Failure analysis on a diesel engine crankshaft that failed due to fatigue was carried out by means of fractography. The value of the applied stress was estimated by the following procedure. First, the crack growth rate was estimated from the percentage of each fracture mode using the reported relationship between them. Then, the stress intensity factor range was estimated from the crack growth rate. Finally, the applied stress was estimated from the stress intensity factor range in a semi-elliptical surface crack. The engine analyzed had been used to drive an emergency electric power supply. Therefore, the total running time was only about 300 hours. The material of the crankshaft was JIS SF55A (normalized carbon steel forged for general use, tensile strength 55kgf/mm2). A fatigue crack formed at the web and grew until the pin finally fell off. A beach-mark was observed on the fractured surface of the web initially damaged. Microfractographs of this surface showed facet-like fracture mixed with intergranular fracture in sequence with fatigue striations. From these observations, the applied stress on the web was estimated to be 8 to 11kgf/mm2. Microfractographs of the second web showed fatigue striations followed by a river pattern. The striation spacing was about 1μm. The stress value was estimated to be 38 to 54kgf/mm2. In the first web, the stress value was more than the web's fatigue limit, about 9kgf/mm2. The failure of the crankshaft was caused by overload during operation.
Zircon powders were prepared by means of a sol-gel method from an equimolar mixed solution of Si(OC2H5)4 and ZrOCl2·8H2O dissolved by hot ethanol, and the crystallite growth and phase change in the powders caused by heat treatment were investigated. The formation mechanism and the thermal-decomposition temperature of zircon powders were discussed on the basis of X-ray diffractometry. The results obtained are as follows: (1) The starting temperature of zircon formation in the as-dried powders without grinding treatment was ca. 1500°C and the rate of formation was very slow. On the other hand, the formation of zircon was accelerated by the grinding treatment of the as-dried powders (the starting temperature was ca. 1200°C), and especially the formation reaction was remarkably promoted in the temperature range of 1300-1450°C with the aid of “Hedvall effect” induced by the crystallization of amorphous silica. (2) It was considered that the promotion effect of zircon formation by grinding was attained by the breakdown of siloxane polymers. The formation of zircon took place via the solid-state reaction of tetragonal ZrO2 and SiO2, and silica changed its phase from amorphous (lower temperature) to transition and then to crystal-like (higher temperature) with heat-treatment temperature. (3) It was pointed out that the thermal decomposition temperature of high purity zircon was higher than that reported in literature (1676°C).
Deformation of inhomogeneous materials such as polycrystals is usually quite complex. The present study attempts to clarify some aspects of deformation behavior during elastic-plastic deformation of polycrystalline materials. A simple plane model of polycrystals was employed, in which three kinds of rectangular and octagonal grains having respective stress-strain relations were assumed to exist regularly. The differences in Young's modulus, yield stress and work-hardening rate were expressed with the parameters of inhomogeneity. The finite element method was used for the numerical calculaion, where the von Mises yield criterion and the Prandtl-Reuss stress-strain increment relation were assumed. The applied load was uniaxial tension. The rotation of grain resulting from the inhomogeneous deformation of surrounding grains and the roughening of free surface were mainly studied, and discussion was made on the influences of the material constants on the rotation and the roughening.
Residual-stress redistribution caused by making a full-depth hole in a plate with biaxial residual stresses was elastically calculated using the Finite Element Method (FEM). The initial residual stresses were introduced into each finite-element through the intermediary of the most suitable eigen-strain distribution in such a way as to accord with the X-ray residual stresses measured prior to the material removal. A new calculation method is also proposed which removes at a time all the finite-elements of the removal-area in the FEM analysis, in contrast to the conventional method in which the elements are removed one by one. The analytical results of residual-stress redistribution were in good agreement with the X-ray residual stresses measured after drilling with an end mill, in the plate with residual stress induced by welding or by water-cooling. These results show that the residual stress change due to the material removal occurs markedly in the vicinity of the removed place.
