Plotting the misorientation distribution

[John-Michael Bradley]

Hello,

I was wondering how I might use DREAM.3D to plot the misorientation distribution of a 2D or 3D data set (single-phase polycrystal). This is a length (2D) or area (3D) weighted frequency plot of the misorientation between grains in a volume. It is a simple idea that is not too hard to code. You just need to look at each point that lies on a grain boundary, calculate the misorientation between the grains, and then bin the information. This leaves you with a simple plot that displays which misorientations have a higher frequency in your sample. However, symmetry operators have to be taken into account (hence my desire to use DREAM.3D which already has this capabilty).

I'd appreciate any thoughts.

John-Michael Bradley

[Michael Groeber]

John-Michael-

I have a question about which version of the MDF you are requesting. Do you want the 1-d, McKenzie plot of only the angles of misorientation or do you want the 3-d MDF with axis and angle? If it is the former, we can do that pretty easily. The latter would take a little more work to slice through the 3-d MDF, but should also be doable from data we already calculate.

Thanks in advance

[John-Michael Bradley]

Right now I am interested in the 1-d McKenzie plot, but in the future I will probably be interested in both.

[Sean Donegan]

Hi John-Michael,

It is possible to get the 1-D misorientation information directly.  The filter you want to run is called "Find Feature Neighbor Misorientations".  This filter will calculate the misorientations between each feature and its neighbors and store the angle.  You can then use the filter "Export Data (ASCII Text)" to write out the misorientation angles.  Note that on the "Find Feature Neighbor Misorientations" filter you will need to check the option to compute the average misorientations per feature and then select this array (AvgMisorientations) to write in the export filter; this is because the misorientation list that is generated cannot itself be written by the export data filter at this time.  Once you write the misorientation angles, you can plot them as a frequency histogram, which should give you the McKenzie plot.

hope that helps,

-Sean 

[Michael Groeber]

Sean

I think this is a little different than what Jon-Michael wants.  The Average Misorientation Per Feature is the average of the misorientations of all neighbors for each feature, so it isn't exactly what you would want to plot for a 1-D MacKenzie style plot.  The Export Data (ASCII Text) filter cannot currently write out the misorientation lists, but the Write Feature Data As CSV File filter can.  If you run that filter, there is an option to Write Out Neighbor Data and if you check that it will write out all the neighbor connectivity lists and the misorientation lists if you have them.  The format of that file is not the easiest to use, but would have all the information John-Michael wants.  The file will have all the scalar and vector attributes of the Features listed first, but if you scroll down below that, there will be blocks of data that are the neighbor information for each feature.  It should be described in the documentation, but you will get a line for each Feature and a list of its neighbors or a list of misorientations it has with those neighbors or a list of the shared surface area with those neighbors, etc.  In something like Excel, you could just highlight all of misorientation values for all of the Features and make a histogram of the selection.

Hope that helps for now and we can work on including a more direct 1-D misorientation calculation filter in the future.

Mike

[John-Michael Bradley]

Thanks for the help. I will see if I can make due with that.

John-Michael