Intemational Journal of Concrete Structures and Materials Vol.t8, No.3E, pp.151~159, December 2006 Test of Headed Reinforcement in Pullout Il: Deep Embedment Dong-Uk Choi (Received February 25, 2006, and Accepted August 30, 2006) Abstract: A total of 32 pullout tests wero performed for the multiple headed bars relatively deeply embedded in reinforced concrete coluinn-like members. The objective wes to determine the minimum embedment depth that was necessary to safely design exterior ‘beamn-column joints using headed bars, The variables for the experiment were embedment depth of headed bar, center-to-centerdis- tance between adjacent heads, and amount of supplementary reinforcement. Regular strength concrete and grade SD420 reinforcing sicel were used, The results of the text the indicated that a headed bar embedment depth of 10d, was not sufficient to have relatively closely installed headed bars develop the pullout strength corresponding tothe yield strength. ll the experimental variables, influenced the pullout strength, The pullout strength increased with increasing embedment depth and head-to-head distance. It also increased with ‘increasing amount of supplementary reinforcement. For a group of closely-spaced headed bars installed in a beam-column joint, it recommended to use column ties at least 0.68% by volume, 1% or greater amount of column main bars, and an embedment depth of 13d, or greater simultaneously, to guarantee the pullout strength of individual headed bars over 125% of, and ductile load-dis- placement behavior. Keywords: headed reinforcement, mule healed bars, pullout, embedment depth, distance between heads, supplementary reinforcement 1. Introduction “The headed reinforcement or headed bar consists of a head and a deformed reinforcing bar, Simpler insalation an ess con seston of reinforcement, and more effective anchorage are the main advantages of using the headed bars over hooks." Cirular, rectangular, or square ste pate (hea) is typically Welded t0 the reinforcing bar using fiction welding or general welding tech- nique while the head can also be fastened to threaded end of the reinforcing bar** Fig. | shows a headed bor embedded in com ‘te with an embedment depth off (distance berween top ofthe head and the concrete fie) subjected to a tensile force, PLAS shuwa in Fig, 1 the head transfers the fore primarily by bearing ‘ofthe head on concrete. Adklitinal force may be transferred along the sem, The stem yields and factues in wension when. the embedment depth and the cover thickness of eanerete tothe head are both suficietly large. When the embedment depth is sulfi- cicaly large while the cover thickness i insufiient, the concrete adjacent tothe head breaks out, resting in aside-blowout failure. (On the other hand, a Headed bar wth an insufficient embadment epeh can abxyply fl resulting in a concrete breakout fire as indicated in Fi. | In this sty, the use of headed bars in the exterior beam-2ol- ‘umn join. was explored by testing multiple headed bar in pallu. As shown in Fig 2of the jin, while the embedment depth can be sullcicaly lage to make a single headed bar {a beam top ba) KCL mando, Dept. oF Arctetral Enginwing, Haskyong Natal Univerty, Anscong 456-240 Koren. mal: cea Consisht © 2008, Korea Covert Insti. All rights reserved, including the making of copce Wier the weiter. pemmision fom uly te copyeaht proprietors yield in tension, a group of closely spaced headed bars (mukiple tbeam top kar} having the same embeiment depth may create one lage concrete breakout cone, resuking in a premature faire. ‘Thus one practical drawback of headed bars canbe the reduction in the pullout strength and a brite failure of « group of closely spaced headed ars is noted that this sty isthe second of three technical papers prepared to explore the use of the headed bars in ‘the exterior beam-column joints and deals withthe pullout behav- ior ofthe headed bars only.” A total of 32 pullout tests were completed on the mile headed bars relatively deeply embedded in reinfneed concrete columns of f,,= 10 LS (in which dis diameter ofthe stem). ‘An experimental program was prepared 0 determine the mini- ‘mum embedment depth that was necessary to safely design exte- soe beam-column joins using the headed bars. Throughout this {> Conerete breakout be N Fig. 1 Force transfer of