Tension-compression fatigue tests were carried out by using high strength steels with different strength levels and the endurance limit diagrams were obtained. The main results are as follows: (1) The shape of the endurance limit curves of various high strength steels differed largely depending on their strength level. It also differed when the strength level was varied by changing tempering temperature for the same material. (2) The endurance limit curves of the steels with tensile strength higher than 150kg/mm2 abruptly decreased with the increase of mean stress in the region from the completely reversed tension-compression to the pulsating tension, and when the mean stress was further increased and exceeded a certain value, the curve became flat at low stress amplitude level, and was independent of mean stress. (3) The endurance limit curves of the steels with tensile strength lower than 150kg/mm2 showed the same tendency as those of the ordinary carbon steels.
Rotating bending fatigue tests were carried out, using the specimens cut out from a rolled steel with remarkable laminated structures in three different directions-rolling, thickness and 45°. The fatigue process on the specimen surface was observed successively by an optical microscope, and then the fracture surface of the same position was observed by a scanning electron microscope. The main results obtained are as follows: (1) The anisotropy of the fracture life is controlled by the difference in the process of crack initiation and propagation up to about 0.5mm. (2) Cracks in the rolling directional specimen originate at grain boundaries or the inside of grains in ferrite layers, while those in the thickness directional specimen originate at boundaries or near-boundaries between ferrite and pearlite in most cases, and the initial fatigue fracture surfaces are almost parallel to the maximum shearing stress plane. (3) The origin of fracture in the material with a large grain size is not so clear as in the case of the material with a small grain size.
Cavitation erosion tests for rimmed and killed steels were performed using a magnetostrictive vibratory facility. In ion exchanged water, the ratios of erosion resistance of Al-killed and Si-killed steels to that of rimmed steel for the annealed specimens were 1.16 and 1.33, respectively. Thus, erosion resistance is improved by deoxidized treatment for the both annealed and quenched specimens, even though the carbon content and crystal structure of steel are the same. In 3% salt water, the weight loss rates of Al-killed and Si-killed steels increased about 10%, but that of rimmed steel did not increase compared with those in ion exchanged water. These behaviors were discussed through the measurements of erosion particle size and/or number and length of cracks observed on the cross section of specimen. Furthermore, erosion resistance was found to show a better correlation with σB2/E of HV2/E.
Powdered resins were prepared from softwood or hardwood spent sulfite liquor (SSL) by acidifying and spray-drying, and they were used as binders for particleboard. The strength properties and dimensional stability of the board were evaluated. The results obtained were as follows. (1) The particleboard with SSL powder showed good dimensional stability, but low modulus of rupture (MOR) and low internal bond (IB). (2) The effects of spray-drying temperature of SSL on the properties of the board were little in the range from 120°C to 150°C. (3) The board with softwood SSL powder was better in both strength properties and dimensional stability than that with hardwood SSL powder.
The source location of acoustic emission (AE) events was performed during the creep of Ohshima granite under uniaxial compression. During the stage of loading up to the creep stress, AE events were randomly distributed throughout the sample. As soon as the transient creep was initiated, abrupt migration and clustering of AE events into several near-surface zones were clearly observed. These migration and clustering strongly suggest the rapid anisotropic development of dilatancy at the very beginning of transient creep. These clusters, though the shape of each one was“volumetric”, formed a shear band at an angle 30∼35° to its loading axis. The spatial distribution pattern of AE events established in this stage was retained until the main fracture. By the end of transient creep. AE events began to concentrate into one of these clusters; on the other hand, the activities of other clusters were reduced gradually. This change was spread broadly and continuously in time until the main fracture. There was no fundamental difference in the spatial distribution pattern of AE events between the stationary and the tertiary creep. In the latter half of the creep, the occurrence of AE events was limited within only one cluster which remained active, and the acceleration of AE activity was seen before the main fracture in this cluster. And its shape was spheroidal whose long-axis was parallel to the loading axis.
Surface strain mapping of Ohshima granite during the creep experiment shows that a uniformity of loading up to the creep setting stress was attained, and that a large change in strain field occurred during the early stage of transient creep. The change of the strain field represented the anisotropic development of dilatancy immediately after the creep initiation. The pattern of strain distribution established in this stage was rarely changed in the subsequent course of the creep. The accelerating increase noted in one of the circumferential strain gages during the tertiary creep documents a strong localized deformation before the main fracture. The clustering region of AE events during the tertiary creep, which was determined by the simultaneous measurement presented in the other paper, was close to the region where the strong localized deformation was observed in this surface strain mapping. The projection of AE hypocenters to the specimen surface revealed that the hypocenters of AE events during the creep were mostly located between the strain gages. This biasing was explained on the basis of the role of water vapor in stress corrosion cracking. The adhesion of strain gage prevents the transport of water vapor from the surface of specimen to the crack tip.
