for Standardization, Brussels, 2010. Derron, M.-H., Jaboyedoff, M., and Blikra, L. H.: Preliminary as- Chau, K. T., Chan, L. C. P., Wu, J. J., Liu, J., Wong, R. H. C., and sessment of rockslide and rockfall hazards using a DEM (Opp- Lee, C. F.: Experimental studies on rockfall and debris flow, in: stadhornet, Norway), Nat. Hazards Earth Syst. Sci., 5, 285–292, One Day Seminar on Planning, Design and Implementation of doi:10.5194/nhess-5-285-2005, 2005. Debris Flow and Rockfall Hazards Mitigation Measures, 115– Descoeudres, F.: Aspects geom´ ecaniques´ des instabilites´ de falaises 128, Hongkong, China, 1998a. rocheuses et des chutes de blocs, Publications de la societ´ e´ suisse Chau, K. T, Wong, R., Liu, J., Wu, J. J., and Lee, C. F.: Shape ef- de mecanique´ des sols et des roches, 135, 3–11, 1997. fects on the coefficient of restitution during rockfall impacts, in: Descoeudres, F. and Zimmermann, T.: Three-dimensional dynamic 9th International Congress on Rock Mechanics, 541–544, Paris, calculation of rockfalls, in: Sixth International Congress on Rock France, 1999a. Mechanics, pp. 337–342, International Society for Rock Me-

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2645

chanics, Montreal, Canada, 1987. Rotterdamm, Lausanne, Switzerland, 1988. Dhakal, S., Bhandary, N. P., Yatabe, R., and Kinoshita, N.: Numer- EOTA: ETAG 027 – guideline for the European technical approval ical investigation of the effects of idealized rock-block shapes of falling rock protection kits, Tech. rep., European Organization and impact points on the performance of Long-span Pocket-type for Technical Approvals, Brussels, 2008. Rock-net, in: 46th National Conference on Geotechnical En- Evans, S. and Hungr, O.: The assessment of rockfall hazard at the gineering, pp. 1185–1186, Japanese Geotechnical Society JGS, base of talus slopes, Can. Geotech. J., 30, 620–636, 1993. Kobe, Japan, 2011a. Falcetta, J.: Un nouveau modele` de calcul de trajectoires de blocs Dhakal, S., Bhandary, N. P., Yatabe, R., and Kinoshita, N.: Con- rocheux, Revue Française de Geotechnique,´ 30, 11–17, 1985. stitutive modeling of friction damper for numerical simulation Fell, R., Ho, K., Lacasse, S., and Leroi, E.: A framework for land- of Long-span Pocket-type Rock-net, in: Annual Conference of slide risk assessment and management, in: Landslide Risk Man- Japan Society of Civil Engineers JSCE, pp. 1185–1186, Shikoku agement, edited by: Hungr, O. and E., F. R. C. R. E., 3–26, Taylor Branch, Kagawa, Japan, 2011b. and Francis, London, 2005. Dimnet, E.: Mouvement et collisions de solides rigides ou Fell, R., Corominas, J., Bonnard, C., Cascini, L., Leroi, E., and deformables,´ Ph.D. thesis, Ecole Nationale des Ponts et Savage, W.: Guidelines for landslide susceptibility, hazard and Chaussees,´ Paris, 2002. risk zoning for land use planning, Eng. Geol., 102, 85–98, 2008. Dimnet, E. and Fremond,´ M.: Instantaneous collisions of solids, Fornaro, M., Peila, D., and Nebbia, M.: Block falls on rock slopes in: European Congress on Computational Methods in Applied – application of a numerical simulation program to some real Sciences and Engineering, 11–17, Barcelona, Spain, 2000. cases, in: 6th International Congress IAEG, Rotterdam, The Dorren, L. and Berger, F.: Stem breakage of trees and energy dissi- Netherlands, 1990. pation at rockfall impacts, Tree Physiol., 26, 63–71, 2006. Frattini, P., Crosta, G., Carrara, A., and Agliardi, F.: Assessment of Dorren, L. and Seijmonsbergen, A.: Comparison of three GIS- rockfall susceptibility by integrating statistical and physically- based models for predicting rockfall runout zones at a regional based approaches, Geomorphology, 94, 419–437, 2008. scale, Geomorphology, 56, 49–64, 2003. Frayssines, M.: Contribution a` l’evaluation´ de l’alea´ eboulement´ Dorren, L. K. A., Berger, F., and Putters, U. S.: Real-size ex- rocheux (rupture), Ph.D. thesis, Universite´ Josef Fourier, 2005. periments and 3-D simulation of rockfall on forested and non- Frehner, M., Wasser, B., and Schwitter, R.: Nachhaltigkeit und Er- forested slopes, Nat. Hazards Earth Syst. Sci., 6, 145–153, folgskontrolle im Schutzwald – Wegleitung fur¨ Pflegemassnah- doi:10.5194/nhess-6-145-2006, 2006. men in Waldern¨ mit Schutzfunktion, Tech. rep., Swiss Federal Dorren, L., Berger, F., Jonnson, M., Krautblatter, M., Moelk, M., Office for the Environment FOEN, Bern, 2005. Stoffel, M., and Wehrli, A.: State of the art in rockfall - for- Fremond,´ M.: Rigid bodies collisions, Phys. Lett. A., 204, 33–41, est interactions, Schweizerische Zeitschrift fur¨ Forstwesen, 158, 1995. 128–141, 2007. GEO: Landslides and boulder falls from natural terrains: interim Dorren, L. K. A.: Rockyfor3D revealed – description of the com- risk guidelines, GEO Report 75, Geotechnical Engineering Of- plete 3D rockfall model, Tech. rep., EcorisQ, http://www.ecorisq. fice, Civiel Engineering Department, Hong Kong, 1998. org, 2010. Gerber, W.: Richtlinie uber¨ die Typenprufung¨ von Schutznetzen Dorren, L. K. A., Maier, B., Putters, U. S., and Seijmonsbergen, gegen Steinschlag, Tech. rep., Bundesamt fur¨ Umwelt, Wald und A.C.: Combining field and modelling techniques to assess rock- Landschaft (BUWAL), Eidgenossische¨ Forschungsanstalt WSL, fall dynamics on a protection forest hillslope in the European Bern, 2001a. Alps, Geomorphology, 57, 151–167, 2004. Gerber, W.: Vergleich zwischen Vertikal- und Schragwurfanlagen¨ Dudt, J. and Heidenreich, B.: Treatment of the uncertainty in a zur Typenprufung¨ von flexiblen Steinschlschutzverbauungen, three-dimensional numerical simulation model for rock falls, in: Tech. rep., Eidg. Forschungsanstalt fur¨ Wald, Schnee und Land- International Conference on Landslides – Causes, Impacts and schaft WSL, Birmensdorf, 2001b. Countermeasures, 507–514, Davos, Switzerland, 2001. Gerber, W.: Einwirkungen bei Steinschlag, in: FAN – Herstkurs Duffy, J. D.: Flexible Wire Rope Rockfall Nets, in: Soils, Geology, 2008, Fachleute Naturgefahren Schweiz, Bellinzona, 2008. and Foundations – Rockfall prediction and Control and landslide Gerber, W. and Volkwein, A.: Different flexible Rockfall Barriers – case histories (Transportation Research Record No. 1343), 30– comparative Results from Type Testing, Geophys. Res. Abstr., 9, 35, Trans. Res. B., 1992. 2007. Duffy, J. D. and Haller, B.: Field Tests of Flexible Rockfall Barri- Ghadimi-Khasraghy, S., Kishi, N., and Vogel, T.: Numerical sim- ers, in: Proc. Transportation Facilities through Difficult Terrain, ulation of consecutive rockfall impacts on reinforced concrete edited by: Wu, J. T. and Barrett, R. K., 465–473, Balkema, 1993. slabs, Tech. rep., 33rd IABSE Symposium, Sustainable Infras- Dussauge-Peisser, C., Helmstetter, A., Grasso, J.-R., Hantz, D., tructure, Environment Friendly, Safe and Resource Efficient, Desvarreux, P., Jeannin, M., and Giraud, A.: Probabilistic Bangkok, Thailand, 2009. approach to rock fall hazard assessment: potential of histor- Giacomini, A., Buzzi, O., Renard, B., and Giani, G.: Experimen- ical data analysis, Nat. Hazards Earth Syst. Sci., 2, 15–26, tal studies on fragmentation of rock falls on impact with rock doi:10.5194/nhess-2-15-2002, 2002. surfaces, Int. J. Rock Mech. Min., 46, 708–715, 2009. Dussauge-Peisser, C., Grasso, J.-R., and Helmstetter, A.: Statistical Giani, G. P.: Rock Slope Stability Analysis, Taylor & Francis, analysis of rockfall volume distributions: Implications for rock- Balkema, 1992. fall dynamics, J. Geophys. Res. Sol. Ea., 108, 1–11, 2003. Glover, J., Volkwein, A., Dufour, F., Denk, M., and Roth, A.: Rock- Einstein, H. H.: Landslide risk assessment procedure, in: 5th In- fall attenuator and hybrid drape systems – design and testing con- ternational Symposium on Landslides, 2, 1075–1090, Balkema, siderations, in: Third Euro-Mediterranean Symposium on Ad-

