Engineering Geology 187 (2015) 84–97

Contents lists available at ScienceDirect

Engineering Geology
journal homepage: www.elsevier.com/locate/enggeo

SPT blow count vs. shear wave velocity relationship in the structurally
complex formations of the Molise Region (Italy)
Silvia Fabbrocino a,⁎, Giovanni Lanzano b, Giovanni Forte b,
Filippo Santucci de Magistris b, Giovanni Fabbrocino b
a
University of Naples Federico II, DiSTAR Department of Earth Sciences, Environment and Georesources, Naples, Italy
b
University of Molise, DiBT Department, Structural and Geotechnical Dynamics Laboratory, StreGa, Campobasso, Italy

a r t i c l e i n f o a b s t r a c t

Article history: The relationship between the shear wave velocity (VS) and the blow number of the Standard Penetration Tests
Received 11 November 2013 (NSPT) was investigated with reference to the structurally complex formations located in a wide sector of the
Received in revised form 4 November 2014 Southern Apennine chain (Molise Region, Italy). Novel NSPT–VS correlations were developed, based on a balanced
Accepted 25 December 2014
number of in situ tests coming from the seismic microzonation of the municipalities of Campobasso District and
Available online 30 December 2014
the seismic vulnerability assessment of strategic road infrastructures. It was proven that available relationships
Keywords:
were not able to fit the data in the area of interest, even making the analysis according to the soil grain size clas-
SPT-N sification. This circumstance was related to the nature of the geological formations, which are chaotically built
Shear wave velocity and affected by complex structural–textural and lithological characteristics. Hence, in this paper, a rational meth-
Structurally complex formations odology for an integrated geological and geotechnical assessment of the available field data based on
Empirical correlations lithostratigraphical features was presented and discussed. Three original empirical NSPT–VS power function rela-
Seismic and geotechnical characterization tionships were derived and distinguished depending on the main lithofacies recognized in the Campobasso Dis-
Molise (Italy) trict, corresponding to Structured Soil Deposit (SSD), All Soils (AS) and Weathered Soils (WS). Their validation
through statistical assessment of data confirmed the relevance of the results and the advantages related to
their use at regional scale, based on an integrated geological and geotechnical approach.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction characterization influences the nature and the magnitude of the seismic
action at the foundation level of the structures (EN1998-5, 2004). In
The solution of any problem in civil engineering usually requires the particular, the profile of the shear wave velocity VS in the ground is
acquisition of a certain amount of data, such as geometrical size and regarded as the most reliable predictor of the site-dependent character-
shape of the construction, extension of the significant area for design istics of the seismic action at stable sites.
and/or assessment of the structures, material properties, intensity mea- Design codes recommend a combination of laboratory tests and in-
sures of natural events and so on. In some circumstances, the data acqui- situ penetration resistance or other geotechnical properties, allowing
sition process is not difficult. Conversely, in the case of geotechnical for the scatter of such correlations. The present paper deals with the re-
applications, the part of the soil that interacts with the construction lationships between the blow number of the Standard Penetration Tests
and/or area of interest needs to be characterized in terms of geometric, (NSPT) and the shear wave velocity (VS), in a complex geological context.
hydraulic and physical–mechanical properties, generally all a-priori un- A careful review of the technical literature demonstrated that an opti-
known and affected by large uncertainties (EN-1997-1, 2004). mal correlation between the above-mentioned parameters does not
Type, quality and quantity of the geotechnical investigations need to necessarily exist and is often site-specific. However, simplified proce-
be planned according to the type of construction, its performance class dures for relevant geotechnical mechanical parameter estimation can
and the complexity of the subsoil (EN-1997-1, 2004; NTC, 2008) and be advantageous in a variety of real applications, especially when they
should lead to definition of a reliable geotechnical model that is one of are distributed on large areas.
the key issues for a successful solution of any problem civil engineering. In the following, a novel methodology was proposed and imple-
This topic is particularly relevant in seismic areas, where the subsoil mented with reference to the lithostratigraphical features of the
Campobasso District, Molise Region (Italy), where structurally complex
formations crop out. A comparative analysis with the conventional
⁎ Corresponding author at: DiSTAR Department, University of Naples Federico II, Via
grain size approach was also discussed in order to validate the outcomes
Mezzocannone, 8, 80134 Napoli, Italy. of the study. Results appear to be encouraging, even though not com-
E-mail address: silvia.fabbrocino@unina.it (S. Fabbrocino). prehensive, as the proposed approach and the resulting empirical

http://dx.doi.org/10.1016/j.enggeo.2014.12.016
0013-7952/© 2014 Elsevier B.V. All rights reserved.
 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97 85

relationships could apply to larger areas located in the Apennine chain proposed correlations for soils with different features on empirical
characterized by similar structurally complex formations. bases. This is possible since the correlated parameters (Vs and NSPT)
have different mechanical meanings. It is commonly established that
2. NSPT–VS empirical correlations: basics and limitations for geomaterials initial stiffness and strength are dependent by the
same intrinsic and state parameter, including their lithological nature
It is known that in a porous elastic solid containing a compressible and void ratio, effective confining stress and stress-history. Other rele-
viscous fluid, the shear wave velocity of elastic waves, propagating in vant aspects related to the significance of these empirical formulations
the solid component, is directly correlated with the shear modulus at are reported below:
small strain of the material G0 (e.g., Biot, 1956) according to the classical
relationship: 1. According to the standard for the execution of the SPT, NSPT is the
total number of blows needed to obtain a sample tube penetration
G0 ¼ ρ
VS
2
ð1Þ of the last 30 cm out of the total 45 cm in each test step. The penetra-
tion of the sampler is obtained from the hammering of a 63.5 kg
where ρ is the soil mass density. weight from a constant free falling height (76 cm) onto an anvil
In scientific and professional literature, books (e.g., Kramer, 1996), mounted on top of the drill rods. Despite this standardization of the
standards (e.g., ASTM D7400-08 for Down-Hole tests, ASTM D4428 for test, the results might be influenced by some features of the equip-
Cross-Hole), guidelines or manuals (e.g., Butcher et al., 2005 for Seismic ment and the procedures to perform the in situ test, as diameter of
Cone Penetration Tests and Seismic Dilatometer Tests) provide specific the borehole, the length of the shafts and the energy content of the
recommendations to perform accurate measures of the shear wave ve- hammer. A way to avoid these false responses of the test is to correct,
locity, especially when in-hole techniques are adopted. and eventually normalize, the number of blows according to the pro-
Instead, the Standard Penetration Test (SPT) is performed during the cedures mainly established for liquefaction analyses (see for in-
soil boring to obtain an indirect measure of the soil resistance, as well as stance, Idriss and Boulanger, 2004);
a disturbed drive sample (split barrel type). The test was introduced by 2. In saturated fine graded soils, the dynamic loading of the blows in-
the Raymond Pile Company in 1902 and remains today as the most duces instantaneous undrained conditions, therefore, test result
common in-situ test worldwide (Mayne et al., 2001; ASTM might depend also on the amount of the excess pore water pressure
D1586-11). The test can be performed in a wide variety of soil types, development. In very dense fine or silty sand usually a correction ap-
as well as weak rocks, even though it is not particularly useful in the plies to NSPT according to the classical recommendations by Terzaghi
characterization of gravel deposits or clays. In the first case the sampler and Peck (1948);
can become obstructed by soil particles giving erroneous high values of 3. The NSPT–VS correlations were generally site dependent: the empiri-
NSPT. In the second case, results might be inconsistent with actual in-situ cal correlations should be obtained in a homogeneous geologic con-
condition (Kulhaway and Maine, 1990) or poorly correlated with phys- text and used only in that considered environment. This is why the
ical–mechanical soil properties (Reid and Taylor, 2010). As a matter of most recent correlations were obtained for specific and limited
fact, standard penetration tests cause the soils to fail and even though areas. However, Anbazhagan et al. (2012) tried to correlate the re-
theoretical analyses of the test were performed (Schmertmann, 1979; sults of SPT test with G0 in any regional context, merging the data-
Schmertmann and Palacios, 1979) sets of empirical correlations lead base employed to obtain the previous correlations. An intermediate
to the estimation of the soil parameters from the knowledge of the solution between these two outermost approaches could be the set-
blow number — see again Kulhaway and Maine (1990) for a compre- ting of empirical NSPT–Vs correlation for specific geologic context in
hensive set of relationship between NSPT and physical and mechanical combination with the widening of the results to similar areas, in
soil parameters. order to extend and validate the equations.
Therefore, a correlation between NSPT and VS represents an indirect
correlation between initial strength and stiffness for a soil, hence it 3. NSPT–VS empirical correlations: literature review
could not always be a good correlation. It relies on the idea that, refer-
ring to a typical stress–strain curve for a soil, the larger the initial stiff- A careful analysis of available empirical correlations in the technical
ness, the higher is the strength. literature has been carried out. The main outcomes are summarized in
However, it is well known that: 1. the initial soil stiffness is strictly Table 1, which extend a similar table reported in Akin et al. (2011).
dependent upon the soil microstructure – see for instance Vinale et al., Here the most common and relevant existing correlations were
1999; Santucci de Magistris and Tatsuoka, 2004 – while its influence reviewed with regard to their key distinctive features: country/location;
on the stress–strain response tends to reduce as the strain increases, al- main geological characteristics of soils; basic geotechnical classification
most disappearing at critical state (e.g., Santucci de Magistris et al., of soils; regression function; total number of measured data; coefficient
1998); 2. the overall stress–strain curve for a soil is affected by the stress of correlation, r or coefficient of determination, R2.
dilatancy and its volumetric behavior; and 3. the overall stress–strain It is evident that the basic idea and the early researches were origi-
curve and particularly the failure conditions are affected by the drainage nally carried out in Japan, but later the topic became of some interest
conditions. in seismic prone regions worldwide. The first attempts to correlate the
In spite of all, Tatsuoka and Shibuya (1991) determined under labo- shear wave velocity Vs (or the initial shear stiffness G0, see Eq. (1))
ratory controlled conditions a set of relationships between the maxi- and the number of blows date back to the early '70s and were based
mum Young's modulus Emax, defined at very small strain and free on several datasets collected in different regions of Japan (Imai and
from bedding errors, and the maximum deviator stress qmax. A correla- Yoshimura, 1970; Ohba and Toriumi, 1970; Shibata, 1970; Ohta et al.,
tion between the ratio Emax/qmax and the deviator stress qmax was ob- 1972; Fujiwara, 1972; Ohsaki and Iwasaki, 1973; Imai, 1977). A process
served. In particular, the higher is qmax, the lower is Emax/qmax; the of improvement introducing other factors as the type of soils or the
latter ranges between about 1000 for uncemented soil (clays, sand depth from the ground surface is also found. Ohta and Goto (1978) ob-
and gravel) and about 500 for cemented materials (soft and hard tained as much as fifteen sets of empirical correlation combining four
rock). Analyses of new data from Tatsuoka and Kohata (1995) con- indexes (NSPT, depth from the ground surface, geological age and soil
firmed the general concepts reported above. type) using about 300 data derived by the characterization of the alluvi-
The correlation between the shear wave velocity, VS, and the num- al deposits in Japan. Table 1 reports the empirical equations relating
ber of blows, NSPT, obtained by the Standard Penetration Tests (SPT), NSPT to VS (r = 0.719), but the function with the higher value of corre-
was largely investigated in the last decades. A large number of Authors lation coefficient is the one that includes all the indexes (r = 0.856).
86 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97

