Combining multiple applications on one instrument to optimize sample characterization
Steel and metal alloys as well as coatings that are deposited on them, are subject to extensive monitoring and testing during and after the production process, due to their various applications in demanding environments. The knowledge of material properties such as phase composition, crystallite size and microstrain, residual stress, as well as texture is essential for the later usage of metal parts. These properties have a direct influence on the metals' capacity to resist loads and other mechanical and physical forces. Hence a multipurpose diffraction platform that can measure all before-mentioned properties of a bulk or coated steel samples is desired in order to determine the product quality as well as to predict potential failure.
Complete solution for steel analysis
The Empyrean equipped with a Co-tube, Eulerian cradle and a PIXcel MediPIX3 detector provides an optimized solution for fast retained-austenite analysis, classical stress and texture measurements (Figure 1).
Figure 1: Overview over the relevant XRD applications for steel analysis
Full flexibility with a 2nd diffracted beam path
A second detector arm equipped with a parallel-plate collimator and a Xe proportional counter allows the automatic switch to parallel-beam geometry for grazing incidence measurements, texture measurements without defocusing and phase analysis and residual stress measurements on irregular or curved surfaces.
Choice of wavelength
Different radiations are commonly used for steel analysis depending on which type of steel is analyzed. Traditionally for residual stress analysis the choice of wavelength is mainly dependent on the 2θ-positions of the used reflections (high 2θ angles desired). Therefore, mainly Cr (λKα = 2.2898 Å) or Mn radiation (λKα = 2.1019 Å) are used, Table 1. For retained austenite analysis, however, neither radiation is well suited because not enough reflections of martensite and austenite can be measured in the angular range that can be covered by an XRD system. Therefore Mo radiation (λKα = 0.7107 Å) is commonly used due to its much shorter wavelength.
In this study Co radiation (λKα = 1.7809 Å) was used as it provides an optimal compromise for XRD steel analysis. First, it allows to access sufficient reflections of both martensite/ferrite and austenite at a significantly higher intensity yield for a fast and reliable retained-austenite analysis but also size & strain as well as texture analysis. Secondly, it results in reflections at sufficiently high 2θ angles to perform a residual stress analysis on both martensite and austenite.
Table 1. Overview of the effect of the used wavelength on the peak positions for both cubic ferrite/martensite as well as austenite
The pre-aligned fast interchangeable X-ray modules for optics and sample stage allow quick and easy switching between applications.
Sample alignment tools
A state-of-the-art alignment camera and a laser alignment tool in combination with motorized positioning stages provide the optimal setup for precise alignment of small sample spots as well as the mapping of larger areas.
Possible system configurations
Several combinations of incident-beam optics, sample stages and radiations are possible depending on the required spot sizes (1 cm to 50 µm), sample shapes and sizes as well as the material of interest and the desired type of metal analysis. On the first diffracted beam path the use of the PIXcel detector allows fast Bragg-Brentano scans for phase analysis (retained austenite) and size and strain analysis as well as fast stress measurements. The 2nd beam path with the parallel plate-collimator provides a true parallel-beam geometry enabling grazing incidence measurements on coatings as well as stress and texture measurements without defocusing and misalignment errors. Analysis of irregularly shaped samples without being sensitive to variations in the sample height are also possible.
Table 2. Summary of the possible combinations of optical modules for steel analysis making use of the unique 2-detector approach.
All analyses in this study were performed with one configuration (indicated in bold orange letters), with no need for user interaction between the measurements.
Two hardened steel samples originating from different steps in the hardening process were analyzed using the above- mentioned solution for steel analysis. Sample-1 was taken right after the quenching step, Sample-2 additionally underwent an annealing step.
Retained austenite, crystallite size & microstrain, residual stress and texture measurements on both samples were performed to identify differences in sample properties resulting from the different stages in the process.
No sample preparation was required prior to the measurements. The sample positioning and height adjustment were done fully automatically using the motorized chi-phi-X-Y-Z stage in combination with the camera as well as the laser alignment tool.
Figure 2. Temperature profile during the hardening process of steel and images of the two steel samples analyzed in this study. Sample snapshots were taken with the Empyrean alignment camera.
Figure 3. Comparison of the scans of the 2 steel samples performed with Co radiation that were used for subsequent phase and size & strain analysis. The Sample-2 dat is shifted vertically for better visibility. They show significant differences between the two samples and most significantly a difference in the austenite content.
Retained austenite analysis
Retained austenite is an instable remainder of the process of hardening of steel. It slowly transforms to the stable martensite phase, which is accompanied by a change in volume. The presence of austenite can significantly compromise the stability of hardened steel components and represents a risk for failure during service life of high- precision steel components. As a consequence the amount of retained austenite must be as low as possible (typically <2%).
Figure 4. Structural differences between austenite and martensite
The Rietveld full pattern based analysis routine used in HighScore Plus 4.1 and RoboRiet 4.1 is in compliance with the ASTM E975 norm for hardened steel and can be fully automated. In addition to the volume fractions of austenite and ferrite/martensite (and if present cementite) texture parameters are also calculated.