The internal stress of epoxide resin coatings cured with aliphatic α, ω-dicarboxylic acids, HOOC(CH2)pCOOH (p=2, 4, 7, 8, 10) in the presence or absence of tertiary amine accelerator and those cured catalytically has been investigated by measuring the changes of sample size and internal stress in the curing and cooling processes. Three types of network structures based on ester bonds, ether bonds or ester-and-ether bonds were formed in these cured systems. The internal stress was caused by shrinkage occurring in the cooling process from the glass-transition temperature to room temperature in all systems. This stress increased with elevating glass-transition temperature in the cured systems, and with decreasing methylene chain length of the curing agents. From the above observations, it was suggested that the lowering of glass-transition temperature was necessary to reduce the internal stress and shrinkage caused by curing.
In the previous paper, the temperature dependence of strength in the flatwise direction of roving glass cloth FRP was reported for the case where the weakest interface is the fiber/matrix interface irrespective of the fiber/matrix ratio and the condition of heat treatment. The present paper reports the case of roving glass cloth FRP having the fiber/matrix ratio of 100/0 and 0/100, when the weakest interface is not affected by the condition of heat treatment. The results obtained are as follows. (1) The temperature dependence of the brittle or ductile behavior of the interface is affected by the thermal properties of the matrix which depend on the condition of heat treatment, but not by the location of the weakest interface. (2) The temperature dependence of the location of the weakest interface is affected by the fiber/matrix ratio, but not by the condition of heat treatment. (3) The temperature dependence of the bond strength is not affected by the fiber/matrix ratio nor by the condition of heat treatment, but it changes in the specified temperature range.
Mechanical properties of steam turbine casing and rotor materials serviced for 13 years were investigated. The results obtained are summarized as follows. (1) The creep rupture strength of the materials used in steam at 538°C showed a little decrease. (2) The value of KIc of the rotor material used at 480°C was 65% of that of the same material used at 70°C. (3) Embrittlement during service occurred at 400∼522°C, but not below 300°C. It is considered that the main factor of embrittlement is the segregation of impurities such as phosphor to the grain boundary. (4) The FATT of embrittled materials was recovered by the de-embrittle treatment at 650°C.
The fatigue crack growth behavior and its mechanism have been investigated for the particulate filled polyvinyl chlorides (PVC). The results obtained are as follows; (1) The fatigue crack growth rate of PVC filled with CCR or MBS particles is much influenced by the kind of stabilizers at the same quantity of particles and the same stress intensity factor range, comparing with the effect of the kind of particles on the crack growth rate. (2) The fatigue crack initiates due to the debonding between matrix and CCR particles or from voids which occurred among aggregate particles of CCR in PVC filled with CCR particles. In PVC filled with MBS particles, it initiates due to the separation of aggregate particles of MBS. (3) The micromechanism of fatigue crack growth in the materials can be guessed from micromorphology of the fracture surfaces.
The rotating bending and the plane bending fatigue tests were carried out on specimens having a small blind hole (diameter: 0.3mm) and smooth specimens of an annealed 0.42%C steel. The difference in fatigue strength between these two types of loading was investigated by the successive observations of crack initiation and crack propagation processes. The results obtained are summarized as follows: (1) The number of cycles to failure in the plane bending test was larger than that in the rotaing bending test under the same stress. (2) There was little difference in fatigue limit and in length of non-propagating cracks between two types of loading. (3) The crack growth rate in the rotating bending test was faster than that in the plane bending test under the same stress and crack length. On the other hand, the crack initiation lives were roughly equal each other. Therefore, the difference in fatigue strength between two types of loading results from the difference of their crack propagation processes.
The effect of precipitation on the resistance to fatigue crack growth was investigated in rotary bending- and cyclic torsional fatigue tests, using a precipitation hardening austenitic steel, SUH 38. Matrix precipitation was found to have no marked effect on the resistance to crack growth in rotary bending fatigue, while in cyclic torsional fatigue the resistance increased considerably with an increase in proof stress or tensile strength due to matrix precipitation. It was also found that coarse grain boundary precipitation decreased the crack growth resistance of a specimen without matrix precipitation, such as an as-solution treated specimen in both fatigue tests, but this effect of coarse grain boundary precipitation disappeared remarkably with matrix precipitation. These results are discussed in terms of Weertman's equation. It is shown that matrix precipitation which increases static strength can improve the resistance to crack growth only in such fatigue having the markedly low level of triaxiality of the stress state, as cyclic torsional fatigue, and that coarse grain boundary precipitation may have no effect on the improvement in resistance to fatigue crack growth.
