Journal of Fluids Engineering,
Journal Year:
2019,
Volume and Issue:
141(6)
Published: March 5, 2019
Vertical
inline
pump
is
a
single-stage
single-suction
centrifugal
with
bent
pipe
before
the
impeller,
which
usually
used
where
installation
space
constraint.
In
this
paper,
three
objective
functions
of
efficiencies
at
0.5
Qd,
1.0
and
1.5
multi-objective
optimization
on
inlet
vertical
was
proposed
based
genetic
algorithm
artificial
neural
network
(ANN).
order
to
describe
shape
pipe,
fifth-order
third-order
Bezier
curves
were
adopted
fit
mid
curve
trend
parameters
cross
sections,
respectively.
Considering
real
computation
complexity,
11
variables
finally
in
optimization.
Latin
hypercube
sampling
(LHS)
generate
149
sample
cases,
solved
by
CFD
code
ANSYS
cfx
18.0.
The
calculation
results
design
utilized
train
ANNs,
these
surrogate
models
for
optimum
using
(MOGA).
showed
following:
(1)
There
great
agreement
between
numerical
experimental
results;
(2)
ANNs
could
accurately
variables.
maximum
deviations
predicted
values
computational
values,
1.94%,
2.35%,
0.40%;
(3)
has
influence
efficiency
part-load
conditions
while
slight
overload
condition;
(4)
Three
optimized
cases
selected
increase
Qd
4.96%,
2.45,
0.79%,
respectively;
(5)
velocity
distributions
outflow
more
uniform
than
original
one.
Micromachines,
Journal Year:
2022,
Volume and Issue:
13(8), P. 1208 - 1208
Published: July 29, 2022
In
this
study,
we
proposed
a
novel
approach
to
improve
centrifugal
pump
performance
with
regard
the
head,
efficiency,
and
power.
Firstly,
establish
constraints,
an
optimal
numerical
model
that
accounted
for
factors
such
as
efficiency
head
was
considered.
The
designed,
artificial
intelligence
algorithmic
applied
before
performing
experiments.
We
considered
set
of
models
by
selecting
parameters
casing
section
area,
interference
impeller,
volute
tongue
length,
angle.
weights
safety
displacement
on
optimization
indices
were
estimated.
matrix
process
less
than
38%
or
greater
62%.
This
guarantees
complicated
multi-objective
problem.
results
show
performances
improved.
Proceedings of the Institution of Mechanical Engineers Part C Journal of Mechanical Engineering Science,
Journal Year:
2016,
Volume and Issue:
231(14), P. 2569 - 2579
Published: March 22, 2016
Optimization
design
of
centrifugal
pumps
involving
multiple
parameters
and
objectives
is
a
complicated
research
topic.
The
orthogonal
method
introduced
in
the
present
study
to
find
high
efficiency
low
cost
way
optimization
process
pump.
A
table
designation
L
16
(4
5
)
established,
which
individuals
impellers
are
generated
with
five
parameters:
blade
wrap
angle,
angles
at
impeller
inlet
outlet,
leading
edge
position,
trailing
lean
varying
four
levels
for
each
parameter.
To
realize
multiobjective
both
pump
cavitation
performance,
an
integrated
factor
considering
weight
two
introduced.
On
basis
validated
computational
fluid
dynamics
(CFD)
technique,
range
analysis
gives
influence
order
also
determines
value
Finally,
optimal
obtained
remarkable
superiority
on
3.09%
rise
performance
1.45
m
promotion
comparison
original
