A composite material, ferrocement, is a thin cement mortar with multiple layers of reinforcing mesh. The strength of composite materials is dependent upon the strength of reinforcing materials. The purpose of this paper is to investigate the fatigue strength of reinforcing meshes used for ferrocement. Specimens with weave angle are extracted from square woven wire cloths, and tested under pulsating load. The woven wires used for these specimens are about 1 mm diameter-cold drawn and galvanized low carbon steel wires. The sizes of mesh openings. are 6, 8 and 10 mm. For the comparison, 2%- and 5%- prestrained woven wires and a straight wire are also tested. The summary of concluding remarks is as follows.
The knuckle points of the specimens are subjected to bending moment as well as axial load. Therefore the S-N curve between the nominal stress S and the number of cycles to failure N does not have linearity. On the high stress level close to the yield point where the weave angle is diminished, the fatigue strength is improved by work hardening, though the shortest fatigue life appears on the middle stress level. Therefore this S-N curve may be difficult to be applied to design purposes.
On the middle stress level of S-N curves, the fatigue life shows minimum cycles when the extreme fiber stress of a knuckle point calculated by elastic analysis has maximum value. The maximum strains, or peak strains, calculated by elastic stress analysis at a knuckle point are used to convert a S-N curve into a ε-N curve. The obtained ε-N curve is given by ε·N^0.148=0. 0206. The equation provides a significant coverage for all the experimental data of the specimens with different weave angles. The fatigue strength at N=2 × 10⁶ cycles becomes ε=0. 0024. This value is close to the proof stress (2% strain), the fatigue strength (2 ×10⁶ cycles) of the straight wire and the 5%-prestrained woven wire. This ε-N curve can be applied easily to design.

抄録全体を表示

In order to evaluate the fatigue life in the stress-concentrated region of structures using the results of strain-controlled fatigue tests, it is necessary to know the cyclic local strain range rather than the monotonous strain behavior. In the present study, a simple method to estimate the local strain range under cyclic loading was investigated.
Namely, the elastic-plastic finite element analyses under cyclic tensile load were conducted on the highly stress-concentrated models besides the results in the first report. Then an investigation was made into the effects of applied cyclic stress condition, elastic stress concentration factor and material properties on the shakedown and the behaviors of local stress, strain ranges and their concentration factors.
As a result, the formulae to estimate the shakedown limit and the cyclic local strain range were derived. Therefore, when the elastic stress concentration factor and the material properties such as cyclic yield stress, workhardening modulus and parameter of the Bauschinger effect are given, the cyclic local strain range can be estimated for a certain applied cyclic stress range. Then, by substituting the estimated strain range into a relationship between the strain range and the failure life obtained by the strain-controlled fatigue test, the fatigue evaluation can be done reasonably for the structural members.
The stress-controlled fatigue test on notched plates and the strain-controlled low cycle fatigue test on small smooth specimens were conducted on 50 kgf/mm² class steel. Then the estimated local strain range and fatigue life were compared to the experimental results. As a result, the estimated values agreed well with the measured values. In addition, the comparisons were made between the estimated fatigue lives on notched plates and the results so far published. Then it was found that the method proposed in this study can be applied to evaluate well the fatigue lives in the wide range from 5 × 10² to 5 × 10⁶ cycles in the notched plates of 41 to 80 kgf/mm² strength levels.

抄録全体を表示

As showed in the previous paper by authors, for the estimation of fatigue life on allowable stress for fatigue failure of structural members and mechanical components, etc., the method based on S-N diagrams obtained from random loading fatigue tests is considered to be much more reliable than the method using cumulative damage laws. But it will take a very long period of time to carry out the above tests completely, and is considered to be impossible actually. Therefore, unavoidably Miner's Rule is generally used at present.
In this paper, a new method of fatigue test was proposed. It is a combined method of random loading fatigue tests and analytical calculation, and is able to obtain the complete S-N diagram under random loading during a very short period of time.
To inquire the applicability and the accuracy of the proposed method, the following random loading fatigue tests were carried out for plane and notched mild steel specimens. They are axial reversed load controlled tests with exponential distribution. Estimated S-N diagrams in the region of under fatigue limit were compared with experimental results and a very good agreement was found quantitatively.
So the proposed method is considered to be very useful for the estimation of fatigue life or allowable stress under random loading.

