Professional Documents
Culture Documents
ABSTRACT
Wood (1973) provided analytical solutions for the response of a rigid wall retaining elastic uniform soil backfill of finite length
subjected to harmonic base excitation. Wu and Finn (1999) proposed a modified shear beam solution and derived closed-form
formulations for computing dynamic soil pressures under harmonic loading. However, for rigid walls retaining sloped backfills,
analytical solutions are not available for computing seismic soil pressures against the walls.
In the current study, seismic soil pressures on rigid walls retaining 2H:1V (27) sloped backfills have been computed using a total
of eight acceleration time histories recorded in past large earthquakes. These records were selected and linearly scaled to three
levels of ground motions with a nominal PGA of 0.26g, 0.48g and 0.71g. Nonlinear time-history analyses were conducted using
the computer program VERSAT-2D which uses a hyperbolic stress strain model to simulate the hysteresis response of soil
under cyclic loads. Soil pressure diagrams are shown in the paper for horizontal backfills (=32), and for sloped backfills with
loose sand (=32) and dense sand (=40) under the three levels of ground motions.
A soil pressure coefficient, K0E, has been introduced to represent the total static and seismic pressures on a rigid (or non-yielding)
wall. It is found that K0E varies from 1.1, 1.7 to 2.2 for horizontal backfills under ground motions of 0.26g, 0.48g and 0.71g,
respectively. K0E increases to 2.7, 3.8 and 4.9 for the 27 sloped backfills under the same three levels of ground motions,
respectively. The point of thrust is at about 0.47H above the base of wall for horizontal backfills, but for sloped backfills it
increases to 0.53H for loose sand and 0.55H for dense sand.
INTRODUCTION
Wood (1973) provided analytical solutions for the response
of a rigid wall retaining elastic uniform soil backfill of finite
length subjected to harmonic base excitation. Due to its
mathematical complexity, the application of Woods
solution is very much limited to a very low vibration
frequency or the so-called static solution. Wu and Finn
(1999) proposed a modified shear beam solution and
derived closed-form formulations for computing dynamic
soil pressures under harmonic loading. For earthquake
loading, they provided design charts of seismic soil
pressures that take into account the effect of vibration
frequency on the thrust. However, these solutions are only
applicable for horizontal backfills.
For rigid walls retaining sloped backfills, analytical
solutions are not available for computing seismic soil
pressures against the walls.
The record search was carried out primarily from the PEER
NGA and the COSMOS record databanks. The Pacific
Earthquake Engineering Research Center maintains the
PEER NGA databank which can be found in
The Consortium of
http://peer.berkeley.edu/smcat/.
Organization for Strong-Motion Observation Systems
(COSMOS) manages the COSMOS virtual data centre that
can be located in http://db.cosmos-eq.org/scripts/default.plx.
The core members of COSMOS include US Geological
Survey, California Geological Survey, US Army Corps of
Engineers and US Bureau of Reclamation.
Due to insufficient number of earthquake records that
satisfy all selection criteria, some selection requirements
were relaxed to include ground motions recorded at an
earthquake source-site distance of 30 km and site shear
wave velocity less than 560 m/s.
