醉歌离人Sheet thickness /
mm
Tensile strength /GPa
Tensile m odulus /GPa
Tensile elongation /%
011113164621516119
114 Test setup
Fig 14show s the loading fram e and part of the instrum en 2
tation used for the lateral load tests .The top and bottom of each colum n are fixed by the test fram e using the bolts and a ball socket bearing is p laced under the bottom of the col 2um n.A xial load,given in Tab 11,is first app lied using hy 2draulic loading equipm ent and rem ains constant during the testing p rocess to si m ulate the dead load on the colum n .The reversed cyclic lateral load is then app lied through t w o one 2w ay hydraulic jacks and m easured by t w o load cells at 2tached to the hydraulic jacks .The disp lacem ent control m ode is used to app ly the p redeter m ined disp lacem ent his 2tory,w hich m eans that the speci m en is subjected to a dis 2p lacem ent of Δ1=1mm for the first cycle follow ed by
three cycles,each of 2
Δ1,4Δ1,6Δ1,?,until the colum n fails,as show n in Fig 15
.
F i g 14 Test setup for colum n
testing
F i g 15 L ateral disp lacem ent history
2 Test R esults
Three control speci m ens (RB 2,RB 3and RC 2),w hich
have different corrosion levels,are tested to failure w ithout any strengthening to establish the behavior against w hich the perfor m ance of the strengthened speci m ens can be eval 2uated .S ince they have si m ilar failure m echanism s,speci 2m en RB 3is selected as a rep resentative sam p le .Speci m en RB 3has four longitudinal corrosion cracks due to rebar cor 2
rosion before loading and the longitudinal cracks continue to develop as the lateral load increases .This is follow ed by the appearance of flexural cracks at the interval w ithin 300
mm from the stub interface,during the second cycle (δm ax
=2Δ1).From the third cycle (δm ax =4Δ1)onw ard,all the cracks continue to grow and lead to initiation of spalling of the concrete near the colum n 2stub interface .D uring the last cycles,the concrete cover falls off com p letely .From this observation on the speci m en after failure,it is concluded that the falling off of the concrete cover is caused by the corrosion of the rebar,w hich reduces the bond bet w een the concrete cover and the core .
Speci m ens RBJ 22and RCJ 22are strengthened w ith steel jacketing only,so speci m en RBJ 22is selected as a rep re 2sentative sam p le .The behavior of speci m en RBJ 22is char 2acterized by stable hysteretic hoops in both directions of loading due to confinem ent of the steel jacket .The m axi 2m um load is 265179kN and the corresponding disp lace 2
m ent is 8mm.From the ninth cycle (δm ax =16
Δ1)onw ard,the connection bet w een steel angle and batten p late near the colum n 2stub interface is rup tured and results in considerable degradation of flexural strength of the colum n .
O ther speci m ens excep t RA 0,strengthened w ith com 2bined CFRP sheets and steel jacket show si m ilar behavior .Hence speci m en RBJ 321is chosen as a rep resentative sam 2p le .The speci m en RBJ 321exhibits m ore stable behavior than the colum ns strengthened only w ith steel jacketing due to the confinem ent of com bined CFRP sheets and the steel jacket .There is a distinct yield p lateau follow ed by a grad 2ual hardening up to a m axi m um load,follow ed by gradual softening after the m axi m um load .D uring the 11th cycle (δm ax =20
Δ1),the rup ture of the steel angle near the loca 2tion of the colum n 2stub interface occurs and results in rap id degradation of the flexural strength of the colum n,and the test is ter m inated .
Fig 16gives the view s of failure of som e speci m ens and Fig 17gives the lateral load against disp lacem ent curves for som e speci m ens .Tab 15summ arizes the results of the cy 2clic tests .P y and Δy are the yield load and corresponding disp lacem ent,respectively .P m ax is the m axi m um app lied lateral load and Δm ax is the corresponding disp lacem ent
.P u is the failure load,w hich is defined as 85%of the m axi 2m um lateral load P m ax ,and Δu is the corresponding failure disp lacem ent .μΔis the disp lacem ent ductility factor,calcu 2lated by Δu /Δy ,E end is the total cum ulative dissipated ener 2gy
.
Fi g 16 V ie w s of failure of the speci m ens .(a )RBJ 22;(b )RBJ 321
8
05L i J inbo,and Gong J inxin
Tab 15 L ist of the test results
Speci m en P y /kN
Δy /mm
P m ax /kN
Δm ax /mm
P u /kN
Δu /mm
μΔ
E end /(kN ?m )
RA 0
16016621101901876110162124715031605123RBJ 1212231712145253160816021515618140715179170RB 215311011901641916130140117618031584144RBJ 222201212175265179811722519218102615578170RBJ 22126516821733161459195268198191907129116198RBJ 2222531142145291117141452471492615010180221124RBJ 22317214321052341314125199116221711110878189RB 313512011601731203110147122417421963173RBJ 32126215731003341681216428414719176615881192RC 2135142216016718081101421631212341708159RCJ 22230158316027916210133237168221906136104187RCJ 221
2721143100
32016412128
27215419149
6147
117.
