On the high frequency heating apparatus used for thermal fatigue tests, the relation between the shape of heating coils and the temperature distribution of specimens was experimentally examined for SUS 304 stainless steel. This report describes the relation between heating coil dimensions and heating rate, the most ideal coil shape for various specimen types, and the effects of temperature range and specimen shape on temperature distribution. The results obtained are as follows: (1) The heating rate increases with decreasing distance from specimen surface to heating coil and with increasing coil tube diameter. (2) For the general cylindrical and hour glass type specimens, the most ideal coil shapes are a separate type and a parallel type coil, respectively, as judged from the point of temperature distribution.
The effects of deformation in pipes and environmental differences around them on creep rupture strength are discussed after conducting creep tests in the uni-axial stress state and other tests under internal pressure of water by using a heat-resisting alloy, 15-15N. The analysis of the deformation in pipes was made based on the uni-axial elongation curves with consideration for the reduction of wall thickness. The environment tests were carried out by using pipes with both non-corroded and corroded innner surface by aqua-regia solution. Although the experiments covered a relatively short period of time, the following results have been obtained. (1) The uni-axial result of 15-15N shows a similar behaviour as that of an ordinary austenitic stainless steel. (2) The result of non-corroded pipes is not in agreement with the uni-axial results calculated by any of the existing creep formulas, but for the corroded pipes, the mean-diameter formula or the creep-common formula represents the result at elevated temperature (700°C) well. This shows that the crack initiation and propagation behaviours of both corroded pipes and bars are alike. (3) The results of numerical calculation of external diametral strain for the creep strain curves agree fairly well with the values obtained in the uni-axial experiments. (4) The notch-strengthening effect of corroded pipes in a short life range are well proven.
Creep rupture tests on tubular specimens under internal pressure at high temperatures give the most reliable information for tube design of boilers, because the tubes for boilers are subjected to multiaxial stresses in actual use but the maximum principal stress is tensile hoop stress. In this study, the creep rupture tests on the tubular specimens of carbon steel under internal pressure at high temperatures were carried out. Furthermore, the burst tests on the thickness-deviated tubular specimens were done in order to investigate the fracture behavior under internal pressure. The results obtained are summarized as follows: (1) The creep rupture properties of the carbon steel tubes produced by the electric resistance welding process, JIS-STB 42 EG and BS-ST 45 ERW, were proven not at all inferior to those of the seamless tubes. (2) In the internal creep rupture tests, all specimens ruptured at the base, not at the welded portion, and the slight deviation in wall thickess of a steel tube affected the location of initial rupture. (3) The thickness-deviated tube ruptured at the position of minimum thickness and showed a kind of the notch-strengthening. (4) The electric resistance weld tubes were recognized to have sufficient mechanical properties for use as boiler tubes from the results of material strength tests, microscopic observations and creep rupture tests under internal pressure on the tubes used for 7954 hours at a 250 MW boiler plant.
After 15 to 22 months' operation of a steam reformer furnace, cracks were detected at the butt weld between reformer tube (HK 40) and top flange (11/4 Cr 1/2 Mo steel) with Inconel 82 filler metal. The tube being jointed was 18mm thick and 144mm outside diameter. The cracks occured circumferentially along the fusion line between the weld deposit and CrMo steel. The metal temperature during the operation was approximately 500°C. The residual welding stress measurement at room temperature indicated that the circumferential stress is considerably higher than the longitudinal stress. Other stresses considered in the present study are the radial residual stress, which is estimated to be as high as the circumferential stress, and the thermal stress caused by the difference in the thermal expansion coefficient between the dissimilar metals. The stress due to the dead weight of materials and the one due to internal pressure can be neglected because of their low magnitudes. Laboratory creep rupture tests using specimens with the dissimilar metal weldment showed that the specimens ruptured at HK 40 base metal in the short time creep duration tests, while the specimens subjected to the longer rupture life tests always ruptured at the fusion line between weld deposit and CrMo steel as observed on the reformer tubes. The failure of butt weld joint can be explained in terms of the progressive change in stresses involved and rupture strength of the joint at various loads, as follows. Though the thermal stress and the residual stress are gradually relaxing during the service exposure, the rate of the stress relaxation is considered to be slower than the damage accumulation and thus the retained rupture strength for the weldment is decreased.
Thermal stress concentration at the slit or the crack in a finite region occurs in many engineering problems. For such problems, the use of the finite difference or finite element technique brings in a large number of linear simultaneous equations, and so solving these equations is very time-consuming. Especially, the case of approximate crack-formed slit is more difficult due to divergency of the solution and it is almost impossible to obtain the correct solution for that case. In this paper, therefore, the analysis is developed by means of the point-matching technique in the elliptic coordinate, and the solutions are obtained on the rate of stress concentration for slits having the ratio of the major axis to the minor axis of ellipse up to 35. Numerical examples are shown for the problems in a square plate with many different shaped types of slit formed by an elliptical hole under heat generation.
In this paper, the discussions are focussed on yield function to be applied to the general method for estimating the deformation due to multiaxial thermal stress ratcheting, which the authors had already proposed before. The yield function employed is the one which enables us to consider the effects of anisotropy and cyclic hardening of material and the Bauschinger's effect as well as the influence of temperature; the anisotropy of material and the cyclic hardening are considered to be caused by the variations of the shape and the magnitude of yield surface, respectively, while the translational change of yield surface contributes to the Bauschinger's effect. Deformations of hollow cylindrical specimens of low carbon steel subjected to cyclic axial thermal stress with steady stresses of combined tension and torsion were detected in order to examine the propriety of the analytical procedure. The effect of torsional stress on the ratcheting strain is discussed analytically, and the predicted ratcheting limit diagram of the material is also included.
