1 s2.0 S0267726122000392 Main

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 13

Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Contents lists available at ScienceDirect

Soil Dynamics and Earthquake Engineering


journal homepage: www.elsevier.com/locate/soildyn

Quantification of energy dissipation demand for buckling-restrained braces


T.Y. Yang a, Muhib Muazzam a, Musab Aied Qissab Al-Janabi b, *, Svetlana Brzev a
a
The University of British Columbia, Vancouver, Canada
b
Al-Nahrain University, Baghdad, Iraq

A R T I C L E I N F O A B S T R A C T

Keywords: Buckling-Restrained Braces (BRBs) have been frequently used as stable energy dissipation devices for seismic
BRB applications. Currently, the performance of the BRBs is validated using displacement-based loading protocol,
Loading protocol which does not account for the energy dissipated by the BRBs. In this paper, a novel procedure is proposed to
Energy dissipation quantification
quantify the energy demand for BRBs at different stories and sites. Hence, the BRB can be tested in the laboratory
Performance evaluation
Seismic design
to ensure that sufficient energy can be dissipated. The proposed method uses the site-specific target spectrum to
quantify the earthquake energy for the structures equipped with BRBs. In addition, new quantification factors,
including the rise time for the energy dissipation and story-wise modification factor are proposed. The proposed
quantification factors are obtained from a large array of time history analyses for a range of structures. The
proposed method was verified on a 5-story prototype model equipped with BRBs. The results show that the
proposed procedure can estimate the energy demand of BRBs with a reasonable accuracy, and can be used as a
reliable method to predict and quantify the energy dissipation demand for BRBs.

1. Introduction was observed or the BRB fractured. Takeuchi [6–8] applied similar
strain-based cyclic loading protocol to examine the local buckling and
Buckling restrained brace (BRB) is a commonly used energy dissi­ out-of-plane stability of BRBs. The loading sequences consists of multi­
pation device, which was first proposed in 1988 by Wada and initially ple cycles of axial strain with amplitudes of 0.10%, 0.50%, 1.00% and
applied to two steel frame structures in Japan [1]. The performance of 2.00%. Each amplitude was repeated 3 times. The axial strain was
the BRBs was first tested on a 0.4 scaled model of a single-story, one-bay calculated using the axial displacement of the BRB divided by the length
moment frame with pinned base and inverted-V BRBs. The result of the of restrained yielding segment of the BRB, as shown in Fig. 1. After a
cyclic testing showed that BRBs can dissipate the earthquake energy loading cycle was completed, the loading was repeated with 3% strain
efficiently and effectively [2,3]. BRBs started to gain popularity after the amplitude until either BRB fractured or instability in the BRB and con­
1994 Northridge and the 1995 Kobe earthquakes [4]. The growing nections was observed. The Building Center of Japan (BCJ) adopted
application of BRBs demanded more rigorous tests. In Japan, Iwata [5] loading protocol proposed by Takeuchi [8], with additional 3 cycles of
proposed a strain-based testing protocol for BRBs. The strain-based yield strains added at the beginning of the loading protocol [1].
loading protocol started with one cycle of loading with amplitude Qu et al. [9] developed BRBs with replaceable steel angle fuses. The
equal to 1/3 of yield strain, followed by another cycle of loading with an proposed BRB was tested considering fuse design and material,
amplitude equal to 2/3rd of the yield strain. Subsequently, the loading debonding material, and loading protocol as variables. The test results
protocol consists of incrementally increasing amplitudes, i.e., 0.25%, show that the proposed BRB has a stable hysteresis response. It was
0.50%, 0.75%, 1.00%, 1.50%, 2.00%–2.50% of the maximum strain demonstrated that the BRBs repaired by replacing the fuses performed in
obtained from the dynamic analysis. In general, each amplitude was a satisfactory manner in the subsequent tests. Hu et al. [10] proposed a
repeated 2 times, except for the 0.25% amplitude was applied only in 1 new lateral force-resisting, self-centering system called SCENARIO in
cycle, while the 1.00% amplitude which was repeated 5 times. Finally, which BRBs were provided based on a design procedure. The nonlinear
the amplitude was increased to 3% of the maximum strain and the response history analysis confirmed that the proposed system, when
testing was terminated when either a significant strength degradation designed considering a response modification factor (R) of 6 or 8

* Corresponding author.
E-mail addresses: yang@civil.ubc.ca (T.Y. Yang), cemuhibm@gmail.com, musab.a.jindeel@nahrainuniv.edu.iq (M. Muazzam), musab.a.jindeel@nahrainuniv.edu.
iq (M.A. Qissab Al-Janabi), sbrzev@mail.ubc.ca (S. Brzev).

https://doi.org/10.1016/j.soildyn.2022.107190
Received 8 September 2021; Received in revised form 17 January 2022; Accepted 2 February 2022
Available online 10 February 2022
0267-7261/© 2022 Elsevier Ltd. All rights reserved.
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Fig. 1. Components of a BRB.

amplitudes equal to 0.5 times the expected maximum displacement


Table 1
(Δbm ), which can be calculated from Equation (2) as follows
Loading protocol for Western Intra-plate event in Canada [17].
( )
Load Steps Cycles Δ Duration (sec) Δbm = min 0.01hs , 5Δy (2)
Normalized Amplitudes (μ ​ = )
Δy

1 8 1.00 3.00 where hs is the story height.


