|
Micro-computed tomography based experimental
investigation of micro- and macro-mechanical response
of
particulate composites with void growth
K. Ramos and K. Matous
Center for Shock Wave-processing of Advanced Reactive
Materials
Department of Aerospace and Mechanical
Engineering, University of Notre Dame, Notre Dame,
Indiana 46556, USA.
Abstract
In this work, we
develop an image-based material testing approach using
micro-computed tomography to understand the influence
of microstructure and local damage phenomenon on the
effective mechanical response of rubber-glass bead
composites. Furthermore, a nondestructive,
three-dimensional image-based analysis protocol which
provides high fidelity of sample testing and data
assessment has been established. An investigation was
performed on various compositions of silicone rubber
reinforced with silica particles. In situ
compression experiments were used to study how the
microscale damage (void creation from debonding)
develops and evolves in the context of four primary
studies: (i) effect of particle volume fraction, (ii)
effect of particle diameter, (iii) local damage
phenomena and its evolution (incremental
loading/unloading), and (iv) effect of surface
treatments on bonding characteristics. A detailed
statistical analysis of the evolution of structural
features through robust image processing strategies at
various stages of loading was conducted. The rich data
analysis collected from the experimental studies
offers an understanding of the complex phenomena
attributing to the material's macro and microscopic
response to loading. Altogether, this framework
results in the development of
microstructure-statistics-property relations.
Furthermore, the mechanical and morphological response
of non-linear viscoelastic materials subjected to
uniaxial compression is investigated.
Conclusions
In this work, we present an
image-based materials testing approach using
microcomputed tomography to observe void formation and
its evolution while varying the microstructure of
rubber-glass bead composites. A material testing
methodology is developed resulting in precise and
reliable force-displacement data using the in situ
material testing stage.
Visualization and analysis software is
used to determine multiscale coupling between the
morphological and mechanical response of the
composites when subjected to uniaxial compression
loading. Microstructure-statistics-property relations
are established. We test multiple samples to ensure
repeatability and use large data sets for the
statistical calculations representing the amount of
microscale damage. The developed image processing
pipeline accurately identifies particles and voids
which are critical for material response assessment.
The investigation of varied
microstructure's effects on microscale damage can be
summarised as follows. When varying the particle
volume fraction, the void growth saturates with higher
filled composites despite having different material
bulk responses to loading. When varying the particle
diameter a non-monotonic like trend in void content is
observed, a result of chain-like clusters that reduce
stress concentrations in the largest particle diameter
composite. When studying void evolution, highly
nonlinear void growth is observed with each
incremental loading step. Lastly, comparisons of
strong and weak bonding agents used for particle
matrix interfacial bonding results in 5x increases in
void sizes.
In addition to the fundamental
understanding, these experiments serve as validation
data sets for multiscale modelling approaches. The
constitutive model development, validation, and
co-designed simulations and experiments are a natural
next step. This analysis and framework yields new
insight of particulate composite materials required
for optimal material design. Furthermore, the novel,
thorough, and reliable experimental and image
processing pipelines are applicable to other material
systems.
Acknowledgment
We gratefully acknowledge Dr. A. Gillman for his numerous
contributions to the development of the experimental
procedures. This work was supported by the Department of
Energy, National Nuclear Security Administration, under
the award number DE-NA0002377 as part of the Predictive
Science Academic Alliance Program II.
Download
paper here
|