scieee Science in your language
[en] (orig)

Characteristics of the pipe junction with 2.4 cross-section area ratio for the case of the flow division.

Abstract

There are presented characteristics of the pipe junction for the case of the flow division in this paper. The pipe junction consists of one straight pipe, with the diameter 50 mm and one adjacent pipe with diameter 32 mm. The characteristics have been measured for five different angles of the adjacent pipe.

Read accessible full text

Characteristics of the pipe junction with 2.4 cross-section area ratio for the case of the flow division.

Author: Štigler, Jaroslav; Šperka, Oldřich
Publisher: EDP Sciences
Year: 2013
DOI: 10.1051/epjconf/20134501112
Source: https://dspace.vut.cz/bitstreams/d7392d1b-2ca0-4b89-ae31-5870109a1079/download
Cha ac e is ics o he pipe junc ion wi h 2.4 c oss-sec ion a ea a io o
he case o he low di ision
J. Š igle 1a and O. Špe ka1
1B no Uni e si y o Technology, Facul y o Mechanical Enginee ing, Ene gy Ins i u e, Vic o Kaplan Depa men o Fluid
Enginee ing, Technická 2896/2, 616 69, Czech Republic
Abs ac . The e a e p esen ed cha ac e is ics o he pipe junc ion o he case o he low di ision
in his pape . The pipe junc ion consis s o one s aigh pipe, wi h he diame e 50 mm and one
adjacen pipe wi h diame e 32 mm. The cha ac e is ics ha e been measu ed o i e di e en angles
o he adjacen pipe.
1 In oduc iona
Many pape s, abou he luid low in he pipe junc ion,
ha e been w i en by hese au ho s. I is possible o use i
o he luid low solu ion in he pipe line ne . The e a e
some o he ma hema ical models bu hey a e based on
he un ealis ic assump ions o hei coe icien s ha e no
any physical meaning. [1, 2] The new ma hema ical
model o he 90° pipe junc ion oge he wi h i s
coe icien s is desc ibed in [3, 4]. The imp o ed
ma hema ical model o he a bi a y angle o he
adjacen b anch is in oduced in he [5]. Some discussion
abou ma hema ical model o he uns eady luid low in
pipe junc ion has been p esen ed in [6]. A lo o
nume ical solu ions and expe imen s o he luid low in
he pipe junc ion ha e been done in pas ew yea s. Fo
example nume ical s udy o luid low has been p esen ed
in he [7, 8]. The PIV measu emen o he luid low in
he pipe junc ion has been desc ibed in [9] The
Compa ison o he PIV measu emen wi h nume ical
solu ion o luid low in [10]. The i s new
cha ac e is ics ob ained by measu ing and i s compa ison
wi h CFD calcula ions we e p esen ed a [11]. Many
pape s ha e been w i en abou luid low in pipe junc ion
as i was men ioned abo e and eade can ge mos o
hem on he in e ne . This pape will be he e o e
dedica ed mainly o he p esen a ion o he coe icien s
ob ained by measu emen s. And mo e o e he discussion
abou he esul s will be included in his pape .
2 Ma hema ical model
The pipe junc ion i is deal wi h is d awn in he igu e 1.
a [email p o ec ed]
The ma hema ical model o he luid low in he pipe
junc ion consis s o h ee equa ions which desc ibe he
ela ionship be ween he low a es and p essu es a he
bo de o he pipe junc ion a ea.
Fig. 1. Pipe junc ion
The bo de o he pipe junc ion a ea is delimi ed by he
c oss-sec ions S(a), S(b) and S(c) placed in such dis ance
om he low di ision domain whe e he low is no
dis u bed by he pipe junc ion. The minimum dis ance o
ensu e his assump ion is abou 10 diame e s a each
b anch.
Fi s equa ion o he ma hema ical model o he pipe
junc ion is he powe equa ion. This equa ion ep esen s
he mechanical ene gy conse a ion law in he pipe
junc ion
0
2
1
2
22
2
2
2
2
2
2
2
2

























