Professional Documents
Culture Documents
A R T I C L E I N F O A B S T R A C T
Keywords: The present study evaluates the seismic performance factors of typical buildings with dual lateral force resisting
Dual steel systems system (LFRS) consisting of special moment frames (SMFs) combined with special concentrically braced frames
Special moment frame (SCBFs) using FEMA P695 methodology. For this purpose, six building archetypes with 2-, 4-, 8-, 12-, 16- and 20-
Special concentrically braced frame
stories, in two short- and long-period performance groups were analyzed by the response spectrum method and
FEMA P695
Seismic performance factors
designed based on the requirements of ASCE 7–16 and AISC 341–16. Validated nonlinear models were developed
Collapse fragility analysis for each building in the OpenSees software platform. Then, the static and dynamic behavior of each building was
investigated up to the collapse stage by nonlinear static analyses and nonlinear time history analyses under 44
far-field ground motions of FEMA P695. In this investigation, the peak and residual drift ratios, hysteresis
behavior of the stories during the earthquake, collapse fragility curves, and seismic performance parameters (R,
Ω, and Cd) were obtained and discussed. The results of nonlinear analyses demonstrated that code-compliant long
period dual SMF-SCBFs (16 stories or higher) performed poorly and can be severely damaged in strong
earthquakes.
* Corresponding author.
E-mail addresses: ehsan.bakhshivand@gmail.com (E. Bakhshivand), H.AhmadieAmiri@gmail.com (H. Ahmadie Amiri), smaleki@sharif.edu (S. Maleki).
https://doi.org/10.1016/j.soildyn.2022.107506
Received 10 May 2022; Received in revised form 31 July 2022; Accepted 16 August 2022
Available online 25 September 2022
0267-7261/© 2022 Elsevier Ltd. All rights reserved.
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
In this dual type of LFRS, SMFs act as a secondary system providing for yielding of primary structural elements could be obtained by applying
ductility and energy absorption needed, while the primary SCBF system the above strategy and designing column sections to remain elastic with
absorbs most of earthquake forces due to its high stiffness. It should be the aid of the Theory of Plastic Mechanism Control (TPMC). Many other
noted that, at low seismic hazard levels, due to the considerable stiffness studies [40–43] have also been conducted on the application of the
of the bracing system relative to the moment frame, most of the lateral TPMC in the design of dual MRF-EBF and MRF-CBF systems.
load is resisted by the primary system; whereas, by increasing the Current codes accept linear analysis methods such as equivalent
seismic hazard level, yielding, buckling, and failure of the braces occur, static lateral force analysis and response spectrum analysis for the
and the participation of the secondary system in seismic energy dissi seismic design of buildings. The accuracy of these methods that are
pation and structural stability becomes more prominent [21]. In dual popular among designers due to their simplicity heavily depends on
LFRSs, different combinations depending on the type of bracing and the seismic performance factors used in the design process. These factors
moment frame used can be created. For example, the combination of were first introduced in ATC 3–06 [44] for conventional LFRSs at that
SMFs with SCBFs, buckling-restrained braced frames (BRBFs), or time. The basis of determining these factors was mostly engineering
eccentrically braced frames (EBFs) are approved for use [1] when the judgment and the performance of buildings under the past earthquakes
SMFs are capable of resisting at least 25% of prescribed seismic forces. [45]. In the late twentieth century, numerous studies [46,47] were
The first attempts to identify the behavior of dual systems date back conducted to estimate these factors. With the research advancements
to the late twentieth century when experimental and analytical studies made over the past two decades in performance-based earthquake en
were carried out to investigate the performance of moment frames gineering, the FEMA P695 [48] methodology was introduced for
combined with concentric and eccentric braces. In these studies, damage quantifying the seismic performance factors and evaluating the effec
distribution between the moment frame and the braced frame, the use of tiveness of existing and new LFRSs in fulfilling the performance objec
25% minimum force for SMFs, and the seismic performance factors were tives of seismic codes. In this methodology, the life safety performance
evaluated, and solutions were presented to improve the performance of objective of a LFRS, which is the basis of seismic requirements of
dual systems [22–25]. The concept of dual system in improving the building codes such as ASCE 7–16 [1], is evaluated by the acceptability
seismic performance of BRBFs was also studied [26–28]. The results of of its collapse probability at the MCE seismic intensity level. In this
these studies showed that the addition of a moment frame to a BRBF can evaluation, various building archetypes of the LFRS are analyzed by
compensate for the low post-yield stiffness of the braces and reduce the nonlinear static and dynamic analyses, and the uncertainties in
residual drift of the entire system significantly. Mehdipanah et al. [29] modeling, analysis, and design are considered. Following the introduc
used a stiffness-based design method to arrive at a satisfactory perfor tion of FEMA P695 methodology, seismic performance factors of various
mance in SMF-BRBF dual system. They obtained the stiffness contribu LFRSs were evaluated by researchers [34,49–51] and the National
tion of the moment frames and the BRBFs to be 35% and 65%, Institute of Standards and Technology (NIST) [52]. However, the
respectively. Maley et al. [30] implemented the displacement seismic performance factors of SMF-SCBF dual system have not been
based-design method as a potential design strategy to improve the per studied using the FEMA P695 methodology in the technical literature so
formance of dual systems incorporating moment-resisting frames far.
(MRFs) and BRBFs. The results of this study demonstrated that the This study aims to evaluate the seismic performance factors of SMF-
adopted design strategy provides good control of drifts for studied SCBF dual LFRSs using the FEMA P695 methodology. This goal is pur
structures. Freddi et al. [31] investigated the optimal combination of sued through the following steps: 1) Selecting the building archetypes
strength, stiffness, and ductility between the BRBFs and moment frames and ground motions and explaining the methodology and design as
in a parametric study. Aukeman and Laursen [32] examined the 25% sumptions, 2) developing nonlinear models of building archetypes using
rule in such dual systems. They showed that the dual system performs recent achievements in the field of nonlinear modeling of braced and
well, even with a 15% share of the moment frame in the base shear. moment frames, 3) performing nonlinear static and dynamic analyses of
Some studies [33–35] have also demonstrated that there is a secondary building archetypes and evaluating the seismic demand parameters, and
moment action (reserve capacity) even in the concentrically braced 4) performing collapse fragility analysis of the building archetypes and
frames (CBFs), which could enhance the collapse performance of evaluating the seismic performance factors of the dual LFRS using the
low-ductility steel CBF systems. This additional lateral force-resisting FEMA P695 methodology.
