Applying inverse symmetry on EBSD IPF for Titanium (Alpha) - 622

[Chris Triantafyllou]

Hi MTEX community!

I'm in the process of transitioning from EDAX's OIM Analysis software to MTEX for greater post-processing flexibility and ease of using a MATLAB-based routine for integrating with my computational work. I am trying to get a colormap that is largely similar to the one from OIM to be consistent with older scans and with colleagues' scans in our group who use OIM. 

In the image I have attached for reference you can see what the intended image looks like on the far left - this was generated using the default IPF settings in OIM which includes 'inverse symmetry' enabled by default. In MTEX 5.1.1, I am importing my EBSD data from *.ang and use

CS = {...

  'notIndexed',...

  crystalSymmetry('622', [2.95 2.95 4.68], 'X||a', 'Y||b*', 'Z||c', 'mineral', 'Titanium (Alpha)', 'color', 'light blue'),...

  crystalSymmetry('432', [3.31 3.31 3.31], 'mineral', 'Titanium (Beta)', 'color', 'light green')};

and

setMTEXpref('xAxisDirection','north');

setMTEXpref('zAxisDirection','outOfPlane');

followed by

oM.inversePoleFigureDirection = zvector;

color = om.orientation2color(ebsd('Titanium (Alpha)').orientations);

plot(ebsd('Titanium (Alpha)'),color)

which of course generated the plot using the default MTEX colouring scheme, shown in the next image to the right. There is nothing wrong with it and if anything, it is more insightful than the original, but for the reasons listed above I would like to similarly apply 'inverse symmetry' or express the orientations in the unit triangle. Effectively, I would like all the points in the darker half to be 'reflected' on the lighter one and I believe this will give me the desired output.

This issue has been brought up before here, where the use of TSLOrientationMapping or ipfTSLKey was suggested. Using

oM = ipfTSLKey(cs);

we get an oM which is visually very close to the desired one, but the resulting IPF is visually very different and the following observation has been made: points in the, say for example, dark green region of the original MTEX plot have not been mirrored to be light green (but light green ones are still light green as expected), but have been "compressed" to fit that oM resulting in their positioning at the top and slight right of the unit triangle subsequently resulting in the dark green points being shown in purple. I don't have the capacity to say whether this is a bug of some sort or improper use of TSL in this case.

Overriding the standard crystal symmetry from 622 to 6/mmm, as shown in the final plot, does not seem to be a good idea either. I have also tried to use 'antipodal' for it and the project2FundamentalRegion function with no success. The default EBSD plot from MTEX is currently the closest match and I have a feeling that reflecting the upper half of the triangle would probably be easy to implement, but anyone's advice on how to do that would be much appreciated!

Kind regards,

Chris

--

Chris Triantafyllou
MEng  AMIMechE
PhD Student in Materials Engineering

James Watt South Building

University of Glasgow

University Avenue, G12 8QQ

[Ralf Hielscher]

The TSL keys look indeed a bit strange when compared to the default maps.

you may try the following

1. you can change 622 to 6mm in the default key by

ipfKey = ipfHSVKey(cs.Laue);
ipfKey.inversePoleFigureDirection = zvector;

2. One can plot the pixel orientations into the color key by

plot(ipfKey)
plot(ebsd.orientations,'add2all')

This may help to find mistakes.

Ralf.

[Chris Triantafyllou]

Good evening Ralf,

Thank you so much for your reply so quickly! Indeed, adding

ipfKey = ipfHSVKey(cs.Laue);

produced the desired image straight away. The produced plot is a great match and a little bit more vivid as well. Plotting some random orientations on the colour key also produces a reasonable result and I can now happily proceed with cleaning!

Your help is much appreciated! Looking forward to exploring more MTEX features in the near future.

Kind regards,

Chris