grain average GND map and Grain average GND density plotted against grain area

[LuqingCui]

Hi all,

Does anyone know how to calculate the grain average GND map? I have obtained the GND map on each pixel for FCC metal based on the code on MTEX website: https://mtex-toolbox.github.io/GND.html.
But how to get the grain average GND map (Fig. 3h) and ### Grain average GND density plotted against grain area (Fig. 3i) as follows:

Thank you very much in advance!
Best regards,
Luqing Cui

[zmichels]

Hi Luqing, If you have a variable for GND at every EBSD pixel, and you have computed grains, you can use accumarray() to compute a mean GND value for each grain: grainGND = accumarray(ebsd.grainId, GND, size(grains), @nanmean); Does that work for you? Cheers, Zach

…

On Dec 19, 2020, at 9:24 AM, LuqingCui ***@***.***> wrote: Hi all, Does anyone know how to calculate the grain average GND map? I have obtained the GND map on each pixel for FCC metal based on the code on MTEX website: https://mtex-toolbox.github.io/GND.html <https://mtex-toolbox.github.io/GND.html>. But how to get the grain average GND map (Fig. 3h) and ### Grain average GND density plotted against grain area (Fig. 3i) as follows: <https://user-images.githubusercontent.com/72967420/102692807-1bea8480-4216-11eb-8964-98a9f64dded7.png> Thank you very much in advance! Best regards, Luqing Cui — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#639>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACC5AU7AFN54NJQ43UFZC33SVTAUTANCNFSM4VCLT43Q>.

[LuqingCui]

Hi Cheers,

Thanks for your reply. I type the code as you suggested, and when the code run to the line of "grainGND = accumarray(ebsd.grainId, gnd_tot, size(grains), @nanmean);"
The error comesout as follows:

The code I used is as follows:

import_wizard
ebsd=ebsd('fe')
% define the color key
ipfKey = ipfHSVKey(ebsd);
ipfKey.inversePoleFigureDirection = yvector;
% and plot the orientation data
plot(ebsd,ipfKey.orientation2color(ebsd.orientations),'micronBar','off','figSize','medium')
% reconstruct grains
[grains,ebsd.grainId] = calcGrains(ebsd,'angle',15*degree);

% remove small grains
ebsd(grains(grains.diameter<=5)) = [];

% redo grain reconstruction
[grains,ebsd.grainId] = calcGrains(ebsd,'angle',15*degree);
% smooth grain boundaries
grains = smooth(grains,5);
hold on
plot(grains.boundary,'linewidth',2)
hold off
% a key the colorizes according to misorientation angle and axis
ipfKey = axisAngleColorKey(ebsd);

% set the grain mean orientations as reference orinetations
ipfKey.oriRef = grains(ebsd('indexed').grainId).meanOrientation;

% plot the data
plot(ebsd,ipfKey.orientation2color(ebsd('indexed').orientations),'micronBar','off','figSize','medium')

hold on
plot(grains.boundary,'linewidth',1.5)
hold off
% denoise orientation data
F = halfQuadraticFilter;

ebsd = smooth(ebsd('indexed'),F,'fill',grains);

% plot the denoised data
ipfKey.oriRef = grains(ebsd('indexed').grainId).meanOrientation;
plot(ebsd('indexed'),ipfKey.orientation2color(ebsd('indexed').orientations),'micronBar','off','figSize','medium')

hold on
plot(grains.boundary,'linewidth',1.5)
hold off

% consider only the Fe(alpha) phase
ebsd = ebsd('indexed').gridify;

% compute the curvature tensor
kappa = ebsd.curvature

% one can index the curvature tensors in the same way as the EBSD data.
% E.g. the curvature in pixel (2,3) is
kappa(2,3)

kappa12 = kappa{1,2};

size(kappa12)

newMtexFigure('nrows',3,'ncols',3);

