0% found this document useful (0 votes)
19 views

7-Hydraulic analysis of gate valve using computational fluid dynamics (CFD)

The article presents a hydraulic analysis of gate valves using computational fluid dynamics (CFD) to understand the physical behavior of fluid flow at various degrees of valve opening. It employs Ansys Workbench software to model and analyze the impact of flow velocities and valve positions on hydrodynamic forces, revealing potential issues like erosion and vibrations. The study aims to enhance the design and performance of gate valves in water supply networks and other industrial applications.

Uploaded by

Ali Ece
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
19 views

7-Hydraulic analysis of gate valve using computational fluid dynamics (CFD)

The article presents a hydraulic analysis of gate valves using computational fluid dynamics (CFD) to understand the physical behavior of fluid flow at various degrees of valve opening. It employs Ansys Workbench software to model and analyze the impact of flow velocities and valve positions on hydrodynamic forces, revealing potential issues like erosion and vibrations. The study aims to enhance the design and performance of gate valves in water supply networks and other industrial applications.

Uploaded by

Ali Ece
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 15

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/346404470

Hydraulic analysis of gate valve using computational fluid dynamics (CFD)

Article in Scientific Review Engineering and Environmental Sciences (SREES) · October 2020
DOI: 10.22630/PNIKS.2020.29.3.23

CITATIONS READS

7 1,115

3 authors:

Elvis Žic Patrik Banko


University of Rijeka University of Rijeka
12 PUBLICATIONS 23 CITATIONS 1 PUBLICATION 7 CITATIONS

SEE PROFILE SEE PROFILE

Luka Lešnik
University of Maribor
31 PUBLICATIONS 442 CITATIONS

SEE PROFILE

All content following this page was uploaded by Elvis Žic on 07 December 2020.

The user has requested enhancement of the downloaded file.


PRACE ORYGINALNE
ORIGINAL PAPERS

Scientific Review – Engineering and Environmental Sciences (2020), 29 (3), 275–288


Sci. Rev. Eng. Env. Sci. (2020), 29 (3)
Przegląd Naukowy – Inżynieria i Kształtowanie Środowiska (2020), 29 (3), 275–288
Prz. Nauk. Inż. Kszt. Środ. (2020), 29 (3)
http://iks.pn.sggw.pl
DOI 10.22630/PNIKS.2020.29.3.23

Elvis ŽIC1, Patrik BANKO1, Luka LEŠNIK2


1
University of Rijeka, Faculty of Civil Engineering
2
University of Maribor, Faculty of Mechanical Engineering

Hydraulic analysis of gate valve using computational fluid


dynamics (CFD)

Key words: gate valve, hydrodynamic analy- 16 bar. Their main disadvantage is the
sis, CFD, Ansys Workbench software package large required number of turns of the
valve opening/closing handwheel.
During the opening or closing of
Introduction the gate valves, considerable forces are
exerted on the valve construction due to
The water supply network consists the leakage of the flow. The hydrodyna-
of a number of interdependent elements, mic forces caused by the high flow ve-
one of which is a gate valve. They re- locities under the gate valve result in a
present machine elements commonly vertical force downwards. As the gate
used to control fluid flow because they valve opens, the velocities increase non-
provide positive seal at high liquid and -linearly in relation to the degree of open-
gas pressures (Fig. 1). They are used in ing. Most flow changes occur near the
various industries such as refineries, pe- valve at a relatively high flow velocity
trochemical plants, power stations, hy- and cause wear on the valve walls and
droelectric power plants, nuclear power bearings. High flow velocities in par-
plants, etc. High flow velocities with tially opened valves can cause erosion
partial opening of the valve can lead to of the valve discs and the bearings them-
erosion of its walls, vibrations and noise selves, and vibrations can cause dama-
(Banko, 2019). They are most commonly ge to the partially opened disc (Quimby,
used for drinking water and wastewa- 2007). When the gate valve is lowered to
ter in the temperature range from –20 to reduce the flow (e.g. by closing), the pres-
+70°C and can withstand flow velocities sure on the lower surface of the valve de-
of up to 5 m·s–1 and pressures of up to creases due to the high flow velocity,

Hydraulic analysis of gate valve using computational fluid dynamics (CFD) 275
a b

