The distributions of stress and strain in some notched plates or bars in creep had been already calculated by means of the finite element method. Based on these previous results, the behavior of stress relaxation at the notch root and the relationship between stress or strain concentration factor of creep and elastic stress concentration factor were examined in the present study. The stress at the notch root relaxes from the initial elastic stress in the course of creep, and reaches its steady state. The time for the stress to reach the steady state is the time period until the magnitude of creep strain becomes equal to the magnitude of elastic strain under the same nominal stress in the plane stress condition. It is also the time period until the creep strain becomes about ten times of the elastic strain in the plane strain condition. The stress and strain concentration factors for the steady state of stress in creep can be estimated through the elastic stress concentration factor by using the Neuber's equation in the case of plane stress. However, the Neuber's equation overestimates the magnitude of the stress and strain concentration factors for notched bars. The following equations seem to be applicable for the relationship between stress or strain concentration factor, Kσ or Kε, and elastic stress concentration factor, Kt. Kσ=√1-k+k2Kt2/(α+1), Kε=(√1-k+k2)αKt2α/(α+1), where k=σzmax/σymax, √1-k+k2=1 in plane stress and 0.866 in plane strain, and α is a creep exponent.
The influences of specimen geometry and temperature upon the creep crack growth rate of 1Cr 1Mo1/4V steel casting have been studied. The main results obtained are summarized as follows. (1) In the SEN specimen without side grooves, the creep crack tends to tunnel at its center, and consequently, it is difficult to measure the creep crack growth rate accurately. However, in the SEN specimen with proper side grooves, the creep crack front is kept straight during its propagation without tunneling, and it is easy to measure the crack growth rate accurately. (2) The relation between the creep crack growth rate and temperature can be described by the equation of so-called Arrehenius type and is similar to the relation between the steady creep rate and temperature in the temperature range of the present experiment.
Low cycle fatigue tests of cobalt-base superalloy HA-188 were conducted at 700°C in the strain rate range of 3.6×10-5 to 2.0×10-2 sec-1. Experimental results were compared with estimated results in consideration of fracture mode. The initiation and the propagation of the main crack was observed during fatigue test by using the surface replica method. The results obtained in this study are summarized as follows. (1) Fatigue life decreased with decreasing strain rate, and the slope of log Δεp-log Nf were-0.66 at any strain rate. (2) In the universal slope method, experimental results were distributed between average band and lowerband, and fallen within a factor of two of the results estimated by an equation based on the Ohtani-Nakamura's equation. (3) Relation between experimental results and estimated results was not affected by fracture mode in the universal slope method, but affected in equation based on the Ohtani-Nakamura's equation. (4) The ratios of Ni/Nf (Ni: crack initiation life) were decreased as the strain rate and the strain range decreased.
Long-time creep rupture tests up to 40000hr have been carried out at 400°-500°C on nine commercial heats of silicon-killed carbon steel for boiler tube (JIS STB42). The data shows a large scatter from heat to heat and from product to product. The main factors affecting the scatter have been studied and the results obtained are as follows: (1) The scatter of ±20per cent in rupture strength has been observed. By taking into consideration that the nine heats are from nominally same steel, the scatter is considerably large. In addition, the scatter doesn't become smaller at higher temperatures nor with increasing time to rupture. (2) All the 100000hr rupture stresses extrapolated by the Larson-Miller method are higher than the average stress calculated from the allowable stress in JIS. (3) Rupture strength at higher temperatures and for longer time becomes higher in proportion to the increase of molybdenum content from 0.005 to 0.019per cent. (4) At lower temperatures and for shorter time, the variations in the molybdenum content and the nitrogen content in solution are the main causes of the scatter. With increasing testing time the nitrogen content in solution is considerably reduced owing to the formation of nitride with a corresponding deterioration in rupture strength. (5) Although the effect of normalizing temperature on rupture strength is slight, tempering causes the rupture strength at lower temperatures and for shorter time to deteriorate due to the decrease in the nitrogen content in solution.