抄録全体を表示

In this study, seven types of load controlled fatigue tests were carried out under variable stress amplitude in order to clarify the stress-strain response and the fatigue strength of SS 41 and STPY 41 plate specimens. The tests were conducted in constant stress amplitude loading, step stress amplitude loading, repeated high-to-low and low-to-high block loading, two types of block-random loading and random loading with exponential distribution. In each test, stress-strain loops were recorded with X-Y plotter from the start of cycling to visual crack initiation by using the strain gauges put them on the test piece.
As a result, the following conclusions were obtained.
(1) S_a-ε_pa (stress amplitude-plastic strain amplitude) curves are different with each variable loading patterns.
(2) Cyclic softening for random loading is greater than that for constant stress amplitude in the conditions of S_a≥σ_y and S_a≤S_aw, while the relation is contrary to above in the conditions of S_aw<S_a<σ_y, where S_aw is a fatigue limit and σ_y is a static yield stress of the material.
(3) S_a-ε_pa curve for random loading with exponential distribution can be obtained approximately by the step stress amplitude loading.
(4) Fatigue life estimations by Miner's rule and Modified Miner's rule make over estimation compared with experimental results for block, block-random and random loadings.
(5) Fatigue life estimations by Manson-Coffin's law do not coincide with the experimental results for block-random loading and random loading with exponential distribution included the large repeated number of stress amplitudes below fatigue limit.
(6) Fatigue life estimations by Corten-Dolan's method fairly coincide with experimental results for various kind of variable stress amplitude loading.

抄録全体を表示

In the previous report, authors proposed a new mechanical model for the fatigue crack initiation and investigated the fatigue life properties under variable stress amplitude.
In this study, using the montecarlo method the fatigue crack initiation lives were estimated for random loads with exponential, weibull and uniform density based on this model.
The outline of the model is as follows. In fatigue test, although great many micro fatigue damages are occured in the material, the number of cracks which can grow up to visual length is limited, because many micro cracks among them can not penetrate the grain boundaries. Actual fatigue crack initiates from the position where the number of cycles for visual crack initiation takes the smallest value in the material.
In the model, the distribution function of the crack initiation life, N_c, is expressed as in equation (3) and (4). In the equations, ƒ_ΔNm (n) is the probability density function PDF of the number of cycles consumed by the micro crack propagation of the m-th number of trans-granular, f_{ΔNm, w} (n) is the PDF of the number of cycles consumed by the crack penetration of m-th number of grain boundaries. And symbol signifies the convolution integral, k is the set of the number of micro fatigue damages on the material surface.
The reliability system of the model is expressed as in Fig. 2. After the PDF of ƒ_ΔNm (n) and ƒ_{ΔNm, w} (n) are determined, the montecarlo simulations are carried out in accordance with the flow chart of Fig. 8.
On the above random loads, the effects of the shape parameter, the scale parameter, the maximum stress and the minimum stress amplitude of the stress frequency distributions are considered. Estimated results are a good agreement with the experimental results obtained by authors and others qualitatively.

抄録全体を表示

In this paper, discussion is made on a problem of determining the inspection interval appropriately to keep the reliability of ship structure at a certain level which may be equal to the initially estimated one at the design stage. Bayesian reliability analysis method is applied to the evaluation of the results of repeating inspections of hull structural members to be incorporated with improvement of the reliability of ship structures. A general formulation is given on the following assumptions that (1) fatigue life of the structural member considered follows a two-parameter Weibull distribution, (2) its scale parameter is unknown, and (3) the uncertainty of inspection is introduced as the average detectability which is determined from the probability distribution function of fatigue crack size and the detectability of certain size of a crack.
Actual inspection records are analysed by the proposed method to proove its effectiveness for estimating the safety of ship structures after inspections.

