Features of crack patterns are explored by performing computer simulations of crack growth in the two-dimensional model systems of triangular lattice. The simulations are concerned with crack growth in brittle materials to which the stain energy is supplied nonuniformly. It is shown that the present model gives different crack patterns and strain energy distributions, which are almost independent of the shape of lattice. The results of simulations have some implications in evaluating crack patterns in real materials which were exposed to thermal shock, radiation and so on; the differences in materials and external influences could be conjectured from the characteristic aspects of observed crack patterns.
Fiber reinforced plastic (FRP) materials composed of glass/unsaturated polyester (GF/UP) are used in many fields because of their good properties. In this paper, the fiber reinforced plastic honeycomb sandwich construction made of GF/UP composites was investigated to examine its applicability as structural parts for underground use by utilizing its high flexural rigidity and high corrosion resistance. The creep tests were carried out by dipping the pamples in alkali buffer solution simulating the condition of under-ground water. The creep properties of FRP honeycomb sandwich construction were examined in flexural mode. The fracture appeared in the flexural or shear mode pattern depending on the loading conditions. The experimental results clarified the creep properties of the material and provided an important information about creep design for the application of FRP honeycomb sandwich structure in underground water.
Cyclic loading-unloading experiments at constant temperature and cyclic heating-cooling experiments under constant stress were carried out on TiNi shape memory alloys, and the cyclic characteristics of deformation behavior were discussed. The results are summarized as follows. (1) The transformation lines on the stress-temperature plane moves toward the lower stress region and the higher temperature region under cyclic deformation. (2) The transformation stress and the transformation temperature vary significantly in the early thermomechanical cycles, but take almost the constant values after a cetain number of cycles. (3) There is a similarity of cyclic behavior between the stress-strain curve at constant temperature and the temperature-strain curve under constant stress.
Effects of carbon, nitrogen and tungsten on the tensile and creep properties of a composition-controlled modified 316 stainless steel for high-temperature use, have been investigated. This steel has a lower carbon and higher nitrogen content than conventional 316 stainless steel (SUS316). The results obtained are as follows. (1) Modified 316 exhibits high creep ductility, which changes little with increasing rupture time. It also has a generally higher creep strength than SUS316, and this advantage increases with longer rupture time. (2) The improved creep strength in the modified 316 is attributable mainly to a lower creep rate in the tertiary creep range, and this reduced creep rate is thought to be explicable in terms of stable solution hardening by nitrogen and suppressed grain boundary embrittlement resulting from the low carbon content. (3) Addition of tungsten increases the creep strength of the modified 316 with little change in creep ductility. A high tungsten content, however, is observed to enhance Fe2(Mo, W) formation. It is thought that the resultant reduction in Mo and W contents in the grain matrix lowers the creep strength as the rupture life increases.
In order to examine the effect of compressive creep on crack propagation behavior, creep-fatigue tests were carried out at 823K and 1073K in air under stress waveforms with tensile and compressive holds (cc-type) and with tensile one (cp-type) using notched specimens of Type 304 stainless steel. Tests of smooth specimen were also conducted for the further discussion on fracture morphology. The results obtained are summarized as follows: (1) There is little difference in the crack propagation rate at 923K between the cc- and cp-type tests, and the rate is correlated well with the creep J-integral range regardless of the compressive creep. (2) At 1073K, the cp-type test shows much faster crack propagation rate than the cc-type test for the same value of creep J-integral range. (3) While there are few small cracks inside the specimen near the macro-crack tested in the cc- and the cp-types at 923K and in the cc-type at 1073K, multiple small cracks are found along grain boundaries in the vicinity of the macro-crack in the cp-type at 1073K. The examination of smooth specimen tested reveals that small inner cracks are found only in the cp-type test at 1073K, which corresponds with the fracture morphology in the crack propagation tests. (4) The small cracks weaken the resistance of material to the macro-crack propagation in the cp-type test at 1073K, and thus brings about the faster propagation rate.
