Non-Linear Seismic Assessment of Steel Braced RC Frames

Transcription

1 Non-Linear Seismic Assessment of Steel Braced RC Frames Rose Maria George #1, Joshma M #2 1 Mtech Student, Vimal Jyothi Engineering College, Chemperi, Kannur, India 2 Assistant Professor, Vimal Jyothi Engineering College, Chemperi, Kannur, India Abstract In present construction industries, reinforced steel structures which plays an important role in performing well under seismic loads. Structures having steel bracing systems have accomplished as efficient seismic retrofits. In this paper Non-linear static and Nonlinear dynamic analysis in Reinforced concrete buildings having different bracing system were performed using Finite Element Package Workbench.16 and Design and analysis software ETABS 2015.For the non-linear static analysis, finite element modeling of ten micro frame models each of having bracing of X, knee and offdiagonal systems were considered. Cyclic loading is applied on each frame models to get the ultimate load capacity and maximum displacements. Off-diagonal and X braced systems having double channel sections have resulted better ductility by showing 17.95% higher ductile for off-diagonal system. Then non-linear dynamic analysis performed on multi-storeyed frame models of 2, 6 and15 height configurations having 2 types of bracing systems. Models were analysed for time history analysis using two real earthquake acceleration data in India, in which models having X bracings were observed by lower story drift values for medium height buildings and Offdiagonal bracings were observed by lower story drift values for tall buildings. Based on the results obtained from the time history analysis, position optimization check of bracing systems in ten G+9 building models under greater earthquake magnitude data were performed. Thus analysis revealed influence of bracings arrangements over the performance under seismic load and model 10 and model 2 were satisfied for economic and efficient seismic retrofits among ten building models. Keywords Nonlinear seismic analysis; Nonlinear static analysis; Time history analysis; FEM; story drifts; ductility; pushover curves. I. INTRODUCTION Bracings are commonly used in steel frames in order to provide resistance to wind and seismic actions. Concentric braces are among the simplest and easiest bracing methods that have been in use [3]. The main disadvantage of concentric braces, when considering earthquake resistance, is their relatively high stiffness to strength ratio and low deformation capacity. Eccentric bracing systems can be considered as an improvement over concentric ones by providing higher deformation capacity, lower stiffness and higher energy dissipation through improved hysteresis loop characteristics. In this paper Non-linear static and Non-linear dynamic analysis in Reinforced concrete buildings having different bracing system were performed using Finite Element Package Workbench.16 and Design and analysis software ETABS For the non-linear static analysis finite element modeling of 10 micro frame models each of having bracing of X, knee and off-diagonal systems. Cyclic loading is applied on each frame models to get the Ultimate load capacity and Maximum displacements. Models were analysed for time history analysis using two real earthquake acceleration data in India, in which models having X, knee and Off-diagonal bracings were observed by lower story drift values for medium and tall buildings. Based on the results obtained from the time history analysis, position optimization check of bracing systems in ten G+9 building models under greater earthquake magnitude data were performed. Thus analysis revealed influence of bracings arrangements over the performance under seismic load and model 10 and model 2 were satisfied for economic and efficient seismic retrofits among ten building models. II. METHODOLOGY In the first phase non-linear analysis being performed on 8 micro 3D RC frame models using finite element package ANSYS.16.WORKBENCH. Analysis is essentially the extension of the lateral displacement procedure of static analysis into non-linear regime. It is carried out under constant gravity loads and monotonically increasing displacement applied on the masses of the structural model. From which the respective maximum displacement and ultimate load capacity of the models can be recorded [6]. Due to the complexity in modeling and analysis of the macro 3D RC frame models in the software, finite element modeling of the single RC frames been carried out [3]. A. Finite Element Modeling Modeling of the frame which follows the detailed experiment parameters held by (Youssef et al. 2007). Ansys.16 (workbench) is used as the 3D finite element modeling and non-linear seismic analysis software. A four-storey building with dimensions of 12.0 m by 12.0 m was considered for the design process [1]. A mid span panel measuring 4.0 m by 3.0 m was isolated from the third floor of each frame. A 2/5 scaled model frame ISSN: Page 93

