CAPACITY DESIGN CRITERIA FOR SEISMIC RESISTANCE OF PRECAST CONCRETE COLUMNS USING STEEL BOX CONNECTION

Transcription

1 Geotec., Const. at. & Env., DOI: ISSN: (Print), (Online), Japan CAPACITY DESIGN CRITERIA FOR SEISIC RESISTANCE OF PRECAST CONCRETE COLUNS USING STEEL BOX CONNECTION *Chayanon Hansapinyo, Ntchanon Siri and Chinnapat Bachart Center of Excellence for Natral Disaster anagement, Department of Civil Engineering, Chiang ai University, Thailand *Corresponding Athor, Received: 30 Jne 017, Revised: 19 Nov. 017, Accepted: 0 Dec. 017 ABSTRACT: A variety of advantages of precast concrete adding to benefits of concrete has made the prefabrication and assembly systems widely adopted in bilding constrction bsiness. However, the connections may diminish the increase in the tilization de to their complication in the installation work and seismic design consideration. This paper presents an easy-to-install steel box for colmn base connection designed based on the standard capacity design criteria. Experimental investigations on the precast concrete colmns sing the steel box connection at the colmn base nder cyclic loading were also made. The test reslts show the satisfactory seismic performance of the precast specimens both in seismic shear capacity, dctility and energy dissipation compared with the identical cast-in-place specimen. The failre of the precast colmns, with proper details, was in the flexral dctile mode. The shear capacity was respectively times higher than the averaged capacity of the cast-in-place colmn. The damping ratio was abot 18-5 percent in the inelastic range, after the drift ratio of 3.0%. Finally, the seismic performance of the designed precast concrete colmn sing the steel box connection based on the capacity design criteria is garanteed withot early brittle failres. Keywords: Precast concrete colm Seismic desig Colmn steel box connectio Lateral cyclic load 1. INTRODUCTION The exposre of bildings to damage from seismic shaking is steadily increasing becase of contining rbanization. However, a significant redction of the risk can be accomplished throgh improved bilding codes and constrction qalities. For the design of bildings in a seismic prone area, failre mechanism of strctral elements in a bilding mst be considered to ensre dctile damage nder a severe earthqake. Colmns and beam-colmn joint are maintained in the elastic range. In other words, a good strctre mst be able to control the failre mode and location of damage. Under a big earthqake, flexral failre indced by yielding of the longitdinal reinforcement is expected to localize on beam elements for absorbing and dissipating energy. This phenomenon can be achieved by a systematic design for the Weak Beam-Strong Colmn failre mechanism sing the Capacity Design criteria [1]. Following this concept, the Beam sideway mechanism (Fig.1(a)) is permitted for the failre mechanism of frame bildings nder a severe earthqake load. Experience of past strong earthqakes has proved considerable fewer losses of the Beam sideways bildings compared with the Colmn sideways bildings (Fig.1(b)). As the colmn and the beam-colmn joint elements are very important for the frame bilding, many past researchers investigated the shear capacity demand of the elements nder the lateral loading [-5]. The colmn capacity models based on key parameters were developed [6,7]. For the precast colmn with the base connectio Rejame et al [8] condcted an experimental investigation on precast colmns in socket fondations with internal smooth interfaces. The tested specimens were failed by the yielding of the longitdinal reinforcement ot of the embedded region. Seismic design of dctile reinforced concrete frames has been developed over a long period of time. A sccessfl design method has been based on the past lessons learned from the failed bildings. However, the design of precast concrete element has been done based mainly on information of the experimental reslt of a prototype element, as a wide variety of precast connection options. As a reslt, innovation integrated with fabricator capabilities is the key to create a sccessfl soltion. This paper presents lateral cyclic load test and seismic design of precast concrete colmns sing steel box connection. The connection was adapted from the colmn-to-fondation connection with bolted socket presented by Negro, and Toniolo [9]. The techniqe permits faster constrction with the se of this bolted connection type, as shown in Fig.. 1

