Dynamic recrystallization phenomenon in the tensile deformation of copper was investigated in order to clarify the deformation and the metallographic conditions under which the dynamic recrystallization appears. Cold-rolled specimens were tensiled preliminarily at various temperatures and a strain rate of 4.2×10^-4sec^-1. The stress-strain curves obtained showed plural maxima of the stress at temperatures above 250°C at which the specimens had already recrystallized statically before the loading. Optical and electron microscopic observations, a hardness measurement, and an X-ray diffraction experiment were carried out for water-quenched specimens after an interruption of the tensile deformation. These results showed that the plural maxima were attributable to strain hardening and dynamic recrystallization which became predominant one after another. A series of experiments on the effect of deformation temperature, strain rate and initial grain size on dynamic recrystallization were carried out for fully annealed specimens, as a function of the strain at the first maximum of the stress, ε_m, in the stress-strain curves. In the present investigation, ε_m decreased with increasing temperature and decreasing initial grain size, while ε_m had a minimum at the strain rate of 4.2×10^-5sec^-1. Finally, Cu-0.064, 0.12 and 0.26wt.% Sn alloys were prepared, from a consideration that Sn atoms suppress the static recrystallization of Cu-rich solid solution. The dynamic recrystallization of Cu-Sn alloys was also suppressed much in comparison with pure copper. This tendency became prominent with an increase in the amount of Sn.
The results obtained in the present study were discussed in reference to the dependence of deformation and metallographic conditions on the static recrystallization during the isothermal annealing.

View full abstract

The main purpose of this paper was to study systematically the development recrystallized textures of cold-rolled sheets of mild steel, such as aluminum killed steel and rimmed steel, by means of a three-dimensional orientation distribution function.
The deformation texture in cold rolling showed sufficient spreads, although {112} <110>, {100} <011> and {554} <225> orientations existed preferentially. The annealing of aluminum killed steel sheet below 600°C resulted in the gradual decrease of {100} <011> and {112} <110> components, but that annealing provided no drastic change in texture: the annealing texture were nearly the same as the deformed one.
The three-dimensional analysis showed that the texture of the sheet annealed above 600°C changed remarkably, i. e. {100} <0vw> components decreased repidly and {111} <uvw> group developed. Especially, the intensity of {111} <011> orientation grew stronger. It was observed during development of those recrystallized textures, that the dispersion of <110> fiber texture being parallel to the rolling direction decreased, and {112} <110> orientation disappeared gradually.
Another purpose of this paper was to carry out a precise analysis of fully recrystallized texture.
It has been said that the preferred orientation of recrystallized sheets of mild steel, especially that of aluminum killed steel sheet, is {111} <011> and {554} <225> according to the pole figure analysis; however, the three dimensional analysis has given no clear evidence that {554} <225> can be the highest intensity region. The discrepancy of the results given by these two analyses was explained by using the “three-dimensional orientation distribution function”, w(ψ, θ, φ).
From its definition, the pole density on “ND-RD axis”, the plane containing the ND axis and the RD axis, in {100} pole figure ought to agree with w(0, θ, φ). (ND is the normal direction, and w(0, θ, φ) is the average of w(ψ, θ, φ) with φ from 0 to π/2, when angle ψ is 0.) The results obtained showed that the values of w(0, θ, φ) agreed precisely with the pole density on the ND-RD axis of {100} pole figure. This fact leads to the clear conclusions that the highest intensity point of {554} <225> poles is due to w(0°, 60°, φ) and that the point is irrelevant to any preferred orientation.

View full abstract

Stress relaxation of NaCl single crystal was measured by the method using transmission of polarized light. It became possible by this method to analyze the time variation of stress in the direction perpendicular to the compression axis. It was found that the mechanism of relaxation of the principal stress [σ₂] in short time range was different from that in long time range. The initial process of stress relaxation was strongly influenced by the strain rate of compression. This influence was found to be strong but continue for a short time when the strain rate was high, and weak but for a long time when it was low. It was observed that σ₂ decayed following the Maxwell type viscoelastic model at the early part of the initial short time range of the stress relaxation, and thereafter σ₂ increased gradually by the pinning of moving dislocations and finally attained at a constant value.

