This paper discusses the influence of irreversible hydrogen on the fatigue crack initiation life in cold drawn eutectoid steel. Fatigue tests were conducted for the sample having an artificial small notch fabricated by focused ion beam (FIB) system. Samples were cathodically hydrogen charged, and internal hydrogen states were changed as follows : (a) virgin sample and (b) the one that contained only irreversible hydrogen. In the notched sample, the fatigue strength was lower than that of smooth sample. The slope of the S-N curves became steeper and the influence of irreversible hydrogen was not observed. This is because the fatigue life was almost dominated by fatigue crack propagation life, resulted from a crack being easily initiated at a notch root owing to high stress concentration. The surface fatigue crack propagation rate, obtained by using fatigue life data for notched sample, was not affected by irreversible hydrogen. Using the obtained fatigue crack propagation rate and the rate in high vacuum corresponding to the propagation in a fish-eye region, the crack initiation life from an internal inclusion for smooth sample was estimated, and it shows that the irreversible hydrogen reduced the fatigue crack initiation life, compared with those of a virgin sample. This is a reason why the fatigue strength of a smooth sample was decreased by irreversible hydrogen.
This study aims to evaluate characteristics of hydrogen embrittlement cracking (HEC) of sensitized duplex stainless steel with simulated heat-affected zone (HAZ). The HEC of the steels, aged for 20hr at 650°C, 750°C and 800°C, were studied using a testing machine which can control the stress intensity factor, K, for the precracked compact type (CT) specimens subjected to hydrogen charging. Fracture mechanisms were studied by acoustic emission (AE) method in conjunction with metallographic observation. Fracture toughness, Kc, for the steel aged at 650°C was higher than those of others. The steel aged at 650°C, with hardened ferrite phase, suffered hydrogen-induced brittle cracks in front of main crack, while austenite phase exhibited ductile fracture. For the steels aged at 750°C and 800°C, both the ferrite and sigma phase suffered hydrogen-induced microcracks.
Due to advantages such as high strength-to-weight ratios, specific castability and recycling efficiency, magnesium alloys are used in a wide variety of industrial applications. Stress corrosion cracking (SCC) tests of AZ31B magnesium alloy in distilled water and 2∼8wt% sodium chloride solutions at 298K were performed with slow strain-rate technique. In a distilled water, the SCC of the magnesium alloy occurs surely and its susceptibility increases with a decrease of strain-rate. This magnesium alloy in sodium chloride solutions is very susceptible to the SCC under open circuit condition, regardless of the strain-rate and solution concentration. The SCC in 4wt% sodium chloride solution under potential-controlled condition occurs near the corrosion potential, which lies to the potential range of hydrogen evolution. Furthermore, the fracture surface is characterized by transgranular quasi-cleavage appearance. The evidence introduced here supports a hydrogen embrittlement for the SCC of AZ31B magnesium alloy in sodium chloride solution, including an anodic dissolution mechanism by chloride ion.
The extreme-value distribution of metastable-pit depths on SUS304 stainless-steel electrodes was examined. The electrode with the surface area of 10mm2 was immersed in a 10mass% NaCl solution at 60, 70 or 80°C, and measured its potential noises for 48hours. The depth of the metastable pits was calculated from the electric quantity of the local anodic currents that were estimated from the measured potential noises by using the RPS method, assuming the shape of the metastable pits as true hemispheres. The maximum pit depths for 7 to 8 samples at 70 and 80°C were fairly approximated to a Gumbel distribution. The minimum surface area required for forming the stable pit was estimated about 10000mm2, under the conditions employed in this study. From the return period equivalent to this surface area, the critical pit depths at 70 and 80°C were estimated 6.4 and 9.4μm, respectively.
Classification of corrosion-AEs from noises is attempted for the source location of acoustic emissions (AEs) from storage tank in service. As the AEs are monitored by resonant-type sensors with center frequencies lower than 50 kHz and amplified by 60dB, the system detects both the corrosion-AEs and noises. Noises are generally minimized by changing the threshold value of the monitoring system or interrupting the monitoring during strong winds and rain drops, but not be separated based on the quantitative waveform analyses. We first studied waveform characteristics of AE events monitored by sensors mounted on annular plate or side wall of the tank, and then separated corrosion-AEs from noises. Percentage of corrosion-AEs to the total events are found to be 30∼50% of the total events monitored by the sensors on the annular plate in windless day, but less than 10% at the monitoring on windy day.
