Cracks caused by cold bending during shipyard construction were found to be longer in mild steel than in high tensile steel. Because mild steel has better elongation, the cause of such cracking was difficult to ascertain. Bending tests of large number of steel pieces to confirm the effects of multiple parameters, nonlinear finite-element analysis to check strain distribution, and fractography to determine the nature of the fracture indicated that ductile fracture initiated where the strain value was highest, as is generally expected. In the bending tests, the fractures were reproducibly longer in alloys with better ductility. One of the reasons for this phenomenon was that the surface was more hardened by plasma-arc cutting in mild steel because of its high carbon content according to hardness tests on the cutting surfaces. We therefore proposed a new equation to estimate crack length, taking into account three factors: strain, elongation and surface hardness. We showed that the relationship between crack length and the values by the equation explains the results of the bending tests. Additionally, we suggested the possibility that inclusions within an alloy affect the crack length. Cold forming is a general procedure used in shipyards. If inspections fail to find a crack, it may propagate to fatal damage after delivery. This study is therefore useful in preventing the initiation of cracks.

In the ship-hull design phase, the distance of a misalignment of a cruciform welded joint is enlarged, by up to 100 mm for example, until its effect is thought to be minimized. To control tolerance, we used a technique to estimate the coefficient of the misalignment effect: the ratio of the stress with misalignment to stress without misalignment. Current methods, created using the two-dimensional (2D) misalignment model and focusing on small misalignments, cannot be applied in the design phase. Misalignment models are needed, as there have been no studies of these joints and it is impractical to do finite element (FE) analysis in each case. We propose an estimation method of the coefficient of misalignment effect in three dimensions (3D). First, we created a new equation to estimate the coefficient in 2D. The equation accommodates larger misalignments. Second, we found the membrane component of a misaligned member stress is reduced as a 3D effect and the coefficient converges given a certain misalignment value. Third, a modified equation taking the effect into account indicates the coefficient in 3D.

In ship hull design, longitudinal stiffeners may be twisted or knuckled to keep the continuity of strength along the vessel. This occurs at angled discontinuities such as discontinuous connections at the transverse web between two longitudinal stiffeners having different angles, e.g., a side shell longitudinal stiffener in the rounded part of the ship’s hull. Although the criteria for selection of “twist”, “knuckle”, or “angled discontinuity” as it is should be strength criteria, there are no generalized criteria or published systematic studies, and ship hull designers usually depend on experimental rules. In this study we evaluate and compare the stress magnification factors and define the relative ratio between the stresses in the discontinuity and continuity structures by using finite element analysis from a strength-of-materials point of view. Our results show that under lateral load conditions the stress factor is larger in the twist structure than in the angled discontinuity structure, although under axial load conditions the stress factor in the knuckle and angle discontinuity structures is larger than in the twist structure. Additionally, we propose a formula that estimates the factor in the knuckle structure based on the strength-of-materials approach.

SummaryIn the ship-hull design phase, the distance of longitudinal discontinuity is enlarged, by up to 100 mm for example, until its effect is thought to be minimized. To control tolerance, we used a technique to estimate the coefficient of the misalignment effect: the ratio of the stress with misalignment to the stress without misalignment. Current methods, created using Strength-of-Materials misalignment model, cannot be applied in the design phase. Longitudinal discontinuity models are needed, as there have been no studies of these joints and it is impractical to do finite element (FE) analysis in each case. We propose an estimation method of the coefficient of discontinuity effect in cruciform welded joint with shell plate. First, we created a new equation to estimate the coefficient in Strength-of-Materials model. Second, we found the membrane component of a discontinuous member stress is reduced because of the shell plate and the coefficient converges given a certain discontinuity value. Third, a modified equation taking the effect into account indicates the coefficient in the joint with shell plate.X, Y, Z X Y Z ( Fig.2 ) L