| Author: | Chen, Wei |
| Title: | Fatigue performance of high strength S690 to S960 steel and their welded sections |
| Advisors: | Chung, Kwok-fai (CEE) |
| Degree: | Ph.D. |
| Year: | 2026 |
| Department: | Department of Civil and Environmental Engineering |
| Pages: | xvii, 245 pages : color illustrations |
| Language: | English |
| Abstract: | Motivation High strength steel (HSS) with yield strengths at 690 and 960 N/mm², i.e. S690 and S960 steel have been used in heavy machinery and lifting equipment for many years due to their large resistances per unit weight. Owing to their remarkable strength to self-weight ratios, structural engineers are eager to exploit their use in buildings and bridges as the cost as well as the time on fabrication, transportation and construction will be significantly reduced together with less energy consumption, when compared with those of normal strength S355 steel. Over the past 30 years, fatigue accounts for about 80% of failure of steel bridges in the world, and damages due to fatigue in these steel bridges have caused huge traffic delay, economic loss, and in extreme cases, problems in public safety. Owing to their martensitic microstructures, HSS possess superior mechanical properties, and their fatigue performance are also expected to be better than those of S355 steel. However, up to the presence, systematic experimental investigations into fatigue performance of HSS and their welded joints are rather limited. Due to a lack of engineering data, some design specifications allow the use of fatigue methods which have been specifically developed for S355 steel to be applied to S690 steel, leading to very conservative and inefficient design. In order to promote effective use of HSS in bridges, there is an urgent need to determine their fatigue behaviour under cyclic actions, and also to develop effective design rules to assess their fatigue performance in bridges. Objectives and scope of work The research project aims to quantify the fatigue performance of S690 to S960 steel and their welded sections through experimental and numerical investigations. Systematic fatigue tests on coupons of various types of high strength steel, and also their welded sections have been carried out to determine their fatigue resistances, in terms of the number of cycles completed before fracture. Hence, any reduction in the fatigue resistances of the S690 and the S960 steel due to welding is readily determined. Then, systematic investigations into both welded T-joints between S690 circular hollow sections and S690 U-stiffened steel decks were also performed to provide engineering data and numerical predictions on their fatigue resistances for a direct comparison with those of the coupons of the S690 welded sections. Based on these experimental and numerical data, applicability of various design recommendations to S690 and S960 steel and their welded sections was carefully established, and the use of stress concentration factors in typical welded T-joints was also verified. Methodology and key findings • Full range S-N curves of base plates In order to obtain the full range fatigue performance of S690-QT, S690-TM and S960-QT steel, a total of 39 fatigue tests have been carried out on cylindrical coupons extracted from their base plates, and these coupons are tested under symmetrical cyclic actions from ± 400 to 900 N/mm² over a range of 10⁴ to 107 cycles. It is found that All of these test data lie significantly above the design curve for fatigue assessment on the base plates given in EN 1993, especially for those design fatigue life larger than 106 cycles. The fatigue limits, i.e. the threshold value of the fatigue stresses that will not cause any damage, of all these steel are found to be more than 4 times the given values in EN 1993. • Full range S-N curves of welded sections In order to obtain the full range fatigue performance of butt-welded sections of the steel, another 39 fatigue tests have also been carried out on cylindrical coupons extracted from their welded sections, and these coupons are tested under symmetrical cyclic actions from ± 400 to 900 N/mm² over a range of 10⁴ to 107 cycles. It should be noted that during welding, a heat input energy of 1.0 kJ/mm was adopted. It is found that the fatigue limits of S690-QT, s690-TM and S960-QT welded sections are reduced to be merely 37.2%, 21.8% and 29.1% of those of the base plates respectively due to the effects of welding. Similarly to those of the base plates mentioned above, all the test data of the welded sections lie significantly above the design curve for fatigue assessment on welded sections given in EN 1993-1-9, especially when the design fatigue life is larger than 106 cycles. The fatigue limits of all the welded sections are also more than 4 times the given values in EN 1993. • Structural behaviour of welded T-joints In order to determine the structural behaviour of welded T-joints between S690 circular hollow sections under in-plane