The mechanisms of damage initiation and growth of random short fiber SMC specimens with a notch during quasi-static loading and the relationship between the damage zone size and the acoustic emission (AE) have been studied. The results obtained are shown as follows: (1) The damage in the notched SMC specimen initiates in the matrix at the tip of notch and it propagates to the fiber dominant area with an increase of load. The damage in the matrix dominant area is observed in a form of crack at the interfaces between resin and particles of calcium carbonate. It is found that the damage in the fiber dominant area is initiated by debonding at the interfaces between fiber and matrix or in fiber strand. (2) The damage zone at a notch tip can be divided into the main-damage and the sub-damage zones. The AE cumulative event counts and the AE total energy correspond to the growth of damage zone. Therefore, it can be concluded that the damage changes from the stable growth to the unstable one according to the AE cumulative event counts or the AE total energy when the sub-damage zone approaches to the critical value. (3) It is clarified that the growth of sub-damage zone is represented by the stress intensity factor KI considering the effect of main-damage zone.
Recently, some welded joints are designed to require a partial joint penetration part for the purpose of optimal design. So, it is very important to evaluate the fatigue crack initiation life at the root of partial joint penetration area. In the present study, the fatigue life prediction for such a welded joint was investigated by paying attention to the crack initiation at the root of partial joint penetration area using compact tension type specimens extracted from the partial joint penetration part of thick plates. The experimental results indicated that the stress ration (R) and the root gap (Gap) did not affect the fatigue crack initiation life of welded joints with partial joint penetration in the range of 0<R≤0.75, 0≤Gap≤1.5mm. It means that the short fatigue cracks growing from the root of partial joint penetration area do not close in the loading range of this study. From an engineering point of view, the fatigue crack initiation life of welded joints with partial joint penetration can be estimated by the linear fracture mechanics analysis based on the effective range of stress intensity factor.
Corrosion-fatigue crack initiation and growth behaviors of two high-strength steels (HT50 and HT80) and a rolled steel for general structure (SS41) were studied by taking a localized corrosion of specimen into account. The tests were carried out in the salt solutions having a concentration from 0wt percent (ion exchanged water) to 3wt percent. The following conclusions were drawn. (1) The average diameter of corrosion pits was almost identical regardless of material and concentration of salt water. But, a wide separation of the corroded area (including many corrosion pits) and the lustrous uncorroded area occurred with an increase in concentration. It is considered that some corrosion pits were joined and led to the early initiation of surface cracks in salt water. (2) The number of surface cracks was remarkably reduced with an increase in concentration of salt water. Localized corrosion along the surface cracks enhanced the growth in salt water. These phenomena influenced the fatigue properties in salt water.
The fatigue crack propagation behavior in fatigue impact at room temperature and 103K was investigated by means of fracture mechanics, X-ray diffraction analysis and fractography for an austenitic stainless steel, SUS 304. The crack growth rate in fatigue impact decreased with decreasing temperature. The crack growth rate at room temperature was scarcely influenced by the microstructure, while at low temperature it was markedly influenced by the microstructure. The effects of microstructure and temperature on the crack growth rate were closely related to the strain-induced martensitic transformation. The martensitic transformation was influenced by the microstructure, the temperature, the fracture morphology and the stress intensity level and resulted in a decrease in crack growth rate with increasing crack opening level.
The effects of ageing at 475°C on the tensile properties and fatigue crack growth behavior were studied for the 35%Cr-5%Ni ferritic steel. The main results obtained are as follows; (1) Both of the 0.2% proof stress and the tensile strength increased considerably with ageing at 475°C. The ductility, on the other hand, decreased remarkably with ageing. (2) The fatigue crack growth resistance increased with ageing at 475°C in the low ΔK range. This is caused by the enhancement of roughness-induced crack closure with ageing. However, the fatigue crack growth resistance decreased with ageing in the high ΔK range. This is due to the change in fatigue crack growth mechanism by ageing. (3) The appearance of the fracture surface of the aged specimens fatigued in the low ΔK range was very different from that of the fractured surface in tension, that is, the former showed the ductile mode and the latter showed the brittle one.
