|Title:||Structural effects of welding onto high strength S690 steel plates and welded sections|
|Advisors:||Chung, K. F. (CEE)|
|Subject:||Hong Kong Polytechnic University -- Dissertations|
Steel, High strength
|Department:||Department of Civil and Environmental Engineering|
|Pages:||xx, 444 pages in various paging : color illustrations|
|Abstract:||Motivation-High performance S690-QTsteels (HPS) are modern steel productswhich possess a yield strength at 690 N/mm², i.e. two to three times those of commonly used structural steel materials, namely S235 and S345 steels. Owing to their excellent strength-to-self-weight ratios, HPS often offer effective structural solutions to heavily loaded structures. In the past decade, HPS have been successfully used in large lifting equipment and machinery as well as structures on offshore platforms. In general, many researchers in structural engineering believe that wide adoption of HPS in construction will have a huge impact on the construction industry worldwide because of their greatly improved structural efficiency, and a significant reduction in self-weights. However, HPS have not been widely adopted in construction owing to a lack of understanding on effects of welding onto mechanical properties of HPS, in particular their ductility in welded sections. Moreover, it is essential to provide suitable residual stress patterns for welded HPS H-sections in order to achieve safe and efficient structural design. Objectives and scope of work-In this project, a systematic experimental and numerical investigation into structural behaviour of HPS has been conducted in order to examine any adverse effect onto structural behaviour of HPS steel plates and welded sections due to welding. The scope of work includes: - Task1 Mechanical properties Stress-strain curves of coupons of HPS steel plates and their welded sections under monotonic and cyclic tests. These tests provide basic "stress and strain" curves of S690 steel plates and welded sections to demonstrate their mechanical properties. - Task2 Temperatures of welded H-sections during welding Experimental and numerical investigations into thermal responses of welded sections, and subsequent formation of residual stresses in welded sections. - Task3 Residual stress patterns of weldedH-sections Confirmation of proposed residual stress patterns in welded H-sections with various plate thicknesses. - Task4 Effects of residual stresses onto welded H-sections Identification of residual stress patterns, and their effects onto both cross-section and member resistances of welded sections. The areas of interest include: - ductility of high strength S690-QT steel plates and their welded sections under monotonic and cyclic actions; - temperature distribution histories of welded sections during welding, and reduction in both strength and ductility of steel plates in the vicinity of the heat affected zones; - variations of temperatures, residual strains and stresses across plate thicknesses in the vicinity of the heat affected zones; - coupled thermal-mechanical finite element modelling; - reduction in cross-section and member resistances due to the presence of residual stresses in welded H-sections. Four different cross-sections of various dimensions and plate thicknesses, namely, Sections C1, C2, C3 and C4 were fabricated with 6, 10 and 16 mm with: - temperature measurements on outer surfaces of the section flanges during welding; - residual strain measurements on outer surfaces of the section flanges; - compression tests on both stocky and slender columns. Research Methodology and Key Findings-Based on a series of coordinated experimental and numerical investigations, all the four tasks were successfully completed, and key findings are presented as follows: - Task 1 Mechanical properties A total of 40 standard tensile tests were conducted under monotonic loadings. Small deformations up to yielding were measured with strain gauges while large deformations with extensive yielding and necking were measured with a digital imaging method. It was found that among all the tested coupons of the steel plates and the welded sections, they were able to satisfy the following ductility requirements stipulated in EN 1993-1-1 and -12: i)fu / fy = 1.05, ii) eu = 10%, and iii)eu = 15 fy / E where fu is the tensile strength, fy is the yield strength, E is the Young's modulus, eu is the elongation limit at fracture. In order to examine hysteretic ductility of the steel plates and their welded sections, a total of 40 low-cycle-high-strain cyclic tests on non-standard funnel coupons were carried out. The targeted maximumstrains were specifiedto be ±2.5%, ±5.0%, ±7.5% and ±10%, and the loadings were applied at four different frequencies, namely, 0.1, 0.5, 1.0 and 2.0 Hz. All the coupons of S690 steel plates and welded sections were demonstrated to be able to exhibit a high degree of ductility under the proposed loading protocols. It should be noted that: i) the coupons of the steel plates were able to complete all 20 loading cycles, and they have attained a maximum strain of ±10%, and ii) the coupons of the welded sections were able to complete atotal of 19 loading cycles, and they have also attained a maximum strain of ±10%. Consequently, these S690 steel plates are demonstrated to satisfy specific ductility requirements stipulated in EN 1993-1-1, and they have good hysteretic ductility under cyclic loads up to ±10%. - Task 2 Temperatures in welded H-sections during welding-In order to collect temperature data, a total of 4 welded H-sections with different cross-section dimensions and plate thicknesses, namely, Sections C1, C2, C3 and C4, were instrumented with thermocouples so that temperature variation histories of the steel plates during welding were acquired for subsequent calibration of thermal finite element models. Advanced thermal finite element models were established with ABAQUS to simulate various welding processes of welded H-sections. A heat source model of a double ellipsoidal heat source model (Goldak, 1984) was adopted. It was found that the predicted temperature distribution histories in these welded sections compared very well with corresponding measured temperatures obtained in Task 2, when