A new numerical solution method for nonlinear shallow water wave problems is developed. It is a version of the TUMMAC method that is based on the MAC method, and employs the donor-cell method in the finite-difference representation of the convective terms, the simultaneous iterative method in the solution of the Poisson equation for the pressure and the arbitrary body boundary condition. This method is suited for the computation of nonlinear characteristics of a solitary wave and a sinusoidal wave in shallow, water with an arbitrary bed topography or an arbitrary floating body. Some computed results of a solitary wave approaching a beach, a.vertical wall and a curved wall are presented.

View full abstract

Some computed results by the TUMMAC-I method for bow waves of steadily advancing wedge models are presented, and the difference due to the differencing scheme of the convective terms of momentum equations is examined using centered and donor-cell differencings. The measurement of pressure distribution on the wedge surface is conducted and the results are compared with the computation. The effect of cell size is also studied in order to know the available degree of accuracy of the TUMMAC-I method.

View full abstract

In the previous report, design problems of the minimum wave resistance hull form with the fairing constraint were carried out by means of the nonlinear programming, and hull forms having normal end shapes without swan necks were obtained by means of the optimization procedure. Those normal end shapes had not been expressed exactly in the usual design procedures based on variational approaches.
In the present paper, some examples of optimization problem for bow and stern shapes of normal type, bluff type and bulb type are introduced under constraints such as constant displacement, fairing and so on. Results for fore and aft asymmetric hull forms are also included.

View full abstract

This paper presents results of mean velocity and turbulence measurements in the three dimensional boundary layers and wakes around a long ship model in a low-speed wind tunnel. The long ship model was obtained by inserting the 4 m length parallel body into the 2 m length normal ship model which had been tested in the 1st report. The effects of the boundary layer thickness on the stern flow field became evident after comparing the results of measurements in the boundary layers of the long ship model with those of the normal ship model.
Moreover, a new analyzing method was developed to measure the mean velocity and the turbulence in three dimensional flow with a single-type hot-wire anemometer.
Comparing the stern flow field of the long ship model with that of the normal ship model, following facts were found.
(1) The longitudinal vortices are located at the same place with the same vorticity.
(2) The ridge-like parts are located at the same position away from the ship model surface on the contour curves of the turbulent kinetic energy.
(3) The non-dimensional normal distance (ζ/θ_ξ) (L/B) gives very good agreements on the profiles of the mean velocity and the Reynolds stresses distributions in the boundary layers between the long and the normal models.

View full abstract

This paper deals with the roughness effects (ΔC_F) on skin friction from a hydrodynamical point of view.
The former part of the paper consisted mainly of the experimental reports, that is, the devices of pipe flow experiments, the friction measurements in the roughened pipes and the analyses of their data.
The main results obtained are as follows;
(a) The measured frictional resistances of variously roughened pipes are shown in Fig. 5. In the figure, the resistance curves of painted pipes show clearly different tendencies from those of sand roughened pipes as mentioned by Prof. Sasajima in 1954.
(b) The hydrodynamical relation between the roughness height, measured by BSRA roughness analyser, and the so called equivalent sand roughness height, determined by comparing Nikuradse's experimental results, is shown in Fig. 6. The relation can be expressed also by
K_s= 1.47 K_BSRA
The Latter part of the paper consisted mainly of the theoretical considerations of C_F and their applications.
The main results obtained are as follows;
(c) Using a logarithmic type of velocity distribution in the boundary layer, the local ΔC_f can be expressed simply by local ΔC_f0 and roughness function ΔU/u* as
ΔC_f∝C_f0^3/2 (ΔU/u*).
(d) The procedures to determine roughness function ΔU/u* from the measured resistance data are explained. It can be done without measuring velocity distribution in the boundary layer.
(e) A new roughness parameter h≡kF_n^2/3 (g/ν²) ^1/3 is proposed, concerning the similarity law for ΔC_F. As the parameter can be applied commonly to both the model and the ship, the new parameter may be a powerful tool for estimating ΔC_F.
(f) As an application of the above stated considerations, the frictional resistances of a tanker are calculated in the various roughness conditions. The results are shown in Fig. 12.

View full abstract

The theoretical calculation of hydrodynamic characteristics of a marine propeller has been remarkably developed by the aid of high speed computers with large memory capacity. It is already useful for practical applications to the design problems on the steady characteristics. But the unsteady characteristics have not necessarily been solved enough. The authors have analyzed some fundamental unsteady problems on the fluctuating thrust and torque of a propeller in the typical harmonic wakes, the pattern modes of which are given by the function cos nθ, where n one to eight, employing the mathematical model of the harmonic-oscillating vortex lattices lifting-surface. The numerical results of this calculation present some interesting facts, such as existence of amplification effect of thrust and torque fluctuations in the certain modes of wake pattern cos 2θ, cos 3θ and cos 4θ, and their progressive phase due to striking influence of the time derivative term of circulations of vortices. On the contrary, the outflow of shed vortices from the trailing edge has effect to reduce the amplitude and to give a lag of the phase. It is pointed out that the non-linear effect of a non-uniform wake field brings about the higher efficiency of propeller than in open water and another fluctuations of thrust and torque of double blade frequency.
The results are compared with other theoretical calculations of Koyama's method and a quasi-steady way.
The performance tests of three model peopellers of 3, 4, and 5 blades were carried out behind the mesh-screens to generate harmonic wakes. The data of experiments were analyzed to be compared with calculations. The agreement of them is to be satisfied when errors in experiments and analyses are considered.

