Measurements of the three-dimensional flow velocity components (u, v, w) are carried out around a 2 m Inuid Model M 21 where a specially developed mini five-hole pitot tube of 6 mm diameter is applied satisfactorily.
The velocity components are analyzed (u, v, w-analysis) in a synthetic co-ordination with the corresponding wave analysis (C-analysis) which was previously reported on the same model (M21).
The study indicates that, i. e.
For potential field ;
(1) The two linearized free surface conditions, i. e. the “atmospheric” condition as well as the “kinematical” condition, are both valid either between ζ (measured) and u (measured) or between ∂ζ/∂χ (measured) and w (measured).
(2) The behaviour of the transverse componentv (y) is the most important, because this, component alone can give the unique informations for sheltering effects which are not included in the wave-analysis.
(3) The results analyzed from u or w are almost similar to those already obtained by wave analysis.
For viscous field ;
(4) The flow field analysis is useful also for the refinement of the wake-survey by enabling the quantitative estimation of the disturbance velocity (u₁) which is fictitiously presumed within the wake belt.

View full abstract

A ship form of minimum wave resistance with bulbs is obtained by an application of the minimum wave resistance theory to the ship form with bulbs. The ship is assumed to be wall-sided except the vicinity of both ends and represented by a combination of doublet distribution on the center plane and point singularities such as doublet, source and sink at fore and aft ends.
The optimum density of doublet distribution and the optimum intensity of point singularities at both ends are determined by means of calculus of variations. The solution corresponds to the optimum sectional area curve and the optimum size of bulbs at both ends.
The numerical study indicates that the optimum ship form with bulbs has considerably less wave resistance than the optimum ship form without bulbs. This theoretical prediction is verified by experiments.

View full abstract

The present analysis is an application of the bow-flow theory developed by Ogilvie (1972) to a flat body on the free surface. Ogilvie's “bow near field” is one on which ∂/∂x =0 (ε^-1/2), whereas ∂/∂y, ∂/∂z=0 (ε^-1), so that exactly the same linear free-surface conditions that are familiar from the classical thin-ship theory are used. In this report flat bodies are represented by a distribution of pressure on the waterplane z=0. An approximate theory representing flow around bow is developed and compared with some experiments. The agreement is encouraging and suggests that the singular behavior of the pressure distribution at the leading edge may be related to the wave-breaking phenomenon around the bow.

View full abstract

When the waves generated in bow region traveling along a long slender-body, they meet deformation at middle body region. This phenomenon is difficult to explain by the classical wave-making theory. So far, the deformation of ship wave has been observed by many researchers, and this is known as sheltering effect of body.
In this paper an attempt for analysing the ship wave deformation is made under the assumption that the wave number v (ratio of gravitational effect to dynamical effect) is of order 0 (ε^-1). The assumption means the waves generated by bow part are of the same order as the beam of ship. The problem is solved by the method of matched asymptotic expansions using slender-body theory.
Incidentally, it becomes clear that as far as the diffraction of ship waves near a slender-body concerned, there are no differences between steady and unsteady ship motions. This is true under the condition that wave length is of the same order as beam. So, the near-field problem coincides with that of Faltinsen's¹⁾.
The result obtained here can explain qualitatively the phenomena which have been discovered in the experiments on an extremely slender-body²⁾.

View full abstract

Bent trailing edge of propeller blades was studied in view of plastic deformation caused by cumulative impact load due to collapse of cavity.
It is found that the geometrical shape and microstructure of bent trailing edge of propeller blades are similar to those by shot blasting. These facts not only throw light on the mechanism of bent trailing edge but also lead to a simulation method.
This simulation method will be used for studying the intensity of load due to collapse of cavity quantitatively and for developing an effective prediction method of both bent trailing edge and cavitation erosion of propellers.

View full abstract

The Report deals with a characteristic flow field around a slender rectangular thin wing, where free vortices, which separate from side edges, roll up and form a symmetrical spiral vortex sheet. The aims are to obtain lift distribution and to grasp the macroscopic flow field around the wing.
Referring to the observed flow pattern around the wing by means of a hydrogen bubble method in a water tunnel, a simple flow model is assumed. The model consists of a pair of core vortices and feed vortex sheets which emerge from side edges to the core vortices.
With the simplification of the flow field, boundary conditions need to be made less detailed. That is, the condition on the wing surface is satisfied only along the center line of it, and the condition for the free vortices is simplified to the condition that total forces acting on the free vortices are zero.
By the method, reasonable results are obtained for normal force coefficient, center of pressure and average location of core vortices.

