[in Japanese]

View full abstract

Despite the extensive studies on the stress-relaxation properties of heat resisting steels, there are several questions left unanswered, in which, for example, the subjects on the extrapolation of stressrelaxation data and effect of re-loading on the stress relaxation properties are included.
The authors have taken interest in these subjects and have been carrying out the stress-relaxation tests on TAF which is 12% chromium heat resisting steel.
Tests were made with the specimens of a 10mm guage dia. and 100mm gauge length by using the automatic type stress-relaxation testing machines of 4 tons capacity, and were processed by the following methods: (1) Applying the same initial stress of 30kg/mm² at the temperature range of 550°∼650°C (this test procedure was termed as“initial-loading 1st”); (2) The specimens subjected to the foregoing procedure being re-loaded to the same initial stress of 30kg/mm² at the same temperature range (termed as“re-loading”); and (3) On the other hand, to another group of specimens applying the initial strain equal to the residual plastic strain left in the“initial-loading 1st”plus initial strain in the “re-loading”(termed as“initial-loading 2nd”).
The results obtained are as follows;
(1) As far as the present tests are concerned, the predicted value of“re-loading”stress-relaxation properties assessed from the data of“initial-loading 2nd”based on the well-known Strain-Hardening Theory deviates significantly from the actual“re-loading”value.
On the other hand, the predicted value calculated from Time-Hardening Theory by using suitable material constants is in comparatively good agreement with the actual one within the range of observed duration.
(2) The equation which correlate plastic strain rate ε_p calculated from“initial-loading”stress-relaxation data for TAF steel, residual stress (kg/mm²), and test temperature T (°K), within the temperature range of 800°∼900°K and the stress range of 24∼14kg/mm², takes the form;
ε_p≈A exp(ασ)exp(-Q'/RT)
where
Q'={124(kcal/mole)……σ_c>σ
267-7σ(kg/mole)……σ_c<σ
where A and α are material constants which are determined separately according to the two stages definend by stress level (corresponding to the two straight portions in log ε_p-log σ curves), and σ_c is the critical value of about 21(kg/mm²).
(3) All the initial loading stress-relaxation data obtained at various test temperatures on TAF steel have been successfully expressed on a single master relaxation curve within the limit range of stress and temperature by using the parameter P (=log Et-Q4.6 T, where E is Young's modulus depending on temperature, t is time, T is absolute temperature, and Q is material constant). This relationship could be also available successfully to other alloys.

View full abstract

The creep properties in bending such as the lateral creep deflection versus time relations of a four-parameter model are investigated for the fluctuated temperature profile alongside the column length. The temperature fluctuation is assumed to be of asymmetric stepwise pattern, incessantly alternating around the average temperature just like an alternating current. The column considered is for simplicity, assumed to have an idealized section. The applied axial load is also assumed to be near an Euler load, and so the deflection curve has a single harmonic sine. The temperature fluctuation treated here is confined to be of small amplitude, so that the modulus of elasticity may be taken as left unchanged. The viscous term, representing the creep phenomenon, is expressed in terms of rate process to take account of temperature effects. The numerical example for polymethylmethacrylate shows a conspicuous degradation of life time even for a trifle amount of temperature fluctuation, say, about 30% reduction for only ±1.75% fluctuation. The temperature fluctuation amplitude therefore, should be exactly detected especially for such a linear viscoelastic column.

View full abstract

It was the aim of the present study to elucidate the influences of both hydrostatic stress and strain history on creep deformation of polycrystalline metallic materials at elevated temperatures. In the present paper, the special interest was taken in studying the correlation between the effect of hydrostatic stress and that of strain history through the study on plastic potential for metallic creep.
From the flow rule based on the yield function f=D(J₁)+H(J₂', ∫dε_ij), the following conclusion was made, J₁ being the first invariant of stress σ_ij, J₂' the second invariant of deviatric stress S_ij, and ∫dε_ij the strain of the material experienced during creep. First regarding the influence of hydrostatic component of stress on the creep, the comparison between both the test data under simple tension and simple torsion was cited as the basic examination. The ratio r of simple tensile creep rate ε_z to shearing creep rate γ at the same equivalent strain ε was given as
γ≡(γ/√3)/ε_z=1-1/4Aε²+3/4A²ε²/3D+1+Aε+3/4A²ε²,
where A is a scalar anisotropic parameter and D a scalar parameter of the volumetric stress. Therefore. it was found that r took the numerical value of r_??_1 under D_??_0 and A_??_0. D≈0 and A>0 was examined in the present types of tests of 0.15per cent carbon steel at 450°C. It was also ascertained that the decrease in the value of r in progress of creep resulted from the development of anisotropy of the material during the creep.
Secondly, under general loadings of rotated principal stress axes and fixed equivalent stress, the development of anisotropy of 0.15per cent carbon steel in the course of creep was remarkable than that under the fixed principal stress axes. It was also much dependent on strain history. As a matter of fact, the numerical values of anisotropic parameter and the parameter of Bauschinger effect of the material varied discontinuously after changing the stress components. Therefore, the translation of location and the form of the yielding curve were also ascertained. This resulted in the discrepancy between the principal direction of stress and that of creep rate, and caused disagreement in creep curves of tests which were performed under different load paths.
Thirdly, regarding the effect of combined loading of hydrostatic pressure p on the creep, the ratio of axial strain rate under the confining pressure ε_z to that at atomosphere ε_zo of the material which experienced the strain history of ε was given as
ε_z/ε_zo=1-9D/3D+1+Aε+3/4A²ε²·p/σ_z,
where σ_z is a fixed axial tensile stress. Therefore, the ratio took the numerical value of ε_z/ε_zo_??_1 under D_??_0 independent of the sign and the numerical value of A. It was also found that the ratio was not only dependent on D but also on the development of anisotropy of the material, that is, Aε.