Crack propagation behavior was examined in creep-fatigue at high temperatures using center cracked plate (CCP) specimens of an oxide (Y2O3) dispersion strengthened (ODS) superalloy, Inconel MA754, manufactured by mechanical alloying and a Ni-base superalloy without oxide dispersion, Nimonic 75. The Inconel MA754 has coarse elongated grains whereas the Nimonic 75 has fine homogeneous ones. The results obtained are summarized as follows. (1) The transverse crack propagation of the Inconel MA754 at 1273K and 1173K was strongly affected by the microstructure. The crack arrested temporarily at the grain-boundaries perpendicular to the propagation direction while it propagated fast at the midway. (2) The average crack length through the thickness of specimen, which included about 15 grains, was measured by means of the DC potential drop technique during the test of Inconel MA754. The length increased smoothly without the crack arrest and the propagation rate was correlated well with the creep J-integral range. (3) Little difference was observed in the dl/dN-ΔJc relations between the Inconel MA754 and the Nimonic 75 (at 1173K and 1073K). (4) The crack propagation of the Inconel MA754 was much slower than that of the Nimonic 75. This high resistance of ODS superalloy against creep-fatigue crack propagation was caused by the excellent creep deformation property due to the oxide dispersion.
The fatigue crack growth behavior was investigated on hot work die steel SKD62, austenitized at 1025°C and subsequently cooled at various rates, and then tempered at 560°C or 630°C. The fatigue crack growth tests were carried out with a ΔK control servohydraulic test machine, using compact-tension specimens. When the specimens were continuously cooled from 1025°C to room temperature, the fatigue fracture toughness, Kfc, for both 560°C and 630°C tempered specimens gradually decreased with a decrease in cooling rate. The decrease in Kfc correlated well with microstructural change; i.e., the formation of bainite in quenching, and the precipitation of coarse carbides along bainite grain boundaries in tempering, and a fracture morphology change; i.e., the occurrence of a larger quasi-cleavage facet. When the specimens were rapidly cooled from 400°C to room temperature after slow cooling from 1025°C to 400°C, Kfc for a 560°C tempered specimen was the same as that for a specimen rapidly cooled from 1025°C to room temperature. This may be attributed to the suppression of bainite formation due to the rapid cooling below 400°C. However, Kfc for a 630°C tempered specimen was lower than that for a specimen rapidly cooled from 1025°C to room temperature, though higher than that for a specimen slowly cooled from 1025°C to room temperature. The result indicates that Kfc for a 630°C tempered specimen is not only affected by the cooling rate below 400°C but also by the cooling rate above 400°C. The fatigue crack growth rate, da/dN, in what is called region 2 was not affected by cooling rate, but slightly decreased with an increase in tempering temperature. This is closely related to the fact that yield strength is not affected by cooling rate, though affected by tempering temperature. As tempering temperature rises, yield strength decreases and crack closure level rises, consequently decreasing da/dN in region 2.
A crystallographic study has been performed on the growth behaviour of microstructurally small fatigue cracks using coarse grained materials of 3%Si iron. The orientation of grains around cracks was determined by analyzing etch pits developed on the specimen surface, and the correlations between the crack path or crack growth rate and the crystallographic parameters such as slip plane and slip direction were examined. Macroscopic growth direction of transgranular cracks tends to be consistent with the primary or secondary slip directions, or the resultant direction of both. Furthermore, the angle between the crack growth direction and the intersection of primary slip plane and specimen surface in adjacent grain exerts a significant influence on the growth behaviour. As the angle is larger than approximately 40°, cracks can not grow, or can grow with a marked deflection, into the adjacent grain. With increasing the angle, cracks show deep decrease in crack growth rate and/or temporary arrest at the grain boundary.
