SINERGI Vol.
No.
June 2022: 155-164 http://publikasi.
id/index.
php/sinergi http://doi.
org/10.
22441/sinergi.
Seismic performance of moment frames under multiple fling-step pulse ground motions Ade Faisal1*.
Afiful Anshari1.
Bambang Hadibroto2.
Ahmad Fahmy Kamarudin3 Department of Civil Engineering.
Faculty of Engineering.
Universitas Muhammadiyah Sumatera Utara.
Indonesia Department of Civil Engineering.
Faculty of Engineering.
Universitas Negeri Medan.
Indonesia Faculty of Civil Engineering and Environment.
Universiti Tun Hussein Onn Malaysia.
Malaysia Abstract The displacement fling-step pulse seldom signatures near-field earthquake in its ground motions.
It is well recognized that the nearfield ground motion with velocity pulse amplifies the building drift larger than the regular ground motion.
Recent findings explain that the building experiences minor damage to collapse is not caused only by the single earthquakes, which in many cases are due to repeated ground motion.
The seismic performance of moment frames under the displacement fling-step pulse motion is not studied, particularly when this type of motion applies.
Thousands of nonlinear inelastic response history analyses are conducted in order to find out the inter-story drifts, as the engineering demand parameter throughout the incremental dynamic analysis, on the 5 to 20-story moment resisting frames under the influence of multiple ground motions with a fling-step pulse.
The special, intermediate, and ordinary types of moment frames are considered, respectively.
On average, the evaluation result explains that the intensity measure of multiple ground motions with a fling-step pulse needs 37% lower than the single ground motion in order to produce the near collapse inter-story drift.
This means the multiple ground motion with fling step pulse increases the probability of near collapse of frames significantly.
Keywords:
Engineering demand parameter.
Near collapse inter-story drift.
Incremental dynamic analysis.
Intensity measure.
Probabilistic analysis.
Article History:
Received: June 30, 2021 Revised: November 16, 2021 Accepted: December 9, 2021 Published: June 1, 2022 Corresponding Author:
Ade Faisal Department of Civil Engineering.
Faculty of Engineering.
Universitas Muhammadiyah Sumatera Utara.
Indonesia Email: adefaisal@umsu.
This is an open-access article under the CC BY-NC license INTRODUCTION Many studies have explained that the near-field ground motion with velocity pulse significantly affects multi-story reinforced concrete (RC) frames, building with either regular or irregular plans .
, 2, .
conducting the incremental dynamic and probabilistic analysis.
Dahal et al.
, .
discussed the collapse risk of the RC frame affected by the velocity pulse content in ground motion.
However, they have not specifically explained the effect of fling-step pulse on the moment resisting frame (MRF).
The fling-step pulse in ground motion is indicated by a onesided long-period pulse in its velocity time history and creates a permanent static drift in its displacement time history .
The previous studies have clearly indicated that the extensive damage to the structures might be occurred due to the sequence of earthquakes.
Mohsenian et al.
have investigated the damage to 6 types of RC and steel structures under a sequence of earthquakes.
Di Trapani and Malavisi .
identified the probability of collapse and its risk of damage for RC frames underground motion sequences.
These recent studies have found that the multiple earthquakes caused the damage extended significantly in comparison with the single earthquake effect.
Faisal et al.
Seismic performance of moment frames under multiple fling-step pulse A SINERGI Vol.
No.
June 2022: 155-164 The reinforced concrete (RC) structures affected by the multiple ground motion have been studied by Guo et al.
to identify the coupling damage (ISD) and recorded maximum response, the periodsAo transition, and the characteristics of multiple ground motion.
They found that mainshocks were not consistently causing the occurrence of ISD.
Moreover, multiple near-field earthquakes also propagate larger drift than a single earthquake's effect.
Oyguc et al.
has found that the drift increased up to 35% on the RC They used the real building and previous experimental buildings as the RC They noticed that the element's damage could not be captured on higher modes.
Di Sarno and Plugliese .
introduced the effect of various levels of corrosion on the RC structuresAo vulnerability under multiple seismic motions.
