《AISC kulak2003Q2.pdf》由会员分享,可在线阅读,更多相关《AISC kulak2003Q2.pdf(10页珍藏版)》请在三一文库上搜索。
1、ENGINEERING JOURNAL / SECOND QUARTER / 2003 / 89 M ost connections use a single fastening device (bolts, rivets, welds) to make the connection between adjoining members. However, it is sometimes necessary, or desirable, to combine different fasteners in a single connec- tion. This is particularly tr
2、ue when making alterations to existing structures, and the usual situation is the need to add one fastener type to a joint that was made using another type. For example, in order to increase the capacity of an existing riveted joint, some rivets can be replaced by high- strength bolts. In another si
3、tuation, fillet welds can be added to an existing bolted joint, again so that an increased load can be carried. Usually, it is necessary to do such alterations while the member is already loaded. Joints that combine two (or more) types of fastening ele- ments can take different forms. For example, a
4、 beam-to-col- umn connection that uses a beam web shear tab welded to the column face and bolted tee-stubs that connect the beam flanges to the column is a combination bolted-welded joint. Such arrangements can be handled by simply considering the forces that must be transferred at each location. It
5、 is not as apparent, however, how to determine the capacity of a joint that uses bolts and welds, for example, that act in the same shear plane. This would be the case if an existing bolted truss member splice were to be reinforced with fillet welds. The discussion presented here will be limited to
6、the particular case where the connectors act in the same shear plane and in which the connection elements are specifically high-strength bolts and fillet welds. Statement of Problem The capacity of each connector type in a shear-splice type of joint is reflected by its shear strength and shear defor
7、ma- tion characteristics. When two or more types of fasteners share the load, it is clear that it is not satisfactory to simply combine the ultimate strength of each individual fastener type. Each of the different fastening elements has a differ- ent ductility and, in general, each reaches its ultim
8、ate strength at a different value of overall connection deforma- tion. The characteristics of a bolt, transverse fillet weld, and longitudinal fillet weld shown in Figure 1 are representa- tive. (It should be noted that the relative strength of the transverse and longitudinal welds is a function of
9、their respective lengths. Similarly, the relative position of the bolt curve depends on the size of the bolt.) In order to determine the ultimate strength of a combina- tion joint, the load vs. deformation characteristics of each fastener type first must be established. The way in which these indivi
10、dual elements interact with each other and how much resistance each contributes to the connection can then be determined. REVIEW OF PREVIOUS WORK Research by Holtz and Kulak A series of tests involving combination joints was per- formed by Holtz and Kulak (1970). In the only relevant por- tion, doub
11、le lap shear splices using bolts and welds in the same shear plane were loaded in axial tension. Each con- nection contained either one or two 20 mm diameter A325 high-strength bolts in combination with fillet welds of 6 mm nominal leg size (E410 electrode). Nine specimens were tested, and there wer
12、e three identical splices in each of the following groups: Strength of Joints that Combine Bolts and Welds GEOFFREY L. KULAK and GILBERT Y. GRONDIN Geoffrey L. Kulak is professor emeritus, department of civil and environmental engineering, University of Alberta, Edmonton, Canada. Gilbert Y. Grondin
13、is associate professor, department of civil and environmental engineering, University of Alberta, Edmonton, Canada. Fig. 1. Representative load vs. deformation characteristics. 90 / ENGINEERING JOURNAL / SECOND QUARTER / 2003 1. Longitudinal fillet welds combined with two high- strength bolts. No cl
14、earance was provided between the bolts and holes, i.e., so-called fitted bolts. Bolts were snug-tightened only. 2. Same as Case 1 except that clearance between bolts and holes was provided at the standard nominal value, 1.6 mm (1/16in.), and the bolts were pretensioned. 3. Transverse fillet welds co
15、mbined with one high-strength bolt. Clearance between bolt and hole was provided at the standard value, 1.6 mm. Bolts were pretensioned. In the first series, the bolts were assumed to be in bearing from the beginning of the test (because of the so-called fit- ted condition), but in the other two ser
16、ies, the bearing con- dition at the start of the test was unknown. Holtz and Kulak made predictions of the load vs. defor- mation behavior of their combination joints by using the Fisher model for component response (Fisher, 1965). The specific deformation vs. strength relationships they used for th
17、eir bolts and welds were the same as those used by Craw- ford and Kulak (1971) and Butler, Pal, and Kulak (1972), respectively. (The Holtz and Kulak test bolts and the Craw- ford and Kulak bolts came from the same lot. The weld electrode specification was the same in both the Holtz and Kulak study a
