This paper is concerned with strain and temperature distributions of a long rod subjected to high longitudinal impact strain. A constitutive equation in Malvern's linear strain-rate dependent theory was used. If dynamic deformation of the rod is adiabatic at a high strain rate, temperature rise caused by plastic work cannot be neglected. The constitutive equation was developed on the following three assumptions with respect to the temperature of the rod.
(1) Dynamic deformation process is adiabatic and all plastic work is converted to heat.
(2) A specific heat is a linear function of temperature.
(3) Material constants in a stress-strain relation of the rod are dependent on its absolute temperature.
A system of basic formulas in a thin rod is hyperbolic, which can be solved numerically by the method of characteristics. The characteristic line is curved due to its dependence on the temperature of rod so that a supplement rectanguler net was used in the numerical analysis.
The numerical results show that the Malvern's theory with the consideration of temperature rise exhibits a steep strain region such as the flow deformation region followed by a strain plateau whose length is a function of strain hardening parameter.
A profile of temperature distribution was similar to that of strain distribution. A region of steep temperature distribution corresponding to the region of flow deformation was followed by a uniform temperature region (temperature plateau) such as the strain plateau.

View full abstract

In order to elucidate the pressure dependence of the flow stress for hexagonal metals, the tensile tests were made under hydrostatic pressure up to 600MPa on polycrystalline cadmium sheets which were rolled at room temperature. The results obtained are summarized as follows.
(1) The rate of increase of flow stress Δσ/σ due to 600MPa increases with an increase of strain, and in every range of the strain it is considerably large in comparison with the rate of increase of shear modulus due to pressure. In the extension along the rolling direction Δσ/σ amounts to about 16%, and in the extension along the transverse direction to about 20%.
(2) Every specimen shows a remarkable yield point on the stress-strain relation. A Lüders' extension following the yield drop occurs mainly due to the activation of the basal slips.

View full abstract

The coupled and the anisotropic features of plastic damage and creep damage in Nimonic 80A were analysed with a special emphasis on the finite deformation and the material spin of the damaged material. In view of the fact that both the plastic and the creep damage are governed by the formation of grain boundary cavities, it was first assumed that the states of plastic damage and creep damage were represented in terms of symmetric second rank damage tensors Ω^p and Ω^c; the sum of these tensors Ω=Ω^p+Ω^c represents the damage state of the material. The evolution equations of these variables were established on the basis of the experimental observations on the nucleation and growth of microscopic cavities. The creep constitutive equation of the material, on the other hand, was formulated by taking account of an acceleration of creep rate due to material damage as well as of the material softening caused by the formation of a dislocation network at particle interfaces.
Finally, the creep damage process at finite deformation of Nimonic 80A at 750°C subjected to prior plastic damage brought about by the plastic prestrain at room temperature was analysed. The numerical results were compared with the corresponding experimental results to discuss the validity of the proposed theory. Though a considerable rotation of the principal damage direction was observed in the process of torsional creep, its effect on the creep damage process was found to be rather small.

View full abstract

The authors' method for breaking a high carbon steel rod into billets is characterized by a process in which cross-sectional cracks develop from a slight peripheral notch by rapid local heating and cooling.
In the previous paper, the residual stress distribution in the notched or cracked section of SUJ-2 notched specimen was calculated by F.E.M., based on the measured residual stress distribution of the unnotched specimen caused by rapid heating and cooling. And it was shown that the length of cracks developing in actual rod breaking tests could be precisely estimated by comparing the calculated distribution of residual stress after cooling near the crack tip in the notched section of the specimen with the true fracture strength of material. However, in this rod breaking method, crack development was found to occur during cooling.
In the present study, the change of stress distribution in the central cross-section of the unnotched specimen was examined with the lapse of time during heating and cooling in order to clarify the stress at crack initiation. Elastic-plastic FEM analysis was adopted for the specimens subjected to the heating and cooling conditions including the optimum one for breaking notched SUJ 2 φ 55 rods by a single heating process.
The results obtained are summarized as follows.
(1) The stress of specimen during heating was compressive near the surface and tensile at the core in both axial and tangential directions. During cooling, the stress at the surface layer changed to tensile, the compressive stress adjacent to the surface layer remained, and the tensile stress at the core gradually changed to compressive. The tensile stress at the surface layer of specimens under the optimum heating condition was higher in value and present in the deeper portion than those under other conditions.
(2) The period when the crack started to develop in the actual rod breaking coincided with the calculated lapse of time when the stress at the surface layer reached the highest and deepest level.

