This paper is intended to give an exemplary review of recent investigations performed in the X-ray laboratory of the Institut für Werkstoffkunde I of the Universität Karlsruhe, FRG, concerning particular problems of residual stresses of heat-treated and fatigued steels. The experimental work was mainly performed with computer-controlled Karlsruhe type ψ-diffractometers. If linear distributions of residual lattice strains occurred the sin2ψ-method was applied to determine residual stresses. The experiments were performed with plain carbon steels of 0.22 and 0.45 wt.-% carbon (German grade Ck 22 and Ck 45) and some low alloyed steels.

The growth of fatigue crack builds up the compressive residual stress with the plastic zone in front of the crack tip. The presence of this residual stress is one of the main causes for the crack closure. The consequence of residual stress is a reduction of the crack tip stress intensity variation during a load cycle. Although this effect is well understood in a qualitative manner, it has not been completely quantified because of a lack of information about the load and residual stress distributions at the crack tip.

One of the most detrimental effects on the accuracy of an X-ray diffraction residual stress analysis, XRDRSA(l), is found in the examination of textured materials. The degree of elastic anisotropy and texture is in general agreement with the extent of the error in the residual stress. Several approaches have been made to correct for the effects of texture, particularly involving experimental techniques. Reviews of such efforts are given by H. D811e(2), v.M. Hauk﹛3) and G. Maeder, J.L. Lebrun and J.M. Sprauel (4), just to mention a few.

A brief chronology of the texture corrections involved in XRDRSA follows. With isotropic materials the d spacing of a crystal lattice, d, is assumed to vary linearly with sin2ψ. With textured materials the d vs sin2ψ relationship is nonlinear. This is due to the anisotropy of the elastic constants and their departure from a random distribution, or taking on a preferred orientation.

It is well knovm that a characteristic plastic flow occurs when two discs are rolled together at loads exceeding the yield point; even in the case of rolling contact without slip, a deformed layer shows marked plastic flow with respect to the direction opposite to the rolling direction. This phenomenon was first found by Crook and Welch. Later, Johnson and his co-workers explained why the forward flow occurs under pure normal stresses.

The aim of the present study is to reproduce the various types of plastic flow under different testing conditions of rolling contact, such as nearly pure rolling contact, rolling contact by friction drive, and rolling-sliding contact. After a test of repeated rolling contact, the mode of plastic flow and the residual stress distribution at the surface of the specimen were observed. Further, the experimental data were compared with the result of analytical calculation by the method of Merwin and Johnson.

X-ray fractography is a new method utilizing the X-ray diffraction technique to observe the fracture surface for the analysis of the micromechanisms and mechanics of fracture. The line broadening of X-ray diffraction profiles and the residual stress are two of the important X-ray parameters. Among them, the X-ray residual stress has been confirmed to be particularly useful for the fracture surfaces of high strength steels, and applied to the fracture surface of fracture toughness specimens and the fatigue fracture surface.

The strength of ceramics or glasses can be increased by placing their surfaces into compression. Techniques include ion exchange, temperature glazing, surface chemical reactions and stress-induced phase transformations. Although most of these techniques are well recognized, little effort has been expended In experimentally determining the magnitude of the compressive stress, and in particular, to use experimental evidence to identify important material and process parameters that need to be controlled. The goal of this investigation was to determine some of the factors that effect the magnitude, profile and depth of the compressive layer introduced by a structural phase transformation. X-ray residual stress measurements were used to directly determine the state of the surface residual stress.

The trend towards smaller devices and packages in the semiconductor industry makes it increasingly important and increasingly difficult to obtain useful X-ray diffraction information from the small areas employed. This study covers applications of a Rigaku micro-diffractometer to measure strain and obtain phase information from areas less than 100μm in diameter. Examples of residual stress mappings between electrical vias only 100μm apart and phase determination on a single electrical via 120μm diameter will be shown.

