246
N. Wade et al.
parameters all the way through to computational models. The approach can easily
introduce new modules and/or additional complexity as needed.
We show how the framework can provide detailed insight into the influence of
data collection parameters, such as resolution, electron beam energy, and dwell
time. The resolution study showed that increasing sample point spacing results in
an approximately linear increase in mismatched volume, and an analytical model
was developed that provides basic insights into this behavior. Additional studies
showed how changes to the interaction volume (i.e., electron beam energy) or
increased levels of random noise (i.e., shorter dwell times) also affect error. It was
also demonstrated that through proper data processing parameters, much of this
error can be mitigated; however, an improperly applied erode/dilate data processing
filter can increase error for certain microstructural types. Finally, an example was
provided in which the framework was used to analyze the propagation of error from
characterization through to a simple finite element model based on the characterized
microstructure. For the particular study here, it was shown that sample size was
more critical to accurate evaluation of mean elastic stress in each grain than either
resolution or integration volume. Such conclusions provide input as to the most
efficient data collection parameters that will lead to accurate results from the
associated computational models.
A natural extension of the framework presented here will be more formal
optimization of data collection parameters, balancing cost, and accuracy. Defining
an objective function based on costs, subject to the constraints of acceptable
error levels, such an optimization is feasible. Furthermore, the framework can be
extended to more detailed physically based characterization parameters and/or to
more challenging computational models, such as those that attempt to predict the
onset and growth of fatigue cracks.
References
1. R.T. DeHoff, Quantitative serial sectioning analysis: preview. J. Microsc. 131(3), 259–263
(1983)
2. M. Uchic, Serial Sectioning Methods for Generating 3D Characterization Data of Grain- and
Precipitate-Scale Microstructures (Springer, Boston, 2011), pp. 31–52
3. X. Lui, D. Furrer, J. Kosters, J. Holmes, Vision 2040: a roadmap for integrated, multiscale
modeling and simulation of materials and systems. Technical report, NASA Glenn Research
Center (2018)
4. M. Groeber, Digital Representation of Materials Grain Structure (Springer, Boston, 2011),
pp. 53–97
5. T. Sakamoto, Z. Cheng, M. Takahashi, M.Owari, Y. Nihei, Development of an ion and electron
dual focused beam apparatus for three-dimensional microanalysis. Jpn. J. Appl. Phys. 37(4R),
2051 (1998)
6. M. Uchic, M. Groeber, D. Dimiduk, J.P. Simmons, 3D microstructural characterization of
nickel superalloys via serial-sectioning using a dual beam fib-sem. Scr. Mater. 55(1), 23–28
(2006). Viewpoint set no. 41 “3D Characterization and Analysis of Materials” Organized by
G. Spanos
N. Wade et al.
parameters all the way through to computational models. The approach can easily
introduce new modules and/or additional complexity as needed.
We show how the framework can provide detailed insight into the influence of
data collection parameters, such as resolution, electron beam energy, and dwell
time. The resolution study showed that increasing sample point spacing results in
an approximately linear increase in mismatched volume, and an analytical model
was developed that provides basic insights into this behavior. Additional studies
showed how changes to the interaction volume (i.e., electron beam energy) or
increased levels of random noise (i.e., shorter dwell times) also affect error. It was
also demonstrated that through proper data processing parameters, much of this
error can be mitigated; however, an improperly applied erode/dilate data processing
filter can increase error for certain microstructural types. Finally, an example was
provided in which the framework was used to analyze the propagation of error from
characterization through to a simple finite element model based on the characterized
microstructure. For the particular study here, it was shown that sample size was
more critical to accurate evaluation of mean elastic stress in each grain than either
resolution or integration volume. Such conclusions provide input as to the most
efficient data collection parameters that will lead to accurate results from the
associated computational models.
A natural extension of the framework presented here will be more formal
optimization of data collection parameters, balancing cost, and accuracy. Defining
an objective function based on costs, subject to the constraints of acceptable
error levels, such an optimization is feasible. Furthermore, the framework can be
extended to more detailed physically based characterization parameters and/or to
more challenging computational models, such as those that attempt to predict the
onset and growth of fatigue cracks.
References
1. R.T. DeHoff, Quantitative serial sectioning analysis: preview. J. Microsc. 131(3), 259–263
(1983)
2. M. Uchic, Serial Sectioning Methods for Generating 3D Characterization Data of Grain- and
Precipitate-Scale Microstructures (Springer, Boston, 2011), pp. 31–52
3. X. Lui, D. Furrer, J. Kosters, J. Holmes, Vision 2040: a roadmap for integrated, multiscale
modeling and simulation of materials and systems. Technical report, NASA Glenn Research
Center (2018)
4. M. Groeber, Digital Representation of Materials Grain Structure (Springer, Boston, 2011),
pp. 53–97
5. T. Sakamoto, Z. Cheng, M. Takahashi, M.Owari, Y. Nihei, Development of an ion and electron
dual focused beam apparatus for three-dimensional microanalysis. Jpn. J. Appl. Phys. 37(4R),
2051 (1998)
6. M. Uchic, M. Groeber, D. Dimiduk, J.P. Simmons, 3D microstructural characterization of
nickel superalloys via serial-sectioning using a dual beam fib-sem. Scr. Mater. 55(1), 23–28
(2006). Viewpoint set no. 41 “3D Characterization and Analysis of Materials” Organized by
G. Spanos