The specific volume and the viscosity of binary mixtures of methanol+, ethanol+, 1-propanol+, and 2-methyl-2-propanol+ water were investigated as functions of temperature, pressure, and composition.
The specific volume was measured by means of a “high pressure burette” apparatus at temperatures ranging from 283.15 to 348.15K and pressures from 0.1 to 200MPa with an uncertainty of 0.09%. The viscosity was obtained by a falling-cylinder viscometer with an uncertainty less than 2% covering the same temperature range and pressures up to 120MPa.
The specific volume data were correlated satisfactorily by the Tait equation, and thermodynamic quantities such as compression, isothermal compressibility, excess quantities, and partial molar volume were calculated. In these systems, the isobars of the thermodynamic quantities versus composition showed a definite minimum or a maximum, caused by the complex interactions between alcohol molecules and water.
The viscosity of alcohols and mixtures increased with pressure, whereas that of water was almost independent of pressure. The viscosity versus composition isobars showed a maximum at 0.2-0.4mole fraction of alcohols. The viscosity maximum was distinct at low temperatures and pressures, but it disappeared gradually with rising temperature and pressure. The viscosity data were correlated with composition and pressure or density at each temperature using polynomial equations.

View full abstract

Viscosity changes of lubricating oils with regard to pressure (Max. 500MPa), temperature (Max. 150°C) and shear rate (Max. 10⁵sec^-1) were investigated by using a falling body type and a capillary type viscometers. The oils tested were mineral oils, animal and vegetable oils, oleic acid and polymer-blended mineral oil.
The viscosities of animal and vegetable oils and oleic acid had similar dependencies on both pressure and temperature. The viscosity of an oil with a large pressure dependency also had a large temperature dependency. Both the pressure and temperature dependencies of viscosity became smaller as the temperature rose, and the temperature dependency of viscosity became greater as the pressure rose. The viscosity abruptly decreased beyond a critical shear rate, while it was nearly constant below the shear rate. The critical shear rate became smaller with an increase of pressure.
By taking viscosity changes into account, the change of oil film thickness at roll bite in steel rolling was analyzed and discussed.

View full abstract

A new set of experimental PVT property and vapor pressure data is reported for the range of temperatures 643-673K, of pressures 20-40MPa and of densities 108-618kg/m³. The experimental uncertainties in the present study are estimated to be ±0.04% for density, ±4mK for temperature, ±5kPa for pressure in the PVT measurements and ±3kPa for vapor pressure data measured.
Some detailed comparison of the measured results have been made not only with those reported by the previous investigators, but also with the equations of state proposed recently. The compared results have shown that the present measurements were in good agreement with an available, single set of extensive measurements by Rivkin et al. for the densities below 500kg/m³, whereas the latter reported for higher densities above 500kg/m³ seemed less reliable because the present results agreed more satisfactorily with another sets of reliable data by Kell et al. and also by Alexandrov et al.
Based upon the present measured vapor pressure data as well as those reported previously, an analytical vapor pressure correlation has been developed for the entire temperature range between the triple point and the critical point. At the same time, a nonanalytical scaled equation of state formulated by Hanafusa based upon the present measured results has been discussed in detail and the critical parameters of water have been recommended with evaluated uncertainties.

View full abstract

Compression of aqueous binary mixtures containing acetonitrile, acetone, and dimethyl sulfoxide has been measured as a function of composition at 298.15K and 101.3MPa. For the aqueous systems with n-propanol, acetone, and dimethyl sulfoxide, the isothermal compressibilities of the mixtures at atmospheric pressure have been further determined from the compression data obtained at pressures up to 150MPa. The composition dependences of compression and isothermal compressibility for the mixtures were discussed from a viewpoint of the effect of groups in the organic compounds.
The dielectric constants of the aqueous mixtures with methanol, n-propanol, acetonitrile, and acetone have been measured at 298.15K and pressures up to 300MPa. The data obtained were fitted to the Tait-type equation. The values of ε^-2(∂ε/∂P)_T were calculated from the results and its composition dependence was compared among the systems.