View full abstract

The equilibrium of an oil layer on flowing water is treated as a two-dimensional fluid motion problem.
Low and high velocity regions are considered separately. In the former two types of water flow was studied and the result is that there is not so large difference between them as to the calculated length of the oil layer. In the latter a simple approximate solution using Lewis form and an alternative one were obtained. Calculated length of oil layer agreed rather well with experiments in both regions. A short description is given to the critical stream velocity for the leakage of oil as well as necessary corrections for the flow velocity.

View full abstract

In the case when the long-term distribution of wave induced variable of a ship is estimated theoretically, it becomes usual to suppose wave condition on the North Atlantic Ocean. It is the reason why the condition is chosen, that accurate wave statistics on the sea zone are prepared and the sea zone is regarded as one of the severest main trade routes, so supposing the condition means estimating the distribution at a safety side. Though it is known that the long-term distribution changes according to supposed wave condition, analytical study on the problem that by what factor of supposed wave statistics the severity of supposition is evaluated has never been made.
From investigation of character of the extreme value of the response variable, author finds that severity of supposed wave statistics is evaluated by such factors as the maximum wave height in each wave period interval. Thus the severity of a given wave condition is shown by drawing a curve which shows change of the maximum wave height along the abscissa of wave period. Comparing a given condition with another one, the condition is regarded severer than another one for smaller ships but moderate for larger ships when the maximum wave height is higher than that of another condition in shorter wave period range and it become lower than that of another condition in longer range. By drawing the curve from the wave statistics by Walden, the curve from those by Roll, and the curve from those by Hogben and Lumb, it is clarified that the estimated extreme value changes largely by supposed wave statistics in some case. Being so high the maximum wave height in the wave period interval from 5 sec. to 7 sec. of the wave statistics by Walden, the effective wave height curve corresponding the extreme value of wave bending moments of similar ships of various length has a hump and hollow when the wave statistics are supposed in estimation.
The height of the maximum wave height is determined by result of only several observations. So the height of each wave statistics is not accurate. By assuming the frequency distribution of wave heights in each wave period interval is described by the Weibull distribution, author corrects the maximum wave height in each wave period interval of wave statistics including those by Walden, those by Hogben and Lumb, those on the sea zone 3, and those observed on board of a Japanese patrol ship is corrected. And the severity of the corrected wave condition on the North Atlantic Ocean are compared with that of the one on the North Pacific Ocean. From the comparison, it is found that the wave condition on the North Pacific Ocean has at least the same severity as that on the North Atlantic Ocean, when wave period is less than 10 sec. And when the period is over 10 sec, it becomes slightly moderate in comparison with that on the North Atlantic Ocean.

View full abstract

The motion is investigated of a box-type floating vessel which is moored among the regular waves in the sea of finite depth by the chains and on which the constant drifting force acts. The present authors assume the floating vessel to be very blunt, and the length-breadth ratio of investigated one is 1.
The aim of the present report is to investigate the following problems :
(1) Is the strip method acceptable for such a blunt body as the above mentioned from the engineering standpoint ?
(2) How much is the nonlinear effect of the chain reaction on the motion of the floating vessel ?
The linearized equations of motion are solved analitically. On the other hand, the nonlinear equations of motion are integrated numerically by Runge-Kutta-Gill method.
The results of the theoretical calculations show the followings :
(1) If the various damping forces of the rolling motion could be estimated in any way, the motion of the blunt body among the waves may be estimated by the strip method.
(2) If the initial tensions of the chains are small, the nonlinear effect of the chain reaction on the stationary harmonic motion of the floating vessel is very small. A rough standard condition for the small tension is that the chains touch horizontally the sea bottom at the anchoring points.
(3) When the unstationary or constant drifting force acts on the moored floating vessel and it drifts from the initial equilibrium position largely, the nonlinear effect of the chain reaction is very large. Hence, the motion in the irregular wave, under the constant drifting force or under the large initial tension should be estimated considering the nonlinearity of the chain reaction.