On the blades of gas turbines and blowers, the vibrating stress superposes on the centrifugal stress which is repeated with the start-stop of machine. This combined stress wave is called the intermittent stress wave, and number of studies have been conducted under this stress wave. However, some unclarified problems are left regarding the intermittent stress fatigue. The present report aims at the fatigue life prediction of a notched specimen made of carbon steel (S35C) under the condition of superposed stress cycles 102-104. The following conclusions were obtained from this experimental study. (1) Miner's rule does not predict the fatigue life under the intermittent stress. (2) The experimental life is a little shorter than the life predicted by the modified Miner's rule. (3) The proposed life prediction method may well predict the crack initiation life by using the interaction coefficient C of 0.2-0.4.
In order to investigate the fatigue behavior of corroded steel plates, fatigue tests and surface roughness analysis of the plates were carried out for two kinds of steel: mild steel JIS SS41 and weathering steel JIS SMA50 which were exposed for 7 years at a materialyard or for 25 years at seaside, respectively. The following results were obtained: (1) Fatigue cracking of the corroded steel plates occurred under the conditions of surface roughness with y/ymax≥0.56 and (n/L)max≤1.2. (2) The average angles at the fringe of corrosion pits were 151°30′00″ and 156°49′54″ for the weathering steel and mild steel, respectively. (3) The fatigue strengths calculated basing on the sectional area through corrosion pits nearly coincided with the fatigue test results. (4) In the case of weathering steel, the fatigue strength of the as-rusted surface for 2×106 stress cycles was higher than that of the rust removed surface. On the other hand, the fatigue strength of the rust removed mild steel for 2×106 stress cycles was nearly equal to that of the as-rusted mild steel. This fact may be caused by the adherent rust on weathering steel surface which has an effect of stress relief and cutting off the groundmetal from atmosphere.
In order to make clear the effect of repetition of loading on Al2O3 ceramics, bending fatigue experiments were carried out under pulsation or reversion of loading with the frequencies of 10Hz, 25Hz and 50Hz, and under static loading in dry atmosphere at 288K. The results are summerized as follows; (1) Frequency dependence of fatigue life was evidently seen in the case of repetition. Time to fracture was shorter for high frequency. (2) The fatigue life under reversion of bending was considerably shorter than the fatigue life under the pulsation of bending. (3) For the same maximum stress level, time to fracture under a pulsation of loading was shorter than that of static loading. At the frequency of 50Hz, the fatigue life was about one tenth of the static case.
The influence of notches on the static tensile strength was studied for selected polymeric materials. The tension tests on the grooved shafts have been carried out for a wide range of notch tip radii and depths. The materials used were the following seven kinds of thermoplastics; PC, PVC, PMMA, ABS, Nylon 66, PP and PE. The sharply notched specimens failed in a brittle manner and the bluntly notched specimens failed in a ductile manner for all the materials except PMMA. All the notched specimens of PMMA failed in a brittle manner. The predominant factor for determining the brittle-ductile transition of the fracture mode was the notch tip radius. The experimental results were discussed in terms of a relation between the maximum elastic stress at the notch tip, σmax, and the notch tip radius. In the case of brittle fracture, it has been verified that the σmax at which the specimen failed is governed by the notch tip radius alone and is independent of notch depth for all the specimens. On the basis of the concept of linear notch mechanics, the experimental results mentioned above could be clearly explained and the limiting condition for fracture of the grooved shafts of rigid plastics was expressed. The condition is applicable commonly to seven kinds of thermoplastics.
Ethylenediamine, hexamethylenediamine, ortho-, meta-, para-phenylenediamines and 2, 2-diamino-4, 4, 6, 6-tetraphenoxy-cyclo-triphosphazene were used as the curing agents of an epoxy resin. Mechanical properties of the cured epoxies were determined by the dynamic mechanical measurements and uniaxial tensile tests as a function of temperature. To increase the Young's modulus and the glass transition temperature of the epoxy resin, aromatic amines are more effective than aliphatic amines, and the amines with short distances between amino groups are also effective. The amino-phosphazene investigated in this paper is effective to increase the room temperature Young's modulus. However, it is not effective to increase the glass transition temperature of the epoxy resin, because of the steric hindrance during the curing reaction. The resistances of the epoxy resin to the immersion into water, HCl and NaOH aqueous solutions were also investigated.