View full abstract

The fatigue crack growth rate was studied for high strength steel of different prior austenite grain sizes, from 34.2 to 238μm, produced by changing austenitizing temperature. The effect of tempering temperature and stress ratio on the growth rate was also examined.
It was found that the variation in prior austenite grain size and the changes of tempering temperature had little effect on the fatigue crack growth rate at the mid-range of ΔK. At low (ΔK<30kgmm^-3/2) and high (ΔK>80kgmm^-3/2) ΔK regions of the materials tempered at 200°C, however, the crack growth rate was increased as the prior austenite grain size became fine. The threshold stress intensity factor, ΔK_th, decreased with reducing grain size, and with increasing yield strength and stress ratio.
Intergranular facets were formed in the both low and high ΔK regions regardless of tempering temperature. In this case, the area percentage of intergranular facets produced at the low ΔK region was correlated with ΔK, whereas the percentage at the high ΔK region was dependent on K_max Morphology of intergranular facets seemed to be divided into four types which correspond to ΔK level. At the low ΔK region, a trace of twin deformation or smooth feature was found on the facets, while dimples and tongues were formed at the high ΔK region.

View full abstract

As a basic model of fatigue cracks in composite materials, the fatigue crack growth behavior was studied for the case that a crack grows perpendicularly to the bonded boundary of two dissimilar materials from the material of lower Young's modulus (E₁) to that of higher one (E₂). A set of fatigue tests were conduced on plate specimens of epoxy-epoxy or epoxy-metal composite. These specimens were selected and prepared for this study to produce a wide range of E₂/E₁ and to have a fully bonded boundary. The fatigue crack growth rate (da/dn) in the material E₁ for the specimens with various E₂/E₁ ratios was measured and compared with that in the homogeneous material (E₁). The stress intensity factor range ΔK_re was calculated experimentally from this comparison of da/dn. The data of fatigue crack growth rate were corrected with ΔK_re/E₁, and the ΔK_re/E₁-da/dn relation was represented by one straight line for various dissimilar materials including dissimilar metals. From these results, it is confirmed that the fatigue crack growth rate in dissimilar materials depends on K and the present testing method is useful as the simulation test of fatigue crack growth in general dissimilar materials.

View full abstract

Constant load fatigue tests with and without a single tensile overload were performed using the center cracked sheet specimens of two different thickness (2mm and 8mm) of a high tensile steel. The effect of specimen thickness on the crack propagation behaviour and the difference between the behaviours at the surface and the center of sheet were discussed. The conclusions obtained are summarized as follows:
(1) The dl/dn under constant load was independent of the specimen thickness in lower ΔK level, but the thick specimen shows higher dl/dn than the thin specimen in higher ΔK level.
(2) The effective stress intensity factor range ratio U measured on the surface under a constant load was not affected by the specimen thickness and U was constant (U=0.7) over a wide range of ΔK.
(3) The dl/dn at the surface of the thick specimen did not differ from that at the center.
(4) The both specimens showed the retardation behaviour following overload and the intensity of retardation was about the same, but the affected crack length l^* was longer in the thin specimen.
(5) The retardation cycles N_d increased with the magnitude of overload and was larger in the thin specimen at a given stress range.
(6) The retardation behaviour at the center differed from the surface. The maximum intensity of retardation at the center was attained immediately after overload but with an interval at the surface, and the l^* is longer at the center.
(7) By using ΔK_eff, the range from the maximum intensity to the end of retardation was able to be arranged on dl/dn-ΔK_eff relation under a constant load for both specimens. But, the range from the application of overload to the maximum intensity of retardation was not able to be arranged. This is because the effect of blunting due to overload exists in that range.

View full abstract

Completely reversed plane bending fatigue tests were carried out using small size plate specimens made of Al-killed low carbon steel in two environments. One was uncontrolled room atmosphere and another was dry atmosphere whose relative humidity was reduced to about 1×10^-4. The effect of the reduction of humidity was clearly observed and the fatigue life as well as the fatigue limit were largely improved in the dry atmosphere. This improvement was due to the delay of the initiation of microcracks on the specimen surface. Moreover, it was observed that in both environments the microcracks originated in the close neighborhood of boundaries of several surface grains having the optimum orientation for slip and its larger portion buried below the surface.

View full abstract

The effects of prior loading conditions on stress corrosion cracking (SCC) of high strength low alloy steel AISI 4340 in 3.5 percent NaCl solution environment were investigated with blunt-notched specimens of compact tension type. Without prior loading of notched specimens, the life to crack initiation at the notch root was found to be determined by the quotient of the stress intensity factor divided by the square root of notch tip radius in the range where the radius is larger than a certain value, i.e. the effective radius for a crack. In this range, it can be concluded that the initiation life is characterized by the maximum stress at the notch root. The effect of prior loading on the delay of crack initiation was examined with the specimens having the radius larger than the effective radius. The amount of delay was found to become larger as the pre-loading level increased. The residual stress formed near the notch root due to pre-loading was calculated by elastro-plasticity. A unique relation was established between the initiation life and the actual stress near the notch root which was obtained by subtracting the compressive residual stress from the applied maximum stress near the notch tip. When the K-value was raised in the growth stage, the growth rate immediately became the rate obtained in constant K-tests. On the other hand, when the K-value was partially reduced from K₂ to K₁, the incubation time was observed before the crack started growing under K₁-value. The threshold stress was also found to increase due to prior high loading. The incubation time t_in was expressed as a second power function of the difference of ΔK=K₂→K₁. The relation between t_in and ΔK was unaffected by the stress variation sequences tested, i.e. K₂→K₁, K₂→0→K₁ and K₁→K₂→K₁. The experimental relation was discussed based on the residual stress distribution due to load variation and the mechanochemical mechanisms of stress corrosion crack growth.

