5h Inernaional Conference on Durabiliy of Concree Srucures Jun 30 Jul 1, 2016 Shenzhen Universiy, Shenzhen, Guangdong Province, P.R.China Behaviour of Corroded Sud Shear Connecors under Faigue Loading Ju Chen, Ao-Yu Jiang, and Wei-Liang Jin Deparmen of Civil Engineering, Zhejiang Universiy, Hangzhou, China ABSTRACT Experimenal invesigaion was conduced on corroded composie push-ou specimens o sudy he behavior of shear suds subjeced o faigue loading. A oal of eigh sandard EC4 push-ou specimens were esed. The expeced corrosion rae of es specimens ranged from 0 o 50%. The main purpose of hese ess was o deermine he faigue life and he reducion effec caused by he corrosion on he faigue life. The effec of corrosion on he faigue crack, load-slip curves, and failure modes was also sudied. In addiion, he es resuls were compared wih curren Eurocode design predicions, which is only for specimen wihou corrosion. I is shown ha he curren Eurocode design predicions are quie conservaive for he es specimens in his sudy. Design equaions were also proposed for faigue life of corroded composie push-ou specimens. 1. INTRODUCTION Normally, shear connecors are used a he seel and concree inerface o provide ineracion beween he concree slab and seel girder. Headed sud shear connecors are he mos common ype of shear connecors and are used in seel concree composie bridges for is economic and fas applicaion. Especially in bridges due o raffic acions, hese suds are subjeced o high-cycle faigue loading. Mos of he daa on connecors have been obained from various ypes of push-ou es. Since he 1960s, various researchers have conduced a grea number of cyclic push-ou ess o deermine he faigue life of shear connecors (Akao, Kuria, & Hiragi, 1982; Feldmann, Gesella, & Leffer, 2006; Feldmann, Hechler, Hegger, & Rauscher, 2008; Hanswille, Porsch, & Usundag, 2007; Johnson, 2000; Leea, Shimb, & Changc, 2005; Nakajima e al., 2003; Oehlers, 1989; Sluer & Driscoll, 1965; Veljkovic & Johansson, 2006) Alhough safey facor in iniial design may delay he occurrence of faigue problems, he priori negligence of corrosion pis in srucural inegriy analysis may lead o significan overesimaion of he damage olerance abiliy of exising seel srucures (Peroyiannis, Kermanidis, & Akid, 2005). There is lack of proecion beween he concree slab and seel girder inerface and also makes corrosion difficul o be deeced. The corrosion will damage he suds and he seel girders. Corrosion faigue, which refers o he join ineracion of corrosive environmen and repeaed dynamic sressing, is more derimenal han ha of eiher one acing separaely (Sephens, Raemi, Sephens, & Fuchs, 2000). Bridges, which serve as a major link 127 componen in he infrasrucure sysem, are exremely vulnerable o he combined acion of corrosion and faigue. The acion could possibly bring major hreas on srucural safey and lead o caasrophic failure, such as he ragedy of he collapse of he Silver Bridge in 1967 (LeRose, 2001). Recen research has been carried ou on he corrosion faigue behavior of reinforcing bar or seel beams in he bridge (Aposolopoulos, 2007; Kashani, Lowes, Crewe, & Alexander, 2014; Xu & Wang, 2015; Zhang & Yuan, 2014). I is shown ha corrosion-induced faigue srengh reducion has a large effec on faigue life. Faigue life could reduce more han 60% for he low corrosion condiion and more han 70% for he medium and high corrosion condiion. However, here is lack of corrosion faigue behavior of shear sud connecor for composie bridges. This paper deals wih he resuls of he experimenal sudies on he behaviour of corroded headed shear suds subjeced o unidirecional cyclic loading. Based on he es resuls, load-slip curves and ulimae srenghs of specimens having differen corrosion raes were evaluaed. In addiion, he suiabiliy of curren design equaions for corrode sud shear connecors was also evaluaed. 