This paper presents a technique to determine the thermal shock resistance of graphite by the basic analysis of nonsteady thermal stress in a circular disk heated by an arc discharge at its central area. The thermal shock resistance, defined as Δ=σθk/Eα, (σθ: circumferential tensile thermal stress, k: thermal conductivity, E: Young's modulus, α: thermal expansion coefficient), can be determined en bloc only by measuring the arc discharge electric power using the specific non-dimensional thermal stress S* which is calculated theoretically as the maximum saturation stress in the nondimensional diffusion time τ>1/4. This technique is also applicable to measure simultaneously the thermal diffusivity κ=k/cρ (c: specific heat, ρ: density). The experimental studies for several kinds of graphite were carried out and the results were compared with the values of σθk/Eα and k/cρ estimated indirectly from these individual mechanical and physical properties.
Temperature dependence of the rate of fatigue crack propagation was examined, and compared with the temperature dependence of tensile ductility. As numerous microcracks initiate and affect the propagation behavior of the main crack at elevated temperatures, the following factors were found to be elucidated; initiation rate of microcracks, reduction of ductility of the material in the vicinity of the main crack tip, and relaxation of concentrated strain by multi-cracks. The results obtained are summarized as follows: (1) During a strain controlled low cycle fatigue test at 1cpm, the rate of crack propagation is largest at the blue-brittleness temperature range (200-300°C) in a low carbon steel. On the other hand, it is largest at above 700°C in austenite stainless steels. The temperature dependence of the rate of fatigue crack propagation is opposite to that of tensile ductility. (2) Microcracks formed in the vicinity of the main crack tip were calculated, by considering the strain concentration and strain cycles imposed. Then, the local fracture strain was evaluated. There found a good correlation between the rate of crack propagation and the local fracture strain.
Observation was made on cracks on the surface and on the cross-sections of austenitic stainless steel specimens (Type 304, 316 and 321) subjected to a low-cycle fatigue test at high temperatures. The specimens were cycled in axial loading under a constant total strain amplitude. The test variables were the total strain range of 0.5 to 2.5%, the strain rate of 0.4 to 40%/min., and the temperature from room temperature to 800°C. The results obtained are summarized as follows: (1) Initiation of cracks on the surface occurred at the twin boundaries at R.T. and 450°C, and at the grain boundaries at 700°C. (2) Surface cracks initiated at the twin boundaries where slip bands were formed parallel to them or impinged on them with angles. (3) Surface cracks observed on the fractured specimen propagated along the grain boundaries when the temperature was higher, the strain rate was lower and the strain amplitude was larger. The relative magnitude of creep damage per cycle showed a correlation with the mode of crack propagation. That is, the larger the creep damage is, the more the mode of propagation tended to be intergranular than transgranular. (4) Cracks observed on the cross-section of failure specimens were dependent on crystal orientations at 450°C and 600°C (ε=40%/min).
As the basis for designing the fuel assembly for a sodium cooled fast breeder reactor and for determining the service condition of the reactor core, it has been required to obtain creep and creep-rupture properties of the domestic seamless tubes of AISI Type-316 stainless steel with the nominal dimension of 6.3mm outside diameter and 0.35mm wall thickness in flowing sodium environment under various conditions. For the examination in sodium environment, creep testing apparatuses were designed and incorporated with the two sodium loops for material tests. The sodium handling technique and the testing method up to 700°C have been established. The effects of oxygen level in flowing sodium on creep-rupture properties of tubular specimens of Type-316 stainless steel have been examined. At the oxygen levels of 5 and 10 wppm in flowing sodium, the values of creep-rupture strength under uniaxial tension at 600°, 650°, and 700°C up to about 2000hr (the longest rupture time) are not significantly different. The rupture strengths under internal pressure and under uniaxial tension are significantly lower in flowing sodium containing 10 wppm oxygen at 700°C than in air at the same temperature. Furthermore, the decrease of rupture strength is greater as the exposure time in the sodium is longer, and it is related to the tendency of the relative increase in the minimum creep rate. This apparent reduction in rupture strength is accompanied by the significant decarburization of specimens tested, and the formation of a degraded layer in the neighbourhood of the specimen surface exposed to sodium.
In order to clarify the long-term stress relaxation behaviour, the stress relaxation tests up to one year have been carried out on a batch of domestic 1Cr-0.5Mo-0.25V high temperature bolted steel used for a steam turbine at the test temperatures of 450°C, 500°C and 550°C for the total strains of 0.10%, 0.15%, 0.20% and 0.25%. The influence of the total strain on the residual stress was discussed, and the relationship between creep and stress relaxation was analysed using the creep constitutive equations based on strain hardening, time hardening, exponential and hyperbolic-sine laws. The results obtained are summarized as follows: (1) The influence of the total strain on the residual stress tended to decrease with increasing test temperature and test period. Within the range of present test temperatures, there was little influence of the total strain on the residual stress when the total strain was raised over 0.15% and time was prolonged over 1000hrs. (2) The estimated stress relaxation curves obtained from the creep data at the transient stage by applying the strain hardening theory almost agreed with the observed ones. However, the estimated residual stresses were slightly weaker than the observed ones at every temperature and every value of the total strains.