2 6 1.50 3.00 Subsequently, the loading protocol includes 4 cycles of amplitudes
3 4 2.50 3.50 equal to Δbm , followed by 2 cycles of higher amplitudes equal to 1.5 Δbm ,
4 3 3.50 4.00 and additional cycles with amplitudes equal to Δbm . Testing continues
5 2 4.50 3.00
until the BRB has achieved a cumulative inelastic axial deformation of
6 2 5.50 3.50
140 Δy . In 2016, AISC 341-16 [15] adopted a similar loading protocol,
Note: Δ = displacement amplitude; Δy = yield displacement. but a different number of loading cycles. The protocol starts with 2
cycles of yield displacement (Δy ), followed by 2 cycles of amplitudes
equal to (0.5Δbm ), and 6 cycles with 3 amplitudes (Δbm , 1.5Δbm , and
(according to the US codes), has a satisfactory seismic performance 2.0Δbm ), where each amplitude was repeated twice. Subsequently, 2
characterized by a very small residual drift. more cycles of 1.5 Δbm were added and the testing continues until the
In the USA, BRBs were first tested at University of California, Ber­ cumulative inelastic axial deformation reaches the value of 200 Δy .
keley [11] using a loading protocol adapted from the SAC project, which AISC341-16 [15] permits application of other testing protocols, pro­
was developed by Clark [12,13]. In 2001, SEAOC-AISC [14] proposed a vided that they are capable of imposing equal or greater severity in
displacement-based testing protocol for BRBs. The loading protocol terms of the maximum and cumulative inelastic deformations in a BRB.
starts with six cycles of yield displacement (Δy ) of the BRB which can be A ductility-based loading protocol for testing BRBs was proposed in
calculated using Equation (1), as follows Canada by Dehghani and Tremblay [16]. Dehghani and Tremblay [17]
further proposed three loading protocols for three different seismic
Fy Lys
Δy = (1) events in Canada, namely Western intra-plate, Eastern intra-plate and
AE
Western inter-plate events. Each protocol was characterized by ductility
where Fy = yield strength of BRB; Lys = length of the restrained yielding amplitudes, number of cycles, and duration of loading. For example,
segment as shown in Fig. 1; E = Young’s modulus of steel core, and A = normalized displacement amplitudes, cycles and duration of loading
area of the BRB. protocol for West intra-plate event are presented in Table 1.
Next, the loading protocol consists of 4 displacement cycles with Existing loading protocols were developed with an objective to
ensure that the BRBs are able to sustain significant inelastic

Fig. 2. The Ed -Ei relationship.

2
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Fig. 4. Comparison of the numerical and experimental results.

premise that during an earthquake, a significant portion of the input


energy, Ei , will be dissipated through damping, Eξ , while the remainder
of the energy will be stored in the structure in the form of kinetic energy,
Ek , and strain energy, Ea . If the structure behaves non-linearly, the en­
ergy dissipated by the structure, ES , can be partitioned into strain en­
ergy, Ee and hysteresis energy, Eh , where Eh indicates energy dissipation
induced by structural damage due to nonlinear behavior. In this paper,
an approach for estimating the energy demand is adopted from the
equivalent energy design procedure (EEDP) proposed by Yang et al.
[19], where the earthquake input energy for the structure can be
calculated by idealizing a structure as an elastic
single-degree-of-freedom (ESDOF) system, as shown in Equation (3):
1
Ei = mSa (T)Sd (T) (3)
2

where m is the mass of the structure, Sa(T) is the spectral acceleration of


the ESDOF, and Sd(T) is the spectral displacement of the ESDOF.
EEDP assumes that the energy to be dissipated by the ESDOF system
is the same as for the nonlinear single-degree-of-freedom (NSDOF) sys­
tem. Hence, the total energy dissipated by the BRB system can be
quantified by Equation (4) [20], as follows
∫u
Ed = fs (u)du (4)
0

where Ed is the total energy dissipated by yielding of a steel structure,


and fs (u) is the resisting force of an inelastic SDOF system. Fig. 2 shows
the Ed and Ei relationship for a specific design spectrum.
The energy quantification factor, γ lp , can be defined as the ratio of
Fig. 3. Prototype buildings: (a) 3-story structure with BRBs [22] (b) 6-story the energy dissipated by the BRB system when subjected to dynamic
structure with BRBs [22], and (c) 8-story structure with BRBs [23]. cyclic loading to the total input energy, as per Equation (5):
Ed
deformations during an earthquake; however, these protocols are not γ lp = (5)
Ei
able to explicitly quantify the actual energy dissipation demand expe­
rienced by the BRBs. In this paper, a novel procedure for quantifying The following steps outline the framework of the proposed energy
energy dissipation demand for structures equipped with BRBs at quantification approach:
different building sites and locations within a building, has been pro­
posed. In addition, a novel parameter intended to quantify the rate at 1 Obtain the building design criteria and seismic hazard parameters for
which the energy is dissipated in the BRBs during an earthquake, has the building site.
also been presented. 2 Determine the fundamental period of the structure and the total
energy from the design spectrum based on the site seismicity.
2. Energy demand for BRBs 3 Determine the energy quantification factor and quantify the total
dissipated energy.
Energy proportioning concept for seismic design of structures was
first introduced by Housner in 1956 [18]. The concept is based on the

