Scientific Reports,
Год журнала:
2024,
Номер
14(1)
Опубликована: Авг. 29, 2024
The
compressive
strength
of
concrete
depends
on
various
factors.
Since
these
parameters
can
be
in
a
relatively
wide
range,
it
is
difficult
for
predicting
the
behavior
concrete.
Therefore,
to
solve
this
problem,
an
advanced
modeling
needed.
aim
literature
achieve
ideal
and
flexible
solution
necessary
develop
new
approaches.
Artificial
Neural
Networks
(ANNs)
have
evolved
from
theoretical
method
widely
utilized
technology
by
successful
applications
variety
issues.
Actually,
ANNs
are
strong
computing
tool
that
provides
right
solutions
problems
use
conventional
methods.
Inspired
biological
neural
system,
networks
now
used
solving
range
complicated
civil
engineering.
This
study''s
target
evaluating
performance
developed
African
vulture
optimization
algorithm
(DAVOA)-Elman
(ENNs)
considering
different
input
self-compacting
strength.
Hence,
once
8
again
get
as
close
possible
prediction
conditions
laboratory,
140
entered
improved
version
Elman
input.
According
results,
element
network
has
lowest
mean
squares
test
error
7
28
days
100
repetitions.
Further,
both
strengths,
grid
with
Logsig-Purelin
interlayer
transfer
function
error,
which
determines
optimal
function.
Moreover,
results
showed
DAVOA
reliable
time
cost
savings
high
power
desired
characteristics.
Also,
7-day
28-day
strength,
built
74.54
70.44%
improvement
over
8-parameter
networks,
respectively,
directly
affects
effect.
Further
considered
rate
properties.
REVIEWS ON ADVANCED MATERIALS SCIENCE,
Год журнала:
2024,
Номер
63(1)
Опубликована: Янв. 1, 2024
Abstract
The
degradation
of
concrete
structures
is
significantly
influenced
by
water
penetration
since
serves
as
the
primary
vehicle
for
movement
harmful
compounds.
process
capillary
absorption
widely
recognized
a
crucial
indicator
durability
unsaturated
concrete,
it
allows
dangerous
substances
to
enter
composite
material.
capacity
intricately
linked
its
pore
structure,
inherently
porous.
main
goal
this
work
create
an
innovative
predictive
tool
that
assesses
porosity
analyzing
components
using
machine-learning
(ML)
framework.
Seven
distinct
batch
design
variables
were
included
in
generated
database:
fly
ash,
superplasticizer,
water-to-binder
ratio,
curing
time,
ground
granulated
blast
furnace
slag,
binder,
and
coarse-to-fine
aggregate
ratio.
Four
distant
ML
algorithms,
including
AdaBoost,
linear
regression
(LR),
decision
tree
(DT),
support
vector
machine
(SVM),
are
utilized
infer
generalization
capabilities
algorithms
estimate
accurately.
RReliefF
algorithm
was
implemented
calculate
significant
features
influencing
porosity.
This
study
concludes
comparison
alternative
techniques,
AdaBoost
method
demonstrated
superior
performance
with
R
2
score
0.914,
followed
SVM
(0.870),
DT
(0.838),
LR
(0.763).
results
evaluation
indicated
binder
possesses
remarkable
influence
on
concrete.
Buildings,
Год журнала:
2024,
Номер
14(11), С. 3419 - 3419
Опубликована: Окт. 27, 2024
In
recent
years,
various
aspects
of
research
related
to
pervious
concrete
(PC)
have
progressed
rapidly,
and
it
is
necessary
summarise
generalise
the
latest
results.
This
paper
reviews
compares
raw
materials
concrete,
examining
elements
such
as
porosity,
permeability,
mechanical
properties,
durability.
According
comparisons,
we
put
forward
an
ideal
aggregate
model
with
Uneven
Surface,
which
may
reinforce
properties.
By
summarising
important
issues
aggregate,
particle
size,
water–cement
ratio,
additives
admixtures,
mixing
ratio
design,
moulding,
other
factors
that
affect
new
design
methods
are
proposed.
A
effective
stress
based
on
continuous
porosity
Terzaghi
developed
fit
principle
better.
Finally,
by
frontiers
key
need
be
addressed
in
future
scientific
raised.
