Proceedings of the Tenth Pacific Conference on Earthquake Engineering

Building an Earthquake-Resilient Pacific

6-8 November 2015, Sydney, Australia

Low damage seismic design of structure using footing uplift

X. Qin & N. Chouw

Department of Civil and Environmental Engineering, the University of Auckland, New Zealand.

ABSTRACT: In conventional seismic design of structures, plastic hinge development

was permitted. In recent earthquake events, it became clear that the damage to structures
due to plastic hinge development can be too costly. Thus damage avoidance seismic
design of structures has been proposed as one of the major design approaches for future
buildings. Allowing a structure to uplift is one of the practices to significantly reduce or
even eliminate the development of plastic hinges in the structure. However, the challenge
of this practice is to accurately calculate the response of structure with uplift. Previous
studies have developed numerical formulas for this propose. This paper presents an
experimental validation of these numerical formulas. Shake table test was conducted on a
model. The response of structure obtained in the experiment was compared against that
obtained using existing numerical formula. An approach to calculate the response of
structure with uplift was developed.

1 INTRODUCTION
1.1 Structure with uplift
The system considered is shown in Figure 1, it is a SDOF model with a base width 2b, a height h and
the dynamic properties (mass m, lateral stiffness k and damping c). It is assumed that the supporting
ground is rigid, i.e. only the positive vertical displacement at the footing can develop (Figure 1(b)).
During the base excitation, if uplift is not permitted or initiated (Figure 1(a)), a fixed base condition
can be assumed. The horizontal displacement at the top of structure relative to the column footing (u)
due to base excitation can be calculated by using an equation of motion. On the other hand, when the
inertia force in the structure produces a moment at o or o (Figure 1(b)) greater than that due to
gravity, rotation about o or o can occur, respectively. Footing vertical displacement v in Figure 1(b)
other than the rotation pivot will develop. The total horizontal relative displacement in the structure is
the combined movement due to the rocking motion (x) and structural deflection (u). Because the
structural deflection (u) determines the seismic performance of the upliftable structure, it is necessary
to be quantified for design.

u x u

o
+v o'
b
(a) (b)

Figure 1: Effect of uplift on the bending moment

In this study, shake tablet test was conducted to validate an empirical formula for estimating the
deformation of an upliftable structure. The formula was originally developed by Psycharis (1981,
1991). For the test, a SDOF model was considered. Bending moment at the column was measured. It
was found that the original formula will underestimate the bending moment in the structure during an

Paper Number 188

earthquake. To improve the formula, free rocking test on the model was performed. Result suggested
that the behaviour of structure during uplift was dependent on the amplitude of footing vertical
displacement. The natural period of an equivalent SDOF model with uplift was determined. By
incorporating the natural period of upliftable model, the empirical formula can predict the bending
moment in the model
1.2 Previous studies
After the Valdivia earthquake in Chile in May 1960, a good seismic performance of several water
towers had been reported by Housner (1963). Structural uplift has been suggested as a possible
earthquake resistance solution for structures. Compared to the conventional approach, the stepping
action of structural footing on the supporting ground due to uplift can consumed part of the earthquake
energy. To understand the uplift behaviour, Psycharis and Jennings (1983) used Winkler and two-
spring foundations to simulate the uplift of a rigid structure. Result suggested that the rocking
frequency of a rigid structure is influenced by the amount of the uplift. Wang and Gould (1993) had
extended the analytical study of structural uplift by including sliding behaviour. Kodama and Chouw
(2002) investigated the effect of soil-foundation-structure interaction on upliftable structures. Hung et
al. (2008) performed a number of quasi-dynamic tests on concrete bridges with allowable uplift. This
investigation concluded that allowing structures to uplift could lead to a reduction of the maximum
deformation and forces activated in the structure. Consequently, the design strength and ductility
demand of a structure could be reduced. Kafle et al. (2011) had conducted a series of shake table test
to identify the peak displacement demand of a rigid structure with various geometrical characteristics.
Loo et al. (in print, 2012) and Ormeno et al. (in print) considered upliftable structures with slip-friction
connectors and fluid-structure interaction, respectively. Ali et al. (2013) investigated the effect of
uplift throughout the structure on the seismic performance of low-damage earthquake-resistant
structures. Qin et al. (2013) investigated the uplift behaviour of structure with soil nonlinearity and
structural plastic hinge. The beneficial effect of this nonlinear structure-foundation-soil interaction
(SFSI) on the response of structure and secondary structure was considered.
Uplift behaviour has been considered in a number of design guideline to control seismic response of
structure. FEMA 356 (2000) had proposed a guideline for rocking rigid structure with allowable uplift
based on a shake table test result (Priestley et al. (1978)). One of the remarkable structures designed
with rocking mechanism is the Rangitikei Railway Bridge built in New Zealand (Beck and Skinner,
1974). In the retrofit programme of the Lions Gate Bridge in Vancouver (Crippen, 2002), structural
uplift was implemented to improve the seismic resistance of the bridge. Although the beneficial effect
of structural uplift has been recognized and a design framework is available, structures with capability
to uplift are still very limited. Also, these works were conducted based on the assumption that the
structure is rigid. Only a very few work has provided an analytical frame work in calculating the
deformation in a flexible structure with allowable uplift e.g. Psycharis (1981 and 1991). Experimental
data has not been used to confirm the accuracy of the existing analytical model.

