Effect of 3D models on seismic vulnerability assessment of deficient RC frame structures.
Pakistan is a developing country and is located in an active earthquake zone. Among the most recent events, a very drastic earthquake of magnitude 7.5 jolted the western zone of Pakistan on 26 October 2015, resulting in severe causalities and damages to structures. Earlier in 2013, Awaran (District of Baluchistan province) faced an intensive earthquake, and in 2005, the most drastic ever in the history swallowed around 80,000 lives and imparted total collapse of several structures in Kashmir area.
Most of the construction which was done before development of Building Code of Pakistan (2007) is not engineered for earthquakes; hence, there is a dire need to investigate the response of these structures under seismic excitations. These structures are termed as 'deficient' by authors in this research, and as a case study, reinforced concrete (RC) frame structures of three, five and eight storey have been selected. The term 'deficient structures' although covers a very broad spectrum of structural deficiencies, but the focus in this research has been put on RC frame structures, lacking in proper detailing for seismic zones.
From literature review, considerable effort has been made in this regard both on global and local levels. Seo et al (2015) have carried out research for seismic performance of RC structures with shear walls, which made the ways helpful to local codes. As an initiative, considerable effort has been made by NED University of Engineering & Technology Karachi with Geo Hazards International under USAID programme by analysing 2 four-story buildings in Muzaffarbad and Karachi. The researchers emphasised on better seismic performance by suggesting retrofitting techniques (Khan & Rodgers 2012, 2014). Naeem et al. (2005) worked over the post-disaster report of 8 October 2005 earthquake in Pakistan. Shezada et al (2011), worked on seismic vulnerability assessment of structures in Pakistan. The researchers primarily focused over the building stock in Abbottabad, a city located in the northern section of Pakistan. A well-renowned effort has been made by Bhatti, Varum, and Alam (2012) for seismic vulnerability assessment and evaluation of high-rise buildings in Islamabad while focusing on improvement suggestions to Building Code of Pakistan. Mushtaq et al. (2015) studied the seismic vulnerability assessment of typically constructed deficient RC structures of three, five and eight storey by using 2D analytical structural models. Ali et al. (2015) have carried out the research for seismic vulnerability assessment of low- to medium-rise structures. The researchers collected data for various types of typical construction and have drawn the probabilistic seismic vulnerability for each case. Ali et al. (2016) have also published a research for seismic vulnerability assessment of masonry infill structures. Seismic vulnerability assessment of different materials has also been conducted for their use in structures. Qureshi, Ahmad, and Salahuddin (2016) investigated the seismic vulnerability assessment of the strengthened glass fibre-RC structures. Also, the effect of confinement on seismic vulnerability assessment has been studied by Salahuddin et al. (2017) by using CFRP (carbon fiber reinforced polymer) strengthened RC structures.
This research is in continuation of previous work by Mushtaq et al. (2015), using 3D structural models for three-, five- and eight-storey-deficient RC structures, typically constructed in Pakistan. The research focuses on difference in trends if sophisticated 3D structural analysis models are used compared to conventional 2D models. From literature review, most of the research has been carried out using 2D models. However, 2D models have certain limitations like under lateral loads they are unable to simulate the relative displacement effect of transverse members to longitudinal. This relative displacement of members can cause induction of biaxial stresses (Wang and Wen 2000), hence change in structural response is obvious. The selected cases studies have symmetric plans to ensure that only structural response in terms of interconnectivity of frame members in orthogonal directions could be studied. PERFORM 3D has been used as analytical tool for structural modelling and analysis. Results of static cyclic analysis are further used to estimate the seismic vulnerability assessment using advance capacity spectrum method (Kyriakides 2008) and are compared with earlier research. Comparison of vulnerability curves has also been carried out with standard Global Earthquake Safety Initiative (GESI) curve for deficient structures.
