Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Polymer grafted carbon nanotube reinforced ultra high molecular weight polyethylene fibre Ma, Yue 2014

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
24-ubc_2014_spring_ma_yue.pdf [ 7.79MB ]
Metadata
JSON: 24-1.0165846.json
JSON-LD: 24-1.0165846-ld.json
RDF/XML (Pretty): 24-1.0165846-rdf.xml
RDF/JSON: 24-1.0165846-rdf.json
Turtle: 24-1.0165846-turtle.txt
N-Triples: 24-1.0165846-rdf-ntriples.txt
Original Record: 24-1.0165846-source.json
Full Text
24-1.0165846-fulltext.txt
Citation
24-1.0165846.ris

Full Text

POLYMER GRAFTED CARBON NANOTUBE REINFORCED ULTRA HIGHMOLECULARWEIGHT POLYETHYLENE FIBREbyYue MaMSC, Donghua University, 2007BSC, Donghua University, 2004A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FORTHE DEGREE OFDOCTOR OF PHILOSOPHYinThe Faculty of Graduate and Postdoctoral Studies(Materials Engineering)THE UNIVERSITY OF BRITISH COLUMBIA(Vancouver)January, 2014? Yue Ma, 2014iiAbstractIn this research, a series of experiments have been conducted to develop a highperformance ultra-high molecular weight polyethylene (UHMWPE) fibre with improvedmechanical properties. A novel process was developed whereby polyethylene grafted multi-walled carbon nanotubes (PE-g-MWCNTs) were used to reinforce UHMWPE fibre. PE-g-MWCNT/UHMWPE fibres with remarkable modulus and tensile strength of 125.5 GPa and 4.0GPa, respectively, were successfully fabricated and showed definite potential for reducing theweight of body armour.A systematic study was carried out to investigate the effects of gel spinning conditions ontensile properties and morphologies of UHMWPE fibre. Spinning parameters, including polymerconcentration, spinning temperature and winding-up speed, were selected and studiedsystematically and the spinning condition of UHMWPE fibre was optimized by design ofexperiment.Intensive experiments were conducted to investigate the feasibility of reinforcingUHMWPE fibre with pristine multi-walled carbon nanotubes (MWCNTs). Various mechanicalmethods include ultra-sonication, ball milling, microfluidizing, etc. were applied for dispersingpristine MWCNTs. Studies on tensile properties and morphologies of formed MWCNT/UHMWPE fibre demonstrated that pristine MWCNTs tend to exist in micro-meter sizeagglomerations and no improvement in tensile properties of the MWCNT/UHMWPE fibres wasfound.Finally, chemical functionalization of MWCNTs using a coupling agent and polymergrafting technology was studied. The effective modulus and strength of MWCNTs werecalculated based on the ?rule of mixture?. Compared to coupling agent functionalization, polymeriiigrafting has been found to be more effective in improving reinforcement of MWCNTs inUHMWPE fibre due to a stronger load transfer on the interface. The reinforcement mechanismof polymer grafted MWCNTs was analyzed based on experimental observations.ivPrefaceThis thesis presents the results of several studies on the preparation of strong UHMWPEfibre reinforced by MWCNTs, which were performed as part of the research project?Lightweight Composite Armour for IED Protection? funded by CRTI (chemical, biological,radiological and nuclear, CBRN, Research and Technology Initiative). The project wasconducted through a collaboration of researchers from the Canada National Research Council,McGill University and Allen Vanguard Incorporation with the objective of developing a strong,stiff fibre that could be used to reduce the weight of armour for IED protection.Chapter 8 reveals results of a study on grafting polyethylene on MWCNT to improve theload transfer between MWCNT and UHMWPE. That study is a refined extension of a methoddeveloped by Dr. Jingwen Guan from the National Research Council of Canada. I extended theconcept by developing a simpler and more effective approach for grafting polyethylene andapplied it as reinforcement for gel spun UHMWPE fibre. A version of Chapter 8 has beenpublished in the Proceeding of the American Society for Composites 27th Technical Conference,15th US-Japan Conference on Composite Material. I organized and conducted all the testing andwrote the manuscript.vTable of ContentsAbstract............................................................................................................................................iiPreface............................................................................................................................................ ivTable of Contents.............................................................................................................................vList of Tables................................................................................................................................... xList of Figures.................................................................................................................................xiList of Abbreviations..................................................................................................................... xvAcknowledgments....................................................................................................................... xviiDedication.....................................................................................................................................xixChapter 1- Introduction....................................................................................................................11.1 Background............................................................................................................................11.2 Fibre with high ballistic performance....................................................................................11.3 Unavailability of stronger and tougher fibres........................................................................41.4 Development of stronger fibre...............................................................................................5Chapter 2- Scope of this study.........................................................................................................72.1 Scope of the research............................................................................................................. 82.2 Organization of the dissertation.............................................................................................82.3 Significance of the study..................................................................................................... 10Chapter 3- Literature review..........................................................................................................113.1 General review.....................................................................................................................113.2 Theory of high strength fibre...............................................................................................123.3 Spinning of high performance fibres................................................................................... 133.3.1 Liquid crystalline spinning of rigid rod polymer..........................................................14vi3.3.2 Super drawing of polymer with flexible molecular chain............................................ 163.4 Techniques to produce UHMWPE fibre..............................................................................193.4.1 Solid state extrusion......................................................................................................193.4.2 Surface growth..............................................................................................................193.4.3 Gel spinning/drawing....................................................................................................213.5 Molecular model of UHMWPE fibre.................................................................................. 233.6 Polymeric composite fibre reinforced by carbon nanotube.................................................253.6.1 Carbon nanotube...........................................................................................................253.6.2 Carbon nanotube fibre and yarn................................................................................... 273.6.3 Carbon nanotube reinforced polymeric composite fibre.............................................. 303.6.4 Carbon nanotube reinforced UHMWPE fibre.............................................................. 313.7 Carbon nanotube reinforcing mechanism............................................................................343.8 Functionalization of carbon nanotubes................................................................................353.8.1 Non-covalent functionalization of CNTs......................................................................363.8.2 Covalent functionalization of CNTs.............................................................................37Chapter 4- Materials and methods.................................................................................................394.1 Materials.............................................................................................................................. 394.2 Experiment and characterization......................................................................................... 414.2.1 Gel spinning..................................................................................................................414.2.2 X-ray photoelectron spectroscopy (XPS)..................................................................... 444.2.3 Transmission electron microscopy (TEM)................................................................... 444.2.4 Raman spectroscopy.................................................................................................... 444.2.5 Fourier transform infra-red (FTIR)...............................................................................45vii4.2.6 Thermogravimetric analysis (TGA)............................................................................. 454.2.7 Tensile test.................................................................................................................... 46Chapter 5- Optimization of gel spun UHMWPE...........................................................................485.1 Introduction..........................................................................................................................485.2 Effect of spinning conditions on morphology and property of UHMWPE fibre................ 485.2.1 Polymer concentration.................................................................................................. 495.2.2 Spinning temperature....................................................................................................495.2.3 Winding speed.............................................................................................................. 505.2.4 Orthogonal experiment design......................................................................................515.2.5 Statistical analysis of orthogonal experimental design results..................................... 535.3 Results and discussion......................................................................................................... 535.3.1 Optimization of spinning process................................................................................. 555.3.2 Optimization of extraction process...............................................................................615.3.3 Optimization of post-drawing process..........................................................................635.4 Validation of optimization of gel spinning process.............................................................695.5 Summary..............................................................................................................................70Chapter 6- MWCNT reinforced UHMWPE fibre......................................................................... 716.1 Introduction..........................................................................................................................716.1.1 Importance of CNT dispersion in UHMWPE fibre......................................................726.1.2 Methods to disperse CNTs in solvent...........................................................................736.2 Dispersing by ultra-sonication.............................................................................................736.2.1 Materials and experiment............................................................................................. 746.2.2 Results and discussion.................................................................................................. 74viii6.3 Dispersing by ball milling and sonication........................................................................... 786.3.1 Materials and experiment............................................................................................. 806.3.2 Results and discussion.................................................................................................. 816.4 Conclusion........................................................................................................................... 83Chapter 7- Coupling agent functionalized MWCNT reinforced UHMWPE fibre........................857.1 Introduction..........................................................................................................................857.2 Experimental........................................................................................................................867.2.1 Functionalization of MWCNTs.................................................................................... 867.2.2 Gel spinning and post-drawing.....................................................................................877.2.3 Characterization of MWCNT....................................................................................... 877.3 Results and discussion......................................................................................................... 877.3.1 XPS............................................................................................................................... 877.3.2 Raman spectrum........................................................................................................... 907.3.3 TGA.............................................................................................................................. 917.3.4 Dispersion of MWCNTs...............................................................................................927.3.5 Tensile properties..........................................................................................................937.4 Conclusion........................................................................................................................... 95Chapter 8- Polymer grafted MWCNT reinforced UHMWPE fibre.............................................. 968.1 Introduction..........................................................................................................................968.2. Experimental.......................................................................................................................978.2.1 Functionalization and polymer grafting of MWCNTs................................................. 978.2.2 Gel spun PE-g-MWCNT reinforced UHMWPE fibre..................................................988.2.3 Characterization............................................................................................................99ix8.3 Results and discussion......................................................................................................... 998.3.1 TEM and SEM..............................................................................................................998.3.2 FTIR............................................................................................................................1018.3.3 Tensile properties........................................................................................................1038.3.4 Morphology and microstructure................................................................................. 1128.4 Reinforcing mechanism of polymer grafted CNTs........................................................... 1128.5 Conclusion......................................................................................................................... 114Chapter 9- Summary and suggestion for future work................................................................. 1169.1 Major achievements of the present thesis..........................................................................1169.1.1 Effect of spinning conditions on fibre tensile properties and morphology................ 1179.1.2 Feasibility of reinforcing UHMWPE fibre by pristine nanotubes..............................1189.1.3 Reinforced UHMWPE fibre by coupling agent functionalized CNTs....................... 1199.1.4 Polyethylene grafted CNTs reinforced gel spun UHMWPE......................................1199.2 Suggestion for future work................................................................................................ 1209.2.1 Characterization of CNT alignment in gel spun UHMWPE fibre..............................1209.2.2 Quantitatively study load transfer between CNT and UHMWPE..............................1219.2.3 Gel spun CNT reinforced UHMWPE fibre using commercial production line......... 121References....................................................................................................................................124Appendices.................................................................................................................................. 145Appendix A Statistical analysis of orthogonal design of experiment......................................145Appendix B Statistical analysis of tensile test data................................................................. 148Appendix C Dispersing MWCNTs through mechanical methods.......................................... 152xList of TablesTable 1. 1 Tensile properties of Kevlar? and Spectra? fibre [4, 5].................................................. 3Table 3. 1 Theoretical and achieved tensile properties of polymeric fibres [35].......................... 14Table 3. 2 Mechanical properties of carbon nanotubes, measured using SPM............................. 26Table 4. 1 Characteristics of polymers......................................................................................... 40Table 4. 2 Characteristics of solvents........................................................................................... 40Table 4. 3 Characteristics of other materials................................................................................ 40Table 4. 4 Post-drawing condition UHMWPE fibre.................................................................... 43Table 5. 1 Factors and levels used in experimental design............................................................51Table 5. 2 L9_3_4 Orthogonal experiment design........................................................................ 52Table 5. 3 Tensile properties of UHMWPE fibre spun at different conditions.............................56Table 5. 4 UHMWPE fibre post-draw condition...........................................................................64Table 5. 5 Tensile strength of UHMWPE fibre spun at optimal conditions..................................70Table 7. 1 Elemental compositions of the COOH-MWCNT and Ti-MWCNT............................ 88Table 7. 2 Analysis of C1s peaks of the COOH-MWCNT and Ti- MWCNT............................... 90Table 7. 3 Intensity of Raman peak of MWCNT and Ti-MWCNT.............................................. 91Table 8. 1 Effective modulus and strength of PE-g-MWCNT................................................... 110Table 8. 2 Effective modulus and strength of CNTs from literature [188]................................ 110Table 8. 3 Effective modulus and strength of Ti-MWCNT........................................................111Table 8. 4 Effective modulus of PE-g-MWCNT as a function of volume fraction....................111xiList of FiguresFigure 1. 1 Expected ballistic performance of fibre with U*(1/3) of 862 m/s....................................4Figure 2. 1 Flow chart sequentially showing the various tasks conducted in this study................. 9Figure 3. 1 Specific strength and modulus of high performance fibres [35].................................12Figure 3. 2 Molecular structure of different types of fibre [35].................................................... 13Figure 3. 3 Liquid crystalline spinning process [35]..................................................................... 15Figure 3. 4 Molecular structure of p-phenylene terephtalamides (PPTA) [49].............................16Figure 3. 5 Polymer chain alignment during spinning [35]...........................................................16Figure 3. 6 Molecular model of polymer fibre [35].......................................................................17Figure 3. 7 Coutte flow apparatus for producing an oriented UHMWPE fibre [65].....................20Figure 3. 8 Shish kebab structure of UHMWPE fibre [67]........................................................... 21Figure 3. 9 Gel spinning of UHMWPE......................................................................................... 22Figure 3. 10 As-spun UHMWPE fibre spun from various UHMWPE concentrations [7]........... 23Figure 3. 11 Fibrilar model of UHMWPE fibre [74].....................................................................24Figure 3. 12 Microfibrilla model of fibrous structure with intrafibrillar [75]............................... 25Figure 3. 13 Computational image of single-walled and multi-walled CNT [79].........................26Figure 3. 14 Schematics of the experimental setup for coagulation spinning [95]....................... 27Figure 3. 15 Carbon nanotube yarn formed from solid state spinning [89].................................. 28Figure 3. 16 Liquid crystalline spinning of CNT fibre [102]........................................................ 29Figure 3. 17 Tensile stress of CNT/UHMWPE strips prepared by ball milling [113].................. 31Figure 3. 18 Wear rate of CNF/UHMWPE composite [114]........................................................ 32Figure 3. 19 Crazing and rupture of a CNT/polymer composite fibre [122].................................34Figure 3. 20 SWCNT/PAN fibres after stretching [123]............................................................... 34xiiFigure 3. 21 PVP-wrapped CNT [124]..........................................................................................35Figure 3. 22 SDS absorbed on the surface of CNT [136]..............................................................36Figure 3. 23 Grafting MMA from MWCNTs via living radical polymerization.......................... 37Figure 4. 1 Preparation of UHMWPE gel..................................................................................... 42Figure 4. 2 Swelled MWCNT/UHMWPE gel...............................................................................42Figure 4. 3 Gel spinning set-up.....................................................................................................42Figure 4. 4 Post-drawing process..................................................................................................43Figure 4. 5 Preparation of single fibre tensile test sample............................................................46Figure 4. 6 Typical stress-strain curve of UHMWPE fibre.......................................................... 47Figure 5. 1 Viscosity of fibre at various spinning temperatures...................................................50Figure 5. 2 3-factor, 3-level orthogonal experimental design...................................................... 52Figure 5. 3 Variation of tensile strength with UHMWPE fibre diameter.....................................54Figure 5. 4 Variation of tensile strength with glass fibre diameter [153].....................................54Figure 5. 5 Morphology transformation during post-drawing......................................................58Figure 5. 6 Morphology of UHMWPE fibre after post-drawing..................................................58Figure 5. 7 Morphology of UHMWPE fibre dran at various winding speeds...............................60Figure 5. 8 Oxidization of UHMWPE gel under atmosphere....................................................... 61Figure 5. 9 Maximum post-draw ratio of UHMWPE fibre with and without mineral oil.............63Figure 5. 10 Effect mineral oil on UHMWPE fibre tensile properties..........................................63Figure 5. 11 Effect of temperature on maximum draw ratio......................................................... 65Figure 5. 12 Morphology of UHMWPE fibre drawn at room temperature...................................66Figure 5. 13 Effect of post-drawing temperature on UHMWPE fibre tensile properties..............68Figure 5. 14 Morphology of UHMWPE fibre spun at optimum condition................................... 69xiiiFigure 6. 1 Morphology of as-produced CNTs [160, 161]............................................................72Figure 6. 2 Tensile properties of MWCNT/UHMWPE fibre prepared by ultra-sonication..........75Figure 6. 3 Morphology of MWCNT/UHMWPE fibre prepared by ultra-sonication...................76Figure 6. 4 MWCNTs protruding from UHMWPE fibre surface................................................ 77Figure 6. 5 CNT morphology during ball milling process [178]...................................................79Figure 6. 6 CNTmorphology after ball milling [179]....................................................................80Figure 6. 7 Retsch ball mill machine............................................................................................. 81Figure 6. 8 Tensile properties of MWCNT/UHMWPE fibre prepared by ball milling................ 82Figure 6. 9 Morphology of MWCNT/UHMWPE fibre prepared by ball mill process................. 83Figure 7. 1 Reaction between coupling agent and COOH-MWCNTs [26]..................................86Figure 7. 2 XPS spectrum of an MWCNT and a coupling agent treated MWCNT..................... 88Figure 7. 3 Deconvoluted high resolution C1s spectra of COOH-MWCNT and Ti-MWCNT...89Figure 7. 4 Raman spectra of (a) COOH-MWCNT (b) Ti-MWCNT..........................................90Figure 7. 5 TGA of COOH-MWCNT and Ti-MWCNT.............................................................. 92Figure 7. 6 CNT suspension in mineral oil...................................................................................93Figure 7. 7 Tensile properties of MWCNT/UHMWPE fibre with various CNT contents...........94Figure 8. 1 Polyethylene grafted maleic anhydride functionalized MWCNT..............................98Figure 8. 2 TEM of COOH-MWCNT and PE-g-MWCNT........................................................100Figure 8. 3 PE-g-MWCNT with thicker grafting layer.............................................................. 100Figure 8. 4 FTIR spectrum of COOH-MWNT, NH2-MWCNT and PE-g-MWCNT.................102Figure 8. 5 Typical stress-strain curve of 3% PE-g-MWCNT/UHMWPE fibre.........................104Figure 8. 6 Effect of adding HDPE on draw ratio of UHMWPE fibre [186].............................105Figure 8. 