You are on page 1of 22

Bulletin of Engineering Geology and the Environment (2021) 80:5347–5368

https://doi.org/10.1007/s10064-021-02288-1

ORIGINAL PAPER

Impact of rock block characteristics on rockfall hazard and its


implications for rockfall protection strategies along Himalayan
highways: a case study
Ashok K. Singh1,2 · J. Kundu2 · K. Sarkar2   · H. K. Verma1 · P. K. Singh3

Received: 30 July 2020 / Accepted: 10 May 2021 / Published online: 17 May 2021
© Springer-Verlag GmbH Germany, part of Springer Nature 2021

Abstract
Rock block characteristics are the controlling factors that define the potential rockfall volume, associated hazard and design
of preventive practices. The present study defines a methodology to demonstrate the impact of block characteristics on rock-
fall hazard and its implications on the design of mitigative measures by analysing a recurring rockfall site along a critical
highway corridor in NW Himalaya, India. The field data were quantitatively characterised for statistical distributions of size
and volume of in situ as well as rockfall blocks. Possible block failures and rockfall initiation zones were predicted using
modified kinematic analysis and distinct element method. The size and volume distributions as well as shapes were simulated
against these predicted initiation zones to determine the cumulative impact of block characteristics on rockfall hazard by
analysing the block kinetics and trajectories. It is found that the block initiation zones majorly govern the relative bounce
height, kinetic energy, run-out distance and impact down the slope respective of block sizes and associated mass coupled
with slope geometry. However, the effect of block shapes is observed to be minor compared to the blocks greater than a
particular size range. Therefore, the design of rockfall protection such as barrier can be more effective when considering the
kinetic energies of such block size range assisted by initiation zones. This study proposed that the knowledge of block size
frequencies and rockfall initiation zones could provide a rational threshold for which a protection strategy can be designed
and optimised to prevent rockfall hazard along the Himalayan highway corridors.

Keywords  Rockfall · Block size distribution · Kinematic analysis · Rockfall barrier · Himalaya

Introduction parts of the world and accounts for substantial economic


loss and deaths (Turner and Jayaprakash 2012; Collins and
Rockfalls, induced by kinematically detachable blocks from Stock 2016; Froude and Petley 2018; Wang et al. 2020).
steep slopes, are rapid to an extremely rapid form of bounc- The unfavourable arrangement of discontinuities with the
ing, sliding, rolling and free fall, always threatening nearby slope orientation is a common cause of rockfalls, although
lives and properties. The associated impact energy of the irregular fractures may significantly affect. At the same time,
rapidly moving rock blocks can be high enough to cause cau- it may get triggered by undercutting of the slope, differential
salities even when the block is very small in size and volume weathering, intense rainfall, earthquake, dynamic loading,
(Shen et al. 2017). Rockfall is a common occurrence in many anthropogenic activities and biological factors (Highland
and Bobrowsky 2008; Wyllie 2014; Ansari et  al. 2014;
Bostjančić and Pollak 2020). The effect of wind also cannot
* K. Sarkar be discounted (Pei et al. 2019). In a mountainous region,
kripamoy@iitism.ac.in
the highway corridors are highly susceptible to rockfall
1
Bilaspur Research Center, CSIR-Central Institute of Mining and further aggravated by engineering activities (Budetta
and Fuel Research, Chhattisgarh 495001, India 2004; Ferlisi et al. 2012; Palma et al. 2012; Ansari et al.
2
Department of Applied Geology, IIT(ISM) Dhanbad, 2016; Geniş et al. 2017) especially in the fragile Himala-
Dhanbad  826004, India yan region (Chakraborty and Anbalagan 2008; Singh et al.
3
CSIR-Central Institute of Mining and Fuel Research, (2015a); Gupta and Tandon 2015; Singh et al. 2016; Vishal
Dhanbad 826015, India et al. 2017). National Highway (NH) - 05 is such a major

13
Vol.:(0123456789)
5348 A. K. Singh et al.

highway and the only linking route in the tectonically active investigation followed by determination through conven-
and fragile Himalayas between the hilly regions and India’s tional and numerical methods. The workflow of the study is
plain area. The highway corridors in the Himalayan region presented in Fig. 1, and the methods adopted to achieve the
have colossal significance in accelerating and maintaining objective of the study are concisely described in the method-
the socio-economic development and livelihood improve- ology section. The current research signifies the cumulative
ment of the residents (Hearn and Shakya 2017). According effects of rock block characteristics such as size, shape and
to Barnard et al. (2001), two-thirds of the landslide activities initiation zones on rockfall hazard parameters and discusses
along highway cuts in lower Himalaya are contributed by the its influence on the design of rockfall protection strategies
housing- and commercial development-related constructions through constraining rockfall hazard along the highway.
adjacent to highways. These human activities pose a mainte-
nance challenge to the concerned authorities as a majority of
the population settle along the road network or in the vicinity Study area
for better convenience. Singh et al. (2014) also mentioned
the adverse effect of steep excavation of road-cut slopes, Socio‑economic significance
inadequate geotechnical investigation and overblasting dur-
ing road widening as the foremost reasons behind small and Rampur-Jhakri region of Himachal Himalaya is highly vul-
large-scale slope failures. nerable to rockfalls and rock failures around habitats and
The size and volume distributions of blocks in a jointed highway cuts throughout the year (Singh 2018). Being a
rock slope derived by spatial discontinuity arrangements are famous tourist spot (especially for Lavi fare) and a central
the controlling parameters in rockfall and related protec- zone for trade to the neighbouring area, the exposed excavated
tion management to reduce the impact of rockfall hazard slopes along the highway corridor in this region poses a threat
(Ruiz-Carulla et al. 2015; Mavrouli et al. 2015; Ferrero to transportations. Potentially detachable rock blocks from
et al. 2016; Mavrouli and Corominas 2017; Corominas et al. the cut slopes disrupt the traffic and impose a considerable
2018; Luo et al. 2019). Laimer (2020) stated that the block risk to the people. Topographically, the study area is highly
size and volume distribution better assist the engineers in rugged with steep slopes and is drained by River Satluj and
dealing with the uncertainties associated with rockfall haz- its tributaries, the third-largest trans-Himalayan river system.
ard and applying the preventive measures effectively. Impact The area has been experiencing some large-scale hydroelec-
positions and associated energies, relative bounce height tric and urbanisation projects in recent years due to the rich-
and runout path of the failed blocks are crucial parameters est hydropower potentiality of the Satluj River valley. Some
that the block characteristics can efficiently predict. Recog-
nition of block detachment zones from the rock wall face
and determination of block dynamics throughout the rock-
fall path is essential to install protection measures at suit-
able places relative to the slope geometry (Pérez-Rey et al.
2019). The number of joint sets and their mutual geometri-
cal arrangements, individual spacing, persistence and rock
mass strength are the factors that majorly govern the size
and volume distribution of in situ blocks from the rockfall
source area (Palmstrom 2005; Elmouttie and Poropat 2012;
Mavrouli and Corominas 2020). The maximum rockfall size
from a slope is constrained by the inherent fracture and local
structural configurations (Corominas et al. 2018).
The present study’s objective is to investigate the impact
of rock block characteristics and initiation zones on rock-
fall hazard at a recurrent rockfall site along a crucial seg-
ment of NH - 05 near Rampur, Shimla of the state Himachal
Pradesh in the Indian Himalayas. Since the Himalayan
highway corridors are frequently threatened by rockfalls
adversely affecting people and infrastructures, protection
strategies have also been suggested based on the observed
rockfall hazard and rock block characteristics. Kinemati-
cally detachable rock blocks and their trajectories along
the slope profile have been predicted from detailed field Fig. 1  The workflow of the present study

