An attempt has been made to obtain the relation between strength and ductility parameters in tensile deformation of various metals and alloys used in engineering fields. In the tensile strength (σB)-uniform elongation (εU) diagram, the data were classified into several curves depending on crystal structure and stacking fault energy. Materials with low stacking fault energy such as stainless steel SUS 304 have an excellent σB-εU relation, while hop materials such as zinc and magnesium have low σB and σU compared with the standard curve obtained for ferrite steels, copper and aluminium. The relation between yield ratio (σY/σB) and εU for all materials investigated in the present study is roughly summarized by a single curve, although the curve does not exactly agree with the theoretical curve generated under the assumption that the stress-strain relation obeys the n-th power law.
Impact tensile fatigue tests were performed on smooth specimens of 0.21% C steel in order to investigate the effect of loading wave shape on the deformation and failure. The shape of impact loading wave was characterised by R-value defined as the ratio of the minimum nominal stress to the maximum one in the damping curve of one cycle. R-values employed in the present experiment were -0.17, -0.44, and -0.68. The main results obtained were as follows: (1) The process of the change in residual strain with the number of cycles was found to be divided into two stages, i.e., macrocrack nucleation and its propagation. The residual strain increased with increasing R-value. The strain at fracture for the case of R=-0.17 was larger than that under static tension, while that for the other cases of R-value was smaller. (2) The compressive residual stress was detected by the X-ray method on the surface of the specimen during impact tensile fatigue. Its absolute value was higher than that induced in static tension, when the residual strains were the same. (3) The following equation was found between the fatigue life Nf and the increment of residual strain per one cycle ΔεRm. ΔεRm·Nfn=C The exponent n was independent of R-value, while C varied with R-value.
A strain controlled low-cycle fatigue test was conducted on an austenitic stainless steel SUS 321 at 500 to 700°C under a fully reversed triangular wave form strain with a strain rate of 10-4 to 10-2 sec-1, to obtain fundamental information on the effects of temperature and strain rate on the low-cycle fatigue behavior of austenitic stainless steels at high temperatures. The test results were examined by observation of fracture surfaces and discussed from view-points of deformation mechanism and fracture morphologies by considering cyclic strain induced precipitation at high temperatures. The main results of the study are as follows: (1) The fatigue life is reduced with decreasing strain rate at a constant temperature and with increasing temperature at a constant strain rate. (2) At 500 and 550°C, the fatigue life Nf can be well correlated with the strain aging parameter, (T/ε)exp(-Q/RT), while not at 650° and 700°C. (3) The fatigue strength and cyclic strain hardening properties of the material are affected by dynamic strain aging, by the precipitation of Ti(C, N), M23C6 and a σ-phase and by creep deformation. At 550°C, dynamic strain aging plays an important role on the reduction of fatigue life caused by the decrease in strain rate, whereas at 650°C the fatigue life is reduced mainly by creep deformation. (4) The failure mode of the specimens tends to change from transgranular type to an intergranular as the temperature becomes high, the strain range large, or the strain rate slow.
Intermittent loading creep tests were carried out on SUS 304 austenitic steel over a broad stress range, and its creep deformation, life and rupture ductility under intermittent loading were compared with those data under constant loading. The creep deformation (minimum creep rate) under intermittent loading decelerated at 600°C and at 700°C as compared with that under constant loading except the case of high stress at 600°C, and its value approached gradually to zero with decreasing stress. In these cases, the fracture mode showed an intergranular nature, and the rupture ductility for intermittent loading was lower than that for constant loading. In the case of high stress at 600°C, the transgranular mode of fracture was observed and the ductility for the both types of loading showed a similar trend. The dependence of life (tR) on the minimum creep rate (εmin) under intermittent and constant loading was shown by the relation logtR/εcp∝logεmin where εcp is the total creep strain.
