You are on page 1of 11

Tenth U.S.

National Conference on Earthquake Engineering


Frontiers of Earthquake Engineering
July 21-25, 2014
10NCEE Anchorage, Alaska

REDUCED PASSIVE FORCE-DEFLECTION


BEHAVIOR FOR SKEWED BRIDGE
ABUTMENTS
K.M. Rollins 1, K. Palmer2 and A. Fredrickson3

ABSTRACT

A series of large-scale passive force-deflection tests were performed on a simulated abutment to


investigate the effect of skew angle on performance. Tests were conducted at skew angles of 0,
30 and 45 with a backwall that was 3.35 m wide and a backfill that was 0.9 m deep and 6.7 m
wide. Test results indicate that the passive force decreases significantly as the abutment skew
angle increases to 30or 45 relative to non-skewed walls. These results confirm observations
from previous lab-scale tests and large-scale field tests where the abutment width to height ratio
was about one-half of that in this series of tests. The results also indicate that the reduced passive
force can be accounted for using a simple adjustment factor that is a function of skew angle. The
initial stiffness of the passive force-deflection curves was relatively unaffected by skew angle.
In addition, the displacement required to develop the peak passive force was about 3% of the
wall height regardless of skew angle.

1
Professor, Dept. of Civil & Env. Engineering, Brigham Young Univ., Provo, UT 84602, rollinsk@byu.edu
2
Staff Geotech. Engr., Kleinfelder, Inc., 6380 Polaris Avenue, Las Vegas, Nevada 89118
3
Grad. Research Asst., Dept. of Civil & Env. Engineering, Brigham Young Univ., Provo, UT 84602

Rollins, K.M., Palmer, K., Fredrickson, A. Reduced Passive Force-Deflection Behavior


for Skewed Bridge Abutments, Proceedings of the 10th National Conference in Earthquake Engineering, Earthquake
Engineering Research Institute, Anchorage, AK, 2014.
Reduced Passive Force-Deflection Behavior
for Skewed Bridge Abutments

K.M. Rollins 1, K. Palmer2, and A. Fredrickson3

ABSTRACT

A series of large-scale passive force-deflection tests were performed on a simulated abutment to


investigate the effect of skew angle on performance. Tests were conducted at skew angles of 0,
30 and 45 with a backwall that was 3.35 m wide and a backfill that was 0.9 m deep and 6.7 m
wide. Test results indicate that the passive force decreases significantly as the abutment skew
angle increases to 30or 45 relative to non-skewed walls. These results confirm observations
from previous lab-scale tests and large-scale field tests where the abutment width to height ratio
was about one-half of that in this series of tests. The results also indicate that the reduced passive
force can be accounted for using a simple adjustment factor that is a function of skew angle. The
initial stiffness of the passive force-deflection curves was relatively unaffected by skew angle. In
addition, the displacement required to develop the peak passive force was about 3% of the wall
height regardless of skew angle.

Introduction

Passive force-deflection behavior must be considered in bridge design to ensure adequate


resistance to both seismically and thermally induced forces. Several researchers have conducted
large-scale field studies to investigate passive force-deflection behavior with densely compacted
granular backfills for non-skewed abutments [1, 2, 3, 4]. The results from these field studies
show that the ultimate passive force may be reasonably predicted using the log-spiral method and
that it develops at displacements of approximately 3% to 4% of the wall height [1, 3]. Owing to
geometrical constraints, bridge abutments are increasingly constructed at a skew to underlying
roadways. The national FHWA bridge database indicates that about 40% of 600,000 bridges in
the US are skewed. Unfortunately, skewed bridges experienced twice the damage rate of non-
skewed bridges during the 2010 Chilean earthquake [5].

