Performance Evaluation of a Seismically-Isolated Bridge

Structure with Adaptive Passive Negative Stiffness

N. Attary
Rensselaer Polytechnic Institute, Troy, NY, U.S.A.

M. Symans
Rensselaer Polytechnic Institute, Troy, NY, U.S.A.

S. Nagarajaiah
Rice University, Houston, TX, U.S.A.

A.M. Reinhorn
University at Buffalo, Buffalo, NY, U.S.A.

M.C. Constantinou
University at Buffalo,Buffalo, NY, U.S.A.

D. Taylor
Taylor Devices Inc., North Tonawanda, NY, U.S.A.

D.T.R. Pasala
Rice University, Houston, TX, U.S.A.

A.A.S. Sarlis
University at Buffalo, Buffalo, NY, U.S.A.

ABSTRACT

A Negative Stiffness Device (NSD) that has been developed for seismic response control of structures is
presented herein. The device is purely mechanical in that it utilizes a pre-compressed spring to push a structure
away from its equilibrium position and thus induce negative stiffness behavior. The negative stiffness is
combined with the positive stiffness of the primary structural framing to produce a system that exhibits an
apparent reduction in both strength and stiffness. In this paper, the NSD devices are described along with their
response to cyclic testing that was conducted in preparation for shaking table tests of a seismically-isolated
bridge model that incorporates the NSDs within the isolation system. In addition, numerical simulation results
are presented to illustrate the effect that the NSDs are expected to have on the isolated bridge.

Keywords: Negative Stiffness Device, Bridge Structure, Adaptive Passive, Seismic Protection Device,
Performance Measure

Introduction: Next generation Seismic Protection Devices

Bridges are one of the most critical structures in our built environment. In order to protect
these vital lifelines, many different types of seismic protection devices have been developed
around the world, assessed numerically, tested experimentally and implemented in many
bridges. In the past few years, the authors have developed the next generation of these
devices, using the concept of negative stiffness, which can significantly reduce the response
of structures. The developed Negative Stiffness Device (NSD) is a completely mechanical
device that exhibits true negative stiffness behavior using a pre-compressed spring (as
contrasted with other similar devices that employ pseudo-negative stiffness (e.g., see Iemura,
2009). The device is regarded as providing "true" negative stiffness since the force produced
by the device is displacement-dependent and in a direction opposite to the imposed
displacement. Further, the device is considered to be an "adaptive passive" device in that it
can mechanically change its behavior base on the deformation of the structure to which it is
attached. The NSD has the effect of reducing the dynamic forces in a structure through virtual
softening behavior. The reduced forces generally come at the expense of increased
displacements, although the displacements can be controlled by implementing a damper in
parallel with the device (Nagarajaiah et al, 2010 and Reinhorn et al., 2009). Two prototypes
of the device have been fabricated by Taylor Devices, Inc. (a seismic protection device
manufacturer) and tested to evaluate their response to cyclic loading. In addition, the effect of
the devices on the seismic response of two different three-story structures (one isolated
structure with linear behavior and one non-isolated with plastic behavior) has been evaluated
via shaking table tests (Sarlis et al, 2011 & 2012 and Pasala et al, 2011, 2012a & 2012b). In
the final stage of this project, the NSDs will be implemented in the isolation system of a
quarter-scale bridge model (see Fig. 1) and tested on one of the shaking tables at the
University at Buffalo NEES (UB-NEES) (Attary et al., 2012).

Figure 1. Virtual 3D view of bridge model with NSDs on seismic shaking table at UB-NEES site

Behavior of Negative Stiffness Device

Properties and characteristics of the NSD devices have been presented in other publications
by the authors (Sarlis et al. 2011 & 2012, Pasala et al. 2011, 2012a, & 2012b, and Attary et
al. 2012). In anticipation of the shaking table tests of the aforementioned bridge model, the
properties of the NSDs were modified to optimize their effect on the structure. The key
feature of the device is its negative stiffness, which is controlled by the stiffness and pre-
compression of the primary spring. The pre-compression force of the primary springs was
reduced to 19.57 kN. In addition, the NSDs are designed to have no effect on the structure
until a specified displacement is exceeded. Beyond that displacement, the primary spring is
engaged and produces negative stiffness. To achieve such behavior, the device is equipped
with two gap-spring assemblies (GSAs) that employ secondary springs. In order to optimize
the behavior of the device for the bridge model testing, the GSAs were redesigned such that
the negative stiffness is engaged after 0.2 in. of displacement. The modified design of the
NSDs results in the analytical force-displacement relations shown in Fig. 2 wherein separate
curves are shown for the following cases: 1) NSDs alone, 2) isolation system without NSDs,
3) isolation system with NSDs, and 4) isolation system with NSDs and in series with the
bridge piers.