2 Seismic design of concrete frame bildings is first described to overview demand capacity for the design of the precast concrete colmns. Following the strength-based design concept or the Capacity Design method, a dctile emlative design can be achieved maintaining the elastic performance of the connection part (Fig.(c)) [10]. The precast concrete colmns with the steel box connectio as shown in Fig.(b) and identical cast-in-pace concrete colmn tests nder lateral cyclic load is explained. From the test, seismic performance of the precast colmns is discssed compared with the cast-in-place one. Eventally, the design of the precast colmns and the steel box connection are described based on the test reslts. (b) Steel box connection Plastic hinge location Strong, elastic colmn connection Beam connection (c) Emlative precast connection design (a) Beam sideways Fig. The precast concrete colmn with steel box connection Plastic hinge (b) Colmn sidesway Fig.1 Failre mechanisms of frame bildings Fig.3 Probable flexral strength of the beam ends and demand shear capacity. SEISIC DESIGN OF PRECAST CONCRETE FRAE (a) Precast concrete colmn in a precast frame (Precast framing beams are ignored for a clear definition of the scope of this stdy on the precast colmn connection) An overview of the seismic design of reinforced concrete frame bildings is described below. Following the steps, damage pattern can be controlled in flexral mode at the beam ends and the collapse of bildings nder a strong earthqake

3 can be avoided based on the Capacity Design criteria..1 Estimation of Seismic Load First, for a reglar bilding, the lateral seismic load on the bilding is estimated according to a local standard. The load depends on seismic parameters inclding local seismic intensity, type of bilding, type of soil, bilding weight and dynamic properties. The calclation starts with base shear force estimation and then distribtes the force to each floor of the bilding. Strctral analysis of the bilding is next performed for the element forces indced by all cases of load combination.. Design of Frame Beam and Capacity Demand From the envelope flexral force of the bilding analysis, flexral strengths reqired for beams are obtained and reqired longitdinal reinforcement can be determined accordingly. The provided beam reinforcement is sed for the capacity demand to avoid ndesirable damage nder a strong earthqake by forming the flexral hinge at the beam ends. The demand shear force V ) at the stage is estimated as shown in Eq.(1) ( e and Fig.3. The beam shear capacity designed following this procedre ensres the avoiding of shear failre in the beam element. V e pr 1 pr w l n (1) l n nc. nb 1 () nc is the sm of nominal flexral strengths of colmns framing into the joint nb is the sm of nominal flexral strengths of beams framing into the joint The flexral capacity of the colmn ( nc ) is then sed to determine the colmn dimension and reinforcement. The reqired shear capacity of the reinforced concrete colmn is also estimated considering the free body diagram as shown in Fig.4(a). The based on the capacity of the designed colmns, the reqired strength of the beam-colmn joint can be determined by considering Fig.4(b) and Eqs.(3). Conseqently, the steel box connection is designed which will be explained next. h l n Vpr1 V pr l h n (a) Colmn shear pr 1 and pr are the probable flexral strength of the beam ends w is the factored load/length of the beam l n is the clear length of the beam.3 Design of Frame Colmn and Beam-Colmn Joint T V j C To protect the colmns against a strong earthqake, design shear and flexral strengths of the colmns mst be higher than the flexral strength of the joining beams. Kntz and Browning [11] have shown that the failre mechanism as shown in Fig.1(a) can only be achieved if the colmn-to-beam ratio is relatively large (on the order of for). According to ACI318 [1], the flexral strengths of the colmns shall satisfy (b) Joint shear Fig.4 Colmn shear and joint shear V T C j T 1.5 f A y s, top, V V (3.1) nj (3.) C 1.5 f A y s, bottom (3.3) j 3

