ORIGINAL PAPER Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks Nicola Nocilla Aldo Evangelista Anna Scotto di Santolo Received: 28 September 2007 / Accepted: 3 March 2008 / Published online: 13 June 2008 Ó Springer-Verlag 2008 Abstract In recent years, rock fall phenomena in Italy have received considerable attention for risk mitigation through in situ observations and experimental data. This paper reports the study conducted at Camaldoli Hill, in the urban area of Naples, and at Monte Pellegrino, Palermo, Italy. The rocks involved are volcanic Neapolitan yellow tuff (NYT) in the former area and dolomitic limestone in the latter. Both rocks, even though with different strength characteristics, have shown a significant tendency towards rock fragmentation during run out. This behavior was first investigated by comparing the volumes of removable blocks on the cliff faces (V 0 ) and fallen blocks on the slopes (V f ). It was assumed that the ratio V f /V 0 decreases with the distance (x f ) from the detachment area by an empirical law, which depends on a coefficient a, correlated with the geotechnical properties of the materials involved in the rock fall. Finally, this law was validated by observation of well- documented natural rock falls (Palermo) and by in situ full-scale tests (Naples). From the engineering perspective, consideration of fragmentation processes in rock fall modeling provides a means for designing low-cost mitigation measures. Keywords Rock falls Hard and soft rocks Fragmentation In situ full-scale tests N. Nocilla Dipartimento di Ingegneria Strutturale e Geotecnica, Universita ` degli Studi di Palermo, Viale delle Scienze, 90128 Palermo, Italy A. Evangelista A. Scotto di Santolo (&) Dipartimento di Ingegneria Geotecnica, Universita ` degli Studi di Napoli Federico II, Via Claudio 21, 80125 Naples, Italy e-mail: [email protected]A. Evangelista e-mail: [email protected]123 Rock Mech Rock Eng (2009) 42:815–833 DOI 10.1007/s00603-008-0006-0
19
Embed
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ORI GIN AL PA PER
Fragmentation during Rock Falls: Two Italian CaseStudies of Hard and Soft Rocks
Nicola Nocilla Æ Aldo Evangelista ÆAnna Scotto di Santolo
Received: 28 September 2007 / Accepted: 3 March 2008 / Published online: 13 June 2008
� Springer-Verlag 2008
Abstract In recent years, rock fall phenomena in Italy have received considerable
attention for risk mitigation through in situ observations and experimental data. This
paper reports the study conducted at Camaldoli Hill, in the urban area of Naples,
and at Monte Pellegrino, Palermo, Italy. The rocks involved are volcanic Neapolitan
yellow tuff (NYT) in the former area and dolomitic limestone in the latter. Both
rocks, even though with different strength characteristics, have shown a significant
tendency towards rock fragmentation during run out. This behavior was first
investigated by comparing the volumes of removable blocks on the cliff faces (V0)
and fallen blocks on the slopes (Vf). It was assumed that the ratio Vf/V0 decreases
with the distance (xf) from the detachment area by an empirical law, which depends
on a coefficient a, correlated with the geotechnical properties of the materials
involved in the rock fall. Finally, this law was validated by observation of well-
documented natural rock falls (Palermo) and by in situ full-scale tests (Naples).
From the engineering perspective, consideration of fragmentation processes in rock
fall modeling provides a means for designing low-cost mitigation measures.
Keywords Rock falls � Hard and soft rocks � Fragmentation � In situ full-scale tests
N. Nocilla
Dipartimento di Ingegneria Strutturale e Geotecnica, Universita degli Studi di Palermo,
Viale delle Scienze, 90128 Palermo, Italy
A. Evangelista � A. Scotto di Santolo (&)
Dipartimento di Ingegneria Geotecnica, Universita degli Studi di Napoli Federico II,
In the last few years, systematic investigations of the downhill propagation of single
large blocks or rock falls have been carried out in two locations in Italy, along the
Camaldoli Hill (Soccavo site) in Naples, and on Monte Pellegrino (Acquasanta site)
in Palermo (Fig. 1). Rock falls involve Neapolitan yellow tuff (NYT) in the former
area, and limestone and calcareous dolomite in the latter area. These rocks have
very different fabric and mechanical behaviors (Evangelista et al. 2000).