www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2646 A. Volkwein et al.: Review on rockfall characterisation and structural protection

vances in Geomaterials and Structures, edited by: Darve, F., Fed´ erale´ de Lausanne, Lausanne, 2004. Doghri, I., El Fatmi, R., Hassis, H., and Zenzri, H., 379–384, Heierli, W.: Viadotto Bosco di Bedrina No2 – Steinschlagschutz: Djerba, 2010. Verhalten von Kies – Sand – Dampfungsschichten,¨ Tech. rep., Gokceoglu, C., Sonmez, H., and Ercanoglu, M.: Discontinuity con- Dipartimento Pubbliche Costruzioni, Ufficio Strade Nazionali, trolled probabilistic slope failure risk maps of the Altindag (set- Bellinzona, 1984. tlement) region in Turkey, Eng. Geol., 55, 277–296, 2000. Herrmann, N.: Experimentelle Erfassung des Betonverhaltens unter Goldsmith, W.: Impact, The theory and physical behaviour of col- Schockwellen, Ph.D. thesis, TH Karlsruhe, 2002. liding solids, Edward Arnold Publishers, Dover, 1960. Hoek, E.: Rockfall: a computer program for prediction rockfall tra- Gottardi, G. and Govoni, L.: Full-scale Modelling of Falling jectories, ISRM News J, 2, 4–16, 1987. Rock Protection Barriers, Rock Mech. Rock Eng., 43, 261–274, Hoek, E.: Strength of rock and rock masses, ISRM News, 2, 4–16, doi:10.1007/s00603-009-0046-0, 2010. 1994. Govi, M.: Photo-interpretation and mapping of the landslides trig- Hoek, E. and Bray, J.: Rock Slope Engineering, E & FN Spon, gered by the Friuli earthquake (1976), Bulletin of the Interna- London, 3rd edn., 1981. tional Association of Eng. Geol., 15, 67–72, 1977. Hopkins, M.: Eiger loses face in massive rockfall, http://www. Grassl, H. G.: Experimentelle und numerische Modellierung des nature.com/news/2006/060717/full/news060717-3.html, 2006. dynamischen Trag- und Verformungsverhaltens von hochflex- Hudson, J. A.: Rock Engineering systems: Theory and Practice, iblen Schutzsystemen gegen Steinschlag, Ph.D. thesis, Eid- Ellis Horwood, Chichester, 1992. genossische¨ Technische Hochschule Zurich,¨ 2002. Hungr, O. and Evans, S.: Engineering evaluation of fragmental Grenon, M. and Hadjigeorgiou, J.: A design methodology for rock rockfall hazards, in: 5th International Symposium on Land- slopes susceptible to wedge failure using fracture system mod- slides, 1, 685–690, Balkema, Rotterdamm, Lausanne, Switzer- elling, Eng. Geol., 96, 78–93, 2008. land, 1988. Gruner, U.: Climatic and meteorological influences on rockfall and Hungr, O., Evans, S., and Hazzard, J.: Magnitude and frequency of rockslides (“Bergsturz”), in: Protection of populated territories rockfalls and rock slides along the main transportation corridors from floods, debris flow, mass movements and avalanches, 26– of south-western British Columbia, Can. Geotech. J., 36, 224– 30 May, 2008, 147–158, 2008. 238, 1999. Guenther, A., Carstensen, A., and Pohl, W.: Automated sliding sus- Hungr, O., Fell, R., Couture, R., and Eberhardt, E.: Landslide Risk ceptibility mapping of rock slopes, Nat. Hazards Earth Syst. Sci., Management, Taylor and Francis, 2005. 4, 95–102, doi:10.5194/nhess-4-95-2004, 2004. Ishikawa, N.: Recent progress on rock-shed studies in Japan, in: Gunther,¨ A.: SLOPEMAP: programs for automated mapping of Joint Japan-Swiss Scientific Seminar on Impact Load by Rock geometrical and kinematical properties of hard rock hill slopes, Falls and Design of Protection Structures, 1–6, Kanazawa, Japan, Comput. Geosci., 29, 865–875, 2003. 1999. Guzzetti, F., Carrara, A., Cardinali, M., and Reichenbach, P.: Land- Jaboyedoff, M. and Derron, M.-H.: Integrated risk assessment pro- slide hazard evaluation: a review of current techniques and their cess for landslides, in: Landslide risk management, edited by application in a multi-scale study, Central Italy, Geomorphology, Hungr, O., Fell, R., Couture, R. R., and Eberhardt, E., 776, Tay- 31, 181–216, 1999. lor and Francis, 2005. Guzzetti, F., Crosta, G., Detti, R., and Agliardi, F.: STONE: a com- Jaboyedoff, M. and Labiouse, V.: Preliminary assessment of rock- puter program for the three dimensional simulation of rock-falls, fall hazard based on GIS data, in: 10th International Congress Comput. Geosci., 28, 1079–1093, 2002. on Rock Mechanics ISRM 2003 – Technology roadmap for rock Guzzetti, F., Reichenbach, P., and Wieczorek, G. F.: Rockfall haz- mechanics, 575–578, Johannesburgh, South Africa, 2003. ard and risk assessment in the Yosemite Valley, California, USA, Jaboyedoff, M., Baillifard, F., Hantz, D., Heidenreich, B., and Nat. Hazards Earth Syst. Sci., 3, 491–503, doi:10.5194/nhess-3- Mazzoccola, D.: Terminologie, in: Prevention´ des mouvements 491-2003, 2003. de versants et des instabilites´ de falaises, edited by Carere, K., Guzzetti, F., Reichenbach, P., and Ghigi, S.: Rockfall hazard and Ratto, S., and Zanolini, F. E., 48–57, 2001. risk assessment along a transportation corridor in the Nera Val- Jaboyedoff, M., Baillifard, F., Philippossian, F., and Rouiller, J.-D.: ley, central Italy, Environ. Manage., 34, 191–208, 2004. Assessing fracture occurrence using ”weighted fracturing den- Habib, P.: Note sur le rebondissement des blocs rocheux, in: Rock- sity”: a step towards estimating rock instability hazard, Nat. Haz- fall dynamics and protective works effectiveness, ISMES publi- ards Earth Syst. Sci., 4, 83–93, doi:10.5194/nhess-4-83-2004, cation no. 90, 123–125, Bergamo, Italy, 1976. 2004. Hamberger, M. and Stelzer, G.: Neue Erkenntnisse aus Tests von Jaboyedoff, M., Baillifard, F., Couture, R., Locat, J., and Locat, P.: dynamischen Seilsperren – Auswirkungen auf die Baupraxis, New insight of geomorphology and landslide prone area detec- Tech. rep., Trumer Schutzbauten, Kuchl, 2007. tion using DEM, in: Landslides: Evaluation and Stabilization, Hearn, G., Barrett, R. K., and McMullen, M. L.: CDOT Flex- edited by Lacerda, W. A., Ehrlich, M., Fontoura, S. A. B., and post Rockfall Fence Development, Testing and Analysis, in: Sayo, A., 191–198, Taylor & Francis, London, 2004b. Soils, Geology, and Foundations – Rockfall prediction and Con- Jaboyedoff, M., Baillifard, F., Derron, M.-H., Couture, R., Locat, trol and landslide case histories (Transportation Research Record J., and Locat, P.: Switzerland modular and evolving rock slope No. 1343), pp. 23–29, Transportation Research Board, National hazard assessment methods, in: Landslide and avalanches, edited Academy Press, Washington, D.C., 1992. by: Senneset, K., Flaate, K. A., and Larsen, J., ICFL, 2005a. Heidenreich, B.: Small and half scale experimental studies of rock- Jaboyedoff, M., Dudt, J. P., and Labiouse, V.: An attempt to refine fall impacts on sandy slopes, Ph.D. thesis, Ecole Polytechnique rockfall hazard zoning based on the kinetic energy, frequency and