4 V, pr1 V are the shear capacity of the pr framing beams calclated based on the maximm probable flexral strength of the beams V is the colmn joint shear considering the col force eqilibrim in Fig.4(a) V is the joint shear j V nj is the redced joint shear capacity [1] T and C are respectively the maximm tension and compression of the left and right joining beam A, and A s, are the top and bottom bottom s top longitdinal reinforcement area, respectively f is the yielding strength of the y reinforcement.4 Design of Connecting Steel Box The factored axial load ( P ), shear ( V col ) and bending moment ( nc1 ) demand for the design of the connecting steel box are shown in Fig.5(a). The strength of a stb colmn (as the 4-angle bilt-p in Fig.5(b)) nder combined axial and bending moment is examined throgh the interaction formla in Eq.(4)[3]. Connecting Steel Box Beam P nc1 nc Beam Threaded anchor bars (a) Axial, shear and bending moment Sectional area of the bilt-p colmn (b) 4-Angle bilt-p steel colmn section (c) Bending of base plate Fig.5 Design of connecting steel box b T For P P : nc1 1 (4.1) P P 9 con con con For P P 0. 0 : nc1 1 P P con con con (4.) P n, and con n, are the redced nominal con axial load capacity and bending moment capacity of the connecting box, respectively. For other types of ltimate capacities, eg. shear and bending of the base plate, shear transfer across the joint plane between the plate and concrete, anchored bolted etc, the design eqations are codified in AISC steel design [13] and sing the force diagram shown in Fig.5(c). 3. TEST OF PRECAST CONCRETE COLUNS 3.1 Specimens and Experimental Setp a P nc1 P T To verify the applicability of the designed steel box connectio tests of 5 precast colmns were carried ot at the Strctral Laboratory, Department of Civil Engineering, Chiang ai University, Thailand [14]. The test reslts were compared with an identical cast-in-place one. All the colmns were 0.5x0.5 m section and 1.7 m in height from colmn base to top. 4-DB16 were sed as the longitdinal reinforcement for all specimens. For the precast specimens, the steel box connection was sed for the base connection. The difference between the precast specimens is based on the differences in constrction method, e.g. groting, welding to nts, welding of the colmn reinforcement to the steel box. Fig.6 shows the specimen details. For the specimen 4

5 nomenclatre, and P stand for the onolithic cast-in-place specimen and the Precast specime respectively. Lateral cyclic loading was applied to the tip of the specime as seen in Fig.7. While the colmn bottom was arranged to be a fixed conditio the application of the load indces a cyclic shear and moment as if it is a half-story colmn in a rigid frame. The deformation control at the loading point, 1.5 m from the colmn base, represented as story drift was adopted as recommended in ACI T [15]. 3. Section of the steel box connection This is the additional design procedre compared with the cast-in-place colmn. Based on ACI318 [1], the calclated ltimate capacity of the reinforced concrete colmn nder yielding of the longitdinal reinforcement is 5.33 kn, as shown in Table 1. The vale of the ltimate capacity can be eqivalent to the flexral capacity at the colmn base of 40 kn-m. (the point of the applying load is at 1.50 m. from the base). The capacity is sed for determining the demand for shear reinforcement in the colmn and the connecting steel box at the colmn base. For the connecting steel box, based on AISC steel design [13], the estimated flexral yielding ( y ) and stiffness (EI) of the box are 57.9 kn-m and 5.88 kn-m respectively, providing 1.5 and.3 times relatively higher than the ltimate flexral capacity and non-cracked flexral stiffness of the reinforced concrete colmn section. With the capacity of the steel, box connection is higher than the colmn above, the connection failre can be avoided. 3.3 Nmber of dowel bar and threaded diameter Colmn shear and combined axial-bending moment indced in the colmn above the connection are directly transferred to the connecting steel box and the anchored bars. For general cases, P , P-100-0, P-50-0 P-100-0L P-150-0L Fig. 6 Test specimens (b) Precast specimens Fig. 7 Lateral cyclic load test the critical loading stage is determined by the existence of tension. Under the tension load on the colm longitdinal reinforcement resists the entire load. Hence, at the steel box connectio nmber and threaded diameter of the dowel bars are calclated based on eqilibrim, as shown in Eq.(5). It is noted that the bildability also decides size and nmber of the dowel bars. For this test, as the reinforcement of the colmn is 4-DB16, 4 dowel bars with the threaded diameter of 0 mm. and same yield strength of the colmn reinforcement was sed. By sing this approach, the design nder tension becomes conservative. In the real sitatio colmns will be sbjected to combined compressive load and bending. The entire colmn section will be crved and the stress of the section is nder distribted compressive stress or compressive-tensile stress. A sc f A f (5) yc sj yj (a) onolithic specimen -100 A sc, f and yc A sj, f are the sectional yj area and yield strength of the colmn reinforcement adjacent to the connecting steel box and dowel bars, respectively. 5