FIGURE 1. Gate valve: a – cross-section (1 – body, 2 – bonnet, 3 – solid wedge, 4 – body seats,
5 – stem, 6 – back seat, 7 – gland follower, 8 – gland flange, 9 – stem nut, 10 – yoke nut, 11 – hand-
wheel, 12 – handwheel nut, 13 – stud bolts, 14 – nuts, 15 – stud bolts, 16 – nuts, 17 – bonnet gasket,
18 – lubricator, 19 – packing); b – model with solid wedge (Banko, 2019)

while the pressure on the upper surface (2013) carry out stress analyses on cri-
of the valve changes only slightly relative tical components of gate valves using
to the static regime. The aim of this paper the FEA technique. The modelling of
is to apply computational fluid dynamics valve components was performed in the
(CFD) to gain insights into the physical CATIA V5R17 software and analysed
quantities for gate valve models within with the FEM method in the ANSYS-
a pipe at characteristic opening degrees. -11 software. The validation of the soft-
By comparing the results of models with ware results is analytically supported by
different degrees of opening of the gate a stress analysis using the classical theory
valve, a more accurate and better quality of solid mechanics. Patil and Gambhire
of the observed pipeline components can (2014) provide a basic methodology for
be guaranteed. the design of gate valve bodies using a
CAD technology where structural FEM
analysis is applied at maximum operating
Previous research pressure. The work involved static, dyna-
mic, thermal, harmonic and electromag-
Numerous studies have been carried netic analyses on a valve using CATIA
out on gate valves, only some of which and Ansys Fluent software. The work of
are listed below. Jatkar and Dhanwe Wang (2014) is based on the CAD/CAE

276 E. Žic, P. Banko, L. Lešnik


system. The influence of factors such as length of 820 mm was taken (300 mm in
fluid flow, flow velocity, wall thickness front of the valve and 520 mm behind the
of the valve body and transverse instal- valve), because the changes are larger
lation was investigated in the paper. and longer in the span behind the gate
Pujari and Joshi (2016) carried out the valve. The DesignModeler software was
analysis and optimization of the design used to generate the network model of
of gate valve bodies using the FEA tech- the gate valve. After mesh generation, it
nique and stress analysis. Katkar, Kul- is necessary to check the quality of the
karni, Patil and Katkar (2017) analysed numerical mesh to ensure that a mean-
the critical components of a gate valve. ingful result is obtained during pro-
The paper gives a detailed overview of cessing (Žic, 2019). It is also necessary
the different techniques used in the de- to define all the contour elements of the
sign of gate valves (developed in CATIA future model (e.g. inlet and outlet pro-
software) and the analysis in the ANSYS file, pipe walls, valve, etc.). The network
Workbench software package using the consists of 101,205 nodes and 502,984
FEM technique. elements. In addition to checking the
quality of the numerical grid, the qual-
ity of the elements was also checked by
Application of numerical models checking the aspect ratios for the tri-
angle, prism and tetrahedron, the Jaco-
For the calculation and hydraulic bian ratio or “Jacobian”, the twist factor,
analysis in this paper the Ansys CFX 19.1 the characteristic length of the element,
and Ansys Fluent 19.1 software with- etc. For processing, it is necessary to
in the Ansys Workbench software pack- define physical parameters for a given
age was used (Ansys CFX 15.0, 2015, numerical model/submodel, including
Žic, 2019). The following part describes the definition of the input variables and
the design of a numerical model of a gate their values, the definition of a model
valve using the Ansys CFX 19.1 soft- type, the definition of the dynamic and
ware and the definition of the water sup- kinematic viscosity and the initial and
ply pipe and the valve around which the boundary conditions. A single-phase
fluid flows. The water supply pipe and problem is selected, which means that
the 3D geometric model of the gate valve only one fluid is defined in the problem
were created in the AutoCAD 2016 soft- (water at 25°C). For the hydrodyna-
ware for a starting position of 20% pipe mic analysis, a stationary flow regime
closure. The water supply pipe has a dia- with a reference pressure of 101,325 Pa
meter of 100 mm, while the thickness of without heat transfer within the model
the pipe and valve flange is 1 mm. Defin- and the so-called k–ε turbulence
ing and importing the pipe system geo- model with standard wall function was
metry is done in the SpaceClaim and De- chosen. The first variable (k) represents
signModeler software packages within the turbulent kinetic energy and the
the software Ansys Workbench packa- second transport variable (ε) refers to the
ge (Banko, 2019). For the initial model dissipation rate of the turbulent kinetic
with a 20% of the valve opening a pipe energy. The transport equation for k is

Hydraulic analysis of gate valve using computational fluid dynamics (CFD) 277
described by the expression (1) and the βc = β − β *ξ * F ( Mt ) (4)
transport equation for ε by the expres-
sion (2) (Ansys CFX 15.0, 2015):
F ( M t ) = ( M t2 − Mt20 ) H ( M t − M t 0 ) (5)

Dk § ∂U j ∂Ui · ∂U j for the values Mt0 = 0.25 and ξ* = 1.5.