Table 1
Significant features of existing empirical correlation NSPT–VS.

Reference Country (location) Geological description Material type Type a b c Data r R2

pairs

Imai and Yoshimura (1970) Japan All soils VS = aNb 76.00 0.330
Ohba and Toriumi (1970) Japan (Osaka) Alluvium All soils VS = aNb 84.00 0.310
Shibata (1970) Japan Sands VS = aNb 32.00 0.500
Ohta et al. (1972) Japan Sands VS = aNb 87.00 0.360 100
Fujiwara (1972) Japan All soils VS = aNb 92.10 0.337
Ohsaki and Iwasaki (1973) Japan All Soils VS = aNb 81.40 0.390 220 0.886
Sands 59.40 0.470
Imai (1977) Japan Quaternary and All soils VS = aNb 91.00 0.337 943
Pleistocene alluvium Sand 80.60 0.331 151
Clays 102.00 0.292 183
Diluvium Sand 97.20 0.323 100
Clays 114.00 0.294 122
Ohta and Goto (1978) Japan Quaternary and All soils VS = aNb 85.35 0.348 289 0.719
Pleistocene alluvium Sands 88.40 0.333 0.719
Clays 86.90 0.333 0.719
JRA (1980) Japan Clays VS = aNb 100.00 1/3
Sands 80.00 1/3
Seed and Idriss (1981) USA All soils VS = aNb 61.40 0.500
Imai and Tonouchi (1982) Japan Quaternary and All soils VS = aNb 97.00 0.314 1654 0.868
Pleistocene alluvium Sands 87.80 0.314 0.690
Clay 107.00 0.274 0.721
Seed et al. (1983) USA Sands VS = aNb 56.40 0.500
Sykora and Stokoe (1983) USA Sand VS = aNb 100.50 0.290 97 0.840
Jinan (1987) China (Shanghai) Soft Holocene All soils VS = a(c + N)b 116.10 0.202 0.320 59 0.700
deposits
Lee (1990) Taiwan Sands VS = aNb 57.40 0.490 22 0.620
Silt 105.60 0.320 44 0.730
Clays 114.00 0.310 22 0.440
Yokota et al. (1991) Japan All soils VS = aNb 121.00 0.270
Lee (1992) Taiwan (Taipei) Sands VS = c + aN 4.74 157.100 0.691
Silt VS = a(N + 1)b 104.00 0.334 0.798
Clays 138.40 0.242 0.695
Kalteziotis et al. (1992) Greece All soils VS = aNb 76.20 0.240 268 0.820
Sands/silts 49.10 0.502 0.740
Clays 76.55 0.445 0.890
Athanasopoulos (1994) Greece All soils VS = aNb 107.60 0.360 221 0.730
Sands 85.30 0.420 0.680
Clays 121.70 0.330 0.770
Raptakis et al. (1995) Greece Loose sand VS = aNb 123.00 0.290
Medium sands 100.00 0.240
Soft clays 105.70 0.330
Stiff clays 184.20 0.170
Gravel 192.40 0.130
Iyisan (1996) Turkey (Erzincan) Deep Alluvial deposits All soils VS = aNb 51.50 0.516 0.810
All soils VS = aNb1 71.20 0.440 0.560
Kayabali (1996) Turkey Alluvium Sands VS = c + aN60 3.75 175.000 57 0.410
Rollins et al. (1998) USA Holocene Gravels VS = aNb60 63.00 0.430 186 0.590
Pleistocene Gravels 132.00 0.320 105 0.480
b
Debris fill VS = aN 222.00 0.060
Pitilakis et al. (1999) Greece (Volvi) Alluvium Sands/silts VS = a(N60)b 145.00 0.178 145 0.700
Clays 132.00 0.271 176 0.750
Jafari et al. (2002) Iran (Teheran) Quaternary alluvium All soils VS = aNb 19.00 0.850 65
Silts 22.00 0.770 27
Clays 27.00 0.730 38
Hasançebi and Ulusay (2007) Turkey (Yenisehir) Quaternary alluvium All soils VS = aNb 90.00 0.309 87 0.730
and detritus Sands 90.82 0.319 39 0.650
Clays 97.89 0.269 48 0.750
All soils VS = aNb60 104.80 0.260 87 0.710
Sands 131.00 0.205 39 0.560
Clays 107.60 0.237 48 0.750
Hanumantharao and Ramana (2008) India (Delhi) Alluvial deposits All soils VS = aNb 82.60 0.430 0.950
Sands 79.00 0.434 0.980
Silt 86.00 0.420 0.940
Koçkar and Akgün (2008) Turkey (Ankara) Quaternary All soils VS30 = aNb30 56.94 0.428 74 0.643
Fluvial All soils 37.05 0.527 37 0.731
Dikmen (2009) Turkey (Eskiçehir) Quaternary alluvium All soils VS = aNb 58.00 0.390 193 0.750
Sands 73.00 0.330 82 0.720
Clays 44.00 0.480 35 0.820
Silt 60.00 0.360 76 0.710
Maheswari et al. (2010) India (Chennai City) Coastal and All soils VS = aNb 95.64 0.301 200 0.840
alluvium deposits
Sands 100.50 0.265 0.840
Clays 83.91 0.358 0.930
All soils VS = aNb60 90.75 0.304 200 0.830
 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97 87

Table 1 (continued)

Reference Country (location) Geological description Material type Type a b c Data r R2

pairs

Maheswari et al. (2010) Sands 96.29 0.266 0.830

Clays 83.27 0.365 0.920
Akin et al. (2011) Turkey (Erbaa) Alluvial soils All soils VS = aNbzc 59.44 0.109 0.426 137 0.890
Sands 38.55 0.176 0.481 63 0.940
Clay 78.10 0.116 0.350 30 0.920
Pliocene soils All soils 121.80 0.101 0.216 30 0.940
Sands 52.04 0.359 0.177 6 0.980
Clay 140.60 0.049 0.232 19 0.890
Tsiambaos and Sabatakakis (2011) Greece Quaternary and All soils VS = a(N60)b 105.70 0.327 416 0.600
Neogene deposits Sands 79.70 0.365 78 0.630
Clays 112.20 0.324 301 0.610
Silt 88.80 0.370 37 0.750