SiC fiber and glass fiber reinforced epoxy resins (SiCFRP and GFRP), which were fabricated by using matching male and female metal molds, were immersed in boiling water for the periods from 5 to 500 hours. The amount of water absorption of SiCFRP was less than that of GFRP, and those of SiCFRP and GFRP immersed for 100 hours were 38.5mg/cm3 and 76.6mg/cm3, respectively. The amount of water absorption at the time when the width of SiCFRP specimen began to increase was 9.5mg/cm3 against 44mg/cm3 of that of GFRP. ILSS retention of SiCFRP immersed decreased with increasing immersion time, while the diameter, weight and tensile strength of SiC fiber were not affected at all by immersion in boiling water up to 500 hours. The ILSS retension of 74% for SiCFRP immersed for 100 hours was almost the same as that of 76% for GFRP. The reason of such low durability of SiCFRP to hot water was interpretated as the bond between the fiber and resin was weak to water because the surface of SiC fiber was not treated.
Friction and wear properties of epoxy resin (Epikote 828) cured with hexahydrophthalic anhydride and filled with different (C2F)n powders (C2F), that is, large particles (C2F-L), small particles (C2F-S), and C2F made from natural graphite (C2F-N), in the range of 0-30wt%, were investigated. The composite filled with 85wt% of the mixture of natural graphite and C2F-L by the weight ratio of 9:1 was also used as the specimen (NG: C2F-L-85). It was observed by SEM photographs that the behavior in the friction and wear process of the composites filled with C2F was similar to that of the composites filled with carbon powders. The pV values of the composites filled with C2F increased more than those of the composites filled with carbon powders. NG: C2F-L-85 showed the largest pV value among the composites filled with C2Fs and carbon powders. The friction coefficient (μ) and the specific wear rate (W) of the composites decreased as the content of the C2F powders increased. The μ value of the composite filled with 30wt% of C2F was lower than that of the composite filled with 30wt% of carbon powders, in particular, in the higher range of pV value. The W values of the composites filled with small size C2F, particularly C2F-N, were similar to those of the composites filled with carbon powders. The μ and W values of NG: C2F-L-85 were superior to those of the composite filled with 85wt% of natural graphite.
Teflon-bonded active carbon electrodes were prepared using Pittsburgh active carbon, acetylene black and manganese oxide consisting of Mn5O8 and Mn3O4, and their characteristics as an air electrode in alkaline solutions were discussed in connection with the resistance and activation polarization accompanying with the oxygen-reduction reaction, and the decomposition rate of HO2-ion. As the results, the active carbon electrode containing 20 wt% of acetylene black exhibited a good performance because of its low resistance polarization. By the addition of manganese oxide to the carbon electrode, the activation polarization was reduced, i.e., the magnitude of exchange current density for the electrode containing manganese oxide was about twice as much as that for the original carbon electrode. The voltage-recovery characteristics of the active carbon electrode was largely improved in the presence of manganese oxide. This seemed to be ascribable to the contribution of the catalyst effect of manganese oxide in the decomposition of HO2-ion.
A simple method for the measurement of Young's modulus of small specimens of brittle materials was proposed and applied to measure Young's moduli of teeth, compact bones, ceramics [Al2O3-TiC, PSZ(ZrO2-Y2O3-Al2O3), Al2O3, SiC and PSZ (ZrO2-NbC)] and a carbon amorphous (glass-like carbon). The principle of the method is the three-point bending in which the deflection at a loading point is measured by Martens' mirror equipment system. In this system, a small amount of load ranging from 0.1 to 10N can be exactly applied and the deflection of about 10μm can be accurately measured. A very good linear relationship between load and deflection was obtained and Young's modulus was calculated from this linear relationship. The error of the present method is estimated to be less than 10% by measuring the materials of which Young's moduli are already known from the previous studies.