This paper deals with solutions for technical questions on the remaining-life assessment of fossil fuel power plants in Japan. The difference in reliability and accuracy between rationalization of regular inspection and life extension of working plants, importance of clinical methods for damage evaluation, difficulties in the quantitative evaluation of various kinds of damage, necessity of crack-and/or cavity-totolerant assessments and standardization of allowable defect size, limitation of reliability and accuracy of present methods are discussed. A methodology based on an inverse analysis of cracks or cavities is presented as a new method of damage evaluation of materials.

View full abstract

The effects of long-term aging on mechanical properties of CF3M duplex stainless steel and D308L weld metal were investigated to predict the remaining life of in-service cast components under operating conditions. This paper reports on a tensile test, a fatigue test, fatigue crack growth test, a Charpy impact test, and a J_IC toughness test conducted on these materials aged for up to 10000 hours at tempertures form 623 K to 743 K. The effects of aging conditions on mechanical strength of duplex stainless steel and D308L weld metal can be evaluated by considering the aging phenomenon to be a thermally activated process described by an Arrhenius relation.

View full abstract

Microstructural changes in the ferritic phase of ferritic and ferritic/austenitic duplex stainless steels during aging have been examined using the atom probe and the position-sensitive atom probe. These provide a means for atomic-scale microstructural and microchemical analysis, and allow the composition variations in the sample to be reconstructed in 3-D with subnanometer resolution. With the progress of aging, the ferritic phase undergoes phase decomposition via the spinodal mechanism, and the initial fluctuation in Cr concentration develops, resulting in the precipitation of Cr-rich particles with modulated three-dimensional interconnected structure. The hardening may be attributed to the spinodal decomposition, that is, the lattice misfit effect, resulting in increase in fluctuation of Cr concentration and Cr-rich particle precipitation in the modulated structure. An increase in hardness is closely related to the volume fraction of Cr-rich particles.

View full abstract

A small, highly sensitive magnetic sensor called SQUID (superconducting quantum interference device) was developed for nondestructive inspection of defects in stainless steel. The SQUID sensor can measure 3-dimensional areas through use of a robot scanning system. The device was success-fully used to detect defects, fatigue damage before crack initiation, and thermal aging degradation at 623K to 723K by means of magnetic characterization measurements.

View full abstract

The influence of thermal aging on fracture properties, transition temperature of V-notch Charpy impact test, elastic-plastic fracture toughness, fatigue crack growth behavior and stress corrosion cracking of the base metal and welded joint of A533B steel were examined. The chemical compositions of the base metal and weld wire were adjusted, referring to those of reactor pressure vessels of old Japanese commercial nuclear power plants constructed before 1970. A step-cooling heat treatment was used as embrittling treatment. The transition temperature of the base metal and welded joint subjected to step-cooling rises approximately 30°C due to the segregation of phosphorus. However, the step-cooling heat treatment has little influence on the fracture toughness, the crack growth behavior and stress corrosion cracking.

View full abstract

For clarifying the dependence of the thickness of a thin metal sheet on fatigue crack propagation in air or in a corrosive environment, the fatigue tests with data-gathering procedures as dictated by ASTM were carried out on specimens of thin 6/4 brass sheets with thicknesses of 0.3 mm, 0.5 mm, 0.8 mm, and 2.5 mm, which are coated by corrosion-proof oil except at the notch. As a result, the relationship of the crack propagation rate, da/dN, to the stress intensity factor range, ΔK, was obtained. It is found that in the second range of the relationship in a corrosive environment, the slope, m, is not affected by the thickness of the specimen, and shows a stable value, but in air, the slope is affected by thickness. Assuming that the plane strain condition is established when the thickness of the specimen is equal to or larger than the range of plasticity at the head of a crack subjected to tension in one direction, the ranges where the slope, m, of each different thickness are stable and parallel to each other, within each limit of ΔK, are determined, and the formula for determining the thickness of the specimen under the condition of plane strain can be derived from the experiment. In this way, reliable data on the relationship of da/dN to ΔK is attained.

View full abstract

The present paper describes an experimental study on a physical meaning of nondestructive creep damage detection for a steam turbine rotor material after long-term service using an ultrasonic technique. As a result of microstructure observation and density measurements, it was made clear that change in ultrasonic parameters such as attenuation and noise value was found to be attributed to nucleation, growth and coalescence of creep cavities with increasing creep damage. The ultrasonic technique was found to be valid for predicting remanent creep life of a fossil fuel plant steam turbine rotor.