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2647

fragmentation degree, Nat. Hazards Earth Syst. Sci., 5, 621–632, Krummenacher, B. and Keusen, H.: Rockfall simulation and haz- doi:10.5194/nhess-5-621-2005, 2005b. ard mapping based on Digital Terrain Modell (DTM), European Jaboyedoff, M., Metzger, R., Oppikofer, T., Couture, R., Derron, Geologist, 12, 33–35, 1996. M.-H., Locat, J., and Turmel, D.: New insight techniques to an- Krummenacher, B., Schwab, S., and Dolf, F.: Assessment of natural alyze rock-slope relief using Dem and 3D-imaging cloud points: hazards by three calculations of rockfall behaviour, in: Interdisci- COLTOP-3D, in: Rock mechanics: Meeting Society’s Chal- plinary Workshop on Rockfall Protection, edited by: Volkwein, lenges and demands. 1st Canada-US Rock Mechanics Sympo- A., Labiouse, V., and Schellenberg, K., 49–51, Swiss Fed. Re- sium, edited by: Eberhardt, E., Stead, D., and Morrison, T., 1, search Inst. WSL, Morschach, Switzerland, 2008. Taylor and Francis, Vancouver, Canada, 2007. Labiouse, V.: Fragmental rockfall paths: comparison of simulations Jacquemoud, J.: Swiss guideline for the design of protection gal- on Alpine sites and experimental investigation of boulder im- leries: background, safety concept and case histories, in: Joint pacts, in: 9th International Symposium on Landslides, 1, 457– Japan-Swiss Scientific Seminar on Impact Load by Rock Falls 466, Balkema, 2004. and Design of Protection Structures, 95–102, Kanazawa, Japan, Labiouse, V., Descoeudres, F., and Montani, S.: Experimental study 1999. of rock sheds impacted by rock blocks, Struct. Eng. Int., 6, 171– Jancke, O., Dorren, L., Berger, F., Fuhr, M., and Kohl,¨ M.: Impli- 176, 1996. cations of coppice stand characteristics on the rockfall protection Labiouse, V. and Heidenreich, B.: Half-scale experimental study of function, Forest Ecol. Manag., 259, 124–131, 2009. rockfall impacts on sandy slopes, Nat. Hazards Earth Syst. Sci., Japan Road Association: Rockfall Handbook, Tokyo, Japan, 1983. 9, 1981–1993, doi:10.5194/nhess-9-1981-2009, 2009. Japan Road Association: Manual for anti-impact structures against Labiouse, V., Heidenreich, B., Desvarreux, P., Viktorovitch, M., falling rocks, Tokyo, Japan, 2000. and Guillemin, P.: Etudes´ trajectographiques, in: Prevention´ des Japanese highway public corporation: Research report on rock mouvements de versants et des instabilites´ de falaises, edited by: falling tests, 1973. Carere, K., Ratto, S., and Zanolini, F., 155–211, Aosta, Italy, Jones, C. L., Higgins, J., and Andrew, R.: Colorado Rockfall 2001. Simulation Program Version 4.0, Tech. rep., Colorado Depart- Lambert, S. and Bourrier, F.: Design of rockfall protection embankment of Transportation, Denver, http://dnr.state.co.us/geostore/ ments: a critical review, Earth Surf. Proc. Land ., in press, 2011. ProductInfo.aspx?productid=MI-66, 2000. Lambert, S. and Nicot, F., eds.: Rockfall engineering, ISBN 978- Jonsson, M.: Energy absorption of trees in a rockfall protection 1-84821-26-5, 464 pages, John Wiley & Sons, ISTE Ltd., New forest, Ph.D. thesis, ETH Zurich,¨ Zurich,¨ 2007. York, London, 2011. Kalberer, M.: Quantifizierung und Optimierung der Lambert, S., Gotteland, P., and Nicot, F.: Experimental study of Schutzwaldleistung gegenuber¨ Steinschlag, Ph.D. thesis, the impact response of geocells as components of rockfall pro- Albert-Ludwigs-Universitat,¨ Freiburg im Breisgau, 2007. tection embankments, Nat. Hazards Earth Syst. Sci., 9, 459–467, Kamijo, A., Onda, S., Masuya, H., and Tanaka, Y.: Fundamen- doi:10.5194/nhess-9-459-2009, 2009. tal test on restitution coefficient and frictional coefficient of rock Lan, H., Martin, D., and Lim, C.: RockFall analyst: A GIS ex- fall, in: 5th Symposium on Impact Problems in Civil Engineer- tension for three-dimensional and spatially distributed rockfall ing, 83–86, 2000. hazard modeling, Comput. Geosci., 33, 262–279, 2007. Kawahara, S. and Muro, T.: Effect of soil slope gradient on mo- Lang, H.-J.: Erdgasleitungen in der Gemeinde Innertkirchen, Tran- tion of rockfall, in: International Symposium on Slope Stability sitgas AG, Zurich, 1974. Engineering, 2, 1343–1348, Matsuyama, Japan, 1999. Lato, M., Diederichs, M. S., Hutchinson, D. J., and Harrap, R.: Kemeny, J., Turner, K., and Norton, B.: LIDAR for Rock Mass Optimization of LIDAR scanning and processing for automated Characterization: Hardware, Software, Accuracy and Best- structural evaluation of discontinuities in rockmasses, Int. J. Practices, in: Workshop on Laser and Photogrammetric Methods Rock Mech. Min., 46, 194–199, 2009. for Rock Mass Characterization: Exploring New Opportunities, Le Hir, C., Dimnet, E., and Berger, F.: Etude´ de la trajectogra- Golden, Colorado, USA, 2006. phie des chutes de blocs en foretsˆ de montagne, Bull. Lab. Ponts Kirkby, M. and Statham, I.: Surface stone movement and scree for- Chaussees,´ 263/264, 85–101, 2006. mation, J. Geol., 83, 349–362, 1975. Lepert, P. and Corte,´ J.: Etude en centrifugeuse de l´ımpact de gros Kishi, N., Nakano, O., Matsuoka, K., and Nishi, H.: Field test on blocs rocheux sur un remblai de protection, in: Centrifuge ’88, absorbing capacity of a sand cushion , J. Struct. Eng., 39A, 1587– 1988. 1597, 1993 (in Japanese). Leroueil, S. and Locat, J.: Slope Movements – Geotechnical Char- Kishi, N., Okada, S., and Konno, N.: Numerical Impact Response acterization, Risk Assessment and Mitigation, in: XI Danube- Analysis of Rockfall Protection Galleries, Struct. Eng. Int., 19, European Conf. on Soil Mechanics and Geotechnical Engineer- 313–320, 2009. ing, edited by: Lisac and Szavits-Nossan, Balkema, Porec, Croa- Kobayashi, Y., Harp, E., and Kagawa, T.: Simulation of Rock- tia, 1998. falls triggered by earthquakes, Rock Mech. Rock Eng., 23, 1–20, Lied, K.: Rockfall problems in Norway, ISMES Publication, 90, 1990. 51–53, 1977. Konno, H., Ishikawa, H., Okada, S., and Kishi, N.: Prototype im- Liniger, M. and Bieri, D.: A2, Gotthardautobahn, Felssturz Gurt- pact test of steel-concrete composite type rock-sheds, in: In- nellen vom 31 Mai 2006, Beurteilung und Massnahmen, in: Pub. terdisciplinary workshop on rockfall protection, edited by Volk- Soc. Suisse Mecanique´ Soles Roches, 153, 81–86, 2006. wein, A., Labiouse, V., and Schellenberg, K., 46–48, Swiss Fed. Lorentz, J., Donze,´ F., Perrotin, P., and Plotto, P.: Experimen- Research Inst. WSL, Morschach, Switzerland, 2008. tal study of the dissipative efficiency of multylayered protec-

www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2648 A. Volkwein et al.: Review on rockfall characterisation and structural protection

