맞대기용접에서 횡구속이 용접잔류응력에 미치는 영향

Abstract

The welding process is widely used not only in the construction industry but also in various industrial fields such as ships, nuclear power, and aviation. However, welding techniques inevitably generate weld deformation and residual stress due to expansion, contraction, and metal phase transformation at high temperatures. In particular, welding residual stress affects fatigue strength, buckling strength and crack propagation path of brittle fracture. Its size and distribution are determined by heating input, thickness, welding process, restraint, and steel grade. These weld residual stresses are applied with secondary loads when evaluating the safety of steel structures, and a reasonable residual stress distribution is required for more accurate safety studies. In the method of measuring welding residual stress, based on the measurement position, we distinguish between surface measurement method and internal measurement method. The surface measurement method used was the hold ring method, the cutting method and the X-ray. The internal residual stress can be measured by the Sachs method, which is easy to measure the axially symmetric residual stress of a cylinder or hollow cylinder, and the residual stress can be calculated from the change in the diameter of the hole by drilling the position to be measured. There are DHD (Deep Hole Drilling Method) method and neutron diffraction method using neutron diffraction. In this study, the thermal elasto-plastic analysis was verified by using the neutron diffraction method. The prediction method of residual stress by the thermo-elasto-plastic analysis method can be easily calculated by improving computer numerical calculation ability and developing general-purpose numerical analysis code. However, with this method, the residual stress of a small test piece can be predicted with an economical calculation time. However, as the structure becomes larger, enormous calculation is required, and a method based on a simple analysis method has emerged. In these simple analysis methods, it is necessary to verify the reliability of the analysis result by an experimental method. The welding residual stress obtained by these thermo-elasto-plastic analysis and measurement is generated under the boundary condition in a free state where there is no constraint allowing only rigid body deformation. However, the actual structure is inevitably constrained by the manufacturing procedure, the constraint itself, and the welding procedure. When assessing the safety of a structure, not only the UK BS7910 [14], which is generally applied, but the yield strength of the material is used as the welding residual stress value if there are constraints in most codes. It is recommended to do so. There are many differences between the unrestrained weld residual stress distribution and the yield stress distribution. When evaluating the safety of large structures using high-strength steel, the influence of the weld residual stress used in secondary loads is significant. There is a limit to practical safety evaluation. Therefore, it is necessary to consider the influence on the welding residual stress according to the size of the actual constraint and present a more rational distribution of residual stress. Therefore, in this study, the effect on the welding residual stress is clarified by changing the thickness, yield stress and degree of constraint by thermal elastic-plastic analysis, and this is analyzed through regression analysis using R through the relationship between the restraining degree and the residual stress. For this reason, since the structure to be studied is thick, the three-dimensional thermal elastic-plastic analysis has a limit because it takes enormous analysis time. Therefore, two-dimensional thermal-elasto-plastic analysis was performed and the conditions of the welding heat source for more accurate temperature analysis were examined. In addition, the reliability was verified by comparing with the result of thermal elasto-plastic analysis that obtained the residual stress distribution obtained by regression analysis.