7 Tensile properties of gel spun UHMWPE and MWCNT/UHMWPE fibre............. 106xivFigure 8. 8 PE grafted MWCNTs bridge the crack of UHMWPE fibre.....................................109Figure 8. 9 Cross section image of PE-g-MWCNT/UHMWPE fibre........................................ 112Figure 8. 10 Interface between PE-g-MWCNT and UHMWPE................................................ 113xvList of AbbreviationsThe following table describes the significance of various abbreviations used throughoutthe thesis. The page on which each one is defined or first used is also given.Abbreviation Meaning PageCNF Carbon nanofibre 32CNT Carbon nanotube 1CRTI Canada Research and Technology Initiative xviiiDecalin Decahydronaphthalene 96EOD Explosive ordnance disposal 1FTIR Fourier transform infared spectroscopy 45IED Improvised explosive device 1MMA Methyl methacrylate 37MWCNT Multi-walled carbon nanotube iiNaDDBS Sodium dodecylbenzenesulfonate 36nm Nano mater 23PAN Polyacrylonitrile 5PE-g-MWCNT Polyethylene grafted multi-walled carbon nanotube iiPLA Polylactic acid 38PMMA Polymethyl methacrylate 37PPD P-phenylene diamineLL 15PPTA p-phenylene terephtalamides 15PVA Polyvinyl alcohol 27xviPVP Polyvinylpyrrolidone 36SDS Sodium dodecylsulphate 37SWCNT single walled carbon nanotube 6TCL Terephthaloyl chloride 16TEM Transmission electron microscopy 40TPU Thermoplastic polyurethane 30UHMWPE Ultra high molecular weight polyethylene iixviiAcknowledgmentsI would like to take this opportunity to thank all those who have helped and inspired meduring my doctoral studies.First and foremost, I would like to express my sincere gratitude toward my supervisorProf. Frank Ko for his enthusiastic supervision and guidance during my studies at the Universityof British Columbia. My interest in textile composite material came as a result of reading thebook Textile Structural Composites written by Prof. Frank Ko while I was an undergraduatestudent at Donghua University in China. I sincerely thank Dr. Ko for providing such a wonderfulresearch opportunity studying nano-composite material at a world-class university. Dr. Koprovided insightful discussions about the research. In addition, he set an example for me as aresearcher through his passion about research.The thesis is based on a research project in collaboration with the National ResearchCouncil (NRC) of Canada. The experience working with researchers from the NRC broadenedmy perspective on the practical aspects in industry. I would like to thank all researchers fromNRC for their technical discussions. In particular, I would like to thank Dr. Jingwen Guan fromfor his advice and unsurpassed knowledge of nano-materials.I would like to thank members of my thesis advisory committee, Prof. Rizhi Wang, Prof.Tom Troczynski and Prof. Reza Vaziri, for their valuable advice on the thesis research and effortsin reviewing the thesis. All the students and researchers of the Advanced Fibrous Material Groupare thanked for help with experimental setup and general advice. In particular I would like toacknowledge the help of Yingjie Li for her generous help. I am also very grateful to Dr. YuqinxviiiWan, Dr. Heejae Yang, Dr. Cagri Ayranci and Dr. Ardeshir Behi for their scientific advice andknowledge and many insightful discussions and suggestions.The generous support from the International Tuition Award of the University of BritishColumbia is greatly appreciated. In addition, I would like to acknowledge the CRTI (chemical,biological, radiological and nuclear, CBRN, Research and Technology Initiative) as my researchwas supported by them, in part, under Grant CRTI-07-121RD.Finally, my deepest gratitude goes to my parents and husband for their love and supportthrough the past five years. I am forever indebted to them for always being supportive. Thisdissertation would have simply been impossible without their understanding, endless patienceand encouragement.xixDedicationTo my parentsChapter-1 Introduction1Chapter 1- Introduction1.1 BackgroundThe increase of terrorist events involving Improvised Explosive Devices (IED) is asignificant threat to all military and law enforcement. Explosive Ordnance Disposal(EOD) bomb disposal suit is worn as the first line of defense once an IED is identified. Currentfull body coverage EOD ensembles provide acceptable protection. However, the weight of theseEOD ensembles is in excess of 30 kilograms which results in increased fatigue, heat stress andreduced mobility[1]. Thus, both military and civil defence organizations are looking for strongerfibre to further reduce the weight and improve the protection body armour provides.Motivated by the need for lightweight IED protective armour, a research program(Lightweight composite armour for IED protection: A carbon nanotube solution) was funded bythe Government of Canada through the CRTI (Chemical, Biological, Radiological-Nuclear, andExplosives Research & Technology Initiative) with the objective of reducing the weight of bodyarmour used in IED protection by 25%. The feasibility of reducing armour weight by theintegration of carbon nanotubes (CNTs) to either new material or currently used materials forIED application was explored.1.2 Fibre with high ballistic performanceThe performance of textile body armour is evaluated by V50, which is the velocity atwhich impacting projectiles are expected to defeat a system 50% of the time. Soft textile ballisticarmour dissipates the impact energy by wave propagating along the fibres. The relationshipbetween material properties and armour performance has been studied by many researchers inChapter-1 Introduction2the past thirty years. Several analytical models and semi-analytical models have been developedthat could effectively simulate the penetration process and predict ballistic performance. In the1990s, Cunniff [2] proposed a set of simple and effective parameters to provide critical guidanceto fibre developers in order to optimize textile-based body armour systems as shown in Equation1.1. Dimensional analysis indicates that the performance index of interest is U*, which is theproduct of fibre toughness and sonic wave velocity as shown in Equation 1.2. The U* allows foran estimation of the performance of an armour system based solely on the quasi-static fibremechanical properties [3].? ?501 3A AV 0Ud ppm?? ?? ?? ?? ?? ??Equation 1.1Where:U* - Fibre ballistic performance indexAd- System areal densityAp- Projectile presented areamp - Projectile mass???? EU 2* ? Equation 1. 2Where:? ?Fibre tensile strengthE- Fibre Young?s modulus?- Fibre tensile strain?- Fibre densityChapter-1 Introduction3As shown in Equation 1. 2, U* is directly proportional to fibre tensile strength, strain,and Young?s modulus, and inversely proportional to fibre density. Thus, fibres with highmodulus, strength and strain are preferred for producing armour with better performance. Thetwo most widely used high performance fibres that possess high enough tensile properties toprotect a solider from ballistic impact are para-aramid fibres such as Kevlar? and UHMWPEfibres such as Spectra?. Typical tensile properties and the ballistic performance index (U*)(1/3) ofthese two fibres are listed in Table 1. 1.Table 1. 1 Tensile properties of Kevlar? and Spectra? fibre [4, 5]Fibre Strength (?)(GPa)Failure Strain (?)(%)Modulus (E)(GPa) U*(1/3) (m/s)Kevlar?KM2 3.40 3.55 82.6 682Spectra? 1000 2.57 3.57 120 801V50 velocity of armour made of Spectra? 1000 fibre with various areal densities weretested as shown in Figure 1.1 and the V50 velocities of armour with other areal densities areapproximated with a regression fit to the data. From the fitting curve, the dimensionlessparameter 100Ad*Ap/mp needed for armour made of Spectra 1000? to obtain a V50 of 550 m/s,which is a typical value for high performance personal armour can be obtained. To reducearmour weight by 25%, the parameter 100Ad*Ap/mp will need to be 25% lower than the value foramour made of Spectra 1000?. By calculating the ratio between the 100Ad*Ap/mp of the twoarmour system, the U*(1/3) of high performance fibre needed to reach the goal of reduce armourweight by 25% while maintaining V50 of 550 m/s can be calculated according to Cunniff model.Chapter-1 Introduction4Figure 1. 1 Expected ballistic performance of fibre with U*(1/3) of 862 m/s1.3 Unavailability of stronger and tougher fibresThe mechanical properties of high performance fibres increased remarkably from the1940s to the 1980s with the invention of Nylon?, para-aramid fibre such as Kevlar?, andUHMWPE fibre such as Spectra? through continuing improvement in the spinning process.However, it seems that the fibres? properties have reached a plateau after years of development.In the last three decades, no significant improvement in fibre mechanical properties has beenmade [6]. In applications such as body armours, where a high specific Young?s modulus andstrength is critical, there is a desperate need to reduce weight by developing stronger and tougherfibre.Chapter-1 Introduction51.4 Development of stronger fibreAlthough, the properties of these high modulus, high strength fibres are still far behindthe theoretical value, few significant breakthroughs in fibre properties have been made in recentyears. One of the most promising newly developed high performance fibres is M-5, which has asimilar chemical structure to Kevlar? but m[odulus and tensile strength as high as 310 GPa and5.8 GPa, respectively [3]. However, the fibre has been unavailable in a pre-commercializedstatus until now. From previous experience, it took twenty to thirty years for thecommercialization of both Kevlar? fibre and Spectra? fibre. Thus, it is expected that M-5 fibrewill not be available on the market in the foreseeable future.Besides developing new high performance fibre, using strong reinforcements to producecomposite fibre is another approach for obtaining fibre with improved mechanical properties. Inrecent years, carbon nanotubes (CNTs) with excellent mechanical properties have attracted theattention of researchers for reinforcement in polymer composite.Mechanical properties of CNTs have been determined by theoretical calculation andexperimental measurements. Large variations in the reported mechanical properties of CNTshave been found due to different production processes and various testing methods. However, itis generally accepted that CNTs have tensile strength of ~300 GPa, Young's modulus ~1000 GPa,and tensile strain as high as ~30% [7]. Because of their superior mechanical properties, CNTshave been considered as ideal reinforcement fillers and are of great interest in ballistic armourapplication.In the past twenty years, various researches have focused on using CNTs to reinforcepolymer fibres, such as Nylon-6? [9-16], polyethylene [17-27], and polyacrylonitrile(PAN) [28-32]. Significant improvements in the mechanical properties of these polymeric fibres have beenChapter-1 Introduction6observed. For example, by adding 1 wt% single wall carbon nanotube (SWCNT), carbonizedSWCNT/PAN fibers exhibited 64% higher tensile strength and 49% higher tensile modulus thanthe carbonized control PAN fibre [8]. However, it has been found that most of these researchesfocused on commodity fibres, which have tensile strength lower than 1 GPa, which is far lowerthan high performance fibres such as Kevlar? and Spectra?, which have strength of around 3GPa [35].In order to develop fibre with a higher U*(1/3) which will lead to a 25% weight reductionin ballistic armour, it is of great interest to focus on incorporating CNTs into high performancefibres. As one of the most widely used high performance fibres in body armour application, it isthus of great interest to investigate the feasibility of producing CNT reinforced UHMWPE fibrewith improved mechanical properties.Chapter-2 Scope of This Study7Chapter 2- Scope of this studyAlthough a fair amount of research has been carried out on reinforcing polymeric fibrewith CNTs, very few studies focused on using CNTs to reinforce high performance fibres, suchas UHMWPE fibre. The aim and scope of this thesis is to explore the feasibility of producingUHMWPE fibre with improved mechanical properties by utilizing the extraordinary mechanicalproperties of CNTs.Compared to commodity fibres, which consisted of partially aligned molecules,UHMWPE fibre with highly aligned molecules has a much different macro-molecular structure.In order to obtain such a structure, UHMWPE fibre is produced through a unique gel spinningprocess in which a low-concentration polymer solution in gel form is used as spinning dope.Thus, a systematic study on the effect of spinning conditions on fibre properties is necessary.Besides the unique spinning process, one of the obstacles in processing CNT-reinforcedUHMWPE fibre is obtaining a uniform dispersion of CNTs in the polymer matrix. However, thedispersion of CNT can be very challenging due to the high viscosity and non-polarity of itssolvent. A thorough study on improving CNT dispersion in UHMWPE is critical for producingCNT-reinforced UHMWPE fibre with improved mechanical properties.Compatibility between CNT and the polymer matrix is the other important factoraffecting reinforcing efficiency. Owing to their inert nature, the interaction between CNTs andUHMWPE molecules is via Van der Waals forces and a hydrogen bond, which leads to a weakinterfacial adhesion. In order to improve the CNT's reinforcing efficiency, the chemicalfunctionalization of CNT will be investigated.Chapter-2 Scope of This Study82.1 Scope of the researchThe organization of the research is listed below:(1) Determine effect of spinning conditions on gel spun UHMWPE fibre properties andmorphology(2) Study dispersion of pristine CNTs in mineral oil and the feasibility of reinforcing UHMWPEfibre by pristine CNTs(3) Seek improvement in CNT dispersion and reinforcing efficiency in UHMWPE throughfunctionalization by a coupling agent(4) Seek improvement in compatibility between CNTs and UHMWPE through polymer graftingon the surface of CNTs2.2 Organization of the dissertationThis study starts with investigating the effect of the spinning condition on gel spun fibreproperties, and then analyzing the feasibility of using pristine CNTs to reinforce UHMWPEfibres. Two chemical methods?including using a coupling agent and polymer grafting tofunctionalize CNTs?were utilized with the purpose of improving CNT dispersion in UHMWPEand the load transfer on the CNT and UHMWPE interface. Various methods were used tocharacterize the degree of dispersion of CNTs, the functionalization on the surface of CNTs, andthe tensile properties, as well as the morphology of composite fibres. A summary of the elementof this study is shown in Figure 2.1.Chapter-2 Scope of This Study9Figure 2. 1 Flow chart sequentially showing the various tasks conducted in this studyTask 1---Optimize spinning process by studying the influence ofvarious spinning parameters on fibre morphology and propertiesTask 2---Investigate the feasibility of reinforcing UHMWPEfibre using pristine MWCNTs methods by different mechanicalmethodsTask 3---Study the effect of coupling agent functionalization onMWCNT dispersion and the reinforcing effect on UHMWPEfibreTask 4---Improve compatibility and load transfer on theinterface between MWCNT and UHMWPE by polymer grafting .(1) Gel preparation(2) Spinning(3) Extraction(4) Post drawingCharacterization of UHMWPEfibre: tensile test, SEM,TEMFunctionalization of CNTs oruse pristine CNTs directly.Characterization ofCNTs by TEM,FTIR, XPS andTGA, etc.Dispersion of MWCNT inmineral oilChapter-2 Scope of This Study10The thesis is organized into nine chapters. Chapters 1 and 2 provide an introduction of thebackground and aim of the study. The literature on UHMWPE fibre, CNT-reinforced polymericfibres, and especially CNT-reinforced UHMWPE fibres are reviewed in Chapter 3. In Chapter 4,the details of the materials, process, and characterization methods are described. Theinvestigation of the effects of various spinning conditions and optimization of the gel-spun fibreare discussed in Chapter 5. In Chapter 6, the possibility of using pristine CNTs to reinforceUHMWPE fibre is investigated. Several mechanical methods (including ultra-sonication, ballmilling, and micro-fluidization) were applied to assist dispersion of CNTs in UHMWPE. InChapter 7, the CNTs were functionalized with a titanium derived coupling agent and themechanical properties and morphology of formed CNT/UHMWPE composite fibre werecharacterized. In Chapter 8, the preparation of polyethylene grafted CNTs and morphology andtensile properties of the composite fibre are shown. The reinforcing mechanism is revealed andthe reinforcement efficiency is calculated based on the ?rule of mixture?. Finally, Chapter 9draws general conclusions and gives suggestions for future research.2.3 Significance of the studyIn this study, the concept of CNT-reinforced UHMWPE is demonstrated with the goal ofproducing fibre with improved ballitistic performance index U*(1/3). UHMWPE fibre is one ofthe most widely used high performance fibres in body armour where high tensile properties andlow weight is critical. Successful demonstration of the reinforcing concept, the CNT/UHMWPEfibre will lead the mechanical properties of high performance fibre to a new level. TheCNT/UHMWPE fibre will significantly reduce the body armour's weight while maintaining ahigh level of ballistic performance, thus enforcing the development of light weight armour forIED protection as well as a wide range of impact protection products for the civilian market.Chapter-3 Literature Review11Chapter 3- Literature review3.1 General reviewDifferent from the traditional commodity textile fibres, high performance fibres areengineered for specific applications that require exceptional mechanical properties, heatresistance, or chemical resistance [35]. For example, fibres with ultra-high strength, stiffnesssuch as carbon fibre, para-aramid fibre, and ultra high molecular weight polyethylene(UHMWPE) fibre have been developed since the 1970s and widely used in various applicationsincluding transportation, aerospace, protective clothing, marine (ropes and sails), automobile andothers [36-40].Compared to other strong materials, high performance fibres such as carbon fibre, para-aramid fibre and UHMWPE fibre have much higher specific strength and modulus due to theirlow density, as shown in Figure 3. 1. Due to their relatively low density and high specificstrength and modulus, high performance fibres find applications in areas where weight is critical,such as advanced composite material in airplanes and ballistic armours, etc. The usage of highperformance fibres in these areas greatly reduces the weight of these products and improves theirperformance. One good example of reducing product weight using high performance fibre is inthe area of life protection, such as ballistic body armour, vehicle and property protection panels.Chapter-3 Literature Review12Figure 3. 1 Specific strength and modulus of high performance fibres [35]3.2 Theory of high strength fibreThese high performance fibres are made from various polymer materials and spinningmethods. However, all of these high performance fibres feature one unique structure, which is ahighly oriented and largely extended molecular chain structure. Figure 3. 2 shows a comparisonof polymer chain orientation between injected molded polymer material, textile grade fibre,industrial grade fibre and high performance fibre. Compared to injection molded polymermaterials, which consist of randomly distributed molecular chains, polymer chains in fibrestructure usually show a certain level of orientation, especially for high performance fibre inwhich molecules are perfectly aligned. Orientation offers great advantages for increasing thestrength and stiffness because tensile strength and modulus along fibre direction are controlledChapter-3 Literature Review13by the intrinsic strength of the carbon-carbon bond in polymer chains and the cross sectional areaof molecular chains. Thus, the strength and modulus of fibre consisting of polymer moleculesthat are stretched and orientated in the direction of the fibre axis will be much higher thanconventional fibre, which is made of partially aligned polymer chains [41-45].Figure 3. 2 Molecular structure of different types of fibre [35]3.3 Spinning of high performance fibresDuring the development of high performance fibres, questions such as "How strong canpolymer fibres be and how can they be produced?? were often raised. Research has been carriedout to investigate the structure required to obtain high strength and high modulus. The theoreticalstrength and modulus of polymeric fibre is calculated as 20-30 GPa and 200-400 GPa by variousmethods [46-48]. However, such an idealized arrangement of macromolecules cannot beobtained since the achievement of a perfect crystalline structure is impeded by the presence ofchain ends, entanglements, and by the tendency of polymers to undergo folding duringcrystallization. As a result, the mechanical properties of experimental fibres are much lower thantheoretical limits. A comparison of theoretical and achieved properties of various fibres is shownin Table 3. 1. As can be seen, the achieved fibre properties are far lower than theoretical values;in particular, tensile strength is only about 1/10 of the theoretical value.Chapter-3 Literature Review14Table 3. 1 Theoretical and achieved tensile properties of polymeric fibres [35]PolymerStrength (Gpa) Modulus (GPa)Theoretical Commercial Theoretical CommercialUHMWPE 32 3.6 240 116Aramid 30 3.3 183 120PA-6 32 0.9 142 6PES 28 1.1 125 14PP 18 0.6 34 6In order to obtain a polymeric fibre with a highly extended molecular chain and highorientation, various methods have been investigated by scientists, including liquid crystallinespinning of rigid rod polymer and super drawing of flexible linear polymer, etc.3.3.1 Liquid crystalline spinning of rigid rod polymerLiquid crystalline spinning was invented in the early 1970s with the development of anaramid fibre by Dupont? [49]. Aramid fibres are made of polymers with 'rigid rod'-likemolecules. Because the molecules are rigid, as the concentration of polymer increases, asaturation level of these rigid rod-like molecules is attained. Continual increase of polymerconcentration to a certain level will lead to a parallel arrangement of these molecules, which istermed 'nematic'. The nematic liquid crystalline solution, which consists of domains in whichpolymer chains are oriented in the direction of spinning, will be extruded through a spinneret.The shear-flow field existing inside the spinneret will cause the rigid molecules to align. Afterspinning, an external force induced by an elongational flow will be applied on the spin line toChapter-3 Literature Review15further orientate polymer chains in the as-spun fibre. A schematic illustration of the spinningprocess is shown in Figure 3. 3.Figure 3. 3 Liquid crystalline spinning process [35]The polymer used in liquid crystalline spinning is prepared by the reaction between anamine group and a carboxylic acid halide group. Simple AB homopolymers can be obtainedthrough the reaction as shown in Equation 3.1 [49]:nNH2-Ar-COCl ? -(NH-Ar-CO)n- + nHCl Equation 3.1Kevlar? is made of p-phenylene terephtalamides (PPTA), which are the simplest form of theAABB aromatic polymer and are usually synthesised via a low-temperature polycondensationreaction based on p-phenylene diamine (PPD) and terephthaloyl chloride (TCL), with a structureas shown in Figure 3. 4.Chapter-3 Literature Review16Figure 3. 4 Molecular structure of p-phenylene terephtalamides (PPTA) [49]3.3.2 Super drawing of polymer with flexible molecular chainThe high mechanical properties of Kevlar? fibre are attributed to its unique rigid rod-likemolecular composition, which could be aligned during the spinning process. Most polymer fibresare composed of flexible molecules that usually show partial or random orientation, even afterpost-drawing, as shown in Figure 3. 5 [50, 51]. It is of great interest to develop a method thatcould align the flexible polymer chains.Figure 3. 5 Polymer chain alignment during spinning [35]Chapter-3 Literature Review17To improve molecular orientation, fibres are post-drawn after extrusion. The draw ratio isused to determine the degree of stretching during the orientation of a fibre or filament, expressedas the ratio of the cross-sectional area of the undrawn material to that of the drawn material. Inthe conventional melt spinning process, the post-draw ratio is usually only around 5, due to thesevere molecular entanglements. The structure of the polymeric fibre from conventional spinningmethods was found to consist of the complex structure of the coexistence of the crystal regionand the amorphous region, as shown in Figure 3. 6. It has been widely accepted that in order toproduce high strength and high modulus fibre from polymers with flexible molecular chains, amuch higher post-draw ratio is necessary. In order to reach a high post-draw ratio, polymers withlinear molecular structure and fewer branches are favoured.Figure 3. 6 Molecular model of polymeric fibre [35]In the 1980s, DSM? researchers started the super drawing (draw ratio>30) ofpolyethylene fibre, which consists of long flexible molecular chains [52-54]. Its molecular chainis connected by a strong carbon-carbon covalent bond which has been estimated to have atheoretical strength of 20-30 GPa and modulus 200-300 GPa [55-59]. However, the interactionbetween the molecular chains is via the weak van der Waals force. In order to transfer the loadChapter-3 Literature Review18between molecular chains through Van der Waals forces, ultra-long molecular chains with a highoverlap length are required. Thus, researchers focused on spinning polyethylene with an averagemolecular weight of one million or more, which is known as ultra-high molecular weightpolyethylene.It is known that polymer viscosity increases with polymer molecular weight. The Mark?Houwink equations can be used to relate the intrinsic viscosity of polymer to its averagemolecular weight as indicated in Equation 3.2. A critical molecular weight, Mc, is defined as themolecular weight above which chain entanglements occur in a polymer melt. In a polymer melt,when molecular weight is lower than Mc, zero shear melt viscosity, ?0, is roughly proportional tomolecular weight. At molecular weights above Mc, ? 0 increases with molecular weight to thepower of 3.4.Mw < Mc ?0 = K1MMw > Mc ?0 = K2M 3.4Equation 3.2Where:K and ? are empirically determined constantswM - molecular weight of the polymercM - critical molecular weightUHMWPE with molecular weight around 3,000,000 to 6,000,000 shows melt viscosity ashigh as 108 Pa?s and a melt flow rate (MFR) close to 0 when conventional testing methodswere employed. A modified testing method to determine the elongational stress instead of MFRwas developed which tests the stress required to elongate the testing polymer by 600% in anChapter-3 Literature Review19elongation time of 10 minutes. The elongational stress values determined for fibre gradeUHMWPE resin is around 0.7 MPa [60].3.4 Techniques to produce UHMWPE fibreConsiderable research has been conducted to produce high modulus, high strength fibreusing polyethylene with ultra-high molecular weight. Various methods, including solid stateextrusion, surface crystal growth and gel spinning were developed.3.4.1 Solid state extrusionIn sold state extrusion, UHMWPE melt has to be extruded through a spinneret underpressures as high as 0.2- 0.3GPa because of its high melt viscosity [61]. The polymer chains inas-spun fibre will be aligned along the fibre axis during the post-drawing process. Severepolymer chain entanglements due to the extreamly high molecular weight have been observed inthe fibre spun from solid state extrusion process and the fibre's tensile strength is only 1.94 GPa[62-63]. Besides the low fibre tensile properties, the stringent requirements for spinningequipment due to the high pressure needed during the spinning process also limited theapplication of the solid state extrusion process in large scale production.3.4.2 Surface growthThe surface growth method was developed by Zwijneburg and A.J. Pennings in the late1970s in which UHMWPE fibres were prepared by flow-induced crystallization in a Couetteapparatus [64]. In order to orient the molecules, a drum rotates in a cylindrical bath of a polymersolution as shown in Figure 3. 