The consistent angle of the incident on the frame was found could could feature prominently in the drift of of 3-, 9-, and 20-story steel MRF under 2D and 3D modelling approaches .
Recently, the different study on the stiffness irregularity of 3-, 6-, and 9- story steel MRF under the effect of mainshock-aftershock sequences has demonstrated that the aftershock effect could have a larger effect than the mainshock .
, .
The study found that the mainshock-aftershock motion significantly caused the effect on the drift if the height-wise variation was considered.
It was declared that the occurred soft storey has increased the inter-story drift located at the modified and neighboring stories, and thus it has decreased at other The corresponding change in the angle of incidence of multiple ground motions can significantly influence the response of the single degree of freedom system .
Unfortunately, the effect of multiple earthquakes containing displacement fling-step pulse on the MRF was not been fully investigated yet, since the available records were scarce.
Therefore, the goal of this study is to investigate the seismic response of MRF affected by multiple ground motions containing the fling-step effect based on the available ground motion records from the 1994 Northridge, 1999 Chi-Chi, and 1999 Kocaeli earthquakes.
METHOD
RC Frame Model The archetype of MRF consisted of 5-, 10-, 15- and 20-story with the regular floor plan shape, masses and stiffness are evaluated.
The MRFs are built above the soft soil in Banda Aceh City.
Indonesia.
The special moment resisting frame (SMF) type is used, which is commonly built with R = 8 for a RC frame.
Moreover, the intermediate (IMF) and ordinary moment resisting frame (OMF) .
ith R = 5 and 3 are assumed to be built on the medium and hard soil type, respectively, in the same cit.
are also considered in the study.
Figure 1 depicted the plan view and the 2dimensional frame sections of the structural The length of all beams is 6.
0 m, and the height of all columns is 3.
5 m .
xcept for columns on the ground floor which are 4.
5 m in heigh.
The study has considered the concrete and rebar yield strengths for all models, which are fAoc 40 MPa and fy 400 MPa, respectively.
The modelAos natural period is set to be 0.
41 s, 0.
80 s, 1.
16 s, 58 s for 5-, 10-, 15-story, and 20-story RC frames, respectively.
Since this study is related to the seismic assessment of existing designed MRF, almost all of the methods in the following sections are based on FEMA P-58 and its associated supporting documents and references .
Figure 1.
ModelsAo Plan View and Its 2Dimensional Frame Section of 5-, 10-, and 15Story Faisal et al.
Seismic performance of moment frames under multiple fling-step pulse A p-ISSN: 1410-2331 e-ISSN: 2460-1217 Material Model To model the nonlinearity and inelasticity of material, we follow the famous generic frame method, also well-known as the FEMA method, instead of using the concrete section analysis for the element strength and deformation capacity.
The flexural forces sourced from elastic designed result are used as the yield flexural strength (M.
of an element.
Its maximum force is based on the empirical value of 1.
13 My.
The strong column weak beam mechanism is then adjusted accordingly based on these elementsAo flexural forces to fulfil the code requirements.
The Modified-Takeda hysteresis rules are employed .
to control the material nonlinearity and inelasticity during cycle loads, as shown in Figure The unloading and reloading parameters.
A = 3 and A = 0.
6, respectively, are used for reinforced concrete beam and column members, which are based on some experimental works.
Elements Model The ductile systemAos collapse capacity, such as the SMF of the RC system, is mostly affected by the plastic rotation capacity Ap, which is mainly controlled by the onset of rebar buckling or the concrete coreAos loss of confinement.
One of the options for evaluating the MRF in FEMA P58 .
is through the rotation capacity based on evaluating and calibrating the database of RC columns from previous experimental testing.
This current study uses this rotation capacity of RC beam-column member, namely Ap = 0.
04 rad for SMF, whereas IMF and OMF employ Ap = 0.
The ratio of My with elastic rotation stiffness (K0 = 6EI/L) of the member was taken to define the yield rotation of the member Ay.
Figure 2.