The most wide-ranging database of Japanese surveys (about 1650 ex- uncorrected shear wave velocity measurements (e.g., measured at any
perimental points, comprehensive of all previous records) was analyzed given depth) to overburden correct SPT blowcount (e.g., reported at
by Imai and Tonouchi (1982), revealing that clayey soils are generally the conventional atmospheric pressure, as for the confining stress).
characterized by higher VS than sands. Generally speaking, the geologic context in which the correlations
Since the early '80s, a similar assessment was carried out with refer- were obtained, was strictly homogeneous. Usually, the considered lith-
ence to several areas in the USA, especially to evaluate the liquefaction ological formations are alluvial or marine deposits, collecting the exper-
susceptibility of alluvium and marine sands in California (Seed and imental data for a single site (Jinan, 1987; Iyisan, 1996; Pitilakis et al.,
Idriss, 1981; Seed et al., 1983; Sykora and Stokoe, 1983; Rollins et al., 1999; Jafari et al., 2002; Hasançebi and Ulusay, 2007; Hanumantharao
1998; Brandenberg et al., 2010). Some of these empirical relationships and Ramana, 2008; Koçkar and Akgün, 2008; Dikmen, 2009;
are based on energy-corrected SPT blow counts N60 and on Maheswari et al., 2010; Akin et al., 2011) or for a set of sites in the
overburden-corrected SPT blow counts (N1)60 (e.g. Seed et al., 1985), same region, district or country (Ohta and Goto, 1978; Imai and
which are generally used for the evaluation of the cyclic strength ratio Tonouchi, 1982; Tsiambaos and Sabatakakis, 2011). An attempt to
CRR in the liquefaction simplified analysis. make a distinction between the different ages of geological deposits is
On the other hand, in the last decade the enhancement of seismic done by some authors (Rollins et al., 1998; Koçkar and Akgün, 2008;
microzonation study refined data field techniques and increased the Akin et al., 2011), but, in most cases, the experimental database was
availability of VS measures. So, the empirical correlations were differen- clustered according to the grain size of the soil layers. In general, the en-
tiated according to the geographic location and the soil type, mainly for tire database was divided in sands and clays, but in some case also silts
alluvial formations in Greece (Kalteziotis et al., 1992; Pitilakis et al., and gravels were considered. Ohta and Goto (1978) used the grain size
1992; Athanasopoulos, 1994; Raptakis et al., 1995; Pitilakis et al., as a significant parameter in correlation and observed that the shear
1999; Tsiambaos and Sabatakakis, 2011), Turkey (Iyisan, 1996; wave velocity VS for gravel resulted larger than the one for sand and
Kayabali, 1996; Hasançebi and Ulusay, 2007; Koçkar and Akgün, 2008; the latter larger than the one for clay. More in general, the empirical cor-
Dikmen, 2009; Akin et al., 2011), India (Hanumantharao and Ramana, relations for cohesive soils are more consistent than those for sandy
2008; Maheswari et al., 2010; Anbazhagan et al., 2013) and Iran (Jafari soils and also showed higher shear wave velocities than sand; the silts
et al., 2002). Most of the empirical laws used for the data regression gave intermediate results and gravels the highest. Some remarks on
were power type, such as: geological features of the area can be finally found in Imai and
Tonouchi (1982), who observed that the greater VS values are in tertiary
b
VS ¼ a
NSPT ð2Þ deposits, lower values pertain to diluvial, and the lowest in alluvial
layers. Ohta and Goto (1978), Rollins et al. (1998) and Akin et al.
in which VS is the shear wave velocity measured at a specific depth z (2011) observed significant difference in the values of VS between the
and NSPT is the uncorrected number of blows of the SPT at the same Holocene and the Pleistocene sands, in particular the former were char-
depth. In some cases, different expressions were employed to fit the acterized by lower values compared with the latter. Sykora and Stokoe
dataset (Table 1). Only in a few cases the linear type correlations were (1983), instead, suggested that the geological age and the type of soil
used (Kayabali, 1996; Lee, 1992), showing, however, low regression co- were not predictive of the shear wave velocity VS.
efficient r. In other cases, the power expression incorporated also a con- In summary, the technical literature review did not prove that the
stant term c (Jinan, 1987; Kayabali, 1996). The most recent correlations grain size approach for empirical correlations is able to provide a full
explored the possibility to add depth (z) dependence (Akin et al., 2011) and reliable description of the soil geological and geotechnical variabil-
again in a power formula. As aforementioned, a rather extensive set of ity. This is the reason why an additional effort seems to be required in
correlation (Kayabali, 1996; Rollins et al., 1998; Pitilakis et al., 1999; order to perform an integration of the geotechnical and geological
Hasançebi and Ulusay, 2007; Maheswari et al., 2010; Tsiambaos and knowledge in view of reliable results. This is particularly true for the ex-
Sabatakakis, 2011) investigated the relationships between the VS with tremely complex geological context of the Italian Southern Apennines,
the well-known energy corrected SPT blow counts N60 and overburden where the Campobasso District area (Molise Region) is located.
stress corrected values (N1)60.
However, with some exception (e.g., Lee, 1990), almost all the stud-
ies showed better correlations in evaluation of VS if uncorrected blow 4. Geological setting of the Campobasso District area (Molise Region)
counts were considered (Hasançebi and Ulusay, 2007; Dikmen, 2009).
In this regard, Tsiambaos and Sabatakakis (2011) argued that the use The Molise Region is part of the Apennines chain, a fold and thrust
of (N1)60 may provide worse regression models, possibly due to uncer- belt, that represents the backbone of the Italian peninsula. According
tainties related to the definition of correction coefficients. Instead, low to Cello and Mazzoli (1999), the architecture of the chain recognizes
regression coefficients are probably due to the attempt to correlate “a complex imbricate structure including different thrust sheets”, in
88 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97

which the orogenic activity juxtaposed several paleogeographical do- the so-called “Samnitic Nappe” deposits. Molise Sannio basin formations
mains, characterized by different lithological properties. A number of (Lias–Messinian) may be identified on the basis of the facies analysis.
authors proposed several models of its late paleogeography, identifying Namely, Frosolone Unit (Ippolito et al., 1975; Sgrosso et al., 1988;
more or less complex systems of carbonate platforms and pelagic basins Antonucci et al., 2002), shelf-edge and transitional sedimentary deposits;
(e.g., Selli, 1962; Ogniben, 1969; D'Argenio et al., 1972; Ippolito et al., Agnone Unit, Tufillo Unit and Daunia Unit, which progressively present
1975; Mostardini and Merlini, 1986; Sgrosso, 1988, 1998; Patacca and more distal facies. Frosolone and Agnone Units consist of calciturbidites
Scandone, 2007). The Campobasso District area, where a large number enriched with siliciclastic deposits composed of clay and marly clay in-
of geotechnical tests are available, occupies the external portion of terbedded with sandstones. These Units are separated by a major thrust
chain and it represents the foreland thrust belt system. The complex geo- surface, but the original contiguity of the Frosolone and Agnone deposi-
logical setting of this sector can be simplified into three main units com- tional domains is attested by geological evidences of gradual transition of
posed of carbonate successions, siliciclastic flysch deposits and marine to facies. They pass gradually to the megaturbidites of the S. Elena Flysch
alluvial mainly clastic successions (e.g., Patacca et al., 1992; Patacca and (Festa et al., 2006). The Tufillo Unit (Langhian–Tortonian) consists of
Scandone, 2007). varicolored shales with subordinate intercalations of calciturbidites,
Fig. 1 shows a simplified map of geolithological complexes recog- hemipelagic marls, limestones and Numidian sandstones (Patacca and
nized in the area of interest. Carbonate platform units crop out extensive- Scandone, 2007). Daunia Unit is composed by the following geological
ly in the south-western part of the Molise Region, in the so-called Matese formations: Faeto Flysch, Toppo Capuana clayey marls and Montecastello
Mountains, and include only the south western edge of the Campobasso Evaporites (Di Nocera et al., 2006; Pertusati and Buonanno, 2009). Faeto
District. The stratigraphic succession is composed of Mesozoic dolomites Flysch is mainly composed of intercalation of marls, calcareous marls and
in the lower part and limestones in the upper one. At the top, the sedi- limestones. In the upper part of the succession the stratigraphic contact
mentary succession is closed by Miocene transgressive terrigenous sedi- to the Toppo Capuana clayey marls (Tortonian) is present. The Messinian
ments in facies of continental shelf and bypass made of carbonate and Evaporites (Montecastello Evaporites) are composed of gypsum and car-
turbiditic deposits (Cusano and Longano formations, Pietraroja flysch) bonate conglomerates in an arenaceous matrix with clay intercalation.
(Patacca and Scandone, 2007; Bonardi et al., 2009). Siliciclastic flysch de- They rest on the Faeto Flysch and are transgressive on the Tona Forma-
posits represent the most widespread geolithological complex, distin- tion (Messinian–Middle Pliocene) that represents the closure of a regres-
guishable from the north-east of the carbonate complex up to the front sive cycle, constituted of marly clay, sands and conglomerates. The Faeto
of the thrust belt, close to the current coast line. It includes all allochtho- Flysch passes by lateral facies heteropy or by tectonic contact to the San
nous siliciclastic deposits of the thrust top and foredeep basins. Such Bartolomeo Flysch Unit that is widespread in the Campobasso District
complex involves Molise Sannio basin formations, the Fortore Unit and too. It was recognized as a piggy-back basin resting on the roof-thrust

