Nickel-base and cobalt-base superalloys for landed and aerospace gas turbines are situated between in ready-made, existing materials and in order-made, advanced materials. Although the development of elemental materials, evaluation of structural materials, design of parts and maintenance of structures have been performed rather independently in each general (or vertical) research and development (R & D) units for the existing materials, they have to be investigated in close connection with each other as an objective (or holizontal) R & D for the advanced materials. In this paper, some required subjects for the objective R & D are pointed out by taking the case of superalloys for gas turbines: necessity of out-of-phase thermal fatigue resistance, crack propagation analysis and coating materials design in the development of materials, importance of inelastic analysis for crack tolerance design, fatigue life evaluation under compressive strain-hold cycling and coating effect on crack initiation in the design of structural materials, and difference between pathological and clinical approaches and applicability of inverse analysis for remaining life prediction in the diagnostics of structures in service.
As a fundamental study of developing a long term creep-fatigue evaluation method of weldment of stainless steel that incorporates the degradation of weld metal, the authors investigated the mechanism of micro damage in the structure of 308 and 316 type weld metals that had been subjected to thermal and mechanical histories using transmission electron micrography. The results obtained are as follows. (1) M23C6 and Laves phase that precipitated in δ-ferrite phase were transformed into σ phase that is more stable under high temperatures after long time heating. (2) While δ-ferrite phase remained in the microstructure at 550°C even after supplying alloying elements to precipitates and changing to Fe-rich composition, at above 600°C it disappeared after transforming completely to σ phase or austenite phase. (3) The rate of dissolution of δ-ferrite phase under cyclic stresses for fatigue or creep-fatigue failure was higher than that under constant stress for creep failure, and was accelerated by 10 to 100 times compared with that under no load. (4) The dissolution of δ-ferrite phase led to initiation and propagation of cracks at the interfaces of σ phase and remaining δ-ferrite phase or austenite phase by making a difference of mechanical properties between the phases.
This paper deals with the estimation of thermal ratchetting strain of cylinders subjected to short axial travelling of temperature distribution. The estimation is made by considering partly pressurized cylinders as a model. It is assumed that the pressurized region is ruled by the deformation theory with the uniform secant modulus determined so as to be exact in the long travelling case. This assumption enables us to utilize a linear elastic solution for the problem. The secant modulus is specified using the analytical solution which was derived by the present authors for the long travelling case by assuming Masing's rule. It is shown that the model based on the partly pressurized cylinders gives good estimates comparable to the finite element analysis of the thermal ratchetting strain as well as to the experiments on 316FR steel cylinders done by Kitade et al. An extension of the model, in which plastic yielding outside the travel region is taken into account, is also discussed.
Long-term stress relaxation behavior was investigated on NCF 800H alloy. Complicated stress relaxation curves such as those with rapid decrease or increase in residual stress were observed. Their relationship between stress relaxation behavior and microstructural change was examined. The rapid decrease in residual stress was caused by the reduction of creep deformation resistance due to coarsening of carbides. Minus relaxation behavior which expresses the increase in residual stress was related to the shrinking of testing material due to precipitation of carbides. Therefore, it is difficult to extrapolate the long-term stress relaxation curve from the result of short-term tests, and the long-term stress relaxation value should be predicted using the aged material.
Modified 9Cr-1Mo steel is the principal candidate material of a steam generator in a fast breeder reactor because of its superior high-temperature property. In this study, the influence of long-term strain hold on the failure life of Mod. 9Cr-1Mo steel was investigated and the applicability of existing creep-fatigue life evaluation methods was disccussed based on the experimental results. Creep-fatigue tests with hold time up to 10 hours per cycle at either tension or compression maximum strain were conducted under strain control of 0.5% and 1.0% at 550°C. Although failure life reduction occurred by introducing short hold period in the compression hold tests, the degree of reduction became smaller with increasing hold period. On the other hand, the failure life reduction became larger with increasing hold period in the tension hold tests. It was found that the failure life reduction in the compression hold tests was mainly due to the accumulation of tensile strain at the center of a specimen and that in the tension hold tests it was caused by intergranular damage in terms of creep cavity nucleation and growth from observation of the failure specimens. The time fraction rule adopted in a design code gave longer life compared to the experimental life, especially at low strain level. The failure life was well predicted by the ductility exhaustion method and the nonlinear damage accumulation model.