View full abstract

In order to measure accurately the ultrasonic speeds in liquids under pressure, a new experimental apparatus was constructed employing a sing-around method.
The ultrasonic speeds in benzene and toluene were measured at temperatures of 293.15, 298.15 and 303.15K, and pressures up to 160MPa. By measuring the pressure variations of ultrasonic speeds in benzene, the reliability of apparatus was confirmed to be better than ±3msec^-1.
From the experimental results of the pressure dependence of ultrasonic speeds, the densities under high pressure were calculated. The calculation method was improved in part from that published previously. The densities thus obtained agreed well with the experimental values reported in the literature within approximately 0.2 per cent over the range of pressure up to 160MPa, and the improved calculation method of density was found to be good, especially in higher pressure region.

View full abstract

A nonlinear fracture mechanics study was made on the low-cyle fatigue crack growth in SUS 304 stainless steel under tension-compression trapezoidal stress waves at 650°C. The high temperature low-cycle fatigue crack growth behavior was classified into two categories: cycle-dependent and time-dependent. The cycle-dependent crack growth rate correlated well with the cyclic J-integral, whereas the time-dependent crack growth rate was characterized by the integrated value of the modified J-integral, J', during tensile hold. Due to the interaction of the elastic and creep strains under small-scale creep conditions, the modified J-integral at tensil hold, J'_hold, was much greater than that calculated on the assumption of the steady-state creep. This high value of J'_hold accelerated crack growth. A parameter β was defined as a ratio of the tensile hold time to the period t_tr, which characterized the transition from the small-scale creep to the large-scale creep. This parameter β was found useful in predicting the transition from the cycle-dependent crack growth to the time-dependent crack growth.

View full abstract

Torsional fatigue tests were carried out for commercial aluminium wires by changing temperature from 77K to R.T. or vice versa, and resistivity was measured as a function of number of cycles at 77K.
After the temperature change from R.T. to 77K, the resistivity increased with number of cycles, and saturated at an advanced stage of fatigue. This phenomenon was explained from the change in lattice defect density given by
ρ¹²=α/β{1-(1-P)e^-(β/2)γ}
C=C_s[1-1-P/K₂-β/2{K₂e^-(β/2)γ-(β/2-K₂P/1-P)e-K₂γ}]
C_s=K_fα²/K₂β, P=ρ_i^1/2β/α
where ρ_i is the dislocation density prior to temperature change, ρ, C and γ is the dislocation density, the concentration of point defects and the cumulative strain after temperature change, respectively, and K_f, K₂, α and β are constants.
The decrease in resistivity induced during fatigue after the temperature change from 77K to R.T. was thought to occur mainly due to dynamic annealing of point defects.

View full abstract

The effect of strain rate on the shear strength of a commercial pure aluminum and alloys similar to 4340 steel (high strength materials of four kinds) was examined by the torsional split Hopkinson bar method using a combined tension-torsion impact testing apparatus. It was clarified for aluminum that the effect of strain rate on the strength was not distinct at small strain, but became more conspicuous at larger strain.
For high strength materials, such as alloys similar to 4340 steel, the shear strength also depended on the shear strain rate. The yield stress at a high strain rate was higher than that at a low strain rate, and the slope of the tangent to the dynamic stress-strain curve in the plastic region at a constant strain rate had a tendency to become gentler as the strain rate became higher. The practical constitutive equation for an elastic/viscoplastic body proposed by one of the authors was applicable to these tested materials.