2 SHAKE TABLE EXPERIMENT

2.1 Experimental setup
Figure 2(a) shows the setup of the shake table experiment. A SDOF frame structure was considered
with a height and a width of 0.83 m and 0.40 m, respectively. The beams of the model were assumed
to be rigid and constructed using aluminium section. The columns of the model were constructed using
PVC and assumed massless. The mass at the top of the model were 29.7 kg. The fundamental period
of the model with an assumed fixed base was 0.34 s. The property of the SDOF model was obtained
and scaled from a six storey prototype structure described in the study by Qin and Chouw (2012).
Shake table test using stochastically simulated excitations based on Japanese Design Spectra (2000)
was conducted (Chouw and Hao, 2005). Three different ground excitations were considered in this
study. Figure 2(b) shows the acceleration (ag) time history of the applied excitation. Figure 2(c) shows
the spectrum acceleration (Spa) of the excitation with a damping ratio of 5%.

2
 0.2

0.1

ag (g)
0

-0.1

-0.2
(b) 0 5 10
Time (s)

0.6

Spa (g)
0.4

0.2
0.1 0.4 0.7 1
(a) (c) Period (s)

Figure 2: Shake table test. (a) Setup, (b) Excitation and (c) response spectrum

Two sets of shake table experiment were conducted to reveal the effect of structural uplift on the
seismic force development in the structure. To obtain the seismic force, strain gauge was attached at
the base of columns. Two Linear Voltage Differential Transformers (LVDTs) were placed at the edges
of the footing to measure the vertical displacement of footing when uplift was permitted (Figure 2(a)).
Sand paper was attached at the interface between the footing and the support to minimize sliding when
uplift occurs.
2.2 Experimental results
Figure 3 shows the time history of bending moment with and without uplift ability. With uplift the
bending moment is smaller. While the maximum bending moment in the upliftable model was 87.4
Nm, the maximum bending moment in the fixed base model was 94.6 Nm. The period of the structural
response increases with the time, if uplift was permitted. Figure 4 shows the time history of footing
rotation. It is found that the change of the period of the structural response occurred when uplift was
initiated.

100
Upliftable
BM (Nm)

50
0
-50
Fixed base
-100
4 6 8 10 12
Time (s)
Figure 3: Effect of uplift on the bending moment

3
 0.8

Rotation (o) 0.4

0
-0.4
-0.8
4 6 8 10 12
Time (s)
Figure 4: Time history of footing rotation during excitation

3 ESTIMATING THE SEISMIC FORCE IN STRUCTURE

When a fixed base was considered, the response of the structure can be calculated using the equation
of motion. When uplift is permitted, the horizontal top displacement of the structure relative to ground
is the combination of the horizontal displacement due to structural deformation (u) and footing
rotation.
Psycharis (1991) proposed an empirical formula to estimate the maximum deformation in an uplifting
structure. By assuming a small horizontal top displacement due to footing uplift, the response of
system with uplift was then assumed to be linear. Equation of motion that governs the lateral response
of upliftable structure was derived. The equation was used to conduct a parametric study to reveal the
influence of the natural frequency, damping ratio and slenderness of structure on the uplift behaviour.
The result obtained from numerical study was used to establish a set of empirical formula for
determining the normalized maximum horizontal displacement relative to the column footing
(Equation (1)).

