Solid state equilibria in iron-nickel alloys containing 0-65 at % nickel were studied by X-ray diffraction techniques, and the gamma/gamma + alpha boundary has been relocated.

Alloys were prepared by thermal decomposition and hydrogen reduction of mixed-crystal iron—nickel formates and by distilling off mercury from mixed ironnickel “amalgams,” both methods of preparation yielding equilibrium alloys in relatively short time at temperatures below 500°C, Alloys were studied at various temperatures by use of a high-temperature diffraction camera, or at room temperature as quenched from the temperature of preparation.

The study showed that the gamma/gamma + alpha boundary lies at generally higher temperatures than in previous diagrams and is concave downwards instead of upwards. The boundary shows three discontinuities which have been located to lie at approximately (a) 3 at % nickel and 760°C, (b) 35–43 at % nickel and S25°C, and (c) 50 at % nickel and 330°C. These discontinuities indicate the possible existence of three isobaric invariant three-phase equilibria, heretofore, undetected in the iron-nickel system. Further studies covering the entire range of iron—nickel alloy compositions are in progress.

Gold wires were vaporized by the exploding-wire phenomenon using a 20 μF capacitor bank charged to voltages up to 14 kV. The resulting condensate, an aerosol or metallic smoke, was collected on membrane filters and subjected to X-ray analysis to determine lattice constant, crystallite size, and behavior with isothermal annealing. Wire explosions were conducted in au ambient atmosphere of ait or nitrogen at barometric pressure. It is estimated that the quench rate for this material is of the order of 108 deg/sec from the melting point although no substrate is involved and it is expected that any effects of epitaxial origin on the structure would be minimized.

Before annealing, diffractograms showed broad peaks apparently shifted to the high-angle side. Line breadth may be attributed primarily to particle size broadening, since ft correlates well with size determined by electron microscopy, (β cos θ) is linear with θ, and [(β cos θ)/λ] is approximately constant for three radiations used. Crystallite size is of the order of 400 Å and is observed to decrease roughly with increasing voltage used for vaporization. The observed lattice decrement, approximately 0.2%, generally increases with voltage used for vaporization, and apparently correlates rather well with the inverse of sise as has been reported in some work on thin gold films. However, studies of colloidal gold particles do not show significant lattice shifts, although the particle size is less than 100 Å so that the decrements observed may be due to factors other than size alone. For this black, particulate material, some lattice decrement apparently persists even after protracted isothermal annealing below the melting point. Crystallite size increases with annealing but remains below about 1000 Å. Results suggest that the lattice decrements observed in condensed gold vapor are due to surface tension effects and the presence of vacancy aggregates.

The asymmetry of diffraction peaks of deformed α-Cu-Ge alloy filings is determined by the center of gravity and Fourier coefficient method with computer calculations. It is shown that it is not possible to describe the asymmetry by a simple value that is a characteristic of the extent of the deformation and the Miller indices of the peak. Instead the asymmetry is highly dependent upon the initial arbitrary conditions for evaluation, which throws some doubt on the reliability of published twin fault and stacking fault probabilities.

Residual stresses were measured in hardened and tempered specimens after unidirectional plastic extension. X-ray and strain gage-layer removal methods were compared. Anomalous residual stresses were found in extended samples at hardnesses of Rc 32–35. The X-ray method indicated compressive residual stresses of nearly constant magnitude through ⅓ the thickness of flat samples, while the strain gagelayer removal method indicated that no macrostress existed. A constant anomalous residual stress was also seen by X-ray through ⅗ the thickness of a cylindrical specimen deformed uniformly in tension. Little or no anomalous stress was found in an extended specimen at Re 55 or in a specimen at Rc 44 after uniform bending.

X-ray diffraction as a means of determining stresses has found increasing application in the last few years. This is primarily because it is the only technique by which stresses can be determined without making measurements on the specimen or structure in the unstressed condition and, consequently, it is the only truly nondestructive technique for determining residual stresses. The principles of determining macrostresses on surfaces with commercially available equipment is quite well known and employs either the X-ray diffractometer or back-reflect ion camera techniques. The diffractometer technique was selected for this investigation because of its accuracy and because it allows both macrostresses and microstra in to be analyzed from the change in position and shape of the diffraction peaks. The X-ray analysis actually consisted of two separate phases. The first dealt with the X-ray determination of the elastic constants (Young's modulus and Poisson 's ratio) for several aluminum alloys, including 5083. These values were compared with the theoretical or published values as determined by standard tensile tests and used later in stress calculations. For these tests, a unique stress stage was used which allowed the specimen to be stressed while positioned in the diffractometer, and also have angular rotation about the diffractometer axis that is independent of the rotation of the counter and receiving slit system. The second phase consisted of analyzing different groups of 5083—aluminum alloy specimens which had been subjected to various degrees of cold working by rolling. This analysis consisted not only of the computation of macrostresses, but also of microstrain and change in particle size as a function of percentage reduction in thickness. The final portion of this phase dealt with electro polishing successive layers from the surface of each sample and relating the measured relaxation to the thickness of the layers removed. In this way, stress distribution in depth was obtainable as a function of cold working.