2. EXPERIMENTAL INVESTIGATION 2.1 Tes specimens Tes specimens were fabricaed according o he sandard push es specimen specified in Annex B of Eurocode (2004). Slabs of 150 mm hickness were used, and bond a he inerface beween he flanges of seel beam and he concree slab was prevened by greasing he flange. Each of boh concree slabs was
128 5TH INTERNATIONAL CONFERENCE ON DURABILITY OF CONCRETE STRUCTURES cas in he horizonal posiion, as is done for composie beams in pracice, and he push es specimens were air cured. The deails of he specimens are shown in Figure 1. The es specimens were labeled ha he exceped corrosion rae could be idenified from he label. For example, he labels SPS-10 and FPS-00-1 define he specimens as follow: The firs hree leers SPS and FPS refer o saic push es specimen and faigue push es specimen, respecively. The following wo digis (00 and 20) indicae he expeced corrosion raes of sud in percenage. The las digi 1, 2, or 3 indicaes he es specimens having he same corrosion rae bu differen load ampliude. Table 1. Maerial properies of concree. Specimen E c (MPa) f cu (MPa) 1 3.31 10 4 43.9 2 3.36 10 4 44.3 3 3.38 10 4 45.2 Average 3.35 10 4 44.5 2.3 Acceleraing corrosion process All specimens, excep he uncorroded one (conrol specimen), were immersed in a 5% NaCl soluion for 3 days afer cured for 28 days, hen he direcion of curren abou 0.2 ma/cm 2 was arranged for acceleraing sud corrosion, suds worked as he anodes, whereas a piece of sainless seel posiioned in he soluion served as cahode, as shown in Figure 2. The I-secion seel beam was isolaed by epoxy resins, so ha corrosion only occurs a he sud and seel concree inerface, as shown in Figure 2. The corrosion ime of each specimen was deermined based on he expeced corrosion rae. The Faraday s heory is used o calculae he corrosion ime. The calculaed resuls are shown in Table 2. I should be noed ha he acual corrosion raes of es specimens may differ from hose expeced corrosion raes. Table 2. Exceped sud corrosion rae and acual corrosion ime of es specimens. Figure 1. Deails of push-ou es specimens. 2.2 Maerial properies and measuremens Three concree cubic specimens were prepared a he ime of push es specimen casing, o deermine he concree srengh of he push es specimens. Table 1 summarizes he maerial properies of concree a 28 days. Suds wih he nominal diameer of 10.0 are used in his sudy. Tensile ess for he sud maerial were conduced. The yield sress from he ensile ess was deermined by 0.2% srain, because he seel for suds generally does no show clear yielding poin. The yield srengh and ensile srengh of sud maerial are 275 and 392.3 MPa, respecively, whereas he elasic modulus is 1.94 10 5 MPa. Qualiy conrol of welding process is a very imporan facor since he effec of welding qualiy may cover he effec of corrosion. Therefore, welding rials were carried ou o obain proper and reliable welding qualiy. Specimen F series FPS0-1 FPS0-2 FPS0-3 SPS-0 FPS10 SPS10 FPS20 SPS20 FPS30 SPS30 FPS40 SPS40 FPS50 SPS50 Elecronic densiy (ma/cm 2 ) Exceped corrosion rae (%) Corrosion ime (days) 0.2 0 0 0.2 10 62 0.2 20 124 0.2 30 186 0.2 40 248 0.2 50 310 Figure 2. Elecronic acceleraing corrosion of push es specimens.
Behaviour of Corroded Sud Shear Connecors under Faigue Loading 129 2.4 Loading es seup and procedure The experimenal program consiss of a oal of wo series es specimens. In he firs series, saic ess were performed o deermine he ulimae saic load P U of he corroded push-ou specimen, as shown in Table 3. The saic esing procedure complied wih he mehod specified in Eurocode 4 Annex B (2004). The ulimae saic load represens he relaive values of loading parameer required for cyclic ess for he second series ess. Using he relaive loading parameers, load-conrolled cyclic es was performed o deermine he faigue life N u of he push-ou specimen. The maximum load (P F-max ) and minimum load (P F-min ) for he cyclic es were 0.65 P U and 0.45 P U, respecively. The chosen maximum and minimum load for each specimen are summarized in Table 3. Table 3. Saic ulimae srengh and corresponding faigue load of es specimens. Saic ulimae srengh Faigue es load Specimen P U (kn) P F-max (kn) P F-min (kn) SPS0 344 222 160 SPS10 326 211 146 SPS20 299 193 134 SPS30 270 176 122 SPS40 241 158 109 SPS50 215 140 97 Corroded push es specimens were loaded in an Insron 8805 faigue esing machine wih loading capaciy of 1000 kn, as shown in Figure 3. The slabs are bedded ono he lower plaen of he esing machine, and load is applied o he upper end of he seel secion. Slip beween he seel member and he wo slabs is measured using LVDTs. The daa were obained by a CRONOS compac 400-08 acquisiion insrumen. The monoonic ess were conduced a a displacemen rae of 0.3 mm/min. The ime aken o reach he ulimae load was abou 40 min. Cyclic ess were conduced wih a load frequency of 7 Hz. During he ess, he ime load from he acuaor load cell, ram displacemen from he buil-in ransducer in he acuaor, longiudinal displacemen beween he concree slab and seel beam, and uplif of he slabs were measured. The ram displacemen included movemen due o he compliance of he es rig, and herefore, i was no used in any subsequen daa analysis. The longiudinal displacemens were measured by wo LVDTs on each seel flange o which he suds were welded, as shown in Figure 3. Figure 3. Tes seup. 3. TEST RESULTS 3.1 Measuremen of sud corrosion rae The corroded suds were rerieved from he failed specimens and he corrosion produc was cleaned using a corrosion-inhibied HCl soluion (Beroa, Simionib, & Saeab, 2008), as shown in Figure 4. I is shown ha he corrosion is uneven disribued. Wih increasing corrosion ime, pis corrosion gradually ransformed o wide-shallow pi (Xu & Wang, 2015). The area loss of he seel sud ( A) was esimaed aferwards by subracing he pos-corrosion area from he measured pre-corrosion area. The pos-corrosion area of sud was calculaed using he measured diameer of he shank of he sud. The measured diameer of he shank was used o calculae he corrosion rae of each sud (y) as: y = (A - A)/A%. The average corrosion rae of eigh suds is aken as he corrosion rae of each push es specimen, as shown in Table 4. I is shown ha he measured corrosion raes of push es specimens are differen from hose expeced corrosion raes. Alhough all hose suds failed a he boom of he shank, he corrosion rae along he whole lengh of he shank provides a reasonable descripion of he corrosion sae. Table 4. Measured corrosion raes of es specimens. Specimen Exceped corrosion rae (%) Measured corrosion rae (%) FPS10 10.0 5.42 SPS10 10.0 5.31 FPS20 20.0 16.87 SPS20 20.0 19.02 FPS30 30.0 25.02 SPS30 30.0 28.40 FPS40 40.0 34.59 SPS40 40.0 36.02 FPS50 50.0 47.04 SPS50 50.0 51.62