Survey techniques and analyses based on terrestrial photogrammetry (Nocilla and
Urciuoli 1997; Nocilla et al. 1999), together with the morphology and structural
characteristics of the tuff and dolomitic limestone, were used to define removable
blocks1 in the detachment area (Evangelista et al. 2002, 2004). Measurements were
taken of the characteristics of the rock falls, and of the size, shape, and location of
the fallen blocks along the slope. At Camaldoli Hill, in situ full-scale tests were
carried out (Nocilla et al. 2003). Detailed measurements were taken of the traces of
a significant rock fall that took place on Monte Pellegrino in 1998.
Fig. 1 Location maps of the study areas
1 It is noted that a ‘‘removable block’’ identifies a standalone block on a slope and a block which has
already been dislodged and detached from the cliff face by movements of the order of as a little as 1 dm.
816 N. Nocilla et al.
123
The data show a reduction in block volume during run out. A similar empirical
relationship of volume reduction versus travel distance was proposed for the two
locations. Firstly, this relationship was evaluated by comparing the population
volumes of fallen blocks along the slope with the population of removable blocks on
the rock face. This relationship was then validated by using the results of in situ full-
scale rock fall tests at Camaldoli Hill and the traces of the landslide that occurred at
Monte Pellegrino.
This paper focuses on the phenomenon of rock fragmentation during rock falls by
means of observations and in situ full-scale tests carried out in two Italian sites with
different geological and geotechnical features. The study points out the relevance of
this phenomenon and its impact on the engineering design.
2 Geomorphological Characteristics of the Rock Fall Areas
2.1 Camaldoli Hill, Naples
The Camaldoli Hill (457.5 m a.s.l.) is the highest peak of the Phlegrean Fields and
originates from the collapse of the Ignimbrite caldera (39 ka) (Fig. 1). It is
composed of a succession of different pyroclastic strata deposited during eruptive
phases. Rock falls start at the cliffs of fractured NYT that are present at the peak and
at lower elevations along the slope. There are two subvertical strips of NYT hill
faces which vary in height between 25 and 50 m, from which blocks tend to detach
and propagate along the slope until they reach the urban areas at the toe of the hill.
The slope below, covered by debris, is steep down to an elevation of about
200 m, after which it becomes gentler and evolves into a succession of small
terraces used for agriculture and flower cultivation. The basal plain, where the
inhabited area is located, is flat and densely populated (districts of Soccavo and
Pianura). The wooded areas that once existed here have almost entirely been wiped
out by fires.
As already noted, NYT is present throughout the city of Naples; it is a soft rock,
with varying structure and mechanical characteristics, which consists of a cineritic
matrix with pumaceous and/or scoriaceous and lithic inclusions (Evangelista et al.
2000). The quantity and quality of the inclusions depend on the specific type of tuff
and, even within the same formation, on distance from the volcanic source. The
minerals in the cineritic matrix play an important role in the properties of the rock.
The most common minerals are phillipsite, cabasite, and sanidine.
The mesostructure of the tufaceous mass is characterized by a system of nearly
vertical discontinuities (set 1) that are normal (set 2 and 4) to the slope surface
(Fig. 2). These discontinuities delimit removable rock blocks with parallelepiped or
slab-like shapes. The base of such blocks consists of subhorizontal discontinuities,
dipping between 5� and 38�, which lead to planar sliding. Block size is determined
by the spacing, which ranges from a few decimeters up to 3–5 m in the subvertically
oriented discontinuities.
Besides the blocks that are removable from the cliff face and that are the result of
the unfavorable positioning of the discontinuities, other phenomena associated with
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks 817
123
instability are caused by particularly strong wind erosion at the toe of the cliff or
along very weak rock strata of the hill.