View full abstract

During the last ten years, more than 70 incidents of creep embrittlement damage of structural components in catalytic reformer reactors which had been in long-term operation have been reported. Creep embrittlement is the loss of creep ductility in the heat-affected zone of welds or base metal. Creep void density, as related to stress concentration, revealed a critical stress intensity for a catalytic reformer reactor that was actually operated for the last 19 years. There was a threshold Larson-Miller parameter in the range of 19.0×10³ to 19.5×10³, above which creep embrittlement occurred. Creep crack growth tests were also performed to evaluate crack propagation and remaining life of the components. For the safety of these components, it was necessary to evaluate the creep void density because of the shortage of long-term crack propagation data.

View full abstract

Analytical discussion was given on the relationships between actual cavity distribution and creep damage parameters such as volume fraction of cavities, effective cross-sectional area and the 'A' parameter. The relationships between cavity distribution and these damage parameters were derived using a very simple probabilistic model where creep cavities on a grain boundary facet were initiated and grew, depending only on the macroscopic stress (or strain rate) acting on the facet. The exact relationships were represented by Eqs. (16), (20) [or (23)] and (26), respectively. These expressions take into account the effect of the size and direction of grain boundary facets. Damage evolution during creep and creep-fatigue can be calculated using these equations.

View full abstract

The schedule of activities for plant life extension should be optimized on a rational basis. This paper describes a new method of optimizing such a schedule from the viewpoints of reliability and cost. Degradation of components and implementation of activities are modeled by means of the Markov chain. Then, a potential function considering the cost of the activities is minimized under the constraint condition of reliability in terms of failure probability, by the use of the Monte Carlo simulation method. Calculation results of 30 components show the effectiveness of the present method.

View full abstract

Statistical fatigue tests in the long-life region were conducted on a medium carbon steel, S45C, in 3%NaCl solution. It was found that corrosion fatigue life distribution followed the three-parameter Weibull distribution and the scatter of fatigue life increased with decreasing stress level. Along with these fatigue tests, the growth behavior of corrosion pits was examined using a laser microscope. The distribution of corrosion pit depth followed the log-normal distribution, and the corrosion pit depth increased with increasing time or the number of stress cycles. It was confirmed that the growth of corrosion pits depended on the stress level and was enhanced with increasing stress level. Based on these results, a growth law of corrosion pits including the effect of stress level was proposed.

View full abstract

The purpose of this paper is to characterize the fatigue damage tolerance behavior of unidirectionally continuous-fiber-reinforced metal matrix composites. The fatigue crack growth tests were conducted on silicon-carbide-fiber unidirectionally reinforced titanium alloy matrix composite SiC_CVD/Ti-15-3 using a single-edge-notched(SEN) specimen. The analysis of the effective stress intensity factor at the fatigue crack tip with reinforcement fiber bridging was also carried out on the basis of the crack closure model. It was found that there is little difference in fatigue crack growth resistance between L- and T-orientations.

View full abstract

The fracture mechanisms in the bending process of a chopped-strand-type CFRP with different degrees of bending fatigue damage are examined and compared with those of the virgin specimen. Observation using SAM and SEM has also been conducted. It is confirmed that the fatigue damage has a distinctive influence on AE events before the initiation of macrocracking. As the degree of fatigue damage increases the events with amplitude from 240μV to 560μV due to debonding between the matrix and fiber increase and the events with amplitude from 600μV to 720μV mainly due to fibers breaking decrease. The characteristic differences in the three-dimensional diagram of power spectrum-frequency-load between the virgin specimen and the fatigued one are described. Distinctive features before and after the macrocracking in the bending fracture process for both the virgin specimen and the fatigued one are also recognized.

View full abstract

Fatigue life of a plain specimen is stochastic and consists mainly of the growth life of small cracks. Therefore, the distribution of fatigue life can be evaluated through the statistical properties in the growth of small cracks. In the present paper, rotating bending fatigue tests were carried out on specimens with two small blind holes in 0.42% carbon steel in order to investigate the distribution characteristics of the fatigue crack growth life based on the small-crack growth law, dl/dN=Cσ_a ⁿ. The distribution of crack growth life of holed specimens can be predicted through the growth law considering the statistical properties of the crack growth rate. Although the crack growth rate in a plain specimen fluctuates markedly due to the influence of the microstructure, scatter of the crack growth rate among the specimens can be evaluated through the distribution of the crack growth rate in the region where the growth behavior is mainly controlled by a mechanical parameter. That is, the distribution of fatigue life in plain specimens can be predicted through the statistical properties of the crack growth rate which are controlled by a mechanical parameter.