1 / 2 1 / 3 35
u max u cr (1 0.73
2
e ) (1)

where ucr is the critical displacement in the structure for uplift to occur, is the slenderness
coefficient, is the damping ratio of fixed base condition, is the ratio between the maximum
horizontal displacement of the fixed base SDOF structure and ucr, and

Tn T
for n 1 and (2a)
To To
Tn
1 for 1 (2b)
To
bg
u cr (3)
h 2

where T is the natural period of the fixed-base SDOF structure and To is the period of the harmonic
excitation or Tmin of an arbitrary earthquake; h and b are the height and half of the base width of the
model, respectively; g is the gravitational acceleration.
The period of a harmonic excitation is constant. On the other hand, when an excitation with a range of
predominant periods are considered (e.g. earthquake), research in the past (Psycharis, 1991 and

4
Chopra and Yim, 1985) has confirmed that the minimum value (Tmin) of the predominant period range
can be used. In this work, Tmin of different excitations are obtained by examining the response
spectrum of the corresponding excitation. In general, Fourier spectrum can be used to find the Tmin.
However, a response spectrum is more common because it is generally available in most of seismic
design documents and thus considered herein.
Equation (1) suggests that the maximum normalized horizontal relative displacement in the model due
to the applied excitation was 1.21. Using the maximum normalized horizontal displacement, the
maximum bending moment can be calculated using Equation (4).
BM max u max k h (4)

Equations (1)-(4) suggests that the maximum bending moment in the model with uplift is 79.5 Nm.
Compare to the result obtained using shake table test (87.4 Nm), it is found that Equations (1)-(4) has
underestimated the maximum bending moment. As found in the time history of bending moment
obtained from the shake table test (Figure 2), the period of the structural response increased when
uplift was permitted. This observation shows that by calculating the response of upliftable structure
using the fundamental frequency of the structure with an assumed fixed base is inappropriate. It is
suggested that when applying Equations (1)-(4), the fundamental period of equivalent model with
uplift should be considered.

4 DETERMINATION OF THE SEISMIC FORCE IN STRUCTURE

To determine the fundament period of an equivalent structure with uplift, free rocking test was
conducted. The experimental procedure involved in giving an initial vertical displacement at one side
of the footing by inserting a rigid block between the footing and the support. The size of the block was
known and the model was tilted with an initial rotation. The block was removed suddenly to create a
free rocking motion of structure. Three different block sizes were utilized in this study (10.3 mm, 12.9
mm and 15 mm.) Figure 5(a) shows the time history of footing rotation during free rocking test when
10.3 mm block was used. Three significant cycles of footing rotation can be found. The response
period of structure with uplift was obtained by finding the time between the peaks of footing rotation
displacement.
0.6
4 1st cycle 2nd cycle

2
Rotation (o)

0.5
TR (s)

0
0.4
-2

0.3 -4
0 0.9 1.8 9 10 11 12 13 14 15
(a) (b)
Rotation ( o ) Time (s)

Figure 5: Free rocking. (a) Rotation amplitude-period relationship and (b) rotation time history

Because of the energy due to the impact between the footing and support, the peak footing rotation at
the beginning is larger than that at the end. The peak footing rotation at the beginning of each cycle is
called initial footing rotation, herein. Figure 5(a) shows the relationship between the initial footing
rotation and the corresponding vibration period (Tv). As illustrated, the greater the initial footing
rotation, the longer the period of structural free rocking motion. Using shake table measurement
(Figure 4), the maximum rotation of the footing was 0.79o. Figure 5(b) suggests that the fundamental
period of the equivalent model with uplift due to the excitation was 0.43 s. This period of the structural
response with uplift was longer than the fundamental period of structure with an assumed fixed base
(0.34 s). The period of model with uplift should be used to replace the T in Equation (2).

5
5 PREDICTING THE MAXIMUM BM IN AN UPLIFTABLE STRUCTURE
Table 1 shows a comparison of the maximum bending moment obtained from Equations (1)-(4) and
shake table test. The results obtained from three different excitations have shown that estimating the
bending moment in the structure using the effective vibration period (Tv) of upliftable model is more
appropriate than using the fixed base fundamental period (T). The average of maximum bending
moment obtained using three different ground motion was 80.2 Nm. The perdition of average
maximum bending moment obtained by Equations (1)-(4) using the fundamental period of fixed base
model and equivalent upliftable model were 71.6 Nm and 80.4 Nm, respectively. The results show that
using the fixed base fundamental period to predict the response of structure with uplift will
underestimate the maximum bending moment. The prediction of maximum bending moment using the
new approach is very similar to the shake table result.

Table 1: Accuracy of Equations (1)-(4) using different structural period

Maximum BM (Nm) Excitation 1 Excitation 2 Excitation 3 Average
Shake table result 87.4 73.7 79.6 80.2
Using T 83.5 65.6 65.9 71.6
Using Tv 79.5 77.9 83.8 80.4

6 CONCLUSIONS
Experimental work was performed to validate an empirical formula for estimating the bending
moment developed in an upliftable structure. The accuracy of the formula improved when the
fundamental period of an equivalent SDOF model with uplift is applied. This fundamental period was
determined using free rocking tests.
This study has revealed that:
1. The greater the footing rotation due to uplift, the larger the footing rocking period.
2. When uplift took place during an earthquake, the period of structural response with uplift
becomes longer, and thus the bending moment in the structure cannot be estimated using the
fundamental period of structure with an assumed fixed base.