View full abstract

For the study of the effects of hydrostatic pressure and temperature on mechanical behaviour of metallic materials, a high-temperature tri-axial testing machine was designed and constructed. By using this machine, it is made possible to carry out tension or compression tests at elevated temperatures up to 600°C under hydrostatic pressures up to 5000kg/cm².
The results of tension tests on carbon steel (0.14%C) and titanium show that the strength of both materials increases with hydrostatic pressure within the range of the tests (room temperature to 500°C, atmospheric pressure to 5000kg/cm²). The effect of hydrostatic pressure on the ductility of titanium is remarkable over the whole range of test temperature. On the other hand, the effect on the ductility of carbon steel is positive up to the temperature of about 300°C, but the effect on elongation decreases beyond that temperature and the effect on reduction in area becomes negative.

View full abstract

Concerning the problems respecting the creep under non-steady multiaxial stress conditions, there are two cases in the state of stress encountered. The first is the case in which the principal stress directions are kept constant, and the second is the case in which those are variable with varying stresses. In order to investigate a creep behavior under such conditions, the creep of thick-walled cylinders of low carbon steel and 21/4Cr-1Mo steel under stepwise changing internal pressure of periodic rectangular wave was studied. This is one of the fundamental problems on non-steady multiaxial stress conditions under constant direction and constant ratio of principal stresses. Furthermore, such a situation is present in boiler tubes and pressure vessels subjected to internal pressure. It is necessary, therefore, to consider the problem with the aim of predicting the creep strength of thick-walled pressurized cylinders under varying stress.
As the results of the experimental and analytical study. It was found that the law of“the mechanical equation of state in solid”was not valid for the creep of the materials tested both in the case of varying uniaxial tension and also in the case of varying multiaxial stresses. The strain rate at the period of reloading was about 20% higher than that under steady load, and the rupture life in the tests under cyclic stress was about 20% shorter than that of the reference tests under constant stress. In the stage of steady state creep, creep recovery came to an almost constant value independent of the amount of strain. These tendencies were quite the same in the creep of the thick-walled cylinders under varying internal pressure. It is considered that residual stresses in the cylinder at the period of pressure removal have little effect on the stress distribution after reloading, and it seems to be appropriate to assume in the analysis that there is no rotation in the principal directions of creep rate and that the material is isotopic under cyclic loading and unloading. This leads to the conclusion that the creep strength under non-steady multiaxial stress condition of constant stress direction and constant ratio of principal stresses can be estimated from the results of the creep tests under varying uniaxial tension by the use of the multiaxial creep theory based on the relationship between the static creep strength of the uniaxial stress and that of the multiaxial stresses.

View full abstract

Thermal fatigue and thermal cycling tests of 18-8 stainless steel were performed to study the effects of cyclic thermal strain and cyclic heating on the change of its hardness during the thermal fatigue. In this study, the thermal fatigue and thermal cycling tests were carried out by means of Coffin's type thermal fatigue testing machine, under conditions of high-frequency induction-current heating, (the heating rate being at 67°C/sec and holding time at maximum temperature 5sec), and water cooling (the minimum temperature being at 20°C). After testing to a certain number of thermal cycles, the hardness of the specimens was measured by means of micro Vickers hardness tester, and the relations between these hardness changes and grain size number were investigated. The conclusion can be drawn as follows.
In the range of temperature variations from 600 to 800 deg (C), the hardness change by thermal cycling or precipitation-hardening showed the values of 5 to 40 per cent of total hardness change of thermal fatigue, and the hardness change owing to thermal strain cycling or the resultant of strain hardening and its softening indicated the values of 50 to 90 per cent. Consequently, in the final stage of thermal fatigue, the hardness change was considered as summing of these hardness changes, but in the initial stage, the effect of softening of the material by recrystalization and grain growth was obvious.

