Theoretical and experimental investigation of die build-up phenomenon in biaxial oriented

Theoretical and experimental investigation of die build-up phenomenon in biaxial oriented
Theoretical and experimental investigation of
die build-up phenomenon in biaxial oriented
polypropylene film production
Wannes Sambaer, BSc
Master Thesis
2008
ABSTRACT (ENGLISH)
In this work, die build-up phenomenon has been investigated theoretically as well as experimentally for two different polypropylene samples used in the biaxial oriented film
production. By the use of digital image analysis, it has been revealed that Sample 1 die
deposit speed is about 1.5 times faster in comparison with Sample 2. Based on the die
drool measurements, rheological study and theoretical analysis it has been suggested that
the main reason for the higher Sample 2 die build-up stability is occurence of slip at the
contamination layer having Newtonian viscosity just about 100 Pa.s, which does not intensively adhere to the die wall surface.
Keywords: Extrusion, die drool phenomenon, die deposit, die lip build up, negative pressure, finite element analysis
ABSTRAKT (ČESKY)
Tato práce se zabýva studiem jevu ‘die build-up’ (spontánní akumulace materiálu na
hraně výstupní štěrbiny při vytlačování) u dvou odlišných typů polypropylenů, které se
používají při výrobě biaxiálně orientovaných fólií. Na základě experimentální a teoretické
analýzy bylo zjištěno, že pokud při toku taveniny polypropylenu dochází k tokem
indukované frakcionaci, jejímž produktem je nízkoviskozní složka mající Newtnoskou
viskozitu kolem 100 Pa.s a současně malou přilnavost ke stěně vytlačovací hlavy, může
na rozhrání obou vrstev dojít ke skluzu a následně k významné redukci nežádoucího jevu
die build-up.
Klíčová slova: Vytlačování, spontánní akumulace materiálu na hraně výstupní štěrbiny
při vytlačování, negativní tlak, metoda konečných prvků
ACKNOWLEDGEMENT
I would like to express my sincere gratitude to all people who supported me during the work on the thesis.
I am especially grateful to my supervisor, prof. Ing. Martin Zatloukal, Ph.D. for
his patience, guidance and support throughout the process of doing the experiments and
writing thesis.
I also want to thank Ing. Katerina Chalopkova, Ph.D. for helping me with the extrusion tests and for her other good advice.
I also want to thank ir. Frederik Desplentere, Ph.D for the assistance of my thesis
in Belgium.
I also want to thank the teachers ir. Hilde Bonte, ir. Luc Monserez and lic. Gwendolyn Rogge to arrange the practical side of the Erasmus project. I also want to thank the
Tomas Bata University for accepting me as Erasmus student and give me the opportunity
to do research with the laboratory material.
Last but not least, I would like to extend my gratitude to my family who support
me for doing this Erasmus project in Zlín.
I agree that the results of my Master Thesis can be used by my supervisor’s decision. I
will be mentioned as a co-author in the case of any publication.
I declare I worked on this Master Thesis by myself and I have mentioned all the used
literature.
Zlín, May 27, 2008
Wannes Sambaer
πάντα ῥεῖ (panta rhei) "everything flows"
Heraclites (c.535-475 B.C.)
TABLE OF CONTENTS
I.
THEORETICAL BACKGROUND ............................................................................ 10
1
FILM CASTING PROCESS ....................................................................................... 11
1.1 BIAXIAL ORIENTED POLYPROPYLENE FILMS ............................................................... 11
1.1.1 BOPP film production ....................................................................................... 11
1.2 SEALING PROCESS ....................................................................................................... 12
1.2.1 Structure of a BOPP film .................................................................................. 12
1.2.2 Sealing the skin layer on the core layer ............................................................ 13
2
DIE BUILD-UP PHENOMENON.............................................................................. 14
2.1 GENERAL .................................................................................................................... 14
2.2 FACTORS HAVING INFLUENCE ON THE DIE DROOL PHENOMENON............................... 17
2.2.1 Die swell and shark skin ................................................................................... 17
2.2.2 Processing conditions (extrusion rate, processing temperature) ...................... 25
2.2.3 Additives (slip agent, antitack, PPA, LMw lubricant) ....................................... 29
2.2.4 Rheology............................................................................................................ 34
2.3 VARIABLES SENSITIVE TO THE DIE DROOL PHENOMENON .......................................... 38
2.3.1 Shear stress ....................................................................................................... 38
2.3.2 Shear elasticity .................................................................................................. 39
2.3.3 Negative pressure .............................................................................................. 40
3
PROBLEM DEFINITION ........................................................................................... 42
4
AIMS OF THE WORK ................................................................................................ 43
II.
EXPERIMENTAL ........................................................................................................ 44
1
EQUIPMENT AND METHODS ................................................................................. 45
1.1.1
1.1.2
1.1.3
Rotational rheometer ARES .............................................................................. 45
Capillary rheometer Rosand RH7-2 ................................................................. 46
Extrusion line for die build-up evaluation ........................................................ 49
III. RESULTS AND DISCUSSION ................................................................................... 53
1
RHEOLOGICAL RESULTS...................................................................................... 54
2
DIE BUILD-UP MEASUREMENTS .......................................................................... 61
2.1.1
3
Digital image analysis of the die build-up phenomenon ................................... 67
THEORETICAL ANALYSIS .................................................................................... 75
CONCLUSION REMARKS ................................................................................................. 83
REFERENCES ....................................................................................................................... 84
LIST OF ABBREVIATIONS................................................................................................ 91
LIST OF FIGURES ............................................................................................................... 92
LIST OF TABLES ................................................................................................................. 95
LIST OF APPENDICES ....................................................................................................... 96 9
IN
NTRODU
UCTION
Durinng polymer extrusion, spontaneous
s
s accumulattion of the m
material at the
t die exit zone can occur. Due to oxygen inn the surrou
unding air and the highh processing
g temperatuure, the mateerial can deegrade due to
t thermal oxidation.
o
A
After
the cerrtain time, this degradaated polymeer start to make
m
the extrudate dirrty, thickneess variationn is increassed which
reeduce opticaal and physiochemical properties of the final product. D
Due to that, the extrusion lines haave to be shhut down, cleaned
c
and
d run again. This unwaanted shut downs
d
are
omenon is
veery expensiive and therrefore any knowledge about this unwanted flow pheno
veery welcom
me. Thereforre, the mainn aim of th
his work is to investiggate theoretiically and
exxperimentallly the die build-up
b
pheenomenon occurring
o
duuring polyppropylene biiaxial film
prroduction.
10
I. THEO
ORETIC
CAL BA
ACKGR
ROUND
D
11
1
FILM CASTING
G PROCE
ESS
1..1 Biaxiaal orienteed polyproopylene fiilms
Biaxiaal Orientedd PolyPropyylene films or shortly BOPP-film
ms are polyp
propylene
fillms that aree stretched in two directions. They are used in a lot of aapplicationss like foot
paackaging inndustry, mulltilayer highh barrier fillms, balanced shrink fi
films, etc. The
T reason
is that these films
f
have better
b
mechhanical and optical
o
propperties than no oriented
d polyproa
i that theree will be a lower perm
is
meability because an
pyylene films. Another advantage
im
mproved Water
W
Vapor Transmission Rate (W
WVTR) whhich is an iimportant and
a highly
deesirable property for foood packaginng applicatiions.
1..1.1
BOPP
P film prod
duction
The BOPP
B
films usually connsist of threee layers whhich are prooduced by continuous
c
cooextrusion technology
t
(1) employying more th
han one extrruder. Usuaally the basee polymer
toogether withh particular additives are
a fed via a feeder intto the singlee screw exttruder and
affter meltingg and mixingg the polym
mer mixture melts are combined
c
inn the flat co
oextrusion
diie (2) to gett multilayer film. In thee consequen
nt step, the film is streetched in lon
ngitudinal
diirection (3) by increasing the speeed of the roles. After this longituudinal stretcching, the
fillm is stretched in trannsversal direection (4) and
a at the ennd of the liine the BOP
PP film is
coollected by rolling
r
it upp (5).
2.
4.
5.
1.
3.
Figuure 1: A typiccal biaxially--oriented pollypropylene film
f
process ffactory layo
out [49]
12
Durinng this proceess, the polyymer in the film undergoes three stages to beecome the
bii-orientated film. In thee first step, the
t polymerr chain degrree of orienntation is verry small, .
Inn the secondd step, the head
h
chainss are stretch
hed in longiitudinal direection and in
i the last
step also the branched chhains are strretched in trransversal direction.
d
Polymer ch
hain
STEP 1
STEP 2
STEP 3
Figuure 2: Changges in the poly
lymer chain orientation
o
d
during
at parrticular step during biaxiall orientationn of PP film.
1..2 Sealin
ng processs
1..2.1
Struccture of a BOPP
B
film
A typpical BOPP film is com
mposed by three layerss as depicteed in Figuree 3. where
coore and skinn layer has thickness15
t
5-40µm and
d 0.4-1µm, respectively
r
y. The core layer
l
consists of homoopolymer with
w additivees to achiev
ve high deggree of orienntation during the biaxxial stretchinng and desiirable stiffneess and optiical properties (for exaample: colorr).
The heat
h seal ressin occurs inn a skin layeer which is consequenttly sealed on the core
laayer by heatting up the layer (hot tack).
t
After that, this skin layer siignificantly improves
thhe BOPP prooperties as seal ability,, good opticcal (for exam
mple: glosss) and surfacce propertiees and it maakes the BO
OPP film priintable.
SKIN LAYER
0,4 to 1
1µm
CORE LA
AYER
15 to 40
0µm
0,4 to 1µm
SKIN LLAYER
Figuure 3: Structuure of a typiccal 3-layer BOPP
B
film
13
1..2.2
Sealin
ng the skin
n layer on th
he core lay
yer
Heat sealing invoolves the inntimate conttact betweeen two semii crystallinee film sur-
faaces [1]. Heat is typicallly applied through
t
seaal bars causiing the meltting of both
h surfaces.
Thhe melted surfaces
s
“w
wet” and coounter- diffu
fuse across the interfacce to entan
ngle molecuules, thus fuusing the tw
wo surfaces. Finally, co
ooling leads to the re-crystallizatiion across
thhe interface.. Generally, this interfface has adeequate entaanglement to perform a singular
likke (simply thicker)
t
struucture. Thiss entire proccess is influuenced by teemperature, time and
prressure and occurs in leess than a seecond(see Figure
F
4).
HEAT
PRESSURE
T
TIME
COOL
TIME
Surface Regio
on of Crystalline Copolymer
Melted Surfacce
“Wettted” Surface
Diffusio
on and Entangleements
R
Recrystallization
Fiigure 4: Scheematic simpllification of the
t heat sealiing process and
a the postuulated moleccular
prrocesses invoolved [1]
14
2
DIE BU
UILD-UP
P PHENO
OMENON
N
2..1 Generral
Die build-up (alsso called as die drool, plate-out,
p
d build up,, die drip orr die peel)
die
is a phenomeenon occurrring in mellt extrusion
n of polyoleefin’s, PVC
C, or filled polymers,
w
which
manifeests itself ass an undesirrable build--up of material, normallly on the liip or open
faaces of extruusion dies.
In com
mmercial exxtrusion proocesses (e.g
g. blown or cast
c film, B
BOPP, etc.), die depositt can have a significannt influencee on the prroductivity, as it requirres to shut down the
prrocessing linne periodiccally to cleaan the die. Furthermore
F
e, die depossit can also affect the
quuality of thee extruded product.
p
Foor example, in film bloowing, wherre the die deposit
d
occuurs on the lip of the annular
a
diee, the film quality
q
willl deteriorate when thee build-up
reeaches a levvel at whichh it is in coontact with the molten film. Thenn it causes transverse
t
linnes, or conttinually breaaks off the die
d and emb
beds itself inn the film.
Die Deposit
Figuure 5: Extrusion of LLDP
PE-A from a pipe
p die withh die lip buildd-up indicateed [2].
15
Die build‐up
a.
b.
Figuure 6: Extrussion of high
h density polyyethylene. 6aa) a clean diee face, 6b) diie lip
buildd-up occurreence [8]
Die build‐up
a.
b.
Figuure 7: Extrusion of polyviinyl chloridee.7a) a clean slit die face,, 7(b) die lip build up
occuurrence [8]
Figurre 8: Die Buuild-Up during PP capill-
Figure 9:
9 Die Build--Up in Nylon
n [50]. Left
lary extrusion
e
[455]. Left sidee – clean diie
side – cllean die facee, right hand
d side – die
face, right hand side – die lip build-up
p
lip buildd-up occurreence.
occurrrence.
16
Theree are many different faactors havin
ng an influence on the die lip builld-up phenoomenon. Soome of them
m are summaarized in Fig
gure 10..
Visccoelasticity
Phyysical factorrs
Rheologiccal considerattions Shearr rate
Pressure fluctuationss
Exit angle
Rough
hness
Pressure drop
p
Die d
design
Land lengtth
DIE DRO
OOL
PH
HENOMEENON
Die sweell
Kind of streesses
Dissim ilar n blends
viscosities in
Unsaaturation
Chemical faactors
Re
esin morpho
ology
L
LMw‐species
Poor disp
persion of pigm
ments
Branched / LLiniear
Other
Thermal staability
Enviironmental factors
Fillers
Additives
S
Slip agents
Antio
oxidants
PPA
A’s
TThermal conssiderations
Oxyggen
Moiisture
Melt temperaturee sensitivity
Fiigure 10: Sum
mmarizationn of all possibble factors leeading to diee build-up phhenomenon [15,
[1 20]
17
2..2 Factors havingg influence on the die
d drool phenome
p
non
In thiss section, thhe main desstabilization
n factors witth respect too die drool phenomep
noon are discuussed in morre detail.
2..2.1
Die sw
well and sh
hark skin
2.2.1.1 Die swell
It has been demoonstrated byy A.C.-Y. Wong
W
[53] thhat the tempperature, thee load and
thhe lenth: diaameter (L:D
D) ratio of a capillary die
d are the primary
p
extternal factors that affeect both the melt flow rate and the extrudate swell ratioo measured from a mellt flow indeexer.
The extrudate
e
sw
well ratio (deefined as as
a die diameeter devidedd by extrudaate diameteer) has beenn found to be
b melt flow
w rate depeendent [53].. The moleccular long-cchain side
brranches werre demonstrrated to be an importan
nt moleculaar property affecting th
he L:D effeect on the exxtrudate sw
well ratio. It has been sh
hown that thhe extrudatte swell ratio of a polyymer havingg highlong-cchain side branches
b
co
ontent, is more
m
sensitivveto the cap
pillary die
L::D ratiothann a polymer having low
w side-chain
n branches content.
c
H.W. Müllner ett al [52] havve investigaated die sweell phenomeenon for rub
bber compoounds by ussing differed L/D capilllary die rattios and sheear rates (seee Figure 11).
1 From
thhis figure, itt can be seeen that the shear
s
strain rate increasse leads to hhigher levell of extrudaate swell.. According
A
t Han [55]], this behaavior is attriibuted to thhe recoverab
to
ble elastic
ennergy increaase.
