nd rd Interntionl Conference on Electricl, Electronics nd Civil Engineering (ICEECE'1) Jnury 4-5, 1 Bli (Indonesi) Experimentl Study, Stiffness of Semi-Rigid -to- s Using Bolts nd Angles Khled M. Amtered El-Abidi nd Budi Suswnto Abstrct A steel structure cn only perform its best in sustining lod when the connections between members re designed dequtely. Thus, the initil slope is of direct significnce from design viewpoint, in tht it cn be used to represent the stiffness of the connection in the nlysis of the complete structurl system, prticulrly under live lod fluctutions. This pper focuses on design nd experimentl results of semi-rigid bem-to-column connections using fsteners. Six specimens consist of bem-to-column connection sub ssemblges with four specimens using double ngles for (SRC- 1A, SRC-1B) nd (SRC-A, SRCB), nd two specimens using topset ngles for (SRC-A, SRC-B) were designed, mde, nd ed using monotonic loding. From the experimentl results, it cn be obtined initil stiffness (k i ), plstic stiffness (k p ), behvior of connections, nd filure mode of structure. Bsed on the experimentl nd nlyticl results reported herein, the following conclusions cn be drwn: the filure mode of SRC1-A due to buckling of double ngle nd SRC-A nd SRC-A due to buckling of column flnge; SRC is the lrgest stiffness compred to other two specimens, nd the stiffness of SRC is lrger thn tht of SRC1, it cn impct to the moment cpcity of ech specimen. column connections, [], [] hve proposed few simple formuls in lst two decdes. But those formuls were developed bsed on either liner contct seprtion line between ngle nd column or rottionlly free boundry condition so tht ccurcy of those formuls needs to be improved. The purpose of this pper is to propose prcticl model regrding semi-rigid connections with ngles tht predicts the initil stiffness nd the plstic stiffness chrcteristics by determining two shpe prmeters, nd to develop the nlytic formultions for evluting the two shpe prmeters for double web ngle connection nd top nd set ngle connection s shown in Fig. 1. Keywords experimentl progrm, initil stiffness, plstic stiffness, semi-rigid connections. O I. INTRODUCTION NE of the most vitl prts in steel structures concerning its strength is the connection between members. A steel structure cn only perform its best in sustining lod when the connections between members re designed dequtely. Therefore, the designers must give proper ttention to the connections in order to ensure tht the structure cn function s intended. When the vlues of the lodings on the bem-to-column rech limiting stte, the member will experience out-of-plne bending nd twisting. This type of filure occurs suddenly in members with much greter in-plne bending stiffness thn torsionl or lterl bending stiffness [1]. Therefore ccurte clcultions of displcements or initil stiffness re needed in design of semi-rigid bem-to-column connections. Regrding to how to estimte initil stiffness of semi-rigid bem-to- Khled M. Amtered El-Abidi is Doctor Student of Philosophy, School of Civil Engineering, Engineering Cmpus Universiti Sins Mlysi (phone: +6014988; fx: +6159484; e-mil: idi89195@yhoo.com). Budi Suswnto is lecturer of Deprtment of Civil Engineering, Institut Teknologi Sepuluh Nopember, Surby, Indobesi. Cmpus ITS Sukolilo, Surby, Indonesi, 60111 (e-mil: budisw00@yhoo.com). () Double web ngle connection (b) top nd set ngle connection Fig. 1 Semi-rigid connection with ngle II. 1BFUNDAMENTAL THEORY A. 5BDetermintion of initil stiffness If the initil stiffness nd the plstic stiffness could be determined nlyticlly, the prediction of the M θ reltionship from the proposed model cn be mde independent of experiment. In order to determine krir krpr, the lod deformtion (P Δ) the behvior of ngle segment in tension needs to be exmined first of ll. 60