Current codes and practices do not distinguish between skewed and non-skewed bridge abutment
geometries in computing passive force. However, in recent years, numerical models [6] and
plane-strain laboratory tests [7] suggested a significant reduction in passive force for skewed
bridge abutments. Using data from these two studies, Rollins and Jessee [7] proposed the
reduction factor, Rskew, given in Eq. 1 as a function of skew angle, (degrees). This proposed

1
Professor, Dept. of Civil & Env. Engineering, Brigham Young Univ., Provo, Utah 84602, rollinsk@byu.edu
2
Staff Geotech. Engr., Kleinfelder, Inc., 6380 Polaris Avenue, Las Vegas, Nevada 89118
3
Graduate Research Asst., Dept. of Civil & Env. Engineering, Brigham Young Univ., Provo, Utah 84602

Rollins, K.M., Palmer, K., and Fredrickson, A., Reduced Passive Force-Deflection Behavior for Skewed Bridge
Abutments. Proceedings of the 10th National Conference in Earthquake Engineering, Earthquake Engineering
Research Institute, Anchorage, AK, 2014.
relationship effectively reduces the non-skewed passive force, Pp, for skewed bridge abutments
relative to a non-skewed abutment of equivalent width. Using Eq. 2, this reduction factor can be
used to obtain the reduced passive force, Pp-skew, for skewed bridge abutments.

= 8.0 105 2 0.018 + 1.0 (1)

= (2)

Because Eq. 1 was based only on small-scale tests and computer models, the need for additional
large-scale testing was apparent. To understand the behavior of skewed abutments better, a
series of large-scale tests have been performed with abutment skew angles of 0, 15, and 30 using
an existing pile cap [3.35 m (11ft) wide by 4.57 m (15 ft) long by 1.68 m (5.5 ft) high] and a
densely compacted sand backfill that was 1.68 m (5.5 ft) thick. Marsh et al [8] report results of
passive force tests with wingwalls perpendicular to the sides of the test abutment while Franke et
al [9] report results from tests using Mechanically Stabilized Earth (MSE) wingwalls running
parallel to the sides of the test abutment. These large scale field tests confirmed the results from
the laboratory scale tests and showed a significant reduction in passive force (about 50%) at a
30 skew angle. Although these results are compelling, it has been noted that the abutment
width to height ratio in both the field and lab tests was only 2.0 whereas this ratio would be
substantially higher for many abutments in the field. To evaluate the effect of the width to height
ratio on the reduction in passive resistance, another series of tests were performed with the same
pile cap (3.35 m wide) and a 0.90 m high backfill. For these tests, the width to height ratio was
nearly twice as high (3.7). In addition, a test was performed with a skew of 45 for which no
field test data was previously available. This paper describes this testing program and the results
relative to previous tests.

Background

As a bridge deflects longitudinally into the soil at an abutment, different interaction forces
develop between the bridge structure and the abutment for a bridge with skewed geometry as
shown in the plan view drawing in Figure 1. At the soil-abutment interface, the longitudinal
force, PL, acts parallel to the bridge structure. This driving force may be broken into two
components: a component acting normal, PLcos, (see Eq. 3) and a component acting parallel,
PT, (see Eq. 4) to the abutment backwall face. To ensure stability, of the bridge structure, the
passive soil resistance, Pp, and shear resistance, PR, (see Eq. 5) provided by the soil backfill and
bridge abutment must resist this driving force. It should be noted from Eq. 5 that the shear
resistance is directly related to the passive resistance so that a decrease in passive resistance
leads to a proportional decrease in shear resistance. Force and moment equilibrium must be
maintained as noted by Eqs. 6 and 7 suggested by Burke [10]. In these equations, is the
backwall skew angle, c is soil cohesion, A is backwall area, is the abutment backwall-soil
interface friction angle, Fs is the factor of safety, and L is the length of the bridge. During an
earthquake, a transverse force will also develop in addition to the longitudinal force.
Components of the transverse force will increase the shear force acting parallel to the abutment
and decrease the normal force on the wall. Both of these forces would decrease the stability of
the wall with respect to sliding and rotation.
Figure 1. Plan view of force interaction between bridge structure and skewed abutment.