headed bar. stFE cota sain tur Been op bie thesded bn) P I | Cconercte z Benn bstom bir N Fig, 2 External beam-column joint ‘stud, its assumed thatthe headed bars are the main reinforce ‘ment of the beam and embedded in the column as shown in Fig. 2. As the beam is subjected (o a negative moment, the hheaded bars becomte subjected to pullout. The pullout perfor- riance of the headed bars is likely dependant upon embedment depth, number of headed bars, cover thickness, distance between the adjacent headed bars, and amount of supplemen- tary reinforcement, in this case, the supplementary reinforee- ment consists of existing column reinforcement such as column main bars and column ties. It is emphasized that, although many research reslts of headed reinforcement have been published, the study describing the pullout behavior of group of headed bats is searce in the literature 2. Preparation for test 21 Head The bearing area of the head can significantly influence the anchorage capacity of the headed bar. ASTM A 970-98, for example, recommended a net head area of 1044 (in whieh is the cross-sectional area of the stem) for a welded head.'” The same document, however, no longer specifies a head area but requires mechanical ts of the head stating in 2004." Is gen erally agreed among researchers thatthe head area should be large enough to provide « soeure mechanical anchorage with reglissble slip ofthe hoad tis tressd tha, incase ofthe beam- column joint where the reinforcing bars are typically congested, the use of very large head is umeaistc. In these applications, smaller heads canbe use, and the anchorage rfes on the Bear- ing sess atthe head combined withthe bond stress over the development length* Small heads were designed and used in this study w prevent the congestion of head in the beam-col- «umn joint. The nt bearing arva ofthe head is only 328 shown in Table 1. A square head was fastened toa threaded end of the reinforcing haras shown in Fig 3. 2.2 Materials Normal sirenuth concrete was used. Concrete compressive strengths of test cylinders were 30.7 MPa or 32.4 MPa as shown in Table 2. Grade SD$20 (reinforing steel of which ‘yield strength, ¢ is 420 MPa) was used for headed bars, col- Table 1 Head geometiy Cross seston Disncior becbicd ‘gam » min) (ram) Die DN 6 Dis 18 Daz Seeger Das 3 a 2 Fig. 3 Headed bar unin main reinforcement, and column ties. 2.3 Test specimens Researchers generally agree that an embedment depth of 8 to lod, is required for a single healed bar embedded in normal strength plain concrete to develop a poilout stength over fi Hence, in eight pullout specimens, v0 D29 headed bars. were «embedded ina reinforeed coneretecolumurlike members using an embedment depth of 10d, Both headed bars were pulled outst once. Test objective was to determine ifthe two relatively closely spaced healed bars (centertowenter distance between adjacent heads, Sa = 6c) develop a pullout strength over J; with an ‘embedment depth of 10d, Several different reinforcing schemes in the columnlike members were used forthe same set of eight pallout specimens, The center-o-center distances between column ties (Sy) were 3d, 6 9, and none (foe the ease of no column ties), Also, two diffrent main renforeement ratios forthe column (= 0.56%, 1.13%) were used as summarized in Table 2 end Fig. 4. The objective was to determine if he pullout strengths and the Joau-tisplacement behaviors ofthe headed bars would change in relation othe different column reinforcement scheme Aadtional 24 pullout tests for D16, D22, D25, and D29 multi- ple headed bars were completed using deeper embedment depths ‘of 13d, and 154, The test vatables other than the emmbedment depth were similar to the previous test sets as summarized in Table 2 and Fig, 4, “The noon fr specimen inex wed inthis is shownas follows Qolumn-tke member 25 etc 2 headedbars le te spacing of 6d. C2513d-20-L —+ low mainreint-<1% 152 | Intemational Journal of Concrete Structuras and Materials (Vol.18 No.3E, December 2006)Table 2 Summary of pullout test variables and test results, 2029 [30.7 [250 (10d) 158 4.75da 174 Gd) | BGd) | 036] 457 | Sk7 | Seamgage Em [ a_i era tae [os wy ase Gv-I0eb2EL_| 2029 | ~307 390 (tay [198 4.754) 174 (dy | 056 [MB [643 fameas above 2D29 [307 [2M AU) 1384 75d} 174 (6d) | 8TGd) | 113 | 473] 87.7 fame as above ‘Ca9-i0ab-2E-M | 2D29_| “30-7 [290 (Ua 138 (4.78dh 4 C6) | 17416) [113] 478] SRO fame as above eis [ae [me aout) tan rey eS, _G | | (C22-15d6-3D-L | 3D22 324 [330(15as) | 1004 Sd,v)] 100 (4.5d,} 200 Ga, [0.65 S48 un GRASGIEL | 322 | 4 _[ 330054) 10014 Sd) [IO ASA 065 | 462] 9a (€22-1$db-3B-M | 3D22, 32.4 1330(15d3}[ 100 (4.54,) | 1004.5, 1004. Sat 129) 620 127 (€22-13db-3D-M | 3D22 324 [330(15d,)| 100 (4.54,) [100 (45d) 200 0d,) | 129 | 609 125, €25-13db-2C-M | 2D25, 324 | 330(13d,)] 100 (44) [200 (8a) |T50 (6c) 129 586 18 Caxisab-2EL | 2029 [324 [435 (Say) 138 ch 75A 1178 (A) | 056 | 664 “jy Seen eer Seu akoetnetericeget 2, peeenmmnrate ats Is noted dat all columm-like reinforced conerete members ‘were cast and tested inthe horizontal positon, and no axial loads ‘were applied during test for the cost concer, The test results, ‘which is determined in the absence of axial load, should be on the conservative side compared to those determined with exial load application, 2.4 Test setup and instrumentation The pullout test setup consisted ofa reaction fame and load- ing assembly shown in Fig. 5. The loading assembly was comn- posed of a ste] box, high-stength stel rod, and 1,000 N= ‘opecty hydraulic eylinder as show in Fig 0). Predsilled holes in the 25-mmethick ste plate located atthe botom side ofthe ste Box allowed the thread end ofthe heades ars to protrude inside the box so that a mut and w washer could be fisened 0 the bar as shown in Fi. 5(b). Multiple beaded bars were subjected ta pullout as the force was slowly applied using the hydraulic eyline ‘er operated by @ hand pump. The displacement was measured on top of the headed bars using LVDTs as shown in Fig. 5(b). The applied forec was measured using a pressure transcer. Signals fom the LVDIS and the pressure tanschicer were recorded using an electrenic data acquisition system with the sampling rte of ten data sels per second. In ight pullout specimens, § mm strain gages were installed on headed bars, column main reinforcement, and column ties as shown in Fig. 6, Strain gges onthe column tes sere installed assuming a 45° failure surfice. Three pairs of gauges were installed forthe pullout specimens with S,, of 34, as show in Fig (6(@), The stain gauges were named as “tio,” “te-2. and so cn tartng from the pair closest tothe headed bar. Two pairs of strain Intemational Joumal of Concrete Structures and Materials (Vol18 No.3E, Decemiver 2006) | 153ois 400 k——>] | Tis, DIO «0 | Ln ta L bo p= 048% (0) 22 eras ie, DIB 600 Main bar, baz par 065% eH 129% © pee 40 kK >| Ties DIB s00 Main bars } i pes pa= L13% Fig. 4 Cross sections of reinforced concrete columns. gauges were installed on the specimens with S,- of 6d, and one pair of stain ganges were installed on the specimen with S,_ of 53d as shown in Fig, 6(0). The postion o strain gages installed on the column main reinforcement was 60mm away from the headed har while the three puits of stain gages were installed on the stem as shown in Fig. 6(b): “bead-1” right above the head, ‘heade3" right below the conerete face, and ‘head-2" inthe middle between *head-I" and “heads Fig. 5 Pullout test setup 3. Test results 3.1 Pullout strength 3.1.1 Pullout tests: fig= 10dy, ‘The results of eight pullout tests for two of D29 headed bars Cy) kl @ Fig. 6 Position of strain gauges. 154 | International Journal of Concrete Structures and Materials (Vol. 18 No.3E, December 2006)with eg= 10d, 290 mm) as summarized in Table 2 reveal that the average pullout strength of two of D29 headed bars is 435 KN ‘or 338 MPa for cach headed bar, which is equivalent ro 80.6% of fu Test results indicate that the embedment depth of 10d, is not, Sufficient to make each of the relatively clesely installed headed ‘bars (Shai 174 unm or 6) develop the strength equivalent tof. In Table 2, the pullout strength (P,) ranges between 348 KN and 478 KN (or 64.5% and 88.6% of .) showing wide variations of the pullout resistance as determined from the same two of D29 hheaded bars because the pullout sttengths are influenced by supplementary reinforcement, i2, existence of the column rein= forcement. The pullout stength increases with decreasing column tiespacing (S,,) as shown in Fig, 7, n Table 2 and Fig, 7, the aver- age P, of two of D29 headed bars with the smallest te spacing (Sye~3dk) is 465 KN while that of two of D29 headed