6 4. TEST RESULTS 4.1 Lateral Load-Deformation Relationship and Failre ode The cast-in-place specimen was failed in flexral failre with yielding of the longitdinal reinforcement at the colmn base. High dctility and energy dissipation were observed. For all the precast colmns, depending on different reinforcement detailing, P and P-50-0 specimens were dctile failre and P-100-0L, P and P-150-0L were the brittle failres. The brittle failre was cased by the improper welding details of the colmn reinforcement on the steel box connection. Under the lateral cyclic loading, the dctile precast concrete colmns with the steel box connection satisfied seismic performances, compared with the cast-in-place colmn. The dctile precast colmns were failed nder flexral mode similar to that of the cast-inplace specimen bt at the section above the steel box connection. Hence, the ltimate lateral load of the dctile precast is higher than the load of the cast-in-place colmn de to the shorter moment arm length. Figs 8 shows the lateral loaddeformation relationship of the cast-in-place concrete colmn and the precast colmns. As seen in Table 1, the ltimate lateral load capacities of the dctile precast colmns are abot times higher than the averaged capacity of the castin-place colmn. (a) -100 (b) P (c) P-50-0 (d) P

7 (e) P-100-0L (f) P-150-0L Fig.8 Hysteresis relation of lateral load-deformation Table 1 Lateral load capacity and damping ratio Specimen P _psh P _pll P _Theory (kn) 3 Failre mode Damping ratio (%) (Story drift >3%) Avg. (P 1 _a ) Flexral -5 P (1.18) (1.19) Flexral 15-1 P (1.4) 9.86 (1.16) Flexral 18- P (0.9) 1.86 (0.50) 5.33 Slip of threaded dowel 3-4 P-100-0L (1.0) (1.0) Fractre of welding 7-15 P-150-0L 7.46 (1.07) (1.79) Fractre of welding P _a the average of psh and pll ltimate capacity of specimen -100 Nmbers in parenthesis and ltimate capacity ratio compared with the P _a 3 P _Theory is calclated based on the ltimate capacity of maximm shear and moment section (at colmn base). The calclated strengths are controlled by flexre 4. Energy Dissipation Capacity The energy dissipation capacity of a strctre indicates the degree of effectiveness of the strctre to withstand earthqake loading. In the present stdy, the qantity of energy dissipation is shown in term of the eqivalent viscos damping ratio (eq), recommended by Chopra [16]. The eqivalent viscos damping ratio is obtained from Fig.9. The eqivalent damping ratio corresponding to drift ratio of all six experiments are shown in Fig.10. The eqivalent damping ratios were between 18-5 percent after drift ratio more than 3 percent, as seen in Table 1. Fig.9 Determination of the eqivalent viscos damping ratio ( eq ) [16] 7

