Hereunder is presented a procedure to analyze elasto-plastic thermal stresses by the finite element method. On the finite element analysis of thermal stresses, the equivalent load matrix due to thermal expansion is employed, as the so-called external load matrix. The stiffness matrix in the plastic region is constructed according to Yamada's [D^p] matrix, which is derived from Prandtl-Reuss equation and von Mises yield criterion. A modified Marcal's method gives appropriate stiffness to the element adjacent to the elasto-plastic interface.
Attension is to be focused on the analysis of thermal residual stresses. The steps in calculation of the residual stresses are as follows:
(1) The elasto-plastic stresses and strains will be determined.
(2) The plastic strains will be estimated.
(3) The residual stresses will be calculated so as to compensate the remaining plastic strains in the body.
Numerical calculation was made of the thermal residual stresses in a cylindrical bar induced by the partial induction quenching. The temperature distribution in the heating and cooling processes was determined by means of the finite difference method. The residual stresses induced in the present case are found to be resultant of two kinds of plastic deformation; one made during the initial heating and the other during the subsequent cooling.
The method of calculating the residual stresses by such a technique is applicable to repeated loading.

抄録全体を表示

The behaviours of dynamic recovery and work-hardening of alpha iron and its molybdenum alloys were investigated by stress relaxation and tensile tests during high temperature creep. It has been found that dynamic recovery rates in the steady-state creep depend strongly on temperature, creep stress, and molybdenum content. On the other hand, the work-hardening rates hardly vary with creep stress and molybdenum content, depending but slightly on temperature through the temperature variation of the elastic modulus. The creep rates which were calculated from the observed dynamic recovery and the work-hardening rates in terms of the idea of recovery creep are in good agreement with the observation. From the results as mentioned above it is concluded that the creep behaviours of alpha iron and its molybdenum alloys are controlled mainly by the dynamic recovery process in which the diffusion of atoms plays an important role.

抄録全体を表示

The relation between creep rupture strength and the electron microstructures of 2 1/4% Cr-1% Mo steel in heat-treatment conditions were investigated. The results are summarized as follows:
(1) Curves of creep stress-time are linear or almost linear in stabilized and semistabilized structures and that of unstabilized structure shows knick point in curve at 600°C, 1000hrs.
(2) There is no change in the microstructure of stabilized specimens during the creep test at 600°C, but that of unstabilized specimens changes to match the gradient of stress-rupture curves.
(3) The interactions between one dislocation and another or between dislocation and the precipitates are important factors of creep rupture strength.
(4) The activation energy of creep calculated by Zener-Hollomon parameter, agrees well with that of diffusion of Cr atom when the structure is stabilized by long time annealing.
(5) The recovery and recrystallization are observed at creep deformed zones preferentially.
(6) From the facts (4) and (5), the creep deformation is supposed to be accelerated by recovery and recrystallization which are processed by dislocation movement accompanied with diffusion of Cr atom.
(7) The long time rupture strength of the test specimen will be easily and accurately estimated by annealing it for a long time and so stabilizing its structure. The plot calculated by Dorn parameter with the experimental values of the stabilized specimen agrees with Dorn master curve.

抄録全体を表示

Hereunder is given an account of the uniaxial tension and internal pressure creep rupture tests conducted of the joints of 2. 25 Cr-1 Mo, 5 Cr-0.5 Mo, 9 Cr-1 Mo low alloy steel tubes, and of 18-8 Mo, 18-8 Ti, 18-8 Nb stainless steel tubes, welded with dissimilar metals, to investigate the effect of the weld defects, and of the decarburized layer, on the internal pressure creep rupture strength of the welded joints, as well as the relationship between the uniaxial and the internal pressure creep rupture strength.
The results are summarized as follows:
(1) When poor penetration is formed at the welded joints, fracture often occurs in the welded metal under the internal pressure, and leaves the internal pressure creep rupture strength poor.
At sound joints, fracture always occurs in the base metal in the internal pressure test, and the strength of the welded joints is equal to that of plain tubes.
(2) A decarburized layer is formed at all dissimillar metal joints, regardless of electrodes, stainless steels or Inconel alloys.
At the joints of 2.25 Cr-1 Mo and stainless steels with 24 Cr-13 Ni stainless steel electrode, fracture occurs in the decarburized layer in uniaxial tests lasting for a longer time than 700∼2000 hours, but in the internal pressure tests, fracture occurs in the base metal even after a longer duration. At other joints, fracture always occurs in the base metal both in the uniaxial and the internal pressure tests.
Thus it is evident that the internal pressure creep rupture strength of joints with dissimilar metals is not affected by the formation of a decarburized layer.
(3) The uniaxial and internal creep rupture data of the plain tubes and the welded joints which fractured in the base metal, are analyged by the statistical method to determine the relationship between the uniaxial and the internal creep rupture strength.
The mean diameter formula is fairly applicable to the result of 2, 25 Cr-1 Mo tubes, and of 9 Cr-1 Mo tubes, either in the received condition or in the stress relieved condition, while the ASME or the Tresca formula is applicable to the result of 5 Cr-0.5Mo steel tubes.