2. Structural models
The three structural models have been analysed using PERFORM 3D as analytical tool. Material properties were defined as concrete with compressive strength of 20 MPa and reinforcing steel with yield strength of 365 MPa (Ali et al. 2015). The authors were able to collect the design data of these structures and the same was incorporated in analytical models. From design details, it is revealed that the structures were designed with enough bar splices but seismic detailing for beam column joint confinement was missing (Mushtaq et al. 2015). To incorporate this seismic deficiency in analytical models, bar pullout model (bar stress vs. slip) of CEB FIP (Eurpean Committe for Concrete--International Federation for Presetressing 1990) and joint shear degradation model by LaFave and Kim (2011) were defined as hysteresis parameters. Figure 1 shows the characteristic backbone curves for bar stress vs. slip and joint shear stress vs. strain, respectively. The former curve is from CEB FIP (1990), and the latter is from LaFave and Kim's (2011) model.
From literature review, the above-shown joint shear degradation model was used by LaFave and Kim (2011) to simulate the cyclic response of beam column joint sub-assemblies. The researchers (LaFave and Kim 2011) derived this curve from earlier work of Krawinkler and Popove (1982) and Paulay and Preistly (1991). The same model was also used by Lowes and Altontash (2003) and Shin and Lafave (2004). Along similar lines, CEF-FIP model code has defined this bar pullout model, being influenced by bar roughness, position, concrete strength, its cover and boundary conditions. The code-specified bar stress-slip curve is considered as statistical mean curve having capabilities to be used in broad engineering perspectives.
Beams and columns were modelled as inelastic fibre sections in PERFORM 3D to ascertain the correct distribution of flexural, shear and torsional stresses with respect to neutral axis (Canbolat 2008). Considering the rigid floors with no axial deformation in the slabs, the same were modelled as elastic elements (Lee and Lee 1988). Connection zone element was used as beam column joints to simulate joint shear degradation and modified steel properties were used to simulate the bar stress-slip response near potential plastic hinge zone of the beams and columns (Canbolat 2008).
Structures were modelled using standard commands and procedure, i.e. defining joint constraints, nodal fixity, supports, drift direction; lateral and gravity loads, etc. The major difference between 2D and 3D modelling lies in defining the nodal constraints and connectivity with adjoining floors and members. Figure 2 shows the real-time pictures of the structures under consideration while their plan and elevations, showing storey height and bay width, are shown in Figures 3 and 4, respectively. The typical joint model is also shown for quick reference (Mushtaq et al. 2015).
The analytical joint model as presented above (Figure 4(d)) is the connection zone element to simulate joint shear degradation in PERFORM 3D. The model is based on Krawinkler and Popov's (1982) approach, which consists of four rigid links and hinged corners with a spring to simulate the response in linear and non-linear range. The model is very well validated experimentally by LaFave and Kim (2011). Mushtaq et al. (2015) have also used the same model to simulate joint shear degradation for deficient RC structures, hence the authors find it convenient to apply the same for this study.
3. Structural analysis and seismic vulnerability assessment
Static cyclic analysis has been performed on 3D structural models for storey drift of 00% to 10%, at an increment rate of 0.05% for each loading, unloading and reloading cycle. Once analysis is complete, the authors derived the backbone capacity curves of each structure through representative hysteresis loops, as shown in Figure 5. For comparison, results of already published 2D structural analysis are also shown for quick reference. Failure modes of the structural models at different stages are also shown for corresponding storey shears and drift. It is important to mention that users cannot extract the failure modes directly from software; rather, the same can be derived through post-processing of analysis files for each cycle separately.
From the above results, a significant difference is observed for 2D and 3D structural models. The trend is obvious since addition of transverse structural members perhaps changes the structural stiffness and crack propagation mechanism.
The backbone capacity curves of 3D structural models are further used to assess the seismic vulnerability using advanced capacity spectrum method (Kyriakides 2008; Kyriakides, Pilakoutas, and Ahmad 2012). The procedure starts from conversion of cyclic backbone capacity curves to spectral acceleration to spectral displacement format, further to bilinear idealisation and calculation of damage index with respect to corresponding peak ground acceleration (PGA). Comparison to 2D structural models is drawn to distinguish the results (Figure 6). Limits states as per FEMA 273 (Federal Emergency Management Agency 1997) are also drawn to understand the level of occupancy at different PGA levels.
