You are on page 1of 15

Soil Dynamics and Earthquake Engineering 31 (2011) 891–905

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering
journal homepage: www.elsevier.com/locate/soildyn

Transient kinematic pile bending in two-layer soil
Stefania Sica a, George Mylonakis b, Armando Lucio Simonelli a,n
a b 

degli Studi del Sannio, Piazza Roma 21, 82100 Benevento, Italy Dipartimento di Ingegneria, Universita Department of Civil Engineering, University of Patras, University Campus, Rio 26500, Greece

a r t i c l e i n f o
Article history: Received 16 September 2010 Received in revised form 25 January 2011 Accepted 2 February 2011 Available online 31 March 2011 Keywords: Piles Soil–structure interaction (SSI) Kinematic interaction Numerical modelling

abstract
The dynamic response of piles to seismic loading is explored by means of an extensive parametric study based on a properly calibrated Beam-on-Dynamic-Winkler-Foundation (BDWF) model. The investigated problem consists of a single vertical cylindrical pile, modelled as an Euler–Bernoulli beam, embedded in a subsoil consisting of two homogeneous viscoelastic layers of sharply different stiffness resting on a rigid stratum. The system is subjected to vertically propagating seismic S waves, in the form of a transient motion imposed on rock outcrop. Several accelerograms recorded in Italy are employed as input motions in the numerical analyses. The paper highlights the severity of kinematic pile bending in the vicinity of the interface separating the two soil layers. In addition to factors already investigated such as layer stiffness contrast, relative soil–pile stiffness, interface depth and intensity of ground excitation, the paper focuses on additional important factors, notably soil material damping, stiffness of Winkler springs and frequency content of earthquake excitation. Existing predictive equations for assessing kinematic pile bending at soil layer interfaces are revisited and new regression analyses are performed. A synthesis of findings in terms of a set of simple equations is provided. The use of these equations is discussed through examples. & 2011 Elsevier Ltd. All rights reserved.

1. Introduction During the passage of seismic waves through soft deposits, embedded piles tend to deform in a different manner with respect to the surrounding soil. The difference between soil and pile motion depends on several factors, notably soil layering, pile–soil stiffness contrast, excitation frequency and kinematic constraints at the pile head and tip [1–4]. As curvatures are imposed to the pile body by the vibrating soil, bending and shearing will develop even in the absence of a superstructure. The associated pile bending moments are, thereby, referred to as ‘‘kinematic’’, to be distinguished from those generated by loads acting at the pile head due to the dynamic response of the superstructure (so-called ‘‘inertial’’ moments). Kinematic and inertial bending moments constitute complementary aspects of a unique phenomenon known as soil–pile–structure interaction (SPSI). Reviews of the subject have been published, among others, by Novak [5], Pender [6] and Gazetas and Mylonakis [7]. Evidence from case histories – as documented by Mizuno [8] and other Japanese researchers [9–12] – or from recent experimental investigations on physical pile models in centrifuge and 1g earthquake simulators [13–18] have elucidated the important

role of kinematic interaction in seismic response of pile foundations. Kinematic bending is significant (as compared to its inertial counterpart) particularly in correspondence to stiff pile caps and soil layer interfaces. The latter may explain the concentration of seismic demand at depths where inertial effects are negligible. The accumulated evidence has generated significant interest in exploring theoretical and analytical aspects of the phenomenon and developing seismic regulations to incorporate it into design procedures [19–21]. Following the early work by Margason and Halloway [22], theoretical investigations of the problem began in the 1980s [1,23–27] and continued into the 1990s and beyond [2–4,28–31]. In 2005, a systematic research effort was initiated in Italy under the auspicious of the ReLUIS project (University Network of Seismic Engineering Laboratories), which has lead to a number of publications [32–43]. The main goal of the project was to produce engineering provisions to be incorporated into the new national seismic code [21], which is compulsory in Italy since July 2009. 1.1. Unresolved issues At present, it appears that many aspects of kinematic pile bending are well understood, whilst others require further research and remain unresolved. First, most of the published results concentrate on flexible piles (i.e., piles whose lengths are greater than the so-called ‘‘active pile length’’ [44], embedded in

Corresponding author. E-mail addresses: stefsica@unisannio.it (S. Sica), mylo@upatras.gr (G. Mylonakis), alsimone@unisannio.it (A.L. Simonelli). 0267-7261/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.soildyn.2011.02.001

n

Mylonakis et al.1. 1). On the basis of these results. System considered. [29] and Mylonakis [62]. without de-convolution or consideration for rock outcrop effects.40.3]. 1. Last but not least. or even neglected altogether. Analysis method and validation The response to vertical S-wave excitation of a single vertical cylindrical solid pile embedded in layered soil is investigated through a Beam-on-Dynamic-Winkler-foundation (BDWF) formulation. corresponding to the Fixed head h1 Vs1. The configuration adopted for the pile and the soil is shown in Fig. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 two-layer soils with the interface placed deep beneath the surface [2–4. With reference to kinematic pile moments at soil layer interfaces.43]. 1. ρ1 D1. with emphasis on frequency content of ground motion.3. Likewise. ρ2 D2. the effect of material plasticity for both soil and pile occurring under stronger earthquakes needs to be clarified. (iii) a sensitivity analysis of the results as function of soil damping employed in the analysis of free-field response. this type of modelling is known to provide satisfactory accuracy as compared to rigorous numerical schemes. contrary to methods that require frequency-independent parameters to obtain the response directly in the time domain. the frequency content of the accelerograms adopted as input motion is quantified through the predominant frequency. which implements a Beam-on-Dynamic-Winkler-Foundation (BDWF) formulation encompassing finite-element-based springs and dashpots distributed along the pile axis. the research groups working as part of the Italian ReLUIS project conducted a comparative study of predictions provided by different numerical tools on conceptual prototypes consisting of single piles embedded in two-layer profiles under the assumption of vertically propagating S waves (Fig. Second.892 S. k ¼ k(o) and c ¼ c(o). Mylonakis [3. new regression analyses are carried out to correlate peak pile bending moments with the above parameters. The work at hand focuses only on some of the aforementioned issues. Such springs and dashpots connect the pile to the free-field soil. there is a lack of simple formulations for assessing corresponding moments at the pile head regardless of layer thickness and pile length. Fourth. fp. which is consistent with a zone of high seismicity according to the Italian seismic zonation of 2003 [53].24. ν1 EpI. As the fundamental aspects of the method are well-known. Although simplified.71] by comparing pile bending moments to those obtained by a variety of continuum approaches solved by FEM and BEM procedures. . The analyses are performed using an extension of the numerical tool developed by Mylonakis et al. [4. the paper presents: (i) a validation of the BDWF procedure against other available solutions. the recordings were scaled at a peak acceleration of 0. as well as between inertial and kinematic effects. The parameter analyses were carried out by varying the dimensionless interface depth (h1/d) while keeping the overall soil depth constant. The analyses were performed by adopting a suite of Italian accelerograms (Table 1) selected from the Italian database SISMA [52]. Available predictive equations are either pure frequency-domain approaches [3]. thereby. Soil–pile interaction is modelled through a set of continuously distributed springs and dashpots. Mylonakis et al. geometric factors and earthquake excitations. have been calibrated against the results of finite-element and boundary-element analyses. the effect of certain modelling assumptions involving soil damping. [29] and Mylonakis [62] reformulated and extended earlier frequency-domain solutions [2] to the time domain using a Discrete Fourier Transform (DFT) approach as described by Veletsos and Ventura [51]. Fifth. solved using the hybrid analytical–numerical algorithm of Fig. stiffness of Winkler springs and wave propagation on kinematic bending moments needs to be better quantified. 2. the depth of the upper layer and the earthquake waveform. Key parameters such as number of earthquake cycles and frequency content are currently accounted for in an approximate way.35 g. or mixed frequency–time domain formulations based on a limited number of accelerograms [4. [29] and Mylonakis [62]. the parameters of which.45–50]. despite recent progress [33.35]. often allowing closed-form solutions to be obtained [1. Less research has been carried out on short piles and/or interfaces located close to the pile head. becomes significant and. Third. Sica et al. For the purposes of this validation the accelerograms were directly applied at the base of the soil column. Validation Recently.4. Finally. a set of conclusions and recommendations are produced for implementing the findings in routine engineering calculations. The approach has been validated in Mylonakis et al.31].62] and Nikolaou et al. Specifically it aims at reporting the results of an extensive parametric investigation carried out on single piles in layered. ν2 h2 d 2. viscoelastic soil accounting for different material properties. the wave-induced motion of the latter serves as the support excitation for the pile–soil system. as done by the other ReLUIS research groups.4. ρp L Vs2. kinematic pile head moments have not been addressed in cases deviating significantly from that of a homogeneous soil. For such systems the interplay between pile head and interface moments. (iv) a comprehensive parametric study on a two-layer soil profile as function of stiffness contrast between the upper and the lower layer. the effect of material nonlinearity in the soil (and the pile itself) needs to be explored in a more systematic way. a proper summation rule is essential for combining the contributions of the individual response components. whereas a number of simplified procedures for estimating kinematic pile bending moments at the interface of two soil layers are available [24. the effect of the transient nature of input motion on the development of kinematic bending moments along the pile requires further investigation. only a brief description is given here. In Table 1. This can accommodate precisely the frequency dependence of the spring and dashpot moduli. (ii) a sensitivity study of pile bending as function of spring stiffness adopted in Winkler models. [29].