2 measuring 1.6 m by 1.2 m was found to be satisfactory. To keep stresses in the scaled model similar to that in the full-scale panel, the forces acting on the panels were also scaled down by a factor of ( ) 2.The respective modeling detailing were explained in Fig 1 for X and off diagonal bracing systems [5]. ` v) RC frame with off centre-braced system having double channel section, e = 0.36 (ODC2). vi) RC frame with off centre-braced system having double angle sections, e = 0.36 (ODA2). vii) RC frame with off centre-braced system having double channel sections, e = 0.39 (ODC3). viii) RC frame with off centre-braced system having double angle section, e = 0.39 (ODA3). ix) RC frame with knee-braced system having double channel section (KDC). x) RC frame with knee-braced system having double angle section (KDA). Fig 1: Finite element modeling details of X-braced and off-centre bracings [2]. Models need to be analysing RC frames which having 3 different types of bracing system, such that X, knee braced and off centre bracing systems have selected for the comparison studies. Modeling of the off-diagonal bracing systems has taken according to the eccentricity values of the member 3 of the bracing. In which the value of eccentricity is considered in between.3 and 0.4.Bracings of section double angle and double channel, which is been arranged for 3 eccentric values as 0.33, 0.36 and 0.39, analysis been performed for the respective models so that best among them can be used for the further analysis to find the seismic performance of the frames [5]. For non-linear static analysis cyclic loading is applied on Braced frame. Push and Pull displacements of 5 mm is applied in first step. Displacement is increased by an increment of 5mm for the consecutive steps. Loading conditions also applied for each steps by an increment of 10kN. Cyclic loading continued till the braced frame cover the plastic limit. Ten micro frame models were analysed for the non-linear static analysis and those bracing systems have analysed for 2 types of cross sections i.e., double channel and double angle sections are listed as, i) RC frame with X-braced system having double channel section (XDC). ii) RC frame with X-braced system having double angle section (XDA). iii) RC frame with off centre-braced system having double channel section, e = 0.33 (ODC1). iv) RC frame with off centre-braced system having double angle section, e = 0.33 (ODA1). Fig 2: The basic micro frame Models for X.knee and offdiagonal braced systems. In order to understand the seismic behaviour of the structure, ductility and stiffness values can find out using the results. The structure ductility µ is defined in terms of maximum structural displacement ( max ) and the displacement corresponding to the idealised yield strength ( y ) as [8]. µ= max/ y... Eqn. (1) Where; µ = Ductility of frame = Pushover displacement capacity max y = Displacement pertaining to the yield point on the idealised elastic perfectly plastic response curve. Stiffness of the frames k can be computed by using the equation, K= V/ max.. Eqn. (2) Where; K = Stiffness of frame.. V max = Base shear (kn). = Maximum displacement (mm). Fig. 3: Parameters used in calculation of ductility [3] ISSN: Page 94

3 B. Non-Linear Dynamic Analysis Inconvenience to do the macro frame model study in the FEA software, non-linear dynamic analysis of the macro models were done in structural design software ETABS. Non -linear dynamic analysis performed on 2, 6 and 15 storeyed buildings on the basis of short, medium and tall is taken. Stories which having frame dimensions of 4m x 3m [4]. To do the time history analysis, real earthquake acceleration data in India are taken. Storeyed frames have been analysed for two seismic acceleration data which of magnitude 4.8 and 7.8. For the non-linear dynamic analysis of the building, earthquake acceleration data are taken from PESMOS earthquake database [10]. From which Indian earthquake acceleration data are extracted. 