In order to evaluate the effect of aluminium oxide on crevice corrosion resistance of carbon steel which is used as a regenerator vessel material in the absorption refrigeration cycle, immersion tests and electrochemical measurements of carbon steel in concentrated lithium bromide solutions were carried out. As a result, when there was mill scale on surface of carbon steel and there existed magnetite as impurity in concentrated lithium bromide solution, it was found that addition of aluminium oxide to the solution was very effective to protect crevice corrosion. It was inferred that aluminium oxide dissolved partially in concentrated lithium bromide solution at high temperature, and deposited into the surface oxide film of carbon steel. Therefore, cathodic reaction on the surface of carbon steel was restrained and the potential of carbon steel became less noble than repassivation potential of crevice corrosion.
Silver, a typical material used in electronic devices, corrodes in sulfur gas emitted from rubber. To evaluate this corrosion damage of silver, we have established a new method for estimating the two emission characteristics of sulfur gas from rubber : initial sulfur-emission concentration and decrease rate of emission. Firstly, we used a long-term silver-corrosion test with rubber at 70–30°C to investigate the emission mechanism and emission characteristics of sulfur gas from free sulfur, i.e., non-linked sulfur, in rubber. This investigation showed that the free sulfur in rubber is consumed by crosslink reaction between rubber chains and by sulfur-gas emission. The sulfur-consumption rate is controlled by the crosslink reaction at 70–50°C, while it is controlled by sulfur-gas emission at 30°C. The emission characteristics at the temperature that the sulfur-consumption rate is controlled by sulfur-gas emission (i.e., 30°C) can be estimated by corrosion analysis considering sulfur-gas diffusion and the corrosion reaction of sulfur gas on silver. Secondly, we investigated the relationship between the amount of free sulfur and the emission characteristics. This investigation found that the amount of free sulfur can be calculated by integrating concentration of sulfur-gas emission. Given that result, we were able to establish a new method for estimating emission characteristics from the amount of free sulfur by chemical analysis and the initial concentration of sulfur-gas emission by a short-term silver-corrosion test with rubber.
Si3N4/SiC composite containing Y2O3 as sintering additive was hot-pressed to examine their crack-healing behavior as a function of oxygen partial pressure (pO2 = 50–21,000Pa). A semi-elliptical surface crack with length of about 100μm was introduced on the center of tensile surface by the indentation method. These specimens were healed at various temperatures and times using tube furnace. After the crack-healing process, the bending strength of each specimen was measured at room temperature and at high temperatures of 800-1,400°C. Even under pO2 = 50Pa, the cracks were completely healed by heat treatment at 1,300°C for 10h. Bending strengths of the crack-healed specimens exhibited almost the same strength as the smooth specimens healed. Moreover, the specimens exhibited an almost constant bending strength (∼800MPa) up to 1,400°C.
Sphere indentation tests were performed to clarify damage properties on surface of Advanced Pore-Free SiC (APF-SiC) with excellent damage tolerance. The ring crack initiation strength properties were discussed from fracture mechanics viewpoint based on FEM analysis and detailed observations of growth behaviors of the ring crack and the cone crack. As compared to conventional SiC, the initiation strength was nearly equal to regardless of sphere sizes. However, its scatter was increased with increasing sphere sizes. Clusters are formed by uniting some in multi-micro cracks which occur near surface with increasing the contact load, one of that from which the ring crack occurs. Therefore, the strength and the scatter are controlled by the cluster size and its distribution, and the size depends on the generating crack density. It can be explained by the mechanics model which is expressed by the relationship between strength ratio and crack density based on the energy equilibrium theory.
In order to understand internal fatigue fracture of a shot-peened spring steel, origins of fish-eyes formed on fracture surfaces were observed and discussed. Most origins of fish-eyes were not nonmetallic inclusions, but were microstructural defects, which shapes on fracture surfaces were polygonal facets or narrow bands. From results of hardness tests and microstructure observations, the defects were considered to be slack quenching structures formed flatly along prior austenitic grain boundaries. The fatigue crack initiated at an interface between the defect and the matrix of tempered martensitic structure. When the square roots of area of the defect of the origin √areaDEFECT were smaller than 15μm where the shapes of defects were narrow bands, fatigue life became more than 107cycles and some areas with rough surface called ODA were observed around the origins. On the other hand, when the square roots of area of the defect were bigger than 15μm, where the shapes of defects were polygonal facets, fatigue life became less than 107cycles and ODA were not observed around the origins. The shape of the origin of the fish-eye on the fracture surface were considered to depend on the relationship between the direction of loading axis and the normal direction of the flat defect of the origin.