)mX(
)X(
)mX(
)P()mc()c(
)c(
)mc(
)mb()b(
)b(
)mb(
)ma()a(
)a(
)ma(
Q
S.
Q
Qp
S
Q
Qp
S
Q
Qp
S
Q




.
(1)
The second equa ion is he momen um equa ion. This is a
ec o equa ion bu we will ake in o conside a ion only
This is an Open Access a icle dis ibu ed unde he e ms o he C ea i e Commons A ibu ion License 2 0 , which
.pe mi s un es ic ed use, dis ibu i
and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed.
on,
EPJ Web o Con e ences
DOI: 10.1051/
C
Owned by he au ho s, published by EDP Sciences, 2013
,
epjcon 201/
01112 (2013)
45
34501112
A icle a ailable a h p://www.epj-con e ences.o g o h p://dx.doi.o g/10.1051/epjcon /20134501112
EPJ Web o Con e ences
he componen in he di ec ion o b anch “b” his
di ec ion is ep esen ed by he uni no mal ec o n(b)i.
This is clea om he igu e 1
0
22
2
2

























)X(
)mX(
i)M(i)c()c()c(
)c(
)mc(
i)b()b()b(
)b(
)mb(
i)a()a()a(
)a(
)ma(
S
Q
nSp
S
Q
nSp
S
Q
nSp
S
Q





.
(2)
This equa ion has o be mul iplied by he ec o n(b)i o
ge he componen equa ion in ha di ec ion.
The las hi d equa ion is con inui y equa ion
0



)mc()mb()ma( QQQ . (3)
The summa ion con ec ion is used in he abo e
equa ions. This ma hema ical model is alid unde hese
assump ions
 S eady luid low
 The g a i y ec o is pe pendicula o he pipe
junc ion plain.
 The densi y is cons an
 The Co iolis numbe s and Bousineques
numbe s on he c oss-sec ions S(a), S(b), and S(c)
a e equal o 1.
I is possible o emo e all hese assump ions in he
gene al ma hema ical model [5 ].
The index “X” in he abo e equa ions has o be eplaced
by he index “a”, “b” o “c”. I depends on b anch o
which he coe icien s

(M)i and

(P) will be ela ed o.
3 Pipe junc ion coe icien s
Two coe icien s appea in he ma hema ical model. The
powe coe icien

(P) can be exp essed om he equa ion
(1). This coe icien ep esen s he ene gy losses in he
pipe junc ion. I is p opo ional o he kine ic ene gy pe
ime in one o he b anches o which he coe icien will
be de i ed.
The momen um coe icien

(M)i can be exp essed om
he equa ion (2). This coe icien is p opo ional o he
o ce which he luid impac s on he pipe junc ion. The
o al o ce can be ob ained by mul iplying o his
coe icien wi h he o al momen um o luid in b anch
o which he coe icien was de i ed. In his case we will
ake in o conside a ion he componen o his coe icien
only in he di ec ion o b anch “b”.
The momen um coe icien depends also on he o al
p essu e a all b anches. I all pa ame e s as shape o pipe
junc ion, low con igu a ion, low a es, di e ences o
p essu es be ween he b anches a e he same hen he
coe icien s a ies wi h o al p essu e. The e o e i is
necessa y o ela es all p essu es o he cons an alue
p essu e in some b anch. In case o his pape he
p essu e alue in b anch “a” is aken as ze o. I means
ha he momen um coe icien is de i ed o case ze o
p essu e le el in b anch “a”. I means ha he momen um
coe icien will be exp essed om his modi ied
momen um equa ion
0
22
22

























)X(
)mX(
i)M(i)c()c()ca(
)c(
)mc(
i)b()b()ba(
)b(
)mb(
i)a(
)a(
)ma(
S
Q
nSp
S
Q
nSp
S
Q
n
S
Q