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5349

of these projects include the largest underground hydroelec- has a stretch of 30 –38 m with 27–31 m height. Four sets of
tric project of Asia known as Jhakri Hydroelectric project joints (J1, J2, J3 and J4) are prominent at the site (Fig. 2e,
of 1500 MW capacity and Rampur hydroelectric project of f and Table 1), making the slope highly vulnerable to dif-
412 MW capacity, functioning and managed by Satluj Jal ferent types of discontinuity-driven rock failures. Based on
Vidyut Nigam (SJVN) Ltd. Another relatively small project the slope angle variation, section A is divided into two seg-
known as the Nogli hydel power project of 2500 kW is also ments, the upper steep scar segment with a mean inclina-
operating in this area. Therefore, a safe highway access is tion of 83° (highly susceptible to kinematic failures) and the
essential to keep the current socio-economic improvement lower segment with a mean inclination of 47° (Fig. 2d). Sec-
drive in this part of the Himalayas. tion A itself has an arcuate shape with the highway whose
dip direction varies from N220° ± 5° to N265°. The Bor-
Previous slope failure investigations der Road Organization (BRO) took a preventative measure
along the entire exposed slope by constructing gabion walls
Heavily jointed rock masses are the characteristics of high- to arrest the falling blocks. However, the gabion wall at sec-
way slopes in the Himalayan region, so a complex com- tion A has failed and bulged at several locations in a recent
bination of rock failures, including rockfalls, is frequently rockfall event (Fig. 2d).
reported. Therefore, major highways such as NH-05 require Geologically, the study area belongs to Rampur series
site-specific investigations with an integrated approach to of Paleoproterozoic and forms a part of the Kullu-Rampur
quantify and reduce rock failure–induced hazards. Research- Window Group (KRWG) in the Lesser Himalayas between
ers have applied different approaches to understand the fail- two major thrust zones of the Himalayan system: Main
ure characteristics and assess slope instabilities along the Central Thrust (MCT) and Main Boundary Thrust (MBT)
highway slopes of these parts of the Himalayas. Methods (Srikantia and Bhargava 1998). Shali Formation underlines
like kinematic analysis and rock mass characterisation Rampur series and the Wangtu Gneiss Complex thrusts onto
approaches have been adopted by Acharya et al. (2020), the quartzite of Rampur Formation along an active thrust
Singh (2018), Sarkar et al. (2016), Kundu et al. (2017), named Jakhri Thrust (Fig. 2e) which is younger than MBT
Mahanta et al. (2016), and Singh et al. (2015b) to assess the (4.5 Ma) (Jain et al. 2000; Pandey et al. 2004). The Ram-
rock mass health in this area. Some improved classification pur series are composed of quartz-mica schist, quartzite and
approaches such as rock mass rating (RMR) through con- volcanic/metavolcanic rocks. Quartzite is highly deformed
tinuous functions and modified slope mass rating (SMR) and forms the core of Rampur anticline where strata strike
method have also been developed and implemented by in NNE-SSW direction and dipping towards East and West
Kundu et al. (2020) and Kundu et al. (2019), respectively. directions (Pandey et al. 2004).
Some researchers have also performed finite element method
(FEM) and discrete element method (DEM) to investigate
the failure mechanism under static as well as dynamic load- Methodologies
ing condition in this region (Singh 2018; Kundu et al. 2016b;
Mahanta et al. 2016; Sarkar and Singh 2008). Gupta et al. Rock block characterisation
(1993) and Kumar et al. (2018) have appraised the regional
hazard zonation in Kinnaur and Shimla areas to categorise Preliminary investigation of a site and joint characterisation
the slopes based on prevailing geological field attributes. suggests qualitative as well as quantitative description of the
rock masses and associated in situ rock blocks (ISRM 1978).
Details of the rockfall site and geological setup Block characterisation also offers an idea about the strength
and mechanical behaviour of rock masses that largely
The studied slope section is situated along NH-5 (at the left depend upon the joint parameters. The said parameters are
bank of river Satluj) near Rampur Bushahr city of Shimla, often controlled by the orientation relationship of joints and
Himachal Pradesh, India. The site is located at a sharp turn blocks relative to the examined rock structures (Smith 2006).
along NH-5 (N31° 27.935′ and E77° 40.532′), making the According to Zhang (2016), the joint sets and their intensity,
slope curved in profile changing dip direction from N10° to spacing, persistence and mutual orientation are the govern-
N265° (Fig. 2a). The investigated highway section comprises ing parameters to determine the potential in situ blocks along
heavily jointed quartzite rock which is slightly weathered in with their dimensions and volume. Palmstrom (2005) also
nature (Fig. 2b). urged that joint spacing and joint volumetric count are sig-
Based on the curvature and rockfall severity, the high- nificant empirical indices to predict the quantile statistics
way section was divided into two sections, sections A and B of in-situ block characteristics. Therefore, to quantitatively
(Fig. 2c, d). The stretch of the slope section A is about 80 m characterise the rock blocks, detailed joint characterisation
with a height varying from 25 to 29 m whereas, section B is essential.

13
5350 A. K. Singh et al.

Fig. 2  a Google Earth imagery


of the highway sections (section
A and section B). b Unfavour-
able orientations of joints
feasible for kinematic blocks
detachment from section A . c
Vulnerable field condition of
section B (notice the location
of gabions with respect to the
slope profile). d Vulnerable
field condition and rockfall
event with fragmented blocks
at section A that damaged the
existing gabion. e Visualisation
of multi-oriented joints. f Lower
hemisphere-equal angle ste-
reographic projection of major
joints. g Geological setting of
the Rampur area (after Pandey
et al. 2004)

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5351

Table 1  Qualitative descriptions Joint properties Range Description


and statistical distribution of
joints parameters Mean orientation J1 24/14 Variability (°) 4/8
(Dip°/Dip direc- J2 38/227 7/5
tion°)
J3 70/107 12/10
J4 84/311 5/6
Persistence (m) 3–20 Low to high persistence
Spacing (cm) Max 83.5 Varying from close to moderate spacing for major joints
Min 2.6
Mean 25.2
Aperture (mm) Max 16 Varying from close to open aperture with no fill to siliceous/
Min 0 soil filling
Mean 3

Joints characterisation (
JRCn = JRC0 Ln ∕L0
)−0.02JRC0
(4)

The joint parameters were collected as per ISRM (1978) )−0.03JRC0


standards which include the orientation of major joint (5)
(
JCSn = JCS0 Ln ∕L0
planes with their persistence, spacing, aperture, infilling,
weathering, surface roughness (joint roughness coeffi- where ­JRC0, ­JCS0 and ­L0 represent the values of 100 mm-
cient (JRC)) and compressive wall strength (joint com- scale samples and J­ RCn, ­JCSn and L
­ n refer to in-situ block
pressive strength (JCS)). Moreover, the included acute sizes.
angles between joints and water condition were also care-
fully observed. Joint volumetric count (­ JV) (or volumetric
frequency) was measured using the method proposed by Assessment of rock block distributions
Palmstrom (1982) (Eq. 1). The Rock Quality Designation
(RQD), a practical tool to quantify the linear frequency of The statistical distributions of rock block parameters have a
joint population data, was calculated using the following substantial influence on rockfall hazard assessment. Block
correlation proposed by Palmstrom (2005) (Eq. 2). sizes and block volume distributions greatly influence the
mobility behaviour of rock blocks and their associated kinetic
energies during rockfall events. Suggestions for protective
( )
∑J 1
Jv = (1)
i=1 S
i measures and their implementation are primarily based on
distributions of in-situ rock blocks that are kinematically
feasible for rockfalls (Windsor 1997; Elmouttie and Poropat
and, RQD = 110 − 2.5Jv (2)
2012; Ruiz-Carulla et al. 2015). Joint spacing and persistence
where ­Si is the average joint spacing for an ith joint set in are the critical parameters to predict the size and volume
meters, J is a total number of joints sets excluding random of in situ as well as detached rock blocks (Elmouttie and
joints. JRC, which is potentially a very significant param- Poropat 2012). In the present study, two types of rock block
eter of joint shear strength, were obtained for each of the distributions were assessed at the rockfall site, which are:
joint sets with the help of a 2D roughness profilometer. The
2D profiles were then visually compared with the stand- 1. In-situ Rock Block Distribution (IRBD), and
ard profiles proposed by Barton and Choubey (1977) and 2. Rockfall Block Size Distribution (RBSD)
assigned a numeric value to get the quantitative roughness.
The Schmidt hammer (N-type) test was carried out on the The volumetric distribution and potential dimensions of
joint surfaces (Aydin 2008), and JCS was calculated using IRBD were statistically quantified through the methods pro-
the correlation given by Deere and Miller (1966) as in Eq. 3. posed by ISRM (1978). The potential block volume ­(Vb) for
the persistent joints are calculated as
UCS(MPa) = 6.9 × 10[0.0087𝛾N+0.16] (3)
Vb = S1 × S2 × S3 ∕Sinγ1 × Sinγ2 × Sinγ3 (6)
where γ is the unit weight in g/cm3 and N is the rebound
value of the N-type Schmidt hammer. The scale corrected where ­S1, ­S2 and S
­ 3 are the spacings of three joint sets 1, 2
equations (Barton and Bandis 1982) as given in Eqs. 4 and 5 and 3, respectively. Sin γ1, Sin γ2 and Sin γ3 are the acute
were applied to quantify the scale effects on JRC and JCS as intersection angles between the joint 1 and 2, 2 and 3, 1

13
5352 A. K. Singh et al.