View full abstract

The numerical solutions of the existing lifting surface theories do not converge at the parabolic wing tip on account of the singularity, which is an obstacle to practical application to marine propellers. The method presented in the first and second reports overcomes the difficulty, and supplies the accurate solution even near the tip of propeller blade.
In this third report, the measurement of the pressure on the propeller blade near the tip is performed. Experimental investigation reveals the following. (1) Characteristics of pressure on propeller blade near the tip are shown by the pressure measurement at γ/γ₀=0.900, 0.962 of a model propeller of the geometry magnifying the tip singularity. (2) Numerical results for lift density on blade obtained by the present method and the existing method are compared with experimental results, which makes clear the fact that the present method removes the defect in the existing methods. (3) The comparison of the pressure distribution on blade between the experimental results and the theoretical results by the combined method of present lifting surface theory and two-dimensional aerofoil theory reveals to what extent the theory can explain quantitatively the actual phenomenon near the tip of blade. (4) It is shown that the three-dimensional correction of thickness effect on pressure should be taken into account in order to improve further the theoretical method. A simple method based on the calculation of ellipsoid is considered to be convenient for the correction. (5) Under heavy load condition, the flow separation occurs at the tip of blade, which causes very low pressure on back side of blade tip. The phenomenon is not dealt with by the present method. In this respect further works are considered to be necessary in future.

View full abstract

On the basis of results of theoretical calculations and model tests on Highly Skewed Propellers (HSP), solid type 46° HSP was designed for a car carrier 'Chijin-Maru' of 3, 200 carrying capacity according to the following primary plan;
1) Skew angle which decreases the vibratory propeller forces within the tolerable limit values is applied.
2) Blade strength of the HSP is higher than that of CP both in ahead operation and in crash astern operation.
3) Propeller efficiency and cavitation characteristics of the HSP are the same as those of CP or better.
Especially, careful attention was paid for the blade strength in such a way that relation between HSP geometry and blade stress is fully examined by using propeller lifting surface theory and finite elements method. The 46° HSP design procedure is described in full detail in this paper.
Model tests were performed to compare the performance characteristics of the 46° HSP with those of CP. Results of the tests demonstrated that the 46° HSP decreased propelle-rinduced fluctuating pressure by 37 percent on the hull surface and decreased sheet cavitation on the blade back without decreasing propeller efficiency.
Then, the 46° HSP was applied to the car carrier and full scale measurements were carried out.
The car carrier cleared design speed, required number of propeller revolution, estimated crash astern performances etc. at the official sea trial. Further, results of full scale measurements proved that propeller-induced fluctuating pressure measured on the hull surface was close to the value estimated by theoretical calculation or the model test and local vibration on the hull was in low level sufficiently in comparing with the other car carriers. Also, from results of full scale measurements it was confirmed that blade stresses measured in ahead operation and in crash astern operation were close to the estimated values and no problem was involved on strength of the 46° HSP.
Adequately designed HSP can deduce vibratory propeller forces without lowering the other propeller performance characteristics and it should be considered to apply HSP to any ships with vibration problems involved.

View full abstract

Lifting-line approximation and calculations are performed to obtain the characteristics of partially cavitating propellers. Main treatments are summarized as follows.
(1) Perturbed flow field by the partially cavitating propeller is described by two kinds of singularity-vortex and source distributions. And then a triple integral equation for determining the singularity distributions is derived from pressure constancy condition on cavity and tangential flow condition on foil surface.
(2) By application of lifting-line approximation to the triple integral equation, it is denoted that the two dimensionality holds for any blade element of propeller with respect to effective angle of attack corrected by induced one.
(3) In order to improve the accuracy of the lifting-line theory of the partially cavitating propeller, curved flow effect, which is of importance in wide blade propellers, is obtained in the form of correction factors to induced angle of attack on the basis of the existing experiments at noncavitating condition.
(4) Two partial cavity theories (thin and thick hydrofoil theory) formerly presented are used for obtaining the two dimensional characteristics of any blade element and optimum model at cavity termination formerly proposed is adopted in these theories. The integral equation of lifting-line is solved by application of these two dimensional characteristic data.
(5) Numerical calculations are performed for three kinds of MAU propeller-
(i) No. 0197 (number of blade 3, pitch ratio 0.4, developed area ratio 0.5)
(ii) No. 0.01 (number of blade 4, pitch ratio 0.73, developed area ratio 0.55)
(iii) No. 0105 (number of blade 5, pitch ratio 0.679, developed area ratio 0.6)
Some notable effects of propeller cavitation numbers on cavity extent, local cavity volume, local lift coefficient, local drag coefficient, thrust coefficient and torque coefficient are clarified concretely.