View full abstract

A method of the estimation of approximate stress intensity factor for a semi-elliptical crack at notch root was proposed. The stress distribution in the vicinity of notch root was divided into two components: uniform and linear components. The linear component was correlated with the notch root radius ρ. The stress intensity factors for these two stress components were calculated using the equations which were made on the basis of numerical results for a semi-elliptical crack at the surface of a semi-infinite body.
First, the accuracy of the proposed method was checked by comparing the present values with the accurate numerical results for a semi-elliptical crack at the inner wall of a pressurized hollow cylinder with an infinite outer radius. And then, the method was applied to a semi-elliptical crack in a pressurized hollow cylinder with a finite outer radius and to a semi-elliptical crack at notch root of CT specimen.
The following closed form formula is proposed for the estimation of stress intensity factors for a semi-elliptical crack at notch root in other problems.
K_{I, D}=K_IT+K_IB
K_IT=σ₁√πbF_IT(λ), λ=b/a
F_IT(λ)=1.122-0.230λ-0.901λ²+0.949λ³-0.280λ⁴
K_IB=(σ₂-σ₁)√πbF_IB(λ)
F_IB(λ)=0.443-0.310λ-0.104λ²+0.206λ³-0.061λ⁴
where σ₂ is the maximum stress at notch root and σ₁ is that at the point corresponding to the deepest point of the elliptical crack front.

View full abstract

The J₂ corner theory proposed by Christoffersen et al. was applied to the interaction problem between a crack tip and voids. CT specimens with voids near the crack tip were analyzed by the finite element method using the finite deformation theory. The results were compared with those by the infinitesimal theory and those by the J₂ flow theory. First, it is pointed out that the finite deformation theory is necessary to apply the J₂ corner theory to the crack problem. It is also shown that the results by the J₂ corner theory give larger displacement fields than those by the J₂ flow theory. Especially, the difference between these two theories is recognized clearly on the interaction pattern between a crack and a void. By changing the distance between a crack tip and a void, the displacement fields were compared with each other. It is noticed that the interaction effects appear differently with the change of the distance. When the distance is small, the effects of two singularities (a crack tip and a void) overlap each other and behaves as a single singularity. It is estimated that the fracture criterion may be satisfied locally in this region before the macroscopic fracture criterion is satisfied. It is concluded that the application of the J₂ corner theory is preferable for the analysis of the localization or the bifurcation problems.

View full abstract

The effect of austenitizing temperature on the fracture toughness of machine structural steels, SCM440 and S45C was investigated. Steels were austenitized at 780°C or 760°C, 1050°C and 1200°C and then quenched and tempered. The elastic-plastic fracture toughness J_Ic tests by the stretched zone method were performed to obtain the fracture toughness. Also the cross section of the stretched zone was observed to evaluate the effect of the microstructure.
The results obtained are summarized as follows.
(1) For SCM440 steel the fracture toughness decreased with increasing the austenitizing temperature, while it increased slightly for S45C steel.
(2) The decrease of the fracture toughness of SCM440 steel at high austenitizing temperature was caused by embrittlement of the grain-, martensite- and acicular ferrite·pearlite-boundaries.
(3) The critical stretched zone width, SZW_c of S45C steel hardly changed because the variation of the strength of pearlite grains was very small under the present heat treatment conditions.

View full abstract

The paper is concerned with the fracture toughness and microfractures of a sheet molding compound polyester composite at room and low temperatures. Fracture toughness tests were performed by using compact tension specimens of the composite at room temperature and liquid nitrogen temperature, 77K. Acoustic emission signals were monitored during the fracture toughness tests. The microscopic observation of the fractured surfaces and the spectrum analysis of the acoustic emission signals were made in order to obtain a reasonable explanation of the fracture mechanism. The results are summarized as follows:
(1) The load-crack mouth displacement curves at liquid nitrogen and room temperatures became nonlinear at about the same load level and the maximum load observed at liquid nitrogen temperature increased by about three times that observed at room temperature. The acoustic emission activity for the specimen tested at liquid nitrogen temperature was higher than that at room temperature.
(2) The fracture toughness K_AE at liquid nitrogen temperature, obtained as the stress intensity factor which corresponds to the onset of abrupt increase of the accumulated acoustic emission energy, was extremely larger than that at room temperature.
(3) At room temperature the fracture toughness K_AE was in good agreement with the fracture toughness K_Q obtained by the 5% offset procedure of ASTM E399, and at liquid nitrogen temperature K_AE was larger than K_Q.
(4) From the microscopic observation, it was found that the fracture of a sheet molding compound composite accompanies the microfractures of four types, i.e. fiber breakage, fiber debonding, resin cracking and delamination.
(5) The spectrum analysis indicated that the acoustic emission signals could be classified into three types for the specimens tested at room temperature and four types for the specimens tested at liquid nitrogen temperature. An attempt was made to assign each type of the frequency spectra to the microfracture, and the fracture mechanism of the SMC composite was discussed.

