Processing of Polyethylene and Polypropylene Foams in Rotational Molding
Fangyi Liu
A thesis submitted in conformity with the requirements for rhe degree of Master of Applied Science Graduate Department of Mechanical & Indusnial Engineering University of Toronto
$3 Copyright by Fangyi Liu (1 998) National Library Bibliothèque nationale 1+1 ,,.da du Canada Acquisitions and Acquisitions et Bibliographie Services services bibliographiques 395 Wellington Street 395, rue WdlingtOn OMwaON KlAON4 OctawaON K1AON4 CaMda Canada
The author has granted a non- L'auteur a accordé une licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/nlm, de reproduction sur papier ou sur format électronique.
The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or otheIWise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation. Processing of Fine Celi Polyethylene and Polypropylene Foams
in Rotational Molding
Fangyi Liu
LM.A. Sc., 1998, Department of d mec ha ni cal and Indusaial Engineering
University of Toronto
Abstract
The goal of thts research is to develop a scientific and engineering basis needed for
production of therrnoplastic foarns in rotomolding using a chernical blowing agent (CBA).
Both polyethyicne (PE) and polpropylene (PP) were used for the foaming process, but a
specid ernphasis was givcn to Pl?, duc to its many advantages. Two rypes of methods were
uulized in the research, narnely the dry blending method and the melt compounding method.
Dsîomposition behaviors of different CBAs were analyzed with a Thermogravimerric hnalvzer (TGA) and a Differential Sca~ingCalorimeter @SC). The zero-shear viscosities
of polymers that determine the sintering behavior of plastic powder particles were studied with a rotaaonal rheometer. The mechanisms that govcrn the ceil nucieation, sintering, and
ce11 distribution for both the dry blending and the melt compounding methods wcre
clarified. The criacal proccssing parameters that govern the mi'cing, foaming and skin
formation were outiïned. Based on these investigations, conclusions were drawn on the
selection of polymer material, application of CBA, and the determination of processing conditions. Acknowledgements
First 1 would iike to espress my sincere gratitude to rny supervisor Professor C. B.
Pzrk for providing guidance and encouragement throughout my research. I wiil never forget the help he gave at my most difficult times. Aiso, 1 would iike to thank WedTech for their funding and support for this project.
h:fy gratitude is estended to the Department of Mcchanical and Industriai Engineering at thc University of Toronto for pro\-iding the Universir}. of Toronto Alaster's Opcn
Fcllowships. It was these financial assistance that enabled me to tinish my study.
1 Special thanks are due Professor B. Benhabib and Brenda Fung for their great help during mv ciifficuit times and throughout the study at the University of Toronto.
1 wodd also like to thank my coiicagues in the hlicroceiiular Plastic manufacturing
Laborator?. for their help and friendship over the past nvo years. They includr Guobin Liu,
Dr. Yucjian Liu, An~lionyYeung, Arnir Bchravesh, Ghaus Rimi, Remon Pop-lliev, Deepak
Fernandes, Hani Naguib, Drnitry Ladin, Simon Park, Haiou Zhang. Also 1 want to acknowledge the assistance from Marianne Chan. Table of Content
.. ABSRACT ...... ri ... Acknowledgements ...... IU
Table of Content ...... iv
List of Tables ...... ?c
List of Figures ...... si
Xomenciature ...... SV
Chapter 1 Introduction ...... 1
1.1 Rotationai Molding ...... 1
1.1.1 Histoncd Review ...... 2
1.1.2 Characteristics and Application of Rotational Molding ...... 3
1.1.3 Rotational Molding Machine ...... 4
1.2 Polyethylene and Polypropylene ...... 5
1.2.1 Polyethylcne ...... 5
1.2.2 Polypropylene ...... 5
1.2.3 Homopolymer and Copolymer ...... 5
1.2.4 Linear and Branched Polypropylene ...... 6
1-3 Plastic Foams ...... 7
1.3.1 Foams ...... 7
1.3.2 Principal Foam Production Techmlogies ...... 7
1.4 Rotational Foarn Molding ...... 8 1.5 purpose Sutement ...... 9
1.6 Overd Strategy ...... 10
1.7 Oudine of the Thesis ...... 10
Chapter 2 Theoretical Background ...... 12
2.1 Polymer properties ...... 12
2.1.1 Viscosiq and Elasticity ...... 12
2.1.2 Melt Flow Index ...... 13
2.1.3 iMelt Strengh ...... 13
2.2 Chernical Blowing Agent ...... 14
2.2.1 Classification of CBhs ...... 14
2.2.2 Endothennic and Esothennic Behaviors ...... 15
2.2.3 Decornposition Rate ...... 16
2.2.4 Sclection of CBAs ...... 16
2.3 &li>dngTechnolo_e)- ...... 17
2.3.1 Dry Blçnding Equipment ...... 17
2.3.2 Melt Compounding Equipmenr ...... 18
Chapter 3 Analysis of Materials ...... 23
3.1 *Material Requirements of PE and PP in Rotational Foam hlolding ...... 23
3.1 -1 Viscosi- and Elasticiq-...... 23
3.1 -2 Plastic Powders ...... 24
3.2 Materials Used ...... 26
3.3.Viscosity Analysis of PE and PP ...... 26
3.4 DSC Analysis of PE and PP ...... 27
3.4.1 Experimend Semp ...... 28 3.4.2 Results and Discussion ...... 29
3.5 Thermal Analysis of CBAs ...... 29
3-51 Experimental Setup ...... 29
3.5.2 Results and Discussion ...... 30
? - 3.6 Selection of CBAs ...... 33
Chapter 4 Rotational Molding of PE and PP Foarns
Mth the Dry Blending Method ...... 36
4.1 Introduction ...... 36
4.2 Overail Processing Technolog)...... 37
4.3 Processing Steps in Rotational Foarn Molding ...... 37
3.3.1 Continuous Phase Formation by Sinteriq
of Powder Particles ...... 37
4.3.2 Ceii Nucleaaon in the Dry Blending Method ...... 38
4.3.3 Ce11 Growth and Cell Coalescence ...... 38
4.3.4 Cell Coarsening...... 39
4.4 Esperimentaaon...... 40
4.4.1 Espenmend Procedure ...... 40
4.4.2 Expenmental Setup ...... 40
4.4.3 G~dingof Plasac Materials ...... 40
4.4.4 Hot Stage iMicroscopic Analysis ...... 41
4.4.5 Rotational Foam Molding ...... 41
4.4.6 CharactcrizaUon of Foams ...... 43
4.4.7 Temperature Profile ...... 44
4.5 Results and Discussion ...... 44 4.5.1 Investigation of the Temperarure Profile
in Rotational Foarn Molding ...... 44
4.5.2 Investigation of Sintering of Powder Particles
and its Relationship with Ceii Nucleation ...... 46
4.5.3 Investigation of Effect of Powder Qualities
on Ceii Morpholog)...... 50
4.5.4 Investigaaon of CDhs on Ce1 MorphoIog): ...... 51
4.5.5. Investigation of Effect of CBA Particle Size
on Ce11 Morpho10 g'...... 52
4.5.6 Investigation of Effect of Zn0 on CeU Morpholog?...... 53
4.5.7 Investigation of Proccssing Pararnctcrs on Ceii Morphology .... 53
Chapter 5 Rotaaonal Molding of PE and PP Foarns
with the Melt Compounduig Method ...... 59
5.1 Introduction ...... 59
5.2 Overview of the Process ...... 60
3.3 Processing Steps in Rotational hlolding
5.3.1. Cell Nucleation ...... 61
5.3.2. Sintering of Pellets and Ceii Growth ...... Gt
5.4 Esperimentation ...... 62
5.4.1 Esperimental Setup ...... 62
5.4.2 Materials ...... 62
5.4.3 Melt Compounding ...... 62
5.4.4. Ro tational Foarn Molding ...... 64
5.4.5 Hot Stage hlicroscope Analysis ...... 64
vii 3.5. Compounding Results and Discussion ...... 64
5.5.1 Effect of Barrei Temperature on Compounding Qualicy ...... 66
5.5.2 Effect of Rotationd Speed of the Estnision Screw
on Compounding Qualit)...... 66
5.5.3 Effect of Screw Cooling on Compounding Quality ...... G7
55.4 Esothermic and Endothermic Effect of CBAs
on Compounding Quality ...... 67
5.5.5 Effect of Heaang rate on Compounding Quaiicy ...... 68
5.6 Foarning Results and Discussion ...... 69
5.6.1 Foarning with Non-Decomposed Pellets ...... 69
5.6.2 Foaming with Pre-Decomposed Pellets ...... 73
56.3 Effçct of the Ovcn Temperature on CeIl Morpho10 gy ...... 74
Chapter 6 Formation of Unfoamed Skin ...... 76
6.1 Introduction ...... 76
6.2 Overall Process ...... 77
6.3 Esperïmentacion ...... 77
6.3.1 Esperimental Setup ...... 77
6.3.3 Procedures ...... 78
6.4 Results and Discussion ...... 80
6.5 Concluding Remarlis ...... 82
Chapter 7 Conclusions and Recornmendations for Future Work ...... 83
7.1 Conclusions on Dry Blending Based Technology ...... 83
7.2 Conclusions on Compounding Base Technology ...... 85
viii 7.3 Condusions on Skin Formation ...... 88
7.4 Suggestions for Future Work ...... 89
References ...... 90 List of Tables
Table 1-1 Density of Different Grades of PE ...... 93
TabIe 3-1 PolyethyIene hIaterials ...... 94
Table 3-2 Polypropylene hlaterials...... 94
Table 3-3 Chernical Blowing Agents Used in the Esperimcnts ...... 94
Table 5- 1 Processing Parameters and Results ...... 95
Table 6-2 EffectofMarerialAmount...... 96 List of Figures
Figure 1. 1 Schematic of Rotational hlolding Process ...... 96
Figure 1-2 Sintering in Rotational Molding ...... 96
Figure 1-3 Examples of Rotational hiolding Machine (Courtesy of Crawford II j) ...... 97
Figurc 1-4 Isomerism for Positions in Poljpropylene ...... 97
Figurc 1-5 Producing a Foam Core with a Dropping Bos ...... 98
Figure 1-6 Overd Stratcgy...... 99
Figure 2-1 Schemaac of MF1 Measurement ...... 100
Fipre 2-2 Effect of Endotherrnic and Esothermïc Behavior ...... 100
Figure 2-3 Characterization of CBAs ...... 101
Figure 2-3 Cylinder Drum hlixer ...... 101
Figure 2-5 Different Shape of Dry Blenders ...... 102
Figure 2-6 Blender ~4thDeflectors ...... 102
Figurc 2-7 Two-Roll Mill ...... 102
Figurc 2-8 Basic Structure of single Screw Esuuders ...... 103
Figure 2-9 Configuration of the Scrcw ...... 103
Figure 2-1 0 Blockhcad hiïxing Head ...... 103
Figure 3-1 Typical Plastic powder shapes ...... 104
Figure 3-2 Rotationai Stress Rheometer ...... 104 Figure 3-3 Rheological Properties of Polypropylenes ...... 105
Figure 3-4 TA291 0 Di fferentid Scanning Calorimeter (Illusmtion Courtesy ofT-1 Instrument) ...... 106
Figure 3-5 Schemaac of DSC Module (~UustrationCourtesy of TA ~nstrument)...... 106
Figure 3-6 DSC Thermogms of PE Materiais ...... 107
Figure 3-7 DSC Thrmograms of PP Materials ...... 108
Figurc 3-8
Figure 3-9 DSC Anaiysis of CBA Decomposition ...... 112
Figurc 3-1 0 Effect of Heating Rate on DSC Thermogram of CBAs ...... 115
Figure 3-1 1 TGA Analysis of CBA Decomposition Behaviors ...... 116
Figure 3-1 2 Effecr of Heating Rate on the Decomposition Behavior ...... il9
Fiprc 3-13 Effect of Zn0 on Decomposition of Celogen AZ3990 119 (Hcating Rate 10 "C/rnin) ......
Effect of CBA Pamcle Size on Decomposition of ADC Series CE3 A ...... ~...... *...... -.-..*...... *...... 120 Figure 3- 15 Effect of CB-A Amount (Sarnple Size) on Decornposition of Celogen OT...... 120 Figure 4-1 Schematic of Sintering and Foarning Process in Rotauonai ~Molding...... 121
Figure 4-2 Csll Coalescence ...... 121
Figure 4-3 Uni-axial Rotational Molding Machine ...... 122
Fiprc 4-4 Grinding...... 123
Fiprc 4-5 Schematic of Cell Population Density Caiculation ...... 123
Figure 4-6 Polyethylene Foams of Three-fold Expansion ...... 124
xii Figure 4-7 Temperature Profile in Rotational Molding ...... 125
Figure 4-8 Cross Sections of Sintered S-mples SC873 (50) + 3% Celogen AZ 3790 (sis-fold) ...... 126
Figure 3-9 Cross Sections of Sintered Sarnples SD242 (30) + 3% Celogen AZ 3990 (sis-fold) ...... 126
Figure 4-10 Cross Sections of Sintered Samples of MT4390 (MFR 20) + 3% Celogen A2 3990 (sis-fold) ...... 126
Figure 4-1 1 Cornparison of Difkrent hfaterials (thrcc-fold) ...... 127
Figure 4-12 Effect of Material Viscosity on Cell Size. Cell Population Density and Volumc Espansion Raao of Three-fold Espanded PP Foams ...... 128
Figure 4- 13 Effect of Polymer Particle Sizes on Cell Morphology ...... 129
Figure 4-14 Effect of Long Tads on Ce11 Morphology ...... 129
Figure 4-1 5 Effect of CBA on the Ce11 Morphology of PE Foams ...... 129
Figure 4-1 6 Effect of CBA Particle Size on the Ceii Morpholog)...... 130
Figure 4-1 7 Effect of Zn0 on Cell Morphology of lvlT4390 Foams ...... 130
Figure 4- 16 Effect of CBA Amount on CeU Morp hology ...... 130
Figurc 4-18 Effect of CBA hmount on CeU Morphology of Thrce-fold Espanded MT4390 Foams ...... 131
Figurc 4-19 Effect of CBA Anount on Ceil Size. Ccil Population Densiq- and Volume Espansion Ratio of Three-fold Expanded MT4390 Foams ...... 132
Figurc 4-20 Effect of Processing Time on Cell Morphology of MT4390 Foams ...... 133
Figure 4-21 Effect of Processing Time on Cell Size. Ceii PopuIation Densiry and Volume Espansion Raùo of Three-fold Espanded MT4390 Foams ...... 134
xiii Figure 4-22 Effect of Oven Temperature on the CeLi Morphology ...... -......
Figure 4-23 Self Heating Behavior of Exothemiic CBAs ......
Fipre 4-24 BalLing Phenornenon .-......
Figure 5-1 Melt Compounding System ......
Figure 5-2 Schematic of Heat Eschange in Estlusion (Courtesy of Stevens [23])...... ,...... -...... -...... -
Figure 5-3 Effect of Residence Time on Decornposiaon of Celogen A23990...... - ......
Figure 5-4 Comparison of the Ceil Morphologies Obtained from the Dq Blending and Compounding Methods ( Three-fold Expansion) (Three-fold Expansion) ......
Figure 5-5 Comparison of Dry Blending and Compounding mree-fold Espansion) ...... -...... -...... -......
Figure 5-6 Comparison of the Cell Morphologies Obtained from the Dn Blcnding and Compounding Methods. The Foams Were to bc Espanded SLY-fold......
Kgurc 5-7 Comparison of Compounded Peiiets of PP and PE (Six of Sample: 2.0 mm s 1.5 mm) ......
Figurc 5-8 Foaming Process of Pre-decornposed Peiiets (LLDPE 8556 + 3% Celogcn OT, Six of Samplc: 2.0 s 1.5 mm) ......
Figurc 5-9 Effect of Oven Temperarure on Cell h.lorphology (PF633 + 2.5% AZ3990, Sis-fold expansion, Scdc: 1:1) ......
Figure 6-1 Effect of Material
Figure 6-2 Effect of Oven Temperature (Sale 1:1) (Skin: 15g SC1355, Foam: 1On PF633 + AZ3990) ......
xiv Nomenclature
b channel width (m)
CBA chernical blowing agent
D diameter of the extruder screw (cm)
DSC Differenaal Scanning Calorimecer
h flight height (cm)
hskin thickness of the skin layer in the formation of skin and foam core
structure (cm)
I edge length of extruder screw(prn)
L length of the mold (cm)
MFR Melting Flow Index (g/lOrnin)
m mass of foamed sample (g)
~CBA amount of CBA material used (g)
mptrllet amount of compounded pellets used for rotomolding in compounding
based method (g)
mpolyrncr weight of pure polymer material used in dry blending based method
(g)
mskin amount of powder materials used for the skin layer in the formation of
skin and foam core structure (g)
N rotational speed
No ce11 density (#/cm3)
nb number of bubbles pressure (Pa)
pitch (cm)
heat flow to the constantan pan of the DSC ceIl (w)
drag flow in a single screw extruder(m3/s)
leakage flow due to drag flow in a single screw extruder (m3/s)
leakage flow in a single screw extruder (m3/s)
pressure flow in a single screw extruder (m3/s)
leakage flow due to pressure in a single screw extruder (m3/s)
first pnnciple radius of curvature (cm)
second principle radius of curvature (cm)
thermal resistance of the constantan pan of the DSC ce11
flight width of a single screw extruder (cm)
crystallization temperature of pol ymer("C)
room temperature (OC)
TGA - Therrnogravirnetrïc Analyzer
volume of the mold (cm3)
volume of gas generated by CBA decomposition (at polymer
crystallization temperature, cm3)
expanded volume (cm3)
volume of gas generated by CBA decomposition (at room
temperature, cm')
extensionai strain rate
xvi shear strain rate
extensional strain
extensional viscosi ty
shear viscosity
shear stress
ex tensional stress
shear strain
helix angle
volume expansion ratio density of n~aterial(~/crn~) volume of sas generated per unit rnass of blowing agent pressure difference between the two cells density of polymer used (g/cm3) volume of gas generated per unit mass of blowing agent at room temperature given by CBA manufacturers (cm3ig) temperature difference between the two pans of the DSC ceIl
xvii Chaptet 1 Introduction
Chapter 1
Introduction
1.1 Rotational Molding
Rotacional molding is a method of produung suess-free parts with complicated shapcs, especially those with hoiiow structures. The concept of the process can be described as foliows: a metai mold, charged with a predetermined amount of cold plastic powders and addiaves, is put into a preheated oven and rotated bi-auiaily. Once in the oven, the metal mold is heated up very quiclcly and the plastic particles begin to melt and sinrer while tumbling in che mold. After the designated tirne for rotaaon in the oven had elapsed, the mold is pded out of the oven. At dis point of cime, the plastic powders have already forrncd a uniform melt suucmre. The mold is then cooled to room temperature, without stopping the rotation, und the part is fuiiy solidified. Finally, the moid is opened and the formed part is removed. Thc rotational molding process is illustrated in Figure 1-1.
hlthough rotational molding has sidaricies with cenuifugal casting, rotational rnolding is a distinct and unique process. The most important ciifference benveen the nvo is the magnitude of the rotationai speed that is applied on the mold during the proccss.
Rotational rnolding is characterized with much lower rotational speeds chan the centrifugai casting. Normally, in rotational molding, the speed is in the range 5-20 RPM. Therefore, as opposed to centrifuga1 casting, the plastic particles remain at the bottom of the rotating cavity of the mold, due to the gravir).. As a result of the higher range of rotational speeds applied. in centrifugai casting, a significant centrifuga1 force is developed. This is why, in centrifuga1 casting, the melted material is forced to flow in the radial direction in the mold. Chapter 1 Introduction
In conuast, wvith rotational molding the particles stay in loose pouder form und the surface of the mold rcaches the temperature level that is hiçh enough for the partides to begin to adhcrc or sinter to the mold and/or to each other, at thc layer nearest to the surface. As the temperature increases, a uniform plasac melt structure is formed. The sintcring process in rotauonal molding is ïiiustrated in Figure 1-2.
1.1.1 Historical Overview [l,21
Thc first work on producing hollow structures by heat and bi-asid rotation is recorded as early as in 1865, but the first equipment alike a rotational molding machine of today appeared in a British parent of 2935 [Il. However, ar chat cime, suitable materials that could be used for the process were not avaiIable, except some ceiiulosics materials. This situation rcmained unchanged since 1961 when finely ground polyethylene by CS1 (Quantum
Corporation of today) \vas used for thc first Ume in rotational moiding 131. With the introduction of polyethylene, the industry esperienced a great expansion. Bu the rnid-1970~~ rotauonai molding was aireadp one of the primary plasacs processing technologies in the industry. After the mid-1Vos, the rate of gro\vth of rotaaonal molding market was much fastcr compared to the rest of the plastic industry. In recent years, che rotational molding market is growing with a rate of 10-12% per annum. Athough the traditional toy indus? is still the biggest part of the marker, more applications arc found in other areas such as tanks and containers, and automotive parts [3].
Nowadays, with the improvement of the technology of materiai manufacniring, polycth ylene (PE), polypropylene (PP), polycarbonate, fluoropol ymers, PVC, Nylon, and othcr materials find broad applications in the rotationai rnolding industry. Among thern, PE Chapter 1 Introduction is sdthe most popular material for rotational molding. PE accounu for about 85% of the application. PP covers only a smaii section of the market [3].
1.1.2 Characteristics and Application of Rotationai Molding
As a unique plastic processing technology, rotational molding has many advamages over many other methods [Il.
1. The process cm be used to produce parts of a wide range of sizes and shapes.
Currendy, products made with this technology can be as small as a doli's cyc or as
big 2s a 79,500 iiter tank [2].