Peak kinematic pile bending strain at interface level as function of depth for A-TMZ270 (top) and A-TMZ000 (bottom) Tolmezzo recordings.5 5.35 g). when normalised by pertinent strain parameters such as soil shear strain or site amplification functions.5 5. different damping schemes (e. [33]) BDWF (Mylonakis et al.27 0.31 0.5 5. [29]) BDWF (Dezi et al.7 7.32 0.8 6. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 893 Table 1 Ground motions employed in the analysis (all records are scaled to PGA ¼ 0.34 0. (ii) it is directly measurable.5 Mg/m3. as defined in [54] on the basis of frequency content.0 4.8 4.5 3. Particularly satisfactory is the comparison between the BDWF approaches of Mylonakis et al. [29].6 A A A A A A A A A A A A A A A A A A Aftershock. [42] and the two continuum approaches (FEM or BEM).0 0.1 8.3 2. D1 ¼ D2 ¼ 10%. Vs1 ¼ 100 m/s. Predominant frequency from response spectra.0 8..3 7.9 6 5. L ¼ 20 m.21 0. From Fig.36 0. n1 ¼ n2 ¼ 0.1 5.g. Sensitivity of BDWF analyses on selection of Winkler springs In the BDWF approach the stiffness k of the springs connecting the pile to the soil is generally defined as k ¼ d Es ð2Þ .19 0.9 6.38 0. as well as different boundary conditions at the pile tip.6 6.2 1. 2.5 εp x 10-3 1.. [33]. ep is plotted as function of the dimensionless embedment factor h1/d. Vs2 ¼ 400 m/s.3 5.5 1.0 2.0 2. Ep ¼ 2.8 6.5 6.4 6.5 5. The continuum approach [32] is implemented through a frequency-domain BEM technique that makes use of the soil stiffness matrices derived by Kausel and Roesset [55] to simulate the response of a horizontally layered deposit. Dezi et al.2. Cairo et al.26 0.7 5.3 5.23 0.1 5.5 12.1 5.3 5.6 3.1 6.1 5.5 Â 107 kPa and rp ¼ 2. The last approach [56] is implemented by means of the commercial finite-element code ANSYS [57] using very fine discretization.6 6.6 5.6 5.5 5. [56]) BDWF (Cairo et al.32 0.0 25.3 5.1 4. other two approaches – one continuum and another BDWF – were developed by Cairo and Dente [32] and Cairo et al. [42] and was validated by detailed 3D finite-element analyses [43] using the ABAQUS platform [70].7 5.8 5.6 3.1 Epicentral distance (km) 23 23 32 32 21 10 10 11 12 12 20 23 23 8 8 8 10 10 PGA (g) fp (Hz) b fm (Hz) c Soil type 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 a b c A-TMZ270 A-TMZ000 A-STU270 A-STU000 A-AAL018 E-NCB090 E.3 4.9 Mg/m3.NCB000 R-NCB090 J-BCT000 J-BCT090 E-AAL108 B-BCT000 B-BCT090 TRT000 C-NCB000 C-NCB090 R-NC2090 R-NC2000 Tolmezzo-Diga Ambiesta Tolmezzo-Diga Ambiesta Sturno Sturno Assisi-Stallone Nocera Umbra-Biscontini Nocera Umbra-Biscontini Nocera Umbra-Biscontini Borgo-Cerreto Torre Borgo-Cerreto Torre Assisi-Stallone Borgo-Cerreto Torre Borgo-Cerreto Torre Tarcento Nocera Umbra-Biscontini Nocera Umbra-Biscontini Nocera Umbra 2 Nocera Umbra 2 Friuli Friuli Campano Lucano Campano Lucano Umbria Marche Umbria Marche a Umbria Marche a Umbria Marche a Umbria Marche a Umbria Marche a Umbria Marche a Umbria Marche Umbria Marche Friuli a Umbria Marche a Umbria Marche a Umbria Marche a Umbria Marche a 06/05/76 06/05/76 23/11/80 23/11/80 26/09/97 06/10/97 06/10/97 03/04/98 14/10/97 14/10/97 06/10/97 26/09/97 26/09/97 11/09/76 03/10/97 03/10/97 03/04/98 03/04/98 0.S.0 0.4. different discretizations of the pile.19 0.0 5. (iii) it can be used to quantify damage.0 0 5 10 15 h1/d Fig. 2a in correspondence to the ratios h1/d ¼ 25 and 32 may be attributed to the selected values of Winkler springs.19 0.0 1.6 10. fm. and through the mean frequency. [33].. 2. The results obtained with the selected approach are compared to the kinematic bending moments provided by two different BDWF formulations and two continuum solutions employed in the ReLUIS project. Average frequency according to Rathje et al. Some of these factors are discussed in the ensuing. [54]. Some discrepancies appearing in Fig.0 1.5 8.3 10.38 1. [32]) FEM (Di Laora. The use of bending strain over other normalisation schemes is desirable in such problems as [3]: (i) it is dimensionless. EpI the bending stiffness of the pile and d/2 the distance from the pile centerline to the outer fibre of the pile cross section. 2 the results obtained by the BDWF method employed in this study are represented in terms of maximum pile bending strain: M d ep ¼ max Ep I 2 ð1Þ Mmax being the maximum kinematic bending moment at the interface. Sica et al.5 6. The first BDWF approach has been developed by Dezi et al.5 εp x 10-3 BEM (Cairo & Dente.31 0. 2 a satisfactory comparison is noted among the results provided by the aforementioned methods. d ¼ 0.18 0. # Record label Station name Earthquake Date (d/m/yr) Magnitude (Mw) 6.33 0.5 0.5 0.. 2. In Fig. r1 ¼ r2 ¼ 1. 20 25 30 35 maximum value of the acceleration response spectrum for 5% damping.60 m.19 0. (iv) it can generate bi-dimensionless transfer functions.0 2. as yield amplitudes do not vary significantly among common structural materials (being of the order of 10 À 3). Rayleigh versus linearly hysteretic). [42]) 1.0 6.

