Estimations Of Stress Concentration Factor For Fatigue Design Of Welded Tubular Connections

Abstract

ABSTRACT The S-N approach with hot-spot stress range concept is a widely used method for the fatigue design of tubular connections in fixed offshore structures. In order to use this method, realistic values of stress concentration factors (SCFs) in the tubular connections are required. A number of alternative parametric equations are available for the calculations of SCFs. However, there is still a paucity of test data on realistically constructed and sized specimens for some basic joint types to validate these equations and to select the most suitable equations for use in design. This paper presents results of a total of 45 elastic tests carried out on 15 realistically constructed and sized tubular joint specimens of either T, Y, X or K configuration. The tests concentrate on joints with high ? (brace to chord diameter) ratios. The SCFs of these specimens under either axial, in-plane bending and out-of-plane bending loads were estimated using strain gauges readings. Three different extrapolation techniques have been used to interpret the measured results. An assessment on the reliability of SCF parametric equations for fatigue design has been carried out in the light of the new test data. INTRODUCTION A jacket type offshore structure consists mainly of tubular connections with brace and chord members. Most of these members are made of steel tubulars with circular hollow section which are joined together by full-penetration welds. One important aspect in design and analysis of the jacket structures is to check for the fatigue damage of these tubular joints under environmental loads. The S-N approach with 'hot-spot' stress range concept is a widely used method for the fatigue assessment. To use this approach, realistic values of stress concentration factors (SCFs) for the joints are required. A number of alternative parametric equations are available for the calculations of SCFs. However, there is still a paucity of data on realistically constructed and sized specimens for some basic joint types to validate these equations especially for joints with high ? (brace to chord diameter) ratios. As part of a joint industry research project, 45 elastic tests were carried out on 15 large scale steel specimens to measure the SCFs for joints with high ? ratio. The SCFs results are presented here and compared with the predictions of the available parametric equations. Adopting the approach of the previous papers(1,2), this paper addresses the reliability of the existing parametric equations in predicting SCFs in the light of the new test data. Implications of the use of SCF equations in design are also commented. Details of the tests described in this paper can be found in a HMSO report(3). TEST PROGRAMME Test specimens The test programme comprises elastic tests on nine T/Y-joint, three X-joint and three K-joint specimens with chord members of 508mm (20in) diameter. For all T/Y and X joints, except for one specimen, the ? ratios are ranging between 0.8 and 1.0. A K joint where ? = 1 is also included in the programme. Details of the specimens and their dimensional parameters are given in Table 1. Three loading modes, namely, axial, in-plane bending and out-of-plane bending, were considered.