You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/263918006

Capacity Design of Retaining Structures and Bridge Abutments with Deep


Foundations

Article in Journal of Geotechnical and Geoenvironmental Engineering · July 2013


DOI: 10.1061/(ASCE)GT.1943-5606.0000825

CITATIONS READS
13 969

2 authors, including:

Sebastiano Rampello
Sapienza University of Rome
92 PUBLICATIONS 1,108 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Seismic Performance of Earth Structures View project

Soil structure interaction View project

All content following this page was uploaded by Sebastiano Rampello on 07 August 2016.

The user has requested enhancement of the downloaded file.


Capacity Design of Retaining Structures and Bridge
Abutments with Deep Foundations
Luigi Callisto1 and Sebastiano Rampello2

Abstract: This paper examines a seismic capacity design approach for retaining structures with piled foundations, which assumes full-strength
Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

mobilization in the soil and in the foundation piles during the earthquake. The plastic mechanism activated by the seismic forces consists of
a horizontal movement of the structure, involving plastic hinging in the piles. This mechanism is triggered when the seismic inertial forces
acting within the structure and the soil mass equal the overall strength of the soil-pile foundation system. The paper describes an iterative pro-
cedure for evaluating the critical seismic acceleration that activates the plastic mechanism. The seismic performance of the structure is expressed
by its permanent displacements and the corresponding curvature ductility demand in the foundation piles. With reference to an idealized bridge
abutment, this procedure is expressed in a fully consistent nondimensional form and is applied to a reference case, to show its potentiality and to
discuss the influence of a number of key parameters, such as the soil strength and the foundation geometry on the seismic performance of the
structure. DOI: 10.1061/(ASCE)GT.1943-5606.0000825. © 2013 American Society of Civil Engineers.
CE Database subject headings: Retaining structures; Pile foundations; Earthquakes; Displacement; Seismic effects; Deep foundation;
Bridge abutments.
Author keywords: Earth retaining structures; Piled foundations; Earthquake-induced displacements; Seismic capacity design.

Introduction Construction codes, such as EN 1998-2 [European Committee


for Standardization (CEN) 2005] and D.M. 14.01.2008 (Ministero
The capacity design of an RC structure implies that during a severe delle Infrastrutture 2008), state that when the attainment of the piles’
earthquake, the strength of some member is fully mobilized: energy- flexural strength is unavoidable, specific consideration should be
dissipating elements of a plastic mechanism are chosen and suitably given to the possible development of plastic hinges in the piles and
detailed for ductility whereas the remaining members are provided ductility detailing should be provided. Despite this, the nature of the
with a reserve of strength capacity sufficient to ensure that the checks that the designer should carry out in this eventuality is still
chosen plastic mechanism is maintained at nearly its full strength unclear. If during the earthquake the maximum internal forces are
during the earthquake. larger than the piles capacity, then the design must be based on an
For piled foundations of RC structures, a common capacity de- explicit prediction of the seismic performance of the structure, which
sign philosophy consists of forcing plastic hinging at the base of the in turn implies a direct evaluation of the seismic displacements and
columns, preventing the foundation piles from fully mobilizing the associated ductility demand.
their flexural strength (Priestley et al. 1996). This approach derives A structure subjected to significant earth pressure, be it a bridge
mainly from difficulties in investigating pile conditions after an abutment or a retaining wall, has nonsymmetrical seismic behavior
earthquake, and in repairing pile damage. as it essentially displaces away from the soil fill; therefore, its
In some cases, this approach may not be feasible because of the seismic performance must be evaluated on the basis of the accu-
very large flexural strength of the structure. This is often the case for mulated final displacements. A common procedure to evaluate these
a conventional bridge abutment with wing walls, for which the displacements is to perform a rigid sliding-block analysis (Newmark
previously mentioned capacity design criterion would result in very 1965) that requires the evaluation of the critical acceleration that
large horizontal forces transmitted to the pile heads and a conse- activates the plastic mechanism and is based on the integration of
quent uneconomical design. In these circumstances, it may be the equation of relative motion. This procedure is based on a rigid-
necessary to take into consideration the full mobilization of the perfectly plastic assumption: permanent displacements occur only
flexural strength of the foundation piles. when the strength of the system is fully mobilized.
In principle, the computation of the permanent displacements
requires a description of seismic loading through the entire accel-
eration time history. Simplified methods do exist (Richards and
1
Associate Professor, Dept. of Structural and Geotechnical Engineering, Elms 1979) that are based on a parametric integration of a number
Univ. of Rome La Sapienza, via Eudossiana 18, 00184 Rome, Italy of seismic records. These methods relate the final displacement to
(corresponding author). E-mail: luigi.callisto@uniroma1.it selected ground-motion parameters, such as the maximum accel-
2
Professor, Dept. of Structural and Geotechnical Engineering, Univ. of eration, and to the seismic resistance of the system under consid-
Rome La Sapienza, via Eudossiana 18, 00184 Rome, Italy.
eration, expressed by its critical acceleration. Regardless of the
Note. This manuscript was submitted on December 6, 2011; approved on
August 27, 2012; published online on August 31, 2012. Discussion period specific procedure adopted, the calculation of the seismic permanent
open until December 1, 2013; separate discussions must be submitted for displacement always requires an estimate of the critical acceleration
individual papers. This paper is part of the Journal of Geotechnical and that activates the plastic mechanism under consideration.
Geoenvironmental Engineering, Vol. 139, No. 7, July 1, 2013. ©ASCE, This paper presents a methodology for evaluating the critical
ISSN 1090-0241/2013/7-1086–1095/$25.00. acceleration ac of a retaining structure with piled foundations,