7. Diluted UHMWPE solution with concentration as low as0.5wt% is used to reduce the number of polymer chain ends and thus the chain entanglement.With the rotation of the roller insides, a thin layer of UHMWPE molecules is adsorbed on theChapter-3 Literature Review20surface of the rotor. Crystal seeds are then added into the solution and longitudinal polyethylenecrystals start to grow from the macromolecules in close contact on the surface of the rotor. Sincefibres are pulled out at the same speed as the crystal growth and in the opposite direction of therotation of the roller, the growing crystal and the rotor surface experiences a continuous strainwhich prevents the polymer chains from returning to the coiled conformation.Figure 3. 7 Coutte flow apparatus for producing an oriented UHMWPE fibre [65]The formed UHMWPE fibres were found to have a shish kebab structure, which is atypical morphology for high polymer weight polymers formed under a force that is often via aflow field, as shown in Figure 3. 8. The shish structure, which appears as the shaft part of thepolymer, is crystallized from extended polymer chains. The shish structure acts as a nucleatingtemplate for the coiled chains to form kebab structures. Fibre obtained from this process hastensile strength and modulus as high as 5 GPa and 150 GPa respectively [66]. The majorChapter-3 Literature Review21disadvantage of the surface growth process is that, until now, the production rate of thelongitudinal crystals is far below commercially viable speeds.Figure 3. 8 Shish kebab structure of UHMWPE fibre [67]3.4.3 Gel spinning/drawingIn order to reduce the entanglements of UHMWPE molecular chains and thus obtain ahigher post-draw ratio, a new spinning process named gel spinning was invented by Smith in thelate 1970s [68]. In this new process, diluted UHMWPE solution in gel form is used as spinningdope, instead of the polymer melt used in solid state extrusion. The main steps in the gelspinning process include: polymer solution preparation, spinning of the solution, coagulation,removal of the remaining solvent in the fibre, and super-drawing the formed fibre. After super-drawing at a draw ratio of over 100, fibres with a high level of macromolecular chain orientation,high crystallinity (>90%) and high tensile properties (3.0 GPa in tensile strength, 90 GPa inYoung?s modulus) can be attained [68]. The gel spinning process is shown in Figure 3. 9.Chapter-3 Literature Review22Figure 3. 9 Gel spinning of UHMWPE fibreIn laboratory, UHMWPE fibres with tensile strength and Young?s modulus up to 7.0 GPaand 200 GPa, respectively, have been produced [69]. UHMWPE gel spinning has receivedconsiderable attention since the 1980s. The effect of processing parameters such as polymerconcentrations [70], winding speed [54] and the post-drawing process [53] on the properties andstructure of UHWMPE fibres has been widely studied. Jian [71] studied the effect of temperatureon the viscosity of UHMWPE gel and found that when the temperature reaches 120?C-140?C,there is a tremendous increase in shear viscosities of UHMWPE gel solutions and the shearviscosity values approach the maximum at 140?C. Yeh [72] studied the influence of draw ratiosof gel-spun UHMWPE fibres on resultant morphologies and tensile properties and found that thecrystal structure of UHMWPE fibre changed from the constrained orthorhombic into hexagonalcrystals during post-drawing. The melt temperature of the fibre increased to 148?154?C, which isthe melting of the hexagonal crystals during post-drawing.As-spun UHMWPE fibres contain a high content of solvent which needs to be removedthrough solvent extraction before post-drawing. Xiao [73] found that phase separation ofUHMWPE and mineral oil is most significant in the first hour and reaches equilibrium state afterChapter-3 Literature Review23about 48 hours. The critical extraction time and the optimum bath ratios of extraction agent togel fibres are 2 min and 10 mL/g, respectively. It was found that fibre spun from highconcentration solution has much denser morphological structure, as shown in Figure 3. 10, whichleads to lower maximum draw ratio and tensile properties.Figure 3.10 (a) Figure 3.10 (b)Figure 3. 10 As-spun UHMWPE fibre spun from various polymer concentrations [7](a)8% , (b) 16%The UHMWPE fibre produced by the gel spinning process was first commercialized byDSM? with the trade name Dyneema? in the 1990s. There are currently two main manufacturersfor UHMWPE fibre besides DSM?: Honeywell? in the US with trade name Spectra?, andToyobo? in Japan with trade name Dyneema?.3.5 Molecular model of UHMWPE fibreBased on experimental observations, it has been observed that UHMWPE fibres obtainedfrom high ratio post-drawing process exhibit a microfibrillar structure. UHMWPE fibre is madeof fibrils with diameters around 100 nanometer (nm), which themselves are composed ofChapter-3 Literature Review24microfibrils 10nm in size. Microfibrils consist of alternating crystalline and non-crystalline zones,as shown in Figure 3. 11.Figure 3. 11 Fibrilar model of UHMWPE fibre [74]Peterlin [75] pointed out that fibre tensile properties, such as modulus and tensile strength,are not the consequence of the crystal lattice orientation but of the connection of the crystals by agreat many taut tie molecules that bridge the amorphous layers which alternate with the crystalsin the axial direction of the fibrous material, as shown in Figure 3. 12. Peterlin also argued thatthe number of taut tie molecules increases during the post-drawing process and, as such, the fibremodulus increases. Penning [76] verified Peterlin?s model through experimental results thatshowed that fibre tensile strength is determined by the fraction of load carrying chains in thedisordered domains. The elongation at break and modulus are largely determined by the ratio ofdisordered domain length to crystal block length.Chapter-3 Literature Review25Figure 3. 12 Microfibrilla model of fibrous structure [75](A) interfibrillar (B) tie molecules.3.6 Polymeric composite fibre reinforced by carbon nanotubeIt is well known that stronger material can be obtained by incorporating reinforcementsinto the matrix material. In order to produce stronger fibre, researchers are constantly looking forproper reinforcement materials. Carbon nanotubes (CNT) have been thought of as one of themost promising reinforcement materials for polymer materials due to their unique properties.3.6.1 Carbon nanotubeCNTs were discovered by Ijima [77] in 1991. The structure of CNTs can be thought of asa rolled-up graphene sheet. The strong sp2 in-plane bonds of the CNTs bind the atoms in theplane and results in the high stiffness and strength of CNTs. Based on their molecular structure,theoretical analysis predicts that CNTs should possess extremely high tensile properties. Thereare two main types of CNTs: multi-walled (MWCNTs) and single-walled (SWCNTs), as shownin Figure 3. 13. MWCNTs consist of several layers of graphite shell, which are in a co-axialcylindrical shape, with an adjacent shell separation of approximately 0.36 nm. SWCNTs consistof a single graphite sheet wrapped into a cylindrical tube with a diameter of about 1.2 nm [78].Chapter-3 Literature Review26Figure 3. 13 Computational image of SWCNT and MWCNT [79]Because of their extremely small size, it is difficult to directly measure the mechanicalproperties of CNTs. In recent years, with the help of scanning electron microscopy (SEM),atomic force microscopy (AFM) and the scanning tunneling microscope (STM), researchers havesuccessfully measured the strength, Young?s modulus and other properties of CNTs, as shown in.As is clear from Table 3. 2, CNT's Young?s modulus is in the range of 800-1300 GPa. Strength ismore difficult to measure than modulus, thus there are few published data for CNT strength. Yu[7] used AFM and SEM to measure the strength of MWCNTs and the result is around 63 GPa.Table 3. 2 Mechanical properties of carbon nanotubes, measured using SPMYoung?s modulus (GPa) Nanotube type Deformation method Reference1300?600 MWCNT Lateral bending [80]1000?600 SWCNT Normal bending [81] [82]870?400 MWCNT Normal bending [81]Chapter-3 Literature Review273.6.2 Carbon nanotube fibre and yarnMotivated by the exciting properties of CNTs, there are intensive efforts to convert CNTsdirectly into linear assemblies [83-93]. However, CNTs do not melt and they are not soluble inorganic or aqueous solvents. Spinning of CNTs using traditional spinning methods has beenproved to be challenging. In the past few years, pure CNT fibres have been spun by four generaltypes of spinning processes: coagulation, solid-state, liquid crystalline and gas state.(1) Coagulation spinningIn coagulation spinning, CNTs are homogeneously dispersed in water by sonication withthe help of sodium dodecylsulphate (SDS) as a surfactant. The aqueous dispersion of CNTs isinjected into a poly-vinyl alcohol (PVA) flowing bath using a set-up shown in Figure 3. 14. Thecoagulated fibre contains PVA in the matrix which will pass through a series of washing stagesto remove it. CNT filaments produced from this method show tensile strength over 1 GPa andmodulus 9-15 GPa [94].Figure 3. 14 Schematics of the experimental set-up for coagulation spinning [95]Chapter-3 Literature Review28(2) Solid-state spinningIn the solid state spinning process, CNTs act as the constituent fibre in the yarn, in asimilar way as conventional fibres such as cotton and wool [89, 90, 96, 97]. CNTs are drawncontinuously from CNT forests. The CNT ?trees? in the forest are over 50,000 times longer thantheir diameter. An array of vertically aligned CNTs is shown in Figure 3. 15...Figure 3. 15 Carbon nanotube yarn formed from solid state spinning [89](3) Liquid crystalline spinningSmalley and other researchers [98-101] found out that CNTs have similar properties tothe rigid rod molecules, such as poly (p-phenyleneterephthalamide) (PPTA), which can form aliquid crystalline solution in sulfuric acid above a certain concentration. Thus, a similar processto produce aramid fibres was used to produce CNT filaments (shown in Figure 3. 16). However,the tensile properties of CNT filaments made from this process are far lower than expected. Forexample, in Erison?s work, the tensile strength and modulus of the formed CNT yarn is only0.116 GPa and 120 GPa, respectively [101].Chapter-3 Literature Review29Figure 3. 16 Liquid crystalline spinning of CNT fibre [102](4) Gas state spinningDuring CNT synthesis through chemical vapor deposition, CNTs become entangled in thegas-phase due to van der Waals forces. The entangled CNTs form black smoke and are called?aero gel? which can be continuously drawn out of the reactor and wound as film or twisted toform a fibre. The strength and modulus of the CNT fibres produced by this method are around1 GPa and 50?100 GPa, respectively [103, 104].While these processes are promising, the properties of individual CNTs are far frombeing fully realized in these CNT-based fibres. For example, the highest strength achieved withany of these spinning methods is around 1 GPa, much lower than the strength of individualSWCNTs which have been measured up to 60 GPa [7]. This could be partly attributed to the lackof sufficient load transfer among individual nanotubes. The load transfer among CNTs is throughvan der Waals interaction which is much weaker compared with the sp2 in-plane bonds, thusCNTs tend to slip under tension before reaching the ultimate strength of individual tube. Tofacilitate stress transfer, the concept of CNT co-extrusion/co-spinning with a polymer to formChapter-3 Literature Review30composite fibres was conceived. A series of CNT-reinforced polymer composite fibres havebeen produced.3.6.3 Carbon nanotube reinforced polymeric composite fibreThe extremely small size and exceptional mechanical properties of CNTs make themideal candidates for tailoring mechanical properties of polymer materials. Significantimprovements in mechanical properties of the polymer composite have been observed with theaddition of a small CNT weight percentage, compared to their unreinforced counterparts [28,105-107]. In addition, from a cost standpoint, the CNT-reinforced polymer composite is anattractive way to make commercially viable CNT-based materials.A number of studies have focused on producing CNT/polymer composite fibres by meltspinning. Chen et al. [108] spun MWCNT-reinforced thermoplastic polyurethane (TPU)elastomeric fibre using a twin crew extruder. The resulted MWCNT/TPU composite fibresshowed a significant increase in mechanical properties. The fibre's Young?s modulus increasedfrom 49.6 MPa of the unfilled TPU fibre to 135.1 MPa of fibre with 17.7% CNT. Andrews et al.[109] prepared SWCNT-reinforced carbon fibre by adding purified SWCNTs into pitch fibre andremoving the solvent after sonication. The fibre was melt spun, drawn, and subjected to heattreatment in air and nitrogen. Carbon fibres with 5wt% of SWCNT showed a 90% increase intensile strength and 150% increase in Young?s modulus. Sreekumar et al. [110] producedpolyacrylonitrile (PAN)/SWCNT composite fibres. Fibres with 10wt% SWCNTs showed 43.5%and 105% higher tensile strength and Young?s modulus, respectively, compared to the propertiesof the pure PAN fibres.Chapter-3 Literature Review313.6.4 Carbon nanotube reinforced UHMWPE fibreWith its low density and excellent mechanical properties, UHMWPE fibre has becomeone of the most widely used high performance fibres. Research has been conducted to reinforceUHMWPE fibre with CNTs. The high viscosity of spinning dope makes reinforcing UHMWPEwith CNTs very challenging. In order to obtain a significant reinforcing effect, a good dispersionof CNTs in the matrix is necessary. Zhang [111] developed a new method to improve CNTdispersion in UHMWPE by spraying an aqueous solution of SWCNTs onto fine UHMWPEpowder. These SWCNTs adhered on the surface of the UHMWPE and the CNT-coatedUHMWPE powders were used in compression molding to make CNT/UHMWPE film. Gao [112]further developed Zhang?s method by using alcohol as a solvent to disperse the CNTs. Afteralcohol evaporated, UHMWPE particles covered with a layer of CNTs were obtained.Subsequently, the powders were compression molded into composite board. Test resultsindicated that the MWCNTs/UHMWPE composites formed a 2-dimension conductive networkat a very low percolation of 0.072 vol%. Martinez [113] incorporated 1, 3 and 5 wt.% pristineMWCNTs into UHMWPE using a ball milling process. The tensile tests showed a 38% increasein the Young? s modulus in the reinforced nano-composites and a small decrease in toughness(5%), as shown in Figure 3. 17.Chapter-3 Literature Review32Figure 3. 17 Tensile stress of CNT/UHMWPE strips prepared by ball milling [113]Weston J. Wood [114] developed a paraffin-assisted melt-mixing method for improvingcarbon nanofibre (CNF) dispersion. A CNT/UHMWPE composite prepared by paraffin oilassisted mixing exhibited much higher wear resistance compared to a composite preparedthrough a dry mixing process, as shown in Figure 3. 18. Maksimkin [115] used a planetaryactivator with steel vials filled with steel balls to produce 0.1% and 1% MWCNT/UHMWPE composite. The mixture was then thermo-pressed and post-drawn to form solidsamples. Remarkable 130% and 20% increases in tensile strength and yield strength wereobtained, respectively, for UHMWPE by adding 0.1% MWCNT.Chapter-3 Literature Review33Figure 3. 18 Wear rate of CNF/UHMWPE composite [114]Most of the above studies focused on CNT-reinforced UHMWPE composite film or plate.Very limited research focused on reinforcing gel spun UHMWPE fibre using CNTs. This couldbe because UHMWPE fibre is composed of highly oriented molecules and has very highcrystallinity (>90%); the requirement for CNT dispersion in UHMWPE fibre is thus morestringent. However, the studies by Ruan?s group at Hongkong Poytechnic University and byWang?s group at Donghua University in China were notable. Ruan et al. [24] investigated thefeasibility of reinforcing the gel spun UHMWPE with pristine CNT using decahydronaphthalene(decalin) as a solvent. By adding 5wt% MWCNTs, ultra-strong fibres with tensile strength of 4.2GPa and strain at break of approximately 5% (which corresponds to a 19% increase in tensilestrength and 15% increase in ductility) could be produced. Wang et al. [26] used 1% of purifiedand coupling agent functionalized MWCNTs to reinforce the UHMWPE fibre and found that thestrength and modulus of the composite fibre increased by about 9% and 14%, respectively.Chapter-3 Literature Review343.7 Carbon nanotube reinforcing mechanismAs various polymer fibres [19-21, 112, 114-120] showed significant improvement afterCNT reinforcing, it is of great interest to investigate the reinforcing mechanism of these CNTcomposites.Ye and Ko [121] pointed out that enhancement in fibre tensile properties arises when theoriented CNTs bridge crazes and the subsequent cracks that propagate normal to the applieduniaxial stress, which is schematically represented in Figure 3.19. By partly replacing the crazingfibrils, CNTs strengthen the weakest part of the fibre. On the other hand, randomly orientedCNTs will allow unimpeded crack propagation across regions that are similar to a neat polymermatrix, thereby the reinforcement of randomly oriented CNTs will not be as high as alignedCNTs. Compared to bulk material, CNTs exhibit much better alignment in polymer fibre due tothe post-drawing process. As shown in Figure 3. 20, CNTs align perfectly along fibre axis.Figure 3. 19 Crazing and rupture of a CNT/polymer composite fibre [122]Chapter-3 Literature Review35Figure 3. 20 SWCNT/PAN fibres after stretching [123]Homogeneous CNT dispersion throughout the polymer matrix and adequate interfacialbonding is critical for load transfer in CNT?polymer composites. However, it has been foundthat CNTs are very difficult to properly disperse in organic matrices in their pristine state due totheir inert nature. Great efforts have been made to improve CNT dispersion through bothmechanical methods and chemical functionalizations.3.8 Functionalization of carbon nanotubesMany studies have been carried out to develop methods that improve CNT dispersion.Such methods can be divided into two categories: mechanical methods and functionalizationmethods (which include physical (non-covalent treatment) and chemical (covalent treatment)functionalization). The mechanical dispersion methods include ultrasonication, ball milling, highshear mixing, etc.. While these methods can significantly improve CNT dispersion, they can alsofragment the tubes and decrease their aspect ratio. More often, these mechanical methods areapplied in combination with chemically functionalized CNTs.Chapter-3 Literature Review363.8.1 Non-covalent functionalization of CNTsIn non-covalent functionalization, the modified polymer molecular chains are physicallybound to the surface of the CNTs by wrapping around the CNTs or adsorbed on the surface. Forexample, polyvinylpyrrolidone (PVP) is able to wrap around CNTs (as shown in Figure 3. 21)and wrapped CNTs can be easily dispersed in polar solvents like water. In addition to wrapping apolymer onto the CNT surface, using a surfactant is another widely used method to improveCNT dispersion.Figure 3. 21 PVP-wrapped CNT [124]The molecule of a surfactant usually consists of both polar and apolar groups whichenable them to be absorbed on the interface between immiscible bulk phases, such as oil andwater, air and water, or particles and solution. Surfactants such as sodium dodecylsulfate (SDS),Triton X-100 and sodium dodecylbenzenesulfonate (NaDDBS) have been widely used toimprove CNT dispersion in an aqueous solvent [125-135]. These surfactants form micelles onthe non-polar CNT surface and result in a greater hydrophilic character of the surface (shown inFigure 3. 22).The advantage of non-covalent functionalization is that it keeps the original CNTstructure intact, thus the mechanical properties do not change. However, the forces between thewrapping molecules and the nanotube surface may be weak and the efficiency of the loadtransfer may be low. Therefore, non-covalent functionalization is not suitable for some load-bearing applications.Chapter-3 Literature Review37Figure 3. 22 SDS absorbed on the surface of CNT [136]3.8.2 Covalent functionalization of CNTsIn the covalent functionalization method, polymer chains are either ?grafting to? or?grafting from? the CNT surface. Grafting long polymer chains onto CNTs can improve theirsolubility in solvents and increase miscibility in polymer matrices. Because of the higher energyof the covalent bond, a more favourable load transfer between CNTs and the polymer matrix canbe obtained through this method. There are two types of polymer grafting functionalization:?grafting to? and ?grafting from? [137-146]. ?Grafting to? is based on the reaction between thepre-formed polymer molecules and the functional group (such as amino, hydroxyl, epoxy, or aradical group) on the CNT surface. In the ?grafting from? method, the polymer grafting isrealized by the covalent immobilization of polymer precursors on the CNT surface andsubsequent polymerization in the presence of monomeric species. One good example of the?grafting from? process was developed Baskaran et al. [147] who functionalized MWCNTs with2-Bromo-2-methylpropionyl bromide, as shown in Figure 3. 22. The functionalized MWCNTscould then react with methyl methacrylate (MMA) to form the polymethyl methacrylate (PMMA)Chapter-3 Literature Review38grafted MWCNTs with good solubility in tetrahydrofuran (THF) and chloroform (CHCl3).Similar methods have also been used to graft polystyrene and some other polymers [147].Figure 3. 23 Grafting MMA from MWCNTs via living radical polymerization [147]The ?grafting to? method has also been used by many researchers for CNTfunctionalization. The carboxyl and hydroxyl group can be attached to the CNT tip and surfacethrough a reaction with a strong acid or oxidant. The hydroxyl group on the CNT surface reactswith the amine and alcohol group through an acylation and condensation reaction to form amideor ester bonds through which polymers can be grafted onto CNTs. Because of the diversity ofreactions between the hydroxyl group and the polymers, various polymers have been graftedonto CNTs. Haddon [148] used acylated MWCNTs to react with octadecylamine. Yoon [149]grafted polylactic acid (PLA) on acylated MWCNTs and found that the molecular weight of PLAhas a significant influence on the coverage of PLLA on the MWCNT surface.These methods generate surface modifications that are chemically bonded to CNTs.Compared to non-covalent bonding using polymer wrapping or a surfactant, the load transferthrough these chemically bonded polymers is much higher [150]. Thus, in the applications wherehighly reinforced efficiency is required, CNT modification through polymer grafting is favoured.Chapter-4 Material and Methods39Chapter 4- Materials and methodsThe materials, experimental methods and characterization methods used in this thesis arelisted and described in detail in this chapter.4.1 MaterialsThe solvents, polymers and other materials used in this research are listed in Table 4.1,Table 4.2 and Table 4.3, respectively. All the materials were used as received unless otherwisestated. Ultra-high molecular weight polyethylene (UHMWPE) from Sigma Aldrich with amolecular weight of 3~6?106 was used in all the spinning experiments in the thesis. Mineral oilwas used as a solvent to dissolve UHMWPE for all the gel spinning experiments. Xylene wasused as an extracting solvent for mineral oil from the as-spun UHMWPE fibre in all extractionprocesses. Carboxyl functionalized MWCNTs (COOH-MWCNT) with a length around 1?m anda diameter about 8 nm was from Cheaptubes Inc. Ethanol was used as a solvent to disperseMWCNTs during functionalization with a titanium-derived coupling agent. Ethylene diaminewas used to react with the carboxyl group functionalized MWCNTs to produce aminofunctionalized MWCNTs. PE-g-MA with 0.5wt% maleic anhydride content was used as apolymer to be grafted on the surface of MWCNTs. N,N-Dimethylformamide was used todisperse CNTs to prepare samples for scanning microcopy (SEM) and transmission electronmicroscopy (TEM).Chapter-4 Material and Methods40Table 4. 1 Characteristics of polymersPolymer Source SpecificationUHMWPE Sigma-Aldrich MW=3?106~6?106PE-g-MA Sigma-Aldrich 0.5wt% MA contentEthylenediamine Fluka absolute, ?99.5% (GC)Table 4. 2 Characteristics of solventsSolvent Source SpecificationEthanol Fisher >99.5%Mineral oil (light) Sigma-Aldrich Cat. No.330779Xylene Sigma-Aldrich ACS reagent, ?98.5% xylenesN,N-Dimethylformamide Sigma-Aldrich HPLC, ?99.9% (Sigma-Aldrich)In order to avoid UHMWPE oxidization at elevated temperatures, 2, 6-di-tert-butyl-4-methylphenol (BHT) was used in this study as an anti-oxidant. A type of titanium-derivedcoupling agent was used to functionalize the surface of the CNTs.Table 4. 3 Characteristics of other materialsRaw material Source SpecificationCOOH-MWCNT Cheaptubes.Inc 95wt%, diameter=5-10nm,length~1?m2, 6-di-tert-butyl-4- methylphenol Sigma Aldrich ?99.0%tri(dioetylpyrophosphoryloxy)isopropyl titanateNanjing Shuguang ChemicalPlant of ChinaCAS # 67691-13-8Product number: NDZ201Chapter-4 Material and Methods414.2 Experiment and characterization4.2.1 Gel spinningIn this thesis, MWCNTs were functionalized by a coupling agent and a polymer graftedby PE-g-MA with the purpose of improving dispersion and reinforcing efficiency in theUHMWPE. Detailed functionalization processes are described in Chapter 7 and Chapter 8.(1) CNT dispersionMWCNTs were dispersed in mineral oil by sonication for about 10 hours using theBranson 2510 bath sonicator. UHMWPE was added to the MWCNT/mineral oil dispersions. Themixtures were sonicated for 1 more hour.(2) Gel preparationThe UHMWPE/mineral oil mixture or the MWCNT/UHMWPE/mineral oil mixtureswere heated up in an oil bath from room temperature to 100?C at 5?C/min, and then from 100?Cto 110?C at 1?C/min under constant stirring and nitrogen protection using a set-up as shown inFigure 4. 1. At 110?C, the UHMWPE started swelling in the mineral oil. Magnetic stirring wasstopped when swelling was observed, as shown in Figure 4. 2. The temperature was kept at110?C for 1 hour. After the UHMWPE fully swelled, the temperature of the oil bath was furtherincreased from 110?C to 160?C with magnetic stirring at 5min/min and kept at 160?C for another2 hours to form uniform UHMWPE/mineral oil or MWCNT/ UHMWPE/mineral oil gel. The airtrapped in the gel was removed by vacuuming it for 24 hours.Chapter-4 Material and Methods42Figure 4. 1 Preparation of UHMWPE gel Figure 4. 