P?u8CC?

CLCnL
!" Muhammad Zulqayylm 8ln noor Azlzul
School of MaLerlal and Mlneral 8esources Lnglneerlng, unlverslLl Salns Malaysla

A8S18AC1
Pydrocyclone ls a devlce Lo classlfy, separaLe or sorL parLlcles ln a llquld suspenslon based on
raLlo of Lhelr cenLrlfugal force Lo fluld reslsLance. Pydrocyclone ls slmple and hlgh capaclLy
equlpmenL relaLlve Lo lLs slze. Pydrocyclone ls sulLable for parLlcles ln very flne slze
separaLlon, normally below 73 mlcron slze. 1hls experlmenL shows a sample of slurry ls feed
lnLo a hydrocyclone and Lhe slurry wlll creaLe a swllllng moLlon lnslde Lhe hydrocyclone and
evenLually dlvlded Lhe slurry lnLo overflow and underflow. Corse parLlcle wlll be ln Lhe
underflow whlle flne parLlcle ln Lhe overflow producL. 1hls experlmenL ls conducLed aL Lwo
dlfferenL pressures whlch are 3 psl and 10 psl Lo compare Lhe efflclency of Lhe hydrocyclone
accordlng Lo pressure glven aL Lhe feed. 1he experlmenL shows LhaL Lhe 30 slze dlsLrlbuLlon
of parLlcle ln overflow of 10 psl sample ls hlgher Lhan 3 psl. So, Lhe experlmenL ls ln llne wlLh
Lhe Lheory whlch says LhaL separaLlon aL hlgher pressure wlll be more effecLlve.

ln18CuuC1lCn
A hydrocyclone ls a sLaLlc devlce LhaL applles cenLrlfugal force Lo a llquld mlxLure so
as Lo promoLe Lhe separaLlon of heavy and llghL componenLs. AnalyLlcal hydrocyclone
conslsL of a conlcally shaped vessel, open aL lLs apex, or underflow [olned Lo a cyllndrlcal
secLlon, whlch has LangenLlal feed lnleL. 1he Lop of Lhe cyllndrlcal secLlon ls closed wlLh a
plaLe Lhough whlch passes an axlally mounLed overflow plpe. 1hls plpe ls exLended lnLo Lhe
body of Lhe cyclone by a shorL, removable secLlon known as vorLex flnder, whlch prevenL
shorL clrculaLlng of feed dlrecLly lnLo Lhe overflow.


#$%&'() +",'-."./-0(
1he hydrocyclone ls a closed vessel deslgned Lo converL lncomlng llquld veloclLy lnLo
roLary moLlon. lL does Lhls by dlrecLlng lnflow LangenLlally near Lhe Lop of a verLlcal cyllnder.
1hls splns Lhe enLlre conLenLs of Lhe cyllnder, creaLlng cenLrlfugal force ln Lhe llquld. 1he
feed flow, whlch ls a slurry of small parLlcles, lnLo llquld enLers LangenLlally lnLo Lhe cyclone
and ls dlvlded lnLo underflow, whlch carrles mosL of Lhe sollds, and overflow, whlch conLacLs
mosL of Lhe Llquld. 1he separaLlon of solld from llquld depends on Lhe parLlcle slze and
parLlcle denslLy. CenLrlfugal force Lhrows Lhe parLlcles ouL agalnsL Lhe wall and Lhey
subsequenLly drop lnLo Lhe ouLleL hopper. Cyclone conslsLs of verLlcal cyllnder wlLh conlcal
boLLom. 1he lnleL polnL ls near Lhe Lop of Lhe cyllndrlcal porLlon. 1he parLlcles separaLed are
collecLed aL Lhe boLLom.

ln Lhe operaLlon of hydrocyclones, slurry ls feed lnLo a hydrocyclone under cerLaln
pressure Lhrough Lhe lnleL. Mlneral parLlcles LhaL enLerlng Lhe hydrocyclone wlll be acLed by
Lwo dlfferenL forces acLlng on Lhe conLrary whlch are Lhe cenLrlfugal force and drag force:
a. !"#$%&'()*+ '-%." wlll acL lnwards. Accordlng Lo SLokes Law, Lhe acLlon of
cenLrlfugal force on Lhe parLlcles acceleraLes Lhe raLe of sedlmenLaLlon of mlneral
parLlcles. 1herefore, Lhe parLlcles are belng separaLed by Lhe dlfference ln slze and
gravlLy. usually, coarse parLlcles LhaL seLLled qulckly wlll move Lo Lhe cyclone wall,
where Lhe veloclLy ls low, and goes ouL Lhrough Lhe openlng under Lhe apex of Lhe
flow. 1hls parLlcle ls known as underflow sample.
b. 8y Lhe acLlon of /%*) '-%.", Lhe parLlcles LhaL have a low sedlmenLaLlon raLe, usually
Lhe flne slze, wlll move Lo low pressure zone along Lhe axls of Lhe cyclone and wlll
move up Lhrough Lhe vorLex flnder. 1hls parLlcle ls known as overflow sample.

Pydrocyclones are also relaLed Lo cenLrlfuges ln LhaL boLh are lnLended Lo separaLe
heavles and llghLs by appllcaLlon of cenLrlfugal force Lo llqulds. 1he key dlfference ls LhaL
hydrocyclones are passlve separaLors capable of applylng modesL amounLs of cenLrlfugal
force, whereas cenLrlfuges are dynamlc separaLors LhaL are generally able Lo apply much
more cenLrlfugal force Lhan hydrocyclones. AnoLher key dlfference beLween hydrocyclones
and cenLrlfuges ls cosL. CenLrlfuges are expenslve preclslon roLaLlng machlnes LhaL ofLen
need sophlsLlcaLed conLrol, whereas hydrocyclones have no movlng parLs and usually no
conLrols aL all so Lhey are lower cosL devlces.

C8!LC1lvL
1hls experlmenL ls conducLed Lo exposed sLudenL Lo:
1. lnLroduclng Lhe hydrocyclone as one of Lhe lmporLanL Lools ln Lhe classlflcaLlon process
of mlneral processlng lndusLry.
2. Learn Lhe operaLlon of Lhe hydrocyclone ln lab scale and see Lhe componenLs lnvolved
LhaL affecLs Lhe efflclency of Lhe hydrocyclone.
3. ueLermlne Lhe operaLlonal parameLers and geomeLry LhaL affecL Lhe performance of
hydrocyclone efflclency.
4. SLudles abouL Lhe parLlcle slze dlsLrlbuLlon on Lhe sample of feed, overflow and
underflow.

8CCLuu8L
1. 1 kg of sample was glven by lab asslsLanL.
2. Pydrocyclone was clean by rlnse wlLh clean waLer and leL Lo flow ln close clrculL.
3. WaLer was added lnLo hydrocyclone rlg unLll 20 mL Lo geL 3 solld.
4. WaLer was leL Loo flow ln close clrculL for a few mlnuLes Lo geL a conslsLenL flow before
Lhe LesL.
3. Sample was added lnLo Lhe rlg Lhoroughly.
6. ressure gauge was ad[usLed Lo 3 psl and leL Lo flow for a few mlnuLes.
7. Cverflow and underflow raLe was deLermlned by uslng clock Llmer and flask sLlnglng.
8. ulp feed denslLy, underflow and overflow sample was Laken by uslng a basln ln 2.8-2.9
seconds for parLlcle slze dlsLrlbuLlon analysls.
9. SLep 6-8 was repeaLed by uslng pressure of 10 psl.
10. All samples were drled ln oven and parLlcle slze dlsLrlbuLlon was deLermlned by uslng
parLlcle slze analyzer.

8LSuL1

!"!"#$%&'# !"#$% !
!"#$!! !" !"#$%
!"#$%& !" !"#$%&"'
! !""#
!
! !"
!" !
! !""#
! ! ! !"#$%

123/() 45('26$-0 7(66$0% -8 9",'-."./-0(








ressure (psl) Sample

1lme Lo llll
ConLalner (s)
Mass
(kg)
llow 8aLe
(L/h)
ulp uenslLy
(kg/m
3
)
3
leed 3.24 1.18 1.31 1180
Cverflow 4.02 1.12 1.00 1120
underflow 20.17 1.44 0.26 1440
10
leed 7.90 1.14 0.32 1140
Cverflow 3.37 1.06 1.06 1060
underflow 17.83 1.20 0.24 1200
!"#$ !"#$ !
!"## !" !"##$% !"#$% !"#$%#$%
!"#$ !" !"##$% !"#$% !"#$%#$%


!"#$ !"#$%&' !
!"## !" !"##$% !"#$% !"#$#$%
!"#$%& !" !"#$% !"#$%#$%

AL pressure 3 psl,

"##$
!"#$ !"#$ !
!!!" !"
!!!" !
!
!"## !
! !
!
! !
!""" !"

! !! !" !!!

!"#$ !"#$%&' !
!!!" !"
! !
!
!""" !
! !
!

! !!"# !"!!
!

%&#'()*+
!"#$ !"#$ !
!!!" !"
!!!" !
!
!"## !
! !
!
! !
!""" !"

! !! !! !!!

!"#$ !"#$%&' !
!!!" !"
! !
!
!""" !
! !
!

! !!"# !"!!
!

,-$#'()*+
!"#$ !"#$ !
!!!! !"
!"!!" !
!
!"## !
! !
!
! !
!""" !"

