The present paper reviews state of the art of theoretical and experimental studies of both the propagation and non-propagation of short fatigue cracks. The limitation of the conventional ΔK-based fracture mechanics approach to short crack growth is clarified, and several useful, alternative mechanical approaches proposed are discussed on the basis of crack growth mechanisms. The significance of short crack studies is emphasized in relation to fatigue-life prediction, fatigue-limit determination, and alloy microstructure design. Possible future developments are suggested.
Recent trends in random data analysis with parametric and nonparametric approaches are reviewed. First, for nonparametric spectrum analysis, the effectiveness of using data windows is emphasized with an example. Secondly, based on a recursive algorithm by Levinson and Durbin, various autoregressive (AR) spectrum estimation techniques by Yule-Walker, Burg and others are explained together with their specific properties. Also, selection criteria for the order of AR model fitting are briefly discussed. As for autoregressive moving average (ARMA) spectrum estimation, the high order Yule-Walker method and its variants are mentioned. Finally, a periodic AR model is introduced for treating random data with periodic structures. An example of applying this model to real temperature data is also given.
Near-threshold fatigue crack growth behavior for SB 42 carbon steel and SUS 304 stainless steel was investigated at elevated temperatures of 200 and 400°C at a frequency of 50 Hz. A new test procedure was atempted in order to avoid crack closure: ΔK was decreased under increasing minimum load. A unique ΔKth-value of about 2.8 MPa·m1/2 was obtained independent of material and temperature by the closure-free test procedure. The crack growth rate, however, was found to be accelerated in the ΔK region higher than 4 MPa·m1/2 for both materials at elevated temperatures. The fractographic features influenced by microstructure and environment were discussed in terms of the mechanism of crack growth.
The fracture behavior under tension of notched plates and grooved shafts of polycarbonate has been investigated for a wide range of notch geometries and dimensions of specimens. Careful observations were made on the fracture process at the notch tip during testing. Bluntly notched specimens failed in a fully ductile manner with large shear bands, and sharply notched specimens failed in a brittle manner after the formation of a small crack at the tip of the plastic zone. By varying the notch tip radius and thickness of plates we have been able to clearly distinguish how both influence the brittle-ductile transition. The critical notch tip radius at the brittle-ductile transition increases with increasing plate thickness. Experimental results were discussed in terms of a combination of the maximum elastic stress at the notch tip and the notch tip radius. It has been established that by applying the fracture criterion derived here, we can make an accurate estimate of the strength of notched polycarbonate specimens of any shape or size.
The constitutive equations of unidirectionally fiber reinforced metals were derived. A modified fraction model concept was applied to the matrix and fiber, separately. Equilibrium and compatibility were considered between the matrix and the fiber for deriving the constitutive equations. The proposed theory is expected to describe the inelastic behavior of FRM and the strain rate sensitivity of stress-strain relations. Static and dynamic tensile tests were performed on unidirectional boron/aluminum composites and cross-ply laminated plates. Static and dynamic stress-strain curves were obtained and the effect of the strain rate on tensile strength was investigated. The tensile strength of the 0° and laminated specimens showed no strain rate sensitivity; but the tensile strength of the 45° and 90° specimens increased as the strain rate increased. The validity of the proposed constitutive equations was confirmed by comparing the calculated results with the experimental ones.
The plane elastostatic problem for two bonded elastic half planes consisting of isotropic and anisotropic materials is considered. It is assumed that the anisotropic half plane contains a straight crack normal to the bimaterial interface, and that the bonded plane is subjected to a constant strain away from, and perpendicular to, the crack. Singular stress fields at the tips of a crack fully embedded in the anisotropic half plane and terminating at the interface are investigated. Mathematical formulations are made using the method of continuous distribution of dislocations and singular integral equation method. Numerical calculations are carried out for the stress intensity factors and the order of stress singularities at the tips of the crack.