bars with ro column tie is 395 KN. The average P,, of the headed bars wit the smallest tte spacing is 17.7% greater than the average P, of those without any coum ties. ‘The test results are consistent in that P, keeps increasing, with eereasing columan tic spacing while this tendency is mare conspic- ‘uous for the pullout specimens with columa-like members of a lower main reinforcement ratio (= 0.56%6). Table 2 shows that, the pullout strengths ofthe two of 29 headed bars with two difer- cent column main reinforcement ratios are not the same. The aver= age P, of four tests with , = 0.56% is 402 KN while that of four tests with 2, = 1.13% is 467 KN, showing a difference of 16.2%, ‘The test results indicate that P, increases with ineresing amount of main reinforcement as well as decreasing column tie spacing. 3.1.2 Pullout tests: hey= 23d, “Thible 2 shows that the average P, determined from six tests oF ‘wo of D25 headed bars with hy; = 13d (330 mm) is 532 KN (or 525 MPa equivalent to 125% of for each headed bar). Although the average pullout strength is 125% of fin all six eorapleted pullout tess, the pullout strengths are smaller than 125% of fin st least two tests (C25-13db-2F-L and C25-13db-2F-M), indicating thatthe employed embedment depth of 13d, is not lange enough for the actual use of the headed bars in the field especially when. the bars are installed closer than 8, (Spa 200 mm or Sy this set of six tests). Table 2 and Fig. 8 show that P, ranges between 408 KN and (605 KN again, showing wide variations in the pullout strengths. In ‘Table 2, the average P, of two of D2S headed bars with the small- est te spacing (S,,=4.5¢}) is S84 KN, and that forthe specimens, without any’ column ties is 444 KN. There is a significant 31.5% difference in the pullout strengths, Test results again reveal that P, increases consistently with decreasing column tie spacing as shown in Table 2 and Fig. 8 ‘The pullout strengths of the two of D25 headed bars with two different column main reinforcement ratios are not the same, The average P, of thre tests with 0, is 507 KN while that of, three tests with 2, = 1.29% is 557 KN, showing a 9.9% difference, ‘The test results again indicate that P, increases with inoreasing, amount of main reinforcement 311.3 Pullout tests: y= 18d Although the headed bar embedment depth of 13d, resulted in an average pullout strength of 125% of f, as determined fiom the above six tes, the results obtained from the test of ‘wo headed bars that were as much as 8c, apart from each other (Siyai~ 8c) Iti possible that multiple headed bars with smaller centcr-ovcenter distance between adjacent heads may result in pullout strengths less than that currealy determined. Consequently, in the next three sets of pullout tests, an embed- ment depth of 15d, was tried for 3 DI6, 3 D22, and 2 D29 headed burs for pallor Table 2 shows that the average pullout strength determined from six tests of three of D16 headed bars with ny 15d, (240 mm) and Siege = Sy (144 man) is 331 KN (or 555 MPa ‘equivalent to 132% off, thr each headed bar). The average pullout strength determined fim six tests of two of 129 headed bars with thg~ VS (3S to) and Saas = Bey (174 mam) is 716 KN (oF ‘558 MPa equivalent to 133% of f, for each headed bat). Also, ‘Table 2 shows that the average pullout strength determined from six tests of three of D22 headed bars wit = 1S (330-mmt) and ds (100 mm) is S73KN (or 493MPa equivalent to Sia 800 700 coo | 90.55% 1.13% 509 400 Py, KN 300 200 100 Sad 6db_ db none Fig. 7 Summary of pullout strength: two of 029 headed bars, y= 100, Syags= Bay, Pep= 0.56% OF 1.13% 700 4+ 0.85% mI 29% 500 500 400 Py. kN 300 200 100 4.506 Bab none Sie Fig. 8 Summary of pullout strength: 2025 headed bars, hae 13d, Spas 8th, 2.s= 0.65% OF 1.28%, Intemational Journal of Concrete Structures and Materials (Vol 18 No.3E, December 2006) | 1550.48% mo 98% 200 300 100 4.5d0 none (@} 3016 headed bats, Se 48% or 0.96% 1900 00 200 58% 0° 1.1 Bio. 65% Bt 29% 600 500 400 200 200 160 6 4.50 900 none (6) 2022 hoscied bars, Sse = 4.8de, pe 0.65% or 1.20% Fig. 9 Summary of pullout strength: fhy= 150b. LI Paot foreach beaded bat), Fig. 9 shows the influence of existing column reinforcements, both main reinforcement ang! ties, on dhe pullout strength of the headed bars. The pullout strengths increase with