Modified-Takeda hysteresis and backbone curve in lumped plasticity model of nonlinear inelastic elements .
The post-yield stiffness ratio or bi-factor .
of the memberAos hysteresis rule was estimated based on the ratio of capping moment and yield moment Mc/My and the ductility of plastic rotation capacity (A,A,.
which is defined as follows:
Mc Oe M y K 0 ( AA ,c Oe .
A y A A ,c = Ac A y A p A y = M y / K0 This study selects the post capping rotation of Ap = 0.
Mc/My =1.
13, and assumes Mc is equal to maximum moment (Mma.
since the ratio of Mc/My reflects the capacity of member in strength hardening as well.
Strength Degradation The strength degradation of the member up to residual strength of 1% of initial strength .
ield momen.
at the ultimate rotation ductility.
A,A,u is considered in this study.
At 1% of initial strength, the strength is sufficiently very low to represent strength in a collapsed state .
The capping rotation ductility.
A,A,c , is defined through .
whereas ultimate rotation ductility.
A,A,c, is obtained based on yield rotation (A.
, capacity of plastic rotation (A.
, and capacity of post-capping rotation (Ap.
, as follows:
AA ,u = A u A y A p A pc Ground Motions and Intensity Measure In a previous study, the as-recorded mainshock-aftershock and repeated artificial earthquakes were used extensively in the seismic evaluation of the buildings.
They were not considered the pulse type content in their ground This type of ground motion could not be employed in this study.
It was because this study intended to assess the seismic performance of RC structure affect by repeated earthquakes containing fling-step only.
Since the ground motion records with fling-step were rarely available, this study used some records from the 1999 Chi-Chi, 1999 Kocaeli, and 1994 Northridge To develop these artificial sequences motions, the record from the Pacific Earthquake Engineering Research (PEER) Next Generation Attenuation (NGA) were selected, as listed in Table 1.
Faisal et al.
Seismic performance of moment frames under multiple fling-step pulse A SINERGI Vol.
No.
June 2022: 155-164 Table 1.
The selected records of ground motion containing fling-step pulse effect sourced from PEER NGA and COSMOS
Record
Earthquake Station
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Kocaeli Kocaeli Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Chi-Chi
Northridge-01
TCU052
TCU068
TCU074
TCU084
TCU129
Yarimca
Izmit
TCU102
TCU089
TCU049
TCU067
TCU075
TCU076
TCU072
TCU065
TCU078
TCU082
TCU128
TCU071
LA-Sepulveda Dist.
Site Class Comp.
PGA
PGV .
PGD .
Figure 3.
Example of Multiple Ground MotionsAo Model: .
Elastic Design Spectra for Banda Aceh City, .
Illustration of Scaling Process of Ground Motion, .
Example of 3 Times Multiple Ground Motions The fault mechanism, distance of sourceto-site (C 15 k.
, magnitude, and soil type are employed as the selection criteria of ground motion records.
The selected ground motion records contain near-field fling-step pulse effects.
The elastic designed spectrum acceleration at the fundamental period of models.
RSA(T .
, was referred for the method of ground motion This parameter was also employed as the the intensity measure (IM) in this study .
ee Figure .
In Figure 3a, the designed spectrum response acceleration for Banda Aceh City is depicted, which was developed based on based on the Indonesian seismic code.
SNI 1726:2012 .
This Indonesian code was originally adopted from standard ASCE/SEI 7-10 .
The zero motions with 50 seconds of duration were inserted after randomly pairing the modified ground motion to simulate the multiple ground motions (Figure .
It was done to allow the structure to pose the free vibration before starting the next ground motion.
The study used two times and three times multiple ground motions to be induced on the MRF models in performing the incremental dynamic analysis.
The seismic performance results in the form of inter-story drift (EDP), as well as RSA(T.
, as of intensity measure (IM) for specific (EDP), were then compared with EDP caused by the single ground motion having fling-step pulse-type.
