 Printer EmailIn storage racks, hook-in end connectors are used to make beam to column connections. The semi-rigid nature of this connection is primarily due to distortion of the column walls, tearing of the column perforation, and distortion of the beam end connector. The storage rack stability depends significantly on the behavior of this connection, and therefore, it is important to have a proper way of predicting it. Designs of these connections vary widely, making it impossible to develop a general analytic model. Instead, beam to column connection tests are usually done to determine the relationship of the moment at the joint M and the change in angle between the column and the connecting beam q. A complete study is performed to find out the flexibility of the beam-column connector with experimental studies and this was followed by full scale frame test to compare the results obtained from the two tests. Finite element analyses of both the tests and also of full scale test was carried out using ANSYS [1] software. The experimental results along with finite element analysis are presented here in this paper. In the past, researchers have investigated various different types of connectors used in pallet racking systems. Baldassino and Bernuzzi [3] have given the influence of beam-to-column joint modeling on the overall frame response in their numerical study. Bernuzzi, and Castiglioni [4] carried out some experiments on the behavior of flexible connections, when subjected to cyclic reversal loading and investigated the effect of joint performance on the overall frame response. Bolts, screws, blind rivets or cartridge fired pins are commonly used in joints between cold-formed sections or for steel sheet connections. Recent research, by Rogers and Hancock [9,10] have studied the extension of actual design specifications to high strength steel sheets. Dubina and Zaharia [6], by means of experimental tests and numerical simulations, have studied the rotational capacity of bolted joints between cold-formed C sections with the objective to use, in the design of latticed beams, the advantage of the semi-rigid behaviour of the joints. Fan et al. [7] have carried out numerical simulations using a finite element program with strong features in the non-linear computations with large inelastic strains and contact elements to improve ourknowledge of the real behaviour of the joints.
The moment carrying capacity of the beam end connector was determined by a simple cantilever (beam-to-column connection) test Markazi et al. [8]. Tan et al. [11] have given a mathematical model to represent the connection behavior after carrying out experimental studies on the non-linear behavior of their connector. There are many types of beam-end-connectors with different geometry of the connected members available in literature. The theoretical approaches to evaluate the performance of such joints are currently available only for few simple connectors. As a consequence, such type of connector design needs proper experimental evaluation and also numerical studies. In this paper, two such experimental set-ups are presented to find out the flexibility of the connector developed, which is verified by a full scale test. Cantilever test The experimental set-up used for this test is shown in Fig. 1. The column is fixed at both the ends, and a box section beam is connected to the column with the help of beam connector. The out-of-plane movement of the cantilever beam with connector is prevented by a using a side plate. Vertical load is applied at the free end of the beam in steps and the corresponding rotations of the beam and the column are calculated from observed deflections using the dial gauges for each load step. Moment acting on the connection is given as the product of applied load P and the lever arm d. In the cantilever test, the connection stiffness F is given as moment per unit rotation. The test and finite element results are compared in Fig. 8. Double cantilever test set-up Fig. 2 shows experimental set-up for double cantilever test. The ends of box section beams are pin connected using two vertical channel sections. The other two ends of beam are connected to the upright section using flexible connectors (Fig. 2). The load is applied in vertical direction gradually on the top of upright in a UTM, and the corresponding rotations of the beams were calculated from observed deflections of dial
gauges. The entire assembly was prevented from moving outof-plane by using a side plate. Strain gauges were fixed on the inner face of the column near the slots to find the actual stresses coming on the upright. The test and finite element results are Frame testing Full scale frame testing is done to asses the behavior of all the components forming the frame with special emphasis on the beam to column connector developed. The aim is to asses the behavior of connector and continuity of the frame in actual conditions. The maximum load was limited to permissible values as per deflection criteria (of beam), moment capacity of the connector, and stress criteria of the beam and upright. The cross sections of column and box beam used are shown in Fig. 4. The frame is of two bay and two stories, braced together by cross bracings. The experimental set-up along with various strain gauge locations during the test is shown in Fig. 3. Total frame height is 3.0 m with each storey of 1.5 m, clear span of each bay is 2.7 m and bay width in transverse direction is 1.0 m (outer to outer). Bracings bolted to the upright are made of ‘C’ section. Four strain rosettes and twelve strain gauges (total of twenty four points) were fixed on the beams and upright (see Fig. 3.). Generally in pallet racking systems the beam is loaded by four point loads. Hence, a four point load is applied in this test (see Fig. 3). MS plates of 3.15 mm thickness and 2.50! 1.25 m size each weighing approximately 740 N are used for loading. Loading was done on bottom storey only. This kind of loading has advantage over conventional sand bag loading, with respect to better load distribution, accuracy and cleanliness. Total of 34 plates were used with 17 in each bay. The test and finite element analysis results are compared in Table 4. Finite element modeling In order to study, the behavior of the connection, non-linear finite element analysis was performed using ANSYS [1]. A finite element model (Fig. 5) that best represents the behavior of the connection was developed. Geometry, boundary and loading conditions of the finite element model were made as close to the cantilever test as possible (Fig. 2). The column was fixed in all degrees of freedom at both the ends to represent the ends of the column that were welded to immovable supports. Contact surfaces were defined using CONTA173 element between the connector plate and the column to represent their interaction. The hook in connector was modeled by the use of SOLID43 element to provide connection between the connector and the column. The column and beam section were modeled using SHELL63 element. The material model was elastic–plastic with strain hardening capabilities and having the properties shown in Table 1. Properties of various finite elements used for the analysis are given in Table 2. A concentrated load was applied at 900 mm. from the connection on the beam to represent the jack load. The load was gradually applied in steps and the rotation of the connector was monitored using four nodes as shown in Fig. 6. The total
load applied here was equal to that applied during the test. The moment–rotation curve resulting from the finite element analysis and the physical test are compared in Fig. 8. The deformed shape in the finite element analysis with Von Mises stress is shown in Fig. 7. Readmore please click here to download paper from Central Library Indian Institute of Technology Bombay Author: K.M. Bajoria and R.S. Talikoti Department of Civil Engineering, Indian Institute of Technology Bombay, Powai, Mumbai 400 076, India |
|||||