130 5TH INTERNATIONAL CONFERENCE ON DURABILITY OF CONCRETE STRUCTURES of each headed sud were examined. The examined fracure surfaces consis of he ypical dull faigue fracure and brigh forced fracure zones, where in he faigue ess he former one is formed by cracks propagaing o a criical lengh and he second one due o forced shear fracure (Van der Walde & Hillberry, 2007). Typical fracure of faigue specimen FPS40 is shown in Figure 6. I is differen from he fracure surface of he sud in saic es, as shown in Figure 5. Figure 4. Corroded suds of es specimens. 3.2 Saic srengh In his sudy, he failure mode of all push es specimens is sud failure. Figure 5 shows ypical sud failure of he es specimens. The ulimae srenghs of es specimen series are shown in Table 3. I is shown ha he ulimae srenghs of es specimens decrease when he corrosion rae increases. The maximum ulimae srengh reducion rae of es specimen SPS series is 62.5%. I means ha he corrosion has significan effec on he ulimae srenghs of es specimens. Figure 6. Sud fracure surface of specimen FPS-40. 3.3.2 Faigue life I can be seen ha he faigue resisances of corroded specimens gradually decreased and heir faigue life ranged from 30 o 80% compared wih he uncorroded specimens, as shown in Tables 5 and 6. This is much beer han he observaions of oher invesigaors (Xu & Qiu, 2013) who repored faigue life reducion of abou six o eigh imes for pied seel specimens under consan ampliude ension ension loading. The reason may be ha he applied cyclic load decreased when he corrosion rae of specimens increased in his sudy. Anoher reason may be ha he fracure posiion always a he boom of he shank, where may no be he place of deepes corrosion pi as specimen in ension. Table 5. Comparison of faigue life predicion wih es resuls of uncorroded es specimens. Figure 5. Sud fracure surface of specimen SPS-20. 3.3 Faigue es resuls and discussion 3.3.1 Failure mode To invesigae he reasons for he reducion of he faigue life, he concree slabs were separaed from he seel beam, and he fracured surfaces a he foo N N R,EC4 R,prop1 Ds n N N R,EC4 N R,prop1 Specimens (N/mm 2 ) (10 4 ) (10 4 ) (10 4 ) N N FPS0-1 149.2 23.9 5.4 23.8 0.226 0.918 FPS0-2 119.4 92.0 32.1 93.0 0.349 1.027 FPS0-3 102.8 183.4 106.3 182.1 0.579 1.055 Mean 0.385 1.000 COV 0.381 0.059 3.3.3 Maximum slip a every load cycle In case of cyclic loading he load deformaion behavior is characerized by an increasing slip and a decreasing elasic siffness. In Figure 7, he maximum slip corresponding o he peak load a every cycle is ploed agains he number of cycles for faigue es
Behaviour of Corroded Sud Shear Connecors under Faigue Loading 131 Table 6. Comparison of faigue life predicion calculaed using nominal sress wih es resuls of corroded es specimens. N N R,EC4 R,prop2 Ds n (N/ N N R,EC 4 N R,prop2 Specimens mm 2 ) (10 4 ) (10 4 ) (10 4 ) N N FPS10 107.8 80 72.8 94.8 0.910 1.185 FPS20 97.8 72 158.0 74.7 2.195 1.038 FPS30 89.5 70.5 320.9 60.1 4.552 0.852 FPS40 81.3 53.5 698.2 47.3 13.050 0.884 FPS50 71.3 32 1985.2 34.3 62.036 1.072 Mean 1.006 COV 0.123 specimens. The beginning and he end of he lifeime are associaed wih a seep increase in he maximum slip wih he number of cycles, whereas in he remaining par of he lifeime a nearly linear increase of he maximum slip occurs wih he number of cycles. The seep increase in he maximum slip a he end of lifeime occurs earlier for specimens having larger corrosion rae. The maximum slip value a each cycle increased wih he increasing corrosion rae. I may be explained ha he corrosion reduce he faigue performance of he sud connecors. Eq. (1) by using curve fiing mehod, as shown Eq. (2). I is shown ha he predicion from he proposed equaion agree wih he es resuls well. N R,prop1 102.2 = 200 τ (2) R 5.540 For he corroded specimens, Eq. (1) was also used o predic he faigue life. However, wo kinds of sress ampliude were used, one is he nominal sress and he oher is real sress. The nominal sress is calculaed by using he cross-secion area of uncorroded sud connecors. The real sress is calculaed by using he reduced cross-secion area of corroded sud connecors. The faigue life of corroded es specimens prediced using he nominal sress was shown in Table 7 and Figure 8. I is shown ha he faigue life increased wih he increasing corrosion rae. This is caused by he reduced nominal sress calculaed using he reduced load bu unchanged cross-secion area of sud connecors. The curve fiing equaion of he nominal sress predicion was shown in Eq. (3), which is only for comparison in Figure 8. N R,prop2 146.0 = 200 τ R 2.460 (3) Figure 7. Maximum slip a each faigue load cycle. 4. DESIGN METHOD OF FATIGUE LIFE Curren Eurocode (2004) provides design mehod for he faigue life of uncorroded sud, as shown in Eq. (1). The design shear resisance of a headed sud in accordance wih Eurocode 4 should be deermined from: m m ( τ ) N = ( τ ) N (1) R R c c where N R is he faigue life corresponding o he load ampliude D R ; D C is he load ampliude a he reference poin, D C = 95.0 N/mm 2 ; N c is he faigue life a he reference poin, N c = 2,000,000; m is he parameer, m = 8. The design srenghs prediced using Eurocode (2004) are compared wih es resuls of uncorroded specimens in Table 6. I is shown ha he design predicions are very conservaive as i is saed in he Eurocode 4. A new equaion is proposed based on Figure 8. Comparison of design predicion wih es resuls. Table 7. Comparison of faigue life predicion calculaed using real sress wih es resuls of corroded es specimens. N N R,EC4 R,prop3 Ds n N N R,EC4 N R,prop2 Specimens (N/mm 2 ) (10 4 ) (10 4 ) (10 4 ) N N FPS10 114.0 80 46.6 86.4 0.583 1.080 FPS20 117.7 72 36.0 71.6 0.501 0.994 FPS30 119.4 70.5 32.1 65.7 0.455 0.932 FPS40 124.2 53.5 23.4 52.2 0.437 0.976 FPS50 134.6 32 12.3 32.6 0.384 1.020 Mean 0.472 1.000 COV 0.142 0.049
132 5TH INTERNATIONAL CONFERENCE ON DURABILITY OF CONCRETE STRUCTURES The faigue life of corroded es specimens prediced using he real sress was shown in Table 7 and Figure 8. I is shown ha he faigue life decreased wih he increasing corrosion rae, which means ha he corrosion reduce he faigue performance of he sud connecors. The curve fiing equaion of he real sress predicion was shown in Eq. (4). I is shown ha he fiing curve agree wih he es resuls well, as shown in Figure 8. I is also shown ha he Eurocode 4 predicions using he real sress are very conservaive. However, he load in he pracical engineering srucures nominal remains consan during he whole life of he srucures. Thus, he real sress will increased wih he increasing corrosion rae of sud connecors, which may resul in he furher reducion of faigue life. N R,prop3 5. CONCLUSION 98.7 = 200 τ (4) R 5.838 Faigue es of seel and concree composie push es specimens wih corrosion deerioraion was conduced in his sudy. Saic es was also conduced for ulimae srengh reference. The es specimens were firs elecronic acceleraing corroded hen loaded o failure. All specimens failed in sud connecors facure failure. Based on he es resuls, he effec of corrosion on he maximum slip a each load cycle and faigue life was sudied. I is shown ha he maximum slip value a each cycle increased wih he increasing corrosion rae. I is also shown ha he faigue life increased wih he increasing corrosion rae a he similar load ampliude. Tes resuls obained from he loading ess were compared wih faigue life prediced by curren Eurocode 4. I is shown ha he faigue life was very conservaive for uncorroded specimens. New design equaion wih reasonably accuracy was proposed, which enables he designer o consider he effec of corrosion on he faigue life. ACKNOWLEDGMENT The research work described in his paper was suppored by research projecs from Science and Technology Deparmen of Zhejiang Province (2015C33005). REFERENCES Akao, S., Kuria, A., & Hiragi, H. (1982). Faigue srengh of sud shear connecors wih concree deposied from differen placing direcions. In: IABSE Faigue, Lausanne, Swizerland. Aposolopoulos, C. A. (2007). Mechanical behavior of corroded reinforcing seel bars S500s empcore under low cycle faigue. Consrucional and Building Maerials, 21, 1447 1456. Beroa, L., Simionib, B., & Saeab, B. (2008). Numerical modelling of bond behaviour in RC srucures affeced by reinforcemen corrosion. Engineering Srucures, 30(7), 1375 1385. EN 1994-1-1. (2004). Eurocode 4: Design of composie seel and concree srucures Par 1-1: General rules and rules for buildings. Feldmann, M., Gesella, H., & Leffer, A. (2006). The cyclic force-slip behaviour of headed suds under non saic service loads experimenal sudies and analyical descripions. Composie Consrucion in Seel and Concree, 5, 564 572. Feldmann, M., Hechler, O., Hegger, J., & Rauscher, S. (2008). Faigue behavior of shear connecors in high performance concree. Composie Consrucion in Seel and Concree, 6, 39 51. Hanswille, G. E., Porsch, M. A., & Usundag, C. (2007). Resisance of headed suds subjeced o faigue loading. Journal of Consrucional Seel Research, 63(3), 475 484. Johnson, R. P. (2000). Resisance of sud shear connecors o faigue. Journal of Consrucional Seel Research, 56(1), 101 116. Kashani, M. M., Lowes, L. N., Crewe, A. J., & Alexander, N. A. (2014). Finie elemen invesigaion of he influence of corrosion paern on inelasic buckling and cyclic response of corroded reinforcing bars. Engineering Srucures, 75(1), 113 125. Leea, P., Shimb, C., & Changc, S. (2005). Saic and faigue behavior of large sud shear connecors for seel-concree composie bridges. Journal of Consrucional Seel Research, 61(11), 1270 1285. LeRose, C. (2001). The collapse of he Silver Bridge. Wes Virginia his soc quar, XV(4). Nakajima, A., Saiki, I., Kokai, M., Doi, K., Takabayashi, Y., & Ooe, H. (2003). Cyclic shear force-slip behavior of suds under alernaing and pulsaing load condiion. Engineering Srucures, 25(4), 537 545. Oehlers, D. J. (1989). A new approach o he design of sud shear connecors in composie bridge beams (Research Repor R82). Universiy of Adelaide. Peroyiannis, P. V., Kermanidis, A. T., & Akid, R. (2005). Analysis of he effec of exfoliaion corrosion on he faigue behavior of he 2024-T351 aluminum alloy using he faigue damage map. Inernaional Journal of Faigue, 27, 817 827. Sluer, R., & Driscoll, G. C. (1965). Flexural srengh of seel-concree composie beams. Journal of Srucural Engineering, ASCE, 71 99.
Behaviour of Corroded Sud Shear Connecors under Faigue Loading 133 Sephens, R. I., Raemi, A., Sephens, R. R., & Fuchs, H. O. (2000). Meal faigue in engineering (2nd ed.). New York, NY: John Wiley & Sons. Van der Walde, K., & Hillberry, B. M. (2007). Iniiaion and shape developmen of corrosion nucleaed faigue cracking. Inernaional Journal of Faigue, 29(7), 1269 1281. Veljkovic, M., & Johansson, B. (2006). Residual saic resisance of welded sud shear connecors. Composie Consrucion in Seel and Concree, 5, 524 533. Xu, S., & Qiu, B. (2013). Experimenal sudy on faigue behavior of corroded seel. Maerials Science and Engineering: A, 584(1), 163 169. Xu, S., & Wang, Y. (2015). Esimaing he effecs of corrosion pis on he faigue life of seel plae based on he 3D profile. Inernaional Journal of Faigue, 72(1), 27 41. Zhang, W., & Yuan, H. (2014). Corrosion faigue effecs on life esimaion of deerioraed bridges under vehicle impacs. Engineering Srucures, 74(1), 128 136.