Polarimetry for spatiotemporal photoelastic analysis

This paper describes a new photoelastic technique for the spatiotemporal stress analysis. In a polarimeter developed, an elliptically polarized signal beam of light, modulated in state of polarization by two-dimensional principal-stress distributions interferes with a reference beam of light consisting of orthogonal linearly polarized two components. A time-sequential series of two-dimensional interference patterns are received one after another by a MOS video camera, followed by a computer. Of the elliptically polarized signal beam, the orthogonal field components along the directions of the principal stresses in a two-dimensional photoelastic sample can be computed from a recorded interference pattern, which offer the data needer for mapping the spatiotemporal principal-stress distribution over the sample. Not only each of the two orthogonal principal stresses but also the principal-stress difference are mapped in a time-sequential diagram. No use of any movable polarization element such as a rotating analyzer allows us to follow a rapid change in stress distribution within the maximum frame rate 2066 s⁻¹ of the MOS video camera.     

On the existence of stress concentrations in loaded bodies made of polycrystalline materials

We study the character of stress distributions near the corner point of the interface between the two joined crystals. The interface forms a dihedral angle. The joined crystals have a cubic symmetry and consist of the same material. They have a single common principal direction of elasticity, which is parallel to the edge of the dihedral angle. The other principal directions do not coincide and are oriented arbitrarily.In the framework of elasticity, we consider problems of out-of-plane and plane strain of the twocrystal. We show that, in the case of longitudinal shear in the direction of the common principal axis of elasticity, there is no stress concentration near the corner point of the interface between the two joined crystals.For the case of plane strain in which all displacements and strains occur only in the planes perpendicular to the common principal direction, we use separation of variables to construct the characteristic equation that determines the stress concentration degree and find the roots of this equation, which determine the order of singularities of the stresses.     

Two new expanding cavity models for indentation deformations of elastic strain-hardening materials

Two expanding cavity models (ECMs) are developed for describing indentation deformations of elastic power-law hardening and elastic linear-hardening materials. The derivations are based on two elastic–plastic solutions for internally pressurized thick-walled spherical shells of strain-hardening materials. Closed-form formulas are provided for both conical and spherical indentations, which explicitly show that for a given indenter geometry indentation hardness depends on Young’s modulus, yield stress and strain-hardening index of the indented material. The two new models reduce to Johnson’s ECM for elastic-perfectly plastic materials when the strain-hardening effect is not considered. The sample numerical results obtained using the two newly developed models reveal that the indentation hardness increases with the Young’s modulus and strain-hardening level of the indented material. For conical indentations the values of the indentation hardness are found to depend on the sharpness of the indenter: the sharper the indenter, the larger the hardness. For spherical indentations it is shown that the hardness is significantly affected by the strain-hardening level when the indented material is stiff (i.e., with a large ratio of Young’s modulus to yield stress) and/or the indentation depth is large. When the indentation depth is small such that little or no plastic deformation is induced by the spherical indenter, the hardness appears to be independent of the strain-hardening level. These predicted trends for spherical indentations are in fairly good agreement with the recent finite element results of Park and Pharr.     

The evolution of caustics from holes to cracks

The method of caustics was used for the study of the evolution of stress concentration around a circular hole, which progressively changes in shape and becomes an elliptic hole, tending to an internal crack. The influence of the amount of ellipticity of the holes and their orientation relative to the axis of the applied external loads at infinity on the form of caustics created around the discontinuity was studied, as the elliptic holes tended to become internal cracks. A series of experiments with tension specimens containing small elliptic holes of any ellipticity and orientation was performed. Comparison of experimentally obtained caustics with theory yielded a good agreement of both results. Finally, the use of small elliptic holes drilled all over a biaxial stress field for the determination of the individual principal stresses and the principal directions at the area of the holes was outlined.     

Measurement of the Residual Stress Tensor in a Compact Tension Weld Specimen

Neutron diffraction measurements have been performed to determine the full residual stress tensor along the expected crack path in an austenitic stainless steel (Esshete 1250) compact tension weld specimen. A destructive slitting method was then implemented on the same specimen to measure the stress intensity factor profile associated with the residual stress field as a function of crack length. Finally deformations of the cut surfaces were measured to determine a contour map of the residual stresses in the specimen prior to the cut. The distributions of transverse residual stress measured by the three techniques are in close agreement. A peak tensile stress in excess of 600 MPa was found to be associated with an electron beam weld used to attach an extension piece to the test sample, which had been extracted from a pipe manual metal arc butt weld. The neutron diffraction measurements show that exceptionally high residual stress triaxiality is present at crack depths likely to be used for creep crack growth testing and where a peak stress intensity factor of 35 MPa√m was measured (crack depth of 21 mm). The neutron diffraction measurements identified maximum values of shear stress in the order of 50 MPa and showed that the principal stress directions were aligned to within ~20° of the specimen orthogonal axes. Furthermore it was confirmed that measurement of strains by neutron diffraction in just the three specimen orthogonal directions would have been sufficient to provide a reasonably accurate characterisation of the stress state in welded CT specimens.     