capacity that is triggered after significant damage to the primary LFRS
have some sources such as beam-column connections that are part of 2. Building archetypes and ground motions
gravity system, braced frame beam-column gusset connections, and
frame action within braced frames. Additional stiffness, ductility, and Applying the FEMA P695 methodology requires the selection of
strength provided by these sources can aid in the distribution of inelastic building archetypes within the design range of the LFRS under study. In
response over multiple stories to decrease concentration of story drift the present study, assuming the use of one type of structural plan, six
following brace buckling or brace connection fracture. Recently, Bradley building archetypes with 2, 4, 8, 12, 16, and 20 stories were defined in
et al. [36] used this concept and proposed design criteria for a concen the two short- and long-period performance groups. Buildings are
trically braced dual frame that can be used as an alternative to con considered short- or long-period depending on their transition period (Ts
ventional low ductility braced frames. Their proposed system consists of in the ASCE 7–16 spectrum). This ratio is defined as the spectral
a stiff low-ductility concentrically braced frame supplemented by a response acceleration parameter at the period of 1 s divided by that at
reserve moment-resisting frame, which can achieve better collapse the short period (Ts = SD1/SDS = SM1/SMS). According to this definition,
performance without the need for additional detailing requirements to buildings with a period of less than Ts are considered short-period and
conventional seismic design provisions. In further studies about the dual vice versa.
systems, Bosco and Rossi [37] proposed a method for designing dual Fig. 1 depicts the typical plan and elevation layout of the building
MRF-EBFs, which could use the capacity of all link beams. Seker [38] archetypes. It should be noted that the plan is selected from the NIST
recently investigated the effects of bracing configuration on the seismic project [52], in which the bay length of the frames is equal to 20 ft (6.1
behavior of two 4- and 10-story buildings with SMF-SCBF dual steel m), the height of the first story is 15 ft (4.57 m), and the other stories are
system and found that the cross X configuration performed better than 13 ft (3.96 m). As can be seen, in each direction, four bays of the frame
the split X, Chevron, and V configurations. Piluso et al. [39] evaluated (two bays on each side) are braced symmetrically with a two-story X
the seismic response of MRF-CBF dual systems equipped with low bracing configuration. Considering the symmetry of the plan and that
damage friction dampers at beam ends, column bases, and chevron the other LFRS is a perimeter moment frame, half of the building is
brace intersections. They showed that the idea of structures free from studied two-dimensionally in the east-west direction. The P-Delta effects
2
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Fig. 1. Typical plan and elevation of building archetypes; a) plan view, b) elevation view.
of the interior gravity frames are considered by rigid leaning columns archetypes, the latter criterion governs the final cross-section of the
without lateral stiffness. The dead load of all floors is 90 psf (4.31 beams and columns of moment frames. In addition, SMF components
kN/m2), the unreduced live load of the floors and the roof are 50 psf (2.4 fulfilled the ASCE 341-16 seismic requirements such as seismic
kN/m2) and 20 psf (0.96 kN/m2), respectively. The load of the exterior compactness and the strong column-weak beam concept. All beam to
cladding walls is 25 psf (1.2 kN/m2). column connections of the SMFs were considered of the type Welded
Building archetypes were designed using the requirements of ASCE Unreinforced Flange-Welded Web (WUF–W) moment connections,
7–16 [1], AISC 360–16 [53], AISC 341–16 [2], and AISC 358–16 [54] in satisfying the requirements of AISC 358–16 [54] for full development of
ETABS software [55]. For this purpose, the buildings were assumed to plastic hinges in beams. The panel zones were designed as strong ele
have an office occupancy, soil site class D, and seismic design category ments and were not considered as dissipative elements in the design
Dmax. Seismic performance factors including response modification process. The braces were also seismically compact with a slenderness
factor (R), overstrength factor (Ω), and the deflection amplification ratio of less than 200. Members of the braced bays, including beams,
factor (Cd) were considered as 7, 2.5, and 5.5, respectively, in the initial columns, and gusset plates, were designed for the capacity-limited
design process. The response spectrum analysis (RSA) method was used seismic load effects. The gusset plates were tapered plates made of
in the design of all building archetypes because of the significant ASTM A572 Gr.50 (Fy = 50 ksi) designed using the method proposed by
participation of higher modes in the seismic response of long-period Astaneh-Asl et al. [56]. The structural sections obtained from the design
buildings. In Fig. 2, the response spectra of the DBE and MCE intensity of building archetypes are summarized in Table 1. In the table, the SCBF
levels, along with the periods used in the design of buildings, are dis bay members (beams and columns) are designated as “Braced” and the
played. In developing these spectra according to FEMA P695 and ASCE SMF bay members as “Unbraced”. Full details of the seismic design
7–16 the parameters used were: Fv = 1.7, Fa = 1.0, S1 = 0.599 g, and Ss procedure of the building archetypes together with the design outcomes
= 1.5 g. are openly shared via the repository at https://github.com/EBakhsh
The American W-sections with ASTM A992 steel (Fy = 50 ksi) were ivand/SMF-SCBF-dual-system.
used for beams and columns. The hollow square structural sections The present study employs 44 far-field ground motions of FEMA
(HSS) for braces were used with ASTM A500 Gr. B (Fy = 46 ksi) steel. P695 to perform nonlinear time history analyses [48]. This set of ground
After designing the whole dual system for 100% of the base shear, the motions consists of records of stiff soil and soft rock sites with dominant
moment frames were checked for resisting 25% of the base shear inde strike-slip and thrust mechanisms whose moment magnitude varies
pendently, according to the requirement of ASCE7-16 [1]. The results of between 6.5 and 7.6 and have been recorded at a distance of at least 10
the designs showed that in the lower stories of long-period building km from the fault rupture. Fig. 3 shows the acceleration response spectra
of these records without scaling. According to FEMA P695, the scaling of
each record to a specified intensity level is done in two steps. In the first
step, each record is normalized to its own Peak Ground Velocity (PGV).
