Sie sind auf Seite 1von 11

Bull Eng Geol Environ (2007) 66:203–213

DOI 10.1007/s10064-006-0063-0

ORIGINAL PAPER

Empirical correlations between shear wave velocity and


penetration resistance for ground shaking assessments
Nilsun Hasancebi Æ Resat Ulusay

Received: 30 March 2006 / Accepted: 7 July 2006 / Published online: 29 August 2006
 Springer-Verlag 2006

Abstract Prediction of the ground shaking response uration du site et de contraintes économiques. Des
at soil sites requires knowledge of the soil, expressed in corrélations entre la vitesse de propagation des ondes
terms of shear wave velocity. Although it is preferable de cisaillement et des données pénétrométriques ont
to measure this dynamic soil parameter in situ, this is été évaluées et comparées avec des corrélations entre
often not economic at all locations. Existing correla- la vitesse de propagation des ondes de cisaillement et
tions between shear wave velocity and penetration des données SPT, ces dernières obtenues à partir de
resistance have been assessed in this study and com- données géotechniques et sismiques issues d’une zone
pared with correlations with SPT values obtained de forte sismicité en Turquie. Les résultats obtenus
based on geotechnical and geoseismic data collected confortent de précédentes études montrant que les
from a first-degree earthquake zone in Turkey. The données SPT apportent un paramètre significatif, tan-
results obtained support the findings of earlier studies dis que le type de sol ne joue pas un rôle important
that blow-count is a significant parameter in these pour les corrélations établies. Des équations de
correlations while type of soil has no important influ- régression ont été établies et permettent de prévoir la
ence. The regression equations developed in this study réponse d’un site à une sollicitation sismique.
compare well with most of the previous equations and
exhibit good prediction performance. It is noted that Mots clés Equation de regression Æ Sondage
better correlations are obtained when uncorrected géotechnique Æ Réfraction sismique Æ Vitesse de
blow-counts are used. propagation des ondes de cisaillement Æ Essai SPT Æ
Turquie
Keywords Correlation equation Æ Geotechnical
borehole Æ Seismic refraction Æ Shear wave velocity Æ
Standard penetration test Æ Turkey Introduction

It has been recognized for a very long time that


Résumé La prévision de la réponse d’un site à une earthquake damage is generally larger over soft sedi-
sollicitation sismique nécessite des connaissances sur le ments than on firm bedrock outcrops. This is particu-
sol, relatives en particulier à la vitesse de propagation larly important because most urban settlements have
des ondes de cisaillement. Il est préférable de mesurer occurred along river valleys over such young and soft
ce paramètre de dynamique des sols in situ. Cependant surface deposits. For this reason, particular consider-
ceci n’est pas toujours possible en raison de la config- ation is paid to the effect of local site conditions when
assessing ground motion characteristics for the seismic
design of buildings and other structures. Shear modu-
N. Hasancebi Æ R. Ulusay (&)
lus, damping ratio and shear wave velocity profiles are
Department of Geological Engineering,
Hacettepe University, Beytepe, Ankara 06800, Turkey important input parameters in site response analysis.
e-mail: resat@hacettepe.edu.tr Prediction of the ground shaking response at soil sites

123
204 N. Hasancebi and R. Ulusay

requires a knowledge of the stiffness of the soil, ex- classification tests, the use of borehole data from the
pressed in terms of shear wave velocity (Vs), which is previous work in the study site and statistical assess-
measured at small strain levels by in situ seismic ments. Based on the statistical assessments and taking
methods. While it is preferable to determine Vs directly into account the type of soil, a series of empirical
from field tests, it is often not economically feasible at equations for the prediction of Vs from SPT-N
all locations. For this reason, a reliable correlation were developed and compared with those suggested in
between Vs and penetration resistance would be a previous studies in order to evaluate the prediction
considerable advantage, reducing the number of field capability of the equations.
verifications required.
In this study, which is an integral part of a research
study on soil amplification (Hasancebi 2005), the var- General setting of the study site
iation of shear wave velocity measured by seismic
refraction and SPT blow-count (SPT-N) was investi- Geology and seismotectonics
gated and new correlations have been developed for
estimating the shear wave velocity. For this purpose, a Yenisehir town is located within an alluvial basin which
site with recorded high seismicity at Yenisehir, foun- is surrounded by ridges both to the north and south
ded on an alluvial plain located in the Marmara Region (Fig. 2). The study site is geologically represented by
of Turkey (Fig. 1), was selected. The investiga- the basement rocks of pre-Neogene age and Neogene
tion programme included SPT borings at different and Quaternary deposits. The basement rocks seen in
locations, seismic refraction studies, laboratory soil the southern part of the site consist of schists and