View full abstract

Crack tip stress fields of cracks propagating at a speed of about 550 m/s are experimentally studied at the time of 18 or 73μs after crack initiation, and the following results are obtained. (1) In the direction along the crack, no effect of higher-order terms appears at either 18 or 73μs after crack initiation, and the singular field is developed as far as 9 mm from the crack tip. Hence, measurement of crack opening displacement always gives the true value of the dynamic stress intensity factor, K. (2) In the direction of 72 degrees from the crack propagation direction, the effect of higher-order terms strongly appears at 18μs after crack initiation, but it disappears at 73μs, and the singular field is developed as far as 5 mm from the tip. It can consequently be said that the higher-order terms affect K measurement by the caustic method if the measurement is carried out within several tens of microseconds after the beginning of rapid crack propagation.

View full abstract

Construction of physical model and numerical analysis were carried out concerning hydrogen diffusion in solid under the elastic and elastic-plastic local stress field. These analyses showed that hydrogen diffuses and concentrates at the site of the elastic-plastic boundary in the direction of crack length, when the gradient of the initial density of hydrogen exists in this region. The characteristic of hydrogen concentration in this region is accelerated when yield stress becomes larger. These analytical results were related to the sensitivity of corrosion cracking.

View full abstract

This paper presents a general method of solving problems of anisotropic semi-infinite thermoelastic body due to a concentrated steady-state source of heat with various surface boundaries such as free, fixed and two sliding conditions. The principal axes of elastic moduli of the anisotropic body under consideration may have arbitrary directions such as out-of-plane directions. General thermal properties of anisotropy which is variable coefficients of heat conduction and thermal expansion are assumed. The thermal surface boundary condition of semi-infinite body is isothermal or adiabatic.

View full abstract

A new method is presented for estimating residual stress produced in a pipe made by surfacing weld. In focusing on the concept and features of inherent strain causing residual stress, the proposed method has the convenience of the Sachs method and a higher degree of accuracy. An experiment on residual stress of steel with brass padding was performed using both the Sachs and the proposed method. The equation of the Sachs method for a composite pipe was derived. The equation for the new method was obtained using the theory of inherent strain. It is found that the proposed method is useful for estimating residual stress of pipe with padding.

View full abstract

This paper describes a method of measurement for the Mode I stress intensity factor (S.I.F), which uses extremely small-sized strain gages consisting of 5 measuring grids. The strain gages used in this study are extremely small consisting of 5 grids with a gage length of 0.15 mm and pitch of 0.5 mm. This paper proposes two kinds of methods of stress intensity factor alalysis. One of them uses only one datum of the nearest gage at the crack tip ("1-point gage method"), and the other uses 5 gage datums which are used in the extrapolation ("5-point gage method"). The experiments are carried out using center-cracked test specimens (CCT) and single-edge-cracked plate tension specimens (SECT). For the SECT specimen, the accuracy of the S.I.F mode I determined by the experiments is within ±13 percent for the "5-point gage methed" and "1-point gage method", compared to the analytical value.

View full abstract

In this study, the stress singularity at the jointed corner tip of a full plane consisting of trimaterial wedges is considered. The eigenequation for the stress singularity is given in an explicit closed form. The eigenequation is reduced from a 4-dimensional determinant. The eigenequations for a single wedge or for a bimaterial wedge can be seen as special cases of the present eigenequation for a trimaterial wedge.

View full abstract

Thermographic NDT is developed for the inspection of the delaminated defects embedded in fiber reinforced plastic samples. The location and shape of the defects are identified from the surface temperature distribution appear on the sample by the heat insulation effect of the defects, when the sample is under heating and/or cooling. The resolution is examined for the artificial delaminated defects both in CFRP and GFRP samples, when the conventional infrared radiation heating is adopted. Experimental results show that the radiation heating is not effective for the accurate defect inspection of the CFRP sample due to its high thermal conductivity. The Joule effect heating, on the other hand, is found to be effective for the similar sample. Two methods of the Joule effect heating, a direct current application and an induction heating, are investigated. The induction heating thermographic NDT is degraded in the resolution of the defects, compared with the direct current thermographic NDT, because of the inhomogeneity in the induced current field. This can be improved by the use of an image processing technique in the analysis of the thermal images. Further the present thermographic NDT technique is applied for the inspection of the GFRP sample with the actual delaminated damage under cyclic straining. The possible detection of the subsurface damage is discussed.