18、nd in the Butler et al. investigation.) Using these characteristics, the investigators calculated the resist- ance contributed by each fastening element for each notional increment of deformation of the combination joint. In the analysis, the deformation of the bolt was adjusted for those connection
19、s in which bolt hole clearance was present. By summing the calculated resistances of the individual fas- tening elements, a prediction of the ultimate resistance of the joint is produced. For the first series, fitted bolts and longitudinal welds, the ratio of predicted ultimate load to test ultimate
20、 load was 1.21, and the standard deviation was 0.08. This poor, and non-conservative, result reflects the inadequacy of the assumption that the bolts contribute resistance right from the start of the test. For the tests in which bolts installed in holes of standard 1.6 mm clearance were combined wit
21、h longitudinal welds, the ratio of predicted load-to-test load was 1.00, and the standard deviation was 0.02. In the last series, bolts in standard holes combined with transverse welds, the ratio was 0.93, and the standard deviation was 0.03. In these last two cases, the ratio of predicted ultimate
22、load to test ultimate load is reasonable and is conservative. Holtz and Kulak concluded that only a small amount of load was carried by the bolts at service loads (taken as about 1/3 ultimate), that the effect of friction when pretensioned bolts are used cannot be predicted reliably, and that bolts
23、do not work very well in combination with transverse welds because of the limited ductility of the latter. Research by Jarosch and Bowman Jarosch and Bowman (1986) performed a series of physical tests on combination joints similar to the tests performed by Holtz and Kulak (1970). All connections use
24、d standard hole clearance, and the bearing condition of the bolts at the start of the test was unknown. Six combination joints were tested, two in each of the following groups: 1. Two high-strength bolts in combination with longitudi- nal fillet welds. 2. Two high-strength bolts in combination with
25、transverse fillet welds. 3. Two high-strength bolts in combination with both longi- tudinal and transverse fillet welds. Jarosch and Bowman used values of fastener ultimate strength and ultimate deformation as obtained in calibration tests, but used the same regression coefficients for both bolts an
26、d welds as developed by Holtz and Kulak. Their predictions of test load were generally satisfactory. Com- bining all six specimens, the ratio of predicted ultimate load to the test load was 1.09, and the standard deviation was 0.07. As had been suggested by Holtz and Kulak, Jarosch and Bowman also r
27、ecommended that transverse welds should not be used in combination with high-strength bolts because their limited ductility does not permit the development of any significant shear load in the bolts. Research by Manuel and Kulak These researchers carried out a series of 24 tests of tension lap splic
28、es that had bolts and welds in the same shear plane (Manuel and Kulak, 2000). The parameters encompassed the following conditions: 1. Bolts pretensioned or snug-tight only. 2. Bolts combined with longitudinal fillet welds only, com- bined with transverse fillet welds only, or combined with both long
29、itudinal and transverse fillet welds. 3. Position of the bolts relative to their holes established at the start of the test. Bolts were either in a location in which no slip was possible, “positive bearing,” or in a location such that maximum slip could take place, “neg- ative bearing.” There were t
30、wo replicate tests in each category. For example, there were two tests of a specimen that contained four 20 mm (3/4-in.) diameter ASTM A325 bolts in positive bearing combined with 560 mm of longitudinal fillet weld of 6 mm nominal leg size deposited using E48018-1 filler metal. In addition to these
31、full-size specimen tests, ancillary tests were carried out to determine the load vs. deformation ENGINEERING JOURNAL / SECOND QUARTER / 2003 / 91 characteristics of the fastener components. Bolt shear tests used bolts from the same lot of bolts subsequently installed in the full-size pieces and plat
32、e taken from the same source as used in the full-size tests. Likewise, the plate used to make up the weld coupon test specimens was taken from the same stock of plate used to fabricate the full-size speci- mens. The weld electrodes used in the coupon tests came from the same stock as used subsequent
33、ly to fabricate the full-size specimens. Manuel and Kulak reported the following conclusions: 1. When transverse fillet welds are used in combination with high-strength bolts, the deformation at the time of weld fracture is such that almost no bolt shear strength is developed. (This was also observe