A general method for the solution of dynamic problems in an elasto/viscoplastic solid is presented. This method reduces the elasto/viscoplastic problem to a sequence of elastic problems with initial strains. The solution of the elastic problem with initial strains is determined by using four displacement functions. Using the foregoing method, a solution is derived for a dynamic elasto/viscoplastic problem in a thick-walled spherical shell subjected to an internal impact load. The numerical results show how the dynamic stresses in a sphere with viscoplastic properties vary with time.
Fully-developed plane Couette flow in the region of a transitional Reynolds number is simulated numerically, without any model which represents the effect of turbulence smaller than the size of computational grids. The upwind scheme proposed by Kawamura and Kuwahara is applied to the direct integration of the three-dimensional, time-dependent Navier-Stokes equation of motion. In this study, the procedure necessary to examine the equivalence of the numerical result to the real flow is considered, and the effect of numerical diffusivity of the upwind differencing is evaluated. From the numerical results, the outstanding features of low Reynolds number wall turbulence are discussed. Furthermore, a model of small scale turbulence in large-eddy simulation is also tested using the results of direct numerical simulation. It is inferred that the Smagorinsky model cannot be applied to the highly anisotropic turbulent shear flow of low Reynolds number without any modification.
Turbulence models applicable to turbulent swirling flow in a straight pipe have been developed. Two models, the standard k-ε model with higher order terms in Reynolds stress equation are applied, and a modified k-ε model with an anisotropic representation of turbulence is proposed. The comparisons between the computed flow distributions and the experimental data show that the modified k-ε model predicts complex flow fields successfully. The magnitudes of the viscosity tensor components in the modified k-ε model are discussed in detail.
An experimental investigation is reported of the zero pressure gradient boundary layer along a 90 degree streamwise corner. Turbulent and non-turbulent conditional sampling measurements in the outer intermittent region of the boundary layer as well as a comprehensive set of conventional measurements were made in a cross section 1.7m downstream from the leading edge. The results of a turbulent zone-averaged secondary velocity field indicate a pair of closed streamwise vortex patterns near the corner, as in conventional measurements. A large difference was found in the outer part of the boundary layer, where the turbulent zone-averaged secondary flow moves from the wall to the free stream direction. On the contrary, the non-turbulent zone-averaged counterpart moves from the free stream to the wall and it includes almost no streamwise vorticity, indicating that the secondary flow of the second kind is generated only when the flow is turbulent.
In order to understand unsteady laminar flow along an accelerating or decelerating disk in a closed cylindrical casing, experimental and theoretical studies were performed. A solution representing the unsteady pressure distribution on a rotating disk and the axial thrust was derived, and compared with that for turbulent flow. As a result, the following points were made clear: (1) Relative to the accelerating flow, the angular velocity of the center core at small radii is considerably smaller than that at large radii. This phenomenon is much different from that in a turbulent flow condition; (2) The time delay of the flow is comparatively small owing to the viscous effect; (3) Calculated results coincide pretty well with ones derived experimentally.
A new method for designing the valve plate of axial piston pumps and motors is presented. By adopting a valve plate with hydrostatic pads, fluid film lubrication can be realized on the sliding part between the plate and cylinder block over a wide range of operating conditions. This is in the rare case of a valve plate with conventional hydrodynamic pads. The effects of the main related dimensions and the operating conditions are numerically clarified, and the supporting of the cylinder block is also discussed.
Fluid flow and heat transfer in a two-dimensional miter-bend were examined in connection with a corrugated wall channel as a means of augmenting a forced-convective heat transfer with a single-phase flow in a heat exchanger. In a previous paper, experimental results were not in perfect agreement with those of theoretical analysis because of the effect of unsteady motion due to the instability of the flow downstream of the bending corner. In the present paper, unsteady motion for Reynolds numbers Re=600, 1000 and 2000 and Prandtl number Pr=0.71 is studied numerically by the finite difference method. The previous experimental results are simulated in outline by numerical calculations, but they do not agree quantitatively with those of numerical calculations for the Strouhal number of fluctuations and the Nusselt number at the concave-side of the wall.