using the results from the finite-element formulation of Blaney et al. εp x 10-3 0 0 δ = 1. [63] and Syngros [64]. 1.4 2.2 to 2.2 1. There. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 where Es is Young’s modulus of the soil and d is a dimensionless coefficient.30 (Kavvadas & Gazetas. Note that for a hollow cylindrical pile. Roesset [60] recommended the value: d ¼ 1:2 ð3Þ [42. among others. respectively. With reference to kinematic pile bending in a two-layer deposit. Es1 the Young modulus of the upper soil layers and G1.78(Mylonakis. Naturally kinematic bending moment increases with increasing d.78 2. L the pile length. [60]) 60 2.894 S. Kavvadas and Gazetas [2] related d to pile and soil properties according to the equation:  À1=8  1=8  1=12  À1=30 3 Ep L h1 G2 d¼ ð4Þ d h2 G1 1Àn2 Es1 where n is the Poisson ratio of the soil material (common in both layers). Mylonakis [3] simplified the above equation for the case of long piles in two-layer soil as follows:  À1=8 E dffi6 p ð5Þ Es1 The effect of Winkler parameter d on kinematic pile bending moments was investigated by means of an extensive parametric analysis.5 1 1. In that region an increase in d from 1. Ep the pile Young’s modulus.62]. [2]) 5 2. 4. 3. including the authors.61. For instance. (Mylonakis.8.20 (Roesset.8 Roesset. Sica et al. h1 and h2 the thickness of the upper and lower layers. Based on these results the following approximate relation was derived:  À1=40 Ep M ðdÞ ¼ ð0:97 þ 0:17dÞ ð6Þ Mð1:2Þ Es1 This equation is in agreement with the solution of Dobry and O’Rourke [24] as to the dependence of pile bending on pile–soil stiffness contrast. Note that for soft piles (Ep/Es1 r 500) the effect regardless of other problem parameters.2 (Ep/Es1 = 500) Kavvadas & Gazetas.4 A-AAL018 A-STU000 A-STU270 ATMZ000 ATMZ270 BBCT000 B-BCT090 C-NCB000 C-NCB090 E-AAL108 ENCB000 ENCB090 J-BCT000 J-BCT090 R-NC2000 R-NC2090 R-NCB090 TRT000 Eq.8 0.78 increases kinematic bending by 20% or so. It can be observed that d has a negligible effect on kinematic bending everywhere.[3]) [ ]) 3 0.2. G2 the shear moduli of the upper and lower layers. .5 2 10 15 z/d 20 25 interface 30 35 Fig. 6 M (δ) / M (1.6 2 δ 2. respectively. Fig. 4. t being the tube thickness. except for the vicinity of the layer interface. Effect of Winkler spring factor d on kinematic bending moments at a soil layer interface for pile–soil configuration S1-6 subjected to the suite of records in Table 1.2) 1. Effect of Winkler spring factor d on kinematic bending moments computed along the pile for the system of Fig. [3] 3. by Dobry et al. 1. [60] 1 Mylonakis. This simple proposal was later adopted by several investigators. subjected to A-TMZ270 input motion (case S1-6 in Table 2).2 Fig. 3 presents results referring to accelerogram A-TMZ270 recorded during the 1976 Friuli earthquake. Following the early work of Novak [58]. which was selected as a reference case. different formulations have been proposed over the years for evaluating the d factor. the increase being of the aforementioned order for d approaching 2. kinematic bending moments computed for different d’s are normalised with the results obtained for d ¼ 1. [2] 0. Young’s modulus should be taken as Ep[1 À (1 À 2t/d)4].8 1. Improvements over the original formula have been presented. A wider set of results referring to the same pile–soil configuration is presented in Fig. [59].

5 1. Elastic bedrock conditions have been assumed at the bottom of the lower layer.05–0.06–0.0 2.4.7 4 2 3 4 2 2.2) Eq.42 0. 3.3 2.10–1.05 0.3 100 S3 15 6 150 470 209 S4 5 25 100 S5 10 20 100 470 S6 19 11 100 . 10% and 20%. All cases considered are listed in Table 2.35 g.and fixed-head piles in a two-layer subsoil subjected to vertically propagating seismic waves.3 1.05 0.6 1.1 1. 6 δ from Mylonakis.86 0. This is evident in Fig.02 0.5 f1/fp.58 0.9 2.7 4 2 3 4 2 3 4 2 3 4 ID parameter case S1-1 S1-2 S1-3 S1-4 S1-5 S1-6 S1-7 S1-8 S1-9 S2-10 S2-11 S2-12 S2-13 S2-14 S2-15 S3-16 S3-17 S3-18 S3-19 S3-20 S3-21 S4-22 S4-23 S4-24 S5-25 S5-26 S5-27 S6-28 S6-29 S6-30 Vs.28 0.8 1.50 0.09–1.04–0. [3] 1.08–1.22 0. where results are presented for the geometry studied in the previous figures.47 0. In Table 2 the ratio f1/fp between the fundamental frequency of the subsoil f1 and the predominant frequency of input motion fp is provided. which cover several cases of practical interest. Effect of Winkler stiffness factor d and Ep/Es1 ratio on interface kinematic bending moments.06–1. The accelerograms were chosen in such a way that their original peak ground acceleration is as close as possible to the reference maximum peak acceleration on soil type A of a seismic zone according to the Italian seismic zonation of 2003 [53].2 2.09–1.3 1.6 2. subjected to A-STU000 and A-TMZ270 records.06–1.7 4 2 3 4 2 2.84 50 1880 S1 15 15 100 470 150 209 100 S2 15 30 150 470 209 2% 10% 20% 1.47 0. Sica et al.05–0.09–1. Table 2 Problem parameters considered.3 1. 66 75 80 133 150 160 200 218 240 133 150 160 200 218 240 160 169 174 235 245 255 171 225 267 150 180 200 122 132 138 30 (m/s) Soil type (EC8) D D D D D D C C C D D D C C C D D D C C C D C C D C/D C D D D Ep/Es1 D f1 (Hz) 0.13–2.05–0. range 0.30 in the upper 30 m of the subsoil.2 1.03 0.05–0.5 1.42 0.80 0.2 M (δ) /M (1.86 0.07–1.7 2. for pile–soil configuration S1-6 subjected to A-TMZ270 and A-STU000 input motions.5 3.5 1. In all analyses the following parameters 1.4 1.10–1.08 0. 6 δ from Kavvadas & Gazetas.S.10–1.97 0.67 0. The soil profile consists of a soft surface soil layer of thickness h1 and shear wave velocity Vs1.99 0.07–1.6 2. L/d ¼ 33. In all cases n1 ¼ n2 ¼ 0.08–1.2 1.99 0.49 0.06–1.06–0.0 2. followed by a stiffer stratum of thickness h2 and shear wave velocity Vs2 (Fig. shear wave velocities of the upper and lower layers were selected in such a way that subsoil profiles correspond to class C or D of the EC8 classification on the basis of the equivalent shear wave velocity Vs.8 0 100 200 300 400 500 600 Ep/Es1 700 800 900 1000 1100 1200 Fig. Scheme h1 (m) h2 (m) Vs1 (m/s) Vs2/Vs1 2 3 4 2 3 4 2 2.8 0.09–1. [2] 1 δ from Roesset.08–1.45 0.03 0. linearly de-convoluted to bedrock level and then propagated upward in the soil to provide the excitation motion of the embedded pile.7 0. For each geometry. Three different values of soil damping were employed: 2%. Vrock ¼ 1000 m/s. Eighteen runs were carried out for each parametric case based on the input motions of Table 1.4 2. the selected accelerograms have been scaled in amplitude to peak acceleration of 0.53 0. [60] A-TMZ270 A-STU000 0.28 0. 5. Parametric investigation An extensive parameter study has been performed by the aforementioned BDWF approach for both free.6 1.05–0. 5.06–1.00 0.54 0.3 1.03–0.10–1.3 1. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 895 of d becomes more significant.03–0.4 Eq.06–0.74 0. 1).6 2. For comparison purposes.57 0.03–0.

soil mass density (rs ¼ 1.60 m). For each reinforcement configuration. Sica et al.5 %o 15 εp = 1.4). M–N interaction diagrams were computed assuming concrete class C20/25 with Rck ¼ 25 N/mm2 and fck ¼ 20 N/ mm2 and steel rebars with fyk ¼ 375 N/mm2 and ftk ¼ 450 N/mm2. From top to bottom: parameter cases S1-4. The shear wave velocity of the elastic bedrock.9 % A) 12 Φ 30 (3.9 Mg/m3). In all figures.5 Mg/m3. was taken equal to 1000 m/s. corresponding to the Italian steel class FeB38K. Grey bands define yield resistance of pile cross sections for different amounts of longitudinal reinforcement. soil Poisson’s ratio (n1 ¼ n2 ¼ 0. were kept constant: pile length (L ¼ 20 m). three zones shown in grey colour are indicated. where envelopes of maximum kinematic bending moments along the pile are plotted for each of the 18 signals. S1-5 and S1-6 (subsoil D) of Table 2. 6.896 S.32]. pile diameter (d ¼ 0.6 % A) z (m) εp = 0.1. corresponding to the range of yield moments for typical reinforcements of the concrete pile cross section (8F16. Effect of geometry and soil impedance contrast Typical results from the parameter study are shown in Figs. 6–8. Effect of shear wave velocity contrast on kinematic bending moments for pile–soil configuration S1. .0 % A) 200 400 600 800 1000 10 8 Φ 16 (0. pile Young’s modulus Ep ¼ 2. Vrock. the lower limit of the grey zone represents the cross sectional yielding moment corresponding to zero normal load while the higher one to a typical normal load N ¼ 1200 kN [normalised axial load nd ¼ N/(fckA) between 0 and 0. 3.5 Â 107 kPa and pile density rp ¼ 2. 24F12 and 12F30) and variable normal loads at the pile head. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 M (kN*m) 0 0 VS1/ VS2 = 1/2 5 12 Φ 24 (1.1 %o VS1/VS2 = 1/3 5 A-AAL018 A-STU270 A-STU000 ATMZ000 B-BCT000 C-NCB000 E-AAL108 E-NCB090 J-BCT090 R-NC2090 TRT000 z (m) ATMZ270 10 B-BCT090 C-NCB090 E-NCB000 15 J-BCT000 R-NC2000 R-NCB090 20 0 VS1/VS2 = 1/4 5 z (m) 10 15 20 Fig.8 %o 20 0 εp = 0.

