arXiv (Cornell University),
Год журнала:
2023,
Номер
unknown
Опубликована: Янв. 1, 2023
Generalisability
and
the
consistency
of
a
posteriori
results
are
most
critical
points
view
regarding
data-driven
turbulence
models.
This
study
presents
progressive
improvement
models
using
simulation-driven
surrogate
optimisation
based
on
Kriging.
We
aim
for
augmentation
secondary-flow
reconstruction
capability
in
linear
eddy-viscosity
model
without
violating
its
original
performance
canonical
cases
e.g.
channel
flow.
Explicit
algebraic
Reynolds
stress
correction
(EARSCMs)
$k-\omega$
SST
obtained
to
predict
secondary
flow
which
standard
fails
capture.
The
is
achieved
by
multi-objective
approach
duct
quantities,
numerical
verification
developed
performed
various
test
cases.
testing
new
guarantee
that
preserve
model.
Regarding
generalisability
models,
unseen
demonstrate
significant
prediction
flows
streamwise
velocity.
These
highlight
potential
enhance
fluid
simulation
while
preserving
robustness
stability
solver.
Metrology,
Год журнала:
2025,
Номер
5(1), С. 11 - 11
Опубликована: Фев. 5, 2025
Ultrasonic
flow
meters
(UFMs)
by
transit
time
are
ubiquitous
in
industrial
applications,
mainly
for
their
versatility
and
practicality.
They
widely
used
gas
liquid
installations,
such
as
the
oil
industry
or
feedwater
systems
nuclear
power
plants.
Computational
fluid
dynamics
(CFD)
techniques
can
be
a
tool
to
potentially
improve
ultrasonic
measurements.
CFD
may
contribute
predicting
velocity
profile
factor
disturbed
flows,
integrating
acoustic
ray,
improving
calibration
of
UFMs,
assisting
design
optimization.
This
communication
presents
working
principle
UFM,
discusses
how
support
improvements,
shows
relevant
trending
fields
that
deserve
further
investigation
promote
significance
on
this
subject.
This
paper
focuses
on
the
use
of
reinforcement
learning
(RL)
as
a
machine-learning
(ML)
modeling
tool
for
near-wall
turbulence.
RL
has
demonstrated
its
effectiveness
in
solving
high-dimensional
problems,
especially
domains
such
games.
Despite
potential,
is
still
not
widely
used
turbulence
and
primarily
flow
control
optimization
purposes.
A
new
wall
model
(WM)
called
VYBA23
developed
this
work,
which
uses
agents
dispersed
near
wall.
The
trained
single
Reynolds
number
(Reτ=104)
does
rely
high-fidelity
data,
backpropagation
process
based
reward
rather
than
an
output
error.
states
RLWM,
are
representation
environment
by
agents,
normalized
to
remove
dependence
number.
tested
compared
another
RLWM
(BK22)
equilibrium
model,
half-channel
at
eleven
different
numbers
{Reτ∈[180;1010]}.
effects
varying
agents'
parameters,
actions
range,
time
step,
spacing,
also
studied.
results
promising,
showing
little
effect
average
field
but
some
wall-shear
stress
fluctuations
velocity
fluctuations.
work
offers
positive
prospects
developing
RLWMs
that
can
recover
physical
laws
extending
type
ML
models
more
complex
flows
future.
International Journal of Heat and Fluid Flow,
Год журнала:
2023,
Номер
104, С. 109242 - 109242
Опубликована: Ноя. 7, 2023
Generalisability
and
the
consistency
of
a
posteriori
results
are
most
critical
points
view
regarding
data-driven
turbulence
models.
This
study
presents
progressive
improvement
models
using
simulation-driven
Bayesian
optimisation
with
Kriging
surrogates
where
is
achieved
by
multi-objective
approach
based
on
duct
flow
quantities.
We
aim
for
augmentation
secondary-flow
prediction
capability
in
linear
eddy-viscosity
model
k−ω
SST
without
violating
its
original
performance
canonical
cases
e.g.
channel
flow.
Progressively
data-augmented
explicit
algebraic
Reynolds
stress
(PDA-EARSMs)
obtained
enabling
secondary
flows
that
standard
fails
to
predict.
The
new
tested
guaranteeing
they
preserve
successful
model.
Subsequently,
numerical
verification
performed
various
test
cases.
Regarding
generalisability
models,
unseen
demonstrate
significant
streamwise
velocity.
These
highlight
potential
enhance
fluid
simulation
while
preserving
robustness
stability
solver.
This
study
employs
wall-resolving
large
eddy
simulations
(LES)
to
investigate
whether
and
how
laminarization
occurs
in
turbulent
pipe
flows
through
annular
gap
fluid
injection.
The
investigations
focus
on
the
effects
of
injection
velocity
ratio
(Uratio)
inflow
bulk
Reynolds
number
(ReB)
flow
dynamics,
examining
key
characteristics
such
as
mean
profiles,
stresses,
production,
friction
factors.
Sustained
laminar
was
observed
at
ReB=3700
for
Uratio
values
between
1.8
2.6
over
a
length
100D,
achieving
significant
45%
reduction
factor
Uratio=2.6.
is
attributed
suppression
turbulence
regeneration
cycle,
evidenced
by
pronounced
downstream
peak
streamwise
fluctuations,
shear
stress,
production.
Conversely,
ReB=5500,
sustained
unattainable
any
tested
Uratio,
persisted
due
suboptimal
conditions
or
adverse
effects.
Nonetheless,
these
cases
with
non-laminarizing
still
exhibited
reductions
coefficient—up
30%,
ReB=5500
Uratio=1.6—over
certain
before
ultimately
reverting
fully
state
further
downstream.
Our
LES
results
demonstrate
good
qualitative
agreement
experimental
findings
Kühnen
et
al.
(Flow
Turbul.
Combustion,
100(4),
2018,
pp.
919–943),
effectively
capturing
dynamics
regeneration.
Physics of Fluids,
Год журнала:
2023,
Номер
35(12)
Опубликована: Дек. 1, 2023
In
the
field
of
data-driven
turbulence
modeling,
consistency
a
posteriori
results
and
generalizability
are
most
critical
aspects
new
models.
this
study,
we
combine
multi-case
surrogate
optimization
technique
with
progressive
augmentation
approach
to
enhance
performance
popular
k−ω
shear
stress
transport
(SST)
model
in
prediction
flow
separation.
We
introduce
separation
factor
into
equation
turbulent
specific
dissipation
rate
(ω)
correct
underestimation
viscosity
by
SST
case
for
two-dimensional
cases.
The
is
optimized
based
on
their
training
cases
including
periodic
hills
curved
backward-facing
step
flow.
Simulation
channel
likewise
included
process
guarantee
that
original
preserved
absence
verified
multiple
unseen
different
Reynolds
numbers
geometries.
Results
show
significant
improvement
recirculation
zone,
velocity
components,
distribution
friction
coefficient
both
testing
cases,
where
expected.
models
test
no
shows
they
preserve
successful
when
not
Flow Measurement and Instrumentation,
Год журнала:
2024,
Номер
96, С. 102543 - 102543
Опубликована: Янв. 20, 2024
The
enhancement
of
two-stand
ultrasonic
flow
meters
relies
upon
obtaining
a
precise
understanding
and
prediction
their
complex
physics
throughout
entire
dynamic
range
operation.
This
study
provides
comprehensive
numerical
experimental
investigation
the
typical
meter
by
industry
standards.
Predictions
based
on
computational
fluid
dynamics
simulations
are
employed
to
obtain
results,
which
validated
through
experiments
laser
Doppler
velocimetry
static
pressure
drop.
Results
indicate
that
no
qualitative
changes
occur
beyond
an
inflow
Reynolds
number
104
in
terms
coherent
structures
dynamics.
Analysis
distribution
across
cross-sections
reveals
stands
most
influential
areas
contributing
In
cases
with
turbulent
inflow,
there
is
noticeable
recovery
following
significant
gradients
stands,
while
such
absent
scenarios
laminar
inflow.
Both
approaches
yield
excellent
agreement
outcomes,
accurately
estimating
axial
velocity
within
meter's
measurement
volume
drop
it,
deviations
uncertainty
ranges
1
3
standard
respectively.
developed
methodology
demonstrates
its
potential
evaluate
internal-flow
systems
similar
features
ranges.
for
wide
operation
shown
detail
both
regimes,
displaying
rolling
vortices,
detached
flow,
recirculation
zones.
Flow Measurement and Instrumentation,
Год журнала:
2023,
Номер
94, С. 102469 - 102469
Опубликована: Окт. 15, 2023
Commercial
flowmeters
are
often
costly
and
complex
but
crucial
to
mapping
flows
calculating
contamination
flux.
Sampling
is
an
easy
method
for
determining
concentration,
without
flow
rates,
it
impossible
find
total
loading.
This
work
devised
a
simplified
circuit
innovative
3D
prints
produce
low-cost
meter.
Interchangeable
impellers
compensate
the
electronics'
limitations
allowing
device
be
highly
accurate
widely
applicable.
The
opensource
impeller/propeller
design
software
OpenProp
was
used
build
several
variations
different
conditions.
enabled
optimisation
of
design.
blades
tested
here
in
open
channel
(though
impeller
designed
fit
inside
2"
pipe).
paper
also
describes
optimise
both
physical
electrical
flowmeter.
Characteristics
including
stability,
sensitivity
accuracy
were
studied.
flowmeter
River
Eden,
Cumbria,
UK.
robust
reproduce,
opensource,
community
educational
groups.