Optimized design conditions of pebble bed regenerative heat exchangers are discussed from the standpoint of minimizing their pressure loss with a minimum reduction in thermal performance. Blow-down experiments for heating argon by hot air or by combustion gas were carried out to measure the pressure loss of argon in a pebble bed. Suitable equations for predicting the pressure loss were selected based on the measured results. The optimum values of pebble diameter and the mass flow rate of a working gas were determined by calculating pressure loss and thermal performance of pebble bed regenerative heat exchangers under cyclic continuous operation. The use of helium as a working gas in closed cycle MHD, instead of argon, proved to be advantageous from the viewpoint of both pressure loss and thermal performance.
Application of the pulsed laser holography technique was made to observe a diesel spray injected into a high pressure bomb. Two different holography techniques, that is, single-pulsed laser holography and double-pulsed holographic interferometry were adopted in this study. Shadowgraphs, schlieren photographs and micrographs of a reconstructed image of the spray were taken. The spray width and equivalence ratio of fuel droplets were microscopically measured. The fuel droplets and vapor around the spray in a high pressure and temperature atmosphere could be observed by single -pulsed laser holography. Interference fringes were observed along the spray edge by double-pulsed holographic interferometry. double-pulsed holographic interferometry made the image of the evaporating spray at high temperature clearer than the single-pulsed laser holography. The spray width had a tendency to decrease as the atmospheric temperature was increased. Some droplets were observed around the spray up to a pressure of 3.1 MPa and a temperature of 773 K.
Through numerical calculations based on a simulation program, the effect of a resonator, which is installed in intake manifolds, has been investigated for the purpose of improving the volumetric efficiency of four cycle diesel engines. The results show that in single cylinder engines a resonator can increase the volumetric efficiency at low engine speeds without decreasing volumetric efficiency in the high speed range. The diameter of the resonator pipe must be chosen to generate an adequate flow loss in the resonator. The optimum sizes of the resonator can not be estimated simply by the natural acoustic frequency. However, the ratio of the natural frequency of the resonator to that of the intake system for the open period of the intake valve may be used to estimate roughly the resonator effect. The resonator effect is considered to be slight in multi-cylinder engines.
A method automatically of measuring acoustic intensity is presented in this paper. This method uses measurements of the cross spectrum of the sound pressures at two closely spaced microphones. The two microphones used are arranged side by side and slid back and forth by a small distance in the direction of the microphone axes which are symmetric with respect to the holder axis. The holder can be turned on its axis by steps of a fixed angle for measuring the three directional components of an acoustic intensity vector. The acoustic intensity at a field point can be automatically measured by microphones arranged in this manner. This method is actually applied to the visualization of the flow pattern of sound energy emitted from a rectangular plate vibrating in resonance modes. As a result, an acoustic intensity probe composed of two microphones, as used in this study, has been proved to be an appropriate sensor for the automatic measurement of spatial acoustic intensity vectors.
A system using ultrasonic sensors was developed to track a moving target. In Part 1, a new technique for sharpening the ultrasonic beam, and for attaining a higher accuracy in the determination of the direction of a moving target was proposed. Furthermore, a new electronic technique for sweeping the emitted beam, for searching more rapidly a moving target in space, was proposed. In Part 2, an outline of the system and its experimental performance are described.
A system using ultrasonic sensor was developed to track a moving target. In Part 1, new techniques for sharpening the ultrasonic beam and for electronically sweeping the axis of emission of the beam, and their theoretical bases were described. In Part 2, the outline of the system is described. The performance verified by experiments is as follows: The maximum trackable velocity of a moving object, at a distance of 1.2m from the ultrasonic sensor system, is 480 mm/sec. The measurement error in the position of a target moving with this maximum velocity is 91 mm.