A total of 318 removable blocks, with volumes exceeding 0.5 m3, as shown in
Fig. 3, have been mapped on the cliff. There is a significant number of blocks, with
volumes ranging from 1 to 10 m3, which are generally cubic in shape. Blocks with
volumes exceeding 10 m3 are parallelepiped in shape, with heights greater than
their length or thickness. The geometric features are such that these blocks will
either slide or topple over.
Fig. 2 Pole density contour plot for Camaldoli Hill (Naples)
Fig. 3 Projection onto a vertical plane sub-parallel to the cliff face of removable blocks located in thecliff face below the Eremo dei Camaldoli
818 N. Nocilla et al.
123
2.2 Monte Pellegrino, Palermo
Monte Pellegrino, which consists of limestone and dolomitic rock, of the Upper
Triassic to the Lower Miocene, is bounded mainly by subvertical or sheer slopes
(Fig. 1). It has a highly complex structure, characterized by folds and faults, reverse
geological sequences, and a dense system of faults. At the toe of the mountain, a
large debris layer has accumulated over coastal quaternary deposits, forming a
35–40� slope near the cliff faces.
On the southern side of the mountain, which overlooks the Acquasanta
neighborhood, Via Bonanno is cut into the debris layer, whose characteristics vary
from area to area. Successful reforestation projects were started in the 1950s here
where the ground consists mainly of gravel and reddish, lime-matrix blocks with the
occasional block exceeding 2 m in diameter. The gravel is generally clumped
together and may even be cemented to varying degrees.
The hill face has an average dip direction of about 85�. There are five
predominant families of discontinuities which can be correlated with tectonic events
represented by N–S, ENE–WSW, and NW–SE oriented fault lines (Fig. 4). Three of
these families are characterized by dips greater than 5�, while the discontinuities of
another two families vary in dip between 30� and 50�.
The rock mass is subdivided into blocks of various shapes and sizes (also
removable) which, in the cortical area of the larger mass, are frequently dislodged
due to displacements of the order of centimetres or even decimetres, or as a result of
rotation around the lower edge.
A total of 554 removable blocks with volumes greater than 0.5 m3 were counted
by terrestrial photogrammetry surveys in the cliff faces above Via Bonanno, the
Fig. 4 Pole density contour plot for Monte Pellegrino (Palermo)
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks 819
123
road providing access to Monte Pellegrino (Nocilla et al. 1999). Figure 5 shows the
blocks identified in the face below the Grattarola Peak.
2.3 Consideration of the Geotechnical Properties of the Two Rocks
The mechanical properties of the NYT and the Palermo limestone have been the subject
of numerous studies carried out at the Geotechnical Engineering Department of the
‘‘Federico II’’ University of Naples, and at the Department of Structural and
Geotechnical Engineering of the University of Palermo. These studies show that the
former are low to very low strength rocks, while the latter are medium to medium-high
strength rocks. The two rock materials, classified by uniaxial compression strength
(rc B 10 MPa for the NYT and rc & 100 MPa for the limestone) and the elastic
modulus (E B 10,000 MPa for NYT and E & 100,000 MPa for limestone), fall within
the medium-to-high modulus ratios according to Deere and Miller (1966), Fig. 6.
3 Rock Fall Phenomena
As described above, the tests and observations carried out show a strong tendency
for the rock blocks to break during run out. The reduction in volume which takes
place is shown by comparing the removable blocks that have already broken away
from the mountain face with those present both along the slopes of Camaldoli Hill
and Monte Pellegrino. A comparison of the volumes of removable and fallen blocks
is given below.
3.1 Rock Fall Propagation at the Soccavo Site (Camaldoli Hill, Naples)
Topographic survey techniques and terrestrial photogrammetry were used to detect
and catalogue 120 fallen blocks along the slope (Fig. 7), with volumes equal to or
exceeding 1 dm3; only two blocks had a volume greater than 30 m3.