View full abstract

Recent advances in the research of ship motion in a seaway have enabled to predict the ship motion. These researches, however, are restricted to those of displacement type ships such as a merchant ship and a destroyer, and as to behaviours of high speed planing boat in a seaway few available researches have been published.
In the present study the towing tank test of a torpedo boat model was carried out in regular head wave and experimental data were discussed along with the results calculated by Ordinary Linear Strip Method (O. S. M.).
The conclusions deduced from these studies are as follows,
1) High frequency components appear in pitching and heaving oscillations at higher encounter frequency, that is, at higher speed and shorter wave. In these regions, each mean value of the pitching and heaving oscillation becomes larger compared with the trim angle and vertical displacement in still water.
2) The motions calculated by the O. S. M. give good agreement with experimental data for Fn< 0.5, but they do not always for higher Froude number.
3) In spite of much higher frequency components in the motions, the amplitudes of pitching and heaving motions are almost proportional to those of the wave, that is, the linearity to wave height is nearly conserved.

View full abstract

When we want to estimate the added mass and the added mass moment of inertia of a ship, the strip method which has proved to be very useful for the calculation of ship motion in waves, seems to be suitable, because it is possible to take the hull form of a ship into consideration in detail. However, since the strip method is based on two-dimensional consideration of the flow field around a three-dimensional body, both the added mass and the added mass moment of inertia obtained by the strip method are always excessively large compared with their exact values.
Therefore in this paper, the correction factors are introduced separately for the added mass as to the swaying motion, and for the added mass moment of inertia as to the yawing motion in an analogous way to the correction factor for the added mass of the vertical ship hull vibration.
Especially, the effect of finite water depth on these correction factors is investigated from the viewpoint of ship's manoeuvrability in restricted waters. As a result, it is found that the correction factors decrease with decrement of the water depth, and that the correction factor for the added mass moment of inertia is smaller than that for the added mass.

View full abstract

The main purpose of an auto-pilot is the course-keeping. The author has determined a performance index for course-keeping from the viewpoint of economic run. It consists of (1) path elongation due to detour, (2) speed reduction due to steering, and (3) fuel consumption increase per unit time due to steering. Full-scale experiments are carried out with M. S. Fukae Maru (360 Gross Ton.) to obtain the dynamic loss factors corresponding to (2) and (3). The performance index is given as follows :
J=lim z→∞∫^z₀1/2 {(θ-β) ²+λδ²} dt
where λ_??_45 for the shortest time of propulsion, and λ_??_67 for the minium of fuel consumption. The optimum control law is derived from the performance index and given below,
δ=- [θ/√λ+θ {√1+2KT/√λ+K_β²/λ-1} /K]
The control character is of PD-control. Some results of computer simulation are shown.

View full abstract

In the previous paper, very violent impact pressures on stem were measured by the experiments in high transient water waves.
In this paper, the authors continued the experiments in order to take a somewhat closer look at some aspects of water impact to clarify the roles of certain mechanisms. Both the impact on stem in full load condition and the impact on bottom in the ballast condition were investigated. Ship model was equipped with relative wave surface probe, so the relative velocity and angle between hull and wave surface were obtained.
The following conclusions can be deduced from these studies :
1) Large impact pressures acting on hull surface can be explained by the fundamental natures that are taken by theories and experiments of the drop of the two dimensional body or the breaking waves on the vertical wall.
2) The maximum value of impact pressure, P_I, is considered to be proportional to square of relative velocity, V_n, and the coefficient of impact pressure, C=P_I/ (1/2ρV_n²), reaches 48 to 66 at the bottom, 17 at the stem.
3) When the relative angle between hull and wave surface at the moment of impact, β, is large, the coefficient C tends to be proportional to the square of cot β, as the Wagner's theory. As decreases, it comes to be proportional to cot β, as the Chuang's experiments.
4) When air is entrapped, large impact pressure occures, but at the stem air is hardly entrapped because of its shape.

View full abstract

The second wave-maker of plunger type has been set up at a right angle to the first one in the No. 1 Ship Model Basin of the Ship Research Institute to generate short-crested waves from two directions, as a part of a co-ordinated programme of long term research based on the view that it is reasonable to expect a fairly complete description of various hydrodynamic forces acting on ships in waves in relation to ship motions to be achieved eventually, and that this will provide a tool capable of solving most structural design problems.
Model tests in short-crested regular cross waves were carried out and analyzed with special emphasis on the roll and pitch motions of an ore-carrier at the full load condition.
The ship responses in long-crested regular waves were simply superposed, and the results were compared with those measured in short-crested regular waves. The varidity of the linear superposition principle was examined to the prediction of ship motions at various headings to the waves.

