Microscale Testing and Characterization Techniques for Benchmarking Crystal. . .
93
A multitude of ICME studies on oligocrystalline structures have been published
on materials that include Zirconium alloys and near-gamma titanium aluminides
[10], polycrystalline columnar Al [6, 8, 11], single and bicrystal stainless steel [6],
hot worked waspaloy-ingot specimens [12], and coarse-grained tantalum [9]. The
ability to model the full sample gage of these samples in 3D had allowed for a
more accurate approximation of the true boundary conditions experienced by the
sample during an experiment and a more accurate prediction of the deformation
response. Another technique that has been utilized when only 2D microstructural
data is available is the use of statistics to extrapolate a 3D microstructure from a
2D Electron Backscatter Diffraction (EBSD) scan, such as the work of St-Pierre
in modeling TiAl and grade 702 zirconium [13]. This novel modeling approach is
however difficult to truly compare one-to-one with experimental results, as even
slight changes in the subsurface microstructure can lead to significant deviations in
the observed behavior at the sample surface.
The most prevalent experimental technique for building a 3D dataset involves
serial sectioning of samples by deconstructing the sample layer by layer and
characterizing each newly exposed surface before performing the next sectioning
step. The acquired 2D scans can be stitched together to provide a 3D representation
of the sample microstructure. Musienko was one of the first to characterize and
model a sample using this methodology on a tested Cu microtension sample [14].
A small subsection of the gage, containing about 100 grains, was characterized in
this manner and used to generate a finite element simulation of the 3D structure.
Although this small subsection did not represent the full physical volume of the
sample required to capture requisite boundary conditions for a true one-to-one
model, this was one of the first times that a modeling effort was performed on a
sample characterized in 3D. Spanos, Lewis, Rowenhorst, and coworkers at the Naval
Research Laboratory combined serial sectioning using a Buehler Minimet system
and characterization with EBSD to develop 3D datasets of stainless steels and Ti
alloys to connect with FEM models [15]. More automated methodologies using
mechanical sectioning, such as the Alkemper-Voorhees micromiller, developed at
Northwestern University [16], or the Robomet.3D system, developed at the Air
Force Research Lab [17, 18], were major advances in streamlining workflows for
3D dataset collection.
Within the last decade or so, the use of more advanced sectioning tools and
methods has enabled significant advances in studying microstructure in 3D. In
dual beam systems equipped with both a scanning electron microscope (SEM) and
focused ion beam (FIB), serial sectioning via FIB and EBSD can be performed and
automated in a routine manner. Uchic et al. and Groeber et al. were some of the first
to demonstrate this technique in sectioning samples of the nickel-base superalloy
IN100 [19, 20]. Shortly thereafter, Zaafarani et al. used FIB serial sectioning to
characterize the microstructure surrounding a nanoindent in Cu and to develop
a finite element model from the collected data [21]. In terms of utilizing this
technique for characterizing and modeling a full sample volume, the applications
have been limited by the low material removal rate of the FIB. One study that
did yield positive results despite such limitations was carried out by Shade et al.
93
A multitude of ICME studies on oligocrystalline structures have been published
on materials that include Zirconium alloys and near-gamma titanium aluminides
[10], polycrystalline columnar Al [6, 8, 11], single and bicrystal stainless steel [6],
hot worked waspaloy-ingot specimens [12], and coarse-grained tantalum [9]. The
ability to model the full sample gage of these samples in 3D had allowed for a
more accurate approximation of the true boundary conditions experienced by the
sample during an experiment and a more accurate prediction of the deformation
response. Another technique that has been utilized when only 2D microstructural
data is available is the use of statistics to extrapolate a 3D microstructure from a
2D Electron Backscatter Diffraction (EBSD) scan, such as the work of St-Pierre
in modeling TiAl and grade 702 zirconium [13]. This novel modeling approach is
however difficult to truly compare one-to-one with experimental results, as even
slight changes in the subsurface microstructure can lead to significant deviations in
the observed behavior at the sample surface.
The most prevalent experimental technique for building a 3D dataset involves
serial sectioning of samples by deconstructing the sample layer by layer and
characterizing each newly exposed surface before performing the next sectioning
step. The acquired 2D scans can be stitched together to provide a 3D representation
of the sample microstructure. Musienko was one of the first to characterize and
model a sample using this methodology on a tested Cu microtension sample [14].
A small subsection of the gage, containing about 100 grains, was characterized in
this manner and used to generate a finite element simulation of the 3D structure.
Although this small subsection did not represent the full physical volume of the
sample required to capture requisite boundary conditions for a true one-to-one
model, this was one of the first times that a modeling effort was performed on a
sample characterized in 3D. Spanos, Lewis, Rowenhorst, and coworkers at the Naval
Research Laboratory combined serial sectioning using a Buehler Minimet system
and characterization with EBSD to develop 3D datasets of stainless steels and Ti
alloys to connect with FEM models [15]. More automated methodologies using
mechanical sectioning, such as the Alkemper-Voorhees micromiller, developed at
Northwestern University [16], or the Robomet.3D system, developed at the Air
Force Research Lab [17, 18], were major advances in streamlining workflows for
3D dataset collection.
Within the last decade or so, the use of more advanced sectioning tools and
methods has enabled significant advances in studying microstructure in 3D. In
dual beam systems equipped with both a scanning electron microscope (SEM) and
focused ion beam (FIB), serial sectioning via FIB and EBSD can be performed and
automated in a routine manner. Uchic et al. and Groeber et al. were some of the first
to demonstrate this technique in sectioning samples of the nickel-base superalloy
IN100 [19, 20]. Shortly thereafter, Zaafarani et al. used FIB serial sectioning to
characterize the microstructure surrounding a nanoindent in Cu and to develop
a finite element model from the collected data [21]. In terms of utilizing this
technique for characterizing and modeling a full sample volume, the applications
have been limited by the low material removal rate of the FIB. One study that
did yield positive results despite such limitations was carried out by Shade et al.