Effect of shear wave velocity contrast on kinematic bending moments for pile–soil configuration S1. 6) and C (Fig. Conversely. 7). For subsoil profiles corresponding to class D (Fig.5 %o 15 εp = 1. Grey bands define yield resistance of pile cross sections for different amounts of longitudinal reinforcement.9 % A) 12 Φ 30 (3.6 % A) z (m) εp = 0. 8). In the latter case. This occurs for both subsoil types D (Fig. S1-8 and S1-9 (subsoil C) of Table 2.1 %o VS1/VS2 = 1/3 5 A-AAL018 A-STU270 A-STU000 ATMZ000 B-BCT000 C-NCB000 E-AAL108 E-NCB090 J-BCT090 R-NC2090 TRT000 z (m) ATMZ270 10 B-BCT090 C-NCB090 E-NCB000 15 J-BCT000 R-NC2000 R-NCB090 20 0 VS1/VS2 = 1/4 5 z (m) 10 15 20 Fig. 7. Maximum bending moment develops close to the interface when the active pile length la is smaller than the thickness h1 of the first layer.5 [2–4. in agreement with earlier research. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 897 M (kN*m) 0 0 VS1/ VS2 = 1/2 5 12 Φ 24 (1. however.S.64].0 % A) 200 400 600 800 1000 10 8 Φ 16 (0.29. Sica et al. the kinematic bending moment at the pile head may exceed that of the interface (Fig. when the active pile length ‘a is equal or smaller than h1.75 and 2.8 %o 20 0 εp = 0.44. peak moments at interface and pile head are much smaller than those computed for subsoil type D. 6) the computed kinematic bending moments may be well above   . This length can be determined from the following empirical formula as  0:25 Ep ð7Þ ‘a ¼ wd Es1 where w is a dimensionless constant varying between 1.62. The following observations may be drawn from these figures:  Interface bending moments increase dramatically when the shear wave velocity contrast between the upper and lower layers Vs1/Vs2 decreases from 1/2 to 1/4. From top to bottom: parameter cases S1-7.

and average frequency. 9 results pertaining to case S1-6 (Table 2) are shown. In Fig.898 S.0 % A) 200 400 600 800 1000 10 8 Φ 16 (0. finput has been determined on the basis of both response spectrum predominant frequency. For subsoil profiles corresponding to soil class C and shear wave velocity contrasts Vs2/Vs1 4 2. especially for low levels of pile reinforcement (Fig. finput. kinematic bending moments may also be considerable. Effect of shear wave velocity contrast on kinematic bending moments for pile–soil configuration S4.8 %o 20 0 εp = 0. 7). 3.9 % A) 12 Φ 30 (3. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 M (kN*m) 0 0 VS1/ VS2 = 1/2 5 12 Φ 24 (1.1 %o VS1/VS2 = 1/3 5 A-AAL018 A-STU270 A-STU000 ATMZ000 B-BCT000 C-NCB000 E-AAL108 E-NCB090 J-BCT090 R-NC2090 TRT000 z (m) ATMZ270 10 B-BCT090 C-NCB090 E-NCB000 15 J-BCT000 R-NC2000 R-NCB090 20 0 VS1/VS2 = 1/4 5 z (m) 10 15 20 Fig. and the predominant frequency of input motion. As the analyses have been performed assuming linear elastic behaviour for the materials.2. Effect of frequency content of input motion For a better interpretation of the analytical results.  the yielding moments for typical concrete reinforcements and normal loads acting along the pile. Sica et al. It is evident that the . the computed moments can be assumed valid until the yield limit is reached. 8. has been established for all cases. Grey bands define yield resistance of pile cross sections for different amounts of longitudinal reinforcement. as described in [54] (Table 1). the ratio f1/finput between the fundamental natural frequency of the subsoil. f1. fm. fp. From top to bottom: parameter cases S4-22 (subsoil D). S4-23 and S4-24 (subsoil C) in Table 2.5 %o 15 εp = 1.6 % A) z (m) εp = 0.

Relation between transient and steady-state interface kinematic bending moments for configuration S1-6 and the input motions of Table 1 (finput ¼ fm). for case S1-6 (Table 2) and the signals of Table 1. Naturally. 10% and 20%).S. in addition to inertial interaction. the following regression formula was derived: M ð DÞ 5 ¼ DÀ1=4 M ð10%Þ 9 ð8Þ Fig.5 finput = fp f1 / finput 1 possible resonance 0. the potential for kinematic bending will be high and kinematic interaction should be considered. the potential of developing significant kinematic pile bending can be established on the basis of the following criterion: . . These results suggest that a threshold value of ratios f1/finput can be defined beyond which kinematic effects are important (Fig. 10. Results refer to case S1-6 and the input motions of Table 1. to the predominant frequency of input motion.if ratio f1/finput is lower than the threshold value.5. 9).3. Ratio of natural frequency of soil profile. Note the critical frequency fc beyond which maximum moment is smaller than static. 9. finput. kinematic bending will typically be minor and only inertial interaction can be accounted for in design.5 0 0 5 10 D (%) 15 20 Fig. which is reminiscent of corresponding expressions in seismic regulations for structural problems [19.if ratio f1/finput is higher than the threshold value. 8 1 0.5 no resonance 0 A-AAL018 E-AAL108 ATMZ000 ATMZ270 C-NCB090 E-NCB090 B-BCTT000 R-NCB090 A-STU000 A-STU270 B-BCT090 J-BCT000 J-BCT090 R-NC2000 C-NCB000 E-NCB000 R-NC2090 TRT0000 Input motion Fig.5 Eq. 2 M (D) / M (10%) 1. a trend obviously related to the development of resonant phenomena in the soil. Conversely. It is observed that soil damping can significantly affect the magnitude of kinematic bending moments. as it controls free-field response and it should be carefully assigned when linear elastic analyses are carried out. Sica et al. Based on these results.4] by frequency–time domain analyses: maximum effects of kinematic bending in piles occur when the frequency of ground motion is near the fundamental period of the subsoil (f1/finput E 1). This finding may provide a practical way for selecting records from a database of seismic events for a particular region: if acceleration time-histories (in addition to peak ground accelerations) are available as part of the design process. The reported results pertain to parameter case S1-6 (Table 2).20] and can be used to quantify the effect of damping in preliminary design calculations. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 899 2 finput = fm 1. accelerograms that induce the highest kinematic moments are characterised by values of f1/fp or f1/fm close to unity. This is in accordance with the findings in [2. 1400 1200 M (kNm) 1000 800 600 400 200 0 0 1 f1 2 fc 3 4 finput (Hz) 5 6 7 Mstatic Mresonance 3. which provides a comparison among the kinematic bending moments computed at layer interface for different levels of material damping in the soil (2%. f1. all the accelerograms that induce low kinematic moments in the pile have f1/finput below 0. kinematic response tends to drop with increasing . 11. 10. Effect of soil damping The effect of soil damping on the response is examined in Fig. Effect of soil material damping on kinematic pile bending moments.