View full abstract

This paper describes the results of a model test on the impulsive water pressures acting on the hull surface of a gigantic ore-carrier in waves. In order to generate impulsive water pressures on the free running model, especially on the ship side, the regular waves from two directions were adopted as the wave environment of the experiment.
From analysis of the test results, it is discussed what parameters affect occurrence of impulsive water pressures and on what part of the hull surface impulsive water pressures will appear.
A new measuring system including a special pressure transducer and recording apparatus was prepared. The system worked well to catch impulsive water pressures even in detail of the peak precisely.

View full abstract

Three-dimensional, semi-cylindrical 1/3-scale bow models were droptested in the water impact generating tower to determine the characteristics of wave impact pressure acting on the bows of very large ships. The models were dropped onto the still water surface from various heights (5 m at the maximum) at small relative angles (0°, 5°, and 15°) and with the total drop weight of 10 tons each time. From the test results, relations between the maximum impact pressure P_max, relative bow-wave impact velocity V₀, relative stem-wave angle α and radius of bow curvature R were obtained. It was found that the maximum impact pressure would increase approximately as the square of the relative bow-wave impact velocity and would be affected more by the relative stem-wave angle than by the radius of bow curvature. No air-entrapment occurred even though the model was dropped at the relative stem-wave angle of 0°. The acoustic pressure ρcV₀ was detected when the model hit the water surface at the relative angle of 0° and at the relative impact velocity of 8 m/sec. The impact pressure generated was the unsteady hydrodynamic pressure and, in the severest impact conditions, its rise was restricted by the acoustic pressure. The maximum impact pressures as measured on the surface of one rectangular section of the curved model plating were not distributed uniformly and simultaneously, as distinct from the uniformly distributed impact pressures, and were not considered to directly contribute to the structural response of the plating.

View full abstract

This research was carried out with purpose of developing a method to estimate the ore load acting on ship structures.
The Coulomb's theory has been used to calculate the soil pressure, but that theory dose not adequate for the present purpose. In this research, it is assumed that a body of ore-pellets can be treated as a continuous medium having nonlinear stress strain relation, the material coefficients of such a medium were determined from the results of compaction tests, triaxial tests and friction tests.
Using a cylindrical test tank, the cariblations for the soil pressure cells used in the present research were performed, then the ore pressure distribution in the tank, the total pressure acting on the bottom of tank and the effects of the cylinder wall on the ore pressure were examined.
The behaviours of the ore in the tank were calculated by the finite element method with the material coefficients obtained from the above basic tests. The calculated results roughly conformed to the experimental results.

View full abstract

This is a report of experiments on ore pressure using pellet. Some of the conclusions of the semi-static behavior of ore pressure are as follows :
(1) There is no evident difference in foundamental soil properties by the kind of pellet.
(2) Ore pressure distribution on ore hold is strictly governed by loading method, and some scattering of ore pressure is observed due to complex behavior.
(3) Ore pressure on longitudinal bulkhead in an ore carrier is greater than the value of the Coulomb's active earth pressure, and could be estimated rather by the value of the Jaky's rest earth pressure than by the value of the Coulomb's active earth pressure. But, it is not exactly estimated by those conventional theories by reason of the variety of coefficient of earth pressure due to ore hold depth.
(4) Ore pressure distribution of increment at rolling on longitudinal bulkhead in an ore carrier is rather uniform than triangle alongside the depth of bulkhead.
(5) Static ore pressure on flexible longitudinal bulkhead in an ore carrier varies widely with void ratio of pellet.
(6) A structure strength analysis in which the compressive spring is substituted for pellet is useful as an approximate solution. However, it will be necessary that an analysis of pellet taking into consideration dilatancy phenomenon.