2.0
%lul 090
%gil 067
%bld 360
1.5
%gaz 000
%tabas LN
SA (g)
%tcu W
1.0
%teresa 225
Ruskin rock site
0.5
0.0
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
Period (sec)
1.2
SA (g), 5% damped
1.0
0.8
0.6
0.4
0.2
0.0
0.00
0.20
0.40
0.60
0.80
1.00
1.20
1.40
1.60
1.80
2.00
Period (sec)
Table 1. Summary of the Eight Earthquake Records for Ground Motions with AEF of 1/2475
Record
#
Earthquake, Station
& Short Name
Magnitude
Mechanism
Duration
(sec)
RRUP
(km)
RJB (km)
6.9
Reverse
Oblique
50
15
14
7.6
7.4
6.8
1994/01/17 US Northridge,
#24157 Baldwin Hills LA,
%bld
1980/05/25 US Mammoth
Lakes, 54214 Long Valley
dam (Upr L Abut), %lul
1979/10/15 US Imperial
Valley, 6604 Cerro Prieto,
%cpe
6.7
6.9
6.3
6.5
Reverse
Normal
Reverse
Normal
Reverse
Normal
Reverse
Normal
Reverse
Oblique
Reverse
Oblique
Strike Slip
90
33
16
40
40
30
64
(1)
30
23
10
15
13
15
15
Vs30
(m/s)
Geomatrix
Class
672
625
767
660
297
730
345
660
Component
(2)
(3)
PGA
(g)
(3)
PGV
(cm/s)
(3)
PGD
(cm)
Scaling
Factor
PGA after
scaling (g)
#225 (S)
0.274
26.2
13.15
1.6
0.438
#315
0.228
20.6
5.27
1.6
0.365
Up/Down
0.209
16.7
4.97
1.6
0.334
0.655
69.4
49.06
0.7
0.459
W (S)
0.567
44.4
13.76
0.7
0.397
0.449
34.8
31.32
0.7
0.314
LN (S)
0.836
97.8
36.92
0.5
0.418
TR
0.852
121.4
94.58
0.5
0.426
UP
0.688
45.6
17.04
0.5
0.344
#000 (S)
0.608
65.4
25.29
0.7
0.43
#090
0.718
71.6
23.71
0.7
0.50
Up/down
1.264
64.2
30.15
0.35
0.44
#090
0.239
14.9
6.17
2.0
0.48
#360 (S)
0.168
17.6
4.79
2.0
0.34
Up/down
0.091
8.4
3.29
2.0
0.18
#067 (S)
0.357
28.6
6.35
1.2
0.43
#337
0.325
22.3
4.59
1.2
0.39
Up/down
0.191
12
5.77
1.2
0.23
#000
0.43
23.6
7.52
1.6
0.69
#090 (S)
0.271
13.9
3.06
1.6
0.43
Up/Down
0.123
8.4
1.72
1.6
0.20
#147 (S)
0.169
11.6
4.25
2.0
0.34
#237
0.157
18.6
7.95
2.0
0.31
Up/Down
0.212
6.8
3.29
2.0
0.42
(1)
RRUP = Closest distance to rupture plane; RJB = Closest Horizontal Distance to rupture plane
(2)
(S) = horizontal component selected to be used in the dynamic time-history analyses
(3)
PGA, PGV, PGD are for original (not scaled) earthquake records
VERSAT-2D BACKGROUND
VERSAT-2D (Wutec Geotechnical International 2001) is a
2-dimensional finite element program that is used to
conduct static and dynamic stress and deformation analyses
of earth structures subjected to base excitation or to dynamic
loads at specified locations (Wu 2001; Wu and Chan 2002;
Wu et al. 2006). The program includes a non-linear
hyperbolic model to simulate the hysteresis response of soil
under cyclic loads. Excess pore water pressures caused by
cyclic loads, if applicable, can also be computed. Large
ground displacements caused by excess earthquake loading
are calculated using the updated Lagrangian analysis.
Structural beam elements and bar elements are used for
modeling soil-structure interaction.
Earthquake firm-ground accelerations are applied at the base
of the model, and displacements relative to the model base
are computed. Inertial forces on the soil mass caused by
base motions are computed using Newtons law, and base
accelerations are used directly in the equations of motions.
The equations of motions describing the incremental
dynamic force equilibrium are given as:
[ M ]{
d 2
d
} [ C ]{
} [ K ]{ } { P }
dt 2
dt
Where
(1)
[M]
[C]
[K]
[]
[d/dt]
[d2/dt2]
[P]
= mass matrices
= viscous damping matrices
= tangent stiffness matrices
= incremental displacement matrices
= incremental velocity matrices
= incremental acceleration matrices
= incremental external load matrices.
m'
Pa
m'
Pa
)m
)n
(2)
(3)
Where
Pa
Kb
Kg
m, n
m'
The relationship between the shear stress, xy, and the shear
strain, , for the initial loading condition is modelled to be
nonlinear and hyperbolic as follows:
xy
Gmax
1 Gmax / ult | |
(4)
Where
ult
Gmax
sand parameters:
unit weight: 19 kN/m3
stiffness: Kg = 868, m=0.5
strength: c=0 = 32
30
0.0 TO 5.5 SEC
10
10
0
-10
-10
-20
AT 3.1 M DEPTH
ELEM 1483
L32_2475_GAZ
-30
-0.2
-20
-0.05
0
0.05
SHEAR STRAIN (%)
-0.1
0.1
10
12
14
-40
-60
-80
1.1 m depth
2.1 m depth
3.1 m depth
4.1 m depth
TIME (SEC)
8
-20
-100
60
50
%bld
40
%teresa
%tabas
%tcu
%gaz
30
%lul
%gil
%cpe
20
1.0
2.0
20
30
40
-50
-150
-200
horizontal backfills,
static Ko=0.47
%bld, max. Koe=1.14
%teresa, max.