View full abstract

Corrosion fatigue crack propagation of 304 stainless steel in boiling MgCl₂ solution was investigated under trapezoidal cyclic stresses. The crack propagation rate at high stresses obeyed the law of linear summation of fatigue and stress corrosion cracking, but deviated from the law at low stresses because of a significant effect of the environment. The value of the exponent m in the equation dl/dn=AΔK^m was considerably smaller in corrosion fatigue than in fatigue in air.
Characteristic fracture patterns appeared in the respective regions in the fracture map, and characteristic striations were observed on the fracture surface in the high stress region. It is suggested that these striations are formed in every stress cycle by the operation of two crack propagation mechanisms, ductile fatigue crack propagation due to slip-off at a crack tip during the stress-increasing period and brittle stress corrosion cracking during the subseqent period of stress-holding. The acceleration of the crack propagation in corrosion fatigue seemed to occur in both of these periods.
A transition phenomenon of the crack propagation was observed when the load was changed from cyclic stress to static stress.

View full abstract

Fatigue properties of 9% Ni steel and its welded joint were investigated at room and liquid nitrogen temperatures using notched and unnotched specimens.
The fatigue strengths of unnotched base metal and welded joint specimens at liquid nitrogen temperature were higher than those at room temperature except for the welded joint specimens broken at the weld bond region, while the notched strengths at liquid nitrogen temperature coincided well with those at room temperature for both the base metal and the weld bond specimens. The extended life of the unnotched specimens at low temperatures may be caused by retarded crack initiation. In the notched specimens, however, such extension of the crack initiation life seems to be cancelled by acceleration in crack propagation rate, so that the final fracture lives at both temperatures appear to be the same.
The effect of mean stress superposition was more distinct in the notched specimens than in the unnotched ones at both room and liquid nitrogen temperatures. This phenomenon may be explained by the fact that the mean stress increased crack propagation rate but gave no effect on crack initiation life.
The fatigue life was not affected much by stress variation in all the cases tested, and the results of the varying stress amplitude tests agreed rather well with those of constant amplitude tests.
The clearcut effect of temperature change during the test was observed. The change from low to high temperatures gave the cumulative cyclic ratio larger than unity, though the reversed temperature sequence gave the ratio of less than unity. This tendency is attributable to late crack initiation and increased crack propagation rate at lower temperatures, together with the crack propagation behavior following the crack propagation curve at the actual temperature.

View full abstract

The purpose of this investigation was to examine what factors exert influence on the dynamic flexural strength of various mortar specimens, which was determined by a drop hammer testing apparatus.
The results obtained were summarized as follows:
(1) The dynamic flexural strength of mortar specimens increased up to the age of 7 days.
(2) When mortar specimens had the same static flexural strength, their dynamic flexural strength was hardly influenced by the ratio of sand to cement, S/C. For the specimens having the same S/C, the ones with a large ratio of water to cement W/C were stronger on dynamic flexural strength than the ones with a small W/C, even though the static flexural strength of the latter was large.
(3) The dynamic flexural strength was not interrelated to the static flexural strength, but interrelated to the first or the second kink point of the load-strain curve of bending specimens with a notch. The dynamic flexural strength depended on the magnitude of the energy absorbed till the first or the second kink point.

View full abstract

The shrinkage and internal stress of bisphenol type epoxide resins cured with aliphatic α, ω-diamines H₂N-(CH₂)_m'-NH₂ (m'=2, 4, 6 and 12) were investigated by measuring the density change and photoelasticity in the curing and the cooling processes.
The internal stress of these systems were caused by the shrinkage occuring in the cooling process from the glass transition temperature (T_g) to room temperature. The shrinkage and internal stress increased with an increase in crosslinking density and T_g of the cured systems, and with a decrease in m' of the curing agents.
From the above observations, it was suggested that the reductions in crosslinking density and T_g were necessary to reduce the shrinkage and internal stress caused by the curing.

View full abstract