and 3, respectively. Another approach to quantify ­Vb was The Fisher distribution function generally applies to describe
proposed by Palmstrom (2000) as in Eq. 7. the angular probability distribution of 3-dimensional orienta-
( )−3 tion vectors over a sphere with the equation provided in Eq. 8.
(7)
( )
Vb = β × Jv , where, β = 20 + 7 Smax ∕Smin
K sin 𝜃eK cos 𝜃
f (𝜃) = (8)
where β is the “block shape factor” ranging from 27 to 100. eK − e−K
­Smax and ­Smin refer to the longest and shortest dimensions of
where θ in degree is the angular deviation from the mean
the blocks, respectively. Along with ­Vb, block size index (lb)
(pole) vector. “K” is the “Fisher constant” or dispersion
has also been measured to predict the average dimension of
factor that delineates the tightness or dispersion of an ori-
typical in situ rock blocks.
entation cluster (Rocsciences 2016a). A tight cluster of ori-
Fragmentation analysis is a very popular and frequently
entated vectors (poles) is represented by a larger K value,
used technique in rock blasting to know the distribution
whereas a smaller K value suggests a more dispersed cluster
and quality of fragmented rocks due to given explosive
of vectors. The K value can be assessed through Eq. 9 as
energy. For RBSD, fragmentation analysis was conducted
through WipFrag v2.7.28 to describe the size distribution N−1
of detached blocks/lumps during recent rockfall activity.
K= (9)
N−R
WipFrag considers the fragmented rock block as an equiva-
lent sphere. Based on equivalent diameters (ED) of blocks, where N defines the number of poles and R expresses the
a gradation (distribution) curve for fragmented rock block magnitude of the resultant pole (or magnitude of the vec-
heap was obtained from the analysis by digitising block tor sum of all pole vectors). However, the exposed slope
boundaries with an appropriate field scale. The detailed profile at the studied rockfall site is curved in nature where
nature and distribution of IRBD and RBSD at the rockfall the vulnerable scar portion depicts a continuous change in
site are discussed in “Assessment of rock block distribution” slope dip direction. Therefore, a modified stereographic
and “Characteristics of IRBD and RBSD”. projection method proposed by Kundu et al. (2016a) was
used to determine the kinematic feasibility over the convex
slope profile. Then, to forecast the potential rockfall initia-
Rockfall hazard prediction
tion zones from the highway slope section, numerical stabil-
ity analysis was performed through a discontinuum method
In the present study, rockfall hazard prediction involved the
(distinct element analysis using UDEC). Discontinuum
appraisal of kinematically removable blocks induced by unfa-
techniques are more resilient when dealing with the inher-
vourable joint orientations against the slope face direction. It pro-
ent discontinuities in the rock masses and possible rupture
vides an idea about the types of block failure that can be involved
within it. Distinct element method (DEM) is based on divid-
in rockfall hazard. The potential zones along which these block
ing rigid or deformable blocks by deformable interfaces.
failures may initiate can be numerically modelled through the
The DEM is popular for solving discontinuum problems
distinct element method. The methods that were applied to pre-
and is extensively used for practical stability analysis pur-
dict the rockfall hazard has been briefly described in this section.
poses of jointed rock slopes (Kveldsvik et al. 2009; Dotta
et al. 2017). Among the discontinuum methods, DEM has
Prediction of kinematically feasible rockfalls and initiation the advantage of adopting both rigid and deformable blocks
zones along with deformable contacts (Cundall and Hart 1992).
This method, being explicit, directly solves the equations of
The kinematic analysis is an effective means to anticipate the motion by time marching scheme (Itasca 2004). DEM is the
most likely failure modes in the primary phase of rock stability most suitable for discontinuous media as it allows large dis-
analysis (Hoek and Bray 1981). According to Priest (1985), the placements along the discontinuities and rotation of blocks
kinematic analysis is an essential precursor in the analysis of (Bobet et al. 2009). For small block sizes, the method allows
rigid blocks as it shows the possibility of physical removal of translational shear followed by rotation; however, only trans-
blocks towards free space without disturbing the mechanical lational shear occurs in case of large block sizes (Bhasin and
stability of adjacent blocks or rock mass. This method is based Kaynia 2004).
on the assumption that friction is the only responsible factor
for the shear strength of the slope surface. This method consid-
ers the geometrical relationship among joints, bedding planes Rockfall trajectory simulation with rock block
and slope orientations along with friction angle to assess the characteristics
types and direction of failures. The Fisher probability density
function of poles (Fisher 1953) was applied to determine the Conventional rockfall trajectory simulation was done
mean discontinuities orientations of the collected joint data. through RocFall v5.0 (Rocsciences 2016b) using the rigid

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5353

body mechanism to find out the critical rockfall parameters Table 3  Variation of J­ V and RQD values
with respect to rock block characteristics. These parameters Parameters JV Volumetric frequency RQD
include maximum runout distance, bounce height, transla-
tional velocity and total kinetic energy of the falling block Mean 17.1 High 67.2
relative to slope profile during the block impact. The maxi- Min 7.0 Moderately high 92.6
mum runout distance differentiates between the safe and Max 94.1 Very high -
unsafe areas across the highway. The cumulative effects of
rock block characteristics and predicted initiation zones on
rockfall hazard were studied, and the performance of a suit- between joints (Fig. 3) that qualitatively depicts very small
able preventive measure was also carried out in predicted to medium-sized in situ rock blocks (ISRM 1978).
rockfall scenarios. The bounce height, total kinetic energy
and impact of falling blocks relative to slope geometry were Assessment of rock block distribution
used to select the height and capacity of the proposed bar-
rier to restrict the falling blocks from reaching the highway. In‑situ Rock Block Distribution (IRBD):  IRBD provides a fair
approximation of kinematically detachable rock blocks that
might come down from rockfall initiation zones. Conven-
Results tional methods were used to quantify the IRBD and rock
block characteristics based on a detailed field survey. The
Rock block characteristics weigh-up distribution of IRBD for the studied rock mass are
summarised in Tables 4 and 5. Depending on the range of
Joint characterisation observed ­JV based on joint spacing data, the predicted block
sizes vary from very small to medium-sized blocks (ISRM
The quantitative and qualitative descriptions of major joints 1978). The maximum joint spacing (83.5 cm) from the cho-
with their mean orientations are summarised in Table 1. sen sampling zone shows that the maximum potential block
Joint set J1 was observed to be persistent greater than 20 m, sizes that represent the volumetric expression of joint den-
whereas J2, J3 and J4 have the persistence of 3 to 15 m. sity will be of moderate size and the potential block volume
The joint opening is categorised as open to close with sili- ­(Vb) will be in the range of 0.03 to 1 ­m3 (Fig. 4). The cal-
ceous/soil filling to no fill. The surface staining observed in culated ­Vb (Eq. 6) for mean, minimum and maximum joint
J1 and J2 shows slight surface weathering condition along spacing were found to be 24.7 d­ m3, 0.18 d­ m3 and 335.29
the joint walls. The estimated J­ RCn for investigated joints ­dm3 respectively (Table 4), which shows dominancy of very
shows slickenside stepped to rough planar and undulating small to moderate block sizes at the rockfall site (Fig. 4).
wall roughness, whereas ­JCSn suggested slightly weathered
joint wall conditions (Table 2). The largest detached blocks were considered to determine
The estimated average values of ­JV and RQD (Table 3) the block shape factor (β), which suggest an average value
suggest a high to very high degree of jointing (Palmstrom of 45.28. The block volume estimated from J­ V and β (Eq. 7)
2005) and fair rock quality (Deere and Miller 1966), respec- lies within the range of 0.05 d­ m3 to 133.62 d­ m3. Volume
tively. Higher ­JV represents highly fractured nature of the distribution of IRBD from various conventional approaches
quartzite rock mass. As per ISRM (1981), the frequency (Table 4) shows quite a variation in the estimated mean V ­ b.
distribution of spacing suggested close to wide spacing Vb depicts the maximum value range when intersecting
acute angles between joints are introduced in ­Vb calculation
(Table 4), which signifies the profound influence of included
Table 2  Quantitative estimations of J­RCn and J­CSn for major joint angle between joints. The block size index (lb) indicates the
sets average dimension of typical in situ rock blocks. Three joints
Parameters Joints Mean SD Description whose average spacings are minimum out of four were cho-
sen to represent the Ib as most widely spaced joint and may
JRCn J1 5.9 2.0 Slickenside
not suggest realistic results and have little effect on actual
J2 3.2 1.2 stepped
to rough observed block dimensions (ISRM 1981). The estimated
J3 4.6 2.4
planar lb ranges from 4.37 to 57.33 cm with mean dimension of
J4 2.4 0.9
21 cm (i.e. very small to moderate sizes) which was quite
JCSn (MPa) J1 77.3 26.4 Fresh to
visible at the rockfall site. Additionally, the dimensions of
J2 65.8 25.2 slightly
weathered some medium-sized detached blocks (the majority being
J3 137.5 27.4
cubical) from recent rockfall events were measured during
J4 170.9 33.4
the field survey which suggests lb range from 30 to 36 cm.