2. The initial investments for moIds and the necessary manufacturing equipment are
relatively low. Since the rotational molding process is often under very 1ow
pressure, the molds are inexpensive and the difticulty of mold manufacturing is
dramaticdy reduced. Espensive high-strength molds, which are essenaal for the
injecaon rnolding process, are not necessary in rotaaonal molding. As a result,
rcducuon of overall production costs can be ensily achjeved.
3. Bccause of the iow-pressure proccssing environment, it is easy ro make stress-free
pans, if necessary prccautions were taken as appropriate.
4. The rotational molding process is especially suitable for making searnless hollow
structures with a very uniform skin diickness. Very complicated profües and
structures, such as core-skin structure, can be easily generated from rotational
molding, especidy for Iarge parts. In addition, parts of different sizes and
structures cm be processed simuItaneously within one batch. Chapter 1 Introduction
However, cotational molding also has disadvantages that resuict its development md application. It is believed that the potenaal of the process has not been fully utilized yet bccause of thcse disadvaritages-
The slow producaon rate, among ail the disadvantages, is the most important one.
Since the mold has to be heated to the processing temperature chat wiil melt rhe
plastic materials and then cooled down to room temperature again, the cycle urne
is long and is not suitable for smali parts that have to bc produced in large
quantines.
The material used for rotauonal molding must be in fine powder form becausc thc
size of the powder plays an important role in deterxnining the part quality. This
increases thc material costs dramatically.
The number of materiais that cm be used for rotaaonal molding is limited.
Currently, polyeth ylene is the most commonly used one.
Soiid rib structures are difficulr to produce using rotational molding. Thercfore,
estra attention should be paid to the design stage for Structures that need saffness.
1.1.3 Rotational Molding Machine
h rotational molding machine has a place where the mold is heatcd and cooled while being rotated. The rotacional motion is usuaily in a bi-axial manncr that can be provided by a straight-arm structure or cranked-arm structure of the mold ami (Figure 1-3). Electricicy, hot oil, or gas can provide the heat The elecmciry opuon is clean but iess efficient The hot-oil opaon is just the opposite, which is efficient but messy and estra maintenance is nccdcd for the equipmenc. The gas option is the most popular method now [Il. Chapter 1 Introduction
1.2 Polyethylene and Polypropylene
1.2.1 Polyethylenc
PE is a kind of thermoplasuc that has a long chah (CH2-CHI) aliphatic hydrocarbon structure. Ir is wax-like and becomes sofrened at about 80 - 130 "C. Polyethylene is a tough material with a moderate tensile strength. There are many ways of characterizing PE, but thc main categorizing feature is the density of the polyethylene [9]. Thc four categones of PE, specified in accordance with ASTM standard, are presented in Table 1-1.
The ciifference between LDPE and LLDPE originates from the ciifferences in their production process. LDPE is produced under a high pressure and possess a very branched poI!rmer structure [9]. The LLDPE is formcd in a low-pressure process that yields a linear structure with short side branchcs bÿ controlling the reactor condiaons. HDPE is aiso developed by low-pressure processes. However, the branching suucmre in HDPE is nchicvcd by adding varying amounts of cornonomers such as butane and hesane during polymerization [l O].
1.2.2 Polypropylene
PP has manp advanmges, compared with polyethylene. Thc wo most important are: the higher melting temperature (usually about 150-175 "C 1121) and the higher tcnsile modulus, while the density is kept low. However, the characteristics of PP with differcnt structures differ from each other dramaticaiiy.
1.2.3 Homopolyrner and Copolymer
PP homopolymer is the result of poiymerizauon of propylene molecules CH,=CHCH, with cataiysts under certain procrssing condiaons. The first appearance of Pol.propylene Chaptet 1 Introduction
chat is vduablc for an industry purpose was inuoduced in 1955 by Natta et al. from organo-
metallic catalyst based on ritanium and aluminurn [S-7. A PP homopolymer consiscs of only propylene segments: (CH2-CHCHJ. Three possible suucrures esist: hcad-to-tail, head-ro- head, tail-to-tail (Figure 1-4) [6]. The steric effect of the methyl group highly favors the
hcad-to-rail suucnire because of the high chernical regularity of the PP chains caused by the structure. \men a CO-monomer is added to the chaïn structure, a copoher is obtained.
Usudy, ethylene is used as CO-monomer, The introduction of ethflene modifies the chcmicd and physical properûes of the copolymer, such as the thermd stability, saffness, and strength. This is due to the introduction of the CO-monomerchat reduccs the regularit). of the chain structure and therefore reduces the material's ability to crystallize. With the dccrease of the crystallizauon abiiity, the PP material becomes less sciff and the brïttlcness becomes reduced. The effect becomes more dramatic with increasing the conterit of cchvlene. \%en the ethylenc amount is higher than 10 percent of moIecular weight, thc matcrial is cded Ethylene-propylene Rubber @PR) 161.
1.2.4 Linear and Branched Polypropylene
As discussed eariier for PE, branching structure cm aiso be introduced into linear PP backbone chains [8,9]. \%en a long chain branch is attachcd to the main chain, the matcrial will bc referred to as High Melt Suength Polypropylene (HMSPP) [IO]. By the branching structure, the HMSPP shows very high melt strength chat is highly favored in processes such as Çoarning, blow molding, and estrusion coaàng. Chapter 1Introduction
1.3 Plastic Foams
1.3.1 Foams
Plastic foams, also known as expanded or sponge plastics, consist of at least nvo phases: a soiid polymer macris and a gaseous phase denved from a blowing agent [IO].
Plastic foams cm be found in a variety of densiaes ranging from 1.6 h/mfl to over 96
[kg/mi. The type of foaming method and the combination of blowing agent and plastic material used for ia production determine the densiq of the foam. The density of the foarn rnainly determines its application. With reference to the density of the foamed structure, foams arc classified into three categories: flexible, rigid, and scmi-rigid. Rigid foams are usually used for load-bearing purposes while low density or flesibie foams are normdy used for packing or thermal insulation.
In foaming processes, a physical or chernical blowing (foarning) agent can be used for nucleauon of the gas bubbles. A Chemicai Blowing Agent (CBA) refers to rhose materials thar decompose under hear and generate either N, or CO, or both of thern. A physical
Blow-ing Agent cm be a gas such as N2 or COb or a liquid Like fluorocarbon or isopentane that has a ver). low boiiing point. The blowing agent dispcrsed in the polymer can be brought out for foaming by reducing the pressure or hcaang up the mixture [11,16]. This is caiicd a nucleation process. Sometimes, a nucleating agent such as talc will be added to the mixture to increase the number of bubble nuclei.
1.3.2 Principal Foam Production Technologies
Among ail the foam production technologies that have been developed by now, the followving are recognized as principal [12]:
00 Reaction Injection 1Molding (Rlbf), usudy by impingement mixing Chapter 1 Introduction
Estrusion foarning bu using espandable beads, pellets or physical blowing agent
P3, 143
Injection molding of cxpandable beads or pellets
Spraying of foarns
Rotational foam molding
Larninaaon of foams, or foarn board producâoii
Compression molding of foams
1.4 Rotational Foam Molding (151
In the rotational foarn molding process, when chc mixture of' polymer powders and chemical blowing agent is charged into the mold and roto-moided, a rotational foam producr is obrained. When obseriing a finished roto-foamed part, a soiid (unfonmed) skin layer and a foamed core strucrure can usuaily be disanputshed. This dual layer structure is achiered by a two-shot process, in which the skin material is chargcd and rnolded first. After the skin is obtained, the polymer and chemical bloWng agent composition is added into the mold either through a vent or by using a special device cded "Drop Bos" (Figure 1-51. The dual layer structure can also be achicved in one-shot process by choosing a proper combination of marerials. This is also one of the research objectives on which this thesis is based.
1.5 Purpose Statement
The objective of the thesis work is to develop a new rotomolding technolog). for manufacniring of fine-cell PE and PP foams with an unfoamed siÿn Iapr. To be able dcvelop a process mode1 to conuol rhe ceU morphology of the PE and PP foams in Chapter 1 introduction rotomolding, it is essential to idenhg the effects of the material propemes and processing conditions on foarn structures in rotomolding. Since the rotational foarning process is under a low pressure, the ceil nucleation and growth processes are different from those in extrusion and injection foam molding. Thercfore, a clear understanding of the foarning mechanism in the process is critical to promote the uniformity of the ceU distribution and to rcducc the celi size.
Above aii, efforts shouid be made to conaol the nucleation process. It is desirable thar a largc number of micro-bubblcs work as the nuclei. Their distribution and population densi. will determine the final foam ceii size and structure. To achieve this goal, promotion of good rnixing techniques, propcr processing conditions, and proper selecaon of marerials arc the essential elements t~ be considered. Prevention of cd deterioration is another crucial factor chat has to be considered in order to achieve fine celi structure. Celi coalescence and ceIl coarsening are major reasons chat reduce the ceii population density dunng the ceii growrh stage. It is also critical to prevent the gas Ioss during the roto foam molding process. Besides the economic factors, the gas loss may completely destroy che gencral foam structure.
To be able to control ali these factors, it is important: (1) to promote a good mising of
PE or PP powders and the chernical blowing agent (CBA); (2) to induce a good werting of rhc polpmer melt ont0 rhe surface of the CBA parricles; and (3) to select proper material and processing conditions to efkctively control ceU deterioration.
Finally, it is the ultimate objecave of the project to develop a new technique for ohtaining a skin-core dual laycr structure, Le., an unfoamed skn and a foarned core, within one-shot rotomoiding by controUing the properties of the skin and core materials. A f~shed Chapter 1 Introduction
part, like the esample shown in Figure14 (b), udi have much better mechanical properties
compaxcd to parts molded in a traditional process.
1.6 Overall Strategy
The research was conducted in wo directions: maceriai control and processing
condition optirnization. Guideiines for choosing proper combinations of CBA and polymer
materials were established. In addition, the mechanism of the process was investigaced and
the effect of processing condition variation was studied. A schematic illusnation is
prcsentcd in Figure 1-7.
1.7 Outline of the Thesis
Since the objective of the thesis is to develop ncw techniques of producing foam in
rotational molding by esperimentation, the foarning rncchanism was investigatcd by
analyzing the obtained results obtained from a series of esperiments. In other words, the
thcsis is based on the esperimental results and the related analysis.
The research was focused on the anaiysis and design of nvo distinct technologies that
wcrc applicd on two differcnt types of polymers. Alore specificaiiy, both dry blending and
melt compounding based technologies were applied for esperimental production of PE and
PP foams, respectively. Comparative analysis of the esperimental results and conclusions are prcsented in thc thesis.
Chaptcr 2 provides the theoretical background of related subjects including the polymer rheological properties, the CBA decomposicion behaviors, a sintering theor). in rotationai molding, and the concepts of rnixing technology. Chapter 1 Introduction
In Chapcer 3, the csperimentd resdts obtained from the andysis of polymer materiais
and CBAs are presented. The proper selection of blowing agents and matri.; poiper
materials is based on these analyses.
Chapter 4 describes the espenmental results obtained while using the dry blending
tcchnolog).. The identified effects of material properties, processing parameters, selection of chcrnical blowing agents, and additives on PE and PI? rotaaonal foam molding prescntcd in thiç chapter were experimentaily verified. Aso, the foaming mechanism for this particuiar process is studied in detd.
Chapter 5 shows the work on the tlsperiments conducted on the mclt cornpounding based technoIogy. In the chapter, the compounding process as wdl as thc roto foaming is dcscnbcd and analyzed. Since the use of a melt compounding based technique changes thc nucleauon mechanisrn, the foaming process is different form that in the dry blending based technology. The foarning mechanisms in the melt cornpounding based technique are thoroiighly studied and discussed in this chapter.
Chapter G concenuares on the research on the skin formation. Bascd on the melt compounding process, preliminary investigation has been conducted to ouclinc the critical conditions and parameters for controUing the skin formation.
In Chapter 7, a summary and conclusions of the research arc made. Some future w-ork recommendations are also suggestcd based on this research. Chapter 2 Theoretid Backeround
Chapter 2
Theoretical Background
2.1 Polymer Properties
2.1.1 Viscosity and Elasticity
When a liquid-iike polymer melt is subjected to a stress, it will deform continuously
as urne goes on. When the stress is released, the iiquid stays in its deforrned state [22,23].
t'nlike that of soiid materiais, the stress and suain relaaonship of iiquids is ame-dependent.
In the case of extension, the strain rate can be expresseci as:
and the estensional viscosiry is
\Wcn a shear force is applied to liquids, the property of the matcrial is indicaced by a shcar
. whcre y is the shear strain rate.
For a Newtonian fluid, the viscosity is constant at a specific tcmperature. However, for a non-Ncwtonian polymer melt, the shear rate is no longer proportional to the shear stress.
The viscosity becomes a hnction of shear rate and temperature. For examplc, for PP and
PE, the viscosity deciines when the shear rate is increased. Therefore, the zero-shear Chapter 2 Theoretical Background
viscosity, defined as the viscosity at the zero shear rate, may be used to indicate the property
of die material.
A clear understanding of the polymer viscosity properties is critical in polymer
processing. In rotational molding, the zero-shear viscosity of the materiai piays an important
role in particle sintering and bubble removal, corner filiing, and some finai mechanical
properucs [Il.
2.1.2 Melt Flow Index
Besides the shear or estension viscosity, the Melt Flow Indes (i\IF?J or the AIelt
Flow Rate (LMFR) is another parameter that indicates the viscosity of polymer melt at a
certain temperature [II. They are defined as the amount of polyrner (in gram) cstruded from plastometer for 10 minutes. h standard measurement can be done with the equipmcnt shown in Figure 2-1, according to the ASTM standard. The material to be tested is put in a heatcd barre1 (190 OC for MF1 and 230 "C for blFR) and a dead weight is applied on the polyer melt through a piston. The estrudate is coilecrcd and weighed.
UsuaiIv, the higher the average molecular weight, the !ower the MF1 or MFR.
Therefore, MF1 or iMFR can be used effectively in choosing materials with a designared viscosity.
2.1.3 Meît Strength
The melt strcngth is an indicator of how much extensional force a plastic melt can stand [23]. In plastic foarning, this parameter reflects the strength of the foam celi walls, which indicates the ability to resist wd collapse [24]. The higher the melt strength, the Chapter 2 Theoreticai Background casier to maintain the ceii population density and a smaii ceii six as d be analyzed in detail in Chapters 4 and 5.
2.2 ChemicalBlowingAgent
A chernical blowing agent (CBA) produces gases through a chernicd rcacuon, which may be a thermal decomposition or a reacaon benveen two or more components [121.
2.2.1 Classification of CBAs
As the most inexpensive and most traditional blowing agent, the CBAs can be classified as inorganic and organic according to the structure of the CBA molecules.
Inor-ganic CBAs
Two major types of inorganic CBAs fdinto this category: ammonium carbonate and carbonates of alkali rnetals 1131. Gcneraiiy, ammonium carbonate is n mixture of
(NHJ2C03H20,NH,HCO,, and ammonium carbonate, NH,COON,. With the presencc of
\vater or an increase in the temperature up to 30 - 40 "C, ammonium carbonate begins CO dccompose at about GO "C and liberates NH, and CO2.
Among carbonates of aikali metals, sodium carbonates and bicarbonates are thc most popular ones [17. After being heated up to the range of 145-150 "C, sodium bicarbonates
(NaHCOJ wiii be completely decomposed within 30 minutes. The decomposition follows the following scheme: Chapter 2 Theoreticai Background
The foaming activity of the sodium bicarbonates is relativcly slow, which can be an advantage when the smoothness of the decomposition is a concern. With sodium bicarbonates, the pressure of generated gas increases slowly and the process is more stable
[131-
The disadvantage of inorganic blowing agents lies in their tendency m disperse in polymer mauis poorly and usually to dccompose at temperatures far below the melting tcmpcraturc of some polymers wvich a high meiung temperature. This rcstricts the application of this type of CBA to a very narrow range.
Ormnic Blowing- Apnts
Compared with inorganic blowing agents, orpic blowing agents have man. advantages. They usually start to decompose or reach their maximum gas yield at rhe temperature that is close to the melting temperature of the polymer [13]. In addition, the
CBAs can be well mixed with the polymer matris. However, the disadvantage of this typc of blowing agent is its relatively high cost and the non-gascous products of the decomposition.
There are many classes of orgnnic blowving agents, including azo and diazo compounds, N-Niuoso compounds, sulfonylhydrazidcs, azides, trianzincs, triazoles and tetrazoles, suifonyl semicarbazides, urea derivaaves, and esters [lq. Among al1 types of organic blowing agents, azodicarboxamide is one of the most effective high-temperature l~lowingagents. When decomposing at 190-240 OC, azodicarbosamide generates COand
N2,the gas sources of the foam. The blowing agent cm be well dispersed in many solvents and can be easily mixed with different polymers. One of the disadvantages of azodicarbosamide is the acid decomposition residuc that may cause darnage to the processing equipment such as mold or exuuders. Cbapter 2 Theoretical Background
2.2.2 Endothermic and Exothennic Behaviors
During the decomposiuon process of CBAs, the chernical reaction cm be either absorbing heat (endo thermic) or releasing heat (esothermic) (Figure 2-2). For foaming processes in rotational molding, which is normally under low pressure, the endothermic or csothermic reaction involved in the decomposition process is vital to the foarn quality according to the research result Therefore, a dear ucderstanding of the endothermic or csothermic effect of the particular blowing agent used is dso very important.
2.2.3 Decomposition Rate
The decomposiaon rate is the speed at which the CBAs generate gases. Ir can be indicarcd by the speed of weight loss, the temperature range of the weight loss, or the speed of the pressure increment (Figure 2-3). hccording to Our research, the decomposition rare is nnother cricical factor in low-pressure foaming processes and it wiii be discussed in detail in the following chapters. Howevcr, it should be menuoned char too high or too low a dccomposition rate is harrnful to the final foam structures.
2.2.4 Selection of CBAs
In Handbook of Pohmenk Foamr and Fonm Techolo~[13], Daniel Klempner and Kurt C.
Frisch mentioned nine principal requirements, while makng dear that no indusuially available CBAs can meet with all the requirements. As summarized in the book, the main criteria by which to judge CBAs are:
mGas numlier or gas yield, which is the volume of gas (in cc) yielded by one gram of
blowing agent per unit of time at the temperature of maximum gas liberation,
*.Tnc onset point of the decomposition of the CBA, Chapter 2 Theoretical Background
+The temperature of the maximum rate of decomposition and the pressure developed
by generated gases,
+The rate and kinetics of the decomposition.
2.3 Mixing Technology
In polymer processing, in addition to the designated form, homogeneity in composition and propemes is aIso necessary, especiaiiy when additives are involved [18, 19,
301. Chris Rauwendaal defined mi-xing as "a process to reduce the nonuniforrnity of a coinposition by inducing physical motion of the ingredients, resulâng in a reduction of the concentration gradients or temperame gradients" [29].
According to the nature and form of the materials in the rnising process, rnising cm be placed into nvo caregories: me/t compoltnning and d~ blending. In some processes, either during the shaping processes or prior to the processes, materials, especially the pdymer material, are in shear conditions in a melt or softening state. This type of process is usudy calied mek compounding or Intennve miking [18,19]. In other conditions, when the mkxïng is adequate without melting or softening the material, it is usudy cded simple mising, or &y llrildzhg in which "components or ingredicnts are physically intermingled without significanc change of the physicai state of the components" [18]. In dry blending, the matcrial musc bc in suitable forms such as powders.
2.3.1 Dry Blenàing Equipment
Therc are many types of dry mising devices that can be used depending on the difficulry of the mising and the requirement for the rnising qualit).. Among them, vibratory or reciprocaang blenders, tumbling blenders, sur mixers, intensive non-flusing misers, ribbon bIenders and related mixers, Z-blade and related double-arm mixers, lough misers, Cbapter 2 Theoretid Background
roii rnills etc. are very cornrnonly used in the indus.. In the project of roto-foaming,
turnble blendcrs are used for the dry blending based method.
In tumbIe blenders, die mising is achieved through the movement of the composition
paracles by pouring, roliing, or falling. The main mising action comes from the gravit).
force. It can be sirnply achieved by partidy fïkng a mmbling drum with che materials and
rotaang it. There are severai ways to improve cross mixing qualiaes includîng different
ways of rnoun~gthe drum and differenc choices in the shape and structure of che drum.
Figures 2-4 and 2-5) [!a]. In Figure 2-4 (a), the cylinder drum is mounted in such a way that
the rotation is eccentric. By such a means, a reciprocating alting motion across the direction
of rotation introduces an element of cross mtuing. In the machines illustrated in Figure 2-4
(a), the rnising in the longitude direction is achieveci by using a V-shape chamber. The cross
mising can aiso be improved by changing the shape of the miuing charnber, as illustrated in
Figure 2-5. Sometimes, deflector plates arc rnounted in the double-cone blenders to further
improve the mixing quality (Figure 2-6) [la].
2.3.2 Melt Compounding Equipment
Mclt compounding can be done in either a batch processing manner or a continuous manncr. A continuous compounding process is more econornical for large-scde proccsses in the plastic indusay and is also more desirable when a high degree of product uniformiy is critical.
Two-Roll MiII
The two-roll rniil is a typical piece of batch compounding equipment that was originally used by the rubber indusq. The structure of the mill is illustrated in Figure 2-7 Chapter 2 Theoretical Background
[19]. In the machine, two rolls made of cîst iron or ailoy steel are mounted on a massive
end frame. men the mivrure of the blended materials passes the adjustablc gap benveen
the nvo roiis, the mkkg is achieved. Usuaiiy, the roiis are equipped with heating and
cooling units to control the temperature. For exarnple, blowing agents and polymers can be
mised in this way. ïhe polymer di be melted by the heat Çrom the heater as weii as by the
friction gcnerated benveen the rollers and îhe materiais.
