Mechanical and morphological characterization of wood plastic composites based on municipal plastic waste
Mémoire
Yasamin Kazemi
Maîtrise en Génie chimique Maître ès sciences (M. Sc.)
Québec, Canada
© Yasamin Kazemi, 2013
i
ii
Résumé Les développements récents de la législation associée aux impacts environnementaux des déchets plastiques d’origine post-consommation ont mené à des efforts sur le développement de techniques viables de recyclage. Ainsi, le but de cette recherche était de produire des composites bois-plastique (WPC : wood plastic composites) à partir de la fraction légère des déchets plastiques municipaux (post-consommation) et de résidus de transformation du bois (sciure). Afin d’améliorer la compatibilité et l’adhésion entre le polyéthylène (PE) et le polypropylène (PP), un copolymère d’éthylène-octène (EOC: ethylene-octene copolymer) a été utilisé pour développer la compatibilité entre les phases polymères tout en agissant comme modificateur d’impact. L’ajout de PE et PP maléatés (MAPE: maleated polyethylene; MAPP maleated polypropylene) a permis de fournir une meilleure compatibilité entre la matrice polymère et la farine de bois. Les effets combinés de tous les composants ont mené à la production de composites présentant des propriétés morphologiques (dispersion et adhésion) et mécaniques (traction, torsion, flexion et impact) intéressantes après l’optimisation de l’ensemble des additifs (mélanges d’agents couplants). Dans un second temps, des composites structuraux à trois couches ont été produits à partir des matériaux composites mentionnés plus haut afin d’étudier l’effet des paramètres de design sur les performances en flexion et à l’impact. Les paramètres étudiés incluent la teneur en bois, l’épaisseur des couches individuelles de composite, ainsi que la séquence et la configuration d’empilement des différentes couches (structures symétriques et asymétriques). Enfin, la théorie classique des poutres a été utilisée avec succès pour prédire le module en flexion et ce, avec un maximum de 10% de déviation pour ces structures complexes.
iii
iv
Abstract Recent legislations associated with environmental impacts of post-consumer plastic wastes have driven substantial attention toward developing viable recycling techniques. Therefore the aim of this research was to produce wood plastic composites (WPC) from the light fraction of municipal plastic wastes (post-consumer) and wood processing residues (sawdust). In order to improve compatibility and adhesion between polyethylene (PE) and polypropylene (PP), an ethylene-octene copolymer (EOC) was used to compatibilize the polymer phases and also to act as an impact modifier. Addition of maleated polyethylene (MAPE) and maleated polypropylene (MAPP) provided improved compatibility between the polymer matrix and the wood flour. The combined effect of all the components was found to produce composites with interesting morphological (dispersion and adhesion) and mechanical properties (tension, torsion, flexion and impact) after optimization of the additive package (blend of coupling agents). In the second phase, three-layered structural composites were produced from the aforementioned composites to investigate the effects of design parameters on their flexural and impact performance. The studied parameters include wood content, thickness of individual composite layers, as well as stacking sequence and configuration (symmetric and asymmetric structures). In addition, the classical beam theory was successfully used to predict the flexural modulus within 10% of deviation for these complex structures.
v
vi
Foreword This dissertation is written in a manuscript-based format and consists of five chapters. The first chapter includes a brief introduction to environmental issues, obstacles and possibilities of plastic recycling. The current plastic recycling techniques are discussed and subsequently mechanical recycling is proposed as the most beneficial technique. The light fraction of recycled plastics (mainly composed of polyethylene and polypropylene) is introduced as the main contributor of plastic waste streams and therefore compatibilization techniques between PE and PP are reviewed in this chapter. In the second chapter, natural fiber composites are discussed as a proposed application for recycled plastic materials. Fabrication and modification techniques of wood-plastic composites (WPC) are discussed and WPC literature based on polyethylene and polypropylene blends is reported. In addition, structural designing is introduced to improve the mechanical properties of these composite materials. The objectives of this research work are presented at the end of this chapter. The following two chapters present the experimental results in the form of submitted articles. My contribution in these manuscripts was to perform all the experimental work, data collection and analysis (including calculations) and to write their first drafts. Chapter 3 discusses the fabrication and characterization of wood-plastic composites of recycled origin. The results of this work are submitted in the following manuscript: [1] Kazemi, Y., Cloutier, A. and Rodrigue, D., Mechanical and morphological properties of wood-plastic composites based on municipal plastic waste, Polymer Composites, Submitted in October 2012. Then, the composites produced in Chapter 3 were exploited in Chapter 4 to study the effect of design parameters on mechanical performance of three-layered structural composites. The manuscript was submitted as: vii
[2] Kazemi, Y., Cloutier, A. and Rodrigue, D., Design analysis of three-layered structural composites based on post-consumer recycled plastics and wood residues, Composites part A, Submitted in January 2013. The last chapter includes a general conclusion on the work performed and recommendations for future works. Nevertheless, this research work includes more results presented in other manuscripts or conference presentations as follows: [3] Kazemi, Y., Cloutier, A. and Rodrigue, D., Natural fiber composites based on postconsumer polyolefins and wood fiber residues: Effect of coupling agent addition, PPS Americas Conference 2012, Niagara Falls, ON, Canada, May 21-24 (2012). [4] Ramezani Kakroodi, A., Kazemi, Y. and Rodrigue, D., Mechanical, rheological, morphological and water absorption properties of maleated polyethylene/hemp composites: effect of ground tire rubber addition, Composites Part B, in press (2012). [5] Ramezani Kakroodi, A., Kazemi, Y. and Rodrigue, D., Impact modification of waste plastic/wood flour composites via structural modification, Submitted to ICCM19, Montreal, QC, Canada, July 28-August 2 (2013).
