The effect of grain size on the increase of flow stress of polycrystalline cadmium due to hydrostatic pressure was investigated. Tensile tests and pressure change tests were carried out under 10MPa and 600MPa on various specimens with different grain sizes. The results obtained are summarized as follows. (1) A strain softening phenomenon was observed on the specimens when hydrostatic pressure was changed from 10MPa to 600MPa. The amount of decrease of the flow stress under 600MPa decreased with increasing prestrain at 10MPa, and it was larger for the 18μm grain size specimen than for the 36μm one. (2) The flow stress at a particular strain value was able to be related to the grain size d by the same form as the Hall-Petch equation, σ=σοε+κεd-1/2 where the intercept σοε and the slope κε are experimental constants. (3) σοε at 10MPa and 600MPa increased with an increase of strain, while κε decreased gradually. σοε at 600MPa was larger than that at 10MPa, while κε at 600MPa was smaller than that at 10MPa. (4) The rate of increase Δσ/σ of the flow stress due to the pressure change to 600MPa increased with increasing strain and grain size. For the specimen with the largest grain size of 111μm, Δσ/σ became the maximum value of 20%. It is considered that this large increase of flow stress due to hydrostatic pressure is caused mainly by the pressure dependence of σοε, and the grain size dependence of Δσ/σ is brought in by the decrease of κε due to pressure.
The corrosion behavior of metal is closely related to its polarization behavior in a corrosive solution. The polarization behavior is said to be useful for understanding pitting corrosion or stress corrosion cracking and hydrogen embrittlement in particular. In this study, in order to elucidate the relation between the catholic potential and mechanical properties, the slow strain rate tension tests were carried out under various cathodic potentials. The main results obtained are summarized as follows. (1) Under tensile stress, the pitting initiation potentials of SUS 430 and SUS 304 shifted towards the less-noble side but SUS 316 did not show such a shift. The hydrogen overpotential of the testing materials shifted towards the noble side with increasing applied stress in the elastic region. Consequently, the hydrogen evolution reaction in 3%NaCl solution was promoted by the presence of stress. (2) In the case of SUS 430, the degradation of mechanical properties (reduction in size and time to fracture, etc.) became more pronounced with increasing setting cathodic potential, but in the case of molybdenum-bearing SUS 316, the degradation of mechanical properties was less than the other materials. The above phenomenon can be also understood from the observation results obtained by S.E.M.. Such a correlation between the mechanical properties and electrochemical properties became more pronounced at slower strain rates.
The path independency of J-integral proposed by Rice has been accepted to be valid in both linear and nonlinear elastic materials. This independency, however, has not been necessarily proven for incremental elasto-plastic materials. In this paper, at first, FEM analyses based on deformation and incremental plasticity theories were performed in order to compare the solutions obtained through these two theories. Some factors governing the path independency of J-integral were examined with an emphasis on the loading condition of an element in the vicinity of crack tip. From the analyses and examinations, the applicability of J-integral according to the incremental plasticity theory was clarified. Next, the results of FEM analyses based on the incremental plasticity theory were investigated to verify the adequacy of analytical assumptions employed to develope suitable formulas estimating J-values from a single load-displacement curve. The applicability and limitation of these formulas were confirmed and advanced. As a consequence of these studies, a new formula which is a modification of Rice's estimation procedure was proposed.
The J contour integral has been applied mainly to homogeneous cracked bodies, but not to heterogeneous cracked ones. In this paper, the FEM type analyses of J contour integral were performed on heterogeneous cracked bodies of linear elastic or elasto-plastic materials, especially those of three-phase materials, such as welded joints. From analytical and numerical standpoints, the revised J-integral formula offered in this paper was verified to be path independent. Furthermore, it was extended to the problem of heterogeneous cracked bodies. Through these results, the validity of the solutions and the effect of the strength ratio of weld metal to base metal on the J-value were revealed. Moreover, the J-integral was proved to control the normal strain ahead of the crack tip of three-phase materials.