Apart fiom the low capacity of this type of mising, the merhod has thc disadvantage of
raising safer). concems. An unguarded structure is very dangerous because it mai catch a
operator's hair, dothes, or even hgers. For smd machines, it is easy to cover the working
section, bur special efforts should be made to avoid the problem when a large machine is
used.
Sin+ Screw Esuuder
A single screw extruder is the type comrnonly used in the industry 1211. The structure
of a typical extruder is shown in Figure 2-8 [22]. Solid materials are added rhrough a hopper
and fcd into the barrel by gravity. hfter entering into the barrel, the material dlpass
through three sections: the feeding, compressing and metering sections. In the feeding
section, the marerial is kept in a soiid form and transponed; in the compressing secaon, the
plastic begins to melt when the channel of the screw becomes shaiiow and the materiai is compressed and sheared; in the metering section, the channel becomes shailower to keep
the output stable. The detailed picture of thc structure of a typical esuusion screw is shown in Figure 2-9 [13]. Chapter 2 Theoretical Background
The mass flow in the extruder channel cm be grouped into four categories. The flow generated directly by the shear or the relative movement of the scrcw surface and barre1 surface is the Drag Fioui, which cm be expressed by the foiiowing mode1 113,301:
Where D is the diameter of the screw, N is the rotational speed in revoluuon pcr second, and h is the chmel depth.
The pressure flow is a flow caused by the backpressure at rhc end of the screw, whose effect is to push the matenal backward to the feeding zone. The mass flow caused by the pressure can Le espressed by the following equation 1131:
In Figure 2-9, the flow chat is perpendicdar to the flow chamel of the screw causes another
The pressure difference and transverse flow generated by shear force cause another flow which goes backvard, the leahgej7utv. The amount of leakage flow due to drag is:
And usually espressed as a fraction of the drag flow [13]: Chapter 2 Theoretical Background
In addition, a certain amount of leakage is also causcd by pressure ciifference, including a longitude pressure gradient and a transverse pressure gradient. The leakage flow can be espressed by the foiiowing equaaon:
The total output fl ow can be expressed approsimately as
QTc,pl = Qu- QI%- QI. (2-9)
Hou-ever, since Q, is too smaU for normal extrusion screw dcsigns, it can bc ncglected in most cases. Therefore, the ourput dlbc
In single screw estruders, the rnising process starrs frorn the material-conveyîng zone, but as has been shown abovc, because the flight clearance is so tight that thcre is veqv little lcakage flow, the materiai is very unlikely to go bachvard. Consequently, the rnising qualin- is very poor. To improve the mising quality, in addition to the common sections, a rnising section cm bc added in the front of the screw. For example, a mising screw wïrh a blockhead mising section as shown in Figure 2-10 [20] is used in the present cxperimcnts.
Thc shape of the rnising head cm be changed according to requircments.
in a mising estruder, the heat that melts the plastic cornes from the band heaters and viscous beating. To baiance the temperature, some heat is carried away by the moving matcrial, bg conduction, by radiation, and sometimes by the cooling of the screw. A proper choice and combination of heaang and cooling conditions is criticai to scabilize the remperature, especiaily when heat sensitive materials such as CBAs are processed. This will bc discussed in detail in Chapter 5. Chapter 2 Theoretical Background
Twin Screw Extruder
Typicaiiy, a nvin screw extruder is for compounduig. Through the CO-rotaung or counter-rotating moaon of the screw, thc shear suaining is increased intensivcly. For this rcason, the mising quality is supex-ior to rhat of single screw exuuders. Howcver, the high shear force makes it difficuir to maintain the plastic melr at the designated tempcranire even rhough the temperature gradient is reduced dramaticaiiy with the polymer melt. This may present a problem for the compounding of temperanxre sensitive materials tike CBh. Chapter 3 Aadysis of Materials
Chapter 3
Analysis of Materials
3.1 Material Requirements of PE and PP in Rotational Foam Molding
As stated in Chapter 2, material properties are essenuai in polymer processing. It wiii
deresmine the processing parameten, product qualities, and applicaaons. For rotational
rnolding, the material properties become even more crucial as uiii be show-n in Chaptcn 4
and 5. In this context, this chapter descnbes in detail the propemes of the mareriais used in
the rotational foam molding esperiments.
3.1.1 Viscosity and Elas ticity
The viscosity of plastic materials determines the abiiity of polymer particles to merge
with each other in rotomolding and the moIten pol~mcr'sability to flow in the mold under
certain temperatures. A low viscosity material is easier to sinter at a relativdy low
ternperaturc, allowing the processing temperature to be lowered and the processing urne to
bc reduced. Ir is also easier for a low- viscosiry material to flow into thc corners or tips of
some parts with complicated forrns and structures.
In the rotational foam molding process, the sintering ability of plastic powders becomes critically important, especially when a dry blending based technology is used. The gas loss, nucleaaon, and the finiil ceii property of the foam di be governed by the sintering qudity. This aspect will be discussed in detail in Chapter 4.
Another essential material parameter in rotomolding foarn processing is the meit elasticity that represents the melt suengrh. \men a materiai with a high melr strength is used, the wall of the gas bubble is relaavely difficult to break during the process of ceIl Chapter 3 Mysisof Materiais espansion. This will reduce the ceii coalescence and increase the ceii population densil of
the foams. T'herefore, high melt strength material should be used and attcnuon should be paid to control of processing conditions in order to maintaining a high melt strcngth of material.
3.1.2 Plastic Powders
Since PE and PP materi& arc of a poor thlsrrnai conducrivit)., che speed at which hcat is transferred into the center of each partide dspends on the size and volume of the particles
[II. For example, a large particle of polymer material melts more slowly than a small one; thcrefore, smaii particle sizes are preferable when a quick sintering is needed. For a normal rotational molding process, chc sintering speed \vil affect the surface finishing, bubble removing and final mechanical propemes. In roto-foaming processes, in addition to the problems related to sincering, the powder quality is also critical for achieving a high rnising degree which wviii also detennine the final foam ce11 size and disuibution. Among the powder properties, the particle size and partick shape are the most cntical oncs in rotauonal foam molding.
Particle Size
Since fine powders will melt more quickiy than coarse powders, it is dcsirable to have powders in a finer grade. Besides the melting rate, the uapped air pockets among particles pose the most serious problem. When charged into a mold, the materials arc a misture of plastic powders, additives, and air. The air will form air bubbles in the polymer melt later. It seems reasonable co espect that the bigger the particles, thc bigger thc air bubbles will be and the more difficult it wiil be to remave them. The exisang bubblcs will affect the Chapter 3 Analysis of Materiais mechanical property and surface quality of the molded part in normal rotaaonal molding processes. On the other hand, with the same amount of powders used, it is not dificuit to understand that the bigger the powder sizes, the fewer the air bubbles. Because the air bubbles will act as ceii nudei in die roto foaming process, their dismbution and population densir). determine the foarn ceii dismbution charactenstics. A fetv large bubbles WUlead to huge cells in the final foamed products. This aspect wiil be discussed in detail in Chaprer 5.
Particle Shape
The shapes of the plastic particles after being ground are usually irregular. Typicd shapes of plasac powders are shown in Fiçure 3-1. There are different opinions on how the particle shapc affects the sintering in rotational molding. Zn his Rota&brza/ ~\lofdnof Pla~tic~,
R. J. Crawford claimed that too regular shapes such as uniforrn sphericai particles would be too easy CO roll around before the melting and sintering started. As a result, dl the problems rclated to poor sintering either in normal roto-molding or rotational foam rnolding wodd occur. In addition, easy roliing \vil1 also cause an uneven distribution of materiais in the mold [Il. However, we believc that the tds or hairs of powdcr pamdes with irregular shapes increase the inter-particle distance and the free volume of powdcrs. Consequendy, the sintering process wiii become slower.
However, when impropcr grinding conditions are applied, the plastic particles are tom and shredded since plastics iike PE and PP are tough materials. When this problem occurs, the pou-der particles will have fibriliation structures as shown in Figure 3-1 in which a tail is atrached to the powder pamcle. Poor sintcring dlresult in because the fibrillation structurc obstructs the contact of neighboring particles. Long fibdation structures will also cause a bridging effect. It is not easy for hairy particles to flow freely around in thc mold. If there is Chapter 3 Analysis of Materiais a narrow tip or corner on the mold, the inteded hairs may block the passage and no powder partides cm flow into ic. Finaliy, there will be a void part existing in the molded pmor even a dcfect on the surface. Whcn foamïng mkes place in the rnold, the air pressure cm push the material into these types of corners.
3.2 Materials Used
PE and PP materials are used in this study. For experirnents on PE, a rotzuonal molding grade chosen as the base For research various matcriais with different grades of viscosity, different molecular structures, and different melt strengths were tested. Detailed information of the materials used is iisted in Tables 3-1 and
3.3 Viscosity Analysis of PE and PP
In currcnc expenments, the melt flow index (LIFI) or melting flou. rate PLFR), and rhe zcro-shear viscositv are used to indicate thc viscosity properties of PE and PP. Ml3 or hIFR cm serve only as an indicaror of the viscosity of the material at the standard testing temperature, which is 190 "C for PE (iLIFI) and 230 "C for PP @[FR). The viscosity decrcases as the processing temperature increases. The \iscosity of PE and Pi? matcriais also depends on ~!eshear rate and the materials used in the expenments are shear-thinning materials of which the viscosity decreases when the shear rate is increased. Since rotational molding is a lour shear process, rhe zero-shear viscosity wouid be the most important rheological property, and thcrefore, it was used to evaluate the material viscosity. Chapter 3 Analvsis of Materials
3.3.1 Experirnentd Setup
A cone-plate type dynamic stress rheometer from Rheometrics Scienàfic (5-200) is used in the esperirnent. The picture of the rheometer is shown in Figure 3-2.
Before the test, standard samples are made using the sample maker of the rheometer. ln order to remove the air pockets that become trapped among the plastic powder particles, the plastic materials are melted in a 3/4" inch Brabender single screw escruder (Brabender
05-25-000). The polymer melt from the cstruder is charged into the standard sample maker and prcssed into the designatcd shape. Then, the sample is loaded in the gap benveen the conc and plate of the testing head. When the sarnple is rnolten and reaches the tesung tcmpcrature, the gap is Çorccd to be a standard testing gap (0.051 mm in the esperiments) and the esperiment is started. To prevent the plastic material from being degraded and oxidized, nitrogen is used as a protecting gas. To test high viscosity matcrials, the iniud tcmpcraturc of the heater should be a lide higher than the testing temperature in order to facilirace the melting process and improve the melt quaiiry. On obcaining the polymer melt, the ïemperature can be iowered to the testing temperature before conducting the esperiment.
In this study, at a certain temperamre, a shear rate sweep starting from 10.' to 1 \vas appiied on the samples and the zero-shear viscosity was determined. In Figure 3-3 (a) the measured zero-shear viscosicy of different macerial at 190 "C was plocted as a function of hlFR that are providcd by suppliers. An esample of the effect of processing temperature on the \iscosiy was shown in Figure 3-3 (b), for SC873 @FR 50). Chapter 3 Anaiysis of Materiais
3.4 DSC Analysis of PE and PP
To obtain the processing temperature window of plastic, it is very important to know the melung point and reIated thermai reaction during the heating process. A Differential
Scanning Calonmeter (DSC) is onc of the most widely used equipment for rhis purpose in the thermal andysis.
3.4.1 Experirnental Setup
DSC works by measunng the heat flow changes into or out of materinls as a funccion of tcmperature. There are nvo most commonly used methods in DSC anal!.sis: the power compensation method and the heat Bumethod. For the TA 2910 DSC ceii (Figure3-4) that is used in this study, the heat fiow flus method is applied. The principle by which the method works is illustrated in Figure 3-5. In the enclosed cd, a reference pan and a pan with sample encapsulated is put on a constantan thermoelectric &SC. Prcheated purging ps is injected into the celi so that a stable and uniform heat environment is created. Once the pans are heated up, the difference between the temperattue of thc reference pan and the temperature of the sample pan is analyzcd. From the tcmperature difference, the hcat flow into or out of the material is cdcuiated according to the following cquation:
where dQ/dt is the heat flow, AT is the temperame difference between the two pans, and
R,, is the thermal resistance of the constantan disc.
During DSC esperiments, plastic materiais averagirig 3-5 mg are encapsulated in the aluminum pan, and a inoderated heating rate such as 10 "C per minute is used to achicve rcsults with proper accuracy and resolution. Before the tests, special attention was given to Chapter 3 AnaJysis of Materials the temperature calibracion. A standard indium specimen with a known melang point was run on the DSC under different condiaons and caiibration curves were obtained. Through the temperature calibration, the temperature enors caused by a thermal lag were minimized.
3.4.2 Results and Discussion
A series of DSC esperiments have bcen conducted on both PE and PP materials.
Results are shown in Figures 3-6 and 3-7, respectively. The resdts were analyzed with the data proccssing software, Universai Anaiysis, which is provided by the TA Instrument. The onsct temperatures were studied to idenafy the stanng point of the melting process. From test results, it was found that LLDPE 8556 and LLDPE 8361 started to melt at about 119°C.
For most of PP materids, the mrlting bcgan at 152-155°C and reached its peak at about
1G5"C. The only escepaon is PF633, the one with the highest melt strength and melt viscosicy. It started to melt at 149 "C and was Mymelted at 160 "C. This would be helpful to rcduce the processing temperature in rotomolding.
3.5Thermal Analysis of CBAs
In order to maintain the highest melt strength, the systcm temperature should bt: maintained at the minimum level that the process allows. To achieve this goai, the decomposition of CBAs shouid occur as early as possible afier a uniform polymer melt has been obtained. Therefore, a proper processing tempcrature can only be set after a clear understanding of the thermai characteristics of the CBAs. Chapter 3 Analysis of Materiais
3.5.1 Experimental Setup
Two types of esperiments were conducted to determinc the thermal activiaes of candidate CBAs. DSC was used to study the endothermic and esothermic bchaviors of the materials, the heat generated by the decomposition, and the decomposition temperature range.
In addiaon to DSC, a Themalgravimetrïc Analyzer (TGA) rA2050, Figure 3-8) was uscd to test the decomposiuon behavior. The TGA moniton the decomposition behavior by measuring the u-eight of the sample chat has not been decomposed. From rhe TGA rcsults, the decomposition eficiency cm be clearly identified, as cm the decomposition temperature and the decomposition rate.
In the esperirnents, about 10 mg of sample marerials was put on a piatinum sample pan and loaded ont0 a very accurate baiance. Then, rhe sarnple was encloscd in an oïcn.
Thc sample was heated up to the designated temperature Mth a certain hcating rate. A stream of Helium gas (40 cc/min) \vas injected into the oven so that thc gencrated esfrom decomposiaon could be purged. Another suearn of Helium gas (60 cc/min) runs through rhe baiance charnber to prevent the influence of gas and heat corning up from the ooen.
Nitrogen or argon can also bc uscd as purging and protection es. But the relauvely low thcrrnal conducavity of nitrogen or argon may increase the measurement error caused by tlicrmai Iag. nierefore, the Hciium that hns a high conducavity is chosen cspecially for high heaang rate esperiments. An additionai water-cooling system surrounding the balance chamber uras used to maintain the balance charnber temperature. The weight of the sarnple was recorded as a function of temperature or tirne. A nurnber of CBAs were tested (see
Tabic 3-3). Chapter 3 Analysis of Materiais
3.5.2 Results and Discussion
The rcsults of DSC analysis are shown in Figures 3-9 and 3-10. Figure 3-9 shows the cornparison of different CBAs, including their decomposition onset temperature, pcak
ccrnperature, and endocherrnic or esothcrmic propemcs. Together ~4thDSC informauon on polymcr materials, this prosides the basis for the initial selecuon of CBAs. In Figures 3-
10, the kinetic properties of selected CBAs as a function of the heating rate are Uustrared.
Thc information developed from the results can be used as guidelines in the control of processing conditions.
The results obtained from the TGA analysis of CBAs are shown in Figures 3-1 1 to 3-
15. From thesc results, the weight loss properaes and decomposition temperatures of diffcrent CBAs were compared, the effect of the heaang rate on the weight loss behavior was investigated, and the effect of catalysts on the decomposition temperature of selected
CBAs was srudied. Since the CBAs are normaily of a low thermal conductivity, the particlc six of the CBA and the size of the testing sarnple may alI have an effect on the dccomposiuon behavior. Espenments werc also conducted to iIlustrate the effect of these nvo factors.
Com~arisonof Different CBAs
Celogen A23970 is a type of an azodicarbosamide blowing agcnt providcd by Uniroyal
Chernical. As shown in Figures 3-9 (a) and 3-11 (a), the decomposition temperature of
Celogen A23990 is 210.4 "C,which is about 50 "C highcr than the meIting temperature of ali the PP materials. Therefore, Celogen AZ3990 may be used for foarning of Pl?, Detailed csperiments were conducted to understand the decomposition behavior of the blowing agent under different processing conditions. Cbapter 3 Ansilysis of Materials
ADC zentv is another type of azodicarboxamidc CBA provided by Bayers. According
to the information Çrom the DSC results in Figures 3-9 @)-(c), and 3-ll(b)-(c), ADC senes
have propemes very similar co those of Celogen A23990 with an onset temperature of 204
"C. They, therefore, can be another choice for PP foams. In addition, there are nvo grades
of particle size available for this type of blowïng agent and this series of CBAs could be used
to illusuate the effect of CBA particle size on the decomposition behaviors of CBAs. Two
ADC sarnples were tes ted: ADC/ hi-Cl and ADC/ F-CS.
It should be noted that the DSC diagrams in Figure 3-9 (a), (b), and (c) are sharp and
righc skewed, especially (b) and (c). This phenornena is caused by the esothcrmic nature of
the azodicarboxarnide blowing agent. The heat generated in decomposition increased the
local temperature around the sarnple pan to a ievel that was higher than the oven
tcmperature. The decomposition was enhmced by the high Iocal temperature and finished
almost instantmeousIy. This was also indicated by the sharp weiglit loss in TGA diagram in
Figure 3-11 (a). However, the temperature of the sampie pan dropped back to the oven
tcmperaturc after the decornposition. This esplains why the temperature fell down after the peak in the DSC diagrams. This self-heating phenomenon and its impact on roto foarning will be discussed in detd in Chapters 4, 5 and 6. Sirnilar decomposition beha~lorcan also bc idcnufied when another azodicarbosamidc blowing agent Celogen OT was tested.
Celogen OT (benzenenesulfonyl hydrazide) is another esothcrmic blowing agent that gcnerates N, and H,O. In Figure 3-9 (d), it can be seen that Celogen OT dccomposed ver). quickly wich an onsec temperature of 159 "C. The blowing agent is suitable for PE foams but cannot be used for PP foams because of its low decomposition temperature.
Sodirrnr bicarbonate is an endothermic blowing agent that stms to decompose at a rclauvely low temperature. As the TGA results of Figure 3-1 1@ indicatcs that the onsec Chapter 3 Anaiysis of Materials dccomposition temperature is about 124 "C,and that the peak gas yield occurs at 172 "C.
Obviously, the decomposition is too early for PP materials and most of the gas will be lost before the melting of polymer stms. However, it is just above the melting temperature of
PE and therefore cm be used for PE foam processing. It is dso identified that the blouing agent decomposes at a uide temperature range of 48"C, which indicates that the dccomposition rate of sodium bicarbonates is very slow.
Effect of Decom~osiaonPromoters
Since it is desirable char the decomposition temperature be kept as low as possible, the
CBA should decompose as soon as the sintering is cornpleted in order to kecp a high polymer meit sucngth. Ir is well known chat cacdysts can be used to lower thc decomposiaon temperame of azodicarboxamide materials. In the esperiments, Zn0 \vas selected to rnodify the decomposiaon behavior of Celogen AZ3390. From Figure 3-13, it can be seen that ~viththe increase of the amount of ZnO, the decomposiaon temperature dccreascs. The percentage of Zn0 to CBA is calculateci in weight ratio and espressed as the amount of Zn0 in 100 parts of Celogen A23990 (200 phr). Whcn 100 phr of Zn0 was added, the decomposition temperature was lowered to 189 "C, which is about 20 "C lower than that of the unrnodified AZ3990.
Effcct of Ueatin~Rate
Thc heating rate of CBA varics greatly under different processing conditions. For examplc, the one in rotomolding is very different from that in the melt compounding proccss. To control the decomposition temperature precisely, the effect of heaang rate on decomposiaon must be stuciied. From the TGA results shown in Figure 3-12, it can be seen Chapter 3 Analysis of Materials that with an increase of heaùng rate, the dccomposition was delayed. This is caused by the thermal lag resuiting from the low thermal conducavity of CBA particles. The higher the heating rate, the more drarnaac the temperature difference between the surface and the center of the CBA particies. \men the center part of the blowing agent particle reaches the decomposition temperature, the surface temperature of the particle, which is the measured tcmperaturc, appears to be higher whcn the iieating rare is higher. However, from the DSC rcsuits shown in Figure 3-10 (a), the esothermïc effect becarne more drarnaac when the heatinç rate increased. This brought a very high local temperature increase. From the figures it was found that the acmal temperame of the area surrounding the decomposed
CBA particles is nomally 10 "C to 20 "C higher when the decomposiaon reached its pcak point. Because of this locai tempcraturc increase, the decomposiaon urilI reach its peak point more quickly. In Figure 3-12, this is indicated by an increased rate of weight loss. This phenornenon becomes more obvious for Celogen OT when a ver). high heaang rate of 200
"C per minute has been used (Figure 3-IO@)).