1086 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013

J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.


relative to a plastic mechanism consisting of a rigid horizontal Only the seismic forces induced by a horizontal seismic co-
translation that entails the development of plastic hinges in the efficient k are considered, as follows:
foundation piles. In this evaluation, a number of reasonable sim- • The inertial force kW applied at the center of gravity of the
plifying assumptions are made, some of which can easily be mod- equivalent structure;
ified by a prospective user. The computations are presented in • A seismic force transmitted at the top of the structure, equal to
a nondimensional form to study the role of the different parameters a fraction kQ 5 zQ 3 k of the vertical force Q; a value of zQ 5 0
and to generalize the results, making them useful for practical represents a frictionless bearing, whereas zQ 5 1 may represent
application. The use of the critical acceleration to evaluate the full-force transmission at the top of the abutment; and
seismic performance of the structures at hand is discussed, and • The resultant of the seismic soil pressure, DSE , evaluated as
the procedure is applied to a reference case for illustrative purposes.
1
DSE ¼ g  H 2  ðKaE 2 Ka Þ ð3Þ
2
Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

Problem Definition
and acting at a vertical distance bSE H from the base. The active
Fig. 1 depicts a schematic layout of the problem in an equivalent seismic coefficient KaE is evaluated with the Mononobe-Okabe
plane-strain representation: a structure of width B retains a soil fill or formula, assuming a vertical soil-wall contact surface, no frictional
an excavation of total height H; the structure as a whole, including soil-wall resistance, and a horizontal backfill
the structural material and the soil directly insisting on the foun-
dation, is fully defined by its equivalent unit weight cos2 ðw9 2 uÞ
KaE ¼ " rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi#2 ð4Þ
sin w9 sinðw9 2 uÞ
geq ¼ W=ðB HÞ ð1Þ cos2 u 1 þ
cos u
where W 5 total weight (per unit length), and by the position of its
center of gravity, expressed by the nondimensional parameters aW where u 5ffi tan21 k is the angle of inclination of the total body force
pffiffiffiffiffiffiffiffiffiffiffiffi
and bW of Fig. 1. W 1 1 k 2 to the vertical.
The structure is founded on equal RC piles of diameter D, The soil fill behind the structure is purely frictional, with an angle
arranged in n rows, with a transverse spacing sT (parallel to B) and of shearing resistance w9fill . The foundation soil can be either fine-
a longitudinal spacing sL . grained with an undrained shear strength Cu or coarse-grained with
The structure is loaded by static and seismic forces. The static an angle of shearing resistance w9. The foundation soil and the fill
part of the earth pressure, S, is taken to act at height bS H from the have the same unit weight g.
base of the structure and is evaluated as A simple elastic-perfectly plastic relationship is assumed be-
tween the bending moment in the piles M and their curvatures c, as
1 depicted in Fig. 2. The yield bending moment My is a function of the
S ¼ g  H 2  Ka ð2Þ
2 axial load P and the structural strength of the pile section, defined by
the following quantities:
where Ka 5 Rankine active pressure coefficient. In addition to S and • Steel yield stress fy ;
W, the top of the structure is loaded by a vertical static force Q, taken • Cylinder compressive strength of the concrete fc ;
to act at horizontal and vertical distances from the toe equal to aQ B • Reinforcement ratio V; and
and bQ H, respectively. • Concrete cover c.

The critical acceleration ac is defined as the horizontal seismic


acceleration that activates the simple plastic mechanism depicted in
Fig. 3: as the seismic forces increase, the total overturning moment
Mov increases, loading the external pile rows and unloading the
internal rows. For a given seismic coefficient k, the axial load acting
at the head of a pile in the ith row, Pi , is found by equilibrium
considerations only, assuming that the pile cap is infinitely stiffer
than the piles, and the axial behavior of each pile is independent
from that of the other piles. Although this last assumption may not be

Fig. 2. Assumed relationship between pile bending moment M and


Fig. 1. Schematic layout of an abutment founded on piles curvature c

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013 / 1087

J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.


Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

Fig. 4. Load-moment interaction diagrams for the pile section studied


in the reference case

Fig. 3. Plastic mechanism activated when the full strength of the system  2=3
Tu My
is mobilized ¼ h 3:676 ð8Þ
Kp gD3 Kp gD4

valid for pile groups loaded under working conditions, it appears In Eqs. (7) and (8), h # 1 is an efficiency factor that accounts for
realistic during the activation of a plastic mechanism, when load group effects, Kp is the Rankine coefficient of passive earth pressure,
concentrations at the pile group periphery have been somewhat and Cu is the undrained shear strength, assumed constant over the
smoothed out by plastic redistribution. It is then limited depth involved in the plastic mechanism (usually smaller
than 10 pile diameters).
sL Mov sL
Pi ¼ ðW þ QÞ þ n d ð5Þ The plastic mechanism is activated when the sum of the hori-
n P 2 i
dj zontal static and seismic forces Eh equals the sum of the horizontal
j¼1 forces that represent the horizontal capacity of the piles in the group.
Therefore, a critical seismic coefficient kc 5 ac =g that satisfies the
where dj 5 distances of each pile axis from the foundation center, condition
and
Pn
Mov ¼ aQ kQbQ H þ kWbW H þ SbS H Eh ¼ zQ kc Q þ kc W þ S þ DSE ðkc Þ ¼ 1 Tui ð9Þ
sL j¼1
B  
þ DSE bSE H 2 Wð2aW 2 1Þ þ Q 2aQ 2 1 ð6Þ
2 is searched iteratively. This iterative procedure requires a number of
subiterations because, for each value of the seismic coefficient, the
is the total overturning moment (per unit length) about the founda- axial forces in the piles of each row vary, and the corresponding
tion center. values of My must be recalculated, as previously described. At the
The yield bending moment Myi for a pile of each row is calculated end of the iterative process, a check can be made that load eccen-
as a function of the axial load Pi , using the Whitney approximation tricity is such that pile yielding is initiated by steel tension, as this
(Park and Paulay 1975) and assuming that yielding of the pile sec- occurrence is favorable for ductility (Fig. 4).
tion is initiated either by steel tension or by concrete compression The procedure gives, in addition to kc , the vertical distance zy
(Fig. 4). This procedure requires iterations to solve the tension and between the two plastic hinges in the piles (Fig. 3); according to
compression equations, because in both expressions, the yield axial Broms (1964a, b) it is
load Py 5 Pi is an implicit function of the load eccentricity e, and
My 5 Py e.
zy 4My
Because My depends on the corresponding pile axial load, it ¼ 2 1:5 ð10Þ
changes continuously during seismic loading, and at a given instant D Tu D
is different for the piles of different rows. The horizontal force Tu
that activates the plastic mechanism for a pile within the group is zy 3My
¼ ð11Þ
in turn dependent on its yield bending moment My , and on the soil D Tu D
strength parameters; within this paper, Tu is calculated using the
simplified formulas proposed by Broms (1964a, b) (but any equiv- for cohesive and frictional soils, respectively.
alent expression can easily be incorporated in the present method); The distance zy is not the same for all the pile rows: it depends
for cohesive and frictional soils it is, respectively on My , that in turn depends on Pi and hence on the position of the
 rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi row. For a given horizontal displacement u, the maximum plastic
Tu My rotation, and therefore the largest bending ductility demand for the
2
¼ h 213:5 þ 182:25 þ 36 ð7Þ
Cu D Cu D3 pile section, corresponds to the smallest value of zy (Fig. 3).

1088 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013

J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.


In the present implementation of this procedure, it is assumed that example, in the following, the attention is restricted to the Italian
the yield bending moment My acting in the two plastic hinges of seismicity: Rampello et al. (2010) performed a parametric integra-
a given pile is the same, implying that the axial load in a pile does tion of the entire set of Italian acceleration time histories included
not change significantly over the depth zy . Further simplifying in the database Site of Italian Strong-Motion Accelerograms (SISMA)
assumptions are as follows: (Scasserra et al. 2009), expressing the permanent displacement u
• The structure width B is directly related to the pile diameter and corresponding to the 94th upper-bound percentile as a decreasing
to the transversal spacing function of the ratio kc =kmax , where kmax 5 amax =g is the seismic
coefficient corresponding to the maximum horizontal acceleration
B ¼ ðn 2 1ÞsT þ 2D amax . In the following, a modified form of the equation proposed by
the authors is used:
• Both the static and the seismic earth pressure resultants act at
a distance from the base equal to one-third of the overall height,  kc