抄録全体を表示

A 48- or 72-pointed creep-rupture tester has been developed, of a sort far more multiple pointed than conventional testers. This is of a spring loaded type in which displacement of spring by elongation of specimens is manually adjustable. Actually the integrated accuracy of test results is identical to, or even better than, those by conventional single type testers, due to its good temperature distribution within the furnace in spite of the fact that there is some fluctuation of load caused by manual adjustments. These results were confirmed by testing five specimens for from 200 to 4000hrs, each under its respective testing condition of carbon, Cr-Mo and 18-8 Ti steels. The testers can be handled much more easily in supplying specimens and taking them off as well as in checking the thermocouples, etc, than the conventional testers.
This type of creep-rupture testers is best suited for test lasting for more than 1000hrs, and makes is possible to reduce required installation space and minimize running cost to a several tenth as much per specimen.

抄録全体を表示

Cr-Mo-V cast steel was subjected to high temperature tensile tests, creep rupture tests and thermal fatigue tests to investigate the effect of aluminium deoxidation, cast block size and heat treatment on its high temperature properties (Its tensile strength, rupture strength and thermal fatigue resistance).
The tensile test data show that the austenitizing temperature and the cooling rate are primary factors for the tensile strength of Cr-Mo-V cast steel. They show also that the strong deoxidation (Al 0.9kg/ton) reduces not only variation of the tensile test data owing to elimination of microporosities, but slightly decreases the ultimate strength and the 0.2% proof stress.
Strong deoxidation impairs the rupture strength and ductility of Cr-Mo-V cast steel, normalized and tempered. Its double step heat treatment, however, improves its rupture ductility, although it reduces its rupture strength.
The result of the thermal-fatigue tests of Cr-Mo-V cast steel shows its superior thermal-fatigue resistance to that of Cr-Mo cast steel.

抄録全体を表示

A transtition from static creep to dynamic creep was observed in the range of frequency of stress cycling from 1 cycle per hour to 1 cycle per minute with metallic materials of different crystal structures and compositions; the creep rate decreased drastically with the increase of frequency of stress cycling from the rate which is equal to the constant stress creep rate to a very low value. This transition was shown to be the result of creep recovery during the low stress period in each stress cycle. In the case of intermittent stressing, the frequency at which the transition occurs is determined by a critical creep strain which is, roughly speaking, common for different materials and independent of stress and temperature; if the creep strain increment during the load-on period in a stress cycle is less than the critical creep strain, it recovers during the following load-off period and the resultant creep rate becomes very low. In the case of cyclic stressing of non-zero lower stress, the transition frequency increases with the increase of the lower stress level. It is suggested that the transition might be related to small amount of impurity in the materials.

抄録全体を表示

A phenomenological theory was developed to describe the static-to-dynamic transition in creep of metals reported in the preceding report with a good agreement with the experiment. The theory is essentially one of interaction of creep and creep recovery, ant it is based on the following two fundamental features: first, the total strain under creep condition can not be regarded as the sum of the three mutually independent components, i. e. elastic, anelastic (creep recovery) and plastic (creep) strains as has usually been done. There is a strong interaction between the anelastic and the creep strains. Secondly, the creep recovery under load-off condition does not appear to be simply the inverse of the anelastic creep under load-on condition as has so far been thought. The delay time for the load-off condition seems to be much longer than that for the load-on condition. It is also to be noted that the rate of creep recovery immediately after unloading is approximately equal to the creep rate during the load-on period. Much more works are desired to be done on the creep recovery not only because of its importance in connection with the creep behavior under cyclic stress condition but also because it is possibly related with the mechanism of creep in general.