2 Swelled MWCNT/UHMWPE gel(3) SpinningA laboratory mixing extruder (LME) from Dynisco Co., Ltd. was used for the gelspinning process. The MWCNT/UHMWPE/mineral oil gel was fed into the hopper and the gelpassed through an annular zone formed by a rotating screw inside the extruder where the actualmelting took place. The temperature in the annular zone was kept at 150?C, 180?C and 210?C,according to the experimental design. As rotation continued, the molten gel driven by shear forceflowed to the spinneret, which had a 3 mm diameter. A 10 cm air gap was used between thespinneret and the water bath. The extruded fibres were then solidified in a cold water bath andwound on a take-up roller. The gel spinning set-up is shown in Figure 4. 3.Figure 4. 3 Gel spinning set-upChapter-4 Material and Methods43(4) ExtractionThe as-spun fibre containing a high proportion of mineral oil was extracted using xylene.The as-spun UHMWPE fibres were wound on a metal frame to avoid fibre shrinking afterextraction. The fibres wound on the metal frame were immersed in 50ml xylene and sonicated 3times at a duration of 15, 10 and 5 minutes, respectively. Fresh xylene was changed after eachextraction run to make sure that all the mineral oil was removed.(5) Post-drawingExtracted fibres were post-drawn using a homemade post-drawing system consisting oftwo rollers and a heated oven, as shown in Figure 4. 4. The oven was kept at a desiredtemperature in the range of 110?C to 140?C with the optimum conditions obtained in Chapter 5.By adjusting the speed of the two rollers, different draw ratios could be applied during the post-drawing process. The draw ratio and temperature used in this study are shown in Figure 4. 6.Figure 4. 4 Post-drawing processTable 4. 4 Post-drawing condition UHMWPE fibrePost drawn condition 1 2 3 5Draw ratio 5 3 2 1.5Temperature 110?C 120?C 130?C 140?CChapter-4 Material and Methods444.2.2 X-ray photoelectron spectroscopy (XPS)X-ray photoelectron spectroscopy (XPS) is a method used to determine the elementalcomposition of a material?s surface. The Perkin Elmer PHI 5600 XPS using the Mg K? X-rayline of 1253.6 eV excitation energy, 300W, 15 kV was used in this study for characterizing theMWCNT functionalization. The vacuum base pressure was approximately 1?10?8 Torr. Thebinding energy of the C1s of graphite, 284.5 eV (?0.35 eV energy resolution of the spectrometerat the settings employed) was taken as the reference. MWCNT samples were set on indium foiland then placed on the sample holder for analysis in the XPS surveys and multiplex. Prior toindividual elemental scans, a survey scan was taken for all the samples in order to detect theelements present. Carbon, oxygen and titanium analysis were carried out on each sample.4.2.3 Transmission electron microscopy (TEM)TEM images were taken on a FEI Tecnai G2 electron microscope at 200 KV. A dropletof the MWCNT/dimethylformamide (DMF) dispersion was deposited on a microscopy coppergrid covered with a holey silicon dioxide film and the solvent was evaporated in air at roomtemperature.4.2.4 Raman spectroscopyRaman spectroscopy is one of most widely used spectroscopic techniques for analyzingnano-material. It is based on inelastic scattering, or Raman scattering, of monochromatic light,usually from a laser in the visible, near-infrared or near-ultraviolet range. The laser light interactswith molecular vibrations and photons of the laser light are absorbed by the sample and then re-emitted. The energy of the laser photons shifts up or down to provide information aboutvibrational, rotational and other low-frequency transitions in molecules. In this study, aChapter-4 Material and Methods45Renishaw 1000/2000 Raman micro-spectrometer equipped with a 514.5 nm argon ion laser asthe excitation source was used for all the analyses. The laser power was reduced to <0.1 mW (50?objective) to prevent overheating of the fibre sample.4.2.5 Fourier transform infra-red (FTIR)In the FTIR test, infrared radiation is passed through a sample and some of it is absorbedby the sample while the other is passed through (transmitted). The resulting IR spectrum createsa molecular fingerprint of the sample with absorption peaks that correspond to the frequencies ofvibrations between the bonds of the atoms making up the material. Because each material is aunique combination of atoms, no two compounds produce exactly the same infrared spectrum.Therefore, infrared spectroscopy can result in a positive identification (qualitative analysis) ofevery different kind of material. This makes infrared spectroscopy useful for materialcomposition analysis. FTIR spectra of MWCNTs were recorded from 400 cm-1 to 4000cm-1witha Perkin Elmer FT-IR spectroscopy instrument by KBr pellets.4.2.6 Thermogravimetric analysis (TGA)TGA is a technique in which the mass of a substance is monitored as a function oftemperature or time as the sample specimen is subjected to a controlled temperature program in acontrolled atmosphere. Measurements are primarily used to determine the composition ofmaterials and to predict their thermal stability at temperatures up to 1000?C. The technique canbe used to characterize materials that exhibit weight loss or gain due to decomposition, oxidation,or dehydration. TGA of a functionalized MWCNT sample was performed using a TA InstrumentQ500TGA. A 5mg sample was heated at 20?C min-1 from room temperature to 600?C in adynamic nitrogen atmosphere (flow rate=50ml/min).Chapter-4 Material and Methods464.2.7 Tensile testThe sample for the single fibre tensile test was prepared by placing the fibre on arectangular paper frame and affixing it with epoxy glue, as shown in Figure 4. 5. According toASTM D3379-75, a gauge length of 25 mm was used for the UHMWPE fibre, which has atensile strain of 3-4%. After placing the sample on the tensile tester, the paper frame was cut toinitiate the testing frame. 25 fibres were tested for each sample and the average strength,modulus, and strain were used for data analysis.Figure 4. 5 Preparation of single fibre tensile test sampleSingle fibre tensile tests were carried out using a KES-G1 tensile testing machine with a50 Newton load cell. Tests were performed at a crosshead speed of 20 mm/min according toASTM D3379-75. The breaking load in Newton and time duration in seconds needed forbreaking the fibre were recorded. The diameter of the fibre was measured by the average of thethree readings from the optical microscopy image. The modulus of the fibre was calculated fromthe slope of the initial linear section of the stress-strain curve. A typical stress-strain curve ofUHMWPE fibre is shown in Figure 4. 6.Chapter-4 Material and Methods47Figure 4. 6 Typical stress-strain curve of UHMWPE fibreChapter-5 Optimization of Gel Spun UHMWPE fibre48Chapter 5- Optimization of gel spun UHMWPE5.1 IntroductionThe tensile properties of gel spun UHMWPE fibre are influenced by spinning conditions,including polymer concentration, spinline draw ratio (draw ratio between spinneret and winder),and spinning temperature. For example, the viscosity of spinning dope is highly sensitive tospinning solution concentration and extrusion temperature because of the high molecular weight.Besides spinning temperature, winding speed has been found to have a significant effect on thefibre properties as well [151]. The probability of introducing structural defects into the as-spunfibre increases with winding speed. If formed, these defects are difficult to remove during the hotdrawing step, thus leading to the production of fibres with poor tensile properties.In order to investigate the influence of the spinning conditions on the morphology andmechanical properties of gel spun UHMWPE fibres, polymer concentration, spinningtemperature, and winding speed were studied systematically using a design of experiment. Theaim of this part of the study is to find the optimum spinning condition that could be used insubsequent spinning experiments for this thesis.5.2 Effect of spinning conditions on morphology and property ofUHMWPE fibreDuring gel spinning, polymer gel was forced through the spinneret and quenched in awater bath. Winding of the as-spun fibre applied initial drawing to the fibre. A high percentageof solvent exists in as-spun fibres and is removed through solvent extraction. The purpose ofthese processes is to prepare the fibre with disentangled molecular chains, thus facilitating theChapter-5 Optimization of Gel Spun UHMWPE fibre49subsequent super-drawing process. During super-drawing (draw ratio>100), these disentangledmolecular chains were extended and aligned along the fibre axis, thus giving the fibre ultra-highstrength and modulus. Throughout the whole gel spinning process, weak polymer gel withrandomly oriented polymer chains was transformed to the fibre with high tensile properties. Itwas expected that the processing parameters would have great influence on the properties of thegel spun fibre. Among various parameters, polymer concentration, spinline draw ratio, andspinning temperature were found to be the three most important factors [152]. In this chapter, theeffects of these parameters on the tensile properties and morphology of gel spun UHMWPE werestudied systematically according to a design of experiment.5.2.1 Polymer concentrationPolymer concentration has a significant effect on the density of the molecular chainentanglements contained in the spinning dope. In order to obtain fibres with a highly alignedstructure, the optimal polymer concentration should be the concentration at which just enoughentanglements exist to maintain spinning and drawing stability. Lower than the optimal polymerconcentration, the intermolecular entanglement network cannot sustain the spinline stresses(which leads to constant breakage). In this study, 2wt% UHMWPE/mineral oil was found to bethe lowest concentration at which gel spinning could be conducted. Based on this observation,polymer concentration ranging from 2wt% to 5wt% were used for analyzing the effects ofpolymer concentration on UHMWPE fibre properties.5.2.2 Spinning temperatureSpinning temperature is one of the most important parameters for the spinning processbecause it has great influence on the viscosity of the spinning dope. Figure 5. 1 shows how theChapter-5 Optimization of Gel Spun UHMWPE fibre50shape of a spinline of a 1.5% UHMWPE/mineral oil gel changes with increasing temperature. Itwas found that when the temperature is too high, the viscosity of the spinning dope decreases tosuch a point that continuous spinning of the fibre cannot be carried out and the spinline willfracture frequently due to the low coherence between the molecules, resulting in serious capillaryfailure. However, a certain temperature is needed because an increase in temperature will reducethe elongation rate at which molecular scission occurs. Based on published results andpreliminary experimental results in our laboratory, three temperature levels (150?C, 180?C and210?C) were selected for this study.(a) 150?C (b) 180?C (c) 210?CFigure 5. 1 Viscosity of fibre at various spinning temperatures5.2.3 Winding speedSpinline fractures can also happen as a consequence of high winding speed due to highstress from rapid deformation of visco-elastic materials. This high stress can lead to stress-induced crystallization, which in turn increases the elasticity of the extrudate. Based on somepreliminary tests, three levels of winding speed levels (5, 10 and 20 feet/min (FPM), whichcorrespond 12.7, 25.4 and 50.8 meter/min, respectively) were selected for this study.Chapter-5 Optimization of Gel Spun UHMWPE fibre515.2.4 Orthogonal experiment designTo understand the effect of spinning condition on fibre properties and morphology and tofind an optimal spinning condition, a 3-factor, 3-three level design of experiment was applied.The factors and levels of the experimental design are shown in Table 5. 1. For a complete 3-factor, 3-level factorial experiment design, twenty-seven experiments are required for allpossible level combinations. As described in last chapter, the entire gel spinning process includesgel preparation, spinning, extraction, post-drawing, and a tensile test. In order to study the effectof the spinning condition on fibre tensile properties more efficiently, a design of experimentmust be considered.Table 5. 1 Factors and levels used in experimental designFactorsLevelsPolymer concentration(wt%)Spinning temperature(?C)Winding speed(FPM)1 2 150 52 3.5 180 103 5 210 20Among various design of experiment (DOE) methods, orthogonal experimental design,which is one type of fractional factorial experiment, was chosen for this study due to itsefficiency and suitability for the various natures of the factors. Orthogonal experimental design isan optimization method that utilizes orthogonal tables to arrange the experiment forsystematically researching various factors and levels. The advantage of using orthogonal designis that only representative tests will be performed according to the orthogonality. As such, it hasthe capability of sampling a smaller amount but representative set of level combinations. Forexample, a 3-level, 4-factor optimization will need to test eighty-one combinations. By usingorthogonal design according to orthogonal array L9_3_4 (the 3-level, 4-factor design array) asChapter-5 Optimization of Gel Spun UHMWPE fibre52shown in Table 5. 2, only nine representative combinations are necessary. In the L9_3_4orthogonal matrix, each row represents a designed level and each column represents a factor. Inan orthogonal matrix, each level is repeated the same number of times and any combination oftwo factors is combined with the third factor at three levels, respectively. As shown in Figure 5.2, the experiment points of orthogonal design are evenly distributed among all factors and levels;for example, A1B1 has a combination with C1, C2 and C3.Table 5. 2 L9_3_4 Orthogonal experiment designExperiment NO. Combination PolymerConcentrationSpinningTemperatureWindingSpeed1 A1B1C1 2% 150?C 5FPM2 A1B2C2 2% 180?C 10FPM3 A1B3C3 2% 210?C 20FPM4 A2B1C2 3.5% 150?C 10FPM5 A2B2C3 3.5% 180?C 20FPM6 A2B3C1 3.5% 210?C 5FPM7 A3B1C3 5% 150?C 20FPM8 A3B3C1 5% 180?C 5FPM9 A1B3C2 5% 210?C 10FPMFigure 5. 2 3-factor, 3-level orthogonal experimental designChapter-5 Optimization of Gel Spun UHMWPE fibre535.2.5 Statistical analysis of orthogonal experimental design resultsThe analysis of variance conducted by an F-test determines if the variance in experimentalresults is coming from the variance of factors or from experimental error and the significance. Ifthe null hypothesis is true, statistical analysis will show that factor has no significant influenceon testing properties. If the null hypothesis is false, then we will reject the null hypothesis andconsider that the factor has no significant effect on the testing results.5.3 Results and discussionUHMWPE fibres spun from gels with different polymer concentrations and at differentspinning conditions were prepared. Tensile properties were tested according to the methodsdescribed in Chapter 4. From the tensile test results, it was found that diameter has a significantinfluence on the tensile strength of gel spun UHMWPE fibre, as shown in Figure 5. 3. A similarphenomenon has been observed by many researchers. The earliest study was conducted byGriffith [153] in the 1920s when he tested tensile strength of glass fibres at different diameters,as shown in Figure 5.4. He pointed out that material strength was greatly affected by the size ofdefects which initiate cracks under tension due to stress concentration. Fibres with a biggerdiameter showed lower tensile strength because larger pre-existing cracks could statistically existin larger samples.Chapter-5 Optimization of Gel Spun UHMWPE fibre54Figure 5. 3 Variation of tensile strength with UHMWPE fibre diameterFigure 5. 4 Variation of tensile strength with glass fibre diameter [153]In the post-drawing of UHMWPE fibre, fibre diameter decreased from 100-200?m to 20-30?m. To avoid the influence of fibre diameter on tensile test results it is necessary to compare aChapter-5 Optimization of Gel Spun UHMWPE fibre55fibre with the same diameter. In this study, great efforts were made to control fibre diameter.However, due to the equipment limitations, variations in fibre diameter still existed as fibres withdiameters in the range of 20-25 ?m were tested, which is one of the reasons for the variance intested fibre properties.5.3.1 Optimization of spinning processThe orthogonal table L9_3_4 (as shown in Table 5. 2) was used to arrange theoptimization experiments. UHMWPE fibres were prepared according to orthogonal experimentaldesign. For each sample, 25 fibres with diameters in the range of 20-25 ?m were tested. Thefibre ballistic performance index U*(1/3) value, which is calculated according Cunniff?s model asshown in Equation 1.2, was used as the response for the optimization since the objective of thisstudy was to obtain fibre with the highest anti-ballistic properties. Tensile test results includingfibre strength, strain, modulus and U*(1/3) are listed in Table 5. 3.Chapter-5 Optimization of Gel Spun UHMWPE fibre56Table 5. 3 Tensile properties of UHMWPE fibre spun at different conditionsExperimentNOPolymerconcentration TemperatureWindingspeed(RPM)ErrorDiameter (?m) Strength (GPa) Modulus (GPa) Strain (%) 1*( )3U(m/s)AVG STDEV AVG STDEV AVG STDEV AVG STDEV1 2% 150?C 5 1 21.97 2.70 1.72 0.37 51.40 6.80 4.40 1.06 656.642 2% 180?C 10 2 21.66 3.19 1.98 0.36 56.01 6.13 3.94 0.64 673.343 2% 210?C 20 3 21.66 3.13 0.94 0.28 21.51 4.38 4.26 0.49 459.974 3.5% 150?C 10 3 20.95 3.10 2.92 0.39 92.06 19.51 4.23 1.38 852.035 3.5% 180?C 20 1 21.55 2.74 2.62 0.37 58.12 7.00 4.79 0.54 794.346 3.5% 210?C 5 2 21.29 3.23 1.96 0.28 63.92 9.48 4.12 0.71 696.807 5% 150?C 20 2 21.71 3.00 1.54 0.23 40.00 9.76 4.91 1.15 629.708 5% 180?C 5 3 20.61 3.13 1.78 0.50 55.22 9.90 4.17 0.47 661.109 5% 210?C 10 1 20.84 3.08 1.06 0.29 24.96 4.93 4.17 0.76 499.51An F-test was done to investigate if these factors had significant influence on U*(1/3).From the F-test results (as shown in Appendix A), it can be seen that among the three factorsinvestigated, polymer concentration and spinning temperature had significant influence on U*(1/3)at 99% confidence, while winding speed had significant influence at 90% confidence. Thus, it isimportant to study the effect of these factors on fibre tensile properties and morphology and findthe optimal level of these three factors so as to spin UHMWPE with the highest U*(1/3). Theoptimum spinning conditions were found to be with a polymer concentration of 3.5%, a spinningtemperature of 150?C, and a winding speed of 5 FPM.5.3.2 Effect of spinning condition on fibre morphologyBesides affect fibre tensile properties, spinning condition also has significant influence onthe morphology of gel spun fibres. Fibre morphology was studied using scanning electronmicroscopy (SEM) to understand its relationship with fibre tensile properties . Post-drawn fibreswere initially used as samples for morphology study. However, it was found that the morphologyof fully post-drawn UHMWPE fibres spun from different conditions looked quite similar eventhough they were spun at various conditions, as shown in Figure 5. 6. During hot drawing, all theporous structures in as-spun fibres are converted into fibres with smooth fibrils. It is known thatthe main factor determining the drawability is the entanglement network in the as-spun fibres,which is strongly affected by the flow conditions in the extruder. Thus, as-spun fibres were usedas samples to investigate the effect of the spinning condition on fibre morphology.Chapter-5 Optimization of Gel Spun UHMWPE fibre58Figure 5. 5 Morphology transformation during post-drawingFigure 5. 6 Morphology of UHMWPE fibre after post-drawingChapter-5 Optimization of Gel Spun UHMWPE fibre59After UHMWPE fibre was extruded from the spinneret, the polymer was immersed in acold water bath to be quenched and crystallized, during which the structure would be completelysolidified. Tension applied on the as-spun fibre during crystallization is determined by thewinding-up speed. Thus, the morphology of as-spun fibres is affected by winding speed as well.From an examination of the morphology of UHMWPE fibres winding up at different speeds itwas found that although the as-spun UHMWPE fibres showed a shish-kebab structure, there wasa distinct difference in the nature of the backbone materials. Figure 5.7(a) shows UHMWPEfibre that was wound up at 5 FPM. It can be seen that all the kebab structures formed closelyfolded molecular chains and no extended fibrils were discerned in between. Meanwhile, inFigure 5.7(b), which is the fibre winding-up at speeds of 10FPM, it can be seen that the fibrillarentities have average lateral dimensions of about 500 nm and the length of the kebab structure isaround 100 nm. When the winding speed was increased to 20 FPM, the lateral dimension of thekebab structure decreased to about 100nm while the length between each kebab became evenbigger, as shown in Figure 5.7(c). This observation implies that winding speed greatly affects themorphology of as-spun fibre, as the elongation of the kebab structure is proportional to thedeformation rate during spinning.Figure 5.7 (a)Chapter-5 Optimization of Gel Spun UHMWPE fibre60Figure 5.7(b)Figure 5.7 (c)Figure 5. 7 Morphology of UHMWPE fibre dran at various winding speeds(a) Winding at 5 FPM (b) Winding at 10 FPM (c) Winding at 20 FPMIt was found that UHMWPE showed certain level of oxidization when heated-up atatmosphere for longer than four hours. As shown in Figure 5. 8, the colour of the spinning gelwas found turning from white to yellow when the spinning temperature was increased to 210?C.UHMWPE oxidization can be avoided in practice by using nitrogen to protect the heating zonein the commercial production condition however was not achievable in lab scale in this study.This could be another reason that a lower spinning temperature is preferred in this study.Chapter-5 Optimization of Gel Spun UHMWPE fibre61Figure 5. 8 (a) before oxidization Figure 5. 8 (b) after oxidizationFigure 5. 8 Oxidization of UHMWPE gel under atmosphere5.3.2 Optimization of extraction processBecause the UHMWPE concentration in the spinning dope used in the gel spinningprocess is only about 2-5wt%, the as-spun fibre contains a high concentration (95-98%) ofmineral oil. The necessity of removing the mineral oil before the hot drawing process has beenargued by researchers such as Roukema [52] and Peterlin [154]. Roukema pointed out that themineral oil could have a similar effect as a plasticizer which has been used in other fibreproduction and is believed to increase the maximum draw ratio through diffusion into theamorphous regions and a decrease in the cohesion between the molecules. On the other hand,Peterlin found that it was much more effective to align polymer chains by post-drawing them inthe solid state rather than the liquid state because the relaxation phenomena counteracting themolecular orientation in the solid state is negligible comparing to those in the liquid state.In this study, gel spun UHMWPE fibres were hot drawn with and without the extractionof the mineral oil. The fibre's drawability and tensile strength was measured to evaluate thenecessity of the extraction process. The extracted sample was prepared according to the methoddescribed in Chapter 4. As-spun fibre was used as an un-extracted sample. Both samples werepost-drawn using a homemade post-drawing system. A very low constant post-drawing speedChapter-5 Optimization of Gel Spun UHMWPE fibre62(1cm/min) was used in the post-drawing process to avoid stress build-up, which can lead topremature breakage of the fibre. The samples were drawn until they broke. The final length ofthe sample was measured and compared with the original length to calculate the draw ratio usingEquation 5. 1.Equation 5. 1It was found that fibre strength and modulus increased significantly with the post-drawratio. This can be explained using Peterlin?s model which relates fibre strength with the numberof taut tie molecules in a cross section area. During the post-drawing process, the moleculesgradually aligned along the fibre axis and the fibre diameter become smaller. Thus, there aremore taut tie molecules in the unit's areal on the fibre's cross section. A fibre with a higher post-drawn ratio is thus more desirable. As evident in Figure 5.9, the fibre extracted before post-drawing showed a maximum draw ratio of about 21, as opposed to 14 for the un-extracted fibre.The tensile strength and modulus of UHMWPE fibre at different draw ratios is shown in Figure5.10 This figure shows how the fibre strength and modulus increase sharply as the diameterdecreases. At the same draw ratio, fibre that is post-drawn without mineral oil showed muchhigher strength and modulus compared with fibre with mineral oil. This could be explained bythe fact that the fibre without extraction contained a high content of mineral oil, and therefore notenough fixing points. Entangled molecules tend to slip over each other instead of beingstraightened when they are pulled under tension. Thus solvent removal in as-spun UHMWPEfibre using solvent extraction is necessary.Chapter-5 Optimization of Gel Spun UHMWPE fibre63Figure 5. 9 Maximum post-draw ratio of UHMWPE fibre with and without mineral oilFigure 5.10 (a) Figure 5.10 (b)Figure 5. 10 Effect mineral oil on UHMWPE fibre tensile properties(a) Tensile strength ( b) Young modulus5.3.3 Optimization of post-drawing processDuring the gel spinning process, UHMWPE gel exits the capillary of the spinneret and isquenched in a water bath. The stretching of the spinline due to the winding up mechanismChapter-5 Optimization of Gel Spun UHMWPE fibre64introduced a limited amount of orientation to as-spun fibre. In order to further straighten andalign polymer chains, a post-drawing process is necessary. During post-drawing, the relativelyweak as-spun UHMWPE fibres were transformed into fibres with improved molecularorientation and tensile properties.Temperature is one of the most important factors affecting the effectiveness of post-drawing [155]. In order to obtain high tensile properties, the fibre has to be drawn at atemperature that is high enough to prevent molecular scissioning, yet low enough to prevent slipand relaxation. From practical experience, the optimum drawing temperature seems to be severaldegrees below the polymer's melting temperature [156, 157]. For the UHMWPE used in thisexperiment, it was found that polymer starts to melt at a temperature of about 145?C, thus, thehighest temperature used in the post-drawing process was set to be 140?C, which is 5?C lowerthan the melting temperature. In order to evaluate the effect of temperature on fibre drawabilityand tensile properties, four different temperatures were applied in the post-drawing process. Thedetailed descriptions of the post-drawing conditions are listed in Table 5.4Table 5. 4 UHMWPE fibre post-draw conditionPost-drawn condition A B C DTemperature 25?C 110?C 140?C110?C for fist draw120?C for second draw130?C for third draw140?C for fourth drawThe maximum draw ratios of UHMWPE fibre post-drawn at different temperatures areshown in Figure 5.11. It is evident that the maximum draw ratio of fibre post-drawn at roomtemperature was much lower compared with fibre drawn at elevated temperatures. This isChapter-5 Optimization of Gel Spun UHMWPE fibre65because room temperature could not provide enough energy for the molecules to disentangle.Post-drawing at room temperature led to the formation of tight knots of entangled molecules asdrawn ratio increased which ten caused stress concentration at certain points and initiated fibrebreakage at low strain.Figure 5. 