Fatigue tests were carried out under axial loading by using 2017-T3 Al alloy plate specimens having three defects through thickness to investigate statistically the crack propagation behavior and their coalescence leading to a failure of the specimen. The important findings in the present study are summarized as follows. There was little difference between the number of cycles to crack coalescence and that to a failure. Based on the distribution characteristics of crack length at an arbitrary number of cycles, probabilistic evaluations were made on the coalescence of fatigue cracks, and the distribution of the numbers of cycles to crack coalescence has been theoretically derived on the basis of crack coalescence probability. The analytical results obtained were in good agreement with the experimental trends.
Fatigue in solder joints in thermal cycling is one of the problems in surface mount technology. This paper proposes the evaluation method of fatigue damage (fatigue crack length) in solder joint by the measurement of residual strength (peeling strength) of solder joint. An attempt was made to evaluate quantitatively the fatigue damage in solder joint, paying attention to the residual strength of cracked solder joint. The fatigue damage in solder joints was given by a four-point bending mechanical fatigue test of the substrate, on which some packages were mounted. After the fatigue test, the peeling test of each solder joint and the observation of each fracture surface by a scanning electron microscope were carried out to investigate the relationship between the fatigue damage and the residual strength. On the other hand, numerical analyses were carried out to estimate the residual strength of the cracked solder joint. The elasto-plastic finite element method was employed for these analyses. Furthermore, the effects of solder material, solder volume and package lead shape on the fatigue damage were investigated. Two kinds of solder materials (35Pb63Sn2Ag and 96.5Sn3.5Ag) were used in these solder joints. It was found that the fatigue damage can be evaluated quantitatively by the residual strength of solder joint, and the residual strength can be estimated by the finite element analysis.
The fatigue strength of silver-filler brazed joints in nickel-copper alloys was studied along with tensile strength with emphasis on the influence of filler metal, base metal and joint clearance. The brazed joint with low-melting-point filler metal contained many large defects such as blow holes, and both tensile, so that its fatigue strength was low. The strength was high for high-melting-point filler metals due to a decrease in defects. However, the fatigue strength was not so high as expected from an increase in tensile strength, because the former was more sensitive to defects than the latter. Base metal with higher elastic constants constrained the deformation of filler metel, causing the fatigue strength to increase. The effect of joint clearance on the fatigue properties did not appear clearly, because the effect of defects was much more that of joint clearance.
Since graphite components in the HTTR (=High Temperature Engineering Test Reactor) are subjected to cyclic stress during reactor start-up, operation or shut-down, the design of the reactor requires the knowledge of cumulative fatigue damage of graphite materials. However, the number of reports on the cumulative fatigue damage of graphites are very limited. In this paper, the results of fatigue test in the mixed fatigue mode of cyclic and static stresses were reported on HTTR components graphite, IG-110, regarding the effects of stress hold time and static fatigue-cyclic fatigue interaction on cumulative fatigue damage of the graphite. The following results were derived: (1) Static fatigue failure of IG-110 graphite was observed at the stress level of more than 0.95 times as large as the mean tensile strength. (2) Fatigue life under cyclic stress was longer than that of mixed fatigue mode of cyclic and static stresses at the stress level of more than 0.85 times as large as the mean tensile strength. (3) Fatigue life of mixed fatigue mode decreased with increasing stress hold time. (4) Damage factor of static stress in the mixed fatigue mode was very small, and cumulative fatigue damage in this was smaller than 1.
For the purpose to clarify the temperature dependence above 1000°C of fracture strength properties of silicon nitride ceramics, which are expected to apply to a gas turbine from the viewpoint of high resistances to heat, corrosion and wear as well as light weight, fast-fracture, cyclic fatigue and static fatigue tests were carried out in 3-point bending at 1000°C, 1100°C, 1200°C and 1300°C. On the basis of the unified estimation method, which is proposed here for the evaluation of ceramic strength property, the temperature dependence of fracture strength of sintered Si3N4 was investigated in the temperature range from 1000°C to 1300°C. The results obtained are summarized as follows; (1) The normalized fracture strength data, which were measured by three kinds of high temperature tests, can be commonly expressed by a formula with two statistical parameters (m, σ0), the crack growth property parameter, n, and the activation energy, Q*. (2) The value of n, which is the index in the formula of crack growth rate, decreases with elevating temperature. It becomes more than twice compared with the value below 1000°C. (3) The temperature-dependence of fracture strength can be expressed well in Arrhenius plots. (4) The strength follows the Weibull distribution with two parameters. So, both fracture strength and fracture life at an arbitrary temperature, Tθ, within 1000°C-1300°C, can be probabilistically estimated, if the effective volume, Veff, and the effective hold time, teff, are given.
