noisy euler angle

[Yao Fu]

Hello, I am new to dream3d and intend to use it to process the experimental EBSD data for MOOSE input (https://mooseframework.inl.gov/source/userobjects/EBSDReader.html).

The problem I have is the very noisy euler angle data which is needed by moose. The recommended ipf color was shown fine in paraview.  I am wondering 1) how to obtain clean euler angle data, 2) the ipf map also seems to lost some resolution compared to the original data (I did a conversion from hex to quad for the ang file to get the Sqr_1.ang).

Attached are the filters I used, sqr_ang file, the output ipf and euler angle file and the original ipf map.

Yao

[Michael Jackson]

Dear Yao,

  To visualize Euler angle data, IPF colors are highly recommended. Coloring purely by Euler angle does not really tell you much as Euler angles with numerically different values can represent the same Orientation. Looking at the IPF color maps the orientation data looks very nice and not noisy at all. 

If you think there was a loss of resolution when you converted from Hex to Square grid, I would go back and double check the values that were used for the conversion. The filter that does the conversion does not have any prior knowledge of the hex file so it is up to you to enter reasonable values to generate the Square grid.

--
Mike Jackson                    mike.j...@bluequartz.net
BlueQuartz Software         www.bluequartz.net

President/Owner               Dayton, Ohio

Principal DREAM.3D Developer

--
You received this message because you are subscribed to the Google Groups "dream3d-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dream3d-user...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dream3d-users/757b9479-a91f-4746-bcf9-801cab0e7244n%40googlegroups.com.

[Yao Fu]

Thank you Mike, I understand your suggestion of IPF map for visualization.  But for moose, it appears to me that they use euler angle for input (https://mooseframework.inl.gov/source/userobjects/EBSDReader.html), and mentioned dream3D to prepare the input file.  So I am not sure how the noisy euler angle would provide an input.  Could you advise on this? 

[Anthony Rollett]

Hi Yao: can you show us a picture of what you mean by “noisy Euler angle”?  Why the adjective “noisy”?  I.e., give us more context.

Thanks, Tony

[Yao Fu]

Tony,

That was provided in the first post, where I included a few figures, one of them is the euler angle after dream3d. The figure was generated in Paraview.  

[Michael Jackson]

Dear Yao,

   Using this link (https://github.com/idaholab/moose/blob/next/modules/phase_field/doc/content/source/userobjects/EBSDReader.md) what MOOSE wants is the output from the “Write INL File” filter. You will need to segment your EBSD data using the “Segment Features (Misorientation)” filter (as a start) and then use the “Write INL File” filter to output the necessary data file for input into MOOSE. 

We should probably rename that filter if MOOSE is the only one using it.

—

Mike Jackson

BlueQuartz Software

To view this discussion on the web visit https://groups.google.com/d/msgid/dream3d-users/1018cff3-20eb-4d7b-afff-14198b2cae68n%40googlegroups.com.

[Yao Fu]

Mike,

I did output to INL file and it is readable by moose. But when I output the file from moose to check the euler angle by paraview, I got similar noisy data.  The “Segment Features (Misorientation)” filter was used but I am not sure exactly the order to place it.  I attached the modified pipeline that includes 'write INL' and euler angle (2nd one of the three) snapshot from the output moose file.   

Could you advise what I should modify further? 

Yao

[Michael Jackson]

Dear Yao, 

    You should segment the features, and then write out the INL file. In fact there is an example pipeline that comes with DREAM.3D “Prebuilt Pipelines->Examples->INL_Export” that is a complete working example.

Also, just to be clear, visualizing the data by coloring by any of the Euler angles really isn’t going to tell you much. IPF Colors is the preferred way to look at the Euler angles. This is because the Euler Angle can have many different numerical values but all represent the same orientation. For instance, this Euler Angle in Space Group 32 (m-3m) has all the following equivalent values (The input is in degrees, the output is in Radians):

 Euler (phi1°, Phi°, phi2°) : 32.324, 10.324, 52.342

  0.73624518786082083       -8.8602731713122870E-002   1.5637402965217925E-002 -0.67070712752531580     

   8.8602731713122870E-002  0.73624518786082083       0.67070712752531580        1.5637402965217925E-002

   1.5637402965217925E-002  0.67070712752531580      -0.73624518786082083       -8.8602731713122870E-002

  0.67070712752531580       -1.5637402965217925E-002  -8.8602731713122870E-002  0.73624518786082083     

  0.58325555737835166       0.45795237252634852       0.48531887174015353      -0.46320424438644914     

  0.50954665127549781      -0.53691315048930288       0.53166127862920232      -0.41160996563729979     

FZ  0.99486552301565134       -7.3708906102853786E-002  -5.1594278749149340E-002   4.6342406889048726E-002

  0.45795237252634852      -0.58325555737835166      -0.46320424438644914      -0.48531887174015353     

  0.53166127862920232       0.41160996563729979      -0.50954665127549781      -0.53691315048930288     

   4.6342406889048726E-002  -5.1594278749149340E-002   7.3708906102853786E-002 -0.99486552301565134     

   5.1594278749149340E-002   4.6342406889048726E-002  0.99486552301565134        7.3708906102853786E-002

   7.3708906102853786E-002  0.99486552301565134       -4.6342406889048726E-002  -5.1594278749149340E-002

  0.46320424438644914      -0.48531887174015353       0.45795237252634852       0.58325555737835166     

  0.48531887174015353       0.46320424438644914      -0.58325555737835166       0.45795237252634852     

  0.53691315048930288       0.50954665127549781       0.41160996563729979       0.53166127862920232     

  0.41160996563729979      -0.53166127862920232      -0.53691315048930288       0.50954665127549781     

  0.73995882206702079       -1.9351037171417840E-002  0.66699349331911584        8.4889097506922917E-002

   3.7136342061999672E-003   6.9251694541704989E-002  0.65135609035389797       0.75559622503223878     

  0.75559622503223867       0.65135609035389797       -6.9251694541704989E-002  -3.7136342061999603E-003

   1.9351037171417840E-002  0.73995882206702079       -8.4889097506922861E-002  0.66699349331911595     

  0.65135609035389797      -0.75559622503223867       -3.7136342061999672E-003   6.9251694541704989E-002

   8.4889097506922861E-002  0.66699349331911584        1.9351037171417840E-002 -0.73995882206702079     

  0.66699349331911584       -8.4889097506922861E-002 -0.73995882206702079       -1.9351037171417888E-002

   6.9251694541704989E-002  -3.7136342061999672E-003  0.75559622503223867      -0.65135609035389797  

So viewing the data by an individual Euler component isn’t really telling you anything about the data.

--
Mike Jackson                    mike.j...@bluequartz.net
BlueQuartz Software         www.bluequartz.net

President/Owner               Dayton, Ohio

Principal DREAM.3D Developer

To view this discussion on the web visit https://groups.google.com/d/msgid/dream3d-users/ab47adce-9325-48f9-94c8-d60d239931ddn%40googlegroups.com.

[Yao Fu]

Thanks Mike for your explanation.  I was worried because Moose will conduct further calculation (such as anisotropic elastic tensor) based on the euler angle so I wanted to make sure they are all correct.  I will explore further.  If different sets of euler angle represent the same orientation then the ’noisy’ euler angle may not be a concern in further calculation in Moose.

Best,
Yao