Pp = PL cos (3)

PT = PL sin (4)

PR = cA + Pp tan (5)

cA + Pp tan
PL sin (6)
Fs

cA + Pp tan L cos
Pp L sin (7)
Fs

Test Layout

As shown in Fig. 2, the backfill zone was approximately 6.7 m (22 ft) wide and extended 7.3 m
(24 ft) longitudinally from the backwall (pile-cap) face for the 0 skew test. However, since an
additional concrete wedge was affixed to the existing pile cap, the backfill extended
approximately 7.3 m (24 ft) longitudinally from the acute corner of the skewed abutment.
Additionally, since the native soil was significantly stronger than the backfill material, the
bottom of the test pit was placed approximately 0.30 m (1.0 ft) below the bottom of the pile-cap
extending 3.1 m (10 ft) longitudinally from the backwall face for both tests. Beyond this region,
however, the test pit tapered, becoming level with the bottom of the pile cap. This prevented any
interference from the native soil in the development of the suspected, log-spiral failure surface.

In addition to attaching the concrete wedge to the pile-cap face for the 30 and 45 skew tests,
the concrete wedge was placed atop a set of rollers resting on a small platform beneath the
concrete wedge. This minimized friction between the concrete wedge and underlying soil and it
also ensured that lateral resistance was due only to the passive soil resistance provided by the
backfill material and the piles beneath the existing pile cap. Preliminary testing with no backfill
showed a negligible increase in lateral resistance as a result of the additional concrete wedge.
7.3 m

4.6m 1.7m 1.2 m

Backfill
Sand
6.7m

3.35-m x 4.57-m x 1.78-m


Concrete Pile Cap

3.35-m x 4.57-m x 1.78-m


Concrete Pile Cap

Backfill Sand 0.9m 1.2m

Figure 2. Plan and profile views of the test abutment, backfill, and loading system.

Backfill Characterization

Backfill soil used for this series of tests consisted of poorly graded sand (SP type soil by the
Unified Soil Classification System or A-1-b type soil by the AASHTO Classification System) at
a moisture content of 7 to 9%. The maximum dry density and optimum moisture content were
determined to be 17.5 kN/m3 (111.5 pcf) and 7.1%, respectively, in accordance with the
modified Proctor compaction test (ASTM D1557). Compaction, however, was found to be much
easier in the field when the moisture content was approximately 9%. Onsite, the target
compaction level was 95% of the modified proctor maximum. This was accomplished using a
vibratory, smooth-drum, roller compactor to compact 6-in (15 cm) lifts of backfill material along
with a plate compactor near the wall face. Throughout testing, a calibrated nuclear density gauge
was used to ensure proper compaction and moisture content. The average compaction of the 45
test was slightly higher thant eh other test but with normal variation for field conditions. A
summary of compaction and water content is provided in Table 1 along with an estimation of
relative density using a correlation by Lee and Singh, [11]. From lab tests, the sand used for
testing had a drained friction angle, , of approximately 41 and an apparent cohesion, c, of
approximately 4.79 kPa (100 psf).

Table 1. Average dry unit weight, relative compaction and relative density for backfills.

Parameter 0 Test 30 Test 45 Test


Avg. Dry Unit weight, d 16.8 kN/m3 16.9 kN/m3 17.2 kN/m3
(107 lbs/ft3) (107.5 lbs/ft3) (109.2 lbs/ft3)
Avg. Moisture Content 9.2% 9.5% 7.1%
Relative Compaction, Rc 96.0% 96.4% 97.9%
Relative Density, Dr 81% 82% 90%

Instrumentation

Hydraulic actuators, fitted with pressure transducers, were used to apply and measure
longitudinal forces during testing. Four string potentiometers attached to the back side of the pile
cap measured longitudinal pile-cap displacement relative to an independent reference beam.
Longitudinal displacement of the pile cap was confirmed using independent measurements from
inclinometer and shape array readings taken at both the north and south ends of the pile cap.

Before testing, a grid of 0.61 m (2.0 ft) squares, refined to a grid of 0.30 m (1.0 ft) squares near
the backwall, was painted on the surface of the backfill. To determine vertical heave, the relative
elevation at grid-intersection points was measured using a total stations survey before and after
each test. This grid also helped to identify the location of surface cracks and relative horizontal
displacement. In addition, the horizontal displacement of the backfill was measured with a total
station survey for two of the tests.