gr1r grr is rd Interntionl Conference on Electricl, Electronics nd Civil Engineering (ICEECE'1) Jnury 4-5, 1 Bli (Indonesi) Typicl double web ngle connection is shown in Fig. 1(). In the initil elstic stte, it is ssumed tht there is no slip nd tht the leg of ngle djcent to the column behves linerly elstic, while the leg of ngle djcent to the bem behves s rigid body [4]. Since the higher restrint between the ngle nd the column fce rises from the clmping ction of bolts, it is ssumed tht the full length of the ngle cn resist the bending moment. The center of rottion is close to the midlength of the connection [5] s shown if Fig.. Using the P e Δ e reltionship of ngle segment in Eq. (1) 1EI g g 1 + Pe = e g g g. (1) 1 1 + 4 where: distnce from the leg s center line to the first fstener-nut edge in leg djcent to the column distnce from the leg s center line to the center of fstener holes in leg djcent to the bem EI bending stiffness of ngle segment per unit length. The reltionship between the moment М nd the rottion θ is obtined s 1 4 0.5l.0.5..0.5 θ. M = P e l l () e where IRαR: the length of ngle Consequently, initil stiffness krir given by 1EI g1 + g k i =. 1 1 4 l () g g + g Fig. Force distribution of double web ngle in elstic stte B. 6BDetermintion of plstic stiffness According to experimentl results reported by Lewitt et l. [6], it cn be seen tht the plstifiction of ngle tkes plce perfectly t ΔRP R= 5.08 mm (0. in) nd the center of rottion hs been ssumed to be t 0.8 of the length of ngle. This ctully vried with the vlue of ΔRPR, the lod distribution in the length of ngle hs tendency to be concentrted on bolts []. Therefore it is ssumed tht the totl tension force in connection ngle is the sum of the product of the pplied lod (FR1R,, FRnR) nd dimeter of the nut (DRnutR). Assuming the deformtion of ngle is proportionl to the distnce from the center of rottion s s shown in Fig., the reltionship between the moment M nd the rottion θ is obtined s M = where: li 0.81 θ. n Ii.. + 0.1 1 0.8 i Pp Dmit I i= I Ii p I Distnce from the center of rottion to the ech center of fstener DRnutR Dimeter of the nut Finlly, plstic stiffness kp is given by where Δp=5.08 mm Fig. Force distribution of double web ngle in plstic stte III. BEXPERIMENTAL PROGRAM Six specimens consist of bem-to-column connection sub ssemblges with four specimens using double ngles for (SRC-1A, SRC-1B) nd (SRC-A, SRCB), nd two specimen using top-set ngles for (SRC-A,SRC-B) re designed nd mde in this pper s listed in Tble 1 nd Tble. In this experimentl progrm, sub ssemblge with column height of 65 cm nd bem length of 65 cm ws designed with semirigid connections using fsteners s shown in Fig. 4. TABLE 1 MATERIAL PROPERTIES OF THE MODEL Yield No Component Size strength Fy (MP) 1 4 5 Model ( A) Model ( B) Model ( A) Model ( B) 151.5x4. 5x5x6.50 151.50x4.5 0x5.5x6.5 Averge Models ( A nd B) Ultimte strength fu (MP) 0.0 505.05 460.5 5.68 81.8 59.6 199x99x5x 1.89 46.19 0.5x10x 5.50x8 Averge Models ( A nd B) (4) (5) 485.1 580.95 408.8 58.5 Bolt ccording to 1 mm 468. 490.6 tensile 61