Crushing load is used as a measure of static load capacity of bearing balls. The purposes of this paper are to make clear experimentally how crushing load of Si3N4 balls for ceramic bearings varies with the ball size, and to interpret the experimental result theoretically. Crushing tests were carried out by employing JIS B 1501 test method, i.e., by compressing two balls of the same size against each other. The ball diameter was changed in the range of 1/8-3/4 inch (3.175-19.05mm). The experiment showed that the relation between crushing load Pc and the ball radius R was expressed as Pc∝ Rλwith λ≅1.8. This implies that the maximum contact pressure at final fracture is not constant but decreases with increase of the ball size. In order to interpret this experimental result, a probabilistic theory of crushing strength of ceramic balls was proposed assuming that the fracture strength of a volume element of a ball follows a two parameter Weibull distribution. The theory leads to the relation of Pc∝ Rλ with λ being smaller than 2, and thus agrees with the experimental result. It was also shown that the crushing strength of the surface-finished ball is almost equal to that of the unfinished ball. This implies that the former can be known from crushing tests of unfinished balls.
Crushing strength is used as a measure of static load capacity of ceramic balls for rolling bearings. As ceramic materials are associated with large scatter in tensile fracture strength, it is natural to anticipate that there is also large scatter in crushing strength of ceramic balls. Thus, the statistical distribution of the crushing load (load at final fracture) of Si3N4 balls for ceramic bearings was investigated in this paper. For this purpose, crushing tests were carried out on a large number of Si3N4 balls with the diameter 3/8 inch (9.525mm), and the crushing load Pc was obtained. According to JIS B 1501, tests were carried out by compressing two balls of the same size against each other. The experiments showed that the crushing load Pc followed a two parameter Weibull distribution approximately, and that its shape parameter was about 13. This result was compared with the probabilistic theory of crushing strength of ceramic balls which was proposed by the present authors in the previous paper. This theory together with the previous experiment on the radius dependence of the crushing load predicts that Pc follows a two parameter Weibull distribution with the shape parameter of about 15, which is in good agreement with the experimental value of about 13. Thus, both of the statistical distribution of Pc and the radius dependence of Pc found by the experiments can be interpretted by the proposed theory unifiedly. Furthermore, the present paper also examined fracture morphology and fracture process in crushing tests in detail.
The effects of added organics on the fluidity of alumina slurry for the doctor blade method and its sintered body were investigated. The homogeneous slurry was prepared by mixing alumina powder, organics, and solvent with a ball mill. The tape casting was performed by the doctor blade method after deairing. Then, the prepared sheets were dried and fired. The viscosity of slurry, and the relative densities of green sheet and sintered body were measured. The results obtained are summarized as follows: (1) The viscosity of alumina slurry increased with increasing organics, and it was possible to obtain the green sheets containing 4.5 wt% to 10 wt% organics by a doctor blade method. (2) The phenomena of shear thinning appeared clearly in the slurry with increasing aging time. (3) The organic solution without alumina powder showed a Newtonian flow. (4) The relative densities of green sheets were gradually increased with increasing organics, however, the packing densities of alumina were little changed. (5) The relative densities of sintered bodies were increased with the higher sintering temperature, while little changed with the organic amount at each sintering temperature.