View full abstract

The long life fatigue failure of a case hardened unnotched steel specimen starts just under the hard surface layer and “Fish eye” is found on the fractured surfaces. In the previous experiment, it was found that the rotating bending fatigue strength of carbonitrided unnotched specimens was fairly higher than the value estimated based on the strength for fatigue crack initiation and propagation at the critical portion in the specimen cross section.
To examine the reason for the difference between the experimental and estimated fatigue strengths, fully reversed bending fatigue tests were made on two kinds of Cr-Mo steel specimens of rectangular test section with surface layers hardened by low temperature gas carbonitriding. Specimen A was provided with hard cases only on the upper and lower surfaces of the test section, and specimen B was carbonitrided on the whole surface.
By using specimens A, crack initiation and propagation on the side surfaces was watched during the fatigue tests. Cracks were initiated just below the hard layer (0.4∼0.6mm distant from upper or lower surface) and propagated to fracture.
At a considerable wide range of stress amplitude (about 10kgf/mm²) below the fatigue fracture limit, cracks were initiated but soon arrested due to compressive residual stress of the hard layer. The minimum stress amplitude where non-propagating cracks were observed, coincided mostly with the estimated fatigue limit based on the strength of the critical portion.
In the fatigue tests of specimen B, “Fish eye” was found on the fracture surfaces of all specimens failed beyond 3×10⁵ stress cycles. The mean depth of the fish eye center from upper or lower surface of test section was 0.47mm. Comparison of the S-N relation of specimen B with that of specimen A, suggested the slow propagation of internal crack in specimen B, but the difference in fatigue limit between B and A was comparatively little.
Three specimens of kind B which endured to 10⁷ cycles at the stress amplitude levels less than 4kgf/mm² below the fatigue limit, were fractured under rather high stress levels to show the internal failure. A non-propagated internal crack was found on one of these forced fracturd surfaces.
The higher fatigue strength of case hardened unnotched specimen than the estimated value is supposed to depend mainly on the following factors.
(1) Arrest of crack initiated in the critical portion due to compressive residual stress in the hard layer.
(2) Slower propagation of the internal crack than the external one.

View full abstract

The effect of interaction between corrosion fatigue and stress corrosion cracking on the crack propagation behaviour of 7278-T6 aluminum alloy has been studied under zero-to-tension low cycle fatigue tests with a hold time in salt water (3.0% NaCl).
The results obtained are summarized as follows:
(1) From the low-cycle corrosion fatigue tests with trapezoidal stress waves, the crack growth rate was found to increase with increasing stress hold time and was approximately proportional to the cyclic Loading frequency f^α. The crack growth rate during the stress hold time was accelerated by the stress cycling as compared with that of stress corrosion cracking.
(2) The tests of alternate loading of stress cycling and holding at a constant stress were conducted and the crack growth rate was determined. The crack growth rate calculated in terms of that with trapezoidal stress waves was larger than that of trapezoidal stress waves itself in the region of short holding time. On the other hand, it decreased with increasing stress hold time and became smaller than that of trapezoidal stress waves in longer holding time because of crack tip blunting.
(3) The crack propagation rate during the stress hold time was approximately proportional to t_h^-β, where t_h is the stress hold time, β≈0.5 for trapezoidal stress waves and 1.4 for alternate loading. It was suggested that β is affected by the sensitivity to environmental attack at crack tip and slip dissolution.
(4) Environmental enhancement for crack growth rate was found in the low stress intensity range, and it increased as the testing frequency was lowered in the case of trapezoidal stress waves. For the alternate loading, a critical frequency was found at which the maximum environmental attack occured.

View full abstract

Appliciability of fracture mechanics parameters to stress corrosion crack propagation of SUS 304 stainless steel in boiling MgCl₂ solution was investigated in consideration of the specimen size effect and the characteristic environmental effect arising from coating of side surface of the specimen.
The stress corrosion crack propagation rate was nearly proportional to the J integral in the medium to high stress region where fracture surface morphology showed the fan-shaped type in the medium stress region and the mixed intergranular-transgranular type in the high stress region. Therefore, it can be understood that the J integral is a dominant fracture mechanics parameter to stress corrosion crack propagation in the medium to high stress region. It may be expected that the elastic fracture mechanics parameters G and K become dominant parameters to stress corrosion crack propagation in the low stress region where fracture surface morphology showed the brittle-facetted type.
Characteristic nonpropagating stress corrosion crack, caused by incomplete penetration of the corrosive solution into the crack tip, was observed on the coated specimen in the low stress region. There was little effect of the coating on the stress corrosion cracking in the medium stress region.