18
Swell value
L/D = 40
0/2
L/D = 20
0/2
L/D = 10
0/2
o
‐1
Shear sttrain rate γ [s ]
Figuure 11:The eff
ffect die desig
ign and shearr strain rate on the rubbeer compound
d swell
valuee [52] ( die diameterD
d
= 2mm, die teemperature equal
e
to 120°°C.)
19
Shuicchi Tanoue et
e al [56] haas shown by
y number off numericall analyses th
hat the die
sw
well will increase if thhe Weissenbberg numbeer (
) also inncreases as shown in
2
2
1
1
0
r/Ro
r/Ro
Fiigure 12.
a. We = 2
1
d. We = 100
1
2
2
‐1
0
1
2
z/Ro
3
4
5
‐1
2
2
1
1
0
b. We = 5
r/Ro
r/Ro
0
0
1
2
z/Ro
3
4
5
0
e. We = 400
1
1
2
2
‐1
0
1
2
z/Ro
3
4
5
‐1
0
1
2
z/Ro
3
4
5
2
r/Ro
1
0
c. We = 10
1
2
‐1
0
1
2
z/Ro
3
4
5
Figuure 12: Theooretically preedicted effecct of Weissennberg numbeer, We, on th
he annular
extruudate swell (in
( terms of streamline
s
fieeld) [56]
20
Hamielec and Vllachopouloss [57] conclluded that extrudate
e
sw
well is causeed by two
foollowingreassons: memoory of entraance (i.e. an
n imaginaryy cylinder ffluid elemen
nt tries to
reegain its origginal shapee when it is emerging th
he capillaryydie as it is shown in Figure
F
13)
annd elastic recovery
r
(normal stressses releasee). When thhe tube is ssufficiently long, the
m
memory
of enntrance fadees completeely.
Reservoir
Extrud
date
Die
Figuure 13: Schem
matic representation of the
t sequence of deformattions of a ma
aterial as it
enterrs, flows trouugh, and emeerges from a die [57]
21
Chai, Adams and Frame [22] have foun
nd that therre is a relattionship between die
sw
well and thee die build--up by doinng experimeent on LLD
DPE. Figuree 14 show
ws average
exxtrudate sweell together with die deeposit level as the functtion of masss flow rate. It is nicelyy visible thaat both die swell and die depositt increase with
w increassing extrusiion output
exxcept that thhe former, distinct
d
from
m die depossit, does nott exhibit a m
maximum value
v
with
exxtrusion ratees.. This unnexpected behavior
b
hass been explaained by noo possibility
y to effectivvely collectt die drooledd material from
f
the diee lip at highh mass flow rate becausse the mateerial is remooving from the die lip at this casee, which maake the die ddeposit meaasurement
prractically im
mpossible. Thus
T
it appeears that diee swell does have a strrong influen
nce on die
lipp build-up.
1.55
30
1.50
Die Deposiit
20
Die Swell [D/Do]
Die Depostit [mg]
Die
e Swell
1.45
10
1.40
0
15
25
35
4
45
55
Exxtruder outp
put [kg/h]
Figuure 14: Extruusion of LLDP
PE from cap
pillary die, shhowing die swell and die deposit as
a funnction of extrruder output [2]
This conclusion
c
has also beeen supporrted by I. Klein
K
[3]. thhat die sweell has the
laargest influeence on die lip
l build-upp.
Promootion role of
o the die sw
well phenom
menon on thhe die buildd-up can be explained
byy the better contact between die exxit wall surfface and exttrudate [4].. However, it appears
thhat this becoomes less faavorable aboove a criticaal (very highh) shear ratte or stress, where the
loow moleculaar weight species
s
mayy no longerr be in the correct
c
geoometry for deposition
d
(i.e., not wettting out) leeading to a decrease in
n observed die
d deposit, even thoug
gh the extruudate increaases, with extrusion ouutput.
22
2.2.1.2 Shaarkskin
Kurtzz [5] explaiined shark skin pheno
omenon by the relaxaation of stored strain
ennergy of polymer layerr closest to the die walll when mellt emerges nnear die lip
p (see Figurre 15). The die lip is a region of critical
c
strettching whenn the melt iimmediately
y changes
from a flow confined
c
byy no slip bouundary to a free boundaary upon exxiting die.
Die
Tansient velocity profile Velocity ofile pro
upsttream
Equilibriu
um velocityy profile Area of A
c
critical strretching
Figuure 15: Flow in the viciniity die exit ba
ased on [9]
Migleer et al [48]] have concluded that sharkskin
s
can occur inn three possible flow
coonditions: no
n slip, weaak slip or sttrong slip. They
T
have found
f
that ccohesive an
nd surface
faailures are the
t main driving mechhanisms for shark skin onset . In m
more detaill, in their
exxperiments, it has beenn revealed thhat under th
he unstable case,
c
the fraacture surfacce divides
thhe material into
i
two reggions, a surrface layer and
a a core region.
r
Thee surface lay
yer bulges
uppwards. Theerefore, the splitting off the polym
mer into coree and surfacce layers an
nd the discoontinuity inn velocity att that interfface have been
b
found to be the pprimary reassons for a
coohesive failuure.
23
Figuure 16: mLLD
DPE extrudaate surface ap
ppearance cllassification at different flow
f
conditioons. (a) Smoooth string, (b)) Slight sharrk skin, (c) Shhark skin. [228]
a.
Die
b.
Acceleration of the polyymer Æ stressses increasee
c.
Surfacce (Arrow No.1)
Cohesive failure (Arrow No.2)
d.
Second faailure
Figuure 17: Sketch of the sharrkskin instability kinetics,, side view [4
[48]
A sidde sketch viiew of sugggested mech
hanism for shark skin developmeents is depiicted in Figgure 17 [48].. Here, thhe velocity rearrangem
ment at the ddie exit reg
gion is depiicted in moore detail. The extenssional stresss at the airr-polymer-ddie wall co
ontact line
buuilds up oveer time untiil the materrial fractures ( see Figuure 17b). Thhe fracture is located
att the air-pollymer surfacce. In Figurre 17c, it iss visible in what way tthe materiall closer to
thhe die face(ssee arrow No.1
N
in Figgure 17c) go
oes toward the surfacee region of the extrudaate while thhe surroundiing materiaal goes towaard the coree region (see arrow No
o.2 in Figurre 17c). It has
h also beeen revealed that
t the extensional strrain rate is m
much lowerr, comparinng with defoormation ratte stage deppicted in Fig
gure 18a, ass the materiaal flows into the core
24
reegion. Theyy also has foound so callled surfacee failure whhich occurs as the results of surfaace size grow
w followed by its peeliing off from
m the die exiit (see Figurre 17d).
Furtheermore, it has
h been revvealed [28] that the onsset of the diie drool pheenomenon
foor mLLDPE
E may be directly
d
connnected with
h this pronoounced die surface deffect called
shhark skin. On
O the otherr hand, the die
d drool disappears if the shark skkin becomees weak or
diisappears coompletely. This
T suggessts that the rupture of the extrudaate free surfface at the
diie exit regioon significaantly contribbutes to the die drool phenomenon
p
n for mLLD
DPE melt.
A visualizatiion of this connection between th
he drool phhenomenon and the sharkskin is
shhown in figuure 18.
Com
mmon line
Shark skin
Accelerattion of the polyymer particaal
Die depo
osit
Figuure 18: From
m sharkskin too die depositt
D.R. Arda,
A
M.R.. Mackley [39]
[
have done intensivve researchh about sharrkskin and
thhey have foound that shharkskin cann be reduceed or eliminnated by thhe incorporaation of a
fluuoropolymeer based Poolymer Proccessing Aid
ds (PPAs), which
w
migrrate during the extrusion toward the
t die surfaace where very
v
thin lay
yer is createed. Due to thhe fact that fluoropolyymers have low surfacee energy, sllip occurs between
b
polymer melt aand PPA lay
yer which
leeads to veloocity rearraangement/sttretch/sharkskin reducttion at the die exit zo
one. The
m
more
detailedd discussionn about the PPA
P
with reespect to die drool phenomenon iss provided
inn Chapter 2.2.3.3.
25
Recenttly, boron nitride (BN) has been found
fo
as thee new proceessing addittive which
significantly delays the sharkskin onset
o
[37]. The
T authorss have foundd that the addition
a
of
sm
mall amounnt of fine BN
N particles eliminates sharkskin phenomenon
p
n as well ass the criticaal shear ratee for the grooss melt fraccture is posttponed [38].
2..2.2
Proceessing cond
ditions (extrrusion ratee, processin
ng temperature)
Inn this sectionn, the effectt of processsing conditio
ons on the die
d drool phhenomenon will be
innvestigated in
i more detail.
2.2.2.1 Extrrusion ratess
Chai, Adams andd Frame [2] investigateed the effectt of extrusioon output (o
or extruder
RP
PM) on thee die deposit level oveer a 30-min
nute period for LLDPE
E. The resu
ult of their
reesearch is deepicted in Figure
F
19. Firstly,
F
they have foundd that all tessts were reprroducible,
inndependent of the orderr of extrudeer output ch
hanges. They also havee found thatt there is a
crritical extruusion rate att which die deposit reaaches a maxximum leveel and beyo
ond which
diie deposit leevel appearss to decreasse with furth
her increasee in extrusioon output raate until it
diisappears coompletely. This
T suggessts occurren
nce short-terrm [6], fast die lip builld-up phenoomenon occcurring at veery high maass flow rates which haas not been detected by
y the techniique which authors useed in their work.
w
Maximum
Die Deposit [mg]
40
30
20
10
0
5
15
25
35
Extruder output [kgg/h]
45
55
5
Figuure 19: The effect
ef
of the extrusion
e
outp
tput on the diie deposit levvel for LLDP
PE
26
2.2.2.2 Proccessing tem
mperature
The effect
e
of thhe processinng temperatture on thee die build--up phenom
menon has
beeen investiggated in [28]].
TEMPERATURE
L
LOW
90°C
95°C
100°C
C
Flakes
HIGH
105°C
Ring flakes
110°C
Powder
Figuure 20: Die drool
d
appearaance for diffe
ferent wall teemperatures [[28]
It hass been founnd that processing temp
perature hass crucial efffect on the die drool
phhenomenon. In more detail,
d
Hinricchs [32] con
ntrolled thee two melt ttemperatures (die and
fillm temperaature). He reported
r
thaat the melt temperature (too low or too high
h) can incrrease die deeposit. Chappman [33] observed
o
diee drool at high
h
processsing temperatures onlyy. On the othher hand, K.
K Chaloupkkova and M.. Zatloukal [28] found that the link
k between
thhe die drool intensity annd temperatture has non
n-monotonicc trend.
hat there are two types oof the die drrool:
Kurtzz and Szanisszko [5] havve found th
•
Sloow die droool: material which
w
is breeaking awayy from the eextrudate att the die
exiit region, is the source for the slow
w die accum
mulation.
•
Fast die drool: material seeparation frrom the polyymer matrixx along the internal
diee walls causses very fastt drool.
The role of the external
e
die exit surfacee cooling onn the die drrool phenom
menon has
beeen experim
mentally invvestigated by
b K. Chalo
oupkova andd M. Zatlouukal [28]. In
n their reseearch, they have used external
e
fann to cool do
own the die/extrudate ssurface from
m the out-
27
side. They haave found that
t
the exteernal coolin
ng promotess the die drrool occurreence. This
ngements annd higher fr
free surface stretches,
caan be explaiined by inccreased veloocity rearran
w
which
resultss in its ruptuure at the die
d exit zonee. Moreover, it also caauses larger extrudate
sw
well, whichh leads to a better extrrudate surfaace contact with the die promotin
ng the die
drrool occurreence as dem
monstrated Figure
F
21.
High
h extruder ou
utput Î
Big die swell
Low extruder output Î
ell
Small die swe
Eq
quilibrium v
velocity profile Equilibrium
m velocity profile Biggerr risk on die d
deposit
a. No exterrnal cooling
b. High e
external coolingg
Figuure 21: The effect
ef
of die swell
s
on die deposit
d
durinng external ccooling; a. Sm
mall die
swelll, b. Big die swell
Anothher possiblee factor promoting the die drool phenomenon
p
n is the theermal degraadation of thhe polymerr. The convventional mo
odel for theermal degraadation [7] consist of
foour stages which
w
are deescribed belllow in moree detail.
a. Initiaation
The innitiation of thermal deggradation in
nvolves the loss of a hyydrogen atom
m from
the poolymer chaiin (shown below as R.H
H) as a resullt of energyy input from
m heat of
the diie. This creaates a highlyy reactive an
nd unstablee polymer ‘ffree radical’’ (R*) and
a hydrrogen atom
m with an unnpaired electtron (H*). The
T strengthh of the carb
bon - fluorine C-F bond inn the long chain
c
backb
bone of the fluoropolym
f
mers processsed (such
as PT
TFE, FEP, PFA, PVDF,, THV, ETF
FE and ECT
TFE) means that there it is much
harder for thermaal degradatiion to initiatte.
b. Propaagation
The propagation
p
n of thermall degradatio
on can involve a varietyy of reaction
ns and
one of these is where
w
the freee radical (R
R*) reacts with
w an oxyggen (O2) mo
olecule to
28
form a peroxy raadical (ROO
O*) which caan then rem
move a hydroogen atom from
f
peroxide (R
ROOH) and so regeneraate the
anothher polymer chain to forrm a hydrop
free raadical (R*).. The hydrooperoxide caan then splitt into two nnew free rad
dicals,
(RO*) + (*OH), which will continue to
o propagate the reactionn to other po
olymer
moleccules. The process
p
can therefore acccelerate deepending onn how easy it
i is to
removve the hydroogen from the
t polymerr chain.
c. Term
mination
The teermination of thermal degradation
d
n is achievedd by ‘mopping up’ the free radicals too create inert products.. This can occur naturally by combbining free radicals
r
or it can
c be assistted by usingg stabilizerss in the plastic.
A genneral mechaanism for thermal degraadation is shhown below
w:
X = Free radic
X*
cal
Initiatio
on
Propagattion
R.H
Heat
R* + H*
R* + O 2
R . O . O*
R . O . O* + R . H
O + R*
R . O . OH
R . O . OH
R . O . O * + OH*
R* + R*
Terminattion
R* + R . O . O*
oducts
Inert pro
R . O . O + R . O . O*
Figuure 22: General mechanissm for therm
mal degradatiion [34]
As can be seen above, thermal degrad
dation mayy leads to loow molecullar weight
coomponent production
p
w
which
can build up on the
t die and extruder suurfaces. Such
h deposits
caan create diee lines and when/if
w
theyy are releassed, defects on the extruudate [59].