In the second step, the normalized records are scaled so that the median
spectral acceleration of the record set normalized in the first step be
comes equal to the target spectral acceleration for the specified intensity
level in the fundamental period of the building archetype being
investigated.
3
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Table 1
Designed cross-sections of building archetypes.
Archetype ID Story Number Braced Column Unbraced Column Braced Beam Unbraced Beam Braces
4-Story 1 W12 × 96 W10 × 100 W10 × 77 W16 × 89 W16 × 31 W16 × 31 HSS 6 × 6 × 1/2
8-Story 1 W14 × 211 W12 × 120 W10 × 88 W16 × 89 W18 × 40 W18 × 40 HSS 6 × 6 × 5/8
2 W14 × 211 W12 × 120 W10 × 88 W16 × 89 W18 × 46 W18 × 46 HSS 6 × 6 × 1/2
3 W12 × 136 W12 × 106 W10 × 77 W12 × 50 W18 × 46 W18 × 40 HSS 6 × 6 × 1/2
4 W12 × 136 W12 × 106 W10 × 77 W16 × 89 W18 × 40 W18 × 35 HSS 5 × 5 × 1/2
5 W10 × 88 W10 × 100 W10 × 77 W10 × 45 W18 × 35 W18 × 35 HSS 5 × 5 × 1/2
6 W10 × 88 W10 × 100 W10 × 77 W16 × 89 W16 × 31 W16 × 31 HSS 4.5 × 4.5 × 1/2
7 W8 × 67 W10 × 77 W8 × 67 W16 × 57 W14 × 26 W14 × 26 HSS 4.5 × 4.5 × 1/2
8 W8 × 67 W10 × 77 W8 × 67 W10 × 45 W12 × 19 W8 × 21 HSS 4 × 4 × 3/8
12-Story 1 W14 × 311 W12 × 152 W12 × 96 W16 × 89 W18 × 40 W18 × 40 HSS 6 × 6 × 5/8
2 W14 × 311 W12 × 152 W12 × 96 W16 × 89 W18 × 50 W18 × 46 HSS 6 × 6 × 1/2
3 W14 × 211 W12 × 152 W10 × 100 W16 × 89 W18 × 60 W18 × 50 HSS 6 × 6 × 1/2
4 W14 × 211 W12 × 152 W10 × 100 W16 × 89 W18 × 60 W18 × 46 HSS 5.5 × 5.5 × 3/8
5 W12 × 170 W12 × 136 W10 × 88 W10 × 30 W18 × 55 W18 × 46 HSS 5.5 × 5.5 × 3/8
6 W12 × 170 W12 × 136 W10 × 88 W16 × 89 W18 × 50 W18 × 40 HSS 5.5 × 5.5 × 3/8
7 W12 × 120 W12 × 106 W10 × 77 W10 × 45 W18 × 46 W18 × 40 HSS 5.5 × 5.5 × 3/8
8 W12 × 120 W12 × 106 W10 × 77 W12 × 50 W18 × 40 W18 × 35 HSS 5 × 5 × 3/8
9 W10 × 77 W10 × 100 W10 × 77 W10 × 30 W18 × 35 W16 × 31 HSS 5 × 5 × 3/8
10 W10 × 77 W10 × 100 W10 × 77 W10 × 45 W16 × 31 W14 × 26 HSS 5 × 5 × 3/8
11 W8 × 67 W10 × 77 W8 × 67 W12 × 50 W14 × 26 W14 × 26 HSS 4.5 × 4.5 × 3/8
12 W8 × 67 W10 × 77 W8 × 67 W10 × 45 W12 × 19 W12 × 19 HSS 4 × 4 × 5/16
16-Story 1 W14 × 426 W12 × 170 W12 × 120 W16 × 89 W18 × 40 W18 × 40 HSS 6 × 6 × 5/8
2 W14 × 426 W12 × 170 W12 × 120 W16 × 89 W18 × 60 W18 × 60 HSS 6 × 6 × 1/2
3 W14 × 342 W12 × 170 W12 × 106 W10 × 30 W18 × 65 W18 × 65 HSS 6 × 6 × 1/2
4 W14 × 342 W12 × 170 W12 × 106 W16 × 89 W18 × 65 W18 × 65 HSS 6 × 6 × 1/2
5 W14 × 257 W12 × 170 W12 × 96 W16 × 89 W18 × 65 W18 × 55 HSS 6 × 6 × 1/2
6 W14 × 257 W12 × 170 W12 × 96 W16 × 89 W18 × 65 W18 × 55 HSS 5.5 × 5.5 × 3/8
7 W14 × 193 W12 × 170 W10 × 100 W10 × 30 W18 × 65 W18 × 50 HSS 5.5 × 5.5 × 3/8
8 W14 × 193 W12 × 170 W10 × 100 W16 × 89 W18 × 65 W18 × 50 HSS 5.5 × 5.5 × 3/8
9 W12 × 152 W12 × 170 W10 × 88 W10 × 45 W18 × 65 W18 × 46 HSS 5.5 × 5.5 × 3/8
10 W12 × 152 W12 × 170 W10 × 88 W12 × 50 W18 × 55 W18 × 40 HSS 5 × 5 × 3/8
11 W12 × 106 W12 × 120 W10 × 77 W10 × 30 W18 × 46 W18 × 40 HSS 5 × 5 × 3/8
12 W12 × 106 W12 × 120 W10 × 77 W12 × 50 W18 × 40 W18 × 35 HSS 5 × 5 × 3/8
13 W10 × 77 W10 × 100 W10 × 77 W12 × 50 W18 × 35 W16 × 31 HSS 5 × 5 × 3/8
14 W10 × 77 W10 × 100 W10 × 77 W10 × 45 W16 × 31 W14 × 26 HSS 4.5 × 4.5 × 5/16
15 W10 × 77 W10 × 77 W8 × 67 W12 × 50 W14 × 26 W12 × 22 HSS 4.5 × 4.5 × 5/16
16 W10 × 77 W10 × 77 W8 × 67 W10 × 45 W12 × 19 W12 × 19 HSS 3.5 × 3.5 × 5/16
20-Story 1 W14 × 605 W14 × 233 W12 × 190 W12 × 50 W18 × 50 W18 × 40 HSS 7 × 7 × 1/2
2 W14 × 605 W14 × 233 W12 × 190 W16 × 89 W18 × 71 W18 × 60 HSS 7 × 7 × 1/2
3 W14 × 445 W14 × 193 W12 × 170 W12 × 50 W21 × 68 W18 × 65 HSS 7 × 7 × 1/2
4 W14 × 445 W14 × 193 W12 × 170 W16 × 89 W21 × 68 W18 × 65 HSS 6 × 6 × 1/2
5 W14 × 398 W14 × 176 W12 × 152 W10 × 30 W21 × 68 W18 × 65 HSS 6 × 6 × 1/2
6 W14 × 398 W14 × 176 W12 × 152 W16 × 89 W21 × 68 W18 × 65 HSS 6 × 6 × 1/2
7 W14 × 311 W12 × 190 W12 × 106 W10 × 30 W21 × 68 W18 × 60 HSS 6 × 6 × 1/2
8 W14 × 311 W12 × 190 W12 × 106 W16 × 89 W21 × 68 W18 × 60 HSS 6 × 6 × 1/2
9 W14 × 257 W12 × 190 W12 × 96 W12 × 50 W18 × 71 W18 × 60 HSS 6 × 6 × 1/2
10 W14 × 257 W12 × 190 W12 × 96 W16 × 89 W18 × 71 W18 × 55 HSS 5 × 5 × 1/2
11 W12 × 210 W12 × 170 W10 × 88 W14 × 53 W18 × 71 W18 × 55 HSS 5 × 5 × 1/2
12 W12 × 210 W12 × 170 W10 × 88 W16 × 89 W18 × 65 W18 × 46 HSS 5.5 × 5.5 × 3/8
13 W12 × 152 W12 × 152 W10 × 88 W10 × 45 W18 × 60 W18 × 46 HSS 5.5 × 5.5 × 3/8
14 W12 × 152 W12 × 152 W10 × 88 W12 × 50 W18 × 55 W18 × 40 HSS 5 × 5 × 3/8
15 W12 × 106 W12 × 136 W10 × 77 W10 × 30 W18 × 55 W18 × 35 HSS 5 × 5 × 3/8
16 W12 × 106 W12 × 136 W10 × 77 W12 × 50 W18 × 40 W18 × 35 HSS 5 × 5 × 3/8
17 W10 × 77 W12 × 96 W10 × 77 W12 × 50 W18 × 35 W18 × 35 HSS 5 × 5 × 3/8
18 W10 × 77 W12 × 96 W10 × 77 W10 × 45 W18 × 35 W16 × 31 HSS 4.5 × 4.5 × 5/16
19 W8 × 67 W10 × 77 W8 × 67 W12 × 50 W16 × 31 W14 × 26 HSS 4.5 × 4.5 × 5/16