Fig. 5 Removable blocks identified through terrestrial photogrammetry and shown in projection onto avertical plane sub-parallel to the mountain face below Grattarola peak
820 N. Nocilla et al.
123
About 30% of the blocks (with volumes ranging from 0.1 to 7 m3) came to rest
on a 40� slope between 300 and 390 m a.s.l., close to the hill face (within 35 m).
About 70% reached elevations below 250 m a.s.l. These blocks can be divided into
the following groups:
• 21% (volumes between 0.2 and 0.8 m3) came to rest between 200 and 240 m
elevation a.s.l., where the slope is about 35�• 26% (maximum volume about 7 m3) came to rest between 160 and 200 m
elevation a.s.l.
• 23% (the remainder, with volumes between 1 dm3 and about 9 m3) remained in
motion until they reached an elevation of about 120 m, propagating to a
maximum distance from the hill face of about 250 m.
The points of rest of the latter group are located in agricultural zones with thickly
wooded areas where the slope has an average dip of 10–15�, with several plain-like
terraces.
Histograms of the removable blocks on the hill face (295 of the 318 identified)
and of the fallen blocks (120) are shown in Fig. 8a, b. The blocks along the slope
tend to be smaller: only about 34% of the entire population had volumes greater
than 2.5 m3. On the hill face, however, removable blocks predominate with
volumes above 2.5 m3, corresponding to about 70% of the sample. Cumulative
frequency diagrams, shown in Fig. 9a, highlight the different volume distribution
on the hill face and along the slope, with smaller volumes being predominant in
the latter.
Fig. 6 Mechanical properties ofthe NYT and of the Palermolimestone
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks 821
123
3.2 Rock Fall Propagation at the Via Bonanno Site (Monte Pellegrino, Palermo)
The points of rest of the fallen blocks not connected to the slope were detected
topographically (Fig. 10). Two main zones were identified:
• Zone 1, below the Volo dell’Aquila Peak and near the S. Maria dei Rotoli
cemetery, with 68 blocks
• Zone 2, below the Grattarola Peak, which embraces the area of the October 1998
landslide, and includes 205 blocks.
Both areas, at the toe of hill, with elevations of 170–190 m and 90–120 m a.s.l.,
respectively, have the following land features:
• slope between 35� and 40� for a distance of about 60 m from the hill face
• slope of about 31�–32� up to 140 m from the hill face
• slope of about 10�–15� beyond (in the vicinity of the inhabited areas).
In Zone 1, below the Volo dell’Aquila Peak, where the rock cliff is 170–190 m in
height, the blocks were distributed along the slope in the following way:
• a few blocks with volumes just below or greater than 1 m3 close to the hill face
• more than 60% of the fallen blocks had rolled to where the hill slope drops to
10–15�• only a few blocks were present in the larger part of the slope, i.e., between the
hill face and its lower section.
Fig. 7 Plan view of the Camaldoli Hill, Soccavo side, with the positions of fallen blocks along the slopeand the location of the field test site
822 N. Nocilla et al.
123
In zone 2, below the Grattarola Peak, where the cliffs reach heights of 90–120 m,
the blocks were distributed almost uniformly along the slope for a distance of 120 m
from the face, with only a few blocks reaching greater distances.
Figure 11a shows the volume histograms for 548 out of the 554 removable
blocks identified on the hill face, while Fig. 11b shows the histogram of the 302
fallen blocks. Figure 11a does not include six boulders with volumes greater than
50 m3, four of which had volumes of about 110, 137, 183, and 300 m3 . It may be
observed that blocks with volumes less than 2.5 m3 predominate on the hill face
(57%) and along the slope (78%). Also at this site, the blocks remaining on the hill
face outnumber those along the slope (fallen blocks).
Fig. 8 Volume histograms of removable blocks on the Camaldoli Hill face, Soccavo side (a) and ofblocks fallen on the slope (b)
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks 823
123
By comparing Figs. 8, 9, and 11 the following observations can be made:
• The population of removable rocks on the hill face is always more numerous
than the population of fallen rock; the difference between the two is larger for
the Camaldoli site.