Fig. 1 Location map of the


study site

123
Connection of shear wave velocity with SPT 205

Fig. 3 Siesmotectonic map of the Eastern Marmara Region


(after Doyuran et al. 2000)

Doyuran et al. (2000) and the State Port and Airport


Directory (DLH 2002). These previous studies in-
cluded a total of 37 boreholes and associated SPTs.
Fig. 2 Geological map of the study site, and locations of Twelve of these boreholes are relevant to the present
geotechnical and geoseismic investigations (modified from study. It is well known that the average shear wave
Doyuran et al. 2000) velocity in the upper 30 m of the ground is an impor-
tant factor for ground characterization (Borcherdt
marbles. The Neogene deposits which appear on the 1994; Dobry et al. 2000). Therefore, for the present
gentle slopes at the north and south consist predomi- study, the boreholes were planned to penetrate to a
nantly of loosely cemented conglomerate, sandstone, depth of 30 m if possible. In fact, 9 extended to 30.45 m
claystone and marl and unconformably overlie the while the others ranged between 4.5 and 17 m. SPT
basement rocks (Genc 1992). The Quaternary deposits tests were conducted at 1 m intervals and the samples
are represented by alluvial soils and detritus and are from SPT tubes were used for laboratory testing. The
observed in the middle of the basin. Based on the data groundwater table in each hole was also measured.
from the records of the boreholes drilled by the State Locations of the previous and recent boreholes and the
Hydraulic Works (DSI), the thickness of the alluvial geoseismic investigations are shown in Fig. 2.
sequence in the basin ranges between 25 and 115 m. The most commonly used seismic methods for
Yenisehir is located within a first-degree earthquake velocity logging are the cross-hole and down-hole
zone of Turkey (GDDA 1996). The study site is sur- techniques. Seismic refraction is largely used in deter-
rounded by a number of active faults as shown in mining the dynamic properties of the underlying layers.
Fig. 3. The southern strand of the North Anatolian In this study, shear wave velocities were measured
Fault Zone (NAFZ) and the Bursa and Inonu– using seismic refraction with the assistance of the
Eskisehir faults are the most important earthquake General Directorate of Disaster Affairs (Dikmen et al.
sources in the study site. The 1999 Kocaeli earthquake, 2004) and Geophysical Engineering Department of
which resulted in extensive loss of life and damage to Ankara University. These measurements were taken at
structures particularly in the Marmara Region, was the locations of nine boreholes drilled during the study.
also felt in Yenisehir and its vicinity. Due to some restrictions at the locations of boreholes
H6 and H7 and the very shallow depth of borehole
SPT soundings, geoseismic investigations H12, seismic studies were not undertaken at these
and laboratory testing locations.
A total of 149 specimens extracted from SPT tubes
In the present and future settlement areas of Yenisehir, were tested in the laboratory to determine their grain
geotechnical studies for the assessment of foundation size distribution and Atterberg limits. The tests were
conditions and a railway route were conducted by conducted in accordance with the methods given by