1) Structural modelling: Modeling parameters of structural members were tabulated for 2, 6 and 15 storey levels along with their bracing connections using steel sections according to IS code. Bracings in the structure have arranged at side bays and lower floor levels, in which the frames of having plan dimension of 4m x 3m. Table 1: modeling parameters of 2 storeyed building Components Floor Dimensions Braces Beam Column X Knee 300mm x 300mm 350mm x 300mm Double channel 250 x 250 x 35mm 1,2 members, Double channel 250 x 250 x 35mm, 3 member box 120 x Double channel 250 x 250 x 35mm, Knee box 120 x Modelling details of the 2, 6 and 15 storeyed models are detailed in the tables 1, 2 and 3 respectively. Structural components are arranged according to the height configurations. Bracing systems were preferred along the side bays of the structure in x-direction, as most of the studies fulfil the better results in structural buildings. In accordance with increase in height of the structure, dimensions of structural components were altered [7]. Table 2: modelling parameters of 6 storeyed building Components Floor Dimensions Beam Column Braces X Offcentre Offcentre Knee second 350mm x 350mm 3-6 floors 300mm x 300mm second floors 400mm x 350mm 3-6 floors 350mm x 300mm second floor Double channel 250 x 250 x 35mm 3-6 floors second floors 3-6 floors second floors and 3-6 floors 1, 2 members, Double channel 250 x 250 x35mm, 3 member box 120 x Double channel 250 x 250 x 35mm, Knee box 120 x While considering the tall framed structures, it is adequacy to provide better attention towards the seismic performance of the structure under minor and major earthquake vibrations. In the case of 15 storeyed buildings complexity of transferring of inertia force in the story heights is more than that of other frames while considers non-linearity of structure. Table 3: Modeling parameters of 15-storeyed building Components Floor Dimensions Beam 650mm x 500mm second, third 4-8 floors 550mm x 450mm 9-15 floors 450mm x 350mm Column 700mm x 650mm second, third floors 4-8 floors 600mm x 550mm 9-15 floors 500mm x 450mm Braces X second, third floor Double channel 250 x 80 x 4-8 floors Double channel 200 x 50 x 9-15 floors Double channel 80 x 50 ISSN: Page 95

4 Knee second, third floors x 4mm 1,2 members, Double channel 250 x 80 x, 3 member box 120 x 4-8 floors 1,2 members, Double channel 200 x 50 x, 3 member box 120 x 9-15 floors 1,2 members, Double channel 80 x 50 x 4mm, 3 member box 120 x second, third floors Double channel 250 x 80 x, Knee box 120 x 4-8 floors Double channel 200 x 50 x, Knee box 120 x 9-15 floors Double channel 80 x 50 x 4mm, Knee box 120 x Fig.4: Modelling of 2, 6 and15 storey frame models with X-bracing system. Modelling of the other models of 2,6,15 storey frames were done with knee and off-diagonal bracing systems. Bracings were arranged in the side bays of the structure which proven better results in the similar studies held by others. C. Position Optimization As an extension to find out the effective positioning of the braced system in the building models, a 10 storeyed building is designed for the analysis and model been assumed to be situated under seismic zone V. Design parameters adopted in the structural models are detailed in the Table. In addition to the loading criteria, earthquake load intensities are taken as per the IS 1893(part 1)2002.For the further step in time history analysis of the consecutive models, load combinations also have been given as per the IS code. Bracing systems are adopted in the respective floor levels in continuous and alternate fashion both in the x and y directions. Some of which are given at side bays of the structure and others arranged with side bays free. In the building, most among the models are equipped with bracing in the lower story levels. Table 4: Modeling parameters of G+ 9 structural models. Components Floor Dimensions Beam Column Braces X second, third Offcentre Offcentre Slab thickness second, third 650mm x 500mm 4-10 floors 550mm x 400mm second, third 700mm x 650mm 4-10 floors 600mm x 550mm Double channel 200 x 50 x 4-10 floors Double channel 200 x 50 x 10 mm second, third 1,2 members, Double channel 200 x 50 x, member 3 box 120 x 4-10 floors 1,2 members, Double channel 200 x 50 x, member 3 box 120 x 150mm for floors and roof. ISSN: Page 96

5 Loading Details Dead load Floor finishes 1kN/m 2 intensities Live load intensities Floor finishes 3kN/m 2 Earthquake LL on slab as per Cl and of IS 1893(part 1)2002 Roof 0 kn/m 2 Floor = 0.75kN/m 2 III. RESULTS AND DISCUSSIONS A. Nonlinear static analysis results Results those obtained for non-linear static and nonlinear dynamic analysis was helped to determine the structural stability of the frames which having steel braced connections. By comparing the results which have obtained from the non-linear static analysis on the basis of maximum displacement and maximum displacement pushover curves were plotted. Table 5: Results obtained for X and Knee braced models Models XDA XDC KDA KDC Ultimate load(kn) Maximum displacement(mm) Ultimate force