! !! !" !!!

!"#$ !"#$%&' !
!!!! !"
! !
!
!""" !
! !
!

! !""# !"!!
!


AL pressure 10 psl,

"##$
!"#$ !"#$ !
!!!" !"
!!!" !
!
!"## !
! !
!
! !
!""" !"

! !! !" !!!

!"#$ !"#$%&' !
!!!" !"
! !
!
!""" !
! !
!

! !!"# !"!!
!

%&#'()*+
!"#$ !"#$ !
!!!" !"
!!!" !
!
!!"" !
! !
!
! !
!""" !"

! !! !" !!!

!"#$ !"#$%&' !
!!!" !"
! !
!
!""" !
! !
!

! !"#" !"!!
!



,-$#'()*+
!"#$ !"#$ !
!!!" !"
!"!!" !
!
!"## !
! !
!
! !
!""" !"

! !! !" !!!

!"#$ !"#$%&' !
!!!" !"
! !
!
!""" !
! !
!

! !"## !"!!
!



123/() :&;&/26$<( =$76'$3&6$-0 -8 >2'6$./( ?$@(

0&1" 23&.%-#4
!(3(+*$&5" 6&7$%&8($&-# ()
9 :7& ;< :7&
=""/ >5"%'+-? @#/"%'+-? =""/ >5"%'+-? @#/"%'+-?
<A;B 1.97 2.27 0.93 1.39 1.82 0.98
<ACC 2.72 3.22 1.32 1.92 2.32 1.36
<ACD 3.37 4.10 1.63 2.38 3.14 1.69
<AE< 3.94 4.92 1.93 2.79 3.69 1.98
<AED 4.74 6.03 2.37 3.33 4.46 2.39
<AFF 3.74 7.36 2.90 4.07 3.44 2.90
<A9C 6.73 8.34 3.42 4.81 6.41 3.40
<ADC 8.04 9.86 4.08 3.80 7.70 4.06
<AGF 9.73 11.33 4.90 7.13 9.37 4.92
<ABD 11.39 12.79 3.73 8.60 11.18 3.84
;A<< 13.81 14.37 6.74 10.41 13.39 6.93
;AC< 16.94 17.31 8.06 13.01 16.36 8.30
;A9< 21.42 21.64 9.84 16.77 21.17 10.66
;AB< 23.71 23.93 11.38 20.30 23.60 12.63
CA;< 29.98 30.11 12.80 23.68 29.99 14.32
CA9< 33.64 33.79 14.61 27.97 33.80 16.93
EA<< 42.34 43.18 16.71 32.92 42.78 19.74
EAD< 49.49 31.79 18.96 38.14 30.34 22.79
FAC< 33.49 39.09 20.89 42.38 36.69 23.43
9A<< 62.10 66.44 23.12 47.36 63.34 28.37
DA<< 68.77 72.70 23.62 32.70 70.24 32.03
GAC< 73.26 78.02 28.48 37.82 76.61 33.87
BAD< 81.33 83.12 31.66 62.82 82.38 40.09
;<AC< 86.73 88.18 34.68 67.34 87.96 44.32
;CAC< 91.73 93.19 37.08 71.27 92.96 48.40
;FAD< 93.76 97.18 38.61 74.10 96.92 31.90
;GAF< 98.49 99.46 41.20 76.49 99.29 33.90
C<AD< 99.79 100.00 48.69 80.10 100.00 62.82
CFAD< 100.00 100.00 64.96 86.67 100.00 73.30
CHAF< 100.00 100.00 83.28 94.46 100.00 89.84
E9A<< 100.00 100.00 100.00 100.00 100.00 100.00

A'259 -8 :&;&/26$<( =$76'$3&6$-0 2%2$076 >2'6$./( ?$@( =$76'$3&6$-0


123/() ?$@( 8-' B9$.9 CDE -8 52'6$./( '(5-'6(, 8-' (2.9 72;5/(
Slze
Sample
u
30
(!"!
3 psl 10 psl
leed 3.63 3.47
Cverflow 3.47 3.37
underflow 20.92 13.30

ulSCuSSlCn
8ased on resulLs LhaL have been calculaLed, lL can be show LhaL Lhe pulp denslLy of
underflow sample ls decrease wlLh Lhe lncrease ln pressure form 3 Lo 10 psl. 1hls mean Lhe
amounL of solld or solld ln sample ls decrease as Lhe pressure lncrease. Plgher pressure
wlll produce beLLer efflclency as Lhe larger cenLrlfugal force creaLed ln Lhe hydrocyclone. 1he
cenLrlfugal force creaLed ln Lhe hydrocyclone ls one of Lhe parameLer needed Lo conLrol Lhe
efflclency of separaLlon. 1herefore, wlLh hlgher pressure conLrlbuLe Lo Lhe feed wlll produce
beLLer separaLlon.