123
206 N. Hasancebi and R. Ulusay

Subsurface conditions

The data from the previous and recent boreholes and


the resistivity study (Dikmen et al. 2004) suggest that
the alluvial sequence generally starts with silty clay.
This clay, with high SPT-N values, is a stiff soil. Below
this, there exists medium dense to loose silty sand.
However, at some localities the silty clay may also ap-
pear below the sandy zone. Occasionally, gravel layers
of variable thickness can be observed in the Quaternary
deposits at shallow depths. A typical cross-section
through the Yenisehir settlement area, some selected
geotechnical logs and two typical seismic refraction
profiles showing the variation of Vs at the locations of
boreholes H3 and H8 are depicted in Fig. 5, respec-
tively. As can be seen from the Vs–SPT(N) versus depth
plots for some selected borehole locations in Fig. 6, Vs
increases with increasing SPT values.
Statistical evaluations of the data from grain size
analyses are given in Table 1. Sand sized material in
the Quaternary deposits is dominant at the southern
part of the site and these soils are represented by SP
and SW, and SC and SM soil classes. Towards the
north, grain size decreases and clays and silty clays with
high plasticity dominate. Most of the tested soils from
Fig. 4 Geological cross sections (a) and some typical engineer-
ing logs (b) illustrating the subsurface ground conditions at the north of the site fall into the CH and CL soil
Yenisehir settlement and its vicinity classes.
The groundwater table in the study site is shallow,
ASTM (1994). Based on these results, the specimens generally ranging between 3 and 8 m. However, based
were classified according to the Unified Soil Classifi- on the data from Doyuran et al. (2000), it is deeper
cation System. (‡14 m) at the north and shallower in the south.

Fig. 5 Seismic refraction


profiles at the locations of
boreholes H3 and H8

123
Connection of shear wave velocity with SPT 207

Table 1 Statistical evaluation of grain size distribution of soil


samples
Grain size Max Min Mean Standard Standard
error deviation

Gravel (%) 0 49 11 12.47 0.084


Sand (%) 4 100 49 27.74 0.186
Silt (%) 0 76 30 23.91 0.160
Clay (%) 0 34 10 9.3 0.066

Fig. 7 Correlations between Vs and SPT-N for all soils (a), sandy
soils (b) and clayey soils (c)
Fig. 6 Variation of Vs and SPT-N with depth at some borehole
locations
Vs measurements at all locations. Many correlations
Proposed empirical correlations for Vs–SPT(N) between Vs and penetration resistance have been
proposed; 17 are quoted in Table 2, the majority based
While it is preferable to determine Vs directly from on uncorrected SPT-N values. Sykora and Stokoe
field tests, it is often not economically feasible to make (1983) suggest that geological age and soil type are not

123
208 N. Hasancebi and R. Ulusay

Table 2 Some existing correlations between Vs and SPT-N


Author(s) Vs (m/s)
All soils Sands Clays

Ohba and Toriumi (1970) Vs = 84N0.31 – –


Imai and Yoshimura (1970) Vs = 76N0.33 – –
Fujiwara (1972) Vs = 92.1N0.337 – –
Ohsaki and Iwasaki (1973) Vs = 82N0.39 – –
Imai (1977) Vs = 91N0.337 Vs = 80.6N0.331 Vs = 80.2N0.292
Ohta and Goto (1978) Vs = 85.35N0.348 – –
Seed and Idriss (1981) Vs = 61N0.5 – –
Imai and Tonouchi (1982) Vs = 97N0.314 – –
Sykora and Stokoe (1983) – Vs = 100.5N0.29 –
0.202
Jinan (1987) Vs = 116.1(N + 0.3185) – –
Lee (1990) – Vs = 57.4N0.49 Vs = 114.43N0.31
Sisman (1995) Vs = 32.8N0.51 – –
Iyisan (1996) Vs = 51.5N0.516 – –
Jafari et al. (1997) Vs = 22N0.85 – –
Pitilakis et al. (1999) – Vs = 145(N60)0.178 Vs = 132(N60)0.271
Kiku et al. (2001) Vs = 68.3N0.292 – –
Jafari et al. (2002) – – Vs = 27N0.73

