X-ray diffraction observation of metal fractures provides fracture analysists with useful information on the mechanisms and mechanical conditions of fracturing. This method is called “X-ray fractography” and has been developed especially in Japan as a new engineering tool for fracture analysis.

In the present paper, X-ray fractography is applied to fracture surfaces of ductile cast iron (JIS FCD 60) which are widely used as machine parts. The fracture toughness tests were conducted at ambient and low temperatures by using compact tension (CT) specimens with blunt notches and three-point bending (TPB) specimens with fatigue pre-cracks. The line broadening of X-ray diffraction profiles was measured on and beneath fracture surfaces of fracture toughness specimens.

The flow and fracture behavior of ceramic and other brittle materials under the influence of contact loading is important to both component fabrication and performance. The ease of machining, severity of residual surface damage and rate of wear during subsequent service are controlled to a large degree by the character and extent of the flow zone and its influence on the fracture mode. This Investigation was undertaken to provide experimental verification of the results obtained through elastic/plastic finite element modeling cf the stress distribution and deformations introduced by static contact loading. Experimentally, X-ray double-crystal diffractometry (DCD) was applied to obtain a mapping of the distortions produced beneath a Vickers indenter, and hence to evaluate the effect of material and geometric parameters on the flow and fracture mechanisms.

An x-ray diffraction study of the reversible deformation inodes associated with the shape memory effect has been carried out on a series of uranium-niobium alloys near the monotectoid composition (6.2 wt. % Nb). Diffraction patterns were measured as a function of strain, in situ, while the specimens were under stress as part of an attempt to explain the “easy-flow”, low-strain plateau in the stress-strain curve. The alloys, consisting of highly twinned, metastable α” (monoclinic) and γ° (tetragonal) phases derived from the high-temperature BCC γ phase, produced broad, overlapping diffraction lines difficult to analyze by conventional techniques. One solution to this problem was to use a segmented step-scan technique so as to apportion the scan time to concentrate on the most difficult regions. This paper discusses data obtained from an α” alloy and a dual-phase α” + γ° alloy.

Synchrotron radiation (SR) has several unique properties which cause it to be used for many purposes in science and technology. It is continuous spectrally, collimated spatially, short-pulsed yet continuously-operating temporally, as well as polarized( 1). Many applications of SR have already been demonstrated. These range from production of fine scale structures by lithography to numerous studies of materials. Determination of atomic structure by diffraction and other scattering experiments and by absorption spectroscopy (EXAFS), and emission spectroscopic studies of electronic structure have been given most attention(2). Only limited attention has been paid to determination of materials composition by x-ray fluorescence analysis (XRF) employing synchrotron radiation. Sparks et al.(3-5) have performed the most notable experiments, examining mica inclusions for the presence of primordial super heavy elements and irradiating two National Bureau of Standards Standard Reference Materials (SRM 1571, Orchard Leaves and SRM 1632, Coal) to measure the fluorescent x—ray intensity from the trace elements.

The performance of a room temperature mercuric iodide X-ray detector was investigated as a function of detector bias, amplifier time constant, and detector temperature. A Mn Kα line of 200 eV FWHM was obtained by using low noise electronics developed for Si(Li) detectors, including a cooled input PET. Measurements of the detector's resolution at various X-ray energies result in a Fano factor of 0.20.

Recent, state of the art, x-ray spectrometers have made use of polarizing the source x-rays by scattering through 90° (1) . One then observes the analyte fluorescence in a direction perpendicular to the scattering plane in which the polarized x-rays are generated. The signal/noise ratio at the detector is much improved. Unfortunately there is a concomitant loss of intensity and analysis times increase. This adversely affects the minimum detection limits.