• The number of fallen blocks with volumes greater than 5 m3 is larger at the
Monte Pellegrino site.
Overall, block fragmentation at the Camaldoli site is greater than that on Monte
Pellegrino, which is consistent with the mechanical properties of the rock materials
involved.
4 Volume Reduction Analysis
The volume distribution of the removable blocks depends essentially on the
mesostructure of the rock mass, a characteristic that can be considered as having
remained virtually unchanged in recent millennia. For this reason, the present
population of rock blocks on the hill face may be assumed to be the same as the
population of fallen blocks on the slope. As already noted, considering the different
Fig. 9 Cumulative frequency diagrams of fallen blocks on the slope and removable blocks on the hillface: a Camaldoli Hill; b Monte Pellegrino
824 N. Nocilla et al.
123
volume composition on the hill face and along the slopes, it follows that the sizes of
removable blocks on the hill face are larger than those estimated at the point of rest.
It may be suggested that the volume reduction is due to impact processes and to
erosion during run out (rolling and/or sliding phases).
The result of these phenomena intensifies with increasing distance between hill
face and point of rest. It can be assumed that the volume (Vf) at the point of rest
depends on the initial volume (V0) on the hill face, and that the Vf/V0 ratio is an
inverse function of the distance from the hill face (xf). The following empirical
reduction law can be suggested:
Vf ¼V0
ð1þ xfaÞ; ð1Þ
where Vf is the volume at a certain point of rest at a distance xf from the hill face,
V0 is the initial volume on the hill face, xf is the distance from the hill face to the
point of rest, and a is a reduction coefficient [L-1] correlated to the mechanical
properties of the materials involved in the rock fall.
Fragmentation intensity increases as a increases. The first approach for
evaluating the a coefficient was obtained as follows:
• Different values were assumed for the coefficient a, and hypothetical initial
volumes V0 were evaluated starting from Vf and xf.
Fig. 10 Plan view of Monte Pellegrino, Acquasanta side, with locations of fallen blocks on the slope anddelineation of the 1998 rock fall zone
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks 825
123
• The cumulative frequency of these volumes was then calculated and compared
to the removable blocks on the hill face to find the best fit.
In Fig. 9 (a cumulative frequency diagram of the fallen blocks along the slope),
in situ removable blocks and those calculated by means of (1) are shown for the two
sites. The calculated V0 values reach a best fit with in situ values for a coefficient aequal to 0.015 m-1 for the calcareous dolomite, and a = 0.02 m-1 for the NYT.
Equation 1 was then validated by observation of well-documented natural rock
falls at Monte Pellegrino (Palermo) and by the results of in situ full-scale tests
carried out along the Camaldoli slope (Naples).
Fig. 11 Volume histograms of removable blocks identified on the hill face of Monte Pellegrino, aboveVia Bonanno and the Acquasanta neighborhood (a), and of fallen blocks present along the slope (b)
826 N. Nocilla et al.
123
5 Full-Scale Testing at Camaldoli Hill (Naples)
The test site was set up along the eastern border of Camaldoli hill (Fig. 6). Distance-
measuring posts and four video cameras synchronized with a longitudinal time code
(LTC) system were placed on the slope. Three of these cameras were positioned
according to the maximum dip of the slope for lateral viewing. The fourth camera
was located further down in order to provide a frontal view of the slope.
Along the slope, which dips 40� at the test site, a pyroclastic debris layer is
present, as well as thick and uniform vegetation reaching a height of about
80–100 cm. There is a copse of beech trees at about 300 m from the hill face.
When blocks were already detached from the hill face, the dislodgement process
was carried out manually by using a lever wedged into a discontinuity (removable
block) or by isolating rock portions with a precut process carried out by placing
expansive mortar in neighboring vertical fissures. The initial volume of the blocks
ranged from 1 to 12 m3. The trajectories down the slope, recorded by the cameras,
were analyzed to determine block velocities, heights of bounce, and run-out
distances. For each major impact of the block, the translational and rotational
velocities before and after the impact, and consequently the restitution coefficients,
were evaluated.