The fatigue fracture of a Ni-base superalloy, IN738LC, coated with so-called MCrAlY alloys by low pressure plasma spraying was studied at 400°C and 800°C in vacuum condition. Most of the protective coatings exhibited the fatigue strength comparable to that of IN738LC substrate, when fatigue fracture was controlled by the propagation of fatigue cracks which initiated from the casting defects in the superalloy substrate. However, when the fatigue crack initiation occurred from the coated layer, the fatigue strength was significantly reduced depending on the MCrAlY alloy systems, compared with that of the substrate alloy. The fatigue crack propagation process was also elaborately investigated at 400°C by employing the special equipment with a cyclic loading stage which fits within a scanning eiectron microscope. Special focus was paid on the fatigue crack growth behavior in the coated layer, as well as near the interface. The fatigue threshold level significantly depended on the MCrAlY alloy systems used. The fatigue crack grew in a very complicated manner near the interface in all of the protective coatings. Based on the measurements of the residual stress and the mechanical properties of the MCrAlY bulk coating alloys, the factors relevant to the specific feature of the crack propagation in protective coatings were discussed from the viewpoint of fracture mechanics.
Creep-fatigue tests of SUS304 stainless steel were conducted and creep cavities initiated on grain boundaries inside the specimen were observed on the longitudinal cross-section by means of a scanning laser microscope. The results obtained were as follows. (i) The shape of cavities was almost spherical when they were small. (ii) There was no preferential direction of grain boundaries for the cavity initiation. (iii) Since cavities preferentially grew and coalesced along grain boundaries perpendicular to the stress axis direction, the shape became slit-like. Based on the experimental determination of cavities, four kinds of creep damage parameters, such as area fraction of cavities on cross-section, fraction of cavities on grain boundary lines, areal cavity density, and A-parameter were evaluated.
In order to clarify effective measures to extend a life of converter shell in a steel production plant, the thermal fatigue crack initiation life was predicted by the strain range partitioning method. A three-dimensional FEM model of the converter which is composed of a shell and bricks was established. An inelastic strain of the shell caused by a thermal cycle during operation was evaluated by an elastic-plastic-creep analysis. Since the thermal expansion of bricks is partly absorbed by a gap between bricks, the inelastic strain of the shell is affected by the gap size. Therefore, in the FEM model the hyperelastic element was used in order to describe the deformation behavior of bricks and gaps, and the stress analysis was conducted taking the gap size into consideration. Based on the analytical results and creep-fatigue properties of a shell material, the crack initiation life of the shell was predicted. The relationship between the shell life and cooling conditions at the outer surface of the shell and the relationship between the shell life and the gap size were examined. The following results were obtained. (1) The predicted shell life corresponded well with that in the actual converter. (2) It is possible to extend the shell life by decreasing the maximum shell temperature which can be attained by increasing the film coefficient of cooling at the outer suface of the shell. (3) It is also possible to extend the shell life by increasing the gap size. Futhermore, the larger the gap size is, the decrease of shell temperature becomes more effective for the shell life extention.
Creep crack growth tests on very large 5T-CT specimens of 1Cr-Mo-V steel have been conducted in order to clarify the specimen size effect on creep crack growth properties and to contribute to the standardization of test method in VAMAS project. The creep crack growth rate could be correlated with C* parameter except for the initial stage of crack growth. In this stage, crack was considered to propagate under small-scale creep conditions. In the accelerating stage, the crack growth rate was independent of specimen width but was dependent on specimen thickness. As the specimen thickness increased, the creep crack growth rate increased due to the deformation constraint. The estimation method of displacement rate for CT specimens from the creep strain curve of round bar specimens was also discussed.
This paper studies the low cycle fatigue crack detection using the A.C. potential drop (ACPD) method for CoCrAlY coated 738LC Ni base superalloy. Push-pull low cycle fatigue tests were carried out at 1123K in air using solid specimens with fast-fast and slow-fast waves. The low cycle fatigue life of the coated specimen was somewhat smaller than that of the uncoated specimen, but there exists no large difference between the two lives. In the fast-fast wave of the coated specimen, cracks were initiated at the specimen surface and propagated inside. In the slow-fast wave test, some cracks were initiated at the interface between the coating and base metal in addition to the crack initiation at the specimen surface. ACPD method was able to detect the cracks initiated at the specimen surface and the interface. Especially low frequency ACPD is more sensitive than high frequency ACPD. The result of the electromagnetic FEM analysis agreed well with the experimental ACPD measurements.