3
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Table 2
Ground motions and scale factors (S.F.) for prototype buildings.
Event Station S.F. (3- S.F. (6-
story) story)

Kobe, Japan (1995) Kakogawa 2.91 2.98


Kocaeli, Turkey Duzce 1.97 1.93
(1999)
Chi-Chi, Taiwan CHY036 1.74 1.68
(1999)
Duzce, Turkey (1999) Bolu 1.01 1.00
Imperial Valley-06 El Centro Array #8 1.56 1.56
(1979)
Yountville (2000) Napa Fire Station #3 1.69 1.71
Northern Calif-03 Ferndale City Hall 2.89 2.75
(1954)
Victoria, Mexico Chihuahua 3.37 3.30
(1980)
Westmorland (1981) Westmorland Fire Sta 1.44 1.42
Taiwan SMART1(40) SMART1 O03 2.99 2.90
(1986)
Coalinga-01 (1983) Pleasant Valley P.P. - yard 1.01 1.01
Parkfield-02, CA PARKFIELD, VINEYARD 3.12 3.13
(2004) CANYON
Niigata, Japan (2004) NIG012 3.68 3.89
Morgan Hill (1984) Halls Valley 3.62 3.64
Chuetsu-oki, Japan Kawanishi Izumozaki 1.82 1.82
(2007)
Chalfant Valley-02 Zack Brothers Ranch 1.36 1.35
(1986)
Whittier Narrows-01 Bell Gardens, Jaboneria 3.17 3.26
(1987)
Darfield, New Zealand LINC 1.68 1.70
(2010)
Superstition Hills-02 El Centro Imp. Co. Cent 1.87 1.86
(1987)
Christchurch, New Styx Mill Transfer Station 2.71 2.67
Zealand (2011)
El Mayor-Cucapah, Westside Elementary School 2.35 2.36
Mexico (2010)
40204628 (2007) San Jose; CHP Field Office 4.37 4.48
Junction Ave; 1-story; ground
level
Northridge-01 (1994) Canyon Country - W Lost Cany 1.58 1.57
Managua, Nicaragua- Managua, ESSO 3.05 3.01
02 (1972)

spacing, and typical story height of 4.0 m (except for the first story
height of 5.5 m). BRBs were installed at the perimeter of a prototype
structure: 2 bays for each exterior frame of a 3-story structure and 3 bays
for each exterior frame of a 6-story structure. The prototype buildings
were designed according to FEMA302 [21] by Sabelli et al. [22] for a Los
Angeles site. The design of these structures was discussed by Sabelli et al.
Fig. 5. Ground motion spectra scaled at (a) T = 0.25 s, (b) T = 0.50 s, (c) T = [22]. The 8-story structure shown in Fig. 3 c) has a floor plan of 21.9 m
1.00 s and (d) T = 2.00 s. × 21.9 m and 7.3 m column spacing, and the building information was
adopted from Choi and Kim [23]. Only one of the three bays of the
4 Determine the energy distribution over building height. The dissi­ exterior frame was equipped with BRBs. Typical story height for this
pated energy demand at any story level can be obtained as a product structure was 3.7 m, except for the first story (5.5 m).
of the total dissipated energy and the energy share of that story.
5 Finally, determine the rise time and the distributed dissipated energy 3.2. Modeling of the prototype buildings
demand. The dissipated energy demand over time can be used to
evaluate the performance observed in experimental investigations. Two-dimensional finite element models were developed using the
OpenSees platform [24]. The columns of the prototype buildings were
3. Modeling and analysis assumed as pin-connected at the base. The beams were rigidly connected
to columns to form a moment resisting frame (MRF) system. Elastic
3.1. Prototype structures beam-column elements were used to model the columns and beams as
they were not expected to undergo plastic deformations. Nonlinear truss
Three prototype moment frame structures (3-story, 6-story, and 8- elements with calibrated Steel02 material model were implemented to
story) equipped with BRBs were considered in this study. The 3-story model the BRBs. Fig. 4 shows the force-deformation response of an BRB
structure shown in Fig. 3a) has a regular rectangular plan shape with obtained from numerical simulation against the experimental data from
horizontal dimensions 37.8 m × 56.1 m, 9.1 m column spacing, and a University of California, Berkeley study performed by Clark et al. [11].
typical story height of 4.0 m. The 6-story structure shown in Fig. 3 b) has The comparison shows that the numerical model can accurately simu­
a square plan with dimensions of 46.9 m × 46.9 m, 9.1 m column late the force-deformation response of the BRB.

4
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Fig. 6. Scaled ground motion spectra for prototype structures: (a) 3-story structure and (b) 6-story structure.

Fig. 7. Proposed fit curve for the energy quantification factor, γlp, versus fundamental period, T, for the entire set of data.

All masses from the tributary areas were lumped at the nodes. A 2%
Rayleigh damping ratio was assigned for the three prototype buildings.