View full abstract

Generally, when designing the ore carrier, the Coulomb's theory of soil mechanics is used to determine the ore pressure which acts on the tank walls. However, the are pressure calculated using the Coulomb's theory is the static are pressure, and, actually, such factors as wave loads, ship motions, hull vibrations, etc., would demand due consideration. The authors, therefore, carried out some model tests to analyze the dynamic ore pressure and thereby formulated an experimental equation in order to estimate the ore pressure in more realistic manner. And, using the experimental equation thus formulated, the authors analyzed the distribution of ore pressure in the cargo tank of a 250, 000 tons deadweight ore carrier. Consequently it was found that the ore pressure acting on the longitudinal bulkhead in way of the wing-tank strut where a considerable amount of dynamic loads were exerted by wave, etc., would be 35 times greater than it was calculated using the Coulomb's theory, as a result of the static ore pressure being superposed by dynamic ore pressure due to horizontal and vertical ship motions and by pressure from the spring action of ore due to deformation of the longitudinal bulkhead.

View full abstract

In this paper, an outline is presented of the total system of computer programme that has been developed by Nippon Kaiji Kyokai for the rational and direct evaluation on the longitudinal strength of hull girder of oceangoing vessels.
This system is consist of an integrated group of individual programmes which include an input data processing and automatic generating programme as well as those concerned with the theoretical analysis on the oscillating motion of ships among waves with use of strip method, structural analysis and stress calculations of hull girders based upon the general bending and torsion theory of thin walled beams, and the statistical analysis on the fluctuating stresses in longitudinal members of the ship. The out put of the computations gives a direct information on the probability of exceeding a certain prescribed value of the stresses caused in the longitudinal members during the whole life time of the ship.
Then, a discussion is made into the detail of a series of calculations which are performed on the longitudinal strength of existing oil tankers, of which the deadweight capacity varies from 50, 000 to 500, 000 tons. It is found from this analysis, that the fluctuating stresses in the longitudinal strength members due to vertical bending of hull girders are almost in the same level among these ships, whereas those due to horizontal bending of the hull girders show a gradual rise with increase of the size of the ships.

View full abstract

In addition to the longitudinal bending and shear deformation of ships in the vertical plane, the effect of longitudinal bending and shear deformation of ships in the horizontal plane coupled with the torsional deformation is particularly considered in the method of theoretical analysis.
This theoretical analysis treats the ship as a beam shell with various cross-section in which the actual structure of ship is rather faithfully considered.
Concerning the structural idealization, ships are divided into substructures with uniform cross-section and then each substructure is also divided into small strip elements.
Further, the experimental works are carried out, using a model of container ship with large hatch openings. And it has been confirmed that the calculated results obtained through this method are in good agreement with the experimental results.
The results of this investigations led to the following conclusions.
In dealing the longitudinal strength of container ships considering the horizontal bending and torsion, the analytical method described here seems to be effective for the practical use in designing of ships.
In the structural analysis of container ships, the cross decks between hatch openings should be treated as a reinforced member arresting the torsional deformation of main hull instead of treating them as a part of it. And it is appropriate to adopt 1.0 as the fixedness of the end connection of cross decks for the practical use in the calculating the torsional strength of the main hull.

View full abstract

There are several papers concerned with vibration analysis of rectangular plates without stiffeners subject to inplane loads. In this paper the authers studied the vibration of reinforced rectangular plates subject to inplane loads. And they found that the frequency of a reinforced rectangular plate under uniform compression, pure bending or pure uniform shear is less than that under no in-plane loads as well as in case of plates without stiffeners. The analysis of plates with reinforcement under any boundary conditions can be treated simply and generally by this method.

View full abstract

It has become important recently to determine the maximum strength of various structures which have initial deformation caused by fabrication.
On stiffened circular cylindrical shell as one of those cases, the authors carried out a study on 2-stage buckling where the shell panels between the stiffeners buckle at first, to be followed by buckling of the stiffeners.
For this purpose, the characteristic equations of elastic buckling were formulated under compression and pure bending load considering the reinforced members discretely.
The buckling load of stiffeners, after the panel has buckled, was determined by iterative computation on the framework model.
The compressive buckling experiments were carried out on three small and two large models with different arrangement of the rings, and the pure bending buckling experiments on a small model. As a result of the experiments, it was found the buckling loads of stiffeners were almost equal to the buckling loads of the panel and the effects of arrangement of rings on stiffener buckling loads were negligible.
The experimental value of buckling loads of stiffener has fallen about 25% in the case of compressed buckling and about 19% in the case of pure bending buckling because of initial deformation and torsional deformation of stiffeners.
In conclusion, the authors can say that the stiffened cylindrical shells which have similar configuration as these models, have appropriate buckling strength if the allowable load are taken to be equal to the buckling load of panel.