抄録全体を表示

A general judgement method to evaluate material properties of advanced composites such as carbon, aramid and hybrid fiber reinforced plastics has not well been established on account of a much broader range of properties and a much stronger anisotropy than conventional glass fiber reinforced plastics. The number of factors to affect a scatter of properties is generally larger in composite materials, which are composed of much different phases of constituents, than in conventional monolithic materials : a variation in defects mixed in and dispersion state of constituents during a composition process tends to make the amount of scatter in mechanical properties of composites larger in addition to a variation in constituent material properties themselves.
Furthermore, as the failure process of fiber reinforced composites is a very complicated accumulation process of damage due to random failure of fibers, matrix and interface, which leads to a catastrophic fracture, it should be necessary to introduce a reliability assessment system to evaluate effectively a decrease in strength due to cumulative damage and defects taking every aspect of variation into consideration in order to understand thoroughly the statistical nature of strength properties of composite materials.
For this purpose, the present paper aims at establishing a general assessment system to predict a descrease in reliabilty of composite materials due to cumulative damage, with a main system to simulate a stochastic failure process of composite materials considering the effect of a scatter in strength of fibers and matrix and defects mixed in during a fabrication process, accompanied with a subsystem of systematic statistical analysis.
In this report, a new failure simulation model of composite materials has been introduced considering the effect of matrix shear failure as well as fiber breaks based on a shear-lag theory in which a failure can occur randomly not only in fiber elements but also in matrix elements : a stochastic tensile failure process has been simulated by means of a Monte Carlo method based on a repeated increment scheme using a finite difference technique. Then a Weibull analysis technique has been applied to determine design allowable properties and the results have been described as 'A' and 'B' allowable levels for unidirectional carbon and glass fiber reinforced plastics.

抄録全体を表示

An experimental investigation was conducted to study the effect of applied stress on the increment of crack opening displacement (COD) under constant load condition.
It has been shown in several investigations that the time-dependent increment of COD is observed at constant load after monotonic loading, when COD at the beginning of load holding is greater than δ_i, which is a critical value for fibrous crack initiation.
In this paper, the effect of applied stress on the increment of COD at constant load was investigated on large sized center cracked specimens as well as on small sized ones. As a result, it was found that the time-dependent increment of COD is affected strongly by the applied stress level rather than the COD level, and that COD increases sharply at constant load beyond general yielding. The mechanism of time-dependent increment of COD under constant load condition was also discussed.

抄録全体を表示

3-dimensional welding residual stress distribution of repair weld was measured and a wide plate test with a surface notch at repair weld was carried out. The brittle fracuture strength from the 3-dimensional crack at the repair weld was evaluated using COD criterion. The results of the present study are summarised as follows.
3-dimensional welding residual stresses in the thickness direction are measured. These values are seemed to be independent of the repair weld depth. CTOD values obtained from wide plate test are equivalent to those from COD test, provided that the microstructure of the notch tip is similar between them. The brittle fracture strength of the 3-dimensional crack at the repair weld is reasonablly estimated by using equivalent 2-dimensional crack, COD design curve and the criterion of δ_R+δ_σ=δ_c, where δ_R, δ_σ and δ_c are CTOD due to residual stress, applied stress (strain) and its critical value, respectively. The effect of welding residual stress is analysed quantitatively by means of the proposed method of the present study, but practically, WES 2805 gives the reasonable and safety-side estimation for it.