JSMS database on fatigue strength of metallic materials was compiled as a joint project of the JSMS Committees on Fatigue and on Reliability Engineering. A large number of numerical fatigue data and the related information offered by many researchers were filed on a magnetic tape after computer processing. The PN data (Statistical fatigue test data) and the ST data (Staircase fatigue test data) of ferrous metals were sorted from the database, and statistically analysed to find the basic nature of the distribution characteristics of endurance limits.
Fully annealed commercial base pure iron plate specimens were tested under completely reversed bending stress at 60Hz. The crack nucleation and its early growth were observed until about 80% of the fracture life at a stress level just above the endurance limit. Then the specimen was removed from the machine and the grain orientation was determined for grains surrounding cracks by the etch pit method. Observation was also made after removal of surface layers to see the crack growth pattern toward the inside from the surface. It was found that cracks originated at grain boundaries on the surface penetrated into the inside along two or three dominant slip systems in one of the grains adjacent at the cracked boundary in the same manner as the pencil glide which was commonly observed in b.c.c. metals.
This paper describes the effect of a small notch at the toe of the reinforcement on the fatigue strength of welded joints. Both large and small notches appeared at the root of the toe. It was clarified from a detailed survey on the toe shape of the joints welded under several conditions that the radius of the large notch was larger than that of the critical radius. The fatigue strength of welded joints was determined by the crack initiating condition when the fatigue strength was evaluated based only on the shape of the large notch at the toe without regard for the effect of the small notch existing at the root of the large notch. The depth of the small notch was about 35μm, and was less than the length of nonpropagating cracks which was in the range of 40-80μm depending mainly on the hardness of the carbon steel used in this study. This means that the small notch does not give any effect on the fatigue strength of welded joints. The out of plane bending fatigue tests revealed the validity of the above estimation on the fatigue strength of welded joints. The fatigue strength of welded joints was not affected by the small notch but depended on the shape of large notch at the toe.
The fatigue crack propagation characteristics at Stage 2 in spheroidal graphite cast iron (s.g. cast iron) have been investigated by fracture mechanics and fractography, and compared with those of steel having almost the same tensile strength. Fatigue tests were carried out at room temperature in laboratory air under stress ratio, R=0.1-0.8. The relationship between fatigue crack propagation rate, da/dN, and stress intensity factor range, ΔK, in s.g. cast iron was divided into Stage 2a, 2b and 2c in the same manner as in general metallic materials. At Stage 2a, the oxide deposits produced by fretting of fracture surfaces were observed on the fracture surface regardless of the stress ratio. The amount of the deposits depended mainly on the roughness of fracture surface, and the da/dN at Stage 2a was lower than that in steel because the roughness-induced crack closure encouraged the oxide-induced crack closure. On the other hand, the da/dN at Stage 2b was faster than that in steel because of the lower Young's modulus. At Stage 2c, the main fracture pattern was cleavage fracture and the da/dN was accelerated.
In order to investigate the effect of axial-torsional straining phase on the elevated-temperature biaxial low-cycle fatigue life, axial-torsional strain controlled fatigue tests were conducted on SUS 304 stainless steel at 550°C under in-phase and out-of-phase conditions between axial and torsional straining. Based on the experimental results, a criterion of biaxial low-cycle fatigue was discussed in this study. The fatigue life under out-of-phase straining was found to be shorter than that under in-phase straining when compared at the same Mises equivalent strain range or maximum shear strain range. From the observation of slip-bands and micro-cracks on the specimen surface by means of a replica method, this appeared to be due to the change in fracture mode arising from the difference in macroscopic strain condition between in-phase and out-of-phase straining. Accordingly, a unified correlation for both the in-phase and out-of-phase biaxial fatigue lives could be obtained by expanding the Brown-Miller's theory of equivalent shear strain which contains the maximum shear strain and the normal strain across the maximum shear plane as the factors controlling the fatigue crack initiation and propagation under biaxial loading.