It is difficult to perfectly predict the fatigue strength of a bolt in joints at the stage of strength design, since many factors influence the load on the bolt. This study treats the possibility of preventing the fatigue fracture of a bolt by detecting the fatigue crack at the thread root at an early stage by means of an ultrasonic flaw detector. A method is discussed in which a crack is initiated at the bolt thread root near the top surface of the nut, in the mating threads, by shortening the bolt pitch slightly against the nut pitch. By using this method, it is hoped that the period from the detection of the crack to the fracture will be lengthened. From the results of fatigue tests, it is clarified that pitch modification is effective in the case of relatively low mean stress.
To clarify the seizure resistance and the frictional characteristics of rollers with various coatings, two roller tests on rollers coated with copper, graphite and molybdenum disulfide were carried out using an Amsler machine, and the flash temperature rise on the contacting surface when the rollers were seized was calculated. Further, the correlation between the test result of seizure resistance in the two roller machine and the previous test results in a power-circulating gear machine and in a Soda-type four ball machine was examined, and the frictional characteristics of the coatings wee estimated. From the experimental results, a correlation between the seizure resistance in the two roller tests and the scoring resistance in the gear tests was recognized. Also, the variation in the coefficient of friction in the two roller tests was different from that in the four ball tests.
The behaviour of tooth mesh frequency sound radiated from a tooth mesh point of a spur gear pair was investigated to see how it is generated. after examining the effect of bottom clearance, torque and material (steel or nylon) upon the radiation, the sound proved to be an aerodynamic sound; i.e., the sound is born from the pumping action of the clearance between meshing teeth, under high speed conditions. Especially towards the exit side of meshing, the radiation is prominent, with a "level trough" feature, which occurs due to the existence of two sorts of sounds having a phase difference of 180° near the meshing point. It has also become clear that this aerodynamic sound can not necessarily be reduced by increasing the clearance, after a consideration based on an analogy to the fluctuation of the flow rate of a gear pump.
The influences of material, heat treatment, tooth profile and surface roughness on the surface durability of surface-hardened gears were investigated. First, the effect of surface roughness on rolling/sliding contact fatigue was studied in detail by using a 2-roller-type contact fatigue testing machine. The results show that the initial surface roughness has a remarkable influence on the surface durability of surface-hardened steels, and that the decrease in surface roughness with running time does not always have an advantageous effect. Secondly, the load-carrying capacities of case-carburized gears and induction -hardened gears were investigated by using a power circulating-type gear testing machine. The phenomenon of "grey-staining", which consists of many micro-cracks and micro-pits, appears on the ground tooth surface under a heavy load. In these tests, it is clear that the surface failure "grey-staining", which causes degradation of the tooth profile and spalling, is most influenced by the initial tooth surface roughness. Furthermore, the effect of initial surface roughness on the surface durability of surface-hardened gears can be estimated quantitatively.
Accurate gears are needed in order to obtain a high load carrying capacity. This study aims at manufacturing highly accurate gears by hobbing. In this report, a method for the calculation of tooth profile errors, caused by geometrical hob errors such as profile errors of the hob tooth and the eccentricity of the hob, is proposed. From a comparison of the calculated tooth profile errors with the measured ones in hobbing tests, it is found that tooth profile errors generated by geometrical errors and the eccentricity of the hob can be distinguished from those by dynamic errors of the hobbing process, i.e. vibrations, built-up edges and so on. It is proved that the accuracy of this calculation method is sufficient to study the accuracy of hobbed gears.
The causes of tooth profile errors in gear hobbing are investigated by comparing computed tooth profile errors with measured ones, for manufacturing accurate gears. Under various hobbing conditions, gears are cut and measured. Spheroidal graphite cast iron is used as the hobbed test material because built-up edges hardly occur. From these comparisons, it is found that tooth profile errors are mainly caused by geometrical errors and the eccentricity of the hob. Tooth profile errors caused by dynamic errors of the hobbing process are very small under these conditions, though they are observed more or less in the case of helical gear hobbing. It is also confirmed that the method of calculating tooth profile error, which is discussed in the first report, is useful in analyzing gear accurcy.