High carbon steel plates (JIS SK5, 0.87%C) were cold rolled and their residual stress was measured by X-ray diffraction method. The plate specimens of 300×190×3.2mm3 were annealed at 600°C for 20min., finished by surface grinder, and then chemically polished to remove the work hardened layers and residual stress. The thickness of the specimens before cold rolling was 2.8mm, the work roll diameter was 180mm, and rolling was carried out under various reduction ratios from 0.33% to 2.53%. The main results obtained are as follows. (1) The reduction ratio was linearly proportional to the roll gaps. (2) The surface residual stress was tensile when the reduction ratio was under 1.0% but was compressive when the reduction ratio became higher. (3) The distribution pattern of the surface residual stress seemed to have a close relation to the rolling pressure.
An investigation of hydrogen diffusion in 21/4 Cr-1Mo steel subjected to quenching and tempering treatments has been carried out at room temperature in an attempt to establish the effect of microstructures developed by the heat treatment on the diffusivity and solubility of hydrogen by means of the electrochemical permeation technique. The results obtained are as follows: (1) The diffusion coefficient in the mixed structure of martensite and ferrite phases obtained by quenching from a low austenizing temperature is higher than that in the as-quenched martensitic structure obtained from a higher austenizing temperature. On the other hand, the solubility of hydrogen shows a behavior opposite to the diffusion coefficient. When tempered at 550°C, the diffusion coefficient decreases but the solubility increases, regardless of austenizing temperature. This is attributed to the precipitation of fine Mo and Cr carbides at the interface of the ferrite phase, thus increasing the interfacial areas. The interface acts predominantly as the hydrogen trapping site, and hence an increase in solubility and a corresponding decrease in diffusion coefficient occur. (2) With regard to the effect of tempering temperature on hydrogen diffusion, the as-quenched martensitic structure has a low value of diffusion coefficient and has the maximum at the tempering of about 350°C. The value of tempered structure decreases to a minimum at near 550°C and increases again at higher temperatures. The solubility of hydrogen is high for the as-quenched martensitic structure and decreases with increasing tempering temperature to a minimum at 350°C, after which it increases until it reaches a maximum around at 550°C and decreases again at higher temperatures. (3) The diffusion coefficient and solubility of hydrogen do not change even though the tempering time is increased up to 100hrs at 350°C. However, in the specimen tempered at 550°C, the diffusion coefficient decreases with increasing tempering time and the solubility of hydrogen increases with it. (4) In the temperature range of 7 to 60°C, the diffusion coefficient and solubility of hydrogen in martensite tempered at 350°C can be described as follows: D=5.86×10-3exp[(-5530±310)/RT](cm2/s) C=7.27×10-17exp[(16670±980)/RT](mol/cm3) The hydrogen absorption occurs by the exothermic reaction.
Tempered martensite embrittlement in low alloy steel was studied in terms of fracture toughness. The effects of plate thickness and notch root radius on crack initiation were analysed from a view point of fracture mechanics. The results obtained are summarized as follows: (1) The decrease in crack initiation stress intensity factor Kinit for the fatigue precracked specimens under plane strain conditions corresponded to the degree of tempered martensite embrittlement. (2) The crack initiation stress intensity factor was correlated to the square root of the notch radius ρ1/2 for the thick specimens. Tempered martensite embrittlement appears in the crack initiation stress intensity factor Kinit of the specimen having a large notch root radius and large thickness.
The fracture surface of stress corrosion cracking of tempered SNCM 8 steel in a 3.5% NaCl solution environment was observed by means of scanning electron microscopy and the X-ray diffraction technique. A quantitative nature of the topography and the residual stress of the fracture surface were found to be more closely related to the crack growth rate than to the stress intensity factor. The relation between the areal fraction φ of prior-austenite grain boundary fracture in the fracture surface and the growth rate da/dt of stress corrosion cracks was obtained as da/dt=exp(A-Cφ) where A is a constant dependent on tempering temperature while C is independent. The residual stress σR measured on the fracture surface by the X-ray method was tensile and changed as a function of the areal fraction of grain boundary fracture: σR=(1-φ)σR The mechanical field near the tip of a stress corrosion crack with microbranching was modeled by a blunt crack with the equivalent root radius ρeq. The equivalent stress intensity factor given by Keq=√ρl/ρeqK was proposed to be a fracture mechanics parameter for characterizing the near-tip field, thus the crack growth rate. The value of ρeq is twice the microbranching width of the crack and ρl is the limiting root radius for a sharp crack. The concept the equivalent stress intensity factor may explain the finding that the fractographic feature is more closely related to the crack growth rate than to the stress intensity factor.