Notch effects on crack initiation in creep were studied using center notched plate specimens of SUS304 austenitic stainless steel and 99.96% Cu. The calculated results by the finite element method were adopted to aid the analysis of creep crack initiation. It was found that the fracture strain on the specimen surface at the notch root is equal to the rupture strain in unnotched specimens. When the stress at the notch root is presented as Kt2/(α+1)σn, where Kt is an elastic stress concentration factor and α is a creep exponent, the time to crack initiation, ti, can be estimated from the unnotch rupture life, tr, i.e., (Kt2/(α+1)σn)αlocti=σnαloctr≡C, where αloc and C are material constants.
In providing the basis of designing structural components for the secondary sodium cooling system of a fast breeder reactor and for determining the optimum service condition of these components, it has been required to obtain creep and creep-rupture properties of 21/4Cr1Mo steel in sodium and also to clarify effects of sodium corrosion on these properties. The creep and creep-rupture tests under uniaxial tension were made using the specimens extracted from seamless tubes of normalized and tempered 21/4Cr1Mo steel of domestic heat at temperatures of 500°, 550° and 600°C in flowing sodium with the oxygen level of 10wppm. The similar tests were carried out at the same temperatures in air for the purpose of comparison. The similar tests were made also in air at 550°C for the specimens exposed to sodium under various conditions. Corrosion rate and carbon loss were examined simultaneously for these specimens exposed to sodium. The results obtained are summarized as follows: (1) At 500° and 550°C, the values of creep-rupture strength in flowing sodium and those in air were not significantly different. The rupture strength at 600°C, however, was significantly lower in the sodium environment than in air. The decrease of rupture strength became greater as the exposure time in sodium was increased, and coincidence was noted with the tendency of the relative increase in the minimum creep rate. This apparent reduction in rupture strength was accompanied by the significant increase in carbon loss due to decarburization of the specimens tested. (2) The creep-rupture strength in air at 550°C for the specimens after exposure to flowing sodium was found to reduce almost proportionally to the carbon loss in the specimens of 0.10 to 0.03w/o. (3) The rate of average metal loss in 21/4Cr1Mo steel was calculated to be 1.2±0.1μm/year from the values of weight loss after the exposure to flowing sodium containing 10wppm of oxygen for about 10000hrs in. the temperature range of 545° to 560°C.
Alloy HH is being used more than other heat resistant steels and especially useful as material for trays. This alloy can be classified into Type I in which ferrite is mixed and Type II of perfect austenite. At present, however, the alloy is unable to be selected for each use according to its use condition. From this point of view, the authors have been examining its basic properties of thermal fatigue. In this report, described are the relation between structural change and thermal fatigue strength and furthermore, studies on crack propagation, the effect of casting defects against thermal fatigue strength and the life estimation. The results obtained are as follows: (1) The effect of maximum temperature appears in Type I, The thermal fatigue life greatly lowers on the high strain range side if the temperature rises as much as 100°C from 850°C. (2) The grain size dependence of thermal fatigue strength of this material is scarcely found when the grain size number changes within the limit of -1.5 to -4.5. (3) In less than about 1×103 cycle range, thermal fatigue life of Type II is longer than that of Type I. This is due to the distribution of eutectic carbides and high temperature strength of matrix. In more than 1×103 cycle range, thermal fatigue life is governed by secondary carbides which make the life of Type I and II equal. (4) The characteristics of crack propagation is formulated by dl/dN=C(Δεp√l)m and the life estimation regarding the slag inclusion size as initial crack length shows favorable agreement with experimental data.
It was shown that thermal fatigue properties of HK40 and 25Cr-35Ni centrifugal cast tubes for petro-chemical industry degrade due to the change of the material properties when used at elevated temperature of about 1000°C. Based on our study performed with the laboratory aged materials, such degradation of thermal fatigue properties of used materials can be estimated from their tensile properties.