抄録全体を表示

The objective of this paper is to clarify the mechanism of crack arrest behavior using surface-notched double tension test proposed in previous papers.
The static and dynamic deformation behavior of the system measured were compared with the results of numerical analysis on centre crack model and single-edge crack model. The full thickness dynamic tear test was carried out to observe fracture phenomena associated with the crack initiation, propagation and arrest. The surface-notched Dynamic Tear test using the specimen with the sharp surface notches located in the crack initiation part and the crack propagation part was also performed to clarify the effect of the deformation at the surface portion of the specimen on the behavior of the crack initiation and propagation.
It is noted that the test results obtained could deny the evaluation of the crack arrest toughness based on the K concept. According to the full thickness Dynamic Tear test, the main factor affecting the crack arrest should be considered the macroscopic plastic deformation and the resulting ductile fracture. The significance of the crack arrest was also confirmed by the surface-notched Dynamic Tear test.
Many similarities in the fracture appearances were observed between in the surface-notched double tension test and in the Dynamic Tear test. It can be clearly seen that there was a close relationship between both test results.
As a result, the full thickness Dynamic Tear test would be a convenient engineering test method in order to evaluate the crack arrest properties of materials.

抄録全体を表示

From the viewpoint of energy balance concept, the resistance of material against fast fracture can be measured by the capacity for absorbing energy dissipated due to crack extension. With respect to brittle fracture of steel, energy absorption is associated with the formation of both flat brittle fracture surface and shear lips.
In the previous report, stress gradient type double tension tests using mild steel were carried out and it was found that shear lips were able to absorb rather large amount of energy and played an important role to arrest behavior of brittle crack. In order to evaluate the arrestability of steel, it is thus necessary to clarify the mechanism of initiation and growth of shear lips.
In this study, experiments are carried out to investigate the factors which may influence the formation of shear lips. Surface-notched double tension specimen using mild steel is adopted. Crack propagates initially in the notched region where shear lips may be restrained, and then propagates into full thickness region where crack may arrest due to fall of driving force and/or shear lip formation.
Data obtained are analyzed on the basis of dynamic fracture mechanics. Shear lips contribute very much to crack arrest phenomena and are influenced by crack velocity, temperature, constraint (stress triaxiality) and applied stress level in this case.

抄録全体を表示

The effect of strain rate and temperature on the dynamic fracture toughness K_Id and the critical COD δ_cd of structural steel has been experimentally studied. Dynamic tensile test was carried out using 1 T, 2 T and 3 T-CT type specimens produced from steel for pressure vessel of 125 mm in thickness (A 533-Cl. 1).
Tensile rates were changed from static (0.2 mm/s) to 8 m/s at loading device in fracture test. Test temperature were selected mainly in which the specimen fractured without unstable fibrous crack at notch tip.
The results are summarized as follows. The values K_Id< and δ_cd decrease with increase in tensile rate (K_I or δ). But it is supposed evidently that there is maximum tensile rate affecting to the fracture toughness behavior. Fracture toughness values obtained by present study have had not so good correlation with the strain rate and temperature parameter T_K·ln (A/ε) excepting the low fracture toughness level. We attempted a modification to strain rate-temperature parameter concerning to the strain rate with which it was appreciated the unification of the effect of strain rate in different strain rate excepting over 2 × 10³ strain.
It is suggested to investigate the local strain behavior at the crack tip corresponding to the load point tensile rate.

抄録全体を表示

For pressure hulls of Deep Submersible Vessel, the most important subject is to perform the maximum load carrying capacity with the minimum weight. From this point, steel with high strength and high toughness is adopted to pressure hulls.
In this case, from the fields of welding, the most important point is to produce the welding joints free from cracking and with enough material characteristics. The authors have carried out a series of research to establish the welding procedure of high strength steel of 80 kgf/mm² yield point of welding condition against cracking and of selection of optimum welding procedure. On the basis of the research, a full scale model was manufactured and evaluated as welding structure. As for weld cracking, the authors reported in another report, now in this report the selection of welding condition and the manufacturing of the full scale model are reported.

抄録全体を表示

A simplified, approximate method of power flow is proposed as a function of space averaged driving point mobilities and space averaged free vibration of machines.
By using the calculated model, with beams which compose a machine and a ship hull, the characteristics of power flow are simulated. In paticular, it is shown that the vertical, horizontal and rotational component have nearly equal contributions, and the above mentioned simplified method is in agreement with the simulated results.
In addition, power flow from a reduction gear to a machine room model ship is measured and the results are compared with the simplified method showing good accordance.
As a convenient method, free vibration of the reduction gear has been estimated by the sound power level, thus demonstrating the effectiveness of this method.