Additionally, pressure was measured normal to the backwall face using six pressure plates
located approximately 0.56m (22 in) up from the base of the wedge and 0.546 m (21.5 in) center
to center across the width of the wedge for the 30 test only. Because some researchers have
suggested a more triangular pressure distribution along the backwall face for skewed geometries
[6, 12], this was done to understand the pressure distribution along the backwall face for skewed
abutments.

Test Results

Passive Force-Deflection

Prior to conducting tests with backfill in place, lateral load tests were performed on the pile cap
without backfill in place to determine the baseline force-deflection relationship produced by
the piles and cap block alone. After placement of the backfill, the longitudinal force measured by
the actuators was reduced by the lateral resistance provided by the piles and cap using the
baseline curves for each test. For the 30 and 45 skew tests, the longitudinal force was then
adjusted to be normal to the backwall face using Eq. 3 to obtain the passive force.

The passive-force deflection curves for the 0, 30, and 45 skew tests are shown in Figure 3. A
comparison of the maximum passive force for the tests on the skewed abutments shows a
significant reduction in passive force as the skew angle increases. For example, the reduction in
passive resistance is more than 50% for the abutment at the 30 skew angle relative to the non-
skewed abutment. Despite this fact, the soil stiffness appears to have remained largely unaffected
by the skew angle of the abutment at small displacements less than about 0.2 inch. However, the
load-deflection curves begin to diverge and show a reduction in stiffness as the ultimate passive
resistance is approached, in comparison to the 0 skew test. For the 0 skew abutment, the
maximum passive force was obtained with a longitudinal deflection of approximately 3.5% of
the wall height; however, for the 30 and 45 skew tests the peak resistance developed at a
somewhat lower longitudinal deflection of 2.75% of the wall height. Nevertheless, the
difference in displacement to reach the peak force is relatively minor. These results are generally
consistent with other large-scale passive force-deflections tests on skewed abutments conducted
previously [8, 9].

Pile Cap Displacement [in]


0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5
900 200
800 180
700 160

Passive Force [kips]


Passive Force [kN]

600 140
0 Degree Skew
30 Degree Skew 120
500
45 Degree Skew 100
400
80
300
60
200 40
100 20
0 0
0 2 4 6 8 10
Pile Cap Displacement [cm]

Figure 3. Passive force-deflection curves for 0, 30, and 45 skew tests with densely compacted
granular backfill.
The post-peak passive force remains relatively constant with displacements up to 9% of the wall
height for the 0 and 30 skew tests, but for the 45 skew test there is a substantial decrease in
the passive force beyond the peak. This behavior is likely a result of the slightly higher backfill
density for the 45 test (see Table 1). With respect to post-peak behavior, a higher backfill
density causes the soil to dilate more during shearing resulting in a decreased soil density. This
reduction in density, in turn, reduces the post-peak shear strength of the soil.

The passive force reduction factor versus skew angle curve proposed by Rollins and Jessee [7] is
presented in Figure 4 along with back-calculated reduction factors from lab tests [7] and
numerical analysis [6]. The measured longitudinal force obtained from each test was used to
compute the passive force using Eq. 3. The passive force for a given skew angle was then
divided by the passive force for the zero skew case to determine the passive force reduction
factor. The results from the tests in this study are also plotted in Figure 4 and the results are in
very good agreement with the proposed curve. The back-calculated reduction factors for the 30
and 45 skew tests are 0.45 and 0.34, respectively, while Eq. 1 predicts values of 0.53 and 0.35,
respectively. If the density of the backfill for the 45 skew test had been a little higher than that
for the other backfill tests, it is conceiable, that the reduction factor for the 45 might have been a
little lower; perhaps as much as 10% lower. Nevertheless, the consistency of these results, with
a backwall width to height (W/H) ratio of 3.7 relative to the previous results with a W/H ratio of
2.0 suggest that the W/H ratio does not strongly affect the observed reduction ratio. If anything,
the reduction in passive force for the tests with the higher W/H ratio is somewhat greater than
was observed for the lab tests with the lower W/H ratio.