11 Effect of temperature on maximum draw ratio(A)25?C (B) 110?C (C) 110/120/130/140?C (D) 140?CThe typical morphology of fibre drawn at room temperature is shown in Figure 5.12. It isclear that large clusters consist of un-oriented molecules that are bridged by a few tie molecules.Because these clusters were covered by more than one bundle of fibrils, they could be easilybroken up into several smaller units as draw ratio increased.Chapter-5 Optimization of Gel Spun UHMWPE fibre66Figure 5.12 (a)Figure 5.12 (b)Figure 5.12 (c)Figure 5. 12 Morphology of UHMWPE fibre drawn at room temperature(a) Draw ratio=1, (b) Draw ratio=3, (c) Draw ratio=6Chapter-5 Optimization of Gel Spun UHMWPE fibre67At room temperature, the fibre elongation was obtained by stretching fibrils that connectthe clusters of folded molecules bundles. These fibrils broke at a low draw ratio and snappedback to random coil conformation immediately, which led to the generation of elastic turbulences.As a result, the connection between these un-oriented clusters was partially destroyed. Thus,fibres post-drawn at room temperature showed poor drawing and tensile properties.Higher draw ratio was observed for UHMWPE fibres drawn at elevated temperatures of110?C and 140?C. Fibre drawn with a stepwise increased temperature (110,120,130,140?C)showed the highest draw ratio. Lower temperature is preferred at the start of post-drawingbecause that lower temperature helps keep the 'fixing points' in the entangled molecular structure.Thus, the molecules in between these fixing points can be straightened. As the post-drawing ratioincreases, the molecules gradually align along the fibre axis and fibre crystallinity increasesaccordingly. At the same time, the melting point of the polymer increases, and as such the highertemperature is needed to provide enough energy for molecules to continue to disentangle. On theother hand, when a higher temperature is used from the beginning, molecules are easily pulledout of entanglements and snap back to their random coil conformation. The breakage of thenetwork gives rise to stress concentrations in neighbouring molecules. Therefore, when thetemperature used in post-drawing is too high, not only does disentangling take place but alsochain scissoring.The tensile properties of UHMWEP fibre post-drawn at different temperatures are shownin Figure 5.13. Similar to draw ratio, fibre post-drawn in three stages by gradually increasing thetemperature showed the highest strength, modulus, and U*(1/3).Chapter-5 Optimization of Gel Spun UHMWPE fibre68Figure 5.13 (a) Figure 5.13 (b)Figure 5.13 (c) Figure 5.13 (d)Figure 5. 13 Effect of post-drawing temperature on UHMWPE fibre tensile properties(a) Tensile strength (b) Modulus (c) strain (d) U*(1/3)Chapter-5 Optimization of Gel Spun UHMWPE fibre695.4 Validation of optimization of gel spinning processThe gel spinning process (which includes spinning, solvent removal and post-drawing)was studied and optimized. The optimum spinning condition was optimized through DOE usingthe orthogonal method. It was found that the optimum spinning condition to obtain higher U*(1/3)was 3.5wt% UHMWPE/mineral oil gel, spin at 150?C and winding at 5 FPM. UHMWPE fibreswere produced under this optimum condition and the morphology and tensile properties of theresulted fibre is shown in Figure 5. 14 and Table 5. 5. UHMWPE fibre processed in theseoptimum conditions showed U*(1/3) of 883, which is significantly higher compared to fibreprocessed in other conditions. This demonstrates that the optimization process performed in thisexperiment was successful.Figure 5. 14 Morphology of UHMWPE fibre spun at optimum conditionChapter-5 Optimization of Gel Spun UHMWPE fibre70Table 5. 5 Tensile strength of UHMWPE fibre spun at optimal conditionsDiameter (?m) Strength (GPa) Modulus (GPa) Strain (%) )31(^U (m/s)Average 22.17 3.37 101.93 3.84% 881STDEV 2.80 0.35 16.29 0.56% 178.55.5 SummaryIn this chapter, the gel spinning process was optimized by studying the influence of thethree most important factors through the design of experiment: polymer concentration, spinningtemperature and winding speed. The influence of these factors on fibre morphology, and theinfluence of extraction and the post-drawing process on fibre post-draw ratio was also studied inthis chapter. UHMWPE fibre was produced under the optimum spinning, extraction and post-drawing conditions and the tensile properties were tested. This optimum spinning condition willbe used for all subsequent experiments in which carbon nanotubes (CNTs) are added to pureUHMWPE fibre as a reinforcement.Chapter-6 MWCNT Reinforced UHMWPE Fibre71Chapter 6- MWCNT reinforced UHMWPE fibre6.1 IntroductionCarbon nanotubes (CNTs) are now widely recognized as excellent reinforcement fillerfor polymer composites because of their outstanding mechanical, thermal and electricalproperties. In recent years there has been an explosive growth in research activity in CNT-reinforced polymer composite materials. It is also widely recognized by researchers that thereinforcement effectiveness of CNTs in a polymeric matrix depends on their volume ratio andtheir dispersion in the hosting medium. In order to utilize the excellent properties of CNTs, oneof the most important challenges is to obtain a uniform dispersion of CNTs in the polymermatrices. In spite of many reports claiming successful reinforcement of polymers by CNTs, thereinforcing efficiency is far below expectations. This is primarily due to the strong interactionforces that cause the formations of bundles and clusters in as-produced CNTs.Although there are various methods of producing CNTs such as arc discharge, chemicalvapor deposition, laser ablation, high pressure carbon monoxide (HiPco) and electrolysis, theCNTs produced from these methods are always physically entangled or adhere to each other inthe form of bundles, as shown in Figure 6. 1. The cohesion energy between two parallel CNTs isabout ?0.095 eV/?, as calculated by Girifalco et al. using a continuum model [158]. Since thethermal energy at room temperature is only about 0.025 eV/? (which is much lower than thecohesion energy between CNTs) the CNT de-bundling from an existing bundle on their ownunder room temperature is unlikely. In some cases, those CNT bundles will further aggregateinto entangled networks, which have been found to act as a obstacle for realizing theextraordinary mechanical and electrical properties of individual tube [94, 159]. Because of theChapter-6 MWCNT Reinforced UHMWPE Fibre72strong inter-tube van der Waals forces, the as-produced CNTs usually align parallel to each otherand pack into crystalline bundles, which are typically composed of 100-500 tubes. Thus,improving CNT dispersion has been widely recognized as a crucial issue for fully utilizing theenhancement potential of the CNT in CNT/polymer composites.Figure 6.1 (a) Figure 6.1 (b)Figure 6. 1 Morphology of as-produced CNTs [160, 161](a) CNT ropes produced by discharge method (b) CNTs grown by plasma-enhanced chemicalvapour deposition6.1.1 Importance of CNT dispersion in UHMWPE fibreA good dispersion of CNTs is especially important for reinforcing UHMWPE fibre. Thisis because the fibrils of UHMWPE fibre, with a size of around 100nm to 1?m, are extended andtightly aligned along the fibre axis during the post-drawing process and any micro-meter sizenanotube agglomeration could form structural defects. As fibre is under great tension duringdrawing, any stress concentration due to structural defects may lead to crack initiation and thusthe fibre's rupture. CNT agglomerates will also lead to premature breakage of UHMWPE fibresduring the post-drawing process. Thus, the molecules in the fibre consisting of theseagglomerates can not be fully drawn to obtain a highly extended and aligned structure. Sincetensile properties of UHMWPE fibre are greatly determined by the fibre's draw ratio andChapter-6 MWCNT Reinforced UHMWPE Fibre73diameter (as discussed in Chapter 5), fibres with lower draw ratio or larger diameter usuallyshow lower tensile properties.6.1.2 Methods to disperse CNTs in solventConsidering the importance of separating CNTs? bundles and agglomeration, there havebeen intensive efforts to develop methods that will enable CNT solubilization, dispersion andseparation. These methods can be organized into two categories: chemical methods andmechanical methods. The chemical approach includes chemical modification of the CNT surfaceby attaching functional groups, or functionalization by wrapping the CNT with long polymerchains. The mechanical method usually refers to nanotube separation using shear mixing or ultra-sonication, which can transfer local shear stress to break down aggregates. Due to its simplicityand efficiency, mechanical dispersion is usually preferred. In this study, the feasibility ofachieving a good dispersion of CNTs in UHMWPE fibres by mechanical methods wasinvestigated. MWCNTs dispersed by ultra-sonication and ball milling were used asreinforcements for UHMWPE fibres. Fibre morphology and tensile properties were used ascriteria for evaluation of CNT dispersion in UHMWPE fibres.6.2 Dispersing by ultra-sonicationUltra-sonication is one of the most widely used methods to disperse nanoscale materialsand has been shown to be very effective in breaking up CNT aggregates and dispersing CNTs[162-169]. CNT dispersion through high-intensity ultrasonic waves is caused by cavitationduring which very high temperatures (approx. 5,000K) and pressures (approx. 2,000atm) can bereached locally [170]. Implosion of cavitation bubbles also creates high velocity liquid jets witha velocity up to 280 m/s that overcomes bonding forces between CNTs and de-agglomerateChapter-6 MWCNT Reinforced UHMWPE Fibre74CNTs [171]. The fluid flow caused by the bubble collapse depends on the solvent's boiling pointand viscosity.6.2.1 Materials and experimentAcid purified MWCNTs with a carboxyl group (COOH-MWCNTs) with a diameter ofabout 8nm and a length of 1?m were used for this study. Detailed information for the materialsincluding the UHMWPE, mineral oil and anti-oxidant used in this study can be found in Chapter4.From the optimization of the gel spinning process in Chapter 5, it was discovered that theoptimum UHMWPE concentration of mineral oil was 3.5wt%. In this study, UHMWPE fibreswith a MWCNT concentration of 0.5, 1.0 and 1.5wt% were prepared. MWCNTs were dispersedin mineral oil by bath ultra-sonicating(Model: Misonix3000) for 10 hours before addingUHMWPE into the mixture. UHMWPE powders and anti-oxidants were then added to theMWCNT/mineral oil mixture. The mixture of MWCNT, anti-oxidant, UHMWPE and mineral oilwas then heated up to form a uniform gel. 3.5wt% UHMWPE/Mineral oil with MWCNTconcentration of 0.5, 1.0 and 1.5wt%, respectively, was extruded into the fibre at 150?C with awinding speed of 5FPM, which was the optimum spinning condition determined in Chapter 5.6.2.2 Results and discussionThe tensile properties of MWCNT/UHMWPE composite fibres with various MWCNTcontents were tested using the methods described in Chapter 4. Pure UHMWPE fibres were alsoprepared and tested under the same conditions and used as reference samples. The tensilestrength, modulus and strain of MWCNT/UHMWPE composite fibres containing 0.5, 1.0 and1.5wt% MWCNTs are shown in Figure 6. 2 (a), (b), (c), respectively.Chapter-6 MWCNT Reinforced UHMWPE Fibre75Figure 6. 2 (a) Figure 6. 2 (b)Figure 6. 2 (c) Figure 6. 2 (d)Figure 6. 2 Tensile properties of MWCNT/UHMWPE fibre prepared by ultra-sonication(a) Tensile strength (b) Modulus (c) Tensile strain (d)U*(1/3)As can be seen in Figure 6. 2(a) and (b), the tensile strength and modulus of UHMWPEfibre decreased significantly after adding MWCNTs, which were dispersed by ultra-sonication.Tensile strain increased slightly after adding 0.5 and 1.0% MWCNTs but decreased when theMWCNT content increased to 1.5wt%, as shown in Figure 6. 2(c). However, the increase instrain is not statistically significant at a confidence level of 95%. Due to the decrease in tensileChapter-6 MWCNT Reinforced UHMWPE Fibre76strength and modulus, the U*(1/3) of MWCNT/UHMWPE fibre is even lower than pureUHMWPE. The decrease in fibre tensile properties after adding MWCNTs could be attributed tothe non-uniform dispersion of MWCNTs, which was examined by optical microscopy andshown in Figure 6. 3. Apparently, MWCNTs were not uniformly dispersed in the UHMWPEmatrix. MWCNT agglomerations with a size of 10-20?m could be found even at 5 ?magnifications. The agglomeration worsened as the CNT concentration increased. According tothe stress concentration theory, these CNT agglomerates will behave as structural defects andcause stress concentration under tension, thus leading to inferior mechanical properties.Assuming that the CNT agglomerations have a circular shape, the stress on the edge of the CNTagglomerations could be at least three times of the tension applied on the fibre during post-drawing. These CNT agglomerates create discontinuity in the UHMWPE matrix, especiallywhen the fibre diameter decreases during post-drawing, thus leading to adverse effects inmechanical performance. MWCNT clusters with a size similar to fibre diameter were foundprotruding from the fibre surface, as shown in Figure 6. 4.(a) (b) (c) (d)Figure 6. 3 Morphology of MWCNT/UHMWPE fibre prepared by ultra-sonication(a) Pure UHMWPE (b) 0.5% (c) 1.0% (d)1.5% MWCNT/ UHMWPEChapter-6 MWCNT Reinforced UHMWPE Fibre77Figure 6. 4 MWCNTs protruding from UHMWPE fibre surfaceDuring ultra-sonication, CNT agglomerates are subjected to local shear stress generatedby the flow of the solvent determines the dispersion of CNTs. The solvent flow is caused bycavitation of the micro air bubbles. In order to successfully debundle CNT agglomerates, thelocal energy density in the solvent generated by sonication must be higher than the bindingenergy that holds CNTs together. The binding forces for CNTs can be evaluated by the van derWaals interaction between them. In order to simplify the calculation, a pair of CNTs wasmodeled by a pair of parallel rods [172]. It was estimated that the energy density ? required toseparate a pair of parallel tubes is around 100MPa [173]. The shear rate generated in a ultra-sonication bath is through the of cavitation collapse of bubbles which could reach as high as 109s-1 [174, 175]. The viscosity of mineral oil at room temperature is around 0.05 Pa?s. The localshear stress in the vicinity of the cavitation should be around 50 MPa. Thus, it was expected thatthere would not be sufficient energy to separate tubes from the bundles during sonication. Inorder to improve CNT dispersion, alternative dispersion methods that can provide a higher levelof shear energy, such as ball milling, were sought.Chapter-6 MWCNT Reinforced UHMWPE Fibre786.3 Dispersing by ball milling and sonicationBall milling is a mechanical process by which high-energy forces by colliding ballsreduce the material to a fine powder. It has been widely used in industry as a key piece ofequipment for grinding materials to produce powders such as cement and silicates. Laboratoryscale ball milling machines usually consist of cylindrical capped milling jars that sit on driveshafts. The milling jars are partially filled with the material to be ground and the grindingmedium rotates around a vertical axis. It has been demonstrated that the size of the CNTagglomerations can be decreased after the ball milling process [176]. However, side effects suchas CNT shortening and destruction have also been reported [177]. Kukovecz [178] performed asystematic study on the morphological and length changes experienced by MWCNTs duringlong-time ball milling. Samples were collected after ball milling for 1, 80, and 140 hours. It wasevident that SWCNT entanglement significantly improved and the SWCNT length wasmaintained after 1 hour of ball milling, as shown in Figure 6. 5 (b). However, longer ball millingdurations such as 40 hours and 140 hours did shorten the SWCNT length significantly, as shownin Figure 6. 5 (c), (d).Figure 6.5 (a) Figure 6.5 (b)Chapter-6 MWCNT Reinforced UHMWPE Fibre79Figure 6.5 (c) Figure 6.5 (d)Figure 6. 5 CNT morphology during ball milling process [178](a) 0 hours (b) 40 hours (C) 80 hours (d) 140 hoursIn another study, N. Pierard [179] investigated the effect of ball milling on the structureof SWCNTs. SWCNTs were ball milled for 1 hour, 8 hours and 50 hours, respectively, and itwas found that there was no significant nanotube destruction for the first 3 hours, as shown inFigure 6. 6 (a). However, longer milling times progressively disrupted the SWCNTs' structureand transformed them into multilayered polyaromatic carbon structures, as shown in Figure 6. 6(b), (c).Figure 6. 6 (a) Figure 6.6 (b)Chapter-6 MWCNT Reinforced UHMWPE Fibre80Figure 6. 6 (c) Figure 6. 6 (d)Figure 6. 6 CNTmorphology after ball milling [179](a) before ball milling (b) ball milled for 1 h (c) ball milled for 8 h (d) ball milled for 50 h6.3.1 Materials and experimentThe ball milling machine from Retsch (Model:Planetary Ball Mill PM 200 ) is shown inFigure 6. 7 with 50ml stainless steel grinding jars and stainless steel grinding balls (2mm indiameter) used in this study. 1.75 grams of MWCNT/UHMWPE mixtures with MWCNTconcentrations of 0.5, 1.0, 1.5wt% were fed into the grinding jars. Grinding balls were fed intogrinding jars until they took about half of their volume. The rotation speed was set to 3000 rpmand milling duration was fixed at 15 minutes. After ball milling, the mixture (including MWCNTand UHMWPE) was dispersed in mineral oil by sonication using a bath sonicator (Model:Misonix3000) under a power level of 80 Walt's for three hours. Anti-oxidant BHT was added tothe MWCNT/UHMWPE mixture and sonication was subsequently continued for another hour.The mixtures of MWCNT/UHMWPE/mineral oil with different MWCNT concentrations wereheated up in an oil bath to form spinning gels. A laboratory mixing extruder was used for the gelChapter-6 MWCNT Reinforced UHMWPE Fibre81spinning process. Detailed information regarding gel preparation, spinning and post-drawingprocesses has been detailed in Chapter 4.Figure 6.7(a) Figure 6.7(b)Figure 6. 7 Retsch ball mill machine(a) planetary ball mill machine (b) grinding jar6.3.2 Results and discussionContrary to the previous experiments which used ulta-sonication as the only method fordispersing MWCNTs, a ball milling process was added before ultra-sonication with the purposeof further improving MWCNT dispersion in UHMWPE. The tensile properties of theMWCNT/UHMWPE composite fibres containing 0.5, 1.0 and 1.5wt% MWCNT were tested andanalyzed, as shown in Figure 6. 8. There was a slight increase in tensile strength, however noimprovement in modulus and strain was observed. Accordingly, the U*(1/3) did not showsignificant change after adding MWCNT into UHMWPE either, as shown in Figure 6. 8(d).Chapter-6 MWCNT Reinforced UHMWPE Fibre82Figure 6.8 (a) Figure 6.8 (b)Figure 6.8 (c) Figure 6.8 (d)Figure 6. 8 Tensile properties of MWCNT/UHMWPE fibre prepared by ball milling(a) Strength (b) Modulus (c) Strain (d)U*(1/3)The CNT dispersion in UHMWPE fibre processed by ball milling and the ultra-sonication process was examined by optical microscopy, as shown in Figure 6. 9. The dispersionof MWCNTs was still not uniform as agglomerates the size of about 10?m were easily observedunder the microscope. As most MWCNTs exist in the form of agglomerates, very few individualnanotubes were dispersed in the UHMWPE matrix to act as reinforcement, thus the reinforcingefficiency was very low.Chapter-6 MWCNT Reinforced UHMWPE Fibre83In conclusion, ball milling could not efficiently disperse UHMWPE in UHMWPE fibre.The reinforcement of UHMWPE fibre using MWCNTs dispersed through ball milling could notbe realized. Other mechanical dispersion methods such as micro-fluidizing and a solvent assistedMWCNT spray method have also been studied and the results are shown in Appendix C.Figure 6.9 (a) Figure 6.9 (b) Figure 6.9 (c) Figure 6.9 (d)Figure 6. 9 Morphology of MWCNT/UHMWPE fibre prepared by ball mill process(a) Pure (b) 0.5% (c) 1.0% (d) 1.5% MWCNT/UHMWPE6.4 ConclusionThe feasibility of reinforcing UHMWPE fibre using MWCNTs without chemicalfunctionalization was investigated. Mechanical methods (including ultra-sonication and ballmilling) were used to disperse MWCNTs in mineral oil. The MWCNT dispersion and tensileproperties of MWCNT/UHMWPE composite fibres with 0.5%, 1.0% and 1.5% MWCNT wereexamined. It was found that both ultra-sonication and ball milling were not able to disperseMWCNTs uniformly in mineral oil, and thus 10?m to 20?m MWCNT agglomerations werefound in the UHMWPE fibre. Compared to using ultra-sonication alone, MWCNTs dispersedChapter-6 MWCNT Reinforced UHMWPE Fibre84through both ball milling and ultra-sonication showed improved dispersion. However, nostatistically significant improvement in tensile properties was observed. Other methods includingthe newly developed microfluidizing technology and a solvent assistant dispersion were alsostudied, as shown in Appendix C. However, again, no significant increase in tensile propertieswas found. Thus, it is concluded that UHMWPE fibre reinforcement using MWCNTs cannot berealized through simple mechanical dispersion. In the following two chapters, the feasibility ofdispersing MWCNTs through chemical functionalization is investigated.Chapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre85Chapter 7- Coupling agent functionalized MWCNTreinforced UHMWPE fibre7.1 IntroductionCoupling agents are chemicals that have two functional groups with different reactivity.One of the two functional groups reacts with organic materials and the other reacts withinorganic materials. The special structure of the coupling agent enables the linkage betweenfillers and a polymer matrix. Traditionally, coupling agents have been widely used in compositeindustry to improve the dispersion of inorganic fillers in a polymer matrix and the mechanicalperformance of some filler/polymer composites [180-183]. In recent years, coupling agents havealso been used in surface modification of nano-particles for improving dielectric properties inorganic?inorganic nano-composites. For instance, Cheng?s group applied a silane couplingagent to MWCNTs to increase the degree of interaction between a polyimide and MWCNTs forfurther improvement in the mechanical properties and the thermal stability of thepolyimide/MWCNT nano-composites [184]. Lin et al. [185] improved the tensile modulus ofthe polypropylene nano-composites by functionalization of nano silicon dioxide using silanecoupling agents.In this study, a titanium derived coupling agent was used to functionalize the surface ofthe MWCNTs. The coupling reaction was achieved in the MWCNTs by taking advantage of thecarboxyl groups on the surface of the acid-treated MWCNTs. These results in the formation ofmatrix-compatible organic monomolecular layers consist of a long alkyl chain on the MWCNTsurface.Chapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre867.2 Experimental7.2.1 Functionalization of MWCNTsThe titanium derived coupling agent isopropyl tri(dioctylpyrophosphate) titanate has thechemical form R0_Ti_(ORT)3,where R0 is a functional group and R is a hydrolysable alkyl group.The titanium derived coupling agent used in this work has the structure. The coupling reaction was achieved by taking advantageof the carboxyl groups on the MWCNTs, resulting in the formation of matrix compatible organicmonomolecular layers consisting of long alkyl chains on the CNT surface, according to thechemical mechanism shown in Figure 7. 1. The reaction was confirmed by XPS and Ramanspectroscopy.Figure 7. 1 Reaction between coupling agent and COOH-MWCNTs [26]Following Wang?s [26] method, the carboxyl group functionalized multi-walled carbonnanotubes (COOH-MWCNTs) reacted with a titanium derived coupling agent by adding COOH-MWCNTs into a coupling agent/alcohol solution. The mixture was stirred with a magnetic stirrerat 60?C for 30 minutes followed by sonication for 3 hours at 65?C. The functionalized MWCNTs(identified as Ti-MWCNTs) were then separated by filtration and dried in a vacuum oven atChapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre87110?C for 1 hour to allow a complete evaporation of alcohol. The resulting Ti-MWCNTs werewashed thoroughly with acetone to eliminate any non-reacted coupling agent.7.2.2 Gel spinning and post-drawingUHMWPE gel with 0.5%, 1% and 1.5% Ti-MWCNTs was prepared by dispersing Ti-MWCNTs and UHMWPE in mineral oil first and then heated up in an oil bath under nitrogenprotection to 160?C. The gel was vacuumed for 24 hours to eliminate air bubbles and then wascut into small pieces. The small pieces of MWCNT/UHMWPE gel were fed into the extruderand spun into fibre. The spinning, extraction, and post-drawing was performed using the methodsdescribed in Chapter 4. Spinning and post-drawing were carried out at the optimum spinningconditions discussed in Chapter 5.7.2.3 Characterization of MWCNTX-ray photoelectron spectroscopy (XPS) and a field emission analytic transmissionelectron microscope (FE-TEM) were used to characterize the changes in MWCNT chemicalstructure and surface morphology during functionalization. Single fibre tensile tests were carriedout using a KES-G1 tensile testing machine and the diameter of the filament was measured usingan optical microscope. The detailed methods are described in Chapter 4.7.3 Results and discussion7.3.1 XPSXPS is a useful tool for chemical surface analysis. Figure 7. 2 depicts XPS survey scans ofa COOH-MWCNT and Ti-MWCNTs in wide scan mode. Compared to the COOH-MWCNT, theChapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre88spectrum for Ti- MWCNTs shows a sharp Ti1p peak at 457 eV, which is attributed to the titaniumelement in the R0_Ti_(ORT)3 coupling agent.Figure 7. 2 XPS spectrum of an MWCNT and a coupling agent treated MWCNTThe elemental compositions obtained from the XPS spectra of the COOH-MWCNT andthe coupling agent functionalized MWCNTs are summarized in Table 7. 1. The COOH-MWCNT has an oxygen-to-carbon (O/C) atomic ratio of 0.19, while the O/C ratio of the Ti-MWCNT is 0.62. The increase in the oxygen-to-carbon (O/C) atomic ratio should be attributedto oxygen in the coupling agent, which confirms the reaction between MWCNTs and thecoupling agent.Table 7. 1 Elemental compositions of the COOH-MWCNT and Ti-MWCNTSample O/C Ti/O Elemental compositions (mol%) Binding Energy(eV)O C Ti O C TiCOOH-MWCNT 0.19 0.00 16.27 83.73 0.00 532.9 284.5 ------Ti-MWCNT 0.62 0.70 30.04 48.79 48.79 532.9 284.4 459.46Chapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre89Figure 7. 3 shows the high resolution C1s spectra of a COOH-MWCNT and a Ti-MWCNT. The C1s peak could be deconvoluted into two fitting curves: C1 at 284 eV representsC-C or C-H and C2 at 285 eV, which corresponds to C-O. These results are summarized in Table7. 2. After treatment by the coupling agent, MWCNTs exhibited a different intensity in the C1speak. The increase in the relative intensity of C1 peak at 285 eV suggests the formation of moreC-O bonds, which further confirms the presence of long alkyl chains on the MWCNT surfaceafter reactions.Figure 7.3 (a) COOH-MWCNTFigure 7.3 (b) Ti-MWCNTFigure 7. 3 Deconvoluted high resolution C1s spectra of COOH-MWCNT and Ti-MWCNTChapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre90Table 7. 