5
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Fig. 8. Proposed fit curve for the energy quantification factor, γlp, versus fundamental period, T.

Fig. 9. Idealized story-wise energy distribution.

3.3. Ground motions corresponding to 0.2 T–2.0 T range for 0.51 s fundamental period.
It is worth mentioning that the effect of ground motion duration on
A total of 24 ground motions were selected from the PEER NGA the energy dissipation demand was not explicitly considered in this
database [25] for the analysis of SDOF system. Since crustal earthquakes study, which is in line with the current seismic design provisions.
are frequent at the west coast of North America, the ground motions However, Hou and Qu [29] found that the duration of spectrally
were selected to represent crustal earthquakes with moment magnitude matched ground motions has a significant impact on the hysteretic en­
(Mw) ranging from 5.0 to 7.9 on the Richter scale. The closest site-fault ergy dissipation demand. Future extension of the proposed procedure
distance associated with the recorded motions was nearly 80 km. should consider the effect of ground motion duration and the ground
Scaling of ground motions was performed according to ASCE7-16 motion selection criteria proposed by previous research studies [30,31].
[26] and NBCC 2015 [27], such that the mean spectral value is 10%
less than the target spectrum for the period range from 0.2 T to 2.0 T, 4. Energy demand quantification procedure
where T is fundamental period of the SDOF system. Fig. 5 illustrates
individual mean ground motion spectra scaled to the target spectrum. Energy demand for a structure located at a specific site can be
The target spectrum represents site class D according to FEMA 302 and determined from the design spectrum as the product of input energy and
10% probability of exceedance in 50 years hazard level adopted from energy quantification factor. Subsequently, the demand is distributed
Sabelli [28]. Table 2 shows the selected ground motions and scale vertically along the building height and horizontally at a story level.
factors. After that, the energy demand of each floor is distributed over some
The same ground motions selected for the SDOF systems were time-steps. The proposed procedure can be used to develop a loading
adopted for the prototype structures. Fundamental periods for the 3- and protocol for assessing the seismic performance of BRBs located at
6-story structures were 0.51 s and 0.78 s, respectively. The target spectra different story levels.
to which the mean ground motion spectra were scaled for analysis were The analysis results for 3-, 6- and 8-story prototype structures were
developed according to FEMA 302 recommendations [21]. The scaled studied to investigate floor-wise distribution. The frames were concen­
spectra are shown in Fig. 6. It can be seen from the figure that for 3-story trically braced BRBs (CBRBs). The recommended procedure is discussed
structure the spectra far matched for period range from 0.1 to 1.5 s, in detail in this section.

6
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Fig. 10. Actual versus predicted energy distribution profiles for prototype Fig. 11. Fitted curves for coefficients: (a) C1 , (b) C2 , and (c) C3 .
buildings: (a) 3-story building, (b) 6-story building, and (c) 8-story building.
earthquake was recorded and plotted against the corresponding funda­
4.1. Energy quantification factor, γlp mental periods. It was observed that the R-factor has no significant effect
on the γ lp magnitude, hence, Equation (6) was proposed based on curve
The main objective is to propose an equation for energy quantifica­ fitting, as follows
tion factor (γ lp) as a function of the fundamental period of the original
γ lp = 0.09T − 2.88
+ 1.96 (6)
structure (T). Four SDOF systems with fundamental periods equal to
0.25 s, 0.50 s, 1.00 s and 2.00 s were considered for this purpose. The The entire data set and the median values were compared with the
yield strength of each SDOF was modified and determined depending on predicted energy quantification factor values obtained from the pro­
their force response modification factors (also known as the R-factors). posed equation for R = 4, 6, and 8, see Fig. 7. Fig. 8 shows a comparison
According to ASCE7-16 [26], for design of steel BRBFs the R-factor is between the predicted energy quantification factor values and the me­
recommended to be taken equal to 8.0. In this study the R-factor values dian values with fundamental period (T).
of 4.0, 6.0, and 8.0 were selected as variables to investigate the effect of
R value on energy demand parameters.
The selected SDOF systems were subjected to the scaled ground 4.2. Story-wise energy dissipation demand
motions, and multiple time history dynamic analyses were performed.
Energy dissipated in the system during an earthquake was calculated Once the energy quantification factor (γlp) is obtained based on the
from the recorded force-deformation hysteresis curves at the end of the fundamental period of the structure, energy dissipation demand can be
analysis. The total energy dissipated during an earthquake was distributed up the building height based on the empirical equation
normalized by the input energy obtained from the target spectrum to proposed in this study. The rationale of the equation is that the overall
obtain (γlp). The energy quantification factor, γ lp, for an individual sum of the story-wise energy distribution terms should be equal to 1. For
an n-story building (n being the number of stories), the sum of energy

7
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Fig. 12. Rise time, tR .

Fig. 13. Median value trends for rise time tR.

Fig. 14. Effects of T and R on median trends of rise time values for different fundamental periods: (a) T = 0.25 s, (b) T = 0.50 s, (c) T = 1.00 s, and (d) T = 2.00 s.

8
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Fig. 15. Proposed fit equations for rise time values, tR.

Fig. 16. A 5-story test frame with BRBs [11].

