Hello everyone,
Recently I started to do EBSD analysis and I am new to the field. I have couple of questions related to the same.
1) During my analysis process I came to know that even there are many factors which affect CI such as polishing, dislocations etc. I am looking forward to work on EBSD of Cu based SMAs. Suppose the sample is polished properly what could be the other reasons for low CI in SMAs? In my case I am getting CI (>0.1) match is
Hi, Abhishek, I find this information for you, maybe helpful.
section 1.8
"The maximum number of indexed bands versus the number of bands that have been considered for indexing a pattern can be used as a measure of the probability that a correct solution has been found. There are various ways of defining a statistical “confidence index,” “confidence level,” or “likelihood” (cf. Field 1997). The quantity is most useful in phase discrimination. The less perfect the diffracting crystal volume, the more diffuse the corresponding Kikuchi pattern. The blur indicates a high density of point defects or dislocations, lattice strain, thermal lattice vibrations due to the Debye-Waller factor, micro-fragmentation of
the lattice, or the superposition of diffraction patterns from several grains sampled at a time by the primary beam spot. A diffuse pattern may also result
from a foreign surface layer, such as carbon, of excessive thickness (which may have been deposited intentionally to avoid specimen charging), a contamination layer due to poor vacuum conditions, a deformation layer from inadequate sample preparation, or a defocused electron beam or EBSD detector. The blur can be expressed as a quantity, named “pattern quality,” PQ, (or Image Quality, IQ) by measuring the sharpness of some band edges (profile analysis) or the height of peaks in Hough space, or by performing a Fast Fourier Transform (FFT) (Schwarzer and Sukkau 2003) of either the original diffraction pattern or the Radon transformation. Pattern quality maps of coarse grain materials clearly display grain boundaries and surface scratches and often show features that look like a dislocation network. Pattern quality can be used to discriminate between
deformed and recovered or recrystallized grains in a microstructure. "
Here is a presentation from Professor Rollett that may help your understanding of Confidence Index and overall EBSD indexing method by Hough transforms. (Since you referred to CI, I assume you have a TSL-OIM system).
Assuming a properly polished, deformation free sample, then playing around with the number of triplets and bands indexed in your Hough pattern, you can significantly change your CI value since it's just a voting scheme of the MOST likely answer based on the Hough transform. If these values are set high, then if I remember correctly, this can result in a low CI value. Also (but I need to confirm as it's been a while since I've looked at this), it means that high symmetry orientations typically have lower CIs due to a large number of possible solutions.
Ultramicroscopy
Volume 225, June 2021, 113269
Factors affecting Confidence Index in EBSD analysis
https://www.sciencedirect.com/science/article/abs/pii/S0304399121000590
Thanks everyone for your answers especially Huajie for detailed explaination. it is really helpful. Hi Brian this presentation is really good and explained many of my doubts as well. Hi Grzegorz i will go through this chapter, thanks for recommendation.
Hello Abhishek,
Above mentioned explanations are very much correct. You can have more insight about confidence index (CI) in EBSD by going through the paper by S.I.Wright et al. for which link is mentioned below:
Article Introduction and comparison of new EBSD post-processing methodologies
Quality of EBSD data is very much dependent upon the fraction of points indexed properly during automated scan. For Oxford, Bruker and JEOL systems quality of EBSD data is given in terms of zero solutions. If you have low zero solutions, then maximum of your points (let's say 80000/90000/130000) captured are properly indexed. However, there exists many possible causes for the zero solutions.
But EDAX systems, an alternative approach triplet voting method also known as triplet indexing technique is being used in which the quality of EBSD data is described in terms of confidence index (CI). You can get an idea about this method from the above paper.
It is not inappropriate to say the factors which lead to zero solutions in Oxford are more or less similar to those which contribute to low confidence index. This may be attributed to poor sample preparation(polishing), heavy lattice distortions(deformation), Incorrect lattice matching(phases are not same as considered) and high speed data collection.
Hope this helps......
Regards
Subha Sanket Panda
IIT Kanpur
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Waves