An X-ray diffraction technique was used to measure residual stresses in Ti-6Al-4V and Ti-8Al-1Mo-1V sheet created by glass-bead peening, sand and aluminum oxide blasting, and a vibratory tumbling treatment. For peening and blasting, the use of larger particle sizes produced greater compressive stresses. In the case of the vibratory treatment, an increase in vibratory frequency or treatment time increased the compressive stress. Glass-bead peening caused a 10% reduction in yield strength while the other treatments had little effect on the tensile properties. Significant stress relaxation occurred in all the treated Ti-6Al-4V coupons exposed at 600° and 800°F.

The relevance of X-ray diffraction techniques to the investigation of the chemical, physical, and crystallographic changes occurring during sliding wear is discussed in relation to existing theories of wear. By-way of illustration, the application of the powder X-ray cylindrical film technique to the unlubricated wear of a low-alloy, medium carbon steel is treated in some detail. The variation of the relative proportions of iron and its oxides in the wear debris with changes in sliding speed is described. It is shown that the results lend partial support to the oxidational hypothesis of the unlubricated wear of steel.

A Warren-Averbach1-4 X-ray line profile analysis was applied to broadened X-ray diffraction peaks from copper deformed in fatigue. The copper specimens were fatigued by four-point bending at peak-strain amplitudes between 0.00105 and 0.00442 in./in., and measurements were made at various fractions of the total fatigue life. The analysis results in an estimation of (a) an average coherently diffracting domain size normal to the diffracting planes and (b) an rms strain distribution function where the strain normal to the diffracting planes is averaged over a given distance at all points in the diffracting crystals and expressed as a function of averaging distance.

Prior to fatigue cycling, the annealed copper exhibited extinction, which reduced the integrated intensity from the low-angle reflections. After fatigue cycling, the integrated intensity increased with increasing strain amplitude of fatigue. The integrated intensities and the rms strains were established during the first few percent of the fatigue life and were found to increase with fatigue strain amplitude. The measured strains were larger in the <100> direction than in the <111> direction, but the absolute values were small. On the basis of transmission electron microscopy of thin foils, these results may be explained by assuming the strains are due to the presence of numerous dislocation dipoles.

In the conventional pole figure, an accurate representation can be attained by correcting the observed X-ray diffraction intensity for any change in diffraction geometry and by comparing this with the correctly established standard intensity. For the intensity corrections, ASTM has prescribed the method of Decker et al. In practice, however, the validity of the correction formula is uncertain, since the prerequisite is difficult to attain for parallelism of an incident beam of sufficient intensity. For the standard intensity, it is desirable to take the intensity obtained with the randomly oriented material of the same composition. However, in most cases an arbitrary unit has been taken because of the difficulty in getting a truly random and uniform sample. Under these circumstances, it is first necessary for an accurate representation of pole figures to make a random sample of uniform thickness for the standard. The authors have obtained satisfactory standard samples by sintering the randomly oriented iron powder made from iron chloride and made use of them to check the method of intensity correction. Satisfactory results are obtained in the randomness tests, such as the comparison of the relative intensity diffracted from the crystal planes parallel to the sample surface and the fluctuation of diffraction intensity during α rotation.

In the reflection case by the Schulz method, the (110) reflection intensity of a random sample, by Co Kα radiation, is independent of the tilting angle up to 50°. The other reflections do not give a constant intensity for the wide range of the tilting angles because of dispersion of the diffracted beam due to the wider separation of Kα doublet in the higher reflection angle and the wider irradiated area at the lower angles. In the transmission case Schulz's correction formula is in good agreement with the observed values for the various diffraction lines and the samples of various µt, while the Decker-Harker formula does not give the absorption change with α-rotation even by an incident beam of ⅙° divergence. In both cases, an accurate determination of pole densities is made by comparing the diffraction intensity of the standard sample substituted in place of the test specimen and by correcting the absorption change due to the difference of µt between the standard and test sample, which affords good coincidence in the overlapped region. The pole figure obtained by the above method furnishes an accurate prediction of plastic and elastic anisotropy in sheet metals.

The back-reflect ion double-scanning diffractometer method, by which lattice parameters can be measured with a reproducibility of one part in 150,000 has been applied at liquid helium temperatures. A cryostat attachment is described which enables diffraction profiles to be scanned on both sides of the primary X-ray beam up to 163°, 2θ. Alignment errors may, thus, be eliminated by measuring the included angle 4θ between respective Bragg reflections. The method is illustrated by measuring the lattice parameters of the I.U.Cr. standard specimens of silicon and tungsten at various cryogenic temperatures.