viii
Acknowledgments This dissertation would not be accomplished without the advices and supports of people whom I am greatly indebted. First and foremost, I wish to express my profound gratitude to Prof. Denis Rodrigue, my supervisor, for his invaluable guidance and assistance. I am incredibly grateful and appreciative for his incessant kindness, patience and understanding. He gave me the confidence to pursue my ideas in this research work which is the most valuable achievement for me. I would like to sincerely thank my co-supervisor, Prof. Alain Cloutier, for his helps, supportive direction and scientific insight. I would like to express my sincere appreciation to my family who have been the source of encouragement and inspiration to me throughout my life. I would like to especially express my gratitude to my husband, Adel, for all his kindness and supports as a husband, colleague and a teacher. I also appreciate the technical assistance of Mr. Yann Giroux, who is not only a capable technician, but also a very good friend. I would like also to thank my colleagues and friends of the chemical engineering department for their amity and supports which made great memories throughout my M.Sc. program. Finally, I acknowledge the financial and technical support of the Natural Sciences and Engineering Research Council of Canada (NSERC) and FPInnovations, as well as Centre de Recherche sur le Bois (CRB) and Centre Québécois sur les Matériaux Fonctionnels (CQMF) for technical and financial help.
ix
x
“Nothing in life is to be feared, it is only to be understood. Now is the time to understand more, so that we may fear less.” Marie Curie
xi
xii
Table of Contents Résumé.................................................................................................................................. iii Abstract ................................................................................................................................... v Foreword .............................................................................................................................. vii Acknowledgments .................................................................................................................ix List of Tables .......................................................................................................................xix List of Figures ......................................................................................................................xxi Nomenclature ...................................................................................................................... xxv Chapter 1. Plastic Recycling ................................................................................................... 1 1.1 Importance of plastic recycling..................................................................................... 1 1.2 Methods for recycling of thermoplastics ...................................................................... 2 1.2.1 Re-use and primary recycling ................................................................................ 3 1.2.2 Energy recovery ..................................................................................................... 3 1.2.3 Chemical recycling (Feedstock recycling) ............................................................ 4 1.2.4 Mechanical recycling ............................................................................................. 5 1.3 Compatibility of polyethylene and polypropylene ....................................................... 7 1.4 Compatibilization methods for PE/PP blends ............................................................... 9 1.4.1 Reactive compatibilization .................................................................................... 9 xiii
1.4.2 Nonreactive compatibilization ............................................................................ 11 Chapter 2. Natural Fiber Composites ................................................................................... 15 2.1 Natural vs. artificial reinforcements ........................................................................... 15 2.2 Characteristics of natural fibers.................................................................................. 17 2.3 Modification of natural fiber composites ................................................................... 20 2.3.1 Surface modification of natural fibers ................................................................. 20 2.3.1.1 Chemical surface treatment of natural fibers ............................................... 20 2.3.1.2 Physical surface treatment of natural fibers ................................................. 23 2.3.2 Modification of polymeric matrix ....................................................................... 24 2.4 Matrices for natural fiber composites......................................................................... 26 2.4.1 Bio-based polymers ............................................................................................. 26 2.4.2 Use of waste plastics ........................................................................................... 28 2.5 Structural design of composite materials ................................................................... 30 2.6 Thesis objectives and organization ............................................................................ 36 Chapter 3. Mechanical and Morphological Properties of Wood Plastic Composites Based on Municipal Plastic Waste .................................................................................................. 39 Résumé ............................................................................................................................. 39 Abstract ............................................................................................................................ 40 3.1 Introduction ................................................................................................................ 41
xiv
3.2 Experimental ............................................................................................................... 43 3.2.1 Materials .............................................................................................................. 43 3.2.2 Processing ............................................................................................................ 45 3.2.3 Morphological observation .................................................................................. 46 3.2.4 Mechanical testing ............................................................................................... 46 3.2.5 Density and Hardness Measurements .................................................................. 47 3.3.1 Blend morphology ............................................................................................... 47 3.3.2 Mechanical characterizations ............................................................................... 49 3.3.2.1 Recycled polymeric matrix ........................................................................... 50 3.3.2.2 Uncompatibilized composites ....................................................................... 51 3.3.2.3 Compatibilized composites ........................................................................... 51 3.3.3 Density and hardness results ................................................................................ 52 3.4 Conclusions ................................................................................................................. 53 Acknowledgments ............................................................................................................ 54 Chapter 4. Design Analysis of Three-Layered Structural Composites Based on PostConsumer Recycled Plastics and Wood Residues ................................................................ 55 Résumé.............................................................................................................................. 55 Abstract ............................................................................................................................. 56 4.1 Introduction ................................................................................................................. 57
xv
4.2 Materials and methods ............................................................................................... 58 4.2.1 Materials .............................................................................................................. 58 4.2.2 Processing............................................................................................................ 59 4.2.3 Sample coding ..................................................................................................... 59 4.2.4 Microscopy .......................................................................................................... 60 4.2.5 Mechanical testing............................................................................................... 60 4.2.6 Theory ................................................................................................................. 61 4.3 Results and discussion ................................................................................................ 63 4.3.1 Microscopy .......................................................................................................... 63 4.3.2 Mechanical characterizations .............................................................................. 65 4.3.2.1 Symmetric structural composites ................................................................. 66 4.3.2.2 Asymmetric structural composites with equal layer thickness ........................ 67 4.3.2.3 Asymmetric structural composites with different thickness of layers ............. 70 4.4 Conclusion .................................................................................................................. 72 Acknowledgements .......................................................................................................... 73 Chapter 5. Conclusions and Recommendations ................................................................... 77 5.1 General conclusion ..................................................................................................... 77 5.2 Recommendations for future works ........................................................................... 79 References ............................................................................................................................ 81 xvi