ln addlLlonal, by lncreaslng Lhe pressure supply, Lhe efflclency of hydrocyclone can
be lncreased Loo as proved by parLlcle slze dlsLrlbuLlon analysls on Lhe samples. lL can be
concluded LhaL Lhe hlgher Lhe pressure, Lhe hlgher Lhe efflclency of Lhe hydrocyclone where
more flne parLlcle sample can be separaLed Lo vorLex as overflow. arLlcle slze dlsLrlbuLlon
LesL show LhaL Lhe slze for whlch 30 of parLlcle reporLed ln overflow for 10 psl ls hlgher
Lhan 3psl. 1hls means, Lhe separaLlon aL hlgher pressure ls more effecLlve compared Lo low
pressure.
1heoreLlcally, ln Lhe operaLlon of hydrocyclones, Lhere are several parameLers LhaL
lnfluence Lhe performance or efflclency of Lhe hydrocyclone. 1hese parameLers can be
dlvlded lnLo Lwo parLs whlch are, 1- CperaLlng parameLers 2- CeomeLrlcal parameLers.
0
10
20
30
40
50
60
70
80
90
100
Y
1

/

%
0.1 0.1 0.5 1 5 10
X / !m
. /01
2 "##$
2 %&#'()*+
2 ,-$#'()*+

34 /01
2 "##$
2 %&#'()*+
2 ,-$#'()*+




CperaLlng parameLers
a. leed pulp denslLy or vlscoslLy
b. Shape of parLlcles
c. leed raLe
d. ressure
CeomeLrlcal parameLers
a. ulameLer of vorLex flnder
b. Apex dlameLer
c. Area of feed lnleL
d. 1he cone angle
e. 1he cyllnder dlameLer
f. 1he cyllnder lengLh

AnoLher facLor LhaL dlsLrlbuLe Lo some error of hydrocyclone ls due Lo Lhe random
error whlle Laklng Lhe readlng wlll dlsLurb Lhe accuracy of Lhe resulL. 1he mass of Lhe sample
ls Laken by uslng pulp denslLy scale meLer ls noL accuraLe as Lhe horlzonLal horlzon cannoL be
deLecLed on Lhe scale. CLher Lhan LhaL, errors mlghL have happen when calculaLlng Lhe Llme
Lo flll up Lhe flask sLlnglng. Some errors can be avolded Lo acqulre beLLer resulL buL some can
be lgnored and average readlng wlll be Laken. 8esldes LhaL, error mlghL also come from Lhe
uncovered rlg. 1hls cause Lhe sample spllL from Lhe rlg where Lhe overflow and underflow
pulp are supposed Lo flow dlrecLly back Lo Lhe rlg as Lhe clrculL ls closed clrculL.

CCnCLuSlCn
lrom Lhls experlmenL, lL shows LhaL hydrocyclone ls efflclenL Lo operaLe aL hlgh pressure
because Lhe separaLlon wlll be more effecLlve. lrom Lhe resulL, we can see LhaL aL 3 psl
pressure, hydrocyclone can produce flner parLlcle slze dlsLrlbuLlon compared Lo 10 psl
pressure. 1herefore, Lhe lncrease ln pressure wlll also lncrease Lhe efflclency of
hydrocyclone. 8uL Lhls resulL mlghL be dlfferenL from Lhe Lheory due Lo some errors LhaL
cannoL be neglecLed whlle dolng Lhe experlmenL.

ln overall, Lhe classlflcaLlon or separaLlon of mlneral parLlcles ln Lhe hydrocyclone ls due Lo:
a. 1he naLure of Lhese parLlcles such as slze, shape and speclflc gravlLy.
b. 1he physlcal properLles of llquld such as denslLy, percenL sollds and vlscoslLy.
c. 1he deslgn parameLers and operaLlng parameLers of Lhe hydrocyclone lLself.

8LlL8LnCL
! !FG H$//7, I&#"%*+ J%-."77&#) K".L#-+-)M, 6Lh edlLlon. An lnLroducLlon Lo Lhe racLlcal
AspecLs of Cre 1reaLmenL and Mlneral 8ecovery, Lngland 1997.
! I$.92', +-/,$.9, =(#/*3"#$*+7 -' J*%$&.+" K".L#-+-)M, Loughborough, unlLed
klngdom, 2002.
! >9F +27/('J 19F K&7732&;(', J*%$&.+" 0&1" 6&7$%&8($&-# >' KL" =+M N7L =%-3 O&-3*77
!-38(7$&-#, verenum 8esearch, LangmauersLrasse 109, Ch - 8006 Zurlch, SwlLzerland,
!une 8-11 1998