2A kmax
Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

as suggested by EN 1998-5 (CEN 2003b) u ¼ u0  kmax  e ð12Þ


1
bS ¼ bSE ¼ where u0 5 reference displacement corresponding to kc 5 0 and
3
kmax 5 1.
• The transversal pile spacing within a row is equal to the longi- The values of A and u0 found by Rampello et al. (2010) are
tudinal spacing between the rows reported in Table 1, grouped according to the EC8 subsoil categories
(CEN 2003a). Values of the coefficient A are in the range of 7.4–8.1,
sL ¼ sT with an average value of 7.7. The reference displacement u0 depends
essentially on duration and frequency content of each acceleration
With these assumptions, dimensional analysis shows that the time history. For the Italian records, the values of u0 associated with
problem is controlled by 18 independent nondimensional parameters, the 94th upper-bound percentile range between 2.14 and 11.8 m,
which may be chosen as follows: with an average value of 6.51 m and a coefficient of variation of
• n number of pile rows; about 40%.
• H=D height to diameter ratio; It is interesting to compare the preceding values with those
• sT =D pile spacing to diameter ratio; resulting from the expression proposed by Whitman and Liao (1984)
• g eq =g unit weight ratio; for the evaluation of seismic displacements of retaining walls. The
• g eq H=fc concrete resistance ratio; Whitman and Liao equation is similar to Eq. (12), with A 5 9:4 and
• fc =fy concrete to steel resistance ratio; the term u0 3 kmax replaced by C 3 v2max =amax , where C is equal to 37
• V reinforcement ratio; and 493 for the 50th and the 95th percentiles, respectively (Whitman
• c=D concrete cover to diameter ratio; 1990). The peak horizontal ground velocity vmax can be obtained
• Q=geq BH external load ratio; from the design elastic response spectrum, as prescribed by D.M.
• kQ =kð 5 zQ Þ seismic coefficient ratio; 14.01.2008 (Ministero delle Infrastrutture 2008)
• aQ horizontal position of external load;
bQ vertical position of external load; TC
• vmax ¼ amax ð13Þ
• h efficiency of the pile group at ultimate horizontal load; 2p
• aW horizontal position of the center of gravity;
• bW vertical position of the center of gravity; where TC 5 period corresponding to the beginning of the constant-
• w9fill angle of shearing resistance of fill; velocity portion of the design spectrum. Hence, the reference dis-
• kc 5 ac =g critical seismic coefficient placement u0 for the 95th percentile can be evaluated for the
and one of the following two: Whitman and Liao expression as
• Cu =gD undrained shear strength ratio (for fine-grained founda- 2
TC
tion soils); and u0 ¼ C  g ð14Þ
• tan w9=tan w9fill angle of friction ratio (for coarse-grained foun- 2p
dation soils).
with C 5 493.
Table 2 lists the values of TC for the EC8-1 subsoil classes,
Computation of Seismic Displacements identified by the equivalent value of the shear-wave velocity over the
depth of 30 m, Vs30 . Table 2 also reports the values of u0 computed
For a given value of the critical seismic coefficient, the displace-
ment u induced by the earthquake can be evaluated assuming Table 1. Coefficients of Eq. (8) for Upper-Bound 94th Percentile
a rigid-perfectly plastic behavior of both the soil and the structure, Permanent Displacements Induced by Italian Seismic Events (Data from
through a simple Newmark integration of the portion of the accel- Rampello et al. 2010)
eration time history that exceeds the critical acceleration ac ð 5 kc gÞ. EC8-1 subsoil EC8-1 subsoil EC8-1 subsoil
This procedure neglects the elastic deformation of the abutment and
the piles, which may be deemed small if compared with the plastic Classes C, D,
deformation produced by the activation of the plastic mechanism. Class A Class B and E
The main difficulty in performing a Nemark integration is that kmax A u0 (m) A u0 (m) A u0 (m)
the seismic motion must be described through acceleration time his-
0.35 7.50 4.83 7.90 4.54 7.40 2.14
tories, rather than by simple ground-motion parameters. An alter-
0.25 7.42 7.16 7.79 6.64 7.54 3.18
native would be to rely on simplified methods, which should
0.15 7.48 6.07 7.86 7.27 8.05 7.73
profitably be based on the parametric integration of earthquake re-
0.05 7.87 7.80 7.86 9.00 8.07 11.80
cords that are representative of the regional seismicity. As an

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013 / 1089

J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.


Table 2. Values of Reference Displacement u0 Obtained Using Whitman
and Liao (1984)
Type 1 Type 2
(Ms . 5:5) (Ms , 5:5)
EC8-1 subsoil
class Vs30 (m/s) TC (s) u0 (m) TC (s) u0 (m)
A .800 0.4 19.6 0.25 7.65
B 360–800 0.5 30.6 0.25 7.65
C 180–360 0.6 44.1 0.25 7.65
D ,360 0.8 78.4 0.30 11.00
E As C and D but with 0.5 30.6 0.25 7.65
Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

thickness ,20 m

using Eq. (14) for two types of design spectrum that CEN (2003b)
proposes, namely, for seismic events with surface wave magnitude
Ms smaller (Type 1) and larger (Type 2) than 5.5, respectively. The
values of u0 found for Ms , 5:5 are close to the upper bound of the
range obtained by Rampello et al. (2010). Conversely, the values of
u0 for the Type 1 spectra are much larger, but this is understandable,
because the Italian seismic events considered by Rampello et al.
(2010) have a magnitude Ms of 4–6.6, which makes the Type 2
spectra more appropriate in the average.
Once the value of the permanent displacement u has been
obtained, either by a direct Newmark integration or through one of
the previously mentioned relationships, the maximum rotation in
the plastic hinges is computed as
 