1.00
0.90 Lab Tests (Rollins and Jessee, 2012)
0.80 Numerical Analysis (Shamsabadi et al., 2006)
Reduction Factor, Rskew

Field Tests (This Study)


0.70
Proposed Reduction Line
0.60
0.50
0.40
0.30
0.20
Rskew = 8x10-52 - 0.018 + 1
0.10
0.00
0 10 20 30 40 50 60 70 80 90
Skew Angle [degrees]

Figure 4. Reduction factor, Rskew, plotted versus skew angle with proposed reduction curve.

Transverse Abutment Displacement and Rotation

Although deflections of both actuators were kept relatively consistent throughout the test to
displace the pile cap longitudinally, rotation and transverse deflection still occurred. For the 0,
30 and 45 skew tests, the pile cap ultimately shifted to the west (in the direction of the skew)
by 0.04, 0.10 and 0.13 inch, respectively. Thus, the transverse displacement progressively
increased as the skew angle increased. In all cases, the pile cap ultimately rotated
counterclockwise with rotations of 0.02, 0.03 and 0.003 for the 0, 30 and 45 skew tests,
respectively. Therefore, rotations were small and unaffected by skew angle.

Backwall Pressure

As previously indicated, for the 30 skew test, the passive pressure normal to the backwall face
was measured directly using pressure plates. As shown in Figure 5, the pressures associated with
deflection increments up to 2.54 cm (1.0 in) were relatively uniform across the width of the cap.
However, at displacements just beyond those necessary to mobilize peak resistance, the pressure
distribution becomes quite non-uniform. The highest pressures develop on the obtuse corner of
the abutment with the lowest presssure near the center of the wall. This increase appears to be
associated with a small counterclockwise rotation of the cap at higher displacements. Franke et
al [9] also reported a pressure distribution with a lower value near the center and higher values at
the edges for a test abutment with MSE wingwalls. However, the wall did not rotate appreciably
and the highest pressures developed on the acute corner of the wall. The measured distribution of
pressures is generally consistent with elastic theory which also predicts that the pressure at the
edges of a wall or footing will be higher than near the center [13].

Distance from West Edge of Pile Cap [ft]


0 2 4 6 8 10
180

160

140

120
Pressure [kPa]

100

80

60

40

20

0
0 0.5 1 1.5 2 2.5 3
Distance from West Edge of Pile Cap [m]

0.64 cm 1.27 cm 2.54 cm 3.81 cm


5.08 cm 6.35 cm 7.62 cm 8.89 cm

Figure 5. Pressure measured normal to the backwall face for the 30 skew test versus transverse
distance from the west corner of the skew.
Vertical Displacement and Crack Patterns

Backfill heave was located primarily in the region associated with a typical log-spiral failure
surface [4.9 to 5.5 m (16 to 18 ft)]. Contour maps of the backfill heave for the 0 and 30 degree
skew tests were created as shown in Figure 6. Generally, the greatest vertical heave was located
between 0.61 and 3.05 m (2.0 and 10 ft) perpendicularly from the backwall face. The maximum
vertical displacement was about 50 mm (2.0 in) which represents a 5.5% heave relative to the 3.0
ft backfill height. Although the crack and heave patterns are quite symmetric for the 0 skew
test, for the 30 degree test the heave and cracks are skewed towards the acute side of the
abutment wall and are more oblong in shape.

6.7m
6.7m

7.3 m

7.3 m

1.22m

1.22m 3.35 m 1.22m

3.35m
1.22m

Figure 6. Vertical backfill displacement (heave) contours in inches for 0 (left) and 30 (right)
skews. [Note: 1 in. = 25.4 mm and gridlines are spaced at 0.61 m (2 ft) intervals.]