View full abstract

Low-cycle fatigue tests were carried out on copper alloy and SUS 304 steel, to observe the initiation of fatigue cracks in smooth specimens and the subsequent growth of the micro-cracks. Their surface crack lengths were examined using cellulose acetate replicas. The results obtained were as follows:
(1) Initiation of a main crack which causes the failure of smooth speimens occurred at the early stage of their lives. The crack initiation cycle ratio N_c/N_f was small for SUS 304 steel in the low-cycle regime, but in the high-cycle regime for copper.
(2) The crack initiation site was almost near the grain boundary triple point and the subsequent growth rate was high at the early stage. The initiation at the slip bands was rarely observed and the crack length did not change for a while.
(3) The experimental crack growth rate was not well correlated in fracture mechanics fashion by ΔK or ΔJ. The crack growth rate at the stable stage was propotional to the crack length, so the nondimensional fatigue crack growth rate k=1/a_e·da/dN represented the data well in this study.
(4) The nondimensional fatigue crack growth rate k was closely connected with the failure lives N_f of smooth specimens, in such a way as k=(9-10)/N_f.
(5) An expression for the damage in low-cycle fatigue was derived on the basis of micro-crack growth behavior.

View full abstract

The strain range partitioning analysis of creep-fatigue cracks is conducted for the normalized and tempered 21/4 Cr-1 Mo steel at 550°C and for SUS 304 stainless steel at 700°C, assuming that creep-fatigue life relations can be simply obtained by integrating creep-fatigue growth rate equations. For Δε_ij straining, crack growth rate equations in air and vacuum are respectively given by (1/a) (da/dN)_ij=ln(a_f/a₀)_ij(Δε_ij/A_ij)^1/m_ij and (1/a)(da/dN)_ij=ln(a_f/a₀)_ij(Δε_ij/α_ijD_i)², where respective partitioned strain versus life relations are given by Δε_ij=A_ijN_ij^-m_ij and Δε_ij=α_ijD_iN_ij^-0.5, and (a_f/a₀)_ij is the ratio of final half crack length to initial one. For (ΣΔε_ij) straining, it is assumed that the crack growth rate is given by Σ(da/dN)_ij.
It is shown based on air data that ln(a_f/a₀)_pp=8.3, ln(a_f/a₀)_pc=10.3, ln(a_f/a₀)_cp=2.6 and ln(a_f/a₀)_cc=4.0 for 21/4 Cr-1 Mo steel, and that ln(a_f/a₀)_pp=13.7, ln(a_f/a₀)_pc=5.5, ln(a_f/a₀)_cp=5.4 and ln(a_f/a₀)_cc=5.2 for SUS 304 steel.
Furthermore, (a₀)_ij and (a_f)_ij of very small size are evaluated on the basis of the obtained values of (a_f/a₀)_ij. It is suggested that four types of inelastic strain range, Δε_ij, are new and useful parameters that will help us to evaluate the crack initiation life, the remaining life and the material damage in future.

View full abstract

The susceptibility to stress corrosion cracking (SCC) and its potential dependence were examined for SUS 304 steel single crystals with the single slip orientation, i.e., 0.50 in Schmid factor.
The SCC tests were carried out in boiling 25-45% MgCl₂ solutions under various constant potentials in the present study.
The corrosion morphology of the specimen surface SCC-tested was dependent on both MgCl₂ concentration and applied potential: The potential range for SCC occurrence of which critical potential was -380--400mV (SCE) increased with increasing MgCl₂ concentration. In this case, the threshold concentration for cracking was about 30%, below which the potential range for pitting appeared. The uneven general corrosion occured as the potential was shifted to more noble values. On the other hand, the cracks mentioned above initiated from the specimen surface which was easy to induce slip-steps by plastic deformation. The fracture surface characterized by river-like pattern was formed by the crack growth along ‹110› direction on {110} plane.

View full abstract

Hydrogen induced cracking (HIC) and hydrogen embrittlement (HE) of ‹001› oriented iron single crystals were studied by cathodic charging tests with or without loading.
The hydrogen induced cracks were produced at six kinds of {110} planes in the annealed specimen, while one kind of (110) cracks which was parallel to the tensile axis was produced in the prestrained one.
When the stress more than the threshold one was applied, many hydrogen cracks were produced at the (001) plane macroscopically, which was perpendicular to the loading axis in both the annealed and prestrained specimens. These hydrogen cracks connected each other with quasicleavage cracks, and then the specimen finally fractured.
It seems that these (001) cracks are not cleavage fracture but quasi-cleavage one from the microscopical observation on the fracture surfaces.