Cemented carbide composites with a WC hard phase and a Co-Ni alloy binder phase have been subjected to monotonic and cyclic deformation and studied using the high resolution General Purpose Powder Diffractometer at the Argonne Intense Pulsed Neutron Source. Upon deformation, relaxation of bulk differential thermal residual stresses, tensile in the binder and compressive in the carbide, is observed to occur as a function of loading mode and plastic strain via shifts in diffraction peak positions. In addition, peak breadth behavior indicates broadening. At low plastic strain this is due primarily to a range of residual stress and at high plastic strain it is attributable to the plastic deformation alone since relaxation is essentially complete. The appropriateness of neutrons is discussed.

At the present time there is considerable interest in the use of adhesive bonded joints for structural applications. The design of such joints is based mostly on finite element calculations and failure tests rather than on strain or stress measurements. To our knowledge no measurements have been made of stresses at or near the adhesive/adherend interface (where many bond failures occur) because of difficulties with accessibility. The purpose of this work was therefore to try to make such measurements using X-ray diffraction.

The approach taken was to gain access to the bond by making one of the adherends in a single lap joint to be relatively transparent to X-rays and the cither relatively opaque. Incident X-rays then penetrated the first adherend and the adhesive and were diffracted from gains in the second adherend adjacent to the adhesive/adherend interface. Stresses resulting from curing of the adhesive were determined first, after which a load was applied and the stresses redetermined.

The interfacing of both straight and curved Braun Position- Sensitive Proportional Counters (PSPC's) to a high resolution Huber-Guinier camera system has been accomplished, resulting in a 10 to 100-fold decrease in data collection times when compared to conventional Guinier (film or scintillation counter) detector techniques.

Various factors causing line broadening were evaluated for both PSPC Guinier systems. The depth of the PSPC gas chamber was found to Have the greatest influence on line profiles. An R0% increase in peak half-widths was observed for PSPC-Guinier data compared to our highest resolution Guinier film data, but still yielded significantly better resolution than conventional powder diffractometer data obtained in our laboratory.

In order to increase the versatility of in-situ X-ray equipment for stress measurements, we decided to design and construct our own easily transportable apparatus. Similar equipments are already commercialized: (i) in Japan (RIGAKU STRAINFLEX, SHIMADZU); (ii) in West Germany (SIEMENS); (iii) in the USA (TEC). After the commercialization of PSD or PSSC, several prototypes were developed among which, the apparatus of Ruud, James, CETIM… Our own apparatus has been developed for two main reasons:

1. (i) For 1971, we have been working to ameliorate our own equipments effectively and frequently used for in situ measurements.

2. (ii) We need a very precise and rapid apparatus amainable for the greatest number possible of applications.

The Pennsylvania State University has developed and tested a miniature x-ray instrument for the measurement of residual stress (strains). The stress head including x-ray source and detection surface is approximately four and one-half inches long and can be inserted into an orifice less than two and one-half inches in diameter. This head is on the end of a several-foot long cable and is extremely manipulatable.

The instrumentation design is based upon a unique position sensitive scintillation detector which ie capable of applying the single-exposure x-ray stress measurement technique. Total lapse time data collection periods for aluminum alloys are less than ten seconds with x-ray tube powers of 120 watts. Results from a four-point bend test on an aluminum alloy specimen showed excellent precision. The configuration of the instrument and the procedures for its application are described.

X-Ray diffraction studies of substances under controlled atmospheres and at different temperatures are of great importance in many research areas. This is especially true in the area of catalysis, where the correlation of structural and catalytic properties is needed. The camera described here was made for this purpose although any sample in the powdered form can be studied as well. Many catalysts are in the form of highly dispersed metal deposited on a granulated support. The content of metal may vary from a few tenths to several percent. In a camera used by Janko and Borodzinski a small amount of catalyst was spread out on a porous silica glass sample holder which then was placed in a high temperature XRD camera with flowing gas of controlled composition.

General Electric (G.E.) and Picker generators are currently installed in many research laboratories all over the United States. These units are old and do not satisfy the modern requirements for X-ray generators. Instabilities of high voltage and tube current complicate the experiment. To maintain G.E. and Picker generators is becoming more and more expensive due to the high cost of vacuum tubes and replacement parts. To replace old generators with new high performance commercial units can cost even more and requires additional adaptations. In the present paper another alternative is proposed, which assumes the replacement of vacuum tube mA and KV controls with solid state circuits using the old G.E. or Picker high voltage transformers. This alternative allows rejuvenation of old generators without difficulties of reinstallation and does not require adaptations to the existing experimental X-ray equipment. The combination of old high voltage transformers with modern solid state control devices provides high performance at low cost. For this purpose the control device was designed and is described below.