00
F i g 17 L oad against disp lacem ent curves for all speci m ens .(a )RB 2;(b )RB J 22;(c )RB J 221;(d )RB J 222
3 D iscussion
311 L ateral load 2disp lacem ent response
D uctility factors and lateral loads are t w o of the m ost
comm on param eters used for the seism ic evaluation of structural com ponents .In Fig 18,the envelopes of the later 2al load 2disp lacem ent response of all test speci m ens are com 2pared w ith rebar corrosion,CFRP layers,axial load level and steel jackets on the ductility factors and lateral loads of the speci m ens .
F i g 18 L ateral force 2disp lacem ent response envelopes
It can be seen from Fig 18that,the enhancem ent in dis 2p lacem ent ductility factors is found to be larger for speci 2m ens RBJ 22and RBJ 221,w hich are tested w ith larger cor 2
rosion degrees,than the corresponding enhancem ents for speci m ens RCJ 22and RCJ 221,w hich are tested w ith low er corrosion degrees .The reason for this behavior is that,al 2though the steel jacket or com bined CFRP and steel jacket reduces the bond degradation of the corroded rebars,but the speci m en RB 2exhibits poorer disp lacem ent ductility than speci m en RC 2due to the greater degree of corrosion . The effect of axial load on the cyclic behavior of the strengthened colum ns w ith com bined CFRP sheet and steel jacketing can also be evaluated in Fig 18.The enhancem ents in the m axi m um lateral load and the ductility factor are high 2er for speci m en RBJ 321,w hich is tested under a high axial load,w ith the m axi m um lateral load value and the ductility factor value increasing 9312%and 12213%,respectively,than those w ith corresponding values of control speci m en RB 3.Corresponding enhancem ents in the m axi m um lateral load value and the ductility factor value for speci m en RBJ 221are 9119%and 10316%,respectively,than for those of speci m en RB 2.This is because under low axial load,both the strengthened and control speci m ens are able to sustain a large num ber of cyclic excursions;the strength 2ened colum ns do so w ith gradual degradation of strength w hile the strength degradation in control colum ns is m ore se 2vere,but for the speci m ens tested under high axial load,the control speci m en undergoes rap id degradation of strength and
9
05Seis m ic strengthening of reinf orced concrete colu mns da maged by rebar corr osi on using combined CFRP and ??
fails m uch earlier than the strengthened speci m en.
Fig18also show s the effect of CFRP sheets on the ductili2 ty and lateral load of speci m ens.B y comparing the results of the speci m ens,it is evident that,strengthening of the col2 um ns w ith com bined CFRP and steel jacketing or steel jack2 eting results in m ore stable responses w ith larger strength and ductility.The i mp rovem ents are found to be m uch greater in the speci m ens strengthened w ith com bined CFRP and steel jacketing than in the speci m ens strengthened only w ith steel jacketing due to CFRP w raps.It is due to the confinem ent of concrete in the p lastic hinge regions.Confining the strength2 ened colum n using steel jacketing w ith one or t w o layers of FRP helped to delay the for m ation of internal sp litting cracks bet w een the corroded colum ns and the steel jacket,thus i m2 p roving the seism ic behavior of the strengthened colum ns. 312 Energy absorp tion capacities
Fig19show s the total cum ulative dissipated energies for the speci m ens.The total cum ulative dissipated energy is calculated from the sum of the energy absorbed in all cycles before the speci m ens failed .
F i g19 Total cum ulative energy absor p tion for all speci m ens
It can be seen from Fig19that,integration of the total hysteretic energy absorbed by the colum ns indicates that the retrofitted colum ns are capable of absorbing m ore energy than that of the corroded colum ns before the colum ns fail. It can also be seen that the CFRP confinem ent substantially i m p roves the energy absorp tion and dissipation capacities of the speci m ens strengthened w ith steel jacketing.Increasing the area of CFRP from zero to t w o layers results in a siza2 ble increase in the energy absorp tion capacity.
4 A nalytical S tudies
411 Yield force and the m axi m um lateral load calculations of the strengthened speci m ens
A ccording to the test results,the strengthened speci m ens fail in the flexural m ode.Thus,the yield force and the m axi m um lateral load are obtained from the flexural strength of the bottom section of the colum n.The tension reinforcem ent in the original colum n is considered by m ulti2 p lying the reduced factor of018.This is because the bond force bet w een the reinforcem ent and concrete is reduced due to the back2filled concrete cover.
The yield m om ent and yield load of the strengthened col2 um ns are calculated by