Appendix A ........................................................................................................................... 91 Appendix B ........................................................................................................................... 93 Appendix C ........................................................................................................................... 95
xvii
xviii
List of Tables Table 1.1: Calorific values of major plastic waste compared with common fuels. ............... 3 Table 1.2: Common separation techniques for plastic recycling. .......................................... 6 Table 2.1: Energy consumption (MJ/kg) for production of different fibers. ....................... 15 Table 2.2: Chemical composition of lignocellulosic fibers (weight %) .............................. 17 Table 2.3: Applications of multi-layered structural composites. ......................................... 31 Table 2.4: Advantages and disadvantages associated with application of multilayered structures.. ..................................................................................................................... 31 Table 3.1: Mechanical properties of the composites without compatibilizer and coupling agent. ............................................................................................................................. 49 Table 3.2: Mechanical properties of the composites with compatibilizer and coupling agent. ............................................................................................................................. 50 Table 3.3: Hardness and density results for compatibilized and uncompatibilized composites. ................................................................................................................... 53 Table 4.1: Flexural and impact properties of symmetric three-layered structural composites. ................................................................................................................... 67 Table 4.2: Flexural and impact properties of asymmetric three-layered structural composites with equal layer thickness .......................................................................... 74 Table 4.3: Flexural and impact properties of asymmetric three-layered structural composites with different thickness of layers ............................................................... 75
xix
xx
List of Figures Figure 1.1: MSW generation and recycling in USA from 1960 to 2010. .............................. 1 Figure 1.2: Life cycle of plastic materials including waste management options . ............... 2 Figure 1.3: Common processes used for chemical recycling of plastic waste and their main products. .......................................................................................................................... 4 Figure 1.4: Schematic representation of plastics separation by a floatation technique. ........ 7 Figure 1.5: LV-STEM of low density polyethylene/PP (80/20) blends (frame width: 9.7 μm) . ................................................................................................................................ 8 Figure 1.6: Common functional groups for chemical compatibilization. ............................ 10 Figure 1.7: SEM micrographs from fractured surfaces of fractured impact specimens of PE/PP: 50/50 blend (a) without and (b) with 10% EPDM. .......................................... 12 Figure 1.8: AFM micrographs of PP/HDPE compounds (a) without compatibilizer and (b) with multi-block EOC (frame width: 20 μm) ............................................................... 13 Figure 2.1: Generic life cycles of (a) glass fiber and (b) natural fiber reinforced composites. ................................................................................................................... 16 Figure 2.2: Chemical structure of a repeating unit in cellulose molecule. .......................... 18 Figure 2.3: SEM micrographs from fractured surfaces of hemp filled PP. ......................... 19 Figure 2.4: Effect of different fillers on water uptake of an epoxy composite. ................... 19 Figure 2.5: Schematic representation of interaction between LDPE and MPS-modified fiber. .............................................................................................................................. 21 Figure 2.6: Effect of cellulosic fibers surface treatment with MPS: a) untreated and b) treated fiber. .................................................................................................................. 21 xxi
Figure 2.7: Surface treatment of natural fibers with maleated polypropylene. ................... 23 Figure 2.8: SEM micrographs of (a) untreated and (b) alkali treated hemp fibers. ............ 24 Figure 2.9: SEM micrographs of PP/agro-fiber composites (a) before and (b) after compatibilization with MAPP. ..................................................................................... 26 Figure 2.10: Current and emerging matrices for natural fiber composites and their biodegradability. ........................................................................................................... 27 Figure 2.11: Elastic modulus (E), tensile strength (TS), elongation at break (EB) and impact strength (IS) of light fraction (LF) based composites with different fillers. .... 30 Figure 2.12: Effect of different cross-sectional designs on properties of composites. Lines depict corresponding fiber placement. Superscript groups are not statistically different for each test. ................................................................................................................. 32 Figure 2.13: Three point flexion test when the composite layer is in extrados or intrados. 33 Figure 2.14: Fractured areas of specimens from flexion test with composite layer placed in the (a) intrados side and (b) extrados side. ................................................................... 34 Figure 2.15: Flexural modulus of two and three layered systems with dissimilar materials. ...................................................................................................................................... 35 Figure 2.16: Flexural modulus of symmetric and asymmetric systems with similar volume fraction of phases (MMC in black and 834 in white). ................................................. 36 Figure 3.1: DSC curve of the municipal plastic waste light fraction used. ......................... 43 Figure 3.2: FTIR spectrogram of the municipal plastic waste light fraction used. ............. 44 Figure 3.3: Typical SEM micrographs of the recycled light fraction plastics: (a) without compatibilizer and (b) with 5 wt.% of EOC. The arrows indicate typical domain sizes in SEM micrographs..................................................................................................... 47 xxii
Figure 3.4: SEM micrograph of composites with 40 wt.% wood flour: (a,c) without coupling agents and (b,d) with additives (5 wt.% of EOC and 5 wt.% (MAPE/MAPP : 80/20)) at different magnifications. .............................................................................. 48 Figure 4.1: Load distribution in a three-layered composite under three point bending test. ...................................................................................................................................... 61 Figure 4.2: SEM micrographs of: a) 40-0-40 and b) 0-20-10 samples. The arrows indicate the position of the interface between two composite layers. ........................................ 64 Figure 4.3: Optical micrographs of fractured samples for: a) 0(3)-40(4)-20(2) and b) 20(4)0(3)-40(2). ..................................................................................................................... 65 Figure 4.4: Three-point bending of a beam with two different directions of flexural load: a) 40(2)-20(3)-0(4) for 3.3% deformation and b) 0(4)-20(3)-40(2) for 4.8% deformation. ...................................................................................................................................... 71 Figure 4.5: Typical flexural stress-strain curves for different configurations of threelayered structural composites........................................................................................ 72
xxiii
xxiv
Nomenclature AFM
Atomic force microscopy
C1
Wood content of layer 1
C2
Wood content of layer 2
C3
Wood content of layer 3
D
Ductility
DCP
Dicumyl peroxide
Dev.
