Although they are kernel based method (Kernel means given area/size, other than grain based method), the calculation method is a little different:
For KAM calculation, the misorientations between the center point of the kernel and all surrounding points in the kernel are calculated and averaged, which gives the local misorientation value of the center point.
For LAM calculation, first calculate the misorientations between all neighbor points within the kernel. Then the local misorientation is assigned to the center point of the kernel as the average value of the misorientations calculated in first step.
Although they are kernel based method (Kernel means given area/size, other than grain based method), the calculation method is a little different:
For KAM calculation, the misorientations between the center point of the kernel and all surrounding points in the kernel are calculated and averaged, which gives the local misorientation value of the center point.
For LAM calculation, first calculate the misorientations between all neighbor points within the kernel. Then the local misorientation is assigned to the center point of the kernel as the average value of the misorientations calculated in first step.
Dear Gert:
You could find it in the book:
E. M. Lehockey, Y. P. Lin and O. E. Lepik (2000). Mapping Residual Plastic Strain in Materials Using Electron Backscatter Diffraction. Electron Backscatter Diffraction in Materials Science. A. J. Schwartz, M. Kumar and B. L. Adams. New York: 247-264.
Dear Dr. Gu,
Thanks for the detailed explanation for KAM and LAM.
Do you have any good ideas for quantifing strain with the ESBD measurements? (Especially for the highly deformed material with broken grains)
Dear Mr. Li, I have applied the FIB technology to the preperation of the samlpe. But it's not suit for a large area. If the chosen area is too large, the surface couldnot be prepared smoothly.
I think, one of the biggest problem is to make sure WHAT are you actually interested in and what you are really measuring with EBSD. To get some results is no problem as long as you get patterns. But what do they contain? It should be clear that FIB is NOT the best way for sample preparation of only the direct surface is investigated as in case of EBSD. For EM no problem, even if the direct surface is amorphous after ion polishing. However, it s well known that FIB changes the surface by ion implementation (strain) and this can even cause unexpected phase transformation. Thus you have to expect that the results may look nice but do not describe your materials property. We checked this about 10 years ago using manually prepared iron meteorite samples. The EBSD patterns collected after 1.5kV Ar preparation (50° tilt) tuned into a considerably worse quality.
The other problem is if you generate a free surface (or even two in FIB), how much this is still representative for your bulk material?
Regarding lattice strain determination using HR-EBSD (CrossCourt) the technique requires excellent patterns and reference patterns with an orientation close to the correlated patterns. In highly deformed materials this is usually not fulfilled. The reference pattern could be alternatively generated by dynamical simulations (e.g. A. Winkelmann) but you need at least a well visible diffraction signal since the visible effects are commonly extremely small and below a pixel resolution of an image with 1600x1200 or 1024x1024 pixels. Moreover, you need a perfectly calibrated projection center which is nowadays perhaps the bigger problem than image distortions generated by electromagnetic fields (magnetic phases or the equipment) , lens distortions (can be post-processed) etc. Do the same measurement at two different systems and you will realize what I am talking about. Relative investigations are quite acceptabe, absolute are tricky...
- Badges
- Science topic
- Similar topics
- Engineering
- Materials Engineering
- Annealing
- Metallurgy