Fig. 17. Scaled ground motion spectra for the period range from 0.29 to 2.85 s.

9
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Table 3 Table 5
Ground motions and scale factors (S.F.) for the test structure. Rise time values for the test 5-story frame.
Event Station S.F. Cumulative Energy Dissipation (Ed) (%) Rise Time (tR) (sec)
Northridge-01 (1994) Playa Del Rey - Saran 2.88 5 1.54
Kobe, Japan (1995) Amagasaki 1.01 25 4.42
Kocaeli, Turkey (1999) Yarimca 1.19 50 7.72
Friuli, Italy-01 (1976) Codroipo 4.42 75 13.12
Friuli, Italy-02 (1976) Buia 3.63 95 21.86
Tabas, Iran (1978) Boshrooyeh 3.65 100 34.48
Chi-Chi, Taiwan (1999) TCU083 2.91
St Elias, Alaska (1979) Icy Bay 2.41
Imperial Valley-06 (1979) El Centro, Meloland Geot. Array 1.09
Hector Mine (1999) Mecca - CVWD Yard 2.61 Table 6
Northern Calif-03 (1954) Ferndale City Hall 1.50 Distribution of story-wise energy demands over rise times.
Chi-Chi, Taiwan-03 (1999) CHY025 2.81
Chi-Chi, Taiwan-04 (1999) CHY015 2.86 Ed (%) tR (sec) Energy Demand Per Story (kN-m)
Westmorland (1981) Parachute Test Site 1.56 1 2 3 4 5
Coalinga-01 (1983) Cantua Creek School 1.77
Taiwan SMART1(40) (1986) SMART1 M02 1.78 5 1.54 45.01 21.19 21.99 17.64 8.82
Taiwan SMART1(45) (1986) SMART1 I02 1.74 25 4.42 225.08 105.95 109.96 88.20 44.10
Montenegro, Yugoslavia (1979) Ulcinj - Hotel Olimpic 1.24 50 7.72 450.15 211.90 219.92 176.40 88.20
Chalfant Valley-01 (1986) Bishop - LADWP South St 4.36 75 13.12 675.23 317.86 329.88 264.59 132.30
Iwate, Japan (2008) Furukawa Osaki City 1.49 95 21.86 855.29 402.62 417.85 335.15 167.58
Superstition Hills-02 (1987) El Centro Imp. Co. Cent 1.37 100 34.48 900.30 423.81 439.84 352.79 176.40
Loma Prieta (1989) Sunnyvale - Colton Ave. 1.72 100 60.00 900.30 423.81 439.84 352.79 176.40
Big Bear-01 (1992) San Bernardino - E & Hospitality 3.17
Point Mugu (1973) Port Hueneme 3.72

which can be simplified as shown in Equation (8b):


Table 4 [
(n − 4)(n − 3)
]
Prediction of story-wise energy demands in the 5-story frame. C1 + 1 + C2 + 3(n − 4)C3 C2 Ed2 = 1 (8b)
2
Story Predicted Energy Distribution Predicted Total Energy Dissipation
Ratio (− ) Demand (Ed) (kN-m) Note that for n ≥ 5, the values of Ed2 and C1, C2 and C3 coefficients
1 0.39 900.30
can be determined by trial and error process. It should be mentioned that
2 0.19 423.81 Equations (8a) and (8b) can only be used when n ≥ 5. For height-wise
3 0.19 439.84 energy distribution in 3- and 4-story structures, Equations (9) and (10)
4 0.15 352.79 should be used, respectively.
5 0.08 176.40
C1 Ed2 + Ed2 + C3 C2 Ed2 = 1 For n = 3 (9)
Next, rise time values were calculated from Equations (12) to determine the
story-wise energy dissipation at six energy levels, as summarized in Table 5.
C1 Ed2 + Ed2 + C2 Ed2 + C3 C2 Ed2 = 1 For n = 4 (10)

dissipation terms Edi along the height can be expressed according to The coefficients play a key role in the applicability of the proposed
Equation (7): empirical equations. Hence, an equation has been proposed for each
coefficient as a function of the number of stories for the structure under

n
Edi = 1 (7) consideration.
i=1 Following the proposed procedure, story-wise energy partition was
determined for the considered structures. According to the shape of
The proposed equation includes three coefficients (C1, C2 and C3)
energy distribution profile for each frame, it was found that the co­
which control the shape of the distribution, which has been idealized as
efficients need to be different for each frame in order to match the actual
shown in Fig. 9 based on the analysis results of the 3-, 6- and 8-story
energy distribution profile. Hence, the equations were proposed
prototype structures. The shapes in transition Zones 1, 2 and 3 are
depending on the number of stories (n). To logically distribute the input
controlled by story coefficients C1, C2 and C3, respectively. Coefficient
seismic energy through the stories of the 3-story structure, the first step
C1 is related to energy distribution between the 1st and 2nd story. In
is to solve Equation (9) by setting the energy attributed to the second
Zone 3, coefficient C3 is related to the distribution at the (n − 1)th and nth story (Ed2 ). C1 , C2 and C3 values are obtained by trial and error process
story with respect to the (n − 2)th story. C2 is related to the energy dis­ such that Equation (9) is satisfied and the distribution profile shape is
tribution in the remaining stories with respect to the 2nd story. Similar acceptably close to the actual shape.
distribution as shown in Fig. 9 could be applied to buildings with Similarly, energy distributions for the 6- and 8-story structures are
different number of stories. obtained by solving for the energy assigned at the second story (Ed2 )
The equation was simplified by expressing distributed energy in using Equation (8), such that Equation (8) is satisfied and the shapes are
other stories in terms of energy assigned to the 2nd story (Ed2). Hence, realistic. A comparison between the actual and predicted energy dis­
the energy share of the first story can be expressed as C1 Ed2 , but from the tributions is presented in Fig. 10. Equation (11) can be used to determine
3rd to the (n − 2)th story, energy distribution is equal to [(n − 2)th − C1, C2, and C3 values based on the number of stories (n). Fig. 11 illus­
2]C2 Ed2 . Finally, the energy shares assigned to the (n − 1)th and nth trates the fitted curves based on which Equation (11) is proposed.
stories are 2(n − 4)C3 C2 Ed2 and (n − 4)C3 C2 Ed2 , respectively. As a result,
C1 = 0.075n2 − 1.035n + 5.43 (11a)
the overall sum of distributed energy terms can be expressed according
to Equation (8a):

( )
C1 Ed2 + Ed2 + C2 Ed2 + ⋯ + (n − 2)th − 2 C2 Ed2 + 2(n − 4)C3 C2 Ed2 + (n − 4)C3 C2 Ed2 = 1 (8a)

10
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

Fig. 18. Actual versus predicted story-wise energy distribution ratio for the 5-story test frame.

Fig. 19. Actual and predicted energy demand on the BRBs at each story level of 5-story test frame.

C2 = 0.058n2 − 0.9417n + 4.3 (11b) median rise time values for 12 SDOFs are shown in Fig. 13.
The effects of fundamental period T and response modification factor
C3 = 0.005n2 − 0.03n + 0.42 (11c) R on the median trends are illustrated in Fig. 14. It is also observed that
the rise time increases as the frequency of the system decreases. More­
4.3. Rise time, tR over, increase in rise time is observed with an increase in response
modification factor R. Since the observed variability trends related to
In this study a novel parameter called rise time, (tR), is defined as a rise time vary for different systems, proposed Equations (12.a) to (12.f)
measure of rate at which energy is dissipated in the BRBs during an enable rise time (tR) to be determined from the fundamental period T of
earthquake. Rise time is defined as the time required to dissipate the the target system depending on the R-factor. The tR values calculated
total energy predicted based on the design spectrum. The rise time from these equations are plotted in Fig. 15.
required to accumulate 100% of energy (tR100%) can be presented at tR5% = − 0.09T + 1.67 (12a)
other levels ranging from 5% (tR5%) to 95% (tR95%), as illustrated in
Fig. 12. The proposed time step-wise distribution of the energy dissi­ tR25% = 0.86T + 3.19 (12b)
pation demand can be converted to loading cycles to develop a loading
protocol. tR50% = 1.95T + 4.94 (12c)
The analysis of SDOF systems was extended to calculate the rise time
(tR). First, each cumulative energy dissipation time history was tR75% = 4.83T + 6.23 (12d)
normalized by the product of the input energy Ei and γ lp. Subsequently,
rise time values required to obtain 5%, 25%, 50%, 75%, 95% and 100% tR95% = 8.74T + 9.39 (12e)
of the energy for each individual earthquake were recorded, and the