Fig. 1. Simplified geolithological complex map of the study area and location of the experimental NSPT–VS pairs.
 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97 89

of the chain (Tortorici, 1975). These flysch deposits are partitioned into the investigation at a test site. Those results were collected and ana-
two members: the composition of the lower is for the most part argilla- lyzed, providing a very dense database of available geotechnical surveys
ceous; the upper is arenaceous-conglomeratic. The San Bartolomeo for this relatively small Regional district. Further data were made avail-
Flysch often overlies the Fortore Unit, composed of pelagic sediments able in the framework of a wide regional program aimed at evaluating
(varicolored clays and shales) intercalated with fine carbonate turbidites. the seismic vulnerability of strategic buildings and road infrastructures
The so-called “Samnitic Nappe” deposits are also noteworthy. Their ori- (Di Carluccio et al., 2009; Evangelista et al., 2011) and from the con-
gin is still object of debate: they are mainly characterized by the “Varie- struction of a 3-D subsoil model in the coastal part of the investigated
gated Clays” Formation interlayered with calcareous quartz arenite regional district (Fabbrocino et al., submitted for publication).
inclusions. Finally, in addition to the cover deposits, in the eastern and As a result, a large database of soil investigations was set with a con-
periadriatic sector of the Campobasso District, Plio-Pleistocene trans- siderable potential in the geological and geotechnical characterization of
gressive–regressive cycles crop out. The whole succession was described the structurally complex formations of the district. Indeed, it was already
due to several stratigraphic logs derived from deep drilling data (ViDEPI used for the construction of a regional EC8 soil class map (Forte et al.,
Project), in which the Apulian carbonate bedrock (Trias–Creta) was 2013a) and for the study of the earthquake-induced landslide suscepti-
reached. A Plio-Pleistocene terrigenous sequence thick more than bility (Forte et al., 2013b). For the purpose of this study, the location of
2000 m is in onlap on the carbonates. It consists of the Montesecco sites where NSPT–VS pairs were collected is shown in Fig. 1. Despite the
Clays Formation (Pliocene–Calabrian?) composed of gray-bluish marly uneven distribution of the dataset, the investigations characterize all
and silty clay, which passes upward to the Serracapriola Sands and the soil types recognizable in the Campobasso District, or, more precisely,
Campomarino Conglomerates. They also represent the closure of the re- all the identified geolithological complexes. Overall, a total of 160 loca-
gressive cycle. tions were considered, where both the DH and SPT tests were performed.
Molise Region is exposed at seismic risk, since it is embedded within The borings reach a maximum depth of 50 m. SPTs were executed during
well-identified seismogenic zones (Meletti and Valensise, 2004; the drilling and, usually, in each borehole, two or three blow count mea-
Morasca et al., 2008; Valensise et al., 2004). The detailed description of surements were taken in a depth range variable from 2 to 30 m. From the
the regional seismicity is out of the scope of the present paper, but the preliminary dataset, 142 NSPT measurements were rejected, because the
interested reader can find further details elsewhere (Caccavale et al., number of blows was greater than 100. On the other hand, the achieved
2010). DH tests recorded the first arrival time of shear wave at 1 m depth inter-
vals. Finally, 286 NSPT–VS couples, all measured at the same depth and at
5. Molise Region geotechnical test database and NSPT–VS relations the same site, were assumed as a reliable reference dataset to assess
novel and specific empirical relationships to fit the Campobasso District
A large number of geotechnical in situ and laboratory tests were geological and geotechnical features.
executed for the study of the seismic microzonation of all the municipal-
ities belonging to the district of Campobasso (Molise Region, 5.1. Empirical correlations NSPT–VS based on the grain size approach
2003–2009). Available datasets include stratigraphic logs, Standard
Penetration tests (SPT), laboratory tests on soil specimens (triaxial UU The first step of the analysis consisted of the superposition of the
or RC tests), and Down-Hole tests (DH). Fig. 2 reports typical results of NSPT–VS pairs for the Campobasso District with the most common and

Fig. 2. Main results of integrated borehole microzonation data (stratigraphic log, shear wave velocity profile and the SPT profile): an example in the village of Mafalda (CB).
90 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97

applicable empirical correlation laws for “all soils” (Ohta and Goto,
1978; Yokota et al., 1991; Kalteziotis et al., 1992; Iyisan, 1996; Jafari
et al., 2002; Hasançebi and Ulusay, 2007; Dikmen, 2009; Akin et al.,
2011), as shown in Fig. 3. It is clear the large dispersion of the experi-
mental data and the inability of any aforementioned correlation rela-
tionship to characterize the terrain of interest. Technical literature
reports several NSPT–VS regression equations related to the different
“soil types”. Hence, the integrated borehole experimental data were an-
alyzed in order to better define the soil grain size classes in the dataset.
The available stratigraphic log descriptions and laboratory tests were
compared in order to validate the results.
In this way, the database was divided according to the main compo-
nent of the grading curves: the larger amount of data (58%) referred to
clayey soil samples, followed by silty (23%) and sandy ones (11%). A
very few pairs of measures (8%) referred to gravelly soils, and so they
were neglected. Then, for each grain size class the correlations were de-
veloped for uncorrected number of blows NSPT, according to the back-
ground technical literature. As shown in Fig. 4, statistical correlations,
between uncorrected number of blows NSPT and relating VS measure-
ments, and their coefficients of determination, R2, were evaluated by a
power law, for sandy (Fig. 4a), silty (Fig. 4b) and clayey soils (Fig. 4c) re-
spectively. In spite of the fact that the Standard Penetration Test is more
appropriate for coarse-graded soil rather than for fine-graded material,
it seems that there is no correlation between NSPT and VS for the sandy
materials, while two relationships can be derived for silty and clayey
soils. This is probably due to the limited amount of sand data available
for the empirical correlation (32 pairs). For medium and fine graded
soils the shear wave velocity (m/s) could eventually be obtained from
the following relationship:


0:192 2
VS ¼ 149:3NSPT R ¼ 0:362–r ¼ 0:571 ð3Þ

based on the non-linear regression of 62 VS–NSPT pairs. For clayey soils

the relationship, based on 178 pairs is:


0:252 2
VS ¼ 110:5NSPT R ¼ 0:369–r ¼ 0:604 : ð4Þ

It is worth noting that a quite large dispersion of data is observed,

and confirmed by low values of R2, thus the soil type could have only
a slight effect on the investigated correlations. Hence, this approach
seems to be not reliable to characterize the Campobasso District terrain,
where, on the analogy with other Italian Southern Apennines areas,
many complex geological and morphological factors control and mark
the mechanical behavior of soils. Fig. 4. VS–NSPT empirical correlations based on grain size approach: a) sandy, b) silty and
c) clayey soils.

5.2. Empirical correlations NSPT–VS based on the lithostratigraphical

approach

The heterogeneity of the sedimentary deposits cropping out in the

Campobasso District reflects the geological features of the whole fault
and thrust Apennine chain. In this sector of the chain, most of the
lithotypes consist of flysch sequences, with the exception of carbonate
olistholites derived from bedrock splitting and embodied in such
marly–clayey and arenaceous–conglomeratic–clayey sequences. They
often show the characteristic structure of the so-called “wild flysch”, a
very chaotically built formation (Maisadze, 2008). Their structural–tex-
tural and lithological features differ in both vertical and horizontal
directions. In general, it is likely to recognize a clayey and lesser clay-
ey–marly and arenaceous–sandy facies in the lower member, while in
Fig. 3. Comparison between experimental values and the most common empirical NSPT–VS the upper member the conglomeratic–sandy facies, with heterometric
correlations for “all soils”. calcareous and calcareous–marly blocks, is dominant.
 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97 91