Generally an ‘‘explosion’’ accompanied the initial impact of each block, which
was characterized by a rapidly expanding dust cloud from which blocks with fan-
like trajectories initiated after a few tenths of a second. The fragments traveled
distances less than those predicted by the numerical analyses (Richards 1988;
Labiouse 2004; Pfeiffer and Bowen 1989); almost all of the rock fragments were
stopped by the beech copse (Nocilla et al. 2003).
Video recordings of the rock falls showed that the impact down the slope gave
rise to the subdivision of each block into fragments and the formation of fan-like
trajectories. The fragments fan out with a width of about 30–35% of the travel
distance. This width is 2–20% greater than the values reported in the literature
(Azzoni and De Freitas 1995).
Imaging techniques made it possible to determine the approximate size of the
blocks prior to and immediately after the impact. For the dislodged blocks, Fig. 12
shows the ratio of the largest volume fragment (V) to the initial volume (V0) versus
the travel distance (x) from the starting point. The same figure highlights the rapid
volume reduction, equal to about 80% of the initial volume, during the first 50 m
travelled. The data were interpolated by using the following relationship:
V
V0
¼ 1
1þ xað Þ ; ð2Þ
which is the same as in Eq. 12, assuming different a values.
The values of a provide a satisfactory interpolation with the experimental data,
ranging from 0.02 to 0.06 m-1 with a mean value of 0.04 m-1. There was only one
2 Equation 1 can be used to determine the initial theoretical volume starting from the final volume of the
blocks along the slope. From Eq. 2 the block progressive volume reduction under propagation can be
derived.
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks 827
123
test, trial number 16, in which the reduction coefficient a was about 0.008 m-1. For
this test, the final volume Vf was found to be 40% of the initial volume V0, which
was the smallest volume reduction process in all the tests performed along the
Camaldoli Hill.
The landslides initiated in the experimental field were simulated with a three-
dimensional numerical code (Scioldo 2000). The numerical analysis conducted,
which simulated the beech copse as damping elements (E = 200 kJ), showed that
masses with constant volume equal to the initial value of V0 would have continued
their path and reached the rockfall protection barrier at the bottom of the hill. At any
rate, the insertion of dissipative elements proves incapable of impeding the masses
when considering the initial volume in the simulations. Good agreement between
the observed and calculated arrest points is obtained only by assuming partially
reduced mass volumes according to Eq. 2.
6 The Monte Pellegrino Landslide (Palermo)
The significant landslide that took place at Monte Pellegrino in October 1998 was
also analyzed. A boulder of 64 m3 broke away from the mountain side at an
elevation of 228 m a.s.l. (Fig. 5). The outline and initial volume V0 of the in-face
blocks were estimated using photographic measurements taken in 1997. The boulder
had already broken down into numerous blocks within the face, the largest of which
had a volume of 38 m3 (Fig. 13).
During the slide, the rock broke into 43 fragments with volumes ranging from 0.2
to 23 m3 (Fig. 14), and numerous smaller fragments with volumes less than 0.2 m3.
Most of the fragments (around 60%) came to rest close to the hill face at a distance
of about 40 m; 24% came to rest near Via Bonanno at about 120 m from the hill
face; nearly 8% reached the slope between two of the road hairpin curves. One
fragment (number 41), with a final volume Vf of 23 m3, and thus a subfragment of
Fig. 12 Volume reduction versus travel distance
828 N. Nocilla et al.
123
the highest initial volume (V0 = 38 m3), reached the edge of an old abandoned
quarry at a distance of about 200 m from the toe of the hill face.
This block followed a straight trajectory, shearing tree trunks and branches in its
path, and left a well-defined circular wake, 1–2 dm deep, and a slight bulge at its
perimeter. Figure 13 shows the trajectory which was inferred from the outline of the
impact craters, from the heights of broken tree trunks or snapped/scraped branches
sheared by the flying boulder, from trees contiguous to the trajectory, and from
traces of sliding in the area immediately before the final point of rest.