29
2..2.3
Addittives (slip agent,
a
antittack, PPA, LMw lubricant)
Addittives are thee ‘pepper’ and
a ‘salt’ fo
or the polym
mer. With aadditives it’s possible
too improve the
t polymer propertiess and also the processing properrties. If thee polymer
prrocessing addditives aree used to suuppress die drool phennomenon, thhe occurren
nce of the
poolymer addiitives may create
c
chem
mical and ab
brasive interractions redducing the PPA
P
functioonality [8].
move the flluoropolymeer coating
For exxample, anttiblock agennts may abrrasively rem
from the die face [5], orr the filler in
i some hig
ghly filled reesins will ddrool onto th
he surface
p
a [12, 13]]. More dettailed inform
aid,
mation are pprovided in [7].
evven with a processing
2.2.3.1 Slipp agents
Polyoolefin films tend to adhhere to them
mselves and metal surfaaces due to their high
cooefficient off friction (C
COF). For processing
p
ease, films need a COF
F near 0,2. Slip additivves [14] cann modify thhe surface properties off a film andd thus lowerr the friction
n between
fillm layers annd other suurfaces. To be effectiv
ve the slip needs
n
to miggrate out off the polym to the suurface and therefore
mer
t
it must have a degree of
o incompatiibility with the polym Fatty accid amides are
mer.
a often ussed as slip additives.
a
D
During
proccessing they
y are solubiilized in thee amorphouus melt, but as the polymer cools and crysttallizes the fatty acid
am
mide is “squueezed” outt forming a lubricating layer at thee polymer ssurface. The addition
off a slip addiitive can preevent film sticking and
d pulling hellping to incrrease throug
ghput.
2.2.3.2 Antiitack (antibblocking ageent)
Polyoolefin and other
o
plasticc films havee a tendencyy to adhere together, often
o
makinng it difficult to separaate layers. This
T adhesiion betweenn film layerrs, called bllocking, is
ann inherent property
p
of some polym
mers. Antib
blocking addditives [13]] can be add
ded to the
pllastic to minnimize this adhesion and
a so loweer the blockiing force beetween layeers. Once
coompoundedd into a plasttic these additives creaate a micro rough
r
surfaace which reeduces the
addhesion betw
ween film layers
l
and lowers the blocking teendency. Tw
wo factors determine
thhe antiblockking effect:
•
Numbber of particcles of antibblock at the film surface.
•
Size of
o the antibllock particlees.
30
The greater
g
the concentratio
c
on of antibllock presennt then the rrougher thee film surfaace produced. However it is impoortant that the
t particless are well ddispersed as agglomeraates reduce antiblockinng performaance. Conv
versely the coarser thee particles th
he further
thhe two film layers are kept
k apart.
2.2.3.3 PPA
A
Polym
mer Processsing Aids baased on a fluoropolym
f
mer or shorttly PPA’s arre a polym that conntains atomss of fluorine. It is charracterized by
mer
b a high reesistance to
o solvents,
accids, and baases.
Greett Dewitte [114] have foound that flluoropolym
mers by natuure are quitte inert to
chhemical reacctions and thermal
t
deggradation. They have a low surfacee energy and
d are generrally incomppatible withh other polym
mers. Micro
oscopic exaamination off a polyethy
ylene (PE)
thhat contains PPA, reveaals discrete micron-sizeed droplet shaped
s
partiicles of the fluoropolyymer (measuured with 3M Method "Optical Microscopy
M
M
Method
forr Dispersion
n Analysis
inn Polyolefinn’s"). Typical PPA usee levels vary
y from 100 to 1000 pppm dependiing on the
appplication. When
W
extruuded, the applied
a
sheaar-field alloows the PP
PA to phasee separate
from the PE matrix and form a thinn, persistentt coating onn the metal ssurfaces of the extrusion equipmeent. Once this
t
coatingg is established, the differential
d
between th
he surface
ennergies of thhe two polyymers allow
ws for reduceed friction during
d
extruusion of thee PE.It has
beeen found thhatthe addittion of fluoropolymer based Polym
mer Processing Aids (PPA) to a
reesin or a com
mpound cann reduce or eliminate
e
diie drool [111, 12, 17, 188, 19, 20 and
d 21]
Chapm
man et al [17] claim thhat there iss no concenntration graddient of fluoroelastom toward to
mer
t the die suurface. How
wever, it can
n be explaiined by stroonger affinitty of PPA
foor metal surrface than thhe resin whhat leads to the depositting of PPA
A onto die face
f
when
exxiting the diie [16 -18]. Other advaantages of PPA
P
during polymer prrocessing arre the supprression of melt
m fracturee [18, 22], thhe reduction
n of entrancce die pressure [23] and
d increase
thhe extrusionn rate. It hass been show
wn that proccessing aid performancce is strong
gly dependeent on the morphology
m
y of fluoroppolymer-pollyethylene blend
b
[24 - 25] and caan exhibit
addverse interraction withh other addditives, with
h the resultt that PPA performancce can be
drramatically altered [12,, 21].
Chan [25] believves that posssible mechaanism of diie drool supppression iss based on
PP
PA coating of metal suurface becauuse the PPA
A migrates tooward to plaace with hig
gher shear
31
raate (the die surface), prroviding a low
l
energy slip layer between
b
thee metal surfface of the
diie and the melt.
m
Chan [25] showeed that PPA
A at high levels (15 wtt%) will itself drool an
nd loosely
addhere to the surface of the
t extrudatte.
The effect
e
of PPA on the diie drool pheenomenon and
a final fillm finish is shown in
Fiigure 23.
Witthout PPA
Un
nstable
Die
Equilib
brium veloccity proffile v =0
wall
Die droo
ol
With PPA
Stable
Die
Equilibrrium velocity profile v >0
wall
Fluoropolym
mer film
Figuure 23: The effect
ef
of PPA on the die drool
d
phenom
menon (lef sidde) and finall film
finishh [14,19]
It hass been found that PPA
As suppress die lip builldup as welll as melt fracture
f
at
m
much
higher concentratiions [20,27]]. Several examples
e
prrovided by B
Blatz [20] shows
s
that
addding a fluooropolymer die drool suuppressant to
t either higgh-density ppolyethylen
ne or ethyleene/butene copolymer
c
i blown film extrusion
in
n will also suppress
s
meelt fracture. Although
thhe concentraation requirred to supppress melt fracture
f
cann be 100 times that reequired to
32
suuppress die lip buildup [26, 20], soome report comparable
c
e concentrattions [28, 11
1, 12], but
noone report loower conceentrations.
It shoould be menntioned thaat fluoropoly
ymer proceessing aids reduces diee pressure
annd increase the extrusioon rate wheereas the meechanical prroperties of the extrudeed product
arre not influeenced [26].
Slip plane
PP
PA
a.
Die wall
Cohesive Slip
Adhesive Slip
No Sliip
Case I
Case II
C
Case
III
Polymerr
b.
Figuure 24: Possiible boundary
ry conditions at the wall-ppolymer inteerface [52]
The Figure
F
24 clarifies
c
the PPA role during
d
the exxtrusion thrrough differrent boundaary conditioons compariison. In thiss figure, Caase I represeents standarrd no slip ap
pproximatioon; Case II, shows slipppage occuuring at the wall-polym
mer interfacce; Case IIII, shows a
finnite layer off polymer (PPA) whichh stucks to the
t wall andd slippage ooccurs in thee polymer
juust beyond this
t layer i.ee. slippage occurs
o
at thee polymer-ppolymer inteerface. [46]
33
2.2.3.4 Low
w molecularr weight lubbricant
Too less stearatte
Too
o much stearate
Polym
mer mattrix
Fillerr
Stearrate
a.
b.
Figuure 25: Visuaalization of thhe statearetee role in polyymer matrix. 25a). steara
ate << critical amount, 25b)
b). stearate >>
> critical am
mount [20]
Low molecular
m
lubricants (ppolyolefin wax,
w siliconn oil, aluminnium stearatte) appear
too have simillar behaviorr as PPA. However,
H
Leee [20] statees that it is nnecessary to
o optimize
thhe amount of
o stearate byy addition in
i highly filled master batch.
b
Whenn the amounnt of lubricaant is below
w the criticall value, therre is not eno
ough stearaate to coverr the surface. This cann lead to hig
gher friction and resullt in higher extent of
deegradation. This degraddation can promote
p
chaain scissionn what can lead to die drool.
d
Too
m
much
stearatee will not coover the filller surface and
a can alsoo lead to diee drool.
34
2..2.4
Rheoology
In thiis section, the effect off the polym
mer rheologyy on the diie drool pheenomenon
w be analyyzed based on the openn literature.. Choon, Chhai, Adamss and Framee [2] have
will
m
measured
thee shear viscoosity at 2700°C for the five
f differennt LLDPE rresins for co
onsequent
coomparison with
w an autooclave LDPE (MFR = 2)
2 (see Figuure 26).
Shear viscosity [Pa.s]
400
LLDPE‐A
LLDPE‐B
LLDPE‐C
LLDPE‐D
LLDPE‐E
LDPE
300
200
100
0
100
0
10
000
Shear rrate [s‐1 ]
10000
Figuure 26: Capilllary shear viscosity
v
data
a measured at
a 270°C for ffive differentt LLDPE
resinns and autocaalve LDPE type.
ty
It hass been founnd that polyymers havin
ng high sheear viscosityy (at high shear rate
raange) createes very inteensive die drool
d
pheno
omenon andd vice versa. This beh
havior has
beeen explaineed by the fllow inducedd fractionatiion which occurs
o
due to intensive shear and
exxtensional stress
s
generration, especcially for highly
h
viscoous polymerr melts, which forces
thhe shorter molecules
m
too migrate tooward the barrel
b
wall and thus prromoting die
d deposit
foormation.
The effect
e
of die geometry on
o the die drool
d
phenom
menon has been investtigated inteensively [288,8]. In more detail, thee following die designss were inveestigated fro
om the die
drrool point of view: straaight, flared,, converging
g and shapeed die (see F
Figure 27)
35
Straight
Flared
Convergging
Shap
ped
Figuure 27: Straigght, flared, converging
c
and shaped diie [8, 32]
Ding et al [29] have
h
found that flared die can redduce the diie drool pheenomenon
byy 94% in coomparison with
w referennce straight die. It also has been foound that co
onverging
geeometry cann yields 655% reductioon of die drrool phenom
menon but polymer deegradation
caan occur in this
t case duue to vortex generation.
Gandeer and Giaacomin [8] speculated that the sttabilization effect of flared die
(F
Figure 28) can
c be due to productioon of stresss undershooot at the die exit zone. Basically,
noonlinear visscoelasticityy causes a shhear stress undershoot
u
for a step ddecrease in shear rate
annd, likewisee, a stress ovvershoot forr a step incrrease in sheaar rate [29]..
Straight
Flared
C
Converging
γ
σ
Figuure 28: Explaaination by Gander
G
and Giacomin
G
abbout the stabiilization effecct of a
flareed die [8]
Anothher conclusion, which can be extrracted from
m the open lliterature is the statem that reccirculation upstream
ment
u
cann initiate deeposit onto the
t land [311].
36
K. Chhaloupkova and M. Zattloukal [15]] have donee experimenntal work with
w differennt die exits with respecct to die drool phenom
menon Theyy have founnd that 15° exit angle
reeduces the deposit
d
amoount from 1000% down to
t 3,3% whhereas the 455°exit anglee has been
foound to supppress the diie drool pheenomenon completely
c
for the testted mLLDP
PE resin at
paarticular proocessing connditions (seee Figure 29
9 in more deetail).
Die Drool Amount DDA [‐]
Die N
No.1
Die No.5
Die No.4
Die exitt angle [°]
Figuure 29: Effectt of die exit angle
a
on the die drool am
mount for mL
LLDPE [15]
They also have found
f
that flared
fl
die is much effecctive than shhapes dies [15].
[
They
ussed viscoelaastic FEM analysis
a
of the die dro
ool experim
ments. It hass been foun
nd that the
diie drool onsset can be predicted
p
byy the help of
o negative (suction) prressure occu
urrence at
thhe die exit zone.
z
In moore detail, thhey have rev
vealed that with increaase in cham
mfer length
(w
which is staabilizing change from the die dro
ool point off view) leadds to lowerr value of
suuction presssure and preessure graddient at the die exit zoone. The ressults their theoretical
t
annalysis is prrovided in Figure
F
30.
37
Die No.2
2
Pressure gradient [Mpa/m]
S i
Suction pressure [Mpa]
[M ]
Die No
o.2
Die No.6
Die No.6
Die No.7
D
Die No.7
Chamfer llenght [mm]
Chamfeer lenght [mm]
Figuure 30: Prediicted suction pressure an
nd pressure gradient
g
as thhe function of the
cham
mfer length [28]
[
It also seemss that surfacce energy of
o the die wall
w has signnificant imppact on the die drool
phhenomenon. It has beeen revealed that a loweer surface ennergy wall induces low
wer levels
off die deposiit [50]. Thiss has been attributed to
o small adhhesion betw
ween the pollymer and
thhe metal surrface reducinng the effecctive shear stress
s
[33]. Another efffect of loweer die wall
suurface energgy is the redduction of the
t shear strress level thhus suppresssing the po
olymer degrradationn annd flow indduced fractioonation. The effect of lower
l
surfaace energy metals
m
can
bee summarized as follow
ws[4-6]:
•
to redduce the sheear related fractionation
fr
n of the polyymer melt thherefore a smaller
s
propoortion of thee low moleccular weightt material would
w
migraate towards the
t extruderr wall decreeasing the liikelihood off depositionn,
•
the appparent sheaar stress appplied to the polymer woould reducee and therefo
fore die
swell reduces,
•
the coontact anglee between thhe polymer melt and thhe surface att the die exiit would
be unnfavorable for
fo depositioon [4].
38
2..3 Variaables sensiitive to th
he die drool phenom
menon
In this section, thhe role of different calcculable variiables such as shear strress, shear
ellasticity andd pressure onn the die drrool phenom
menon is disscussed in m
more detail.
2..3.1
Shearr stress
Increaasing shear (and extennsional) defformation onn polymer melt is kno
own to in-
duuce partial molecular
m
f
fractionation
n in the meelt stream [44, 28, 34] eessentially leading
l
to
thhe lower moolecular weeight moleccules migratting and cooncentratingg at the inteerface betw
ween the poolymer mellt and the extruder
e
waall. For polymers withh high poly
ydispersity
(bbroad MWD
D) the fractioonation will show a greeater affect..
Shelbby and Cafliisch [10] haave been hy
ypothesized that a high shear stress gradient
innduced durinng polymer processingg could causse an increased migratioon of oligom
mer to the
suurface, thereeby leadingg to the droool effect. To
o test this hypothesis, ppoly (ethyleeneterephthhalate) (PET
T) and polyy (ethylenee-co-cycloheexylenedim
methylene teerephthalatee) (PETG)
w extrudeed into film,, and the moolecular weeight probedd as a functiion of the depth
were
d
from
thhe surface byy incrementtal milling.
To annalyze theirr results, thhey made a graph of the weight averaged molecular
m
w
weight
as thee milling deepth functioon (see Fig
gure 31). Thhe Figure 331 clearly shows that
thhe moleculaar weight is lowest at thhe surface (depth
(
= 0m
mm) and higghest at the centerline
Mw [g/mol]
(thhe measured differencee was aboutt 6%).
6%
%
Dep
pth [µm]
Figuure 31: Moleccular weightt data as a fu
unction of miilling depth ((depth = 0 reefers to
the outer
o
edge) for
fo the extrudded PET film
m [10]
39
In theeir experimeental studies for PET die
d extrusionn they havee shown a gradient
g
in
thhe concentraation of low
w moleculaar weight cy
yclic oligom
mer across the thickneess of the
shheet. The cooncentrationn was foundd to be abou
ut 18% highher at the ouutside surfacce in compaarison with the centerlline. They also
a
mentio
oned that the results arre in good agreement
a
w a theoreetical modell based on shear-induce
with
s
ed diffusionn that takes into accoun
nt the tendeency for thee longer poolymer chainns to movee toward thee center of the flow field where
shhear stresses are lowesst. They havve concludeed that sheaar diffusionn is probably a major
faactor of die drool and other
o
polym
mer processiing-related phenomenaa such as diie lubricatioon, melt fraacture reduction, and soo forth.
