The buckling characteristics of cylindrical shells with a circular hole, under axially compressed loads, have been analyzed and the results have been compared with existed experimental results. Deflection function with decay factor is assumed, and stress distribution around a circular hole in tensioned infinite plate is used for formulating buckling energy function. Applying Rayleigh Ritz procedure to this energy function, characteristic equation of eigenvalue problem is determined. Buckling load is defined by the minimum value of eigenvalues calculated according to several decay factors, and as the radius ratios of a circular hole (a/R) and shell thickness ratios (R/t) are varied, the reducing characteristics of buckling load are studied. As a result, buckling loads are reduced by about 50% according to some radius ratios ($a/R{\geq}0.15$) of circular hole and are not nearly affected by shell thickness ratio(R/t).

Generally the stiffened plate in the ship structure is subjected to not only axial load but shear load. With respect to those combined loads buckling analysis in necessary. In this paper, buckling strength is analyzed by using Finite Element Method when the stiffened plate with hole is under loading conditions mentioned above. To obtain the higher buckling strength, we need some reinforcement. The methods of reinforcement are attaching doubler around hole and stiffeners in the arbitrary directions For the sake of convenience those arbitrary directions were selected paralleled($0^{\circ}C$), vertical($90^{\circ}C$)and oblique($45^{\circ}C$) to the edge. Two kinds of method mentioned above are investigated, it is clarified that which of the two is more effective reinforcement. From the viewpoint of buckling strength, following conclusions were obtained. When external load direction is unknown, doubler reinforcement is more effective than those of parallel and vertical stiffener. And oblique stiffener reinforcement is more effective than that of doubler when external load direction is know.

This paper deals whth the buckling as well as postbuckling analysis of axisymmertric shells taking the initial deflection effects into account. Incremental equilibrium equations, based on the principle of virtual work, were derived by the finite element method, the successive step-by-step Newton-Raphson iterative technique was adopted. To define the transition pattern of postbuckling behavior from the prebuckling state more accurately, a simple solution method was developed, i.e. the critical load was calculated by the load extrapolation method with the determinant of tangent stiffness matrix and the equilibrium configuration in the immediate postbuckling stage was obtained by perturbation scheme and eigenvalue analysis. Degenerated isoparametric shell elements were used to analyse the axisymmetric shell of revolution. And by the method developed in this paper, the computer program applicable to the nonlinear analysis of both thin and moderately thick shells was constructed. To verify the capabilities and accuracies of the present solution method, the computed results were compared with the results of analytical solutions. These results coincided fairly well in both the small deflection and large deflection ranges. Various numerical analyses were done to show the effect of initial deflection and shape of shells on buckling load and postbuckling behavior. Futhermore, corrected directions of applied loads at every increment steps were used to determine the actual effects of large deflection in non-conservative load systems such as hydrostatic pressure load. The following conclusions can be obtained. (1) The method described in this paper was found to be both economic and effective in calculating buckling load and postbuckling behavior of shell structure. (2) Buckling and postbuckling behavior of spherical caps is critically dependent upon their geometric configuration, i.e. the shape of spherical cap and quantities of the initial deflection. (3) In the analysis of large deflection problems of shells by the incremental method, corrections of the applied load directions are needed at every incremental step to compensate the follower force effects.

In ship's structure, deck and bottom plate are main strength member subjected to the inplane load due to longitudinal bending, i.e. tensile and/or compressive load. The deck and bottom plate are subdivided into many plate members by stiffeners and girders longitudinally and transversely. Since the plate members are thin, it is likely to be collapsed under compressive load, and when we consider the local strength of deck and bottom, the plate members play an important role in the longitudinal strength. Therefore the precise analysis of their compressive ultimate strength is required for the optimal design of ship's structures. In this paper, the modified analytical method using the incremental form of principle of virtual displacement is introduced to determine the compressive ultimate load of plate members. The results by the present method is satisfactory, and the present method is more effective and economical than the finite element method.

A computer program was developed for the analysis of the coupled transverse vibrations of a multi-supported shaft system. The program, based on the theory of Transfer Matrix Method, was written including the system parameters such as the entrained water and gyroscopic effects of the propeller, the rigidity of bearing combined with the oil film effect, and the whirling frequency of the shaft. The program was used to calculate the resonance frequency of the shafting system of the ship Hanbada. The results show good agreement with the measured values. The results are also compared with those of F.E.M. from the comparison, it is found that both results agree well with each other.

For the analysis of vertical vibrations of a ship's hull, the Timoshenko beam analogy is accepted up to seven or eight-node modes provided that the system parameters are properly calculated. As to the shear coefficient, it has been a common practice to apply the strain energy method or the projected area method. The theoretical objection to the former is that it ignores lateral contraction due to Poisson's ratio, and the latter is of extreme simplifications. Recently, Cowper's and Stephen's shear coefficient formulas have drawn ship vibration analysts' attentions because these formulas, derivation of which are based on an integrations of the equations of three-dimensional elasticity, take Poisson's ratio into account. Providing computer programs for calculation of the shear coefficient of ship sections modeled as thin-walked multicell sections by each of the forementioned methods, the authors calculated natural vibration characteristics of a bulk carrier and of a container ship by the transfer matrix method using shear coefficients obtained by each of the methods, and discussed the results in comparision. The major conclusions resulted from this investigation are as follows: (1) The shear coefficients taking account of the effects of Poisson's ratio, Cowper's $K_c$ and Stephen's $K_s$, result in higher values of about 10% in maximum as compared with the shear coefficient $K_o$ based on the conventional strain energy methods; (a) $K_c/K_o{\cong}1.05\;and\;K_s/K_o{\cong}1.10$ for ships having single skin side-shell such as a bulk carrier. (b) $K_c/K_o{\cong}1.02\;and\;K_s/K_o{\cong}1.05$ for ships having longitudinally through bulkheads and/or double side-shells in the portion of the cargo hod such as a container carrier. (2) The distributions of the effective shear area along the ship's hull based on each of $K_o,\;K_c\;and\;K_s$ are similar each another except the both end portions. (3) Natural frequencies and mode shapes of the hull based on each of $K_c\;and\;K_s$ are of small differences as compared each other. (4) In cases of using $K_c\;or\;K_s$ in ship vibration analysis, it is also desirable to have the bending rigidity be corrected according to the effective breadth concept. And then, natural frequencies and mode shapes calculated with the bending rigidity corrected in the above and with each of $K_o,\;K_c\;and\;K_s$ result in small differences as compared each another. (5) Referring to those mentioned in the above (3) and (4) and to the full-scale experimental results reported by Asmussen et al.[17], and considering laboursome to prepare the computer input data, the following suggestions can safely be made; (a) Use of $K_o$ in ship vibration analysis is appropriate in practical senses. (b) Use of $K_c$ is appropriate even for detailed vibration analysis of a ship's hull. (6) The effective shear area based on the projected area method is acceptable for the two-node mode.