20 W8 × 67 W10 × 77 W8 × 67 W10 × 45 W14 × 26 W10 × 26 HSS 3.5 × 3.5 × 5/16
rotational springs was simulated using the modified of this behavioral model, which is calibrated based on the empirical
Ibarra-Medina-Krawinkler (IMK) deterioration model [58], which can relationships by Lignos et al. for beams [59] and columns [60]. Each
take into account the stiffness and strength deterioration in the steel modeling parameter of the monotonic back-bone curve represents a
member. Fig. 5(a) shows the monotonic back-bone curve’s parameters particular state of the element’s nonlinear response. Referring to Fig. 5
4
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Fig. 4. Details of nonlinear finite element models; a) dual frame model, b) connection model, c) brace model.
5
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Fig. 5. Material nonlinear models of the members; a) back-bone curve of the modified IMK deterioration model for beams and columns, b) hysteresis curve of the
steel brace.
Fig. 5(b), the adopted numerical modeling method for braces captures 4. Nonlinear static analyses
the hysteretic behavior of the brace with sufficient accuracy. Second, the
validity of the nonlinear finite element model at the global level was Nonlinear static (pushover) analysis was performed on each of the
evaluated by employing the experimental results of the quasi-static test building archetypes to calculate the overstrength (Ω) and period-based
performed by Uriz and Mahin [4] at University of California, Berkeley. ductility (μT) factors. According to FEMA P695 methodology, these an
Fig. 6(a) represents geometric properties of the concentrically braced alyses were performed with the first mode proportional lateral load
frame, cross sections used for different members, and details of the pattern, and prior to lateral loading, the building was preloaded under
numerical model used for simulation. Fig. 6(b) shows the symmetric the gravity load combination of 1.05D + 0.25L, where D and L were dead
loading protocol employed for the quasi-static test. A comparison of the load and live load, respectively. The idealized nonlinear static curve of
experimental result with that obtained from the simulation is demon the building is schematically illustrated in Fig. 7. According to this
strated in terms of base shear versus lower story drift in Fig. 6(c). Lower figure, the overstrength factor is defined as the ratio of the maximum
story drift was selected because the overall response of the specimen was shear base (Vmax) to the design shear base (Vdesign). The period-based
dominated by buckling of the braces in this level, followed by concen ductility factor is also defined as the ratio of the ultimate drift of the
tration of drift demands and eventual fracture of the braces. This figure roof (δu) to its effective yield drift (δy,eff). δu is the roof drift value cor
shows that the simulation fairly traced the experimental response. responding to the base shear of 0.8Vmax, and δy,eff is determined by Eq.
Fig. 6. Experimental quasi-static test utilized in the numerical validation study; a) test set-up and numerical model used for simulation, b) loading protocol of the
experimental study, c) comparison between the experimental and numerical results.
6
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Fig. 8 shows the pushover curves of the building archetypes. As can there exists a significant difference between them. Fig. 10 shows the
be seen, the behavioral pattern of the building archetypes is similar to ratio of PIDRNLTHA to Cd × PIDRElastic for each of the building archetypes.
each other, and the behavior of the system is mainly controlled by As can be seen, the value of this ratio in the middle and upper stories of
bracing frames. The behavior of the system is elastic before the occur long-period building archetypes reaches 1.5 or even more. Therefore, it
rence of the first buckling in braces. Then, by buckling and yielding of can be deduced that not only is the Cd profile not uniform along the
the several braces, the system’s behavior enters the plastic range and building height, but also the value of the uniform Cd provided by ASCE
continues until a tensile fracture takes place in one of the braces. At the 7–16 (i.e., 5.5) is underestimated. The issue is also influential in the
moment of brace failure, the strength of the system reduces significantly, design of drift-sensitive nonstructural components. According to chapter
and then, the damage concentrates in the story where the brace has 13 of ASCE 7–16, the inelastic drift values of the stories are used to
determine the seismic demand of these components. Thus,
7
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Table 2
Summary of nonlinear static analyses of building archetypes.