important parameters in determining Vs, while the and 1:2 slopes, with smaller Vs values (Vs < 250 m/s)
SPT-N value is of prime importance. However, as can falling close to the line 1:1.
be seen from Table 2, some investigators have also The correlations from the present study are plotted
proposed correlations between Vs and SPT-N for dif- in Fig. 9 to assess the effect of soil type. Figure 9
ferent soils, such as clays, silts and sands. In addition, suggests that the correlations for different soil catego-
the soil properties considered in the regression analy- ries yield similar values of Vs indicating that soil type
ses for some correlations included stress-corrected Vs, has little effect on these correlations. This is consistent
energy-corrected SPT-N, energy- and stress-corrected with the findings of Iyisan (1996). However, Iyisan
SPT-N, depth (D) and fines content (FC). (1996), who also studied gravely soils in Turkey, indi-
In this study, 97 data pairs (Vs and SPT-N) were cated that the situation for gravels is different because
employed in the assessments. The correlations were the correlation for gravels estimates higher Vs values
developed using a simple regression analysis for the when compared to those from other soils, due to the
existing database. In the first series of analyses, new effect of grain size and cementation.
relationships were proposed between uncorrected Vs Equations 1–3 are plotted in Fig. 10a, c together
(m/s) and corresponding SPT-N values in three cate- with several of the earlier regression equations given in
gories, i.e. for all soils, sandy soils and clayey soils Table 2. Except the relationships of Ohsaki and Iwa-
(Fig. 7). Because few data from the silty layers were saki (1973), Seed and Idriss (1981), Sisman (1995),
available, this category was not included in the evalu- Iyisan (1996), Jafari et al. (1997) and Kiku et al. (2001)
ations. The following relationships with their correla- in Fig. 10a (which were recommended for all soils), all
tion coefficients (r) are proposed between Vs (m/s) and the equations including the equation of the present
SPT-N values for the three different soil categories. study (Eq. 1) yield similar Vs values. There is only a
slight difference between Eq. 1 and those developed by
Vs ¼ 90N 0:309 ðr ¼ 0:73Þ; All soils ð1Þ Ohba and Toriumi (1970) and Imai and Yoshimura
(1970); Eq. 1 proposed in this study estimates Vs values
considerably closer to those derived from most of the
Vs ¼ 90:8N 0:319 ðr ¼ 0:65Þ; Sandy soils ð2Þ existing equations.
Similar comparisons made for sands (Fig. 10b)
indicated that except the equation developed by Lee
Vs ¼ 97:9N 0:269 ðr ¼ 0:75Þ; Clayey soils ð3Þ (1990), the proposed equation (Eq. 2) compares well
with the other equations for the prediction of the Vs of
Comparisons between the measured Vs and Vs pre- sands. Based on the distribution of the plotted data, the
dicted from Eqs 1–3 are presented in Fig. 8. The equation of Lee (1990) generally over-predicts Vs for
plotted data are scattered between the lines with 1:0.5 N > 20 and under-predicts Vs for N £ 20.

123
Connection of shear wave velocity with SPT 209

Fig. 10 Comparisons between proposed and previous correla-


tions for Vs and SPT-N for all soils (a), sandy soils (b) and clayey
soils (c)

The comparison for clays given in Fig. 10c suggests


that the equations developed by Imai (1977) and Lee
(1990) predict higher Vs values when compared to
Fig. 8 Measured versus predicted shear wave velocities for all those from Eq. 3 of the present study. The equation of
soils (a), sandy soils (b) and clayey soils (c)
Jafari et al. (2002) differs from the other three equa-
tions and yields under-predicted and over-predicted Vs
values for SPT-N £ 20 and SPT > 20 conditions,
respectively. The specific geotechnical conditions of
the studied area are probably the main cause of this
while the quantity of the processed data, the SPT
procedure and the different methods of shear wave
velocity measurements employed in various studies
may be other causes of difference.
In addition to the comparisons shown in Fig. 10, the
scaled percent error (Eq. 4) versus cumulative fre-
quency graphs have also been drawn in Fig. 11.

Scaled percent error ¼ ½ðVsc  Vsm Þ=Vsm   100 ð4Þ

where Vsc and Vsm are the predicted and measured


Fig. 9 Effect of soil type on Vs–SPT(N) relationship shear wave velocities, respectively.