In this work we plan to describe a new simple technique for exciting characteristic X-rays for X-ray fluorescence. This technique was originally designed for use with β-emitting radioisotopes (1) but now has been extended to other types of isotopes.The technique involves mixing directly with the sample (about 20mg) a small amount of a radionuclide (about 100 μCi). An equal atomic mixture of the following elements was prepared: Ti, Mn, Zn, BrfAg, Sn,and Ba. This mixture was investigated with the following isotopes: Ni-63, S-35, Pm-147, Na-22, Co-137, I-125, Co57, Cd-109, Zn-65, Fe-55 and Am-241. A composite table of the corrected integrated Kα—lines of the elements in the mixture and Graphs showing the ratio of the Ka line intensity of the element of interest divided by the Kα line intensity of Ti will also be presented.

Earlier, I reported in this conference the results obtained with (i) a three crystal spectrometer (ii) two curved crystal spectrometer of Cauchois type and (iii) spherically bent crystal spectrometer. These instruments were developed to obtain higher resolution that helped us to observe fine structures of Kα1, Kα2 lines of transition elements. Our work has been followed in other laboratories.

Here I present the essential points of a very simple spectrometer hitherto unreported. The most attractive feature of the spectrometer is its simplicity of adjustment in a minimum time.

A series of FORTRAN IV programs have been written to aid in qualitative x-ray fluorescence (XRF) analysis. These programs have been implemented on a DEC PDP 11/34 computer with an RSX-11M operating system and access the NIH elemental data base (1). The programs and their application to XRF analysis in our laboratory will be described.

Two principal mathematical methods are used for quantitative XRFA: fundamental parameter calculations and the evaluation of empirical parameter equations. A comprehensive computer program based upon fundamental parameter equations was introduced in 1976 by D. Laguitton and M. Mantler (LAMA-I) and improved by T. C. Huang in 1979 (LAMA-II). The present paper describes the features of the theoretical background of a computer program using a new type of empirical (alpha*-) parameter equations. It is essentially designed for convenient analysis of compounds including those containing chemical elements, that cannot be directly measured by conventional X-ray spectrometers, such as oxides, nitrides, and others. The program also communicates automatically with LAMA in order to establish theoretical tables of alpha*-coefficients as well as conventional alpha-coefficients.

The x-ray fluorescence (XRF) method is well known for its capability to perform fast and accurate quantitative analysis for all elements with atomic numbers greater than ten. Energy dispersive x-ray fluorescence (EDXRF) adds to this capability the benefit of quick qualitative analysis, due to its simultaneous sensitivity to all the elements. The method has the potential for rapid and complete chemical analysis of any sample which arrives on the analytical chemist's doorstep. Although the method has been a productive tool for fast and accurate repetitive analysis of similar samples, its applicability to unique unknowns has been rather limited. The limitation arises from the usual need to calibrate the instrument's response with a set of 6 to 12 standards, whose compositions must be similar to the unknown sample. Anyone who has struggled to develop and maintain such a suite of accurately certified standards knows that a great deal of effort and expense is involved. This effort is well justified when the analyst expects to analyze the same type of material frequently over an extended tine period. However, for a unique sample analysis, the task of developing a suite of similar standards simply makes the analysis impractical. What is needed is a method that requires minimal standards, or uses no standards at all.

The XRF11 program by John Criss and the EXACT program are two commercially available fundamental parameters programs for energy dispersive x-ray fluorescence spectrometry (EDXRF). These programs are both based on the same underlying equations but use different approaches to the calculations. The EXACT program assumes monochromatic excitation, and the XEF11 program models polychromatic excitation sources. There are also great differences in how the two programs approach the iterations in the calculations and in how the data from standards are used in the two programs for calibration.

To produce the monochromatic excitation neeiled by the EXACT program, secondary targets have been the preferred method. But it is also possible to approximate monochromatic excitation by using filtered direct excitation. The purpose of this study is to compare the data obtained from both secondary targets and direct filtered excitation as processed through both XRF11 and EXACT.