It is clear that the same geolithological complex can be character- marly–clayey–calcareous facies. The top layer, on the contrary, was
ized by a high variability of geotechnical properties, while the grain characterized by an arenitic Qz-rich facies (e.g., Pertusati and
size classification plays a minor role in the improvement of the cor- Buonanno, 2009). All experimental NSPT–VS data pairs shown in Fig. 6
relations between NSPT and VS. Conversely, a relevant role may be as- were correlated to the upper member of the Faeto Flysch formation,
cribed to the lithostratigraphical features of the geolithological but the data from Mafalda village B5 described the arenitic Qz-rich tur-
complexes. An insight into the integrated borehole data and particu- bidites interbedded with massive marls and marly limestones. The data
larly into stratigraphic logs—see Fig. 2—was made to identify the geo- of Santa Croce di Magliano B3 and Casacalenda B2 identified the marly–
logical formation related to each experimental pair of NSPT and VS clayey–calcareous facies, composed of regular alternations of massive
values. They were allocated on the 1:100,000 Molise Geological Map marls, marly limestones, calcilutites and white and gray clays. In partic-
by Festa et al. (2006), which represents the existing geological outline ular, the corresponding NSPT–VS value pairs were evaluated for the clay-
of the Campobasso District area at a fit scale of representation. The rec- ey layers. The location of the Casacalenda investigation in displaced
ognized geological formations are summarized in Table 2. The number materials of landslide is also noteworthy.
of experimental NSPT–VS data pairs associated with the identified geo- This approach pointed out that the evident dispersion of experimen-
logical formations is shown in Fig. 5. All geolithological complexes tal NSPT–VS data pairs recorded in Fig. 3 could be reasonably related to
were characterized with the exception of the detrital and carbonate the different lithofacies observed in the study area. In particular, three
ones—see Fig. 1—such as the most widespread formations, namely main lithofacies could be distinguished in the Campobasso District and
flysch sequences, and were significantly represented. The combination related to the geotechnical behavior of the examined structurally com-
of such data and information derived from the background literature plex formations, in view of the optimization of the empirical NSPT–VS
as well as from a geological survey was able to distinguish within the correlation formulas (Fig. 6):
examined formations the structural–textural and lithological features
of each horizon related to analyzed NSPT–VS value pairs. a) Structured Soil Deposit (SSD): this facies is typical of complexes and
Particular attention was paid to local conditions of outcropping sed- formations (or only of some horizons of them) in which the rock
imentary deposits, such as weathering processes, tectonic setting, hy- interbeddings are not negligible and highly affect the soil behavior
drodynamic models, and anthropic impact, which can modify intrinsic (Fig. 6—Mafalda B5);
properties of recognized layers and influence the geotechnical charac- b) All Soils (AS): this facies involves complexes and formations (or only
teristics of the same horizon. For instance, the integrated boreholes de- some horizons of them) in which the argillaceous fraction is domi-
rived from the investigations in the Mafalda—see also Fig. 2, Santa Croce nant. Its behavior could be considered similar to the sensu-strictu
di Magliano and Casacalenda villages, were correlated with the succes- soils, because it encompasses the frictional and cohesive materials
sion of the Faeto Flysch (Fig. 6) as it is representative of the main (Fig. 6—S. Croce di Magliano B3);
lithofacies detectable in the Campobasso District. The Faeto Flysch con- c) Weathered Soils (WS): this facies includes the same complexes and
sists of a calcareous marly turbiditic succession. The lower part was formations of the so-called “All soils”, but the geotechnical proper-
characterized by a calcareous–marly facies; the upper part acquired a ties are affected by both natural and anthropic factors (Fig. 6—

Table 2
Summary of the lithological descriptions of geological formations cropping out in the investigation sites.

Geological Description Outcropping sites

formations

Alluvial and lacustrine deposits Gravels with sand intercalation inserted in a silty matrix. Bojano, Trivento, Termoli, Campochiaro, Lucito
Lacustrine and marsh deposits Varved clay bounded upwards and downward by clastic complexes Campobasso, Bojano, Termoli, Campochiaro, Campomarino,
in a clayey matrix Guglionesi, Vinchiaturo
Campomarino conglomerates Lenses and beds of gravels and conglomerates with intercalation of Termoli, Campomarino
sands and clays
Montesecco clays Gray-bluish marly and silty clay Petacciato, Campomarino, Ururi
Gessopalena gypsums Gypsum and carbonate conglomerates in an arenaceous matrix with Montecilfone
clay intercalation
Tona formation Gray-bluish silty clay with cemented layers and intercalation of Sant'Angelo Limosano, Larino, Castellino del Biferno
sandy clay
Campobasso conglomerates Conglomerates constituted by carbonate marly–calcareous and Campobasso
sandstones elements. They are contained in a matrix of clayey sands
Sant'Elena Flysch Graded sandstones in banks with intercalation of clay deposit Toro, Campodipietra, Spinete
Pietraroja Flysch Intercalation of clay and siltstones passing upward to sandstones. At Ripalimosani, San Paolo Matese, Campolieto, Guardiaregia
the top of the formation are present elements of varicolored clays
and quartz sandstones
Frosolone Unit Clay, marls, gray-bluish marly siltstones with intercalation of San Massimo, Frosolone
calcarenites and turbidite sandstones
Tufillo Unit Thick interbedding of limestones, white marly limestones, clayey Palata, Salcito, Montagano, Petrella Tifernina, Guardialfiera,
marls, siltstones with intercalation of bioclastic carbonates Montemitro, Provvidenti
Fortore Unit Interbedding of yellow bioclastic sandstones with clays, marls and Campobasso, Ripalimosani, Busso, Matrice, Monaciloni, Montagano,
gypsum passing upwards to gray clays with lenses of carbonate Ferrazzano, Mirabello Sannitico, Macchia Valfortore, Gambatesa,
conglomerates Baranello, San Giuliano del Sannio, Vinchiaturo
Faeto — S. Bartolomeo Flysch Calcarenites and brecciated limestones with abundant organic San Felice del Molise, Montefalcone del Sannio, Trivento, Mafalda,
fragments and chert nodules, interbedded with lenses of yellow Roccavivara, Colletorto, Santa Croce di Magliano, Bonefro,
carbonate sandstones, marls and marly clays Casacalenda, Montorio nei Frentani
Cercemaggiore formation Interbedding of carbonate breccias, biocalcarenites with chert Castropignano, Oratino, Sepino
nodules and marly clays. At the base of the deposit is present a layer
of red-greenish clays
Variegated clays Thick intercalation of marly clays and cherts. It is a chaotic mélange Trivento, San Biase, Pietracupa, Gildone, Torrella del Sannio, Sant'Elia
with presence of micritic limestones, quartz siltstones and a Pianisi, Jelsi, Montagano, Lupara, Lucito, Pietracatella, Limosano,
radiolarites Campodipietra, Larino, Provvidenti
92 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97

Fig. 5. Number of experimental NSPT–VS data pairs per each geological formation.

Casacalenda B2). This group is mostly composed of very shallow relationships between the shear wave velocity and the SPT blow-
measurements, composed by eluvium, colluvium or talus deposits. number compared with the soil grain size classification; and 2) the reli-
ability and significance of the original relationships based on the
lithostratigraphical features of the area.
Then a non-linear regression analysis was carried out to define three
original empirical functions between NSPT and VS, driven by the lithos- 5.3.1. Issue #1
tratigraphical features of the outcropping geological formations (Fig. 7): A preliminary check on NSPT data quality was performed. Firstly, for

 each available measurement, the normalized value N1 was computed to
0:385 2
SSDÞVS ¼ 104:55
NSPT R ¼ 0:681–r ¼ 0:825 ð5Þ remove the dependence of the blow-number from the overburden
stress. Specifically, following Idriss and Boulanger (2004):


0:317 2
ASÞVS ¼ 90:35
NSPT R ¼ 0:748–r ¼ 0:865 ð6Þ
!0:784−0:0768
pffiffiffiffi
N
ffi
1
pa

 N1 ¼ 0
NSPT ð8Þ
0:272 2 σv
WSÞVS ¼ 78:59
NSPT R ¼ 0:607–r 0:774 : ð7Þ

In all the cases, the developed analytical functions represent a better in which, pa is the atmospheric pressure and σ′v is the overburden pres-
fit of data, as shown by the higher correlation coefficients, compared sure. The latter was evaluated by assuming the two boundary condi-
with those provided by Eqs. (3) and (4). tions related to fully saturated soils and fully dry soils. Then, the
Some remarks can be proposed, based on the comparison between normalized SPT blow-number data were grouped into three different
the novel and available empirical correlation formulas (Fig. 8): soil classes, respectively clay, silt and sand. The corresponding box-
chart is shown in Fig. 9. A general increasing trend of the normalized
• The SSD relationship shows higher values of VS compared to other blowcount depending on the increase of the average soil grain size
existing expressions; this could be associated with the prominent can be observed. Furthermore, the data present a quite large dispersion
contribution of the rock component of the deposit; that becomes more significant for the coarser materials. Such distribu-
• For the case of AS, the empirical correlation is quite similar to the Ohta tion of experimental normalized blow number N1 is consistent with
and Goto (1978) relationship, because the soils investigated by the the remarks reported in the previous sections. The grain size classifica-
Authors are mainly composed of clay and silty clay successions, with tion has slight or negligible significance in order to improve the evalua-
sandy intercalations; tion of correlations between the two examined parameters (NSPT–VS).
• The correlation for WS is similar to the expression of Dikmen (2009);
this empirical evidence was likely due to similar lithofacies in which 5.3.2. Issue #2
the analyses were done. The lithostratigraphical analysis of geolithological complexes docu-
mented in the Campobasso District provided an improvement of the in-
vestigated mathematical relations. It was confirmed not only by higher
5.3. Discussion of the results values of correlation coefficients, but also by the detailed statistical rep-
resentation given by the additional parameters in Table 3 that are:
A simple statistical treatment of data was carried out to assess: (a) The number of samples; (b) The coefficient of determination, R2;
1) the capacity of the lithostratigraphic analysis in establishing suitable (c) The Mean Absolute Error of the prediction (MAE); (d) The Sum of
 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97 93

Fig. 6. Correlation between integrated borehole data and lithofacies. An example for the Faeto Flysch.

