Design charts for connection capacity with one 

Context in source publication

Context 1

... beam-to-column connections are generally subjected to axial and shear forces in addition to bending and torsional moments. Design equation for end plate thickness of these connections has been extensively investigated [e.g. 1-3]. Furthermore, behaviour of end plates and bolts were also numerically and experimentally investigated [e.g. 4-6] to enhance the design equations and investigate the effect of prying force. It was argued [1,7] that pretensioned bolts increase the stiffness of the connection at levels of bending moment below the ultimate values. In connections with thick end plates, the influence of pretensioning of the bolts on their early stiffness is more than that of a thin end plate. Then, a guide for the design and analysis of flushed and extended end-plate moment connections is presented [8]. Despite all these extensive investigation, current methods of estimating bolt force and end plate capacity for these connections still do not consider the interaction between bolt elongation and plate rotation. Here, simple charts and equations considering the said interaction are introduced for the design of flushed end plate connections with full details of the investigation leading to these equations are given elsewhere [9]. For these connections, a finite element model has been developed and verified against the experimental results. Then, a parametric study is carried out for the connection to study the effect of different parameters on the connection behaviour. A comparative study between the investigation results and those obtained via the AISC design guide No. 16 provisions [10] is performed. An experimental programme has been conducted to investigate the behaviour of flushed end plate rigid connections subjected to pure bending moment. The test specimens consist of a column connected to two constant beams using high strength bolts. Four full-scale test specimens with the configurations shown in Table 1 are performed. Tested connections are subjected to pure bending moment via the loading configuration shown in Fig. 2. Electrical strain gauges are connected to the bolt and to the end plate. A load cell is used as an indicator of the applied load and a dial gauge of accuracy 0.01 mm is used to measure the vertical displacement of the columns. Specimens are loaded incrementally by a rate of 9.81 kN/min (1.0 t/min) until failure takes place. The material properties of the component plates of beams, column and end plates are derived by conducting standard tensile tests on three test coupons taken from the manufactured specimens. The bolts’ material properties are derived for each bolt diameter by conducting six tests. For each bolt diameter, three tensile tests are executed on coupons from the bolts, while the other three tests are conducted on a complete bolt. From these tests, the steel yield and ultimate stress were found to be 274 MPa and 372 MPa, respectively. The ultimate stress of M12, M16 and M20 bolts were found to be 947 MPa, 1109 MPa and 1099 MPa, respectively. Specimen F1, showed an excessive elongation in the tension side bolts followed by bolt rupture (Fig. 3). No deformation in the column flanges was recorded. Specimen F2 showed a considerable flange deformation during loading. In this specimen, plate deformed gradually due to biaxial bending moments, while bolt prying force appears at the end of the horizontal axis of the inner bolt. The bolts in the tension side suffered from large rotations due to plate rotation causing excessive stresses. Specimen F3 showed the same behaviour of specimen F2, but with different value of failure load. In specimen F4, failure took place due to bolt failure associated with small plate deformations. A comparison between the failure modes recorded from the experimental investigation and that of finite element model is shown in Fig. 3 and Table 2. Furthermore, the finite element model results have also been verified against three connection specimens of another the experimental work [11] which have double rows of bolts on each side. The model outcomes only deviated by ±5% from the experimental results [9]. Hence, the proposed finite element model shows a good agreement with the experimental analyses. M24 bolts and edge distance 3.0 times the bolt diameter, two more types of steel are studied: Steel 24/35 and steel 28/44 to spot out the change in behaviour and moment capacity of the connection for lower grade steel. Also for specimen with M12 and M24 with edge distance 1.5 the bolt diameter, grades 8.8 bolts are also investigated [9]. Beams of 400mm depth are investigated for connection having one row of bolts at tension side of connection, while beams with 600mm depth are studied for connection with and or two rows of bolts at tension side. Two failure modes are generally recorded in the numerically analysed connections: bolt failure for connections with relatively thick end plate and plate failure for relatively thin end plate. For the later one, the force in the tension bolt increases with the increase of the end plate thickness due to the increase of the end plate resistance. For relatively thick end plates, where the failure turns to be in bolts, the tensile forces in the bolts increase although the moment capacity of the connection remains constant. Axial forces in bolts do not reach their nominal axial resistance. This is because a bending rotation in the bolt shank is recorded and the bolt suffers bending moment as well as tensile force as it rotates with the end plate rotation (Fig. 5). Thus, the bolts suffer a combination of axial tensile Fig. 5. Plate deformation/bolt rotation force and flexure, especially for thin end plates which exhibit large deformations. For specimens with relatively thick end plates and relatively small bolt diameters, very small deformation of the end plate is recorded and hence, no bolt’s rotation is encountered; consequently, bolts are only subjected to axial tensile force. For specimens with two- rows of bolt and relatively thick end plates, forces in outer bolts are larger than those of the inner bolts as the plate rotates as a rigid body without noticeable deformation. The centre of rotation is found to be at the centre line of the beam compression flange. Figure 6 presents the sample results for the moment capacity of the analysed connections. Regardless of the end plate thickness, it is found that the larger the edge distances the smaller the ultimate moment. Each curve of Fig. 6 is divided into two main segments: the first segment represents end plate failure where the moment capacity increases linearly with the increase of the plate thickness while the second segment represents bolt rupture where the moment capacity is almost constant. For specimens with M12 bolts and only one bolt row, the first segment of the curve nearly does not exist because as this bolt diameter is small enough to maintain the connection failure by bolt rupture even for specimens with thin head plates. Figure 6 shows the variation of the connection moment capacity for a constant edge distance (e=3d) with changing other parameters. It is evident that for thin end plates, and despite have a failure mode governed by plate failure, bolt diameter has a significant effect on the moment capacity of the connection. This is because the increase of bolt resistance decreases the plate deformations. For thicker plates, the ratio between ultimate moments for each bolt is nearly the same as the ratio between the bolts areas while for thin plates, the previous relation is not valid. Full details of the analysis results are given elsewhere [9]. Figure 7 shows typical sample of the comparative study [9] performed on the connection capacity resulting from the finite element analysis and that obtained using AISC Design Guide No. 16 provisions: AISC provisions assume that there is no induced prying force except when the stress in the plate reaches 90% of its yield strength. It is evident that the AISC provisions overestimate the capacity for connections that fail by bolt rupture and underestimate it for connections fail by plate yielding. This discrepancy for connections failing by bolt rupture is attributed to the fact that unlike the numerical analysis, the connection capacity obtained using the AISC provisions neglect the The results of the numerical analysis are used to build a series of design charts for the capacity of the connections with one and two bolt rows (e.g. Fig. 8). The charts provide an effective, easy and accurate technique to choose the bolt diameter and the corresponding end plate thickness for a given moment capacity M and beam height h. The effects of the prying force, the bolt elongation, the bolt rotation and the head plate deformation are all included in these charts [9]. Another, semi-graphical design aid is developed where the connections capacity M is calculated as ...