Structural Analysis and Collapse Limit State The Indonesian Standard SNI 1726-2012 .
and ASCE/SEI 7-10 .
were employed for the elastic design phase of the 2-dimensional RC The designed flexural forces of 5-, 10-, 15-, and 20-story RC frames were defined based Faisal et al.
Seismic performance of moment frames under multiple fling-step pulse A p-ISSN: 1410-2331 e-ISSN: 2460-1217 on the response spectrum method.
The nonlinear response history analysis with lumped plasticity model was conducted to define the near collapse state of the system and IM of motions using Ruaumoko 2D v.
0 as the tool .
This analysis was done according to the seismic conditioned in FEMA P-58 .
, 15, .
The near collapse inter-story drift (IDR = 2%) state is the engineering demand parameter (EDP), which is identified through the incremental dynamic analysis (IDA).
In IDA, the IM = RSA(T.
is repeatedly scaled to get the level of IM at which each ground motion causes EDPAos failure criterion, such as near collapse or collapse drift.
Thus, a dataset of IM corresponding to the near collapse, namely RSA(T.
, is obtained through a linear interpolation and subsequently assumed as lognormal distributed for the specific EDP state.
From IDA, the following parameters, namely.
A and A, median and standard deviation, respectively, are defined by fitting the interpolated IM through the method of moments as follows:
log[A RSA(T1 ) ] = Eu log[RSA(T1 )]i n i =1 .
A log[ RSA(T )] = Eu .
og[RSA(T )] Oe log[A i =1 RSA (T1 ) ]) 2 N Oe1 In order to develop the probability function for specific near collapse EDP, as discussed in the next section, this dataset was then fed to the fragility function.
Fragility function for Near Collapse The fragility function was commonly used to express the probability function of any limited state of interest.
In this study, the 5% damping considered structuresAo period was employed as the IM.
This IM resulted from nonlinear dynamic structural analysis and was then executed using statistical procedures to develop the probability Combining this fragility function with a ground motion hazard function could predict the mean annual rate of near structural collapse.
lognormal cumulative distribution function was commonly employed to develop the fragility function, as follows:
E log RSA(T.
/ A RSA(T1 ) E
Alog[RSA(T1 )] Au E P[ EDP C IDRmax | R = RSA(T.
] = AIE
Where P [EDPC IDRmax.
R = RSA(T.
] is the probability of reaching or exceeding near collapse state EDP .
o-called probability of near collaps.
while a ground motion induces the structure with RSA(T.
AI(.
) is the function of standard lognormal cumulative distribution.
ARSA(T.
is the median of IM that would cause near collapse.
and Alog[RSA(T.
] is the standard deviation of the IM that would cause near collapse EDP, in the form of maximum interstory drift ratio.
IDRmax.
The IDAAos result was not all the time could reach the targeted limit state of collapse in developing the fragility function.
Baker .
has developed a procedure to fix the dataset to predict the fragility function.
The study also adopts the recommendation of FEMA P-58 guidelines to increase the logarithmic standard deviation .
y adding Au = It is done so since the uncertainty in the analytically-based fragility curve could not adequately and accurately represent the true variability .
RESULTS AND DISCUSSION In the incremental dynamic analysis (IDA) and collapse probability function, the employed structural model and its ground motion and the number dataset to be tested played an important In this section, the IDA result is presented based on the median value of maximum interstory drift (IDRma.
, selected as the engineering demand parameter (EDP), and the median value of intensity measure IM = RSA(T.
Moreover, global elastic stiffness and global post-elastic stiffness are discussed as well.
The global elastic stiffness is represented by initial linear lines in the IDA curve, whereas after-turning-point lines represent the global inelastic stiffness.
The changes in the direction of these lines are caused by the occurrence of plastic hinges in the element of structures due to the decrement of IM.
The median IDA curves for the MRF with 5-, 10-, 15 and 20-story induced by a single and repeated 2x and 3x earthquakes were depicted in Figure 4 and Figure 5.
These earthquakes were considered as 1GM, 2GM, and 3GM.
For concise and simplicity, the next paragraphs use IM and EDP to explain RSA(T.
and near collapse IDR, respectively.