Effective and relative permeabilities of anisotropie porous media

A model composed of a three-dimensional orthogonal network of capillary tubes was used to simulate the flow behavior in an unsaturated anisotropic soil. The anisotropy in the network's permeability was introduced by randomly selecting the radii in the three mutually orthogonal directions of the network tubes from three different lognormal probability distributions, one for each direction. These three directions were assumed to be the principal directions of anisotropy. The sample was gradually drained, with only tubes smaller than a certain diameter remaining full at each degree of saturation. Computer experiments were conducted to determine the network's effective permeability as a function of saturation. The main conclusion was that the relationship between saturation and effective permeability depends on direction. Consequently the concept of relative permeability used in unsaturated flow should be limited to isotropic media and not extended to anisotropic ones.     

Stress measurement using vertically polarized shear waves

This paper proposes a method to evaluate surface stresses in orthotropic materials by the use of vertically polarized shear waves (SV waves). It is assumed that the normal to the surface coincides with one of the axes of anisotropy, and that the material anisotropy is not necessarily small. The speeds of the waves are expressed in terms of the material properties, the stress, the rigid body rotation and the propagation direction. From the expressions, and for the following two cases: (1) the rigid body rotation is known, (2) the anisotropy is weak, it is possible to determine the components of the surface stress by measuring the speeds of SV waves propagating in several directions. When the anisotropy is weak, the acoustoelastic birefringence for SV waves is also derived, to separate the material anisotropy and the difference of the principal stresses. The theory can be applied to stress evaluation by using ultrasonic Lamb waves whose speeds are nearly equal to those of SV waves.     

QUASI-PRINCIPAL AXIS METHOD IN FINITE DEFORMATION

ＱＵＡＳＩ－ＰＲＩＮＣＩＰＡＬＡＸＩＳＭＥＴＨＯＤＩＮＦＩＮＩＴＥＤＥＦＯＲＭＡＴＩＯＮＺｈｅｎｇＱｉａｎｓｈｕｉ（郑泉水）（ＤｅｐａｒｔｍｅｎｔｏｆＥｎｇｉｎｅｅｒｉｎｇＭｅｃｈａｎｉｃｓ，ＱｉｎｇｈｕａＵｎｉｖｅｒｓｉｔｙ，Ｂｅｉｊｉｎｇ１０００...     

A dynamic photoelastic study of stress-wave propagation through an inclusion

A photoelastic study of the elastodynamic-stress fields around a circular, elastic inclusion (Solithane 113) embedded in an elastic plate (Hysol 4485) is presented. The edge of the plate was loaded by an explosive charge, which produced a plane, compressional stress wave of triangular shape. Isochromatic-fringe patterns were obtained, which give the maximum shear stresses, both inside the inclusion and in the surrounding medium. The principal stresses on the axis of symmetry were determined through the use of the oblique-incidence method. It was found that small tensile stresses are generated at the interface on the shadow side of the inclusion. The focusing effect inside the inclusion predicted by ray theory was not observed. Finally, the shape of the wavefront as the wave passes the inclusion was determined.     

New methods in photoelasticity

The principle of polarization of scattered light is applied to determine the principal stresses in the interior of a model. On the basis of the theorem which states that “a series of birefringents is equivalent to a unique birefringent, followed by a medium endowed with rotational power,” it can be assumed that, if the characteristics of a series of birefringents are known, it is possible to find the characteristics of an interior section. The measurement of the characteristics of a birefringent (eventually following a medium endowed with rotational power) can be accomplished by means of the new methods, making use of a photomultiplier, a constant-speed rotating analyzer and a servomechanism These new methods of measurement are applicable to two-dimensional photoelasticity.     

