Grain Size Measurement from EBSD Data of IN718 Alloy Fabricated by Powder Bed Laser Fusion Process

[SunnySharma200]

I have EBSD data from Inconel 718 samples processed by the Laser deposition process. Please tell me how to calculate the average grain size.

Below is the snippet of the code I am using for the calculation of grain size.

%% Grain Size Analysis - Simple Histogram
% Compute grains and their sizes
[grains, ebsd.grainId] = calcGrains(ebsd, 'angle', 1*degree);
% Ensure grainSize is numeric
grainSize = grains.area.^(1/2); % Grain size in microns
if ~isnumeric(grainSize)
    error('grainSize must be numeric.');
end

[lfgmorales]

hi you may want to read this https://mtex-toolbox.github.io/ShapeParameters.html also, your angle threshold for the definition of individual grains is rather small, so you may want to consider something a little larger cheers Luiz

…

On 8 Jan 2025, at 15:36, SunnySharma200 ***@***.***> wrote: I have EBSD data from Inconel 718 samples processed by the Laser deposition process. Please tell me how to calculate the average grain size. Below is the snippet of the code I am using for the calculation of grain size. %% Grain Size Analysis - Simple Histogram % Compute grains and their sizes [grains, ebsd.grainId] = calcGrains(ebsd, 'angle', 1*degree); % Ensure grainSize is numeric grainSize = grains.area.^(1/2); % Grain size in microns if ~isnumeric(grainSize) error('grainSize must be numeric.'); end — Reply to this email directly, view it on GitHub <#2289>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ALE5IN2FCG2SMT7D7JUWVPD2JUZWVAVCNFSM6AAAAABU2CFKAKVHI2DSMVQWIX3LMV43ERDJONRXK43TNFXW4OZXG44TOOJUGU>. You are receiving this because you are subscribed to this thread.

[SunnySharma200]

Hello Luiz,

Thank you for sharing the URL.
I understand that equivalentRadius is suitable for primarily equiaxed grains. However, in my case, most of the grains are elongated (columnar).

What would be the best way to measure such grain morphologies?

[kilir]

Hi,

grain size is a thing.

The concept of the equivalent radius is actually that it compensates for all sorts of shape differences since it only requires the sectional grain area; so strictly it is not only suitable for equiaxed grains. But you can take any other measure e.g. the longes axis, the perpendicular to the shortest, some inverse of the grain boundary density...

More importantly, what single number do you mean with "the average grain size"? The arithmetic mean of the frequency, the area or volume weighted mean, median, the mode, the RMS or soomethign you can compare to length weighted line intercepts? Comparing samples with different grain property distributions, some/many of these single numbers are mostly (somewhat) meaningless or at least not very telling about the distribution. There is the https://mtex-toolbox.github.io/grain2d.histogram.html function will provide by default at least an area weighted histogram (which is usually much more telling than looking at frequency histograms). You can alos weight by any other property. Also, but in this case rather applicable to rather low aspect ratio grains theres https://mtex-toolbox.github.io/stripStar.html for a proper stereological correction providing you with volume weighted measures.

Cheers,
Rüdiger