ρ = μt ¨ + ¸ +
¨ ∂xi ∂x j ¸ An initial inlet flow velocity of 1.0 m·s–1
Dt © ¹ ∂xi
is defined for the inlet profile on the
∂ ­ ∂k ½
+ ®( μ + μt σ k ) ¾ − ρε + Gb − YM + Sk surface of the entire inlet profile, while
∂xi ¯ ∂xi ¿ a relative pressure of 0 Pa is defined on
the outlet profile. This means that at the
(1)
last profile of the water supply pipe the
Dε § ε · ª § ∂U j ∂Ui · ∂U j º pressure is equal to the pressure outside
ρ = C1ε ¨ ¸ « μt ¨ + ¸ + C3ε Gb » + the pipe (atmospheric pressure). In the
Dt © k ¹ «¬ ©¨ ∂xi ∂x j ¸ ∂xi
¹ »¼
post-processing part of the numerical
∂ ­ ∂ε ½ § ε2 · modelling, arbitrary transverse and lon-
+ ®( μ + μt σ ε ) ¾ − C2ε ρ ¨¨ ¸¸ + Sε
∂xi ¯ ∂xi ¿ © k ¹ gitudinal profiles are selected, on the ba-
(2) sis of which changes of certain physical
quantities within the obtained model can
The turbulent viscosity μt is defined be represented. The gate valve was ana-
lysed by four positions: 20, 40, 60 and
k2
by the expression μt = ρ Cμ where ρ 80% of the valve closed. For each of
ε these submodels a hydrodynamic analy-
is the density of the liquid. The veloci- sis of the fluid flow around the valve at
ties Ui and Uj define the velocities in the an inflow velocity of 1.0 and 1.5 m·s–1
longitudinal and transverse cross section was performed.
of the flow. The coefficients σk, σb, C1ε,
C2ε, C3ε and Cμ are the empirically de-
fined constants. With the marks Gb, Hydrodynamic analysis
YM, Sk and Sε within the expressions (1) and research results
and (2) are presented the values of the
variables with which we can model the The processed variants were com-
turbulence. The compressibility effects pared for each physical quantity, name-
are denoted by YM, the buoyancy force ly flow velocity (v), relative pressure (p)
by Gb and user-defined sources by Sk and and turbulence kinetic energy (k). Each
Sε. The compressibility effects are mainly of the physical quantities is calculated
due to large changes in the properties and using the same eight transverse (Fig. 2)
characteristics of the fluid. Their influ- and nine longitudinal profiles. The trans-
ence is described by the coefficients βc verse profiles are arranged in such a way
and β*c as a function of the Mach num- that the first one is halfway between the
ber by the following expressions (Decaix start of the pipe and the valve, the second
& Goncalvès da Silva, 2013): one directly in front of the valve, the
fourth profile runs through the middle of
βc* = β * (1 + ξ * F ( M t )) (3) the valve, the next three profiles are di-

278 E. Žic, P. Banko, L. Lešnik


TP7 TP8

TP1 TP2 position of gate valve


FIGURE 2. Arrangement of the transverse profiles (TP) in relation to the gate valve

rectly behind the valve and the last one last four models with an inlet velocity of
halfway between the valve and the end 1.5 m·s–1.
of the pipe. The longitudinal profiles are The figure shows that a vortex flow
positioned so that the middle fifth pro- is observed in the area behind the gate
file is in the middle of the pipe and the valve at 80% closure, which is a conse-
four longitudinal profiles are symmetri- quence of the abrupt narrowing of the flow
cally arranged at equal distances on both cross-section under the valve, which also
sides. causes the greatest increase in flow velo-
city (red colour in Fig. 3). The stream-
lines of each model are shown at local
Fluid flow velocity values, i.e. the colours are not univer-
sal and are not the same on each of the
Figure 3 shows a longitudinal view of models, therefore the flow velocities on
the gate valve model at various degrees the model cannot be compared with each
of opening based on the 150 streamlines. other depending on the colour tones,
The first four models show models with but only individually (the legends gi-
an inlet velocity of 1.0 m·s–1 and the ven in Fig. 3 refer to a gate valve with a

b
c

d
e

f
g

h
FIGURE 3. Model view of gate valves with streamlines: a – model with 20% gate closure (inlet ve-
locity v = 1 m·s–1); b – 40% closure (v = 1 m·s–1); c – 60% closure (v = 1 m·s–1); d – 80% closure
(v = 1 m·s–1); e – 20% closure (v = 1.5 m·s–1); f – 40% closure (v = 1.5 m·s–1); g – 60% closure
(v = 1.5 m·s–1); h – 80% closure (v = 1.5 m·s–1)

