264
D. Patel et al.
elastic stiffness constants and the initial slip resistance parameters obtained using the
inverse solution approach were within 5% of the typical values reported in literature
for Fe-3%Si.
4 Challenges
The availability of data and the use of data-driven protocols allows us to objectively
quantify the uncertainty associated with the information gathered and knowledge
gained. The quantity of experimental data generated by a single material research
group is typically relatively small and may not be enough to tackle the engineering
material’s design problems.
This framework has been demonstrated on a variety of problems. Structureproperty relations were developed capable of the prediction of the extreme value
elastic stresses hypothesized to be associated with transverse crack formation at the
scale of the filaments in ceramic matrix composites. The elastic and inelastic bulk
properties of titanium polycrystals were predicted, including the elastic strain fields
in cubic and hexagonal polycrystals and high-cycle fatigue predictions of fatigue
indicator parameters (FIPs). The transport properties of porous microstructures used
in fuel cell applications were linked to the microstructural features of polymers.
Additionally, the framework has been applied to inverse design problems to extract
single crystal elastic-plastic properties of a polycrystalline metal.
The practical realization of the framework is only feasible with the accumulation
of substantially large libraries of materials data that capture the relevant multiscale
spatiotemporal information about materials internal structures for a very broad
class of materials. It is practically impossible for a single research group or an
organization to take on the monumental task given the information required. Further,
the domain expertise needed to mine and curate the materials knowledge needed to
accelerate the material development of new and improved materials lie well outside
the traditional skill sets of materials scientists and engineers. Currently, the materials
community does not have the skill set needed to fully enable the materials data
science revolution. Hence, an intimate collaboration of teams with specific domain
knowledge are needed to facilitate the ICME approach.
5 Summary
A novel workflow template is presented to extract process-structure linkages in
microstructure evolution problems through the utilization of advanced data science
techniques. The presented workflow is scalable, expandable, and can be applied
to a broad variety of microstructure evolution datasets. This workflow consists of
four modular steps: (1) data preprocessing, (2) microstructure quantification, (3)
dimensionality reduction, and (4) extraction and validation of process-structure
D. Patel et al.
elastic stiffness constants and the initial slip resistance parameters obtained using the
inverse solution approach were within 5% of the typical values reported in literature
for Fe-3%Si.
4 Challenges
The availability of data and the use of data-driven protocols allows us to objectively
quantify the uncertainty associated with the information gathered and knowledge
gained. The quantity of experimental data generated by a single material research
group is typically relatively small and may not be enough to tackle the engineering
material’s design problems.
This framework has been demonstrated on a variety of problems. Structureproperty relations were developed capable of the prediction of the extreme value
elastic stresses hypothesized to be associated with transverse crack formation at the
scale of the filaments in ceramic matrix composites. The elastic and inelastic bulk
properties of titanium polycrystals were predicted, including the elastic strain fields
in cubic and hexagonal polycrystals and high-cycle fatigue predictions of fatigue
indicator parameters (FIPs). The transport properties of porous microstructures used
in fuel cell applications were linked to the microstructural features of polymers.
Additionally, the framework has been applied to inverse design problems to extract
single crystal elastic-plastic properties of a polycrystalline metal.
The practical realization of the framework is only feasible with the accumulation
of substantially large libraries of materials data that capture the relevant multiscale
spatiotemporal information about materials internal structures for a very broad
class of materials. It is practically impossible for a single research group or an
organization to take on the monumental task given the information required. Further,
the domain expertise needed to mine and curate the materials knowledge needed to
accelerate the material development of new and improved materials lie well outside
the traditional skill sets of materials scientists and engineers. Currently, the materials
community does not have the skill set needed to fully enable the materials data
science revolution. Hence, an intimate collaboration of teams with specific domain
knowledge are needed to facilitate the ICME approach.
5 Summary
A novel workflow template is presented to extract process-structure linkages in
microstructure evolution problems through the utilization of advanced data science
techniques. The presented workflow is scalable, expandable, and can be applied
to a broad variety of microstructure evolution datasets. This workflow consists of
four modular steps: (1) data preprocessing, (2) microstructure quantification, (3)
dimensionality reduction, and (4) extraction and validation of process-structure