Effect of CBA Amount
In Figure 3-15, it is provcd that when a different amount of CBA materials is used, the decomposition behaviors are of little difference. However, for esotherrnic blowing agents like Celogen AZ3990, when a larger amount of blowing agent is added, more heat is
çcncrated from decomposition. Because the urne for the heat to escape is limited, with a large amount of heat reIeased from the reacaon, the hcating rate in the local area is higher.
As a result, as shown in Figure 3-15, the sample of 30gram decomposed more quickly thm the one of 10 gram. Effect of CBA Particle Size
In the esperiments on ADC series, the two samples are of the same composition and produced with the same method (LMethod Registration No. 123-77-3). The main difference bcnveen the nvo sarnples is the particle size of the CBAs. r\DC/XI-Cl is sirnilar to Celogen hZ3770 with an average size of 4.5 Fm, whiIe the size of IU)C/F is 15.5 Fm. The cspcriments on ADC sarnples show that the decomposition behavior was not significantly affected by the particle size.
3.6 Selection of CBAs
Based on the thermal behaviors of the plastic materials and CBAs invesugated using
DSC and TGA, a preliminq selecuon of CBAs was made. Accordingly, sodium bicarbonate and Celogen OT were chosen for processing of DE foam because they decompose after the melang of the polymer and rheir decomposition temperature is not veq high so chat a proper polymer melt sumgth can be mYntained. Howcver. cheir dccomposition temperature is below the melung temperature of PP materials and cannot bc used for PP foam processing. Therefore, Celogen A23990 was chosen for PP foaming. The cffects of heating rate, endothermic and esothermic behavior, and other factors on the foarn structure were investigated by changing the processing and material parameters in the rotauonal foam molding experiments, which wiii be discussed in dctail in Chapters 4 and 5. Chapter 4 Rotational Molding of PE and PP Foams witb the Dry Blending Method
Chapter 4
Rotational Molding of PE and PP Foams with the Dry
Blending Method
4.1 Introduction
It is well known that the mechanical properties of rotaaond moldcd parts are weak bccause of their hollow structure. Some practices have been developed for using polprethane foam to fil1 the part. However, because polyurethane is not compatible uith rhe rherrnoplasüc materials commonly used in rotomolding, the interface benveen the skin and the foam is very weak. It would be desirable to have a foam core that is made of the same material as the skin layer. Needham introduced a chemical blowing agent (sodium bicarbonate) into poiyethylene-based materiais in rotaaonai molding [25]. Uniroyal
Chcmical 1261 used Celogen OT as the chernical blowing agent to develop low-density polothylene foams in rotornolding. However, the obtained cell suucture is not saasfactory because of a large ceU size and a low ceU density. In addition, the foarning mechanism in rotomolding has not been clearly understood and very little research on the foarning in roromolding and its applications has been published in the literature.
Since polyethyienc is the most popular marerial in rotanonal molding, iniaal efforts were made in developing high quality low-density polyechylene foams in Our research.
However, due to the outstanding mechanical properties of polypropylene materials, research was conducted on developing a technolog). for making fine-ce1 low-density polypropylcne foams as weii. The high melting point and high mechanicd suength make PP a good candidate for high suength foams that cm be used for structurai materials. In contrast,
36 Chapter 4 Rotatiod Molding of PE and PP Foams with the Dry Blending Mead because of the weak melt strength of PP materials, it is difficult to achieve high quality foam ceus. Due to the technological difficulties, PP foams cypicdy have very large celis with a ver): low ceii-population density, and consequently, ver). few applications cm be found.
4.2 Overall Processing Technology
For the dry- blending based technology, the polymer materiais musr be in a powder
Form. It is easy to get PE powders skce they are the mosc often used rfiaterials in the rotational molding industry. However, PP is normaiiy supplied in a pellet form and must be ground.
The PE or PP powders are rnixed with CBAs and other additives before they are charged into a rotational mold. The mold is put into a hot ovcn and rotated. After the predctermined processing tirne elapse, the moid is being pulied out of the oven and cooled whiie stili rotating.
4.3 Processing Steps in Rotational Foam Molding
4.3.1 Continuous Phase Formation by Sintering of Powdcr Particles
In Figure 4-1, a schematic of the rlpical rnorphology change in rotationai foam processing is presented. The powder misture of polymer particles and CBAs is tumbled inside the mold while the mold rorates in the oven. The polymer powder particles will melt and fuse with each other. Ulamately, the powder particles wïii be My fused and a conunuous phase will be formed. This is refereed to as the sintenng process. Chapter 4 Rotational Molding of PE and PP Foams with the Dry Blending ~Method
4.3.2 Ceil Nucleation in the Dry Blending Method
There are two possible mechanisms of nucleation related to the dry blending method,
\men the sintered polÿmer powders become fully molten, the polymer melt should flow and
wet the surface of the dispersed CBA particles. If the polymer melt can wet the CBA
particles weil before the decomposition of the blowing agent, or the CBA particles are
enarely isolated in a uniform polymer manu, the bubbles generated from the fine CBA
partides wili act as celi nudei for the foaming proccss. In contrast, if the polymer melt does
not wct the CBA pamcles wi-ell before the decornposiaon, chere \vil1 be air pockets uapped
benveen the CBA pkcles and the polymer mamk. These air pockets d work as ce11 nuciei
later and dorninate the ioaming process.
4.3.3 Ceil Growth and Ceii Coalescence
As soon as the decomposiaon of the CBA begins, gas is generated, and conscquendy, cc11 growth occurs. As a result of the conunuaiiy generated gas, the pressure in the nucleated ceiis increases. At the samc tirne, the pressure of the polÿmer matris in rotomolding is maintained alrnost at the atmospheric pressure at aii times. The resulting pressure difference is the main driving force for the growth of the cells. The high pressure gas in the cells causes them to espand. The rate of ceIl growth is determined by the gas generauon rate, the hcaung rate, the processing time, the amount of gas that diffused into the polymer mamx, and the strengrh of the polymcr matris.
As the ceiis grow, adjacent cells wiil be in contact with each other. There wiU be a thin polymer waii formed between the neighboring ceils (Figure 4-2(a)). The thickness and the strength of the waii wvill decrease as the ceUs grow. When the waii is broken, nvo cells becorne one big bubble (Figure 4-2(b)). This phenomenon is called ce11 coalescence. Chapter 4 Rotational Moiding of PE and PP Foams with the Dry Blending Method
------
According to thermodynamics, the wo adjacent ceils tend to merge due to the fact that,
through coalescence, the reduced surface area of ceils wili reduce the surface energy [27-291.
In other words, through cell coalescence, the total free energy of the polymer system will be
reduced. Since plastic foarns of a smder ceii size and a narrower size distribution eshibit
bctter mechanical properties [38], it would be desirable CO prevent celi coalescence and
prevent the ceii density deterioration. The melt suength, which is a degree of resistance to
cstensiond flow of the cell waii durhg the drainage of the poIymer in the cell wall when the
volume expansion occurs, should be high enough to prevent ce11 coalescence. It is known
that the meIt strength decreases while the temperaturc increases [23], so mainraining a low
processing temperature during rotationai foam molding is criucally important in reducing ceU
coalescence and thereby for the production of high quality foams.
4.3.4 Ceii Coarsening
During ceU growdi, when nvo bubbles of different ce11 sizes are adjacent to each other,
the gas in the smaller one tends to diffuse into the larger one. This phenomenon is cded celi
coarsening. The six of the smailer bubbles will be further reduced und it collapses. As a
rcsult, the nvo cells will merge inco one and the ce11 density will be lowered. This is because
the pressure is higher in the smder bubble, dut. to the surface tension effcct [28]. The
pressure difference benveen the nvo adjacent celis can be espresscd as a function of the
surface tension on the polymer layer that separate them [39]:
where R, and R, are the radii of ~henvo celis, Y is the surface tension, and Al?' is the pressure di fference. Chapter 4 Rotational Molding of PE and PP Foarns with the Dry Blending Method
In order to reduce ce1 coarsening, ceii nucleation at al1 Iocations should occur as simultaneously as possible. Therefore, CBAs having a higher decomposition rate are pre ferred.
4.4 Experimentation
4.4.1 Experimental Procedure
Whjle perforrning roto foam molding experiments, the polymer materiais used in che cxpcriments are requixed to be in a powder form. PE is available in a powder form and is convenicnt to use, whereas PP is usually in a pellet form and must bc ground before processing. The polymcr powder is blended with CBAs and other additives, and the mixture is molded in an in-house designed rotational molding machine and foarned samples are produced. In order to understand the temperature variation in the moid, esperiments were aIso conducted for measuring the temperature profiles in the mold.
4.4.2 Experimental Setup
The uni-axial rotomolding machine used for the rotational foam moléing esperiments is presented in Figure 4-3. It consists ofa cylindrical aluminum mold, a motor, a hollow shaft, an oven, a temperature controller, and a \vater cooling system. A thermocouplc is uscd in the center of the hollow shaft to monitor the temperature in the center of the rnold. The diameter of the rnold is 1.25" and its length is 4.00".
4.4.3 Grinding of Plastic Materials
PP pellets must be ground before use. The grinding was done in the facilities of
\Y!edTech, the indusrrial sponsor of the project. A schematic of the small scak grinding Chapter 4 Rotational Molding of PE and PP Foams with the Dry Blending Method
machine is shown in Figure 4-4. Polypropylene pellets are fed into the grinding head at the
center. There are two cutting plates with radial cumng blades. One of the plates cm be
changed so that the partide size can be varied. The ground smd partides axe collected at the
bottom edge of the grinding head by a vacuum purnp. The paracles are sieved in a mesh
with a specified mesh size. Those conforming to the size requirernent are coilected rhrough
the coilecting channel. The remaining particles that are too big are recycled to the grinding
hcad to be reground.
PP materials must be frozen before they are ground. Othentise, the frictional heat gets
accumulated and the temperature of the cutting blade and the surrounding air quickly
rcaches the polymer melting point. Thercfore, before grinding, the materiais were placcd in
a freezer n-ith a temperature of 40"C. In this way, the materials becorne more bride and easier to grind.
4.4.4 Hot Stage Microscopie Analysis
The ceii nucleation, growth, coalesccnce, and coarsening phenomena were obsewed
using an opticd rnicroscopic system (Olympus BH2) with a hot stage (Mettler FP 80 HT).
The mismrc of polymers and CBAs were put on the hot stage and heated up to 200 "C w-ith a heating rate of 10 "C/rnin.
4.4.5 Rotational Foam Molding
Rotational foam molding experirnents were carried out on the uni-asïal rotational molding machine iüustrated in Figure 4-3. A series of experiments were conducted to determinc the effect of processing parameters on the cell morphology of different material foams. The oven temperature was varied from 200 "C to 350 OC. When Celogen AZ was Chapter 4 Rotational Molding of PE and PP Foams witb the Dry Blending Method used as a CBA for PP foams, Zn0was used to promote the decomposition of the CBA and its amount was changed to adjust the desired decornposition temperature. The mount of polymer materials and chernical blowing agent were determined by the desired volume cspansion ratio, e.g., three-fold and sis-fold expansion based on the gas yield of CBA (see
Section 4.5.8).
The oven must be turned on and wmed up 15 minutes before the esperiment to cnsure chat the temperarure reaches the set temperawc. At &c same ame, the mold should also bc prepared. Firçc, the venulation cube must be checked to makc sure it is not blocked by polymer material and is replaced if necessary. Othenvise, the accurnulated (hliilcr-Stephenson, MS-122N/COJ is sprayed into thc moid so that the foamed part does noc stick to the surface of the mold. Aftcr the release agent has dricd, the misturc of CBh (additives) and poIymer is charged into the mold. Before mounung the mold onto the rotationai shaft, the mold should be shaken in the longitudinal direction to cvenly distribute the powder materials in the mold. The motor that rotates the mold is mmcd on and the specd is adjusted to about 5 rpm (revolution per minute). Then the mold is inserted into the ovcn and the processing rimer is started. After the processing urne has elapsed, the mold is pullcd out of the ovefi and cooled with running tap water. Ar this stage, the mold should kcep rotaung so that a uniform cooling and thereby a uniform ceii stnicrurc can be obtained. The cooling lasts about IO minutes, and the tap water cm chen be shut off. Finally, the mold is removed to demold a foamed part. Chapter 4 Rotational Molding of PE and PP Foams with the Dry Blending Method By stopping the experiment at fixed intervals during the process, the suttering, cell nuclcaaon, and ce1 growth behaviors of each materiai were monitored. Sarnples were collected and the structure of the samples was investigated using a Scanning Electronic hlicroscope (SEM). The foaming process was investigated by studying the SEM images. To ensure rhat the results were comparable. sarnples used for analysis were coliccted from the sarne position of the rnolded parts. In addiaon, esperïments were repeated to confimi the repeatability of the results obtained. 4.4.6 Characterization of Foams The ceii structures of the roto-moIded foams were characterized in terrns of the ce11 dcnsiry, the volume expansion ratio and the average ceii size. Referring to Park (331, the cell density (or ccii population densiry) is defined as the nurnber of cells (or bubbles) per cubic cenumeter of unfoarned original material. Since the nurnber of celis per square centimeter of foarned materiai cm be easily counted, the ceii densin. can be detemiined from this \due followed by cornpensating for the volume expansion. First, the number of bubbles, n,, in a square of edge length, f, is counted from a micrograph of the sample. The vdue observed is then nomialized to the number of cells in a 10 Pm square, n,. This normalized value is converted to the number of bubbles per volume of 1 cm'. Finaiiy, this converted numbcr is multiplieci by the volume espansion ratio to dctcrmine the ceil density, N,,, which can bc espressed by the foilowing equation: N,,= I nb x (l0wj2I~~ x 109 x O Chapter 4 Rotationai Molding of PE and f P Foanrs with the Dry Blendine Method wherc whcre p = density of materid V, = initial volume of unfoamed solid plastic. A schemaac of the calculation procedure is iliustrated in Figure 4-5. 4.4.7 Temperature Pronle In order to determine the temperature profile wirhin the mold during heating 2nd cooling, a thermocouple was inserted into the mold through the hoiiow shaft. The tip of the thermocouple was positioncd in the cenrer of the mold cavity. The meesured temperature was recorded every 15 seconds. Diagrms of temperature as a funaion of urne were plottcd. 4.5 Results and Discussion Scveral factors chat affect the foam quality werc studied through die esperiments. For pal\-ethrlcnc,. foams of sis-fold expansion with a fine ceii structure were obtaincd when Celogen OT and sodium bicarbonate were used as the biowing agents (sec Figure 4-6). In the case of polypropylene, Celogen A23990 was used as a blowing agent ivith Zn0 as a promoter for three-fold expansion. Foams of six-fold cspansion were dso achieved. Chapter 4 Rotationai Molding of PE and PP Foams with the Dry Blending Method 4.5.1 Investigation of the Temperature Protile in Rotational Foam Molding In Figure 4-7, temperature promes in the mold under different conditions were shown. First, an empty mold was put in the oven and went through the normal molding process. In rhe figure, the cuve "Outside, empty moid" indicates the temperature variation at the point close to the mold surface and the cunre "Inside, empry mold" illustrates the temperature change at the ccnter position of the mold. Then, the temperature curve \vas plotted when 15 grams of hlT4390 (PP) was charged in the rnold, which is marked as "Pure hIT4390" in Figure 4-7. In the end, temperature profiles at the center of the mold were obtained when 1% 50Zn0/50hZ and 5% 5OZnO/5OhZ were added. From these profiles, the foam process in rotationai molding can be classified into five stages. In fact, the first stage rcpresents the pure heating of the rnixmres in the mold. This is why the obtained heaung cunre is almost linear after the process stabilizes. For this stage, there is not much difference bcnveen heating an empty mold and a charged one. If cornparhg the heating cun7es,the one developed from heating an empty mold is almost the same as the one from heaung a chargcd mold within this stage. During the second stage, the speed of the temperature increase slows down. Nso, the heating curve for the charged mold becomes lowcr rhan the one for the empty mold. It is beheved rhat polymer sintering and melting begin in this period. On one hand, the polymer layer developed on the inner surface of the mold reduces the heat transfer to the powders. On the other hand, because the sintering and the melting of polymers is an endochermic processes, some heat is absorbed during melting. During the rhird stage, the CBA is dccornposed. Since the CBA used in the esperiments was an esothermic blowing agent, Cclogen AZ, a large amount of heat was generatcd during the decornposition. Graphicaily, it was indicated by a rapid jump in the temperature profile. In abour 15 minutes, the decomposition of CBA is completed and the system enters into the Chapter 4 Rotationai Molding of PE and PP Foams with the Dry Bknding Method fourth stage of ceil growth. When the Corn structure is deveIoped, the heat transfer rate was rcduccd due to the insulacion effect of foams. After it reaches the designated heating time, thc mold is pulled out and cooled with mnning water, which reprcsencs the fifth stage of the entire process. 4.5.2 Investigation of Sintering of Powder Particles and its Relationship ~4thCeU Nucieation Sis-Fold Es~ansion In processing of PP foams of sis-fold expansion, three materiais with different viscosity were chosen: SC873 (50 MFR), SD242 (30 MFR), and MT4370 (20 MFR). Fully espanded foams were successfully produced from SC873 and SD242 while no good foam structure could be achieved from MT4390. To esplain the different results in roraaonal foam molding of thesc three rnaterials, the heaang was stopped at 13, 14, and 15 minutes and sarnples were taken from the foarned parts to invesagate the sintering quality of polymer powder particles. In order to veri+ the repeambility of the esperiments, three sets of csperiments were conducted ac the samc conditions for each matcrial. Figures 4-8, 4-9, and 4-10 show the SEhI picmres of the selected foam sarnples of SC873, SD242, and MT4390. In Figure 4-8, we can see that when SC873 with a MFR of 50 dg/rnin was used, powder particles began to form a uniform polymer melt after being heated for 13 minutes in the oven, although some unsinterrd particles were stiii found. At 14 minutes, the sintenng was completed although a number of smaii air pockets still esisted in the polymer matrix. These air pockets are believed to be the enuapped air arnong the polymcr particles as indicated bv Liu et al. [35]. Due to the low-pressure processing condition of rotomolding, Chapter 4 Rotational ~Moldingof PE and PP Foams with the Dry Blending Methoà thesc air pockets were very difficult to remove when polymer particles were dry blended with CBA particles. At 15 minutes, the dccomposition started. When SD242 with a MFR of 30 dg/rnin was used, the sintering is not as good as that wich SC873. At 14 minutes, several large air pores still existed in the polymer matrk (see Figure 4-9). \Wen hfT4390 with a MFR of 20 dg/min was rotomolded, the sintering qualit). bccame worse. As shown in Figure 4-10, at 13 minutes, polymer partides were sri11 in the early stage of sintcring and individual polyrner particles were clearly idenafied. At 24 minutes, chcre were a large number of entrapped air pockets in the polymer matris, some of ufhich were connected with each other. \men it rcached 15 minutes, huge caviaes u-cre formed. Cornparison of the sintering behaviors of Pl? materials with various MFRs indicntes that the sintering of powder pdcles plays an important role in the ceIl nucleation proccss as in the case of PE processing [35] in the dry blending method. \men the sintering of the polymer powder particles is good, the encrapped pockets \vil1 bc isolated in the molten polymer matris bcforc the decomposition of CBAs. These bubbles dlbehavc as cell nuclei in the foam processing [35]. If the sintering of polymer powder is poor, the enuappcd air pockets will not be easily isolated before the decomposition of CBAs. These co~ectedair pockcts are problematic in the foam processing. As the CBAs begin to decompose, the cgas gcncratcd from the decomposiuon of CBAs will inflate the connected caviacs, forrning hugr cm-ities or losing all the gis rhrough these connccted cavities eventually to a ventilation holc. So, when MT4390 with a relaavely low MFR of 20 dg/rnin was used in the sis-fold cspanded foam processing, huge cavities wcre observed in the final foam structures because of the poor sintering behavior of plastic powder phdes. Chapter 4 Rotational Molding of PE and PP Folars wiîh the Dry Blending Method ------ Therefore, in order to achieve a fine-cell foam structure, formation of big air pockets çhould be prevented. Since