tive structure against rockfall impact, Revue europeenne´ de genie´ Matsuoka, N.: Frost weathering and rockwall erosion in the south- civil, 10, 295–308, 2006. eastern Swiss Alps: Long-term (1994-2006) observations, Geo- Lorentz, J., Perrotin, P., and Donze,´ F.: A new sandwich design morphology, 99, 353–368, 2008. structure for protection against rockfalls, in: Interdisciplinary Matsuoka, N. and Sakai, H.: Rockfall activity from an alpine cliff workshop on rockfall protection, edited by Volkwein, A., Labi- during thawing periods, Geomorphology, 28, 309–328, 1999. ouse, V., and Schellenberg, K., Swiss Fed. Research Inst. WSL, Mavrouli, O. and Corominas, J.: Vulnerability of simple rein- Morschach, Switzerland, 2008. forced concrete buildings to damage by rockfalls, Landslides, Loye, A., Jaboyedoff, M., and Pedrazzini, A.: Identification of po- 7, 169–180, http://dx.doi.org/10.1007/s10346-010-0200-5, tential rockfall source areas at a regional scale using a DEM- 10.1007/s10346-010-0200-5, 2010. based geomorphometric analysis, Nat. Hazards Earth Syst. Sci., Mazzoccola, D. and Sciesa, E.: Implementation and comparison of 9, 1643–1653, doi:10.5194/nhess-9-1643-2009, 2009. different methods for rockfall hazard assessment in the Italian Luckman, B. H.: Rockfalls and Rockfall inventory data: some Alps, in: 8th International Symposium on Landslides, 2, 1035– obervations from Surprise Valley, Jasper National Park, Canada, 1040, Balkema, Rotterdam, Cardiff, UK, 2000. Earth Surf. Proc., 1, 287–298, 1976. Mazzoccola, D. F. and Hudson, J. A.: A comprehensive method of Lundstrom,¨ T.: Mechanical stability and growth performance of rock mass characterisation for indicating natural slope instability, trees, Ph.D. thesis, University of Bern, Bern, 2010. Q. J. Eng. Geol. Hydroge., 29, 37–56, 1996. Maegawa, K., Yoshida, H., Fujii, T., Shiomi, M., and Ohmori, Meissl, G.: Modellierung der Reichweite von Felssturzen:¨ Fall- K.: Weight falling tests on the rock-shed composed of CFT- beispiele zur GIS-gestutzten¨ Gefahrenbeurteilung, Ph.D. thesis, members, in: Tubular Structures X, edited by: Jaurrieta, A. C., Institut fur¨ Geographie. Univ. Innsbruck, 1998. Alonso, A., and Alonso, A., 533–540, Swets & Zeitlinger, Lisse, Meissl, G.: Modelling the runout distances of rockfall using a ge- 2003. ographic information system, Z. Geomorphol., 125, 129–137, Maerz, N., Youssef, A., and Fennessey, T. W.: New Risk- 2001. Consequence Rockfall Hazard Rating System for Missouri High- Monnet, J., Mermin, E., Chanussotz, J., and Berger, F.: Tree top ways using Digital Image Analysis, Environ. Eng. Geosci., 11, detection using local maxima filte-ring: a parameter sensitiv- 229–249, 2005. ity analysis, in: Silvila-ser 2010, 10th International Conference Magnier, S.-A. and Donze,´ F.: Numerical simulations of impacts on LiDAR Applications for Assessing Forest Ecosystems, 1–9, using a discrete element method, Mech. Cohes.-Frict. Mat., 3, Freiburg, Germany, 2010. 257–276, 1998. Montani-Stoffel, S.: Sollicitation dynamique de la couverture des Malamud, B., Turcotte, D., Guzzetti, F., and P., R.: Landslide in- galeries de protection lors de chutes de blocs, Ph.D. thesis, Ecole´ ventories and their statistical properties, Earth Surf. Proc. Land., Polytechnique Fed´ erale´ de Lausanne, Lausanne, 1998. 29, 687–711, 2004. Murata, S. and Shibuya, H.: Measurement of impact loads on the Margreth, S.: Snow Pressure Measurements on Snow Net Systems, rockfall prevention walls and speed of falling rocks using a mid- in: Acte de colloque, 241–248, Chamonix, 1995. dle size slope model, in: 2nd Asia-Pacific Conference on Shock Masuya, H.: Design Method of Structures under Impact Action by & Impact Loads on Structures, 383–393, Melbourne, Australia, Concept of Performance Based Design, Japan Society of Civil 1997. Engineers, Committee of Structural Engineering, Subcommittee Mustoe, G. G. W. and Huttelmaier, H.: Dynamic Simulation of a concerning Performance Based Design of Structures against Im- Rockfall Fence by the Discrete Element Method, Microcomputer pact Action, 2007. in Civil Engineering, 8, 423–437, 1993. Masuya, H. and Kajikawa, Y.: Numerical analysis of the collision Nakano, O., Sato, M., Kishi, N., Matsuoka, K., and Nomachi, S.: between a falling rock and a cushion by distinct element method, Full scale impact tests of PC multi-girder with three-layered ab- Computer Methods and Advances in Geomechanics, 493–498, sorbing system, in: 13th International Conference on SMiRT, IV, 1991. 201–206, 1995. Masuya, H. and Nakata, Y.: Development of numerical model com- Nakata, Y., Masuya, H., Kajikawa, Y., and Okada, T.: The Analy- bining distinct element and finite element methods and applica- sis of Impact Behaviour of Rock-Shed by Combination of Dis- tion to rock shed analysis, in: Proc. Japan Soc. Civil Eng., 710-I, tinct Element Method and Finite Element Method, in: 2nd Asia- 113–128, Japan, 2001. Pacific Conference on Shock & Impact Loads on Structures, Masuya, H., Tanaka, Y., Onda, S., and Ihara, T.: Evaluation of 403–410, Melbourne, Australia, 1997. Rock falls on slopes and Simulation of the Motion of Rock Falls Nicot, F.: Etude du comportement mechanique´ des ouvrages sou- in Japan, in: Joint Japan-Swiss Scientific Seminar on Impact ples de protection contre les eboulements´ rocheux, Ph.D. thesis, Load by Rock Falls and Design of Protection Structures, 21–28, Ecole Centrale de Lyon, 1999. Kanazawa, Japan, 1999. Nicot, F., Nouvel, P., Cambou, B., Rochet, L., and Mazzoleni,´ G.: Masuya, H., Ihara, T., Onda, S., and Kamijo, A.: Experimental Etude du comportement mecanique des ouvrages souples de pro- Study on Some Parameters for Simulation of Rock Fall on Slope, tection contre les eboulements rocheux, Revue française de genie´ in: Fourth Asia-Pacific Conf. on Shock and Impact Loads on civil, 3, 295–319, 1999. Structures, 63–69, Japan, 2001. Nicot, F., Cambou, B., and Mazzoleni,´ G.: Design of Rockfall Re- Masuya, H., Amanuma, K., Nishikawa, Y., and Tsuji, T.: Basic straining Nets from a Discrete Element Modelling, Rock Mech. rockfall simulation with consideration of vegetation and appli- Rock Eng., 34, 99–118, 2001. cation to protection measure, Nat. Hazards Earth Syst. Sci., 9, Nicot, F., Gay, M., Boutillier, B., and Darve, F.: Modelling of In- 1835–1843, doi:10.5194/nhess-9-1835-2009, 2009. teraction between a Snow Mantel and a Flexible Structure using