抄録全体を表示

It is required that the creep rupture strength of steels for high temperature services will be estimated upon many data for production of one given kind of steels by statistical procedure. When the stress-rupture time diagram is used, there are two methods of estimating 10⁵h creep rupture strength, one by finding the regression curve of rupture time for stress and its confidence limit, and the other by calculating the mean value and confidence limit from the strength of each material. In the present report discussion is made of the validity of the two above mentioned methods based upon the long time data for production of six charges of high strength boiler tubes. Of the two methods in question it is found that the latter has proved preferable in the majority of cases.

抄録全体を表示

Hereunder is given a brief account of stress relaxation tests performed of 19 Cr 9 NiWMoNb steel within the temperature range from 600°C to 700°C with changing initial stress or strain. The experimental results were analyzed by using the parametric method proposed in our perevious papers.
The conclusions obtained are as follows.
(1) The stress relaxation experiment with uniformal initial stress or strain at a specific temperature range can be well combined into a single correlated curve termed as“Master relaxation curve”, using parameter P=LogEt-Q/4.6T
(2) The stress relaxation experiment with uniformal testing temperature, but with a specific range of initial stress are as well correlated into a Master relaxation curve, using parameter π=Logt+mLogσ.
(3) Temperature dependence of relaxation rate ε_rp defined by the so-called rate process with an apparent activation energy about 138kcal/mole within the range of 625-700°C. But ε_rp at 600°C is one 30th as large as the value extrapolated from the above mentioned activation process. This is thought to arise from the heavy locking action of precipitates to the dislocations under these low temperature region.
(4) The so-called Smith's relationship, that is linear relationship between logarithm of time for a specific stress fall and temperature, is also deduced from our parametric relationship between P and X, disregarding the temperature dependence of Young's modulus.

抄録全体を表示

In the present monograph is described a method to predict thermal fatigue resistance of metals under multiaxial thermal stresses, and the multiaxial and uniaxial thermal fatigue tests of low carbon steel and high carbon chromium steel are reported.
Two types of specimens were employed to produce multiaxial thermal fatigue; one was a solid cylinder cooled from one end, and the other was a thick walled cylinder cooled from the inner surface. The elasto-plastic thermal stresses in each case were calculated by using the finite element method and the method of successive approximations.
It is concluded that the lifetime for crack initiation in multiaxial thermal fatigue can be predicted from the conventional thermal fatigue test results. Attention was focused on the difference in the mode of crack propagation in the two given cases.
The results were applied to making estimation of the thermal fatigue strength of the rolls used in the hot strip mills.

抄録全体を表示

In this paper is described the effect of change in property of the materials for gas turbine blade, such as Inconel 700, Inconel X, etc., due to their heating for a long period of time, on their thermal fatigue strength.
The thermal fatigue test was carried out by applying the burner heating method on knife-edge type specimens.
The results are summarized as follows:
(1) In prolonged heating, the material showed overage-softening after a certain period of age-hardening, but its ductility and toughness decreased gradually. In this case, its thermal fatigue strength increased during age-hardening, but decreased during overage-softening just as in the case of experiment with its stress rupture strength.
(2) From above mentioned relation between hardness and strength it was found that tracking of changes in the hardness of these materials is a useful means for the evaluation of the durability and the time for replacement of the turbine blade in use.
(3) The correlation between hardness and thermal fatigue strength mentioned above is considered to depend on the proportional relation between the change of plastic strain range and that of 0.2% proof stress, but this is not the only reason to explain these facts, and also there is some room to examine the relation between thermal fatigue strength and reduction of area in tensile test. It seems that precise examination of thermal fatigue strength needs further experiments.
(4) The experimental results decided by the estimated strain in the edge of the test specimen dispersed from 1 to 1/10 of the calculated life according to modified Coffin-Langer's formula which corresponded to 1 to 1/2 of the calculated strain range. It was the aim of this testing method to make qualitative comparison of thermal fatigue life of materials, and it is also applicable to making their quantitative examination approximately.

抄録全体を表示

It is assumed that the occurrence of forging defects is accounted for by examining the amount of critical compression and by determining the values of this factor, and so by judging whether the material used was of sound quality and whether the working conditions were right. To justify this reasoning, the notched specimens prepared from carbon, low alloy and alloy steels, were compressed by impact energy, and the relationship between the behavior of forging defects produced at the notch bottom and the value of compression amount was investigated.
As the result, it has been ascertained that the critical impact compression amount has a close relation with individual metallurgical factors, such as material and conditions, depending on the external factors such as the working temperature and the working speed. Furthermore, this notch compression method has proved to be an effective means for understanding the tendency of flaw generation in the forging.

抄録全体を表示