these the big caviaes are gencrated by the poor sintering of polymer powder partides, it will be criücai to have a gooci sintering in thc rotomolding of PE and PP foams. Three-Fold Es~ansion ,As presented in Figures 4-8 and 4-9 the sintering qualiry was improved when low viscosity PP materials with high &ER were used for sis-fold espanded foams. hlthough it is cstrcmely difficult to remove the entrapped air completely in rotauonal molding, a good sintering can easily isolate these air bubbles and prevent formation of huge cavities. As a rcsult, better nudeaaon is forrned and a better foam structure is obtained. The improvement of ce11 suucturc by using a high MFR PP was further verified by the detailed espcriments on three-fold expansion foams. From the ce11 morphology shown in Figure 4-1 1 (a), when PF633 of a high viscosity (MFR 3-6 dg/min) was used, no fdy cspanded foam structure was obtained, dthough a propcr amount of blowving agent was added. It seemed chat the viscosity was too high to promote good sintering, and as a consequcnce, most of the gas was lost. Whcn SD812 u-ich a higher hIFR of 16 dg/min was used, a uniforrnly distributed fine-ce11 structure was obtained as shown in Figure 4-11 (b). \Tiith an increase of MFR (or equivalently a decrease of viscosity), the sintering was improved and better foam quaiiry was achieved. In Figure 4- Il (c), the cell stxucture was fürther improved, especially the discribuuon of the cells, when AIT4390 (MFR 20 dg/min) \vas iised. Figure 4-12 (a) shows chat the ccU population density of hIT439O foams is four ames as high as chat of SD8lZ foams. Chapter 4 Rotational Molding of PE and PP Foams with the Dry Blending Method However, a low viscosicy is also detrimental to the foam structure. If the viscosity is too low, the melt strength becomes low (see Section 3.3) and ce11 coalescence ma. govem the foam structure. In the ocher words, when the MFR is too high, the celi nucleation behavior wiU be sztisfactory because of good sintering. However, because of the weak melt strength, severe ceIl coalescence cm be promoted and the high nuclei density may not be successfuiiy preserved during ceii growth. As a result, big bubbles may Le formed in the foam structure. Therefore, there is an optimum viscosity ievel (or opamum hIFR level) co achieve the higl-iest ce11 population densiv. Since the sintering behavior of plastic powder particles suongly depends on the nmount of the plastic powder and CBA charged into the mold, the optimum due of \-iscosity to achieve the highesr ceii density is a funcuon of the desired volumc espansion raao of foam. For esample, for processing of three-fold espanded foams, a MFR higher chan 15 dg/min is necessary to achieve a good sintering behavior and the optimum level of viscosity was observed to be 20 dg/min (see Figure 4-12). On the other hand, for processing of sis-fold espanded foarns, the sintering behavior of SD812 (MFR 16) and AIT4390 @IFR 20) were not satisfactory. A MFR higher than 30 dg/min was necessaq* to produce a uniformiy disuibuted and Uyespanded foam structure. It is interesting to note chat the sintenng of behavior of PP powder material was bctter for three-fold expansion than for sis-fold espansion. For example, when 25 g of MT4390 powders with 1% of Celogen AZ were chargcd into the roto-mold for three-foId espansion, good sintering was observed at the foaming stage. However, when 13 g of MT4390 powders with 2.5% of Celogen AZ were charged into the roto-mold for sis-fold espansion, the sintering behavior of the ~owderswas poor at the onset of decomposiaon of CBA. Chapter 4 Rotational Molding of PE and PP Foam with the Dry Blending Metbod It was not dear why the sintering was better when more plastic materials with relatively less amount of CBA were used in the rotational foam molding. However, it is believed rhar rhe larger thermal inertia of the greater amount of plastic materials was responsible for this diffcrence. When a large arnount of materials were charged into the mold, the heating rate \vas lower because of an increased thermal inema. As a consequence, the decomposition of the erothermic blowving agenr wiU be deiayed (see Figure 3-1 2). In other words, more rime \viU be given for the sintering of plastic powder materials before the decomposition of CBA. On the other hand, when a less amount of materials was charged into the mold for sis-fold cspansion, the heating rate wouid be higher, and the decomposition of esorhermic CBA would be eariier. In other words, the plastic powder parades did not have enough timc to sin ter together before the decomposition of CBAs. Anorher reason for the poor sintering of plastic powders may be due to the incrensed parucle-to-pamcle disrance of plastic powders for the larger espansion composition of materials. It should be noted that for larger espansion, the relative CBA amount is grearer, and therefore, the powder-to-powder distance of plastic powder parricles is greater. 4.5.3 Investigation of Effect of Powder Quaiities on Ceil Morphology Figures 4-1 3 and 4-14 show the effects of physical shapes of polyrner powders on the cc11 morphology of Pi? foams. Figure 4-13 is the foaming results of SD812 for two different particle sizes (see Figure 3-1). Figure 4-14 shows the foam morphology of samples made from SD812 powders with and without long &S. The long tuls of powders wcre observed ro be formed when the pellets were ground into powders without being frozen. Effcct of Particle Size Chapter 4 Rohtional Molding of PE and PP Foams with the Dry Blending Method Figure 4-13 shows that Ligger poljmer powder partides yielded a bigger ce11 size and lotver expansion. Since the ceII-to-cell distance is deterrnined by the average particle site in the dry blending technique, the celI size would increase as the power particle size increases. Furthcrmore, it seems to bc more difficult for larger particles to sinter each other. Because of the poor sintering, the re1eased gas easily escaped and fully espanded foam structure could not be obtained. These results imply that a smaller particle size is favorable for deveioping a fine-ce11 structure in the dry blending based method. Unfortunately, the available parride size was limited and further investigation could not be made. ECfcct of Particle Shape Figure 4-14 illustratcs the effect of long tails attached co the powders on the ce11 morphology. Poor foam cells were generated from those particles with long tails. The hair or tail of these particles caused a poor contact benveen parricles and made it difficult to achieve good sintering. As a result, the generated gswas lost and huge cells tvere obtained. 4.5.4 Investigation of Effect of CBAs on CeU Morphology In Figure 4-15, the effect of the CBAs on the foam qualil of sis-fold PE foams is shown. Celogen OT ("OT") was used in (a) while in @) sodium bicarbonate ("SB") was the blowing agent used for the foaming. %%en Celogen OT was used, a finer ce11 structure was obtained. It was not clear whctiier the bcner ce11 structure of PE foams using Ceiogen OT was due to the rapid dccomposiaon (see Figure 3-11) of Celogen OT or due to the higher decomposition temperature (see Figure 3-10). Since the sodium bicarbonate starts to decompose at 133 "C,the sintering behavior of PE particles at the foaming stage may not be as good as in the case of using Celogen OT. Therefore, the cell nucleation wdlbe better in Chapter 4 Rotationai Molàing of PE and PP Foams with the Dry Blending Method the case of Celogen OT for PE foam processing. Furthemore, ceIi coarsening could have affectcd the foam strucnue. During ce11 nucleation and g-rowth, the bubbles chat nucleated first were inevitably larger than those that were developed lacer. The smder bubbles will evcntudy merge into the bigger ones and cdcoarsening becomes severe. In addition, ce11 coalescence may also occur if the decomposition lasts long. As a result, the cell populaaon dcnsiry is rcduced. Therefore, for the CBhs that decompose over a wide temperature range such as sodium bicarbonate, it wodd be difficuit to obtain fine ceIl structures. On the other hand, for blowing agents such as Celogen OT, the decomposition occurs vev quiclrly and most of the cells nucleate simulraneously. Cell coarsening is relatively low, which dows cclls to grow uniforrnly. Since it was not clear which of the two dorninated the celi forniauon, a future study is required. 4.5.5 Investigation of Effect of CBA Particle Size on Ceii Morphology Figure 4-16 shows the results of SD812 foams obtained whcn another azodicarbosamide CBA ADC series from Bayer was used. The information contained in this figure is helpM to undersrand the effect of CBA particle size on the foam quality. Figure 4- 16 (a) presents the foam gencrated by hDC/hZ-Cl (the CBA uith a smdl particle size of 4.5 Pm), while Figure 4-16 @) presents the foam generated by ADC/F that hns a particlc six of 15.5 Pm, three times larger than that of ADC/M-CI. As show in the figure, no significant difference was identified benveen the morphologies of the two sarnples. It seems that a blowing agent with a smailer particle size is easier to disperse into the polyrncr powder and therefore a better degree of mixing during dry blending cm be acl-iicvcd. Consequently, a bettcr Çoarn structure will be produccd. Howcver, sincc the nuclcation in dry blending method more depends on the sintering quality of the pol)*mer Chûpter 4 Rotational Motding of PE and PP Foams with the Dry Blending Method - - - powder particles and the CBA particle sizes were much smder than that of the plasac powder particles, the effect of the CBA particle size on the resultant foam structure was not significant. It should be noted that therc is no big ciifference benveen the decomposiaon behaviors of ADC/M-Cl and ADC/F (sec Figures 3-9 and 3-1 1). 4.5.6 Investigation of Effect of Zn0on Cell Morphology Figure 4-17 iiiustrates the effect of Zn0 on the ccil morphologies of PP (hiT4390) foarns when AZ3990 was uscd as the blowing agent. From previous work 1351 and the TGA test on the effect of Zn0 (see Figure 3-13), it is known chat the decomposition temperature of azodicarbosamide blowing agent is lowered with the presence of ZnO. The rational bchind using this caraiyst as a decomposiuon promoter is that the melt strength of PP will be increased by lowering the processing temperature via the dccomposition temperature control. When the melt strength of PP was increased by having early dccomposition of CBA, cc11 coalescence wiil bc reduced and improvement of ceil structure is espected. The cfiect of Zn0 on ceii morphologies is depicted by the nvo estreme rcsults shown in Figure 4-17. With only a 20 "C decrease in the dccomposition onset temperature (see Figure 3-13), the average ceil size was reduced from 3 mm to 0.7 mm (?O0 pm). Howevcr, it should be noted that the effect of lowering the decomposition temperature on the ce11 morpholog wi11 be good only whcn the sintering of the plastic powder particies is good at decomposiuon. If the decomposiaon is roo early, the resultant ccii structure will not be good because of the unfinished sintering of plastic powder particles. 4.5.7 Investigation of Effect of Processing Parameters on CeU Morphology Chapter 4 Rotational Molding of PE and PP Foants witb tbe Dry Blending Method Effect of Blowineenr Amount A foamed product is made up of a solid polymer mauis and a gas phase. The gas is generatcd from the chemical blowing agent. To achieve a desired expar.sion raao, a propcr amount of chemical blowing agent should be used. Theoreacally, this cm be estimated using the following equations. wherc - mp<>tw- weight of pure polymer material used, g, m,,,, = weight of blowhg agent, g, V= total volume of the mold, cm3 O= espansion raao, cP = volume of gas generatcd per mir mass of blowing agent, cc/g, p = densic). of rhe polymer materid, g/cc. Thc percentage @y weight) of CBA used can be expressed as hccording to the data provided by the CBA manufacrurers, the amount of gas gcnerated from Celogen A23990 is 220 cc (STP)/g and the density of PP is around 0.9 g/cc. Ir is Chapter 4 Rotational Molding of FE and PP Foams with the Dry BIending Method worth noang that the gas yield data provided by the suppliers is at room temperamre. A typical schematic of a measurement system for CBA characterizauon is given by Klempner and Frish 1131. However, the espansion of foam occurs at a very high temperature, chat is, abovc the polymer melting temperature. As the temperature decreases during thc cooling process, the maximum possible gas volume will be reduced. Since the polymer will be frozen at the cqstallizaàon temperature, the structure of the foam is fixed at this temperamre. In other words, the void fraction of the foam is determined by the occupied gas volume at the crystallization temperature. Since the volume of gas at this high tcmpcrature is larger than thc volume at room temperature, the ps yicld data provided by the manufacturer shodd be corrected in calculating the required CBA amount in Equation 4-6: (4-7) where V,,,,, V,:, T,,, and T,, are the gas volume at the room temperamre, thc 9svolume at the crystailization temperature, the absolute room temperature and die absolutc cnstallization temperature, respectivelly. The room temperature is 25 "C,or 298 K, and the cryscalJization temperature of PP is approsimately 235 "C,or 408 K. Thercbre, for the 80cc mold used in the experiment, ir is estimated that the amount of chernical blowing agent nccdcd is 0.625O/0 for three-fold espansion and 1.5% for sis-fold espansion. The effect of varying the chernical blowing agent amount on the foam structure was invcstigated. For thrre-fold expansion, the expenments were conducted while varying the CBA from 0.5% to 1.5%. In the case of sis-fold expansion, the arnount of CBA was varied from 1.5% to 4.0%. Zn0 was used to mode the decomposition behavior when it was ncccssary and possible. Chapter 4 Rotational Molding of PE and PP Foams with the Dry Blending Method Figure 4-18 shows the ce11 morphologies as a function of the blowing agent amount and Figure 4-19 illustrate the effcct of CBA amount on ceU size, celi population density and expansion ratio. For the three-fold expansion samples of MT4390, it was clearly obsened that, too little an amount of CBA resulted in big ceii size due to a deficient arnount of gas and the bubbles were over espanded to fil1 the mold cavity. Too much gas also resulted in a poor cc11 structure because of too high gas pressure. Ir was also beiieved that when a large amount of CBA was used, the decomposicion occurred carlicr before the completion of sintering. As a consequence, the ce1 structure was not desirable. The esperimental results indicated chat an optimum arnount of CBA should be applied according to pamcular product requirement. For esample, 0.625% AZ3930 together with 0.625% Zn0 is found to yield best results for three-fold expansion in Our esperiments. On the other hand, a much Iargcr CBA amount was required for sis-fold expansion. For sis-fold espansion, approsimately 2.5%-3.0% of A23990 gave the best result. It should be note that this is much higher than the cheoreticaiiy predicted value @y Eq. 4-6) because of the lower blowing agent efficiency. First, the gas loss rnay be high when less polgmcr is used. In addition, the CBA may not havc been fÜUy decomposed during foaming. During cspcriments on sis-fold expansion, the optimum amount of Zn0 was reduced to 05'0 to incrcase the decomposition temperature so that there was enough ame left for achicving bettcr sintering of the polymer particles. With chc reduced amount of decomposition promoter, the AZ may not have been fully decomposed at the particular time/temperature whcn the espenments were stopped. Figure 3-13 shows that with a reduced amount of ZnO, thc residue increases. For esample, at 250°C, there is sd 20 percent rcsidue for pure Celogen AZ3990, but only less chan 10 percent residue is Ieft at 190°C when Zn0 was added. Therefore, more CBA materials were consumcd wich ZnO. Chapter 4 Roîational Molding of PE and PP Foams with the Dry Blending Methoci Effect of pro ces sin^ Time Figures 4-20 and 4-21 illustrate the effect of proccssing ume on the foam morphologies of three-fold espanded MT4300 (PP) foams. The oven temperature used for che sarnples was 350 "C, and the blowing agent was 0.625% of Celogen AZ together u-ith 0.625% of ZnO. The processing time was varied from 10 minutes to 25 minutes. It tumed out that the processing time is another criacal parameter in the roto-molding of PP foams. When the processing tirne was too short, the foam was not fully espanded. Figure 4-20 clearly shows chc unfoarned yet sintered polymer laycr where the temperature \vas still below the decomposition temperature. However, when the processing urne is too long, a coarse foam structure was obtained. This seemed to be due to ceU coalescence. As the processing time increases, the temperature of the polymer becomes higher because of the longer esposure Ume of oven. The rnelt suength and viscosity of PD drop dramaticd!* as thc tcmpcrature increascs over the mclting temperature (see Figure 3-3). Thc pcssibility of cc11 coalescence and ce11 coarsening increases as the melt strength decreascs. Therefore, it would be desirable char the processing be stopped as soon as the espected volume espansion is obtained. Effect of Oven Tem~cramre The effect of oven temperature on the ce1 morphology of PP foams \vas also studied using SC873. Four different oven temperatures of 350 "C, 325 "C, 300 "C,and 275 "C were tested in the esperiments. In al1 the esperiments, 13 g of SC873 powdcrs and 3% riz3990 were charged into the mold for six-foId expansion. In Figure 4-22 shows the foams obtained under oven temperatures of 275 "C and 350 "C. For the same amounts of plastic Chapter 4 Rotational Molding of PE and PP Foams with the Dry Blending hlethod and CBA rnaterials, the higher the oven temperature, the higher the heating rate. It was found that too high a heaang rate yielded very coarse structures becausc of the unique thermal behavior of the exothermic CBAs under a high heating rate. From DSC analysis rcsults shown in Figure 3-10, it can be scen that much more heat is generated when the hcating rate is increased for Celogen m. The heat generated from the reaction will incrcase the local temperature, dmost instantaneously. For esample, in Figure 4-23, the effcct of this csotherrnic hcating is dearly depicted. The acnial temperature cari be 20.52 "C higher than the thcoreucal temperature under a heating rate of 76.35 "C/rnin. Consequently, the mclt strcngth dlbe lowered at interface of the polymer mclt and the growing gas bubbles. This u-il1 promote severe ceIl coalescence and reduce the ceii population density. However, a low oven temperature, or a low heating rate, is noc pracacal in production duc to cconomical rcasons. hocher important factor is that, under a iow heating rate, polymer pamcles tend to sück to each other locaiiy, inscead of forming a uniform polyrner lnycr on the waii of the mold (Figure 4-24). This ''bding phenomenon" also leads to non- uniform structures because of non-uniform distribution of materials. Tnerefore, setang a proper oven temperature is also critical in achieving good quality foams. For the particular mold used in Our esperiments, it is detemiined that 300-350 "C is suitable for three-Çold expansion and 275-300 "C for sis-fold expansion. Chapter 5 Rotational Molding of PE and PP Foams with the Melt Compounding Method Chapter 5 Rotational Molding of PE and PP Foams with the Melt Compounding Method 5.1 Introduction A dr). blended mixture of a chernical blowing agent (CBA) and polymer resins \vas used in the prcviously presented research for the manufaxuring of fine-ceIl, low-densin PE and PP foarns using rotaaonal molding. As a result of this research, PE and Pl? foams of up to six-fold expansion have been successfuliy produced, and the foarning rnechanism and the critical factors chat affect the foam quaiity have been clarified. In dry blending, ce11 nucleation and growth are primarily derennined by the sintering behavior of the plasac resin pamcles during the rotational molding process. Since the sintering ability of the polymer material is directly related to the viscosity, the selecuon of materiais wns limited to those having a higher melùng flow rate @ER) at the espensc of melt strength. In a poor sintering process, the napped air pockets formed among the polymer particles resuit in a relativcly coarse foam structure. It is also difficdt to obtain a large expansion ratio due to the poor sintcring properties or che weak melt suength of such materials. Melt compounding is an alternative method for material preparation in the rnanufacturi~~of fine-ccli PE and PP foams. Melt compounding refers to the process of rnising the chernical blowing agent and the poljmer resin using an extrusion compounder. The estruded mixture is cut into pellets and used for the rotational molding In the melt compounding process, the chcrnical blowing agent cm be bener distributcd in the polgmer mauis due to the high shear force produced in the compounder. This could bc of great hclp Chapter 5 Rotational Molding of PE and PP Foams with the Melt Compounding Method in achieving a uniforni ceii disuibution later in the process. In addition, the melt compounding process ensures better wemng of the polymer melt on the CBA pamcles and removes al1 the trapped air pockets. Moreover, the CBA(s) are weU distributed in the polymer matrix after compounding. Therefore, the sintering behavior of the polymer particlcs is no longer the dominant force in the foaming process. Under these circumstances, the nuclei wiiI bc the CBA parucles, and cell nucleaaon WU be totally controiied by the decornposition behavior of the chernical blowing agent(s) and other additives. Since ceii nucleation no longer depends on the sintering qualit). or the viscosity of the material, a much wider range of materials cm be used for the production of PP foms in rotationai moIding. Furthemore, by using high melt scrength materials foarn suuctures of large espansion ratios cm be achieved; such structures were noc easily obtained with the dry blending rnethods. 