5 finput f1 Φ1 = 1. s being the standard deviation.5 o finput/f1 o 1. g1.0 0. Mstatic. 12.68 finput f1 −1.5 1. 4.2 mean + 1σ mean mean-1σ η = 0. The corresponding . This occurs because the input accelerograms are characterised by much higher dominant frequencies (4–8 Hz—Table 1).5. With reference to the mean value. Naturally. bringing the resonance bending moment.5/10 ¼ 0.5 mean + 1σ mean mean-1σ Fig.900 S.6 η 0. Simple formulas for transient pile bending at a layer interface Mylonakis [3] developed a simple formulation for predicting kinematic bending moments at a layer interface under lowfrequency excitation (o-0). an effect which is not captured by Eqs.5 in conventional design spectra versus a peak resonant amplification of 1/(2D) ¼ 10 for 5% damping. In these equations.] On the other hand. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 η= maximum possible moment Mresonance Φ1 = M (t)max Mresonance M (t)max Mstatic 1. Interestingly. ep. [4.8 0. Eqs.0 Φ1 1.71] and Mylonakis [3]. is equivalent to an Z factor of 2. Factors Z (top graph) and F1 (bottom graph) versus frequency ratio (finput/f1) for different pile–soil configurations and input motions.6 Hz). Definition of response factors Z and F1 for a given pile–soil configuration. parameter F1 may be interpreted either as an amplification or as a de-amplification factor to bring the bending moment computed for static conditions.0 0. Frequency versus time domain bending 0. parameter Z can be interpreted as a ‘‘de-amplification’’ factor. for most signals time-domain moments are lower than the static value Mstatic obtained for zerofrequency excitation. material damping.5. 13. M(t)max. the decrease being of the order of 20% for an increase from 5% to 10%. 3.4. (11) and (12) can be used to quantify the effect of transient nature of input motion on kinematic response of piles. In Fig. respectively.94 finput f1 −1. ep/g1. Sica et al. 13.25 for a simple oscillator. [As an example. (mean+ s) and lower (mean À s) estimates. dashed lines have been employed in Fig. M(t)max. than the fundamental frequency of the subsoil at hand (f1 ¼ 1. 13 for 0. Mresonance. 11 interface kinematic bending moments determined in the frequency domain are compared to corresponding moments computed in the time domain for each selected accelerogram. response is known to depend on number of excitation cycles [4]. spectral amplification of 2.0 resonance range 3.4 0. For this reason. An enlarged version of Fig. the mean values of a numerical regression analysis performed for Z and F1 are displayed with corresponding upper which are valid for long piles and finput/f1 ratios greater than approximately 1.0 In Fig. (9) and (10) are determined from a numerical BDWF analysis—not from approximate expressions such as those provided in the original publications. For finput/f1 E 1 the results exhibit significant dispersion. referring to parameter case S1-6 (Table 2). Only cases corresponding to long piles are shown. which is simply the ratio between peak pile bending strain. (11) and (12). to match the peak timedomain moment.3 2. factors Z and F1 in Eqs.0 0 static resonance 1 2. The kinematic bending moment may be derived from a strain transmissibility parameter. In the ensuing and except if specifically otherwise indicated. Time-domain moments naturally lie below the value associated with resonance condition (Mresonance). both parameters tend to drop with increasing finput/f1 that is for conditions far from resonance.5 maximum transient moment for a given input motion M(t)max moment for ω = 0 Mstatic 0. In that frequency range. the following simple regression relations were derived:  À1:5 f Z ¼ 0:68 input ð11Þ f1 Z¼ M ðt Þmax Mresonance Mðt Þmax Mstatic F1 ¼ 1:94 ð9Þ  À1:3 finput f1 ð12Þ F1 ¼ ð10Þ which were first employed in Nikolaou et al. 11 is schematically represented in Fig. and free-field soil shear strain at the interface. to match the peak time-domain moment. 12 to define the following factors correlating frequencyand time-domain results: 0 1 2 3 4 5 6 7 8 9 10 11 12 finput / f1 Fig. de-amplification is observed and kinematic effects consequently diminish. For this frequency range.

30 for conditions far from resonance and 1.20 for layer stiffness contrast Vs1/Vs2 in the range 1/4–1/3 (Fig. The reasons for this discrepancy are worth investigating. Both these estimates are lower than those reported by Maiorano et al. By linear regression analysis.71] introduced the following empirical equation to compute the pile kinematic bending moment at the interface in resonant steady-state conditions:  0:30  0:65  0:50 Ep L Vs2 Mresonance ¼ 0:042tc d3 ð16Þ d E1 Vs1 where tc is a characteristic shear stress at the interface: Fig. It is noteworthy that F2 is practically independent of (finput/f1). It was found that F2 is equal to approximately 1.5 2. (13). it does not account for the transient.5 0.25. the kinematic bending moment at the layer interface is not expected to be important.5 1. F in Eq. Only cases corresponding to long pile are shown. Recall that parameter b is not equivalent to parameter Z in Eq. Factor F2 versus frequency ratio (finput/f1) for pile–soil configurations S1 and S5. (14) refers exclusively to SSI effects. Indeed. Maiorano et al. Accordingly. The authors introduced a parameter Z (Eq. As Eq. (g1)dyn.0 Vs1/Vs2 = 1/4 ÷ 1/3 Φ2 1. To address this limitation.S. (10) relates static to transient response accounting for all relevant dynamic phenomena.0 0. k1 the Winkler spring modulus associated with the upper layer. as computed from a 1-D wave propagation analysis [67]. F2 relates the exact transient moment with an approximate moment that is neither purely static nor purely dynamic.053. A theoretical weakness of the above formula is that the predicted moment tends to increase without bound for very slender piles.5 1.0 2. (9). To investigate frequency-to-time domain response relations. b can be interpreted as an ‘‘average dynamic coefficient’’ that links the maximum transient bending moment to the transient peak soil shear strain at the interface. Maiorano et al. Using nonlinear regression analysis based on the numerical data obtained in this study. It should be noticed that. Results are shown in Fig. 15. transient kinematic bending moment with the value obtained for resonant conditions. (16). Maiorano et al. to correlate the In this approach. [35].0 0.0 2. 14). This value is close to the generic recommendation by Mylonakis [3]. however. [35] was employed to estimate the parameters F2 and b by adopting the M(t)max provided by the BDWF approach of Mylonakis et al. (ii) the approximate solution for (ep/g1) given by Eq.15 whereas b is around 0. for the denominator in Eq. For relatively soft piles and low-frequency input (oinput d/Vs1 o 0. Contrary to factor F1 in Eq. thereby. (10) F2 is associated with: (i) the peak dynamic strain (g1)dyn. (9)). referring partially to dynamic and SSI effects. large pile–soil stiffness contrasts and large velocity contrasts between the two soil layers.5 2.1). [35] found that parameter F2 is approximately 1. whereas Eq. [29]. (15).0 0 1 2 3 4 5 6 7 8 9 10 11 12 finput / f1 F2 ¼ 2Ep I ep ðg Þ d g1 o ¼ 0 1 dyn M ðt Þ   max ð15Þ for which symbol F2 is employed in this article to avoid confusion with the definitions provided earlier. . (14) is not identical to F1 in Eq. In the same spirit as before.71] equation in the form: b¼ M ðt Þmax G1 b ðg1 Þdyn ð18Þ where parameter b is given by b ¼ d3  0:3  0:65  0:5 Ep L Vs2 d Es1 Vs1 ð19Þ tc ¼ as r1 H1 ð17Þ in which as is the free field acceleration at soil surface. Eq. [4. 14. 14). yet lie beyond the scope of this paper. the parameter F2 was determined to be approximately 1.07 on average. For this scope they re-arranged Mylonakis [3] solution in the form: Φ2 1. [4.5 0. encompass the frequency dependence of the free-field response. reference [3] recommends that parameter F lies in the range 1–1. dynamic nature of the phenomenon. in Eq. Higher values of 1. In this case. This is understood given that both M(t)max and (g1)dyn. Maiorano et al. Vs1/Vs2 ¼ 1/4–1/3 (top graph) and Vs1/Vs2 ¼ 1/2 (bottom graph). As an alternative to the mechanistic model by Mylonakis [3]. cancels out from the ratio. from Eq. although conceptually related. (13) has been obtained for static conditions. (h1/d) the embedment ratio.40 near resonance. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 901 analytical solution is )  À1 ("  1=4   #   ep 1 h1 k1 h1 À1 cðcÀ1ÞÀ1 ¼ 4 ðc2 Àc þ 1Þ 3 g1 o ¼ 0 2c d Ep d ð13Þ where c ¼ (G2/G1) is the layer stiffness contrast.0 Vs1/Vs2 = 1/2 g1 dyn g1 o¼0 which accounts for the effect of frequency on kinematic pile bending.0 3. [35] adopted the kinematic bending moment M(t)max computed in time domain by the FEM code VERSATP3D [65] and the shear strain at the interface (g1)dyn provided by a one-dimensional EERA analysis [66]. [35] re-arranged the Nikolaou et al. [35] report b to be about 0.3 on average were obtained for lower stiffness contrast Vs1/Vs2 ¼ 1/2 (Fig. (10). which. Mylonakis [3] introduced a correction function F of the form:     e ep F¼ p ð14Þ ¼ 3. (18) is not an exact peak steady-state moment. Nikolaou et al. a trend that has not been observed. Sica et al. In the present work the procedure followed by Maiorano et al.

