| doi:10.3850/978-981-08-6218-3_SUS-Th028 |
 Final Paper PDF
|
INVESTIGATION AND EXTENSION OF FATIGUE LIFE OF STRINGER-CROSS GIRDER COPE CONNECTIONS
S. Browna, J. Hwangb and A. Parkerc
Structural Dynamics, Heggies Pty. Ltd., Lane Cove, NSW.
asteve.brown@heggies.com
bjoon-pil.hwang@heggies.com
candrew.parker@heggies.com
EXTENDED ABSTRACT
Many steel girder bridges suffer from fatigue problems at its stringer-cross girder joints, in particular with respect to cracks initiating from its cope detail. Heggies carried out an investigation of fatigue life at a typical detail on Kempsey Bridge, with modal/strain measurements and numerical simulations. Based on these results the reasons behind typical fatigue issues were able to be better understood, and a practical, cost-effective solution was able to be developed and deployed on many bridges that suffer from similar problems.
1. INTRODUCTION
Steel girder, composite or non-composite concrete deck bridges are one of the most common types of bridge structures in Australia, and many of these structures were constructed approximately 50 to 70 years ago. In recent times many of these bridges were found to suffer from failures at the cope and more recently these have been identified as fatigue failures.
Failures were first observed in riveted cope connections of Kempsey Bridge in 2004 when a number of coped stringer-to-cross girder connections were found to have lost the top two rows of rivets and/or were found to have cracks in the stringer web or stringer end flange welds. The lost rivets were replaced with high strength bolts and following inspections revealed that the replacement bolts were failing in as little as 6-months. Some failed bolts were recovered and submitted for metallurgical examination. This examination revealed characteristic evidence of fatigue failure. Review of available published literature revealed that fatigue failures of stringer/cross girder joint details were commonly found in these types of structures.
2. FINITE ELEMENT (FE) MODEL
A finite element (FE) model of the truss span was constructed using the commercial FE software package, MSC.Patran/MSC.Nastran (MSC.FEA), based on the original RTA drawings of the bridge.
3. ON-SITE MEASUREMENTS
3.1. Modal Measurements
Heggies carried out measurements of modal characteristics, static strains and dynamic stresses on site. By simultaneously measuring a reference vibration response and ‘roving’ vibration responses at a large number of responses over the span, the relevant mode shapes of this span were calculated. The results of the FE model closely matched the measured vibration characteristics of the span.
3.2. Strain Measurements
Strain measurements were also carried out on site by Heggies to obtain static and dynamic strains at various points throughout the stringer-cross girder connections. The following conclusions were made on the basis of the strain measurement results:
- It is clear that at the mid span stringer location the neutral axis is near the geometric centre of the beam. This indicates that the concrete deck is providing little composite effect, and they effectively act as separate beams.
- At the cope connection the neutral axis is almost at the bottom flange of the stringer. This indicates a complete lack of composite effect as well as a limited ability for the joint to transfer moments. The cope cut-out reduces the second moment of area by nearly a factor of 4 (3.6), which increased the corresponding stress by a similar amount.
4. FATIGUE ANALYSIS
Heggies also conducted a fatigue study based on the above obtained data, using Miner’s Rule for cyclical accumulation and the S-N curve as provided in AS5100.4-2004. The results of this study indicated that joint details near the cope cut-outs of the stringer-cross girder connections had very little fatigue life remaining, if not used up already.
Heggies was able to incorporate a modification into the FE model aimed at extending the fatigue life of the cope details, involving replacement of the top one to three rows of bolts with fastener sets with vastly decreased axial stiffness, which therefore reduces the out of plane rotational stiffness of the cope connection.
The introduction of compliant fasteners has a clear advantage of much lower cost than the stiffening options, and could be achieved with access from under the bridge, thereby negating the need for costly and disruptive lane/road closures. Furthermore, the system does not significantly affect the shear capacity of the joint which reduces the modifications that are required to be made in comparison to outright removal of the bolts. In further modelling, the introduction of these compliant fasteners to joints with cracks on its cope which have not progressed past the first row of bolts will probably not propagate once the compliant connection modification is fitted and the connection should exhibit a significant increase in fatigue life. Heggies was able to introduce the compliant joint system described above on many bridges in conjunction with the RTA, with success, with the bridges not reporting any recurrence of fatigue problems as of current date.