u
v ¼ tan21 ð15Þ
zy,min

where zy,min 5 minimum depth of the second plastic hinge, which


develops in the piles of the internal row. The maximum plastic
curvature cp can be evaluated from v
v
cp ¼ ð16Þ
lp

where lp 5 equivalent length of the plastic hinge, which can be Fig. 5. Reference case: bridge abutment resting on foundation piles
calculated using a number of empirical expressions. For instance, arranged along five rows
Mattock (1967) suggested the simple form

lp ¼ 0:5ðD 2 cÞ þ 0:05 h ð17Þ


external load ratio was chosen to produce, under static conditions,
in which h 5 distance of the critical section from the point of a ratio of the nominal normal stress to the concrete cylinder strength
counterflexure, which in the present case may be taken equal to sc =fc not larger than 0.2 (see Fig. 4). This preliminary computation
zy =2. If cy is the curvature at yielding, then the curvature ductility was carried out using the earth coefficient at rest K0 5 1 2 sin w9fill
demand x is (Jaky 1944) to evaluate the soil pressure acting against the abutment,
because the static horizontal displacement is typically insufficient to
cp mobilize active limit conditions in the backfill.
x ¼1þ ð18Þ The choice of the pile-group efficiency (h 5 0:55) was based on
cy
the study by Mokwa (1999), that, analyzing the observed resistance
of pile groups to horizontal loads, obtained a relationship between
which must be compared with the ductility capacity of the RC
h and sT =D of the type
section.
h ¼ c1 þ c2 ðsT =D 2 1Þ ð19Þ
Evaluation of the Reference Case
with c1 5 0:25 and c2 5 0:15, corresponding to an average estimate
of the group efficiency.
Problem Layout
The cases of fine-grained and coarse-grained foundation soils are
Fig. 5 shows the layout of an example bridge abutment resting on studied separately, evaluating kc and zy =D as a function of Cu =gD
foundation piles arranged along five rows. The values of the non- and tan w9=tan w9fill , respectively. For a given horizontal displace-
dimensional parameters for this case are reported in Table 3. The ment, the largest rotation in a plastic hinge is associated with the

1090 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013

J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.


smallest depth of the plastic hinge zy,min , which occurs for the piles A value of zQ 5 0 means that no horizontal load is transferred to the
in the internal row, which are subjected to the smallest axial load. abutment by the bridge structure; conversely, zQ 5 1 accounts for full
load transfer in the hypothesis that the structure response is very stiff.
Increasing zQ from 0 to 1 has the obvious effect of decreasing the
Critical Acceleration and Plastic Mechanism critical seismic coefficient; for the present case, such a decrease is of
Figs. 6(a and c) show the variation of the critical seismic coefficient the order of 30–40% [Figs. 6(a and b)]. On the other hand, an increase
kc and the minimum nondimensional depth of the second plastic in the reinforcement ratio from 0.5 to 1.0% produces an increase of the
hinge, zy,min =D, as the normalized undrained shear strength of seismic coefficient of about 20–30%. The depth of the plastic
a fine-grained foundation soil varies from 1 to 10. In Figs. 6(b and d), kc mechanism is only slightly influenced by the aforementioned factors.
and zy,min =D are plotted as a function of the strength ratio tan w9=tan w9fill For a given seismic loading, the increase of kc produces a re-
that varies from 0.4 to 1.4 in the assumption of a coarse-grained duction of the permanent displacement, but the corresponding de-
foundation soil. The reference case is indicated by a bold continuous crease of zy,min =D, evidenced by Fig. 6, may produce contrasting
Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

line, whereas thinner and dashed lines were obtained varying the effects on the plastic rotation [Eq. (15)]. However, the combined
reinforcement ratio V from 0.5 to 1% and the seismic coefficient ratio effect of an increase in kc and a decrease in zy,min =D is always
zQ from 0 to 1. a decrease of the plastic rotation, because the reduction of the
In all cases, kc increases and zy,min =D decreases with the nor- displacement u with kc [Eq. (12)] has typically a large gradient
malized soil strength. For a coarse-grained foundation soil these if compared with the decrease of zy,min =D shown in Fig. 6.
variations are almost linear, whereas for a fine-grained foundation The effect of the pile group geometry was studied varying the
soil in undrained conditions the curves have a decreasing gradient, number of pile rows from 3 to 7 for the abutment of Fig. 5 and adjusting
and kc and zy,min =D become about stationary at large values of the pile diameter to maintain a constant B=H ratio. Therefore, a larger
Cu =gD. As a general result, the depth of the plastic mechanism is number of pile rows correspond to a smaller pile diameter.
from 5 to 8–12 times the pile diameter, the largest values being Figs. 7(a and b) show the variation of the critical seismic co-
obtained for fine-grained foundation soils with a very small non- efficient as a function of the number of pile rows n. As n increases
dimensional undrained shear strength. from 3 to 7, the critical seismic coefficient kc is seen to decrease by an