4. Results and discussion
Trend difference for backbone capacity curves can be summarised as follows:
(1) 3D structural models have shown cracking at earlier stage than 2D models (Figure 5). The unbalanced moments around a beam column joint in 3D model cause relative rotation of the transverse and longitudinal members. This phenomenon promotes initiation of cracks at earlier stage compared to 2D models, where joints are only subjected to unbalanced moments along longitudinal direction.
(2) Joint shear failure of 3D models is occurring at higher values of storey shear than 2D models (Figure 5). Orthogonal structural members around a beam column joint tend to increase the confinement of concrete core. Although cracking of 3D structural model is initiated earlier to their 2D counterparts, but enhanced confinement effect due to transverse (perpendicular) members at a beam column joint leads to relatively improved joint shear response.
(3) A common trend in both modelling techniques, that is, 2D and 3D. is occurrence of bar pullout failure after joint shear degradation. From literature review, seismically deficient but structures with proper splices of reinforcing steel exhibit enhanced bar pullout strength since bondage to concrete and reinforcing steel is not eliminated unless joint concrete is cracked (Loves et al 2003). The repetitive reversals of the moment tend to initiate earlier cracking in concrete. As discovered, the structures are engineered hence have sufficient bar splices with a better resistance to bar pullout failure. The bar is allowed to pull out once all the stresses are transferred to the splices (after joint shear degradation).
(4) Increase in structural stiffness in direction of application of force makes it brittle. The trend is well versed from comparison of models for eight-storey building. This structure has a relatively more longitudinal stiffness than lateral. Hence, the structure has higher frequency and lower time period. This phenomenon is not addressable through 2D structural models.
Apart from theoretical interpretation of the results, the authors find it more relevant to characterise the results such that readers can easily implement or improve the structural design of respective buildings with respect to various seismic zones of Pakistan. Pakistan has been categorised into four seismic zones, that is, 1, 2A, 2B, 3 and 4, as per zoning defined by Building Code of Pakistan (2007). The different zones are shown in Figure 7.
GESI was developed in January 2000 under joint collaboration of Geo Hazards international and Disaster Management Planning Office of United States. The purpose was to establish a mechanism to promote the seismic risk assessment in developing countries including Pakistan. Hence, comparison to seismic vulnerabilities of structural models has also been drawn with standard GESI curve of deficient structures. Vertical lines in Figure 8 show the threshold of each seismic zone of Pakistan (BCP 2007) to easily differentiate the levels of occupancies for a particular case.
Following conclusions have been drawn from the results and discussion
(1) From vulnerability curves of 3D structural models, generally speaking, higher damage index appears for lower PGA levels than the corresponding 2D models. Hence, 3D structural models are more vulnerable.
(2) 3D structural models show relatively brittle behaviour than the corresponding 2D models. The trend is well versed from observations since the relative displacement of transverse structural members to longitudinal tends to cause early development of micro-cracking in concrete with certainly enhanced energy dissipation. The joint constraints for 2D structural models lead the energy dissipation concentrated only over the longitudinal members.
(3) GESI curve (Figure 8) shows the damage index trend with increase in PGA. From comparison, it is observed that all the studied structural models behave less ductile (more vulnerable) compared to the representative behaviour for deficient structures under GESI. It is established that already designed structural models are brittle and can cause catastrophic failure in case of seismic activity with PGA beyond 0.35.
(4) From comparison to GESI curve (Figure 8), the selected cases are good for construction from zone 1 to 2B. The portion of curve for eight-storey building extending towards zone 4 is to be neglected since it has crossed the damage control limit state well before the threshold for zone 4.
Received 2 August 2017
Accepted 20 May 2018
No potential conflict of interest was reported by the authors.
Notes on contributors
Arslan Mushtaq is structural engineer from National University of Sciences & Technology and is Lecturer at Structural Engineering Department of the University. The author has six years of diversified industrial and research experience in field of structural engineering. The author has 3 earlier publications in field of seismic hazard assessment and retrofitting of RC structures.