This paper describes a newly developed analytical method of evaluation of creep-fatigue strength of stainless weld metals. Based on the observation that creep-fatigue crack initiates adjacent to the interface of sigma-phase/delta-ferrite and matrix, a mechanistic model which allows the evaluation of micro stress/strain concentration adjacent to the interface was developed. Fatigue and creep damage were evaluated using the model which describes the microstructure after exposed to high temperatures for a long time. Thus it was made possible to predict analytically the long-term creep-fatigue life of stainless steel metals whose microstructure is degraded as a result of high temperature service.
The fracture toughness KJC values of small specimens in the transition temperature region were analyzed to confirm the offset method standarized by JSPS 129 Committee. The offset method is a very useful method for the determination of lower-bound fracture toughness KJCi of material especially in the lower shelf temperature region where the stable crack growth is smaller than about 0.2mm. The KJCi values obtained from the offset method showed a good agreement with the KIC values of large specimens per ASTM E399 and the KJCi value from the fractography method standarized by JSPS 129 Committee.
The rolling contact fatigue of TiC- and TiN-coated steels was investigated. The damage was observed with a Scanning Acoustic Microscope (SAM). Measurements of X-ray residual stress and surface leaky wave velocity were also made to evaluate the damage. The SAM observation was useful to reveal the behavior of delamination and crack propagation below the surface. The results obtained are as follows. (1) In the TiC-coated steel, cracks initiated at micro pits generated, in the CVD process and propagated in a parallel direction to the rolling direction of balls at the interface between TiC film and the substrate. This resulted in the delamination and finally the peeling of TiC film. (2) In the TiN-coated steel, micro delamination appeared at spherical defects between the film and the substrate, and grew into macro delamination. The delamination expanded and finally TiN film on the delamination peeled off the substrate. (3) Both TiC and TiN films had an extremely large compressive residual stress. The progress of the film delamination caused an additional compressive stress in the tangential direction of the rolling direction. The residual stress measurement by X-ray was found to be effective for the evaluation of the contact fatigue damage of TiC- and TiN-coated steels.
In general, the drying shrinkage of polymer-modified mortars using redispersible polymer powders is much higher than that of unmodified mortar. The purpose of this study is to develop effective powdered shrinkage-reducing agents for reducing the drying shrinkage of polymer-modified mortars using a redispersible poly (ethylene-vinyl acetate) (EVA) powder, which is widely used for the manufacture of premixed-type polymer-modified mortar products at present. Polymer-modified mortars using the redispersible EVA powder with three powdered shrinkage-reducing agents were prepared with various polymer-cement ratios and shrinkage-reducing agent contents, and tested for drying shrinkage and strength. The conclusions obtained from the test results are summarized as follows: (1) The drying shrinkage of the redispersible EVA powder-modified mortars with a powdered shrinkage-reducing agent, polyethylene glycol with a molecular weight of 3000 is remarkably reduced with increasing shrinkage-reducing agent content, and becomes approximately a half of that of the redispersible EVA powder-modified mortars with the same polymer-cement ratios and without the shrinkage-reducing agent at a shrinkage-reducing agent content of 6%, (2) The drying shrinkage of the redispersible EVA powder-modified mortars with the polyethylene glycol can be predicted as a function of the shrinkage-reducing agent content, polymer-cement ratio, water-cement ratio, air content and surface tension (of polymer-modified paste extracts), and (3) The addition of the polyethylene glycol to the redispersible EVA powder-modified mortars causes a reduction of 28% or less in the flexural strength and a reduction of 20% or less in the compressive strength.
This investigation was directed towards the experimental analysis of bending properties of injection molded PC/ABS blend and GFPC/ABS blend. To study the internal structures of injection molded articles, three point bending tests of as-received and shaved skin layer specimens were carried out at various temperatures and strain rates, and time-temperature superposition principle was applied to bending properties. From the results of bending test, two different straight lines of activation energy were obtained when plotting the temperature dependence of time-temperature shift factor. The activation energy varied by the change of specimen thickness in the case of PC/ABS blend. On the other hand, the activation energy exhibited constant value in spite of the change of thickness in the case of GFPC/ABS blend. It can be considered that PC/ABS blend injection molding has heterogeneous internal structure through thickness, whereas GFPC/ABS blend injection molding has uniform internal structure. Furthermore, the morphology of each article was also examined by scanning electron microscopy. It was obvious that PC/ABS blend had heterogeneous phase distribution through thickness direction, whereas GFPC/ABS blend had uniform distribution. Therefore, the activation energy of PC/ABS blend molding varied dramatically by shaving its skin layer. From the above all results, it appears that the internal structures of polymer blend injection moldings can be understood by using the activation energy obtained by time-temperature superposition.