Model standard deviation
DSC
Differential scanning calorimetry
E
Tensile modulus
E1
Flexural modulus of layer 1
E2
Flexural modulus of layer 2
E3
Flexural modulus of layer 3
Ee
Experimental modulus of flexion
Ef
Flexural modulus
EOC
Ethylene-octene copolymer
EPDM
Ethylene-propylene-diene-monomer
EPM
ethylene-propylene copolymer
EPR
ethylene-propylene-rubber
xxv
Et
Theoretical modulus of flexion
Etm
Torsion modulus
EVA
Ethylene-vinyl-acetate
FTIR
Fourier transform infrared
HDPE
High-density polyethylene
HDS
hexadecyltrimethoxy-silane
I1
Moment of inertia for layer 1
I2
Moment of inertia for layer 2
I3
Moment of inertia for layer 3
MA-EPDM
Maleated ethylene-propylene-diene monomer
MAPE
Maleated polyethylene
MAPP
Maleated polypropylene
MIR
Mid infrared
MPS
Methacryloxypropyltrimethoxy
MRPS
Mercaptoproyltrimethoxy
MSW
Municipal solid waste
NIR
Near infrared
PE
Polyethylene
PET
Polyethylene terephthalate
PHA
Polyhydroxyalkanoate
xxvi
PLA
Poly(lactic acid)
PP
Polypropylene
PS
Polystyrene
PVC
Polyvinyl chloride
SBS
Styrene-butylene-styrene block copolymer
SEBS
Styrene-ethylene-butylene-styrene tri-block copolymer
t1
Thickness of layer 1
t2
Thickness of layer 2
t3
Thickness of layer 3
WPC
Wood plastic composite
y0
Neutral axis position
ΔG mix
Energy of mixing
ΔH mix
Heat of mixing
ΔS mix
Entropy of mixing
Ε
Tensile elongation at break
ε max
Flexural strain at maximum stress
σ max
Flexural strength
xxvii
xxviii
Chapter 1. Plastic Recycling 1.1 Importance of plastic recycling During the past decades, production of plastic based materials has attracted increasing attention due to their relatively low price, low density, easy processing conditions, durability and good mechanical characteristics. Plastics are used in a number of applications such as coating, wiring, packaging, as well as automotive and construction industries. In 2011, around 280 million tons of plastic were produced and almost 50% of this material was for single-use disposable applications such as packaging, agricultural films and disposable consumer items [1]. As a result, recycling of polymers has recently emerged as a global concern. Figure 1.1 shows the trends of municipal solid waste (MSW) generation and recycling in USA during the past decades. Plastics represent a significant portion of MSW. In USA, plastic waste represents 12.4% (31 million tons) of the waste stream in 2010 [2]. However, only 8% of plastic waste generated in 2010 was recycled.
Mass (million tonnes)
300 250
MSW Production MSW Recycling
200 150 100
50 0 1960
1970
1980
1990
2000
2010
Figure 0.1: MSW generation and recycling in USA from 1960 to 2010 [2].
1
In the past, landfill disposal used to be one of the most common methods to deal with plastic wastes. However, increasing cost and diminishing spaces in landfills have driven some considerations to find alternative methods [3]. Furthermore, landfill disposal creates health and safety hazards, as well as damage to the environment. This is why the next section of this document is devoted to different recycling methods in opposition to landfill disposal. 1.2 Methods for recycling of thermoplastics Several plastic recycling routes can be used regarding type and quality of plastic waste and market demands. Some methods include energy recovery from waste plastic, while in others plastic materials are recovered. Figure 1.2 shows a schematic representation for the life cycle (including fabrication, service life and recycling) of plastic materials [4].