Dr. Shuakat Ali Khan is an Associate Professor at Department of Civil Engineering at Abbasyn University, Peshawar Pakistan. He is PhD from University of Sheffield UK and have 18 years of structural consultancy and research experience.
Mr. Junaid Ahmad is structural engineer from University of Engineering & Technology, Taxila Pakistan. He is faculty at Department of Structural Engineering at National University of Sciences & Technology, Islamabad with 6 years of structural site and research experience.
Mr. Muhammad Usman Ali is Structural Engineer at Department of Structural Design and Consultancy Services at Government of Pakistan. He has 7 years of structural design and research experience.
Ali, M. U., S. A. Khan, M. Y. Anwar, A. Aleem, S. Raza, and U. Maqbool. 2016. "Seismic Performance of Deficient Masonry Infill Reinforced-Concrete Structures." Proceedings of the Institution of Civil Engineers-Structures and Buildings 170 (2): 143-155. doi: 10.1680/jstbu.15.00115.
Ali, M. U., S. A. Khan, M. Y. Anwar, and H. F. Gabriel. 2015. "Probabilistic Application in Seismic Vulnerability Assessment of Deficient Low-To Medium-Rise Reinforced Concrete Buildings in Pakistan." Arabian Journal for Science and Engineering 40 (9): 2479-2486. doi:10.1007/s13369-015-1684-z.
BCP. 2007. Building Codes of Pakistan, "Seismic Provisions" (2007): Government of Islamic Republic of Pakistan. Islamabad: Ministry of Housing and Works.
Bhatti, A. Q., H. Varum, and Z. Alam. 2012. "Seismic Vulnerability Assessment and Evaluation of High Rise Buildings in Islamabad." In 15th WCEE, World Conference on Earthquake Engineering, Lisbon.
Canbolat, B. B. 2008. "Structural Applications of a Reinforced Concrete Beam-Column-Slab Connection Model for Earthquake Loading." Proceedings of the 14th World Conference on Earthquake Engineering, Beijing, China, October 12-17.
CEB-FIP Model code 1990, Chapter 3, "Bond Stress Strain Relationship". Comite Euro-International du Beton 1991, first published in 1993 by Thomas Telford.
FEMA 273. NEHERP. 1997. Guidelines for the Seismic Rehabilitation of Buildings, Federal Emergency Management Agency Report no. 273. Washington, DC: Applied Technology Council.
Khan, R., and J. Rodgers. 2012. "Four Storey Office Building in Muzafarabad, A Case Study of Seismic Assessment and Retrofit Design". A Joint Venture by GeoHazards international and NED University of Engineering & Technology, Karachi.
Khan, R., and J. Rodgers. 2014. "Four Storey Academic Building, A Case Study of Seismic Assessment. NED University of Engineering & Technology, Karachi. Accessed April 13, 2014 (www.earthquake.usgs.gov/ earthquakes/eqarchives).
Krawinkler, H., and E. P. Popov. 1982. "Seismic Behavior of Moment Connections and Joints." Journal of the Structural Division 108 (2): 373-391.
Kyriakides, N. 2008. "Vulnerability of RC buildings and risk assessment for Cyprus." PhD diss., University of Sheffield.
Kyriakides, N., K. Pilakoutas, and S. Ahmad. 2012. "Vulnerability Curves for RC Substandard Buildings." 15th World Conference on Earthquake Engineering (WCEE), Lisbon, Portugal.
LaFave, J. M., and J.-H. Kim. 2011. "Joint Shear Behavior Prediction for RC Beam-Column Connections." International Journal of Concrete Structures and Materials 5 (1): 57-64. doi:10.4334/ IJCSM.2011.5.1.057.
Lee, D.-G., and S.-Y. Lee. 1988. "An Efficient Model For Prediction Of The Torsional Effect Of Multistorey Building Structures." In Proceedings of the World Conference on Earthquake Engineering, vol. 1, p. 67. Japan Association for Earthquake Disaster Prevention.