The probabilistic analysis in this study produced the standard deviation of IM AIM = 0.
33 for all considered MRF.
Porter et al.
found that commonly AIM = 0.
2 to 0.
6, after including the uncertainty factor Au, whereas others explained that commonly A = 0.
4 was used to develop the fragility function without uncertainty factor .
Basone et al.
indicated their dataset achieving A = 0.
29 to 0.
Faisal et al.
Seismic performance of moment frames under multiple fling-step pulse A SINERGI Vol.
No.
June 2022: 155-164 dataset's quality is high if the A or A differences are found to be C 20%.
when assessing the seismic fragility curve of RC buildings with T1 =0.
34 s.
They evaluated the RC building up to the collapse state, not near collapse state.
Porter et al.
also explain that the .
Figure 4.
Average Maximum Inter-Story Drift Ratio of 5- to 20-Story SMF (R=.
Affected by Single Ground Motion .
GM) and Multiple Ground Motion .
GM and 3GM) Faisal et al.
Seismic performance of moment frames under multiple fling-step pulse A p-ISSN: 1410-2331 e-ISSN: 2460-1217 .
Figure 5.
Probability of Near Collapse for 5-, 10-, 15-, and 20-Story SMF and OMF Induced by MGM with Fling-Step: .
1GM Effect on the SMF, .
Maximum MGM Effect on the SMF, .
1GM Effect on the OMF, .
Maximum MGM Effect on the OMF This study found A and A the difference was the same as indicated by Porter et al.
Therefore, it can be concluded that the high quality of fragility functions in this study was well defined and For all types considered MRF, the increase of EDP at elastic conditions caused by 2GM is 3% higher than the EDP caused by 1GM.
This elastic condition is slightly different from EDP due to 3GM, 26.
80% higher than the EDP response due to 1GM.
The change in factor R is not visible at this condition, as indicated in Figure 4.
The significance of the response of 2GM and 3GM is clearly detected when the magnitude of IM increased and arrived at the inelastic condition, which is posed at near- and after-line of near collapse EDP.
The 2GM has increased the response of 5story OMF by 28.
64% earlier than the IM of 1GM.
This IM could be lower at 45.
95% than the IM of Faisal et al.
Seismic performance of moment frames under multiple fling-step pulse A SINERGI Vol.
No.
June 2022: 155-164 1GM when 3GM affect the OMF.
In this postelastic stiffness region, several of R's effect on the response appears, as indicated in the IM of 3GM when induced to the SMF.
It has produced an IM of 89.
11% lower than the IM of 1GM for near collapse EDP.
This is almost two times larger than the IM for the EDP response of OMF.
The percentage of decrement IM for the near collapse EDP under the influence of multiple motions has showed not largely different under various story types .
arious fundamental It is clearly indicated in 20- and 5-story EDP (Figure .
, which exhibited the EDP earlier with IM of 42.
70% and 45.
60% lower than IM of 2GM, respectively.
These IM effects would more likely decrease up to 79.
40% and 71.
40% if 3GM induced SMF, respectively, compared with IM of 1GM.
The maximum response to the near collapse state caused by RE was exhibited on the 10-story MRF, which needed an IM of 90% lesser than the IM of 1GM.
While the 20story MRF experienced minimum response to near collapse state in this study when RE with IM 50% lower than IM of 1GM induced to the In general, it was found that all frames might pose the near collapse EDP early with IM lower of 45.
15% and 68.
37% than IM of 1GM, when 2GM and 3GM were induced to the frames.
For SMF, the 2GM and 3GM might cause the near collapse EDP reached early with IM of 40% and 76.
80% lower than IM of 1GM.
The 2GM and 3GM were made the near collapse EDP occurred more likely for the IMF.
These were needed IM of about 61.
70% lower than the IM of 1GM.
A similar trend was also found for OMF under 2GM and 3GM, which was achieving near collapse EDP with IM of 32.
60% and 42.
60%, respectively, lower than IM of 1GM.
From probabilistic analysis, it was found that the median IM for near collapse EDP of single ground motion .