Hydraulic analysis of gate valve using computational fluid dynamics (CFD) 279
valve closing degree of 80% at velocities velocity of 1.0 m·s–1 the average valve
of 1.0 and 1.5 m·s–1). The maximum, mini- flow velocity is 1.03 m·s–1, for models
mum and average values of flow veloci- with 40% closure 1.40 m·s–1, with 60%
ties for each of the submodels and both closure 2.39 m·s–1 and with 80% clo-
inlet flow velocities are shown in Table 1. sure the value is 6.50 m·s–1. The maxi-
The average and maximum flow velo- mum flow velocity of 10.4 m·s–1 occurs
cities within the model increase expo- at the fifth profile (directly behind the
nentially as a function of the percentage valve) for models with an inlet velocity of
closure of the gate valve. The increment 1.0 m·s–1 and 15.6 m·s–1 for models with
percentages coincide with the second an inlet velocity of 1.5 m·s–1. Maximum
decimal place and are 115.5% from 20 flow velocities with lower valve clo-
to 40% closed, 133% from 40 to 60% sure occur at a greater distance behind the
closed and 175% from 60 to 80% closed valve, while models with a higher valve
valve for the average values. The percen- closing percentage have maximum val-
tages for increasing the maximum values ues of flow velocity closer to the valve
of the flow velocities are in the same or- due to the abrupt narrowing of the flow
der: 162, 175 and 240%. Table 2 shows area. The nine longitudinal profiles are
the maximum (bold values) and average defined at regular intervals, starting from
values of flow velocities for all positions the centre of the pipe towards the edges
of valve closure with inlet velocities of (the centre of the fifth profile intersects
1.0 and 1.5 m·s–1 up to eight transverse the centre of the valve, seen perpendic-
profiles (Fig. 2). The positions of the lar- ular to the valve). They show most clear-
gest maximum and average flow velocity ly the change in flow velocity along the
values vary depending on the percentage pipe and the transient flow velocity from
of valve closure. the beginning of the pipe system through

TABLE 1. View of the maximum, minimum and average values of the flow velocities [m·s–1] for each
of the gate valve models

Percentage of gate v = 1.0 m·s–1 v = 1.5 m·s–1


valve closure
[%] max min avg max min avg
20 1.564 0.007 1.035 2.337 0.021 1.553
40 2.533 0.003 1.195 3.797 0.004 1.795
60 4.415 0.004 1.594 6.633 0.003 2.390
80 10.585 0.002 2.780 15.884 0.004 4.220

It is also noticeable that the values the valve to the recovery of the flow ve-
of maximum and average flow velocities locity at a certain distance behind the
for all profiles in the immediate vicin- valve. Table 3 shows the maximum and
ity of the valve increase exponentially average flow velocities for all positions of
with the percentage of closure. For mod- valve closure with inlet velocities of 1.0
els with 20% closure and an inlet flow and 1.5 m·s–1 for nine randomly selected

280 E. Žic, P. Banko, L. Lešnik


TABLE 2. View of the maximum and average values of the flow velocities [m·s–1] at a gate valve on
the corresponding transverse profiles

Cross 20% valve closure 40% valve closure 60% valve closure 80% valve closure
section
profile max avg max avg max avg max avg
1 1.03 1.00 1.03 0.99 1.03 0.99 1.03 0.99
2 1.15 0.99 1.37 1.02 1.72 1.08 2.34 1.14
3 1.29 1.01 1.77 1.25 2.76 1.64 5.99 2.40
4 1.45 1.03 2.13 1.40 3.75 2.39 9.93 6.49
5 1.53 1.02 2.28 1.20 4.03 1.59 10.40 2.68
6 1.56 1.17 2.43 1.40 4.26 1.80 10.34 2.62
7 1.51 1.16 2.53 1.57 4.41 1.97 9.84 2.81
8 1.16 0.98 1.59 0.97 2.51 1.14 4.78 2.04
Cross 20% valve closure 40% valve closure 60% valve closure 80% valve closure
section
profile max avg max avg max avg max avg
1 1.54 1.49 1.54 1.48 1.54 1.48 1.54 1.48
2 1.73 1.49 2.05 1.54 2.59 1.62 3.51 1.72
3 1.92 1.51 2.64 1.87 4.13 2.46 8.99 3.60
4 2.17 1.54 3.20 2.10 5.63 3.58 14.91 9.75
5 2.28 1.51 3.43 1.80 6.04 2.39 15.59 4.02
6 2.34 1.73 3.65 2.12 6.39 2.71 15.50 3.93
7 2.26 1.72 3.79 2.37 6.63 2.96 14.76 4.22
8 1.74 1.48 2.39 1.46 3.79 1.71 7.14 3.02

longitudinal profiles. The maximum av- model with an inflow velocity of 1.0 and
erage flow velocities on the defined lon- 2.67 m·s–1 for the model with an inflow
gitudinal profiles are 1.79 m·s–1 for the velocity of 1.5 m·s–1. Figure 4a shows
a b