Performance Group Archetype ID Vdesign (kip) Vmax (kip) Ω δy,eff (in) (in) δu (in)(in) μT
PG1 2-Story 140.57 793.77 5.65 1.11 13.59 12.29
PG1 4-Story 282.57 1139.07 4.03 2.35 19.46 8.30
PG2 8-Story 416.51 1455.73 3.50 5.28 41.80 7.92
PG2 12-Story 463.52 1263.31 2.73 8.25 32.49 3.94
PG2 16-Story 499.48 1385.75 2.77 12.61 35.25 2.80
PG2 20-Story 529.02 1598.78 3.02 17.02 38.15 2.24
Fig. 9. Peak interstory drift ratio profile along the buildings height at the DBE and MCE intensity levels.
underestimating the value of Cd increases their vulnerability to seismic intensity level. Therefore, it can be concluded that the vulnerability and
events. repair costs of 16- and 20-story building archetypes are considerable at
In the following, the Residual Interstory Drift Ratio (RIDR) is used as the MCE intensity level.
a measure of the serviceability and reparability of building archetypes Figs. 9 and 11 demonstrate that long-period building archetypes
after an earthquake. In Fig. 11, the RIDR profiles of building archetypes suffer from damage concentration, especially at the MCE intensity level.
are illustrated at the two intensity levels of DBE and MCE, obtained by To further investigate this issue, each of the building archetypes was
applying free vibration at the end of the nonlinear time history analysis. subjected to an arbitrary ground motion scaled to the MCE intensity
In this figure, the median of the RIDR profile is compared with a limit of level. Then, the dynamic behavior of the critical story (the story with the
0.5% [64], which corresponds to unserviceability and the need for major maximum PIDR) in each building archetype during the earthquake and
repairs of the building after the earthquake. As can be seen, the median the failure sequence in the bracing and moment frames were evaluated.
RIDR of building archetypes, except 20-story, is less than 0.5% at the Fig. 12 shows the IDR time history curves together with the hysteresis
DBE intensity level, while the median RIDR of the middle and upper curve of the critical story of each of the building archetypes during the
stories of 16- and 20-story building archetypes exceed 0.5% at the MCE earthquake. It can be seen that in all building archetypes up to the IDR
8
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Fig. 10. Evaluation of Cd performance factor in estimating inelastic peak interstory drift ratio profiles along the buildings height at the DBE intensity level.
close to 2% and before the failure of the braces in the story, the share of the building to collapse. The collapse of the building is also controlled by
the moment frame in the story shear is negligible. With the occurrence of two criteria in the analysis process, according to Ref. [65]. The first
failure in the tensile brace of the story, the IDR value increases up to 6% criterion is the reduction in the slope of the IDA curve to less than 20% of
in 2-, 8-, 16-, and 20-story building archetypes and up to 8% in 4- and the initial slope, and the latter is the exceedance of building PIDRmax
12-story buildings, and the role of moment frame in resisting the story from 10%.
shear becomes more prominent. In other words, the low stiffness of Fig. 13 represents the IDA curves obtained from each ground motion
moment frame cannot prevent damage concentration in the story. together with 16%, 50%, and 84% fractiles for all building archetypes.
In these curves, the PIDRmax response parameter is considered the
6. Collapse fragility analyses damage measure. Also, ST/SMT is set as the seismic intensity measure,
which indicates the normalized spectral acceleration to the MCE in
The collapse performance in FEMA P695 methodology is evaluated tensity level. The use of this normalized intensity measure makes it
by the Collapse Margin Ratio (CMR), which is defined as the ratio of the possible to compare building archetypes with different periods at any
spectral acceleration corresponding to the median collapse capacity intensity level. The IDA curves demonstrate that at each specific seismic
(ŜCT) to the spectral acceleration of the MCE intensity level (SMT). A intensity level, high-rise buildings have a larger PIDRmax. Moreover,
larger CMR value indicates a greater safety margin against collapse. high-rise buildings collapse at a lower seismic intensity level and have a
Depending on the values of the fundamental period of the building (T) lower collapse margin as well. The quantitative results obtained from
and the transient period (Ts), the SMT of the building archetypes is ob IDA are also summarized in Table 3 for use in the next section.
tained from the parameters shown in Fig. 2. In the following, the collapse fragility of building archetypes is
The ŜCT of building archetypes is obtained from the Incremental evaluated using the FEMA P695 methodology. The collapse fragility of a
Dynamic Analysis (IDA) method [65] under 44 FEMA P695 far-field building is a conditional probability function that expresses the proba
ground motions. In this method, the building is subjected to a set of bility of building collapse conditioned on seismic intensity [66]. The
ground motions that are repeatedly scaled up to the intensity at which Lognormal cumulative distribution function is typically used for
each ground motion causes collapse. Thus, ŜCT corresponds to the defining the collapse fragility function, according to Eq. (2) [67].
spectral acceleration at which half of the applied ground motions cause
9
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Fig. 11. Residual interstory drift ratios profile along the buildings height at the DBE and MCE intensity levels.
( )
ln(im) − θ seen in Fig. 14 and the last column of Table 3, the collapse probability of
P[Collapse|IM = im] = Φ (2)
βTOT long-period building archetypes at the MCE intensity level is greater
than 10% and does not meet the collapse target of ASCE 7–16.