Recent development of numerical control, N/C, systems for digital X-ray powder diffractometry opens the full potential for rapid and detailed machine processing of digital diffraction data. This capability may prove to be as large an incremental technological advance as the earlier shift from film to counter detector techniques. Our N/C system has the capability of performing essentially all operations possible by manual methods. Data output is on magnetic tape and carries identification, and some of the instructions necessary for processing. The data tapes are processed by programs that (a) edit and check, (b) filter through a controlled shape high frequency filter (square, triangular, normal, etc.), (c) remove the minimum background from the diffraction pattern, (d) analyze data points close to background and test for statistical significance deleting those points within limits defined to constitute noise, (e) recognize peaks and record their position and intensity, (f) identify phases, (g) calculate phase concentrations by various methods, including internal standard and mutual standard techniques, and (h) output any desired portion of the data in a variety of digital and analog formats suitable for display, further analysis, and storage in an information retrieval system.

It is possible to record an entire diffraction pattern for a phase on two or three inches of magnetic tape. This record carries detailed peak area and shape information now missing from the ASTM reference system. Furthermore, the new magnetic tape information is directly amenable to computer processing to prepare search oriented record systems of diffraction data that can be inexpensively updated and edited. Currently, an ASTM committee is preparing a magnetic tape format diffraction pattern file. It is suggested that a second generation, computer index system be generated based on digital records of actual diffraction patterns. This type of system should prove more versatile than the old system and possibly be the only way to keep abreast of the flood of new diffraction pattern information generated from the vast number of new organic and inorganic compounds being synthesized.

The aluminum K X-ray emission lines and bands from a series of aluminum binary alloys and other binary compounds were investigated using 6 kV electron excitation and a flat crystal vacuum spectrometer. The overall shape of the emission band and its energy position as a function of alloy composition was determined. It appears from the data that the aluminum K band undergoes changes in shape and energy position which are dependent on the electronic configuration of the element with which the aluminum is chemically bonded. These band changes can be interpreted as indicating a change in the bonding character between the metal atoms. In the Al-Ni system, for instance, the Al K band becomes more symmetrical and shifts to lower energy as the nickel content is increased, indicating that perhaps the bonding on the aluminum atoms is becoming less metallic and more covalent in nature as the nickel to aluminum ratio is increased.

The aluminum Kα4/Kα3 satellite line intensity ratio also varies in an orderly manner in aluminum binary compounds. In general, these satellite line changes go hand-in-hand with the K band changes. If the K band shifts to lower energy, the Kα4/Kα3 intensity ratio will always increase in value.

Multilayer soap film structures, particularly the lead stearate variety have been used for several years as a dispersing element in soft X-ray spectrometers. These structures have a high scattering power, and if a high order of diffraction is used for the shorter wavelengths they provide good resolution throughout the 10-80 Å range. Structures having a 2d spacing smaller than that of lead stearate (100 Å) would provide greater dispersion and, hence, resolution in the first order for radiation in the 10-40 Å range. Details concerning the conditions required to build multilayer structures from the soaps of shorter fatty acids such as lead myristate, lead laurate, lead caprate, etc. are given. The various members of the soap film family are compared regarding their diffracting power both as a function of wavelength and the order of diffraction. Information is given regarding the dependence of the diffracting power, the width of the diffraction peak at half maximum and the peak to background ratio as a function of the number of double layers in a structure. The absorption occurring within a lead stearate and a lead laurate structure has been experimentally measured. Observed spectra can, thus, be corrected for the filtration caused by the soap film structure. The soap film family is evaluated as a dispersing element by comparing the various Structures with single crystals such as EDDT and KAP.

Production efficiencies for the excitation of characteristic soft X-ray emission spectra by proton bombardment were investigated. Results were obtained for K-, L-, and M-shell lines of various elements as a function of proton energies ranging up to 100 keV.

For comparison putposes, helium, nitrogen, and argon ions also were accelerated and used as bombarding particles to yield efficiency data for some of the same spectral lines.

The data demonstrate that:

• 1. The production efficiencies of spectral lines excited by protons are, in the first approximation, dependent on wavelength only and are independent of the atomic shell with which they are associated.

• 2. Production efficiencies increase with the atomic weight of the bombarding ions for some of the spectral lines investigated.

X-ray generation by proton bombardment is a powerful method for producing high-intensity monochromatic radiation for wavelengths above approximately 10 Å, The same technique can be used for analytical purposes to determine the concentration of low atomic number elements.

Although the phase SrSi2 has been previously recognized, it has been described only briefly in the literature.