GM) was in the range of 56 to 0.
36, which was meant that as the story of MRF increased, the median IM for near collapse EDP was decreased.
The standard deviation of IM, which could skew the diagonal line of the fragility curve, was found within the 29 Ae 0.
33 randomly.
The effect of multiple ground motion (MGM), a maximum of two- or three-times ground motions, was depicted in Figure 5 compared to the effect of single ground motion .
GM).
The figure clearly indicated that the IM of MGM has caused the SMF to exhibit near collapse earlier than the effect of 1GM on the SMF.
The median of IM for MGM was within the range of 0.
25 to 16, which was about 13.
81 %, increasing the probability of near collapse of SMF.
A similar trend was also found for OMF under the influence of MGM, compared with 1GM.
However, the increment of probability was significantly larger than SMF.
One of the OMF might be reaching 52% of IM of MGM, which was earlier than the effect of 1GM in achieving near collapse EDP.
This result was slightly less than the influence of MGM on the IMF.
Overall, the effect of MGM on the considered MRF in this study could cause the required IM of 12.
61% lower than the required IM of 1GM.
CONCLUSION
The seismic performance evaluation of the moment-resisting frame (MRF) has been The assessment was using the single, two times and three times multiple ground motions (MGM) with the displacement fling-step This type of pulse was not commonly incorporated in the motion records for seismic evaluation of MRF since the data available was Four archetype RC frames were considered, namely 5-, 10-, 15-, and 20-story with reduction factors R = 8, 5, and 3, which represent special (SMF), intermediate (IMF), and ordinary MRF (OMF).
These R factors were affecting the strength capacity of beam-column elements of Mmax/My =1.
13 and the rotation capacity of 0.
04 rad for SMF and 0.
02 for IMF and OMF, which is based on experimental testing by others.
The nonlinear inelastic response history analysis was conducted incrementally to develop the incremental dynamic analysis curve and calculate the median and standard deviation.
Therefore, the following insight can be On average, two- and three-times multiple ground motions .
GM and 3GM, respectivel.
have increased the engineering demand parameter (EDP), which was inter-story drift IDR in this study, up to 23.
53% compared with the EDP caused by single ground motions .
GM).
These 2GM and 3GM effects were measured under the same intensity measure ( IM = RSA(T1 ) ) with the single earthquake .
GM) and were within the linear elastic Therefore, the difference in EDP caused by 2GM and 3GM was found not The 2GM and 3GM have made IM shift more than half earlier than IM of 1GM in producing near collapse inter-story drift (IDR = 2%), which was selected near collapse EDP in this The 2GM has propagated the IM of near collapse shifted earlier than IM of 1GM for all MRF considered in the study.
Both Faisal et al.
Seismic performance of moment frames under multiple fling-step pulse A p-ISSN: 1410-2331 e-ISSN: 2460-1217 might produce the IM of near collapse 37% lower than the IM of 1GM.
Moreover, the 2GM and 3GM have caused special MRF to produce the near collapse EDP, with IM reaching 76.
76% lower than the IM of 1GM.
A similar trend was also found for intermediate and ordinary MRFs, which might be reached the IM of 85.
71% and 42.
earlier than the IM of 1GM, respectively.
The probabilistic analysis shows that the multiple ground motionAos IM for near collapse EDP of SMF decreases compared to single ground motion.
The probability of near collapse due to MGM is increased, either caused by 2GM or 3 GM, in comparison with 1GM.
Similar findings were also demonstrated in the IMF and OMF under MGM.
However.
IMF produces a maximum response of MGM, compared with SMF and OMF Indeed, the findings herewith might also be due to the variations in considered story heights.
R = 3 to 8, and rotation capacity, which has also contributed to the critical effect on the seismic performance of the structure, besides the multiple ground motions containing filing-step pulse.
ACKNOWLEDGMENT
The internal grant supported by Universitas Muhammadiyah Sumatera Utara is greatly Furthermore, the author wishes to thank my undergraduate students involved in this research as the numerator.
REFERENCES