Figure 2 Analytical force-displacement relations for components of isolated bridge model

Having redesigned the NSDs, it was necessary to conduct system identification tests to verify
their behavior relative to the analytical model and to calibrate the numerical model to be used
within the numerical simulations of the isolated bridge model. To this end, the NSDs were
installed within a load frame at the UB-SEESL laboratory wherein they were anchored at the
bottom and allowed to displace laterally at the top. A hydraulic actuator was attached at the
top of the NSD and was used to impose harmonic motion. A range of tests were performed to
consider the effect of frequency and amplitude of motion. The undeformed and deformed
shape of the NSD during one of the tests is shown in Figure 3 and, as an example, the force-
displacement relation for one of the experimental tests is shown in Figure 4. Note that the
data in Figure 4 represents 4 cycles of harmonic motion.

Figure 3 Undeformed and deformed shape of negative stiffness device

 Figure 4 Force-Displacement relation of NSD (North) from harmonic testing
(frequency = 0.01 Hz, amplitude = 3 in.)

Numerical Simulations and Performance Measures

Using the software SAP2000, a detailed 3D numerical model of the bridge test specimen was
developed (see Fig. 5) and subjected to a number of historical earthquake records, some of
which were scaled based on limitations of the anticipated shake table tests (see Table 1). The
simulations were performed for four different bridge configurations: 1) Isolated Bridge (IB),
2) Isolated bridge with the NSDs (IB-NSD), 3) Isolated bridge with passive viscous dampers
(IB-PD), and 4) Isolated bridge with NSDs and passive viscous dampers (IB-NSD-PD).

Figure 5. Numerical model of isolated bridge with NSDs (model developed using SAP2000)
 Table 1 Ground motions used in numerical simulations

No. Earthquake Record Mw

1  Northridge, 1/17/1994 637-270 6.69
2  Loma Prieta, 10/18/1989 CAP-000 6.93
3  Kocaeli, Turkey, 8/17/1999 DZC-270 7.51
4  Northridge, 1/17/1994 NWH-090 6.69
5  Chi-Chi, Taiwan, 9/20/1999 TCU-129-E 7.62
6  Chi-Chi, Taiwan, 9/20/1999 TCU-065-E 7.62
7  San Fernando, 2/9/1971 PCD-254 6.61
8  Cape Mendocino, 4/25/1992 PET-090 7.01
9 Kobe, 1/16/1995 KJM-000 6.90
10 Northridge, 1/17/1994 SYL-00 6.69

In order to evaluate the effectiveness of the NSD, the following six Performance Measures
(PM) (Reigles and Symans, 2005) are defined and used to compare the response of the
system with NSDs relative to the response without the NSDs:

               | |
                     | |

                     | |
                 | |

                     | |
             | |

  .              | |
  .          | |

  .                    | |
  .                | |

  .                    | |
  .            | |

Note that the deck displacements defined in the performance measures are defined as
displacements relative to the ground. Of course, for an isolated bridge, it is expected that
these displacements are similar to displacement across the isolation plane since the lateral
stiffness of the bridge piers is much higher than that of the isolation system. A summary of
the PM values for each ground motion in Table 1 is shown in Fig. 6. Note that a PM value
smaller than unity indicates that the NSDs were effective in reducing a particular response
quantity. As expected, the effect of the NSDs is to reduce the base shear for all cases
considered while the deck displacement is increased in a number of cases. In general, the
effectiveness of the NSDs in reducing base shear depends on the degree to which they were
activated which depends on the displacement demand on the devices. The effect of the
viscous dampers in controlling displacement is evident by comparing the plots for PMD1 and
PMD3 where the only difference is the addition of the dampers.