2 Analysis of C1s peaks of the COOH-MWCNT and Ti- MWCNTSampleAnalysis of C1s peaks(%) Binding energy(eV)C1 C2 C1 C2COOH-MWCNT 54.81% 45.19% 284.49 285.36Ti-MWCNT 42.99% 57.01% 284.28 285.067.3.2 Raman spectrumMWCNT defects were analyzed by Raman spectroscopy. Raman spectra of a COOH-MWCNT and a Ti-MWCNT are shown in Figure 7. 4(a) and (b). In both cases, two bands at1310 cm-1 (D band) and 1589 cm-1 (G band) appear. The shift of peak at 1585 cm? 1 can beassigned to the G band which is related to the vibration of sp2-bonded carbon atoms in a twodimensional hexagonal lattice, indicative of the formation of crystalline graphite phase. The Dband, at 1310 cm? 1 is usually attributed to the structural disorder or the presence of amorphouscarbon in the MWCNTs. Thus, the intensity ratio of the D and G bands has been used as anindication of the number of defects on the carbon nanotube structure.Figure 7. 4 Raman spectra of (a) COOH-MWCNT (b) Ti-MWCNTChapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre91As can be seen, there is no significant change of intensity ratio of the D/G bands betweenthe COOH-MWCNT and Ti-MWCNT, which indicates that the functionalization of MWCNTsby a titanate coupling agent does not introduce extra defects onto the MWCNT structure.Table 7. 3 Intensity of Raman peak of MWCNT and Ti-MWCNTCNT Intensity of D peak Intensity of G peak D/G ratioMWCNT 2696.9 2499.5 1.08Ti-MWCNT 663.94 635.80 1.047.3.3 TGAThermogravimetric analysis (TGA) is a method of thermal analysis in which changes inphysical and chemical properties of materials are measured as a function of increasingtemperature. TGA is one of the most straightforward methods to characterize CNT reinforcedpolymer composites. CNTs cannot be easily decomposed at high temperatures, whereas mostpolymers will completely decompose at temperatures over 500 ?C. Therefore, TGA can be usedto provide quantitative information about nanotube functionalization.Figure 7. 5 shows a representative TGA spectra under the nitrogen atmosphere of theCOOH-MWCNTs and the coupling agent treated MWCNTs. The TGA curve of the COOH-MWCNTs is steady, without significant weight loss below 600?C. The slight mass loss (5 wt%)below 600?C was caused by the decomposition of amorphous carbon and free carboxylic andhydroxyl groups on the sidewall of the MWCNTs at elevated temperatures. In comparison, theTGA curve for the Ti-MWCNTs shows a distinct weight loss of 3wt% from the beginning to120?C, followed by a gradual mass loss to 20 wt% as the temperature reached 260?C. A moresignificant mass loss from 20 to 58% was observed between 260?C and 600?C, which can beattributed to the decomposition of the coupling agent covalently attached to the MWCNTs. NoChapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre92distinguishable weight loss was observed below 120?C for coupling agent treated MWCNTs dueto the covalent bonding formed between the coupling agent and the carboxyl groups on thesurface of the MWCNTs.Figure 7. 5 TGA of COOH-MWCNT and Ti-MWCNT7.3.4 Dispersion of MWCNTsThe dispersion of 0.05wt% MWCNTs and Ti-MWCNTs in mineral oil was prepared bysonication for 10 hours. These two solutions were allowed to stand for 30 minutes. It was clearthat MWCNTs without coupling agent functionalization completely precipitated from themineral oil while the dispersion of Ti-MWCNTs was still stable after 30 minutes of precipitation,as shown in Figure 7. 6. This observation demonstrated that the functionalization by a couplingagent significantly improved MWCNT suspension stability in mineral oil. This could beChapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre93explained by the improvement in affinity of MWCNTs with mineral oil through the alkyl chainon the MWCNT surface.Figure 7. 6 CNT suspension in mineral oil7.3.5 Tensile propertiesTi-MWCNT/UHMWPE composite fibres with 1%, 2%, and 3% Ti-MWCNTs wereprepared according to the gel spinning process described in Chapter 4 under the optimumprocessing conditions obtained from the design of experiment detailed in Chapter 5. Tensilestrength of UHMWPE with 3wt% Ti-MWCNTs increased from 3.38 GPa to 3.73 GPa, whichcorresponds to a 10.5% increase. This increase is statistically significant at 95% confidence level.The increase in tensile strength alone could be explained by the improvement in fibre structureand crystallinity by adding MWCNTs. It has been found that with good compatibility betweenCNTs and UHMWPE, CNTs could help UHMWPE molecules to disentangle and align along thefibre axis during post-drawing. However, no significant improvement in modulus and strain wasobserved, which suggested a lack of good load transfer on the interface of CNTs and UHMWPE.Chapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre94Figure 7. 7 (a) Figure 7. 7 (b)Figure 7. 7 (c) Figure 7. 7 (d)Figure 7. 7 Tensile properties of MWCNT/UHMWPE fibre with various CNT contents(a) Strength (b) Modulus (c) Strain (d) U*(1/3)To further improve the tensile properties of UHMWPE fibre, higher Ti-MWCNT contentis needed. A 4wt% MWCNT/UHMWPE mineral oil solution was prepared but MWCNTs werefound starting to precipitate on the bottom of the bottle. Thus, MWCNT concentration inUHMWPE fibre through coupling agent functionalization was limited to 3wt%.Chapter-7 Coupling agent Functionalized MWCNT Reinforced UHMWPE Fibre957.4 ConclusionCompared to mechanically dispersed COOH-MWCNTs, functionalization of MWCNTsusing a coupling agent significantly improved MWCNT dispersion in mineral oil. Anenhancement of the mechanical properties of the UHWMPE composite fibres can be obtained atlow nanotube loading, which indicates a uniform dispersion of MWCNT in the UHMWPEmatrix. As a representative example, a 10.5% improvement in tensile strength was achieved bythe addition of only 3 wt% functionalized nanotubes.However, considering the extraordinary properties of MWCNTs, the reinforcingefficiency of MWCNTs in UHMWPE fibre is still relatively low. Also, it was found that higherconcentrations could not be obtained since the MWCNTs started to precipitate at concentrationshigher than 3wt%. Thus, it is difficult to further improve the mechanical properties of UHMWPEfibre using coupling agent functionalized MWCNTs. The reinforcement in polymer compositedepends on both CNT dispersion in a polymer matrix and the load transfer through the interface.In order to further improve the tensile properties of UHMWPE fibres, the load transfer on theMWCNT and UHWMPE interface must be improved.Chapter-8 Polymer Grafted MWCNT Reinforced UHMWPE Fibre96Chapter 8- Polymer grafted MWCNT reinforced UHMWPEfibre8.1 IntroductionIn recent years, there have been many attempts to use CNTs to reinforce UHMWPE fibre.One of the most successful works came from Ruan?s group [24]. In their study, the strength andmodulus of UHMWPE fibre were increased by 19% and 12%, respectively, when reinforcedwith 5 wt% MWCNTs. However, the solvent used in their study was decahydronaphthalene(decalin), which is not only flammable but also toxic. Efforts have been taken to developMWCNT/UHMWPE composite fibres using mineral oil as a solvent, but the result wasdisappointing as the tensile strength of the composite fibre decreased from 4 GPa to around 1GPa when 5 wt% MWCNTs were added. According to Ruan, the main reason is that theMWCNT dispersion in mineral oil is not as uniform as in decalin due to mineral oil's highviscosity. Large MWCNT clusters were found in composite fibres which prevent MWCNTsfrom exfoliating and thus compromise their significant potential. It has been found by manyresearchers that CNTs dispersion in the matrix and the adhesion between the nanotube and thematrix are the two critical factors in capitalizing the strength of the nanotubes. In Chapter 7,CNT dispersion in mineral oil was improved by functionalizing with a titanium derived couplingagent. A better dispersion of CNTs in a UHMWPE matrix was obtained. However, there is nosignificant improvement in ballistic performance U*(1/3) of composite fibre containing 3wt%functionalized MWCNTs, compared to pure UHMWPE fibre. The low reinforcing efficiency isattributed to the inadequate interfacial load transfer between the CNTs and the polymer matrix,which is crucial for the properties of a CNT reinforced composite. In order to transfer theChapter-8 Polymer Grafted MWCNT Reinforced UHMWPE Fibre97properties of CNTs to the composite material, stronger bonding between CNTs and UHMWPE isneeded. Among others, grafting long polymer chains onto CNTs has been regarded as apromising method for improving the miscibility of CNTs in polymer matrices. Through the highenergy of the covalent bond formed during polymer grafting, a higher interfacial strength andthus a more favorable load transfer between the CNTs and the polymer can be obtained.8.2. Experimental8.2.1 Functionalization and polymer grafting of MWCNTsComposite science has proven that the lack of compatibility between reinforcement andthe polymer matrix inhibits the load transfer from the matrix to reinforcement. This is also truefor carbon nanotube reinforced polymer composite. Great efforts have been made to improve thecompatibility between CNTs and the polymer matrix. In this study, the carboxyl groups on thesidewall of the CNTs enable the grafting of molecules onto the CNT surface so as to increase thecompatibility and realize the load transfer between the matrix and the CNTs [149].100mg of as-received carboxyl group functionalized MWCNTs (COOH-MWCNTs) werereacted with 100ml ethylene diamine to form amino group functionalized MWCNTs (NH2-MWCNT). The COOH-MWCNTs were added into ethylene diamine with the presence of acoupling agent. The mixture was sonicated for 4 hours at 50?C. After the reaction, the mixturewas filtrated to separate the resulting NH2-MWCNTs. These NH2-MWCNTs were washedextensively with methanol until the solution was neutral to get rid of the ethylene diamine. TheNH2-MWCNTs were then dried overnight in a vacuum oven at 80?C.The covalent grafting procedure is shown in Figure 8. 1. 100mg NH2-MWCNTs weredispersed in 10ml toluene and sonicated for 2 hours to break up MWCNT agglomerations.Chapter-8 Polymer Grafted MWCNT Reinforced UHMWPE Fibre98500mg PE-g-MA was dissolved into 10ml toluene under a nitrogen atmosphere at 120?C. NH2-MWCNTs/toluene solution was then mixed with a PE-g-MA/toluene solution and then stirred for24 hours. The grafting reaction was accompanied by the formation of bubbles, indicating therelease of water vapour during imidization reaction. The reaction was complete after theformation of bubbles ceased. The grafted MWCNTs were separated by filtration through a 0.8?m polytetrafluoroethylene (PTFE) membrane and thoroughly washed with methanol severaltimes. The grafted CNTs were finally dried overnight in a vacuum oven at 80?C to remove theresidue solvent. Through this process, polyethylene grafted multi-walled carbon nanotubes (PE-g-MWCNTs) were obtained and were ready to be used as reinforcement for UHMWPE fibre..Figure 8. 1 Polyethylene grafted maleic anhydride functionalized MWCNT8.2.2 Gel spun PE-g-MWCNT reinforced UHMWPE fibreThe PE-g-MWCNTs were used to reinforce UHMWPE in the gel spinning process withvarious concentrations. PE-g-MWCNTs were dispersed in mineral oil by sonication. 3.5wt%UHMWPE was added into MWCNT/mineral oil dispersions and heated up to form the spinningChapter-8 Polymer Grafted MWCNT Reinforced UHMWPE Fibre99gel. UHMWPE fibres with different PE-g-MWCNT concentrations were spun under theconditions described in Chapter 4.8.2.3 CharacterizationTransmission electron microscopy (TEM) and Fourier transform infrared spectroscopy(FTIR) were used to study the polymer grafting process. Single fibre tensile tests were carriedout using the KES-G1 tensile testing machine. The diameter of the filament was measured usingan optical microscope. Detailed characterization methods can be found in Chapter 4.8.3 Results and discussion8.3.1 TEM and SEMThe TEM images of the COOH-MWCNTs as well as PE-g-MWCNTs are shown in Figure8. 2. The diameters of the COOH-MWCNTs measured in Figure 8. 2 (a) are 8-10 nm. Afterpolymer grafting, it was found that the diameter of PE-g-MWCNTs increased to around 20 nm,which is about twice the diameter of COOH-MWCNTs. The thickness of the coated polymer onthe surface of the MWCNTs was about 6 nm, as shown in Figure 8. 2 (b) which is directevidence that the polyethylene chains were grafted efficiently to the nanotube surface.Chapter-8 Polymer Grafted MWCNT Reinforced UHMWPE Fibre100Figure 8. 2 (a) Figure 8. 2 (b)Figure 8. 2 TEM of COOH-MWCNT and PE-g-MWCNT(a) MWCNT before polymer grafting, (b) polymer grafted MWCNTBy controlling the amount of grafting, carbon nanotubes with different grafting thicknesswere obtained. For example, a much thicker polymer grafted carbon nanotube was observedunder SEM, as shown in Figure 8. 3.Figure 8. 3 PE-g-MWCNT with thicker grafting layerChapter-8 Polymer Grafted MWCNT Reinforced UHMWPE Fibre1018.3.2 FTIRThe FTIR spectroscopy of the COOH-MWCNT, NH2-MWCNT and PE-g-MWCNT areshown in Figure 8. 4. The spectrum of COOH-MWCNT presenting a peak at ~1716cm-1 isattributed to the C=O stretch of the carboxylic group. The spectrum of the amide-functionalizedMWCNT samples shows the disappearance of the peak at 1716 cm-1 and a correspondingappearance of the peak with lower frequency at 1635 cm-1 , which can be attributed to the stretchmode of the amide carboxyl C=O. In addition, the peaks at 1415 cm-1, 2849 cm-1 and 2918 cm-1 ,which belong to the stretch and scissor mode of C-H, respectively, further confirm the presenceof the amide functional group. In the spectrum of PE-g-MWCNTs, the presence of a new peak at1260 cm-1 represents the stretch mode of the methyl group from the end of the graftedpolyethylene chain. The two new peaks at 800 cm-1 and 717.8 cm-1 represent the bending modeof aromatic C-H formed from the maleic anhydride group in PE-g-MA. In addition, the peaks at1463 cm-1, 2849 cm-1 and 2918 cm-1 corresponding to the scissor and stretch mode of alkane C-Hbecame stronger due to the grafting of the polyethylene chain. These results clearly indicate thatpolyethylene polymer chains were successfully grafted on the surface of the MWCNTs.Chapter-8 Polymer Grafted MWCNT Reinforced UHMWPE Fibre102Figure 8. 4 FTIR spectrum of COOH-MWNT, NH2-MWCNT and PE-g-MWCNTChapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT1038.3.3 Tensile propertiesThe tensile properties of UHMWPE fibres reinforced by PE-g-MWCNT withconcentrations of 1wt%, 2wt% and 3wt% were tested and compared with pure UHMWPE fibre.As shown in Chapter 5, fibre diameter has a significant influence on tensile properties. Duringthe post-drawing process, it was found that the UHMWPE fibres reinforced by PE-g-MWCNTsshowed better drawability and were easier to be drawn into a smaller diameter compared to pureUHMWPE fibres and Ti-MWCNT reinforced UHMWPE fibres. For example, the minimumdiameter of pure UHMWPE fibres and Ti-MWCNT/UHMWPE fibres was around 20-25 ?mafter post-drawing, which corresponds to a draw ratio of around 45. However, PE-g-MWCNTreinforced UHMWPE fibres showed a much higher maximum draw ratio. For example, 3wt%PE-g-MWCNT/ UHMWPE fibre could be drawn to a draw ratio of 125 with a diameter around15 ?m. The tensile strength, modulus and strain to failure of the 3wt% PE-g-MWCNT/UHMWPE fibre is about 5.0 GPa, 150 GP and 3.7%, respectively, which is almost50% higher than commercial UHMWPE fibres such as Dyneema? and Spectra?. A typical stress-strain curve of the small diameter fibre is shown in Figure 8. 5(a).More interestingly, fibre with a diameter as small as 11 ?m could be obtained when thetemperature in the fourth pass of post-drawing was increased from 140?C to 150?C. As fibretensile properties increased sharply with the decrease in diameter, significantly higher tensileproperties were observed on those small diameter fibres. Tensile strength as high as 7.3 GPawere observed as shown in Figure 8. 5 (b).Chapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT104Figure 8. 5(a) Fibre with diameter of 15?m Figure 8. 5(b) Fibre with diameter of 11?mFigure 8. 5 Typical stress-strain curve of 3% PE-g-MWCNT/UHMWPE fibre(a) diameter of 15?m, (b) diameter of 11?mThe increase in the draw ratio of PE-g-MWCNT/UHMWPE fibre could be attributed tothe improvement in molecular disentanglement during the post-drawing process. It has beendemonstrated by other researchers that the draw ratio of UHMWPE fibre can be improved byadding polyethylene with a smaller molecular weight?e.g. high density polyethylene (HDPE)?as shown in Figure 8.6 [186]. This is because polymers with shorter molecular chains are easierto disentangle during the post-drawing process, thus avoiding the breakage of UHMWPEmolecules due to stress concentration. By adding PE-g-MWCNTs which have compatiblesurface with UHMWPE and thus can be uniformly dispersed in the UHMWPE matrix, a similarphenomenon is expected.Chapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT105Figure 8. 6 Effect of adding HDPE on draw ratio of UHMWPE fibre [186]To analyze the reinforcement of PE-g-MWCNTs on UHMWPE fibre, the tensileproperties of 3wt% PE-g-MWCNT/UHMWPE fibre were compared with pure UHMWPE fibre.As shown above, fibre diameter has great influence on tensile test results, too. Thus, only fibreswith the same diameter (20 to 25 ?m) such as pure UHMWPE fibre were used for thecomparison so as to investigate the influence of CNTs on fibre tensile properties. Twenty fibreswere tested for each sample using the test method described in Chapter 4 and the results areshown in Figure 8. 7.It was found that both tensile strength and modulus of the composite fibres increasedsignificantly as the MWCNT loading increased from 1wt% to 3wt%. At 3wt%, the tensilestrength and modulus of PE-g-MWCNT/UHMWPE fibre reached 4.02 and 125.5 GPa,respectively, which represented a 19% and 23% increase compared to pure UHMWPE fibre. Thefibre's ballistic performance index U*(1/3) calculated from Cunniff?s model also increased to 996.4Chapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT106m/s, which is higher than the requirement for the objective of a 25% weight reduction in bodyarmour.0 1% 2% 3%90100110120130140Young's Modulus (GPa)CNT weight fractionFigure 8. 7 (a) Figure 8. 7 (b)0 1.0% 2.0% 3.0%0.0300.0350.0400.0450.050StrainCNT weight fractionFigure 8. 7 (c) Figure 8. 7 (d)Figure 8. 7 Tensile properties of gel spun UHMWPE and MWCNT/UHMWPE fibre(a) Strength (b) Modulus (c) Strain (d) U*(1/3)In order to compare the reinforcement of MWCNTs to UHMWPE fibre quantitatively, thereinforcing efficiency was calculated using ?rule of mixture? equations which are based on theChapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT107weighted contribution of the fibre and the matrix. The ?rule of mixture? was first developed forfibre reinforced composite and is widely used to predict the modulus and strength of fibrereinforced composites. The simplest form of the ?rule of mixture? is shown in Equation 8. 1,which assumes that load is perfectly transferred between the fibre and the matrix. The density ofcomposite c? can be calculated based on the volume fraction of reinforcing fibre fV andmatrix mV .mmffc VV ??? ?? Equation 8. 1Where the density of reinforcing fibre and matrix f? and m? are known. For unidirectionalcomposites, the Young?s modulus of the composite cE can be calculated using Equation 8. 2.mmffc VEVEE ?? Equation 8. 2In this equation, the Young?s modulus of reinforcing the fibre and the matrix, referredas fE and mE , respectively, are known. However, since CNTs are discontinuous and are notperfectly aligned, a modified version of the ?rule of mixture? is needed. Cox [187] introducedthe concept of the effective Young?s modulus for short fibre embedded into a matrix by definingthe length efficiency factor l? and the orientation factor 0? as the reduction ratio of the filler?sintrinsic Young?s modulus. According to Cox?s modification, the modulus of CNT reinforcedcomposite can be calculated by Equation 8. 3. A similar equation formulated for strength isshown in Equation 8. 4.? ?CNTmCNTCNTlc VEVEE ??? 10?? Equation 8. 3? ?CNTmCNTCNTlc VV ??? 10 ????? Equation 8. 4Chapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT108The CNT length efficiency factor l? can vary between 0 and 1. The orientation factor 0? isequal to 1 for fully aligned CNTs. Equations 8. 5 and 8. 6 can be used to evaluate thecontributions of CNTs to the composite properties if the properties of the matrix, the compositeand the CNT volume fraction are known. According to the equation, the effective modulus,strength can be defined as:0??lCNTeff EE ? Equation 8. 50???? lCNTeff ? Equation 8. 6The orientation of PE-g-MWCNTs in gel spun UHMWPE fibre was analyzed by SEM. Asshown in Figure 8. 8, polymer grafted MWCNTs were very well aligned along the fibre axis inthe UHMWPE fibre. From the evidence of the SEM picture, an assumption that the orientationfactor l? of PE-g-MWCNTs is equal to 1 has been made. Hence, an effective nanotube modulus(or nanotube strength), which incorporates only the length efficiency of the nanotubes, can becalculated as lCNTeff EE ?? and lCNTeff ??? ? .A good orientation of PE-g-MWCNT that existed in gel spun UHMWPE fibre can beexplained by two reasons.(1) The extended-chain conformation of polyethylene matrix.In gel spun UHMWPE fibre, the linear polyethylene polymer chains are perfectly alignedalong fibre axis which are rare in polymer materials. During the post drawing process with drawnratio as high as over 100 times, polymer chains gradually aligned along fibre axis which createopportunity for carbon nanotubes with similar diameter as polyethylene micro-fibrils to changetheir orientation as well. Compared with melt spun textile fibre which consists of partialChapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT109orientated polymer chains, orientation of carbon nanotube is much easier in gel spun UHMWPEfibre.(2) The compatibility between the polymer grafting layer on CNT and polyethylenematrixThe compatibility between the polymer grafting layer on CNT and polyethylene matrixallows the polyethylene chains to interact with PE-g-MWCNTs more readily and nucleate onPE-g-MWCNT surfaces due to epitaxy. For this reason, PE-g-MWNTs are potentially easier toalign parallel to the axis direction compared with normal MWCNTs.5umFigure 8. 8 PE grafted MWCNTs bridge the crack of UHMWPE fibreThe effective modulus and strength of polymer grafted MWCNTs are calculated andsummarized in Table 8. 1. Effective CNT modulus and strength was found to be equal to 1244GPa and 35 GPa, respectively, by using Equations 8. 5 and 8. 6. The experimental measurementsof Young?s modulus and strength by scanning probe microcopy were in the range of 400-1900GPa and 11-63 GPa, as shown in Table. 8. 2 [188]. Thus, it is reasonable to use the ?rule ofmixture? model to calculate reinforcing efficiency.Chapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT110Table 8. 1 Effective modulus and strength of PE-g-MWCNTFibre tensile properties Modulus StrengthProperty increase 22.9% 19.0%From 102.12 to 125.51GPa From 3.375 to 4.02 GPaPE-g-MWCNT effective properties 1244.4 GPa 34.7GPaTable 8. 2 shows the effective modulus and strength of various types of carbon nanotubessummarized from various literature [188]. Compared to effective modulus and strength of CNTsfrom the literature, the polymer grafted MWCNTs obtained in this study showed higher values.This demonstrates that a high reinforcing efficiency to UHMWPE was achieved through polymergrafting on MWCNTs.Table 8. 2 Effective modulus and strength of CNTs from literature [188]Type of CNT de [nm] Eeff [GPa] ?eff [GPa]SWCNT 1.4 nm 971 GPa 126 GPaMWCNT 15 nm 929 GPa 121 GPaTo compare the reinforcing efficiency of polymer grafted MWCNTs with the couplingagent functionalized MWCNTs studied in Chapter 7, the effective modulus and strength ofcoupling agent functionalized MWCNTs were calculated and shown in Table 8. 3. The effectivemodulus and strength of coupling agent functionalized MWCNTs were calculated to be 240 GPaand 21 GPa, respectively, which are much lower than polymer grafted MWCNTs. From thiscomparison, it was demonstrated that a higher reinforcing efficiency to UHMWPE fibres can berealized by MWCNT polymer grafting.Chapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT111Table 8. 3 Effective modulus and strength of Ti-MWCNTFibre tensile properties Modulus StrengthProperty increase 2.76% 10.5%From 102.1 to 104.9 GPa From 3.375 to 3.73 GPaPE-g-MWCNT Effective properties 239.8 GPa 20.7 GPaPE-g-MWCNT/UHMWPE fibres with different MWCNT volume fractions were preparedand the slope of increase in modulus versus CNT volume fraction CNTVddY proposed byColeman was used in this study to quantify the enhancement in Young?s modulus [189]. Byadding 0.67 v% (1 wt%) of PE-g-MWCNTs in to UHMWPE, a dramatic increase in Young?smodulus from 102.1 GPa to 117.3 GPa was observed and CNTVddY was calculated to be as highas 2246 GPa. However, CNTVddY was found to decrease as higher CNT volume fractions wereapplied (as shown in Table 8. 4). The decrease in CNT reinforcement efficiency with volumefraction can be explained by the deterioration of CNT dispersion as volume fraction increased.Researchers have shown that the enhancement in mechanical properties of CNT reinforcedcomposites is strongly dependent upon the level of CNT dispersion. It is thus believed that thedispersion deteriorates when the CNT volume fraction is higher than 0.67%, which slows thatincrease in composite?s tensile properties.Table 8. 4 Effective modulus of PE-g-MWCNT as a function of volume fractionCNT volume fraction 0.67 v% (1 wt%) 1.35 v% (2 wt%) 2.05 v% (3 wt%)CNTVddY (GPa) 2246 1588 1142Chapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT1128.3.4 Morphology and microstructureFigure 8. 9 shows a SEM image of fracture surfaces of the PE-g-MWCNT/UHMWPEfibre after the tensile test. It is evident that nanotubes coated by thick layers of polymer protrudefrom the composite fracture surfaces. Fractured PE-g-MWCNTs were found pulling out from thematrix, which suggests a strong interface bonding exists between the polyethylene grafted carbonnanotube and the UHMWPE matrix.Figure 8. 9 Cross section image of PE-g-MWCNT/UHMWPE fibre8.4 Reinforcing mechanism of polymer grafted CNTsIn this study, polyethylene molecular chains were successfully grafted onto the surface ofMWCNTs to improve the compatibility between MWCNTs and the UHMWPE matrix. PE-g-MWCNTs exhibited higher reinforcing efficiency than coupling agent treated MWCNTs. Theimprovement in the tensile properties arises from several factors, including improvement of CNTdispersion and interface bonding.Chapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT113(1) Improvement in CNT dispersionDue to their high surface energy, CNTs tend to form bundles, as was found in Chapter 5where pristine MWCNTs were used as reinforcement for UHMWPE fibres. With PE molecularchains grafted on the MWCNT surface they have better affinity with the UHMWPE matrix andcan thus be better dispersed in the matrix, allowing a more uniform load distribution. As can beseen in Figure 8. 10, individual nanotubes can be found in the fractured surface of the UHMWPEfibre. Good alignment along fibre axis was observed which is critical for higher reinforcementefficiency.(2) Improvement load transfer on interfaceIt is expected that the PE molecular chains grafted onto MWCNTs could enable a moreefficient load transfer from the matrix to the nanotubes. In Chapter 7, a coupling agent was usedto functionalize MWCNTs and improvement in mechanical properties was found. Compared tothe short alkyl chains grafted on MWCNTs through the coupling agent used in Chapter 7, themolecular chains of polyethylene grafted on the MWCNT surface are much longer. Thus, it isexpected that the load transfer layer, which contains the entanglements of molecules from graftedpolymer and matrix polymer, has stronger bonding and will provide more efficient load transferon the MWCNT and UHMWPE interface.Figure 8. 