5.2 Overview of the Process In compounding based rotauonal foarning, rhe first step is the dry blending of the polymer materiais, CBA(s), and other additives. Good premixîng helps improve the rnising qualit). in the meIt compounding stage. Afcer the matcriais enter the compoundtr, the high shear force in the extruder barre1 induces further rnising. However, chernical blowïng agents are heat sensitive rnaterïals. The viscous heat generated by the high shear rnay cause the dccomposition of the CBA(s) at the estrusion compounding stage, which shouid be prevented. Therefore, the shear force should be maintained in a proper range. In the second stage, the inisture is esuuded and then cut into pellets with a peileutcr. The final stage is the foaming of the compounded peiicts in the rotaaonal rnolding machine. Chapter 5 Rotational Molding of PE and PP Foams with the Mdt Compounding Method 5.3 Processing Steps in Rotational Molding 5.3.1 Ceii Nucleation In the mrlt compounding process, air pockets among the polyner particles or on the surface of the chemicai blowing agent panides are removed dunng compounding, therefore, controliing the ceU nudeation during foarning is simpler than with die dry blending method. As a result, the ceU nuclei corne only from die chemical blowing agent particles. The size of rhc CBA partides and their dismbution in the polyner rnacrix determine the ce11 population density and distribution. The shear Force involved in the process of compounding produces a berter quality mixture. In addition, rince CBA particles tend to agglomerate together, ir is desirable thar a high shear force be applied to break the agglornerared particles in the compounder. Therefore, the distribution of CBA(s) is better in melt compounding. Gcnerally speaking, because of the eliminaaon of rrapped air pockets and the better distribution of CBA(s), a better ce11 nuclcation is obtaincd. 5.3.2 Sintering of Pellets and Ceil Growth The compounded pellets are loaded into a roto-mold, and the mold is placed into an oven and rotated. Afier being heated up, the compounded pellets begin to sinter together. \Vhen the temperature reaches thc decomposition temperature of the CBA(s), dccomposition begins. The esbubbles generated by decomposition behave as ceii nuclei. Thc ce115 continue to grow und the CBA(s) are fdy decomposed. The pressure from the generated gas helps the pellets sinter better. Unlike in the dry blending based method, the sintering quality of the pellets affects only the stnictural and mechanical properties of the final products, with no significant effect on the foam quality. The ceii size and population Chapter 5 Rotaiional Molding of PE and PP Foams with tùe Melt Compounding Method density wiii be completely deterrnined by the decomposition behavior of the chernical blowing agent(s) and by celi coalescence. 5.4 Experimentation 5.4.1 Experimentd Setup Figure 5-1 shows the compounding system used in the esperïment. Part (a) is a schematic of the system and part (b) is a photograph of the acrual equipment. A Brabender single screw extruder (05-25-000) that is driven by a DC motor carries out the melt cornpounding. The estrudate from the esuuder is cooled in a watcr bath (WedTech). Afrer the cooiing section, the extrudate is dned and pelleltized in a peiletizer (Brabender 12-72- 000). The pellets are collected and used for the rotational molding. The rotauonai molding is conducted on the same machine as shown in Figure 4-3. 5.4.2 Materiais Esperiments were conducted on both PE and PP. For PE, LL8556 was chosen as the base polymer material. Both Celogen OT and Sodium Bicarbonate were tested as CBAs for PE foams. For PP foams, efforts were concenuated on medium and Iow hIFR materials in ordcr to investigate the possibility of using a wider range of materials in roto foaming. The materids used were lMT4390' SD812, and PF633. Celogen AZ 3990 was used as blowing agent for the PP foams. 5.4.3 Melt Compounding The melt compounding proccss involved several steps. Chapter 5 Rotatiod Molding of PE and PP Foams with the Melt Compounding Method - - First, the heater was tumed on to warm up the system to the designated processing temperature. This temperature is just above the melting temperature of PE or PP and lower than the decomposiaon temperature of the selected blowing agents. According to Sterenr [23],the temperature protile dong the barre1 should be varied purposely so as to produce optimum meltïng as weU as good material transponation. The temperature of the first section, i.e., the material transport section, should be lower chan the rnelting temperature so that the materid wilI not slide in the bard But the temperature should not be too low to makc melting difficult. Second, after the system temperature srabilized, the system was purgcd wirh pure polymer materiai at least for 10 minutes so that contaminancs did not affect the quality of chc misturc. At the sarne urne, the system was re-stabilized under running conditions, including csuusion speed setungs. Third, the dry blended polymer and CBA(s) were put into the hopper and the materiais wcrc compounded in the bard The nmount of CBA was detemiincd according to Equations 4-5 and 4-6. Fourth, the estrudate was ~dedthrough the water bath and put inro the peiletizer. The estrudate was dried before being cut. The pulling speed (or the feeding specd of the pellctizcr) was adjusted so that peiiets of proper diameter and length were obtaincd. A fast fccding rntc results in a smder pellet size. The coilected pellets were used for rotofoaming. Chapter 5 Roîational Molding of PE and PP Foams with the Melt Compounding ~Method - -- 5.4.4 Rotational Foam Molding Cornpounded pellets were roto-molded with the same rotational molding machine used in the dry blending method. Processing parameters were studied and the final foam samples were analyzed. The amount of pellet used was calculated as: - m,,*,- mpl\-m'-,+mcba=mFI~mCr(~+O/oCBA) (5-1) ayhcre m,,,, is the amount of polymer mîterial and m, is thc mass of CBA, which are determincd by Equations 4-4 2nd 4-5, respecuveiiy. 5.4.5 Hot Stage Optical Microscope Analysis Two samples were chosen to be andyzed with a hot stage opacai microscope. Pellets of MT4390 wvith 0.5% of CeIogen AZ 3990, with pre-decomposiaon, were studied first. Then, a studv was carried out on pre-decomposed pellets of LLDPE 8556 and 3% CeIogen OT. A thin layer of the misture was cut from the compounded peiiets and heaced at the hot stage of the microscope. The nucleaaon and ceii growth phenornena observed. 5.5 Compounding Results and Discussion The compo~ndingqudity is very important to the final foam qudty in rotational foaming rnolding. Among ail the issues of concern in the compounding process, the decomposition of CBA is the most serious problem that affects the compounding qualit).. Duc to the shear and the heating environment in compouoding, the CBA maIr decompose during this process. This is referred to as pre-decomposition in this thesis. One of the objcccivcs of the compounding is to prevent the pre-decomposition, or to control the pre- decomposition throughout the process. Chapter 5 Rotational Molding of PE and PP Foams with tbe Melt Compounding Method Several factors that affect the pre-decomposition have been invesagated during thcse csperiments. Howcver, the rnost significant parameters are the system temperature and the rotational speed. The pre-decornposition behavior at different settings of these parameters \vas investiga ted. Figure 5-2 gives a schematic of the heat exchange process for a r'picd extrusion system from which the key factors for melt compounding cm be outiined. The input of cnergy cornes from two sources: the band heater and thc mechanical power from the motor. The mechanical energy is converted to heat by viscous shear. The heat is lost through rhree channels: (1) some heat will be lost from the surface of the barrel through radiation and convection; (2) when the polymer melt Ieaves the barrel, a certain mount of heat is carried away in the melt; and (3) heat is lost through the esrernal cooling system. Compressed air can be Forccd around the barrel to absorb heat, and coohg water is run nround the feeding chroat. Sometimes, screw cooling is also necessary. Only a proper balancc benveen hem input and output can maintain system temperatures at a stable level. Table 5-1 outlines the results of the experimenr when the temperature setangs and rocauonal speed were changed. From diese results, optimum processing condiaons for the particular esperimental system were obtained. For PP materials, pellets wlthout decornposition were successfdy obtained. For PE, when Celogen OT was compounded with LLDPE 8556, the decomposition could not be eliminated although carcfui attention \vas paid to temperature control and rotational speed setup. However, when sodium bicarbonate was used, pellets without decomposition were successfully generated (see section 5.5.4). The mechanism behind the results arc surnrnarized in the foUowing sections. 5.5.1 Effect of Barre1 Temperature on Compounding Quality Chûpter 5 Rotational Molding of PE and PP Foams witb the Melt Compounding Method To compound the chernical blowing agent particles in the polymer melt, the processing temperature musr be above the melting temperature of the polymer. However, the processing temperature must be below the decomposition temperature of the chernical biowving agent to prevent pre-decomposition. If pre-decomposition occurred in the cornpounding process, the bubbles that are generated would adversely affect ce11 nucieation in the foaming process [33].In addition, a certain arnount of the blowing agent is Iost. Since it is impossible to know the amount ofgas lost, ic is difficdt to control the amount of CBA used in the process. 5.5.2 Effect of Rotational Speed of the Extrusion Screw on Compounding Quaiity Since a lot of heat is generated from the shear force in the extruder duc to the screw motion, many local hot spots may exist in the bard Even though the barrel temperature is set low, the actual local temperature wiii be high if the rotaaonal speed of the screw is too high; this wiii result in pre-decomposition during compounding. The rotaaonal speed of the csuuder screw should be low enough to prevent any unnecessanr viscous hcat from bcing gcncratcd by the screw motion. However, the screw speed should dso bc kept wïthin a certain range. With a decrcase in rotauonal speed, the flow rate drops and the Ume that moltcn polymers stay in the barrel (which is referred to as the residence urne) is longer. A longcr residence time may cause pre-decomposition of CBAs in the extruder even at a low proccssing temperature as shown in Figure 5-3. Funhermore, when rnising qudity is a concern, it is desirable to ha\-e a high extrusion speed because a high rotational speed of the screw results in a highcr shear force and better mi'ring quality. In addition, a higher screw speed results in a high flow rate. Therefore, there is an optimum rotationai speed of the screw for compounding. A proper extrusion speed should be carcfully selected for each Cbapter 5 Rotatioad Molding of PE and PP Fo~mswith the Melt Compounding Method processed material. For the cspcriments we performed, a speed in a range of 30 to 60 RPM and a processing temperature in the range of 160 "C to 165 "C werc found to be appropriate to prevent decomposiaon in the compounding of PP and Celogen AZ. A different speed may be required if the polymer and CBA are changed. 5.5.3 Effect of Sctew Cooling on Compounding Quality For a Iarge-scale conunuous proccss, it is not sufficient to control only the barrel temperature and the rotational spced of the screw. When the generated friction overheats thc screw, hot spots may be created and decomposition \vil1 again appear in an unconuoIiable manner. To prevent ttus, screw cooling is neccssary. However, in order to install a screw coohg system, the screw has to be at least 3.81cm [36]. In addition, the dcgrcc of rhe screw cooling must be within a reasonable range because a cooled screw niII increase the shear forcc and reduce the flow rate. 5.5.4 Exothermic and Endothermic Effect of CBAs on Compounding Quality \,en PE (LLDPE 8526) was compounded with Celogen OT and sodium bicarbonate separately, very different results were obtained. Athough the onset decomposiaon temperature of Celogcn OT is about 160 "C (which is about 30 "C higher than the set cnvironmcnt temperature), severe decomposition could not be prevented. Whcn 3-5% of sodium bicarbonate is compounded with PE, with a barrel temperature of 130 "C and a scrcw speed of 40 RPM, peiiets with no pre-decomposition were successfdy generaced. Although the decomposition temperature of sodium bicarbonate is 133 "C (oniy 3°C higher than the cnvironment temperature), the compounding pmcess was much easier to concroi than that whcn OT was used. It is believed that the different thermal characteristics of OT Chapter 5 Rotational Molding of PE and PP Foams with the Melt Compounding Method and sodium bicarbonate were the reason for this difference. As Figure 3-9 (d) indicates, the decomposition ~f Celogen OT is an exothermic process. Since there are rnany hot spots at regions within the barrel with high shear force, local decomposiaon may be inevitable. \.men the dccomposiuon occurs, a large arnount of heat is generated and local temperamres can casily increase by 15-20 "C instantaneously, as shown in Figure 4-23. This leads to a series of decompositions in the neighboring area and propagace the decomposition, which makes the whole system unstable. However, when sodium bicarbonate is compounded, the decomposition is an endothcrmic process. Even if decomposition occurs, the temperamrc of the neighboring region is lowered and the decomposition does not propagace. 5.5.5 Effect of Heating Rate on Compounding QuaLity As the material is processed in the extruder, the temperature of the miscure increases from room temperacure to the melting temperature of polymer \rithin the time it takes for thc materid to travel from the feeding section to the melting section. In the esperimcnts, a umcr was started when the materid was put into the feeding throat (25 "C)and was sropped whcn the materid esited the third heating zone (mclting zone, 165 "C) without a die actachcd. The time was about 63 seconds when the screw rotationai specd was set at 40 RPM. To prevent any material residing in the barrel from affccting the measuremenr, a colored material is added. The exact time of cravehg can be calculated [23],but the simple mcasurcment in the experiment implies chat the rime requircd to heat the polymer from room temperature to melüng temperaturc is less chan one minute. This means thar the hcacing rate of the CBA and polymer is esuemely high. For esample, for PP materids, with a rotaaonal speed of 40 RPM, the temperature WU rise from 25 "C to 165 "C within one minute, and the heating rate is above 100°C per minute. This high heaang rate wiii Chapter 5 Rotational Molding of PE and PP Foams with the Melt Compounding Method ------dramaacally increase the heat generated by exothermic CBhs such as AZ3990. In Figure 3- 10, it is cIcarly shown that the temperature jump is higher under a higher heaung rate. This \r-as why the system became unstable and was difficult to conuol when compounding PP urith Celogen AZ or PE with Celogen OT. 5.6 Foaxning Results and Discussion During rotational foam molding, the influence of melt compounding on the material sclection, foam quality, and cspansion ratio was studied. It is believed that the foarning mcchanism for compounding based roto foaming is different frorn that of the dry blending based foam processing. The foaming mechanism also depends on the compounding quality. \.en there are air bubbles generated by decomposition in the compounded pellets, the nuclcation process is not the same as when pre-decomposition is completel). prevented. 5.6.1 Foarning with Non-Decomposed Pellets \,%en peiiets without pre-decomposition are rotofoamed, only the disperscd CBA parucles can be the source of ce11 nucleaaon. When cumbling in the mold while bcing hcated, the pellcts begin to sinter with each oti~er.Athough some CBA pamcles may decompose during this period, d the generated gas is cntrapped in the polymer matris and the amount of gas lost is ven smd. Ir should be noted that the ceii nucleauon rate is detcrrnined by the decomposiaon rate and dispersion of CBAs. Since the cerl distribution dcpends encrely on the position of CBA particles, the effect of the mising quality is more important. The striation thickncss of die polymer and CBAs, i.e., the average distance benvcen CBA particles in the polper matris, is much smaller in the compounding based process than the average distance benveen the entrapped air bubbles in the dry blending Chapter 5 Rotational Molding of PE and PP Foams with the Melt Compounding Method bascd foarning process. As a result, the nuclei densir). is higher, and therefore, a finer and more uniform structure is espected from the compounding based processing. However, if the nucleated ceiis coalesce during the ceil growth stage in the rotational foam molding, the final celi density of the foams dlbe lower. Therefore, the ce11 morphology of the rotomolded foams obtained in the compounding based foam processing may be governed Ly the degree of celt codescence. When the nudeated ceiis grow, the volume of each pellet increases and this volume cspansion of pellets enhances the sintering behavior of the pellets in rotomolding. Although the normal plastic peilets without any blotving agent do not sinter weli even for the high MFR materials according to Our experirnents, the sintering behavior of the foarnable pellets is outstanding because of the soft foam structure and espanding behavior of the pellets. Cnlike the espanded PS or PP bead processing [38], no welding lines were obscned benrrcen the espanded pellets. Figure 5-4 shows the cornparison of the results from the dry blending method and thosc from the compounding method for three different materials of hlT4390, SD812, and PF633. For MT4390 (Figure 5-4 (a)), no improvement was f~und.Because the sintering ability of LIT4390 is relatively good in the dry blending method, there was not much differencc bcnveen the foam structures obtained in the dry blending technolog. and compounding tcchnology. Since it was cspected that the nuclei density would be higher for the compounding based technology, cell coalescence during foarning must have occurred to a certain degree. Furthcr study is required to understand the ceii coalescence behavior during rotational foam molding in the compounding based method. Chapter 5 Rotational Molding of PE and PP Foams with the Melt Compounding Meîhod In addition, because of the reduced thermal conductivity when pellets are used for roto molding, the processing urne is increased. For esarnple, in the experiments of threc-fold espansion of PF633 (with 1.4% 50 AZ/50 ZnO), ir took 17 minutes to develop a f~illy espanded structure wich the dry blending based method, but it took 21 minutes when compounded peiiets were used. As n rrsdt, the compounded pellet sample as a whole rcnded to be overheated and the cell structure was poor as shown in Figure 5-5. The average cell six \vas 531 Pm in dry blcnding based foiuns and 539 Pm in compounding based foams. Thc cell population density was 261 84/crn3 and 261 76/cm3, respectively. Figure 5-4 (b) presents the foam morphologies of the high melt suengdi (HbïS) PP. SD812 (Monteil), with a MFR of 16. SD 812 has long chain branching in the polyrner structure and has a high melt strength according to the manufacturer of the resin. Bccausc of its high melt suength, the degree of ceil coalescence with SD812 was cspected to be low during foam processing. Furthcrmorr, because of the higher MFR of this materid, the viscosin- was lower and the sintering behavior of the plasac powders \vas adequate for producing a fine ceIl suucture with the dry blending based mechod. \%en the compounding rnethod was used, a finer ce11 structure was obtained (see Figure 5.4 @)). This implies chat the bubble nucleation behavior with the compounding method for SD812 \vas betrer than chat with the dry blending based method, as espected. After the compounding process, the nuclcation was governed only by the decomposition bchavior of the chemical blowing agent, and polymer viscosity had relatively littic influence on nucleaaon. The improved xrixing quality of the chemical blowing agent partides in the polymer maair also played an important role. The advantage of the compounding merhod over the dry blcnding method was evident using this HMS PP material. Chapter 5 Rotational Molding of PE and PP Foams with the Melt Compounding Method The advantage of the compounding method was more distinct when HAIS PP PF633 with a very low MFR of 3-6 g/10 min was tested. From Figure 5-4 (c), it can be seen that a fully espanded foarn structure was not obtained in the drj blending case using PFG33. Because of the high viscosin. of the matenai, the sintering process was too slow to fom a molten polymcr matris and to isolate the CBA particles before the decomposition of the blowing agents. As a result, most of the gases escaped through the connccted air channcls chat wcre formed due to poor sintering. Because of the smd amount of gas left, foams with LreqTlow expansion were obtained. On the other hand, afier compounding, the chemicai blowing agent parucles were well isolated in the polymcr rnauix, and very iiîde gas \vas lost. Consequently, a uniform fine-ce11 foam structure was achieved. The esperiments with PF633 indicate that the appropriate range of hlFR of PP materials for rotational foam molding is rnuch wider in compounding-based processing. The hypothesis thac the melt strength of Pi? materials dominates the foam qualiry in the compounding based roraaonai foarn molding process was further venfied by the csperimental results of sis-fold cspansion. Figures 5-6 gives the cornparison benveen sis- fold expansion foams of SD812 and PF633. With the complete removd of pre- dccomposiaon, the problems reiated to poor sintering and the resultant nucleation behavior in the dry blending method have been eliminated in the compounding method. The properties of the materiai govern the foarning: a higher meIt strength resulted in a smaller ceIl six and improved ce11 distribution by preventing ce11 coalescence. Chapter 5 Rotationai Molding of PE and PP Foams with the Melt Compounding Methal 5.6.2 Foaming with Pre-decomposed Pellets During the compounding process, some extrudates were observed to havc siight pre- decornposition of CBAs. Foaming expetiments were conducted with the peiiets that had micropores induced by the premamre decomposition of CBAs. When peiiets with micropores generated by pre-decomposition were roto foamed, thc qualiry of PE foarns \ras improved dramatically compared to the rcsults from dry blending based processes. The average ceil six uTas reduced drarnaacaiiy and the ceil clistribution was much betrcr. On the other hand, no good foarn structure could be obtained from the compounded PP peiiets with micropores. Only large hoiiow structures were produced. A hot stage opacal microscope andysis (see Figure 5-7) was used to esplain the different result from PE foarning and PP foarning. In Figure 5-7 (a), it can be secn chat a Ixgc nurnber of tiny bubblcs esist in the PE marris. They arc vecy finc and uniformly discribured. On the connary, rhc bubbles from pre-decomposed CBA in PP pellets are fav in number and are poorly distributed as shown in Figure 5-7 @). To further understand how the bubbles in the compounded pellets affect the fooaming process, more hot stage optical microscope obsenwions were conducted on PE pcllcts. A sarnple of LLDPE 8556 with 3% OT was placed on the hot stage and heared from 100°C to 200°C with a heating rate of 10 "C/min. A picture of the morpholog). of the sample was taken whenever a significant change was observed. The stage temperatures were also recorded. Figure 5-8 shows the bubble formation procedure at each temperature mcasured from the optical microscope. From the figure, it can be seen that when the tcmperature was below 157 "C, no decomposition occurred and only purc ce11 growth could be found. Above 157 "C, decomposition started and the ceiis began CO move because of Chaptcr 5 Rotational Molding of PE and PP Foams with the Melt Compounding Method volume expansion. At 162"C, the decomposiâon reached its peak and the movement was so large that it was difficult to focus. Aftcr 168 "C, the celis growth dominared the process again. lwen then temperamre reached 180 "C the cdcoalescence was so severe that the ceii six and population density decreased dramaâcaily. From these observations, we found that the bubbles formed from pre-decomposition worked as nudei in the compounding based rotational foaming method. Those bubbles gencrated fiom the Iater decomposition in rotomolding difhsed into these esisting cells and bçcame the source of growth of these ceiis. For the sample of LLDPE8556 with Celogen OT, the population density of bubbles from pre-decornposition is \.es high and resuln in a foam structure of high ceii population dcnsity. For PP pellets of MT4390 with Celogen AZ 3990, the estremely large ceiis and low population densicy from pre-decomposition govern the nucleation and lead to unacceptable strucmres. 