Table 3. Values of Nondimensional Factors for Reference Case of Fig. 5


n H=D sT =D g eq =g g eq H=fc fc =fy V c=D Q=g eq BH kQ =k aQ bQ h aW bW w9fill
5 12 3 0.9 0.012 0.05 0.5% 0.06 0.523 0 0.18 0.78 0.55 0.55 0.46 35

Fig. 6. Reference case: (a) and (b) critical seismic coefficient kc ; (c) and (d) minimum nondimensional depth of the second plastic hinge as a function
of the normalized strength of the foundation soil; plots (a) and (c) are for fine-grained soils; plots (b) and (d) are for coarse-grained soil

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013 / 1091

J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.


amount, which is almost independent of the soil normalized strength, values of the maximum acceleration varying from 0.2 to 0.4g. The
equal to about 0.07 for the fine-grained soils and 0.1 for the coarse- seismic performance is quantified through the final permanent
grained soils; in addition, because the pile diameter become smaller, displacement u [Eq. (12)] and the corresponding plastic curvature cp
the plastic mechanism becomes shallower producing for a given [Eq. (16)].
horizontal displacement, a larger plastic curvature. Hence, it appears Figs. 8(a and b) show the permanent displacement computed
convenient for a good seismic performance to use a limited number using Eq. (12) with the coefficients of Table 1 (classes C, D and E)
of pile rows with a larger diameter, rather that many rows of small- and values of the maximum seismic coefficient kmax varying from 0.2
diameter piles. to 0.4. In the investigated range of normalized strengths and
maximum acceleration, the permanent displacements are mostly
Seismic Performance smaller than 0.1–0.2 m. Figs. 8(c and d) show that the corre-
For the reference case of Fig. 5 and Table 3, the seismic performance sponding plastic curvatures, evaluated using Eqs. (16) and (17), do
was evaluated with the simplified aforementioned procedure, using not exceed 0.01–0.02 m21. Fig. 9 depicts the moment-curvature
Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

Fig. 7. Influence of the number of pile rows on the critical seismic coefficient: (a) fine-grained foundation soils; (b) coarse-grained foundation soils

Fig. 8. (a) and (b) displacements and (c) and (d) plastic rotations computed for the reference case of Fig. 5

1092 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013

J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.


Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

Fig. 9. Complete and simplified moment-curvature diagram for the


foundation piles of the reference case Fig. 10. Maximum forces acting on the abutment during the activation
of the plastic mechanism

diagram of the pile section obtained from a rigorous analysis


(software Response-2000; Bentz 2000) and the elastic-perfectly
plastic simplified behavior used in the current study. The plot of
Fig. 9 was obtained for an axial load P 5 4:8 MN, which is the
average pile load under static conditions. Applying an energy
balance criterion to the first part of the M-c curve, a curvature at yield
cy 5 2:1 3 1023 m21 is found. This value of cy can be used, to-
gether with Eq. (18), to mark on the plots of Figs. 8(c and d) the
curvatures corresponding to several values of ductility demand x.
Inspection of Figs. 8(c and d) reveals that, for the present reference
case, the curvature ductility demands do not exceed the value of 10,
and for most practical cases, it is smaller than 3-4. Moreover, for the
case at hand, the pile yielding is always initiated in tension, resulting
in a ductile behavior of the RC section in bending. Therefore, it
appears that the computed curvature demand can be easily met by
a proper detailing of the RC section. In other words, there seems to be
a good match between the values of the permanent displacement that
one would accept as a reasonable performance of the structure under
a severe earthquake (at least a few centimeters) and the corre-
Fig. 11. Normalized maximum force on the bridge abutment plotted
sponding ductility demand on the RC sections of the piles.
as a function of the critical seismic coefficient
In the present design approach it is assumed that during a severe
earthquake, the strength is attained in the foundation piles only,
whereas the superstructure remains elastic. To satisfy this re- Conclusions
quirement, a check should be made that the strength of the structural
members is larger than the maximum internal forces. As the seismic Foundation piles are used essentially to carry large vertical loads;
acceleration increases, so do the internal forces; however, when the when a pile foundation is subjected to lateral loads, as the ones
strength of the plastic mechanism is reached, that is, when k 5 kc , originated by a severe earthquake, the flexural behavior of the piles
the internal forces cannot increase further, because all the terms comes into play and the foundation tends to show a limited capacity.
in Eq. (9) depend on the acceleration of the abutment that cannot In capacity design, flexural yielding is preferred because of its
exceed kc . ductility. However, plastic hinging in the superstructure may become
When k 5 kc , the internal forces (per unit length) in the super- difficult to achieve in many typical bridge abutments that are char-
structure are proportional to the forces depicted in Fig. 10. Bending acterized by significant flexural strength, because this would result in
moments and shear forces acting in the structural members can be very large forces applied to the foundation piles. Conversely, de-
evaluated assuming suitable distributions of these forces. Hence, the signing for plastic hinging in the piles that have a limited flexural
sum of the forces Eh-c is an effective indicator of the magnitude of strength may result in reasonable internal forces in the superstructure.
the internal forces induced by the earthquake. Once Eh-c is known, Presently, there is a tendency to require that important structures
the earth thrust S 1 DSE ðkc Þ can be evaluated from Eq. (9), as do not collapse for low-probability seismic events related to a large
depicted in Fig. 10. The variation of Eh-c , divided by the weight of life span, that is, for catastrophic events, possibly never actually
the abutment, is shown in Fig. 11 as a function of kc : the steady recorded at the specific structure location. This no-collapse re-
increase of Eh-c with the critical seismic coefficient evidences that if quirement may be fulfilled only by allowing a significant energy
the abutment is designed to have a large critical seismic coefficient, dissipation and a corresponding damage in the structural elements:
to undergo small seismic displacements, the superstructure will in this case, designing for a plastic mechanism that mobilizes the
have to be designed for large internal forces. piles’ flexural strength may contribute to a cost-effective design.