Squared Errors of the prediction (SSE); and (e) The Root Mean Squared data collected in Table 3, a satisfactory fitting is observed for the three
Error of the prediction (RMSE). lithofacies derived from the lithostratigraphical approach. Fig. 10 re-
The correlation value (Value), the standard error (SE) and the upper ports the bi-logarithmic plots of the proposed relationships for the
and lower boundaries of the 95% confidence interval (CI) are reported Structured Soil Deposits (SSD) (Fig. 10a), for All Soil (AS) (Fig. 10b)
for multiplier (a) and exponent (b)—see Eq. (2). The parameters MAE, and for the Weathered Soils (WS) (Fig. 10c), together with the 95%
SSE and SMSE are error estimators, so they can be helpful to measure
how close predictions are to the outcomes: the lower the parameter,
the more the model approaches experimental data. According to the

Fig. 8. Comparison between the novel empirical relationships derived from the
Fig. 7. Novel VS–NSPT empirical correlations based on lithostratigraphical approach. lithostratigraphical approach and the existing ones.
94 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97

where Vs,c is the shear wave velocity calculated by Eqs. (5)–(7) and Vs,m
is the experimental measured value. Fig. 11 shows that Eq. (5) for SSD
provides about 90% of the predicted values within a 20% error margin;
Eq. (6) for AS gives about 97% of predicted values within a 19% error
margin; finally Eq. (7) for WS leads to about 90% of values predicted
within a 20% error margin. It is thus confirmed that the three proposed
correlations offer a good estimation of investigated parameter. The
comparison between Vs,c and Vs,m derived from the normalized consis-
tency ratio Cd strengthens such comment. The normalized consistency
ratio, see Fig. 12, was evaluated as:

Vs;c −Vs;m
Cd ¼ ð10Þ
NSPT
:

A reliable estimation of VS values is also observed in the case of the

AS and WS correlations. The Cd values are generally close to zero, except
for small NSPT values (generally lower than 15). But, the lower values
were measured under particular local conditions, such as weathered
or fully saturated soils, which affect the results. As expected, a larger
variability was observed for SSD, even if the predictive capability of its
mathematical relation may be considered reliable.

6. Conclusions
Fig. 9. Box chart of normalized SPT blow number N1 per soil type in dry and wet
conditions. The present paper reports a study focused on the assessment of reli-
able relationships between the shear wave velocity VS and the number
of blows derived from SPT (NSPT), in structurally complex formations
confidence interval of the expected values. Due to the good ability to the characterizing the geological terrains of the Southern Apennine chain
power model to fit the data, relatively narrow bands can be seen in the (Italy). Based on a relative large number of in situ tests, available for
graphs. the Campobasso District in the Molise Region (Central Italy), novel
Finally, in order to check the capability of the three novel empirical power-type relationships were presented. They were derived following
correlations for shear wave velocity prediction, the relative error Er, in an original methodology that combines geological and geotechnical
terms of percentage, was evaluated as shown below: knowledge and aims at overcoming the limitations shown by the con-
ventional grain size approach. In particular, a lithostratigraphical ap-
Vs;c −Vs;m
Er ¼ 100
ð9Þ proach was developed in order to account for the structural textural
Vs;c and lithological features of the examined chaotically built formations,

Table 3
Main statistical parameters for the proposed regression equations.

Relation Samples R2 MAE SSE RMSE Coefficient a (m/s) Exponent b

(m/s) (m/s)2 (m/s)
Value SE Lower CI Upper CI Value SE Lower CI Upper CI

WS 63 0.607 17.938 34,023.174 23.617 78.593 8.753 61.122 96.081 0.272 0.030 0.212 0.331
AS 118 0.748 23.636 97,012.695 28.919 90.353 6.175 78.141 102.565 0.317 0.018 0.281 0.353
SSD 105 0.681 43.957 305,425.534 54.455 104.552 10.328 84.093 125.000 0.385 0.028 0.330 0.441

Fig. 10. Novel NSPT–VS correlations and relative 95% confidence interval of expected mean values: a) SSD; b) AS; c) WS).
 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97 95

Fig. 11. Scaled relative errors of VS for the novel correlations based on the lithos-
tratigraphical approach.

mostly consisting of flysch sequences. Three distinct correlations were

obtained depending on the main evident lithofacies, namely Structured
Soil Deposit (SSD), All Soils (AS), and Weathered Soils (WS). The first
correlation (SSD) applies to formations (or only to some horizon of
them) for which the influence of the rock interbedding is dominant;
the second (AS) refers to formations (or only to some horizon of
them) for which the “classical” soil mechanics applies as the argilla-
ceous fraction is very relevant; the third (WS) is useful to predict VS
for the same formations classified as AS, but for which the geotechnical
properties decay either for natural factors or for anthropic impact that
increases the water content and modifies the flow conditions. The sta-
tistical analysis of the data confirmed the validity of the proposed
relationship.
The proposed approach appears to be fully consistent with the ex-
perimental procedures used to obtain the two values—NSPT and VS—
that are here correlated. The first data is a punctual measurement that
is well affected by the specific soil characteristics where the test is per-
formed; the second is representative of more global information. This
means that the SPT blow count in structured soil deposits usually de-
tects the strength of the weaker part of the material, as the testing de-
vice is not able to penetrate into the most lithic part of the soil. On the
other hand, the shear wave velocity, even obtained at a given depth, is
a more global measurement accounting for both the stiffer part and
the softer part of the soil formation. Then, the SSD correlation law is lo- Fig. 12. Normalized consistency ratio Cd for (a) Structured Soil Deposit (SSD), (b) All Soils
(AS) and (c) Weathered Soils (WS).
cated in the upper part of the VS–NSPT plane. Conversely, the WS corre-
lation is in the lower part of the plane because the local SPT blow count
is there mostly influenced by the remolded and the fully saturated na-
ture of the soil belonging to the formation. Even if only a sector of the References
Apennine chain is concerned, the proposed correlations could reason- Akin, M.K., Kramer, S.L., Topal, T., 2011. Empirical correlations of shear wave velocity (Vs)
ably be applied in different sectors of the same chain or in similar geo- and penetration resistance (SPT-N) for different soils in an earthquake-prone area
logical terrain whenever the uncertainties and the features of this (Erbaa-Turkey). Eng. Geol. 119, 1–17.
Anbazhagan, P., Parihar, A., Rashmi, H.N., 2012. Review of correlation between SPT-N and
simplified tool fit the aim of the analysis and the scale of the engineering shear modulus: a new correlation applicable to any region. Soil Dyn. Earthq. Eng. 36,
problem of interest. In view of the use of the proposed original tools, it is 52–69.
worth noting that the knowledge of regional geology is strongly recom- Anbazhagan, P., Kumar, Abhishek, Sitharam, T.G., 2013. Seismic site classification and cor-
relation between Standard Penetration Test N value and Shear wave velocity for
mended for the identification of the lithofacies corresponding to the Lucknow city in Indo-Gangetic Basin. Pure Appl. Geophys. 170, 299–318.
geological formations (or only to some horizon of them) in which the Antonucci, A., Di Luzio, E., Lenci, F., Scrocca, D., Tozzi, M., 2002. Mesoscopic structural
in situ tests are performed. Furthermore, the stratigraphic logs and styles of deformation within the Frosolone unit multilayer (Molise Region, Central
Italy). Boll. Soc. Geol. Ital. 1, 641–648.
eventually the geological survey can be recommended as a guide in set-
ASTM D1586-11, 2011. Standard Test Method for Standard Penetration Test (SPT)
ting out the intrinsic properties (structural–textural and lithological fea- and Split-barrel Sampling of Soils. ASTM International, West Conshohocken,
tures) of the recognized geological formations and in assigning the PA.
experimental blow count value to the most suitable correlation law. In ASTM D4428/D4428M-14, 2014. Standard Test Methods for Crosshole Seismic Testing.
ASTM International, West Conshohocken, PA, USA.
this way, the thickness of soil or lithofacies layers can also be derived, al- ASTM D7400-08, 2008. Standard Test Methods for Downhole Seismic Testing. ASTM In-
though it is not a priori known. ternational, West Conshohocken, PA, USA.
96 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97