View full abstract

Comparisons between model-and full-scale propeller cavitation patterns sometimes show noticeable differences in both extent and kind.
This paper describes an improved method for cavitation tests of model propellers in a cavitation tunnel by the application of roughness at the leading edge of blades.
In the basic study, the flow visualization on the propeller blade and propeller open water tests were conducted.
Model tests of cavitation were carried out in the cavitation tunnel. Investigations were made into the correlation between the cavitation pattern and the flow on the blade as well as the effects of roughness and air content on cavitation.
Finally the model test results were evaluated in comparison with the full-scale observations, and the present method was implied to be very useful for cavitation study of the propeller in the full-scale conditions.

View full abstract

The problem on course stability of a towed vessel is of importance and interest for the ship design as well as safety of navigation. The stability of towed vessel will be affected by some factors such as the size of towed vessel or the length of tow line. Specially the wind effect is the most important factor in towing.
In this paper, the relationship between the course stability of towed vessel and wind direction and wind speed are discussed by the theoretical way. The results on a cargo ship and a ferry boat as towed vessel are as follows.
The stability of towed vessel under wind pressure depends on the inherent stability of towed vessel. However, the ship with large exposed area to wind such as a ferry boat is greatly affected by wind direction and velocity even if the inherent stability is stable.
In general, the towed vessel will be much unstable for the range from beam to quarter wind in higher wind speed than ship speed. The length of tow line has also great infuluence upon the stability, long tow line will be able to stabilize for lower wind speed than ship speed.

View full abstract

Traffic flow in the crossing points is the important factor for the traffic capacity as well as the safety of navigation. In order to remove the dangerous encounter of the ships, the authors studied the time sheared traffic control at the crossing points highlighted on the parallel route crossings (Fig. 1). Digital simulation were made to evaluate the effectiveness of the control system, the distance for time adjustment (DA) and the distance between the crossings (DC) are the controlling factors of the system. Two control system were considered in this paper, one is the priority in order of arrival (first come first service system), the other is signal system. Whether a ship is able to adjust the time or not is decided by its characteristics of the speed control. Minimum time control theory was employed as a base index of this adjustable time.
The major conclusion obtained are;
(1) The distance for the time adjustment relate with the uniformalizing of the arrival time at the crossing point. Increasing from 10, 000m to 20, 000m, the rate of the unallocatable ships are decreased 1020%.
(2) Longer distance between the No. 1 and No. 2 crossing point reduces the interference between the crossing points. In the signal system, the signal interval is dependent on the distance. Increasing the distance, the degree of freedom in the sellectting the signal interval become greater.
(3) In the priority system in order of arrival, 210 vessels/day are able to pass the crossing under the condition of DA=10, 000 m, DC=3, 000 m. Signal system has the greater capacity of 300 vessels/day if DA≥20, 000m.

View full abstract

In the former report, the author analysed many sea trial's data under the AR autopilot system developed by the authors. Emphasis of the report was mainly put on the evaluation of the course keeping quality.
The purpose of this paper is to analyse the same data from the view points of steering law subjecting to the AR autopilot system and the conventional system. One needs to have a robust method not influenced by an existence of feedback loops to attack such a complicated time series related mutually. An answer to this formidable problem is the multi-dimensional auto regressive model decision procedure proposed by Dr. Akaike.
The author proposes a new method to analyse actual ship trial data, gained at sea and uses it throughly.
He will report in this paper that such an analysis method, especially the noise contribution, the impulse response and the frequency response function calculated from the AR model give important informations to make clear the differences between the optimal steering law, the conventional autopilot steering one and a manual steering one.

View full abstract

In the paper, basic idea and some results of a central control system for two dimensional moving vehicles, e. g. ships or free-course carriers in a factory, are shown, where optimum courses for each of a group of vehicles rushing into a limited area are solved, avoiding collisions and yet keeping the minimum distance course for each vehicle as close as possible.
In the group control problems, proper functions of miss distances and deviations from minimum distance courses are formulated as criterion functions, and optimizing algorithm and characteristics of optimum trajectories are analysed for various cases.
The problems can be easily solved by small-size mini-computers for several vehicles, and time and energy losses in collision avoidance can be decreased to the minimum by taking such optimum courses, where all vehicles clear predetermined miss distances to other vehicles effectively.
A back up system for each vehicle to assure safety is also investigated in the case the central control system should go down while it controls the traffic. Such back up system is indispensable to realize the above central control system, because effectiveness in the optimum course decreases the margin for the safety. The back up system is based on a simplified coordinating algorithm of two vehicular collision avoidance solutions, and from some numerical examples, it is shown that it can work well even in the crowded case.

View full abstract

A method for quantitative evaluation of the effect of seakeeping performance on ship's mission is presented.
A concept of mission-effectiveness is introduced which is used as a quantitative index of the effect of seakeeping performance for ship system capability. In order to evaluate the mission-effectiveness, a formulation, commonly used in the reliability engineering, is applied, considering that the ship system is presented by the sum of subsystems onboard, grouped under their functions. For the determination of the effectiveness of subsystems, performance degradation of subsystems due to ship responses are made and/or assumed by using results of research works about seakeeping quality of ships so far.
The present method is applied to evaluate the effect of seakeeping performance on mission-effectiveness of ship system and performance effectiveness of ship subsystems for several kinds of patrol boats which are assumed to be operated under the same mission, “Surveillance of a ship in distress at sea”.