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013 / 1093

J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.


Although damages in foundation piles may be considered difficult Ms 5 magnitude of surface waves;
to detect and repair, seismic remediation of the isolated foundation of n 5 number of pile rows;
a bridge abutment may be relatively manageable (Zelinski et al. 1995). P 5 axial force at pile head [Eq. (5)];
In the present perspective, the adequacy of the foundation is related Q 5 vertical force at the wall head (per unit
to the predicted seismic displacements and the corresponding damage length);
undergone by the energy-dissipating structural elements, which in turn S 5 static earth thrust (per unit length) [Eq. (2)];
can be expressed by the curvature ductility demand. In this paper, sL 5 longitudinal pile spacing;
a simple methodology has been presented for the evaluation of these sT 5 transverse pile spacing;
quantities, based on the study of a plastic mechanism that entails the TC 5 period at beginning of the constant-velocity
simultaneous achievement of the soil and structural strength during
portion of the EC8 spectrum [Eq. (13) and
strong motion. This methodology can be used directly to evaluate
(14)];
the permanent displacements of a structure through the computation
Tu 5 horizontal force activating the plastic
Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

of the critical acceleration, which activates the plastic mechanism


and a subsequent Newmark-type integration. As an alternative, sim- mechanism in a pile [Eq. (7) and (8)];
plified relationships can be used that relate the seismic displacement to u 5 permanent horizontal displacement
the critical acceleration and to ground-motion parameters. The plastic [Eq. (12)];
curvature in the pile sections can then be computed from the position of u0 5 reference horizontal displacement [Eq. (12)];
the lower plastic hinge, which is provided directly from the analysis. v 5 ground velocity [Eq. (13)];
The sum of the horizontal forces carried by each pile when the W 5 weight of the retaining structure (per unit
plastic mechanism is activated provides an upper limit to the inertial length);
forces that can act in the superstructure, and can be used to derive an zy 5 depth of the plastic mechanism [Eq. (10) and
upper bound of the internal forces in the structural members. (11)];
The application to a reference case has shown that (1) the seismic a 5 horizontal distance ratio of a force from the
performance can be improved by increasing the pile diameter and wall toe;
reducing the number of pile rows; (2) even a severe seismic event b 5 vertical distance ratio of a force from the wall
may be accommodated by the structure with a relatively small base;
damage; and (3) permanent displacement generally tolerable for g 5 unit weight of soil;
a retaining structure corresponds to curvature ductility demands that geq 5 equivalent unit weight of the retaining
may be satisfied easily by RC pile sections that are properly detailed. structure [Eq. (1)];
Although these findings are relative to a specific case, the non- DSE 5 seismic increment of the earth thrust (per unit
dimensional formulation ensures that they can be extended to an length) [Eq. (3)];
entire class of structures. Different cases can be readily analyzed zQ 5 horizontal inertia ratio for Q;
implementing the simple procedure described in the paper.
h 5 efficiency of the pile group for horizontal
loads;
Notation sc 5 average compressive stress in the piles before
the earthquake;
The following symbols are used in this paper: w9 5 angle of shearing resistance of the foundation
A 5 coefficient for the evaluation of seismic soil;
displacements [Eq. (12)]; w9fill 5 angle of shearing resistance of the soil located
a 5 k g horizontal acceleration; behind the retaining structure;
B 5 width of the retaining structure; u 5 inclination of the resultant body forces;
C 5 Whitman and Liao (1984) coefficient for the x 5 curvature ductility demand [Eq. (18)];
evaluation of seismic displacements; c 5 pile curvature [Eq. (16)];
Cu 5 undrained shear strength of the foundation V 5 reinforcement ratio of the pile section; and
soil; v 5 plastic rotation in the pile section
c 5 concrete cover of the reinforcing steel; [Eq. (15)].
d 5 distance of pile axis from the foundation Subscripts
center; c 5 critical conditions;
Eh-c 5 total horizontal force activating the plastic i 5 relative to ith pile;
mechanism (per unit length) [Eq. (9)]; p 5 plastic;
fc 5 concrete cylindrical compressive strength; Q 5 relative to the force transmitted at the wall top;
fy 5 yield stress of the reinforcing steel; S 5 relative to the static earth thrust;
H 5 height of the retaining structure; SE 5 relative to the seismic earth thrust increment;
Ka 5 static coefficient of active earth pressure; W 5 relative to the structure’s weight; and
KaE 5 seismic coefficient of active earth pressure y 5 yield conditions.
[Eq. (4)];
Kp 5 coefficient of passive earth resistance;
References
k 5 horizontal seismic coefficient;
lp 5 dimension of plastic hinge [Eq. (17)];
Bentz, E. C. (2000). “Sectional Analysis of Reinforced Concrete Members,”
M 5 bending moment; Ph.D. Thesis, Department of Civil Engineering, Univ. of Toronto, Toronto.
Mov 5 total overturning moment (per unit length) Broms, B. B. (1964a). “Lateral resistance of piles in cohesive soils.” J. Soil
[Eq. (6)]; Mech. Found. Div., 90(2), 27–63.