Athanasopoulos, G.A., 1994. An empirical correlation VS0–NSPT and an evaluation of its Kalteziotis, N., Sabatakakis, N., Vassiliou, I., 1992. Evaluation of dynamic characteristics of
reliability. In: Savidis, S.A., Balkema, A.A. (Eds.), Proc. of the 2nd International Confer- Greek soil formations. Proc. 2nd Hellenic Conference on Geotechnical Engineering,
ence on Earthquake Resistant Construction and Design, Berlin,1994, pp. 219–226. Thessaloniki vol. II, pp. 239–246 (in Greek).
Biot, M.A., 1956. Theory of propagation of elastic waves in a fluid-saturated porous solid. I. Kayabali, K., 1996. Soil liquefaction evaluation using shear wave velocity. Eng. Geol. 44
Low-frequency range. J. Acoust. Soc. Am. 28 (2), 168–178. (1), 121–127.
Bonardi, G., Ciarcia, S., Di Nocera, S., Matano, F., Sgrosso, I., Torre, M., 2009. Carta delle Koçkar, M.K., Akgün, H., 2008. Development of a geotechnical and geophysical database
principali unità cinematiche dell'Appennino Meridionale. Nota illustrativa. Ital. for seismic zonation of the Ankara Basin, Turkey. Environ. Geol. 55, 165–176.
J. Geosci. 128 (1), 47–60 (in Italian). Kramer, S.L., 1996. Geotechnical Earthquake Engineering. Prentice Hall (653 pp.).
Brandenberg, S.J., Ballana, N., Shantz, T., 2010. Shear Wave Velocity as a Statistical Func- Kulhaway, F.H., Maine, P.W., 1990. Manual on estimating soil properties for foundation
tion of Standard Penetration Test Resistance and Vertical Effective Stress at Caltrans design. Electric Power Research Institute, EPRI EL6800, Project 1493-6, Final Report.
Bridge Sites. PEER Report 2010/03. Lee, S.H.-H., 1990. Regression models of shear wave velocities. J. Chin. Inst. Eng. 13,
Butcher, A.P., Campanella, R.G., Kaynia, A.M., Massarsch, K.R., 2005. Seismic cone 519–532.
downhole procedure to measure shear wave velocity — a guideline. ISSMGE TC10, Lee, S.H.-H., 1992. Analysis of the multicollinearity of regression equations of shear wave
Proceedings of the International Symposium on Geophysical Testing in Geotechnical velocities. Soils Found. 32 (1), 205–214.
Engineering, May, 2006, Osaka, Japan. Maheswari, U., Boominathan, A., Dodagoudar, G.R., 2010. Use of surface waves in statisti-
Caccavale, M., Mancini, M., Convertito, V., Fabbrocino, G., 2010. Seismic hazard disaggrega- cal correlations of shear wave velocity and penetration resistance of Chennai soils.
tion in the Molise region, Italy: the case study of Campobasso. Proceedings of the 14th Geotech. Geol. Eng. 28, 119–137.
European Conference on Earthquake Engineering. Ohrid, Republic of Macedonia, 30th Maisadze, F., 2008. Event deposits in chaotically built formations. Bull. Georgian Natl.
August–3rd September. Acad. Sci. 2 (3), 79–87.
Cello, G., Mazzoli, S., 1999. Apennine tectonics in southern Italy: a review. J. Geodyn. 27, Mayne, P.W., Christophe, B.R., De Jong, J., 2001. Manual on Subsurface Investigations. Na-
191–211. tional Highway Institute Publication No. FHWA NHI-01-031 Federal Highway Admin-
D'Argenio, B., Pescatore, T., Scandone, P., 1972. Schema geologico dell'Appennino istration Washington, DC.
Meridionale (Campania, Lucania). Atti del Convegno: Moderne vedute sulla geologia Meletti, C., Valensise, G., 2004. Zonazione Sismogenetica ZS9-App. 2 al Rapporto
dell'Appennino. Accad. Naz. Lincei Quad. 183, 49–72 (in Italian). Conclusivo con contributi di Azzaro R, Barba S., Basili R., Galadini F., Gasperini
Di Carluccio, A., Fabbrocino, G., Santucci de Magistris, F., Todisco, F., Fabbrocino, S., 2009. P., Stucchi M. & Vannucci G. Gruppo di Lavoro per la redazione della mappa
Approccio metodologico alla valutazione di vulnerabilità per infrastrutture della pericolosità sismica INGV (Istituto Nazionale di Geofisica e Vulcanologia)
distribuite: aspetti geologici e geotecnici. Proc. XIII Convegno Nazionale “L'Ingegneria (in Italian).
Sismica in Italia”, ANIDIS Bologna (in Italian). Molise Region, 2003–2009. Linee Guida per la Microzonazione Sismica dei Comuni della
Di Nocera, S., Matano, F., Pescatore, T.S., Senatore, M.R., Torre, M., 2006. Schema geologico Provincia di Campobasso con particolare riferimento a quelli colpiti dal terremoto
del transetto Monti Picentini Orientali Monti della Daunia meridionale; unità del 31.10.2002. 52 pp. World Wide Web address:, http://www.regione.molise.it/
stratigrafiche ed evoluzione tettonica del settore esterno dell'Appennino meridionale. web/assessorati/serviziogeologico.nsf/(InfoInternet)/
Boll. Soc. Geol. Ital. 125, 39–58 (in Italian). 37E7DDE91EC393FDC12572D00033D834?OpenDocument (OpenDocument – Last
Dikmen, U., 2009. Statistical correlations of shear wave velocity and penetration resis- access May 2014 – (in Italian).).
tance for soils. J. Geophys. Eng. 6, 61–72. Morasca, P., Zolezzi, F., Spallarossa, D., Luzi, L., 2008. Ground motion models for the Molise
EN1997-1:2004, 2004. Eurocode 7. Geotechnical Design — Part 1: General Rules, CEN region (Southern Italy). Soil Dyn. Earthq. Eng. 28, 198–211.
European Committee for Standardization, Brussels, Belgium. Mostardini, F., Merlini, S., 1986. Appennino centro meridionale. Sezioni geologiche e
EN1998-5:2005, 2004. Eurocode 8. Design of Structures for Earthquake Resistance Part 5: proposta di modello strutturale. 73° Congr. Soc. Geol. It. Roma 30.9/4.10.86,
Foundations, Retaining Structures and Geotechnical Aspects. Comité Européen de pp. 1–59 (in Italian).
Normalisation, Brussels, Belgium. N.T.C., 2008. Norme tecniche per le costruzioni (Italian Building Code). Gazzetta Ufficiale
Evangelista, L., Fabbrocino, S., Lanzano, G., Todisco, F., Santucci de Magistris, F., della Repubblica Italiana (n. 29 del 4 febbraio 2008 – Supplemento Ordinario n. 30,
Fabbrocino, G., 2011. Integrated geotechnical characterization of distributed sites in 2008 (in Italian)).
the Molise Region (Italy) for seismic vulnerability analysis. Proc. 5th International Ogniben, L., 1969. Schema geologico introduttivo alla geologia del confine calabro lucano.
Conference on Earthquake Geotechnical Engineering 5ICEGE. January 2011, 10–13. Mem. Soc. Geol. Ital. 8, 453–763 (in Italian).
Santiago, Chile. Ohba, S., Toriumi, I., 1970. Dynamic response characteristics of Osaka Plain. Proceedings of
Fabbrocino, S., Forte, G., Lanzano, G., Paduano, P., Santucci de Magistris, F., Todisco, F., the Annual Meeting, A.I. J (in Japanese).
Fabbrocino, G., 1936. Engineering geology model for seismic vulnerability assessment Ohsaki, Y., Iwasaki, R., 1973. On dynamic shear moduli and Poisson's ratio of soil deposits.
of distributed infrastructures. 34th International Geological Congress Australia 2012, Soils Found. 13, 61–73.
Brisbane, Abstract #. Ohta, Y., Goto, N., 1978. Empirical shear wave velocity equations in terms of characteris-
Festa, A., Ghisetti, F., Vezzani, L., 2006. CARTA GEOLOGICA DEL MOLISE 1:100000 (in Italian). tics soil indexes. Earthq. Eng. Struct. Dyn. 6, 167–187.
Forte, G., Lanzano, G., Santucci de Magistris, F., Fabbrocino, S., Silvestri, F., 2013a. Validazione Ohta, T., Hara, A., Niwa, M., Sakano, T., 1972. Elastic shear moduli as estimated from N-
di approcci semplificati per la stima della classe di sottosuolo a scala territoriale: il caso value. Proc. 7th Ann. Convention of Japan Society of Soil Mechanics and Foundation
studio della Regione Molise. IARG, Perugia (Settembre 2013 (in Italian)). Engineering, pp. 265–268.
Forte, G., Fabbrocino, S., Santucci de Magistris, F., Silvestri, F., 2013b. Seismic permanent Patacca, E., Scandone, P., 2007. Geology of Southern Apennines. In: Mazzotti, A., Patacca,
ground deformations: earthquake triggered landslides in Molise Apennines. Rend. E., Scandone, P. (Eds.), CROP-04. Boll. Soc. Geol. Ital., Spec. Issue, No. 7, pp. 75–119.
Online Soc. Geol. Ital. 24, 134–136. Patacca, E., Scandone, P., Bellatalla, M., Perilli, N., Santini, U., 1992. La zona di giunzione tra
Fujiwara, T., 1972. Estimation of ground movements in actual destructive earthquakes. l'Arco appenninico settentrionale e l'arco appenninico meridionale nell'Abruzzo e nel
Proceedings of the Fourth European Symposium on Earthquake Engineering, Molise. Studi Geologici Camerti, vol. spec. CROP 11 (1991/92), pp. 417–441 (in Italian).
London, pp. 125–132. Pertusati, S., Buonanno, A., 2009. Structural evolution of a foreland basin succession: the
Hanumantharao, C., Ramana, G.V., 2008. Dynamic soil properties for microzonation of Dauna Unit in the Sannio-Molise sector of the Southern Apennines. Ital. J. Geosci.
Delhi, India. J. Earth Syst. Sci. 117 (S2), 719–730. 128 (2), 551–564.
Hasançebi, N., Ulusay, R., 2007. Empirical correlations between shear wave velocity and Pitilakis, K.D., Anastasiadis, A., Raptakis, D., 1992. Field and laboratory determination of
penetration resistance for ground shaking assessments. Bull. Eng. Geol. Environ. 66, dynamic properties of natural soil deposits. Proceedings of 10th World Conf. Earth-
203–213. quake Engineering, Rotterdam, pp. 1275–1280.
Idriss, I.M., Boulanger, R.W., 2004. Semi-empirical procedures for evaluating liquefaction Pitilakis, K., Raptakis, D., Lontzetidis, K., Tika-Vassilikou, T., Jongmans, D., 1999. Geotechni-
potential during earthquakes. In: Doolin, et al. (Eds.), Proc.11th ICSDEE & 3rd cal and Geophysical Description of Euro-Seistests, using field and laboratory tests and
ICEGE, Berkeley, CA, USA 1, pp. 32–56. moderate strong ground motions. J. Earthq. Eng. 3 (3), 381–409.
Imai, T., 1977. P and S waves velocities of the ground in Japan. Proceeding of IX Inter- Raptakis, D.G., Anastasiadis, S.A.J., Pitilakis, K.D., Lontzetidis, K.S., 1995. Shear wave veloc-
national Conference on Soil Mechanics and Foundation Engineering 2, ities and damping of Greek natural soils. Proc. 10th European Conf. Earthquake Engg.,
pp. 127–132. Vienna, pp. 477–482.
Imai, T., Tonouchi, K., 1982. Correlation of N-value with S-wave velocity and shear modu- Reid, A., Taylor, J., 2010. The misuse of SPTs in fine soils and the implications of Eurocode
lus. Proceedings of the 2nd European Symposium of Penetrating Testing, Amsterdam, 7. Ground Eng. 43 (7), 28–31.
pp. 57–72. Rollins, K.M., Diehl, N.B., Weaver, T.J., 1998. Implications of Vs-BPT (N1)60 correlations for
Imai, T., Yoshimura, Y., 1970. Elastic wave velocity and soil properties in soft soil. liquefaction assessment in gravels; GSP 75. J. Geotech. Geoenviron. Eng. ASCE 124,
Tsuchito-Kiso 18 (1), 17–22 (in Japanese). 506–517.
Ippolito, F., D'Argenio, B., Pescatore, T., Scandone, P., 1975. Structural–stratigraphic units Santucci de Magistris, F., Tatsuoka, F., 2004. Effect of moulding water content on the
and tectonic framework of Sothern Apennines. In: Squyres, C. (Ed.), Geology of stress–strain behaviour of a compacted silty sand. Soils Found. 44 (2), 85–101.
Italy, Earth Sci. Soc. Lybian Arab Republ, pp. 317–328. Santucci de Magistris, F., Silvestri, F., Vinale, F., 1998. The influence of compaction on the
Iyisan, R., 1996. Correlation between shear wave velocity and in-situ penetration test re- mechanical behaviour of a silty sand. Soils Found. 38 (4), 41–56.
sults. Digest 96 pp. 371–374 (December). Schmertmann, J.H., 1979. Statics of SPT. J. Geotech. Eng. Div. 105 (5), 655–670.
Jafari, M.K., Shafiee, A., Razmkhan, A., 2002. Dynamic properties of fine-grained soils in Schmertmann, J.H., Palacios, A., 1979. Energy dynamics of SPT. J. Geotech. Eng. Div. 105
south of Tehran. J. Seismol. Earthq. Eng. 4, 25–35. (8), 909–926.
Jinan, Z., 1987. Correlation between shear wave velocity and the number of blows of SPT Seed, H.B., Idriss, I.M., 1981. Evaluation of liquefaction potential sand deposits based on
and depth. Selected papers from the Chinese Journal of Geotechnical Engineering, observation of performance in previous earthquake. ASCE National Convention
pp. 92–100. (MO), pp. 481–544.
JRA, 1980. Japan road association, specification and interpretation of bridge design for Seed, H.B., Idriss, I.M., Arango, I., 1983. Evaluation of liquefaction potential using field per-
highway-part v. Resilient Design 14–15 (in Japanese). formance data. J. Geotech. Eng. ASCE 109, 458–482.
 S. Fabbrocino et al. / Engineering Geology 187 (2015) 84–97 97