View full abstract

Course stability of a hydrofoil sailing boat, which is equipped with two surface-piercing dihedral foils and an inverted “T” foil, are examined as an extension of the previous works.
For the analysis of the course stability in case of manual steering, linearized equations of motion are used for surge, sway, roll, yaw and rudder deflection. The analysis indicated that the boat was stable without rudder control, except for the cases of extremely higher boat speed and following wind condition. These unstable behavior, however, will be easily improved by helmsman's rudder control.
Then the course stability with application of a windvane control system were also analyzed in terms of the same procedure as the next step. The windvane used here turns around a vertical shaft and operates an auxiliary rudder by gear linkage. Through the analysis, the optimum size of windvane was determined. The effect of the size of the windvane on the course stability are also examined through the frequency response analysis.
Finally, using the non-linear equations of motion, the yaw and other motions of the boat were simulated numerically assuming a realistic variation of wind direction, which was regenerated from a spectrum of wind direction fluctuation. The numerical simulations, as the results, necessitated slightly larger size of windvane for stabilization compared with those of the stability analysis, especially for following wind condition.

View full abstract

The ideal aim of the ship motion reduction by anti-pitching fin in waves might be a heaving-less and/or pitching-less state. Equations of motion at this state give conditions to be satisfied by fin's location, area and aspect ratio. These equations mean that wave-exciting force and moment of main hull are cancellated by fin's ones.
Calculating such fin's particulars with respect to a container-cargo ship, conclusions are :
(1) heaving-less and pitching-less state may be possible by two-pairs of fairly large fin installed at appropriate location.
(2) pitching-less or heaving-less state may be obtained by one-pair of fin. The fin area is smaller, the location be nearer F. P. in general. Optimum aspect ratio of the fin, especially for anti-pitching, is very small and lies out of assumption, so that another theoretical approach might be necessary.
The experiments in head waves are carried out and confirmed theoretical conclusions. The present results are of a container-cargo ship, there might not be much difference in general tendencies but in details for the other types of ship.

View full abstract

The Kaimei is a wave power absorber developed as a large scale device at the Japan Marine Science and Technology Center. This device is a long floating body and has many air-chambers over which are arranged the air-turbine generators. The Kaimei is specified as an attenuator and a water column oscillator because it is moored in the head sea and the wave power is gradually absorbed through the rise and fall of the water in each air-chamber.
In this paper the authors investigate the optimum design procedure of the Kaimei type wave power absorber. The optimum design means here maximizing the absorbable wave power by adjusting the external load distribution under the prescribed design period.
The Kaimei is modelled as the long floating boxes with different arrangement of air chambers. Two types of ships, that is, a catamaran buoyancy room type and a separation buoyancy room type are mainly examined. It is assumed that the water in the air chamber is treated as a column of constant mass. The hydrodynamic forces are estimated by the Green-function integral equations technique.
The following results are concluded.
(i) In estimating the performance of the converter, a simple method which assumes the uniform load distribution over all the air chambers is useful as the practical design procedure.
(ii) The performance is preferably insensitive to the variation of the external load condition around the optimum value.
(iii) The design period should be long enough because both of the frequency response characteristics and of the efficiency.
(iv) The separation type excels the catamaran type in wave power absorbability, and the performance of both types can be improved when the ship is free at some design period.
(v) The Kaimei is the middle configuration of the above two types, but the performance tends to reduce to each one.

View full abstract

In 1968, the author and his colleagues developed the new mechanism of plastic hinge based on the incremental theory of plasticity and derived the elastic-plastic stiffness matrices for one dimensional members. In 1979, extending the basic idea of the plastic hinge, the authors developed the new method of plastic analysis of plates and solid bodies using the ordinary finite element method and derived the elastic-plastic stiffness matrices. Later, the new method of plastic analysis was named “the plastic node method”.
In this paper, the basic theory of the plastic node method was further developed for analyses of thermal elastic-plastic and dynamic behaviors of structures and derived stiffness equations. Using the new theory, several examples were analyzed and the result indicated successful application of the method.

View full abstract

Most steel shell structures are built up with welding. The residual stresses and deformations due to welds have bad effects on the static and fatigue strength of shell structures. And the residual stresses often cause stress corrosion. Accordingly, many researches are brought out; on circumferential welds of pipes, straight seam welded pipes and so on. However, each of these researches separatly treats a shell in particular, and there have been little studies on residual stresses due to the weld of general shells.
This paper deals with the analytical method on residual stresses in welded built-up shell structures. The method is based on 1) the concept of 'inherent stress' due to weld shrinkage, 2) the extension of Duhamel's analogy, and 3) the geometrical characteristics of curved weld beads. Using the method the welding residual stresses in general-shaped shell structures are able to be calculated. The analytical results of the residual stresses in welded built-up shell structures-circumferential welds of pipes, circumferential welds of spherical shells and straight seam welded pipes-are shown. The results are compared with published experimental results. The analytical results show good agreement with the experimental results. The final part of this paper deals with the formula and the estimating procedure of 'inherent stress'; that depends upon the welding parameters, the geometrical characteristics of weld beads and the material property (yield stress).