The relation between high-cycle fatigue behavior and tensile properties was examined at room temperature and 750°C on two kinds of material with different ductility: Ni base superalloy (forging) and Fe base superalloy (casting). The high-cycle fatigue tests were carried out at room temperature and 750°C, and the effects of tensile strength, ductility, stress ratio, notch and frequency on fatigue strength were examined. Moreover, the relation between fatigue behavior and fracture morphology was discussed based on fractographical observations. The main results obtained are as follows. (1) The high-cycle fatigue strength of two kinds of material with different ductility was dependent on tensile strength but not on ductility, except for the test result at 750°C under the stress ratio of 0.5. (2) The influences of the ductility of the materials on the stress ratio effect, notch effect and fatigue deformation were recognized at 750°C under the stress ratio of 0.5. (3) A good correspondence between the change of fracture morphology and the parameter of fracture mechanics (Kmax/E) was observed independently of test temperature. The effect of the ductility of the materils on fracture morphology was recognized and the fracture surface of Fe base superalloy was influenced by its dendrite structure. (4) The values of fracture mechanics parameter (Kmax*/E) corresponding to the critical length (as) of fatigue stable crack growth were almost constant for the materials independently of test temperature, if the fatigue deformation was small.
Tensile tests were conducted on specimens of plain woven glass cloth/epoxy laminates with a notch. The effect of the notch root radius ρ on fracture was investigated. In the previous study, the three-point bending tests were carried out to make clear the effect of the notch root radius on fracture. In this study, the experimental results were compared with those of the previous study.The F.E.M. simulation was performed, and compared with the experimental results. The following results were obtained: (1) As the notch root radius increased, the smaller ρ specimen showed the fracture behavior which was caused mainly by the propagation of a stable crack, but the larger ρ specimen showed the fracture behavior which was caused mainly by the spread of the damage. (2) The spread of the damage zone was different from that of the three-point bending test, but the effect of the notch root radius on strength remained unchanged for the tension and the bending test. (3) The F.E.M simulation explained the experimental result of the larger ρ specimen well.
The commercial silver-Ni, or W, or CdO contacts have been produced mainly by the powder metallurgy, which seems to be rather expensive from a metallurgical viewpoint. In order to develope a more economical method, the infiltration method used to prepare fiber-reinforced metal composites was applied in the present work to make silver composite contacts with Ni continuous thin wire, alumina whisker SAFFIL, and carbon chopped fiber. Ag-Ni composite was successfully prepared by an unidirectional solidification, and Ag-alumina whisker and Ag-carbon fiber by a centrifugal infiltration method. The Ag-30wt%Ni (No.2) composite contact was found to have a superior contact property than the traditional Ag-15wt%Ni contact, and to be able to save a considerable amount of silver. The Ag-30vol% carbon composite contact showed a comparable property to Ag contacts, though its contact resistance was slightly larger.
Characteristics of several kinds of I.R.P. using optical fiber were investigated. The fundamental structure of these pyrometers is that the infrared flux radiated from an object is accepted and transmitted to an infrared detector by an optical fiber, and converted into electric signals by a detector. Therefore, this type of pyrometer is suitable for measuring the temperature of a very small object whose temperature changes rapidly. Besides, the flexibility of optical fiber makes it possible to measure the inner temperature of the object by drilling a microscopically fine hole in it and inserting the fiber. In the present study, SiO2 and fluoride fibers were used as optical fibers and Ge, PbS, InAs and InSb cells were used as IR-detectors. The results are as follows. These two kinds of optical fiber had the upper limit wavelength of light transmission depending on their core materials, so that this type of I.R.P. had the lower limit of measurable temperature. By using SiO2 fiber, PbS-I.R.P, was useful for the temperature above 300°C, and InAs and InSb-I.R.P. which have a high speed response were suitable above 400°C with a good degree of accuracy. Ge-I.R.P. had a high speed response but was inferior in sensitivity. InSb-I.R.P. using a fluoride fiber whose spectral transmission loss is much smaller in the range of 2-4μm than that of SiO2 fiber was useful for the temperature range above 150°C. The optical fiber with a core of larger diameter was effective for improving the sensitivity of I.R.P., but had a little effect on the lower limit of measurable temperature.