2..3.2
Shearr elasticity
F. Dinng et al havve tested thhe Gander-G
Giacomin [88] hypothesis outlined in Figure
300. Specificaally, they waanted to seee if a stress undershoott must occurr near the die
d exit for
a flared die too suppress die
d drool. For
F this aim
m, they have uses constiitutive equaations propoosed by Waagner [40] annd Liu [41] to calculatee wall stresses in flaredd dies.
Figurre 32 showss the average build-up
p ratios for flared, straaight and co
onverging
diies. For the same exit gap,
g the diee flaring deccreased die lip build-upp by 94%. Thus,
T
flarinng the die liips greatly suppresses die drool. The converging die aalso decreasses die lip
buuild-up by 65%.
6
Presuumably, thiss is becausee the lower shear rate uupstream of the conveerging section lessens fractionation
f
n.
2.00
BP x 10
6
1.80
1.00
0.64
0.11
0.00
Flared
Straightt
Convverging
Figgure 32: Die lip build-up ratio for flarred, straight and convergging dies in film
f
blowing experiments [299]
40
They found theorretically thaat shear streess at the diee exit wall iis not propeer variable
w
which
can bee used to unnderstand thhe die drool onset. On the
t other haand, they haave found,
thhat the Nl unndershoot (tthough welll before the exit) helps reduce the extrudate swell,
s
thus
suuppressing die
d lip buildd-up.
2..3.3
Negative pressu
ure
8] investigaated the effeect of die drool
d
pheK. Chhaloupkova and M. Zaatloukal [28
noomenon on the pressurre field at thhe die exit region for metallocene
m
e based mL
LLDPE by
ussing
Virtual Extrrusion Labooratory™ (C
Compuplast®)). They
FEM simulation software (V
haave found that
t
so calleed external die drool can be expplained by tthe negativee pressure
(ssee Figure 33)
3 occurrinng at the diee exit region
n where thee free surfacce of the ex
xtrudate is
crreated. Theyy have conccluded thatt a significaant suction effect, due to negativee pressure
occcurrence, can
c explainn the tendenncy of posssible low molecular
m
w
weight comp
ponents to
m
migrate
from
m the polym
mer matrix toward the die
d wall as well as maay explain accumulaa
Negative prressure
Die wall
Flow direction
Free surface
tioon tendencyy of these coomponents at the die ex
xit zone.
Figuure 33: The predicted
p
preessure field for
fo the extrussion of mLLD
DPE [28]und
der die
droool conditions
41
They have foundd that critical negativ
ve pressure value (-1.66MPa) exissts for the
m
mLLDPE
maaterial based on theoreetical analyses of die drool
d
experriments wheere stable,
stable-unstabble, unstablee were careefully determ
mined for different
d
tem
mperatures and mass
Suction pressure p [Mpa]
floow rates (seee Figure 344)
‐1
Mass flow rate m [kg.h ]
Figuure 34: Suctioon pressure at
a the end off the die as thhe function oof mass flow rate and
tempperature [28]]
To exxplain the orrigin [28] off the negativ
ve pressure, it has to be pointed ou
ut that the
m elasticityy and stream
melt
mline curvaature may leead to the noormal stresss generation
n that conseequently caauses the noonmonotoniical (local pressure deecrease) in the pressu
ure profile
duuring the poolymer mellt flow. It seems
s
that the
t negativee pressure (local presssure minim
mum)
seems to ‘‘help’’ to the meltt with the veelocity rearrrangement at the die exit
e region
w
which
promootes flow indduced fractiionation.
42
3
PROBL
LEM DEF
FINITIO
ON
This work
w
deals with die buuild-up pheenomenon for
f two diffe
ferent polyp
propylenes
(teerpolymers consisting of butadienne, ethylene and propyllene) whichh have differrent sensitivvity to the die
d build-upp phenomennon. The po
olypropylenne sample w
which has beeen found
too produce die
d build-upp very quicckly (usuallly after sevveral days oon the com
mmercially
BOPP line) is
i called heere as the ‘Sample 1’ whereas ‘S
Sample 2’ is very stab
ble sample
v
from die droool point of view.
Main aim of thiss work is too provide pllausible expplanation foor different Sample 1
annd Sample 2 behavior from
f
the diee drool poin
nt of view. For
F this purrpose, dies drool
d
phenoomenon will be analyzed theorettically as well
w as expeerimentally by using laboratory
l
exxtrusion linee equipped by the speccially design
ned die. Parrtial aims arre summarized in the
neext section.
43
4
AIMS OF THE WORK
The master
m
thesiss partial aim
ms can be diivided in theese free areas:
•
Rhheological characterizaation of tw
wo different polypropyllenes (Sam
mple 1 and
Saample 2) having
h
diffeerent sensittivity to die drool phhenomenon in BOPP
prrocess and description
d
of the meassured data by
b the suitabble constitutive equatioon.
•
Die drool evvaluation forr these two
o resins on the
t laboratoory extrusio
on line by
ussing digital image analyysis.
•
ment method analysis of die droool experim
ments perViscoelastic finite elem
foormed on the laboratoryy extrusion line.
44
II. EXPE
ERIME
ENTAL
L
45
1
EQUIP
PMENT AND
A
MET
THODS
In this section, experimental set-up for co
onsidered rhheological ttesting and die drool
evvaluation foor both polyppropylene samples
s
are introducedd in more deetail.
1..1.1
Rotattional rheoometer ARE
ES
The rootational rhheometer, Advanced
A
Rh
heometric Expansion
E
S
System (AR
RES), used
inn this work is depicted in Figure 35.
3 This kin
nd of rotatioonal rheomeeter is capaable to applly a dynam
mic (sinusoidal) or steaady (linear)) shear straain (deform
mation) to th
he testing
poolymer sam
mple with coonsequent shear/norma
s
al stress meeasurementss by the 2 K FRTN1
annd 2 K FRT
TN2 transduucers (proviided by TA
A Instrumentts) with a loower resolu
ution limit
off 0.02g/cm.
Sensibilitty transduceers
Torq
que
transducers
Oven
Sample
Figuure 35: Descrription of thee main ARES
S parts
46
The samples
s
werre measuredd in oscillattion mode with
w range of frequenccies (0.1 –
1000 rad.s-1). The measuurements weere perform
med with paarallel platee geometry (diameter
255mm) and the
t thicknesss of the saamples was 1mm. Thee rheologicaal propertiess for both
teested samplles have beeen measurred for thee followingg temperatuures: 180°C
C, 200°C,
2220°C, 240° and 260°C..
Air temp
perature sen
nsor
Sample
Mettal temperatu
ure sensor
Figuure 36: Detaiil view of thee sample loca
ation betweenn two paralleel plates.
1..1.2
Capilllary rheom
meter Rosand RH7-2
In genneral a capiillary rheom
meter is a un
niversal visccosimeter w
which can be used for
thhe determinnation of shhear viscosiity, elongattional viscoosity, first/ssecond norm
mal stress
diifferences and
a wall slipp. The rheoometer can be
b in severaal configuraations. The first contaains a singlee – bore barrrel, the second a twin – bore (duaal) barrel. T
The twin – bore
b
types
saave the testing time beccause they are
a able to obtain
o
two results
r
in onne test cyclee. Another
tyype is on – line capillaryy rheometerr which is mounted
m
on an extruderr.
In thiss work, we have used twin
t
bore caapillary rheeometer Rossand RH7-2
2, which is
deepicted in Figure
F
37. It
I is equippped with tw
wo (orifice and
a long) capillary diees. Due to
tw
win-bore tecchnology, shhear and exxtensional data
d can be obtained w
without usin
ng a set of
vaarious capillary dies.
47
Pisston holder
3 heating 3
belts
Left pressure sensor
Up/down operation
Right pressure se
ensor
Figuure 37: Descrription of thee Rosand RH
H7-2 capillary
ry rheometer
The measuremen
m
nts were performed in a constant piston speeed mode at the shear
raate range off (20-2000) s-1. In our measuremen
m
nts we havee used presssure transdu
ucers (Dyniisco, USA) in ranges off (10000) PSI (68.9476
6 MPa), (15500) PSI (100.3421 MPaa).
The capillary
c
diees used in this
t work arre depicted in Figure 338 and they
y have the
foollowing dim
mensions:
- longg capillary die:
d L=16 mm,
m D =1 mm
m
- shorrt capillary die:
d L = 0 mm,
m D =1 mm
m
W
Where
L is capillary
c
diee length andd D stands for
f capillaryy die radius.. Note, that the barrel
raadius was eqqual to 15m
mm. Moreovver, the exttensional viscosity wass calculated
d from entraance pressuure drop meaasurements by using well
w known Cogswell
C
m
model.
The rhheological properties
p
f both testted samples have been measured for
for
fo the folloowing tempeeratures: 1880°C, 200°C
C, 220°C, 24
40° and 2600°C. It also should be mentioned
m
thhat two prehheating stepps (for 3 andd 6 minutess) and two compressioon steps (to 0.5 MPa)
48
w applied before eachh experimennt cycle in order
were
o
to achhieve homogenous poly
ymer melt
att desirable teemperature with no airr bubbles.
Figure 38: Long die (lefft side) and short
s
die (rig
ght hand sidee)
49
1..1.3
Extru
usion line for
f die build
d-up evalua
ation
The extrusion
e
linne used forr the die bu
uild-up anaalysis consists of the Brabender
B
drriving unit, single screw
w extruder for
f polyoleffins, extrusiion die, coooling box an
nd temperatture and preessure regulaation boxess (see Figuree 39).
Drivingg unit with temperrature box
Extruder with ho
opper
Heating supply fo
or the part between thee e
extruder and the
e die
T4
Tem
mperature sensor (T5) for part between
n extruder and die
T3
T2
T1
Temperature 1, T2, sensors for T1
T3, T4 and T5
Turrning set speeed (better)
Die
e
Digital set speed
Outlet cooling
Prressure sensor on
n the end of the exxtruder
Inllet co
ooling
ON/
OFF
Ho
opper an
nd cooling
Extruder
Heating systtem T6
Set coolingg temperatu
ure
Coolingg box
Heating sup
pply for the die
Teemperature seensor (T6) for th
he die
Heating supply for the die
Inlet cooling
Pressurre box
O
Outlet cooling
Set temperature for T6
Actual temperature fo
or T6
Actu
ual pressure on th
he end of the extuder
Figuure 39: Equippments used for
f the die build-up evaluuation
Special extrusionn dies havinng 2 heating
g parts and 1 cooling paart depicted in Figure
400 was emplooyed. Here,, the first heeating elemeent occurs at
a the extra part betweeen extruderr and die whheatear the second onee is located at the die. With the aiim to achiev
ve the die
buuild-up phennomenon att low mass flow rates and short tiimes, the caapillary end
d has been
alllowed to bee cooled dow
wn by usingg of silicon oil.
50
Temperature senso
or
Heating elements
Insert
OUT
IN
Cooling
Speciall die
Between part
Heating supplyy
Figuure 40: Extruusion die connstuction
Temperaature senso
or
Pow
wer supp
ply
Heaating un
nit
Incomin
ng cooling cable
D
Die insert
Figuure 41: Partss of the annullar extrusion
n die
Outggoing coolingg cable
Connection with extruder
51
The Figure
F
41 shhows real viiew of the capillary
c
diee with remoovable capilllary insert
haaving sharp edges at thhe die exit reegion, which
h is describbed in Figurre 42 in morre detail.
Ø 3 mm
Ø 1.6 mm
Silicon oil
Silicon oil
2mm
m
12 mm
2mm 2mm
2
33 mm
Figuure 42: Dimeensions of thee extrusion die
d insert (floow direction iis from rightt to left)
The Figure
F
43 shows
s
in what
w
way th
he extrusionn die is connnected to the
t single
sccrew extruder. In more detail, the main die in
nsert part is heated throough heating
g unit gettinng power frrom the maain power suupply whereeas for the die exit coooling, the siilicone oil
is used and controlled
c
by cooling box.
b
The pro
ocess regulaation is giveen by the usse of tempeerature senssor and presssure transduucer which are connectted to the exxtrusion diee.
Note, that one reegulation deevice (PL20
000) was neeeded to conntrol 4 extru
uder heatinng zones (T1 Æ T4) toogether withh the part between
b
singgle screw eextruder and
d die (T5)
w
whereas
the second
s
reguulation device was need
ded for extrrusion die oonly. It also should be
m
mentioned,
thhat a “Dynnisco 1290” pressure box
b has beeen used to m
measure thee pressure
beetween singgle screw exxtruder and die, especiaally due to safety
s
reasoons. The pro
oblem can
occcur if the polymer
p
sollidification at the die exit
e region occurs due to intensive cooling.
Thhis may leaads to signnificant abruupt pressuree rise causiing the die or the sin
ngle screw
daamage.
52
Coo
oling box
Extrud
der with hoppe
er
Driving unit with te
emperature sen
nsors
Die
Heatting system T6
Temp
peratures
Mechanical connectio
ons
Pressure box
Presssure
Cooling
Heatting
Figuure 43: Equippment arranggement for th
he die build-up measurem
ments.
53
IIII.
RE
ESULTS
S AND D
DISCU
USSION
N
54
1
RHEO
OLOGICA
AL RESU
ULTS
Fiigures 44-445 summariized the shear and exttensional viscosity
v
datta for Samp
ple 1 and
Saample 2 obbtained from
m rotationaal rheometeer ARES annd capillaryy rheometeer Rosand
RH
H7-2. For the
t theoreticcal analysiss, it is necesssary to desscribe the m
measured daata mathem
matically.
