Elsevier

Engineering Structures

Volume 33, Issue 5, May 2011, Pages 1710-1722
Engineering Structures

Blind bolted moment connection to unfilled hollow section columns using extended T-stub with back face support

https://doi.org/10.1016/j.engstruct.2011.02.008Get rights and content

Abstract

This paper presents the results of an experimental and analytical study into a new bolted moment connection between unfilled hollow section columns and open section beams, referred to herein as the extended T-stub connection with back face support. The connection is comprised of T-stubs connecting the top and bottom flanges of the beam to the face of the column and channels connecting the T-stubs to a backplate at the back face of the hollow section column. The addition of the channels to a standard T-stub connection helps to distribute the beam flange tension load to the back face of the hollow section column thus reducing demand on the flexible column face which often will compromise the stiffness of the connection. The extended T-stub connection with back face support was found to be comparable to a previously tested side connection and is much stiffer than a typical face connection. Comparison of the test results and three dimensional finite element (FE) model indicates that FE analysis can be used to predict the connection behaviour with sufficient accuracy. The experimental results and analytical sensitivity analysis demonstrate that this new connection can offer a robust and reliable moment connection for application in the low rise construction industry.

Introduction

The use of hollow sections as columns in low rise structures has become increasingly popular due to their structural efficiency. However it is not possible to provide fully bolted connections on site for hollow section members with standard structural bolts due to the difficulty in obtaining access to tighten nuts of the standard bolts inside these members. Hence welding remains the common practice used to enable connection to hollow section members. The development of blind bolts in recent years has provided an alternative means of connection for the hollow section members. The HUCK blind bolt [1], Flowdrill [2] and Lindapter Hollobolt [3] are among some of the blind bolting systems that have been adopted in the United States of America, Europe and Japan. France et al. [4], [5], [6] conducted intensive investigations on the behaviour of simple and moment end plate connections with the Flowdrill system. In more recent years, Elghazouli et al. [7] and Wang et al. [8] investigated blind bolted connections with the Lindapter Hollobolt.

In Australia, Ajax Engineered Fasteners has developed a type of blind bolting system called the ONESIDE [9]. Details of the blind bolt and its simple installation procedure are given on the Ajax Engineered Fasteners’ website [9] and in Lee et al. [10]. The main current application of ONESIDE is in reinforcing telecommunication towers made of thin walled hollow steel sections in the United States of America [11]. Stiffening ribs are blind bolted to the tower providing an effective means of strengthening the tower without disruption to the operation of the tower. However, the application of ONESIDE in building construction in Australia and internationally is still developing.

This paper is part of an ongoing research project in collaboration with Ajax Engineered Fasteners and Australian Tube Mills to develop various blind bolted connections to unfilled hollow section columns with the Ajax ONESIDE blind bolt. Several tests have been conducted, initially to investigate the behaviour and performance of a simple T-stub connection to the column face [10]. It was found that the inherent flexibility of the column face compromises the connection stiffness. The next stage of the experimental programme focused on transferring loads from the beam flanges to the side faces of the column to avoid the flexible column face deformation. This was achieved using channel sections bolted to the beam flanges which are in turn bolted to the sides of the column using straps (side plates) [12]. It was found from both experimental and analytical work that the channel side plate connection is much stiffer than the T-stub connection. The channel side plate connection requires all bolts to be fully pretensioned in order for the connection to be effective. Otherwise, slip in the connection will occur at an early stage and this would compromise the stiffness of the connection. Whilst the procedure to pretension standard structural bolts has been well-established in the Australian Steel Standard AS4100 using the part turn method [13], further work needs to be done to check whether the same procedure can be used on the ONESIDE blind bolt. Alternatively load indicating washers may be used to achieve a higher reliability in the applied pretension load in the bolts.

An alternative configuration is proposed in this paper, referred to herein as the extended T-stub connection with back face support. Testing of the proposed connection and subsequent finite element (FE) analysis have been carried out to determine the level of stiffness of this connection and to compare its performance with the previously tested T-stub connection [10] and alternative side connection [12]. The proposed connection is intended for use in low rise structures with small gravity loading (light-weight construction) and when the lateral loading is dominated by wind. Therefore the testing of the proposed connection was conducted under monotonic loading. There are other types of connections between hollow section columns and open section beams such as the widened flange column-tree connection [14] and the reduced flange plate moment connection [15]. Such connections are for heavier construction in high seismic regions and often involve extensive fabrication and welding work. For seismic design, the proposed connection can adopt certain features from other seismic connections such as the reduced flange plate moment connection. However this is beyond the scope of this paper. Further details of the connection are discussed in Section 2. Fig. 1 shows the various configurations of blind bolted connections to unfilled hollow section columns that have been investigated to date.

Section snippets

Test specimen

A full scale beam-to-column extended T-stub connection with back face support was tested under static loading. Details of the specimen are shown in Fig. 2. The specimen comprised a 310UB32 universal beam (approximately 310 mm deep, 32 kg/m section) connected to a 150×150×6 mm SHS column which are the same member sizes adopted in the previous test regime. The channels and T-stubs used in the connection can be made by cutting readily available commercial sections. The channel sections used in

General behaviour and failure load

The magnitude of the point load at the end of the cantilever beam was gradually increased until the test specimen failed. As the load was increased, the beam started to rotate in the direction of the applied load. Slip started to occur at an applied load of approximately 40 kN between the beam flange and T-stem as indicated by the onset of non-linearity at this load level in Fig. 4, Fig. 5, Fig. 6, Fig. 7. It is the stiffness of the connection up to the slip load that is of most interest in

Finite element analysis

A three-dimensional finite element (FE) model was created using the general purpose software ANSYS to represent the tested connection. The FE model takes into account material and geometric non-linearities and complex contact interactions between the various elements. Surface to surface contact elements with friction coefficient, μ, of 0.15 (which is the commonly adopted μ for painted surfaces) were employed in the FE model. μ was also validated by separate friction tests conducted earlier as

Design models

In this section, simplified design models for the extended T-stub connection with back face support to predict the strength and stiffness are discussed and summarised in Fig. 22.

Summary and conclusions

A new blind bolted connection to unfilled SHS column has been tested. This new connection is a modification of a typical T-stub connection, with additional channels connected to an extended endplate. The channels help to transfer loads to the back face of the SHS column in bearing; thereby reducing demand on the flexible column face. This results in a much stiffer connection in comparison to the typical T-stub connection. The findings from the test and FE modelling may be summarised as below:

  • (1)

Acknowledgements

This research is supported by the ARC, Linkage Project No. LP0669334. The authors would like to acknowledge the invaluable contribution of the industry partners Dr. Saman Fernando from Ajax Engineered Fasteners and Mr. Hayden Dagg from Australian Tube Mills. The authors also wish to thank Mr. Rodney Wilkie, Mr. Shane Clee, Mr. Justin Fox and Dr. Huang (Jack) Yao for their excellent assistance with the experimental testing.

References (20)

There are more references available in the full text version of this article.

Cited by (57)

View all citing articles on Scopus
View full text