Epoxy resin is usually used as an encapsulation of IC chips for electronic parts. As for the epoxy resin, a state changes from liquid to solid by the chemical reaction. Hereby, the residual stresses and warp deformation occure in the electronic parts. In this report, the warp deformation for two laminated bodies consisting of epoxy resin and steel, epoxy resin and printed board caused by chemical reaction and thermal load was examined by experiment and theory. In other words, the theoretical content is the thermo-viscoelastic analysis based on the linear viscoelasticity theory and simple prediction solution about the laminated bodies consisted of above-mentioned materials. As a result, it was clarified that the warp deformation for the laminated bodies generated by chemical reaction and thermal load could be predicted by using the combination of thermo-viscoelastic analysis and simple prediction solution which was derived from the bending theory of beam, and that the percentage of the warp deformation caused by chemical reaction was about 15∼20% of the total warp deformation for the laminated bodies.
This study was conducted to investigate damage conditions of a large GFRP tank which was employed in the chemical industry to store mixture of nitric acid and phosphoric acid during 19 years. Further, an attempt was made to establish a damage inspection method based on ultrasonic echo measurement for GFRP products. Detailed optical observations were carried out on the inside surface and cross-section of four typical positions of the tank. The results showed that the acid environment invaded the inside of GFRP through cracks and pin holes made on the inside surface of the tank, and then the environment gave remarkable damage to chopped strand mats existing under a surfacing mat. Degree of the damage was adequately predicted, using ultrasonic echo data. Although the above inspection method could predict decrease in the strength of the roof part which suffered severe damage, further investigation was needed to predict remaining strengths of parts where damage was relatively mild.
Slurry scattering technique has been widely accepted to fix moving sand. By using organic slurry that has high water retention capacity, it seems that the vegetation on the desert may be realized. However, the organic material in the slurry, for example hydrophilic polyurethane, is easily decomposed by ultraviolet irradiation. In addition, the strength of ultraviolet ray is extremely strong on the desert. This paper is to examine the ultraviolet resistance of sand stabilized with organic slurry containing hydrophilic polyurethane in terms of its strength and decomposition rate. Ultraviolet resistance of the stabilized sand is improved by the addition of emulsified asphalt. Although all the asphalt emulsions used in the experiment have similar improving effects, the nonion emulsified asphalt is considered to be one of the best solution due to its neutrality. It is confirmed that the nonion emulsified asphalt is one of the best solution with neutrality. Moreover, the ultraviolet resistance of the stabilized sand can be controlled by changing the mixing ratio of emulsified asphalt. By mixing the emulsified asphalt in the slurry with the ratio of 5%, 6-8%, and 9%, the weight loss rate due to the ultrasonic irradiation was 1.5–1.8times, 2.1–5.4times, and 4.7–18.2times smaller respectively.
Three kinds of microstructural finite element (FE) models are constructed based on the fracture surfaces with different level of porosities. The tensile and shear loadings are applied on three kinds of FE models. The fracture criterion used in the present FE analysis is the ductile fracture locus formulated in the space of the stress triaxiality and the plastic strain to fracture. Based on the FE simulation results, a linear relationship between the material ductility and the area fraction of the defects is tentatively constructed. It is found that the shear ductility decreases at a slightly faster rate than the tensile ductility with the increasing area fraction of the defects. Finally, the effective plastic strain as a function of the stress triaxiality up to fracture at the crack initiation point is compared for three kinds of microstructural FE models under tensile and shear loadings. As a result, it appears that the increase of the area fraction of the defects enhances the evolution of the local stress triaxiality under both shear and tensile loadings. In addition, the local stress triaxiality at the crack initiation point changes in the smaller range under shear loading than under tensile loading. It is found that the difference of the evolution of the local plastic strain and local stress triaxiality should lead the different sensitivity of the shear and tensile ductility to the area fraction of the defects.