View full abstract

In designing land based structures and offshore structures, as well as ships including LNG carriers, it is very important to accurately predict the buckling strength and collapse strength of axially compressed cylinders, especially with initial imperfections.
Many researches have been reported with respect to the elastic buckling strength of initially imperfect long cylinders, but few have been made of the buckling strength of initially imperfect short cylinders such as tank skirts of Moss-type LNG carriers, especially in the plastic range.
This paper investigates effects of the initial imperfections on the collapse strength of axially compressed short cylinders, utilizing the finite element method (FEM).
Main conclusions obtained from this study are as follows :
(1) Effects of initial imperfections on the buckling strength of short cylinders, not only in the elastic range but also in the plastic range, have been presented.
(2) It is important to precisely calculate pre-buckling deformations of short cylinders even if they are in the elastic range, since the non-linear behavior would greatly affect the buckling strength.
(3) When cylinders collapse by axially compressed forces in the plastic range, the extent of their plasticity may be expressed by parameter(Rσ_y) /(tE).
(4) In the plastic range, the outer initially imperfect cylinders give lower buckling load than the inner ones. This is contrary to the case of elastic buckling. Further, in the plastic range, the cylinders with Batdorf parameter Z=14 are most sensitive to the initial imperfections.

View full abstract

The algorithm for the discrete limit analysis of general thin-walled structures proposed in the previous reports is applied to the dynamic collapse problems of impulsively loaded plates and shells. The present algorithm allows the discontinuity of displacements on the interelement boundaries and it enables us to solve the problems on the limit strength of arbitrarily shaped thin-walled plate and shell structures taking geometrical as well as material nonlinearities into consideration.
In the impact dynamics of structures the static analysis by the present algorithm is a great help to the investigation on the relation of the external loading energy (or the internal plastic work) and the magnitude of permanent displacements as supposed from the conventional rigid-plastic analysis, however, in the present report the transient response analysis by the direct time-integration of the equation of motion is carried out in order to illustrate the usefulness of the present technique to the general nonlinear dynamic analysis.
As numerical examples the dynamic collapse behaviors of the impulsively loaded strip, rectangular plates, circular cylindrical shells, hemispherical shell and cylindrical panel were simulated, the results of which were compared with the finite element and finite difference solutions, theoretical rigid-plastic solutions and experimental results presented in the literatures. These numerical studies made it clear that the present algorithm can give qualitatively satisfied solutions for the dynamic plastic problems and treat about 130 spring-time steps per CPU second on a HITAC M-280 H in the implicit time integration code.

View full abstract

The air cushioning action during the horizontal impact of a flat plate on a water surface is investigated both theoretically and experimentally. In the presented theory, flow of the air layer between the flat plate and the water surface is analyzed by means of the characteristic curve method for the one-dimensional compressible fluid, while the water surface is determined with the aid of the boundary element method by disregarding the gravitational effect and the compressibility. In the calculation of air flow, the initial values and the boundary condition at the plate edge are investigated by introducing acceptable assumptions. The calculated time histories of impulsive pressure are in good agreement with experimental results.

View full abstract

A great deal of fatigue damage due to severe vibration arises at after ship body, because there are some exciting sources. In order to avoid the resonanse between the ship structure and the exciting forces, it is necessary to estimate exactly the natural frequency of ship structure. The side shell structure of after ship body which recieves directly the propeller surface force is not the flat plate as a general rule, therefore it is not easy to calculate the natural vibration of side shell structure.
In this paper, a method to calculate the natural vibration of rectangular and trapezoid shallow shell was presented. And the results by present method were compared to the results by the finite element method. The natural vibration of shallow shell can be fairly good estimated by using the present method, but the magnitude of maximum initial deflection which this method can be applied to is about 20% to the side length. It was also clarified that the plate thickness-maximum initial deflection ratio is one of the parameter which decides the natural frequency of the shallow cylindrical shell. According to the calculated results for the trapezoid shell, in the case that the upper side-lower side ratio c/a is about 0.91.0, the natural frequency is almost equal to that of rectangular shell.

View full abstract

A linear response analysis method of taut moored platforms subjected to the regular waves was proposed in the previous paper. In this method, the superstructure is assumed a three dimensional elastic frame and each structural member is divided into two kinds of elements; a hull element and a beam element, and each mooring leg is replaced by a mooring element. The hull element is an element for estimating the inertia force, the hydrodynamic force and the hydrostatic restoring force, while the beam element is for estimating the elastic rigidities of the structural members and the mooring element is for estimating the linear restoring force of the mooring leg. The applicability of this analysis method was confirmed by comparison with the experimental results using a so-called footing type of small scale model. Unexpected dynamic responses with a large amount of bending deformation, however, were observed and it was concluded referring the calculated results that these phenomena were due to the structural instability as a truss of the superstructure.
In this paper, therefore, another footing type of small scale model whose deck structure is rigid enough to neglect the instability is tested and the experimental results are compared with the calculated ones. Moreover, two kinds of small scale models, which are modified from the model in the previous paper by adding several truss members in order to maintain the structural stability, are analyzed. The structural responses of the superstructure with the sufficient rigidities and their effects on the tension fluctuation of the legs are comprehensively discussed and several important structural response characteristics are induced.