13
5354 A. K. Singh et al.

Fig. 3  Frequency histogram plot


showing joints spacing distribu-
tion at the rockfall site (ISRM
1981)

Thus, the collective observations from various indices infer the heap were digitised (Fig. 5a, c) to obtain the equivalent
that the majority of small dimensions of in situ rock blocks diameter (ED) of each block sizes followed by the deter-
and IRBD favours the higher percentage of small-sized mination of frequency distribution based on the passing
blocks and volume distribution. Considering the relatively percentage (gradation curve) of the measured rock blocks.
small effect of intersection angle on ­Vb as compared to the A total of 3456 detached blocks from recent rockfall event
joint spacing (Cai et al. 2004), ­Vb has also been approxi- were digitised to obtain the gradation curve (Fig. 5b). The
mated through ­S1 × ­S2 × ­S3 for practical purpose. Out of four maximum frequency of ED ranges from 130 to 50 mm, with
major joints, three consecutive joint spacings were consid- a passing percentage range from 65 to 26%. The blocks hav-
ered at a time for ­Vb calculation. Hence, there are four pos- ing ED greater than 130 mm also constitute a decent propor-
sible combinations (4C3), which approximate the potential tion (almost 25%) of the total failed blocks. The minimum,
range of block volume without considering included angles maximum and mean block sizes as observed from the analy-
(Table 5). sis are 5.1 mm, 625.21 mm and 69.22 mm, respectively, with
a standard deviation of 47.12 mm. The ­D10, ­D25, ­D50, ­D75
Rockfall Block Size Distribution (RBSD):  The frequency dis- and ­D90 are 5.33 mm, 45.42 mm, 93.68 mm, 154.92 mm
tributions of size, volume and mass percentage of detached and 333.37 mm, respectively. It suggests that 50% of the
blocks at rockfall site were quantified through fragmenta- fragmented blocks are greater than 93.68  mm, whereas
tion analysis. The detached blocks were deposited as a heap 25% of them are having ED greater than 154.92 mm with
of rock blocks at the base of the slope. The blocks from a small proportion (10%) of blocks greater than 333 mm.

Table 4  Estimated IRBD Statistical Vb = ­S1 × ­S2 × ­S3 Vb = β × ­(JV)−3 Vb = ­S1 × ­S2 × ­S3/Sin Mean block


through various indices distribution ­(dm3) ­(dm3) γ1 × Sin γ2 × Sin γ3 ­(dm3) size index (­ lb)
(cm)

Mean 14.70 9.01 24.70 20.95


Min 0.12 0.05 0.18 4.37
Max 235.81 133.62 335.29 57.33

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5355

Table 5  Distribution of Distributions Vb = ­Si × ­Sj × ­Sk


potential IRBD considering
spacing of four major joints at S1 × ­S2 × ­S3 ­(dm3) S2 × ­S3 × ­S4 ­(dm3) S3 × ­S4 × ­S1 ­(dm3) S1 × ­S2 × ­S4 ­(dm3)
rockfall site
Mean 19.23 15.70 14.84 9.04
Min 0.14 0.19 0.071 0.06
Max 260.19 165.00 356.04 162.03

The gradation curve also signifies the dominancy of small- In this section, the obtained volume-frequency distribu-
sized blocks (Fig. 5d), which could be the result of subse- tions of IRBD and RBSD were analysed and compared.
quent fragmentation during blocks transportation. Based on For IRBD assessment, the volumes (Eq. 6) were obtained
the fragmentation results, the failed blocks were classified probabilistically for 1000 samples from measured spac-
according to their size and percentage distribution. Table 6 ing distribution considering major joints. The volumes of
shows the calculated rock block masses corresponding to the RBSD was estimated by considering ED from fragmenta-
volume percentage within each size range. The calculated tion analysis that represents a fair approximation of equiva-
size and associated mass distributions (Table 6) were used in lent block volume, which is frequently practised in rock
rockfall trajectory simulation to evaluate the rockfall hazard blasting. It ought to be noticed that the highway section is
along the highway. typically characterised by very small to small-sized blocks
(high jointing frequency) which is also evident from the
Characteristics of IRBD and RBSD volume-frequency plot (Fig. 6a). The RBSD particularly
shows control of very small block volume (Fig. 6a) mani-
The volume-frequency relation of rockfalls satisfies festing the subsequent fragmentation effect. Consideration
certain distribution characteristics (Hungr et  al. 1999; of very small blocks can cause underestimation of the pro-
Corominas et al. 2018). Volume-frequency distribution of tection designs and barrier capacity. Santana et al. (2012)
rockfalls is generally approximated by the inverse-power also mentioned the censoring of small volumes in fre-
law (Mavrouli and Corominas 2017; Ferrero et al. 2016; quency-volume distribution due to under-sampling or phys-
Ruiz-Carulla et  al. 2015; Mavrouli et  al. 2015; Hantz iographic limitations. Therefore, to eliminate the rollover
2011; Santana et al. 2012; Dussauge-Peisser et al. 2002). effect of very small block volumes on distribution fitting,

Fig. 4  Distribution of block
sizes and block volumes for
observed spacing ranges at the
rockfall site

13
5356 A. K. Singh et al.

Fig. 5  (a) Rockfall heap. (b) Gradation curve obtained from the fragmentation analysis. (c) Digitisation of failed rock blocks. (d) Field view size
distribution of fragmented rockfall blocks

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5357

Table 6  Block size statistics derived from fragmentation analysis Prediction of rockfall hazard
3 3
Block Approxi- Density (kg/m ) Volume ­(m ) Mass (kg)
size mated Prediction of kinematically feasible rockfalls and their
(mm) percentage initiation zones
30 25% 3220 3E-05 0.09
50 15% 3220 1E-04 0.4
For the determination of credible failure modes from the
100 15% 3220 0.001 3.22
slope sections, kinematic analysis has been performed using
150 20% 3220 0.003 10.87
the stereographic projection method (Hoek and Bray 1981).
300 15% 3220 0.027 86.94
The slope has been identified with potential wedges and flex-
500 5% 3220 0.125 402.5
ural toppling failures (Singh 2018). As the slope is curved
625 5% 3220 0.244 786.1
with face direction varying from N265° to N10°, the kin-
ematic technique for curved slope (Kundu et al. 2016a) has
All block sizes below 30 mm are considered as 30 mm as they have been implemented to understand the failure potentiality with
very less impact from the rockfall point of view changing slope direction. It can be interpreted from Fig. 2
that the slope section facing towards 265° is more affected
by the discontinuities induced failures. This can be well cor-
the block volumes < 0.001 m ­ 3 were excluded. The volume- related with the kinematic analysis (Fig. 7) which shows the
frequency distribution of IRBD and RBSD fit well with influence of wedge failures in two directions (one towards
inverse power functions with a high R2 of 0.91 and 0.99, 185° and the other towards 225°) in the westerly oriented
respectively for block volumes > 0.001 ­m3 (Fig. 6a). The slope profile (slope section-A in Fig. 7). A similar reason
volumetric distributions obtained in IRBD and RBSD were is also applicable to the less severely affected slope face
compared relative to the cumulative frequency (Fig. 6b), towards 10° (section B in Fig. 7) which has an influence of
suggesting a value of cumulative frequency of about 40% wedge failures in only one direction (towards 35°).
of IRBD and corresponds to RBSD. This can help in pro- Although the kinematic analysis of the overall slope
viding the minimum block size threshold for designing explains more acute failure towards 265°, the advancement
rockfall preventive practices along the highways in the of failure both towards the east and north direction is struc-
region. The upper limit of the distribution defines the com- turally controlled in the particular section. Figure 8 explains
paratively lower volume distribution for RBSD compared the advancement of failure in the east direction that created
to IRBD, which signifies the effect of fragmentation during an open face towards the southwest direction, resulting in a
block mobilisation along its path (Corominas et al. 2018). concave face.
This volume distribution behaviour is generally observed Kinematic analysis of the concave section suggested a
in rockfall having small to medium block size distribution strong influence of both the wedges towards the south-facing
(Ferrero et al. 2015). The results suggest rational valida- slope. Thus, rock blocks are getting released kinematically
tion of block size distribution condition in the study area. from the south-facing section, hence advancing towards

Fig. 6  a Volume-frequency distribution of IRBD and RBSD. In both cases, it fits well with the inverse power functions for block vol-
umes > 0.001 ­m3. b Comparison of cumulative frequencies relative to volume distributions for IRBD and RBSD

13
5358 A. K. Singh et al.

Fig. 7  Kinematic analysis of
slope section A and section B
considering the curved face

the north direction. The kinematic feasibility analysis sug- DEM, cross-sectional outlines were selected from the most
gests a high possibility of frequent joint-driven block failures vulnerable slope profile based on kinematic results and
that eventually turn into rockfalls and may disrupt the traffic field conditions. 2D plain strain distinct element rock slope
services along the highway. Thus, the assessment of probable models were simulated with Mohr-Coulomb (MC) mate-
zones of block initiation becomes essential, which was car- rial properties and MC slip criterion for the rock matrix
ried out employing discontinuum slope stability modelling. and interfaces, respectively (Singh 2018). The properties
The discontinuum approach is very effective in deter- incorporated into the model are given in Table 7.
mining the block detachment from a jointed rock slope. The The slope mass contains four joint sets, out of which the
kinematic analysis signifies that the highway segment is three most vulnerable joint sets were considered (based on
vulnerable to block detachments due to unfavourable joint- kinematic analysis) in the model geometry (Fig. 9a). The
slope orientation. The 2D Distinct Element (DE) analysis reduced joint set helps to avoid complexity and save com-
of the exposed highway slope identifies block detach- putation time. The model slope geometry has a height of
ment locations, which help to detect the probable rock- 40 m with an overall slope angle of 83° except at the base,
fall initiation zones. For the stability analysis through 2D where the slope is relatively gentle (47˚). The 2D profile of

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5359

Fig. 8  Illustration of kinemati-
cally induced wedges in the
main scar portion (towards the
south-facing slope section A)

the slope and relative joint orientations with locations for contour forms. Results of the stability analysis were repre-
monitoring points (1–5) are shown in Fig. 9a. sented in terms of displacement, shear strain and records of
To simulate static loading, rollers are applied at the lateral block detachment. Five monitoring points (1–5), as indicated
boundary restricting its motion along the x-axis and pins at in Fig. 9a, were set from top to bottom along the slope profile
the bottom to fix its base. The model was brought to an equi- to record the displacement history along the most probable
librium state under its own weight. Once the model reached block detachment zone based on observed field conditions.
equilibrium, results were obtained in graphical, pictorial and Results show a maximum displacement of 1.225  m after
100,000 cycles (6.28 s) observed towards the base of the
slope (Fig. 9b) where the hillslope changes its gradient from
Table 7  Input parameters used in numerical stability simulation steep to gentle. The deformation mechanism of the rock mass
Parameters Intact rock Joints involves joint opening, block deformation and block detach-
3
ment (Fig. 9b). According to Jaeger (1971), this type of slope
Density (kN/m ) 31.5 -
mass arrangement signifies the build-up of localised stresses.
Cohesion (MPa) 14.9 -
Slight dislodgement may change the relative position of rock
Friction angle (0) 32 -
block arrangement that may result in high localised stress and
Intact rock elastic modulus (­ Ei) (GPa) 18.06 -
can cause indirect tensile stress-induced individual block fail-
Poisson’s ratio 0.28 -
ures. This type of rock block behaviour was observed during
Residual friction angle for joints (ϕr) J2 - 27
the field survey. There was an increase in x displacement from
J3 - 26
the top to the bottom of the slope face (Fig. 9c). Shear strain
J4 - 26
contours in fig. 9d represent the amount of shear strain induced
Normal stiffness (MPa/m) J2 - 10,908
due to the self-weight of the slope mass. High concentrations
J3 - 10,434
of the shear strain (zoomed portions in Fig. 9d) correspond
J4 - 6315
with the high displacement zones. Thus, the numerical stabil-
Shear stiffness (MPa/m) J2 - 1095
ity assessment provided sufficient shreds of evidence of block
J3 - 1007
failures induced by adverse ordination of joints relative to
J4 - 625
highway slope profile, ultimately generating rockfall.

13
5360 A. K. Singh et al.

Fig. 9  (a) A schematic representation of the slope profile along with placement of the detached rock blocks at monitoring points. (d) Shear
monitoring points to record displacements along the x-direction. (b) strain contours representing the amount of shear strain induced due to
Displacement along the slope in the x-direction (joint opening, block self-weight of the slope mass (readjustment of the joints)
deformation and block detachment are observed). (c) Horizontal dis-

Rockfall trajectory simulation with rock block and finally colliding to the slope surface at gentler portion.
characteristics At the study site, bouncing is prominent rather than roll-
ing along the ramp portion, probably due to the high slope
Based on the field data and conventional and numerical gradient (more than 45°) with an almost barren quartzite
stability analysis, rockfall trajectory simulation was done surface. Vishal et al. (2017) have also pointed out the impor-
for the most susceptible highway profile (section A) con- tance of slope geometry on rockfall trajectories in lesser
sidering the predicted block detachment zones. The poten- Himalayan terrains. As discussed in previous sections, the
tial slope section is characterised by a high gradient (mean highway slope is highly jointed and facilitates detachment
slope dip of 83°) of the scar with a relatively gentler ramp of rock blocks of varying sizes, therefore being extremely
and a mean gradient of 47°. So, as per the observations of prone to rockfall hazard. The observed characteristics of
Ritchie (1963), the upper portion is likely to be more vul- IRBD and RBSD represent the potential blocks exposed in
nerable for free fall after block detachment. Subsequently, the vulnerable slope section that has great significance for
rock blocks would undergo bouncing, rolling and sliding accurate prediction of the falling block dynamics such as

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5361

bounce height, runout path and kinetic energies. Therefore, Table 9  Seeder properties used Type Line seeder
the block distribution was rationally used in rockfall simu- in rockfall simulation
lation to analyse the rockfall hazard and design a barrier of Shape Table 10
adequate energy at a suitable place along the rockfall runout Mass Depending upon
block size
path. The input parameters to define the slope surface and (Table 6)
seeder properties in rockfall analysis are given in Table 8 Density 3220 kg/m3
and Table 9, respectively.
Block shape is an important parameter that ultimately
affects the kinetic energy and bounce height depending
upon the nature of point of contact along the slope profile. just after the detachment of the blocks from the uppermost
So, four-block shapes (Table 10) were considered for the seeder (scar portion). However, the bounce height subse-
analysis based on field observations. The gabion wall, which quently got reduced along its path to an average value of
already exists at the site, has been taken into consideration 4 m as the blocks encountered the gentler slope profile. The
to analyse the effect of rock block characteristics (Fig. 10a). total kinetic energy graph of the detached blocks along the
The slope profile used for rockfall simulation was slope profile suggests a maximum value of 132 kJ which
obtained from the field using a laser distance meter and is successively reduced up to 60 kJ near the existing gabion
shown in Fig. 10a. The positions of 3 line seeders, which (Fig. 11b). The total kinetic energy is observed to be sig-
represent the rockfall initiation zones, were selected based nificantly associated with the bounce height of the detached
on the positions of considerable rock block detachment rock blocks from seeder 1 and 2 (Fig. 11a, b).
zones observed in the DEM stability analysis. The rockfall
trajectory simulation results are presented in Fig. 10b, c
which show varying trajectory paths created by blocks of dif- Discussion
ferent shape and sizes from selected seeders along the slope
profile. The analysis suggests that the block shape hardly Impact of rock block characteristics on rockfall
influences the rockfall hazard parameters such as bounce hazard
height, runout and kinetic energy relative to size and detach-
ment positions (Fig. 10b, c). A relative proportion of block The rockfall hazard parameters such as bounce height, block
sizes was considered from the entire range to normalise each movement and related kinetic energy are observed to be
seeder’s corresponding mass distribution based on observed highly governed by the size, associated mass distribution and
size and volume distribution (Table 11). The results of rock- release position of the detached blocks (Figs. 10b, c and 11).
fall analysis show that the runout of all the simulated block However, the impact of block shape is found to be incon-
sizes (greater than 300 mm too) from lowermost seeder 3 sequential on rockfall hazard parameters as compared to
are effectively captured by the existing gabion (Fig. 10c). the size and detachment positions (Fig. 10b, c). To observe
Unlike blocks from seeder 3, detached blocks from seeder the impact of the size distribution of detached blocks on
1 and 2 produce an average bounce height of 3 to 4 m with rockfall hazard, a particular size range of overtopped blocks
a mean kinetic energy of around 40 kJ. The runout result from existing gabion were analysed with respect to the total
of block sizes having ED greater than 300 mm proposes number of overtopped blocks (Table 11) since these blocks
that the maximum overtopped blocks ended in the valley are can be fatal to highway traffic. Overall, 16.5% of the blocks
from seeders 1 and 2 (Fig. 11a). The results presented in ranging from 30 to 625 mm are detected that overtopped the
Fig.  11a indicate a maximum bounce height of 8–9  m gabion with a large proportion of small-sized (30–150 mm)
blocks. However, the mass percentage of medium-sized
Table 8  Input parameters for slope surface materials in rockfall simula-
(> 300 m) overtopped blocks is very high as compared to
tion the total overtopped blocks (Fig. 12). The observed weight
Material types Normal Tangential Dynamic Rolling
restitu- restitution friction resistance Table 10  Block shapes Shape Illustration
tion considered in rockfall
simulation Super ellipse^1:1
Quartzite slope mass 0.3 0.8 0.57 0.55
Talus cover 0.32 0.8 0.42 0.75 Super ellipse^2:3
Road 0.4 0.82 0.58 0.4
Soil 0.39 0.56 0.58 0.65 Super ellipse^5:6
Concrete 0.48 0.53 0.58 0.5
Super ellipse^1:2
Gabion wall 0.35 0.72 0.58 0.4

13
5362 A. K. Singh et al.