123
210 N. Hasancebi and R. Ulusay

Fig. 11 Scaled percent error


of Vs predicted for all soils
(a), sandy soils (b) and clayey
soils (c)

As seen in Fig. 11a, c, about 85% of the Vs values for striking energy during the test employed in this
predicted from Eqs. 1 to 3 for all soils, sands and clays study (donut-type hammer raised and dropped by two
respectively, are within 20% of the scaled percent er- turns of rope). The developed relationships for differ-
ror, indicating a better estimate than those from the ent soils are given in Fig. 12a, c. When the correlation
existing equations. coefficients obtained from Vs to N60 relationships are
The relationship between Vs and energy-corrected compared to those obtained from Eqs. 1 to 3, the
SPT-N(N60) was also investigated and equations for all equations based on uncorrected SPT-N values provide
soils, sands and clays were established. These were a somewhat better fit than the equations based on
compared to those suggested by Pitilakis et al. (1999) energy-corrected measurements. This situation is also
who previously investigated N60–Vs relationships for seen from Fig. 12d, f. The equations given in Fig. 12b, c
clays and sands. The SPT blow-counts were corrected are plotted in Fig. 13a, b respectively, together with the

123
Connection of shear wave velocity with SPT 211

Fig. 12 Vs–N60 relationships


for all soils (a), sandy soils (b)
and clayey soils (c), and
measured versus predicted
shear wave velocities for all
soils (d), sandy soils (e) and
clayey soils (f)

regression equations developed by Pitilakis et al. purposes. The results obtained from the study support
(1999) for sands and clays. As shown in Fig. 13a, the the findings of earlier work suggesting that blow-count
equation in Fig. 12b compares well with the regression is a significant parameter in Vs–SPT(N) correlations,
equation of Pitilakis et al. (1999) for sands. However, while the type of soil has little influence.
the equation of Pitilakis et al. (1999) for clays yields The regression equations developed in this study
considerably higher Vs estimations when compared to compare well with most of the previous equations and
those from the equation developed in this study exhibit a good prediction performance. The equations
(Fig. 13b). It appears from these assessments that the based on uncorrected SPT-N values provide a some-
equations based on uncorrected SPT-N values are what better fit than the equations based on energy-
preferable for indirect estimations of Vs. corrected SPT-N values. Therefore, the use of an
equation developed for all soils based on uncorrected
blow-counts is recommended for practical purposes.
Conclusions The regression equations developed provide a viable
way of estimating Vs from SPT blow-count for pre-
In this study, based on the geotechnical and geoseismic liminary regional ground shaking mapping and site-
data from the Yenisehir settlement situated in the specific response analysis. The differences between
Marmara Region of Turkey, an attempt was made to existing and proposed equations are mainly due to the
develop new relationships between SPT-N and Vs to specific geotechnical conditions of the studied sites, the
indirectly estimate the Vs to be used for practical quantity of processed data and the procedures used in