various heat transfer parameters and physical thermal properties of the steel plates and the welding materials were selected properly. Moreover, the measured and the predicted temperatures of the heated affected zones of the welded sections of both fillet and butt welds with i) single runs, and ii) multi-runs of welding agreed very well with each other. Moreover, images of Scanning Electron Microscope (SEM) were captured to examine variations in the micro-structures of the heat-affected zones. In general, highly localized temperatures were found in the surfaces of the steel plates in the heat affected zones under direct contact with the welding materials. However, owing to the nature of transient heat transfer during welding, there was always a large temperature gradient across plate thicknesses, and hence, the temperatures of the undersides of those steel plates were rather low. This was considered to have a significant effect on the distributions of residual strains and stresses across the plate thicknesses in the welded sections.|
Task 3 Residual stresses of welded H-sections In order to obtain measured residual strains, Sections C1 to C4 were instrumented extensively with circular rosettes along their outer surfaces of section flanges so that the residual strains were measured with the use of a hole drilling method according to ASTM E837. Moreover, coupled thermo-mechanical analyses on these four sections carried out according to the corresponding measured temperature distribution histories obtained in Task 2. It was shown that the predicted residual stresses of the outer surfaces of various section flanges compared very well with the measured strains obtained in Task 2 for these four sections. It should be noted that the predicted residual stresses in the other sides of the steel plates in the vicinity of the heat affected zones were found to be very low. Summarization of these residual stresses across plate thicknesses provided an average residual stress pattern of the four sections, and force equilibrium was successfully verified in models of these four sections through adding up tension and compression forces developed in various parts of the flanges, the webs, and more importantly, the flange-web junctions. Owing to the increased yield strengths of S690 steel materials, a significant area in the vicinity of the flange-web junctions was found to remain to be elastic under large residual stresses. Hence, the residual stresses in S690 welded H-sections were found to be proportionally less pronounced when compared with those in S355 welded H-sections of similar dimensions. In order to propose a residual stress pattern applicable to welded H-sections of various dimensions and thicknesses, a comprehensive parametric study was carried out in which the plate thickness ranged from 10 to 50 mm. For easy comparison, codified thermal properties and bi-linear stress-strain curves of the steel plates were adopted. Moreover, both single pass and multi pass weldingprocedures were included in the parametric study. It should be noted that, when the plate thickness increased, variations of temperatures as well as residual strains and stresses became more apparent, and averaged values of the residual stresses at the outer surfaces and those of the inner surfaces decreased accordingly. Moreover, there were remarkable reductions in the areas of yielded regions in the flange-web junctions in S690 welded sections, when compared with those in the flange-web junctions in S355 welded sections. - Task 4 Effects of residual stresses onto welded H-sections In order to verify applicability of the proposed residual stress patterns for S690 welded sections, a total of 4 stocky columns and 4 slender columns were concentrically loaded to determine their deformation characteristics under compression. Moreover,advanced numerical models with i) zero residual stress, ii) measured residual stresses, and iii) predicted residual stresses were established. By careful comparison with measured load-deformation curves of these columns, it was found that the numerical models with the proposed residual stress patterns were able to predict deformation characteristics of these columns not only in the initial deformation ranges, but also in the large deformation ranges. Key findings and their significance-The major academic merits of the research work are: - Confirmation of satisfactory structural performance of S690 steel plates and welded sections under monotonic and cyclic actions, and a maximumtargeted strain at ±10% was readily achieved. - Investigation of the weldability of S690-QTsteel materials is varied out, and Only welded connections with a heat energy of 1.0 kJ/mm perform equal strengths with parent plates, whilst applicable reduction factors of yield strength are found to be 0.90 and 0.86 respectively for welded connections with heat energies of 1.5 and 2.0 kJ/mm. - Establishment of advanced thermal modelling and the use of appropriate welding parameters in simulating temperature distribution histories in welded sections after i) single-pass welding, or ii) multi-pass welding. Significant temperature gradients were found within the plate thicknesses in the vicinity of the heat affected zones. - Residual stress patterns within welded H-sections were established through coupled thermal-mechanical analysis, and generic residual stress patterns were proposed for welded H-sections of different cross-section dimensions and plate thicknesses. - Multi-passwelded sections would experience a significantly small magnitude of residual stresses, when compared with those of single-pass welded sections. - Through experimental and numerical compression tests on both stocky and slender columns of Sections C1 to C4, the effect of residual stresses onto load-deformation characteristics of the columns have been properly assessed. Incorporation of the proposed residual stress patterns into finite element models enabled a good prediction to the experimental results. Hence, validity of the proposed residual stress patterns was established.
|Rights:||All rights reserved|
Files in This Item:
|991021959945803411.pdf||For All Users||18.71 MB||Adobe PDF||View/Open|
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: