The fatigue crack growth tests of WC-Co Cemented Carbides were carried out in a wide range of fatigue crack growth rate covering the threshold stress intensity factor range ΔKth. The effects of the stress ratio, Co volume fraction and the phase transformation of Co on the fatigue crack growth characteristics were investigated on the basis of fracture mechanics and fractography. The crack growth rate was measured using 3-point bending specimens. Crack growth tests were carried out at 10Hz and the stress ratio R=0.1 and 0.5. The main results obtained are as follows; (1) The Paris rule can be applied between da/dN and ΔK and it is shown to be da/dN=C(ΔK)m. (2) The fractography of the fracture surface, shows that brittle fracture occurs in the Co binder phase at the stress rate R=0.1 but ductile fracture occurs in the Co binder phase at R=0.5. (3) The relation of da/dN-ΔKeff/E of WC-Co cemented carbides shows the characteristic in the middle of the other metals and the alumina ceramics. (4) The Co binder phase undergoes phase transformation by repeated deformation and so it affects the characteristics delicately.
Fatigue tests were carried out under constant stress amplitude, using a non-magnetic high manganese Mn-Cr steel. X-ray fractography was applied on the fatigue fractured surface to investigate the relationship between stress intensity factor and residual stress or half-value breadth of the X-ray diffraction profile. The fatigue crack propagation rate of this non-magnetic Mn-Cr steel had the same tendency as in the ordinary structural ferritic steels. The relationship between stress intensity factor and the residual stress or half-value breadth of the steel was almost the same as that of the ferritic cyclic work hardening steels. No stress induced transformation was observed on the fracture surface, but the residual stress on the fractured surface was compressive in the high stress intensity factors range, which is typical in the cyclic work hardening steels. The half-value breadth on the fractured surface increased with increasing effective stress intensity factor range. The relationship between the half-value breadth and stress intensity factor range was represented by a linear line regardless of the stress ratio. Therefore, the acting stress intensity factor range at the time of fracture can be estimated from the half-value breadth. The depth of monotonic plastic zone was estimated from the distribution of half-value breadth beneath the fractured surface. The relationship between the maximum stress intensity factor and half-value breadth was expressed by the equation ωm=α(Kmax/σy)2, where the value of α was 0.025. This is about one sixth of the value for ferritic steels, and the fact shows the severe work hardening occuring in the plastic zone in this manganese steel.
There are several investigations on the behavior of residual stress during fatigue process. It is, however, difficult to detect the crack initiation from the change in residual stress because of no significant change in residual stress through the fatigue process. The purpose of this investigation is to detect the fatigue damage from the change in actual stress measured by X-ray diffraction technique during fatigue test. In this paper, a cantilever type rotary bending test machine was developed for an X-ray stress analysis based on Single Exposure Technique. The actual stress at the maximum tensile applied stress was measured to failure during cyclic loading. The actual stress was keeping stable with increasing number of cycles until crack initiation, and this behaivior was independent of stress amplitude. After the crack initiation, it decreased suddenly with crack propagation because of stress relaxation at the crack. These results show that the change in actual stress at the maximum tensile applied stress has much higher sensitivity to detect crack initiation than one in residual stress. In-situ X-ray observation of actual stress during fatigue test is useful to detect fatigue damage.
The cystallographic texture and elastic constants have been measured on heat-treated Zr-2.5%Nb alloy pressure tubes by neutron diffraction. The texture of heat-treated Zr-2.5%Nb alloy is not so strongly oriented compared with that of non-heat-treated Zr-Nb alloy. However, the pole figures show that the appropriate diffraction planes for the axial direction strain are 10·0, 00·2 or 10·1 diffractions, those for the hoop direction strain are 00·2, 10·1 or 11·0 diffractions, and those for the radial direction strain are 11·0 or 10·2 diffractions. The diffraction plane dependency of elastic constants almost agrees with that obtained from Kröner's single cristal model. However, the diffracttion plane dependency of elastic constants is not so strong. From the residual stress measurements by the neutron diffraction, the values of Young's modulus for 10·0, 10·1 and 00·2 diffractions are found to be 96GPa, 95GPa and 104GPa, respectively, and Poisson's ratios are 0.34, 0.36 and 0.32, respectively. Finally, the residual stress evaluation method for the material with strongly oriented texture is proposed.