View full abstract

The pitting corrosion potentials of SUS 304, SUS 316, and E-Brite 26-1 stainless steels under both isothermal and heat transfer conditions in NaCl solutions were measured. The NaCl concentration was varied in the range 0.01-26% and the solution temperature was varied from 50°C to the boiling point under normal atmospheric pressure. The pitting potentials thus obtained were correlated well with the surface temperature of the specimen under heat transfer, but independent of the solution temperature. The effect of NaCl concentration in the vicinity of the surface was also investigated mostly in dilute NaCl solutions under the heat transfer conditions at the temperature higher than the boiling point.

View full abstract

Crack growth tests under cyclic loading were performed in various organic agents by use of compact tension and pure bending specimens of PMMA. It was found that the fracture surfaces were covered with new type striations named as “wavy striation” because of their sinusoidal shape. This paper describes the details of the observed results of the striation and infers its initiation mechanism. A formation model for the wavy striation is proposed based on the meniscus instability that, when viscous fluid flows through a narrow gap between crack planes, the fluid head becomes unstable due to a small perturbation. The proposed model indicates that the wave length of the wavy striation increases with an increase in aparameter P=T(ΔK)²/ηf where ΔK is the stress intensity factor range, f is the cyclic frequency and T and η are the surface tension and the viscosity of the agent. This model is in good agreement with the experimental results.

View full abstract

The tensile strength of polystyrene decreases with increasing temperature and with decreasing molecular weight, showing a S mark curve with the change of temperature or molecular weight.
As amorphous polymers such as polystyrene maintain their strength with the entangled bonds of molecular chains, the following experimental equation representing the relation between the temperature T and the tensile strength T_s has been introduced from the change of density of effective molecular chain with the temperature based on the entangled bonding of molecular chains.
T_s=S_b[1-3/(T_m-T_b)√2∫TT_bexp{-9(T-T_m)²/2(T_m-T_b)²}dT]
where S_b is the tensile strength at the temperature of brittle fracture point, T_m is the maximum temperature of dispersion of the distribution of density of effective molecular chain and T_b is the temperature of brittle fracture point.
The experimental data of polystyrenes of various molecular weights indicate a sufficient agreement with the experimental T_s curves.

View full abstract

Elastic moduli of the crystalline region of polytrimethylene terephthalate (PTT) have been measured by the X-ray diffraction method. The elastic modulus parallel to the chain axis, E_l, was 2.59GPa, and this value of E_l is the smallest as compared with those already reported. This is reasonable, judging from the crystal structure of PTT which consists of a highly contracted conformation of molecular chain. The chain extension is considered to occur mainly by the internal rotation of methylene groups. The elastic modulus of the bulk specimen increased with increasing draw ratio and was found to reach to the value of E_l at the draw ratio over 3.3.
On the other hand, the elastic modulus perpendicular to the chain axis, E_t, was 3.8GPa. Since many other polymers have been found to exhibit larger E_l than E_t, PTT is unique in this respect, and it has a characteristic crystal structure in which the intermolecular force has larger effect against deformation than the force holding the contracted molecular conformation.

View full abstract

The comparison of the modulus E_p evaluated from the ultrasonic propagation speed measurements with the static modulus E_s from the tensile tests and the examination of change in E_p with tensile deformation for polyethylene terephthalate (PET) fibers and glass beads filled polystyrene composites revealed the following peculiarities in E_p.
(1) E_p was not linearly dependent upon the crystallinity but dependent upon the cube root of crystallinity, while E_s was directly dependent upon the crystallinity. In the case of PET fibers, the moduli, E_p and E_s, were satisfactorily described by a parallel two-phase structure model with crystalline and amorphous regions.
(2) The two-phase structure model also enabled to be applied to composite materials. E_p of the glass beads filled polystyrene composites was not linearly dependent upon the volume fraction of fillers but was dependent upon the cube root of volume fraction of fillers. In the case of glass beads filled polystyrene composites in which the adhesion between fillers and matrix was improved by silan-treating, the modulus E_s was described by the series model. In the case of untreated composites, E_s was constant.
(3) In the case of low oriented PET fiber, E_p decreased by drawing in spite of an increase in birefringence. It is suggested that the disorder of secondary bonds are closely related to the ultrasonic propagation mechanism.
(4) E_p for PET fibers increased in stress relaxation after a small deformation but decreased in stress relaxation after a large deformation.

View full abstract

Strain measurement using moire by shifting a model grating is a useful method for full field measurement of strain distribution. We can obtain moire fringes which show a strain distribution by means of only superposing two copies of a deformed model grating shifted one with respect to the other. However, its disadvantage in applications has been the photographic process which involves time-consuming development of the photographic film. Also it has been difficult to shift a model grating exactly.
This paper presents a method using a TV camera and a digital image processor which employs a personal computer whereby high speed and exact measurements of strain distribution are possible. Using this system, we present a method for strain analysis by means of superposing two copies of a deformed cross grating. In addition, the principle and limitation of this method are described and some applications are shown.

View full abstract