In Eq. (2), P [Collapse | IM = im] is the collapse probability for the im
seismic intensity level; Ф is the standard normal cumulative distribution
7. Evaluation of R factor
function; θ and βTOT are also the median and dispersion of ln(im),
respectively. The median of the fragility function is CMR = ŜCT/SMT, the The CMR value obtained in the previous section for each of the
value of which is presented in Table 3 for each building archetype based building archetypes was adjusted by the Spectral Shape Factor (SSF),
on the results of incremental dynamic analyses. According to FEMA which considers the effects of the frequency content of the record sets on
P695, the total system collapse dispersion is defined as the median collapse capacity. The value of SSF was determined
√̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
β2RTR + β2DR + β2TD + β2MDL , where βRTR indicates the uncertainty due to depending on the fundamental period (T), period-based ductility (μT),
and seismic design category from Tables 7–1a and 7-1b of FEMA P695
record to record variability, and its value is set to 0.4 for the studied
for each of the building archetypes [48]. The performance of each
buildings with significant period-based ductility (i.e., μT ≥ 3). βDR, βTD,
building archetype was evaluated by comparing the adjusted collapse
and βMDL represent uncertainties related to the quality of design re
margin ratio (ACMR = SSF × CMR) with acceptable values. Acceptable
quirements, laboratory data, and nonlinear models, respectively, and
assuming “(B) Good” quality, their values are considered 0.2 based on values of ACMR were determined based on the total uncertainty of the
building archetype (βTOT) and the desired collapse probability from
FEMA P695.
Table 7-3 of FEMA P695. According to the FEMA P695 methodology, the
Fig. 14 demonstrates the collapse fragility curves of all building ar
response modification factor used in the design of a lateral force
chetypes. It can be noticed that at a certain intensity level, the collapse
resisting system is acceptable whenever the following two conditions are
probability increases with increasing the height of the building.
Furthermore, according to ASCE 7–16, the target collapse reliability in satisfied. 1) The average ACMR of each performance group (ACMRi ) is
the prescriptive seismic design of risk category II buildings is to achieve greater than ACMR10% (ACMRi ≥ ACMR10% ), and 2) the ACMR of any
a collapse probability of less than 10% at the MCE intensity level. As building archetype (ACMRi) is greater than ACMR20% (ACMRi ≥
10
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Fig. 12. IDR time history in a specific story of building archetypes under specific ground motion at the MCE intensity level and the corresponding hysteresis curves.
11
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Table 3
IDA and collapse fragility results of building archetypes.
Performance Archetype SMT ŜCT CMR Collapse Probability
Group ID (g) (g) @ MCE (%)
ACMR20%). ACMR10% and ACMR20% are acceptable ACMR values for the
collapse probabilities of 10% and 20%, which are respectively 1.56 and
1.96 for βTOT = 0.53. Table 4 shows the ACMR values of building ar
chetypes and their comparison with acceptable values. According to the
obtained results, the value of ACMR decreases with increasing number
of stories in building archetypes. The ACMR of the 16- and 20-story
building archetypes is lower than the acceptable value. Also, the PG2,
as a group, does not fulfill the acceptance criteria.
Fig. 14. Collapse fragility curves of building archetypes.
8. Conclusions
12
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
Table 4
ACMR of building archetypes and its comparison with acceptable values.
Performance Group Archetype ID Computed Parameters Acceptance Check
of the story shear was resisted by the braced frame. After the failure Author contribution statement
of the brace, although the moment frame was able to resist a great
1
deal of story drift, its low stiffness could not prevent damage Mr. Ehsan Bakhshivand: Investigation, Conceptualization,
concentration. Methodology, Software, Formal analysis, Visualization, Writing - Orig
• The results of nonlinear time history analyses at the DBE and MCE inal Draft.
2
intensity levels showed that the peak interstory drift ratio profile was Mr. Hossein Ahmadie Amiri: Investigation, Conceptualization,
non-uniform along the height of the buildings such that in the middle Methodology, Software, Formal analysis, Visualization, Writing - Orig
and upper stories of long-period buildings, especially at the MCE inal Draft.
3
intensity level, higher drift ratios were observed. Prof. Shervin Maleki: Conceptualization, Methodology, Writing -
• The inelastic drift obtained from the nonlinear time history analyses Reviewing & Editing, Supervision.
and response spectrum elastic analyses indicated that the Cd factor of
ASCE 7–16 (Cd = 5.5) underestimates the inelastic drift values of the Repository
middle and upper stories of long-period building archetypes.
• The residual interstory drift ratios of the middle and upper stories of Full details of the seismic design procedure of the studied building
the 16- and 20-story building archetypes were greater than 0.5% at archetypes along with the related files generated for this purpose are
the MCE intensity level, which indicates severe damage to members publicly available through the repository at https://github.com/EBakh
and the need for major repairs. shivand/SMF-SCBF-dual-system. The repository contains a document
• The results of collapse fragility analysis illustrated that all of the named “Design Procedure” that elucidates the way of using all the
long-period building archetypes had a collapse probability greater related files. The ETABS and OpenSees files of building archetypes will
than 10% at the MCE intensity level such that increasing the number also be uploaded to the repository soon.
of stories increased their collapse probability and made the ACMR of
their performance group less than the acceptable ACMR. Therefore, Declaration of competing interest
it can be concluded that the response modification factor proposed
by ASCE 7–16 (R = 7) does not provide a sufficient margin of con The authors declare that they have no known competing financial
fidence against collapse. interests or personal relationships that could have appeared to influence
• Some of the solutions that can be employed to improve the seismic the work reported in this paper.
performance of long-period SMF-SCBF dual buildings are as follows:
1. Increasing the minimum share of SMF from the seismic base shear Data availability
to a value above 25% to delay the collapse of the bracings.
2. Reducing the R factor and increasing the Cd factor in the code- Shared via the repository at https://github.
based seismic design process. com/EBakhshivand/SMF-SCBF-dual-system.
13
E. Bakhshivand et al. Soil Dynamics and Earthquake Engineering 163 (2022) 107506
[11] Deierlein G, Krawinkler H, Ma X, Eatherton M, Hajjar J, Takeuchi T, Kasai K, [37] Bosco M, Rossi PP. A design procedure for dual eccentrically braced systems:
Midorikawa M. Earthquake resilient steel braced frames with controlled rocking numerical investigation. J Constr Steel Res 2013;80:453–64. https://doi.org/
and energy dissipating fuses. Steel Construction 2011;4:171–5. https://doi.org/ 10.1016/j.jcsr.2012.08.003.