Fig. 10  a 2D profile with probable rockfall initiation zones (seeders) for rockfall simulation. b Rockfall trajectories from all the three seeders. c
Rock blocks trajectories from seeder 3

proportions of overtopped blocks of 625-mm, 500-mm and the chance of impact of medium-sized blocks on the high-
300-mm sizes are 64.98%, 22.18% and 14.37%, respectively way traffic is much higher than small-sized blocks and can
(Fig. 12). Comparing the mass percentage of overtopped turn into a fatal hazard. The results also suggested that the
blocks to total simulated blocks also signifies an increas- release positions (initiation zones) highly govern the relative
ing percentage of mass with increasing block sizes. Hence, bounce height, block mobilisation (run out) and impact of

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5363

Table 11  Size and mass distribution of simulated blocks that overtop the gabion wall
Sl. no. Size (mm) Mass per Number of Total weight of each Number Weight of blocks Weight % w.r.t Weight % w.r.t
block rocks used for block size (kg) of blocks crossing gabion total no. of total mass of
(kg) simulation crossed (kg) blocks crossing simulated blocks
gabion gabion

1 30 0.09 100 8.69 15 1.30 0.02 0.004335


2 50 0.40 60 24.15 7 2.82 0.04 0.009365
3 100 3.22 60 193.20 11 35.42 0.49 0.117735
4 150 10.87 80 869.40 11 119.54 1.65 0.397356
5 300 86.94 60 5216.40 12 1043.28 14.37 3.467832
6 500 402.50 20 8050.00 4 1610.00 22.18 5.351593
7 625 786.13 20 15,722.66 6 4716.80 64.98 15.6785
Total = 400 Total Total no. of Total
weight = 30,084.50 blocks = 66 weight = 7529.16

the blocks down the slope respective of block sizes, shapes lowermost detachment zone (seeder 3) due to the relativity
and associated mass coupled with a slope profile (Figs. 10 short released height of blocks with a gentler slope profile,
and 11). It is shown in Fig. 10c that the existing gabion which results in block movements with low bounce height
fully captures the blocks of all sizes generated from the and kinematic energy. However, the detached blocks from

Fig. 11  (a) Variation in bounce


height (maximum) along with
the slope profile for the blocks
crossing gabion (sizes greater
than 300 m) and distribution
of rockfall blocks path end
locations. (b) Variation in total
kinetic energy (maximum)
along with slope profile for the
blocks crossing gabion (sizes
greater than 300 m)

13
5364 A. K. Singh et al.

Fig. 12  Graphical representa-
tion of mass percentage distri-
bution of block sizes crossing
gabion to the total mass of
simulated blocks

seeders 1 and 2 show fairly high bounce heights with high goal of the simulation was to arrest the maximum number of
kinetic energy (Figs. 10b and 11), hence needing extra pre- blocks from the upper scar (uppermost seeder) and middle
vention measures along with the existing gabion. The jagged portion of the slope in the future rockfall scenario. Other-
profile of the slope near its middle portion gives rise to the wise, the detachable small to moderate-sized blocks may
extra bounce and long offset distance of impacted blocks result in shooting stones which can be very hazardous to the
along its runout paths (Fig. 10b). So, it is ascertained that the traffic. Based on the analysed results, a vertical barrier of
block trajectories of varying sizes are related to its released 150-kJ capacity with 2.5 m height was suggested to mount
position from the slope. Therefore, to encounter the imposed over the existing gabion (Fig. 13), which would be sufficient
threat due to rockfall on the highway traffic, medium-sized to capture the future rockfall events.
blocks detached from the upper and middle parts of the slope Singh (2018) has studied the joint and rock block charac-
should be considered in designing the appropriate protective teristics of highway slopes around the Rampur-Jhakri geo-
measure. structural region and suggested the dominance of high to
moderate joint frequency in the area. He has also signified
Implications for rockfall protection strategies the persistence of small to moderate-sized in situ blocks in
the exposed highway slopes which are kinematically sus-
To check and optimise the appropriate protection strate- ceptible to rockfall hazards. The comparative study of
gies for the highway section, a rockfall barrier coupled with cumulative frequency of in-situ and detached rock blocks
the existing gabion wall would be a feasible option at the also suggests that the prevailing in situ blocks correspond
study site. The trajectories of blocks, with size greater than to the detachable blocks irrespective of the fragmentation
300 mm have a high potential to reach the highway and were effect. So the knowledge of in situ block frequencies gener-
analysed to suggest a protection barrier of appropriate capac- ated from high to moderate joint frequency can provide a
ity. To achieve that, a parametric analysis was done to evalu- rational threshold range for which a protection strategy can
ate the performance of the proposed protection barrier with be established and optimised for vulnerable rockfall sections
existing gabion, considering the bounce heights and associ- along the highways. Therefore, the methodology signifying
ated kinetic energies of the simulated blocks. Various com- the determination and analysis of rock block characteristics
binations of barrier characteristics such as capacity, length, on rockfall hazard in the present study would help to design
height, elongation and inclination were simulated for vary- preventive practices along the highway sections, particularly
ing rock block sizes in the expected rockfall scenario. The along NH-5 in the higher Himalayan terrain.

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5365

Fig. 13  Simulated rockfall bar-


rier with properties

Conclusions the designer to set an appropriate block size threshold for


designing rockfall protection measures.
The present study delineates a comprehensive methodology Wedge detachments are found  to be kinematically
to investigate the rock block characteristics effect on rock- feasible and prominent at the site which advanced along
fall hazard and associated protection strategies for highway slope strike posing a high threat to running traffic in the
slopes. The influence of block characteristics such as size, form of rockfalls. These block failures are caused by the
shape, mass and release position from the slope on rockfall localised stress build-up zones along the slope profile,
hazard and design of protection barrier has been demon- which induce shear strain in the rock face that will initi-
strated through a case study in the Himalayan highway slope. ate future rockfalls. The block displacement mechanism
The rock block characterisation is performed by proper inves- suggests that joint opening is prominent near the crest
tigation and statistical analysis of rock joints’ parameters. region while block deformation and detachment occur in
The joint characteristics confirm the dominance of close to the middle and scar portions of the slope. The predicted
wide joint spacing with low to high persistence and signify a rockfall hazard parameters such as bounce height, block
high volumetric frequency that yields small to medium-sized movement and related kinetic energy are observed to be
in situ blocks. The estimated RQD and ­JV have also validated highly governed by the size, associated mass distribution
the result caused by a high to a very high degree of jointing. and release position of the rockfall blocks. However, the
The probabilistically estimated IRBD by empirical indices impact of block shape is found inconsequential. Though
signifies the occurrence of small to medium-sized in situ a large proportion of small-sized (30–150 mm) blocks
rock blocks with V ­ 3. However, the
­ b ranging from 0.03 to 1 m can hit the highway, the mass percentage of medium-
gradation curve obtained from fragmentation analysis speci- sized (> 300 mm) blocks is observed to be high which
fies a gap-graded size distribution of RBSD with significant can be fatal to the highway traffic signifying the block
small-sized blocks, manifested the effect of subsequent frag- size control on rockfall hazard. The study suggests that
mentation. The volume-frequency distributions of IRBD and the release positions (initiation zones) of detachable rock
RBSD also validate typical characteristics of very small to blocks highly control the relative bounce height, run-out
medium-sized blocks that well fitted with inverse power func- distance and impact of the blocks down the slope respec-
tions for block volumes greater than 0.001 m ­ 3. A compari- tive of block sizes and associated mass coupled with the
son suggests a value of cumulative frequency of about 40% slope geometry. To provide an effective design for rock-
of IRBD follow and corresponds to RBSD that will assist fall protection at the study site considering block sizes