for the model XDC thus obtained as 412kN for the displacement of 9.46 mm. Cracks in model was observed near the corner brace connections in the reinforced concrete frame model. Considering XDA, from the pushover curve obtained from the analysis noticed that ultimate load was observed as 282kN.Since the maximum displacement value found as 4.28mm which is lesser than that of the XDC. By computing ductility for the XDC model by the Eqn (1) will get as 9 and Eqn (2) can be utilized to get the stiffness of the model as kg/mm. Stiffness and ductility values of the model XDA obtained as per the equations 1 and 2 as, kg/mm and 4.08 respectively. While comparing XDC by XDA model, percentage difference between the stiffness values for the XDA by XDC model is 33.9 % and comparing the percentage difference in the ductility values of XDC to XDA is 54.66%.It implies that XDC is more ductile than that of XDA even its less stiffer. Pushover curve of XDC model is plotted in graph 1.while comparing the results of knee braced systems, both ductility and stiffness values are comparatively lesser than that of X-braced micro models. Eccentric bracing connections like off diagonal bracings are moreover depends on the positioning and sections of braced connections in the RC structures. In order to get the effective frame with off-centre braced system it s mandatory to check for eccentric values for member 3 of the system. So have taken 3 eccentric values in between and for each of them have selected double angle and double channel steel sections for bracing connections. ODC1 models are of eccentric value of member 3 having 0.33, i.e. member 3 is at distance of 0.66m from corner bracing connection diagonally. By computing the values of ductility and stiffness for ODC1 and ODA1 models by the equations 1 and 2, it was obtained as and kg/mm respectively for ODC1 and those of 8 and kg/mm respectively for ODA1 model. The percentage difference between the stiffness values of ODC1 and ODA1 model is 17.64%.While comparing the values for percentage difference between ductility factor of ODA1 and ODC1, ODC1 is 27% higher ductile than that of ODA1.Correlating the above, ODC1 model is stiffer and ductile than that of ODA1. ISSN: Page 97

6 ODC1 ODA2 ODC2 ODA3 ODC3 International Journal of Engineering Trends and Technology (IJETT) Volume 38 Number 2- August 2016 Table 6: Results obtained for off-diagonal braced models. Models Ultimate load (kn) Maximum displacement(mm) Ductility and stiffness of the model ODC2 has obtained by applying equation (1) and (2) were and kg/mm. Also which is applied on ODA2 model and result obtained as stiffness of 8 and ductility of 1848kg/mm. The percentage difference between stiffness of ODC2 and ODA2 is 20%.In relation to the ductility factor for above models, ODC2 is 27.07% more ductile than that of ODA2.Correlating above results, ODC2 model is stiffer and ductile. In the case of OD brace models having double channel section, Ultimate loads have been increasing for successive eccentric values but maximum displacement was more observed on ODC2 model. Thus we could conclude that off-diagonal braced system which having double channel sections of e value of member 3 as 0.36 can behave efficient during seismic excitations. Whereas considering whole 8 frame models, OD2 and XDC models were highlighting of good non-linear seismic characteristics. In which the systems have proven the good ductile, flexible and energy dissipative characteristics with relative higher displacement values. Correlating the results of frames having double angle steel braced sections, ductility is more observed in off-diagonal frames and KDA. Generalising on behalf of ductility values computed for bracings of double angle sections, obtained values showing significant difference. Though OD models are 36.87% more ductile. While reviewing through stiffness values, models which are significantly highlighted greater stiffness for X braced and Knee braced frames having double angle cross sections. The percentage difference between the stiffness values for X and Knee-braced frame is 39.3%. B. Time History Analysis Results From the results obtained for OD3 models, ductility values and stiffness had calculated using the equations (1) and (2).Ductility values for the model ODC3 and ODA3 models are 9.98 and 8 respectively. Similarly stiffness values were found out as kg/mm for ODA3 model and kg/mm for ODC3. Comparing the stiffness values for the frame, Percentage difference of ODC3 and ODA3 is 19.88%. In the case of ductility values of both the models, ODA3 model is 24.75% less ductile than ODC3 model. Its conveying that ODC3 model is more ductile and stiffer. From the obtained results