1094 / JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013

J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.


Broms, B. B. (1964b). “Lateral resistance of piles in cohesionless soils.” Priestley, M. J. N., Seible, F., and Calvi, M. (1996). Seismic design and
J. Soil Mech. Found. Div., 90(3), 123–156. retrofit of bridges, Wiley, New York.
European Committee for Standardization (CEN). (2003a). “Eurocode 8: Design Rampello, S., Callisto, L., and Fargnoli, P. (2010). “Evaluation of slope
of structures for earthquake resistance—Part 1: General rules, seismic actions performance under earthquake loading conditions.” Rivista Italiana di
and rules for buildings.” EN 1998-1, CEN, Brussels, Belgium. Geotecnica. Pàtron Editore, Bologna, 44(10), 29–41.
European Committee for Standardization (CEN). (2003b). “Eurocode 8: Design Richards, R., and Elms, D. (1979). “Seismic behavior of gravity retaining
of structures for earthquake resistance—Part 5: Foundations, retaining walls.” J. Geotech. Eng. Div., 105(4), 449–464.
structures and geotechnical aspects.” EN 1998-5, CEN, Brussels, Belgium. Scasserra, G., Stewart, J. P., Kayen, R. E., and Lanzo, G. (2009). “Database
European Committee for Standardization (CEN). (2005). “Eurocode 8: for earthquake strong motion studies in Italy.” J. of Earthquake Eng.,
Design of structures for earthquake resistance—Part 2: Bridges.” EN 13(6), 852–881.
1998-2, CEN, Brussels, Belgium. Whitman, R. V. (1990). “Seismic design and behaviour of gravity retaining
Jaky, J. (1944). “The coefficient of earth pressure at rest.” J. Soc. Hung. walls.” Design and performance of earth retaining structure: Geo-
Eng. Arch., 355–358 (in Hungarian). technical special publication, No. 25, P. C. Lambe and L. A. Hansen,
Downloaded from ascelibrary.org by UNIV LA SAPIENZA DI ROMA on 06/12/13. Copyright ASCE. For personal use only; all rights reserved.

Mattock, A. H. (1967). “Discussion on ‘Rotational capacity of reinforced eds., ASCE, New York, 817–842.
concrete beams’ by W.G. Corley.” J. Struct. Div., 93(2), 519–522. Whitman, R. V., and Liao, S. (1984). “Seismic design of gravity retaining
Ministero delle Infrastrutture. (2008). “Nuove norme tecniche per le cost- walls.” Proc., 8th World Conf. Earthq. Eng., Vol. 3, Earthquake
ruzioni.” Gazzetta Ufficiale della Repubblica Italiana, 29, Rome (in Italian). Engineering Research Institute Staff, International Association for
Mokwa, R. L. (1999). “Investigation of the resistance of pile caps to lateral Earthquake Engineering, San Francisco, 533–540.
loading.” Ph.D. thesis, Virginia Tech, Blacksburg, VA. Zelinski, R., Roblee, C., and Schanz, T. (1995). “Bridge foundation re-
Newmark, N. M. (1965). “Effect of earthquakes on dam and embankment. mediation consideration.” Earthquake-induced movements and seismic
The Rankine Lecture.” Geotechnique, 15(2), 139–159. remediation of existing foundations and abutments: Geotechnical spe-
Park, R., and Paulay, T. (1975). Reinforced concrete structures, Wiley, cial publication, No. 55, S. L. Kramer and R. Siddhartan, eds., ASCE,
New York. New York, 49–68.

JOURNAL OF GEOTECHNICAL AND GEOENVIRONMENTAL ENGINEERING © ASCE / JULY 2013 / 1095

View publication stats J. Geotech. Geoenviron. Eng. 2013.139:1086-1095.

You might also like