The integrity of coating can exert a major influence on the first-stage bucket life of gas turbines. This paper clarifies the role of coating and then describes the degradation mechanisms of Platinum-modified aluminide diffusion coating and vacuum plasma sprayed Co-Cr-Al-Y coating based on the investigated results of macro- and micro-structural degradation behaviors for the buckets exposed in service for 11000 and 25000 hours. The life exhaustion concept of the first-stage bucket with that of the coating are also described.
In order to attempt the nondestructive detection of creep damage by ultrasonic technique, the changing behavior of the energy attenuation coefficient of ultrasonic shear wave with creep damage was investigated by using SUS316 steel. Two kinds of energy attenuation coefficients, αst and αsL, were measured, in which the ultrasonic shear stress acted in the parallel and normal directions to the loading axis, respectively. The energy attenuation coefficients changed in a reasonable manner in which some intrinsic types of creep damages, such as the deformation of grains, the change of dislocation structure, the progress of aging, and the nucleation of microcracks, were reflected. Moreover, the relative ratio of αst/αsL monotonically increased with creep damage. From the above results, it was concluded that the absolute and relative properties of ultrasonic attenuation may be useful for the nondestructive creep damage detection of materials with high creep rupture ductility, as a potential measure.
In order to investigate the applicability of a creep damage evaluation method based on damage mechanics, creep tests on the specimens from five kinds of Cr-Mo-V steel differing in impurity content were carried out at 550°C-21kgf/mm2 and 575°C-16kgf/mm2. The results obtained are as follows: (1) When calculation was carried out on the basis of the generalized damage equations derived from the Kachanov-Rabotonov equations, the calculated results corresponded relatively well to the experimental results for both the creep curve and creep damage parameters. (2) The generalized damage equations in (1) are widely applicable to the damage evaluation of actual components under intergranular fracture conditions. (3) Among the creep damage parameters, the A parameter value corresponded best to the calculated results of ω in comparison with the other two parameters, i.e. the cavity number and fraction of area. Therefore, the A parameter value is the suitable quantitative parameter for cavities in arranging them with t/tr. (4) Consequently the equation to estimate the creep life fraction from the A parameter value is given by: t/tr=1-(1-A)η+1 and: η+1=nλ/(λ-1) where n and λ are material constants and A is the A parameter value. Furthermore, in consideration of dependence on material composition, λ is denoted by: λ=10(1-Y)where Y is the creep ductility parameter, defined as the Y-factor. On the basis of the above equations, a creep damage evaluation method to estimate the creep life fraction from the A parameter value is proposed.
Creep-fatigue rupture life and its elastic stress concentration dependence for circumferentially notched 1CrMoV steel bars have been investigated. The elastic stress concentration Kt dependence of creep-fatigue life under 30 minutes tensile hold cyclic loading at 550°C for the long term used material was different from the unused material. The notch strengthening behavior was observed on the both materials as in the case of creep loading conditions. The maximum rupture life was observed at Kt=1.9 for the unused material, while it was at Kt=2.6 for the long term used material. The distribution of hardness across the minimum notch section corresponded to the distribution of cavity density. Many cavities were observed at the softening part. The hardness of the long term used material was lower than the unused one. The fracture strain of the notched specimen under creep-fatigue loading had an inversely proportional relationship with the rupture life. The high triaxial stress state specimen of the unused material (Kt=2.6 or 3.5) exhibited low fracture strain in the intermediate stress range, about a half of the strain of others. The fracture strain of the long term used material was higher than the unused one. In the lower stress range, the fracture strain of the notched specimen was relatively low for both of the materials. Bridgman effective stress which considers the stress triaxiality seemed useful to well evaluate the rupture life of circumferentially notched specimen under creep-fatigue conditions for both the long term used and unused materials.
To confirm the safety and reliability of the long-term operating thermal power plants, various non-destructive residual life assessment methods of component materials have been developed in the past. In this paper, in order to develop a non-destructive creep-fatigue damage assessment method, a series of fundamental experiments were carried out. The experimental condition was as follows: Test material was 2.25Cr-1Mo simulated HAZ. Low cycle fatigue tests and creep-fatigue tests with tensile strain hold were carried out using smooth specimens at 570°C. The strain range employed was from 0.4% to 1.6% and the length of strain hold-time was 10 minutes. To observe the small crack growth behavior during the test, the tests were interrupted and replicas were taken from the specimen surface. The following results were obtained. (1) The creep-fatigue life was smaller than the low cycle fatigue life. Transgranular cracking was observed in the low cycle fatigue specimen, while intergranular cracking and creep voids were observed in the creep-fatigue specimen. The significant life reduction caused by the strain hold resulted from the accumulation of creep damage at grain boundaries. (2) Many small cracks were observed on the specimen surface in the creep-fatigue test. These small cracks grew and joined to make large cracks. If the creep-fatigue life was defined as the cycle when the maximum crack length reached about 2mm, the creep-fatigue life was predictable by the maximum crack length.