PMV1 PMD1
1.2 1.6

1 1.4
1.2
0.8
1

PMD
PMV

0.6 0.8
0.6
0.4
0.4
0.2 0.2

0 0

1-637-270

2-CAP000

7-PCD-254

8-PET090

9-KOBE
3-DZC270

6-TCU065
4-NWH090

10-Sylmar00
5-TCU-129-W
1-637-270

2-CAP000

7-PCD-254

8-PET090

9-KOBE
3-DZC270

4-NWH090

6-TCU065
5-TCU-129-W

10-Sylmar00

Ground Motions Ground Motions

PMV2 PMD2
1.2
1.8
1 1.6
1.4
0.8
1.2
PMV

PMD

0.6 1
0.8
0.4
0.6
0.2 0.4
0.2
0 0
2-CAP000

7-PCD-254

9-KOBE
1-637-270

4-NWH090
3-DZC270

6-TCU065

8-PET090

10-Sylmar00
5-TCU-129-W

2-CAP000

7-PCD-254

9-KOBE
1-637-270

4-NWH090
3-DZC270

6-TCU065

8-PET090

10-Sylmar00
5-TCU-129-W

Ground Motions

Ground Motions

PMV3 PMD3
1.8
1.2
1.6
1
1.4
0.8
1.2
PMV

0.6 1
PMD

0.4 0.8
0.2 0.6
0 0.4
7-PCD-254

8-PET090
3-DZC270

6-TCU065
1-637-270

2-CAP000

4-NWH090

5-TCU-129-W

10-Sylmar00
9-KOBE

0.2
0
9-KOBE
2-CAP000

4-NWH090

8-PET090
1-637-270

5-TCU-129-W

7-PCD-254
3-DZC270

6-TCU065

10-Sylmar00

Ground Motions
Ground Motions

Figure 6. Values of force and displacement performance measures for various ground motions
It should be noted that the analytical model of the NSD used in the simulations is idealized in
that it does not account for the dynamics of the device, friction at the pinned connections, and
the flexibility of the steel framing members within the device. As demonstrated by Sarlis et
al. (2011), these effects help to explain details of the device behavior but have limited
practical significance. As shown in Figure 4, cyclic testing does reveal some degree of
hysteretic response, which may be attributed to friction at various locations within the device.
The hysteretic response may be viewed as a positive feature of the device in that it results in a
device that provides both negative stiffness to reduce forces and damping to limit
displacements.

As indicated above, it can be difficult to simultaneously reduce forces and displacements in

the isolated bridge structure. As a means of illustrating the effectiveness of the various
isolation systems in simultaneously reducing both of these quantities, the base shear and
displacement performance measures can be combined in a single plot (see Fig. 7). In such a
plot, the best systems are those which remain under unity for both axes. As shown in Fig. 7,
the NSDs are effective in reducing forces in the bridge piers (base shear) but they may
increase the deck displacement in some cases. The addition of viscous dampers (without
NSDs) can be used to reduce the displacements but may lead to somewhat increased forces.
Using both NSD's and viscous dampers in parallel can, in some cases, result in simultaneous
reduction of forces and displacements.
1.8 1.8
1.6 1.6
1.4 1.4
1.2 1.2
PMD1

PMD2

1 1
0.8 0.8
0.6 0.6
0.4 0.4
0.2 0.2
0 0
0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2
PMV1 PMV2

1.8
1.6
1.4
1.2
PMD3

1
0.8
0.6
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1 1.2
PMV3

Figure 7. Simultaneous evaluation of force- and displacement-related performance measures

Another approach to quantifying the performance of the NSDs in terms of simultaneous

consideration of force and displacement response is via a Combined Performance Measure
(CPM) (Reigles and Symans, 2005) as defined below.
  | |  | |  | |  | |
 
 | |  | |

 | |  | |  | |  | |
 
 | |  | |

 | |  | |  | |  | |
 
 | |  | |

Note that, if the value of CPM is less than zero, the system with NSDs produces an overall
improvement in performance relative to the system without NSDs. As shown in Fig. 8, the
NSDs reduce the base shear in the bridge model significantly as compared to the isolated
bridge without the NSDs. Although it might be expected that adding the NSDs to the
isolation system would generally increase the displacements, in some cases the displacements
are reduced. Thus, as noted previously, it is possible for the NSDs to simultaneously provide
a reduction in forces and displacements in the system. The addition of passive dampers (PD)
in parallel with the NSDs results in increased forces and reduced displacements relative to the
isolated bridge with NSD alone. Although the base shear increases, it is still less than for the
case where the bridge only employs isolation bearings. Thus, the isolation system that
employs both NSDs and PDs is regarded as providing good performance with regard to both
forces and displacements (the primary function of the NSDs being to limit forces while the
PDs limit displacements).
CPM1
0.6
CPM2
0.4 0.6
0.2 0.4
0 0.2
-0.2
0
-0.2
CPM

CPM

-0.4
-0.4
-0.6
-0.6
-0.8
-0.8
-1 -1
-1.2 -1.2
-1.4 -1.4
1-637-270

7-PCD-254

9-KOBE
5-TCU-129-W
3-DZC270

6-TCU065

10-Sylmar00
2-CAP000

4-NWH090

8-PET090

2-CAP000

7-PCD-254

9-KOBE
1-637-270

4-NWH090
3-DZC270

6-TCU065

8-PET090

10-Sylmar00
5-TCU-129-W

Ground Motions Ground Motions

CPM3
0.6
0.4
0.2
0
-0.2
CPM

-0.4
-0.6
-0.8
-1
-1.2
-1.4
6-TCU065
2-CAP000

4-NWH090

5-TCU-129-W
1-637-270

10-Sylmar00
7-PCD-254
3-DZC270

8-PET090

9-KOBE

Ground Motions

Figure 8. CPM values for different ground motions

Conclusions

In this paper, the performance of a seismically-isolated bridge structure that incorporates