Dynamic mechanical properties, compressive strength, Vickers hardness, and SEM observation for epoxy resin filled with various graphite fluoride powders of (C2F)n(C2F) and (CF)n(C1F) in the range of 0-30wt% were studied. The graphite fluoride powders were distributed randomly in all the specimens. The glass transition temperature (Tg) and the storage modulus (E') of all the specimens became large with increasing content of graphite fluoride. The effect of filler on the E' was in the order of C2F·N, C1F>C2F·S>C2F·L. The affinity of graphite fluoride powders to epoxy resin was larger than that of carbon powders. The compressive strength (σc) and the Vickers hardness (Hv) were influenced strongly by the morphological factors of graphite fluoride powders and their concentrations. The σc values of the C1F-filled specimens were larger than those of the C2F-filled specimens. The Hv values of all the specimens, especially in the case C1F-filled specimens, increased remarkably.
The effects of three types of polymer modification on both hydration characteristics and physical properties of the gypsum pastes and mortars originated from anhydrite plaster have been studied at various polymer-gypsum ratios. The polymer dispersions used were ethylene-vinyl acetate (EVA), polyacrylic ester and styrene-butadiene rubber. The degree of hydration of the gypsum pastes and mortars was examined by X-ray diffraction, scanning electron microscopy, differential scanning calorimetry and porosity measurement. The setting time of fresh gypsum mortars (polymer-modified and unmodified) as well as the physical properties of hardened gypsum mortars were also measured. The hydration was complete in about 3 days, and the degree of hydration of the polymer-modified gypsum mortars was smaller than that of unmodified mortar. The setting time of gypsum mortars was retarded by the addition of polymer, and was affected by the chemical structure of polymer. The flexural and tensile strengths of the EVA-modified mortar with a polymer-gypsum ratio of 5% were higher than those of unmodified mortar. Its elastic modulus was the smallest of all, and further EVA-modified mortars showed small linear expansion. From the above test results, the EVA-modified gypsum mortar with a polymer-gypsum ratio of 5% was shown to be one of the candidates of the better plastering materials for polystyrene and polyurethane foams.
Polymerization of methyl methacrylate (MMA) initiated by Ce (IV) ion was studied in the presence of hydroxyapatite (HAP) and H2SO4 in aqueous solution at 30°C. The rate of polymerization (Rp) was proportional to [Ce(IV)]1.91. The effect of H2SO4 was also studied in the concentration range from 0 to 7.65×10-3mol/l.The maximum conversion of MMA was given around at 3.6×10-3mol/l of H2SO4. The degree of polymerization of the free homopolymer (PMMA) was 4000-8000 and the efficiency of grafting of PMMA onto HAP was 19-25%.
Fretting friction experiments for pure copper in contact with itself were carried out in air and in argon. Variation in coefficient of friction μ was continuously measured before and after the rapid change in relative humidity RH from 0% to the desired RH levels, namely RH jump. The μ in air at a steady state just after the RH jump increased with jump range ΔRH from the minimum value of ΔRH=0%. The μ in argon was quite higher than that in air but less changed with ΔRH. The physical conditions of rubbing surfaces can be kept constant even after the RH jump in air because of the short testing time. Therefore, the μ in air is affected only by the variation in surface functional group of the rubbing surfaces. The increase in μ with ΔRH can be explained by an adsorption model of oxygen and water molecules on the rubbing surfaces. Variation in μ during fretting without RH jump was also measured at various RH levels. The steady μ in air changed complicatedly against RH and was smaller than the μ with RH jump. Roughness of the rubbing surfaces was varied with RH. Therefore, the variation in physical factors affecting μ should be taken into consideration in the fretting experiments at RH levels without RH jump.
A three-dimensional boundary element method was developed for optimizing the locations and impressed currents of electrodes in a cathodic protection system. The electrodes were regarded as the sources of current, and the potential in the electrolyte was described by the Poisson's equation with the boundary condition, in which the polarization of metal to be protected was taken into account. The Poisson's equation was solved by the boundary element method, and the optimization was performed by minimizing the power necessary to keep the potential on the metal surface below a critical value. An effective method is proposed for obtaining the derivatives of potential, which are needed in the optimizing procedure. In order to demonstrate the usefulness of the method, some example problems are presented.