Given the vast superiority of synchrotron radiation sources for EXAFS and near-edge measurements, one might assume that conducting such experiments in the laboratory is a waste of time and resources. Upon more careful consideration of the realities of everyday research, however, one can easily see that this is not true. In fact, many people have come to this realization, so much so that a whole conference has been devoted to laboratory EXAFS facilities, their role in research and their relation to synchrotron facilities. The concensus after the conference was that laboratory instruments can and have been developed, with adequate performance for a variety of nontrivial experiments. They usually employ dedicated focusing spectrometers, rotating anode generators and fairly expensive monochromator crystals (see articles in ref. 1 and 2).

We have explored the merits of using a Hgl2 spectrometer as a detector in x-ray diffraction systems instead of a proportional gas counter, or a scintillation counter. The full width at half maximum energy resolution of the HgI2 spectrometer used was about 1.1 keV for the CuKα line (8.1 keV), and about 1.5 keV for the MoKα line (17.4 keV), The energy resolution was utilised to eliminate x-ray fluorescence background from powder diffraction spectra. We demonstrate the suppression of Fe x-ray fluorescence in diffraction patterns of ErFe03 obtained with a Cu x-ray tube, and of Y x-ray fluorescence in diffraction patterns of Y2O3 obtained with a Mo x-ray tube. The peak height to background ratios were improved by about an order of magnitude in both cases.

Digitized data acquisition and subsequent peak-profile fitting are used in aligning and maintaining the Guinier diffraction system in this laboratory. Utilization of a stepping motor driven goniometer along with profile analysis allows rapid and reproducible alignment while providing a means to investigate the types and magnitudes of errors introduced by various misalignments of the optical system.

A wide range of structural studies have been carried out in high pressure diamond anvil cells using x-rays. The most common experimental geometry is shown in Fig. 1a. The incident x-ray beam passes axially through the first diamond and enters the sample, typically 100-300 μm in diameter and 20-100 μm thick; the diffracted x-rays exit via the second diamond. Energy-dispersive detection techniques (EDXRD) have been used. However the intensity of diffracted radiation from the sample is weak, so that typical exposure times with a conventional, fixed anode, x-ray source are typically one to several days.Accordingly, higher intensity radiation from synchrotron sources has been used for these experiments.

Interest in the reference intensity ratio CRIR) method continues as a result of the potential capabilities of the technique for rapid multi-component quantitative analysis. The theoretical basis for the RIR technique is now well established (Chung, 1974; Hubbard et al., 1976; Davis, 1980, 1981; Davis and Johnson, 1982), Major areas for which the method can still be greatly improved include the methods used for sample preparation, the measurement of accurate intensities, and the use of an internally consistent set of reference intensity constants (designated RIR, or ki). In the methodology developed at the Institute of Atmospheric Sciences (IAS), South Dakota School of Mines and Technology, sample preparation centers about the suspension of the pulverized sample into an aerosol and collection onto filter media. Because of this step, intensities must be corrected from their raw values to intensities representative of “infinite thickness“ and volumetrically constant conditions of normal sample diffraction. The measurement and correction of intensities and sample preparation methodology is the subject matter of the present paper.

Our experience using Mo Kα radiation in coal related research indicates that minerals can be detected and quantitatively measured at the 0.2% level in an amorphous substance such as coal. Since the linear absorption coefficient for most elements is about a factor of eight lower for Mo radiation than it is for Cu radiation, the matrix affect is greatly reduced and the intensity is a more linear function of concentration.

Although there is considerable overlapping of peaks when complex mixtures are analyzed using Mo radiation, the use of second derivative techniques to locate peaks, and least squares techniques to fit peaks can resolve many of these problems. The reduced time needed to collect a diffraction pattern, and the increased intensity obtained more than compensate for the extra computer time needed to analyze the data.