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2649

a Discrete Element Method, in: Proc. Num. Models in Geome- enbamkments – a numerical discrete approach, in: 9th Inter- chanics NUMOG VIII, edited by: Pande, G. N. and Pietruszcak, national Congress on Numerical Models in Geomechanics NU- S., 699–703, Swets & Zeitlinger, 2002a. MOG, 609–614, Ottawa, Canada, 2004. Nicot, F., Gay, M., and Tacnet, J.: Interaction between a Snow Man- Plassiard, J.-P. and Donze,´ F.-V.: Rockfall Impact Parameters on tel and a Flexible Structure: A new Method to Design Avalanche Embankments: A Discrete Element Method Analysis, Struct. Nets, Cold Reg. Sci. Technol., 34, 67–84, 2002b. Eng. Int., 19, 333–341, 2009. Nicot, F., Gotteland, P., Bertrand, D., and Lambert, S.: Multi-scale Raetzo, H., Lateltin, O., Bollinger, D., and Tripet, J.: Hazard as- approach to geo-composite cellular structures subjected to im- sessment in Switzerland – Code of practice for mass movements, pact, International Journal for Numerical and Analytical Meth- B. Eng. Geol. Environ., 61, 263–268, 2002. ods in Geomechanics, 31, 1477–1515, 2007. Rammer, W., Brauner, M., Dorren, L., Berger, F., and Lexer, M.: Norrish, N. and Wyllie, D.: Landslides – Investigation and mitiga- Validation of an integrated 3D forest – rockfall model, Geophys. tion, in: Rock slope stability analysis, edited by: Turner, A. and Res. Abstr., Vol. 9, 04634, Vienna, 2007. Schuster, R. L., Transportation Research Board, special report Rammer, W., Brauner, M., Dorren, L. K. A., Berger, F., and 247, 673, National Academy Press, Washington, DC, 1996. Lexer, M. J.: Evaluation of a 3-D rockfall module within a for- Oppikofer, T., Jaboyedoff, M., and Coe, J. A.: Rockfall hazard at est patch model, Nat. Hazards Earth Syst. Sci., 10, 699–711, Little Mill Campground, Uinta National 150 Forest: Part 2. DEM doi:10.5194/nhess-10-699-2010, 2010. analysis, in: 1st North American Landslide Conference, edited Rapp, A.: Recent development of mountain slopes in Karkevagge by: Schaefer, V. R., Schuster, R. L., and Turner, A. K., AEG and surroundings, northern Scandinavia, Geogr. Ann., 42, 65– Special Publication no. 23, 1351–1361, Vail, Colorado, USA, 200, 1960. 2007. Ritchie, A.: Evaluation of rockfall and its control, Highway re- Oppikofer, T., Jaboyedoff, M., and Keusen, H.-R.: Collapse at the search record, 17, 13–28, 1963. eastern Eiger flank in the Swiss Alps, Nat. Geosci., 1, 531–535, Rochet, L.: Development of numerical models for the analysis of 2008. propagation of rock-falls, 6th Int. Congress on Rock Mech, 1, Paronuzzi, P.: Probabilistic approach for design optimization of 479–484, 1987a. rockfall protective barriers, Q. J. Eng. Geol., 22, 175–183, 1989. Rochet, L.: Application des modeles` numeriques´ de propagation a` Peila, D. and Ronco, C.: Technical Note: Design of rockfall net l’etude´ des eboulements´ rocheux, Bulletin de liaison des labora- fences and the new ETAG 027 European guideline, Nat. Hazards toires des ponts et chaussees,´ 150–151, 84–95, 1987b. Earth Syst. Sci., 9, 1291–1298, doi:10.5194/nhess-9-1291-2009, Romana, M.: Practice of SMR classification for slope appraisal, in: 2009. 5th International Symposium on Landslides, Balkema, Rotter- Peila, D., Pelizza, S., and Sassudelli, F.: Evaluation of Behaviour of damm, Lausanne, Switzerland, 1988. Rockfall Restraining Nets by Full Scale Tests, Rock Mech. Rock Romana, M.: A geomechanical classification for slopes: slope mass Eng., 31, 1–24, 1998. rating, Comprehensive Rock Engineering, Pergamon, Oxford, Peila, D., Castiglia, C., Oggeri, C., Guasti, G., Recalcati, P., and 1993. Rimoldi, P.: Testing and modelling geogrid reinforced soil em- Rouiller, J.-D. and Marro, C.: Application de la methodologie´ bankments subject to high energy rock impacts, in: 7th Interna- MATTEROCK a` levaluation´ du danger lie´ aux falaises, Eclogae tional conference on geosynthetics, 2002. Geologicae Helvatiae, 90, 393–399, 1997. Peila, D., Oggeri, C., and Castiglia, C.: Ground reinforced embank- Rouiller, J. D., Jaboyedoff, M., Marro, C., Phlippossian, F., and ments for rockfall protection, design and evaluation of full scale Mamin, M.: Pentes instables dans le Pennique valaisan. Matte- tests, Landslides, 4, 255–265, 2007. rock: une methodologie´ dauscultation´ des falaises et de detection´ Perret, S., Stoffel, M., and Kienholz, H.: Spatial and temporal rock des eboulements´ majeurs potentiels, Rapport final du PNR31, fall activity in a forest stand in the Swiss Prealps – a dendrogeo- vdf Hochschulverlag AG an der ETH Zurich,¨ Zurich,¨ Switzer- morphological case study, Geomorphology, 74, 219–231, 2006. land, 1998. Pfeiffer, T. and Bowen, T.: Computer Simulation of Rockfalls, Bul- Santi, M. P., Russel, C. P., Higgins, J. D., and Spriet, J. I.: Modi- letin of the Association of Engineering Geologists, 26, 135–146, fication and statistical analysis of the Colorado Rockfall Hazard 1989. Rating System, Eng. Geol., 104, 55–65, 2008. Pichler, B., Hellmich, C., and Mang, H.: Impact of rocks onto Sasiharan, N., Muhunthan, B., Badger, T., Shu, S., and Car- gravel – Design and evaluation experiments, Int. J. Impact Eng., radine, D.: Numerical analysis of the performance of 31, 559–578, 2005. wire mesh and cable net rockfall protection systems, Eng. Pierson, L. A., Davis, S. A., and Van Vickle, R.: Rockfall Hazard Geol., 88, 121–132, doi:10.1016/j.enggeo.2006.09.005, http: Rating System Implementation Manual, Oregon, 1990. //www.sciencedirect.com/science/article/B6V63-4M69JSG-1/2/ Piteau, D. and Clayton, R.: Computer Rockfall Model, in: Meet- 74d8147926832f9eb71fccc4859396f4, 2006. ing on Rockfall Dynamics and Protective Works Effectiveness, Sato, M., Kishi, N., Iwabuchi, T., Tanimoto, T., and Shimada, T.: no. 90 in ISMES Publication, 123–125, Bergamo, Italy, 1976. Shock Absorbing Performance of Sand Cushion, in: 1st Asia- Piteau, D. and Clayton, R.: Discussion of paper Computerized¨ de- Pacific Conference on Shock and Impact Loads on Structures, sign of rock slopes using interactive graphics for the input and 393–400, Singapore, 1996. output of geometrical databy¨ Cundall, P., in: Proceedings of the Schellenberg, K.: On the design of rockfall protection galleries, 16th Symposium on Rock Mechanics, Minneapolis, USA, 62– No. 17924, ETHZ, Institute of Structural Engineering, Zurich, 63, 1977 2008. Plassiard, J., Donze,´ F., and Plotto, P.: High energy impact on Schellenberg, K.: On the design of rockfall protection galleries –

www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 2650 A. Volkwein et al.: Review on rockfall characterisation and structural protection