FIGURE 4. Graphical view of the maximum flow velocities for a gate valve model with 80% closure
based on transverse profiles (a) and longitudinal profiles (b)

Hydraulic analysis of gate valve using computational fluid dynamics (CFD) 281
TABLE 3. View of the maximum and average values of flow velocities [m·s–1] at a gate valve at the
corresponding longitudinal profiles

Longitudinal 20% valve closure 40% valve closure 60% valve closure 80% valve closure
profile max avg max avg max avg max avg
1 1.561 1.015 2.378 1.019 3.806 0.994 2.021 0.976
2 1.559 1.002 2.522 1.027 4.160 1.110 9.816 1.325
3 1.549 0.982 2.532 1.040 4.356 1.195 10.198 1.635
4 1.551 0.981 2.529 1.042 4.383 1.208 10.347 1.702
5 1.549 0.981 2.529 1.052 4.413 1.243 10.437 1.787
6 1.545 0.981 2.528 1.044 4.390 1.213 10.264 1.700
7 1.547 0.986 1.527 1.035 4.333 1.182 10.098 1.566
8 1.553 1.002 2.521 1.027 4.160 1.116 9.911 1.337
9 1.563 1.016 2.405 1.026 3.946 0.997 2.022 0.971
Longitudinal 20% valve closure 40% valve closure 60% valve closure 80% valve closure
profile max avg max avg max avg max avg
1 2.333 1.529 3.584 1.541 5.712 1.497 2.963 1.453
2 2.326 1.505 3.787 1.546 6.245 1.668 14.730 1.970
3 2.309 1.472 3.796 1.562 6.540 1.793 15.295 2.438
4 2.311 1.470 3.793 1.565 6.582 1.813 15.517 2.541
5 2.308 1.469 3.788 1.578 6.627 1.864 15.650 2.670
6 2.302 1.470 3.792 1.565 6.592 1.819 15.390 2.541
7 2.306 1.479 3.792 1.553 6.506 1.773 15.146 2.339
8 2.318 1.506 3.788 1.542 6.248 1.674 14.871 1.998
9 2.336 1.530 3.617 1.539 5.918 1.498 2.987 1.456

a graphical representation of the flow ve- Relative pressure


locities for a gate valve model with 80%
closure at an inflow velocity of 2.0 m·s–1, The maximum, minimum and aver-
compared with the same model with an age values of relative pressures [Pa] for
inflow velocity of 1.0 and 1.5 m·s–1. The all submodels of gate valves based on
maximum velocity value on the fifth the k–ε turbulent model are shown in
profile at an inlet velocity of 2.0 m·s–1 is Table 4. The maximum, minimum and
20.80 m·s–1. Figure 4b shows the values average relative pressure values increase
of maximum flow velocities per longi- exponentially when the valve is closed.
tudinal profile for the model with 80% The average relative pressure of a valve
valve closure for inlet velocities of 1.0, 80% closed is approximately 75 times
1.5 and 2.0 m·s–1. higher than in the case of a valve 20%

282 E. Žic, P. Banko, L. Lešnik


TABLE 4. View of maximum, minimum and average relative pressures [Pa] for each of the gate valve
submodels

Percentage of gate v = 1.0 m·s–1 v = 1.5 m·s–1


valve closure
[%] max min avg max min avg
20 895 –1 053 171 1 983 –2 501 355
40 1 989 –2 689 407 4 433 –6 057 886
60 7 223 –8 209 1 897 16 195 –18 347 4 228
80 56 948 –46 156 12 890 127 831 –103 401 29 080

closed. The values to be analysed when submodels of gate valves and both inlet
dimensioning the valve as a function of velocities. The highest relative pressures
pressure are maximum and minimum and the lowest negative pressures occur
pressures, since extreme maximum and at both inlet flow velocity variants for
minimum pressures can cause the pipe it- the same profiles. The maximum relative
self to expand or twist, which can lead to pressure values are 56,942 Pa for the in-
its damage and cracking. The upper row let velocity of 1.0 m·s–1 and 127,817 Pa
in Figure 5 shows the changes in relative for the inlet velocity of 1.5 m·s–1, which
pressures at the first four transverse pro- occur for partial models with 80% valve
files (a), b), (c) and (d) and the bottom closure on the third profile 7 cm in front
row shows the changes in relative pres- of the disc surface of gate valve, seen in
sures at the last four transverse profiles the direction of flow. The lowest nega-
(e), (f), (g) and (h) for the gate valve sub- tive pressures also occur in submodels
model at 80% closed (at 1.0 m·s–1). with 80% valve closure on the fourth
Table 5 shows the maximum, mini- profile, which is located at the back of
mum and average values of the relative the valve disc.
pressures at the transverse profiles for all

a b c
d

TP 1 TP 2 TP 3 TP 4

e f g h

TP 5 TP 6 TP 7 TP 8
FIGURE 5. Distribution of the relative pressures on transverse profiles of gate valve submodels with
80% of valve closure and inflow velocity of 1.0 m·s–1