The results show a significant decrease of the retained austenite content from 10 vol % in Sample-1 to 0.7 vol % in Sample-2 due to the subsequent annealing step (Figure 5). In addition, a high texture for austenite is indicated that exceeds the acceptance defined in the ASTM E975 norm.
Figure 5. Rietveld refinement plot of Sample-1 (including difference plot below) and the combined result outputs generated by the automatic retained austenite analysis in HighScore Plus 4.1 and RoboRiet 4.1
Size and strain analysis
Crystallite size and microstrain are important parameters that can influence microstructural properties such as stability and elasticity in polycrystalline materials. From an XRD pattern both parameters can be calculated from the 2θ-dependent peak broadening. The size & strain analysis can be performed together with the retained austenite analysis on the same measurements.
Figure 6. Effect of the crystallite size and the microstrain on the peak width
Figure 7. Bragg-Brentano scans showing the crystallite size & microstrain-induced peak broadening of the (222) reflections in the 2 steel samples compared to a W-powder standard that was used to define the instrument-related peak broadening
Table 3. Calculated K-factors for both samples
Figure 8. Williamson-Hall plot showing the annealing-induced changes in crystallite size and microstrain
The crystallite size and microstrain were determined using the peak information obtained from the same scans and Rietveld fits used for the retained austenite analysis. The derived peak widths are then plotted in Williamson-Hall plots (Figure 8) which show a 2θ-dependent structural broadening trend (total peak broadening minus the instrumental broadening with sinθ). From the linear fits of the extracted peak data both the microstrain and the crystallite size of the respective phases can be calculated.
The resulting data shows that the annealing of the quenched steel is mainly accompanied by a significant reduction of the structurally stored microstrain from 0.13% to 0.08% in combination with a slight increase in crystallite size (662 to 705 Å).
Texture analysis of steel is vital for understanding the distribution of crystal-specific physical properties (e.g. mechanical wear, hardness, elasticity, conductivity, thermal expansion) within a polycrystalline material.
The ODFs calculated with the Texture software package as well as the extracted fiber blocks show a similar result confirming that sample-2 has a slightly enhanced texture compared to sample-1.
Table 4. Calculated K-factors for both samples
The pole figures of both samples show the presence of a strong rolling texture with Sample-2 exhibiting slightly stronger texture features. The stronger texture also results in larger deviations of the respective calculated K-factors from the value 1 (K-factor of 1 equals texture-free).
Figure 9. Recorded pole figures of the ferrite (110), (200), and (211) reflections of Sample-1 and Sample-2
Orientation distribution function – Sample 2
Figure 10. 2D and 3D view of the ODF calculated for Sample-2 as well as the derived α- and γ-fiber blocks from both samples
The ODFs calculated with the Texture software package as well as the extracted fiber blocks show a similar result confirming that Sample-2 has a slightly enhanced texture compared to Sample-1. Fiber blocks represent specific directions in Eulerian space that are significant for the respective materials, in this case body-centered cubic ferrite/ martensite. Both the α- and γ-fiber of Sample-2 show stronger features indicating that the annealing step results in an enhancement of the already existing rolling texture.
Residual stress analysis
Residual stress in a material is the result of either externally applied forces or residual internal strain resulting from certain production processes.
Knowledge about the stress state of a material is crucial for the evaluation of the stability of materials under mechanical and physical forces. Stress cannot be directly measured – it is calculated via the strain ε that can be measured with XRD.
Figure 11. Full tensor residual stress measurement for Sample-1
Figure 12. Example of a surface plot and peak fit in Stress Plus SW
The isoline surface plot on the right shows the evolution of the ferrite (220) reflection with increasing sample tilt at different φ-rotations used for the determination of the residual stress.
The overall peak shift is minor, however a large change in intensity is observed, which indicates that the sample is strongly textured.
Figure 13. Sin2 (psi) plots and the calculated residual stresses along the rolling direction for both samples as obtained in Stress Plus
Residual stress - Sample-1
The respective residual stress analyses performed with the Stress (Plus) 2.2 software package show low unidirectional, compressive residual stresses along the rolling direction in both samples. The residual stress level in Sample-1 (before annealing) is -54 MPa along the rolling direction and zero perpendicular to it.
Residual stress - Sample-2
Sample-2 only shows a unidirectional residual stress of -34 MPa along the rolling direction. This result indicates a relaxation of the strain induced by the initial production process during annealing.
A complete overview over the collected results shows significant differences between the two steel samples. The annealing process seems to trigger a recrystallization of the the polycrystalline steel matrix that results in:
- Transformation of austenite into ferrite/martensite
- Crystal growth and release of microstrain
- Preferred orientation if the transformed ferrite/martensite along the existing rolling texture
- Relaxation and release of residual stress
This study demonstrates the Empyrean multipurpose diffraction platform as a fast and complete analytical solution for the analysis of steel and metal alloys. The flexibility of the measurement setup enables analysis of all different types of material (bulk samples, coated steels) with different sizes, shapes and analytical requirements (e.g. small spot analysis). Phase composition, crystallite size & microstrain, texture and residual stress can be determined on the same platform without the requirement of changing optical components or special sample preparation. In combination with dedicated software packages most of these analyses can even be automated.