Koe=1.18
%tabas, max.
Koe=1.21
%tcu, max. Koe=1.06
%gaz, max. Koe=1.00
3.0
4.0
10
10
0
Horizontal stresses (kPa)
20
(5)
50
5.0
TIME (SEC)
-50
-150
-200
0.0
Horizontal backfills,
static K0=0.47
%bld, max. Koe=1.7
1.0
3.0
4.0
The finite element model is same for Case 1 and Case 2, and
it consists of 7487 soil elements. The left side boundary is
fixed to the model base in both vertical and horizontal
motions, while the right side boundary is free in vertical
motion. The slope is 25 m high, and the slope crest is 50 m
from the top of the wall (Fig. 11).
2.0
5.0
Ground Displacements
-50
-150
-200
0.0
horizontal backfills,
static Ko =0.47
%bld, max. Koe=2.1
1.0
2.0
3.0
4.0
5.0
2.50
2.00
1.50
Wood (1973) static
1.00
VERSAT-2D
nonlinear analyses
Series4
0.00
0
0.2
0.4
0.6
0.8
PRE-SHAKING
GROUND SURFACE
RIGID
WALL
-100
-300
0.40
0.0
0.00
-0.40
X - ACCELERATION
-0.80
X - DISPLACEMENT
-1.20
10
20
30
TIME (SEC)
40
1.0
2.0
3.0
50
5.0
I D #: S27phi32_MCE_teresa
-100
-300
-400
0.0
1.0
2.0
3.0
4.0
5.0
0.0
sloped backfills,
static Ko=1.5
%bld, max.
Koe=2.8
%teresa, max.
Koe=2.7
%tabas, max.
Koe=2.8
%tcu, max.
Koe=2.9
%gaz, max.
Koe=2.8
%lul, max.
Koe=2.8
%gil, max.
Koe=2.8
%cpe, max.
Koe=2.9
average curve,
Koe=2.8
sloped backfills,
static Ko=1.5
%bld, max.
Koe=3.7
%teresa, max.
Koe=3.6
%tabas, max.
Koe=3.9
%tcu, max.
Koe=4.2
%gaz, max.
Koe=3.5
%lul, max.
Koe=3.8
%gil, max.
Koe=3.6
%cpe, max.
Koe=3.9
average curve,
Koe=3.8
4.0
-1.60
-400
1.0
2.0
3.0
4.0
5.0
-100
-300
-400
sloped backfills,
static Ko=1.5
%bld, max.
Koe=4.7
%teresa, max.
Koe=4.8
%tabas, max.
Koe=4.8
%tcu, max.
K0e=4.8
%gaz, max.
Koe= 4.2
%lul, max.
Koe=4.8
%gil, max.
Koe=4.5
%cpe, max.
Koe= 5.3
average curve,
Koe= 4.8
K0E for the average curves are 2.6, 3.8 and 5.0 for ground
motions with 0.26g, 0.48 g and 0.71g, respectively.
-400
sloped backfills,
static Ko=1.1
%bld, max.
Koe=2.7
%teresa, max.
Koe=2.4
%tabas, max.
Koe=2.6
%tcu, max.
Koe=2.7
%gaz, max.
Koe=2.5
%lul, max.
Koe=2.4
%gil, max.
Koe=2.8
%cpe, max.
Koe=2.8
average curve,
Koe=2.6
1.0
2.0
3.0
-100
-300
0.0
-400
sloped backfills,
static Ko=1.1
%bld, max.
Koe=4.0
%teresa, max.
Koe=3.7
%tabas, max.
Koe=3.8
%tcu, max.
Koe=3.8
%gaz, max.
Koe=3.8
%lul, max.
Koe=3.4
%gil, max.
Koe=4.3
%cpe, max.
Koe=3.9
average curve,
Koe=3.8
1.0
2.0
3.0
4.0
5.0
-100
-300
-400
0.0
1.0
2.0
3.0
4.0
-300
5.0
4.0
-100
0.0
Depth from top of wall (m)
K0E for the average curves are 2.8, 3.8 and 4.8 for ground
motions with 0.26g, 0.48 g and 0.71g, respectively.
5.0
sloped backfills,
static Ko=1.1
%bld, max.