.
(4)
Whe e
)a()b()ba( ppp



(5)
and
)a()c()ca( ppp



. (6)
Now he unknowns a e h ee mass low a es Q(ma), Q(mb)
and Q(mc) and absolu e p essu e p(a) and wo p essu e
di e ences

p(ba) and

p(ca) ins ead o h ee absolu e
p essu es p(a), p(b) and p(c).
Bo h coe icien s can be di ided in o wo pa s. Fi s pa s
will be ela ed o he ic ion losses in he pipe junc ion.
They will be called ic ion coe icien s. Second pa s
will be ela ed o he shape o pipe junc ion. They will be
called geome y coe icien s. I can be exp essed his way
)PG()PF()P(





(7)
and
i)MG(i)MF(i)M(





. (8)
The o al coe icien s can be ob ained by he
measu emen s. The ic ion coe icien s can be ob ained
om he known ic ion losses, ep esen ed by he
p essu e d op, in s aigh pipes. The ic ion powe
coe icien can be exp essed as ollow


)mX(
)mX(
)X(
)c()c()c(
)b()b()b()a()a()a()PF(
Q
Q
S
LQi
LQiLQi
2
2
22




.
(9)
The ic ion momen um coe icien can be exp essed as
ollow


2
)mX(
)X(
i)c()c()c()c(
i)b()b()b()b(
i)a()a()a()a(i)MF(
Q
S
nSLi
nSLi
nSLi





.
(10)
The componen o ic ion momen um coe icien in
di ec ion b anch “b” will be ob ained by mul iplying
p e ious equa ion wi h uni no mal ec o n(b)i.
The quan i ies i(a), i(b), and i(c) a e p essu e d ops in he
b anches “a”, “b” and “c” wi h low a es Q(a), Q(b) and
Q(c) wi hou pipe junc ion in luence.
The geome y coe icien s can be hen e alua ed om he
exp essions (11) and (12)
)PF()P()PG(





(11)
and
01112-p.2
EFM 2012
i)MF(i)M(i)MG(





. (12)
Ad an ages o he geome y coe icien s a e
 They a e independen on he o al low a e in he
pipe junc ion
 They a e he same o he all geome ically simila
pipe junc ions
Bu i is necessa y o emembe ha hese coe icien s
depend on he low con igu a ion in he pipe junc ion and
on he low a e a io be ween wo b anches. In case o
low di isions he h ee low con igu a ions a e possible.
I is appa en om a igu e 2.
“Di a”, (Da) “Di b”, (Db) “Di c”, (Dc)
Fig. 2. Flow con igu a ion o he low di ision.
I is necessa y o know he unc ion o momen um
geome y coe icien and he unc ion o powe
geome y coe icien . I means i is necessa y o know
wo unc ions o cha ac e is ics o each low
con igu a ion.
4 Expe imen desc ip ion
The es ing ci cui was buil in he Vic o Kaplan’s
Depa men o Fluid Enginee ing labo a o y. The pipe
junc ion was d illed in plas ic blocks. I ensu ed he sha p
edges a he di ision domain. Fi e pipe junc ions, each
o he di e en angle o he adjacen b anch, we e
measu ed. The angles o he adjacen b anch we e 30°,
45°, 60°, 75°, 90°. The gauged pa ame e s was as ollows
Flow a es a each b anch gauged by he magne ic low
me e s, absolu e p essu e gauged by he p essu e
ansduce , p essu e di e ences be ween b anches “b”-
“a” and “c”-“a”, p essu e di e ences was gauged bo h by
di e en ial p essu e manome e and also by U- ube
manome e . I is necessa y o say ha o he
cha ac e is ics e alua ion he alues om he U- ube
manome e we e used. The las gauged pa ame e was
wa e empe a u e.
Fig. 3. The diag am o he measu ing ci cui .
The diag am o he es ci cui is d awn in he igu e 3.
The low a e was con olled by pump equipped by he
equency con e e and by he al es placed a each
b anch.
The luid low in some low con igu a ions was uns able
he e o e all pa ame e s we e gauged h ee imes. Ele en
low a e a ios we e measu ed o each low
con igu a ion. Each con igu a ion o all low a e a ios
was measu ed wice o ensu e ha he esul s a e co ec .
The e is a pic u e o es ing ci cle in he igu e 4 and he
pic u e o PIV measu emen s a he igu e 5.
Fig. 4. The es ci cui o he pipe junc ion cha ac e is ics
measu ing.
Fig. 5. The pic u e o he PIV measu ing.
5 Cha ac e is ics o pipe junc ion
The esul an cha ac e is ics will be p esen ed in his
chap e . The cha ac e is ics o nine di e en angles will
be d awn in he cha s. The cha ac e is ics o he angles
bigge han 90° a e ecalcula ed om he cha ac e is ics
o he angles less han 90°.
.
5.1. Flow con igu a ion “Di a”
The coe icien s