Hydraulic analysis of gate valve using computational fluid dynamics (CFD) 283
TABLE 5. The view of maximum, minimum and average values of the relative pressures [Pa] at the transverse profiles TP1 to TP8 in gate valve
model

Cross v = 1.0 m·s–1


section 20% valve closure 40% valve closure 60% valve closure 80% valve closure
profile max min avg max min avg max min avg max min avg
1 443 439 441 1 538 1 534 1 536 6 773 6 769 6 671 56 499 56 495 56 497
2 665 261 403 1 836 1 076 1 446 7 103 5 769 6 583 56 844 54 647 56 160
3 870 123 –367 1 978 517 1 137 7 216 3 526 5 546 56 942 40 241 52 472
4 48 –58.2 –230 178 –1 548 –791 1 834 –3 759 –1 782 17 326 –13 637 –3 976
5 –21 –623 –343 –185 –1 430 –1 035 141 –3 221 –2 530 793 –9 522 –7 905
6 –110 –838 –443 –674 –1 602 –1 272 –1 736 –2 412 –2 970 –5 821 –9 588 –8 699
7 –174 –591 –373 –1 185 –1 844 –1 589 –3 098 –3 822 –3 579 –8 856 –10 293 –9 819
8 97 90 94 –39 –68 –54 –669 –735 –69 –3 730 –3 922 –3 807
Cross v = 1.5 m·s–1
section 20% valve closure 40% valve closure 60% valve closure 80% valve closure
profile max min avg max min avg max min avg max min avg
1 951 944 949 3 406 3 398 3 402 15 170 15 160 15 165 126 880 126 850 126 850
2 1 459 552 869 4 084 2 375 3 208 15 919 12 920 14 750 127 593 121 977 126 100
3 1 924 245 789 4 407 1 119 2 492 16 178 7 837 12 420 127 817 90 249 117 820
4 81 –1 377 –568 335 –3 549 –1 827 4 067 –8 558 –40 71 38 779 –30 978 –9 124
5 –74 –1 435 –804 –461 –3 254 –2 375 –259 –7 321 –5 755 1 715 –21 557 –17 870
6 –269 –1 830 –993 –1 563 –3 682 –2 925 –3 962 –7 750 –6 749 –13 093 –21 718 –19 650
7 –415 –1 298 –843 –2 698 –4 244 –3 636 –7 016 –8 658 –8 111 –18 727 –23 214 –22 110
8 188 171 180 –188 –122 –154 –1 538 –1 686 –1 599 –8 267 –8 721 –8 446
Cavitation can occur on the part of the sum of the variance (the square of the
the pipe behind the gate valve due to ne- standard deviations) of the velocity com-
gative pressures. The highest average re- ponents. Figure 6 shows the values of the
lative pressures are much higher in sub- turbulent kinetic energy on the cross pro-
models with higher closure percentages files of the submodels at 80% valve clo-
(60 and 80%) than 20% closure valve. sure and an inflow velocity of 1.5 m·s–1.
The value of the highest average relative The upper part of Figure 6 shows three
pressure at 80% closed valve is almost profiles in front of valve (a), b), (c) and
120 times higher than the 20% closed one at valve (d), and the lower part of
valve submodel. As the inlet velocity and the figure shows profiles (e), (f), (g), (h),
the valve closure percentage increase, an which are located behind the gate valve.
additional increase in relative pressures The figure shows that the maximum val-
can be expected up to a certain closure ues of the turbulent kinetic energy occur
percentage when the maximum pressure at the valve itself and beyond, extending
value decreases from that of the previ- from the bottom of the valve wall to the
ous valve closure percentage. For proces- upper half of the pipe along the flow
sed submodels, the maximum absolute behind the valve. The maximum value
pressure to be expected within the that appears is 2.66 m2·s–2 on the last
pipeline is 2.29 bar at 80% closure and cross-sectional profile. The maximum
an inlet velocity of 1.5 m·s–1, which is a value of the turbulent kinetic energy of
fully acceptable pressure for water pipes. 5.52 m2·s–2 does not appear on any user-
The maximum relative pressure for both -defined profile, but directly behind the
flow velocities occur at the middle profile last profile (h). Table 6 shows the max-
and have values of 56,948 Pa for the inlet imum (values in bold), minimum and
velocity of 1.0 m·s–1 and 127,831 Pa for average values of the turbulent kinetic
the inlet velocity of 1.5 m·s–1 (submodel energy [m2·s–2] on the transverse pro-
with 80% of valve closure). The highest files for all numerical submodels and both
pressures occur in the vicinity of the input velocities of 1.0 and 1.5 m·s–1.
second and penultimate longitudinal pro- With the increase of the valve clo-
file, which are 15 mm from the pipe wall. sure degree, the maximum values of the
turbulent kinetic energy move further
away from the valve. This is due to the
Turbulent kinetic energy increase in the variations in flow velo-
cities caused by moving away from the
In fluid dynamics, the turbulent ki- valve in submodels with a smaller clos-
netic energy – TKE (k) is a measure of ing degree compared to a larger closing
the kinetic energy per unit mass associ- degree (e.g. 60% of the valve closing
ated with eddy currents in turbulent flows. degree). For this reason, the maximum
According to the RANS equations (Rey- values for submodels with higher clo-
nolds-averaged Navier–Stokes equa- sure percentages occur behind the last
tions), the turbulent kinetic energy can user-defined cross-section profile in the
be calculated according to the turbulence direction of water flow.
model. It is generally calculated as half