Seed, H.B., Tokimatsu, K., Harder, L.F., Chung, R.M., 1985. Influence of SPT procedures in Terzaghi, K., Peck, R.B., 1948. Soil Mechanics in Engineering Practice. J. Wiley, New York
soil liquefaction resistance evaluations. J. Geotech. Eng. 111 (12), 1425–1445. (549 pp.).
Selli, R., 1962. Il Paleogene nel quadro della geologia dell'Italia centro-meridionale. Mem. Tortorici, L., 1975. Osservazioni geologiche sul flysch di San Bartolomeo nell'area
Soc. Geol. Ital. 3, 737–789 (in Italian). compresa tra il fiume Trigno e Benevento (Italia Meridionale). Boll. Soc. Geol. Ital.
Sgrosso, I., 1988. Nuovi elementi per un più articolato modello paleogeografico nell' 94, 1791–1809 (in Italian).
Appennino Centro-Meridionale. Mem. Soc. Geol. Ital. 41, 225–242 (in Italian). Tsiambaos, G., Sabatakakis, N., 2011. Empirical estimation of shear wave velocity from in
Sgrosso, I., 1998. Possibile evoluzione cinematica miocenica nell' Orogene centro-sud situ tests on soil formations in Greece. Bull. Eng. Geol. Environ. 70, 291–297.
appenninico. Boll. Soc. Geol. Ital. 117, 679–724 (in Italian). Valensise, G., Pantosti, D., Basili, R., 2004. Seismology and tectonic setting of the 2002 Mo-
Sgrosso, I., Amore, F.O., Ruggiero Taddei, E., 1988. L'inizio della sedimentazione terrigena lise, Italy, Earthquake. Earthquake Spectra 20, 23–37.
nel bacino molisano. Boll. Soc. Geol. Ital. 107, 73–79 (Roma (in Italian)). ViDEPI Project, d. Visibility of petroleum exploration data in Italy. World Wide Web Ad-
Shibata, T., 1970. Analysis of liquefaction of saturated sand during cyclic loading. Disaster dress:, http://unmig.sviluppoeconomico.gov.it/videpi/en/ (Last access May 2014).
Prev. Res. Inst. Bull. 13, 563–570. Vinale, F., d'Onofrio, A., Mancuso, C., Santucci de Magistris, F., Tatsuoka, F., 1999. The pre-
Sykora, D.W., Stokoe, K.H., 1983. Correlations of in-situ measurements in sand of shear failure behaviour of soils as construction materials. Proceedings of the Second Inter-
wave velocity. Soil Dyn. Earthq. Eng. 20, 125–136. national Symposium on Pre-failure Deformation Characteristics of Geomaterials; To-
Tatsuoka, F., Kohata, Y., 1995. Stiffness of hard soils and soft rocks in engineering applica- rino; Italy; 28 September 1999 through 30 September 1999, pp. 955–1007.
tions. Keynote Lecture. In: Shibuya, et al. (Eds.), Proc. of Int. Symposium Pre-Failure Yokota, K., Imai, T., Konno, M., 1991. Dynamic deformation characteristics of soils deter-
Deformation of Geomaterials vol. 2. Balkema, pp. 947–1063. mined by laboratory tests. OYO Tee. Rep. 3, p. 13
Tatsuoka, F., Shibuya, S., 1991. Deformation characteristics of soils and rocks from field
and laboratory tests. Keynote Lecture for Session No. 1, Proc. of the 9th Asian Regional
Conf. on SMFE, Bangkok vol. II, pp. 101–170.