International Journal of Mechanical Sciences,
Journal Year:
2024,
Volume and Issue:
282, P. 109551 - 109551
Published: July 25, 2024
Capturing
and
predicting
the
effective
mechanical
properties
of
highly
porous
cellular
media
still
represents
a
significant
challenge
for
research
community,
due
to
their
complex
structural
interdependencies
known
size
effects.
Micromorphic
theories
are
often
applied
in
this
context
model
inelastic
deformation
behavior
foam-like
structures,
particular
incorporate
such
effect
into
investigation
structure–property
correlations.
This
raises
problems
formulating
appropriate
constitutive
relations
numerous
non-classical
stress
measures
determining
corresponding
material
parameters,
which
usually
difficult
assess
experimentally.
The
present
contribution
therefore
alternatively
employs
hierarchical
micromorphic
multi-scale
approach
within
direct
FE2
framework
simulate
irreversible
solids.
predictions
Cosserat
(micropolar)
fully-micromorphic
theory
compared
with
conventional
results
numerical
simulations
(DNS)
loading
scenarios
elastic,
elastic–plastic,
creep
deformations.
Therein,
modes
microstructure
resulting
from
introduced
kinematics
visualized,
as
macroscopic
hyperstresses
Applied Mechanics Reviews,
Journal Year:
2024,
Volume and Issue:
unknown, P. 1 - 51
Published: Aug. 2, 2024
Abstract
In
the
framework
of
solid
mechanics,
task
deriving
material
parameters
from
experimental
data
has
recently
re-emerged
with
progress
in
full-field
measurement
capabilities
and
renewed
advances
machine
learning.
this
context,
new
methods
such
as
virtual
fields
method
physics-informed
neural
networks
have
been
developed
alternatives
to
already
established
least-squares
finite
element-based
approaches.
Moreover,
model
discovery
problems
are
emerging
can
also
be
addressed
a
parameter
estimation
framework.
These
developments
call
for
unified
perspective,
which
is
able
cover
both
traditional
novel
approaches
state
variables
or
structure
itself
inferred
well.
Adopting
concepts
discussed
inverse
community,
we
distinguish
between
all-at-once
reduced
With
general
framework,
large
portion
literature
on
computational
mechanics
--
identify
combinations
that
not
yet
addressed,
two
proposed
paper.
We
discuss
statistical
quantify
uncertainty
related
estimated
parameters,
propose
two-step
procedure
identification
complex
models
based
frequentist
Bayesian
principles.
Finally,
illustrate
compare
several
aforementioned
mechanical
benchmarks
synthetic
data.
Computer Methods in Applied Mechanics and Engineering,
Journal Year:
2024,
Volume and Issue:
430, P. 117208 - 117208
Published: July 11, 2024
Applications
of
soft
materials
are
customarily
linked
to
complex
deformation
scenarios
and
material
nonlinearities.
In
the
bioengineering
field,
typically
mimic
low
stiffness
biological
matter
subjected
extreme
deformations.
Computational
frameworks
surge
as
a
versatile
tool
assist
design
functional
applications.
The
constitutive
model
lies
at
core
such
frameworks.
this
regard,
customary
non-linear
behavior
elastomers
poses
an
additional
challenge
thoroughly
capture
behavior.
Here,
data-driven
methodologies
hold
considerable
promise
for
enhancing
modeling
when
contrasted
with
phenomenological
approaches.
investigation,
we
introduce
data-adaptive
method
tailored
hyperelastic
finite
strains.
Specifically,
our
substitutes
priori
chosen
strain
energy
function
by
flexible
interpolant
defined
on
discretized
invariant
space.
Within
framework,
interpolation
values
assume
role
parameters
determined
through
element
updating
conform
measured
experimental
data
—
comprising
full-field
displacements
coming
from
Digital-Image-Correlation
global
reaction
forces.
We
validate
uniaxial
tests
elastomers,
encompassing
ELASTOSILTM,
DOWSILTM,
V
HBTM.
Overall,
aim
establish
new
route
construction
functions,
untethered
any
predefined
existing
models
or
assumptions
regarding
shape
energy.
Mechanics of Advanced Materials and Structures,
Journal Year:
2024,
Volume and Issue:
unknown, P. 1 - 19
Published: Dec. 17, 2024
Machine
learning
(ML)
models
are
widely
used
across
numerous
scientific
and
engineering
disciplines
due
to
their
exceptional
performance,
flexibility,
prediction
quality,
ability
handle
highly
complex
problems
if
appropriate
data
available.
One
example
of
such
areas
which
has
attracted
a
lot
attentions
in
the
last
couple
years
is
integration
data-driven
approaches
material
modeling.
There
been
several
successful
researches
implementing
ML-based
constitutive
instead
classical
phenomenological
for
various
materials,
particularly
those
with
non-linear
mechanical
behaviors.
This
review
paper
aims
systematically
investigate
literature
on
materials
classify
these
based
suitability
non-linearity
including
Non-linear
elasticity
(hyperelasticity),
plasticity,
visco-elasticity,
visco-plasticity.
Furthermore,
we
also
reviewed
compared
that
have
applied
architectured
as
groups
designed
represent
specific
behaviors
might
not
exist
conventional
categories.
The
other
goal
this
provide
initial
steps
understanding
modeling,
artificial
neural
networks
(ANN),
Gaussian
processes,
random
forests
(RF),
generated
adversarial
(GANs),
support
vector
machines
(SVM),
different
regression
physics-informed
(PINN).
outlines
collection
methods,
types
data,
processing
approaches,
theoretical
background
ML
models,
advantage
limitations
potential
future
research
directions.
comprehensive
will
researchers
knowledge
necessary
develop
high-fidelity,
robust,
adaptable,
flexible,
accurate
advanced
materials.
International Journal for Numerical and Analytical Methods in Geomechanics,
Journal Year:
2025,
Volume and Issue:
unknown
Published: Jan. 15, 2025
ABSTRACT
Most
of
the
robust
artificial
intelligence
(AI)‐based
constitutive
models
are
developed
with
synthetic
datasets
generated
from
traditional
models.
Therefore,
they
fundamentally
rely
on
rather
than
laboratory
test
results.
Also,
their
potential
use
within
geotechnical
engineering
communities
is
limited
due
to
unavailability
along
model
code
files.
In
this
study,
data‐driven
using
only
databases
and
deep
learning
(DL)
techniques.
The
database
was
prepared
by
conducting
cyclic
direct
simple
shear
(CDSS)
tests
reconstituted
sand,
that
is,
PDX
sand.
stacked
long
short‐term
memory
(LSTM)
network
its
variants
considered
for
developing
predictive
strain
(
γ
[%])
excess
pore
pressure
ratio
r
u
)
time
histories.
suitable
input
parameters
(IPs)
selected
based
physics
behind
generation
(%)
liquefiable
sands.
predicted
responses
agree
well
in
most
cases
used
predict
dynamic
soil
properties
same
modeling
framework
extended
other
sand
compared
existing
AI‐based
verify
practical
applicability.
summary,
it
observed
though
trained
histories
reasonably
well;
however,
struggled
hysteresis
loops
at
higher
cycles.
more
research
needed
enhance
predictability
future
before
them
practice
simulating
response.