Stress corrosion tests were conducted on the blunt-notch compact tension specimens of quenched-and-tempered SNCM 8 steel having several different prior-austenite grain-sizes in a 3.5% NaCl solution environment. The effects of grain-size on the nucleation and growth of stress corrosion cracks were analysed from a viewpoint of fracture mechanics. The susceptivity of material to stress corrosion cracking was characterized by five quantities, i. e., the time to crack nucleation at the notch root, tn, the threshold stress intensity factor, KISCC, the stress corrosion fracture toughness, KSC, the crack growth rate, (da/dt)II, and the stress intensity factor, (K)II, at the middle point of the region II in crack growth curves. The results obtained are summarized as follows: (1) The crack nucleation life tn was determined by the stress intensity factor K0, the yield strength σY and the notch root radius ρ as tn=C[2K0/(σY√πρ)]m where C and m are constants independent of prior-austenite grain-size. The value of effective notch root radius obtained by substituting K0=KISCC and tn=240hrs into the above equation was found to increase with grain size. (2) When the plastic zone ω was smaller than the grain size d, the KISCC value was found to take a nearly constant value of 35 to 45kg/mm3/2. For the cases of ω/d>1, the values of KISCC and (K)II increased with decreasing yield strength while the value of (da/dt)II decreased. The larger grain sized material showed lower (K)II and higher (da/dt)II values when compared at the same yield strength. (3) The values of KISCC and (K)II were higher and (da/dt)II was lower in thin specimens than in thick specimens. The amount of microbranching of cracks was suggested to be a significant clue to the effects of grain size and specimen thickness on growth kinetics.
Acoustic emission during the fracture toughness tests of compact tension specimens of structural low-carbon steel was characterized in terms of crack-tip plasticity and correlated to J integral. The plastic deformation occurred inhomogeneously in the form of Lüders bands. These bands first appeared from the crack tip and propagated circularly. Straight bands were formed from the side edge in the compressive zone in the latter stage of the toughness test. The size of the circular plastic zone near the crack tip was in agreement with the BCS model analysis when the zone was fully constrained. An abrupt increase in total emission count and a high peak of rms voltage were coincident with the formation of new Lüders bands from the crack tip. The propagation of Lüders bands in other places were not accompanied with large emission. As far as the circular Lüders bands continued to generate from the crack tip, the total count increased proportionally with the size of the plastic zone and the proportional constant was independent of the initial crack length. Thus, two distinct stages were found in the relationship between total count and J integral and the transition point coincided with the point that the Lüders band stopped to formed at the crack tip.
The elastic constants of nickel base alloy at high temperatures were studied by the X-ray method with a new apparatus consisting of a horizontal type tensile testing machine and a vacuum furnace. Plate specimens of nickel base alloy (Inconel X) were used in the experiments. After heat treatment for precipitation hardening, all the specimens were finished and then electropolished before being exposed to X-ray. The characteristic X-ray of CrKβ was irradiated on them through a thin beryllium foil, and the strain was determined by measuring the diffraction from (311) atomic plane by means of the conventional sin2ψ method using the counter technique. From the slopes of 2θ-sin2ψ diagrams at several applied stresses, the M-σm curve was drawn by using the method of least square. The elastic constant was calculated from the slopes of the M-σm curves at various temperatures up to 300°C. The conclusions of the present study are as follows. (1) The elastic constants of Inconel X obtained by the X-ray technique are in good agreement with those measured mechanically within the experimental errors. The values obtained, however, vary slightly with temperature. (2) The lattice spacings or lattice constants changed about 3.25 degree in 2θ from room temperature to 300°C. Consequently, it is necessary to take great care of slight change in temperature for stress measurement by X-ray.
In the previous paper, the value of constant KX(-34.10±1.7kg/mm2/deg) needed for the X-ray stress measurement of 18 Cr-8 Ni austenitic stainless steel was determined with high accuracy from γ(311) diffraction of CrKβ radiation by the ψ0 oscillation method. However, the study on stainless steel which has been subjected to severe plastic deformation has not been made so far. In this study, the X-ray elastic constants of tensile pre-strained 18-8 stainless steels were measured by the ψ0 oscillation method with CrKβ γ(311) diffraction, and by the η oscillation method with CrKα γ(220) diffraction, and the accuracy of the measurement was examined. The results are summarized as follows. (1) From a viewpoint of accuracy, it is desirable to apply the ψ0 oscillation method for γ(311) diffraction of CrKβ radiation and the η oscillation method for the CrKα γ(220). (2) In the CrKβ γ(311) ψ0 oscillation method, the absolute value of the constant KX increases slightly with pre-strain in the range of small pre-strain but it reaches a certain saturated value when pre-strain becomes large. In the CrKα γ(220) η oscillation method, it decreases gradually with increasing pre-strain. The latter results can be qualitatively explained by taking the tensile deformation texture into analysis, though in the former case a clear explanation has not been made. (3) For the X-ray stress measurements of some unknown pre-strained stainless steels of which the diffraction intensity curve from CrKβ γ(311) diffraction is known to be practically uninfluenced by CrKα α(211) diffraction, the K-value for 18-8 stainless steel reported previously (KX=-34.10 kg/mm2/deg) is considered appropriate in practical use, and the error of measurement is at most 10per cent.