View full abstract

A structure consisted of shallow shell with stiffners is usually analyzed by the method simulated to folded plates or three dimensional frameworks in actual ship-hull-design. These methods of analysis cannot take account of the effect of curvature of shell strictly. One of universal and powerful method for this subject is finite element method. However, finite element method has inconvenience such that extensive calculation due to a large number of freedom is usually needed and in addition increasing number of freedom does not necessarily assure to increase accuracy except by a few particular cases.
This paper presents an analytical method on strength of shallow shell with stiffners by improved finite strip method proposed by author to lighten the volume of calculation. Displacement functions of strip elements are employed the approximate function obtained by Kantorovich method. Those give exact solutions for some special cases and which satisfy the equilibrium equation of shallow shell or very close ones to them for other cases and this procedure is effective to make decrease of discretization errors.
Some examples are shown for the cases of analyzing bending of shallow shell without and with stiffners of which all edges simply supported. The results prove that this method is one of effective means for analyzing of shell structure.

View full abstract

The main purposes of this reseach are to find out some conditions in which bending cracks initiate and propagate and to propose a new reasonable and practical method of bending test.
The behaviour of wide thick steel plate under uniform bending of large displacement and in large plastic strain were calculated as plane strain state with a new displacement function θ and with plastic strain increment Δεθ=ln√ B- (AB/r) ². Using the Tresca's strain hardening law and stress-strain diagrams of tensile test and compressive test, or tortional test, the behaviour of bending test was analysed by the method with good accuracy.
In order to find some practical bending test method, bending test under concentrated load were carried out. Relations between bending moment and bending strain from the test agreed with theoretical values, if the frictional force at supported ends were considered. Effects of this frictional force on the magnitude of bending moment of specimen are pretty large, but the effect on the initiation and propagation of bending cracks are very small.
For these tests, the effects of the shape of specimen, radius of mandrel and critical bending angle were found and some reasonable method of bending test for thick wild steel was proposed.

View full abstract

The effects of plastic constraint and temperatures on fracture initiation from a notch were investigated by using various types of notched specimens, at the temperatures ranging from room temperature to liquid nitrogen temperature.
The results show that, when fracture initiates by fibrous crack, the COD at fibrous crack initiation is almost independent of plastic constraint of the specimens or of the test temperatures, while the COD at maximum load or at final fracture is largely dependent on them. When fracture initiates by cleavage crack, the critical COD also shows considerable dependence on those parameters above mentioned. These behaviors of critical COD are explained by the fracture mechanism in each mode of fracture.

View full abstract

From critical COD δ_c of HT 80 welded bonds, brittle fracture initiation temperatures at the residual stress field are calculated using COD concept. These temperatures are compared with the temperature Ty, which is calculated from K_c using the equation K_c =√ Eσ _yδ_c· T_y indicates the temperature at which the brittle fracture initiation stress reaches to the yield stress of the material. It is clarified that the temperature T_y can be substituted approximately for the brittle fracture initiation temperature of welded joints with the residual stress of the longitudinally welded bead. The validity of T_y is certified by an example of cross-welded notch tests.
This paper proposes to employ this temperature T_y for the quality assessment of welded joints of high strength steels. T_y can be obtained from bending COD tests and tensile tests at low temperature.
An experimental example on HT 80 welded bonds shows a good correlation between T_yand Charpy transition temperature _v T_s, consequently giving the relation between the Charpy requirement and the defect size at HT 80 welded bonds.

View full abstract

Many authors, Yamamoto et al., Byskov, Rao et al., Walsh, Tong et al., etc. have discussed the availability of finite element methods to the calculation of stress intensity factors of cracked plates. Some of them successfully tried the use of special finite elements at the crack-tip-region in order to obtain a higher accuracy. But these are not always very easy to apply to the practical problems because of the complexity of calculation models. The present paper surveys some general relations among variables in finite elements, and proposes a substructure model, as a general extention of Walsh's method, which has a large liberty of choosing crack-tip-element characteristics and reduces the dimension of stiffness matrix. Some simple numerical examples are given to verify the utilities of this model and to examine the influence factors to the accuracy of solutions.