13
5366 A. K. Singh et al.

greater than 300 mm, a vertical barrier of 150 kJ capacity advances for rock mechanics applications. J Geotech Geoenvi-
with 2.5 m high was suggested to mount over the existing ron Eng 135(11):1547–1561
Bostjančić I, Pollak D (2020) Rockfall threat assessment along rail-
gabion which would be sufficient to capture the future ways in carbonate rocks in Croatia. Bull Eng Geol Environ.
rockfall events. Based on the rock block characteristics in https://​doi.​org/​10.​1007/​s10064-​020-​01822-x
the study area, the adopted methodology can be useful to Budetta P (2004) Assessment of rockfall risk along roads. Nat Haz-
define a rational threshold range of block size distribution ards Earth Syst Sci 4:71–81
Cai M, Kaiser PK, Uno H, Tasaka Y, Minami M (2004) Estimation
to efficiently design and optimise the protection strategies of rock mass strength and deformation modulus of jointed hard
for vulnerable rockfall sections along highway corridors rock masses using the GSI system. Int J Rock Mech Min Sci
in Himalayan terrains. 41(1):3–19
Chakraborty D, Anbalagan R (2008) Landslide hazard evaluation
Acknowledgements  The authors are highly thankful to the editor and of road cut slopes along Uttarkashi-Bhatwari road, Uttaranchal
anonymous reviewers for the constructive remarks that significantly Himalaya. J Geol Soc India 71:115–124
improved the quality and clarity of the paper. Collins B, Stock G (2016) Rockfall triggering by cyclic thermal
stressing of exfoliation fractures. Nat Geosci. https://​doi.​org/​
10.​1038/​ngeo2​686
Author contribution  Conceptualisation, methodology, writing—original Corominas J, Mavrouli O, Ruiz-Carulla R (2018) Magnitude and fre-
draft preparation, formal analysis, visualisation: A K Singh. Formal analy- quency relations: are there geological constraints to the rockfall
sis, validation, writing—reviewing, editing, visualisation: JKundu. Valida- size? Landslides. https://​doi.​org/​10.​1007/​s10346-​017-​0910-z
tion, writing—reviewing: K Sarkar. Validation and reviewing: H. K. Verma Cundall PA, Hart RD (1992) Numerical Modelling of Discontinua
and P. K. Singh Engcomput 9:101–113
Deere D, Miller R (1966) Engineering classification and index prop-
Funding  Grant and support were provided by the Indian Institute of erties for intact rock. Tech. Report No AFWL - TR-65-116, Air
Technology (Indian School of Mines) Dhanbad to the first author dur- Force Weapons Lab., Kirtland Air Base, New Mexico
ing his doctoral work. Dotta G, Gigli1 G, Ferrigno F et al (2017) Geomechanical characteri-
zation and stability analysis of the bedrock underlying the Costa
Declarations  Concordia Cruise Ship Rock Mech Rock Eng 50 9 2397 2412
Dussauge-Peisser C, Helmstetter A, Grasso JR, Hantz D, Desvarreux
P, Jeannin M, Giraud A (2002) Probabilistic approach to rock
Conflict of interest  The authors declare no competing interests. fall hazard assessment: potential of historical data analysis. Nat
Hazard Earth Syst 2:1–13
Elmouttie MK, Poropat GV (2012) A method to estimate in situ
block size distribution. Rock Mech Rock Eng 45:401–407
Ferlisi S, Cascini L, Corominas J, Matano F (2012) Rockfall risk
References assessment to persons travelling in vehicles along a road: the
case study of the Amalfi coastal road (southern Italy). Nat Haz-
Acharya B, Sarkar K, Singh AK, Chawla S (2020) Preliminary slope ards 62:691–721
stability analysis and discontinuities driven susceptibility zonation Ferrero AM, Umili G, Migliazza MR (2015) Some open issues on
along a crucial highway corridor in higher Himalaya. J Mt Sci the design of protection barriers against rockfall. In 49th US
India. https://​doi.​org/​10.​1007/​s11629-​019-​5524-6 Rock mechanics/geomechanics symposium. American Rock
Adnan Aydin (2008) ISRM Suggested method for determination of the Mechanics Association
Schmidt hammer rebound hardness: Revised Version. R. Ulusay Ferrero AM, Migliazza MR, Pirulli M et al (2016) Some open issues
(ed.), The ISRM Suggested Methods for Rock Characterization, on rockfall hazard analysis in fractured rock mass: problems and
Testing and Monitoring: 2007–2014 prospects. Rock Mech Rock Eng 49:3615–3629
Ansari MK, Ahmad M, Singh R, Singh TN (2016) Rockfall hazard rat- Fisher R (1953) Dispersion on a sphere. Proc Royal Society London
ing system along SH-72: a case study of Poladpur-Mahabaleshwar A217:295–305
road (Western India), Mahabaleshwar. India Geomat Nat Hazards Froude MJ, Petley DN (2018) Global fatal landslide occurrence from
Risk 7(2):649–666 2004 to 2016. Nat. Hazards Earth Syst Sci. https://​doi.​org/​10.​
Ansari MK, Ahmad M, Singh TN (2014) Rockfall risk assessment 5194/​nhess-​18-​2161-​2018
for pilgrims along the circumambulatory pathway, Saptashrungi Geniş M, Sakız U, ÇolakAydıner B (2017) A stability assessment
Gad Temple, Vani, Nashik Maharashtra. India Geomat Nat Hazard of the rockfall problem around the Gökgöl Tunnel (Zonguldak,
Risk 5(1):81–92 Turkey). Bull Eng Geol Environ 76:1237–1248. https://​doi.​org/​
Barnard PL, Owen LA, Sharma MC, Finkel RC (2001) Natural and 10.​1007/​s10064-​016-​0907-1
human-induced landsliding in the Garhwal Himalaya of northern Gupta V, Tandon RS (2015) Kinematic rockfall hazard assessment
India. Geomorphology. https://​doi.​org/​10.​1016/​S0169-​555X(01)​ along a transportation corridor in the Upper Alaknanda valley,
00035-6 Garhwal Himalaya, India. Bull Eng Geol Environ 74:315–326.
Barton N, Choubey V (1977) The shear strength of rock joints in theory https://​doi.​org/​10.​1007/​s10064-​014-​0623-7
and practice. Rock Mech 10(1):1–54 Gupta V, Sah MP, Virdi NS, Bartarya SK (1993) Landslide hazard
Barton NR, Bandis SC (1982) Effects of block size on the shear behav- zonation in the upper Satluj valley, district Kinnaur. Himachal
iour of jointed rock. 23rd U.S. Symp on rock mechanics, Berkeley, Pradesh Journal Himalayan Geology 4(1):81–93
pp. 739–760 Hantz D (2011) Quantitative assessment of diffuse rock fall hazard
Bhasin R, Kaynia AM (2004) Static and dynamic simulation of a along a cliff foot. Nat Hazard Earth Syst 11(5):1303–1309
700-m high rock slope in western Norway. Eng Geol 71:213–226 Hearn GJ, Shakya NM (2017) Engineering challenges for sustainable
Bobet A, Fakhimi A, Johnson S, Morris J, Tonon F, Yeung MR road access in the Himalayas. Q J Eng Geol Hydrogeol. https://​
(2009) Numerical models in discontinuous media: review of doi.​org/​10.​1144/​qjegh​2016-​109