View full abstract

The behavior of the tank vertical supports of Independent Prismatic Tank System of Liquified Gas Carrier is very important because the supports are the main members which transfer the tank movements to hull structure and vice versa.
Measurements of reaction force of tank vertical supports were carried out during tank test condition and actual voyage of about 2 years. in cooperation with Nippon Kaiji Kyokai by 80, 000 m³ LPG carrier which was constructed in November 1980 by IHI.
The results were compared with analytical values and following conclusions are obtained.
1) The accuracy of analytical method applied to tank design is confirmed to be sufficient.
2) The assumed phase lag of internal and external pressure is within acceptable range and the design maximum value of supporting forces have sufficient margin against the maximum pressure estimated by measurement.

View full abstract

Stiffened plate structure is adopted to tank structure for IMO type B LNG carrier. Fatigue tests were conducted on various types of welded joint specimens and the fatigue properties of the welded elements were investigated. And the results were used to assess the fatigue strength of the structural elements of the tank. Through the course of analysis, fatigue life of the welded joints with respect to stress cycling was found to form a small scatter band when the local stress at location of crack initiation is obtained and standardized to a given stress ratio taking account of the yield stress of the material and the welding residual stress. A rational S-N curve for fatigue assessment is proposed for the local stress amplitude considering the variance of the experimental data.
The fatigue crack growth rate and the threshold stress intensity factor range were also investigated to see the effect of stress ratio R with an emphasis on a high stress ratio behavior. Tests were carried out at room temperature and cryogenic temperature (-196°C). The analysis of experimental data yielded that for 0≤R≤0.5 the base metal showed a stress ratio dependency whereas the weld metal had small effect of the stress ratio and the crack growth rate was slightly higher as compared to the base metal. But the fatigue crack growth rate for R>0.5 the dependency of R becomes small and the crack growth for the two materials are more or less identical. The fatigue crack growth rate was appreciably low at cryogenic temperature indicating that the materials have a higher resistance to crack extension at low temperature. For the basis of design and safety assessment of the tank structure, an expression is proposed which is an upper envelope curve for da/dN vs. ΔK for high R tests conducted at room temperature.
Full scale model of a cutout section of the tank namely, the bottom transverse section and frame end bracket section were fabricated and fatigue tested. It was confirmed that the fatigue strength and the crack growth of all the full scale models were conservative when they are compared with the proposed S-N curve and the fatigue crack growth curve for fatigue assessment obtained from small specimens.
A general approach for the safety assessment of the tank for the LNG carrier is also discussed.

View full abstract

The LNG tank insulation technology is one of the key technologies in LNG ship building. We developed our own insulation system, called “Kawasaki Panel System”, for a large LNG ship with spherical tanks, and applied this system to our first built LNG ship “GOLAR SPIRIT”. In her unprecedented cooling test using LN₂ and gas trial, it was verified that the performance required for LNG tank insulation was satisfactory and that “No Cold Spot” criterion we aimed at was fulfilled.
This paper summarizes the development of our insulation system, describing the outline of the major large scaled model tests carried out to select insulation materials and insulation system and to grasp the thermal efficiency, strength and workability of the system as a whole, and also referring to the specifications of the system determined through these tests and to the outline of the quality control established.

View full abstract

Slow speed air flow in an air conditioned cabin is investigated in the light of non-isotropic turbulent flow subjected to gravitational effect. The terms of Reynolds stress and turbulent heat flux appeared in the time averaged transport equations should be treated taking into account of non-isotropic turbulence in this case.
Using approximate equations presented by Launder et al, the Reynolds stress and heat flux terms considering buoyant effect are assumed to be able to be algebraically superposed that effect on eddy kinematic viscosity and eddy diffusivity of isotropic turbulent flow.
Flow velocity and air temperature in an air conditioned chamber of container type are measured experimentally to investigate the distributions of isothermal zone and heat convecting zone. Further, three dimensional analyses of heat convection by finite element method are carried out for the models of air conditioned chamber used for measurements and the results of theory and experiment are compared.
The relations between Archimedes number and buoyant effect to turbulent flow in air conditioned chamber are clarified which are useful for the air conditioning design.