View full abstract

Abstract-In this paper, we consider the transient thermal-mechanical stresses of G-10CR glassepoxy laminates with temperature dependent properties. The layered composite is made of a layer bonded between two layers of different physical properties, and it is suddenly cooled at the surfaces. The composite material in generalized plane strain is assumed. Numerical results on the transient temperature and thermal-mechanical stress distributions are obtained and are presented in a graphical form.

View full abstract

The impact force generated by high-velocity impact of a dry ice cylindrical flier, which was accelerated with a nitrogen gas gun, was measured by a sensor which we developed. This technology is expected to be useful in many kinds of machining and civil engineering. The tensile stresgth of dry ice and the other material constants were evaluated by radial compression test (Brazilian one) and the finite element analysis. Failure views of dry ice during the impact were taken with a strobovision scope.

View full abstract

Considering that dry-ice is a pure solid material of carbon dioxide and easily sublimed, we intended to utilize it as a high-speed projectile for material processing such as deep drawing and forging of engineering materials. In this article, an experimental investigation is reported on the deep drawing of thin circular plates of stainless steel (SUS304) which are subjected to the direct impact of dry-ice projectiles. The relationship between the kinetic energy of the projectiles and the plastic deformation of the plates is discussed as the centeral topics.

View full abstract

The present paper describes a theoretical treatment of the internal time theory of plasticity from the viewpoint of the internal variable theory, where Helmholtz free energy is expanded by total strain and internal variables. Based on thermodynamics, the relation of elastic work to the free energy expansion and the relation of plastic work in internal dissipations are derived. By regarding an internal variables as plastic slip deformation in the slip system of crystal, the internal variable theory is shown to be identical to crystal plasticity, in which a slip model incorporates the influence of elastic strains, then Onsager's reciprocal relation is used to obtain the evolutional law of the slip model. The macroscopic constitutive equation of plasticity is derived based on the assumption of isotropic property of plastic strain, and is derived in the rate form to incorporate the temperature change.

View full abstract

Coupled equations for mass and energy transportation incorporated with fluid/solid transformation are formulated in the Eulerian coordinate system. The latent heat generation associated with the mushy zone and the heat due to mechanical work of fluid flow are introduced in the heat conduction equation, and the prediction of a solidifying boundary is evaluated by considering the mixture theory of solid and fluid by use of a lever rule in the phase diagram. Assuming the viscoplastic constitutive equation for both solid and liquid, a finite element scheme is formulated. As an application, the thin slab casting process by the twin roll technique in a steady state is simulated to evaluate the fields of temperature velocity and stress/deformation. In order to verify the procedure, calculated temperatures are compared with the measured data, and the effect of a jet stream from the nozzle on the flow pattern is discussed.

View full abstract

The purposes of this paper are to elucidate fundamental material properties of modified 9Cr-1Mo steel as a candidate material for next-generation fast breeder reactors and to obtain information for the formulation of a practical unified constitutive model. For these purposes, monotonous tension, uniaxial symmetric cycles, uniaxial symmetric cycles with hold time, creep, stress control uniaxial symmetric cycles, cyclic relaxation and circular cyclic loading tests are performed. Then these results are simulated by two unified constitutive models, i.e. the Chaboche model and Bodner-Partom model. The results of simulation show that the Chaboche model can describe all the experimental results with relatively high accuracy, and that the identification of material constants of the model is easy. The Bodner-Partom model can similarly describe the results, but the determination of material constants is difficult.

View full abstract

This paper describes the development of an automatic finite element mesh generation method using a shape recognition technique for surface models. The major advantage is that a surface model is divided into only quadrilateral meshes by a single data input. In this method, one or more numbers are input to determine the number of divisions of edges. An arbitary plane derived from a surface model is analyzed, and squares are put together to construct a similar model based on human thought processes, expressed by membership functions defined in fuzzy set theory. Quadrilateral meshes are then generated by mapping the squares into each plane. We develop a new method for assigning division numbers for all edges because there are some restrictions on the number of divisions. These restrictions result from the condition that a surface model should be divided into only quadrilateral meshes. The performance of the method is demonstrated through automatic mesh generation for some mechanical parts.

View full abstract