抄録全体を表示

A lage quantity of high pour point oil which tends to be solidified at a normal temperature is carried by the oil tankers that are equiped with oil heating system for purpose of keeping in low viscosity of crude oil. As cooling bottom plate by sea water, the crude oil near bottom is solidified in spite of heating up and the solidified layer is likely as insulation against heat flow through bottom. The solidified oil blocks up oil courses bored at bottom structural members and prevents from gathering oil at the final period of pumping up crude oil. Besides, the tank capacity is decreased since solid oil at bottom is just unpumpable.
In this paper, the predicted method on thickness of solidified oil at bottom is proposed in comparison with the measured data on an actual tanker and the numerical analysis on Stefan-problem is carried out by using the enthalpy with delta-sequence function in order to investigate the solidifying process of high pour point oil. Furthermore, the experiment of melting solid paraffin by heating up in the tank in the shape of bottom structure on the scale of one-third is executed in addition to calculate by use of physical properties on Minas crude oil. By the results of measurement, it has been recognized to be practically difficult for melting the solidified oil between frame space without heating coil and some useful consequences for design of oil heating system are obtained.

抄録全体を表示

Expansion bellows are widely to absorb thermal expansion of tubes in pipeline system. Considering narrow piping space and small number of special positions where pipes are strongly fixed, the designing of pipelines for ship service requests the expansion bellow to have shorter effective length and smaller spring rate.
The accurate stress-strain analysis concerning the deformation of bellows is necessary in order to assure the safety of the pipeline system. However, the mechanical behaviour of multilayer bellows has not been sufficiently analysed.
Stress and strain applied to a monolayer bellow and a multilayer bellow during loading of axial deflection and internal pressure were experimentally measured with strain gage, and also calculated with the aid of elastic finite element method. The measurement and calculation were conducted with a 200 A-diameter bellow made of type 304 stainless steel.
In the course of the finite element analysis, the gap/friction elements were developed to model the inter-surface between each layer of the mutilayer bellow. The friction coefficient was assumed to be 0 or 10000.
When axial deflection was applied to a monolayer bellow, it was found that the calculated results from the finite element method were in good accordance with the experimental results from actual measurement. In case of a multilayer bellow, the calculated results coincided well with the observed results when the friction coeffecient was 0. It was noted that strain and stress generated in a multilayer bellow were much smaller than those in a monolayer bellow.
In case of internal pressure, the finite element results of a monolayer bellow fitted to the measured results. The measaured values of stress and strain in a multilayer bellow were nearly equal to those in a monolayer bellow. Whereas the calculated values of stress and strain in a multilayer bellow with the friction coefficient of 0 were a little larger than the observed values. However, it was confirmed that the finite element method was more suitable for the stress analysis of multilayer bellows than the EJMA equation.
Multilayer bellows have an advantage over monolayer bellows with respect to endurance because smaller amount of stress is generated through loading, in the former than the latter.

抄録全体を表示

The cathodic protection is an effective anticorrosion technique for offshore structures below the water line. Therefore, it is important that the potential distributions around structures are made clear in process of structural design. In conventional methods for numerical analysis, iterative technique had been applied to determination of potential distributions in corrosion cells.
In this report, a more convenient numerical method to analyse the potential distribution under the cathodic protection is proposed for the purpose of a rational anticorrosion plan. In this method, the potential distribution can be obtained by the application of FEM once for all, on the assumption that polarisation characteristics of electrodes in elctrolyte are linear.
Although polarisation characteristics in a real cell are nonlinear, the present method can be applied to such a real cell by means of linearisation of polarisation characteristics in each domain of potential. The potential distribution of the galvanic cell, in which mild steel in 3% NaCl solution was protected with sacrificial anode of zinc, was analysed by the present method and was measured also. The numerical solution of the potential distribution was in good agreement with measured values.

抄録全体を表示