View full abstract

It is well known that the creep strength of low carbon steels is affected by N, Mn and Si contents as well as heat treatment. But the mechanism by which these factors affect the creep strength of carbon steels is still somewhat unclarified.
The authors made creep rupture tests of low carbon steels with various chemical compositions to study the effect of Al, N and Mn on the creep strength of low carbon steels, and observed the structural changes during the creep by transmission electron microscopy.
The behaviour of N during creep was also investigated with the internal friction method and the mechanism by which Al, N and Mn affect the creep strength of low carbon steels was discussed.
Results are summarized as follows:
(1) Creep rupture strength:
(a) Effect of Al:
Al addition drastically reduces the creep rupture strength of low carbon steels, but Al not exceeding 0.015% lowered it only moderately. An Al-killed steel with 0.039% Al had the lowest creep rupture strength.
(b) Effect of N:
A vacuum melted steel with 0.002% N showed the lowest strength only second to the Al-killed steel.
The higher the N contents were increased the higher the creep rupture strength rose, but the addition of N exceeding 0.005% had relatively small effect.
(c) Effect of Mn:
The increase in Mn contents from 0.5% to 1.2% caused small decrease in the creep rupture strength.
(d) Effect of Mn-N:
When both Mn and N contents simultaneously increased, the creep rupture strength was greatly improved. A steel with 1.23% Mn and 0.02% N showed the highest creep rupture strength among the steels investigated.
(2) Transmission electron microscopy:
(a) Vacuum melted, low N steel:
The cell structure was formed during the creep deformation. The progress of the cell formation was different from one grain to another, from ones in which cells were formed in the primary stage to the others in which cells were not formed until the tertiary stage. The number of grains in which cells were formed increased with the increase in creep strain, and when the specimen ruptured, cells were formed in all the grains.
The cell size decreased to about μ in the primary stage and almost unchanged during the secondary stage, followed by small decrease in the tertiary stage.
The dislocation density within a cell was nearly constant during the secondary stage and rapidly increase with the steep increase in strain during the tertiary stage. The dislocation density within a cell was lower than that within a grain in which cells were not formed, suggesting that some dislocations were annihilated at the cell boundaries during the cell formation.
(b) Al-killed steel:
Fine, coherent precipitates, about 150Å in size and presumedly AlN, were observed in a normalized state. Although, at room temperature, the precipitates seemed to contribute to the strength of this steel, during creep at elevated temperatures, the precipitates were rapidly coursened and seemed to have no contribution to the creep strength.
The precipitation decreased in the amount of N in solid solution and badly reduced the creep strength.
(c) High Mn-high N steel:
The cell structure was not observed even when the specimen ruptured.
Fine, coherent precipitates, about 100Å in size, were observed after the early half of the secondary stage of creep and also after about 800 hours of aging at 450°C.
Similar precipitates were observed in a low Mn-high N steel and a high Al-high N steel.
(3) Internal Friction measurements:
Snoek damping of the high Mn-high N steel was measured using about 400cps vibration during creep.
During the primary stage, while the creep rate decreased, the Snoek peak was as high as the initial value, suggesting that almost all of N remained in solid solution.
The Snoek peak began to decrease in the early half of the secondary stage, not accompanied with any marked change in the creep rate.

View full abstract

Beryllium has long been noted as nuclear reactor materials due to its excellent nuclear properties and thermal and mechanical properties for high temperature. In submitting beryllium to the serious stress conditions at reactor core, however, the brittleness of this material even at comparatively lower temperature has been the most urgent problem with beryllium as fuel cladding material in the high temperature gas cooled reactor. However it does not seem to have been made clear yet what can contribute to solving the problem so as to bring out its strength and ductility to full utility. In this study, test specimens were produced from the sheets of hot pressed and unidirectionally rolled beryllium. Then two factors, i.e. the three grain sizes in the powder metallurgy of typical electrolytic flake beryllium and the anisotropy by the hot rolled fabrication were considered as affecting the mechanical properties of beryllium at elevated temperatures. The experiments were conducted to measure and discuss the various mechanical properties as a function of temperature and neutron irradiation effect on the two factors.

View full abstract

The hot extruded beryllium of reactor grade purity was studied at elevated temperatures with respect to its tensile, hardness, elastic modulus and creep properties. The thus obtained results have been discussed in comparison with out previous report regarding hot rolled beryllium.
The hardness was regarded as a macroscopic measure for the temperature dependencies of critical slip mechanisms in single crystal.
The elastic modulus up to 1000°C were determined by means of an ultrasonic pulse propagation speed.
The creep tests were carried out in two ways, namely the stress increment method at constant temperatures among the room temperature to 700°C and the temperature cycling method of the triangular temperature variation between 350 and 600°C at constant loads. In the stress increment creep, the steady creep rate and the rupture life in the Larson-Miller diagrams agree reasonably with the corresponding data at The Brush Beryllium Co. by ordinary method. The relation of steady creep rate ε and the cycles-to-rupture N_R in the temperature cycling creep was expressed as εN_R^0.70=constant. This equation is similar to the formura on strain controlled fatigue. The considerations in the temperature cycling creep of this material have been discussed in comparison with constant temperature creep.

View full abstract