An analytical approach for a performance based design, SVH Statham, I.: A simple dynamic model of rockfall: some theoreti- Verlag, 2009. cal principles and model and field experiments, in: ISMES: In- Schellenberg, K. and Vogel, T.: Swiss Rockfall Galleries – Impact ternational Colloquium on Physical and Geomechanical Models, Load, in: Structures and Extreme Events, 302–303 and CD– 237–258, Bergamo, Italy, 1979. ROM file LIS099.PDF, 1–8, IABSE Symposium Lisbon 2005, Statham, I. and Francis, S.: Hillslope processes, in: Influence of IABSE, Zurich, 2005. scree accumulation and weathering on the development of steep Schellenberg, K. and Vogel, T.: A Dynamic Design Method for mountain slopes, edited by: Abrahams, A., Allen and Unwin, Rockfall Protection Galleries, Struct. Eng. Int., 19(3), 321–326, Winchester, 1986. 2009. Stevens, W. D.: Rocfall: A tool for probablistic analysis, design Schellenberg, K., Volkwein, A., Roth, A., and Vogel, T.: Large- of remedial measures and prediction of rockfalls, Master’s the- scale impact tests on rockfall galleries, in: 7th Int. Conference sis, University of Toronto, http://www.rocscience.com/library/ on Shock & Impact Loads on Structures, 17–19 October 2007, pdf/rf 3.pdf, 1998. 497–504, Bejing, 2007. Strahler, A. N.: Quantitative geomorphology of erosional land- Schellenberg, K., Volkwein, A., Denk, M., and Vogel, T.: Falling scapes, in: Compt. Rend. 19th Intern. Geol. Cong., 13, 341–354, weight tests on rock fall protection galleries with cushion layers, 1954. in: Interdisciplinary Workshop on Rockfall Protection, edited by: Straub, D. and Schubert, M.: Modelling and managing uncertainties Volkwein, A., Labiouse, V., and Schellenberg, K., Swiss Fed. in rock-fall hazards, Georisk, 2, 1–15, 2008. Research Inst. WSL, Morschach, Switzerland, 2008. Stronge, W. J.: Impacts mechanics, Cambridge University Press, Schneuwly, D. M. and Stoffel, M.: Changes in spatio-temporal pat- Cambridge, 2000. terns of rockfall activity on a forested slope – a case study using Studer, C.: Simulation eines Bremsrings im Steinschlagschutzsys- dendrogeomorphology, Geomorphology, 102, 522–531, 2008. tem, Master’s thesis, Eidgenossische¨ Technische Hochschule Scioldo, G.: Slope instability recognition, analysis, and zonation, Zurich,¨ diplomarbeit am Institut fur¨ Baustatik und Konstruktion, in: Rotomap: analisi statistica del rotolamento dei massi, 81–84, ETH Zurich,¨ 2001. Milano, 1991. Sturzenegger, M., Stead, D., Froese, C., Moreno, F., and Jaboyed- Scioldo, G.: User guide ISOMAP & ROTOMAP – 3D surface off, M.: Mapping the geological structure of Turtle Mountain, modelling and rockfall analysis, http://www.geoandsoft.com/ Alberta: A critical interpretation of field, Dem and LIDAR manuali/english/rotomap.pdf, 2006. based techniques, in: Rock mechanics: Meeting Society’s Chal- Selby, M. J.: A rock mass strength classification for geomorphic lenges and demand. 1st Canada-US Rock Mechanics Sympo- purposes: with tests from Antartica and New Zealand, Z. Geo- sium, edited by: Eberhardt, E., Stead, D., and Morrison, T., 2, morphologie, 24, 31–51, 1980. Taylor & Francis Ltd, Vancouver, Canada, 2007a. Selby, M. J.: Controls on the stability and inclinations of hillslopes Sturzenegger, M., Yan, M., Stead, D., and Elmo, D.: Applica- formed on hard rock, Earth Surf. Proc. Land., 7, 449–467, 1982. tion and limitations of ground-based laser scanning in rock slope Shu, S., Muhunthan, B., Badger, T. C., and Grandorff, characterisation, in: Rock mechanics: Meeting Society’s chal- R.: Load testing of anchors for wire mesh and cable lenges and demands. 1st Canada-US Rock Mechanics Sympo- net rockfall slope protection systems, Eng. Geol., 79, sium, edited by: Eberhardt, E., Stead, D., and Morrison, T., 1, 162–176, doi:10.1016/j.enggeo.2005.01.008, http://www. 29–36, Taylor & Francis, London, Vancouver, Canada, 2007b. sciencedirect.com/science/article/B6V63-4FWKDWV-1/2/ Tajima, T., Maegawa, K., Iwasaki, M., Shinohara, K., and 24d31d12a77ea868edf8184abb781f6d, 2005. Kawakami, K.: Evaluation of Pocket-type Rock Net by Full SIA261: Einwirkungen auf Tragwerke, Tech. rep., Schweizerische Scale Tests, in: IABSE Symposium Bangkok 2009: Sustain- Ingenieure und Architekten, Bern, 2003. able Infrastructure. Environment Friendly Safe and Resource Ef- Soeters, R. and Van Westen, C.: Slope instability recognition, anal- ficient, 96, IABSE reports, International Association for Bridge ysis, and zonation, in: Landslides – Investigation and Mitiga- and Structural Engineering, 2003. tion – Special Report 247, edited by Turner, A. and Schuster, R., Teraoka, M., Iguchi, H., Ichikawa, T., Nishigaki, Y., and Sakurai, S.: 129–177, Trans. Res. B., National Research Council, National Analysis of motion for rock falling on a natural slope by using Academy Press, Washington, D.C., USA, 1996. digital video image, in: 5th Symposium on Impact Problems in Sonoda, Y.: A study on the simple estimation method of impact Civil Engineering, 87–90, Japan, 2000. load by the one dimensional stress wave analysis, in: Joint Japan- Thommen, R. A.: Testing of various types of rockfall flexible wire Swiss Scientific Seminar on Impact Load by Rock Falls and De- rope mitigation barrier: an overview of testing to date, in: 59th sign of Protection Structures, 43–50, Kanazawa, Japan, 1999. Highway Geology Symposium, Santa Fe, 2008. Spang, R. and Bolliger, R.: Vom Holzzaun zum Hochenergienetz – Tonello, J.: Gen´ eralit´ es´ et approche de modeles` simples, in: Stage die Entwicklung des Steinschlagschutzes von den Anfangen¨ bis paravalanches, E.N.P.C., Paris, 1988. zur Gegenwart, Geobrugg Schutzsysteme, Romanshorn, 2001. Tonello, J.: Couverture pare-blocs structurellement dissipante, Spang, R. and Rautenstrauch, R.: Empirical and mathematical ap- Tech. rep., METL/DRAST, Label IVOR 01.1. Mission proaches to rockfall protection and their practical applications, Genie´ Civil, http://www.equipement.gouv.fr/recherche/incitatif/ in: 5th International Symposium on Landslides, 1237–1243, ivor, 2001. Balkema, Rotterdamm, Lausanne, Switzerland, 1988. Toppe, R.: Terrain models – A tool for natural hazard mapping, Spang, R. and Sonser,¨ T.: Optimized rockfall protection by “Rock- IAHS, Publication, 162, 1987a. fall”, in: 8th Int. Congr. Rock Mech., 3, 1233–1242, Tokyo, Toppe, R.: Avalanche formation, movement and effects, chapitre 1995. Terrain models – a tool for natural hazard mapping, IAHS Publi-

Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011 www.nat-hazards-earth-syst-sci.net/11/2617/2011/ A. Volkwein et al.: Review on rockfall characterisation and structural protection 2651

cation, 162, 629–638, 1987b. Wentworth, C. M., Ellen, S. D., and Mark, S. D.: Improved analy- Turner, R., Duffy, J. D., and Turner, J. P.: Post Foundations for sis of regional engineering geology using GIS, in: GIS’87, San Flexible Rockfall Fences, in: Proc. 60th Highway Geology Sym- Francisco, California, 1987. posion, 2009. Wieczorek, G. F., Morrissey, M. M., Iovine, G., and Godt, J.: Rock- Ujihira, M., Takagai, N., and Iwasa, T.: An experimental study on fall potential in the Yosemite Valley, California, Tech. rep., U.S. the characteristics of the impact load of falling rock, International Geological Survey, 1999. Journal of Surface Mining and Reclamation, 7, 81–89, 1993. Wienberg, N., Weber, H., and Toniolo, M.: Testing of flexible Urciuoli, G.: Sperimentazione sulla caduta di blocchi lungo un barriers – behind the guideline, in: Interdisciplinary workshop pendio nella formazione calcareo-dolomitica della Penisola Sor- on rockfall protection, edited by Volkwein, A., Labiouse, V., rentina: Convengo sul tema, in: Convengo sul tema: Cartografia and Schellenberg, K., 114–116, Swiss Fed. Research Inst. WSL, e monitoraggio dei movimenti franosi, 35–54, Bologna, Italy, Morschach, Switzerland, 2008. 1988. Woltjer, M., Rammer, W., Brauner, M., Seidl, R., Mohren, G., Urciuoli, G.: Giornata di Studio su La protezione contro la caduta and Lexer, M.: Coupling a 3D patch model and a rockfall mod- massi dai versanti rocciosi, 29–36, Torino, Italy, 1996. ule to assess rockfall protection in mountain forests, J. Environ. Ushiro, T., Shinohara, S., Tanida, K., and Yagi, N.: A study on Manag., 87, 373–388, 2008. the motion of rockfalls on Slopes, in: 5th Symposium on Impact Wong, R., Ho, K., and Chau, K. T.: Shape and mechanical proper- Problems in Civil Engineering, 91–96, Japan, 2000. ties of slope material effects on the coefficient of restitution on Van Dijke, J. and van Westen, C.: Rockfall hazard: a geomorpho- rockfall study, in: 4th North American Rock Mechanics Sympo- logical application of neighbourhood analysis with ILWIS, ITC sium, 507–514, Seattle, Washington, USA, 2000. Journal, 1, 40–44, 1990. Wong, R. H., Ho, K., and Chau, K. T.: Experimental study for rock- Van Westen, C.: Geo-information tools for landslide risk assess- fall simulation, in: Construction challenges into the next century, ment: an overview of recent developments, in: 9th International 92–97, Hong-Kong, China, 1999. Symposium on Landslides, Balkema, 2004. Wu, S.: Rockfall evaluation by computer simulation, Transportation Vangeon, J.-M., Hantz, D., and Dussauge, C.: Rockfall predictibil- Research Record, 1031, 1–5, 1985. ity: a probabilistic approach combining historical and geome- Wu, T. H., Wilson, H. T., and Einstein, H. H.: Landslides – Investi- chanical studies, Revue Française de Geotechnique,´ 95/96, 143– gation and mitigation, 1996. 154, 2001. Wyllie, D. C. and Mah, C. W.: Rock slope engineering: Civil and Varnes, D. J.: IAEG Commission on Landslides & other Mass Mining, Spon Press, 4 edn., 2004. Movements, in: Landslide hazard zonation: a review of prin- Yang, M., Fukawa, T., Ohnishi, Y., Nishiyama, S., Miki, S., Hi- ciples and practice, 63, UNESCO Press, Paris, 1984. rakawa, Y., and Mori, S.: The application of 3-dimensional DDA Vogel, T., Labiouse, V., and Masuya, H.: Rockfall Protection as an with a spherical rigid block for rockfall simulation, Int. J. Rock Integral Task, Struct. Eng. Int., 19(3), 321–326, 2009. Mech. Min., 41, 1–6, 2004. Volkwein, A.: Numerische Simulation von flexiblen Stein- Yoshida, H.: Movement of boulders on slope and its simulation, schlagschutzsystemen, Ph.D. thesis, Eidgenossische¨ Technische Recent studies on rockfall control in Japan, Tech. rep., 1998. Hochschule Zurich,¨ 2004. Yoshida, H.: Recent experimental studies on rockfall control in Volkwein, A., Roth, A., Gerber, W., and Vogel, A.: Flexible rockfall Japan, in: Joint Japan-Swiss Scientific Seminar on Impact Load barriers subjected to extreme loads, Struct. Eng. Int., 19, 327– by Rock Falls and Design of Protection Structures, Kanazawa, 331, 2009. Japan, 1999. Voyat, I., Roncella, R., Forlani, G., and Ferrero, A. M.: Advanced Yoshida, H., Masuya, H., and Ihara, T.: Experimental Study of Im- techniques for geo structural surveys in modelling fractured rock pulsive Design Load for Rock Sheds, Struct. Eng. Int., P-127/88, masses: application to two Alpine sites, in: Symposium on Rock 61–74, 1988. Mechanics (USRMS), American Rock Mechanics Association Zaitsev, A., Sokovikh, M., and Gugushvily, T.: Field testing of net ARMA, 2006. structures for railway track protection in rocky regions, in: 7th Wagner, A., Leite, E., and Olivier, R.: Rock and debris-slides risk International Conference on Physical Modelling in Geotechnics mapping in Nepal – A user-friendly PC system for risk mapping, (ICPMG 2010), Zurich, 2010. in: 5th International Symposium on Landslides, edited by Bon- Zinggeler, A., Krummenacher, B., and Kienholz, H.: Stein- nard, C., 2, 1251–1258, A. A. Balkema, Rotterdam, Lausanne, schlagsimulation in Gebirgswaldern,¨ Berichte und Forschungen Switzerland, 1988. der Geographisches Institut der Universitat¨ Freiburg, 3, 61–70, Wendeler, C.: Murgangruckhalt¨ in Wildbachen.¨ Grundlagen zu Pla- 1990. nung und Berechnung von flexiblen Barrieren, Ph.D. thesis, In- stitute of Structural Engineering, ETH Zurich, diss ETH No. 17916, 2008.

www.nat-hazards-earth-syst-sci.net/11/2617/2011/ Nat. Hazards Earth Syst. Sci., 11, 2617–2651, 2011