DE102012219998A1 - Method for identifying crystalline phases in polycrystalline sample used in e.g. airplane, involves determining total quality from quality factors, so as to associate crystal structure with highest total quality to measuring point - Google Patents

Abstract

The method involves comparing normalized vector for chemical composition at each measuring point of sample with normalized vector of each suspected crystal structure, for outputting quality factor for respective correspondence of vectors. The diffraction tape determined at each measuring point is compared with diffraction tape of each suspected crystal structure, for outputting quality factor for correspondence of diffraction tapes. The total quality is determined from two quality factors, so as to associate the crystal structure with highest total quality to measuring point. A list of crystal structures suspected in the sample is provided and data regarding chemical composition and electron diffraction image of the crystal structures is stored. An independent claim is included for computer readable storage medium storing computer program for identifying crystalline phases in polycrystalline sample.

Description

The invention relates to a method for the identification of crystalline phases in a mono- or polycrystalline sample.

Technological background

The identification of crystallographic microstructures is of considerable importance in many fields of technology, for example, in metallic workpieces subjected to high loads, such as in aircraft and automobiles.

As a rule, the chemical composition of the sample to be examined is first determined. A standard method of material analysis used for these purposes is energy dispersive X-ray spectroscopy (EDX). For excitation, an electron beam of uniform energy is directed to the respective measuring point of the sample and the resulting X-ray emission is detected. The recorded characteristic X-ray radiation provides information about the elemental composition of the sample.

Electron backscatter diffraction (EBSD) is a structure-elucidation technique used to identify crystals in a sample. In this method, the diffraction of electrons at the crystal lattice (the so-called diffraction pattern) is evaluated for purposes of phase analysis or crystal structure analysis. A diffraction pattern consists of a series of diffraction bands whose position depends on the crystal structure at the sample location and on the local orientation of the crystal. The evaluation of the diffraction images therefore necessarily requires knowledge of the existing crystal structure. From the knowledge of this crystal structure, it is predicted how the diffraction bands would have to lie in the diffraction image at a given orientation.

The structural data of many thousands of known crystal structures necessary for the prediction of the diffraction bands are compiled in databases and serve as references for the identification of hitherto unknown phases of a measured sample. In practice, based on certain selection criteria, mostly the chemical composition, crystal structures are pre-selected in the database, which are suspected in the sample. By means of suitable search methods, the orientation is then determined on the basis of the diffraction pattern of the unknown sample, in which the prediction best matches the measurement. If you have chosen the correct crystal structure, you get a good match, otherwise the match is bad.

The degree of agreement is primarily indicated by the number of diffraction bands successfully declared within a certain tolerance. Thus, a certain deviation of the theoretical diffraction bands from the actually measured ones is accepted. The degree of this deviation - the angle error - is a secondary measure of agreement. If several crystal structures in the sample are suspected, the evaluation procedure is repeated for all candidates. The candidate with the best match is identified with the sample location.

Summary of the invention

• a) Bereitstellen einer Liste von Kristallstrukturen, die in der Probe vermutet werden, wobei zu jeder Kristallstruktur zumindest seine chemische Zusammensetzung und ein Elektronenbeugungsbild mit einer Vielzahl von Beugungsbändern oder zur Vorhersage der Beugungsbänder nötige Strukturdaten hinterlegt sind;
• b) Für jede vermutete Kristallstruktur, Bestimmung eines normierten Vektors p(i) für die chemische Zusammensetzung, dessen Koordinaten proportional zu den einzelnen Elementhäufigkeiten sind;
• c) An jedem Messpunkt der Probe, (i) Aufnahme eines Spektrums mittels energiedispersiver Röntgenspektroskopie (EDX-Spektrum) und Bestimmung der chemischen Zusammensetzung und (ii) Aufnahme eines Elektronenbeugungsbildes und Bestimmung der Beugungsbänder;
• d) Bestimmung eines normierten Vektors v für die chemische Zusammensetzung am Messpunkt, dessen Koordinaten proportional zu den einzelnen Elementhäufigkeiten sind;
• e) Vergleich des normierten Vektors v für die chemische Zusammensetzung am Messpunkt mit jedem der normierten Vektoren p(i) der vermuteten Kristallstrukturen unter Ausgabe eines Bewertungsfaktors s(i) für die jeweilige Übereinstimmung der Vektoren;
• f) Vergleich der am Messpunkt bestimmten Beugungsbänder mit den Beugungsbändern der vermuteten Kristallstrukturen unter Ausgabe eines Bewertungsfaktors n(i) für die Übereinstimmung der Beugungsbänder; und
• g) Bestimmung einer Gesamtgüte aus den beiden Bewertungsfaktoren s(i) und n(i) und Zuordnung der Kristallstruktur mit der höchsten Gesamtgüte zum Messpunkt.