13
Impact of rock block characteristics on rockfall hazard and its implications for rockfall… 5367

Highland LM, Bobrowsky P (2008) The landslide handbook-a guide to LIDAR-generated digital elevation models. Rock Mech Rock
understanding landslide. Reston, Virginia, U.S. Geological Survey Eng 48(4):1589–1604
Circular 1325 Palma B, Parise M, Reichenbach P, Guzzetti F (2012) Rockfall haz-
Hoek E, Bray JW (1981) Rock slope engineering, 3rd edn. Inst Min ard assessment along a road in the Sorrento Peninsula, Campa-
Metall London nia, southern Italy. Nat Hazards 61:187–201
Hungr O, Evans SG, Hazzard J (1999) Magnitude and frequency of Palmstrom A (1982) The volumetric joint count—a useful and simple
rock falls and rock slides along the main transportation corridors measure of the degree of rock jointing. In: Proc. 41st Int. Con-
of southwestern British Columbia. Can Geotech j 36:224–238 gress Int. Ass. Eng. Geol., Delhi, vol. 5. pp. 221–228
ISRM (1978) Suggested methods for the quantitative description of Palmstrom A (2005) Measurements of and correlations between
discontinuities in rock masses. International Society for Rock block size and rock quality designation (RQD). Tunn Undergr
Mechanics, Commission on Standardization of Laboratory and Space Technol 20(4):362–377
Field Tests. Int j Rock Mech Min Sci Geomech Abstr 15:319–368 Palmstrom A (2000) Recent developments in rock support estimates
ISRM (1981) Suggested methods for the rock characterization. ISRM by the RMi. J Rock Mech Tunn Tech 6(1):1–19
Commission on Testing Methods, Pergamon Press, Oxford, Test- Pandey AK, Sachan HK, VirdI NS (2004) Exhumation history of a
ing and Monitoring shear zone constrained by microstructural and fluid inclusion
Itasca (2004) UDEC Version 4.0: Universal Distinct Element Code. techniques: an example from the Satluj valley, NW Himalaya.
Itasca Consulting Group Inc. India Journal of Asian Earth Sciences 23(3):391–406
Jaeger JC (1971) Friction of rocks and stability of rock slope. ­11th Pei X, Luo J, Huang R (2019) Failure mechanisms of wind-induced
Rankine Lecture. Geotechnique 21(2): 97–134 post-seismic rockfall hazard. Bull EngGeol Environ 78:5707–
Jain AK, Kumar D, Singh SK, Kumar A, Lal N (2000) Timing, quan- 5725. https://​doi.​org/​10.​1007/​s10064-​019-​01509-y
tification and tectonic modelling of Pliocene Quaternary move- Pérez-Rey I, Riquelme A, González-deSantos LM, Estévez-Ventosa
ments in the NW Himalaya: evidence from fission track dating. X, Tomás R, Alejano LR (2019) A multi-approach rockfall
Earth Planet Sci Lett 179:437–451 hazard assessment on a weathered granite natural rock slope.
Kumar V, Gupta V, Sundriyal YP (2018) Spatial interrelationship of land- Landslides. https://​doi.​org/​10.​1007/​s10346-​019-​01208-5
slides, litho ‐ tectonics, and climate regime, Satluj valley, Northwest Priest SD (1985) Hemispherical projection methods in rock mechan-
Himalaya. Geol J. https://​doi.​org/​10.​1002/​gj.​3204 ics. Allen and Unwin, London
Kundu J, Mahanta B, Tripathy A, Sarkar K, Singh TN (2016a) Stabil- Ritchie AM (1963) Evaluation of rockfall and its control. Highway
ity evaluation of jointed rock slope with curved face. In: Proc of Research Record 17, Washington, DC
Indorock-2016, IIT Bombay, Mumbai, pp. 971–978. Rocsciences (2016a) Dips v7.0, Graphical and statistical analysis of
Kundu J, Sarkar K, Singh AK (2016b) Integrating structural and orientation data. Rocscience Inc., Toronto
numerical solutions for road cut slope stability analysis—a case Rocsciences (2016b) RocFall v5.0, Computer program for risk analy-
study, India. Rock dynamics: from research to engineering: pro- sis of falling rocks on steep slopes. Rocscience Inc., Toronto
ceedings of the 2nd international conference on rock dynamics Ruiz-Carulla R, Corominas J, Mavrouli O (2015) A methodology to
and applications. CRC Press, Boca Raton, pp 457–462. obtain the block size distribution of fragmental rockfall depos-
Kundu J, Sarkar K, Singh AK (2019) EasySMR: a computer program its. Landslides 12(4):815–825
to check kinematic feasibility and calculate Slope Mass Rating. Santana D, Corominas J, Mavrouli O, Garcia-Sellés D (2012) Magnitude-
Geophys Res Abstr Vol. 21, EGU2019–1540. frequency relation for rockfall scars using a Terrestrial Laser Scan-
Kundu J, Sarkar K, Singh AK, Singh TN (2020) Continuous functions ner. Eng Geol 145:50–64
and a computer application for Rock Mass Rating. Int J Rock Sarkar K, Singh TN (2008) Slope failure analysis in road cut slope by
Mech Min Sci. https://​doi.​org/​10.​1016/j.​ijrmms.​2020.​104280 numerical method. ISRM Int. Symp, Tehran; pp. 635–642.
Kundu J, Sarkar K, Tripathy A, Singh TN (2017) Qualitative stability Sarkar K, Singh AK, Niyogi A, Behera PK, Verma AK, Singh TN
assessment of cut slopes along the National Highway-05 around (2016) The assessment of slope stability along NH-22 in Rampur-
Jhakri area, Himachal Pradesh, India. J Earth Sys Sci. https://​doi.​ Jhakri Area. Himachal Pradesh J GeolSoc India 88(3):387–393
org/​10.​1007/​s12040-​017-​0893-0 Shen WG, Zhao T, Crosta GB, Dai F (2017) Analysis of impact-
Kveldsvik V, Kaynia AM, Nadim F, Bhasin R, Nilsen B, Einstein induced rock fragmentation using a discrete element approach.
HH (2009) Dynamic distinct-element analysis of the 800 m high Int j Rock Mech Min Sci 98:33–38
Aknes rock slope. Int J Rock Mech Min Sci 46:686–698 Singh AK (2018) Landslide vulnerability analysis around Rampur
Laimer HJ (2020) Determination of rockfall design blocks in Upper area, Himachal Pradesh, India. PhD thesis, unpublished, IIT(ISM)
Triassic limestones and dolomites (Dachstein Formation, North- Dhanbad, Jharkhand, India.
ern Calcareous Alps). Bull Eng Geol Environ. https://​doi.​org/​10.​ Singh PK, Kainthola A, Panthee S, Singh TN (2016) Rockfall analy-
1007/​s10064-​019-​01640-w sis along transportation corridors in high hill slopes. Env Earth
Luo G, Xiewen H, Yingjin D et al (2019) A collision fragmentation Sci. https://​doi.​org/​10.​1007/​s12665-​016-​5489-5
model for predicting the distal reach of brittle fragmentable rock Singh PK, Kainthola A, Singh TN (2015a) Earthquake-induced rockfall
initiated from a cliff. Bull Eng Geol Environ 78:579–592. https://​ along cut slopes - a case study. In EUROCK 64th Geomech Col-
doi.​org/​10.​1007/​s10064-​018-​1286-6 loquium, Schubert Kluckner, pp. 1127–1132.
Mahanta B, Singh HO, Singh PK, Kainthola A, Singh TN (2016) Sta- Singh PK, Kainthola A, Singh TN (2015b) Rock mass assessment
bility analysis of potential failure zones along NH-305. Natural along the right bank of river Sutlej, Luhri, Himachal Pradesh.
Hazards, India. https://​doi.​org/​10.​1007/​s11069-​016-​2396-8 India Geomat Nat Hazards Risk 6(3):212–223
Mavrouli O, Corominas J (2017) Comparing rockfall scar volumes Singh R, Kumar R, Singh TN (2014) Stability evaluation of road-cut
and kinematically detachable rock masses. Eng Geol 219:64–73 slopes in the Lesser Himalaya of Uttarakhand, India: conventional
Mavrouli O, Corominas J (2020) Evaluation of maximum rockfall and numerical approaches. Bull Eng Geol Environ 73:845–857
dimensions based on probabilistic assessment of the penetra- Smith JV (2006) Geological contributions to engineering investigation
tion of the sliding planes into the slope. Rock Mech Rock Eng. of rock mass. ASEG Extended Abstracts 2006:1–4. https://d​ oi.o​ rg/​
https://​doi.​org/​10.​1007/​s00603-​020-​02060-z 10.​1071/​ASEG2​006ab​164
Mavrouli O, Corominas J, Jaboyedoff M (2015) Size distribution for Srikantia SV, Bhargava ON (1998) Geology of himachal pradesh. Geo-
potentially unstable rock masses and in situ rock blocks using logical Society of India, Bangalore

13
5368 A. K. Singh et al.

Turner AK, Jayaprakash GP (2012) Introduction. In: Turner AK, Windsor CR (1997) Rock reinforcement systems. Int J Rock Mech Min
Schuster RL (eds) Rockfall characterisation and control. Transpor- Sci 34:919–951
tation Research Board, Nat Aca of Sci, Washington DC, pp. 3–20 Wyllie DC (2014) Rock fall engineering. CRC Press, Taylor & Francis
Vishal V, Siddique T, Purohit R, Phophliya MK, Pradhan SP (2017) Grp, Boca Raton, FL
Hazard assessment in rockfall-prone Himalayan slopes along Zhang L (2016) Engineering properties of rocks, 2nd edn. Elsevier,
National Highway-58, India: rating and simulation. Nat Hazards London
85(1):487–503
Wang X, Frattini P, Stead D, Sun J, Liu H, Valagussa A, Li L (2020)
Dynamic rockfall risk analysis. Eng Geol. https://​doi.​org/​10.​
1016/j.​enggeo.​2020.​105622

13

You might also like