View full abstract

On the large structures such as ships's hull, tank structure and etc., various loads are acting. In some structural members, therefore, out-of-plane stress due to bending often may become larger than in-plane stress due to stretching.
In this report, the characteristics of brittle fracture initiation of mild steel plate with a through crack under the combined action of stretching and plate bending have been studied both experimentally and theoretically using the linear elastic fracture mechanics (L. E. F. M.).
As the results of this investigation, the following conclusions are obtained.
(1) Effect of bending load on the characteristics of brittle fracture initiation can be evaluated over a fairly wide range by the criterion proposed by authors using the concept of energy balance based on L. E. F. M.
(2) From the results of eccentrically loaded and center-notched Deep Notch Tests, it is extremely safe for design that the maximum surface stress is considered as the equivalent fracture stress at higher temperature than -180°C.
(3) The evaluation of fracture phenomena using the equivalent K-value proposed by authors is confirmed to be fairly good for expressing the behaviour of the material such as mild steel plate.

View full abstract

Fatigue crack propagation experiments were carried out on a mild steel, its heat-affected-zone (HAZ) and weld metal. A full thickness side notch of radius 0. 5 mm, width 1 mm and depth 14 mm was machined on both sides of a specimen 10 mm thick and 70 mm wide. For HAZ and weld metal specimens notch center was made to locate at the centerline of the material, that was made by manual welding of right-angled triangular type edge preparation. Zero-to-tension repeated axial stress, which was less than 10. 0 kg/mm², was applied on a specimen, and fatigue crack growth rate was measured on both surfaces of a specimen with the aid of a magnifying glass.
Macroscopic crack growth rates showed a dependency on stress intensity range, ΔK, following the relation of C (ΔK) ^m. The mean value of the power m could be determined to 1. 7 with little deviation, while constant C was observed as a function of stress range. No significant difference in fatigue crack growth rate behaviour was distinguished among base metal, HAZ and weld metal. Comparison of the present results on the dependency of log C on m with the relations proposed by Kitagawa, Koshiga and Gurney-Maddox showed a crossing tendency to the previous relations.
Fractographic investigation on fatigue fracture surfaces revealed the following : (1) Distributions of striation spacings are followed to normal distribution, (2) No significant difference in topography is observed among base metal, HAZ and weld metal, (3) Featureless topography is observed at a point approximately 0. 2 mm inside from a specimen surface, (4) Striation spacing defined at a point approximately 0. 4 mm inside from the surface is quite same as that in the midthickness, and (5) The value of 1. 3 is obtained as an approximate value of the exponent in the relation s=C_s (ΔK) ^ms.

View full abstract

Fatigue crack growth characteristics were investigated using center notched plate specimens. The results showed that the mean stress has effect on the growth rate and the growth rate can be expressed using following formula which takes the mean stress effect into account,
da/dN=C (K_max/ΔK) (ΔK) ^m
Experimental work was also done to investigate fatigue crack propagation behavior of semi-elliptical surface cracks subjected to combined tensile and bending stresses. Off-centered plate specimens with mechanically introduced surface cracks were used for this test. Estimation on fatigue crack growth was then made on such cracks applying the results of theoretical analysis done by Shah and Kobayashi on stress intensity factor for semi-elliptical surface cracks. From the analysis, it was found that the larger the amount of bending the longer the extension is along the surface. Experimental results showed similar trend to analytical results with respect to fatigue crack front configurations.
These results were taken into account to discuss safety in liquid containers on the basis of leak before failure concept.

View full abstract

In the previous paper, a cumulative damage law for prediction of fatigue life of unnotched plate was proposed and in this report, using the proposed cumulative damage law, fatigue crack initiation life of notched plates subjected to cyclic loading in low cycle range was discussed.
It is assumed that initiation of an engineering size crack at a tip of notch was not only dependent on the local cyclic plastic strain Δε_p and cyclic creep rate Δε_m at the notch root, but on strain ratio |ε_x|/ε_y at the notch root and plastic strain concentration factor K_ε, which represent some mechanical condition near the notch root.
As fatigue damage by Δε_p and Δε_m was assumed to be ΣΔε_p^a and ΣΔε_m respectively, so the effect of |ε_x|/ε_y and K_ε are assumed to be represented by a following form.
a_N =a_p · F (|ε_x|/ε_y, K_ε)
where a_p and a_N are exponent for unnotched plate and for notched one.
In the above expression, decrease of |ε_x|/ε_y and K_ε shorten the initiation life.
In order to decide an experimental formula of function “F”, notched specimens of mild steel were tested under constant repeated axial load and function “F” was found as follows
a_N = a_p {0.483 exp (1+|ε_x|/ε_y) -1.01} {-1/K_ε^0.12+2}