View full abstract

The objective of this paper is to investigate the relationship between fracture toughnesses of 9% Ni steel at brittle crack initiation and at brittle crack arrest. It is the authors' point of view that the difference between these fracture toughnesses is due to the difference in strain rates. Based on this assumption, the effect of strain rate on the fracture initiation toughness was investigated by three-point bend test, while the fracture arrest toughness was investigated by Compact Crack Arrest (CCA) test. On comparing fracture initiation toughness, K_c, with fracture arrest toughness, K_ca, it was found that K_ca was almost equal to or higher than K_c. The cause is considered to be the effect of ductile lips left below the plate surfaces. This led the authors to the conclusion that an initiated brittle crack can inevitably be arrested after a short extension even if external force does not fall.
To verify the finding, a notched very wide plate specimen was tested. The result showed that only a pop-in type of fracture could occur, that is, an initiated brittle crack was arrested after the extension comparable to plate thickness. The notched wide plate test also verified that the reinitiation of brittle fracture from the arrested crack could not take place without the substantial increase in external load.

View full abstract

In this report, the dynamic fracture toughness K_D defined by the dissipated energy was obtained, as to several common ship hull steels.
Using the double tension type large size specimen, crack propagation test was made and the crack velocity was measured. By this results, dynamic fracture toughness K_D was calculated using dynamic numerical analysis. So that, the expression on K_D was experimentally obtained and the crack propagation characteristics were compared.

View full abstract

Experimental study has been carried out on the electron beam welds of high strength steel with mechanical and metallurgical heterogeneity.
The main attention of the present study is paid not only to the role of the locally embrittled region on the initiation behaviors of cleavage and fibrous cracks, but also to the fracture characteristics of EB-welds of high strength steel. Using specimens extracted from EB-welds of HT 80 steel, 3-point bending COD tests are conducted.
Fracture toughness of EB-weld metal of HT 80 steel used is the same as that of base metal, with reference to cleavage fracture initiation. On the contrary, fracture toughness of grain coarsened HAZ of EB-welds used is extremely embrittled regardless of fracture mode.
In the tests of specimens with mechanical and metallurgical heterogeneity along notch front, cleavage/fibrous crack initiates at first in the locally embrittled region, i. e. grain coarsened HAZ. Moreover, crack initiation toughness of those specimens are almostly the same as that of specimen whose notch is located in the grain coarsened HAZ. In the particular case that locally embrittled region is located in the vicinity of specimen surface, however, toughness measured for cleavage fracture initiation are considerably scattered, compared with the case that embrittled region is located in the middle of the thickness.
Accordingly, for the purpose of fracture toughness evaluation for materials with locally embrittled region, it is important to take into account of the effects of embrittled region not only on the crack initiation, but also on the scattering of fracture toughness.

View full abstract

Since elastic-plastic fracture toughness J_IC was proposed by Landes and Begley, considerable work on it has been carried out and systematization of the J_IC test method has been accomplished actively. ASTM E 813-81 in USA and JSME S 001-1981 in Japan, which are concerned with standard test methods for fracture toughness J_IC, have been published. From the industrial and practical point of view, it is as a matter of course that a more economical and convenient test method is desired. In the present study, a significantly simple test method for evaluating J_IC by using a Charpy impact testing machine without instrumentation has been investigated. The materials used in the present study were SM 50 B steel, A 533 B Class 1 steel, HT 80 Steel, AISI 4340 steel, Type 304 stainless steel and 7075 Aluminum alloy. It was found that the maximum load point coincides with the ductile crack initiation point in the 25% side-grooved Charpy-size specimen with the crack length to width ratio a/W of 0.6. The ratio of the absorbed energy up to maximum load E_m to the total absorbed energy E_t was found to correlate with the material constant : E_m/E_t= 58.2 (σ_flow/Eε_ƒ). From the foregoing results, a quite simple and convenient test method for estimating J_IC has been proposed as follows : (1) Prepare a 25% side-grooved Charpy-size specimen with the crack length to width ratio a/W of 0.6. (2) Carry out a Charpy impact test and measure the total absorbed energy E_t. (3) Determine the J_IC-value by substituting the E_t-value into the following equation : J_IC=58.2 (σ_flow/Eε _ƒ) × (2E_t/B_effb). Since the scatter of J_IC estimated by using the present test method was considered to remain in the narrow range of ±20%, in practical use it is not a serious problem.

View full abstract

A series of fatigue crack propagation experiments were carried out in order to examine the effects of stress ratio R (=σ_min/σ_max) upon crack propagation rates, together with the closure and opening behaviours of fatigue cracks. Fatigue tests were conducted by using center notched specimens under cyclic axial loading in various stress ratios ranged from -5 to +0.8. Crack opening stress levels were determined by the unloading elastic compliance method.
The variation of crack propagation rates under different stress ratios was well interpreted by the concept of effective stress intensity factor range ΔK_eff, defined as the product of ΔK and effective stress range ratio U. An expression of fatigue crack propagation rates, which is to be available under wide ranged stress ratios, was proposed on the basis of the present experiments and compared with those of other authors. The crack propagation behaviours near the threshold conditions are as well discussed.