ACKNOWLEDGEMENTS
The research is funded by the Ministry of Business, Innovation and Employment through the Natural
Hazards Research Platform under the Award UoA 3703249.

REFERENCES:
Ali, M., Briet, R., Chouw, N. (2013). Dynamic response of mortar-free interlocking structures, Construction and
Building Materials. 42, 168-189
Beck, J.L., and Skinner, R.I. 1974. Seismic response of a reinforced concrete bridge pier designed to step.
Earthquake Engineering and Structural Dynamics. 2 (4). 343-358.
Chopra, A.K. and Yim, S. C-S. 1985. Simplified earthquake analysis of structure with foundation uplift, Journal
of structural engineering. ASCE, 111 (4). 906-930.
Chouw, N. and Hao, H. 2005. Study of SSI and non-uniform ground motion effect on pounding between bridge
girders, Soil Dynamics and Earthquake Engineering. 25. 717-728.
Crippen. K. 2002. North viaduct to Lions Gate Bridge. Canadian Consulting Engineer. 43 (7). 34-35.
Federal Emergency Management Agency 2000. FEMA 356: Prestandard and Commentary for the Seismic
Rehabilitation of Buildings. Washington, D.C.
Housner, G.W. 1963. The behavior of inverted pendulum structures during earthquakes. Bulletin of the

6
 Seismological Society of America. 53 (2). 403-417.
Hung, H.H., Chang, K.C. Liu, K.Y. and Ho, T.H. 2008. An experimental study on rocking response of bridge
spread foundations. Proceedings of the 14th World Conference on Earthquake Engineering, Beijing China.
12-17 Oct.
Japan Society of Civil Engineering (JSCE), 2000. Earthquake resistant design code in Japan, Maruzen, Tokyo.
Kafle, B., Lam, T.K.N., Gad, E.F., and Wilson, J. 2011. Seismic displacement controlled rocking behaviour of
rigid objects, Earthquake Engineering and Structural Dynamics. 40 (15). 1653-1669.
Kodama, T., Chouw, N. 2002. Soil-structure interaction with uplift due to near-source earthquakes. Proc. of
European Conf. on Structural Dynamics, T. U. Mnchen, Germany, 1375-1380.
Loo, W.Y., Kun, C., Quenneville, P., Chouw, N. (in print). Experimental testing of a rocking timber shear wall
with slip-friction connectors, Earthquake Engineering and Structural Dynamics. DOI: 10.1002/ eqe.2413
Loo, W.Y., Quenneville, P., Chouw, N. (2012a). A numerical approach for simulating the behaviour of timber
shear walls, Structural Engineering and Mechanics. 42(3). 383-407
Loo, W.Y., Quenneville, P., Chouw, N. (2012b). A numerical study of the seismic behaviour of timber shear
walls with slip-friction connectors, Engineering Structures. 34. 233-243
Ormeo, M., Larkin, T., Chouw, N. (in print). The effect of seismic uplift on the shell stresses of liquid storage
tanks, Earthquake Engineering and Structural Dynamics.
Priestley, M. J. N., Evison, R. J. and Carr, A. J. 1978. Seismic response of structures free to rock on their
foundations, Bulletin of the New Zealand National Society for Earthquake Engineering. 11 (3). 141-150.
Psycharis, I.N. 1981. Dynamic behaviour of rocking structure allowed to uplift, PhD thesis of California
Institute of Technology. Pasadena, California.
Psycharis, I.N. and Jennings, P.C. 1983. Rocking of slender rigid bodies allowed to uplift, Earthquake
Engineering and Structural Dynamics. 11. 57-76.
Psycharis, I.N. 1991. Effect of base uplift on dynamic response of SDOF structure, Journal of Structural
Engineering, ASCE. 117 (3). 733-754.
Qin, X. and Chouw, N. 2012. Shake table test on uplift behaviour of a shear frame in earthquakes, 2012 World
Congress on Advances in Civil, Environmental and Materials Research, 26-29 August 2012, Seoul, Korea
Qin, X., Chen, Y and Chouw, N. 2013. Effect of uplift and soil nonlinearity on plastic hinge development and
induced vibrations in structures, Advances in Structural Engineering. 16 (1). 135-147.
Wang, X.F. and Gould, P.L. 1993. Dynamics of structures with uplift and sliding, Earthquake Engineering and
Structural Dynamics. 22. 1085-1095.