Figure 0.2: Life cycle of plastic materials including waste management options [4]. 2
1.2.1 Re-use and primary recycling Considering the short service life of most plastic materials, especially in packaging applications, re-use is an option to extend the life cycle of such materials. Re-using consumes less energy compared to other recycling techniques, which makes it a more preferable approach [4]. Primary recycling (re-extrusion) includes introduction of single polymer plastic waste to an extrusion process in order to fabricate similar plastic parts. This type of recycling is only available when the waste is clean (or semi-clean) or made of plastic parts with similar formulations. These criteria made primary recycling an unpopular option for the industry. One example of primary recycling is re-extrusion of plastic parts that do not meet the desired specifications (quality control) of final products in industries. 1.2.2 Energy recovery This method includes combustion for energy production in the form of heat, steam and electricity. Currently, incineration is considered as the prevailing outlet of waste management in plastic recycling. Table 1.1 presents the calorific value of some plastics in comparison with common fuels. It is shown that waste plastics have high calorific values which introduces them as a convenient energy resource. It is also reported that incineration of plastic wastes leads to significant (90-99%) reduction in their volume. However, environmental dilemma associated with this approach, mainly emission of certain air pollutants such as CO2, NOx and SOx, is an incentive to find other options for material recycling [5-6]. Table 0.1: Calorific values of major plastic waste compared with common fuels [7,8]. Item Polyethylene Polypropylene Polystyrene Kerosene Gas oil Heavy oil Petroleum Household PSW mixture
Calorific value (MJ/kg) 43.3–46.5 46.50 41.90 46.50 45.20 42.50 42.3 31.8 3
1.2.3 Chemical recycling (Feedstock recycling) Chemical recycling includes advanced technology processes that are used to convert plastics into lower molecular weight materials such as liquids or gases. The products can be used as feedstock for production of new plastic materials or as fuel. In this method, the chemical structure of polymers is altered through de-polymerization process which results in minimum amount of waste, as well as high product yield [3]. De-polymerization processes include pyrolysis, hydrolysis, gasification, liquid-gas hydrogenation, viscosity breaking and steam or catalytic cracking. Figure 1.3 presents common processes for chemical recycling of plastic waste and their main products. The main advantage of chemical recycling is the ability to use several (heterogeneous) polymers with limited need for pre-treatment.
Figure 0.3: Common processes used for chemical recycling of plastic waste and their main products [8]. 4
1.2.4 Mechanical recycling This type of recycling includes the process of plastic waste recovery for reprocessing and production of new plastic materials through mechanical methods. Many products (grocery bags, pipes, gutters, window and door panels, etc.) are currently being produced using this type of recycling. Mechanical recycling of plastics involves several treatments and steps. For example, Aznar et al. [9] developed one of the most general schemes including the following steps: -
Cutting/shredding of large plastic parts to form small flakes,
-
Separation of dust and other contaminants which is usually done in a cyclone,
-
Separation of different types of plastic wastes,
-
Washing and drying of plastic flakes. Washing is usually performed in water (chemicals can also be used in case of contaminants such as glue, oils, etc.),
-
Gathering and sorting of the products (for further processing),
-
Extrusion, quenching and pelletizing of recycled plastics.
In mechanical recycling, the plastic waste is preferred to contain as few types of polymers as possible to provide products with good homogeneity and characteristics. In case of waste streams with high contamination contents or high diversity of plastics, it is more difficult to use mechanical recycling. Therefore, separation is a critical step in fabrication of highquality products via mechanical processing. In this case, different separation techniques are developed including manual sorting: triboelectric, mid infrared (MIR), near infrared (NIR), selective dissolution/precipitation technique, and density segregation [10-11]. Nevertheless, industrial applications of these techniques appear unlikely to meet the ideal separating expectations in terms of precision and economic efficiency. Some of the most important separation techniques are summarized in Table 1.2.
5
Table 0.2: Common separation techniques for plastic recycling [5]. Method
Separation property
Comments
Manual sorting
Only for large items
Very labor intensive, bad working environment
Triboelectric
Based on electrostatic charge
Only for clean, dry and nonsurface-treated products
Mid infrared (MIR)
Fundamental vibrations
Surface sensitive, can measure black items, expensive
Near infrared (NIR)
Fundamental vibrations
Not applicable for dark or black products, expensive
Density sorting
Large difference in density
Fillers may alter density. Low cost
For the moment, sorting techniques based on density segregation in a float-sink tank or hydrocyclone are considered as the most industrially efficient methods in recycling process [12]. Using this technique, the plastic flakes are separated in two fractions with respect to water density. Generally, the light fraction comprises polyethylene and polypropylene, while the heavy fraction mainly contains polymers with higher densities including polyvinyl chloride, polystyrene and polyethylene terephthalate (Figure 1.4). Floatation is a suitable method for waste plastic separation for three reasons: 1) it is an easy and fast procedure, 2) separated plastics (especially PE and PP in light fraction) have chemical similarity and 3) absence of solvents and simplicity of this technique makes it a favorable method from environmental and economic points of view. Nevertheless, owing to the similar properties of PE and PP, it is technically uneconomical to perform additional separations on the light fraction [13].
6
Figure 0.4: Schematic representation of plastics separation by a floatation technique [14].
1.3 Compatibility of polyethylene and polypropylene Although the chemical structure of the plastics in the light fraction is considered relatively similar, the mechanical properties of the recycled blend (of PE and PP) are still lower than for neat plastics [15]. Blends of PE and PP are reported to have low tensile elongation at break and impact strength. This behavior is related to the fact that most polymers are basically immiscible (or have limited miscibility) which is caused by differences in their chain configuration [16]. It is reported that even different grades of polyethylene do not show complete miscibility when blended [17]. Lednicky [18] reported that introduction of 7
20% PP in low density polyethylene led to high phase separation in the blend. Low voltage scanning transmission electron microscopy (LV-STEM) of such blends (Figure 1.5) showed that the PP phase formed spherical domains which proved high level of interfacial tension.
Figure 0.5: LV-STEM of low density polyethylene/PP (80/20) blends (frame width: 9.7 μm) [18].