The boulder mostly flew through the air, bouncing and hitting the slope. The
separation of fragments from the initial rock volume (V0 = 38 m3) probably
occurred near the site of first impact at the toe of the hill face, which is covered by a
debris layer consisting primarily of rock blocks. Numerous rock fragments with
signs of recent fractures were identified within a 25–30 m radius.
The available data include the initial volume V0 of the blocks involved in the
slide as well as the total volume V0tot, the initial elevation, the point of rest of the
main fragments, and their corresponding volumes (see the table in Fig. 14).
Fig. 13 Monte Pellegrino: trajectory of the V0 = 38 m3 fragment from the 1998 rock fall event
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks 829
123
Starting from fragment number 41, with a volume of 23 m3, an estimate was
made of the different volumes of the block during its trajectory by assuming:
Vi ¼ Vi�1 � ti; ð3Þ
where ti is the volume of the fragment found along the trajectory, Vi-1 is the
volume of the block before impact i, and Vi is the residual volume after the impact.
Figure 15 shows the Vi values normalized with respect to V0tot = 64 m3, versus
the distance of the point of rest from the hill face. The data may be interpolated
Fig. 14 Monte Pellegrino: location of the fragment points of rest from the 1998 rock fall event
830 N. Nocilla et al.
123
using Eq. 1 by assuming a = 0.01 m-1. The volume of the same fragment number
41 has been normalized with respect to its probable initial volume V0 = 38 m3.
Equation 1 yields an a value equal to 0.0035 m-1.
In the first case (V0tot = 64 m3), a volume reduction process occurs which is
slightly smaller than expected by comparing the rock populations on the slope with
those on the hill face (a = 0.015 m-1, Fig. 8b). The application of the value
a = 0.0035 m-1 to volumes Vf of blocks along the slope obtained in the second
case (V0 = 38 m3) would yield a theoretical population of removable blocks on the
hill face which is only significant for the larger volumes, for example, those greater
than 10 m3.
7 Conclusions
The analysis of the populations of blocks on Camaldoli Hill in Naples, and on
Monte Pellegrino in Palermo shows a progressive reduction in size during rock fall,
from detachment to the final point of rest. The phenomenon appears to be well
described by the reduction law given in Eq. 1. The final volume is found to depend
on travel distance and on the coefficient a, which can be correlated with the
mechanical characteristics of the rock mass and of the hill slopes.
Full-scale rock fall tests carried out on the NYT (rc B 10 MPa) at Camaldoli
Hill, and at a landslide which occurred in 1998 in the Monte Pellegrino limestone
(rc * 100 MPa), were analyzed. In both cases the rock systematically shows a high
level of boulder fragmentation right from the initial impacts with the hill face and
talus debris layers, in addition to a fan-like trajectory of particle dispersion along the
slope. The fragments, whose volumes become smaller as travel distance increases
and which reach sizes that are only 10–20% of the initial volume in the case of
NYT, are generally distributed over a broad front.
Fig. 15 Variations in the volume of the cluster of blocks versus travel distance during the 1998rockslide. The curve where a = 0.0035 m–1 refers to the initial volume V0 = 38 m3. The curve wherea = 0.01 m-1 refers to the total fallen volume V0
tot = 64 m3
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks 831
123
The data from the Camaldoli test site (NYT) are in agreement with Eq. 1, but
highlight a non-negligible variation in the parameter a, between 0.06 and 0.008 m-1.
The back-analysis of the Monte Pellegrino landslide (limestone) confirms the
validity of Eq. 1 and indicates that the fragmentation processes, with values of aranging from 0.01 to 0.0035 (depending on the assumed initial volume), are generally
much less intense than those recorded for the NYT.