005 Mmax / b = 0.6 0.015 (γ1) dyn Fig. Synthesis: new proposed approaches As previously illustrated. (g1)o ¼ 0 is the static shear strain at the interface. no numerical site response analysis is necessary and the approach can be fully analytical. This approach is also attractive as dynamic effects due to frequency content of the input motion are incorporated into parameter (g1)dyn.5 Mmax / a = 1. Numerical examples are provided in Appendix I.902 S. Three such approaches are suggested in this paper. The analyses were performed using a numerical tool developed by Mylonakis et al. (13) and dynamic modifier F2 is equal to about 1. whereas F2 is a constant. 15.15 (γ1)dyn ±5. does not require dynamic analysis. which is often not known with sufficient accuracy in practice. 13. 5. The main conclusions of the study are: 1.0 0. It was found that: a) Severe kinematic interaction develops at the interface of two soil layers having sharply different stiffness and pertaining to subsoil of class C and D according to EC8 classification.0 0 0.71] can be rewritten in the form: M ðt Þmax ¼ Z Mresonance ð23Þ 1. (6) and (8). Adjustments due to damping and spring stiffness can be made with the help of Eqs. Graphical representation of peak kinematic bending moments using two different normalisation schemes (modified from [35]). geometric factors and earthquake excitations. A sensitivity analysis of pile bending moments as function of spring stiffness k adopted in Winkler models and damping ratios D employed in free field response analyses was carried out. [69] suggested a simple method to obtain the peak steady-state acceleration at ground surface as. To this end. . [4. A disadvantage is that parameter F1 is sensitive to variations of (finput/f1) ratio. different simple approaches may be followed for determining the transient kinematic pile bending at the interface of two soil layers. accordingly. b) A threshold value of ratio (f1/finput) can be determined beyond which kinematic effects are considerable. Note that Eq. was reported. Eqs (6) and (8) were developed to account for these effects in a simple manner. the other two are dynamic and. knowledge of surface acceleration as at resonance is required.0 (Mmax/a) x 102 mean+1σ mean 2) Dynamic approach based on F2 The procedure by Mylonakis [3] may be written in the alternative form as   2Ep I ep ðg Þ F ð22Þ M ðt Þmax ¼ d g1 o ¼ 0 1 dyn 2 where the static shear strain (g1)o ¼ 0 has been replaced by its dynamic counterpart (g1)dyn.4 0. Z is sensitive to variations of frequency ratio (finput/f1). [29] and Mylonakis [62] based on a properly calibrated Beam-on-Dynamic-Winkler-Foundation (BDWF) model. This approach is attractive for engineering applications as it does not require a free-field site response analysis. 3) Dynamic approach based on Z The original formulation by Nikolaou et al.053 (γ1)dyn ± 2.2 0.5 2. which can be easily established by the following equation: ðg1 Þo ¼ 0 ¼ a h1 2 Vs 1 ð21Þ 6.20–1. Sica et al. In this way. 2.8 × 10-5 0. A disadvantage is that a free-field site response analysis is required to establish the value of (g1)dyn.01 0. (ep/g1)o ¼ 0 is given by Eq.0 (Mmax/bG1) x 103 0. As in the case with parameter F2.5 mean-1σ 1. (ii) depth of upper layer. 1) Static approach based on F1 This is the simplest procedure and can be implemented without carrying out a dynamic analysis. are more demanding from a computational viewpoint.2 1. (13) and F1 is given by Eq.025 where Z can be determined from Eq. Mresonance may be obtained from Eqs. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 2. 13. In these cases the computed kinematic bending moments at the interface may be well above the yielding moments of the pile cross section computed for typical reinforced concrete piles and axial loads. 0. Conclusions Results from an extensive parametric analysis were reported. carried out on single vertical elastic solid piles in layered soil. based on the peak acceleration ar on soil type A and the fundamental period T1 of the subsoil. (12) and Fig. Cairo et al. This limit provides a possible criterion for selecting significant where a ¼ as ¼ ar is a uniform pseudostatic seismic acceleration in the profile. (iii) waveform of input motion and (iv) soil damping adopted in the free-field site response analysis. One is purely static and. the procedure by Mylonakis [3] is reformulated as   2Ep I ep ðg Þ F ð20Þ M ðt Þmax ¼ d g1 o ¼ 0 1 o ¼ 0 1 where (ep/g1)o ¼ 0 is provided by Eq. To this end.02 0.8 0. (16) and (17). The latter can be determined by a harmonic site response analysis or by a fully analytical approach assuming harmonic wave propagation in a two-layer subsoil.25 (Fig. A comprehensive parametric study on a two-layer profile as function of: (i) stiffness contrast between upper and lower layer. accounting for different material properties. (21) is equivalent to the familiar Seed and Idriss [68] procedure with the depth factor rd taken equal to 1. (11) and Fig. 14).0 1. thereby.1×10-4 0.

842 (Eq.5 g. respectively. In the ensuing the following parameters are considered: L ¼ 20 m. 3 rp ¼ 2.622. 1 is embedded in soil profile S1-4. (12). new regression analyses were carried out for computing the transient pile bending moments at the soil layer interface. The three simplified approaches proposed in the paper will be hereafter applied to the above case. (11) yields Z ¼ 0. Acknowledgements The work herein described is part of the ReLUIS research Project ‘‘Innovative methods for the design of geotechnical systems’’. D ¼ 10%. For the case in hand. or by a fully analytical approach [69]. parameter F1 is estimated at 0. (12). Specifically: a) The approach based on F1 is perhaps the most attractive for engineering purposes. as it does not require a free-field site response analysis. From Eq. The authors wish to thank ReLUIS and AGI research coordinators.0029.053 obtained in this study: M ðtÞmax ¼ bbG1 ðg1 Þdyn ¼ 0:053 Â 67:5 Â 19000 Â 0:0029 ¼ 197kN m ðA:8Þ Both these results are close to the exact value of 194 kN m. As in the case of parameter F1. is lower than the static estimate in Eq. knowledge of surface acceleration as at resonance is required. Therefore. rs ¼ 1. the authors suggest to apply the simplified formulas choosing between fp and fm (or any other rational estimate) the value closer to the fundamental frequency f1 of the soil profile. if the acceleration time-histories are provided for a given seismic zone (in addition to conventional peak ground acceleration). (23): Mðt Þmax ¼ ZMresonance ¼ 0:18 Â 1187 ¼ 216 kN m ðA:7Þ Following the Maiorano et al. Sica et al.2. A disadvantage is that parameter F1 is sensitive to variations of (finput/f1) ratio. c) The approach based on Z requires knowledge of Mresonance.9 Mg/m3.20. (21) the static shear strain is likewise obtained: ðg1 Þo ¼ 0 ¼ 0:35 Â 9:81 Â 15 1002 ¼ 0:00515 ðA:2Þ To account for the transient nature of the phenomenon. (16). Eq. At the layer interface a BDWF analysis according to reference [29] provides maximum kinematic bending moments in the time domain equal to 194 and 248 kN m for the A-TMZ000 and A-AAL018 records.9.4.60 m. [35] approach. The shear wave velocity of the elastic bedrock. as suggested in this paper. was taken at 1000 m/s. (17). Z is sensitive to variations of frequency ratio (finput/f1). Dynamic approach based on F2 In this approach a free-field site response analysis should first be carried out to obtain (g1)dyn. Finally. as is found to be equal to 1. (20): Mðt Þmax ¼ 2 Â ð2:5 Â 107 Þ Â 0:00636 Â 0:115 Â 0:00515 Â 0:622 ¼ 195 kNm 0:6 ðA:3Þ which is almost identical to the ‘‘exact’’ value of 194 kN m. b) The approach based on F2 is also attractive as dynamic effects due to frequency content of the input motion are incorporated into parameter (g1)dyn. Vrock. subjected to Italian records A-TMZ000 and A-AAL018 (Tables 1 and 2). Numerical examples The vertical pile shown in Fig. the transient kinematic bending moment is calculated from Eq. whereas F2 is a constant. For the sake of conservatism. For the case in hand. (16) the estimate of the resonant moment is Mresonance ffi 0:042 Â 419 !0:65    20 0:30 2:5 Â 107 400 0:50 ¼ 1187kN m 0:6 100 53 200 ðA:5Þ Appendix I. which can be realized by means of Eq. (12)). the potential of the site with the associated waveforms to induce significant kinematic bending in piles may be established. frequency parameter F1 needs to be determined by means of Eq.S. the ratio (finput/f1) is equal to 2. Record A-TMZ000 Â 0:6 3  ðA:6Þ For (finput/f1) ¼ 2. (6). (A.5 Mg/m . which is often not known with sufficient accuracy in practice. d ¼ 0. if required. Large differences often appear in the parameters representing the frequency content of input motion (fp or fm. Assigning F2 the value 1. applying Eq.(13) as   ep ¼ 0:115 ðA:1Þ Static approach based on F1 Repeating the above analysis yields the same strain transmissibility (ep/g1)o ¼ 0 and static shear strain (g1)o ¼ 0 as before.35 g. Applying Eq. Table 1). n1 ¼ n2 ¼ 0.2). Finally. Ep ¼ 2.5 Â 107 kPa. promoted and funded by DPC (Civil Protection Department) of the Italian Government and coordinated by the AGI (Italian Geotechnical Association).4. Dynamic approach based on Z To apply this approach. On the basis of the above results. From Eq. remarkably. we have tc ¼ as r1 H1 ¼ 1:5 Â 9:81 Â 1:9 Â 15 ¼ 419 kPa From Eq. which. scaled at a peak ground acceleration of 0. This may be obtained from a free-field site response analysis. For a ratio (finput/f1) ¼ 1. Three alternative procedures were outlined to solve the problem in the realm of routine engineering calculations. 4.4. Eq. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 903 records from a ground motion database. with the parameter b ¼ 0. g1 o¼0 . 3. F1 ¼ 0. (22) provides M ðt Þmax ¼ 2 Â ð2:5 Â 107 Þ Â 0:00636 Â 0:115 Â 0:0029 Â 1:20 ¼ 220 kN m 0:6 ðA:4Þ which is about 10% higher than the exact value. Record A-AAL018 Static approach based on F1 The strain transmissibility is computed from Eq.18. A potential disadvantage is that a free-field site response analysis is required to establish the value of (g1)dyn. Additional adjustments to the above estimates can be made with reference to d and D values (Eqs. These values were computed assuming a spring stiffness d equal to 1. For the problem at hand an EERA analysis provided the value (g1)dyn ¼ 0. (8)).