Koe=5.1
%teresa, max.
Koe=4.8
%tabas, max.
Koe=5.1
%tcu, max.
Koe=5.1
%gaz, max.
Koe=4.8
%lul, max.
Koe=4.6
%gil, max.
Koe=5.3
%cpe, max.
Koe=5.1
average curve,
Koe=5.0
Table 2 Ground Displacements (m) at the Slope Crest as Estimated from VERSAT-2D Nonlinear Dynamic Analyses
1/2475
1/475
1/2475
1/10,000
1/475
X-LATER Y-DOWN X-LATER Y-DOWN X-LATER Y-DOWN X-LATER Y-DOWN X-LATER Y-DOWN X-LATER Y-DOWN
%bld
1.40
0.84
0.75
0.50
0.22
0.15
0.43
0.16
0.18
0.07
0.05
0.02
%teresa
1.53
1.00
0.91
0.62
0.30
0.23
0.54
0.29
0.17
0.10
0.03
0.01
%tabas
1.54
0.91
0.96
0.64
0.33
0.24
0.63
0.29
0.25
0.12
0.06
0.02
%tcu
3.70
2.28
2.12
1.48
1.00
0.79
1.20
0.47
0.50
0.20
0.10
0.04
%gaz
1.29
0.82
0.70
0.44
0.14
0.10
0.51
0.23
0.21
0.10
0.04
0.02
%lul
1.72
1.02
0.97
0.65
0.36
0.25
0.53
0.24
0.16
0.07
0.03
0.01
%gil
1.46
1.13
0.88
0.70
0.27
0.21
0.42
0.20
0.19
0.08
0.04
0.02
%cpe
2.52
1.58
1.58
1.03
0.52
0.38
1.03
0.40
0.40
0.15
0.06
0.02
Average
1.90
1.20
1.11
0.76
0.39
0.29
0.66
0.29
0.26
0.11
0.05
0.02
passive
failure line
= 40
passive
failure line
= 32
0.26g
0.71g
0.48
0.48g
0.26
3
0.71g
0.0
Horizontal backfills
27 Sloped Backfills, = 32
27 Sloped Backfills, = 40
Fig. 19. Normalized soil pressure diagram from VERSAT-2D nonlinear analyses of a 5 m high wall with horizontal and 27
sloped backfills
SUMMARY
Seismic soil pressures on rigid walls retaining 2H:1V (27)
sloped backfills have been computed using a total of eight
acceleration time histories recorded in past large
earthquakes. These records were selected and linearly
scaled to three levels of ground motions with a nominal
PGA of 0.26g, 0.48g and 0.71g. Nonlinear time-history
analyses were conducted using the computer program
VERSAT-2D which uses a hyperbolic stress strain model
to simulate the hysteresis response of soil under cyclic
loads. Soil pressure diagrams are shown in Fig. 19 for
horizontal backfills (=32), and for sloped backfills with
loose sand (=32) and dense sand (=40) under the three
levels of ground motions.
A soil pressure coefficient, K0E, has been introduced to
represent the total static and seismic pressures on a rigid (or
non-yielding) wall. It is found that K0E varies from 1.1, 1.7
to 2.2 for horizontal backfills under ground motions of
0.26g, 0.48g and 0.71g, respectively. K0E (average for loose
sand backfill and dense sand backfill) increases to 2.7, 3.8
and 4.9 for the 27 sloped backfills under the same three
levels of ground motions, respectively.
REFERENCES
Bommer, J.J. and A.B. Acevedo [2004]. The Use of Real
Earthquake Accelerograms as Input to Dynamic Analysis,
Journal of Earthquake Engineering, Vol. 8, Special issue 1,
pp. 43-91.
Finn, W.D.L, K.W. Lee and G.R. Martin [1977]. An
Effective Stress Model for Liquefaction, ASCE Journal of
Geotechnical Engineering, Vol. 103, pp. 517-533.
Adams, J. and G. Atkinson [2003]. Development of
Seismic Hazard Maps for the Proposed 2005 Edition of the
National Building Code of Canada, Canadian Journal of
Civil Engineering, Vol. 30, pp. 255-271.
Wood, J. H. [1973]. Earthquake-induced Soil pressures on
Structures, Ph.D thesis, the California Institute of
Technology, Pasadena, California, USA
Wu, G. and S. Chan [2002]. Design of the Russ Baker Way
Overpass on Liquefiable Sand - Vancouver Airport
10