EBSD statistically significant data



Hi all:


I'm working on EBSD scan data collection and bumping against
a road block about whether the data is statistically
significant to be used for data analysis. This basic data collection
method and analysis will be used for several dozens of specimens.
They are all IC metal interconnects.

In my current situation, I am analyzing Al thin films
for grains characteristics. The basic values of interest
are total grains, edge grains (those not used in analysis) and
average grain size. So, how would one converge on a
reasonable range of total grains and edge grains given the
largest grain size and a good confidence index?

As an example, I call your attention to a recent article in
Microscopy Today. "Using EBSD to Map
Domain Structures in Ferroelectrics," September 2008, pg 18-19.

This article has maps with huge data points and poorly defined
edges. Is this due to poor data collection (scan summary data was
not provided) or is it due to analysis in poor vacuum systems?

My work is all high vacuum so I see these reported results
as failures if I did them. But is this typical of ESEM results?
And regardless of the vacuum, what constitutes statistically
significant data?

Any thoughts on this?


gary g.



Kiss French. Drink California.

gary at gaugler dot com
.



Relevant Pages

  • Re: EBSD statistically significant data
    ... I'm working on EBSD scan data collection and bumping against ... for grains characteristics. ... There is an ASTM standard for measuring grain size, ...
    (sci.techniques.microscopy)
  • Re: EBSD statistically significant data
    ... I'm working on EBSD scan data collection and bumping against ... for grains characteristics. ... There is an ASTM standard for measuring grain size, ...
    (sci.techniques.microscopy)
  • Re: EBSD statistically significant data
    ... for grains characteristics. ...  Is this due to poor data collection (scan summary data was ... distribution and misorientation compared to select an optimum if not ... for each EBSD data point, then use inverse-FFT to locate the most ...
    (sci.techniques.microscopy)