Hydraulic analysis of gate valve using computational fluid dynamics (CFD) 285
a b c d

TP 1 TP 2 TP 3 TP 4
e f g h

TP 5 TP 6 TP 7 TP 8
FIGURE 6. Distribution of turbulent kinetic energy at transverse profiles of a gate valve submodel with
80% of valve closure and an inlet flow velocity of 1.5 m·s–1

Conclusions maximum values of velocities, pressures


and other physical quantities occur in
In this paper a hydraulic analysis of models with a lower valve opening de-
gate valve models was performed using gree. The maximum values of the physi-
the commercial softwares Ansys CFX cal quantities in the analysed models oc-
19.1 and Ansys Fluent 19.1. The ana- cur mainly in the valve area or behind it.
lyses were performed for 4° of opening This paper shows that the implementation
of the gate valve with inlet velocities of of hydrodynamic analysis is possible for
1.0 and 1.5 m·s–1. After the hydrodyna- different forms of valve geometry. Cor-
mic analysis it was found that all models rect numerical modelling through CFD
show vortices in the area behind the gate technology allows the obtained results to
valve, especially at smaller opening de- be used to improve the valve characteri-
grees. The appearance of the vortex and stics in its design and operation.
its movement along the pipe is clearly
visible on the given central longitudinal Acknowledgements
profiles of the pipe system. In the case of This paper is the result of a project
the gate valve with 40% closing degree on the Development of Research Infra-
and an inlet flow velocity of 1.0 m·s–1, structure at the University Campus in Ri-
the maximum velocity occurring is jeka (RC.2.2.06-0001), co-funded by the
2.53 m·s–1, whereas for the same model European Regional Development Fund
and an inlet flow velocity of 1.5 m·s–1 (ERDF) and the Ministry of Science and
it is 3.80 m·s–1. The analysis shows that Education of the Rep. of Croatia.

286 E. Žic, P. Banko, L. Lešnik


TABLE 6. The view of the maximum, minimum and average values of turbulent kinetic energy [m2·s–2] on the transverse profiles TP1 to TP8
of a gate valve

Cross v = 1.0 m·s–1


section 20% valve closure 40% valve closure 60% valve closure 80% valve closure
profile max min avg max min avg max min avg max min avg
1 0.01144 0.00014 0.00311 0.01051 0.00014 0.00314 0.01081 0.00014 0.00313 0.01089 0.00014 0.00314
2 0.01356 0.00010 0.00308 0.01716 0.00011 0.00322 0.02147 0.00012 0.00350 0.03406 0.00014 0.00378
3 0.01497 0.00011 0.00243 0.02404 0.00016 0.00273 0.04333 0.00034 0.00384 0.14914 0.00033 0.00949
4 0.02208 0.00011 0.00565 0.05912 0.00022 0.01691 0.16759 0.00092 0.03915 0.88934 0.01319 0.21620
5 0.01968 0.00013 0.00347 0.05966 0.00033 0.01707 0.13125 0.00206 0.03468 0.53885 0.01338 0.10860
6 0.02064 0.00015 0.00482 0.09068 0.00094 0.02597 0.17000 0.00644 0.05021 0.53124 0.01730 0.13810
7 0.02233 0.00021 0.00503 0.18801 0.00256 0.04415 0.29017 0.01401 0.08991 0.69996 0.02282 0.20790
8 0.01337 0.00012 0.00317 0.03227 0.00321 0.01127 0.20445 0.01421 0.06462 1.14978 0.06886 0.35820
Cross v = 1.5 m·s–1
section 20% valve closure 40% valve closure 60% valve closure 80% valve closure
profile max min avg max min avg max min avg max min avg
1 0.02268 0.00027 0.00623 0.02073 0.00027 0.00628 0.02135 0.00027 0.00626 0.02151 0.00027 0.00628
2 0.02666 0.00021 0.00617 0.03358 0.00024 0.00642 0.04186 0.00027 0.00698 0.06617 0.00030 0.00751
3 0.02969 0.00023 0.00489 0.04692 0.00035 0.00546 0.08415 0.00074 0.00762 0.28910 0.00071 0.01871
4 0.06786 0.00025 0.01495 0.12699 0.00048 0.03366 0.33137 0.00200 0.08358 2.11641 0.02867 0.47990
5 0.06669 0.00028 0.01783 0.13641 0.00070 0.03853 0.30955 0.00449 0.08088 1.22364 0.03166 0.26990
6 0.13371 0.00040 0.02447 0.19733 0.00206 0.05792 0.40030 0.01428 0.11610 1.20911 0.04175 0.34190
7 0.12998 0.00075 0.02473 0.40703 0.00555 0.09827 0.68619 0.03434 0.21010 1.75139 0.05895 0.52140
8 0.02658 0.00087 0.00666 0.07491 0.00718 0.02506 0.47935 0.03283 0.15270 2.66295 0.16141 0.84520
References Summary
Ansys CFX 15.0 (2015). Manual. Southpointe, Hydraulic analysis of gate valve using
Canonsburg: ANSYS Inc. computational fluid dynamics (CFD). As
Banko, P. (2019). Hydraulic analysis of gate and a very important element of most water
butterfly valves using Computational Fluid supply systems, valves are exposed to the
Dynamics (graduate thesis). University of effects of strong hydrodynamic forces. When
Rijeka, Rijeka. exposed to large physical quantities, the
Decaix, J. & Goncalvès da Silva, E. (2013). Com- valve and piping can be damaged, which
pressible effects modelling in turbulent cavi- could endanger the performance of a water
tating flows. European Journal of Mechanics supply system. This is the main reason why
– B/Fluids, 39, 11-31.
it is necessary to foresee and determine the
Jatkar, K.H. & Dhanwe, S.S. (2013). Finite Ele-
maximum values of velocity, pressure and
ment Analysis of Gate Valve. Asian Review
of Mechanical Engineering, 2(1), 44-49.
other physical quantities that can occur in the
Katkar, N.Y., Kulkarni, R.S., Patil, P.A. & Katkar, system under certain conditions. Predicting
S.E. (2017). A review on design and analysis extreme conditions allows us to correctly
of gate valve using various computer aided size the valve for the expected conditions to
techniques. International Journal for Inno- which the valve might be exposed, which is
vative Research in Science and Technology, also the main objective of this paper. One of
3(10), 218-220. the methods for predicting and determining
Patil, P.B. & Gambhire, V.R. (2014). Structural extreme values on a valve is to perform a sim-
analysis of gate valve body using FEA. In- ulation with computational fluid dynamics
ternational Journal of Engineering Research (CFD). This is exactly the method used in
& Technology, 3(6), 1815-1818. the preparation of this paper with the aim of
Pujari, A.A. & Joshi, G.S. (2016). Analysis and gaining insight into the physical magnitudes
design optimization of 8’’- 600# gate valve for models of gate valves positioned inside
body using FEA and stress analysis. Inter- a pipe under characteristic degrees of valve
national Journal of Research in Engineering closure. The Ansys CFX 19.1 and Ansys Flu-
and Technology, 4(2), 440-443. ent 19.1 software was used to simulate the
Quimby, B. (2007). Hydrodynamic loads. Re- hydrodynamic analysis and obtain the re-
trieved from: https://www.bgstructuralen- quired results. The hydrodynamic analysis
gineering.com/BGASCE7/BGASCE7006/ was performed for four opening degrees of
BGASCE70604.htm
gate valve.
Wang, D. (2014). Research on CAD/CAE system
of gate valve based on UG. Applied Mechan-
ics and Materials, 541-542, 618-621. Authors’ address:
Žic, E. (2019). 3rd Summer School of Compu- Elvis Žic
tational Fluid Dynamics (materials from (https://orcid.org/0000-0002-5626-8394)
lectures). Novi Sad: Faculty of Technical University of Rijeka
Sciences in Novi Sad. Faculty of Civil Engineering
Radmile Matejčić, 3, 51000, Rijeka
Croatia
e-mail: [email protected]

288 E. Žic, P. Banko, L. Lešnik

View publication stats

You might also like