Dear Ahmad,

I suppose that you already have a SEM equiped with an EBSD detector and the corresponding software to identify the different orientations. Then, you can do it to obtain a map of orientations (for instance in colour), the second step is to use an image analysis software to quantify the area fraction of each colour-orientation. Very often, this software is included (or an option) of the EBSD software.

Finally, if you need the volume fraction, you probably should do several analysis, by removing a certain amount of material (by polishing) and repeat the EBSD analysis. Like a tomography.

I suggest to ask the company of your EBSD detector/software, about this methodology.

Good luck

Prof. Jose San Juan

Dear Ahmad,

I suppose that you already have a SEM equiped with an EBSD detector and the corresponding software to identify the different orientations. Then, you can do it to obtain a map of orientations (for instance in colour), the second step is to use an image analysis software to quantify the area fraction of each colour-orientation. Very often, this software is included (or an option) of the EBSD software.

Finally, if you need the volume fraction, you probably should do several analysis, by removing a certain amount of material (by polishing) and repeat the EBSD analysis. Like a tomography.

I suggest to ask the company of your EBSD detector/software, about this methodology.

Good luck

Prof. Jose San Juan

Ahmad,

Beware: The "fraction of orientation" can only be described correctly using statistical distributions. When metallurgist measure the fraction of crystals with a certain orientation, they mean the fraction of crystals whose orientation is close to the said orientation, often without specifying how close. The "fraction of orientation" is a practical but approximate description. The statistical distribution language (ODF) is accurate, but less practical. As long as you are aware of this, then you can use the "fraction of orientation" when it is appropriate and the ODF in other cases.

Dear Jose San Juan,

Thank you so much, 

I found out the fraction of each orientation using EBSD software (TSL). Image analysis was also included there. However still I don't understand the meaning of pole figure's intensity, while the intensity of a given orientation perpendicular to the observed surface has some how same intensity as the one parallel to the observed surface. Just in pole figure we can see they have different directions or orientations. It seems to me intensity of pole figure is not related to their contribution in the analyzed surface. I am not sure but it seems that I should say it is some how meaningless.

Many thanks

Hi,

I can't give you a complete introduction to texture in metals here, but I may be able to give you a few pointers. Nomenclature:- roughly speaking texture refers to the intensity of crystallographic orientations which have some relationship to the sample or real world frame of reference. In former times we would use pole figures generated by X-Ray diffraction for texture work. We would set the XRD kit up so that it would specifically look for the Bragg reflections from one type of crystal plane, e.g. {100}, in the sample. The sample itself would be set up in the holder so that its key faces and directions, (say those faces parallel to the rolling plane/ those directions parallel to the rolling direction), were known and fixed. The X-Ray source and detector would then move to sample 'all' possible directions in (say) the upper hemisphere for X-Ray reflections, (NB the sample might be rotated slowly and the X-Ray source and detector just alter their angles, often easier to engineer) . Look in a standard text for pole figure construction, if this is not clear. Maybe webpages by Bruker addresses this ? The X-Ray intensities so produced can then be plotted on a pole figure, again see std texts. However, pole figures are ambiguous, e.g. for a given fixed orientation of a cubic crystal there are a set of three, {100}, reflections which will appear on the pole figure, if a single crystal was involved, great, we would know its exact orientation. The problem is that when we produce a pole figure from a polycrystal there are many such sets of three and we have no idea which set of three a particular reflection belongs to. However, Bunge, Roe and Williams, independently developed the Orientation distribution function, ODF, method which can take 3 or 4 pole different pole figures, say {100}, {110} and {111}, from the same sample, and derive such unambiguous single crystal orientations present in the polycrystalline sample, (see Bunge's book, 'Texture Analysis in Materials Science - Mathematical Methods' for the details of the spherical harmonic way of accomplishing this, and the importance of Euler angles in this description). Now the nature of EBSD is quite different to XRD, i.e. we can (normally) determine the crystal orientation of the individual grains/crystals directly.

What exactly do we mean by a texture ? well for a specific rolling texture, it would be that a certain crystal plane has a much higher than random frequency of being parallel to the rolling plane, and a direction lying in that crystal plane has a higher than random frequency of being parallel to the rolling direction. So the format describing such textures is {hkl}, note it has two components in its description.

How to describe the volume fraction of textures is a subject in itself, however, briefly it is possible to segment, the orientation space, the ODF, into Euler angle 'boxes', then quantify the number of orientations from the sample which fall into that particular box. That number can be compared against that in the same 'box' for a random or uniform distribution, and thus after appropriate normalisations, (and accounting for the known distortions in such a Euler space, see Bunge's book), one can quote volume fractions for the texture components. That's an approach one might take coming at the problem from XRD and say using the van Houtte texture software MTM-FHM.

[NB For a wire drawing texture normally only those crystal directions which are parallel to the wire drawing direction are of interest/have meaning, as we have rotational sample, (real world), symmetry in plane perpendicular to the wire drawing direction. Thus here a single descriptor of texture suffices. ]

In EBSD software packages after collecting the EBSD data normally you can interrogate the system to find out the fraction of crystal directions parallel to the main sample directions, (within a certain angular tolerance), and you can also interrogate the system to find out the fraction of crystal orientations of the type you have specified, (again within a certain angular tolerance). Online/in journals you will find papers discussing the effect of that angular tolerance on the results one obtains for the %volume fraction of texture components in the same sample.

So you don't have to resort to image analysis of the colours, but often in presenting the results of EBSD work it's useful to let the human eye do that and produce coloured IPF grain maps etc. in addition to tabulated numerical data of texture volume fractions.