8 Eqivalent viscos damping ratio, ζeq (%) International Jornal of GEOATE, ay, 018 Vol.14, Isse 45, pp REFERENCES Fig.10 Eqivalent damping ratio at different drifts 5. CONCLUSION This paper presents the lateral cyclic load test and seismic design of precast concrete colmn sing steel box connection based on the capacity design concept. First, seismic design of reinforced concrete framing beam is evalated for the demand capacity. The the gideline on how to obtain the seismic design force for the colmns is made. The design of precast concrete colmn with the steel box connection is explained. To illstrate the applicability of the designed steel box connectio lateral cyclic load tests of precast concrete colmns were made and the reslts were compared with a resemble reinforced concrete colmn. The test reslts show the good seismic performances of the designed steel box connection. The colmns with connection possessing the flexral strength and stiffness higher than the concrete colmn above the connection performed the dctile behavior. The flexral yielding strength and stiffness of the box were 57.9 kn-m and 5.88 knm respectively, providing 1.5 and.3 times relatively higher than the ltimate flexral capacity and non-cracked flexral stiffness of the reinforced concrete colmn section. It is noted that welding of the colmn reinforcement to the steel box need to be careflly made. The ltimate capacity of the precast colmn is higher compared with the castin-place concrete colm de to the relocation of the failre section. Hence, based on the capacity design concept, the weak beam-strong colmn can be easily achieved. 6. ACKNOWLEDGEENTS 100 P P P-50-0 P-100-0L P-150-0L Drift ratio(%) The athors wold like to express their sincere appreciation to Chiang ai University for the financial spport. [1] Park R. and Palay T., Reinforced Concrete Strctres, John Wiley & Sons. New York, pp [] Jstin A.. and ehrdad S., Seismic Shear- Axial Failre of Reinforced Concrete Colmns VS. System Level Strctral Collapse, Engineering Failre Analysis, Vol. 3, 013, pp [3] L, X., Urkap, T. H. Li, S. and Li F., Seismic Behavior of Interior RC Beam- Colmn Joints with Additional Bars Under Cyclic Loading, Earthqakes, and Strctres, Vol.3, No.1, 01, pp [4] Bindh K.R., Jaya K.P. and anicka Selvam, V.K., Seismic Resistance of Exterior Beam- Colmn Joints with Non-conventional Confinement Reinforcement Detailing, Strct. Eng. ech., Vol.30, No.6, 008, pp [5] Han S. W. and Jee N. Y., Seismic Behaviors of Colmns in Ordinary and Intermediate oment Resisting Concrete Frames, Engineering Strctres, Vol.7, No.6, 005, pp [6] Elwood K.J. and oehle J.P., Drift Capacity of Reinforced Concrete Colmns with Light Transverse Reinforcement, Earthqake Spectra, Vol. 1, Isse. 1, 005, pp [7] Elwood KJ, oehle JP., An Axial Capacity odel for Shear-Damaged Colmns. ACI Strctral Jornal; Vol. 10, Isse 4, 005, pp [8] Rejame. F. C., Eimair B. E. and Ana L. H., Analysing the Base of Precast Colmn in Socket Fondation with Smooth Interfaces, aterials and Strctres, Vol.4,Isse 6, 009, pp [9] Negro P. and Toniolo G., Design gidelines for connections of precast strctres nder seismic actions. Report EUR 5377 EN, Eropean Commissio 01. [10] Ghosh, S. K. and Hawkins N.., Seismic Design Provisions for Precast Concrete Strctres in ACI318, PCI Jornal, Vol.46, Isse 1, 001, pp.8-3. [11] Kntz G. L. and Browning J., Redction of Colmn Yielding Dring Earthqakes for Reinforced Concrete Frames, ACI Strctral Jornal, Vol. 100, Isse 5, 003, pp [1] ACI318, Bilding Code Reqirements for Reinforced Concrete, American Concrete Institte, Detroit, ichiga 014. [13] AISC, Steel Constrction anal, American Institte of Steel Constrction (4 th Edition), 011. [14] Hansapinyo C., Bachart C. and Wongmetha P., Cyclic Performance of Precast Concrete Colmns Using Steel Box Connectio 8

9 International Jornal of Civil Engineering, Vol. 15, Isse 4, 017, pp [15] ACI T1.1-01, Acceptance Criteria for oment Frames Based on Strctral Testing, American Concrete Institte, 001. [16] Chopra, A. K., Theory and Applications to Earthqake Engineering: Dynamic of strctre (3 rd Edition), Prentice Hall, Englewood Cliffs, NJ, Copyright Int. J. of GEOATE. All rights reserved, inclding the making of copies nless permission is obtained from the copyright proprietors. 9