Andrew Deal

Characterizing Ultrafine Grained Material using EBSD

Ultrafine grained (UFG) alloys, those with grain sizes significantly below 1 micron, are of general interest from both a property and processing standpoint. Improved mechanical properties have been reported, and many materials are under investigation [1,2]. The nanoscale nature of these materials, however, provides a significant number of characterization challenges. Here we discuss the use of Electron Backscatter Diffraction (EBSD) to characterize the grain size of ultrafine grained Ti-6Al-4V (UFG Ti64). UFG Ti64 was produced by near-isothermal Multi-axis Forging (MAF) [3]. Samples of this material were heat treated to study the static coarsening of the alpha phase, and hot compression tests were performed on UFG Ti64specimens to examine dynamic coarsening behavior. After heat treatment or compression, samples were metallographically prepared in conductive 1.25 inch mounts according to a modified version of a published technique [4]. To understand the coarsening kinetics, it was essential to quantify the average alpha grain size of the material. A high spatial resolution and ability to examine large areas for statistical purposes were critical requirements for this task, making a FEG-SEM with EBSD the instrumentation of choice. Since large EBSD measurements are on the order of hours, even with modern camera speeds, maintaining mechanical and thermal stability of the SEM was critical. Accordingly, the FEGSEM used for UFG grain size measurements was enclosed to isolate it from thermal oscillations inherent to HVAC systems. A chiller with precise PID temperature control was installed to keep the lens cooling stable. These modifications helped keep the short-term temperature fluctuations below 0.5C. To minimize long term mechanical or thermal drift, the sample was positioned for analysis under appropriate beam conditions and then rested for a minimum of 3 hours prior to the EBSD measurements. Long term temperature drift was typically less than 1C. Figure 1 shows the band contrast for a small region of tested UFG Ti64 material. Small maps such as these were used to examine the resolution of the EBSD measurements and get a sense of the grain size. Reasonable grain size statistics, however, were accumulated through EBSD line scans along the tilt axis. For each measurement, the step size within each line was 20nm, and lines were spaced between 1 and 50 microns apart. Scan times were on the order of 14 hours. Initial results of the static coarsening experiments are shown in Figure 2, compared with conventional Ti64 material analyzed optically. The EBSD measurements show that the alpha grain size of the UFG Ti64 is reasonably stable with respect to static coarsening at 650C. It remained submicron after an hour and remained below 3 microns after 77 hours.

3D Investigation of Cracking Behavior in a Ni Superalloy

The high temperature fatigue performance of Ni-base superalloys is critical to gas turbine applications and as such, requires a more fundamental understanding when designing and producing turbine components. To investigate the relationship to local microstructure, a fatigue specimen was cycled under conditions designed specifically to result in intergranular propagation. Prior to failure, the test was interrupted and a 3D data set was reconstructed destructively from optical and EBSD slices taken from around the tip of the growing crack. The data set was investigated to understand the character of grain boundary planes along the crack front with respect those of the bulk material.

Handling Misalignment and Drift in 3D EBSD Data Sets

GE Global Research is exploring the 3D reconstruction of high temperature materials to understand material behavior in lifing applications. Large volumes of material are required to generate the appropriate statistical understanding of grain morphologies, grain boundary types and distributions, and residual plastic strain. Consequently, GE has adopted mechanical sectioning coupled with EBSD analysis as the methodology for investigating regions of interest. Among the major factors that affect the accuracy of such a reconstruction are thermal or mechanical drift during EBSD measurements and the precision and accuracy of sample alignment. To correct for these unavoidable issues when reconstructing the final volume, an algorithm based on grain center-of-mass and various shape factors has been developed and applied to a data set. Results show a significant improvement in slice-to-slice registration.