With the aid of the method according to the invention for the identification of crystalline phases in a mono- or polycrystalline sample, one or more disadvantages of the prior art can be eliminated or at least reduced. The method comprises the following method steps for this purpose:

• a) Provision of a list of crystal structures suspected in the sample, wherein for each crystal structure at least its chemical composition and an electron diffraction pattern with a plurality of diffraction bands or structure data necessary for the prediction of the diffraction bands are deposited;
• b) For any suspected crystal structure, determination of a normalized vector p (i) for the chemical composition whose coordinates are proportional to the individual element abundances;
• c) At each measuring point of the sample, (i) recording a spectrum by means of energy dispersive X-ray spectroscopy (EDX spectrum) and determining the chemical composition and (ii) taking an electron diffraction pattern and determining the diffraction bands;
• d) determining a normalized vector v for the chemical composition at the measuring point whose coordinates are proportional to the individual element frequencies;
• e) comparing the normalized vector v for the chemical composition at the measurement point with each of the normalized vectors p (i) of the suspected crystal structures and outputting a weighting factor s (i) for the respective correspondence of the vectors;
• f) comparison of the diffraction bands determined at the measuring point with the diffraction bands of the presumed crystal structures with the output of an evaluation factor n (i) for the conformity of the diffraction bands; and
• g) Determination of a total quality from the two weighting factors s (i) and n (i) and assignment of the crystal structure with the highest overall quality to the measuring point.

The invention is based on the finding that the identification of unknown crystalline phases in a polycrystalline sample can be considerably accelerated if not whole data sets are calibrated, but normalized vectors are used and determined instead. First, a list of crystal structures suspected in the sample is created. The corresponding datasets for each of these suspected crystal structures contain, in addition to the structural information, information on the chemical composition. The relevant diffraction bands are predicted on the basis of stored structural information. Alternatively or additionally, measured electron diffraction patterns can also be read in.

In step b) of the method according to the invention, a normalized vector p (i) for the chemical composition is then determined whose coordinates are proportional to the individual element frequencies. The proportionality factor should be chosen such that the sum of the squares of the coordinates is one.

In step c) of the method according to the invention, the sample is measured, with an EDX spectrum and an electron diffraction pattern being recorded at each measuring point. The EDX spectrum is used to determine the elemental composition of the sample at the measurement point.

In step d) of the method, a normalized vector v for the chemical composition at the measuring point is now determined whose coordinates are in turn proportional to the individual element frequencies. Again, the proportionality factor should be chosen so that the sum of the squares of the coordinates is one. Subsequently, in step e), only the normalized vectors of the measurement point and of the suspected crystal structures are compared, this comparison leading to the output of an evaluation factor s (i) for the respective correspondence of the vectors. Here, as the weighting factor s (i), it is possible in particular to use the reciprocal of the distance of the normalized vector v of the chemical composition at the measuring point from the respective normalized vector p (i) of the chemical composition of the suspected crystal structures. It is also conceivable to use the reciprocal of the square of the distance to determine the weighting factor. In the cases mentioned, the weighting factor s (i), when well matched, assumes high values.

In step f), a comparison of the measurement point-determined diffraction bands with the diffraction bands of the presumed crystal structures follows, with the output of an evaluation factor n (i) for the conformity of the diffraction bands. The value of n (i) is the number of measured diffraction bands successfully explained by the crystal structure. On the basis of the two weighting factors s (i) and n (i), an overall quality is finally determined in step g). The crystal structure with the highest overall quality is assigned to the measurement point.

Alternatively, first the diffraction patterns (step f)) and then only the chemical information (steps d) and e)) can be evaluated.

The invention further relates to a computer program which enables a data processing device, after it has been loaded into storage means of the data processing device, to carry out a method for identifying crystalline phases in a polycrystalline sample as described above.

The invention also relates to a computer-readable storage medium on which a program is stored which, after having been loaded into storage means of the data processing device, enables a data processing device to perform a process for identifying crystalline phases in a polycrystalline sample as described above.

Claims (6)

A method of identifying crystalline phases in a polycrystalline sample, comprising the steps of: a) providing a list of crystal structures suspected in the sample, wherein each crystal structure has at least its chemical composition and an electron diffraction pattern with a plurality of diffraction bands or for the prediction of the diffraction bands necessary structural data are stored; b) For any suspected crystal structure, determination of a normalized vector p (i) for the chemical composition whose coordinates are proportional to the individual element abundances; c) At each measuring point of the sample, (i) recording a spectrum by means of energy dispersive X-ray spectroscopy (EDX spectrum) and determining the chemical composition and (ii) taking an electron diffraction pattern and determining the diffraction bands; d) determining a normalized vector v for the chemical composition at the measuring point whose coordinates are proportional to the individual element frequencies; e) comparing the normalized vector v for the chemical composition at the measurement point with each of the normalized vectors p (i) of the suspected crystal structures and outputting a weighting factor s (i) for the respective correspondence of the vectors; f) comparison of the diffraction bands determined at the measuring point with the diffraction bands of the presumed crystal structures with the output of an evaluation factor n (i) for the conformity of the diffraction bands; and g) determining an overall quality from the two weighting factors s (i) and n (i) and assigning the crystal structure with the highest overall quality to the measuring point. The method of claim 1, wherein the weighting factor s (i) of step e) is determined as the reciprocal of the distance or square of the normalized vector v of the chemical composition at the measurement point from the respective normalized vector p (i) of the chemical composition of the suspected crystal structures. The method of claim 1, wherein step f) is performed prior to steps d) and e). The method of claim 1, wherein in step b) the normalization is performed such that the sum of the squares of all element frequencies is one. A computer program which enables a data processing device, after having been loaded into storage means of the data processing device, to carry out a method for identifying crystalline phases in a polycrystalline sample according to one of claims 1 to 4. A computer-readable storage medium having stored thereon a program enabling a data processing device, after having been loaded into storage means of the data processing device, to perform a method of identifying crystalline phases in a polycrystalline sample according to any one of claims 1 to 4.