rd Interntionl Conference on Electricl, Electronics nd Civil Engineering (ICEECE'1) Jnury 4-5, 1 Bli (Indonesi) TABLE STRUCTURAL STEELS AND BOLTS USED OF THE SPECIMENS Specimen Angle SRC-1A SRC-1B SRC-A SRCB SRC-A SRCB x5x5x x5x5x x5x5x x5x5x x5x5x Type of Angle Number of Bolts Dimeter of Bolt (mm) L 50x65x5 Double 9 1 L 50x65x5 Double 9 1 L 50xx6 Double 1 1 L 50xx6 Double 1 1 L xx10 L x5x5x xx10 SRC-1 Top-set 16 1 Top-set 16 1 x5x5x The structurl steels of wide flnge section re used for column nd bem element, the bolts nd ngles re used for connection. The geometry of specimens is shown in Fig. 4, the mteril properties for the model components re shown in Tble 1. The structurl steels nd bolts used re listed in Tble. The bem-to-column connections were designed bsed on the type of connection scheme nmed of (SRC-1A, SRC-1B), (SRC-A, SRCB) nd (SRC-A, SRC-B) s shown in Fig. 5. For (SRC-1A, SRC-1B), the connections used double ngles with 6 bolts on column flnge nd bolts on bem web; for (SRC-A, SRCB), the connections used double ngles with 6 bolts on column flnge nd 6 bolts on bem web; nd for (SRC-A, SRC-B), the connections used top-set ngles with 8 bolts on column flnge nd 8 bolts on bem flnge. 11 6 6 x5x5x 11 6 6 x5x5x DL 50xx6 Lod P 615 0 () Specimen (SRC-1A, SRC-1B) SRC- x5x5x Lod P 600 0 (b) Specimen (SRC-A, SRCB) SRC- x5x5x 4 6 40 40 6 x5x5x 0 0 0 0 4 48 88 () (SRC-1A &1B) (b) (SRC-A nd B) (c) (SRC-A,SRC-B) Fig. 5 Detils of bem to column connection () Specimen SRC-1A (b) Specimen SRC-A (c) Specimen SRC-A (d) Specimen SRC-1A Lod P 0 610 (c) Specimen (SRC-A,SRC-B) Fig. 4 Side view of specimens (e) Specimen SRC-A (f) Specimen SRC-A Fig. 6 Test set up of specimens 6
nd rd Interntionl Conference on Electricl, Electronics nd Civil Engineering (ICEECE'1) Jnury 4-5, 1 Bli (Indonesi) For the set up, the specimens were ed under monotonic loding with distnce of round 600 mm from the fce of the column using hydrulic jck. Dt collection softwre ws used to check the reding of ll connected chnnels to the instruments on the specimen, such s strin guge, LVDT, nd inclinometer. After initilizing the instrumenttion system, the specimen ws loded by 5 kn incrementl pplied lod until s substntil deflection of the bem cn be observed. For ech loding step, set of reding ws tken for deflections, rottions, nd strin of steel section. The set up of specimen is shown in Fig. 6. Fig. 8 M Δ reltionship from result () Specimen SRC-1A (b) Specimen SRC-A The prediction of the M θ reltionship from the proposed model cn be mde independent of experiment. In order to determine krir krpr, the moment deformtion (M Δ) the behvior of ngle segment in tension cn be exmined fter ing of specimens. From the results, the stiffness of connection cn be clculted ccording to mteril nd section properties of bem, column, nd connection. The clcultion of initil stiffness (krir) nd plstic stiffness (krpr) re listed in Tble to Tble 6. TABLE CALCULATION OF INITIAL STIFFNESS OF SRC1 AND SRC (c) Specimen SRC-A (d) Specimen SRC-1A Specimen gr1 gr gr trа DRnut lr kri SRC-1A 15.9.9 10.1 6 19.6 10. 94,06 SRC-1B 16.5 1.5 9.5 19.6 101 9,864 SRC-A 16 50.6 8 8 19.6 10,54 SRC-B 16.4.9.6 9 19.6 101. 