% cycle through all components of the tensor
for i = 1:3
for j = 1:3

nextAxis(i,j)
plot(ebsd,kappa{i,j},'micronBar','off')
hold on; plot(grains.boundary,'linewidth',2); hold off

end
end

% unify the color rage - you may also use setColoRange equal
setColorRange([-0.005,0.005])
drawNow(gcm,'figSize','large')

alpha = kappa.dislocationDensity
alpha(2,3)
dS = dislocationSystem.fcc(ebsd.CS)
% size of the unit cell
a = norm(ebsd.CS.aAxis);

% in bcc and fcc the norm of the burgers vector is sqrt(3)/2 * a
[norm(dS(1).b), norm(dS(end).b), sqrt(3)/2 * a]
nu = 0.33;
% energy of the edge dislocations
dS(dS.isEdge).u = 7.01317231;

% energy of the screw dislocations
dS(dS.isScrew).u = 4.698825451;
dS(1).tensor
dSRot = ebsd.orientations * dS
[rho,factor] = fitDislocationSystems(kappa,dSRot);

% the restored dislocation density tensors
alpha = sum(dSRot.tensor .* rho,2);

% we have to set the unit manualy since it is not stored in rho
alpha.opt.unit = '1/um';

% the restored dislocation density tensor for pixel 2
alpha(2)

% the dislocation density dervied from the curvature in pixel 2
kappa(2).dislocationDensity

% we may also restore the complete curvature tensor with
kappa = alpha.curvature

% and plot it as we did before
newMtexFigure('nrows',3,'ncols',3);

% cycle through all components of the tensor
for i = 1:3
for j = 1:3

nextAxis(i,j)
plot(ebsd,kappa{i,j},'micronBar','off')
hold on; plot(grains.boundary,'linewidth',1); hold off

end
end

setColorRange([-0.005,0.005])
drawNow(gcm,'figSize','large');

factor

close all
plot(ebsd,factorsum(abs(rho . dSRot.u),2),'micronbar','off')
mtexColorMap('jet')
mtexColorbar

set(gca,'ColorScale','log'); % this works only starting with Matlab 2018a
set(gca,'CLim',[1e12 1e16]);

hold on
plot(grains.boundary,'linewidth',1)
hold off

% total
gnd_tot = factorsum(abs(rho . dSRot.u),2);
grainGND = accumarray(ebsd.grainId, gnd_tot, size(grains), @nanmean);
%total grain average GND map
plot(ebsd,grainGND,'micronbar','off')
mtexColorMap('jet')
mtexColorbar

set(gca,'ColorScale','log'); % this works only starting with Matlab 2018a
set(gca,'CLim',[5e12 1e16]);

hold on
plot(grains.boundary,'linewidth',1)
hold off

Is there anything wrong with the using of accumarray?

[zmichels]

Yes. You need to have existing grains for each existing ebsd.grainId. But after you get rid of some grains, you then have a different number of grain ids than those stores in ebsd.grainId. *After* removing grains and recomputing (“redo”) grains, try adding a line of: ebsd = ebsd(grains) Then proceed with the rest of your code and see if that works. Cheers, Zach