from the time history analysis of the frames, story drift values for the each frames which have resulted significant behaviour fewer than two accelerograms data used for the analysis. Story drifts Values for the respective 2 storeyed frame models have enlighten how the bracings acts effective for small storeyed buildings under less and more earthquake magnitude records. In which knee and X braced systems in structure was much effective under lesser magnitude record, though knee braced system have shown good result in both seismic records. All those values which came out as linear aggression towards the middle stories and the comparable lesser story drift values in the top floor levels than that of the soft storeys. The comparison study over the results of models under two seismic acceleration records is represented in Fig.5 & Fig.6. Behaviour of the medium height frame models with bracings under ground motion records have been well influenced by the arrangement and Nonlinear off-centre braced systems in the analysis ISSN: Page 98

7 X braced systems in the models behaved differently under the two earthquake records, which of showing lesser story drift values in kameng accelerograms data and relative higher values for the off-coast nicobar seismic data. Story drifts in both analysis resulted relatively higher values in the mid storey levels. Among those models which possess off-centre braced system are having good ductile characteristics in sight of non-linearity in its geometry and the story drift values. Story drift values are well portrayed for the models in Fig.7 and Fig.8. In sight of small and medium storeyed models, tall building models (say 15 storeyed) having x and off-diagonal braced systems well performed under medium and higher earthquake intensities. Among them off-centre and X braced systems came out with lesser story drift values. Each models have took a common attention in incremental story drift values towards mid-floor levels, which showing less ductile reinforced concrete sections. While designing the RC framed structures with or without bracing systems it is very essential to assure the ductile behaviour and response of the building under nonlinear criteria. While observing the story drift values for nicobar seismic data, story drift values which are of greater values due to high intensity earthquake. For X braced system it shown a greater story drift value as 2.83cm at third story level. Similarly in the case of models having off-diagonal braced system, story drift value at third story level observed as 0.75cm. ISSN: Page 99

8 C. Position Optimization Results Models have executed for time history analysis using the real earthquake records and maximum story drifts for respective models in X direction have been tabulated. Values which show lesser story drifts in the soft stories and a linear aggression in the values for higher stories which shown much effective system in higher seismic intensities. Evaluating the results of first four structural models, type 4 model which results significant improve over other models by unique behaviour of elasto-plastic property of the ductile braced system. Over which model exposed greater story drift values towards the lower floor levels and decreasing of values towards the higher floor levels. As well as it conveys the good ductility of the building frame with braced system against seismic forces. Model having the lowest level story drift of 1.05cm and 0.09cm for the top most story level of the building. Also model 2 and model 3 have also resulted comparatively better than model 1 which have highly influenced by the positioning of the bracings in the building. In sight of reviewing the results of previous types of models, type 2, type 8, type 6 were excellent among all 10 types of models analysed for non-linear dynamic analysis. So that model 9 and model 10 have modeled for getting the most reliable bracing arrangement by considering the previous models and their influence on the better results. However model 10 highlighted of better results, which have modeled of X-braces over the side bay story levels and off-centre braced system along the mid bay story levels in alternative fashion. In which the model have shown highest of its story drift value of 1.56cm over the sixth story. Both two types bracing system have performed well under the higher seismic intensity and were highly influenced by the arrangements of braced connections. Model 9 is of varying story drift values because of the X- braced connections are of less in number in the arrangements. Fig. 15: model 2, model 9 and Model 10 used for analysis The comparative discussion made on first four models type 1, type 2, type 3 