5.6.3 Effect of the Oven Temperature on CeU Morphology \men foams of a larger expansion ratio such as six-fold expansion arc to be devcloped, the ce11 wall is thinner because there is less polymer materiai and more gas to tiIl the cavity compared to the three-fold espanded foam processing. The thin wd makes the cc11 coalescence phenornenon more sensitive to the processing conditions, especiaily the temperature fzctors. It was found that the oven temperature is one of the most important parameters that determine the celi morphoiogy of the foarns. Although nucleation of ceils cm be improved significanciy using the compounding technique, the ce11 structure of PP foarns may not be good becausc of cell coalescence if the temperature is too high, or the processing time is too long. Especiaiiy whcn an esothcrrnic Chapter 5 Robtiod Molding of PE and PP Fo- with the Melt Compoundibg Method blowing agent such as Celogen AZ 3990 was used, a high heating rate or high oven temperature \vas observed to be detrimental to the foarn structure in Our previous study of the dr)- blending based technology. 4s in the case of the dry blcnding mcthod, we found thar a high oven temperamrc induced a high heating rate and the decomposition heat under such a high heaung race increased the temperature of the neighboring area and promoted ce11 coalcscencc in the compounding mcthod. As a result, large ceiis and a low cell populaaon dcnsit). urere developed under a high oven temperature. Figure 5-9 shows the effect of vqing the oven temperature on the six-fold espandcd Pl? foarns. Chapter 6 Formation of an Unfoamed Skin Chapter 6 Formation of an Unfoamed Skin 6.1 Introduction As mentioned in Chapter 1, a dud layer scnicture with a foamed core and an unfoamed skin cm be developed in rotational molding- Some curent techniques use a txo-step procedure. After the skin is formed, the foam materials are charged into the caviry and roto foarned again. Uniroyal Chernical developed a method using a drop box structure (Figure 1- 6) [27. IVith such a method, a foamable composition consisting of a blowing agent and polymer powders is pur into a box in the center of the mold. The bos opens at a certain urne and the foamable composition drops into the mold and develops a foam structure. This nvo-shot technique has two drawbacks, however. For devices such as a drop box, special considerations have to be given ro the mold design, which may be difficult for a cornplicared structure with thin wds. hloreover, the low production rare makes hem inefficient for industriai applications. In the research that this thesis is based on, for the first tirne an unfoamed skn Inyer wns produced in rotationai molding of Pl? foams with one shot. The concept of this new dual layer manufacturing technology was venfied through a senes of esperirnents. The cffect of major processing parameters was investigated in the research, and the feasibility of the tcchnology has been proven. However, the research is stiii in its preiiminary stages, and n systematic study is required to ideno'$ the effects of materials and processing parameters on the Çomed skin and foam stnic~s.It should be noted that by forrning a skin, the Chapter 6 Formation of an UnfdSkin optimum processing parameters in the compounding based rotationai foarn molding tcchnology wvill be different from those in the processing with no skin. In previous chapters, the research efforts were focused on the producing of fine-cell foams to be filled in the core of rotornolded producrs. The investigation on skin forninaon will be described in detail in Chapter 6. 6.2 Overall Process Ir is wdrecognized that during normal rotaaonal molding, phdes of different sizes tend to form layers [If. Because of the tumbling effect and the vibration when the mold rotates, finer particles wiii be filtered down to the mold surface and coarser marerials wiil risc CO the top. As a result, if partides of two different sizes are used, two layers of material wvill bc formed before the melt stans. In this one-shot method, materiais used for the skin should bc in a fine powder fom so that they cm move to the mold surface area. The material for the foam core should be in a big pellet or granulate form in order to delay the melting and foaming. The smder particles that are closer to the mold surface will be heated first and mclt first. Ideaily, they wdi form a unifom layer of polyrner melt before the pellets begin to melt. Through this method, a skin-core structure can be obtained in one shot. It is espected that the more dramatic the difference benveen the sintering ability of the skin and core matcrials, the more distinct the layer ui1! be. 6.3 Experimentation 6.3.1 Experimental Setup The rotauonal molding machine shown in Figurc 4-3 was used in the esperiments on skin formation. Thc pellets used were gcnerated from the compounding system shown in Figure 5-1. Chapter 6 Formation of an Unfoamed Skin 6.3.2 Materials Two materials with a high and Iow MFR were used to produce the foam core: SDS12 (AIFR 16) and PF633 (iMFR 3-6). For the skin, an esu-usion coating grade PP material of SC1335 @fFR 18) was chosen. In order to investigate the effect of material viscosity of the skin maceriai, SC873 (MFR50), a material with a very low viscosicy, was aiso tested for the purpose of skin formation. 6.3.3 Procedures Comnounding In addition to the temperature scttings and rotationai speed of the screws, thc feeding speed of the peiietizer and the diameter of the die exit should aiso be controiied, because it is believed that the size of the pellets plays a role in determining the skin quality. The feeding speed should be lirnited so that the diarneter of the esuudate is not too thin due to the estcnsional force applied on the polgmer mclt. If necessan7, the diarneter of the die should bt: eniarged. Roto Molding First, a specified amount of powder and pellets were charged into the mold scparately. Thc amount of materials added was determined by the designated expansion ratio and skin thickness. For the cylindrical mold used in the esperiment, thc amount of skin material, m,,,, was determined by Chapter 6 Formation of on Untoamed Skin -- msbn - nLp~'" [d2 (d - th )2], 4 skin where L is the length of the mold, h is the thickness of the skin, d is the diameter of the mold, and pSLnis the density of the skin material. For the pellets used for the foaming purpose, the arnount was determined by foiIowing equation: - "' pellet - rn piyrner where n~,,,~,= weighr of pellets needed, g - mp~Ta- weight of pure polyrner material used, g m,,,;, = weight of blowïng agent, g V = total volume of the mold, cm' O= expansion ratio, and cp = gas yield of bloMng agent cm3/g Since the rotauonai molding machine in Figure 4-3 can only rotate dong one zuis, the mold was shaken dong the longitudinal direction in order to achicve a uniforrn distribution of material. The mold was then mounted on the shaft driven by the motor and turned. The motor was run for 10 minutes at 20 RPiM to separate the powder and pellets before thcy wcrc heated. After 10 minutes, the rotational speed was reduced to the normal rotational molding Ievel of 5 RPM for these expenments, and the rotaung mold was inserted in the pre-hcated oven. The mold was left to rotate in the oven unal the designated time lapsed. Thc mold is then pded out and cooled in che running tap water. Once the mold was coolcd, thc sample was taken out and cut The cross section of the resdang foam was invesagated. Chapter 6 Formation of an Unfoamed Skin 6.4 Results and Discussion During the esperiments, several factors were investigated, incIuding the oven tcmperature, the pellet six, and the relative amount of skin material to core material. In Table 6-1, a detailed description of the esperimental conditions and results is given. 6-41 Effect of the Relative Amount of Powder and Pellets When too much material was added to the mold, no obvious skin was fomed (see Figure 6-1, Tables 6-7 to 6-4). The more material in the mold, the more difficdt it was for thc powvders to travel to the surface, in curn making it more difficult for the skin to form. The skin should be thick enough so thar the molten skin layer has adequate strength and pellets cm be restricted inside this layer and not peneuate to the surface (see Figure 6-1 (b)). 6.4.2 Effect of Oven Temperature Based on the analysis described in Chapters 4 and 5, the higher the oven temperature, the more distincr the effect of thermal lag when larger particles such as pellets are used. As a rcsult, in the skin formation process, the urne difference benveen chc melting of the skin material and the foaming of the pellets is increased. The effect was helpful in achieving a more distinct skin layer as shown in Figure 6-2. In addition, we espect that the oven tcmperamre should Le higher than the optimum oven temperanise used in the pure compounding based rotofoaming process wvithout any skin Iayer. This is due to the fact chat the fomed skin layer di retard the heat transfer. Howcver, it should also be noted that the oven temperature should be Limited. Otherwise, the esotherrnic effect of the blouing agent wiil cause severe ceU coaiescence (see Sections 4.5 and 5.5) Chapter 6 Formation of an Unfoamed Skin 6.4.3 Effect of Material Viscosity In Figure 6-3, the effect of material viscosity on the slün structure is shown. \.en SC873 (MFR 50), a material with an esuemely Iow viscosity was used, no distinct skin was formed. From the esperimencal results in Figure 6-3 (b), we can see that the skin was formed initially during the process. Howcver, due to the low viscosity of SC873, the pellets pcnctrated into the skin layer after foaminç and espansion staned. In the end, the forrned skin layer was destroyed. Whcn SC1355 (MFR18) was used, the relatively high viscosity made the skin stronger and a clear skin was maintained (Figure 6-3(a)). Another reason for having a higher viscosicy material in the skin is the mechanical properties; a higher viscosit). material eshibits bener mechanical properties and it wodd be desirable to have suonger phj.sicd properties of the skin in the foam products. However, use of a high viscous powder material in the skin might cause poor sintering of powdcrs which may affect che mechanical properties and aesthetic aspect thc final product. Thcrefore, the viscosity of the skin material should be carefully selected wïthin a specific range. A further study is required to determine the proper viscosity range for the skin layer. 6.4.4 Effect of Peiiet Size In Figure 6-4, it can be clearly seen chat the skin is more distinct for the sample made from largcr-sized pellets. The reason for this difference lies in the fact that when the pellet six is increased, the thermal conducavity and sintering behavior of the peilets becorne worse, leading to a delay in the foarn formation. The clelay dows a sufficicnt tirne for the plastic powders to Çom a uniform layer. But when the processing cime is too long, then the skin temperature is too high and it becomcs soft enough to be penetrated by foarned pellets. Chaptcr 6 Formation of an Unfoamed Skin Therefore, the pellet size also plays a role in determining the skin quality. However, the pcllet six has to be kept within a specific range. Otherwise, it wiii be difficult for the pellets to rcach corners or aps of some complicated structures and a void section d be developed. A furthcr study is required to determine the proper pellet size range for the skin layer. 6.5 Concluding Remarks Thc feasibilit). of forming of an un foamed integral skin layet for the roto-molded PP foam products has been successfully demonstrated. The effects of the skin matcrial pararneters and processing parameters on the skin layer quaiir). and foam structure wxc invesugated. Several criacal pararneters that affect the skin layer qualicy have been idencified. It was observed that the optimum ranges of thc processing puameters in the compounding based technology have been chmged because of the fomed skin layer wïth a high rhennal ineïtia. Although the skin laver formation was well demonstrated, the achieved fonm structures were not satisfacton. A systematic further study is required to investigate the effects of processing pararneters on the foam structure with the skin layer. Chapter 7 Conclusians and Future Work Recomrnendations Chapter 7 Conclusions and Recornrnendations for Future Work 7.1 Conclusions on Dry Blending Based Technology During the research upon which this thesis is based, fine ce11 foarns with an expansion ratio up to six-foid were successfuily developed for both PE and PP using a dry blending bascd rotational foarn rnolding process. Based on a fundamental investigation of the dry blending based foarning proccss in rotational molding, the foiiowing conclusions can be drawn: 1. fintering Mechani~m.The nucleation mechanisrn is determined by the sintering behavior of the polymer powder particles. If the enuapped air pockets cmbe completely removed in a weli-sintered polymer ma&, the nudei di be the decomposed chernical blouïng agenr particles. Thereforc, the ce11 distribution and population density is detemined onlp by the distribution of the CBA particles. Howcver, it was esperimentaiiy verified that most marerials could not show such a perfect sintering behavior. \Vhm entrapped air pockets are formed during sintering, these cavities wiU dorniiiatc the nudeation and ce11 growth. If the formed cavities are separated from each other by good sintering of polymcr powders before the onset of CBA decomposition, rhen thc isolated cavities become ce11 nuclei and the gas generated from the decomposed CBA \vil promote growth of these cavities. 1f the formed cavities are not separated from each other because of inadequate sintering before the onset of CBA decomposition, the connected cavities wili form large cells and/or di form channels for gas to escape, resulûng in Chaptct 7 Conclusions and Future Wock Rccornmendations poor espansion. Therefore, the polymer material should be weil sintered before the decomposition of the CBA. 2. Mutenkf ~e/ection. Material selection is critical to the foam quality espeuaiiy when foarns of a large expansion ratio are to be developed. (a) Due to the importance of the sintering ability of the polymcr material, materiais with a large zero-shcar viscosity or a high MFR are desirable. Ho\vcver, to maintain high melt strength, the MFR should not be too high. Otherwise, ceU coalescence and ceii coarsening will occur. @) The partide size of the polyrner powden should be as smaii as possible. With smder parricles, better sintenng can be obtained and fewer air pockers wiii devclop. With reduced nurnber of air pockets, ceil nucleation is more controllable and the cell structure can be improved. (c) The shape of the plastic powder particles is aiso an important clement to Lie considered when dealing with raw plastic materiais. Long tail structures should bc reduced to improve the sintenng ability. Good powder shapes rcduce problems that arc related to sintering. CBA Sdection. Many other factors have to be considcred when selecting CBAs and some conclusions have been drawn about the decomposition characteristics of CBAs. Some guidelines on the selection of CBAs have aiso been made: (a) The decomposition behavior of the chernical blowing agent is essentid to the foam qualiry. Since an early decomposition d lead ro poor sintenng and late dccomposiaon wiil cause a weak polymer melt strength at high temperature, it is desirable that the CBA decompose as soon as the sintering is finished. Addiavcs such as Zn0 can successfully control the decomposition temperature. Therefore a Chaptcr 7 Conclusions and Future Work Recommendations proper amount of such additives shodd be detemiined according to the rheological properties of the pmicular material used in the process. (b) The CBA materid should have a quick decomposiuon rate so that nucleaaon and cdgrowth can start uniformiy tvithin the polyrner matrix (c) An endotherrnic CBA is preferred if other decomposition characteristics can satisS the processing requirements. Due to the Iocal heating effect caused by the heat gcnerated t'rom the decomposition, exothermic CBAs tend to decrease the melt strength. (d) The CBA particle size does not play a large role in determining the cc11 morphology in dry blending based technology. Since the sintcring ability governs the nucleauon and cell growth, the effect of CBA particle size is negligible when the dry blending method is applied. 4. Ouen temperatnre. Thc oven temperature is another important parameter that affects the finai foam structure. A high oven temperature will resuit in a high heating rate, which will accelarate the decomposition of esothermic CBA. If the decomposiuon occurs before the sintering of powders, poor foam cells will be generatcd. On the other hand, a low oven temperature will cause a "balling phenomenon" that \vil1 reduce the uniformir). of the material and ceii disuibuaon. Therefore, a proper oven temperature is esscnaal to the foaming proccss. 7.2 Conclusions on Compounding Based Technology Very fine cd foams were successfulIy produced using compounding based rotational foam molding. The following concIusions cmbe made: Chaptcr 7 Conclusions and Future Work Recomrncndations 1. Nztcfeahon Mechi~m.Cell nucleation in the compounding based foarning technology is dominated by the dismbuaon and decomposition bchavior of the chernical blowing agents. Compared to the dry blending based technology, the zero-shear viscosity of polymer matenais has a rclatively lower effect on ce11 nucleation when the compoundicg technology is used. Therefore, to get a large ce11 population densicy and uniforrnly distributed cells, the polyrner and CBAs should be weii nilued initiaily. 2. P/axiric matenal reiect2on. The melt suength of the PP matends is another important factor in the cornpounding based foaming process. The degree of celi codescence wiil determine die final cddensity. Therefore, materiais of high melt strength are dcsirable if a fine ceii structure is expected. In addition, a foam strucnire with large volume espansion can be achieved when high melt suength materials arc used. Since most PP and PE materials can be foamed through melc compounding bascd technology, the application of rotational molding is estcnded to a larger range of materials. 3. CBAz. The endothermic and esothermic characteristics of the CBAs affect thc compounding quaiity. An esothcrmic CBA wili lead to an unscable mising system temperature and may cause unconuoUable prc-decomposition in the compounding process. The high heating rate involved in the estnisjon process makes this heat more severe. On the other hand, the decomposition temperature should not be too high in order to maintain a high melt strength of polyrner. Therefore, the properties of CBA materials are important in determining the ceii quality. 4. Conlpozmding ParPrneters. Careful attention should be paid to the processing parameters of the compounding process. The processing temperature and the Chaptet 7 Conclusions and Future Wotk Rccommcndations estrusion speed (flow rate of polymer) are key factors of the process. The temperature should be above the melting temperature and below the decomposition temperature. Within this range, the temperature should be as low as possible. The rotaaonai speed of the extrusion screw should be high cnough to improve che miving quaiity and reduce the residence time. It shouId also be as low as possible to prevent the decomposition caused by the local heat generated by viscous heating. 