…

On Dec 20, 2020, at 9:11 AM, LuqingCui ***@***.***> wrote: Hi Cheers, Thanks for your reply. I type the code as you suggested, and when the code run to the line of "grainGND = accumarray(ebsd.grainId, gnd_tot, size(grains), @nanmean);" The error comesout as follows: The code I used is as follows: import_wizard ebsd=ebsd('fe') % define the color key ipfKey = ipfHSVKey(ebsd); ipfKey.inversePoleFigureDirection = yvector; % and plot the orientation data plot(ebsd,ipfKey.orientation2color(ebsd.orientations),'micronBar','off','figSize','medium') % reconstruct grains [grains,ebsd.grainId] = calcGrains(ebsd,'angle',15*degree); % remove small grains ebsd(grains(grains.diameter<=5)) = []; % redo grain reconstruction [grains,ebsd.grainId] = calcGrains(ebsd,'angle',15*degree); % smooth grain boundaries grains = smooth(grains,5); hold on plot(grains.boundary,'linewidth',2) hold off % a key the colorizes according to misorientation angle and axis ipfKey = axisAngleColorKey(ebsd); % set the grain mean orientations as reference orinetations ipfKey.oriRef = grains(ebsd('indexed').grainId).meanOrientation; % plot the data plot(ebsd,ipfKey.orientation2color(ebsd('indexed').orientations),'micronBar','off','figSize','medium') hold on plot(grains.boundary,'linewidth',1.5) hold off % denoise orientation data F = halfQuadraticFilter; ebsd = smooth(ebsd('indexed'),F,'fill',grains); % plot the denoised data ipfKey.oriRef = grains(ebsd('indexed').grainId).meanOrientation; plot(ebsd('indexed'),ipfKey.orientation2color(ebsd('indexed').orientations),'micronBar','off','figSize','medium') hold on plot(grains.boundary,'linewidth',1.5) hold off % consider only the Fe(alpha) phase ebsd = ebsd('indexed').gridify; % compute the curvature tensor kappa = ebsd.curvature % one can index the curvature tensors in the same way as the EBSD data. % E.g. the curvature in pixel (2,3) is kappa(2,3) kappa12 = kappa{1,2}; size(kappa12) newMtexFigure('nrows',3,'ncols',3); % cycle through all components of the tensor for i = 1:3 for j = 1:3 nextAxis(i,j) plot(ebsd,kappa{i,j},'micronBar','off') hold on; plot(grains.boundary,'linewidth',2); hold off end end % unify the color rage - you may also use setColoRange equal setColorRange([-0.005,0.005]) drawNow(gcm,'figSize','large') alpha = kappa.dislocationDensity alpha(2,3) dS = dislocationSystem.fcc(ebsd.CS) % size of the unit cell a = norm(ebsd.CS.aAxis); % in bcc and fcc the norm of the burgers vector is sqrt(3)/2 * a [norm(dS(1).b), norm(dS(end).b), sqrt(3)/2 * a] nu = 0.33; % energy of the edge dislocations dS(dS.isEdge).u = 7.01317231; % energy of the screw dislocations dS(dS.isScrew).u = 4.698825451; dS(1).tensor dSRot = ebsd.orientations * dS [rho,factor] = fitDislocationSystems(kappa,dSRot); % the restored dislocation density tensors alpha = sum(dSRot.tensor .* rho,2); % we have to set the unit manualy since it is not stored in rho alpha.opt.unit = '1/um'; % the restored dislocation density tensor for pixel 2 alpha(2) % the dislocation density dervied from the curvature in pixel 2 kappa(2).dislocationDensity % we may also restore the complete curvature tensor with kappa = alpha.curvature % and plot it as we did before newMtexFigure('nrows',3,'ncols',3); % cycle through all components of the tensor for i = 1:3 for j = 1:3 nextAxis(i,j) plot(ebsd,kappa{i,j},'micronBar','off') hold on; plot(grains.boundary,'linewidth',1); hold off end end setColorRange([-0.005,0.005]) drawNow(gcm,'figSize','large'); factor close all plot(ebsd,factorsum(abs(rho . dSRot.u),2),'micronbar','off') mtexColorMap('jet') mtexColorbar set(gca,'ColorScale','log'); % this works only starting with Matlab 2018a set(gca,'CLim',[1e12 1e16]); hold on plot(grains.boundary,'linewidth',1) hold off % total gnd_tot = factorsum(abs(rho . dSRot.u),2); grainGND = accumarray(ebsd.grainId, gnd_tot, size(grains), @nanmean); %total grain average GND map plot(ebsd,grainGND,'micronbar','off') mtexColorMap('jet') mtexColorbar set(gca,'ColorScale','log'); % this works only starting with Matlab 2018a set(gca,'CLim',[5e12 1e16]); hold on plot(grains.boundary,'linewidth',1) hold off Is there anything wrong with the using of accumarray? — You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.

[kilir]

Hi,
you can also use the function EBSD/grainMean which is basically a wrapper around accumarry and has some option. See help grainMean.

Cheers,
Rüdiger