Waves in Random and Complex Media,
Год журнала:
2023,
Номер
unknown, С. 1 - 28
Опубликована: Фев. 6, 2023
AbstractThe
deep
neural
networks
(DNNs)
technique
is
one
of
the
best
computational
methods
with
low
error.
However,
due
to
their
superior
performance,
such
as
reducing
cost
complexity,
this
kind
machine
learning
has
attracted
much
attention.
Therefore,
in
present
paper,
for
first
time,
wave-dispersion
results
an
inhomogeneous
system
are
compared
outcomes
DNN
techniques
obtaining
exact
mean
squared
error
(MSE)
parameter
future
research.
In
mathematical
modeling
section
report,
improved
three-dimension
higher-order
theory
(3D-HSDT)
considering
effect
thickness
stretching
governing
equations
multi-phase
hybrid
nanocomposites
reinforced
rectangular
microplate
been
presented.
Phase-velocity
current
micro
structure
obtained
by
employing
modified
couple
stress
(3D-MCST).
This
non-classical
model
capable
capturing
small-size
impact
on
behavior
plates.
Also,
aid
COMSOL
multiphysics
finite
element
simulation,
verified,
and
new
improving
stability
presented
studied.
Finally,
show
that
software
can
be
verified
specific
MSE
parameter.KEYWORDS:
simulationDNN3D-MCSTmulti-phase
reinforcementwave
propagation
Disclosure
statementNo
potential
conflict
interest
was
reported
author(s).
Composite Structures,
Год журнала:
2023,
Номер
330, С. 117840 - 117840
Опубликована: Дек. 22, 2023
In
this
study,
a
comprehensive
study
on
the
vibration
behaviour
of
functionally
graded
thick
microplates
with
material
imperfections
is
presented
for
free
and
forced
vibrations
within
quasi-3D
model
modified
couple
stress
(MCS)
theory.
Axial,
transverse,
rotation,
stretching
motions
are
considered
modelling
microplate
in
framework
power-law
scheme
MCS
theories.
The
deformation
assumed
to
be
infinitesimal
modelled
using
linear
strain–displacement
relationships.
Using
virtual
work
method,
governing
motion
equations
derived.
For
fourfold
coupled
(axial-transverse-rotation-stretching)
characteristics,
partial
differential
components
discretised
via
trigonometric
expressions
corresponding
natural
frequencies
time
histories
(time-dependent
deflections)
determined
numerically.
methodology
initially
validated
through
comparative
analysis
between
macrolevel
simplified
structure
those
obtained
finite
element
software.
Additionally,
simulation
compared,
validation
purposes,
other
versions
from
literature.
Once
model's
validity
confirmed,
an
investigation
into
conducted
new
model.
results
demonstrate
considerable
effect
thickness
imperfection
characteristics
poroelastic
microplates.
Applied Sciences,
Год журнала:
2024,
Номер
14(6), С. 2233 - 2233
Опубликована: Март 7, 2024
Vortex-induced
vibration
(VIV)
of
long-span
bridges
can
be
large
amplitude,
which
influence
serviceability.
Therefore,
it
is
important
to
predict
the
response
vortex-induced
aid
management
bridges.
A
novel
data-driven
model
proposed
time
history
dynamic
VIV
events.
Specifically,
consists
gated
recurrent
unit
(GRU)
neural
networks
and
Newmark-beta
method.
GRU
perform
accurate
sequential
prediction,
method
complement
physical
meaning
middle
output
model.
To
prediction
amplitude
events,
employs
weighted
mean
square
error
as
loss
function,
put
more
emphasis
on
amplitude.
The
validated
measured
events
a
suspension
bridge.
absolute
percentage
Pearson
correlation
coefficient
trained
indicate
effectiveness
Case Studies in Construction Materials,
Год журнала:
2024,
Номер
unknown, С. e02823 - e02823
Опубликована: Янв. 1, 2024
This
study
addresses
burgeoning
environmental
concerns
amid
rising
global
coffee
consumption
by
exploring
the
viability
of
utilizing
waste
ground
ash
(SCG)
as
a
sustainable
fine
aggregate
in
concrete
production.
Significant
improvements
are
noted
investigating
impact
SCG
on
shear
strength
standard
and
enhancing
properties
through
activated
carbon
reinforcement
derived
from
via
physical
activation.
Experimental
results
reveal
that
incorporating
up
to
1.5%
weight
markedly
boosts
early
strength.
Moreover,
composites
with
small
inclusion
demonstrate
superior
power
during
curing
compared
conventional
mixes.
The
employs
Teaching-Learning
Optimization
(TLBO)
Extreme
Learning
Machine
(ELM)
techniques
forecast
accurately.
ELM
emerges
superior,
showcasing
heightened
accuracy,
lower
RMSE,
r
R2
values
TLBO.
Conclusively,
leveraging
not
only
presents
an
effective
method
for
but
also
underscores
promising
pathway
toward
management
reduced
impact,
reflecting
pivotal
stride
harmonizing
performance
enhancement
eco-friendly
practices.