Measurement of Residual Stresses in Nuclear-grade Zircaloy-4(R) Tubes—Effect of Heat Treatment

Nuclear-grade Zircaloy-4(R) tubes are produced by a unique manufacturing process known as pilgering, which leaves the material in a work-hardened state containing a pattern of residual stresses. Moreover, such tubes exhibit elastic anisotropy as a result of the pilgering process. Therefore, standard equations originally proposed by Sachs (Z Met Kd, 19: 352–357, 1927; Sachs, Espey, Iron Age, 148: 63–71, 1941). for isotropic materials do not apply in this situation. Voyiadjis et al. (Exp Mech, 25: 145–147, 1985) proposed a set of equations for treating elastically anisotropic materials, but we have determined that there are discrepancies in their equations. In this paper, we present the derivation for a set of new equations for treating elastically anisotropic materials, and the application of these equations to residual stress measurements in Zr-4(R) tubes. To this end, through thickness distribution of residual stress components in as-received and heat treated (500°C) Zr-4(R) tubes was measured employing the Sachs’ boring-out technique in conjunction with electrochemical machining as the means of material removal, and our new equations. For both as-received and the heat treated materials, the axial and tangential residual stresses were significantly higher than the radial and shear residual stresses. The largest residual stress was the tangential stress component in the as-received material, showing a tensile value at the outer surface and a compressive value at the inner surface. At high values of von Mises equivalent stress, the principal directions of residual stress coincided with the principal axes of the tube for the as-received material, as well as for the material heat treated at 500°C.     

Analysis of residual stresses in cylindrically anisotropic materials

Metal-forming operations leave residual stresses in formed parts due to nonuniform deformation occurring during the process. An exact method of determining the longitudinal, radial and circumferential (tangential) residual stresses in axisymmetric specimens was proposed by Mesnager¹ and further developed by Sachs². The boring-out technique can be complemented by a similar procedure in which strains are measured on the inner surface of the tube when material is removed from the outer surface.The work proposed in this paper extends previous analyses of residual stresses to the case where the material exhibits cylindrical elastic anisotropy, i.e., the principal axes of anisotropy correspond to the longitudinal, radial and circum-ferential directions of the tube. In addition, the present analysis considers the case in which a residual-shear stress, developed by twisting the tube about its axis, exists in the tube. When such shearing stresses are present, the principal axes of the residual-stress distribution are not parallel to the principal axes of the tube.     

Validation of Mechanical Strain Relaxation Methods for Stress Measurement

Most validation studies of mechanical strain relaxation (MSR) methods for residual stress measurement rely on using the saw-tooth residual stress distribution resulting from four point bending and elastic–plastic deformation. Validation studies using simple applied stress profiles in rectangular steel beams are used in this work, together with beams subjected to elastic–plastic bending. Two MSR methods are explored, deep-hole drilling (DHD) and incremental centre hole drilling (ICHD). As well as a series of experiments, finite element analyses are conducted to determine the accuracy in the inversion of measured deformation to reconstruct stress. The validation tests demonstrated that apart from the applied stresses, the initial residual stresses also contribute even when samples are expected to be stress free. The uncertainty in measurement for the two MSR methods is determined, with the uncertainty in near surface measurement found to be significantly larger than uncertainty for interior measurement. In simple loading cases (and simple stress profiles) the uncertainty in measurement and hence the degree of validation is shown to be within about ±50 MPa for steel for “known” stress up to about 140 MPa. However, if the residual stress distribution is more complex there arises increased uncertainty in the predicted residual stress and lack of confidence between measurements methods.     

Acoustical birefringence and the use of ultrasonic waves for experimental stress analysis

Stress-induced optical birefringence in transparent materials has long been a common technique of stress analysis. Although stress-induced acoustic birefringence was discovered more than 20 years ago, its development and actual applications are still limited. This paper will look at the similarities and differences between the propagation of light waves in photoelastic materials and the propagation of ultrasonic waves in deformed solids. Critical comparisons of the experimental methods employed in photoelasticity with those available in modern ultrasonic measuring technique show why previous studies on ultrasonic measurement of stresses were not very successful. A new experimental technique is devised for using ultrasonic waves for stress analysis. The technique employs a single rotatable 10-MHz shear transducer as the transmitter and receiver of ultrasonic pulses. The enlarged display of the 10-MHz modulated-pulse pattern of reflected echoes provides a convenient way to determine the directions of principal axis of the stress within ±3 deg. The pulse-echo-overlap method is used to measure the absolute velocities of the two principal shear waves. The difference in principal stresses is then calculated from the velocity measurements. Test results of common structural-aluminum and steel specimens under uniaxial compression show a linear relation between the velocity changes and the applied stress. Ultrasonic measurements of stress distribution in a 6.35-cm diameter, 1.9-cm-thick aluminum disk under diametric compression are also reported. Paper was presented at Third SESA International Congress on Experimental Mechanics held in Los Angeles, CA on May 13–18, 1973.     