Metallic-cations were exchanged on the Brønsted acid site structure of silica-alumina and the product was fired on steel as a glaze and the adhesive mechanism was investigated by an electron probe microanalyzer and by tension tests. The behaviors of exchanged metallic-cations and iron-ions in the fired silica-alumina glass-steel interface were investigated by changing the amount of exchanged metallic-cations and the firing time. The results showed that the metallic-cations exchanged on the acid site structure diffused to the glass-steel interface and iron-ions from steel to glass section. The alloy structure consisting of exchanged-metal and iron was found to form on the glass-steel interface. On the other hand, the adherence strength on the glass-steel interface increased with the amount of metallic-cation exchanged. It was concluded that the adhesive effects and the behaviors of exchanged-metal and iron at the glass-steel interface contribute to the formation of alloy structure.
A fatigue experiment has been carried out on the glass fiber reinforced polycarbonates under the load condition of combined cyclic and mean stresses to study the effect of mean stress on the crack propagation behavior. The results obtained are summarized as follows: (1) It was not possible to clarify which one, the maximum stress (σmax) or the stress amplitude (σa), affected the fatigue crack propagation more. But under a given condition of constant σa, the crack propagated faster with increasing mean stress due to an increase in σmax. In spite of the difference in fiber content, all the materials showed a similar tendency. On the other hand, under a given condition of constant σmax, the crack propagated slower with increasing mean stress due to a decrease in σa. (2) In the case of change in mean stress under the condition of variable σmax and constant σa, the relation between the crack propagation rate and the stress intensity factor range followed the Forman's formula at low mean stresses, but at high mean stresses the relation follows the formula only when the corrected fracture toughness values were used. In the latter case, the constant C' depended on the mean stress. Moreover, the fracture surfaces were examined macroscopically as well as microscopically. The different tear line patterns were observed, by changing the mean stress, on the macro-fractographs of the nonreinforced material. The change in pattern due to the debonding between fiber and matrix and the plastic flow in the matrix were observed, by changing the mean stress, on the micro-fractographs of the reinforced ones.
The transition behaviours of fracture have been examined on lamellar pearlite steel containing proeutectoid ferrites. The results obtained are summarized as follows: (1) The types of initiation of micro-cracks are divided into the following classes according to the temperature range. (a) Below TC, micro-cracks occur at the boundary of pearlite colony (type a). (b) Between TC and TB, in addition to type a micro-cracks initiate in the pearlite colony due to the stress concentration caused by the deformation twin in free ferrite. (c) Between TB and TD, micro-cracks initiate due to the stress concentration caused by the slip in free ferrite. (d) In the vicinity of TD, micro-cracks may be formed by linking cracks in cementite films in the pearlite colony. (2) When the micro-cracks which have grown and propagated in the pearlite colony start to propagate into the adjacent grain, the following types of propagation are observed depending upon the range of temperature. (a) Below TC, the micro-crack propagates by separating the interface of ferrite-cementite in the pearlite colony (type A), by cracking cementite and ferrite films (type B) and by cleavaging free ferrites (type C). (b) Between TC and TB, type B and C are mainly observed and micro-crack rarely propagates along the boundary of free ferrite (type D). (c) Between TB and TD, the micro-crack propagates along the boundary of free ferrite. (d) Above TD, the micro-crack in the pearlite colony propagates into the adjacent pearlite through free ferrite because of large ductility of free ferrite. (3) It is considered that TB is the temperature at which free ferrite loses its ductility and TD is the temperature at which the cementite film loses its ductility.
The disk test, in which a circular specimen with an internal crack is subjected to diametral compression, has been used to investigate the fracture criterion for the combined mode of mode I and mode II. This method has several advantages especially of allowing systematic measurements of mode I-, mode II- and the combined mode-fracture toughness under the same condition. In this paper, polynomial appoximating equations for the stress intensity factors, were established with high accuracy by taking the effect of Hertzian contact into consideration. Reactor graphite and plaster were examined to obtain the fracture toughness values KIc, and KIIc and the combined mode fracture criterion.