Fatigue crack propagation tests of mild steel (JIS SM41B) and high strength steel (JIS SNCM439) tempered at 600°C and 200°C were conducted under stress ratios of 0.1 and 0.5. The distribution of the half-value breadth of X-ray diffraction profiles beneath the fracture surface was measured. In mild steel and 600°C tempered steel, the half-value breadth measured near the fracture surface was larger than the initial value measured before the fatigue test. On the other hand, the half-value breadth decreased near the fracture surface in 200°C tempered steel. From the distribution, the depth of the maximum plastic zone ω was determined. It is related to the maximum stress intensity factor Kmax by the relation: ω=α(Kmax/σY)2(σY=the monotonic yield strength), where α=0.09 for mild steel and 0.22 for tempered steels. The difference in α between the materials was used to evaluate the actual yield strength of the material in the plastic zone around the fatigue crack tip.
Fatigue tests under rotating bending and reversed torsion were carried out in air, distilled water and 3 percent saltwater, using smooth specimens of high-strength low alloy steel (Cr-Mo steel). The initiation and growth behavior of small fatigue cracks in each environment were evaluated based on detailed observations, and the effect of corrosive environment was also discussed. The fatigue strength decreased with increasing aggressiveness of test environment. This decrease in corrosive environment was due to the earlier fatigue crack initiation. From the observed locations at which small fatigue cracks began, it was considered that the crack initiation was primarily governed by hydrogen embrittlement in distilled water and also affected by corrosive dissolution in 3 percent saltwater. The validity of the application of linear fracture mechanics for small fatigue cracks was established. The growth rate of small fatigue cracks was higher than that of large through cracks, but not accelerated by the corrosive environment. Moreover, the fatigue life in the corrosive environment was estimated by using the crack growth characteristics in air.
In order to investigate the behavior of elevated temperature low cycle fatigue crack propagation in Type 304 stainless and 2.5Ni-0.5Mo-0.1V ferritic steels, load-controlled low-cycle fatigue tests were carried out on notched specimens under four kinds of loading waveform (i, e., PP-, PC-, CP- and CC-types). The results obtained were summarized as follows: (1) Type 304 stainless steel The crack propagation rate, dl/dN, was correlated well with the fatigue J-integral range, ΔJf, in P-type (i, e., PP- and PC- types) fatigue tested at 973K and 823K. The striations, of which width was found to be nearly equal to dl/dN, were observed on their fracture surfaces. On the other hand, dl/dN in C-type (i. e., CC- and CP-types) fatigue tested at 973K was correlated well with the creep J-integral range, ΔJc. The fracture mode in this case was of intergranular type. However, good correlation of dl/dN with ΔJc as well as ΔJf wasn't obtained in C-type fatigue at 823K, where the fracture mode was of mixed (i, e., intergranular and transgranular) type. This appeared to be caused by creep-fatigue interaction. (2) Ni-Mo-V steel In the same way of the above stainless steel, dl/dN in P- and C-type fatigue at 823K were correlated well with ΔJf and ΔJc, respectively. However, there was no difference in fracture mode between these two types, both of which exhibited the transgranular fracture mode.
The mechanochemical reaction of SBR (Styrene-Butadiene Rubber) was investigated at room temperature up to 10kbar static pressure by using a self-made high pressure apparatus capable of generating simultaneous shear deformation. The shearing curve at high static pressure rapidly reached an equilibrium value. The shearing curve at low static pressure, however, slowly approached it after the gradual increase of shear strength. The maximum shear strength at high static pressure was observed on SBR with 23wt% styrene content which corresponds to widely available SBR. The mechanochemical gelation of SBR was influenced by applied static pressure and was probably related to the occurrence of the cross-linking in styrene region.
A suitable surveillance test program was developed for evaluating the change in fracture toughness of material after long service operation using a retired petroleum reactor pressure vessel. The surveillance program is based on the correlation derived in advance from the standard Charpy absorbed energy (CV) and fracture toughness (KIc, KId) data for pre-service material and on the Charpy energy (CV) obtained from the periodic surveillance after service operation. It is here demonstrated that KId(CV) estimated from the Charpy data shows a good agreement with valid KId determined experimentally for all the base metal, weld deposit and heat affectted zone (HAZ) of the material. The validity of the surveillance program was also discussed.
This paper presents a mechanical model of paper showing nonlinear stress-strain relation and its strain rate dependency, which is applicable both to tension and creep of paper. To clarify the mechanical property of paper, at first the model was applied on the experimental results of metal. Then, the property of paper was measured and the difference between paper and metal was obtained. Since the mechanical property of paper is very sensitive to moisture content, the following relation between Young's modulus, E and the moisture content, w was also considered. E=E0e-c1w. The experimental results obtained were as follows. (1) The stress of paper, σ, is expressible as a function of strain γ and strain rate γ by, σ=Fγnγm (2) The comparison in physical property between paper and metal shows that both n and m of paper are larger than those of metal, while F and E of metal are larger than those of paper. The present model and the physical property data of paper were successfully applied to the machine design of an actual high-speed printer.