An alloy comprised of FeSi2, silicon, and the SrSi2 phase was examined on the electron microprobe. Optical and electron scanning images are correlated with X-ray scanning images for iron, silicon, calcium, and strontium. Quantitative microprobe analyses were made and corrected for absorption and the atomic number effect by several noncomputer methods. The corrections include as variables the X-ray intensities measured at several accelerating voltages. The effects of varying the electron beam size and accelerating potential are included. The results are compared to chemical analyses. The advantages of varying the accelerating voltage when correcting intensity data, increasing the beam size when surface preparation is a factor, and the importance of surface preparation at low accelerating voltages are discussed.

An X-ray diffraction examination showed that the phase SrSi2 has a cubic unit cell with an a0 of 6.515 Å. There are four molecules per unit cell, and the most probable space group appears to be P213. The density was calculated as 3.45 (observed density >3.3).

Metallographic observations with ordinary and polarized light and microhardness measurements are included.

Transient (nuclear) heating experiments were conducted with uranium carbide fuel rods to study the failure characteristics of typical rod designs. Extremely unusual changes in microstructure were observed and the electron microprobe was employed to establish the disposition of materials resulting from the meltdown experiments. The probe results indicated substantial fuel-clad interactions and permitted the resolution of several uncertainties regarding the course of fuel rod failure and material redistribution. The electron microprobe represents a unique capability in the post-test analysis of such meltdown tests.

Standards for microprobe analysis can be made to serve two purposes: (a) proposed correction models can be tested with them, and (b) analysis can be performed more accurately in the system which includes the standard. Few microprobe standards presently are available because they must be homogeneous on the micron scale and their composition must be known accurately. A modified Duwez splat cooling method is described which enables the investigator to prepare suitable standards in most cases. The apparatus which is relatively simple and inexpensive is described in detail. The systems Au-Si and Al-Mg were chosen as test cases. Suitable standards were prepared at different concentrations in each system. The analytical results for all compositions in Al-Mg are presented and discussed.

In the microprobe analyzer, a portion of the high energy electrons impinging on the surface are backscattered from the sample and re-emitted at high energy levels. Low energy (less than 50 eV) or secondary electrons also ate emitted. Both the electron backscatter yield and the secondary electron yield are related to the mean atomic number of the target material and, hence, may be used to provide information about the target composition. Unfortunately, however, the secondary electron yield is very sensitive to the surface condition of the specimen and various instrument parameters. This complicates the otherwise simple linear relationship between sample composition and electron backscatter yield.

It is shown that the effects due to secondary electrons can be minimized by biasing the sample, and that good results can be obtained in the analysis of binary systems. The limitations and utility of the method are discussed, and backscatter yields are determined.

Four types of slit apertures—vertical, horizontal, inclined, and sirtgle-edge—for selected-area analysis with commercial X-ray secondary-emission (fluorescence) spectrometers are described, evaluated, and compared. The vertical and single-edge slits already have been described by the writer and Rizzo. The other two are described here for the first time. The vertical and horizontal slits are secondarybeam apertures; the other two lie in both the primary and secondary beams. AH the slits are mounted on the specimen drawer. No further modification of the spectrometer is required except replacement of the soller collimators with open tunnels or simple slits, and replacement of the flat crystal with a fixed-radius curved crystal to increase sensitivity. The accessories are inexpensive and can be made in even a modest machine shop. They can be installed on or removed from the spectrometer in about 5 min. The accessories are applicable ordy to linear selected areas where composition varies in the direction normal to but not along the line, for example, linear inclusions and sections of plated surfaces, diffusion couples, and interfaces. The four slits were evaluated and compared for resolution, sensitivity, and spectral-line width. Techniques for use, advantages, limitations, and means for improvement are discussed for each of the slits. The horizontal slit is of little value, but each of the others has features which permit analysis of very narrow selected areas for which pinhole apertures would not be sufficiently sensitive. Slit widths as narrow as 0,00012 in. (0.003 mm) have been used in favorable cases. The inclined slit, used with a pulse-height analyzer, is probably the most useful of the four apertures, combining high sensitivity, high resolution, and narrow spectral-line width.

Ferrotitanium ores and residues may vary widely in composition. Because of this, interelemerxt effects are very large. An X-ray spectrographic method will be described for the analysis of the light elements in these ores and residues. The method uses a standard helium path Norelco spectrograph with a tungsten target tube. A KAP analyzing crystal is used for all the elements investigated. Appropriate emission lines of the elements magnesium, aluminum, silicon, calcium, titanium, and iron are measured and the results are calculated on a small computer using the empirical corrections for interelernent effects of Lucas-Tooth and Pyne. The FORTRAN program for these calculations will be given.

The problems of sample preparation, standardization, interferences and instrumental variables will be discussed. A comparison of X-ray results with chemical analysis is made.

The method is especially adaptable to automated analysis and some of the systems needed will be outlined. The method may also be modified to analyze light elements in other matrices.