View full abstract

In order to investigate the behavior of creep crack propagation in a boiler superheater tube (Type 321 stainless steel) after long service, the unused and 120000 hours used materials produced at the same lot were made into test specimens and creep tests on the smooth specimens and creep crack propagation tests on the center notched plate specimens were carried out.
The results obtained were summarized as follows:
(1) The crack propagation rate, dl/dt, of the used material was faster than that of the unused material at higher stress levels. However, this difference in dl/dt decreased as the applied stress decreased. These results arose from the difference in creep deformation property of two materials.
(2) Roughly, the crack propagation rate, dl/dt, was correlated well with the creep J integral (i.e., C^* parameter or modified J integral), J_c, and the correlation was identical for both materials.
(3) Strictly speaking, the crack propagation rate, dl/dt, was accelerated at lower stress levels, because of decrease in creep fracture strain, ε_f. Therefore, better correlation was achieved between dl/dt and the fracture strain modified creep J integral, J_c/ε_f.

View full abstract

The characteristics of fractured surface of SCM 435 steel due to hydrogen embrittlement were studied as a function of hydrogen content, stress intensity factor and crack propagation rate, and the following results were obtained.
(1) The fractured surfaces of steels having 115, 150 and 170kg/mm² tensile strength revealed almost intergranular, while that of 100kg/mm² steel showed quasicleavage.
(2) The grain size of intergranular facet was almost equal to that of prior-austenite. No segregation of Cr, Mo, S or P was recognized at the grain boundary from the observation by an X-ray microanalyzer.
(3) The surface fractured near K_ISCC was intergranular. However, its percentage decreased with increasing K value and the surface became quasicleavage at K value more than 80MNm^-3/2.
(4) With increasing hydrogen content, both the crack propagation rate and the percentage of intergranular facet increased.
(5) The intergranular fracture was thought to occur at some portion of crack front at first during the crack propagation stage and the remained part of crack front was fractured in quasicleavage.
(6) The percentage of intergranular facet was thought to be determined by the hydrogen content at grain boundary.

View full abstract

In order to study the relations between loading conditions and fractographic features of alumina-particulate-filled epoxide composite, the fracture surfaces of notched specimens subjected to impact bending and creep rupture under various temperature conditions were examined.
As the result of microscopic observation, the fracture surface was found to take four types of appearances ranging from the flat and featureless one to the dimple-like one depending upon the test temperature, the applied stress level and the velocity of crack propagation. To clarify the fracture surface formation mechanism, finite element analysis was performed on the interaction between crack tip stress field and thermal stress in the vicinity of an alumina particle, and the direction of crack propagation was evaluated based on the maximum tangential stress criterion and the strain energy density criterion.

View full abstract

Residual stress and half-value breadth of the fatigue-fractured surfaces of various steels-such as WT-80c, SUS 403, SPV 50, SCMV 4 and S35C-were measured by X-ray diffraction. Fatigue tests were performed under the stress ratio R of 0 to 0.4. The half-value breadth b was found to increase with increasing the maximum stress intensity factor K_max but the residual stress σ_r usually decreased with increasing K_max.
The residual stress on the fatigue-fractured surface is said to be decomposed into four parts: that is (1) tensile residual stress due to monotonic deformation, (2) the release of stress due to fractured surface roughness, (3) compressive stress due to crack closure, and (4) compressive stress due to reversible plastic deformation. The present results were discussed in relation to the latter three causes.
From the measurement of the roughness of fractured surface, it was shown that the main reason why σ_r decreased with increasing K_max was the effect of roughness.
Under the constant K_max condition, the larger the R value was, the larger the σ_r was. This may be related to the reversible plastic deformation and/or to crack closure.
Other factors which influence the relative position of σ_r-K_max curve, such as yield stress, fracture mechanism, were also discussed.