The investigation on the residual stress of titanium carbide films in terms of the X-ray diffraction method is our purpose. However, it is hard to measure the stress by the commonly used X-ray stress measurement, because the sin2ψ method demands macroscopic isotropy from the specimen but the ‹110› orientation is observed for our evaporated TiC film by CVD. Therefore, in this paper, the X-ray stress measurement for ‹110› oriented films was formulated by introducing the weighted average method for the crystallite orientation distribution. The formulation derived showed that the relation between the stress of the specimen and the measured strain by X-ray diffraction depended on the measured diffraction planes. Then, the stress calculation was made and discussed based on the loading experiments.
The X-ray stress analysis was examined for an arc-melted Ni3Al intermetallic compound, which consists of coarse grains and exhibits strongly preferred orientation. The imaging plate was used as an X-ray detector, and continuous Debye-Scherrer rings were obtained from such coarse grained material by applying an X-Y plane oscillation method. Mechanical bending stress was applied to the specimen for measuring the X-ray elastic constants and stress constant. An ideal orientation method was adopted to obtain these parameters. It uses the diffraction lines which belong to the same zone axis. These experimental elastic constants were compared with the theoretical values calculated from the elastic compliance of Ni3Al single crystal. The results obtained are summarized as follows; (1) The X-ray elastic constants measured from Ni3Al 220 diffractions which appeared at ψ=0.537° and 60.537°, were EX/(1+νX)=176GPa, EX=220GPa and νX=0.253. (2) The theoretical Young's modulus E0 at ψ=0° was calculated as 207GPa from the elastic compliance of Ni3Al single crystal. This value means the elastic constant in the direction of the tensile axis for the grains which contribute to the diffraction. Young's modulus E60 at ψ=60° was similarly obtained as 228GPa. The experimental value of X-ray elastic constant EX (220GPa) was intermediate between the theoretical values.
The high velocity oxygen-fuel (HVOF) system in air has been established for producing various coatings that are extremely clean and dense. It is thought that the HVOF sprayed MCrAlY (M is Ni, Co or Fe) coating can be applied to protect the oxidation and corrosion in the hot section of gas turbines. Also, it is well known that thicker coating can be sprayed by this system for improving the residual stress in comparison with any other thermal spraying system. However, the mechanical properties of HVOF coating have not always been clarified. In this paper, the mechanical properties were measured in order to compare with a low pressure plasma sprayed MCrAlY coating. In case of the as sprayed coating, the Young's modulus and bending strength of HVOF MCrAlY coating were inferior in comparison with those the low pressure plasma sprayed MCrAlY coating. However, the experimental results suggest that the strength of HVOF MCrAlY coating can be improved by a diffusion heat treatment. These phenomena may be attributed to the fact that the HVOF MCrAlY coating is built up by semi-fused sprayed particles.
A constitutive model for describing the inelastic deformation of unidirectional and symmetric angle-ply CFRP (Carbon Fiber Reinforced Plastics) laminates is developed. The kinematic hardening creep flow law of Malinin and Khadjinsky and the evolution equation of Armstrong and Frederick are extended to describe the creep deformation of anisotropic materials. In order to express anisotropic hardening, back stress taking account of anisotropic inelastic strain sensitivity is incorporated into the creep constitutive equation. Then, the resulting model is applied to analyze the time-dependent inelastic deformation of symmetric angle-ply laminates. Comparison between the prediction and the experimental observation shows that the present model can describe well the time-dependent inelastic behavior under different loadings.
The Al2O3-WC-Co/TiC/Ni/TiC/Al2O3-WC-Co FGMs with a symmetrically graded structure were fabricated by SHS/HIP. It is possible to induce the strong residual compressive stress in the outer ceramic layers of the ceramic/metal/ceramic symmetric FGMs due to the thermal expansion mismatch between the outer and inner layers. The doping of WC-Co into the outer Al2O3 layers further increased the residual stress comparing with the Al2O3/TiC/Ni/TiC/Al2O3 system. The toughness and the strength of the outer layer increased to 13.1MPam1/2 and 950MPa by doping 20vol% of WC-Co, respectively. Stress analysis was carried out for the macro and micro structures of FGMs by means of X-ray diffraction method and finite element method. The correlation between the surface stress and the mechanical properties was discussed.