decreasing col- ‘uma tle spacing. The test results aw consistent as shown in Figs, ‘9{a) anal (e), andl the deviation from: this tren for the case of wo (of 129 is probably an outlier in the test data in Fig. 96). Also, ‘Table 2 and Fig, 9 show thatthe pullout suengths of the headed bars installed in column-like members with greater amount of ‘main reinforcement are consistently higher than those installed in ccolumnike members with smaller amount of the main rinforce+ ‘ment. The difference is 7.2%, 11.3%, and 11.5% for dee of D16, 3122, and tvo of D29 headed bars, respectively: 3.2 Pullout load vs. displacement ‘The existing column reinforcement, both main bars and cob umn tes coitbute not only to increase the pllout strength of the add bars but also to improve the dase behavior in plou as shown in Fig. 10. Fig. 1D shows the laad-displacement plots of tree of D22 headed bars (atest set with the smallest centerto- comer distance tween adjavent ead. In Fig. 10(a). only one specimen named “C22-1Sdb-3E-L” develops the pullout sreneth equivalent to 125% of f, Two her specimens develop strength smaller than 125% off and also show less dae behaviors than ‘C22-1Sdb-3B-L", On the ser hand, Fig. 10(0) shows that the pullout stengihs practically equivalent to 125% off are attained inal tee specimens. All specimens also show ductile behaviors. The column main reinforcement must have infueneed he pullout behavions because, between the (Wo different sets ofthe pullout tests, only p, was changed from 0.65% t0 1.29%, The loads placement plos clearly reveal that column reinforcement is impor tant in improving the strength an the ductile pullout behavior of the ead bars. 00 022182 22-15 a 10 20 Dispiecemant, mn (ai 1 25Abety = B10KN ——622-1500-3E- eee . © 2 © © 50 Disalscemant, am (0) Fig. 10 Pullout load vs. displacement: 3D22, fy 5, 156 | Intemational Journal ef Conerete Stuctures and Matenals (Vol 18 No.3E, December 2008),3.3 Failure mode Th all pullout tests, the headed bars were typically pulled out very slowly asthe load was increased. A brite concrete breakout vwas not observed. No bar factures were encountered neither except for one test of DI6 bars where one of three headed bars ffactared in tension, Fig II shows a eack pattem obiained fom a pallot test specimen after the completion of test (to of D29). The first cracks developed at the head location andl progressed vprvard at an inclination approsimately between 35* and 45°. As the load was father inerased, spliting along the stem appeared. ‘The splting crack sometimes extended below the head postion at ‘aad close to the timate as shown in Fig. 11. 3.4 Strain gauge readings ‘The strain gauges were installed on the headed bar stem, col- ‘umn main reinforcement, and column tes fora set of eight pullout specimens as described previously, Readings ffom the strain ‘ress n stem, MPa 1,008 © 1,000 2,000 3,000 4.000 Miro stain (@) Cour tes 0 1,009 2,000 3,000 4,000 Moro atain (0) React bar Fig. 12 Stress in stem vs. strains in col. ies and headed bar; (629-10d-28-L, ‘gauges are shown in Figs, 12 and 13. The x-axis represents the strains developed in the column ties for Fig, 12(a) and stem fir Fig, 12{b) while the y-axis represents the stresses developed inthe ssem in Fig, 12, All column ties do not exhibit large stains until about 70% of the peak sires is reached as shown in Fig. 12a) ‘Column ties cose to the headed har (eI, te-2) develop large strains and contribute to increase P, while a column te far away from the headed bar (tie-3) develops very small stains and hence Provides no contribution oP, In Fig. 12(b), strains developed in ‘the upper part of the stem (close fo the applied force) are lrger than those developed inthe lower part (close to head). The differ ‘ences inthe strain values indicate that the bond between concrete ‘and stem is not broken even at the peak load, and the anchorage relies on the bearing stress at the smal! head combined with the bond tess over the sem. Un Fig 13, the sans represensthe stains developed inthe col tuna tes while the y-axis represents the stesses developed in the stem. The stress-strain plots determined ftom the three tess agin show that cofumn tes close tothe headed bar develop large strains and contribute t increase P,, In Fig. 14, the positions ofthe column tes, that developed lange strains and