and type 4.Models those have checked for influence of bracing system in Y-direction have resulted very less effect in analysis. i.e., there is no characteristic upgrade in the story drift values for such types of models can identify in the results. By considering the models 5, 6, 7 and 8, model which having braced systems throughout the height level elucidates more ductility characteristics, even the magnitude of the values are bit higher when it was analysed for the higher seismic acceleration record. Type 8 model have justified ductile behaviour of the building by showing higher story drift values beyond the mid floor levels than at the upper story levels. While model 6 moderately satisfied the real ductile behaviour of frame model than that of model 8, though model 7 which poorly performed even which have lesser story drift values. IV. CONCLUSIONS a) Seismic performance study of the frame models were done by using non-linear static and dynamic analysis incorporating soft wares. From the results obtained from the non-linear static analysis, X braced with double channel steel section and ODC2 micro models resulted better ductility as compared with other models. b) There by it reveals that bracings having steel sections of double channel sections are stronger under seismic acceleration in reinforced concrete structures. ODC2 is 17.95% higher ductile than XDC and also observed that ODC2 model was best among the six off-diagonal models. c) While comparing the story drift values for 2 storey frames models, knee braced system performed well under both low and high seismic acceleration data than X and off diagonal braced systems. d) Comparisons were made between the knee and off-diagonal braced models for 6 story buildings. For both seismic acceleration data, models having X and off-diagonal braced systems were performed for lesser story drift values compared ISSN: Page 100

9 to Knee braced systems. Similar results had observed for 15 storeyed buildings, in which X and off-diagonal braced models performed well. So structures having lesser story drift values are stiffer than those having higher values. e) Based on the results obtained from the time history analysis, a position optimization of ten G+9 modes were done using seismic acceleration data. Models were arranged with X and offdiagonal braced systems along the storey heights. Among them Model 2 and Model 10 gave satisfactorily better results, which revealed that if bracings are more arranged in the side bays,better will be the performance under greater earthquake magnitudes. f) It s observed that Performance was enhanced by, equal arrangement of bracings in the lower story levels of models 9 and 10. Maximum story drift value observed for model 10 was less than 1.5cm. Continuous arrangements of bracings in the side bays of the have also strengthen the structure by showing lesser story drift values. In model 2, irregular arrangements of bracings connecting adjacent frames have resulted lower values. [10] Viswanath K.G, Prakash K.B.,Anant Desai, Seismic Analysis of Steel Braced Reinforced Concrete Frames IJCE(2010),Vol.1,issue.1,pp ACKNOWLEDGEMENT The authors would like to thank to Assistant professor Joshma M., faculty in vimal jyothi Engineering college chemperi, Kannur, India, for giving all the encouragement needed which kept our enthusiasm alive. This research was completed during Master Degree project of first author at Vimal jyothi engineering college, chemperi, Kannur, Kerala. REFERENCES [1] ACI Committee 318. Building code requirements for reinforced concrete,(aci ). Detroit (MI): American Concrete Institute; [2] A Kadid, D.Yahiaoui, Seismic Assessment of Braced RC Frames Procedia Engineering 14 (2011) [3] Keyvan Ramin and Mitra Fereidoonfar, Finite Element Modeling and Non-linear Analysis for Seismic Assessment of Off-Diagonal Steel Braced RC Frame International Journal of Concrete Structures and Materials Vol.9, No.1, pp , March [4] M.A.Youssef, H. Ghaffarzadeh, M. Nehdi (2007), Seismic performance of RC frames with concentric internal steel bracing Engineering Structures 29 (2007) [5] Mahmoud R. Maheri, R. Akbari, Seismic behaviour factor, R, for steel X-braced and knee-braced RC buildings Engineering Structures 25 (2003) [6] Mohammad Eyni Kangavar, Seismic Propensity of Knee Braced Frame (KBF) As Weighed Against Concentric Braced Frame (CBF) Utilizing ETABS and OPENSEES, (IJEAT) ISSN: , Volume-5 June [7] PESMOS earthquake database India. [8] Nitin N.Shinde,R. M. Phuke, Analytical Study of Braced Unsymmetrical RCC Building ISSN (2013): [9] Umesh.R.Biradar Shivaraj Mangalgi seismic response of reinforced concrete structure by using different bracing systems eissn: ISSN: Page 101