Origami
structures
have
the
advantages
of
foldability
and
adjustability,
with
applications
spanning
numerous
engineering
fields.
However,
there
remains
a
dearth
intelligent
convenient
methods
that
can
effectively
tackle
both
potential
energy
prediction
design
problems
on
origami
structures.
This
study
proposes
novel
physics-informed
neural
network
(PINN)
for
predicting
performing
A
sorting
operation
is
developed
PINN
to
address
challenge
model
converging
local
optima.
Given
boundness
variables,
constraints
them
are
enforced
during
process.
Two
loss
functions
physical
connotation
customized
problems,
respectively.
The
accuracy
curves
predicted
by
demonstrated
through
comparison
reference
exhaustive
method.
Furthermore,
two
cases
Kresling
structures,
matching
target
curve
set
points,
performed
show
applicability
in
inverse
problems.
presented
physics-driven
approach
without
labelled
data
offers
an
innovative
tool
learning
ability
predict
In
addition,
code
shared
online.
Journal of Applied Mechanics,
Год журнала:
2024,
Номер
92(1)
Опубликована: Ноя. 1, 2024
Abstract
Deployable
structures
are
extensively
used
in
engineering.
A
bistable
panel
structure,
inspired
by
multistable
origami,
is
proposed,
capable
of
deployment
and
folding
powered
air
pressure.
Prototypes
were
manufactured
using
planar
laser
etching
technology
based
on
geometric
design.
Mechanical
behavior
under
out-of-plane
compression,
in-plane
bending
loads
was
analyzed
through
experiments.
The
foldable
showed
superior
mechanical
performance
highlighting
its
potential
as
an
ideal
energy-absorbing
material.
In-plane
compression
along
the
direction
exhibited
lower
strength
due
to
foldability,
with
failure
modes
involving
rigidity
loss
from
folding.
structure
demonstrated
good
energy
absorption
characteristics
during
compression.
As
angle
unit
increased
bending,
improved,
but
mode
shifted
fracture.
In
perpendicular
direction,
enhanced,
failed
Origami
structures
have
the
advantages
of
foldability
and
adjustability,
with
applications
spanning
numerous
engineering
fields.
However,
there
remains
a
dearth
intelligent
convenient
methods
that
can
effectively
tackle
both
potential
energy
prediction
design
problems
on
origami
structures.
This
study
proposes
novel
physics-informed
neural
network
(PINN)
for
predicting
performing
A
sorting
operation
is
developed
PINN
to
address
challenge
model
converging
local
optima.
Given
boundness
variables,
constraints
them
are
enforced
during
process.
Two
loss
functions
physical
connotation
customized
problems,
respectively.
The
accuracy
curves
predicted
by
demonstrated
through
comparison
reference
exhaustive
method.
Furthermore,
two
cases
Kresling
structures,
matching
target
curve
set
points,
performed
show
applicability
in
inverse
problems.
presented
physics-driven
approach
without
labelled
data
offers
an
innovative
tool
learning
ability
predict
In
addition,
code
shared
online.