The dependency of stress intensity factor range, ΔK, on crack growth rate, da/dN, at the early stage, which initiated from a small notch with notch length of 50μm or 200μm and the notch root radius of about 25μm, was examined using a high strength aluminum alloy. The effect of notch length on crack propagation cycles was discussed by integration of da/dN -ΔK curves. The results are as follows. (1) The da/dN of a high strength aluminum alloy can be expressed by the power law as a function of ΔK. The da/dN-ΔK curves can be separated into three stages. (2) In the stage 1 (20>ΔK>8.7kg/mm3/2), the da/dN is discontinuous due to the formation of non-propagating cracks. (3) There is a well defined minimum apparent stress intensity factor for crack propagation from a notch. For R=0.1, this factor is constant regardless of the length of the notch. (4) The apparent crack length for non-propagating cracks for the smooth specimen is 32μm under R=0.1. (5) When the length of a crack from the notch tips, aP, is almost equal to or larger than the notch length, aN, the number of cycles required for a crack to propagate to aP, NP, increases with aN under constant ΔKap and ρ, where ΔKap is the apparent stress intensity factor range and ρ is the notch root radius.
A study was made on the effects of stress increasing time or stress rise time T1, maximum stress holding time T2, stress decreasing time T3 and minimum stress holding time T4 in a wave on fatigue crack growth for a low alloy carbon steel (σu=91.8kg/mm2, σy=81.5kg/mm2) in 3% NaCl solution. Measurements of the effective stress intensity range ratio U and observation of crack tip were performed to clarify the cause of waveform effect. The results were summarized as follows. (1) T1 had a strong accelerating effect due to corrosive dissolution of fresh surface of the crack which was formed during T1. The crack growth rate was enhanced as T1 increased and reached to a constant value (about 3 times than that in air) after T1=10sec. The crack growth rate at low ΔK, however, decreased as T1 increased (more than T1=1sec). (2) T2, T3 and T4 decreased the crack growth rate, respectively. The extent of decrease not only depended on the period of T2 (or T3, T4), but also on ΔK and T1. (3) The crack growth law which was derived by considering the effects of waveform and frequencies in the previous paper, is also valid to first approximation for the present results.
The third order elastic constants of polycrystalline aluminum (purity of 99.79%) and copper (purity of 99.98%) were determined by measuring their ultrasonic velocities under uniaxialy applied stress and the results were discussed by comparing with the previous data. The travel time of ultrasonic waves (5MHz) in a specimen of 10mm thickness was measured precisely with a sing-around system. The longitudinal and shear waves polarized parallel or perpendicular to the uniaxial stress axis were propagated perpendicular to the axis. Since the traveling time under zero stress varied slightly in the experiment, the traveling time was measured before and after periodical stress was applied and released rapidly. Elastic strains were below 200×10-6 in tension and compression and no plastic strains were observed. The results are summarized as follows: (1) The traveling time did not change linearly with applied stress. The deviation toward longer travel times was very large for copper both in tension and compression. However, the deviation for aluminum was a little. This is inferred to be the dislocation contribution. (2) To separate the measured traveling time change into two parts, one caused by the lattice anharmonicity and the other by dislocations, the measurements in both tensile and compressive conditions are needed. (3) The values of third order elastic constants measured for aluminum were in good agreement with the previous data, but not for copper. The present values, however, were most close to those calculated from the data of single crystal without containing dislocation contributions.
In the previous paper, the effects of hardness of samples and feed weight on the grinding rate constants were studied. In the present paper, the effect of feed size on the rate constant of undersize increase is examined. The samples used were silica glass, borosilicate glass, soda glass, quartz, feldspar, limestone, marble, gypsum and talc. The pebble mill consisted of a 12.8cm-diameter by 13.2cm-long cylinder and 40 balls of 3cm diameter having the total weight of 1180g. The tests were done at the speed of 88.1% of critical one with several feed size ranges of 14∼20, 20∼28, 28∼35, 35∼48 and 48∼65 mesh. As the result, the following equation was obtained. Kx=0.024HV-0.47·(WS/WB)-1.5·(x/xf)·(xf/x0)0.7 where Kx is the zero order increasing rate constant of weight fraction less than size x, HV, the dimensionless Vickers hardness, WS, the feed weight of sample, WB, the ball weight, xf, the feed size, and x0, the optimum feed size.
In recent years, particle dispersed composite materials are widely used as structural materials like fiber reinforced composites. These materials with brittle fracture behavior are generally characterized by a large amount of scatter in the results of fatigue life tests, so that the reliability concept is required for prediction of their fatigue lives. The main results obtained in this work are summarized as follows; (1) The fatigue life distribution of particulate composites consisting of matrix and n kinds of particles and voids can be obtained with the assumptions that the weakest link theory can be applied to the specimen composed of an aggregate of cross sectional elements in series, and that the Weibull distribution is adoptable for the distribution of fatigue life of the element in composite materials. (2) The roles of each element played in the fatigue life of particulate composites were well-defined. This study also clarified all of the S-N curves for matrix and each of the particles as well as the shape parameters of each element in Weibull distribution which determines the amount of scatter.