View full abstract

Recently one sided automatic welding has widely been prevailed in the assemble stage of Japanese shipyards. As one of the problems of welding procedure, as well as any other welding techniques, cracking which sometimes occurs near the end of weld, has become an important problem.
The authors have proposed the calculation technique in the seam welding. The most significant feature of the method differing from the ordinary welding thermal stress analysis is that the conditions of calculation are simulated to be those of more realistic welding procedures.
Then, the following assumptions are imposed in order to calculate stress and strain during welding.
(1) Before welding, specimens are connected by tack welds and prepared edges are free boundaries.
(2) Molten region containing arc tip is free from stress.
(3) After the arc passed by, the plates are connected by solidifying of the molten pool, that is, free prepared edge changes to supported one as temperature is cooled down.
(4) When the arc passed through a tack weld, it should be melted away and free from stress.
In order to calculate the thermal stress analysis during welding, firstly we solved un-stationary heat conduction problems and obtained temperature distribution during welding. Based on the calculated temperature distribution, we analysed uncoupled thermal stress problems by using the finite element method, and conclusively showed that one of the main causes of end cracking is the transient tensile stresses generated near the end part of weldment when the stress re-distribution is taken place due to the melting of the last tack weld.
To endorse the numerical calculation by the proposed method, experiments are carried out and it is confirmed that the method could be applicable to this kind of transient welding stress-strain state with time depending varying boundary conditions.

View full abstract

Succeeding to the previous 1 st report, in which the erection stage was stated to be planned to optimum, the author has introduced another part of SASP (Sasebo Scheduling Programme) which deals with the assembly stage. As the assembling subsystem of the consistent total shipbuilding system naturally takes its own behavior, he has also analyzed its general features, and has concluded that it is especially the essential condition in assembly stage for grasping the process and promoting the efficiency to have workers engage in both fitting and welding works.

View full abstract

At the time of hydraulic test and drydocking, there are possibilities that the bottom shell of large vessel and the foundation of dock are locally damaged.
As an example for the former, we measured the reaction in the blocks and deflection in the bottom shell, etc. at the hydraulic test of No. 3 C. O. T. (C) of 125, 000 DWT tanker in the building dock, and calculated the reaction and deflection on modelled conventional framed structure, considering the hull to be supported by springs composed of wooden blocks and foundation.
We had little defference between measurement and our calculation and it proved that our assumed calculation was correct and proper.
As an example for the latter, we have selected the case as our countermeasure against the damage (her bottom shell had partial plastic deformation by block reaction), which we experienced at the final docking of 215, 000 DWT type tanker.
We have changed our previous practice of softwood liner touched with bottom shell to the practice of sandwiching it in between hardwood blocks.
We could, consequentially, succeed to prevent any damage on and after.

View full abstract

Mechanical and thermal properties of several kinds of plywoods, which were trially manufactured for the thermal barrier of the LNG carrier, were experimentally proved.
Material tests were carried out at room temperature, -78°C (dry-ice) and -196°C (liquid nitrogen).
The results are as follows :
(1) The obtained coefficients of variation are for the most part distributed within 20 percent and those of Young's modulus around 10 percent.
(2) Breaking strength and Young's modulus are proportional to the specific gravity.
(3) Breaking strength, proportional limit and Young's modulus have tendency to increase in cryogenic temperature region.
(4) Regarding size effect of test specimen, the mean values do not show evident difference, but the standard deviations decrease evidently according to the increase of size.

View full abstract

In designing the supporting system of the self-supporting tank of LNG carrier, it is very important to evaluate the frictional force acting at the sliding parts. But there is only few available data for such an object.
The authors carried out serial tests for measuring the coefficient of friction of some materials which are supposed to be used for the supporting system, under cryogenic temperature and obtained some available data and useful informations for the actual design.

View full abstract