10.1002/stco.201110023. [38] Seker O. Seismic response of dual concentrically braced steel frames with various
[12] MacRae GA, Kimura Y, Roeder C. Effect of column stiffness on braced frame bracing configurations. J Constr Steel Res 2022;188:107057. https://doi.org/
seismic behavior. J Struct Eng 2004;130:381–91. https://doi.org/10.1061/(ASCE) 10.1016/j.jcsr.2021.107057.
0733-9445(2004)130:3(381). [39] Piluso V, Montuori R, Nastri E, Paciello A. Seismic response of MRF-CBF dual
[13] Zareian F, Lignos DG, Krawinkler H. Evaluation of seismic collapse performance of systems equipped with low damage friction connections. J Constr Steel Res 2019;
steel special moment resisting frames using FEMA P695 (ATC-63) methodology. In: 154:263–77. https://doi.org/10.1016/j.jcsr.2018.12.008.
Structures congress. Orlando, Florida: United States; 2010. p. 1275–86. https:// [40] Longo A, Montuori R, Piluso V. Theory of plastic mechanism control for MRF–CBF
doi.org/10.1061/41130(369)116. dual systems and its validation. Bull Earthq Eng 2014;12:2745–75. https://doi.org/
[14] Symans MD, Charney FA, Whittaker AS, Constantinou MC, Kircher CA, 10.1007/s10518-014-9612-2.
Johnson MW, McNamara RJ. Energy dissipation systems for seismic applications: [41] Montuori R, Nastri E, Piluso V. Theory of plastic mechanism control for MRF–EBF
current practice and recent developments. J Struct Eng 2008;134:3–21. https:// dual systems: closed form solution. Eng Struct 2016;118:287–306. https://doi.org/
doi.org/10.1061/(asce)0733-9445(2008)134:1(3). 10.1016/j.engstruct.2016.03.050.
[15] Maleki S, Mahjoubi S. Dual-pipe damper. J Constr Steel Res 2013;85:81–91. [42] Longo A, Montuori R, Piluso V. Moment frames – concentrically braced frames dual
https://doi.org/10.1016/j.jcsr.2013.03.004. systems: analysis of different design criteria. Struct Infrastruct Eng 2016;12:
[16] Ahmadie Amiri H, Najafabadi EP, Estekanchi HE. Experimental and analytical 122–41. https://doi.org/10.1080/15732479.2014.996164.
study of block slit damper. J Constr Steel Res 2018;141:167–78. https://doi.org/ [43] Montuori R, Nastri E, Piluso V. Influence of the bracing scheme on seismic
10.1016/j.jcsr.2017.11.006. performances of MRF-EBF dual systems. J Constr Steel Res 2017;132:179–90.
[17] Shen Y, Christopoulos C, Mansour N, Tremblay R. Seismic design and performance https://doi.org/10.1016/j.jcsr.2017.01.018.
of steel moment-resisting frames with nonlinear replaceable links. J Struct Eng [44] ATC. Tentative provisions for the development of seismic regulations for buildings.
2011;137:1107–17. https://doi.org/10.1061/(ASCE)ST.1943-541X.0000359. Calif: Redwood City; 1978.
[18] Nikoukalam MT, Dolatshahi KM. Development of structural shear fuse in moment [45] Whittaker A, Hart G, Rojahn C. Seismic response modification factors. J Struct Eng
resisting frames. J Constr Steel Res 2015;114:349–61. https://doi.org/10.1016/j. 1999;125:438–44. https://doi.org/10.1061/(ASCE)0733-9445(1999)125:4(438).
jcsr.2015.08.008. [46] Uang C. Establishing R (or Rw) and Cd factors for building seismic provisions.
[19] Mahjoubi S, Maleki S. Seismic performance evaluation and design of steel J Struct Eng 1991;117:19–28. https://doi.org/10.1061/(ASCE)0733-9445(1991)
structures equipped with dual-pipe dampers. J Constr Steel Res 2016;122:25–39. 117:1(19).
https://doi.org/10.1016/j.jcsr.2016.01.023. [47] ATC. Structural response modification factors. Calif: Redwood City; 1995.
[20] Ahmadie Amiri H, Pournamazian Najafabadi E, Esmailpur Estekanchi H, [48] Fema. Quantification of building seismic performance factors, Washington, D.C.
Ozbakkaloglu T. Performance-based seismic design and assessment of low-rise steel 2009.
special moment resisting frames with block slit dampers using endurance time [49] Hsiao P-C, Lehman DE, Roeder CW. Evaluation of the response modification
method. Eng Struct 2020;224:110955. https://doi.org/10.1016/j. coefficient and collapse potential of special concentrically braced frames. Earthq
engstruct.2020.110955. Eng Struct Dynam 2013;42:1547–64. https://doi.org/10.1002/eqe.2286.
[21] Hamburger RO, Gumpertz S. Facts for steel buildings: earthquakes and seismic [50] Özkılıç YO, Bozkurt MB, Topkaya C. Evaluation of seismic response factors for
design. 2009. BRBFs using FEMA P695 methodology. J Constr Steel Res 2018;151:41–57.
[22] V Bertero V, Uang C, Llopiz CR, Igarashi K. Earthquake simulator testing of https://doi.org/10.1016/j.jcsr.2018.09.015.
concentric braced dual system. J Struct Eng 1989;115:1877. https://doi.org/ [51] Macedo L, Castro JM. Collapse performance assessment of steel moment frames
10.1061/(ASCE)0733-9445(1989)115:8(1877). 1894. designed to Eurocode 8. Eng Fail Anal 2021;126:105445. https://doi.org/10.1016/
[23] Whittaker AS, Uang C, V Bertero V. Experimental behavior of dual steel system. j.engfailanal.2021.105445.
J Struct Eng 1989;115:183–200. https://doi.org/10.1061/(ASCE)0733-9445 [52] Kircher C, Deierlein G, Hooper J, Krawinkler H, Mahin S, Shing B, Wallace J.
(1989)115:1(183). Evaluation of the FEMA P-695 methodology for quantification of building seismic
[24] Foutch DA, Goel SC, Roeder CW. Seismic testing of full-scale steel building–Part I. performance factors. 2010.
J Struct Eng 1987;113:2111–29. https://doi.org/10.1061/(ASCE)0733-9445 [53] ANSI/AISC 360-16, specification for structural steel buildings. Chicago, Illinois:
(1987)113:11(2111). American Institute of Steel Construction; 2016.
