Different results of ODF calculation from MTEX3.4.1 and MTEX4.1.0

[Vicky Liu]

Hello,

I use to deal with all my results from EBSD measurements and VPSC simulations with MTEX3.4.1 and Matlab2012b. Now I try to do the same operations with MTEX4.1.0 and also Matlab2012b. I try to find out the same calculation methods and results from these two MTEX versions, of course, with different represent of codes, however, the results seem not so good.

First, for EBSD measurement results, after importing data and grains re-construction, the calculated ODFs (phi2=45*degree section) are:

from MTEX3.4.1: , and

from MTEX4.1.0: .

Their texture-indexes are: 

MTEX3.4.1-odf-textureindex_ini: 2,55622315273035

MTEX4.1.0-odf-textureindex_ini: 2,50466702577914

These results are not so much different from each other. So it's ok for me, even though I can't find the maximum and minimum values for ODF plotting in MTEX4.1.0.

But turn to individual orientations, the results are so different.(the input data and whole codes are attached.)

the input 5000 orientations are simulated from the calculated ODF of EBSD measurement using the function "calcEBSD(odf)" .

I use functions "ODF" and "calcODF" in MTEX3.4.1, and functions "calcKernelODF" and "calcODF" in MTEX4.1.0. The chosen kernel functions are:

psi3.4.1=kernel('de la Vallee Poussin','halfwidth',3*degree);

psi4.1.0=deLaValeePoussinKernel('halfwidth',3*degree);

I think they should get almost the same results. But:

function "ODF" calculation result plotting with MTEX3.4.1:

, and its texture-index is: 2,57435103415383.

function "calcODF" calculation result plotting with MTEX3.4.1:

, and its texture-index is: 2,56704432556101.

function "calcKernelODF" calculation result plotting with MTEX4.1.0:

, and its texture-index is: 10,0338639786863.

function "calcODF" calculation result plotting with MTEX4.1.0:

, and its texture-index is: 10,0338639786863.

We can see that two calculation method for ODF in MTEX4.1.0 are same, but their results are different with MTEX3.4.1. And the intensities and texture-indexes of ODF figures in MTEX4.1.0 are quite different with the initial ODF of EBSD measurement. So I can't go on analyzing with my simulated-output orientations from VPSC with MTEX4.1.0

Could you help me check the codes I used in both MTEX version? What's the difference in them leading to the different results?

By the way, how to show the the maximum and minimum values in plotODF/PDF in MTEX4.1.0?

Thanks in advance!

Vicky Liu

[Ralf Hielscher]

Hi Vicky Liu,

this is indeed a bug in MTEX 4.1. If you look closely you see that the results differs about the factor 2 and accordingly the texture index differs by the factor 4. A fix will be available with MTEX 4.1.4. You can display min / max by

plot(testodfRot1,'phi2',45*degree,'resolution',2.5*degree,'contourf','minmax');

I hope this helps,

Ralf.

[Vicky Liu]

Hi Ralf Hielscher,

Thanks for the update of MTEX4.1.4, the texture-index problem has been solved.

And 'minmax' also work well for displacement of max/min values. Thanks again.

Best regards,

Vicky Liu