bending, a total of 10 monotonic tests were carried out on various configurations of S690 welded T-joints to determine local distributions of strains in the vicinity of the brace-chord junctions. An advanced measurement method of digital image correlation was adopted to determine local strain fields, and hence, hot spot stresses. Moreover, a total of 10 fatigue tests were carried out on the welded T-joints with one typical configuration to determine directly the fatigue behaviour of these welded T-joints under in-plane cyclic bending. It is found that all the experimental results lie well above the design curves given in EN 1993, CIDECT and API RP 2A for fatigue assessment of these welded T-joints. Hence, the experimental S-N curves with nominal stresses and hot spot stresses are found to be able to assess the fatigue behaviour of S690 welded T-joints with a high degree of accuracy. • Fatigue behaviour of U-rib stiffened steel decks In order to determine the fatigue behaviour of S690 U-rib stiffened steel decks under axial compression, a total of 14 fatigue tests were carried out on the S690 rib-to-deck welded joints with one typical configuration. It is found that all the experimental results lie well above the design curves given in IIW for fatigue assessment of these S690 U-rib stiffened steel decks. Hence, the experimental S-N curves with nominal stresses and hot spot stresses are found to be able to assess the fatigue behaviour of S690 U-rib stiffened steel decks with a high degree of accuracy. • Advanced numerical modelling In order to predict the fatigue behaviour of the welded T-joints between S690 circular hollow sections, an integrated numerical modelling approach is developed. Firstly, a thermal analysis is performed to determine thermal responses, and hence, temperature distributions induced during welding of the T-joints. Secondly, a thermomechanical analysis is then performed to determine mechanical responses, and hence, residual stress distributions induced by differential cooling in the vicinity of the welded collars of the T-joints. Thirdly, a stress analysis is then performed to determine structural responses, and hence, mechanical stress distributions throughout the T-joints under cyclic in-plane bending actions. Finally, a fatigue analysis is performed to determine the fatigue life of critical parts of the T-joints under various loading protocols. It is demonstrated that the fatigue behaviour of these welded T-joints, including the fatigue life and the locations of crack initiations, was successfully simulated through the proposed modelling approach with a satisfactory accuracy. • Verification on applicability of current design methods EN 1993 recommends a number of S-N curves to determine fatigue life for various constructional details under a wide range of cyclic stress ranges for up to S460 steel. For those welded sections of S690 to S960 steel, and also welded joints of S690 steel covered in the present investigation, it is found that EN 1993 often gives very conservative predictions, especially in the range of high cycle fatigue. Hot spot stress method adopted in both CIDECT, API RP 2A and IIW assesses the fatigue life according to the critical stress which is induced by geometric variations in structural members. Through a comparison between the experimental results obtained in the present project and the design results of the recommendations, it is found that these recommendations give very conservative results, especially for S690 welded joints in the range of high cycle range. Significance A series of experimental and numerical investigations into the fatigue behaviour of various high strength steel and their welded sections were performed. A comprehensive database on both experimental and numerical data on the fatigue behaviour of these steel and their welded sections is established. Based on these data, the fatigue performance of these steel, their welded sections as well as those welded joints covered in the present project was demonstrated to be superior to those of the normal strength steel. Enhancements to current design recommendations are also proposed. |
| Rights: | All rights reserved |
| Access: | open access |
Copyright Undertaking
As a bona fide Library user, I declare that:
- I will abide by the rules and legal ordinances governing copyright regarding the use of the Database.
- I will use the Database for the purpose of my research or private study only and not for circulation or further reproduction or any other purpose.
- I agree to indemnify and hold the University harmless from and against any loss, damage, cost, liability or expenses arising from copyright infringement or unauthorized usage.
By downloading any item(s) listed above, you acknowledge that you have read and understood the copyright undertaking as stated above, and agree to be bound by all of its terms.
Please use this identifier to cite or link to this item:
https://theses.lib.polyu.edu.hk/handle/200/14430