Foor this aim, modified White-Metz
W
zner model has been used becausee only six
addjustable moodel parameters have to
t be identified; the moodel providded analytical expressions for steaady shear and
a steady uniaxial
u
exttensional viiscosity whhich make laast square
m
minimization
n for modell parameterr identificattion simple;; finally, thhe model iss stable in
FE
EM even for
f strong flows
f
wherre both, sheear and exxtensional fl
flow compo
onents are
m
mixed
togethher. The moodified Whiite-Metzner viscoelastiic model is given by th
he followinng set of equuations:
∇
τ + λ (II D )τ = 2η (II D )D
η (II D ) =
λ (II D ) =
[1 + (K
η0 f
1f
1− n
a a
II D )
λ0 f
1 + K 2 f II D
f (T ) = e −b(T −Tr )
(1))
]
(2))
(3))
(4))
wheree Ea is the activation
a
ennergy, R is the
t gas constant, Tr is tthe referencce temperatture, T standds for the teemperature, D represen
nts the rate of
o deformattion tensor, IID means
∇
thhe second innvariant of the
t rate of deformation
d
n tensor, τ iss the stress tensor, τ is the upper
coonvected tim
me derivativve of stresss tensor, λ((IID) means the deform
mation rate-d
dependent
reelaxation tim
me and η(IIID) stands for
fo the deforrmation rate-dependennt viscosity, where η0
reepresents Newtonian viscosity,
v
λ0, K1, K2, n and a are constants w
whereas b stands
s
for
teemperature sensitivity
s
55
S ample 1
Sample 2
Carreau mod
C
del
η0 =
72
296
λ
λ =
0,0818
n =
0,1
a =
0,4249
Carreau mode
el
η0 =
780
00
λ=
0,3238
n =
=
0,2543
a =
=
0,5099
Relaxtion fu
R
unction
λ0 =
62
284,3
K1 =
10
0043
Pa
a.s
s
s
s
Exponential model
1//C
b =
0,0192
Tr =
20
00
C
Re laxtion fun
nction
λ0 =
0,1647
K1 =
0,2152
Pa.ss
s
s
s
Exp
ponential m
model
b =
=
0,0173
1/C
Tr =
200
0
C
Tablle 1: Modifieed White Mettzner parameeters for Sam
mple 1 and Saample 2
56
105
Shear and extensional viscosities (Pa.s)
Shear and extensional viscosities (Pa.s)
104
103
180°C
C - Measurements
s shear viscosity
180°C
C - Fitting shear vis
scosity
220°C
C - Measurements shear viscosity
220°C
C - Fitting shear vis
scosity
260°C
C - Measurements shear viscosity
260°C
C - Fitting shear vis
scosity
180°C
C - Measurements extensional viscosity
180°C
C - Fitting extension
nal viscosity
220°C
C - Measurements extensional viscosity
220°C
C - Fitting extension
nal viscosity
260°C
C - Measurements extensional viscosity
260°C
C - Fitting extension
nal viscosity
102
10
1
10
-1
10
0
1
2
10
10
Shear aand extentional rates (1/s)
3
10
4
10
Figuure 44: Compparison betw
ween measurred and preddicted temperrature depen
ndent shear
and extensional viscosities
v
foor Sample 1.
57
105
Shear and extensional viscosities (Pa.s)
Shear and extensional viscosities (Pa.s)
104
103
180°C
C - Measurementss shear viscosity
180°C
C - Fitting shear visscosity
220°C
C - Measurements shear viscosity
220°C
C - Fitting shear visscosity
260°C
C - Measurements shear viscosity
260°C
C - Fitting shear visscosity
180°C
C - Measurements extensional viscosity
180°C
C - Fitting extension
nal viscosity
220°C
C - Measurements extensional viscosity
220°C
C - Fitting extension
nal viscosity
260°C
C - Measurements extensional viscosity
260°C
C - Fitting extension
nal viscosity
102
101
10-1
100
101
102
Shear aand extentional rates (1/s)
103
104
Figuure 45: Compparison betw
ween measurred and preddicted temperrature depen
ndent shear
and extensional viscosities
v
foor Sample 2.
58
F
46 shows
s
sheaar and exten
nsional visccosity data for both samples at
The Figure
2220°C. Based on the modified
m
Whhite-Metzneer parameteers and Figgure 46, we can concllude that thee Sample 2 is more elastic than Sample
S
1 duue to its higgher macroscopic relaaxation timee (see Table 1).
105
Shear and extensional viscosities (Pa.s)
104
103
10
2
10
1
10
Sample 1 - Measurements Exxtensional viscosityy
Sample 2 - Measurements Exxtensional viscosityy
Sample 1 - Fit Extensional visccosity
Sample 2 - Fit Extensional visccosity
Sample 1 - Measurements Sh
hear viscosity
Sample 2 - Measurements Sh
hear viscosity
Sample 1 - Fit Shear viscosity
Sample 2 - Fit Shear viscosity
-1
10
0
1
2
10
10
0
Shear and e
extentional ratees (1/s)
3
10
4
10
Figuure 46: Com
mparison betw
ween shear and extensional viscosiities for Sam
mple 1 and
Sampple 2 at 220°°C.
W the aim
With
m to see wheether some flow induceed fractionaation occurss during thee capillary
teest, the folloowing proceedure has been
b
followed. Firstly (Stage 1), H
HDPE Chev
vron Phillipps Chemicals film bllowing graade has been extrudeed through the clean long die
(D
D=1mm, L=
=16mm) at 170oC withhin wide app
parent shearr rate rangee (20 - 2000
0 1/s). The
coorrespondinng pressure profile
p
for this
t test is depicted
d
in Figure 47. As clearly visible, at
appparent sheaar rate abouut 140 1/s pressure oscillations occcur as the rresult of the slip-stick
59
phhenomenon which is typical
t
for HDPE poly
ymers. Seccondly (Stagge 2), Sam
mple 1 has
beeen extrudeed at the sam
me temperaature and sh
hear rate rannge as HDP
PE polymerr and correesponding pressure
p
respponse has been
b
recordeed. Note, thhat the capilllary die hass not been
clleaned afterr the test fiinished. Thiirdly (Stagee 3), the HDPE
H
was eextruded th
hrough the
caapillary die at apparentt shear rate 140 1/s to check whetther the pressure oscillations occuur or not. Fiinally (Stagge 4), if applicable, the HDPE polyymer has beeen extrudeed through
thhe capillary die by appparent shearr rate equall to 1000 1//s until the pressure osscillations
w
were
recoverred at 140oC. This fouur stages procedure haas been appplied for bo
oth tested
saamples.
2500
30
HDPE - Press
sure
HDPE - Shear rate
25
2000
︶
1500s
/
1
e
t
a
r
r
a
e
h
1000S
︵
Pressure (MPa)
Pressure (MPa)
20
15
10
500
5
0
0
200
40
00
600
Time (sec)
Figuure 47: Presssure fluctuatiions of HDPE
E at 170oC
800
0
1000
60
The pressure
p
proofile for HD
DPE extrussion throughh the clean as well ass contaminaated dies is depicted inn Figure 48. It is clearly
y visible thaat contaminnation level is smaller
(oor contaminnates are noot hardly linnked to thee die surfacce) for Sam
mple 2 in co
omparison
w Sample 1 because pressure osscillations are
with
a recovereed at 140 11/s for Samp
ple B test
(ssee green linne in Figurre 48 in com
mparison with
w orange line). Alsoo, if the pressures are
coompares bettween for Sample
S
1 annd Sample 2 tests at higgh shear ratte ranges, th
he Sample
B contaminaates seems to behaves as the pro
ocessing aidd because thhe pressuree is lower
coompare to Sample
S
1 in this case. Moreover,
M
th
he experimeents revealeed that 4x 45
5,5 cm³ of
H
HDPE
(extruuded by 10000 1/s) wass needed to clean Sam
mple 2 contaaminated diie, but for
Saample 1, 7 x 45,5 cm³ of HDPE was
w not eno
ough. This suggests
s
thaat Sample 1 contaminaates are morre intensiveely adhered to the capillary die in comparison
c
with Samp
ple 2.
30
2400
clean HD
DPE
Sample 1 contaminated die
Sample 2 contaminated die
Shear rate
2000
25
20
︶
s
/
1
e
1200t
a
r
r
a
e
h
S
︵
Pressure (MPa)
Pressure (MPa)
1600
15
800
10
400
5
0
50
10
00
150
Relative timee
200
0
250
Figuure 48: HDP
PE pressure profiles durring extrusioon through cclean as well as Sample1//Sample2 contaminated capillary
c
diees.
61
2
DIE BU
UILD-UP
P MEASU
UREMENTS
In thee first step, Sample 1 has been chosen to find out proccessing con
nditions at
w
which
very pronounced
p
die build-uup phenomen
non occurs.. The follow
wing processs parameteers have beeen change for
f this purppose: the ro
otation speeed of the sccrew (n), tem
mperature
prrofile along the screw (T
( 1 – T4), thhe temperatu
ure of the part
p betweenn the extrud
der and the
diie (T5), tem
mperature of the die (T
T6) and tem
mperature of
o the silicoon oil in th
he cooling
chhannel (Texiit). In more detail, Texitt temperaturre was varieed from 65 °C up to 260°C
2
and
thhe mass flow
w (via screw
w speed) froom less than
n 0.25 kg/h up to 4 kg/hh. All tests have
h
been
doone for a minimal
m
timee of 20 minn so there was
w time to produces
p
poossible die deposit at
thhe die exit region.
r
An overview
o
abbout the inv
vestigated temperature
t
es profiles is given in
Taable 2. It has
h been fouund that veery pronoun
nce die droool occurs aat temperatu
ure profile
giiven by last line in Taable 2. Theerefore, for the conseqquent experriments, T1
1-T6 were
fixxed whereaas the mass flow
f
rate annd Texit weree varied.
Ex
xperiment
1
2
3
4
5
6
7
8
9
T1
80
80
80
80
80
80
80
80
80
T2
T
2
200
2
200
2
200
2
200
2
200
2
200
2
200
2
200
2
200
T3
T
2
240
2
220
2
240
2
240
2
240
2
240
2
240
2
240
2
240
T4
4
26
60
23
30
26
60
26
60
26
60
26
60
26
60
26
60
26
60
T5
5
26
60
23
30
26
60
26
60
26
60
26
60
26
60
26
60
26
60
T6
26 0
23 0
20 0
20 0
20 0
20 0
20 0
20 0
20 0
Tex
xit
‐
‐
‐
80
100
0
110
0
120
0
60
77
Tablle 2: Overvieew of the set temperatures to get the processing
p
w
window
All peerformed exxperiments under thesee conditionss are providded in Tablee 3. Here,
it should be mentioned
m
t the exit temperaturres are meassured, not adjusted.
that
62
Massflow [kg//h]
from
m
to
≤0,25 0,2
25 0,5
0,5
5
0,75 1
0,75 1
1,25 1,5
1
1,25
1,5
1,75 2
1,75 2
2,2
25 2,5
2,25 2,5
5
2,75 3
2,75 3
3,25 3,5
3,25 3,5
3,75
5 4
3,75
≥4
70
65‐7
Low temperatures
70‐7
75
75‐8
80
80‐8
85
85‐9
90
95‐1
100
100‐‐105
105‐‐110
110‐‐115
115‐‐120
120‐‐125
High temperatures
Temperatures [°C]
90‐9
95
125‐‐130
130‐‐135
135‐‐140
140
60
‐ 16
175
0
‐ 200
≥250
0
No die dep
posit (t < 30 m
min)
Die d
deposit (t < 30
0 min)
Tablle 3: Effect of Texit temperature and mass
m
flow rate on the diee build-up ph
henomenon
(corrresponding T1-T6
T
valuess are provideed in Table 2, experimentt No.9).
As shhown in Tabble 3, die buuild-up phen
nomenon waas attainable only at low die exit
teemperatures. The blackk line in Tabble 3 shows the stabiliity contour.. The area above
a
this
linne represennts unstablee processingg condition
ns from diee drool poinnt of view and vice
veersa.
f
processing
p
c
conditions
hhave been chosen
c
for
Basedd on the Taable 3, the following
thhe
Sam
mple
1
and
Sample
2
die
buuild-up
analysis:
T1 = 80°C | T2 = 2000°C | T3 = 240°C | T4 = 260 °C
C | T5 = 2260°C | T6 = 200°C
Teemperature of the die exit
e (cooledd with silico
on oil): Texit = 77°C (see Figure 49
9 temperatuure profile clarification)).
63
T4
T3
T2
T1
T5
Texit
T6
Figuure 49: Set teest conditionss
The speed
s
of thee extruder was
w fixed at
a 5/70, whiich means tthat the digiital signal
w 5 and thee more corrrect turning meter was set on 70, what
was
w correspponded with
h the mass
floow rate.
Figuure 50: Screw
w speed reguulator
Sampple 1 and Saample 2 weere extruded
d for 1 houur under theese processiing conditioons and thee corresponding exit pressure and
d die exit teemperature was measu
ured every
m
minute
to see the evoluution of theese two varriables. Morreover, twoo tripods with digital
viideo cameraa and digitaal photo cam
mera were lo
ocalized in front of thee extrusion line
l under
diifferent anglles to captuure the die build-up
b
dev
velopment properly
p
intoo account.
64
e temperrature and thhe pressuree for both samples
s
is
The time evolutiion of the exit
prrovided in Figure
F
51. Note
N
that acctual valuess of T2, T3 and
a T4 weree practically
y identical
w the adjuusted values and thus thhey are not provided
with
p
in this Figuree.
240
24
4
Samp
ple 1 - Temp. 1
Samp
ple 2 - Temp. 1
Samp
ple 1 - Temp. 5
Samp
ple 2 - Temp. 5
Samp
ple 1 - Temp. 6
Samp
ple 2 - Temp. 6
Samp
ple 1 - Pressure 1
Samp
ple 2 - Pressure 1
20
0
︶
a
P
M
160
e
r
u
s
s
e
r
P
︵
Temperature (°C)
Temperature (
C)
200
16
6
120
12
2
80
0
20
40
60
Time (min
n)
Figuure 51: Timee evolution of
o pressure and
a temperaature during die drool analysis for
Sampple A and Saample B at chhosen processsing conditioons
It can be seenn that the T1, T5, T6 andd Texit temp
peratures aree almost thee same for both
b
materiaals (percenttage faults were
w
foundd to be follo
owing: T1 = 2,08% | T5 = 0,50%
% | T6 =
1,,64% | Texxit = 0,06%). On the othher hand, th
he pressure generation is approxim
mately 1.5
tim
mes higher for Samplee 2 in compparison with
h Sample 1 as visible in Figure 51. In the
exxtreme casee, this can leads
l
to pollymer melt leakage froom the extrrusion die which
w
can
reeduce overaall mass flow
w rate from
m 0.39 kg/h down to 0.331 kg/h (seee Figure 52
2 for more
deetails).
65
Backpressurre
IN
OUT
Backpressu
ure
Me
elt
Figuure 52: Polym
mer melt leakkage from th
he die at veryy high extrussion pressurees for Sample 2.
2
Analyysis of the Sample
S
1 annd Sample 2 extrudatess for these pprocessing conditions
c
mooth with
haas also beenn performedd. It has beeen found thaat the stringg for Samplle 1 was sm
onnly a small texture on the
t extrudatte surface whereas
w
the Sample 2 sstring was wavy
w
with
sinnus functioon like surfa
face defects as shown in the figure 53. Morreover, the Sample 2
exxtrudate collor was darrker than thhe input graanulates. Thhis color chhange was not
n visible
foor Sample 1.
1 It also shoould be meentioned thaat during thee equipmennt cleaning procedure
p
(aafter the 1 hour
h
of extrrusion) a thhin layer off oil like layyer, betweeen the plastiic and the
m
metal,
was obbserved forr Sample B. It was foun
nd that the cleaning prrocess was much
m
easieer if this layeer was preseent on the steel
s
walls.
66
Little
surface
faults
SAM
MPLE
1
No
odes
SAM
MPLE
2
Small lines
in the melt
surface
Approximate
ely 2 cm
Figuure 53: Sampple 1 and saample 2 extrrudates for the
t identical processing conditions
chossen for die buuil-up evaluaation
In thee next sectiion, die builld-up pheno
omenon willl be analyzzed by the digital
d
imagge analysis.