(PG) and

(MG)1 a e d awn as a unc ion
o low a e a io q(ca) in he igu e 4 and 5
)a(
)c(
)ca( Q
Q
q
.
(13)
The e a e some in e es ing esul s. I is appa en , om he
igu e 6, ha lowes losses a e o low alues q(ca) o
angle 30°, o he middle alues o q(ca) a e lowes losses
o angle 45° and o high q(ca) a e he lowes losses o
01112-p.3
EPJ Web o Con e ences
angle 60°. The in e es ing esul is ha he losses a e no
lowes o whole ange o q(ca) o pipe junc ion wi h
angle 30°.
In case o momen um coe icien i is in e es ing, ha he
coe icien s o he angles 30° and 150°, 45° and 135°,
60° and 120°, 75° and 105° a e a he close o each o he
and he maximal alues a e o he angle 90°.
Fig. 6. The

(PG) coe icien s o low con igu a ion “Di a” as a
unc ion o low a e a io q
(ca)
.
Fig. 7. The

(MG)i coe icien s o low con igu a ion “Di a” as
a unc ion o low a e a io q
(ca)
.
One will expec ha he bigges momen um coe icien
will be o he angle 150°. Bu his phenomenon can be
explained his way. In case o he angle 150° he p essu e
in b anch “b”, o big alues o q(ca), inc eases a lo . This
p essu e ac s agains o ce caused by he low di ec ion
change. The e o e he esul ing o ce is less han he o ce
o he angle 90°.
5.2. Flow con igu a ion “Di b”
The commen s o his case a e almos iden ical o he he
commen s in case o low con igu a ion Di a. The
a iable o he pipe junc ion coe icien s e alua ion is
low a e a io q(cb) in his case
)b(
)c(
)cb( Q
Q
q
.
(14)
The powe geome y coe icien is iden ical o he one o
he low con igu a ion Di a.
In case o momen um geome y coe icien one will
expec ha he cu es o he low con igu a ion “Di a”
and angle 30° will be symme ical abou ze o alue wi h
he low con igu a ion “Di b” angle 150° and simila ly
o he o he pai s o he angles. Bu his is no ue. I is
caused by he e e ence ze o p essu e in b anch “a”.
I would be symme ical in case when all p essu es will
be ela ed o he ze o p essu e p(b) in case o low
con igu a ion “Di b”. This is he case whe e i is
possible o see he in luence o he absolu e p essu e in
b anches.
Fig. 8. The

(PG) coe icien s o low con igu a ion “Di b” as a
unc ion o low a e a io q
(ca)
.
Fig. 9. The

(MG)i coe icien s o low con igu a ion “Di b” as
a unc ion o low a e a io q
(ca)
.
5.3. Flow con igu a ion “Di c”
The a iable o he pipe junc ion coe icien s e alua ion
is low a e a io q(ac) in his case
)c(
)a(
)ac( Q
Q
q
.
(15)
Fig. 10. The

(PG) coe icien s o low con igu a ion “Di c” as
a unc ion o low a e a io q
(ca)
.
01112-p.4
EFM 2012
The esul s o he powe geome y coe icien look a he
easonable.
Fig. 11. The

(MG)i coe icien s o low con igu a ion “Di c”
as a unc ion o low a e a io q
(ca)
.
This low con igu a ion was e y di icul o measu e,
because he low was e y uns able o he high alues o
q(ac).
The cu es o he pai s o he angles (30°-150°, 45°-135°
and so on) a e no symme ical abou he ze o alue o
momen um coe icien . The eason is he same as in he
case o low con igu a ion “Di a” and “Di b”.
6 Cha ac e is ics o pipe junc ion as a
polynomial unc ions
Cha ac e is ics o he pipe junc ion as a polynomial
unc ions will be p esen ed in his chap e . The
polynomial unc ions we e d awn o app oxima e he
measu ed alues. The gene al o m o he polynomial
unc ion is as ollow