123
212 N. Hasancebi and R. Ulusay

Borcherdt RD (1994) Estimates of site depending response


spectra for design methodology and justifications. Earth-
quake Spectra 10(4):617–654
Dikmen U, Mirzaoglu M, Sipahi O (2004) Geophysical investi-
gations at Yenisehir (Bursa) settlement area. General
Directorate of Disaster Affairs of Turkey, Ankara (in
Turkish, unpublished)
DLH (2002) Report on investigation and engineering works for
Ayazma-Inönü railway route (in Turkish, unpublished)
Dobry R, Borcherdt RD, Crouse CB, Idriss IM, Joyner WB,
Martin GR, Power MS, Rinne EE, Seed RB (2000) New site
coefficient and site classification system used in recent
building code provisions. Earthquake Spectra 16(1):41–67
Doyuran V, Kocyigit A, Yazicigil H, Karahanoglu N, Toprak V,
Topal T, Suzen ML, Yesilnacar E, Yilmaz KK (2000)
Geological and geotechnical investigation at Yenisehir
settlement area. Middle East Technical University, Ankara
(in Turkish, unpublished)
Fujiwara T (1972) Estimation of ground movements in actual
destructive earthquakes. In: Proceedings of the fourth
European symposium on earthquake engineering, London,
pp 125–132
Genc S (1992) Geology of the Bursa region. In: Proceedings of
the international symposium on the geology of the Black
Sea region, Guide Book, Ankara, pp 22–25
GDDA (1996) Earthquake zonation map of Turkey. Ankara
Hasancebi (Okan) N (2005) Geotechnical evaluation of soil
amplification at Yenisehir (Bursa) district. M.Sc. Thesis,
Hacettepe University, Geological Engineering Department,
Ankara (in Turkish, unpublished)
Imai T (1977) P-and S-wave velocities of the ground in Japan. In:
Proceedings of the IX international conference on soil
mechanics and foundation engineering, vol 2, pp 127–132
Imai T, Yoshimura Y (1970) Elastic wave velocity and soil
properties in soft soil (in Japanese). Tsuchito-Kiso 18(1):17–
22
Fig. 13 Comparisons between N60 and Vs equations proposed in Imai T, Tonouchi K (1982) Correlation of N-value with S-wave
this study and by Pitilakis et al. (1999) for sand soils (a) and velocity and shear modulus. In: Proceedings of the 2nd
clayey soils (b) European symposium of penetration testing, Amsterdam,
pp 57–72
Iyisan R (1996) Correlations between shear wave velocity and in-
undertaking the SPTs and geoseismic surveys. In view situ penetration test results (in Turkish). Chamber of Civil
Engineers of Turkey, Teknik Dergi 7(2):1187–1199
of this, these empirical equations should be carefully Jafari MK, Asghari A, Rahmani I (1997) Empirical correlation
used and wherever possible checked against measured between shear wave velocity (Vs) and SPT-N value for south
Vs values. of Tehran soils. In: Proceedings of the 4th international
conference on civil engineering, Tehran, Iran (in Persian)
Acknowledgments This study was supported by Project No. Jafari MK, Shafiee A, Ramzkhah A (2002) Dynamic properties
0302602008 of the Research Projects, Division of Hacettepe of the fine grained soils in South of Tehran. JSEE 4(1):25–35
University. The authors would like to thank the General Jinan Z (1987) Correlation between seismic wave velocity and
Directorate of Disaster Affairs and the geophysical team of this the number of blow of SPT and depth. Selected papers from
organization for their cooperation and the geophysical surveys; the Chinese Journal of Geotechnical Engineering, ASCE,
the Municipality of Yenisehir for providing the logistic support; 92–100
the Geophysical Engineering Department of Ankara University Kiku H, Yoshida N, Yasuda S, Irisawa T, Nakazawa H, Shimizu
for seismic refraction equipment and interpretation of the mea- Y, Ansal A, Erkan A (2001) In-situ penetration tests and
sured data; and the General Directorate of State Hydraulic soil profiling in Adapazari, Turkey. In: Proceedings of the
Works (DSI) for permission to use the borehole logs. ICSMGE/TC4 satellite conference on lessons learned from
recent strong earthquakes, pp 259–265
Lee SHH (1990) Regression models of shear wave velocities.
J Chin Inst Eng 13:519–532
References Ohba S, Toriumi I (1970) Dynamic response characteristics of
Osaka Plain. In: Proceedings of the annual meeting AIJ (in
ASTM (1994) Annual book of ASTM standards—soil and rock, Japanese)
building stones. Section 4, Constructions, vol 04.08. ASTM Ohsaki Y, Iwasaki R (1973) On dynamic shear moduli and
Publications, Philadelphia Poisson’s ratio of soil deposits. Soil Found 13(4):61–73

123
Connection of shear wave velocity with SPT 213

Ohta Y, Goto N (1978) Empirical shear wave velocity equations previous earthquakes. ASCE National Convention,
in terms of characteristics soil indexes. Earthquake Eng Missouri, pp 81–544
Struct Dyn 6:167–187 Sisman H (1995) An investigation on relationships between
Pitilakis K, Raptakis D, Lontzetidis K, Tika-Vassilikou T, Jong- shear wave velocity, and SPT and pressuremeter test results.
mans D (1999) Geotechnical and geophysical description of M.Sc. Thesis, Ankara University, Geophysical Engineering
Euro-Seistests, using field and laboratory tests, and moderate Department, Ankara (in Turkish, unpublished)
strong ground motions. J Earthquake Eng 3(3):381–409 Sykora DE, Stokoe KHII II (1983) Correlations of in-situ mea-
Seed HB, Idriss IM (1981) Evaluation of liquefaction potential surements in sands of shear wave velocity. Soil Dyn Earth-
sand deposits based on observation of performance in quake Eng 20(1-4):125–136

123

Das könnte Ihnen auch gefallen