Clemons [19] studied the effects of different concentrations of virgin polypropylene and polyethylene on the mechanical properties of their blends. He reported that inclusion of only 25% of polypropylene reduced tensile elongation at break of polyethylene from over 1200% to less than 550%. Notched Izod impact energy of PE also decreased from 125.8 J/m to 35.3 J/m after incorporation of PP. These observations were considered as a result of low compatibility between both phases in the blend. It is reported in the literature that low compatibility between PE and PP is due to their positive energy of mixing (ΔGmix). Energy of mixing can be calculated as [20]: ΔGmix = ΔHmix - T ΔSmix
(1.1)
When blending two high molecular weight polymers, the gain in entropy (ΔS mix) is negligible. Thus, to have a negative energy of mixing, the heat of mixing (ΔH mix) must be 8
negative (the mixing must be exothermic). Negative heat of mixing only occurs in case of specific interactions between the blend components. Three types of blends can be distinguished regarding phase compatibility: -
Completely miscible blends: in this case, the heat of mixing is negative due to high level of interaction between the phases and homogeneity is observed at the nanometer scale or even molecular scale.
-
Partially miscible blends: a small part of polymers is dissolved in the other. In this type of blends, the homogeneity (and characteristics) of the blend are relatively high, an interface is observed between phases and interfacial adhesion is good.
-
Fully immiscible blends: this type of blends has low homogeneity and poor interfacial adhesion between the phases. The interface is sharp and easy to distinguish, and mechanical properties are poor. PE/PP blends are well-known examples of such blends.
1.4 Compatibilization methods for PE/PP blends Compatibilization of plastic blends includes modification of the interface to reduce the interfacial tension and phase separation in the melt. Several compatibilizers and impact modifiers can be used to upgrade waste plastics blend during mechanical recycling. Such compatibilizers can reduce the interfacial tension between each plastic through chemical (reactive) and physical (nonreactive) effects [13]. 1.4.1 Reactive compatibilization In reactive compatibilization of plastic blends, functional or reactive additives are used to interact with both components. Usually a polymer, which is chemically identical with one of the components, is functionalized to create chemical reactivity with the other phase. Chemicals with such chemical groups are presented in Figure 1.6. Chemical bonding between phases in this technique allows the components to be held together by covalent bonds. PE, PP and ethylene-propylene-diene monomer functionalized by maleic anhydride or acrylic acid or poly(ethylene-co-propylene) grafted with succinic anhydride are 9
examples of this technique. However, these grafted polymers are preferred for compatibilization of hydrocarbon polymers (such as PE and PP) with polar polymers (such as polyethylene terephthalate) [21].
Figure 0.6: Chemicals with functional groups for chemical compatibilization [4].
Another approach for reactive compatibilization of plastic blends is through incorporation of suitable monomers and initiators to perform compatibilization reactions by reactive extrusion. Free radical initiators such as organic peroxides are a common example for this method [22]. During processing of PE/PP blends with peroxides, PE tends to crosslink while PP degrades. In this condition, introduction of co-reactants is believed to promote production of PE-PP copolymers. Cheung et al. [23] studied the effects of compatibilization of linear low density polyethylene/PP (50/50) blends through injection of an organic peroxide
(2,5-dimethyl-2,5-bis-(t-butyl-peroxy1)hexyne-3)
during
extrusion.
Using
scanning electron microscopy, they reported that maximum domain size of the dispersed 10
phase (PP) decreased from 4 μm to less than 2 μm after addition of 0.25% by weight of initiator. Elongation at yield was reported to increase by 37%, while impact strength and yield strength decreased by 17 and 54%, respectively. 1.4.2 Nonreactive compatibilization In this method, compatibilization is achieved through addition of graft or block copolymers containing segments identical or compatible with blend components. Efficiency of such compatibilizers depends on their preferential location at the interface. Some examples of nonreactive compatibilizers include styrene-ethylene-butylene-styrene tri-block copolymers (SEBS) [24], ethylene-propylene-rubber (EPR) [25-28], ethylene-propylene-dienemonomer (EPDM) [19,29,30] and ethylene-vinyl-acetate (EVA) [24,31]. According to the literature, diblock copolymers containing segments identical to the blend components are more suitable in comparison with triblock and graft copolymers [5]. Nonreactive compatibilizers have been extensively studied for modification of PE/PP blends regarding their low cost, easy processing conditions and high performance. Clemons [19] studied the effects of EPDM inclusion as a compatibilizer for blends of virgin PE with different concentrations of PP. From SEM observations, he reported that blends of PE/PP (50/50) had a co-continuous morphology in which the PE domains could be easily distinguished from PP in blend without compatibilizer (Figure 1.7). Inclusion of 10% EPDM was shown to increase the compatibility between PE and PP significantly. The domains were not distinguishable for compatibilized blends. Incorporation of 10% EPDM also resulted in 43% enhancement in tensile elongation at yield of the blend. It was also reported that tensile strength and modulus of PE/PP (50/50) blends decreased by 17% and 27%, respectively, due to the elastomeric nature of the EPDM phase (rubber).
11
(a)
(b)
Figure 0.7: SEM micrographs from fractured surfaces of fractured impact specimens of PE/PP: 50/50 blend (a) without and (b) with 10% EPDM [19].