N
i
i
)xy(i qA
0

.
(16)
The quan i y q(xy) is he low a e a io ela ed o he
pa icula low con igu a ion.
Table 1. Polynom coe icien s o he coe icien

(PG) o he
low combina ion “Di a”. Va iable is q
(ca)
Agl. A
(0)
A
(1)
A
(2)
A
(3)
A
(4)
30° 0.03996 0.21542 - 1.98045 5.96808
45° 0.00494 0.69511 - 2.64166 5.30444
60° - 0.03018 0.87583 - 2.18938 4.62091
75° - 0.02644 0.54269 0.22227 2.77669
90° - 0.01502 0.72022 - 0.50585 3.34952
105° - 0.04975 0.13346 2.99545 2.13002
120° - 0.04764 - 0.11965 4.90077 1.62379
135° - 0.05694 0.02440 5.34726 2.24393
150° - 0.04610 0.77084 3.13437 5.61312
Table 2. Polynom coe icien s o he coe icien

(MG)1 o he
low combina ion Di a. Va iable is q
(ca)
Agl. A
(0)
A
(1)
A
(2)
A
(3)
A
(4)
30° 0.06386 - 0.03327 0.50067 -0.30472
45° 0.04349 0.30411 0.01003 -0.17258
60° 0.01982 0.52247 - 0.15851 -0.07909
75° 0.00073 0.67438 0.13419 -0.31179
90° - 0.01433 0.77718 0.18345 -0.27213
105° - 0.04652 0.78045 0.14711 -0.15718
120° - 0.06140 0.90204 - 0.22418 0.08071
135° - 0.07597 0.92531 - 0.75895 0.43393
150° - 0.06522 0.59853 - 0.65970 0.26504
Table 3. Polynom coe icien s o he coe icien

(PG) o he
low combina ion “Di b”. Va iable is q
(cb)
Agl. A
(0)
A
(1)
A
(2)
A
(3)
A
(4)
30° - 0.04610 0.77084 3.13437 5.61312
45° - 0.05694 0.02440 5.34726 2.24393
60° - 0.04764 - 0.11965 4.90077 1.62379
75° - 0.04975 0.13346 2.99545 2.13002
90° - 0.01502 0.72022 - 0.50585 3.34952
105° - 0.02644 0.54269 0.22227 2.77669
120° - 0.03018 0.87583 2.18938 4.62091
135° 0.00494 0.69511 - 2.64166 5.30444
150° 0.03996 0.21542 - 1.98045 5.96808
Table 4. Polynom coe icien s o he coe icien

(MG)1 o he
low combina ion “Di b”. Va iable is q
(cb)
Ang. A
(0)
A
(1)
A
(2)
A
(3)
A
(4)
30° - 0.02744 - 0.15783 0.48191 -0.35726
45° 0.00565 - 0.55647 0.51469 -0.42874
60° 0.01008 - 0.61420 - 0.02193 -0.03182
75° 0.01988 - 0.62134 - 0.28020 0.18022
90° 0.01433 - 0.77718 - 0.18345 0.27213
105° 0.02839 - 0.83435 0.00595 0.28218
120° 0.03744 - 0.83957 0.41874 0.03667
135° 0.04040 - 0.73097 0.34852 0.10066
150° 0.04242 - 0.49984 - 0.07477 0.23415
Table 5. Polynom coe icien s o he coe icien

(PG) o he
low combina ion “Di c”. Va iable is q
(ac)
Ang. A
(0)
A
(1)
A
(2)
A
(3)
A
(4)
30° 6.7239 1.10915 - 19.578 24.26525 -7.34314
45° 7.4166 10.5006 - 81.598 118.1849 -51.384
60° 8.475 3.53416 - 49.05 71.86643 -31.063
75° 4.8848 18.66 - 76.133 108.396 -52.067
90° 4.7174 18.2622 - 57.23 76.319 -37.769
105° 3.7407 16.6857 -63.345 99.87026 -52.067
120° 3.76249 3.21864 - 19.829 52.38555 -31.063
135° 3.12062 3.67499 - 35.345 87.34959 -51.384
150° 2.95850 - 3.1574 9.15858 5.10732 -7.3431
Table 6. Polynom coe icien s o he coe icien

(MG)1 o he
low combina ion “Di c”. Va iable is q
(ac)
Ang A
(0)
A
(1)
A
(2)
A
(3)
A
(4)
30° 0.45341 - 1.165 2.17698 1.82076 -2.9947
45° 0.05468 - 2.706 13.51148 - 17.397 6.70083
60° - 0.184 - 0.857 7.98778 - 12.606 5.75276
75° 0.00097 - 0.2549 0.33661 -0.02042
90° - 0.0566 0.07725 0.03645
105° - 0.1144 0.59046 -0.30261
120° - 0.0394 - 0.6178 2.42304 -1.10727
135° - 0.0125 - 0.1635 - 2.57935 10.44418 -6.9974
150° 0.04895 - 1.9576 7.45342 - 6.42768 1.49592
7 Conclusion
The cha ac e is ics o he pipe junc ion o low di ision wi h
he i e di e en angles o he adjacen b anch we e p esen ed
in his a icle. All cha ac e is ics we e de i ed calcula ed unde
he nex assump ions
 S eady luid low
 The g a i y ec o is pe pendicula o he pipe
junc ion plain.
 The densi y is cons an
 The Co iolis numbe s and Bousineques
numbe s on he c oss-sec ions S(a), S(b), and S(c)
a e equal o 1.
01112-p.5

EPJ Web o Con e ences
 The e e ence p essu e is p(a)=0
The discussion abou he pipe junc ion coe icien s and hei
explana ion was also included in his pape . The in e es ing
esul s was ob ained o he minimum hyd aulic losses o low
con igu a ion “Di a” and “Di b”. The coe icien s o
polynomial unc ions o all measu ed cha ac e is ics we e
lis ed a he end o he pape .
8 Acknowledgemen
The au ho s a e g a e ul o he G an Agency o Czech
Republic (GACR) o unding his esea ch unde p ojec
“Ma hema ical and Nume ical Modeling o Flow in Pipe
Junc ion and i s Compa ison wi h Expe imen ”.
Regis a ion numbe o his p ojec is 101/09/1539.
9 Re e ences
1. D. S. Mille , In e nal low, A guide o losses in pipe
and duc sys ems,
2. K. Oka, H. I o. J Fluid Eng. 127, 110, (2005)
3. J. Š igle , J. Mech. Eng. 5,249-262, (2006)
4. J. Š igle , J. Mech. Eng. 5,263-270, (2006)
5. J. Š igle , IOP Con . Se .:Ea h En i on. Sci., 12,
12101, (2010)
6. J. Š igle , Scien i ic Bule in o he “Poli echnica”
Uni e si y o Timisoa a, Romania T ansac ions on
Mechanics, 6, 83-92, (2007)
7. L. Beneš, P. Louda, R. Kesle o á, J. Š igle , J. Amc.,
04,074, (2011), (In p ess)
8. P. Louda, K. Kozel, J. Příhoda, L. Benes, J.
Comp luid., 12, 003, (2010)
9. M. Ko ek, V. Kopecký, D. Jašíko á, Recen
Resea ches in Mechanics, P oc. o he 2nd
In e na ional Con e ence on Fluid Mechanics and
Hea and Mass T ans e 2011 (FLUIDSHEAT11)
WSEAS Co u,169-172, (2011)
10. J. Š igle , R. Klas, M. Ko ek, V. Kopecký, J.
P oeng., 39, 19-27, (2012)
11. J. Š igle , O. Špe ka, R. Klas, Eng. Mech. 2011,
607-610, (2011)
01112-p.6