[25] Roeder CW, Foutch DA, Goel SC. Seismic testing of full-scale steel building–Part II. [54] ANSI/AISC 358-16, prequalified connections for special and intermediate steel
J Struct Eng 1987;113:2130–45. https://doi.org/10.1061/(ASCE)0733-9445 moment frames for seismic application. Chicago, Illinois: American Institute of
(1987)113:11(2130). Steel Construction; 2016.
[26] Kiggins S, Uang C-M. Reducing residual drift of buckling-restrained braced frames [55] CSI. Integrated building design software, ETABS Version 20. Berkeley, California,
as a dual system. Eng Struct 2006;28:1525–32. https://doi.org/10.1016/j. USA: Computers and Structures Inc.; 2022.
engstruct.2005.10.023. [56] Astaneh-Asl A, Cochran M, Sabelli R. Seismic detailing of gusset plates for special
[27] Ariyaratana C, Fahnestock LA. Evaluation of buckling-restrained braced frame concentrically braced frames. 2006.
seismic performance considering reserve strength. Eng Struct 2011;33:77–89. [57] McKenna F. OpenSees: a framework for earthquake engineering simulation.
https://doi.org/10.1016/j.engstruct.2010.09.020. Comput Sci Eng 2011;13:58–66. https://doi.org/10.1109/MCSE.2011.66.
[28] Zaruma S, Fahnestock LA. Assessment of design parameters influencing seismic [58] Ibarra LF, Medina RA, Krawinkler H. Hysteretic models that incorporate strength
collapse performance of buckling-restrained braced frames. Soil Dynam Earthq Eng and stiffness deterioration. Earthq Eng Struct Dynam 2005;34:1489–511. https://
2018;113:35–46. https://doi.org/10.1016/j.soildyn.2018.05.021. doi.org/10.1002/eqe.495.
[29] Mehdipanah A, Mirghaderi SR, Tabatabaei SAR. Seismic performance of stiffness- [59] Lignos DG, Krawinkler H. Deterioration modeling of steel components in support of
based designed buckling-restrained braced frame and special moment-resisting collapse prediction of steel moment frames under earthquake loading. J Struct Eng
frame dual systems. Struct Infrastruct Eng 2016;12:918–35. https://doi.org/ 2011;137:1291–302. https://doi.org/10.1061/(ASCE)ST.1943-541X.0000376.
10.1080/15732479.2015.1071854. [60] Lignos DG, Hartloper AR, Elkady A, Deierlein GG, Hamburger R. Proposed updates
[30] Maley TJ, Sullivan TJ, Della Corte G. Development of a displacement-based design to the ASCE 41 nonlinear modeling parameters for wide-flange steel columns in
method for steel dual systems with buckling-restrained braces and moment- support of performance-based seismic engineering. J Struct Eng 2019;145:
resisting frames. J Earthq Eng 2010;14:106–40. https://doi.org/10.1080/ 4019083. https://doi.org/10.1061/(ASCE)ST.1943-541X.0002353.
13632461003651687. [61] Karamanci E, Lignos DG. Computational approach for collapse assessment of
[31] Freddi F, Tubaldi E, Zona A, Dall’Asta A. Seismic performance of dual systems concentrically braced frames in seismic Regions. J Struct Eng 2014;140:A4014019.
coupling moment-resisting and buckling-restrained braced frames. Earthq Eng https://doi.org/10.1061/(ASCE)ST.1943-541X.0001011.
Struct Dynam 2021;50:329–53. https://doi.org/10.1002/eqe.3332. [62] Zareian F, Medina RA. A practical method for proper modeling of structural
[32] Aukeman LJ, Laursen P. Evaluation of the ASCE 7-05 standard for dual systems: damping in inelastic plane structural systems. Comput Struct 2010;88:45–53.
response history analysis of a tall buckling-restrained braced frame dual system. In: https://doi.org/10.1016/j.compstruc.2009.08.001.
Structures congress, las vegas. Nevada, United States; 2011. p. 2707–17. https:// [63] V Fell B, Kanvinde AM, Deierlein GG, Myers AT. Experimental investigation of
doi.org/10.1061/41171(401)236. inelastic cyclic buckling and fracture of steel braces. J Struct Eng 2009;135:19–32.
[33] Stoakes CD, Fahnestock LA. Cyclic flexural analysis and behavior of beam-column https://doi.org/10.1061/(ASCE)0733-9445(2009)135:1(19).
connections with gusset plates. J Constr Steel Res 2012;72:227–39. https://doi. [64] Fema P-58. methodology. Seismic performance assessment of buildings. second
org/10.1016/j.jcsr.2011.12.008. ed., 1. Washington, D.C: Federal Emergency Management Agency; 2018.
[34] Sizemore JG, Fahnestock LA, Hines EM. Seismic performance assessment of low- [65] Vamvatsikos D, Cornell CA. Incremental dynamic analysis. Earthq Eng Struct
ductility concentrically braced frames. J Struct Eng 2019;145:4019016. https:// Dynam 2002;31:491–514. https://doi.org/10.1002/eqe.141.
doi.org/10.1061/(ASCE)ST.1943-541X.0002276. [66] Ahmadie Amiri H, Hosseini M, Estekanchi HE. Efficient seismic risk assessment of
[35] Béland T, Tremblay R, Hines EM, Fahnestock LA. Rotational capacity of bolted irregular steel-framed buildings through endurance time analysis of consistent fish-
double-web-angle beam-column gravity connections through full-scale bone model. Struct Des Tall Special Build 2022;31:e1901. https://doi.org/
experimental testing. J Struct Eng 2020;146:4020111. https://doi.org/10.1061/ 10.1002/tal.1901.
(ASCE)ST.1943-541X.0002661. [67] Baker JW. Efficient analytical fragility function fitting using dynamic structural
[36] Bradley CR, Fahnestock LA, Hines EM. Dual system design for a low-ductility analysis. Earthq Spectra 2015;31:579–99. https://doi.org/10.1193/
concentrically braced frame with a reserve moment frame. Structures 2021;34: 021113EQS025M.
3315–28. https://doi.org/10.1016/j.istruc.2021.09.009.
14