Souza and Demarquette [24] also added different compatibilizers including EPDM, EVA and SEBS to the blends of PE/PP. SEM micrographs revealed that all blends showed droplet dispersion morphology type. Small amplitude oscillatory shear analysis was also performed to determine interfacial tension between PP and PE. Increase in compatibilizer concentration led to an exponential decrease in both interfacial tension between the phases and average droplet radius in the blends. EPDM was also shown to be more efficient in comparison with other compatibilizers. Recently, ethylene-octene-copolymer (EOC) has been synthesized by chain shuttling technology. These copolymers consist of crystallizable ethylene/α-olefin blocks with very low comonomer content and high melting temperature, as well as amorphous ethylene/αolefin blocks with high comonomer content and low glass transition temperature. It is believed that the crystalline and amorphous segments of EOC are compatible with polyethylene and polypropylene phases, respectively. Such characteristic has introduced EOC as a promising compatibilizer for PE/PP blends [32]. Lin and co-workers [33] investigated the effects of incorporation of several compatibilizers on blends of PP/HDPE (70/30). The compatibilizers used were a multi-block ethylene– octene copolymer (EOC), two statistical ethylene-octene copolymers, two ethylene12
propylene rubber (EPR) and a styrenic block copolymer (SBC). They reported that addition of multi-block EOC resulted in best combination of low brittle-to-ductile transition and high toughness. Morphological observations through atomic force microscopy (AFM) showed that the compatibilizer was preferentially located at the interface between polypropylene and high density polyethylene (Figure 1.8). Dark regions around HDPE domains indicate the presence of the compatibilizer at the interface.
(a)
(b)
Figure 0.8: AFM micrographs of PP/HDPE compounds (a) without compatibilizer and (b) with multi-block EOC (frame width: 20 μm) [33].
13
14
Chapter 2. Natural Fiber Composites Polymers usually have relatively low mechanical properties, modulus and strength, compared to other solids. One of the most acceptable methods for enhancement of their properties is through inclusion of a dispersed phase with high stiffness and strength. The dispersed phase is usually in form of long/short fibers or even small (micro or nano size) particles [34-36]. 2.1 Natural vs. artificial reinforcements Artificial reinforcements (glass, talc, calcium carbonate, etc.) are proven to have higher mechanical properties compared to natural reinforcements (specifically plant or vegetable fibers). Natural fibers, on the other hand, are less expensive and are considered eco-friendly materials. Table 2.1 presents the energy consumption for production of selected natural fibers in comparison with glass fiber [37]. It is clear that production of natural fibers, such as flax, consumes significantly less nonrenewable energy (9.55 MJ/kg) than glass fiber (54.7 MJ/kg). Table 0.1: Energy consumption (MJ/kg) for production of different fibers [37]. Glass fiber mat
Flax fiber mat
China reed fiber
Raw materials
1.7
Seed production
0.05
Cultivation
2.50
Mixture
1.0
Fertilizers
1.00
Transport plant
0.40
Transport
1.6
Transport
0.90
Fiber extraction
0.08
Melting
21.5
Cultivation
2.00
Fiber grinding
0.40
Spinning
5.9
Fiber separation
2.70
Transport fiber
0.26
Mat production
23.0
Mat production
2.90
Total
54.7
Total
9.55
Total
3.64
15
Joshi et al. [37] reviewed the comparative life cycle assessment studies of selected natural fiber and glass fiber composites. The study covered the environmental aspects of both composites through their life cycle considering their production, use and end of life management options (i.e. recycling, incineration and disposal). They suggested two different generic life cycles for glass fiber and natural fiber reinforced composites as presented in Figure 2.1. They concluded that natural fibers are environmentally superior compared to glass fiber because: 1) production of natural fibers leads to lower environmental impact compared to glass fiber, 2) natural fibers can be used at high concentrations (due to their low price and density) which results in lower amount of polymers (petroleum based) in the final products, 3) lower density of natural fibers results in better fuel efficiency in automotive applications, and 4) end of life incineration of natural fibers leads to energy recovery.
Figure 0.1: Generic life cycles of (a) glass fiber and (b) natural fiber reinforced composites [37].
It is notable that some natural fibers (such as pineapple and banana leaf fibers) are not even considered as products. Such materials are agricultural waste which can be purchased at 16
very low prices. Such advantages, combined with low density and low abrasiveness, have recently driven a considerable number of researchers to develop engineered products containing naturally made fillers [38-40]. 2.2 Characteristics of natural fibers Lignocellulosic fibers are usually grouped into three types depending on their source: 1) seed hair (cotton), 2) bast fibers (jute and flax fiber), and 3) leaf fibers (sisal and abaca). Mechanical properties of natural fibers vary with chemical composition which depends on their type, age and climatic conditions. Components of natural fibers include cellulose, hemi-cellulose, lignin, waxes, water soluble components and pectin. Each natural fiber is a bundle of cellulose fibers which are covered with hemi-cellulose and lignin as matrix. Table 2.2 shows chemical compositions of selected lignocellulosic materials [41]. Table 0.2: Chemical composition of lignocellulosic fibers (weight %) [41]. Fiber
Cellulose
Lignin
Hemicellulose
Pectin
Ash
Hemp
57-77
3.7-13
14-22.4
0.9
0.8
Jute
41-48
21-24
18-22
-
0.8
Flax
71
2.2
18.6-20.6
2.3
-
Sisal
47-78
7-11
10-24
10
0.6-1.0
Kenaf
37-49
15-21
18-24
-
2-4
Cellulose is considered the main component of all natural fibers. The chemical structure of a typical cellulose molecule is shown in Figure 2.2 [42].
17
Figure 0.2: Chemical structure of a repeating unit in cellulose molecule [42].