The critical strains εc for crazing of polyarylate, polycarbonate, and polysulfone in a variety of organic liquids have been measured by means of three-point bending jig. The εc of these polymers were small in methyl ethyl ketone, carbon tetrachloride, and toluene. In linear alcohols εc decreased in the order of methyl, ethyl, propyl, and butyl alcohol. In the case of polyarylate, the correlation of εc with solubility parameter of the crazing liquid was poor. The εc were rather closely correlated with the degree of swelling by crazing liquid. On the basis of these results, the solvent crazing of polyarylate could be explained on the plasticization mechanism proposed by Kambour.
In spite of many works on the initiation and growth of crazes, few studies on plastic deformation caused by crazing inherent in high polymer solids have been performed. In this paper, as the first step to describe plastic deformation in high polymer quantitatively, the craze density, i.e., the number of crazes per unit surface area, and the craze growth rate were investigated, and their statistical nature were examined. The experiments were performed under the action of a crazing agent, kerosene, at room temperature (20±1°C) with a constant load type testing machine on the specimens of polycarbonate and polyvinylchloride. The dependences of craze initiation and growth on applied stress and testing time were determined by photographs taken through an optical microscope without stopping the test. The main results were: (1) the variation of craze density with time can be described successfully by a simple rate theory, (2) the distribution of the craze growth rate, which has essentially a statistical nature, is approximated by a normal distribution function and (3) the dispersion of craze length at a certain time can be explained by a statistical theory which takes into account both the distribution function of growth rate and the density variation of crazes with time.
The so-called NOL ring test has been usually used as a testing method to measure the longitudinal tensile properties of filament-wound composites; however, it has a disadvantage to accompany an appreciable bending moment occurring at the split in the dee-fixture where the fracture takes place. The authors have proposed an improved ring having a form of race track (RT) with straight parts, as Dow noted prior to us. The tensile tests have been carried out to compare both types of ring specimens of carbon fiber reinforced plastics. They show a remarkable bending effect and a decrease of tensile strength with an increase in thickness. On the other hand, the analytical stress distributions in both types of ring specimens under separation of split-dee fixture are obtained by the finite element method. The correction by the numerical results leads to the nearly equal tensile strength and modulus, and the RT ring specimens are shown to be desirable to have the reliable tensile data with a substantial reduction in bending moment. Furthermore, the nonlinear analysis is done by taking into consideration the changes of ring configuration and of contact location with split-dee during elongation, which explains the nonlinear relation between load and bending strains.
There are many cases that the alternating voltage applied on the practical electrical machines contains not only constant amplitude sinusoidal waves but also distorted waves or high frequency-superimposed waves. Therefore, the real appraisal of arc resistance of insulating materials for such applications is difficult to make on the basis of arc resistance determined by the conventional test under sinusoidal wave voltage. In this paper, the arc resistance under distorted wave voltage and high frequency-superimposed wave voltage were investigated. The test method was similar to the high voltage low current arc resistance test like the ASTM method. However, the test method using the triac was tried in order to search for a reliable arc resistance test which incorporates the merits of the electrical static type interrupter of circuit and compensates the weakness of the mechanical intermittent interrupter. The applied voltage, arc current and arc intermittent time of the fundamental waves for the high voltage low current arc test were the same as used in the ASTM method. The electrode type was the rod type and the electrode gap length was 6.4mm. To generate distorted wave voltage and high frequency-superimposed one, the voltage from the oscilator was amplified separately as in Fig. 2 and was superimposed to the fundamental sinusoidal wave at the secondary side of the transformer. The results of the tests show that the arc resistance decreased with increasing frequency of the superimposed wave and also with increasing crest value of the waves at a constant frequency, when the high frequency voltage was superimposed and its crest value was about 4-20% of the fundamental sinusoidal wave voltage. Furthermore, the above arc resistance was lower than the one at the smooth sinusoidal wave voltage with the same maximum and minimum voltages.