A life assessment method is proposed for severely damaged components such as heavy-duty gas turbine nozzles. The method named “Multi-factor damage simulation analysis” is introduced as a sort of inverse problem analysis for actual complex damage phenomena influenced by temperature-stress distribution, microstructural inhomogeneity, environment, interaction of multiple cracks, and so on. The microstructural inhomogeneity model of Co-base nozzle superalloy FSX414 is constructed by using fractal geometry of random percolation cluster model. Multiple crack intiation and growth simulation is performed with a probabilistic damage parameter which is defined as the transition probability of crack initiation and growth in the form of monotonically increasing function of cycle. This method was applied for high temperature low cycle fatigure testing of FSX414 at 923K, 1023K, 1123K and 1223K by assigning relatively simple values for resistance R and driving force F. Arrhenius type temperature dependence was postulated for resistance R by taking account of the experimental results. The crack morphology obtained by the simulation and the crack length distribution were quite similar to the observation results of the failed low cycle fatigue specimen surface. Moreover, the temperature dependence of the rate of life expenditure (1/N25) coincided with the experimental results. Therefore, it is concluded that this method is a promising way to estimate the life of severely damaged component under complex situations.
For the purpose of clarifying the effect of strain waveform on creep-fatigue property under variable straining, two-step variable straining (high-low and low-high) tests were conducted on Mod. 9Cr-1Mo steel at 600°C in air under PC (fast-slow) and CC (slow-slow) type strain waveforms, and the experimental results were compared with the previously reported results obtained under PP (fast-fast) and CP (slow-fast) type strain waveforms. In both high-low and low-high tests, 1.0% was chosen as the higher strain range, and 0.8% and 0.4% as the lower strain ranges. The main results obtained in the present study are as follows. (1) Independent of strain waveform, the high-low and low-high test results could be predicted by the linear damage rule (LDR) when the lower strain range was 0.8%, whereas the deviation of the experimental data from LDR prediction was found when the lower strain range was 0.4%. (2) Such a deviation of the experimental data from LDR prediction was similar to that in CP type strain waveform, but was much smaller than that in PP type strain waveform. (3) The previously proposed creep-fatigue life prediction model under variable straining for the PP and CP data can be successfully applied to PC and CC type strain waveforms, suggesting that in Mod. 9Cr-1Mo steel 2(a0)pp=2.5×10-4mm and 2(a0)cp≈2(a0)pc≈2(a0)cc≈10-3mm.
Failure induced by creep-fatigue damage during operation, is an important failure mode to be avoided in high temperature structural components of thermal and nuclear power plants. Prediction of failure mechanism under creep-fatigue conditions in actual plants, and evaluation of the validity of creep-fatigue life prediction methods need to be done on the basis of long-term creep-fatigue test results. In this study, long-term creep-fatigue tests with up to 10 hours strain peak dwell were conducted using 304 stainless steel and the subsequent failure mechanism and creep-fatigue life prediction method were discussed. From the detailed observations of the failure specimens using a scanning electron microscope, many creep cavities and microcracks at grain boundaries were observed inside specimens. It was indicated that the main crack initiated on the specimen surface and propagated favorably on the cavitated grain boundaries under long-term creep-fatigue conditions with a tension hold period of over 30min. A creep-fatigue life evaluation method was proposed considering the interaction between fatigue and creep damage based on the failure observation. The experimental data of long-term creep-fatigue life in this study and existing literatures were compared with the predicted life by the proposed method and it showed good agreement.
The long term creep-fatigue life of type 304 stainless steel was evaluated by the creep-fatigue life prediction method based on a linear damage fraction rule. The displacement controlled creep-fatigue tests were carried out, and the time to failure of longer than 10000 hours was obtained. The creep damage of long term creep-fatigue was evaluated by taking into account the stress relaxation behavior with elastic follow-up during the hold period. The relationship between life reduction of creep-fatigue and fracture mode was provided by the creep cavity growth. The results of this study are summarized as follows; (1) The long term creep-fatigue data can be reasonably evaluated by the present method. The predicted lives were within a factor of 3 of the observed ones. (2) The present method provides the capability to predict the long term creep-fatigue life at lower temperatures as well as that at the creep dominant temperature. (3) The value of creep damage for the long term creep-fatigue data increased by elastic follow-up. The creep-fatigue damage diagram intercepted between 0.3 and 1 can represent the observed creep-fatigue damages. (4) The cavity growth depends on the hold time. The fracture of long term creep-fatigue is caused by the intergranular cavity growth. The intergranular fracture of creep-fatigue is initiated by the cavity growth and followed by the microcrack propagation along grain boundaries starting from creep cavities.
