Structural Concrete,
Год журнала:
2023,
Номер
24(5), С. 6761 - 6777
Опубликована: Июнь 1, 2023
Abstract
Strain‐hardening
cement‐based
composites
or
engineered
cementitious
(ECC)
is
concrete
produced
using
randomly
distributed
short
polymer
fibers.
It
very
ductile
compared
to
conventional
concrete.
Compressive
strength
(CS)
a
critical
property
used
as
quality
control
tool
evaluate
the
of
implemented
in
structural
provisions
and
mix
designs.
Accordingly,
save
cost
time
for
testing,
it
essential
provide
predictive
model
forecast
CS
mixtures
machine
learning
modeling
techniques.
In
this
study,
different
tools
are
propose
analytical
models
predict
ECC
mixtures,
such
linear
regression
(LR),
multi‐expression
programming
(MEP),
artificial
neural
network
(ANN),
Gaussian
process
(GPR).
A
total
210
data
were
collected
from
literature
train
test
developed
model.
The
fly
ash‐to‐cement
ratio
ranged
0
5.6,
water
binder
0.19
0.56,
superplasticizer
fiber
content,
curing
times
1
180
days.
Based
on
evaluation
models,
ANN
superior
other
with
high
coefficient
determination
(
R
2
),
root
mean
squared
error
(RMSE),
absolute
(MAE),
scatter
index
(SI).
sensitivity
analysis
input
parameters'
effect
prediction
indicates
that
forecasting
ECC's
CS.
Heliyon,
Год журнала:
2024,
Номер
10(4), С. e25997 - e25997
Опубликована: Фев. 1, 2024
Tire
rubber
waste
is
globally
accumulated
every
year.
Therefore,
a
solution
to
this
problem
should
be
found
since,
if
landfilled,
it
not
biodegradable
and
causes
environmental
issues.
One
of
the
most
effective
ways
recycling
those
wastes
or
using
them
as
replacement
for
normal
aggregate
in
concrete
mixture,
which
has
high
impact
resistance
toughness;
thus,
will
good
choice.
In
study,
135
data
were
collected
from
previous
literature
develop
model
prediction
rubberized
compressive
strength;
database
comprised
different
mixture
proportions,
maximum
size
(1-40
mm),
percentage
(0-100%)
replacing
natural
fine
coarse
aggregates
among
input
parameters
addition
cement
content
(380-500
kg/m
Engineering Structures,
Год журнала:
2024,
Номер
309, С. 118061 - 118061
Опубликована: Апрель 22, 2024
Cold-Formed
Steel
Lipped
(CFSL)
channels
are
susceptible
to
a
localized
failure
mechanism
known
as
web
crippling,
triggered
by
concentrated
loads
or
reactions
applied
the
of
section.
These
induce
buckling
and
distortion
in
web,
ultimately
leading
member's
collapse.
It
is
challenging
task
accurately
determine
crippling
capacity
CFSL
channel
due
its
complexity
various
influencing
factors.
This
paper
presents
hybrid
soft
computing
techniques
for
predicting
subjected
two
flange
load
cases.
The
developed
combine
Artificial
Neural
Networks
(ANN)
with
either
Genetic
Algorithms
(GA)
Particle
Swarm
Optimization
(PSO)
improve
computational
efficiency
accuracy.
finite
element
models
validated
experimental
results
then
employed
generate
database,
which
used
train
machine
learning
models,
including
ANN,
GA-ANN,
PSO-ANN.
Analysis
undertaken
on
reliability
existing
design
formulas
determining
channels.
shown
that
PSO-ANN
model
outperforms
other
terms
prediction
codes
not
reliable
estimating
However,
proposed
yields
good
correlation
analysis
results.
A
user-
friendly
graphical
interface
tool
practical
cold-formed
steel
lipped