An energy-dispersive x-ray fluorescence method has been developed for the direct determination of major and minor elements in infinitely thick samples of paint. Matrix absorption and enhancement corrections are iteratively calculated from a knowledge of tabulated fundamental parameters and the unknown weight fractions. An estimate of the significant light element fraction of the bulk sample required for the calculation of matrix attenuation is obtained using the scatter peaks of the silver secondary target. Relative elemental calibration constants and calibration factors for the coherent and incoherent peaks are determined experimentally using either thin-film standards or standards of known total composition. For routine analysis only one absolute standard is required. The method has been applied to different types of paint with a relative standard deviation better than 5% provided the counting statistics are not the limiting factor. The accuracy has been tested by comparing own results with those obtained either from the formulation or from, instrumental neutron activation analysis.

A new approach to fundamental parameter calculations has been devised which makes use of incoherent and coherent backscatter intensities from the excitation radiation. The backscatter results from all sample constituents, and thus provides information on total sample mass as well as on bulk sample composition. By appropriately using the scatter intensities in fundamental parameter matrix calculations, accurate analyses of “unknown“ samples can be obtained without prior knowledge of the sample matrix. The backscatter with fundamental parameter (BFP) method is especially advantageous for biological, geological and environmental samples because of their bulk quantities of carbon, oxygen, and other light elements which cannot be explicitly determined for traditional fundamental parameter calculations.

Low-energy x-rays can undergo considerable absorption in the sample material when EDX spectrometry is used in multi-element analysis of aerosol-loaded filters. An expression which corrects for the effects of absorption, based on counting aerosol filters and standards on both sides, is derived. The method is experimental verified by using the absorption of the 60 keV gamraa-ray of 241Am in aluminium and also by comparison with PIXE results for environmental filters.

The standard addition method is often useful for singleelement analyses in matrices of unknown composition or those difficult to reproduce. Results are good at low concentrations where concentration vs. intensity is approximately linear, but serious errors occur at higher concentrations. A method is shown here for greatly extending the range of the standard addition approach by combining it with with matrix correction models. Expressions incorporating the well-known alpha-or beta-correction coefficients are derived for the observed intensities before and after addition of the analyte. K-ratio measurements on the saraple plus a single standard addition allow the calculation of the correction coefficient and analyte concentration. Several standard addition levels can also be used and concentration obtained from non-linear regression analysis. A computer program has been developed to perform these calculations. Analyses of CuO on silica and of four metals in a metal oxide catalyst over a wide range of concentrations gave results with a relative accuracy generally within ±5%.

Considering all fields of analysis by X-ray fluorescence spectrometry geochemical analysis is probably the most complex. The geochemical analyst is often requested to analyse samples of unknown composition and origin. While these samples need to be analysed for a number of elements of economic or geochemical significance it is often impossible or uneconomic to analyse for all elements that could possibly be present in the sample.

Many analysts use a wet or dry oxidation treatment prior to determination of minor and trace elements in non-fossilised vegetation. This stage is often unnecessary for XRFS since the technique is sufficiently sensitive for many purposes to permit the analysis of unashed material. Indeed, inaccurate results may be obtained if volati1es are lost, if no correction is made for loss on ignition, or if such correction is of large magnitude.

Strontium appears in the earth's crust at a concentration of about 400 ppm and in sea water about 8,1 ppm. Consequently, all living forms have evolved in the presence of this alkaline earth and have incorporated it in their tissues. The exact role of Sr in the human body remains ambiguous (1), even though Sr has been extensively studied with reference to the radiation hazard from 90Sr (2).

Metabolic behavior of Sr, in some aspects, is similar to that of Ca (3), but substantial dissimilarities between these two are also well recognized (1,4). It has been suggested that Sr stimulates the formation of bone matrix. At higher concentrations, however, it interferes with calcification mechanism of the bone matrix (5).

In the last few years, at the Denver X-Ray Analytical Conference, the author and co-workers presented two papers which described the principle and applications of carbon analysis by X-ray fluorescence based upon a monochromatization technique consisting of total reflection and filtering. Instead of the wavelength dispersive method based on Bragg reflection, this monochromatization, combining total reflection by a selected mirror and an appropriate filter, offered an alternative approach for the purpose of increasing measured X-ray intensity. The analytical performance of quantitative determination of carbon content in steel, cast iron and coal were reported.