The short-time tensile strength and creep-rupture strength of butt welded joints for 304 stainless steel by electron beam welding (EBW), narrow-gap gas tungsten arc welding (GTAW) and submerged arc welding (SAW) were evaluated. Three types of SAW joints, were prepared by using a conventional 308 wire and two modified wires with the addition of Nb and V. Short-time tensile and creep-rupture tests were conducted on 308 weld metal specimens (GL=30mm, Dia.=6mm) and welded joint specimens (GL=100mm, Dia.=10mm) at 500, 550, 600, 650 and 700°C. Little difference in short-time tensile strength was observed among conventional and modified 308 SAW weld metals. On the other hand, the creep-rupture strength of modified 308 weld metals was higher than that of conventional 308 weld metal. However, the creep rupture ductility of the 308 weld metals was not improved by the addition Nb and V. First, the creep-rupture data of the welded joints and the base metal were evaluated by the Larson-Miller parametric method. Then, the stress rupture factor (SRF) of the welded joints for creep-rupture strength were computed, and were compared with that for weldments given in Code Case N-47-28 of ASME Boiler and Pressure Vessel Code. The relationship between the SRF and the time dependent allowable stress Smt was discussed. Generally, the SRF for EBW, GTAW and Modified SAW joints were larger than 1.0, which means overmatching joints. But the SRF of SAW joint with conventional 308 wire decreased with increasing time or temperature, becoming as low as 0.67 in 100000h at 650°C and below the value of CC N-47. Besides, it was shown that the allowable stress Smt of the weldment became lower than that of CC N-47 when the SRF was less than 0.83.
This paper describes the high temperature low cycle fatigue life of high nickel steel (HNIS)-SUH35 friction welded specimens. High temperature low cycle fatigue tests were carried out on HNIS, SUH35 base metal specimens and HNIS-SUH35 friction welded specimens. HINS base metal specimens had superior fatigue strength in comparison with SUH35 specimens. The fatigue life of the welded specimens was equivalent to that of SUH35 specimens. The fracture point of the welded specimens was at SUH35 steel, 2-3mm apart from the weld interface. The fatigue life of the welded specimens was estimated from the strain concentration calculated in the finite element analysis. The fracture point of the welded specimen was well explained by the Mises' equivalent strain concentration at SUH35 steel. The low cycle fatigue life of a welded specimen was predicted from the total strain range obtained in the finite element analysis. The predicted low cycle fatigue life was three times smaller than that in the experiments. The difference between the predicted and experimental fatigue lives was attributed to the compressive mean stress occurring in SUH35 steel. The life prediction using the parameter which took account of the tensile stress amplitude successfully predicted the low cycle fatigue life of the welded specimen.
Prediction methods of macroscopic and local stress-strain behaviors of perforated plates in plastic and creep regime are proposed in this paper, and are applied to the creep-fatigue life prediction of perforated plates. Both equivalent-solid-plate properties corresponding to the macroscopic behavior and the stress-strain concentration around a hole were obtained by assuming the analogy between plasticity and creep and also by extending the authors' proposal in creep condition. The perforated plates which were made of Hastelloy XR were subjected to the strain-controlled cyclic test at 950°C in air in order to experimentally obtain the macroscopic behavior such as the cyclic stress-strain curve and creep-fatigue life around a hole. The results obtained are summaried as follows. (1) The macroscopic behavior of perforated plates including cyclic stress-strain behavior and relaxation is predictable by using the proposed method in this paper. (2) The creep-fatigue life around a hole can be predicted by using the proposed method for stress-strain concentration around a hole.
This paper describes the crack growth direction and fatigue life behavior in creep-fatigue conditions under biaxial stress states. In order to examine creep-fatigue interactions, Mises' type equivalent stress controlled tests with the hold time of 0, 10, 30 and 60min. at the peak strain were carried out using the hollow specimens of austenitic stainless steel SUS304 at the strain rate 0.1%/s at 923K in air. The conclusions obtained were as following. (1) The main crack growth direction in the no-hold time test changed from the direction perpendicular to the maximum principal strain to the maximum shear direction at the principal strain ratio -0.7<φ<-0.8. However, in the tests with hold time the transition point of crack growth direction shifted to the reversed torsion side. (2) The effect of strain hold time on push-pull mode was more pronounced than that of reversed torsion. (3) By applying the linear damage rule obtained from the stress relaxation curve of Mises' type equivalent stress basis to biaxial creep-fatigue data, the fatigue lives of the reversed torsion tests were predicted and found significantly conservative because creep damage is over estimated. But those by COD based equivalent stress basis provided satisfactory prediction.
Crack propagation tests of sintered silicon nitrode were carried out under sinusoidal cyclic load at elevated temperatures. The stress ratios used were 0.5 and 0.1, while the test temperature ranged from 1073K to 1273K. The main results obtained are as follows: (1) Although the crack propagation rate with R=0.1 was higher than that with R=0.5 at 1073K, this tendency faded out at 1223K. (2) The crack propagation rate at R=0.5 increased with temperature, while that at R=0.1 did not change with temperature. (3) When the maximum stress intensity factor KImax decreased stepwise with R=0.1, the crack propagation rate decreased gradually, although KImax increased with crack length at each KImax revel. (4) Slight oxidation started at 1073K. The grain boundary glassy phase softened at 1173K and Si3N4 crystal was attacked severely by oxygen in the atmosphere at 1273K.