Conclusions

1. Results of this large-scale field study confirm that passive force decreases significantly
as the abutment skew angle increases to 30 and then 45 relative to non-skewed walls
as observed in small-scale lab test results [7] and numerical models [6] for densely
compacted granular backfill.
2. The proposed passive-force reduction curve proposed by Rollins and Jessee [7] appears
to provide a reasonable estimate of the measured passive force reduction in densely
compacted granular material despite the larger backwall width to height ratio of 3.7 used
in these tests relative to the ratio of 2.0 used in previous testing. These results suggest
that W/H ratios do not strongly influence the reduction in passive force.
3. Despite the reduction in passive force attributable to the increase in skew angle, the
initial stiffness of the passive force-deflection curves was relatively unaffected by skew
angle. In addition, the displacement required to develop the peak passive force was
about 3% of the wall height for the 0, 30 and 45 degree skew tests.
4. Backwall pressure distribution was nonlinear. The largest pressures were located near
the edges and lowest pressures were located near the center of the backwall face.
Acknowledgements
Funding for this research was provided by FHWA pooled fund study TPF-5(264), which was
supported by Departments of Transportation from the states of California, Minnesota, Montana,
New York, Oregon and Utah along with FHWA. Utah served as the lead agency, with David
Stevens as the project manager. This support is gratefully acknowledged; however, the
opinions, conclusions and recommendations in this paper do not necessarily represent those of
the sponsoring organizations.

References

1. Cole, R., and Rollins, K. "Passive Earth Pressure Mobilization during Cyclic Loading." Journal of Geotechnical
and Geoenvironmental Engineering, 2006; 132(9): 1154-1164.
2. Duncan, J., and Mokwa, R. "Passive Earth Pressures: Theories and Tests." Journal of Geotechnical and
Geoenvironmental Engineering, 2001; 127(3): 248-257.
3. Lemnitzer, A., Ahlberg, E., Nigbor, R., Shamsabadi, A., Wallace, J., and Stewart, J. "Lateral Performance of
Full-Scale Bridge Abutment Wall with Granular Backfill." Journal of Geotechnical and
Geoenvironmental Engineering, 2009; 135(4): 506-514.
4. Rollins, K., and Sparks, A. "Lateral Resistance of Full-Scale Pile Cap with Gravel Backfill." Journal of
Geotechnical and Geoenvironmental Engineering, 2002; 128(9), 711-723.
5. Toro, F., Rubilar, F., Hube, M.A., Santa-Maria, H., and Cabrera, T. "Statistical Analysis of Underpasses During
2010 Chile Earthquake". Procs. 7th National Seismic Conf. on Bridges and Highways, FHWA, 2013. 12 p.
6. Shamsabadi, A., Kapuskar, M., a Skend Zand, A. "Three-Dimensional Nonlinear Finite-Element Soil-Abutment
Structure Interaction Model for Skewed Bridges." Fifth National Seismic Conference on Bridges &
Highways, 2006; San Francisco, CA, 14.
7. Rollins, K. M., and Jessee, S. J. "Passive Force-Deflection Curves for Skewed Abutments." Journal of Bridge
Engineering, 2013; 18 (10): 1086-1094
8. Marsh, A.K., Rollins, K.M., Franke, B., Smith, K., and Palmer, K. "Passive Force-Deflection Behavior for Zero
and Thirty Degree Skewed Abutments", Transportation Research Record, FHWA, 2013.
9. Franke, B.W., Rollins, K.M., Marsh, A.K., Smith, J., Palmer, K. Passive Force-Deflection Behavior for Skewed
Abutments with MSE Wingwalls", Procs. 7th National Seismic Conf. on Bridges and Highways, FHWA,
2013. 12 p.
10. Burke, M. P., Jr. "Semi-Integral Bridges: Movements and Forces." Transportation Research Record: Journal
of the Transportation Research Board, 1994; 1460, 1-7.
11. Lee, K. L., and Singh, A. "Relative Density and Relative Compaction." Journal of Soil Mechanics and
Foundations Design, 1971; 97(7): 1049-1052.
12. Sandford, T. C., and Elgaaly, M. "Skew Effects on Backfill Pressures at Frame Bridge Abutments."
Transportation Research Record: Journal of the Transportation Research Board, 1993; 1-11
13. Hegger, J., Ricker, M., Ulke, B., & Ziegler, M Investigations on the Punching Behaviour of Reinforced
Concrete Footings. Engineering Structures, 2007; 29(9), 2233 2241.

You might also like