hence were effective in inreasing Pare shown. Fig, 14 = Micro strain (0) 029-100024M, = € 6 Micro stain (b)c2810db 208 & = o g om Micro strain (eyc2-odb-200 Fig. 13 Stress of stem vs. strain in column ties. Intemational Joumal of Concrete Structures and Materials (Vol.18 No.3E, December 2006) | 157reveals that the column tes that are located within a distance equal 100.45 * hy fom the healed bar contribute to P,- 3.5 Analysis of test results Fig. 15 shows a relationship between the column ties and dhe pullout strength of the headed bars. Fig. 15(a) shows the resus of eight eas of two of D29 headed bars with an emibeciment depth of Tay: xeaxis represents the volumetric ratio of the column ties 10 concrete. The pullout strength ineeases with the use of increasing amount of column ties as shown in Fig. 15(@). Fig. 15(b) shows the results of 24 tests with an embedment depth of 13d (2D2: and 15h, (D16, 3D22, and «wo of D29). The pallout strengths tie 290mm Zz 18 7290 * 0.4 Fig. 14 Position of effective column ties: C29-10db-2A. not effective & 1257- 0 opo2 ao 0060008 ont columm tes / concrete, ratio by vol (8) 2020 (ra= 104) # © 002 0008 9008 ong Opt concrete, ratio by vol (©) 2025(ru= 184, 8016, 9022, 2028 tha = 154) Fig. 15 Infuence of column ties on pullout strength. again inerease with the use of inereasing amount of column ties 08 shown in Fig. 15(0). In Fig. 15(b), « dashed line inate that the pullout strengths are always over 125% off, when the volumetic ratio ofthe column testo concrete is 0.6% or arate, Test results reveal that by employing an emibedment depth of 13d, or greater and, atthe sare time, nsng the column tes with a volumetie ratio (of 0.6% and grenter the headed bars develop pullout strength over 125% of J, the minimarn strength specified by the ACT code. ° Ie should be noced that, in Fig, 15(b) there are scatters in the pullout sents whea the volumetric ratios mul. The strengths vary because they must rely on less predictable behaviors of concrete only when no column ties are used. Wt is noted that the current conclusion is applicable only to headed bats installed in normal sirength concrete when the een- terio-center distance between adjacent heads is 4S, and greater. ‘The pullout behavior ofa group of headed bars with smaller cen- terto-cenler distance between adjacent heads must be investi gated in the follow-p research Fig. 16 shows a relationship betwoon the column main ret forcement and the pullout strength of the headed! bars. The pullout strengths consistently increase with increasing main teinforcement ratios as shown in Fig. 16. The average slope of the linear regres sion lines in Fig. 16 is 21.3 MPa of main bars while the mini- rum is 182 MPa% of main bers. Test results suggest that the pullout strength ofthe headed bars can incwease by about 18 MPa for one percent increment of column main hs. I shouk! be noted here that the main reinforeement ratio up to 1.29% was tested in the curent study. Finally, Table 2 shows that, for the specimens with an cembedment depth of 13d, or 15d and with both column ties and a column main reinforcement ratio of 0.96% and greater, the pullout strength is always over 125% of ,. Consequently, It is recommended to use at least 0.6% of column ties by volume, 1% or greater amount of eofurnn main bars, and an embedment depth of 13, or greater to guarantee the pullout strength of individual headed bars over 125% of fan the ductile behav ior of muliple headed bars. 4. Conclusions Pullout tests were performed on the multiple headed bars rela- o 08 1 is 2 Main reinforcement rato, % Fig, 16 Infuence ofthe main reinforcement on pullout strength, 158 | Intemational Journal of Concrete Structures and Materials (Vol.18 No.3E, December 2006)lively deeply embedded in reinforced concrete columns with an objective to determine the minimum embedment depth that was necessary 10 safely design exterior beam-column joints using headed bars. twas assumed thatthe headed bars were the main reinforcement ofthe beam and embedded in the cokuman a shown in Fig. 2, The pullout test variables were embedment depth, con- ‘er4o-center distance between adjacent heads, column tie spacing, snd main reinforcement rato. Pullout test results of headed bars insalled in normal strength concrete revealed the fllowings 1) The embedment depth of 10d is noe sulicent