11
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

tR100% = 2.36T + 31.11 (12f) 2.54% respectively. However, the procedure is able to efficiently and
quite accurately predict the demands for all stories (see Fig. 19).
Therefore, the proposed procedure can be used to predict story-wise BRB
5. Application of the proposed energy quantification method
energy demand in a structure.
5.1. Test structure
6. Conclusions
To study the applicability of the proposed energy quantification
As seismic design philosophy is shifting toward achieving perfor­
method, a 5-story test frame which was previously designed by Bruneau
mance objectives at different earthquake intensities, the need for simple
[11] and subjected to a different target spectrum than prototype struc­
and reliable procedures to quantify the demands is realized. In this
tures was analyzed. The BRB force-deformation hysteresis relationship
paper, a novel procedure was proposed to quantify energy dissipation
was obtained from experimental tests conducted by Merritt et al. [32] on
demands for BRBs. The procedure, which is based on energy equivalence
Star Seismic BRBs at the University of California, San Diego. The test
between SDOF inelastic and elastic systems, takes into account the
model was regular in plan and elevation and was adopted from Bruneau
fundamental period of the structure as well as the response modification
[11]. The structure was located in an area of moderate to high seis­
factor (R) as the main inputs, and allows to predict energy demands for
micity. The plan dimensions of the structure were 45.7 m × 45.7 m with
BRBs at any site location and for any story of the building without
6.1 m column spacing. The typical story height of the structure is 4.0 m,
conducting time history analysis.
except the first story (5.5 m). Two BRBs bays were provided in each of
To determine the energy quantification factor at any site, an equa­
the perimeter frames of the structure and the braces were inclined in
tion was proposed as a function of the fundamental period of the
opposite directions. The BRBs in both directions consisted of diagonal
structure. Further, an empirical relation was proposed for distribution of
bracings. Floor plan and the frame elevation are shown in Fig. 16.
the energy demand at different stories in the building. As this procedure
The design was performed according to ASCE 7–2010 [33] and AISC
is developed to set the demands for quantification tests, equations to
341–2010 [34] provisions. The target spectrum representing the build­
obtain the time required to dissipate the total energy demand in terms of
ing site was developed according to ASCE 7. The detailed design infor­
six time-steps were proposed which are also functions of the structure’s
mation and frame member properties can be found in Bruneau [11].
fundamental period. Nonlinear dynamic analyses were conducted to
A two-dimensional finite element model of the test frame was
assess the applicability of the proposed quantification procedure. The
developed using OpenSees platform. The supports are fixed at the base
results indicate that the energy demand can be successfully approxi­
and the columns were modeled as nonlinear elements designed to allow
mated with outcomes acceptably close to the median of the actual de­
formation of plastic hinges. The beams were connected rigidly to the
mands, which can lay a solid background for future development of
columns. Fiber sections and force-based beam-column elements were
quantification procedures and loading protocols for BRBs.
used to model the beams and columns. The BRBs were modeled as
nonlinear truss elements and were designed to have the potential of
yielding both in tension and compression. The masses corresponding to Declaration of competing interest
the tributary areas were lumped at the nodes. Modal damping ratio of
2% was assumed for the analysis. The authors declare that they have no known competing financial
A set of 24 ground motions was selected and scaled such that the interests or personal relationships that could have appeared to influence
mean spectrum does not fall below the target spectrum by more than the work reported in this paper.
10% for the period range from 0.2 T to 2.0 T, corresponding to 0.29
s–2.85 s for the fundamental period, T, of 1.43 s (see Fig. 17). Table 3 Acknowledgments
shows the selected ground motions and scale factors for the test
structure. The authors would like to acknowledge the funding provided by the
Quantification of energy demand requires the determination of the International Joint Research Laboratory of Earthquake Engineering
fundamental period of the structure. Therefore, the fundamental period (ILEE), National Natural Science Foundation of China (NSFC
of the 5-story frame (n = 5) was calculated as 1.43 s. The input energy in 51778486), State Key Laboratory of Disaster Reduction in Civil Engi­
the system was calculated from Equation (3) as 1150 kN-m while the neering and Canadian Institute of Steel Construction. Any opinions,
energy quantification factor for the 5-story frame was calculated from findings, and conclusions or recommendations expressed in this paper
Equation (6) as 1.992. Subsequently, energy dissipation demand are those of the authors.
imposed on the structure of 2291 × 103 kN-m is the product of these two
values and shall be distributed over building height, and the corre­ References
sponding coefficient values were determined from Equation (11) as C1
[1] Takeuchi T, Wada A. Buckling-restrained braces and applications. The Japan
= 2.13, C2 = 1.04 and C3 = 0.40. For example, equation (8) was used to Society of Seismic Isolation (JSSI), Tokyo, Japan; 2017.
determine energy distribution ratio in the second story E2 = 0.19. The [2] Fujimoto M, Wada A, Saeki E, Takeuchi T, Watanabe A. Development of unbonded
predicted energy distribution ratios and total energy dissipation de­ braces. Quart Column 1990;115:91–6.
[3] Wada A, Iwata M, Watanabe A, Kuribayashi H. Properties of brace encased in
mands for all stories are given in Table 4.
concrete-filled steel tube. In: IABSE Symposium. 60; 1990. p. 493–4.
The prediction of energy demand for each story was investigated by [4] Aiken ID, Kimura I. The use of buckling-restrained braces in United States. In:
distributing the total energy dissipation from Table 4 and using the rise Proceedings of Japan passive control symposium. Yokohama, Japan: Tokyo
time values presented in Table 5. The distribution of story-wise energy Institute of Technology, Yokohama, Japan,; 2001.
[5] Iwata M, Kato T, Wada A. Buckling-restrained braces as hysteretic dampers. In:
demands over rise times is shown in Table 6. Fig. 18 illustrates a com­ Proceedings of the third international conference. Montreal, Canada: STESSA;
parison between actual and predicted story-wise energy distribution 2000.
demands. It can be observed from the figure that for the 2nd and 5th [6] Takeuchi T, Hajjar J, Matsui R, Nishimoto K, Aiken I. Local buckling restraint
condition for core plates in buckling restrained braces. J Constr Steel Res 2010;66
stories the prediction is quite close to the reality, whereas for the 1st (2):139–49.
story was overestimated and underestimated for the 3rd and 4th stories. [7] Takeuchi T, Hajjar J, Matsui R, Nishimoto K, Aiken I. Effect of local buckling core
It can be seen from the figure that the proposed procedure under­ plate restraint in buckling restrained braces. Eng Struct 2012;44:304–11.
[8] Takeuchi T, Matsui R, Mihara S. Out-of-plane stability assessment of buckling-
estimated the demands for the BRBs at the 3rd, 4th, and 5th story levels restrained braces including connections with chevron configuration. Earthq Eng
of the test structure by 12.75%, 16.04% and 0.54%, respectively while it Struct Dynam 2016;45:1895–917.
overestimated the demands at the 1st and 2nd story level by 15.63% and [9] Qu B, Liu X, Hou H, Qiu C, Hu D. Testing of buckling-restrained braces with
replaceable steel angle fuses. J Struct Eng 2018;144(3). 04018001.