67
2..1.1
Digitaal image an
nalysis of th
he die build
d-up phenoomenon
Digitaal image annalysis for die
d deposit time evoluttion for botth tested sam
mples has
beeen perform
med simultaaneously froom two diffferent frontt angles byy using two
o different
innstruments. In more dettail, digital photo cameera (Panasonic Lumix F
FX12, resolution 7M
piixels, shot interval
i
1 min.)
m
and annalog video camera (Soony Handyccam, resoluttion 2.8M
piixels, framee rate 25 fpps) located at
a right diee side and at
a the left ddie side, resspectively,
w utilizedd (see Figuree 54).
were
Video vie
ew
Photo view
y
y
x
z
x
z
Figgure 54: Two front anglee views for digital
d
image analysis
Too measure the
t die depoosit amountt in a good way it is neecessary to convert thee 3D perspeective of thhe two view
wpoints to a 2D front view
v
i.e. it is necessarry to conveert the 3D
cooordinate syystem to a 2D coordinnate system
m as shown in Figure 555. For thiss purpose,
‘D
Didger® 3.003’ softwaree has been used.
u
68
Diffferent view po
oint
Front view
y
x
z
y
x
y
x
z
Figure 55: Coordinate system trransformatioon
The Figure
F
56 shhows in whhat way the die deposit amount (inn terms of die
d deposit
suurface area) for particuular case (diie deposit for
fo Sample 2 after 10 m
minutes) is quantified
q
byy the use off above desccribed transsformation technique.
t
I this way, all videos and photo
In
shhots were annalyzed andd the relativve Die Depo
osit Amounnt (DDA) w
was determin
ned as the
diie deposit suurface area divided byy the capillaary die surfaace area. It should be mentioned
m
thhat taking thhe die depossit surface as
a the die deposit
d
intennsity measuure is accepttable only
iff die depositt predominaantly occurss in one plaane and not so much inn the thickn
ness directioon, i.e. onlyy at the die build-up
b
onset.
69
Transform
mation
y
y
x
x
z
Figgure 56: Cooordinate systtem transform
mation for diie deposit
Figurre 57 showss digital phhoto cameraa shot collection (1 slide per 5 seecond) for
Saample 1 andd Sample 2 for the ideentical proccessing condditions desccribed abov
ve. Surprisinngly, the diie build-up onset for both
b
samples was foundd to be pracctically iden
ntical (see
Fiigure 57 forr more detaiils).
0 secc
5 secc
10 ssec
15 sec
20 sec
25
5 sec
S
SAMPLE
1
S
SAMPLE
2
Figuure 57: Die build-up
b
onseet evaluation for Sample 1 and Samplle 2
30
0 sec
70
Fiigure 58 reepresents video cameraa collection for time evvolution off die build-u
up phenom
menon
for Saample 1 andd Sample 2 in large tim
me scale too evaluate ddie build-up character
inntensity.
Startt
1 min
5 min
n
6 min
Startt
4 min
20 m
min
26 m
min
2 m
min
3 m
min
4 min
26 min
6 min
46
5 m
min
7 m
min
11 min
29 m
min
36
6 min
51 min
S
SAMPLE
1
8 m
min
S
SAMPLE
2
Figuure 58: Photoographs to fiind out the dif
ifferent shape of the bothh samples
It cann clearly be seen that die deposit iss, firstly, much
m
faster ffor Sample 1 in coma secondlly, its charaacter and shhape is diffeerent. In mo
ore detail,
paarison with Sample 2 and
Saample 1 andd Sample 2 die depositt is similar to
t half of moon
m
and a rring of flakees, respectivvely (see Fiigure 59 forr more detaails). Moreo
over, the Sam
mple 2 die deposit colo
our seems
allso be differrent in com
mparison witth Sample 1.
1 In more detail, the S
Sample 2 die
d deposit
seeems to be brown
b
sugggesting posssible samplee degradatioon, occurrennce of somee low moleecular weighht componennts or PPAss or their vaariants comiing out the ddie.
71
P
Photographs
SAMPLE
Sketch
Start po
oints
Half moon
m
sha
ape
1
Start points
SAMPLE
2
Ring shape
s
Figuure 59: Die build-up
b
chaaracter for Sample
S
1 andd Sample 2 aat the same processing
p
condditions
Anothher way to find imporrtant data iss to play both videos near each other and
seearch for differences. The
T most im
mportant moments
m
thaat could be seen on theese videos
arre summarizzed in the Table
T
4. Thhe main con
nclusion is that that occcurrence of the melt
innstability (cuurl effect) leeads to highher die build
d-up intensiity. Surprisiingly, Samp
ple 1 (havinng low elastticity) is moore sensitivee to the curll effect thann Sample 2 (having hig
gh elasticityy). This cann be explainned by the possible
p
slip
p occurrencee during Sam
mple 2 extrrusion due
too depositionn of Samplle 2 contam
minates at the die waall. This is consistent with the
rhheological measuremen
m
nts made in this work (ssee Figure 48)
4
72
Start
0:23:00
0:02:01
0:03:01
0:03:36
0:03:56
0:04:21
0:04:31
0:05:06
0:06:46
0:08:41
0:09:46
0:11:16
0:11:46
0:15:36
0:17:21
SAM
MPLE 1
Sta
able flow
Sttart
The
e die depossit is taken by the me
elt
0::00:30
e deposit on
n different places Die
0::01:30
butt taken by tthe melt
w: curl effecct → fast Unstable flow
acccumulation
0::03:20
Sta
able flow
e underside
e and the De posit at the
upperside is b
becoming lo
oss
0::03:40
e underside
e is taken De posit at the
by the string
0::04:53
Strong curl efffect
e underside
e
De posit at the
Sta
able flow
0::06:50
Curl effect
Sta
able flow
Curl effect
Sta
able flow
Curl effect
Sta
able flow un
ntill the end
d
0::30:00
SAM
MPLE 2
Stab
ble flow
Littl e unstable
Stab
ble flow
Strin
ng with nod
des
Strin
ng with much nodes
Curl effect and the die de posit is Curl effect and string with
h nodes
Die deposit rel ease from tthe die
Strin
ng with more nodes an
nd die 0::40:00 deposit increasse
Tablle 4: Time annalyse for 1hour Sample 1 and Samplle 2 extrusionn
nt as the fuunction of tiime for both
h samples
Fiigure 60 shhows calculaated die depposit amoun
from both, photo
p
camerra and videeo camera data.
d
Here, the relativee die depossit amount
w calculateed up to 10 minutes onnly where th
was
he die depoosit thicknesss is small enough
e
to
juustify 2D appproximatioon applied for die deposit amouunt definitioon (see notte above).
Frrom this graaph, it is poossible to deetermine die build-up speed as the relative die
d deposit
chhange with the time. Clearly,
C
thee die deposiit amount speed
s
for S
Sample 1
tim
mes is fasteer than for Sample 2 (iif the video
o camera daata are usedd – see Figu
ure 61 for
m
more
details)). Also, onee can clearlly see that the
t obtainedd trends aree consistentt for both,
phhoto and viddeo camera data.
73
4
DDA = 0.349
9 time
Relative DDA
Relative DDA
3
DDA = 0.227 tim
me
2
Sample 1 - Ph
hoto
Sample 2 - Ph
hoto
Sample 1 - Video
Sample 2 - Video
Sample 1 - Fitt
Sample 2 - Fitt
1
0
0
5
10
0
15
20
25
Time (min)
Fiigure 60: Meeasured Relaative Die Droool Amount as
a the functioon of time forr both samplles
0
0 min
2 min
4 min
6 min
8 min
10
0 min
SAMPLE
1
SAMPLE
2
Figure 61: Analyzed
A
vid
deo shots betw
ween 0 and 110 min
Fiinal commeent in this experimenta
e
al part is reelated to thhe unexpectted die droo
ol amount
deecrease for Sample 2 between
b
6 and
a 8 min for
fo video caamera data ((see Figure 60). This
unnexpected behavior
b
is caused by other possible errors which can occur duriing digital
im
mage analyssis if 3D efffects are neglected.
n
This
T
is visuualized in F
Figure 62 where
w
it is
74
deemonstratedd that the saame piece of
o the materrial can havve ‘differentt’ 2D surfacce area, if
thhe rotation inn the 3D sppace is not properly
p
taken into accoount.
First time ment
mom
Second time m
moment
y
x
z
Figure 622: Rotation off the die deposit in 3D sppace
75
3
THEORETICA
AL ANAL
LYSIS
In this section, thhe die buildd-up experiiments for Sample
S
1 annd Sample 2 are fol-
loowed by thee viscoelastiic finite elem
ment analyssis with thee aim to moore deeply understand
u
thhe die build--up phenom
menon for booth investig
gated samplees. For this purpose, th
he recently
prroposed theeory for diee drool onseet predictio
on (based onn the negattive pressurre idea) is
uttilized [28]. In more deetail, it has been found
d that the die
d build-up phenomenon can be
coorrelated wiith the negaative pressurre value occcurring at thhe die exit rregion wherre the free
suurface is devveloped. Inn other wordds, higher value
v
of the negative ppressure pro
omotes the
diie build-up phenomeno
p
on and vice versa.
The finite
f
elemennt method is
i a numerical techniquue used by engineers, scientists,
annd mathemaaticians to obtain
o
soluttion for partticular set of
o differentiial equation
ns that desccribe certainn process which
w
is off interest. In
n this technnique, the cconsidered domain
d
is
suub-divided into
i
a seriees of smalleer regions and
a the soluution is appproximated by
b simple
poolynomials (usually linnear for presssure, quadrratic for velocity) for eaach element of an FE
grrid using vaariational prrinciples. Thhe approxim
mation errorrs (residuals) at the nodal points
arre minimizeed during the
t elementt assembly process. Algebraic
A
soolution of very
v
large
nuumbers of equations
e
is necessary for
f the deterrmination of
o the unknoown variablees. Again,
iteerations aree needed forr the solutioon of nonlin
near probleems. In this work, com
mmercially
avvailable finnite elementt based Viirtual extrusion laboraatory ™ sooftware pro
oduced by
Compuplast Internationa
I
al, Inc. is ussed for the theoretical
t
a
analysis.
Dimension detaills for the exxperimentally used die are provideed in Figuree 63, whereeas the bounndary conditions used are
a summarrized in Tabble 5. Here, the mass fllow rate is
eqqual to 0.39 kg/h, the air
a temperatuure 22°C, th
he heat transfer coefficcient of 25 W kg-1 K1,
T1 = 121°C | T2 = 138°°C | T3 = 777°C were ad
djusted for Sample 1 w
whereas T1 = 120°C |
T22 = 134°C | T3 = 77°C
C were usedd for Samplle 2. Note, that
t
the tem
mperature prrofiles are
slightly different for botth samples, because theey were takken from thee real experriments as
thhe average value.
v
It alsso should bee mentioned
d that the free
fr surface length wass carefully
chhosen to be 100 mm which
w
was found
f
to haave no effecct on the caalculated freee surface
shhape and diee swell leveel.
76
k
100
Free surface
Free surface
l
m j
n i
o
p
q
f
e
s
d
t
Ø3
117,8
85
65
83
Die
102
104
c
Point
a
b
c
d
e
f
g
h
i
j
k
l
m
n
o
p
q
r
s
118
r
106
Die insert
Die insert
Ø0,8
h
g
a
b
Ø9.5
COORDINATES
X [m m]
Y [mm
m]
0
0
9,5
5
0
9,5
5
65
83
3
85
3
102
3
2
3
104
4
0,8
8
106
6
0,8
8
117,,8
0,8
8
118
8
0,8
8
218
8
0
218
8
0
118
8
117,,8
0
0
106
6
0
104
4
0
102
2
0
85
0
83
Figuure 63: Experrimental die dimensions
Line
fro
om
a
b
e
h
j
k
a
Condition
to
o
l
e
h
j
k
l
b
Tablle 5: Boundaary conditionns
Axis
Wall wi th temperatu
ure 1 [°C]
ure 2 [°C]
Wall wi th temperatu
ure 3 [°C]
Wall wi th temperatu
Free surface
e
Output
Input [kg/h]
77
Sppecial attenntion has beeen paid to the finite element
e
messh type whiich is capab
ble to miniimize possible errors which
w
can arise due to not enouugh mesh deensity in th
he regions
w
where
the caalculated vaariable as such
s
as velo
ocity, presssure, stress and temperature are
chhanged signnificantly (cconvergion regions
r
or die
d exit regiion where thhe free surfface is deveeloped). Wiith the aim to fulfill thhe above staated requireements, the calculation
n has been
m
made
in two steps. In more
m
detail,, firstly, thee shape andd the free ssurface locaation have
beeen determiined by the use of autoomatically generated
g
m
mesh.
With the aim to minimize
innterpolation errors (whhich is typiccal for auto
omatically generated
g
m
meshes duee to rather
raandom elem
ment distribbution) for velocity, pressure
p
strress, tempeerature variaables, the
Die
g
si n
rea ns
Inc cis i o
pre
Free surface
Free surface
strreamlined mesh
m
has beeen utilized (see Figurees 64 and 655).
Figuure 64: Sketch of the streaamlined mesh at the end of the die
78
Element
Die
Die
Free surface
Free surface
Free surface
Grid
Figuure 65: Streamlined meshh used for thee theoretical analysis
Thhe Figure 66
6 shows caalculated preessure field for the Sam
mple 1 proceessing cond
ditions. As
exxpected, thee maximum
m value of thhe negativee pressure occurs
o
at thee area wherre the free
Free surface
suurface is creeated.
‐0.85 MPa
Die
0.00 MPa
Figuure 66: Positiion of the neegative pressu
ure at the beeginning of thhe die swell
79
The most
m importtant variables such as die swell leevel (extruddate diameteer divided
byy capillary diameter) and
a maxim
mum value of
o the negaative pressurre for Sam
mple 1 and
Saample 2 aree summarizeed in Table 7.
Samp
ple 1
Samp
ple 2
Die sw
well level [ ‐ ]] Pressure [[Mpa]
‐0,8446
1,1087
75
‐2,11
1,167
75
Tablle 6: Calculaated variablees for both investigated saamples.
Intereestingly, thee predicted die swell and
a negativee pressure vvalues are higher
h
for
c
n with Sampple 1. This should
s
indiccate that Sam
mple 2 is more
m
sensiSaample 2 in comparison
tivve to the diie build-up onset than Sample 1. However,
H
this theoretiical conclussion is not
inn agreementt with the experimentaal work wheere the oppoosite trend hhas been fo
ound. This
suuggests thatt the model used in thee theoreticaal analysis is
i too simpllistic. In orrder to get
beetter agreem
ment betweeen experimeental and th
heoretical annalysis, it w
was necessaary to take
innto account the reason that the coontaminatio
ons layer duuring Sampple 2 extrusion is not
inntensively adhered to thhe die walll (see Figurre 48 and consequent ddiscussion).. This can
bee theoreticaally be donee by consideering second inlet in thhe flow dom
main for thee contaminaation layer as depictedd in Figuree 67. The only unknow
wns for the contaminaation layer
innlet determiination are mass flow rate and Newtonian
N
v
viscosity
(teemperature has been
asssumed to be
b the same as the meltt temperatu
ure at that reegion). The boundary conditions
c
foor the contamination laayer has beeen found by
y varying itss Newtoniaan viscosity and mass
floow rate unttil the maxim
mum negative pressuree value at thhe die exit region was lower for
Saample 2 in comparisonn with Sam
mple 1, i.e. if contaminnation New
wtonian visccosity and
m
mass
flow raate were 100 Pa.s and 0.01 kg/h,, respectiveely. Consideering the prresence of
suuch contamination layeer for the reference
r
prrocessing coonditions, tthe die swell was reduuced from 0.934
0
mm down
d
to 0.91 mm and the
t negativee pressure vvalue from -2.11
MPa
doown to -0.6354 MPa. This
T can be explained by
b the slip occurrence at the contamination
laayer-polymeer melt inteerface for Saample 2.