to have the relatively closely installed headed bars develop the strength core- sponding tof. 2) The pullout strength of individual headed bars is reduced ‘when a ckasely spaced headed bur group i subjected to plot 3) The pullout strengths of headed bars increase with the use of | supplementary reinforcement In this study, the pullout strengths inreased with increasing amount of column ties (or reinforeement «signed to run in a direction paral! to the headed bars, refer to Fig.2) 4) The pallout siengihs of headed bars also increase with ineresing amount of main rvinforcement (or reinforcement designed to run perpendicular to the headed bas, refer to Fig. 2). 5) For a group of closety.spaced headed brs, itis recom- mended to use at east 0.6% of column ties by volume, 1% or sreater amount of column main bars, and an embedment depth of 13d, or greater simultaneously to guarantee the pullout strength of individual headed bars over 125% of f, and the ductile load-ls- placement behavior 6) The above conclusion is applicable when the center-o-center distance between adjacent heads is 4.5, or greater. The pullout ‘behavior of a group of headed bars with smaller center4o-enter distance between adjacent heads shouldbe investigated in the fl- low-up research, In addition, readings obtained from the sain gauges installed con the headed bar stem, column main reinfarcement, and column ties reveal the followings: 1) The bond between concrete and stem was not broken at the peak load, an! the anchorage can rely on the bearing stress at the ‘small head combined with the bond stress over the stem. 2) Most cohunn tes donot exhibit lage strains until about 70% ofthe peak sires is reached, Thus, column tis ar likely 1 con- ‘wibue to increasing the ultimate sength of the headed bars and donot playa signifcant roe inthe service stage. 3) Only those colin tes that are located within a distmoe of | (0.45 times the embedment depth fiom the headed bar eontibute to increase the pullout strength, References 1. Ghali, A. and Youakim, 8. A., “Headed Studs in Concrete State of the Art.” ACI Structural Journal, Vol.102, No.5, 2008, p-657-667. 2. DeVries, R. A., Anchorage of Headed Reinforcement in Concrete, Ph.D. Dissertation, The Univ. of Texas at Austin, 1996, 293pp. 3, Bashandy, T. R B., Application of Headed Bars ku Com- crete Members, Ph.D. Dissertation, The Univ, of Texas at Austin, 1996, 302pp. 4. Thompson, M. K., The Anchorage Behavior of Headed Reinforcement in CCT Nodes and Lap Splices, PhD. Disser- tation, The Univ. of Texas at Austin, 2002, $01pp. 5. Choi, D. U., Hong, S. G, and Lee, C. ¥, “Test of Headed Reinforeement in Pullout” KCI Conerete Journal, Vol. 14, No.3, 2002, pp.102~110. 6, Dilger, W. H. and Ghali, A., “Double-Heead Studs as Ties in Concrete.” ACI Concrete International, 1997, pp-59-66. 7. DeVries, RA., Jitsa, L0., and Bashandy, T, “Anchorage Capacity in Concrete of Headed Reinforcement with Shallow Embecments.” ACI Sructural Journal, Vol96, NoS, 1999, p.728-736. 8. Thompson, K. H,, Ziehl, M. J, Jia, J, 0. and Breen, J. Bx “CCT Nodes Anchored by Headed Bars-Part 1: Behavior of "Nodes," ACI Structural Jona, Vol 102, No, 2005, pp.808-815. 9, Wallace, J. W., “Headed Reinforcement A Viable Option,” ACT Concrete ternational, 1997, pp47~$3. 10, Wallace, J. W., MeCoanell,S. W., Gupia, and Cote, PA, “Use of Headed Reinforcement in Beam-Column Joints Sub- jected to Earthquake Loads,” ACI Structural Journal, Vol95, No.5, 1998, pp.390-606. IL, Patk, H. K., Yoon, ¥. S., and Kim, ¥. H, “The effect of Fread plate dewils onthe pull-out behaviour of headed bars." Mag- azine of Conorete Research, Vol.58, No.6, 2003, pp485~496, 12, ASIM A 970/A 970M-98, Sundard Specification for Welded or Forged Headed Bars for Concrete Reinforcement, American Society for Testing. and Materials, 1998. 13. ASTM A 970/A 9T0M-04, Standard Specificarion for Headed Steel Bars for Conerete Redrfarcement, American Sock ‘ty for Testing and Materials, 2004, 14. ACI Commitee 349, Code Requirements for Nuclear Safety Related Concrete Structures (349-97), American Concrete Institute, 1997, 15. ACI Committee 318, Building Code Requirements for Structural Concrete (318M-05) and Commentary (318RMIS), American Concrete Institute, 2005, International Journal of Concrete Structures and Materials (Vol 18 No.3E, December 2006) | 159