View full abstract

A fatigue crack propagation model with consideration of crack opening/closure phenomenon and the inferior limit stress for hysteresis loop generated at a crack tip during part of stress cycle was already reported in the previous paper. In order to study the effectiveness, fatigue crack propagation tests are carried out by using center notched tension test specimens under various stress ratios and constant loading with over/under-spike loads. These test results are in good agreement with the simulated calculation.
In addition, the following phenomenon become clear from both the experimental results and the predicted results.
1) As a stress ratio increases, fatigue growth rates and m-value of Paris's equation slightly increase.
2) Spike overloading leads to acceleration and delayed retardation of crack growth rates. On the other hand, spike underloading leads to small delayed acceleration after acceleration and delayed retardation. An underload applied after an overload causes the crack to grow faster than when the underload occurs before the overload.

View full abstract

This paper discusses an evaluation technique for fatigue life of multiple parallel surface cracks at one sided surface of a plate under axial and out-of-plate bending loads.
This technique is a revision of the technique for multiple series surface cracks proposed in the previous paper, taking into account the experimentally-determined coalescence condition. If the initial ratio of the vertical distance H to the horizontal distance S between parallel surface cracks is 0.5 or less, or if H is 5 mm or less, these cracks propagate to unite substantially. In addition, there is no mutual influence between parallel surface cracks before coalescence in the virtual projected plane. After coalescence, these cracks propagate outside as a single crack. The fatigue crack in-going rate drops abruptly, in comparison with the out-going rate.
As a result of case studies, it was found that the present technique can give us reasonable estimated fatigue life to coalescence and penetration.

View full abstract

In order to investigate the mechanisms and the factors to govern the brittle fracture initiation during the fatigue crack propagation at low temperatures,
(1) Fracture toughness tests under wide range of loading rate
(2) Fatigue tests at low temperature
(3) Fracture toughness tests suffered from pre-loading were performed. And detailed observations and analyses were carried out on the load, the load-line-displacement and the strain near crack tip at around the fracture initiation, and on the fractography by means of the scanning electron microscope.
The main results obtained were as follows,
(1) The fracture load was always lower than the maximum load of cyclic loading, therefore, the loading rate was non-zero value at the brittle fracture initiation.
(2) Since the residual stress due to the cyclic loading was less effective on the brittle fracture initiation if it were compared with the loading rate effect, the fatigue fracture toughness K_fc was estimated as equivalent value as the fracture toughness K_c under the monotonic tensile loading. If the K_fc and the K_c were analyzed as a function of the K (=dK/dt), there is no significant difference between them.
(3) The residual stress due to the cyclic loading, though, showed large effect on the fatigue crack extension as usual manner.
The calculated fatigue life based on the Paris' formula with taking into consideration of the residual stress showed good one-by-one correspondence with the observed fatigue life up to the brittle fracture initiation.
Therefore, a design curve to determine the fatigue life at low temperature was preliminarily proposed.
(4) Comparing with the results obtained from the fracture toughness tests after having been given monotonic tensile pre-load, the pre-loading effect of cyclic loading at low temperature was shown to be explained by the ρ⁺ theory.

View full abstract

The effect of full-scale model tests, or comparable ones, to the reliability assessment of a structure was discussed assuming that (1) the strength of the underlying structure follows a three-parameter Weibull distribution, (2) its characteristic value is unknown, (3) the test load is limited to a certain value because of the capacity of testing facility or the testing scheme, and (4) a Bayesian analysis is applied to the test results.
General trends of the Bayesian reliability obtained by such a procedure were discussed focusing on the effect of the maximum testing load, number of samples avairable and the prior estimation of the structural reliability.
Main remarks are :
(1) Tests have, regardless of the sample size, no value if the test load is too low compared with the expected service load.
(2) It may be expected that the higher the test load is, the more effective it is. When sample size is very small, this may not be true.
(3) To draw effective informations from the test, including the sellection of an ample combination of the sample size and the test load, a reasonable, neither conservative nor unconcervative, prior distribution of the unknown parameter, characteristic value in the present study, is to be chosen.
(4) On chosing a prior distribution, the uncertainty of the other parameters, if any, should be taken into consideration as well as the required reliability.
(5) A rather small shape factor is recommended if there is any uncertainty on it, since misestimation of a shape factor possibly reduces the effect of a test.
(6) Any number of successive successes in supplementary tests, which must be done after getting a fail at the first test to demonstrate enough reliability can be expected, have no effect at all if test load is too low.
(7) A supplementary test with a higher load than the previous test may produce sufficient results even if the successive successes are few after the fail in the first test.

View full abstract

Static uniaxial tensile tests and water leakage tests were carried out on two types of ferrocement plates; type A and type B specimens were defined by DnV “Tentative rules for the construction and classification of ferrocement vessels”. Type A specimen consists of 4 to 12 layers of the same type of netting. Type B specimen consists of one layer of reinforcing rods and 6 or 12 total layers of netting. The main reinforcing netting used in specimen was a woven wire cloth with mesh width of 8 and 10 mm. Two series of experiments using equal contents of the mortar were tested.
The tensile strength and the displacement behaviors of ferrocement plates due to tensile load, such as elastic modulus, Poisson's ratio, and elongation, were discussed in connection with the volume fraction and the specific surface of reinforcement. Leakage behaviors were studied from the viewpoints of the crack width and the ratio of applied stresses to the ultimate tensile stress of a ferrocement plate. Basic data for design were obtained in this study.

View full abstract