View full abstract

It has been recognized that fatigue crack propagation is closely related with a crack closure caused by residual stress near the crack. So far, however, the quantitative relation between the residual stress distribution around fatigue crack and the crack closure has hardly been discussed based on the experimental facts.
In the present study, the X-ray oscillating stress measurement method was applied to measure the residual stress distribution around the fatigue crack which was generated under different conditions of stress amplitude, crack length and material. Furthermore, a linear fracture mechanics model was applied to discuss the effect of residual stress near the crack on the fatigue crack propagating rate quantitatively. The main results obtained are as follows.
It was found that the compressive residual stress measured on the surface behind the tip of fatigue crack could be estimated as the contact stress on the crack surface, and that the effective stress intensity factor range calculated by using this residual stress distribution was useful in understanding the fatigue crack propagation.

View full abstract

For the purpose of the strength evaluation and quality control of adhesive joints, it seems important to obtain the systematic knowledge on the influence of the bonding defects existing in the adhesive on the strength of adhesive-bonded joint, or, in other word, to do research on how the strength behavior of adhesive butt joint is affected by such defects. In this paper, the effect of dispersed bubbles in the adhesive on the fatigue strength of adhesive joint is investigated experimentally under a full reversed push-pull fatigue load condition by using butt joint specimens bonded with epoxy-polyamide adhesive.
In order to examine the effect of dispersed air bubbles, two sorts of adhesive manufacturing processes were adopted; one was given degassification treatment to remove bubbles in the adhesive and another was not given degassification treatment to disperse bubbles in the adhesive layer. From the strength behavior on S-N diagram for the adhesive bonded butt joints with these two processes, it was revealed that the fatigue strength of the adhesive joint whose adhesive does not contain bubbles was higher than that of the bubbles dispersed adhesive joint in a relatively low stress cycles range, but the difference of fatigue strength between both joints diminished in a relatively high stress cycles range, and the 10⁷ time strength of the both joints showed close agreement.
Furthermore, in order to evaluate such fatigue characters from the view-point of inelastic strain behavior, the total and inelastic strain ranges of the adhesive layer were measured for these two sorts of adhesive joints.
As the results, it was found that the values and the variation trends of these strain ranges of the adhesive joint with no bubbles were somewhat different from those of the joint of bubbles dispersed adhesive layer, and detailed discussions from such a view-point were made in this paper.
Then, it was confirmed that the cumulative hysteresis energy could not play a role of the fatigue failure criterion, when the adhesive layer condition was different with each other.

View full abstract

By using the specimens of press-fitted part of axles which were cold surface-rolled under various rolling pressures, the relations among its mechanical properties, surface residual stress and fatigue strength, and the change in surface residual stress during the process of fatigue have been studied.
The results obtained are as follows.
(1) The fatigue strength increased linearly with the increase in rolling pressure. For the specimen of mean Hertz pressure 2.83×10³MPa, the fatigue strength was 277MPa and its increasing rate relative to that of the normalized specimen was 216%.
(2) The fatigue strength increased as the work-hardening volume and surface compressive residual stress increased, and linear correlations existed between the fatigue strength and these quantities.
(3) The surface compressive residual stress decreased with stress cycle. It decreased linearly to the logarithm of number of stress cycles at the outside of the fitted part, and decreased curvilinearly at the inside of the fitted part. The decreasing rate at the inside of the fitted part was over two times faster than that at the outside. As cyclic stress increased, the decrease of surface compressive residual stress became more notable.

View full abstract