By taking the rate effect of fracture toughness into account, fracture toughness behaviour under stable crack extension in DCB specimens was studied by numerical simulation for 4 kinds of testing with an abrupt change of defomation velocity during its testing (incremental, decremental, reverse (unloading and reloading) and relaxation tests). The simulated results were verified by some experimental works on PMMA, and the load P-displacement δ curves obtained experimentally in these tests were well explained by the simulation including the rate effect in fracture toughness. Both the simulation and the experimental work show that P gets larger (or smaller) in the incremental (or decremental) test, showing viscous characteristic in P-δ relation. The reverse test shows that crack continues to extend even in the unloading process, and it extends conspicuously in the relaxation test as the effect becomes dominant.
Stable fracture behavior in a Ti-24Al-11Nb matrix composite reinforced unidirectionally with continuous SiC fibers, SCS-6, was studied by using the smooth and notched specimens with different notch length and specimen width. Special attention was paid to the understanding and the evaluation of the relationship between stable crack growth and damage tolerance. It was found that the stable crack growth occurred according to the following procedures: (i) first matrix cracking and/or interface debonding near the notch tip, which were induced and promoted by the residual stress and the β-phase depleted zone near the interface, (ii) the increase of density of matrix cracking, resulting in crack face bridging by intact fibers, (iii) the break of some bridging fibers, and finally (iv) unstable fracture, accompanying with the coalescence of cracks and with fiber pull-out. A significant increase in fracture resistance with stable crack growth was observed, in which fiber bridging was found to play the most important role. Not only the critical stress intensity factor but also the critical J-integral value at which unstable fracture occurred were inadequate as fracture criteria, because they did not show a unique value, or varied with the initial crack length, ai and with the relative ratio, ai/W. But instead, the significance of the slope in R-curve, which must give the developing rate of fiber bridging with stable crack growth, was identified as a useful material parameter which represents the degree of damage tolerance of composites.
This paper describes the fatigue life estimation of the welded joint having an unpenetrated portion in the pressure vessel by use of the fracture mechanics. The internal pressure fatigue tests of the pressure vessel were carried out at room temperature and the stress intensity factor of welded joint was analyzed by the finite element method regarding the unpenetrated portion as a crack. The simple method for the probabilistic life prediction of the fatigue crack growth is presented by applying the first-order sensitibity analysis in the deterministic fracture mechanics approach. In the calculating procedure of life estimation, the size of the unpenetrated portion and the fatigue crack gowth rate obtained from the da/dN tests are regarded as the uncertain factors. Comparing the analytical results with the tests results, the validity of the present method is confirmed.
For the estimation of fatigue damage of plain specimens under complex loading, it is important to study the effect of stress change through the behavior of a small crack, because the fatigue life of specimens is determined mainly by the growth life of a small crack. However, such studies performed in corrosion environments are very few. In this study, using 3% NaCl solution, corrosion fatigue tests of plain specimens were carried out under the constant-amplitude, low-to-high block loading and high-to-low block loading. The behavior of cracks was detected by the plastic replica method. The cumulative cycle ratio Σ(N/Nf) was calculated for the variable-amplitude loading tests. The values of Σ(N/Nf) for high-to-low loading were smaller than unity, whereas they were neary unity for low-to-high loading. The difference in Σ(N/Nf) between two loading patterns can be explained based on the small crack propagation characteristics and the density of long cracks which led to fracture.
This paper aims to establish a practical way for predicting the size of chipping at work corner by grinding force. This is based on the idea that the grinding force on an abrasive grain must act as the concentrated loading force for the extension of well developed crack causing the chipping. To find the relationship between the size of chipping and grinding force, the grinding experiments with a single diamond point of three dimensional shape were carried out on optical glass in a wider range of grinding conditions (wheel depth of cut, work feed rate and tip radius of the point). An analysis is performed to estimate the grinding force on the model that the force integrates the traction on the contact face between grain and work during a traverse of the grain that is passing over a prior groove. The experimental results show that the 3/2 power of the chipping size is proportional to the maximum grinding force during a traverse of grain. The grinding force calculated from the grain depth of cut based on the grinding geometry is not always consistent with the real force measured in the experiment, because the real grinding process includes the fracture type material removal (over-cut groove) and the large backward deflection of the grinding system due to grinding force (under-cut groove). The real grinding force is smaller than the calculated one in the over-cut groove, while it is larger in the under-cut groove. More exact evaluation for the grinding force, or the chipping size can be made by compensation of grain depth of cut using some experimental results and the analytical model proposed here.