5. PR-decon'poriton. Preliminary conclusions have been made on the effect of pre- dccomposition in the compounding process. Ic is believed chat air bubbles gcncrated from pre-decomposition act as nucIei. Therefore, the distribution and ceii populaaon density of the bubbles generated from prc-decomposition determine the final foarn quality. In a high pressure and high shear environrnenr, it is easier to achieve a high population densiry if che melt strength of the polymer is adequaie, as sho\vn by the dramaac improvement in cell structure in Our esperiments. However, nudri bubbles shouid not be generatcd by paraai decomposition of the dunng compounding. There is no way to control how much CBA should be decomposed in the compounding process and hou? much should be left for the roto molding process. When a small percentage of CBA decomposes in compounding, ver?. few bubbles wiiI be generated, and thereforc, the population density of nuclei tôr the roto foaming will be very low. As a rcsult, the cell population density of the foam will be low. On the other hand, if a large percentage of CBA decomposes in compounding, there will not bc enough amounts CBA left for the gas needed for ceil growth in the roto molding process. Therefore, pre-decomposition should be avoided. Chapter 7 Conclusioas and Future Work Recommcndations 6. Rotakonaf Mdding Parameters. Processing cime, oven temperature, and the arnount of CBA used are as important as in dry blending based processes. Proper selcction of these factors is critical to the foam structure. 7.3 Conciusions on Skin Formation 1. i\da~endSefertion. nie skin materid should havc an appropriate vkcosity so chat the skin materiai is strong cnough to prevenc the toarned pellets from penetrating the skin layer and destroying the skin. On the other hand, the viscosity of skin materiai should be low enough to reduce the size of the enrrapped air pockets in the skin layer. 2. Amount afmatenh! The amount of marerial used should be determined based on the desired skin thickness. But if too much material is charged into the mold it wiii be difficuit for the skin material, which is in a powder fom, to travel to the surface to forrn the skin layer. As a result no distinct skin layer \vil be formed. 3. Pe/M Si~e.The pellets used for the core should not be roo small. The particle size difference between the powder for the skin and the pellets for the core plays an important role in differenaaang the thermal conducavity and the sintering urne. However, it is more difficult for larger pellets to travel to tips and corners, resulting in voids or large ceiis at these places. Therefore, the pellet size should be within a certain range. 4. Oven Tenp-atrm A proper ovcn temperature should be chosen to improvr the distinction benveen the skin and core according to the size and shape of the mold used (For esample, 350 "C was the optimum for the mold in this research). But it Chapter 7 Conclusions and Future Work Recommcndations should be noted that the oven temperature variations should remain in a reasonable range so that foarn quality can be rnaintained. 7.4 Suggestions for Future Work Based on the results achieved during the present research project and considering the problems that have been encountered, the following suggesaons are made for future work: (1) Although prelimliary conclusions have been reached on the effect of pre- decomposition, further investigation using a hot stage microscope is necessary. (2) Prc-decornposiaon cm be introduced into the compounding process to gcneratc bubblc nuclei before the rotational molding. \men tiny cells with an esuemely large ceil population density produced in the compounded pellets act as cdnuclei during foaming in roto-molding, a very fine ceii structure can be espected according to the conclusions prcsented in Chapter 5. Two potenciai methods cm be applied: (a) Two CBAs with different decomposition temperatures can be used. The one with a Iower decomposition temperature can be used to produce a rnicroporc structure in compounding extrusion. The CBA with a higher decomposition temperature can be used to induce ceii growth in the later rotational foam rnolding process. (b) h physical blowing agent such as CO2 can be used as a substitute for the first blowing agent in the compounding stage. (3) A systcmatic smdy on the mechanism of skin formation should be carricd out to identifi the effect of material and processing parameters on the formed skn and foamed structure. References 1. R. J. Crawford, Rotational Moldinp of Plastics, Second Ediaon, Research Studies Press Ltd., 1996 2. G. L. Beall, Rotational ll.io/din& Todoy und T'mw, ANTEC '97, Volume III, 1997 3. .P. J. ~Mooney,Rotan'onalMo!ding: Ar the Take-OfStage of Grolvrh, Rotation, Volume IV, Issue 3, 1995. 4. hl. J . Wright, A. G. Spence and R J. Crawford, An Anabni of'heating eflne~xcy in Rofanotral illddhg, ANTEC '97, Volume III, 31843191 5. C. Vasile, and R. B. Seymour, Handbookof Polvolefins, Marcel Dekker, Inc. 1993 6. J. Karger-Kocsis, Polv~ro~vleneStructure. Blends and Com~osites,Volume 1 Structure and MorphoIogy, Chapman & Hall, 1995 7. E. P. Moore, Jr., Polypropylene Handbook, Hmser Pubiishers, Munich, 1997 8. J. A. Brydson, Plastic Materials, Butterworth-Heinemann, Fifth Edition, 1989 9. Ser Van Der Van, Polvmo~vlene and Other Polvolefins-PoIvmerization and Characrerizaaon, Elsevier, 1990 10. V. V. Demaio and D. Dong, The Effec. of Chain Jtmctr,te on Me/t Shngth o/P&mp~/eneand Po~eth$ene, ,ANTEC '97, 1512-1517 11. J. Brandrup and E. H. Immergut, Po!ymer handbook, Second Edition, John \Xriley & Sons Inc., 1975 12. C. K. Andrews, and G. klooney, Performance Evaluation of Chemical Poarning Agens by Gas Evoluaon and Rate Analysis., Foam Conference96, 1996 13. D. IUempner and K. C. Frisch, Handbook of Polymeric Foarns and Foarn Technolog;ll, Hanser Publishers, 1991 References 14. C. B. Park, D. F. Baldwin, and N. P. Suh, "Axiorn Design of a Microcellular Filament Esmsion System", ASh.IE Transactions, Journal of iManufactz~ringStienre and Enginernirg floumal of Engtneenngfor Inhsty), Vol. 8, No.3 , 166-177, 1996 15. A. H. Behravesh, C. B. Park, L. K Cheung, R. D. Venter, " Estrusion of Polypropylene Foams with Hydrocerol and Isopentane," Journal of Vin. and Additive Technology, Vol. 2, No. 4, pp. 349-357,1996 16. T hl. Ponuff, "Examinacion of Foam Rotaaonal hioiding Cornpounds" Foam Conference 96, 1996 17. A. H. Landrock, Handbook of Plastic foams. Twes. Pro~erties.'lanufactures and A~dications,Noyes Publications, 1995 1S. George Matthews, Polvmer Mixins Technolo~y,Applied Science Publishcrs, 1982 19. Chris Rauwendaal, LMi,uine Marcel Dekker, Inc., 1991 20. Nicholas P. Chercmisinoff, Polvrner hlisino and Estrusion Technology, Marcel Dekkcr,., 1987 21. R. T. Fe~ner,Extruder Screw Desim, London ILIFFE Books, 1970 22. N. P. Suh, The Princi~lesof Desim, Osford University Press, 1990 23. M. J. Stevens, Extruder Principles and O~erations,Elsevier Applied science Pubiishcrs, 1985 24. C.B. Park and L.K. Cheung, "A Study oc Ce11 Nudeation in the Extrusion of Polypropylene Foams," Poher Engineering adSR'ence, Vol. 37, No. 1,.PP 1-10, 1997 25. N. G. McCrum, C. P. Buckley, and C. B. Bucknall, Princi~lesof Polymer Eneineering, Osford University Press, 1997 26. D.G. Needham, U.S. parent 5,366,675, 1994 Refetences 27. Uniroyal Chemicais' report on "Celogen Biowing Agents, Rotational MoIding of Foamed Plastics", 1996 28 AYV. Adarnson, Pbyn'caf Chemi~hyofsurface~, Fi fth Ed., \?dey Interscience, 1 990. 29 J.H. Saunders, In: Handbook of Polynlen'c Foam and Foam Tecbnology, D, Klempner and K.C. Frisch, eds, Hanser Publishers, 5, 1391 30 E. P. Gy ftopoulos and G.P. Bere tta, Themodyuanzic~: Foundatiom and Appkcatiom, Macmillan, 1991. 31 E. G Fisher, Estrusion of Plasacs, Newnes-Butterworth, London, 1976 32 S. Lam, G. Liu, Interna/ Report, Mcrocellular Plastic Manufacturing Lab, University of Toronto, 1996 33 C. B. Park, D. F. Baidwin and N. P. Suh, "hiornatic Design of a Microccllular Filament Extrusion System", Research in Engzneenlrg Dengn, VoIume 8, PP 166-177, 1936 34 C. B. Park The mie of poIr,n~er/ga.r~olr~tions in contizu~or/sprocessing o/ nzimcei'ldar pohnrerr, Ph.D. Thesis, iMIT, 1993 35 G. Liu, C. B. Park, and J. A. Lefas, "Rotational Molding of Low-Density LLDPE Foams," ASME, IMECE, CAE and Intelhgent Pmcesnhg of Pobmerir Matenafs,H.P, Wang, L.-S. Tumg, and J.-LM.Marcha1 eds., hlD-Vol. 79, ASME, NY,pp. 33-49, 1997 36 C. B. Park, Personal Communicaaon, 1997 37 Sam Duva, WedTech Inc., Personai Comunicaaon, 1997 38 Plastic Encyclopedia 39 C. B. Park, Internai Proposal, 'licrccellular Plastic Manufactunng Lab, University of Toronto 40 K. Cheung, Procesnng of Fine-cefi Po&ropy/ene Foams in Estnuion, M. A. Sc. Thesis, University of Toronto, 1996 Table 1-1 Density of Different Grades of PE Category Density (g/cc) L J Low Density PE (LDPE) 0.91 - 0.925 Linear Low Density PE (LLDPE) 0.926 - 0.94 High Density Copolymer (LLDPE) 0.94 - 0.959 1 High Density PE (HDPE) 0.96 Table 3-1 Polyethylene Materials hlaterials MFR &/l O min.) Supplier LL8S56 6.8 Nova LLS361 5 -2 Nova Table 3-2 Polypropylene Materials 1 Supplier 1 2o 1 Quantum l Table 3-3 Chemical Blowing Agents Used in the Experiments CBAs Onset Temperature CC) Thermal Property Supplier CcIogen AZ 3990 207.07 Esothermic Uniroyal Chernical ADC/LM-CI 204.21 Esothermic Bayers ADC/F 204.37 Esothermic Bayers ~p Sodium Bicarbonate 133.35 Endothermic Exson Celogcn OT 165.80 Esotherrnic Uniroyal Chemical -- H ydrocerol 149.48 Endothermic BI Chemical Table 6-1 Effect of Material Amount Heating Tic Skin Foam T,lwn(.JC) Remarks (hl Fine foam structure, no clear skin Fine foam suucnire, no clear skin Coarsc foam ceJi, distinct skin Fine foam swcnire, no clmskin Table 6-2 Effect of Materiai Viscosity Weating Tie SLn Foam Tm, CC) Remarks (min) Distincr skin formed, no SC1355 (1 5 g)-JLFR18 PF633+39/0 .\Z(15g ) 13 325 decomposition starts Sicm 1s desrroyed No distinct skin obtaincd SC1355 (15 g)-AfFR18 PFG33+3"h .\Z(15g ) 18 325 Distinct skin Table 6-3 Effect of Oven Temperature r Heaeing Time Skin Foam TCxrn(C) Remarks (min) Fine foam suucnirc, no SC1355 (15 g) PF633-+-3% .\Z(15g ) 33 250 ciear skin SC1 355 (15 g) PFG33+3% .\Z(lOg ) 18 335 Clearer skin laver I'cq dis~ctlayer, fuie SC1355 (15 g) PFG33+S0/o .\Z(lOg ) 16 350 foam cciIs Table 6-4 Effect of Pellet Size (T0,,,=325"C) Hcrung Time Skin Foam Pellet Sizc Remarks (min) Xo distinct skin is SC1355 (15 g) PF633+3*'0 ;\Z(15g ) 18 Smd formed I i SC1355 (15 g) PF633 +3% AZ(1Og ) 18 Large CIearer skin laver fonned Heata Rotanon'a (a) Charging (b) Heating Part Removal Rotation ' -& id) Part Removal Figure 1-1 Schematic of Rotational Molding Process Sintered Plastic Mold Particles Loose Phstic Parades Figurel-2 Sintering in Rotational Molding Primary (a) Stmight Atm Figure 1-3 Exarnples of Rotational Molding Machine (Courtesy of Crawford [l]) Figure 1-4 Isomerism for Positions in Polypropylene (a) Fkt Charge (b) Skin Formation (c) Second Charge (d) Final Product Figure 1-5 Producing a Foam Cote with a Dropping Box Unfoarned Skin Unfoamed Skin amed Corc (a) Samples From Traditionai (b) Samplcs From Roto Foaming Process Process Figure 1-6 Example Products: Food Container - tIIm- --U II...I.I C..œ(t.l tumœ*lœ W. Figure 2-1 Schematic of MF1 measurement *' 1 . .I -" .- .I I - .-.- ,TI --l n W.. m hi- @) Exothemiic Effect (a) Endothcrmic Effect Figure 2-2 Effect of Endo thermic and Exothermic Behavior (a) Weight Lose Rate (b) Pressure Incrcase Rate (After Kiempncr and Frisch [13]) Figure 2-3 Characterization of CBAs Figure 2-4 Cylinder Drwn Mixer Figure 2-5 DifTerent Shapes of Dry Blenders Figure 2-6 Blender with Deflectors Figure 2-7 Two-Roii Mill HOPPER Figure 2-8 Basic Structure of single Screw Extruders FLOW - Figure 2-9 Configuration of the Screw Figure 2-10 Blockhead Mixing Head Figure 3-1 Typical Plastic powdet shapes Host Computer C -Testhg Unit Figure 3-2 Rotational Stress Rheometer MFR (dl0 min) (a) MFR vs. Zero-sheat Viscosity (190 OC) 0.001 0.01 o. 1 Shear Rate (s") @) Effect of Temperature on Viscosity (Testhg Sample: SC873 (MFR50) Figure 3-3 Rheologicd Properties of Polypropylenes Figure 3-4 TA2910 Dinerentid Scanning Calorimeter (IUusmtion Courtcsy of TA Instrument) Figure 3-5 Schematic of DSC Module (Iilustmtion Courtesy of TA instrument) File: O:. . .\PUSTIC\L-361 .a1 Operator: F Liu (a) LU3361 S.mple: LL- File. O: ...\PLASTIC\URS6.001 Site: 3.9OûO mg DSC Operator: F Liu M.thod: hr 10 Run Datr: 1-Mar-9B 17.10 Cornent: HRlO (b) US56 Figure 3-6 DSC Themograms of PE Matedals 1O8 Sample: SC1173 Sire: 2.80001g Operator: F Liu Yethod: hr 10 DSC Coonmnt: HRlO (a) SC873 DSC Opcrrtor F Liu (b) PF6ll Figure 3-7 DSC Thermograms of PP Materials Surple: 50242 . - -- -.-. Sire: 3.5000ig Operator: F Ciu Uethod: hr 1C DSC Comment: WlO 1.5, (c) SD242 File: O: ...\PUSTIC\WT4390.001 Opetatar: F Liu DSC RunDate;l-Mar-98 11:29 -0.1, 1 , 120 140 160 110 29 ero Tempera ture (OC) Universal VI. 1 1A TA Xns trwents (d) MT4390 Figure 3-7DSC Thermograrns of PP Materials (cont'd) - DSC ~perator:F Liu (e) SDSî.2 Smple: PF6U Sizm: 7.4000 mg Operator: F Liu Mwthod: hr 10 DSC Comment; Hl10 (f) PF633 Figure 3-7 DSC Analysis of PP Materials (cont'd) Balance Heat Exchanger \ Control Computer Figure 3-8 TA 2050 Thermogravimetric Analyzer Sarpl.: AL k10 File: D:\ATT~DSC\AZWlO.OlS Size: 3.1000ag Operator: F Liu Yethod: hr 10 fast Run Dite: .-Mar4 15: 10 Comment: HRIO Temperature (OC) Universal VI. 11A TA Instruments (a) Cetogcn AZ 3990 Srnple: adcmcl Fi 1.: O .\Af12\DSC\AOOCl .O1 1 Size: 5.0000 mg Operator: F LXU Wethod: hr 10 fast DSC @) ADC/M-cl Figure 3-9 DSC Analysis of CBA Decomposition S~ple:ADC/F Sire: 3.qOfiO mg Mcthod: hr 10 fast Saapfr: CELtXEN UT Fife: 0:.. .\DSC\CBA\QT.Oll Sire: 1O.WOQ ag Operator: F. Liu Mcthod. hr 10 fast DSC Cornent: hrlO 25 7 179.06- 20- h p 75- Y 1 " LL u : 10- * s- - 165.a0-C - O 1 A% 100 150 200 2SO 300 ex0 Teaperatue ('C) Universal Vl. t 1A TA Instruwnts (d) Celogen OT Figure 33- DSC Analysis of CBA Decomposition (cont'd) Sanple: Hycrl C File: O: ...\CnC\WA\HCRLOl.OOl Size; 13.4004 mg Operatw : Ghaus Methad; hr tO fast DSC Cornent: hrTO (e) Hydrocerol Compound Sample: Baking Soda File: O:\ATTZ\OSC\BS.OlI Size: 11 .SOOO mg Opiratw: F Liu Yithad: hr 10 fast DSC 1 (f) Sodium Bicarbonate Figure 3-9 DSC Analysis of CBA Decomposition (cont'd) 1 II II 1. II: il; II: fi: 1-1 : I:i Tmper8turo (OC) Universal V1.11A TA Instruents (a) Celogen AZ 3990 Figure 3-10 Effect of Heating Rate on DSC Therrnogram of CBAs Sample. Pwc Az-test1 File. O....\E#-ZM\PUR€.OO3 Size: 8.6970 mg Operator: F. Liu C R. Pop-Ilrev Yothod: HeatratelO TGA Ccaamnt; Pure At. Heatrng R 10 Tew~eraturo(OC) Universal VI. 11A TA Instmaents (a) Celogen AZ 3990 Smplo: ADC/F. purm. test1 File. O.\ATr\M(XAI.011 Si28. 12.013Q mg 008ritor. F. LIU Methad: HtatratelO TGA Camerit. MC/F. put. hearrat8 10. test1 ,- i O 206 .30°C 100- 1 M- a u -6 .O Y. 1 I x-. d - Y 5 60- l.7 % -4 2 - 1 s I I "w 1 40- I L - -2 8 20 - -O - 0 -( r -2 O 50 100 1% zoo zso ma Tmmraturw (OC) Universal VI .11A TA Instruents @) ADC/F Figure 3-11 TGA Anaiysis of CBA Decomposiaon Behaviors Suple: ADC/n-Cl. purm. test1 File: O:\ATT\MLOOCl.Oll Size: 24.2170 rg Operator: F. Lau Wethod: HratratelO TGA Corirent: =/+Cl. pur.. h.atrrt8 10. test1 1 1 29: ! -1 O 50 100 150 200 250 300 Temperature (OC) Unxversal VI. 11A TA Ins trrrwnts srnple: nydrcrl C-test2 Size: 55.w6arg Op~ritor: nwthod: For tœsting Hydrocerol TGA Comment; Hydr-1 Cap. gh2 ln)-, - 2.9 ' lüô.63~ 110- - 2.4 rrq rar 100- - 146.22't - O -œ - 1.9 *, x " Y 6 90- L .# l Ia ze 186.07T -1 4 i: CI 'B 80- i r I I - L - I -0.9 83 70 - i 1- 0.4 60- - 50, , 4.1 sa raa 150 200 ZSQ 300 r>o T.npmratur. (OC) Universil Vl .llA TA Inrtru.rntr (d) Hydrocctol Cornpouad Figure 3-11 TGA Analysis of CBA Decomposition Behaviors (cont'd) Siaple: Celogen QT File: O:\A~FINALDIT\OT.O11 Sizr. io.n«r ig Oprrator: F Lm nethod: nratrat.10 TGA Comment: k10 1rmpera cure (OC) Unrversal Vl. llA TA Xnstrumts (e) Celogen OT Saple: Biirng Sod8 File: 0 :\ATT\W-TCA\BS .O1 1 Size: 17.4810 ig Operator: Fangy r cru Method: hrlO TGA Comment: -10 (0 Sodium Bicarbonate Figure 3-i.2 Effect of Heating Rate on the Decomposition Behavior 120 r - Pure Celogen AZ - 10 phrtnO -30 phr Zn0 50 phr Zn0 - 100 phr Zn0 Temperature CC) Figure 3-W Effect of Zn0 on Decomposition of Celogen A23990 (Heating Rate 10 "C/min) Tampeca turr ( OC) Univerial VI. 11A TA Instrwents Figure 3-14 Effect of CBA Particle Size on Decomposition of ADC Series CBA (Heating Rate: 10 "C/min) Figure 3-15 Effect of CBA Arnount (Sarnple Size) on Decomposition of Celogcn OT (Heating Rate 10 "C/min) Me!ting Front Line Deconposition T = Tm rront Line Dtstonce frac Powder Btends Tdec.2!2~ (poiyner agents) the Mold Figure 4-1 Schematic of Sintering and Fovning Process in Rotational Molding Coalesced Big Cells Figure 4-2 Ceii Coalescence (a) Schematic of Rotomolding Machine (b) Uni-wial Rotomolding Machine Figure 4-3 Uni-axial Rotational Molding Machine Figure 4-4 Grinding (IUusmtion Councsy of Cmwford (11) The Cell Density Figure 4-5 Schematic of Ceii Population Density Calculation [34J (a) LLDPE 8556+5% Celogen OT (b) WPE8556+5% Sodium Bicarbonate Figure 4-6 Polyethylene Foarns of Three-fold Expansion (Scale 151) Figure 4-7 Temperature Profile in Rotational Molding (a) l3 min @) 14 min (c) 15 min Figure 4-8 Cross Sections of Shtered Samples of SC873 (MFR 50) + 3% Celogen AZ 3990 (Six-fold) (a) U min (b) 14 min (c) 15 min Figure 4-9 Cross Sections of Sintered Samples of SD242 (MFRSO) + 3% Celogen AZ 3990 (Six-fold) (a) umin (b) 14min Figure 4-10 Cross Sections of Sintered Samples of MT4390 (MFR 20) + 3% Celogen AZ 3990 (Six-fold) (a) PF633 (MFR 3-6) (b) SD8l.2 (MFR 16) (c) MT4390 (MFR 20) Figure 4-11 Cornparison ofDifiérent Materials (Three-fold) O 10 20 30 40 MFR (gi10 min) (a) CeU Sizc O 10 20 30 40 50 60 O tO M 30 40 5060 MFR (gil0min) MFR QI10 min) (c) Volume Expansion Ratio @) Ceii Population Density Figure 4-12 Effect of Material MFR on the Ceii Size and Cell Population Density and Volume Expansion Ratio Of Three-Fold Expanded PP Foams (a) Pdcte size 1.5 mm (b) Particle Size 0.4mm Figure 4-13 Effect of Polymer Particle Size on CeU morpho log^r (SDS12 + 0.625% A23990 + 0.625% ZnO, Three-Fold, cale 1:l) (a) Ftom patticies witb (b) From particles without Long Tails Long Tails Figure 4-14 Effect of Long tails on CeU Morphology (SD812 + 0.625%AZ3990 + 0.625% ZnO, Three-Fold, Scale 1:l) (a) LL8556+OT (Six-fold) @) LU556 + SB (Sk-fold) Figure 4-15 Effect of CBAs on the Ceii Morphology of PE Foarns Figure 4-16 Effect of CBA Particle Size on the Ceii Morphology (Scale 1:I) of SD812 Foams in the Dry Blending Method llmm (a) 1.0% AZ (b) 1.0% AZ + 1.0% Zn0 Figure 4-17 Effect of Zn0 on the Ceii Morphology of MT4390 Foams 1Imm (a) 0.5% AZ + 0.5% Zn0 (b) 0.625% AZ + 0.625% Zn0 (c) 0.8% AZ + 0.8% Zn0 (d) 1.0% AZ + 1.0% Zn0 (e) 1.1% AZ + 1.1% Zn0 (f) 1.2% AZ + 1.2% Zn0 Figure 4-18 Effect of CBA Amount on Ceil Morphology of Three- Fold Expanded MT4390 Foams (AZ = Celogen AZ3990) O 1 2 3 Percentage of Blowing Agent (wt%) (a) CeU Size j 2'- O 1 2 3 O 1 2 3 Percentage of Blowing Agent (wt %) Percentage of Blowing Agent @) Ccii Population Dcnsity (c) Volume Expansion Ratio Figure 4-19 Effect of CBA Amount on the Ceil Size, Ceii Population Density and Volume Expansion Ratio of MT4390 Foams (a) 15 min @) 1s min 11 mm (c) 20 min (d) 25 miri Figiue 4-20 Effect of Processing The on CeU Morphology of MT4390 Foams (MT4390 + 0.625% A23990 + 0.625% ZnO, Three-Fold) 14 16 18 20 Time (min) (a) Ceii Size 14 16 18 20 22 24 26 14 16 18 20 22 24 26 Time (min) Time (min) (b) Ceii Population Density (c) Volume Expansion Ratio Figure 4-21 Effect of Processing Time on the Ceii Size, Ceii Population Density and VoIume Expansion Ratio (MT4390 + 0.625% A23990 + 0.625% ZnO, The-FoId) (a) 350 "C Figure 4-22 Effect of Oven Temperature on the Ceii Morphology (SD242 + 3% AZ3990, Six-fold, Scale 19) Sanple: at hr2OO File: D:\AlT2\DSC\OTZ00.201 Size: 8.6000 mg Operator: F Liu Method: hr200 DSC Cornent: HR2ûO 210 - 200 - U - 190 L 1 .%!min a 3 1 -56nin u -18s .agDc La Thcorcticd Temperature -180 t l- - 170 1 1 T I I I , 160 1.40 1.45 1.50 1.55 1 .60 1.65 1.70 1.75 Time (min) Universal V1 .11A TA Instruments Figure 4-23 Self Heating Behavior of Exothermic CBAs Figure 4-24 B alling Phenomenon otor - Water Buth I I Control Panel (a) A Schematic of the Melt Compounding System Motor Water (b) Actuai Compounding System Figure 5-1 Melt Compounding System Radiatln & Convection \ / conduc tion Mechanical r-7 A Screw Cooling :rial Transfer Figure 5-2 Schematic of Heat Exchange in Extrusion (Courtesy of Stevens 1231) 92 ! 1 O 5 10 15 20 25 Tir* (sin) Universal Y1 .IlA TA Instrur~n~s Figure 5-3 Effect of Residence Tirne on Decomposition of Celogen A23990 Dry Blending Method Compounding Method (a) MT4390 + 0.6% A23990 + 0.6% Zn0 Dry Blending Method Compoundiag Method (b) SDSl2 + 0.6 % A23990 + 0.6% Zn0 Dry Blending Method Compounding Method Figure 5-4 Cornparison of the Cell Morphologies Obtained frorn the Dry Blending and Compounding Methods (Three-fold Expansion) - - +Dry Blending Method +Meit Compounding Method MFR @Il0 min) (a) CeU Population Density ------+Dry Blending Method +Meit Compounding Method -- - - .. 1 e+3 -. ----- 2 4 6 8 10 12 14 16 18 20 22 MFR (gil Omin) MFR (1 0 gi min) (b) Expansion Ratio (c) Ceii Sue Figure 5-5 Cornparison of Dry Blending and Compounding (Three-fo1d Expansion) Dry Blending Method Compounding Mechod (a) SD812 + 2.5% A23990 Dry Blending Method Compounding Method (b) PF633 + 2.5% AZ395)0 Figure 5-6 Cornparison of the CeU Morphologies Obtained from the Dry Blending and Compounding Methods. The Foams Were to be Expanded Six-fold. (Scale 1:l) (a) MT43W (PP) + 1% AZ3990 (b) LLDPE 8556 + 3% Celogen OT Figure 5-7 Cornparison of Compounded Pelîets of PP and PE (Size of Sample: 2.0 mm x 1.5 mm) (a) 100 OC (b) 125 OC (c) 129 OC (d) 155 OC (0162 OC Ce11 Coatescence Figure 5-8 Foaming Process of Pre-decomposed Pellets (LLDPE 8556 + 3% Celogen OT,Size of Sample: 2.0 x 1.5 mm) (a) 350 oc (c) 300 OC (d) 275 OC Figure 5-9 Effect of Oven Temperature on Ceiî Morphology (PF633 + 2.5% AZ3990, Six-fold expansion, Scale: 1:l) Skin Skin (a) Slon: 10 g SC3355 (b) Skin: 15 g SCUS5 Foam: 15g PF633 + 3% A23990 Foam: log PF633 + 3% A23990 Figure 6-1 Effect of Material Amount (Scale 1:l) No Distinct Skin 3- (b) 320 OC (c) 350 OC Figure 6-2 Effect of Oven Temperature (Scale 1:l) (Skin: 15g SC1355, Foam: log PF633 + AZ3990) 7" Penctrated Foam Structure (a) Skin: SC1355 (MFR16) (b) Skin SC873 (MFR50) (c) Skiu SC873 (MFRSO) (SbFonncd Initialiy) (Skin Dtstroyed Latet) Figure 6-3 Effect of Materid Viscosity (Scale 1:l) (a) Smlll pellet (14 min) @) Big pellet (17 min) Figure 6-4 Effect of Pellet Size (Scale 1:l) Oven Temperature 325 OC (Sicin: 15g SC1355, Foam: PF633 + 3% AZ3990)