51(5): 425–40. Mylonakis G. Journal of Geotechnical and Geoenvironmental Engineering 2005. [16] Chau KT. (23): M ðt Þmax ¼ ZMresonance ¼ 0:26 Â 1187 ¼ 308kN m ðA:11Þ Following the Maiorano et al. ASCE. 2008. cd-rom. vol. . Shaking table tests of seismic pile–soil–pier–structure interaction. [10] Tokimatsu K. Madrid.30:119–32.. [11] Horikoshi K. In: Dynamic response of pile foundations experiment. Dente G. Gazetas G. 11. Wakahara T. Leoni G. retaining structures and geotechnical aspects. 2008. Spain. Journal of Geotechnical and Geoenvironmental Engineering 2009. Italy: Pa p. [39] Sica S. Soils and Foundations 2001. 2433–56. In: Proceedings of the international conference on performance-based design in earthquake geotechnical engineering. Melville. p.122(1):46–53. paper no. in Italian]. vol. 649–56. [44] Randolph MF. Sica et al. [41] Simonelli AL.LL. [29] Mylonakis G. Soil Dynamics and Earthquake Engineering 2009. 1987. In: Proceedings of the 14th European conference on soil mechanics and geotechnical engineering. In: Dynamic response of pile foundations: experiment. [22] Margason E. [7] Gazetas G.053 obtained in this study yields the estimate: M ðt Þmax ¼ 0:053 Â 67:5 Â 19 000 Â 0:0036 ¼ 244 kN m ðA:12Þ References [1] Flores-Berrones R.26(1): 49–160. [15] Wei X. [4] Nikolaou S. with the parameter b ¼ 0. In Proceedings of the 12th world congress on earthquake engineering 2000. p. paper 6. [19] NEHRP. O’Rourke MJ. Universita’ degli Studi di Napoli Federico II. 2009. Ohtsu H. Coupled soil–pile–structure interaction analysis under seismic excitation. Piles under dynamic loads: state of the art.20(4):19–36. Simplified model for seismic pile bending at soil layer interfaces. Earthquake Engineering and Structural Dynamics 1997. Shen CY. In: Proceedings of the fourth international conference on recent advances in geotechnical earthquake engineering and soil dynamics. Bulletin of the New Zealand National Society for Earthquake Engineering 1993.38(11):1281–305. Pile bending during earthquakes. Mandolini A. In: Proceedings of the fourth international conference on earthquake geotechnical engineering. CEN European Committee for Standardization. p. Oda K.18. Aversa S. 1977. Geotechnical Earthquake Engineering and Soil Dynamics III. Recommended provisions for seismic regulations for new buildings and other structures (FEMA 450). Geotecniche in Condizioni Sismiche. 1995 South-Hyogo earthquake by using the seismic deformation method. paper no. Aspects of EC8. Soils and Foundations 1980. Non linear seismic soil–pile–structure interactions: shaking table tests and FEM analyses. IS-Tokyo. Mylonakis G. IS-Tokyo. M. Reggio Calabria. p. vol. [38] Moccia F. [32] Cairo R. paper 9. Suzuki H. 53–78.34(1):44–52. [8] Mizuno H. Simplified procedure for evaluating earthquake loading on piles. 1998.PP.108(4):554–69. 2007. In: Dakoulas P. A method for assessing kinematic bending moments at the pile head. Design of structures for earthquake resistance: foundations.31:247–59. Seismic response of end-bearing piles.9.LL. [3] Mylonakis G. Maugeri M. Earthquake Engineering and Structural Dynamics 2010. Geo-Institute ASCE Conference. In: Proceedings of the fourth international conference on recent advances in geotechnical earthquake engineering and soil dynamics. [35] Maiorano RMS. Gazetas G. as before. Building damage associated to geotechnical problems. [14] Meymand P. 2007. Mylonakis G. code provisions. 1991. Fondazioni profonde sotto azioni sismiche. Sica S. Seattle. Taormina.26:337–59. [2] Kavvadas M. 2001. PhD dissertation. 2005. Geotechnical Special Publication. Norme Tecniche per le Costruzioni (NTC2008). Detailed investigation of piles damaged by Hyogoken Nambu earthquake. [34] Castelli F. Whitman RV. The response of flexible piles to lateral loading. [24] Dobry R. Washington. Mandolini A. Guo X. 1989.109:778–81. Tateishi A. Part I: soft soil. 1984. Some remarks on the kinematic vs. Sica S. Shimizu K.131(10):1243–51. Fan L. [28] Kaynia AM. [18] Moccia F. p. Napoli. ISSMGE-ERTC 12 Workshop Geotechnical  tron Editore. 2477 (in CD-ROM). Gazetas G. Berkeley. Conte E. Simonelli AL. [26] Conte E. Mylonakis G. and Whitman R. Dynamic Elastic Analysis of pile foundations using finite element method in the time domain. 2008. editors. Simonelli AL. Meerut. Mizuno H. [45] Wu G. Kinematic bending moments in pile foundations.135(10):1440–51. Journal of Geotechnical Engineering Division. San Francisco. 1. Simplified approach for the seismic response of a pile foundation. Pile damage during earthquake in Japan (1923–1983). vol. Special Issue Soils Foundation 1996:189–200. In: Proceedings of the second international conference on recent advances in geotechnical earthquake engineering and soil dynamics. 2009. Forces in pile foundations under seismic loading. Special Issue Soils Foundation 1996:189–200. Iiba M. Shaking table scale model test of non linear soil–pile–superstructure interaction in soft clay. 1990. p. Dente G.18. Finally. due to the hypothesis of linearly viscoelastic soil behaviour. Wu X. 1119–74. F2 to be equal to 1. [27] Mineiro AJC. New York: American Institute of Physics. vol. Tazoh T. IS-Tokyo. inertial interaction for piled foundations. PhD dissertation. p. Geotechnical Special Publication. 3. (22) yields M ðt Þmax ¼ 2 Â ð2:5 Â 107 Þ Â 0:00636 Â 0:115 Â 0:0036 Â 1:20 ¼ 263 kN m 0:6 ðA:10Þ Dynamic approach based on Z The surface acceleration as at resonance. Building Seismic Safety Council. Halloway DM. Kinematic pile bending in layered soils: linear vs. Bologna. Finn W. In: Proceedings of the international conference on performancebased design in earthquake geotechnical engineering. 2009. Kitagawa Y. II. Kinematic pile bending during earthquakes: analysis and field measurements.29:300–10. Banerjee PK. Bruxelles. Discussion on Seismic response of end-bearing piles by Flores-Berrones R. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 The transient kinematic bending moment is obtained from Eq. [21] NTC2008—M. [12] Luo X. [40] Sica S. [5] Novak M. Opere  tron Editore. Kinematic bending of piles: analysis vs. Holtz RD. tc and Mresonance are the same as before. Eq. Canadian Geotechnical Journal 2009. In: Proceedings of the sixth world conference on earthquake engineering. Geotechnique 1981. Fixed-head pile bending by kinematic interaction and criteria for its minimization at optimal pile radius. Eq. [37] Di Laora R. 29. paper 1674. 3. Carbonari S. [31] Saitoh M. [25] Tazoh T. 1690–6. Journal of Structural Engineering 1998. Journal of Geotechnical Engineering Division. p. Belgium. Kinematic response analysis of piled foundations under seismic excitation. Il comportamento sismico del palo di fondazione in terreni eterogenei. [35] approach. [42] Dezi F. 11. 137–45 [in Italian]. (11) yields Z ¼ 0. IS-Tokyo. Nonlinear seismic response of single piles. Assigning. 602–9. Seismic observations and analysis of grouped piles. editors. 1998.V. Seismic analysis of pile foundations damaged in the January 17. [33] Cairo R. [36] De Sanctis L. [9] Matsui T.20. [23] Kagawa T. Italy. Leoni G. 2001. 2009. Earthquake Engineering and Structural Dynamics 2009. Aversa S. Greece. Geotechnique 1993. vol. Seismic soil–structure interaction: new evidence and emerging issues. Moraci N. 1987. Soil–pile–bridge seismic interaction: kinematic and inertial effects.39:1133–54. Canadian Geotechnial Journal 1997. In: Proceedings of the international conference on performance-based design in earthquake geotechnical engineering.43(2):207–22. In: Proceedings of the international conference on performance-based design in earthquake geotechnical engineering. equivalent-linear analysis. Seismic pile foundation design analysis. Kakurai M. 309–34 [Chapter 11. Kinematic interaction: a parametric study on single piles. In: Proceedings of the eighth world conference on earthquake engineering. A model for the 3D kinematic interaction analysis of pile groups in layered soils. Proceedings of the seismic engineering conference on commemorating the 1908 Messina and Reggio Calabria Earthquake (MERCEA’08). [13] Mizuno H. Kinematic seismic response and bending of free-head piles in layered soil. [20] EN 1998-5. Maiorano RMS. Seismic soil pile interaction: experimental evidence. (20) as Mðt Þmax ¼ 2 Â ð2:5 Â 107 Þ Â 0:00636 Â 0:115 Â 0:00515 Â 0:842 ¼ 264 kN m 0:6 ðA:9Þ Dynamic approach based on F2 A free-field site response analysis using A-AAL018 record as input provides (g1)dyn ¼ 0.124:434–44. Bologna: Pa 13. Gazzetta Ufficiale della Repubblica Italiana. Journal of Engineering Mechanics. De Sanctis L. St. Carbonari S. Lateral load–deflection relationships of piles subjected to dynamic loads. Mahzooni S. 1–20. Murono Y. [17] Tokimatsu K. In: Santini A. vol. Lisbon: De Mello Volume.46(5):571–84. Simonelli AL.41(4):47–58. 2003. Yegian MK. analysis and observation. For (finput/f1) ¼ 1. Foundation damage of structures.26. analysis and observations.904 S. Geotechnique 2001. [43] Dezi F. Kinematic interaction analysis of piles in layered soils. ASCE 1983. DM 14 gennaio 2008. Dente G. India. Soil Dynamics and Earthquake Engineering 2010. Seismic soil–pile–structure interaction based on large shaking table tests. University of California. DC. ASCE 1982. In: Proceedings of the XVII Convegno Nazionale di Geotecnica. [30] Guin J. applying Eq. 2009. [6] Pender M. ASCE 1996. Louis. Nikolaou A.0036. Shaking table testing of seismic building–pile– two layered–soil interaction. 1. Kraft LM.