12
T.Y. Yang et al. Soil Dynamics and Earthquake Engineering 155 (2022) 107190

[10] Hu S, Wang W, Qu B. Seismic evaluation of low-rise steel building frames with self- [22] Sabelli R, Mahin S, Chang C. Seismic demands on steel braced frame buildings with
centering energy-absorbing rigid cores designed using a force-based approach. Eng buckling-restrained braces. Eng Struct 2003;25:655–66.
Struct 2020;204:110038. [23] Choi H, Kim J. Energy-based seismic design of buckling-restrained braced frames
[11] Bruneau M, Uang C-M, Sabelli R. Ductile design of steel structures. second ed. New using hysteretic energy spectrum. Eng Struct 2006;28:304–11.
York: McGraw-Hill Companies, Inc.; 2011. [24] OpenSees Navigator 2.5.0. Open system for earthquake engineering simulation
[12] Clark P, Aiken I, Kasai K, Ko E, Kimura I. Design procedures for buildings (OpenSees) framework. In: Pacific earthquake engineering research center (PEER)
incorporating hysteretic damping devices. Proceedings of 68th Annual Convention center; 2013. https://openseesnavigator.berkeley.edu/.
of the Structural Engineers Association of California, Santa Barbara, California; [25] PEER. Ground motion database – BETA version. Pacific Earthq. Eng. Res. Cent.;
1999. 2013. http://peer.berkeley.edu/products/strong_ground_motion_db.html.
[13] Clark P, Frank K, Krawinkler H, Shaw R. Protocol for fabrication, inspection, [26] ASCE/SEI 7-16. Minimum design loads and associated criteria for builidngs and
testing and documentation of beam-column connection tests and other other structures. American Society of Civil Engineers; 2016.
experimental specimens. In: SAC steel project background document; 1997. [27] NBCC. In: National building code of Canada, vol. 1. Ottawa, Ontario: National
October, Report No. SAC/BD-97/02, Redwood City, California. Research Council Canada; 2015.
[14] Uang C-M, Nakashima M. Steel buckling-restrained braced frames. In: Earthquake [28] Sabelli R. Research on improving the design and analysis of earthquake-resistant
engineering from engineering seismology to performance-based engineering. steel-braced frames. The 2000 NEHRP professional fellowship report,. Department
Chapter 16, CRC Press LLC; 2004. of Civil and Environmental Engineering, University of California, Berkeley; 2001.
[15] American Institute for Steel Construction (AISC). Seismic provisions for structural [29] Hou H, Qu B. Duration effect of spectrally matched ground motions on seismic
steel buildings. Chicago, IL: American Institute for Steel Construction; 2016. ANSI/ demands of elastic perfectly plastic SDOFS. Eng Struct 2015;90:48–60.
AISC 341-16. [30] Tian L, Ma R, Qu B. Influence of different criteria for selecting ground motions
[16] Dehghani M, Tremblay R. Standard dynamic loading protocols for seismic compatible with IEEE 693 required response spectrum on seismic performance
qualqification of BRBFs in eastern and Western Canada. In: Proceedings of the15th assessment of electricity transmission towers. Eng Struct 2018;156:337–50.
world conference on earthquake engineering; 2012. Lisbon, Portugal. [31] Tian L, Yi S, Qu B. Orienting ground motion inputs to achieve maximum seismic
[17] Dehghani M, Tremblay R. Development of standard dynamic loading protocol for displacement demands on electricity transmission towers in near-fault regions.
buckling-restrained braced frames. In: STESSA 2012: behaviour of steel structures J Struct Eng 2018;144(4). 04018017.
in sesmic areas. Santiago, Chile: CRC Press; 2012. p. 61–7. [32] Merritt S, Uang C-M, Benzoni G. Subassemblage testing of star seismic buckling-
[18] Housner GW. Limit design of structures to resist earthquakes. In: Proceeding of the restrained braces. Department of Structural Engineering, University of California,
1st world conference on earthquake engineering; 1956. California, U.S.A. San Diego; 2003. Report No. TR-2003/04.
[19] Yang TY, Tung P, Li Y. Equivalent energy design procedure for earhtquake resilient [33] ASCE/SEI 7-10. Minimum design loads and associated criteria for builidngs and
fused strucutres. Earthq Spectra 2018;34(2):795–815. other structures. American Society of Civil Engineers; 2010.
[20] Chopra AK. Dynamics of structures: theory and application to earthquake [34] American Institute for Steel Construction (AISC). Seismic provisions for structural
engineering. 2012. Englewood Cliffs, N.J: Prentice Hall. steel buildings. Chicago, IL: American Institute for Steel Construction; 2010. ANSI/
[21] FEMA 302, NEHRP. In: Recommended provisions for seismic regulations for new AISC 341-10.
buildings and other structures, Part 1. Washington, D.C.: Building Seismic Safety
Council; 1997. Provisions.

13

You might also like