Artifacts created during Pole Figure to ODF inversion

[Emanuel Benatti]

Hi,
I've been using MTEX 3.5.0 in Matlab 2013a for a while now, and I've been having some problems in working with some X-ray difraction pole figures. I've been working with some pole figures obtain from a transmission diffraction experiment using a synchrotron X-Ray source (lambda = 0.014235 nm). In the .mtex files is the raw data. The sample is a 1050 aluminum sheet laminated to a reduction of 70 %. When I plot this raw data I get the pole figures shown in Raw_pole_figure.png, which shows the classical lamination texture. However when I recalculate the same pole figures from the ODF I get what is shown in Recalculated_pole_figure.png where it can be clearly seen the appearance of some spurious maxima.
To solve this problem, I've been trying the following options from the calcODF routine:

• Change the 'RESOLUTION' parameter
• Change the 'HALFWIDTH' parameter
• Applied the zero range method
• Remove the edges of the experimental data and calculate the ODF only with low angle data (theta < 60 degrees)

None of this options helped to remove this spurious maxima. The image shown is the one I get using calcODF with its default values:

AlSR70_pcfn_odf = calcODF(AlSR70_I_pfcn_rot(1:3));

I'm also attaching the m-file I used to generate the pole figures and calculate the ODF.

Any help/advice will be apreciated.
Regards
Emanuel

[Ralf Hielscher]

Hi Emanuel Benatti,

I agree that the result is very unsatisfactory. I'm working on this.

Ralf.

[Ralf Hielscher]

Hi Emanuel Benatti,

something funny happened to you pole figure data. Every pole figure position exists twice in every pole figure. Therefore you have 3300 point which is twice as much as needed for this resolution. Furthermore, the double points show different pole figures the first 1620 points build up one pole figure and the remaining points a second one. I have split all your data and plotted the data. Same Miller indice indicates that the data where formerly mixed in one pole figure but invisible due to overlapping of the points. I think this explains your problems.

[Emanuel Benatti]

Hi,
 Thanks Ralph for taking the time of checking the data. The problem was precisely that. I can't believe I didn't noticed it before. Now that I have separated de data everything works just fine.

Thanks a lot!
Emanuel