Simonelli AL. Berkeley. doi:1-2/RIG 2011. PhD dissertation. Sica et al. ABAQUS Inc. Italian Geotechnical Journal.11:574–91. [60] Roesset JM. Gould PL. 1972. Earthquake Engineering and Structural Analysis 1984. Vicente E. Rosset JM. G. Soil Dynamics and Earthquake Engineering 1996. Curras CJ. 2. Mylonakis G. Kavvadas M. ASCE 1999.dsg.S. USA: SAS IP. Aversa S. [58] Novak M. Sica S. [49] Maheshwari BK. [63] Dobry R. Universita’ degli Studi di Napoli Federico II. Kausel E.E.No. 2004. 70-10. ¨ [55] Kausel E. [62] Mylonakis G. 220 pp. Wu G. Nonlinear analysis of dynamic lateral pile response. vol. Reggio Calabria. Seismic evaluation of pile foundations with three different methods based on three-dimensional elasto-plastic finite element analysis. El Naggar MH. Stewart JP. State University of New York at Buffalo. 2009. Primi elementi in materia di criteri generali per la classificazione sismica del territorio nazionale e di normative tecniche per le costruzioni in zona sismica. [70] ABAQUS. Stiffness matrices for layered soils. Report NCEER-95-0022. In: Proceedings of the fourth international conference on earthquake geotechnical engineering. [61] Gazetas G.12:521–36. [68] Seed HB. SISMA (Site of Italian Strong-Motion Accelerograms): a web-database of ground motion recordings for engineering applications. Bulletin of the Seismological Society of America 1981. In: Proceedings of the second international conference on numerical methods in geomechanics. Melville. Kinematic bending moments in seismically stressed piles. 1976. Journal of Geotechnical and Geoenvironmental Engineering. [50] Maiorano RMS. Seismic pile– group–structure interaction. Journal of Geotechnical and Geoenvironmental Engineering 1998. U. User’s manual. Ground motions and soil liquefaction during earthquakes. National Center for Earthquake Engineering Research. [67] Schnabel B. Truman KZ. Abghari A. no. 2008. EERA—a computer program for equivalent-linear earthquake site response analyses of layered soil deposits. Effects of soil non-linearity on bending moments in piles due to seismic kinematic interaction. Soils and Foundations 2000. In: ASCE specialty conference. Canadian Geotechnical Journal 2004. p. SHAKE: a computer program for earthquake response analysis of horizontally layered sites. Civil Engineering and Nuclear Power. New York: AIP.40(5):113–32. [52] Scasserra G. Lysmer J. 1980. Dynamic stiffness and damping of piles.A.htmlS. p.R. The use of simple models in soil-structure interaction. CA: Earthquake Engineering Research Institute. Wilson DW. Thessaloniki. [71] Nikolaou A. 1001–12. Seismic soil–structure interaction for pile supported system. O’Rourke MJ. [69] Cairo R. NY. . Bray JD. Gazetas G. Idriss IM. 2006. ASCE 1982. [54] Rathje M. Shimizu K. In: Proceedings of the seismic engineering conference commemorating the 1908 Messina and Reggio Calabria Earthquake. 2000.108(GT3):439–59. User’s manual. 1995. Knoxville. Fan K. Canada: Wutec Geotechnical International.. Buffalo. Makris N. vol. vol. [53] OPCM 3274. Journal of Geotechnical Engineering Division. University of Southern California.S.41(1):118–33. VERSAT-P3D: quasi-3D dynamic finite element analysis of single piles and pile groups. 34(ASCE):56–93. and DS. Piles under dynamic loads.itS.simulia. Simplified frequency content estimates of earthquake ground motions. Dente G.com/support/documentation. [66] Bardet JP. Seismic soil–pile– structure interaction experiments and analyses. 2. Efficient analysis of dynamic response of linear system. Roesset JM.125(9):750–9. Contributions to static and seismic analysis of piles and pilesupported bridge piers. New York. Ichii K. [48] El Naggar MH. Abrahamson NA. uniroma1. version 2006. Earthquake Engineering Research Center. TN..71(6):1743–61. / Soil Dynamics and Earthquake Engineering 31 (2011) 891–905 905 [46] Boulanger RW. Novak M.15:233–44. PhD dissertation. State University of New York. Seismic response of piles and pile-supported bridge piers evaluated through case histories. 105 8/5/2003. Houston. Greece. [57] ANSYS. in press. Seed H. D’elia B. Tazoh T. 1982 134 pp. Lin CH. [56] Di Laora R. Ventura CE. [47] Kimura M. Soil-pile kinematic interaction: from research to practice. Lanzo G. 1995. Roesset JM. 3D FEM nonlinear dynamic analysis of pile groups for lateral transient and seismic excitations. Napoli. [59] Blaney GW. Report E. On line documentation /http://www. Horizontal stiffness and damping of single piles. 2003 [in Italian]. [51] Veletsos AS. University of California. paper 1574. Dynamic stiffness of piles.124(2):150–9. PhD dissertation. [65] Wu G. 19th ed. 2.U. The City College and the Graduate Center of the City University of New York. Canadian Geotechnical Journal 1974. 2007. Kutter BL. Zhang F. 1649–56 /http://sisma. Berkeley.C. [64] Syngros C. Geotechnical Special Publication 1992.