The rotating bending fatigue tests were performed on the notched carburized steel specimen at room temperature, and their properties of fatigue crack propagation were investigated. Furthermore, the relation between fatigue crack behaviours and microscopic structure was discussed. All the SCM21 specimens were carburized at 930°C, under 0.9% carbon potential atmosphere for 4hrs, and oil-quenched from 830°C. Tempering was carried out at 180°C for 2hrs. The results obtained are as follows. (1) The non-propagating crack was observed at 107 cycles in the case of SCM21 carburized steel. The maximum notch radius of ρ=0.1mm to 0.08mm was essential for the existence of non-propagating cracks at 107 cycles. (2) In the case of ρ=0.08mm, the crack propagation rate was almost constant or slightly decreasing during a considerable part of its life. That tendency could be understood by considering the distribution of compressive residual stress in the surface layer. (3) In the case of ρ=0.5mm, the crack propagation rate increased proportionally to the crack length. (4) In the case of ρ=0.08mm, the fatigue crack initiated at the grain boundary of prior austenite. It propagated along the grain boundaries of prior austenite at a high stress amplitude level and through the grains of prior austenite at a low stress amplitude level.
In order to prevent brittle fracture of a structure, it is important to know how much the brittle transition temperature of a steel structure increases with cyclic loading. In a past, a number of investigations were carried out, for example, by means of the Charpy impact test. However, quantitative evaluation of the influence of fatigue stressing and cycling on the brittle transition temperature of steel have not yet been made clear. In this paper, the brittle transition temperature of a structural steel subjected to various degrees of fatigue was measured by the hardness testing method in which the brittle transition temperature can be defined as the intersection of the two straight lines in the plot of the logarithmic Vickers hardness vs. the reciprocal of absolute temperature. The results obtained are summarized as follows: (1) When a constant fatigue stress S is repeated, the transition temperature of hardness ΔT* increases with increasing number of stress cycles N and saturates to a certain value. (2) When a constant number of stress cycles N is repeated, ΔT* increases with increasing repeating stress S. (3) The maximum increase of brittle transition temperature ΔT*max appears in the fatigue limit. (4) An experimental formula relating ΔT* with S and N is found to satisfy fairly well the measured data.
The behavior of fatigue crack initiation in a deep notched specimen of Ni-Cr-Mo high strength steel is studied using the apparent stress intensity factor Ka which is obtained by substituting the notch length for the crack length in the equation of stress intensity factor K. The results obtained are as follows: (1) When the notch root radius ρ is constant, the number of cycles at fatigue crack initiation N is decided by the range of apparent stress intensity factor ΔKa regardless of the notch length a0. (2) At constant ΔKa, N decreases with decreasing ρ. The thickness of specimen, B, has no effect on the relation between ΔKa and N. (3) The notch opening displacement Φ depends only on the apparent stress intensity factor Ka, but not on a0 and ρ. (4) The fatigue crack initiation life of a notched specimen with arbitrary a0, ρ and B under arbitrary ΔKa can be predicted from the relation between the strain range and N of smooth specimens, by relating Φ with the strain at the notch root.
This paper is concerned with the time- and temperature dependences of the flexural strengths of the plain fabric glass cloth reinforced laminates (GRP) and the unsaturated polyester resin used as the matrix of the GRP. These properties were measured by constant strain rate tests and creep rupture tests in the temperature range from 60°C to 150°C. The results obtained are summarized as follows: (1) The master curves of the flexural strengths for both the matrix and GRP can be constructed, using their thermo-rheological simple properties. As for the time- and temperature shift factors, they are in good agreement each other quantitatively, and also coincide well with those obtained in the previous paper concerning with the creep compliances of these materials. (2) It was found that the relation between the flexural strengths of the matrix and GRP can be uniquely determined, independent of the conditions of their circumstance such as temperature and strain rate (time). (3) The master curve of the flexural creep rupture strength of GRP vs. time to failure is in good agreement with that of the flexural strength vs. strain rate. (4) From the above-mentioned results (1) and (3), the creep rupture strength of GRP is easily estimated from the time-temperature shift factors measurement for the matrix and the constant strain rate tests for GRP.