34、d by Holtz and Kulak and by Jarosch and Bowman.) 2. The term “slip-critical,” used in the design of bolted joints, has no meaning when high-strength bolts are combined with fillet welds. When welds are present, there can be no slip that places the bolts uniquely into bearing. Once the welds fracture
35、, the situation simply reverts to that of a bolted joint with no welds. 3. If the bolts in a combined bolted-welded joint are known to be in a condition of negative bearing, the ultimate strain of even longitudinal fillet welds will not be suffi- cient to mobilize any significant bolt shear strength
36、. Manuel and Kulak modeled their physical test results using the load vs. deformation characteristics of the indi- vidual components of the joints and by making an estimate of the frictional resistance present. Regression analysis of the individual bolt shear tests and weld shear tests was used to d
37、evelop the necessary relationships (Fisher, 1965). Once these expressions were obtained, the predictions for the physical tests were based ona: where R ult joint=ultimate strength of combination joint, kN Rfriction=plate friction resistance, kN Rbolts=bolt shear resistance, kN R trans=transverse wel
38、d shear resistance, kN Rlong=longitudinal weld shear resistance, kN At any particular value of the joint deformation, each fas- tening element potentially will contribute to the load resist- ance, and the least ductile fastener component in a connection will govern the deformation at which the initi
39、al fracture occurs. In the case of the bolts, the condition of negative bearing or positive bearing was taken into account. As already noted, the contribution of bolts in negative bear- ing was taken as zero and the friction contribution when snug-tightened bolts were used also was neglected. Using
40、this approach, the following results were obtained by Manuel and Kulak: 1. Longitudinal welds + bolts (various conditions of bolt bearing and bolt pretension); predicted load/actual test load = 0.98 (8 cases) 2. Transverse welds + bolts (various conditions of bolt bearing and bolt pretension);predic
41、ted load/actual test load = 1.02 (6 cases) 3. Both longitudinal and transverse weld + bolts (various conditions of bolt bearing and bolt pretension);predicted load/actual test load = 1.09 (6 cases) (Certain tests in the Manuel and Kulak study had instru- mentation problems and were rejected for purp
42、oses of this summary analysis.) The results of the analysis showed that, overall, the ulti- mate strength of a combined bolted-welded joint can be cal- culated with a high degree of accuracy when the characteristics of the individual components are known. However, it is unlikely that a designer will
43、 have the load vs. deformation fastener characteristics that are necessary to make the ultimate load prediction. Recognizing this, Manuel and Kulak used their results to provide simplified rules. (Because their tests included specimens with trans- verse welds, this case is included in the Manuel and
44、 Kulak model. However, they recommended that transverse fillet welds not be used in conjunction with bolts, i.e., the strength of such joints would be taken as the larger of the bolt strength or transverse weld strength alone.) Frictional Forces Friction between the plates at the time of joint ultim
45、ate load is quite variable, as would be expected. However, according to the Manuel and Kulak tests, a reasonable lower bound is where Pslip=the slip resistance of a bolted joint, kN (It can be noted that frictional force will be present as the ultimate capacity of the combined bolted-welded joint is
46、 reached because some bolt pretension still exists at this stage. It is only later, at the time of bolt ultimate shear, that the bolt pretension has decreased to a negligible value.) Transverse Weld Shear When transverse welds are used in a combination joint, the shear resistance of the joint is the
47、 largest of the transverse weld shear strength or the bolt shear strength. For a given Rult joint= R friction+ R bolts+ R trans+ R long (1) 0.25 frictionslip RP=(2) aAlthough SI units are used in this paper, the concepts and conclusions are independent of the system of units used. 92 / ENGINEERING J
48、OURNAL / SECOND QUARTER / 2003 Struik, 1987) is that about one-half a hole clearance of slip is likely, i.e., 0.8 mm. Using this assumption, Manuel and Kulak estimated that the bolts contribute about 49 percent (rounded to 50 percent) of their ultimate shear strength. The equations for connection re
49、sistance contributed by bolt shear for n bolts are then as follows: For combination joints consisting of bolts and transverse welds (with or without longitudinal welds) It should be noted that, once the transverse weld has fractured in a joint that combines bolts and transverse welds, the joint is no longer a combination joint and the strength of the joint is now limited to that of the bolts. Thus, the strength of a joint that combines bolts and transverse weld is effectively the maximum of the trans- verse weld strength or the strength of the bolts. For combination joints with only longit
链接地址:https://www.31doc.com/p-3730783.html