10 Interface between PE-g-MWCNT and UHMWPEChapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT1148.5 ConclusionIn this study, PE-g-MWCNT/UHMWPE composite fibres with significantly improvedtensile properties have been successfully prepared through the gel spinning process. FTIRcharacterization indicated that polyethylene molecular chains were covalently grafted toMWCNTs. TEM images of PE-g-MWCNT showed that the thickness of the grafted PEmolecules is controllable to as thin as 6 nm. The SEM of the cross section of PE-g-MWCNT/UHMWPE fibre confirmed a strong bonding between MWCNTs and the UHMWPEmatrix on the interface.UHMWPE fibre with improved drawability was observed after adding PE-g-MWCNTs.Maximum post-draw ratio increased from around 45 to 125 and the diameter of fibre decreasedfrom 25 ?m to 15 ?m. 3wt% PE-g-MWCNT/UHMWPE fibre with a diameter of 15 ?m wasobtained under the same post-drawing conditions as pure UHMWPE fibre and showed tensilestrength and modulus of 5.0 GPa and 150 GPa, respectively. By optimizing post-drawingconditions, fibre with a diameter as small as 11?m can be obtained. The tensile strength of theresultant fibre is over 7 GPa, which is the highest tensile strength of UHMWPE fibre from gelspinning process that has been reported.To investigate the reinforcement of CNTs on UHMWPE fibre, PE-g-MWCNT/UHMWPE and UHMWPE fibres with the same diameter were tested and compared.The Young's modulus and tensile strength of composite fibre kept increasing up to at least 3 wt%and an enhancement of 23% and 19% was observed. The effective Young?s modulus andstrength was calculated based on the ?rule of mixture?. Compared to the effective modulus andstrength calculated from coupling agent functionalized MWCNTs (as shown in Chapter 7), PE-g-MWCNTs showed much higher properties. The significantly higher reinforcing efficiencyChapter-8 Strong UHMWPE Fibre Reinforced by Polymer Grafted MWCNT115suggests that an improved dispersion and load transfer achieved in the PE-g-MWCNT/UHMWPE composite fibre and the surface modification on MWCNTs by polymer grafting is asuccessful method for producing strong MWCNT/UHMWPE composite fibre. The fibre'sballistic performance index U*(1/3) also increased to 996.4 m/s. According to theoreticalcalculation, U*(1/3) of only 862 m/s or higher is needed to reach the objective of a 25% weightreduction in body armour. Thus, reinforcing UHMWPE fibre with PE-g-MWCNTs is a feasiblemethod to reach the goal of the CRTI project.Chapter 9- Summary and Suggestion For Future Work116Chapter 9- Summary and suggestion for future work9.1 Major achievements of the present thesisThis thesis is part of the research project for developing strong fibre material with theobjective to reduce the weight of IED body armour by 25%. In order to reduce the weight ofbody armour, high performance fibres with higher tensile properties are needed. Motivated bythis urgent need, an extensive study has been conducted to develop and optimize new processesto improve CNT dispersion and compatibility between CNTs and the UHMWPE matrix.In the beginning of the study, the feasibility of using pristine CNTs to reinforceUHMWPE fibre was discussed. However, no statistically significant improvement in tensileproperties was found after adding up to 1.5% MWCNTs. These results demonstrated thenecessity of chemical functionalization to improve MWCNT dispersion in the UHMWPE matrix.A coupling agent (which is one of the most commonly used substances to improve the capabilitybetween fillers and the polymer matrix in the composite industry) was applied to CNTs with asimilar purpose. The UHMWPE fibre reinforced by a 3% titanate coupling agent treatedMWCNT showed a 10.5% increase in tensile strength. However there is no improvement in theballistic performance index U*(1/3 compared to pure UHMWPE fibre.An innovative polymer grafting MWCNT was used to reinforce UHMWPE fibre for thefirst time, with the objective to further improve interface compatibility and reinforcing efficiency.It was found that the drawability of UHMWPE fibre was significantly improved by incorporatingPE-g-MWCNTs. Compared to pure UHMWPE fibres with a diameter of 25?m, fibres with amuch smaller diameter of 15 ?m can be obtained. The tensile strength, modulus and strain of3wt% PE-g-MWCNT/UHMWPE fibre is 5 GPa, 150 GPa and 3.7%, respectively. Since fibreChapter 9- Summary and Suggestion For Future Work117diameter has a strong effect on fibre tensile properties, only PE-g-MWCNT/UHMWPE fibrewith the same diameter as pure UHMWPE fibre were compared to investigate the reinforcingeffect of CNTs. At 3% load level, the tensile strength, modulus and strain increased by 19%,23% and 9%. The reinforcing efficiency of MWCNTs was significantly improved by polymergrafting, compared to pristine MWCNTs and coupling agent functionalized MWCNTs. Theballistic performance of the 3wt% PE-g-MWCNT/UHMWPE reached around 966.4 m/s, whichis much higher than required to reduce the weight of body armour by 25%, according toCunniff?s model. Thus, the objective of the CRTI project was successfully achieved. The majorachievements of this thesis are summarized in detail in the following:9.1.1 Effect of spinning conditions on fibre tensile properties and morphologyUHMWPE fibre is produced through the gel spinning process due to its extremely highpolymer chains. It was found that the fibre properties and morphology were very sensitive tospinning conditions. It is therefore very important to discuss the effects of various spinningparameters on the tensile properties of gel spun fibre and to find the optimum spinningconditions for spinning conducted using laboratory scale spinning systems.In this study, the effect of the three most important spinning parameters on fibreproperties were studied: polymer concentration, spinning temperature and winding speed. Theoptimum spinning conditions were found to be with a polymer concentration of 3.5%, a spinningtemperature of 150?C and a winding speed of 5 FPM. Under these conditions, as-spunUHMWPE fibre showed an open shish kekab structure which was found to have adequatestrength to endure the tension during the post-drawing process and loose enough to be easilyunfolded and straightened. The optimization process is very important because it helped tounderstand the relationship between UHMWPE tensile properties and morphologies. Also,Chapter 9- Summary and Suggestion For Future Work118finding the optimum spinning conditions builds up a good calibration foundation for futurespinning experiments.9.1.2 Feasibility of reinforcing UHMWPE fibre by pristine nanotubesCompared to chemical functionalization, physical methods are preferred in someapplications because they will not cause the disruption of CNT structure. In addition, mechanicalmethods are relatively easy to apply and the cost is usually lower than chemical treatment. SinceUHMWPE fibre is a widely used high performance fibre, the treatment of CNTs asreinforcement for UHMWPE fibre should be easy to apply in industry, which is why themechanical method was taken into consideration in the first place. In this study, various widelyused mechanical methods for dispersing particles were studied including sonication, ball milling,and microfluidizing. In addition, an innovative water spray method was studied. Results showedthat ultra-sonciation was not a sufficient method for dispersing pristine CNTs, as largeragglomerates found in fibres and tensile properties dropped after incorporating these CNTs. Ballmilling is more effective in decreasing the size of CNTs agglomerates than ultra-sonciation astensile properties of the MWCNT/UHMWPE showed a slight increase after adding 1%MWCNTs. However, statistical analysis showed that these improvements in tensile propertieswere not significant. Other methods such as microfluidizing and solvent assistant dispersionwere similarly unable to obtain uniform and stable dispersion. The conclusion is therefore thatchemical functionalization of carbon nanotubes is necessary for better dispersion and higherreinforcing efficiency.Chapter 9- Summary and Suggestion For Future Work1199.1.3 Reinforced UHMWPE fibre by coupling agent functionalized CNTsCNT dispersion and reinforcement in UHMWPE fibre was improved after beingfunctionalized by a coupling agent. MWCNTs with a concentration as high as 3wt% can bedispersed in UHMWPE. Tensile strength of MWCNT/UHMWPE composite fibre increased by10.5% by adding 3w% Ti-MWCNTs, compared to using pure UHMWPE fibre. However, noimprovement in Young?s modulus and strain was observed, which led to little improvement inU*(1/3).9.1.4 Polyethylene grafted CNTs reinforced gel spun UHMWPEThis is the first study focusing on using a polyethylene grafted CNT to reinforce gel spunUHMWPE fibre. Through grafting polyethylene molecular chains onto CNTs, the CNTcompatibility in the UHMWPE matrix greatly improved. Compared to pristine CNTs (whichformed agglomerates the size of micrometers), individual nanotubes were found after polymergrafting, indicating a uniform dispersion. PE-g-MWCNT reinforced UHMWPE fibres werefound to have tensile strength, modulus and strain approximately 19%, 23% and 9% higher,respectively, than pure UHMWPE fibre at 3wt% CNT loading. The effective modulus andstrength of polyethylene grafted MWCNTs was calculated from the ?rule of mixture? and wasfound to be 419% and 667.5% higher than coupling agent functionalized MWCNTs, whichdemonstrated a significant improvement in reinforcing efficiency by polymer grafting. Theballistic performance index U*(1/3)was calculated to be 966.4 m/s which is much higher than whatis needed to reach the goal of reducing the weight of body armour by 25%.More significantly, the maximum draw ratio of UHMWPE fibre was found to greatlyincrease from 45 to 125 by incorporating 3wt% PE-g-MWCNTs. Fibre with tensile strength andmodulus of 5 GPa and 150 GPa, respectively, were obtained under the same spinning and post-Chapter 9- Summary and Suggestion For Future Work120drawing conditions as pure UHMWPE fibre. By further improving the post-drawing conditions,fibres with a diameter as small as 11?m were obtained with tensile strength over 7 GPa, which iswithin the same level of the strongest UHMWPE ever developed by researchers.Commercially available UHMWPE fibres are produced by two different processes thatuse decalin and mineral oil, respectively, as solvents. It is therefore of great interest to developprocesses to reinforce UHMWPE fibre through CNTs using the two solvents. Ruan?s [24] groupproved that acid purified MWCNTs significantly improved the mechanical properties ofUHMWPE fibre when decalin was used as a solvent. However, similar results could not beobtained when mineral oil was used as a solvent. Due to the high viscosity, it is more difficult todebundle and disperse CNTs in mineral oil. The current study developed an innovative processthat could significantly improve CNT dispersion and load transfer on the CNT/UHMWPEinterface. Significant improvement in the tensile properties of UHMWPE fibre was observed byreinforcement with PE-g-MWCNTs using mineral oil as a solvent. The methods introduced inthis thesis can be easily adopted on an industry scale production of UHMWPE fibre withimproved tensile properties.9.2 Suggestion for future work9.2.1 Characterization of CNT alignment in gel spun UHMWPE fibreIn this study, preliminary evidence of CNT alignment was found in UHMWPE fibresthrough SEM images. However, it will be of great interest to study how the alignment of CNTsin UHMWPE was obtained during the post-drawing process using characterization methods suchas polarized Raman spectroscopy.Chapter 9- Summary and Suggestion For Future Work1219.2.2 Quantitatively study load transfer between CNT and UHMWPEIn Chapter 8 of this study, it was found that by grafting MWCNTs with PE molecularchains the tensile properties of the composite UHMWPE fibre significantly improved. Throughthe SEM analysis of the fractured surface of the composite fibre it was observed that strongadhesion was achieved on the interface. In the future, a quantitative measurement of the loadtransfer on the CNT and UHMWPE interfaces could give more direct information forimprovement in compatibility between CNTs and UHMWPE by assessing different surfacefunctionalization properties.9.2.3 Gel spun CNT reinforced UHMWPE fibre using commercial production lineAll the experimental results presented in this thesis are based on a laboratory scaleextruder and homemade post-drawing equipment. Gel spinning is a complex spinning processthat depends a great deal on the processing equipment. Thus, more efforts are needed toreproduce the results on a production line. It is expected that with better processing control,composite fibres with heightened properties could be obtained by applying a similar method on alarge-scale commercial production line.124References1. Gaura, E.I., Kemp, J., Thake, C.D., Increasing Safety of Bomb Disposal Missions: ABody Sensor Network Approach. IEEE Transactions on Systems Man and CyberneticsPart C (Applications and Reviews), 2009. 39(6).2. Cunniff, P.M., Dimensionless Parameters for Optimization of Textile-Based Body ArmorSystems. in Proceeding of the 18th International Symposium on Ballistics, 1999.3. Auerbach, M.A., Vetter, E. and Sikkema, D.J., High Performance ? M5 ? Fiber forBallistics/Structural Composites, 23rd Army Science Conference, 20044. Dupont.Inc, Kevlar@ Aramid fiber Technical Guide, Dupont.Inc, Editor.5. Honeywell Spectra? fiber 900 high-strength, light-weight polyethylene fiber ProductInformation Sheet.6. Mera, H. and Takata, T., High-Performance Fibers, in Ullmann's Encyclopedia ofIndustrial Chemistry2000, Wiley-VCH Verlag GmbH & Co. KGaA.7. Yu, M.-F., et al., Strength and Breaking Mechanism of Multiwalled Carbon NanotubesUnder Tensile Load. Science, 2000. 287(5453): p. 637-640.8. Chae, H.G., M.L. Minus, A. Rasheed, and S. Kumar, Stabilization and carbonization ofgel spun polyacrylonitrile/single wall carbon nanotube composite fibers. Polymer. 48(13):p. 3781-3789, (2007)9. Baji, A., et al., Mechanical behavior of self-assembled carbon nanotube reinforced nylon6,6 fibers. Composites Science and Technology, 2010. 70(9): p. 1401-1409.10. V?ctor J., et al., Nanocomposites based on plasma-polymerized carbon nanotubes andNylon-6, Polymer Journal (2012) 44, 952?958;12511. Mahfuz, H., et al., Enhancement of strength and stiffness of Nylon 6 filaments throughcarbon nanotubes reinforcement. Applied Physics Letters, 2006. 88(8): p. 083119-083119-3.12. Jeong, J.S., et al., Fabrication of MWNTs/nylon conductive composite nanofibers byelectrospinning. Diamond and Related Materials, 2006. 15(11?12): p. 1839-1843.13. Saeed, K., et al., In situ Polymerization of Multi-Walled Carbon Nanotube/Nylon-6Nanocomposites and Their Electrospun Nanofibers. Nanoscale Research Letters, 2009.4(1): p. 39-46.14. Saeed, K. and S.-Y. Park, Preparation of multiwalled carbon nanotube/nylon-6nanocomposites by in situ polymerization. Journal of Applied Polymer Science, 2007.106(6): p. 3729-3735.15. Chen, G.-X., et al., Multi-walled carbon nanotubes reinforced nylon 6 composites.Polymer, 2006. 47(13): p. 4760-4767.16. Jose, M.V., et al., Morphology and mechanical properties of Nylon 6/MWNT nanofibers.Polymer, 2007. 48(4): p. 1096-1104.17. Liu, L., L. Zhang, and J. Lua, Branched carbon nanotube reinforcements for improvedstrength of polyethylene nanocomposites. Applied Physics Letters, 2012. 101(16): p.161907-161907-5.18. Merlyn, X.P., et al., Medium density polyethylene composites with functionalized carbonnanotubes. Nanotechnology, 2009. 20(19): p. 195602.19. Bakshi, S.R., J.E. Tercero, and A. Agarwal, Synthesis and characterization of multiwalledcarbon nanotube reinforced ultra high molecular weight polyethylene composite by126electrostatic spraying technique. Composites Part A: Applied Science and Manufacturing,2007. 38(12): p. 2493-2499.20. Gorrasi, G., et al., Structure?property relationships on uniaxially oriented carbonnanotube/polyethylene composites. Polymer, 2011. 52(4): p. 1124-1132.21. Lisunova, M.O., et al., Percolation behaviour of ultrahigh molecular weightpolyethylene/multi-walled carbon nanotubes composites. European Polymer Journal,2007. 43(3): p. 949-958.22. Mezghani, K., et al., Influence of carbon nanotube (CNT) on the mechanical properties ofLLDPE/CNT nanocomposite fibers. Materials Letters, 2011. 65(23?24): p. 3633-3635.23. Morcom, M., K. Atkinson, and G.P. Simon, The effect of carbon nanotube properties onthe degree of dispersion and reinforcement of high density polyethylene. Polymer, 2010.51(15): p. 3540-3550.24. Ruan, S., P. Gao, and T.X. Yu, Ultra-strong gel-spun UHMWPE fibers reinforced usingmultiwalled carbon nanotubes. Polymer, 2006. 47(5): p. 1604-1611.25. Sulong, A.B., et al., Process optimization of melt spinning and mechanical strengthenhancement of functionalized multi-walled carbon nanotubes reinforcing polyethylenefibers. Composites Part B: Engineering, 2011. 42(1): p. 11-17.26. Wang, Y., et al., Study on the preparation and characterization of ultra-high molecularweight polyethylene?carbon nanotubes composite fiber. Composites Science andTechnology, 2005. 65(5): p. 793-797.27. Zhang, Q., et al., Low percolation threshold in single-walled carbon nanotube/highdensity polyethylene composites prepared by melt processing technique. Carbon, 2006.44(4): p. 778-785.12728. Chae, H.G., et al., Carbon nanotube reinforced small diameter polyacrylonitrile basedcarbon fiber. Composites Science and Technology, 2009. 69(3?4): p. 406-413.29. Chae, H.G., et al., Stabilization and carbonization of gel spun polyacrylonitrile/singlewall carbon nanotube composite fibers. Polymer, 2007. 48(13): p. 3781-3789.30. Chae, H.G., et al., A comparison of reinforcement efficiency of various types of carbonnanotubes in polyacrylonitrile fiber. Polymer, 2005. 46(24): p. 10925-10935.31. Liu, Y., et al, Gel-spun carbon nanotubes/polyacrylonitrile composite fibers. Part I: Effectof carbon nanotubes on stabilization. Carbon, 2011. 49(13): p. 4466-4476.32. Zhang, Q., et al., Hierarchical composites of carbon nanotubes on carbon fiber: Influenceof growth condition on fiber tensile properties. Composites Science and Technology,2009. 69(5): p. 594-601.33. Deng, L., et al., Characterization of the adhesion of single-walled carbon nanotubes inpoly(p-phenylene terephthalamide) composite fibres. Polymer, 2010. 51(9): p. 2033-2039.34. O'Connor, I., et al., High-Strength, High-Toughness Composite Fibers by SwellingKevlar in Nanotube Suspensions. Small, 2009. 5(4): p. 466-469.35. J.W.S.Hearle, High Performance Fibres, 2002: CRC Press. p. 329.36. Green, A.K. and L.N. Phillips, Non-aerospace applications of carbon and other high-performance fibre materials and their hybrids. International Journal of Materials inEngineering Applications, 1978. 1(2): p. 59-65.37. Corinaldesi, V. and G. Moriconi, Mechanical and thermal evaluation of Ultra HighPerformance Fiber Reinforced Concretes for engineering applications. Construction andBuilding Materials, 2012. 26(1): p. 289-294.12838. Park, S.H., et al., Tensile behavior of Ultra High Performance Hybrid Fiber ReinforcedConcrete. Cement and Concrete Composites, 2012. 34(2): p. 172-184.39. Curtis, P.T. and S.M. Bishop, An assessment of the potential of woven carbon fibre-reinforced plastics for high performance applications. Composites, 1984. 15(4): p. 259-265.40. Farhat, F.A., et al., High performance fibre-reinforced cementitious composite(CARDIFRC) ? Performance and application to retrofitting. Engineering FractureMechanics, 2007. 74(1?2): p. 151-167.41. Kumar, S., Structure and properties of high performance polymeric and carbon fibers ?an overview. Composites, 1989. 20(6): p. 602-603.42. Cheng, S.Z.D., et al., A high-performance aromatic polyimide fibre: 1. Structure,properties and mechanical-history dependence. Polymer, 1991. 32(10): p. 1803-1810.43. Huang, B., P.A. Tucker, and J.A. Cuculo, High performance poly(ethylene terephthalate)fibre properties achieved via high speed spinning with a modified liquid isothermal bathprocess. Polymer, 1997. 38(5): p. 1101-1110.44. Lim, J.G., B.S. Gupta, and W. George, The potential for high performance fiber fromnylon 6. Progress in Polymer Science, 1989. 14(6): p. 763-809.45. Mackley, M.R. and S. Solbai, Swell drawing: A new method of manufacturing highperformance polyethylene structures. Polymer, 1987. 28(7): p. 1115-1120.46. Perepelkin, K.E., Theoretical and maximum attainable strength of various uniaxiallyoriented polymers (film, fibers). Polymer Mechanics, 1966. 2(6): p. 536-545.47. Crist, B., The Ultimate Strength and Stiffness of Polymers. Annual Review of MaterialsScience, 1995. 25(1): p. 295-323.12948. Smith, K.J., The breaking strength of perfect polymer fibers. Polymer Engineering &Science, 1990. 30(8): p. 437-443.49. Wright, W.W., Kevlar aramid fiber. H. H. Yang. John Wiley & Sons, Chichester, 1993.pp. 200, price ?39.95. ISBN 0-471-93765-7. Polymer International, 1994. 33(4): p. 438-439.50. Jiang, H., R.K. Eby, and W.W. Adams, Rigid-Rod Polymers, in Encyclopedia of PolymerScience and Technology2002, John Wiley & Sons, Inc.51. Han Gi Chae, S.K., Rigid-Rod Polymeric Fibers. Journal of Applied Polymer Science,2008. 100(1).52. Penning, J.P., et al., On the theoretical strength of gelspun/hotdrawn ultra-high molecularweight polyethylene fibres. Polymer Bulletin, 1990. 23(3): p. 347-352.53. Penning, J.P., A.A. Vries, and A.J. Pennings, The effect of fibre diameter on the drawingbehaviour of gel-spun ultra-high molecular weight polyethylene fibres. Polymer Bulletin,1993. 31(2): p. 243-248.54. Pennings, A.J., et al., High-speed gel-spinning of ultra-high molecular weightpolyethylene. Polymer Bulletin, 1986. 16(2-3): p. 167-174.55. Bastiaansen, C.W.M., Tensile strength of solution-spun, ultra-drawn ultra-high molecularweight polyethylene fibres: 1. Influence of fibre diameter. Polymer, 1992. 33(8): p. 1649-1652.56. Bastiaansen, C.W.M., et al,. Tensile strength of solution-spun, ultra-drawn ultra-highmolecular weight polyethylene fibres: 2. Influence of propylene comonomer content.Polymer, 1992. 33(8): p. 1653-1655.13057. Ulcay, Y., B. Pourdeyhimi, and I. Block, Mechanical performance of ultra-high-strengthpolyethylene fibers. Composites Engineering, 1991. 1(3): p. 145-156.58. Wang, J. and K.J. Smith Jr, The breaking strength of ultra-high molecular weightpolyethylene fibers. Polymer, 1999. 40(26): p. 7261-7274.59. Zhirnov, N.I., Y.G. Koryak-Doronenko, and G.M. Bartenev, Estimation of the theoreticalstrength of orientated polyethylene. Polymer Science U.S.S.R., 1969. 11(6): p. 1396-1401.60. Zachariades, A.E., W.T. Mead, and R.S. Porter, Recent developments in ultraorientationof polyethylene by solid-state extrusion. Chemical Reviews, 1980. 80(4): p. 351-364.61. N. J. Capiati and R. S. Porter, J. Polym. Phys. Ed., 13, 1177 (1975).62. Zachariades, A.E., M.P.C. Watts, and R.S. Porter, Solid state extrusion of ultra highmolecular weight polyethylene. Processing and properties. Polymer Engineering &Science, 1980. 20(8): p. 555-561.63. Bassett, D.C., Chain-extended polyethylene in context: a review. Polymer, 1976. 17(6): p.460-470.64. Zwijnenburg, A., et al., Longitudinal growth of polymer crystals from flowing solutionsV.: Structure and morphology of fibrillar polyethylene crystals. Colloid and PolymerScience, 1978. 256(8): p. 729-740.65. Zwijnenburg, A., Longitudinal growth, morphology and physical properties of fibrillarpolyethylene crystals, 1978, University of Groningen.66. Smook, J., et al., Ultra-high strength polyethylene by hot drawing of surface growthfibers. Polymer Bulletin, 1980. 2(5): p. 293-300.67. Somani, R.H., et al., Flow-induced shish-kebab precursor structures in entangled polymermelts. Polymer, 2005. 46(20): p. 8587-8623.13168. Postema, A.R., et al., Effect of chlorosulfonation of ultra-high strength polyethylenefibres on mechanical properties and bonding with gypsum plaster. Polymer Bulletin,1986. 16(1): p. 1-6.69. Smook, J., J.C.M. Torfs, and A.J. Pennings, Hot drawing of surface growth polyethylenefibers, 2. Effect of drawing temperature and elongational viscosity. Die MakromolekulareChemie, 1981. 182(11): p. 3351-3359.70. Gao, P. and M.R. Mackley, Effect of presolvent loading on the ultimate drawability ofultra-high molecular weight polyethylene. Polymer, 1991. 32(17): p. 3136-3139.71. Jian, T., et al., Spinning and drawing properties of ultrahigh-molecular-weightpolyethylene fibers prepared at varying concentrations and temperatures. PolymerEngineering & Science, 2003. 43(11): p. 1765-1777.72. Yeh, J.-T., et al., Investigation of the drawing mechanism of UHMWPE fibers. Journal ofMaterials Science, 2008. 43(14): p. 4892-4900.73. Xiao, M., et al., Effect of UHMWPE concentration on the extracting, drawing, andcrystallizing properties of gel fibers. Journal of Materials Science, 2011. 46(17): p. 5690-5697.74. Sheiko, S.S., et al., Nanofibrillar surface structures of gel-drawn UHMW-polyethylenetapes. Acta Polymerica, 1996. 47(11-12): p. 492-497.75. Peterlin, A., Plastic deformation of polymers with fibrous structure. Colloid and PolymerScience, 1975. 253(10): p. 809-823.76. Dijkstra, D.J. and A.J. Pennings, The role of taut tie molecules on the mechanicalproperties of gel-spun UHMWPE fibres. Polymer Bulletin, 1988. 19(1): p. 73-80.77. Iijima, S., Helical microtubules of graphitic carbon. Nature, 1991. 354(6348): p. 56-58.13278. Liu, M. and J.M. Cowley, Structures of carbon nanotubes studied by HRTEM andnanodiffraction. Ultramicroscopy, 1994. 53(4): p. 333-342.79. Reilly, R.M., Carbon Nanotubes: Potential Benefits and Risks of Nanotechnology inNuclear Medicine. The journal of Nuclear Medicine, 2007. 48(7): p. 4.80. Wong, E.W., P.E. Sheehan, and C.M. Lieber, Nanobeam Mechanics: Elasticity, Strength,and Toughness of Nanorods and Nanotubes. Science, 1997. 277(5334): p. 1971-1975.81. J.-P. Salvetat, J.-M.B., N.H. Thomson, A.J. Kulik, L. Forr?o, W. Benoit, L. Zuppiroli,Mechanical properties of carbon nanotubes. Appl. Phys. A, 1999. 69: p. 6.82. Salvetat, J.-P., et al., Elastic and Shear Moduli of Single-Walled Carbon Nanotube Ropes.Physical Review Letters, 1999. 82(5): p. 944-947.83. Atkinson, K.R., et al., Multifunctional carbon nanotube yarns and transparent sheets:Fabrication, properties, and applications. Physica B: Condensed Matter, 2007. 394(2): p.339-343.84. Bogdanovich, A.E. and P.D. Bradford, Carbon nanotube yarn and 3-D braid composites.Part I: Tensile testing and mechanical properties analysis. Composites Part A: AppliedScience and Manufacturing, 2010. 41(2): p. 230-237.85. Fugetsu, B., et al., The production of soft, durable, and electrically conductive polyestermultifilament yarns by dye-printing them with carbon nanotubes. Carbon, 2009. 47(2): p.527-530.86. Ghemes, A., et al., Fabrication and mechanical properties of carbon nanotube yarns spunfrom ultra-long multi-walled carbon nanotube arrays. Carbon, 2012. 50(12): p. 4579-4587.13387. Jakubinek, M.B., et al., Thermal and electrical conductivity of array-spun multi-walledcarbon nanotube yarns. Carbon, 2012. 50(1): p. 244-248.88. Miao, M., Electrical conductivity of pure carbon nanotube yarns. Carbon, 2011. 49(12): p.3755-3761.89. Miao, M., Production, structure and properties of twistless carbon nanotube yarns with ahigh density sheath. Carbon, 2012. 