The Figures
F
68 and 69 shoows detailed
d effect of contaminattion layer Newtonian
N
viiscosity andd its mass flow
f
rate onn the negatiive pressuree value gennerated at th
he die exit
reegion for thhe referencee processing condition
ns. This clooser view ddemonstratees that the
chhoice of thee Newtoniaan viscosityy has much higher inflluence on tthe negativee pressure
vaalue calculaation than thhe mass flow
w rate of thiis layer.
Free surface
Free surface
80
Slip layeer
c ut
Input coontamination
lay
yer at 260°C
Die
Die insert
Die insert
Zo
oom
Inp
put material at 260°C
Figuure 67: Modif
ified model w
with real walll slip, figure based on sim
mulation resu
ults
81
-0.3
Pressure (Pa)
Pressure (Pa)
-0.2
-0.1
Sample 1 - No slip wall
Sample 2 - Slip wall
w
0
0
100
0
2
200
300
Newtonian viscosity of thee slipfluid (Pa.s)
400
500
Figuure 68: The effect of coontamination
n layer Newttonian viscoosity on the calculated
negaative pressurre value at thhe end of thee die for Sam
mple 2 refereence processsing conditionss.
82
-0.4
Vis
scosity: 50 Pa.s
s
Vis
scosity: 100 Pa.s
Vis
scosity: 150 Pa.s
Vis
scosity: 300 Pa.s
Vis
socosity: 500 Pa.s
Sa
ample 1
Pressure = -0.615 MFR
R - 0.270
-0.3
Pressure (MPa)
Pressure (MPa)
Pressure = -0.518 MFR - 0.203
-0.2
R - 0.13
Pressure = -0.292 MFR
R - 0.1
Pressure = -0.216 MFR
-0.1
Pressure = -0.14 MFR - 0.068
0
0.01
0.02
0.03
M
Mass flow rate (kg/h)
0.04
0.05
Figuure 69: The effect
e
of conttamination la
ayer mass fllow rate on tthe calculateed negative
presssure value att the end of the
t die for Sa
ample 2 referrence processsing conditio
ons.
83
CON
NCLUSIO
ON REMA
ARKS
In thiss work, twoo different polypropylen
p
ne samples used in thee biaxial orieented film
prroduction were
w
investiggated from die build-u
up phenomeenon point oof view. Th
he followinng conclusioons can bee extracted from the experimenta
e
al and theooretical anaalysis perfoormed in thiis work.
•
It has beeen revealed that the diee build-up phenomenon
p
n can be ob
btained for
both inveestigated saamples on th
he small laaboratory exxtrusion line equiped
by the sppecially dessigned die allowing
a
coontroled diee exit regio
on cooling
even if thhe processinng temperatu
ures were hiigher than 2200oC.
•
Based onn the digital image anallysis, it has been experrimentally found
f
that
the Sampple 1 die deeposit speeed is about 1.5 times ffaster in co
omparison
with Sam
mple 2.
•
Based onn the die drrool measurements, rhheological sstudy and theoretical
t
analysis it
i has been suggested
s
th
hat the mainn reason forr the higher Sample 2
die build--up stabilityy is occuren
nce of slip att the contam
mination lay
yer having
Newtoniaan viscosityy just aboutt 100 Pa.s, which doess not intenssivelly adhere to thhe die wall surface.
s
84
REFER
RENCES
[1]
Exxon moobil chemiccal, “Seal-In
Initiation Teemperature “,Tip from
m Technology, vol. 10, p.1, www.exxonm
w
mobilchemiical.com [oonline] [cit. 2008-0225],
available
at
WWW:
<http://ww
ww.exxonm
mobilchemiccal.com/Pubblic_Files/P
Polyethylenee/Polyeth
ylene/Woorldwide/Teechnology_ttip_10.pdf>.
[2]
CHOON K. CHAI , G. ADAM
MS AND J. FRAME
F
:“
“Polyethylen
ne die dem and routess to reductio
on” , SPE
posit – meeasurementt, formation mechanism
ANTEC Tech.
T
Paperrs, vol.47, p.401-405,
p
2
2001.
[3]
I. KLEIN
N, “Die Droool: What Causes It And
A How too Avoid It”
”, Plastics
World, p.112-113, 19981.
[4]
P.K. DHO
ORI, R.S.JE
EYASEELA
AN, A.J.GIA
ACOMIN aand J.C.SLA
ATTERY,
“Commonn Line Mottion III: Imp
plications in
i Polymer Extrusion”
”, J. NonNewtoniaan Fluid Meech., Vol.71, p.231-2433, 1997.
[5]
S.J. KUR
RTZ and S.R
R. SZANIS
SZLO, “Reeduction in Die Drooll in Filled
Resins annd Productt Improvem
ment”, U.S. Patent 50008056, U.S. Class:
264/130, IPC: B29C 047/12, Ap
pril 16, 1991.
[6]
L. VAN DEN BOSC
CHE, K. FO
OCQUET, G. DEWIT
TTE, “Die Build-Up
P
Prrocessing: A Matter off Polymer D
Design and FluoropoF
During Polyolefin
lymer-Bassed processsing Additivves”, Dyneoon Internal P
Paper, 1997
7.
[7]
J.D.GAN
NDER and A.J.GIACOM
A
MIN, “Reviiew of Die L
Lip Build-Up
Up in Plastics Extruusion”, Polyy. Eng. Sci,, Vol.37, Noo.7, p.1113 - 1126, July
y 1997.
[8]
J.D.GAN
NDER and A.J.GIACOM
A
MIN, “Reviiew of Die L
Lip Build-Up
Up in Plastics Extruusion”, AN
NTEC 96, To
oronto, p.11113-1117, M
May 1996.
[9]
S. J. KUR
RTZ, “Die geometry solutions
s
to sharkskin m
melt fracturre”. Proc.
IX Intl. Congress
C
on Rheology, Mexico, p.3399-407, 19984.
[10]
M. DAVIID SHELBY
Y and GEO
ORGE B. CA
AFLISCH, “Shear Fielld Induced Difffusion and Molecular Weight Fraactionation D
During Poly
lymer
Processinng”, Polymeer engineeriing and scieence, ISSN 0032-3888,,
vol. 44, no7,
n
p.1283--1294, 2004
4
85
[11]
P. FERRA
ADEZ and D.
D E. PRIESTER, “Neew Generatiion of Moree Economic Processs Aids for the Polyoleefin Industry
ry”, World Polyethylen
ne Conference, SP
P 92- Zurichh, vol.2, p.1--15, 1992.
[12]
D.E. PRIIESTER and C. W. ST
TEWART, “New Proccessing Add
ditives for
Polyolefinns Minimizee Formulatiion Elastom
mer Process Aids”, SPE
E ANTEC
Tech. Pappers, vol. 388, p.2024, 1992.
[13]
Well Plasstics, www.w
wellplasticss.com [onlinne] [cit. 20008-02-25], available
a
at WWW
W:
<http://ww
ww.wellsplaastics.com/h
httdocs/resoources/Wellls_Slip_Antiiblock_D
ownLoadd.pdf>
[14]
GREET DEWITTE,
D
“Polymer processing
p
additives too enlarge th
he process
window for
f polyoleffin extrusion
n”, [online]] [cit. 20088-02-25], av
vailable at
WWW:
<httpp://www.tap
ppi.org/conntent/enewslletters/eplacce/2004/8-
2dewitte.ppdf>, 2003..
[15]
K. CHAL
LOUPKOV
VA and M. ZATLOUK
KAL, “Theoretical an
nd Experimental Innvestigationn of Die Drool
D
Phenoomenon”,Phh.D. Thesiss K. Chaloupkova, p.9-18, 20007.
[16]
C. M. CH
HAN, “Effecct of surfacee coating onn the viscossity and the formation
f
of die droool at the polymer-met
p
tal surface”
”, Inter. Polym. Pros., vol. 3, p.
200- 203,, 1995.
[17]
G. R. CH
HAPMAN, D.
D E. PRIE
ESTER, R. D.
D SOUFFIIE, “Advances in the
Use of F1uoroelas
F
stomer as Processingg Aids forr Polyolefm
ms”, SPE
RETEC, Houston,
H
p.. 271-284, 1987.
1
[18]
P.S. BLA
ATZ, “Extruudable com
mposition coonsisting off a polyolef
efin and a
fluorocarrbon polymeer”, U.S. Paatent 31255447,1964.
[19]
Compuplast Internattional, Inc., “Referencce book foor short Co
ourse Feb
2003 Edittion”, Polym
mer rheolog
gy and extruusion, 2003.
[20]
CH. D. LEE,
L
“Die Lip
L Build-up
p in The Hiighly Filledd Low Denssity Polyethylene”, SPE ANTE
EC Tech. paapers, vol. 48,
4 p.264-2668, 2002.
86
[21]
D. E. PR
RIESTER, D.
D E. STIK
KA, K. M. CHAPMA
AN, R. S. MCCINN,
M
P.FERNA
ANDES, “Q
Quality con
ntrol techniqques for prrocessing additives”,
a
SPE ANT
TEC, New Orleans,
O
p.2
2528-2533, 1993.
[22]
D. BIGIO
O, M. G. MEILLON,
M
ARCHENKO
O, D. MOR
RGAN, H.
S. B. KHA
ZHOU, S.
S R. ORIAN
NI, C. W. MACOSKO
M
O, K. B. MIIGLER, “C
Coating kinetics of fluoropolym
f
mer processsing aids foor shark ski
kin eliminatiion”: The
role of droplet
d
sizee, J. Non-N
Newtonial Fluid
F
Mechh., vol.131,, p.22-31,
2005.
[23]
RIANI, G. R.
R CHAPMA
AN, “Funddamentals off melt fractture elimiS. R. OR
nation using fluoropoolymer proccess aids”, ANTEC Paapers, p.22-2
26, 2003.
[24]
G. R. CH
HAPMAN, S.
S R. ORIAN
NI, “Processs aid optim
mization in uni-modal
u
HDPE bloown film”, ANTEC Paapers, p.2719-2723, 20003.
[25]
C. M. CH
HAN, “Effecct of surfacee coating onn the viscossity and the formation
f
of die drrool at the polymer-meetal surfacee”, Inter. P
Polym. Procc., vol. 3,
p.200- 2003, 1995.
[26]
D. E. PRIIESTER annd G. R. CH
HAPMAN, “Reducing Die Build-Up in Extrusion Applications
Ap
s”, SPE Pollyolefins IX
X RETEC C
Conferencee Proceedings, Houuston, 1995..
[27]
D. E. PRIESTER, ”The Effecct of Die Suurface Charracteristics. on Proc” SPE
essing Addditive Perrformance”
AN
NTEC, Sann Francisco
o, vol.40,
p.2059-20062, 1994.
[28]
K. CHAL
LOUPKOV
VA and M. ZATLOUK
KAL, “Theeoretical an
nd experimental annalysis of the
t die drool phenomeenon for m
metallocene LLDPE”,
Polym. Enng. Sci., vool.47, p.871--881, 2007.
[29]
F. DING, L. ZHAO
O, L., A. J.. GIACOM
MIN A. J. aand J. D. GANDER,
G
“Flaring Dies to Sup
uppress Die Drool”, Polym. Eng. Sci. vol.40
0, p.21132123, 20000.
[30]
J. M. DEALY and K.
K F. WISSB
BRUN, “M
Melt Rheologgy and Its in
i Plastics
Processinng”, p.206, Van Norstrrand Rein- Role
R hold, N
New York, 1990.
1
87
[31]
D. A. HO
OLTZEN, J.
J A. MUSIIANO, “Exxtrusion Anomolies - Plate
P
Out,
polymers,, laminationns and coa
ating “, ww
ww.tappi.orgg [online] [cit. 200802-25],
availablee
at
WWW:<
ww.tappi.orgg/s_tappi/do
oc_bookstorre.asp?CID=
=5208&DID
D=50874
http://ww
6 >, 19999
[32]
HINRICH
HS D. R. (R
REYNOLD
DS METALS
S COMPAN
NY). “Diee plate for
plastics extruder”,
ex
U Patent 3933958,
U.S.
3
19976.
[33]
CHAPMA
AN G. R., PRIESTER
P
D. E., STE
EWART C. W., TARN
NEY R. E.
(E. I. DU
UPONT DE NEMOUR
RES COMPA
ANY). “Fluuoropolymeer process
aids contaaining functtional group
ps”, U.S. Patent 51323368, 1992.
[34]
ZEUS, Technical
T
whitepaperr, “Thermaal Degraddation of Plastics”,
P
www.zeuusinc.com [online]
[
[ciit. 2008-022-25], availlable at WWW:
W
<
http://ww
ww.zeusinc.ccom/pdf/Zeus_Thermaal_Degradattion.pdf >
[35]
C. M. CH
HAN, A. NIXON,
N
S. VENKATR
V
RARNAN, “Effects off the LowSurface-E
Energy Com
mponent on the Viscosity of Poly((ether etherr keytone)
and Polyttetraflouroeethylene”, J. Rheology,, vol.36, p.8807-819, 19
992.
[36]
C.W. STE
EWART, J.M. DEALY
Y, “Techniccal note: neew informatiion on the
mechanissm of actionn of the adheesion promooter reporteed in wall sllip of molten high density pollyethylene”
”, J. Rheoloogy, vol.35, p.497, 19
991; C.W.
RT, J.M. DE
EALY, “Teechnical noote: new infformation on
o the meSTEWAR
chanism of
o action off the adhesiion promoteer reported in wallslip of molten
high denssity polyethyylene”, J. Rheology,
R
vool.36, p.9677–969, 1992
2.
[37]
F. YIP, S.G. HATZIIKIRIAKOS
S, T.M. CLE
ERE, “A neew processing aid for
the extrussion of polyyolefins”, J. Vinyl Addditive Technnology, vol.6
6 , p.113–
118, 20000.
[38]
E.E. RO
OSENBAUM
M, S.K. RANDA,
R
S
S.G.
HATZ
ZIKIRIAKO
OS, C.W.
STEWAR
RT, D.L. HENRY,
H
M.
M BUCKM
MASTER, “
“Boron nitrride as a
processinng aid for the extrussion of pollyolefins annd fluoropolymers”,
Polym. Enng. Sci., vool.40, p.179–
–190, 2000..
[39]
D.R. ARD
DA, M.R. MACKLEY
M
Y, “The effeect of die exxit curvaturee, die surface rougghness and a fluoropolyymer additiive on sharkkskin extrussion insta-
88
bilities inn polyethyleene processsing”, Journnal of non--newtonian fluid mechanics, ISSN
I
0377-0257, vol.126, p.47-611 ,2005.
[40]
M. H. WAGNER,
W
“
“Analysis
of time-depeendent non--linear streess-growth
data for shear
s
and elongational
e
l flow of a low-density
l
y branched polyethylp
ene melt”
” Rheol. Acta, vol.15., p.136, 19766.