19,0 TABLE 4 CALCULATION OF PLASTIC STIFFNESS OF SRC1 AND SRC (e) Specimen SRC-A (f) Specimen SRC-A Fig. Filure modes of specimens After ing, the filure mode of SRC1-A due to buckling of double ngle nd SRC-A nd SRC-A due to buckling of column flnge, but for SRC1-B, SRC-B, nd SRC-B, the filure mode due to tension filure of bolts s shown in Fig.. From results, it cn be obtined M Δ reltionship of ech specimen s shown in Fig. 8. SRC1-A nd SRC1-B performs the lowest moment cpcity compred to other specimens, followed by SRC-A nd SRC-B. It cn be ssumed tht semi-rigid connection with double ngle connection using one verticl lyer of bolt on bem web cn contribute low cpcity in moment rther thn using two verticl lyer of bolt. SRC-A hs the highest moment cpcity, becuse top-set ngle cn contribute high cpcity in moment. Specimen gr1 gr gr nrbolts drb σry krp SRC-1A 15.9.9 10.1 0 6,94 SRC-1B 16.5 1.5 9.5 460 55,10 SRC-A 16 50.6 8 0 16,69 SRC-B 16.4.9.6 460 196,9 TABLE 5 CALCULATION OF INITIAL STIFFNESS OF SRC Specimen gr1 gr gr tr DRnut nrbolts kri SRC-A 54.6 49.6 50.6 19.6 1,44 SRC-B 44.9 4.5 40.6 19.6 8,9 6
rd Interntionl Conference on Electricl, Electronics nd Civil Engineering (ICEECE'1) Jnury 4-5, 1 Bli (Indonesi) TABLE 6 CALCULATION OF PLASTIC STIFFNESS OF SRC Specimen g 1 g g t d b σ y k p SRC-A 54.6 49.6 50.6 8 0 61,65 SRC-B 44.9 4.5 40.6 8.8 460.5 5,186 Urbn: Engineering Experiment Sttion, University of Illinois, Jnury, 1969. [6] Lewitt CW, Chesson E Jr, Munse W. Restrint chrcteristics of flexible riveted nd bolted bem to column connections. Bulletin No.500, Urbn: Engineering Experiment Sttion, University of Illinois, Jnury, 1969. [] Sherbourne AN, Bhri MR. Finite element prediction of end plte bolted connection behvior. II: nlytic formultion, J Struct Eng ASCE 199;1():69 81. From nlyticl solution of semi-rigid connections, the vlue of initil stiffness (k i ) nd plstic stiffness (k p ) of SRC is the lrgest stiffness compred to other two specimens, nd the stiffness of SRC is lrger thn tht of SRC1, it cn impct to the moment cpcity of specimen s shown in Fig. 9. Moment-Rottion Reltionship 1600 1400 10 Moment (kg.m) 0 800 600 400 0 SRC1-A SRC1-B SRC-A SRC-B SRC-A SRC-B 0 0.00 1.00.00.00 4.00 5.00 6.00.00 Rottion (Degree) Fig. 9 M-θ reltionship from nlyticl solution IV. CONCLUSIONS Bsed on the experimentl nd nlyticl results reported herein, the following conclusions re drwn: 1. The filure mode of SRC1-A due to buckling of double ngle nd SRC-A nd SRC-A due to buckling of column flnge, but for SRC1-B, SRC-B, nd SRC-B, the filure mode due to tension filure of bolts. The bem mteril of SRC-A series is lower yield strength i.e. 0.0 MP thn tht of SRC-B series i.e. 460.5 MP. And the column mteril of SRC-A series is lso lower yield strength i.e. 1.89 MP thn tht of SRC-B series i.e. 485.1 MP.. From nlyticl solution of semi-rigid connections, the vlue of initil stiffness (k i ) of SRC is the lrgest stiffness compred to other two specimens, nd plstic stiffness (k p ) of SRC is the lrgest stiffness compred to other two specimens. The lrgest of initil stiffness cn impct to the lrgest of moment cpcity of specimen. REFERENCES [1] Trhir, NS (199), Flexurl-Torsionl Buckling of Structures, E & FN Spon, London [] Kishi,N. nd Chen,W.F., (1990). Moment-rottion reltions of semirigid connections with ngles, Journl of Structurl Engineering, 116:, 181-184. [] Azizinmini,A., Bordburn,J.H. nd Rdziminski,J.B., (198). Initil Stiffness of Semi-rigid Steel -to- s, Journl of Constructionl Steel Reserch, 8, 1-90. [4] Wu FH, Chen. A design model for semi-rigid connections. Eng Struct 1990;1;1:88 9. [5] Lewitt CW, Chesson E Jr, Munse W. Restrint chrcteristics of flexible riveted nd bolted bem to column connections. Bulletin No.500, 64