Glass fiber reinforced vinyl ester resins (GFRP) were immersed in weak acid solution (1% sulfuric acid water) at 40°C for various periods from 0 to 12 months. After drying at room temperature for one week, the tensile and the bending tests were carried out on them. The tensile strength of GFRP decreased slowly with increasing immersion time, and the bending strength was degraded rapidly with increasing immersion time in the early stage. From the test results, it was considered as follows: (1) The decrease in strength of GFRP was due to the damage of glass fibers. (2) The damage of glass fibers proceeded from the surface to the inside of GFRP. In order to confirm this mechanism of decrease in strength, the inside of GFRP was observed nondestructively with a scanning acoustic microscope (SAM). From the observation results, it was confirmed that the damage of glass fibers was caused by corrosion by weak acid solution, and the corrosion of glass fibers was produced in GFRP. Therefore, the mechanism of decrease in strength of GFRP due to immersion in weak acid solution was clarified by this study.
Internal stress induced during curing process is investigated analytically and experimentally for the adhesively bonded parts of epoxy resin and aluminum alloy. As the parts, two types are considered: two-layered model, which has a molded resin layer on metal surface, and adhesive model, which has two metal plates bonded with resin. The viscoelastic properties of the resin during curing process are experimentally examined. The constitutive relations for the resin are formulated on the basis of the experimental results by using viscoelastic models. The internal stress induced during curing process is analyzed by using the finite element method. The stress distributions are obtained at the adhering interfaces of the two models. It is shown that the stress is remarkably concentrated at the adhering edge for both models. The method of laser Raman scattering is applied to measure the distribution of the internal stress near the adhering edges. The stress is measured by the method at short intervals along adhering interface from the adhering edge. It is confirmed that the stress singularity exists near the adhering edges.
LaxY1-xBa2Cu3Oy powders were synthesized by the aqueous chelate-mix technique using ethylenediamine-tetraacetic acid (EDTA), nitrilo-tri-acetic acid (NTA) and imino-di-acetic acid (IDA) as chelating agents. The intermediate products obtained in each formation process were investigated by thermogravimetry, X-ray diffraction technique and infrared spectroscopy. It was found that the formation of the 123 phase from precursor-gel proceeds in the following four steps; (1) decomposition of ammonium acetate and Cu-EDTA and formation of metal-copper at 120-220°C, (2) decomposition of carboxyl group of chelate at 340-410°C, (3) combustion of the residual organics and formation of BaCO3 and other oxides at 410-480°C, and (4) decomposition BaCO3 and formation of 123 phase around 800°C.
Characterization of black core in a brick body obtained by sintering the incinerated ash of sewage sludge was conducted to clarify its formation mechanism. The ash was pressed into a rectangular body under 100MPa and the body was fired at 1000°C. Under some firing conditions, the formation of porous black core accompanied by extraordinary expansion occurred in the brick body. TG and ESR analyses revealed that iron oxide in the black core was mainly in the form of FeO. The mass-spectrometric analysis of gases released during heating the ash in an oxidizing atmosphere revealed that CO2 was the only gaseous product and it had to be brought about from the residual carbon. These results indicate that Fe2O3 in the ash was reduced to FeO by this carbon. Furthermore, it was assumed that cavities in the body were formed due to the blockade of CO and CO2, formed at the reduction of Fe2O3 to FeO, by the densely sintered surface.
The tapping method to measure Young's modulus of wood is useful for non-destructive evaluation of sawn lumbers. The dynamic MOE values are calculated from the resonance frequency of the tap tone with a FFT spectrum analyzer. However, if a glulam is used for a bending member such as a beam, it is not certain that the static MOE value of the glulam is equal to the dynamic MOE value. In case of glulams composed of sawn lumbers from juvenile wood, some adjustment to the dynamic MOE values may be necessary for use. In the present study, the relationship between static MOE (Es) and dynamic MOE (Ed) was investigated, since it was expected that the variation of MOE in the cross section of homogeneous-grade glulams would be relatively small. Two types of specimens were used: glulams laminated horizontally (H-type) and vertically (V-type). Dynamic MOE was measured by longitudinal vibration (El) and flexural vibration, whose directions were horizontal (Eh) and vertical (Ev) in the direction of the adhesive face of the glulam. The results indicated that Eh/Es in H-type and Ev/Es in V-type were not influenced by the number of laminae. As it was supposed that El/Eh should be influenced by the number of laminae, the relationship between El and Eh was estimated by simple simulation for different numbers of laminae. The estimated ratios of El, Eh and Ev to Es were almost identical with the experimental data.