[41]
T. Y. LIU
U, D. S. SOONG.
S
an
nd M. C. WILLIAM
MS, “Time-d
dependent
rheologiccal propertiies and tren
nsient strucctural statees of entang
gled polymeric liquuids – a kinnetic networrk model”, Polym. Engg. Sci, vol.2
21, p.675,
1981.
[42]
T. OHH
HATA, H. TASAKI, T. YAM
MAGUCHI,, M. SHIINA, M.
FUKUDA
A, and H. IK
KESHITA,” Die plate for plasticss extruder”
”, U.S. Patent 5417907, 1995.
[43]
J. A. RA
AKESTRAW
W and M. G.
G WAGGO
ONER,” Poolymer extrrusion die
and use thhereof”, U.S. Patent 54
458836, 19995.
[44]
D. G. BA
AIRD and D.
D I. COLLIIAS. “Polyymer Processsing: Princciples and
Design”, Butterwortth-Heineman
nn, Boston,, p.197–2033, 1995.
[45]
S. E. AM
MOS, A. V. POCIUS, “The
“
applicaation of pollymer proceessing additives (P
PPAs) for PP film”, Dynamar Polymer P
Processing Additives
nr.26, p.33, 2001
[46]
K. B. MIIGLER, C. LAVALLE
EE, M.P. DIILLON, S.S
S. WOODS
S and C.L.
GETTING
GER, “Visuualizing thee eliminatioon of sharkkskin trough
h polymer
processinng additivess (PPAs)” , Dynamar Polymer P
Processing Additives
nr.31, p.1- 10, 2001.
[47]
G. WYPY
YCH, “Plassticizers in various
v
proocessing meethods”, Noy
yes Publications, ISBN-10: 08815514964,, Ch.7, 20033.
[48]
K.B. MIG
GLER, Y. SON,
S
F. QIA
AO and K. FLYNN,
F
“E
Extensionall deformation, coheesive failuree, and boundary condittions duringg sharkskin melt fracture”, Thhe Society of Rheology, vol.46(2), p.383-400,, 2001.
89
[49]
KTRON process grroup, “App
plication exxample biaxxially- orieented film
productioon”, www..ktron.com [online] [cit.
[
2008-002-25], avaailable at
WWW:
<http:://www.ktron.com/indusstries_serveed/Plastics//Biaxially-
Oriented__Film_Prodduction_BO
OPP.cfm>, 2007.
2
[50]
Special Chem,
C
www
w.specialch
hem4polymeers.com [onnline] [cit. 2008-0227],
available
at
WWW:
ww.specialcchem4polym
mers.com/tcc/polymer-pprocessing<http://ww
additives//index.aspx?id=die_bu
uildup>
[51]
R. DANG
GTUNGEE, S. S. DES
SAI, S. TA
ANTAYANO
ON, P. SUP
PAPHOL,
“Melt rheeology and extrudate swell
s
of low
w-density poolyethylene//ethylene–
octene cop
opolymer bleends”, Poly
ymer Testingg, vol.25(7)), p.888-895
5, 2006.
[52]
H. W. MULLNER,
M
, J. EBERH
HARDSTE
EINER, W.. FIDI, “Rhheological
characterrization of the
t die swelll phenomennon of rubbber compou
unds”, Polym. Testiing, vol.31, p.1041-1048
8, 2007.
[53]
A.C.-Y. WONG,
W
“F
Factors affecting extruddate swell aand melt flo
ow rate” ,
Journal of Materials Processing Technologyy, vol.79, pp.163-169, 1998.
1
[54]
A.C.-Y. WONG,
W
J.Z
Z. LIANG, “Temperatture and prressure effeccts on the
melt flow
w rate and melt density
ty of high-ddensity polyyethylene”, J. Mater.
Process. Technol.,
T
vool.43, p.293
3, 1994.
[55]
C.D. HA
AN, “Rheoloogy in Poly
lymer Proceessing”, Accademic Prress, New
York, 19776.
[56]
S. TANO
OUE, T. KA
AJIWARA,, Y. IEMO
OTO and K
K. FUNATS
SU, “High
Weissenberg Number Simulatio
on of an Annnular Extruudate Swell Using the
Differentiial Type Coonstitutive Equation”,
E
P
Polym.
Engg. Sci., vol.3
38, p.409,
1998.
[57]
HAMIEL
LEC and VL
LACHOPOU
ULOS, “Inffluence of llong chain branching
b
on extruddate swell of
o low-density polyethyylenes”, Jouurnal of Ap
pplied Polymer Sciience, vol.28, p.2389-2
2392, 1983.
[58]
DELVAU
UX V., CRO
OCHET M. J. 'Numerrical simulaations of deelayed die
swell', Rhheologica Acta,
A
vol.29, p1-10, 19990.
90
[59]
Well Plasstics, www.w
wellplasticss.com [onlinne] [cit. 20008-02-29], available
a
at WWW
W:
http://ww
ww.wellsplasstics.com/htttdocs/resouurces/Wellss_Process_A
Aids_Dow
nLoad.pddf
91
LIST OF ABBR
REVIATIIONS
BOPP
P
BiiOriented PolyPropylen
ne
MWD
D
M
Moleculair
W
Weight
Distrribution
GPC
Geel Permeatie Chromato
ografie
PPA
Poolymer Proccessing Aids
92
LIST OF
O FIGUR
RES
Fiigure 1: A typpical biaxiallly-oriented polypropylen
p
ne film proceess factory laayout [49] ................. 11 Fiigure 2: Chaanges in the polymer
p
chaiin orientation
n during at particular
p
steep during bia
axial
orrientation off PP film. ....................................................................................................................... 12 Fiigure 3: Struucture of a typ
ypical 3-layerr BOPP film ......................................................................... 12 Fiigure 4: Scheematic simpllification of the
t heat sealiing process and
a the postuulated moleccular
prrocesses invoolved [1] ...................................................................................................................... 13 Fiigure 5: Extrrusion of LLD
DPE-A from a pipe die with
w die lip buuild-up indiccated [2]. .................. 14 Fiigure 6: Extrrusion of high
gh density pollyethylene. 6a)
6 a clean diie face, 6b) ddie lip build--up
occcurrence [8]]................................................................................................................................... 15 Fiigure 7: Extrrusion of poly
lyvinyl chloriide.7a) a clea
an slit die face, 7(b) die llip build up
occcurrence [8]]................................................................................................................................... 15 Fiigure 8: Die Build-Up duuring PP cappillary extrussion [45]. Lef
eft side – cleaan die face, right
r
hand
sidde – die lip build-up
b
occuurrence. .................................................................................................... 15 Fiigure 9: Die Build-Up in Nylon [50]. Left side – clean
c
die face, right handd side – die lip buildupp occurrencee. ................................................................................................................................... 15 Fiigure 10: Sum
mmarizationn of all possibble factors leeading to diee build-up phhenomenon [15,
[1 20] . 16 Fiigure 11:Thee effect die deesign and shear strain ra
ate on the rubbber compouund swell va
alue [52] (
diie diameterD
D = 2mm, die temperaturee equal to 12
20°C.) ................................................................ 18 Fiigure 12: Theoretically predicted
p
effeect of Weisseenberg number, We, on thhe annular exxtrudate
sw
well (in termss of streamlinne field) [56]] ........................................................................................... 19 Fiigure 13: Schhematic reprresentation of the sequence of deform
mations of a m
material as itt enters,
floows trough, and
a emergess from a die [57]
[ ...................................................................................... 20 Fiigure 14: Extrusion of LL
LDPE from capillary
c
die,, showing diee swell and ddie deposit as a
fuunction of exttruder outputt [2] .......................................................................................................... 21 Fiigure 15: Floow in the vicinity die exitt based on [9
9]....................................................................... 22 Fiigure 16: mL
LLDPE extruudate surfacee appearancee classificatioon at differennt flow cond
ditions. (a)
Sm
mooth string,, (b) Slight shhark skin, (c)
c) Shark skin.. [28] ................................................................ 23 Fiigure 17: Skeetch of the shharkskin insttability kinetiics, side view
w [48] ........................................... 23 Fiigure 18: Froom sharkskinn to die depoosit ........................................................................................ 24 Fiigure 19: The effect of the extrusion output
o
on thee die deposit level for LLD
DPE ......................... 25 Fiigure 20: Diee drool appeearance for different
d
walll temperaturees [28] ......................................... 26 Fiigure 21: The effect of die swell on diie deposit du
uring externaal cooling; a. Small die sw
well, b.
Biig die swell ....................................................................................................................................... 27 Fiigure 22: Geeneral mechaanism for theermal degrad
dation [34] ........................................................ 28 93
Fiigure 23: The effect of PP
PA on the diee drool phen
nomenon (leff side) and finnal film finish [14,19]
............................................................................................................................................................ 31 Fiigure 24: Possible bounddary conditioons at the wa
all-polymer innterface [52]]................................ 32 Fiigure 25: Vissualization of the stateareete role in po
olymer matriix. 25a). steaarate << crittical
am
mount, 25b). stearate >>
> critical amoount [20].............................................................................. 33 Fiigure 26: Cap
apillary shearr viscosity daata measured
d at 270°C for
fo five differrent LLDPE resins
r
and
auutocalve LDP
PE type. ....................................................................................................................... 34 Fiigure 27: Strraight, flaredd, convergingg and shaped
d die [8, 32] ....................
.
................................. 35 Fiigure 28: Expplaination byy Gander annd Giacomin about the staabilization efffect of a flarred die [8]
............................................................................................................................................................ 35 Fiigure 29: Eff
ffect of die exit angle on thhe die drool amount for mLLDPE
m
[155] ............................. 36 Fiigure 30: Preedicted suctiion pressure and pressure gradient ass the functionn of the cham
mfer length
[228] .................................................................................................................................................... 37 Fiigure 31: Moolecular weigght data as a function off milling depthh (depth = 0 refers to th
he outer
eddge) for the extruded
e
PET
T film [10] ................................................................................................ 38 Fiigure 32: Diee lip build-upp ratio for flaared, straigh
ht and converrging dies inn film blowing
g
exxperiments [229] ............................................................................................................................... 39 Fiigure 33: The predicted pressure
p
fieldd for the extrrusion of mLLDPE [28]uunder die dro
ool
coonditions ........................................................................................................................................... 40 Fiigure 34: Sucction pressurre at the endd of the die ass the functioon of mass floow rate and
tem
mperature [228] ............................................................................................................................... 41 Fiigure 35: Deescription of the main AR
RES parts .............................................................................. 45 Fiigure 36: Deetail view of the
t sample loocation betw
ween two paraallel plates. .................................. 46 Fiigure 37: Deescription of the Rosand RH7-2
R
capilllary rheometter ................................................ 47 Fiigure 38: Lonng die (left side)
s
and shoort die (right hand side) ....................................................... 48 Fiigure 39: Eqquipments useed for the diee build-up evvaluation........................................................... 49 Fiigure 40: Extrusion die constuction
c
................................................................................................ 50 Fiigure 41: Parts of the annnular extrusiion die ................................................................................. 50 Fiigure 42: Dim
mensions of the extrusionn die insert (flow
(f
directioon is from rigght to left) ................. 51 Fiigure 43: Eqquipment arraangement forr the die builld-up measurrements. ....................................... 52 Fiigure 44: Coomparison beetween measuured and preedicted tempeerature depeendent shear and
exxtensional visscosities for Sample 1. ................................................................................................. 56 Fiigure 45: Coomparison beetween measuured and preedicted tempeerature depeendent shear and
exxtensional visscosities for Sample 2. ................................................................................................. 57 Fiigure 46: Coomparison beetween shearr and extensional viscositties for Sampple 1 and Sam
mple 2 at
2220°C. ................................................................................................................................................ 58 94
Fiigure 47: Preessure fluctuuations of HD
DPE at 170oC ....................................................................... 59 Fiigure 48: HD
DPE pressuree profiles duuring extrusio
on through clean
c
as well as Sample1//Sample2
coontaminated capillary diees. ............................................................................................................. 60 Fiigure 49: Sett test conditioons ........................................................................................................... 63 Fiigure 50: Scrrew speed reegulator..................................................................................................... 63 Fiigure 51: Tim
me evolution of pressure and tempera
ature during die drool annalysis for Sa
ample A
annd Sample B at chosen prrocessing connditions ............................................................................... 64 Fiigure 52: Polymer melt leeakage from the die at veery high extrrusion pressuures for Samp
ple 2. .... 65 Fiigure 53: Sam
mple 1 and sample
s
2 extrrudates for th
he identical processing
p
cconditions ch
hosen for
diie buil-up evaaluation ....................................................................................................................... 66 Fiigure 54: Tw
wo front anglle views for digital
d
imagee analysis .......................................................... 67 Fiigure 55: Cooordinate sysstem transforrmation ................................................................................. 68 Fiigure 56: Cooordinate sysstem transforrmation for die
d deposit ......................................................... 69 Fiigure 57: Diee build-up onnset evaluatiion for Samp
ple 1 and Sam
mple 2 ........................................... 69 Fiigure 58: Photographs too find out thee different shape of the booth samples .................................. 70 Fiigure 59: Diee build-up chharacter for Sample 1 an
nd Sample 2 at
a the same pprocessing co
onditions
............................................................................................................................................................ 71 Fiigure 60: Meeasured Relaative Die Droool Amount as
a the functioon of time forr both samplles ......... 73 Fiigure 61: Analyzed videoo shots betweeen 0 and 10 min .................................................................. 73 Fiigure 62: Rotation of the die deposit in
i 3D space ......................................................................... 74 Fiigure 63: Expperimental die
d dimensionns ......................................................................................... 76 Fiigure 64: Skeetch of the sttreamlined mesh
m
at the en
nd of the die .................................................... 77 Fiigure 65: Strreamlined meesh used for the theoreticcal analysis ....................................................... 78 Fiigure 66: Position of the negative preessure at the beginning off the die sweell .............................. 78 Fiigure 67: Moodified modeel with real wall
w slip, figu
ure based on simulation rresults ....................... 80 Fiigure 68: The effect of coontaminationn layer Newto
onian viscosiity on the callculated negative
prressure valuee at the end of
o the die forr Sample 2 reeference proccessing condditions. ...................... 81 Fiigure 69: The effect of coontaminationn layer mass flow
f
rate on the calculatted negative pressure
p
vaalue at the ennd of the die for
f Sample 2 reference processing
p
coonditions. ..................................... 82 95
LIST OF
O TABLES
Taable 1: Modif
ified White Metzner
M
paraameters for Sample
S
1 andd Sample 2 .................................... 55 Taable 2: Overvview of the set
s temperatuures to get th
he processingg window ...................................... 61 Taable 3: Effecct of Texit tempperature andd mass flow rate
r on the diie build-up pphenomenon
(ccorresponding T1-T6 valuues are proviided in Tablee 2, experimeent No.9). ..................................... 62 Taable 4: Time analyse for 1hour Sampple 1 and Sam
mple 2 extrussion............................................... 72 Taable 5: Bounndary conditiions ........................................................................................................... 76 Taable 7: Calcuulated variabbles for both investigated
d samples. ......................................................... 79 96
LIST
T OF APP
PENDICE
ES
1. DVD withh video, phooto materiall, graphs and PDF version of this m
master thesiis
Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement