how to analyze the GND data

[王强]

Dear Ralf and MTEX users,

    As a new user to MTEX, I am asking for some help, e.g. commands, documents, etc. to understand the result from the GND calculation. A 340*395 matrix was obtained though. My material is Hexagonal Zirconium. I am using the new MTEX-4.3.1. 

Thanks,

Qiang

[Ralf Hielscher]

Have a look here

https://github.com/mtex-toolbox/mtex/issues/164

Ralf.

[Matthew Vaughan]

Hey Ralf,

I have been able to calculate GND for single phase ice. Is there a syntax for making maps to visualize this data? Cheers,

[Thomas Simm]

Have you tried something like this?

ebsd2=ebsd(phase_).gridify;

        gnd=calcGND(ebsd2);

        plot(ebsd2,gnd)

Normally you can plot a variable of ebsd like plot(ebsd, ebsd.bc), as a map or plot(ebsd.bc) for values

so ebsd.gnd is worth a go but above works fine for me

Tom

[Matthew Vaughan]

Thanks Tom this works. The dynamic range is dominated entirely by grain boundaries. I wonder if there is an approach for calculating GNDs for grain interiors? 

Cheers

[Thomas Simm]

Hi,

This now becomes more of a programming problem.

The way I may start something like this, although the answer is not complete and the use of logicals would be better

lets say you have a grain no 978- 

% all points in the grain

xx=grains(978).x;

yy=grains(978).y;

% and perimeter points

px  = grains(978).boundary.x; 

py  = grains(978).boundary.y;

% then apply some critiria 

dist_min=3; 

h=1;

for n=1:length(xx) %probably better ways than a for statement but easier to display my point

    dist= min( ((xx(n)-px).^2 + (yy(n) - py).^2 ).^.5 );%distance between each point and perimeter

    if dist > dist_min

        [xuse(h) yuse(h)] = [xx(n) yy(n) ]; h=h+1;

 pos_go(h)= n;

    end

        

end 

You now have all the points within a grain that are some distance from the perimeter and you can calculate the GND from these points instead of all points with a grain

Tom

[Matthew Vaughan]

How are the units given for GNDs?  I get values in the matrix from 0.0001 to 0.1 but I don't know what they mean. Is there any documentation on this?

Cheers

On Saturday, March 26, 2016 at 10:41:26 AM UTC+13, Ralf Hielscher wrote:

[Thomas Simm]

Hi,

You'll probably have to read the paper Pantleon Scripta Materialia, 2008 

• DOI: 10.1016/j.scriptamat.2008.01.050 

As I don't think there is much documentation on it in the MTEX files

Tom

[Richard Hugo]

Hi - I've been trying to work this out as well. As far as I can tell, the calcGND code just calculates the curvature tensor - the orientation gradient with respect to distance in the specimen coordinate frame. The full 3-d tensor has been calculated by assuming that the orientation gradients in the z direction are negligible, thus setting the z-direction components to zero.

Yes? So units would be radians/micrometer?

Or am I missing something? I don't see that the Nye dislocation density tensor is calculated from curvature, or that the slip system data (Burgers vectors/slip planes) necessary to convert the DD tensor to GND are stored, or that any type of dislocation energy/line length optimization is employed. I am using MTEX version 4.4.alpha.2.

Not that this is a complaint - the curvature tensor is quite useful!

Rick

--
If you want to reduce the number of emails you get through this forum login to https://groups.google.com/forum/?fromgroups=#!forum/mtexmail, click "My membership" and select "Don't send me email updates". You can still get emails on selected topics by staring them.
---
You received this message because you are subscribed to the Google Groups "MTEX" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mtexmail+u...@googlegroups.com.
Visit this group at https://groups.google.com/group/mtexmail.
For more options, visit https://groups.google.com/d/optout.

[Thomas Simm]

Think you're probably right the code says:

  % as we do not know the texture gradient with respect to the z-direction

  % we have to do something here TODO!!

  alpha = kappa - trace(kappa)*eye(3);

The software below has something similar, would be interesting to compare, although documentation seems to only reference two papers. But I think it has an axis label. Fairly sound software, well made, quick and free. Although because it's a full application not the flexibility of MTEX

http://www.atom-software.eu/

GND is on my to-do list so I'm not a lot of help at the moment, but future me is very interested

You received this message because you are subscribed to a topic in the Google Groups "MTEX" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/mtexmail/Caz48tEI4e8/unsubscribe.
To unsubscribe from this group and all its topics, send an email to mtexmail+u...@googlegroups.com.