The experimental observations lead to the conclusion that the forecast of rock fall
phenomena can be described more realistically by considering the reduction in
volume of single blocks versus their travel distance. It is suggested that the
reduction law to be adopted, when field tests are not feasible, may be that obtained
by comparing the volumes of the fallen blocks present along the slopes with those
detached but still lying on the hill face. The range of a values to be considered is
given in Fig. 16, where higher-intensity fragmentation processes were not taken into
account (a[ 0.02 for NYT and a[ 0.01 for limestone).
It is noted that passive protection barriers may be designed more rationally and
economically by considering the fragmentation of the blocks that occurs during
impacting and subsequent sliding and rolling. In other words, based on our
experience, in design procedures, smaller volumes can be considered than the
volumes of removable blocks lying on the hill face. Given our current level of
knowledge in this respect, the choice of the a reduction coefficient to be used needs
to be based on field tests and cautious evaluations.
Acknowledgment This research was partly funded by MURST (Ministry for Universities and
Research), Italy, within PRIN 2005 2005085322_004 (National Research Project, 2005).
References
Azzoni A, De Freitas MH (1995) Experimentally gained parameters, decisive for rockfall analysis. Rock
Mech Rock Eng 28(2):111–124
Fig. 16 Ranges for the reduction coefficient a for the NYT from Camaldoli hill and for the calcareousdolomite from Monte Pellegrino
832 N. Nocilla et al.
123
Deere DU, Miller RP (1966) Engineering classification and index properties for intact rock, Tech. Rept.
N0.AFWL-TR–65–116, Air Force Weapons Lab., Kirtland Air Force Base, New Mexico In: Stagg
KG, Zienkiewicz OC (eds) Rock mechanics in engineering practice. Wiley, New York. 1972
Evangelista A, Aversa S, Pescatore TS, Pinto F (2000) Soft rocks in southern Italy and role of volcanic
tuffs in the urbanization of Naples. In: Proceeding of 2nd symposium on geotechnics of hard soils-
soft rocks, Naples 1998, pp 243–1267
Evangelista A, Scotto di Santolo A, Lombardi G, Nocilla N, Danzi M, Vivano D (2002) Sui fenomeni di
instabilita nei costoni di tufo giallo napoletano. In: Atti del XXI Convegno Nazionale di Geotecnica,
L’Aquila, 11–14 Settembre 2002, pp 57–164
Evangelista A, Scotto di Santolo A, Nocilla N, Messineo S, Feola A (2004) Sperimentazione ed
osservazioni in sito sulla caduta massi dai costoni di tufo giallo napoletano. In: Atti del XXII
Convegno Nazionale di Geotecnica, Palermo, 22–24 Settembre 2004, pp 99–106
Labiouse V (2004) Fragmental rockfall paths: comparison of simulations on Alpine sites and
experimental investigation of boulder impacts. In: Proceeding 9th international conference on
landslides, Rio de Janeiro, 1:457–466
Nocilla N, Urciuoli G (1997) Stabilita dei pendii in roccia: Rilievi strutturali e spostamenti ammissibili.
Hevelius Edizioni, Benevento
Nocilla N, Viviano D, Aversa S (1999) Applicazione della fotogrammetria al rilievo strutturale di costoni
rocciosi. In: Atti del XX Convegno Nazionale di Geotecnica, Parma 1999, pp 231–238
Nocilla N, Evangelista A, Scotto di Santolo A, Susani A (2003) Analysis of rockfall phenomena in soft
and hard rocks. In: Picarelli L (ed) Proceeding of international conference on fast slope movements
prediction and prevention for risk mitigation, Naples, May 11–13 2003, pp 379–382
Pfeiffer TJ, Bowen TD (1989) Computer simulation of rockfalls. Bull Ass Eng Geol 26(1):135–146
Richards LR (1988) Rockfall protection: a review of current analytical methods. In: Proceeding of 2nd
conference on rock mechanics and rock engineering, MIR 88, Turin, pp 1–13
Scioldo G (2000) ISOMAP & ROTOMAP 3D Surface modelling & Rockfall analysis. Geo&Soft, Torino
Fragmentation during Rock Falls: Two Italian Case Studies of Hard and Soft Rocks 833