50(13): p. 4973-4983.90. Min, J., et al., High performance carbon nanotube spun yarns from a crosslinked network.Carbon, 2013. 52(0): p. 520-527.91. Tran, C.D., et al., Improving the tensile strength of carbon nanotube spun yarns using amodified spinning process. Carbon, 2009. 47(11): p. 2662-2670.92. Zhang, Q., et al., Dry spinning yarns from vertically aligned carbon nanotube arraysproduced by an improved floating catalyst chemical vapor deposition method. Carbon,2010. 48(10): p. 2855-2861.93. Zhang, W., et al., Self-assembly of single walled carbon nanotubes onto cotton to makeconductive yarn. Particuology, 2012. 10(4): p. 517-521.94. Vigolo, B., et al., Macroscopic Fibers and Ribbons of Oriented Carbon Nanotubes.Science, 2000. 290(5495): p. 1331-1334.95. Xue, P., et al., Electrically conductive yarns based on PVA/carbon nanotubes. CompositeStructures, 2007. 78(2): p. 271-277.96. Miao, M., et al., Effect of gamma-irradiation on the mechanical properties of carbonnanotube yarns. Carbon, 2011. 49(14): p. 4940-4947.97. Cai, J.Y., et al., An improved method for functionalisation of carbon nanotube spun yarnswith aryldiazonium compounds. Carbon, 2012. 50(12): p. 4655-4662.13498. Dierking, I., et al., Aligning and Reorienting Carbon Nanotubes with Nematic LiquidCrystals. Advanced Materials, 2004. 16(11): p. 865-869.99. Zakri, C. and P. Poulin, Phase behavior of nanotube suspensions: from attraction inducedpercolation to liquid crystalline phases. Journal of Materials Chemistry, 2006. 16(42): p.4095-4098.100. Zhang, S. and S. Kumar, Carbon Nanotubes as Liquid Crystals. Small, 2008. 4(9): p.1270-1283.101. Ericson, L.M., et al., Macroscopic, Neat, Single-Walled Carbon Nanotube Fibers.Science, 2004. 305(5689): p. 1447-1450.102. Song, W., I.A. Kinloch, and A.H. Windle, Nematic Liquid Crystallinity of MultiwallCarbon Nanotubes. Science, 2003. 302(5649): p. 1363.103. Motta, M.S., Moisala, A., Kinloch, I. A. & Windle, A. H.,, The role of sulphur in thesynthesis of carbon nanotubes by chemical vapour deposition at high temperatures.Journal of Nanoscien and Nanotechnology, 2008. 8.104. Vilatela, J.J., et al., Structure of and stress transfer in fibres spun from carbon nanotubesproduced by chemical vapour deposition. Carbon, 2011. 49(13): p. 4149-4158.105. Davis, D.C., et al., A strategy for improving mechanical properties of a fiber reinforcedepoxy composite using functionalized carbon nanotubes. Composites Science andTechnology, 2011. 71(8): p. 1089-1097.106. Shen, Z., et al., The effects of carbon nanotubes on mechanical and thermal properties ofwoven glass fibre reinforced polyamide-6 nanocomposites. Composites Science andTechnology, 2009. 69(2): p. 239-244.135107. Rahmanian, S., et al., Carbon and glass hierarchical fibers: Influence of carbon nanotubeson tensile, flexural and impact properties of short fiber reinforced composites. Materials& Design, 2013. 43(0): p. 10-16.108. Abdullah, S.A., A. Iqbal, and L. Frormann, Melt mixing of carbon fibers and carbonnanotubes incorporated polyurethanes. Journal of Applied Polymer Science, 2008. 110(1):p. 196-202.109. Andrews, R., et al., Nanotube composite carbon fibers. Applied Physics Letters, 1999.75(9): p. 1329-1331.110. Sreekumar, T.V., et al., Polyacrylonitrile Single-Walled Carbon Nanotube CompositeFibers. Advanced Materials, 2004. 16(1): p. 58-61.111. Zhang, Q., D.R. Lippits, and S. Rastogi, Dispersion and Rheological Aspects of SWNTsin Ultrahigh Molecular Weight Polyethylene. Macromolecules, 2005. 39(2): p. 658-666.112. Gao, J.-F., et al., CNTs/ UHMWPE composites with a two-dimensional conductivenetwork. Materials Letters, 2008. 62(20): p. 3530-3532.113. Mart?nez-Morlanes, M.J., et al., Effects of gamma-irradiation on UHMWPE/MWNTnanocomposites. Composites Science and Technology, 2011. 71(3): p. 282-288.114. Wood, W.J., R.G. Maguire, and W.H. Zhong, Improved wear and mechanical propertiesof UHMWPE?carbon nanofiber composites through an optimized paraffin-assisted melt-mixing process. Composites Part B: Engineering, 2011. 42(3): p. 584-591.115. Maksimkin, A.V., et al., Ultra-high molecular weight polyethylene reinforced with multi-walled carbon nanotubes: Fabrication method and properties. Journal of Alloys andCompounds, 2012. 536, Supplement 1(0): p. S538-S540.136116. Liu, Y. and S.K. Sinha, Wear performances and wear mechanism study of bulkUHMWPE composites with nacre and CNT fillers and PFPE overcoat. Wear, (0).117. Samad, M.A. and S.K. Sinha, Mechanical, thermal and tribological characterization of aUHMWPE film reinforced with carbon nanotubes coated on steel. TribologyInternational, 2011. 44(12): p. 1932-1941.118. Wannasri, S., et al., Increasing wear resistance of UHMWPE by mechanical activationand chemical modification combined with addition of nanofibers. Procedia Engineering,2009. 1(1): p. 67-70.119. Xu, G., et al., Carbon nanotubes induced nonisothermal crystallization of ultrahighmolecular weight polyethylene with reduced chain entanglements. Materials Letters,2012. 89(0): p. 272-275.120. Xue, Y., et al., Tribological behaviour of UHMWPE/HDPE blends reinforced with multi-wall carbon nanotubes. Polymer Testing, 2006. 25(2): p. 221-229.121. Ko, F., et al., Electrospinning of Continuous Carbon Nanotube-Filled Nanofiber Yarns.Advanced Materials, 2003. 15(14): p. 1161-1165.122. Haihui Ye, H.L., Nick Titchenal, Yury Gogotsi and Frank Ko, Reinforcement and rupturebehavior of carbon nanotubes?polymer nanofibers. APPLIED PHYSICS LETTERS 2004.85(10): p. 3.123. Chronakis, I.S., Novel nanocomposites and nanoceramics based on polymer nanofibersusing electrospinning process?A review. Journal of Materials Processing Technology,2005. 167(2?3): p. 283-293.124. O'Connell, M.J., et al., Reversible water-solubilization of single-walled carbon nanotubesby polymer wrapping. Chemical Physics Letters, 2001. 342(3?4): p. 265-271.137125. Duan, W.H., Q. Wang, and F. Collins, Dispersion of carbon nanotubes with SDSsurfactants: a study from a binding energy perspective. Chemical Science, 2011. 2(7): p.1407-1413.126. Alpatova, A.L., et al., Single-walled carbon nanotubes dispersed in aqueous media vianon-covalent functionalization: Effect of dispersant on the stability, cytotoxicity, andepigenetic toxicity of nanotube suspensions. Water Research, 2010. 44(2): p. 505-520.127. Bai, Y., et al., Aqueous dispersion of surfactant-modified multiwalled carbon nanotubesand their application as an antibacterial agent. Carbon, 2011. 49(11): p. 3663-3671.128. Bystrzejewski, M., et al., Dispersion and diameter separation of multi-wall carbonnanotubes in aqueous solutions. Journal of Colloid and Interface Science, 2010. 345(2): p.138-142.129. Clark, M.D., S. Subramanian, and R. Krishnamoorti, Understanding surfactant aidedaqueous dispersion of multi-walled carbon nanotubes. Journal of Colloid and InterfaceScience, 2011. 354(1): p. 144-151.130. Jiang, L., L. Gao, and J. Sun, Production of aqueous colloidal dispersions of carbonnanotubes. Journal of Colloid and Interface Science, 2003. 260(1): p. 89-94.131. Kumar, P. and H.B. Bohidar, Aqueous dispersion stability of multi-carbon nanoparticlesin anionic, cationic, neutral, bile salt and pulmonary surfactant solutions. Colloids andSurfaces A: Physicochemical and Engineering Aspects, 2010. 361(1?3): p. 13-24.132. Rastogi, R., et al., Comparative study of carbon nanotube dispersion using surfactants.Journal of Colloid and Interface Science, 2008. 328(2): p. 421-428.138133. Vaisman, L., H.D. Wagner, and G. Marom, The role of surfactants in dispersion ofcarbon nanotubes. Advances in Colloid and Interface Science, 2006. 128?130(0): p. 37-46.134. Yu, J., et al., Controlling the dispersion of multi-wall carbon nanotubes in aqueoussurfactant solution. Carbon, 2007. 45(3): p. 618-623.135. Zhang, J. and L. Gao, Dispersion of multiwall carbon nanotubes by sodium dodecylsulfate for preparation of modified electrodes toward detecting hydrogen peroxide.Materials Letters, 2007. 61(17): p. 3571-3574.136. Hu, Y., et al., Electromechanical Actuation with Controllable Motion Based on a Single-Walled Carbon Nanotube and Natural Biopolymer Composite. ACS Nano, 2010. 4(6): p.3498-3502.137. Choi, J.-Y., et al., In situ grafting of carboxylic acid-terminated hyperbranchedpoly(ether-ketone) to the surface of carbon nanotubes. Polymer, 2007. 48(14): p. 4034-4040.138. Farzi, G., et al., Effect of radical grafting of tetramethylpentadecane and polypropyleneon carbon nanotubes' dispersibility in various solvents and polypropylene matrix.Polymer, 2009. 50(25): p. 5901-5908.139. Ghanem, M.A., et al., Covalent modification of carbon nanotubes with anthraquinone byelectrochemical grafting and solid phase synthesis. Electrochimica Acta, 2012. 68(0): p.74-80.140. Lee, H.Y. and B. S. Kim, Grafting of molecularly imprinted polymers on iniferter-modified carbon nanotube. Biosensors and Bioelectronics, 2009. 25(3): p. 587-591.139141. Lin, T.S., et al., Percolated network of entangled multi-walled carbon nanotubesdispersed in polystyrene thin films through surface grafting polymerization. MaterialsChemistry and Physics, 2005. 94(2?3): p. 438-443.142. Liu, P., Modifications of carbon nanotubes with polymers. European Polymer Journal,2005. 41(11): p. 2693-2703.143. Mountrichas, G., S. Pispas, and N. Tagmatarchis, Grafting-to approach for thefunctionalization of carbon nanotubes with polystyrene. Materials Science andEngineering: B, 2008. 152(1?3): p. 40-43.144. Shen, J., et al., Dispersion behavior of single-walled carbon nanotubes by grafting ofamphiphilic block copolymer. Composites Part A: Applied Science and Manufacturing,2008. 39(10): p. 1679-1683.145. Song, Q., et al., Increasing mechanical strength retention rate of carbon/carboncomposites after graphitization by grafting straight carbon nanotubes radially onto carbonfibers. Materials Science and Engineering: A, 2013. 560(0): p. 831-836.146. Yan, D. and G. Yang, A novel approach of in situ grafting polyamide 6 to the surface ofmulti-walled carbon nanotubes. Materials Letters, 2009. 63(2): p. 298-300.147. Baskaran D, M.J.W., Bratcher M S. Angew, Chemical Int. Ed, 2004. 116.148. Chen, J., et al., Solution Properties of Single-Walled Carbon Nanotubes. Science, 1998.282(5386): p. 95-98.149. Li, Q.-F., et al., Dispersions of carbon nanotubes/polyhedral oligomeric silsesquioxaneshybrids in polymer: the mechanical, electrical and EMI shielding properties. Journal ofMaterials Science, 2011. 46(7): p. 2324-2330.140150. Rozenberg, B.A. and R. Tenne, Polymer-assisted fabrication of nanoparticles andnanocomposites. Progress in Polymer Science, 2008. 33(1): p. 40-112.151. Smook, J., Preparation and properties of ultra high stength polyethylene fibres, 1984,Rijksuniversiteit Te Groningen. p. 159.152. Tao, J, et al,. Spinning and Drawing Properties of Ultrahigh-Molecular-WeightPolyethylene Fibers Prepared at Varying Concentrations and Temperatures PolymerEngineering & Science, 2004. 43(11): p. 12.153. Griffith, A.A., The Phenomena of Rupture and Flow in Solids. Philosohical Transactionsof the Royal Society of London, Series A, Containing Papers of a Mathematical orPhysical Character, 1921. 221.154. Peterlin, A., Folow-induced crystallization. Polymer Systems, Midland MacromolecularMonografts, ed. G.a.B. G.R.L.Miller1979.155. Kanamoto, T., et al., Super-drawing of ultrahigh molecular weight polyethylene. 1. Effectof techniques on drawing of single crystal mats. Macromolecules, 1988. 21(2): p. 470-477.156. Djezzar, K., et al., Tensile drawing of ethylene/vinyl-alcohol copolymers. Part 1.Influence of draw temperature on the mechanical behaviour. Polymer, 1998. 39(17): p.3945-3953.157. Mahendrasingam, A., et al., Influence of temperature and chain orientation on thecrystallization of poly(ethylene terephthalate) during fast drawing. Polymer, 2000. 41(21):p. 7803-7814.158. Girifalco, L. A., et al., Van der Waals binding energies in graphitic structures. PhysicalReview B, 2002. 65(12540).141159. Kwon, Y.K.S., Tomanek, D, Phys. ReV. B, 1998. 58(R13314).160. Erik I. Waldorff, A.M.W., Peretz P. Friedmann, and Michael Keidar, Characterization ofcarbon nanotubes produced by arc discharge: Effect of the background pressure. Journalof Applied Physics 2004. 95(5): p. 6.161. Morjan, R.E., et al., High growth rates and wall decoration of carbon nanotubes grown byplasma-enhanced chemical vapour deposition. Chemical Physics Letters, 2004. 383(3?4):p. 385-390.162. Blanch, A.J., et al., Parametric analysis of sonication and centrifugation variables fordispersion of single walled carbon nanotubes in aqueous solutions of sodiumdodecylbenzene sulfonate. Carbon, 2011. 49(15): p. 5213-5228.163. Cardenas, J.F., Protonation and sonication effects on aggregation sensitive Ramanfeatures of single wall carbon nanotubes. Carbon, 2008. 46(10): p. 1327-1330.164. Gkikas, G., et al., Effect of dispersion conditions on the thermo-mechanical andtoughness properties of multi walled carbon nanotubes-reinforced epoxy. CompositesPart B: Engineering, 2012. 43(6): p. 2697-2705.165. Heller, D.A., et al., Sonication-induced changes in chiral distribution: A complication inthe use of single-walled carbon nanotube fluorescence for determining speciesdistribution. Carbon, 2005. 43(3): p. 651-653.166. Park, C., et al., Dispersion of single wall carbon nanotubes by in situ polymerizationunder sonication. Chemical Physics Letters, 2002. 364(3?4): p. 303-308.167. Sobolkina, A., et al., Dispersion of carbon nanotubes and its influence on the mechanicalproperties of the cement matrix. Cement and Concrete Composites, 2012. 34(10): p.1104-1113.142168. Yu, H., et al., Optimizing sonication parameters for dispersion of single-walled carbonnanotubes. Chemical Physics, 2012. 408(0): p. 11-16.169. Zaragoza-Contreras, E.A., et al., Evidence of multi-walled carbon nanotubefragmentation induced by sonication during nanotube encapsulation via bulk-suspensionpolymerization. Micron, 2009. 40(5?6): p. 621-627.170. Abbaszadeh, J., et al., Increasing the efficicency of ultrasonic dispersion system with useof control loop to automatic frequency adjusting. International Journal on smart sensingand intelligent systems 2011. 4(2): p. 19.171. Jacob M. Wernik, S.A.M., Recent Developments in Multifunctional NanocompositesUsing Carbon Nanotubes. Applied Mechanics Reviews, 2009. 63: p. 40.172. Huang, Y.Y. and E.M. Terentjev, Dispersion of Carbon Nanotubes: Mixing, Sonication,Stabilization, and Composite Properties. Polymers, 2012. 4(1): p. 275-295.173. Hamaker, H.C., The London?van der Waals attraction between spherical particles.Physica, 1937. 4(10): p. 1058-1072.174. Lohse, D., Sonoluminescence: Cavitation hots up. Nature, 2005. 434.175. Nguyen, T.Q., Liang, Q.Z., and Kausch, H.-H., Kinetics of ultrasonic and transientelongational flow degradation: a comparative study. Polymer, 1997. 38(15): p. 3783-3793.176. Huh, S., et al., The Ball Milling with Various Rotation Speeds Assisted to Dispersion ofthe Multi-Walled Carbon Nanotubes. Nanoscience and Nanotechnology Letters, 2012.4(1): p. 20-29.177. Smart, S.K., Cheng, H.M. and Lu, G.Q., Shortened double-walled carbon nanotubes byhigh-energy ball milling International Journal of Nanotechnology, 2007. 4(5).143178. Kukovecz, ?., et al., Long-time low-impact ball milling of multi-wall carbon nanotubes.Carbon, 2005. 43(5): p. 994-1000.179. Rubio, N., et al., Ball-Milling Modification of Single-Walled Carbon Nanotubes:Purification, Cutting, and Functionalization. Small, 2011. 7(5): p. 665-674.180. Iijima, M., M. Tsukada, and H. Kamiya, Effect of particle size on surface modification ofsilica nanoparticles by using silane coupling agents and their dispersion stability inmethylethylketone. Journal of Colloid and Interface Science, 2007. 307(2): p. 418-424.181. Zhou, Z., et al., Functionalization of multi-wall carbon nanotubes with silane and itsreinforcement on polypropylene composites. Composites Science and Technology, 2008.68(7?8): p. 1727-1733.182. Nakatani, H., et al., Effect of chemical structure of silane coupling agent on interfaceadhesion properties of syndiotactic polypropylene/cellulose composite. Journal ofApplied Polymer Science, 2011. 119(3): p. 1732-1741.183. Ma, G., et al., Effect of the addition of silane coupling agents on the properties ofwollastonite-reinforced poly(ether ether ketone) composites. Polymer Engineering &Science, 2011. 51(6): p. 1051-1058.184. Cheng, Y.-Y., S.-C. Chou, and J.-H. Huang, Preparation and characterization ofpolyimide/silane coupling agent modified multiwall carbon nanotubes composites.Journal of Applied Polymer Science, 2012. 124(2): p. 1137-1143.185. Lin, O.H., H.M. Akil, and Z.A. Mohd Ishak, Surface-activated nanosilica treated withsilane coupling agents/polypropylene composites: Mechanical, morphological, andthermal studies. Polymer Composites, 2011. 32(10): p. 1568-1583.144186. Yang, G, et al,. Post drawing properties of gel spun HDPE/UHMWPE fibre. ChineseJournal of Colloid and Polymer, 2009. 27(3): p. 3.187. Cox, H. L., The elasticity and strength of paper and other fibrous materials, Brit. J. Appl.Phys. 3, 1952, 72.188. Wang, Z.J., Reinforcing efficiency of Carbon Nanotubes in PVA, Ph.D thesis, QueenMary, University of London, November, 2007189. Coleman, J.N. et al., Enhancement of modulus, strength, and toughness in Poly(methylmethacrylate)-based composites by the Incorporation of Poly(methyl methacrylate)-functionalized nanotubes, Adv. Funct. Mater.2006.16(12): p.1608-1614.145AppendicesAppendix A Statistical analysis of orthogonal design of experimentThe result of orthogonal design of experiment was analyzed by calculating followingparameters as shown in Equation A.1-A.4.???miiyT1Equation A.1nyCTmii21?????????Equation A.2nCTkrQmiijj ?? ??121 Equation A.3jjj fQS ?2 Equation A.4Table A.1 Analysis of orthogonal optimization using U*(1/3) as responseExperimentNO Concentration Temperature Winding speed Error1*( )3U (m/s)1 1 1 1 1 656.642 1 2 2 2 673.343 1 3 3 3 459.974 2 1 2 3 852.035 2 2 3 1 794.346 2 3 1 2 696.80146ExperimentNO Concentration Temperature Winding speed Error1*( )3U (m/s)7 3 1 3 2 629.708 3 2 1 3 661.109 3 3 2 1 499.51K1j 1789.94 2138.36 2014.54 1937.48T=5923.42K2j 2343.17 2128.77 2011.87 1999.84K3j 1777.30 1643.28 1884.01 1973.10K1j2 3203901.44 4572596.12 4058353.65 3753836.63K2j2 5490457.64 4531680.63 4047634.75 3999359.71K3j2 3158795.00 2700367.90 3549486.06 3893104.11*Kij: sum of test result of fibre spun with factor j at level i.The influence of each factor on the objective will also be determined. Total variance TQ ,degree of total freedom Tf , square deviation for factor 2factorS , square deviation for error 2eS canbe calculated from Equations A.5 to A.9.efactorT QQQ ?? Equation A.5nxxQniiniiT?? ???? 1212)(, )...,2,1(12112 kjnxKrQniimiijj ??????????? ??Equation A.6efactorT fff ?? Equation A.7147factorfactorfactor fQS ?2 Equation A.8eee fQS ?2 Equation A.9Where factorQ is the variance for factor, eQ is the variance for error, factorf is the degree offreedom for factor, ef is the degree of freedom for error, 2factorS is the square deviation for factor,and 2eS is the square deviation for error. The F value for each factor is calculated using EquationA. 10.22efactorfactor SSF ? Equation A.10An F test was applied to analyze the significance of the influence of factors on fibre1*( )3UTable A.2 Factor significance analysis using F-testFactor Sum of squaresSSiDegrees offreedom Mean square F F? SignificanceA 67969.95 2 33984.97 167.01 F0.01(2,2)=99F0.025(2,2) =39F0.05(2,2) =19F0.1(2,2) =999%B 50638.63 2 25319.31 124.43 99%C 4109.71 2 2054.86 10.10 90%Error 406.98 2 203.49148Appendix B Statistical analysis of tensile test dataTable B.1 F-test and T-test of tensile properties of UHMWPE fibre reinforced by sonicationdispersed MWCNTs01%239 49You39ong9Y'ns M9g1'9dl(1 G 04Pa)d l(1 G (dCNTnG (d4NTnGweih(a tftrc .ftct.fc5ihS&4'weih(a (f).* .f()* .f.)* +f)..c+a,(.)5ihS&4'weih(a (f-rr .f(t+ .f.tt cfc(c.a,)))fc5ihS&4'weih(a (f((. .f(-. .f.** (f*c+)(a,)-i/0131o M9g1'9dl(1 G 04Pa)d l(1 G (dCNTnG (d4NTnGweih(a ).(f)(. )-f.*t.fc5ihS&4'weih(a -*f.c* )+fc.) .ftt*t tf.+a,.)5ihS&4'weih(a *cf).- ))f+-) .f.-tt .f...)r*.+))fc5ihS&4'weih(a c)fc+r rf-c. .f... )f)t)-(a,)*0ng1uY M9g1'9d5G 04Pa)d5G (dCNTnG (d4NTnGweih(a .f.t* .f..-.fc5ihS&4'weih(a .f.+( .f..- .ft).c .f.-((r*c+)5ihS&4'weih(a .f.t* .f..* .f.--. .frt+-)-.rc)fc5ihS&4'weih(a .f.t. .f..r .f)+(. .fr*.)*)c++149Table B.2 F-test and T-test of tensile properties of UHMWPE fibre reinforced by sonication andball mill dispersed MWCNTs01%239 49You39ong9Y'ns M9g1'9dl(1 G 04Pa)d l(1 G (dCNTnG (d4NTnGweih(a tftrc .ftct.fc5ihS&4'weih(a tfct. .ft.* .f(-. .f)tt)5ihS&4'weih(a tfc-) .fct. .f.)t .f)(.)fc5ihS&4'weih(a tf()( .f))t .f(-- .f)(.i/0131o M9g1'9dl(1 G 04Pa)d l(1 G (dCNTnG (d4NTnGweih(a ).(f)(. )-f.*t.fc5ihS&4'weih(a .cf.* )-f*. .f+tc .f)-.)5ihS&4'weih(a ).tf(r. (cfct* .f.(- .fr.r)fc5ihS&4'weih(a *-f+. ).f)( .f.(( .f...(*0ng1uY M9g1'9d5G 04Pa)d5G (dCNTnG (d4NTnGweih(a .f.t* .f..-.fc5ihS&4'weih(a .f.tc .f..+ .f.tr .f.)-)5ihS&4'weih(a .f.t .( .f.)(+ (ft)ra,.c .fr+t)fc5ihS&4'weih(a .f.tr .f..r .f))* .f+--150Table B.3 F-test and T-test of tensile properties of UHMWPE fibre reinforced by coupling agentfunctionalized MWCNTs01%239 49You39ong9Y'ns M9g1'9dl(1 G 04Pa)d l(1 G (dCNTnG (d4NTnGweih(a tftrc .ftct)5ihS&4'weih(a tf+-- .ft+* .f+-* .ft+t(5ihS&4'weih(a tfc-t .f+.t .f.t+ .f)(rt5ihS&4'weih(a tfrt. .ft+t .f+)c .f...)01%239i/0131o M9g1'9dl(1 G 04Pa)d l(1 G (dCNTnG (d4NTnGweih(a ).(f)(. )-f.*t)5ihS&4'weih(a ).(fr.. )tfttc .f)t* (f...(5ihS&4'weih(a ).(fr.( )tfc-+ .f)c. )f..*t5ihS&4'weih(a ).+f.+. ))f)*. . f.)t* (f.))01%2390ng1uY M9g1'9d5G 04Pa)d5G (dCNTnG (d4NTnGweih(a .f.t* .f..-)5ihS&4'weih(a .f.t. .f..+ )f*tc (f..)(5ihS&4'weih(a .f. +t .f.. . .fc+r (f..)t5ihS&4'weih(a .f. t- .f..- .fcc+ )f..c151Table B.4 F-test and T-test of tensile properties of UHMWPE fibre reinforced by polymer graftedMWCNTs01%239 49You39ong9Y'ns M9g1'9dl(1 G 04Pa)d l(1 G (dCNTnG (d4NTnGweih(a tftrc .ftct)5ihS&4'weih(a tf-)r .f(.( .f()++(*.f.(.-*c(5ihS&4'weih(a tfrr. .ft(* .ft*-ct( .f...+ctt5ihS&4'weih(a +f.)r .ft(. .ftr(r(* )f+.a,.-01%239i/0131o M9g1'9dl(1 G 04Pa)d l(1 G (dCNTnG (d4NTnGweih(a ).(f)(. )-f.*t)5ihS&4'weih(a ))rft.c ))fc(r .f.*)+(r .f..)t+t(5ihS&4'weih(a )(tf-cc rftct .f... ( )ft-a,.rt5ihS&4'weih(a )(cfc)t )cf+)+ .f+c*c+) -fr.a,.c01%2390ng1uY M9g1'9d5G 04Pa)d5G (dCNTnG (d4NTnGweih(a .f.t* .f..-)5ihS&4'weih(a .f.+) .f..+ .f.t(.)( .f.r(t)((5ihS&4'weih(a .f.t. .f..+ .f).rcr( .fc(t.).t5ihS&4'weih(a .f.+( .f..- .f(-tr*) .f.r(.+)152Appendix C Dispersing MWCNTs through mechanical methods1. Dispersing MWCNT in mineral oil by microfluidizingMicrofluidizing is a high shear fluid process that has been used in recent years forparticle de-agglomeration, dispersion and size reduction. In this process, high pressures are usedin order to generate high shear which forces CNT dispersion to go through specially designedmicro channels at high flow rates to pull CNT agglomerates apart. The process pressure and thechannel geometry control the velocities inside the channels, and therefore modulate the energydissipation.1.1 Material and experimentMWCNT/mineral oil mixtures samples with MWCNT concentrations of 0.5wt% wereprepared and sent to Microfluidics Co., Ltd. for processing. The feed material was driven atconstant pressure through the interaction chamber which has micro-channels with unique fixed-geometry inside the microfluidizer. The microfluidizer processor (model M-110P) was used toprocess the MWCNT/mineral oil dispersion. Different levels of shear forces were applied to thesample by varying the pressure from 10,000 to 30,000 psi and passed from 1 to 3 during theprocessing, as listed in Table. C.1.Table C.1 Processing condition of MircofluidizerSample Pressure(Psi) PassLow shear 10K 1passMedium shear 30K 1passHigh shear 30K 3pass1531.2 Results and discussionThe dispersion of MWCNT in mineral oil after treatment by microfluidizer was examinedby optical microscopy. As can be seen in Figure C.1, big MWCNT agglomerations with a size ofabout 10?m can be found after treatment under low shear conditions. These extra-largeagglomerates disappeared after two more passes in the microfluidizing machine at the samepressure level.(1) Low shear mode (2) Medium shear mode (3) High shear modeFigure C.1 MWCNTs dispersion in mineral oil processed by microfluidizerThe particle sizes of MWCNTs processed by the microfluidizer under differentconditions were estimated by photon correlation spectroscopy (PCS) using a Zetasizer?(Malvern Instruments Worcestersher, UK). It was observed that with when low shear force wasused, the size of CNT agglomerates was almost the same as CNTs dispersed by ultra-sonication.As shear force increased, the size of MWCNT agglomerates decreased slightly from about3400nm to 2560 nm under the high shear mode treatment. The size of MWCNT agglomeratesprocessed by microfluidization was about 1/3 smaller than those processed by ultra-sonication,as shown in Table.C.2. However, neither process is effective enough to debundle MWCNTagglomerates.154Table C.2 Average particle size in MWCNT/UHMWPE solutionSample treated by Average Diameter (nm)Sonication 3400MicrofluidizerLow shear 3063Medium shear 2915High shear 2596.2. Dispersing MWCNT in mineral oil by solvent assistant spraying methodIn this method, an aqueous solution of carbon nanotube with surfactant was made andsprayed onto the surface of fine powder UHMWPE. The carbon nanotubes were adsorbed on thesurface of the UHMWPE powders. The UHMWPE that adsorbed the CNTs were then dissolvedin solvent and the MWCNTs/UHMWPE composites (in the form of film) were prepared from thesolution. Similar methods were used in this study to disperse MWCNTs in UHMWPE to producecomposite fibre.2.1 Material and experimentThe UHMWPE and MWCNTs used in this study are the same as those listed in Chapter 3.Distilled water was used as the solvent to disperse MWCNTs with sodium dodecylbenzenesulfate (SDS) as a surfactant. 10 mL solutions containing 12.5, 25 and 37.5 mg of MWCNTs and1 wt % SDS based on H2O were ultrasonicated using a bath sonicator for 3 hours and thencentrifuged at 3000 rpm for 20 min. The MWCNT coated UHMWPE powders were then driedovernight in a vacuum oven under 80?C. 5wt% MWCNT coated UHMWPE powders were155dissolved in mineral oil and heated up to prepare gel for gel spinning. The gel preparation,spinning, and post-drawing processes are detailed in Chapter 3.2.2 Results and DiscussionThe surface morphology of UHMWPE powder used in this work was observed underSEM. Instead of solid particles with smooth surfaces, the UHMWPE powder consisted of manynano-scale particles and fibrils. The diameter of the nanoparticles was around 800-900nanometers while the diameter of the nanofibrils was around 100 nanometers, as measured fromFigure C.2. Because of the above microstructure, many micro-voids form on the surface of theUHMWPE particles. These microvoids provide the channels for MWCNT/water solution topenetrate the whole particle, thus forming an even distribution of MWCNTs in the particle.Furthermore, the micro particles' and fibrils' structures greatly increased the surface area of theUHMWPE particle, which made it easy to adsorb MWCNTs.(a) Before (b) AfterFigure C.2 UHMWPE powder before and after spraying with MWCNT156However, it was found that after spraying only 0.1wt% of MWCNTs, the surface of theUHMWPE particle was covered completely, as shown in Figure 2(b). Thus, theMWCNT/UHMWPE composite material made through this method is too low to improve themechanical properties.

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.24.1-0165846/manifest

Comment

Related Items