Lowes, L. N., N. Mitra, and A. Altoontash. 2003. "A beam-column joint model for simulating the earthquake response of reinforced concrete frames."
Mushtaq, A., S. Ali Khan, H. Farooq Gabriel, and S. Haider. 2015. "Seismic Vulnerability Assessment of Deficient RC Structures with Bar Pullout and Joint Shear Degradation." Advances in Civil Engineering 2015: 1-10. doi:10.1155/2015/537405.
Naeem, A., Q. Ali, M. Javed, Z. Hussain, A. Naseer, S. M. Ali, I. Ahmed, and M. Ashraf. 2005. "First Report on the Kashmir Earthquake of October 8, 2005." EERI Special Earthquake Report.
Paulay, T., and M. J. Nigel Priestley. 1992. "Seismic design of reinforced concrete and masonry buildings."
Qureshi, L. A., J. Ahmad, and H. Salahuddin. 2016. "Seismic Vulnerability Assessment of Strengthened Glass Fiber Reinforced Concrete (GFRC)." KSCE Journal of Civil Engineering 1-10. doi:10.1007/s12205-016-0819-4.
Salahuddin, H., S. A. Khan, M. U. Ali, and A. Mushtaq. 2017. "Seismic Vulnerability Assessment of CFRP Strengthened RC Structures." Gradevinar 69 (02): 121-130. doi:10.14256/JCE.1555.2016.
Seo, J., J. W. Hu, and B. Davaajamts. 2015. "Seismic Performance Evaluation of Multistory Reinforced Concrete Moment Resisting Frame Structure with Shear Walls." Sustainability 7 (10): 14287-14308.
Shahzada, K., B. Gencturk, A. N. Khan, A. Naseer, M. Javed, and M. Fahad. 2011. "Vulnerability Assessment of Typical Buildings in Pakistan." International Journal of Earth Sciences and Engineering 4 (6): 208-211.
Shin, M., and J. M. LaFave. 2004. "Modeling of Cyclic Joint Shear Deformation Contributions in RC Beam-Column Connections to Overall Frame Behavior."
Wang, C. H., and Y. K. Wen. 2000. "Seismic Response of 3D Steel Buildings with Connection Fractures." In Memorias, 12th World Conference on Earthquake Engineering, Aukland, New Zealand.
Arslan Mushtaq (a), Shaukat Ali Khan (b), Junaid Ahmad (a) and M. Usman Ali (c)
(a) Department of Structural Engineering, NUST Institute of Civil Engineering, National University of Sciences and Technology (NUST), Islamabad, Pakistan; (b) Civil Engineering Department, Abasyn University, Peshawar, Pakistan; Structural Design and Consultancy Department, Government of Pakistan, Islamabad, Pakistan
CONTACT Arslan Mushtaq ([mail]) email@example.com ([??]) NUST Institute of Civil Engineering, National University of Sciences and Technology (NUST), Islamabad, Pakistan
Caption: Figure 1. Characteristic backbone curves for hysteresis models.
Caption: Figure 2. Pictorial views of structures under consideration.
Caption: Figure 3. Planar views of structural models.
Caption: Figure 4. Elevations of structural and joint models for analysis (courtesy of Mushtaq et al. 2015).
Caption: Figure 5. Pushover backbone capacity curves.
Caption: Figure 6. Comparison of vulnerabilities.
Caption: Figure 7. Seismic zones of Pakistan (BCP 2007).
Caption: Figure 8. Comparison of vulnerability with GESI curve--zone wise (BCP 2007).
|Printer friendly Cite/link Email Feedback|
|Author:||Mushtaq, Arslan; Khan, Shaukat Ali; Ahmad, Junaid; Ali, M. Usman|
|Publication:||Australian Journal of Structural Engineering|
|Date:||Jul 1, 2018|
|Previous Article:||Efficiency of image correlation photogrammetry technique in measuring strain.|
|Next Article:||Structural performance of riveted connections in cold-formed steel wall stud.|