Local stress/strain field analysis of die-casting Al alloys via 3D model simulation with realistic defect distribution and RVE modelling
Engineering Failure Analysis,
Год журнала:
2024,
Номер
unknown, С. 109104 - 109104
Опубликована: Ноя. 1, 2024
Язык: Английский
Elongation Prediction of Die-Cast Aluminum Alloy Based on 3D Convolutional Neural Network Model
SAE International Journal of Materials and Manufacturing,
Год журнала:
2025,
Номер
18(4)
Опубликована: Апрель 9, 2025
<div>This
study
aims
to
predict
the
impact
of
porosities
on
variability
elongation
in
casting
Al-10Si-0.3Mg
alloy
using
machine
learning
methods.
Based
dataset
provided
by
finite
element
method
(FEM)
modeling,
two
algorithms
including
artificial
neural
network
(ANN)
and
3D
convolutional
(3D
CNN)
were
trained
compared
determine
optimal
model.
The
results
showed
that
mean
squared
error
(MSE)
determination
coefficient
(R<sup>2</sup>)
CNN
validation
set
0.01258/0.80,
while
those
ANN
model
0.28951/0.46.
After
obtaining
prediction
model,
was
used
experimental
specimens.
values
obtained
experiments
FEM
simulation
with
for
samples
smaller
than
9.5%,
both
accuracy
efficiency
surpassed
simulation.</div>
Язык: Английский
Study on the Influence of Injection Velocity on the Evolution of Hole Defects in Die-Cast Aluminum Alloy
Materials,
Год журнала:
2024,
Номер
17(20), С. 4990 - 4990
Опубликована: Окт. 12, 2024
Aluminum
alloy
die
casting
has
achieved
rapid
development
in
recent
years
and
been
widely
used
all
walks
of
life.
However,
due
to
its
high
pressure
high-speed
technological
characteristics,
avoiding
hole
defects
become
a
problem
great
significance
aluminum
production.
In
this
paper,
the
filling
visualization
dynamic
characterization
experiment
was
innovatively
developed,
which
can
directly
study
analyze
influence
different
injection
rates
on
formation
evolution
flow
patterns
gas-induced
defects.
As
speed
increased
from
1.0
m/s
1.5
m/s,
average
porosity
7.49%
9.57%,
marking
an
increase
number
size
pores.
According
comparison
with
Anycasting,
simulation
results
show
that
liquid
metal
when
front
vs.
previous
rate
caused
fractures
at
same
distance.
Therefore,
degree
broken
splash
is
more
serious.
Combined
analysis
transport
mechanics,
fracturing
wall-attached
jet
effect
process.
It
difficult
for
adhere
type
wall
order
fuse
subsequent
form
cavity
With
velocity,
microgroup
volume
formed
via
breakage
decreases;
thus
air
entrapment
increases,
finally
leading
Язык: Английский