The effect of low-frequency biaxial loads on stress-concentration factors in plates

It is generally recognized that stress-concentration factors under stress-wave loading are lower than those under static loads. In this work, the effect of low-range frequency of biaxial sinusoidally varying alternating stresses on the stress-concentration factors for circular and elliptical holes in Plexiglas plates is investigated. The experiments have been performed on a specially designed and built “biaxial cyclic-stress machine” and the results are presented in the form of curves. In the case of biaxial alternating stresses, the stress-concentration factor is defined as the ratio of amplitude of the maximum alternating stress around the geometrical discontinuity to the larger of the amplitudes of the two principal alternating stresses which would occur at the same point, if the geometrical discontinuity was not present. Both values are considered over a stress cycle. The results indicate a slight decrease in the values of stress-concentration factors with increase in frequency.     

The effect of residual stresses on hardness measurements

The RockwellC hardness,RC, was measured as a function of position on steel rings with different residual-stress profiles through the thickness. An experimental correlation between residual stress andRC was obtained. A relationship between the average pressurep of a spherical indenter, the yield strengthS _y and the residual stress of the material was conceived and used in fitting the experimental data. In order to model the effects of residual stresses on the measured hardness, the von Mises-Hencky (power) yield criterion was utilized, together with an adaptation for residual stresses of the expression for the stress state under a spherical indenter, given in Shaw, Hoshi and Henry. A parameter α was introduced in our calculations to account for the effect of the nonperpendicularity of the residual stresses to the pressurep of the spherical indenter. The proposed model in large measure fits experimental hardness versus residual stress data, and results are consistent with different samples. This model can be used as a basis for the measurement of residual stresses in steel or other materials.     

Graphical methods for determining the resulting photoelastic effect of compound states of stress

The aim of this work is to give a simplified representation of an interesting subject, namely, the calculation of the photoelastic effect behind a system in which the principal stresses or secondary principal stresses rotate about the direction of the propagation of light. The methods are described in a popular form for understanding the essential features and the necessary graphical operations for solving these complicated problems. Further, the relations between the different methods are shown. Thej-circle technique is improved in order to simplify the operations and a new possibility in applying Wulff's grid is introduced in photoelasticity. The graphical tools can then be applied as so-called “rotation rules.” Some examples related to recent papers which present theoretical results or new methods are given in order to study the function of these “rotation rules” and to recognize the power of these methods.     

Iteration method in linear elasticity of random structure composites containing heterogeneities of noncanonical shape

We consider a linearly elastic composite medium, which consists of a homogeneous matrix containing a statistically inhomogeneous random set of heterogeneities of arbitrary shape. The general integral equations connecting the stress and strain fields in the point being considered with the stress and strain fields in the surrounding points are obtained for the random fields of heterogeneities. The method is based on a recently developed centering procedure where the notion of a perturbator is introduced and statistical averages are obtained without any auxiliary assumptions such as, e.g., effective field hypothesis implicitly exploited in the known centering methods. Effective elastic moduli and the first statistical moments of stresses in the heterogeneities are estimated for statistically homogeneous composites with the general case of both the shape and inhomogeneity of the heterogeneities moduli. The explicit new representations of the effective moduli and stress concentration factors are built by the iteration method in the framework of the quasicristallite approximation but without basic hypotheses of classical micromechanics such as both the EFH and “ellipsoidal symmetry” assumption. Numerical results are obtained for some model statistically homogeneous composites reinforced by aligned identical homogeneous heterogeneities of noncanonical shape. Some new effects are detected that are impossible in the framework of a classical background of micromechanics.     

Nonlinear buckled states of rectangular plates

A constructive method is developed to establish the existence of buckled states of a thin, flat elastic plate that is rectangular in shape, simply supported along its edges, and subjected to a constant compressive thrust applied normal to its two short edges. Under the assumption that the stress function and the deformation of the plate are described by the nonlinear von Kármán equations, the approach used yields information regarding not only the number of buckled states near an eigenvalue of the linearized problem, but also the continuous dependence of such states on the load parameter and the possible selection of that buckled state “preferred” by the plate. In particular, the methods used provide a rigorous approach to studying the existence of buckled states near the first eigenvalue of the linearized problem (that is, near the “buckling load”) even when the first eigenvalue is not simple.