negative stiffness devices was examined via numerical simulations and using various
performance measures. The numerical simulations demonstrated that the use of NSDs within
the isolation system can significantly reduce the peak base shear. Although in some cases the
NSDs produce a reduction in peak deck displacements, it is expected that they will generally
increase the displacements. To address this issue, passive dampers can be added in parallel
with the NSDs. The passive dampers decrease the displacements but will generally increase
the base shear in such a way that the base shear is still less than the case in which the bridge
only employs isolation bearings, thus providing good overall performance with regard to both
forces and displacements. Furthermore, this study has demonstrated that NSDs are effective
in cases where the isolation system that incorporates the NSDs is supported on a flexible
layer (i.e., the bridge piers) rather than being directly connected to the foundation of the
structure.

Acknowledgements

This work was supported by the National Science Foundation (NSF) under Grant No. CMMI-
0830391. Any opinions, findings, and conclusions or recommendations expressed in this
material are those of the author(s) and do not necessarily reflect the views of NSF. Technical
assistance provided by Mr. John Metzger (Engineering Manager) of Taylor Devices, Inc., in
relation to design of the negative stiffness devices for bridge model testing, is gratefully
acknowledged. In addition, Mr. Paul Tegnazian, working as an undergraduate research
assistant, provided assistance with development of design drawings of the bridge model.

References

Attary, N., Symans, M.D., Nagarajaiah, S., Reinhorn, A.M., Constantinou, M.C., Taylor, D.,
Sarlis, A.A. and Pasala, D.T.R. (2012). “Application of Negative Stiffness Devices for
Seismic Protection of Bridge Structures," Proc. of 2012 ASCE Structures Congress, Chicago,
IL, March.

Iemura H. and Pradono M.H. (2009). “Advances In The Development Of Pseudo-Negative-

Stiffness Dampers For Seismic Response Control,” Structural Control and Health
Monitoring, 16, 784–799.

Nagarajaiah, S., Reinhorn, A.M., Constantinou, M.C., Taylor, D., Pasala, D.T.R., Sarlis, A.A.
(2010). “Adaptive Negative Stiffness: A New Structural Modification Approach for Seismic
Protection," Proc. 5th World Conference on Structural Control and Monitoring, Paper No.
5WCSCM-103.

Reigles, D.G. and Symans, M.D. (2005). “Systematic Performance Evaluation of Smart
Seismic Isolation Systems,” Proc. of 2005 ASCE Structures Congress, New York, NY.

Reinhorn, A.M., Lavan, O. and Cimellaro, G.P (2009), “Design of Controlled Elastic and
Inelastic Structures,” Earthquake Engineering and Engineering Vibration, Special issue on
“Advances in Seismic Response Control of Structures,” 8(4), 469-479.
Sarlis, A.A, Pasala, D.T.R, Constantinou, M.C, Reinhorn, A.M, Nagarajaiah, S., and Taylor
D. (2011). “Negative Stiffness Device for Seismic Protection of Structures – An Analytical
and Experimental Study,” COMPDYN 2011, Proc. of 3rd ECCOMAS Thematic Conference
on Computational Methods in Structural Dynamics and Earthquake Engineering, Corfu,
Greece.

Sarlis, A.A, Pasala, D.T.R, Constantinou, M.C, Reinhorn, A.M, Nagarajaiah, S. and Taylor
D. (2012). “Negative Stiffness Device for Seismic Protection of Structures,” Journal of
Structural Engineering, doi:10.1061/(ASCE)ST.1943-541X.0000616

Pasala, D. T. R., Sarlis, A. A. S., Nagarajaiah, S., Reinhorn, A. M., Constantinou, M. C.,
Taylor, D., (2011) "Adaptive Negative Stiffness: A New Structural Modification Approach
for Seismic Protection," Proceedings of 2011 ASCE Structures Congress, Las Vegas,
Nevada.

Pasala, D.T.R., Sarlis, A.A., Nagarajaiah, S., Reinhorn, A.M., Constantinou, M.C. and Taylor
D. (2012a). “Adaptive Negative Stiffness: A New Structural Modification Approach for
Seismic Protection” Journal of Structural Engineering. doi:10.1061/(ASCE)ST.1943-
541X.0000615

Pasala, D.T.R., Sarlis, A.A.S., Nagarajaiah, S., Reinhorn, A.M., Constantinou, M.C. and
Taylor, D. (2012b). "Negative Stiffness Device for Seismic Protection of Multistory
Structures," Proceedings of 2012 ASCE Structures Congress, Chicago, IL.