Thermosets, thermoplastics and even elastomers are commonly used as matrices for natural fiber composites. Among these, thermoplastics have attracted more attention because of their ease of reprocessing and recycling. Other advantages of thermoplastics include their design flexibility and simple processing methods. Thermosets and elastomers, on the other hand, have crosslinked structures which do not allow them to be reprocessed by conventional methods. Among thermoplastics, however, only a handful can be used as matrix for natural fibers. Since natural fibers are prone to thermal degradation at high temperature, thermoplastics with high processing temperatures (higher than 200°C) cannot be used easily as matrix. This is why polyethylene and polypropylene (low melting points) are the most commonly used matrices for natural fiber composites [43]. Presence of high concentrations of hydroxyl groups on different components of natural fibers (especially cellulose) results in their hydrophilic behavior which leads to low surface adhesion with hydrophobic polyolefin matrices (such as polyethylene and polypropylene). Main disadvantages of low surface interaction are reduced mechanical properties (due to phase separation and also low homogeneity in composite) and high water absorption in natural fiber composites [44-46]. Rachini and coworkers [47] studied the effects of hemp fibers in polypropylene. Via morphological observations of fractured surfaces, they concluded that hemp has low surface interaction with PP. It is clearly shown in Figure 2.3 that the surface of natural fibers is completely clean from matrix and gaps exist at the interface between both phases. 18
Figure 0.3: SEM micrographs from fractured surfaces of hemp filled PP [47].
Sgriccia et al. [48] reported that adding only 15% of natural fibers led to significant increase in water absorption of epoxy based composites, while glass filled composites showed lower water uptakes (Figure 2.4). Hydrophilic behavior of natural fibers results in higher water uptake. Low compatibility between matrix and fibers also increases the ability of water molecules to penetrate through the composite.
Figure 0.4: Effect of different fillers on water uptake of an epoxy composite [48].
19
2.3 Modification of natural fiber composites Several methods have been developed to modify the surface interaction between cellulosic fibers and polyolefins: 1) surface modification of the fiber, 2) modification of the matrix phase (through grafting active groups), and 3) addition of a third (compatibilizer) phase [40]. 2.3.1 Surface modification of natural fibers Surface of natural fibers can be modified via different (chemical or physical) methods. 2.3.1.1 Chemical surface treatment of natural fibers Silane treatment is one of the most frequently used chemical modifications to increase surface adhesion between lignocellulosic fibers and polymers. Silanes are chemical compounds (with SinH2n+2) and are also commonly used to compatibilize glass fiber with polymers. In the presence of moisture, silanols are produced through hydrolyzation of alkoxy groups. Silanol groups then react with hydroxyl groups on the surface of natural fiber and create stable covalent bonds with the cell walls. Therefore, hydrophilic behavior of the fiber decreases which leads to enhanced fiber/matrix interaction. An example of silane treatment of natural fibers is as follows [41]: CH2CHSi(OC2H5)3 + H2O → CH2CHSi(OH)3 + 3 C2H5OH
(2.1)
CH2CHSi(OH)3 + Fiber-OH → CH2CHSi(OH)2O- Fiber + H2O
(2.2)
Abdelmouleh et al. [44] studied the reinforcement of low density polyethylene (LDPE) and natural rubber using different types of cellulosic fibers. Cellulose fibers were added to the matrices before and after chemical treatments using three silane coupling agents namely: γmethacryloxypropyltrimethoxy
(MPS),
γ-mercaptoproyltrimethoxy
(MRPS)
and
hexadecyltrimethoxy-silane (HDS). They proposed the following schematic illustration (Figure 2.5) to explain the interaction between LDPE and cellulosic fibers treated by MPS.
20
Figure 0.5: Schematic representation of interaction between LDPE and MPS-modified fiber [44].
Increase interaction between matrix and cellulosic fiber was confirmed by morphological and mechanical characterizations. Morphological observations revealed that surface treatment with silanes increase the compatibility between LDPE and cellulosic fibers (Figure 2.6). It is seen in Figure 2.6 that the surface of treated fiber is completely covered with the polymeric matrix.
(a)
(b)
Figure 0.6: Effect of cellulosic fibers surface treatment with MPS: a) untreated and b) treated fiber [44].
21
Enhancement in surface interaction between lignocellulosic fibers and polyolefins can also be achieved via introduction of an acetyl functional group (CH3COO-) to the surface of the fibers. Acetylation of natural fibers with acetic anhydride (CH3-C(=O)-O-C(=O)-CH3), for instance, substitutes hydroxyl groups on cellulose molecules with acetyl groups which leads to hydrophobic behavior of fibers [41]. The reaction is as follows: Fiber – OH + CH3-C(=O)-O-C(=O)-CH3 → Fiber – OCOCH3 + CH3COOH
(2.3)
Acetic acid (CH3COOH) is produced as a by-product of the reaction which must be removed from the fiber before introduction into polyolefins. Rong et al. [49] studied the effect of sisal fibers acetylation to reinforce epoxy resins. Acetylation was performed by a 50% acetic acid aqueous solution for 5 minutes (fiber/solution ratio: 1/25). The authors claimed that improved interfacial bonding is due to creation of hydrogen bonds between acetyl groups (on fiber surface) and hydroxyl or amine groups in the epoxy resin. Maleated coupling agents can also be used for surface covering of natural fibers in order to compatibilize them with polymers. In this approach, natural fibers are soaked in a solution of maleic anhydride or maleated polymers. Maleic anhydride groups react with hydroxyl groups on the surface of cellulosic fibers which results in decreased hydrophilic behavior. Figure 2.7 represents the reaction of maleated polypropylene (MAPP) with cellulose molecules [50].
22
Figure 0.7: Surface treatment of natural fibers with maleated polypropylene [50].
Modified surface of lignocellulosic fibers allows better interaction with thermoplastic matrices through decreased fiber hydrophilic behavior and also physical entanglement of PP chains with the matrix molecules. Mohanty and Nayak [51] investigated the effects of natural fibers surface covering with MAPP in PP/jute composites. They immersed jute fibers in MAPP/toluene solution at 100°C. They reported that surface covering of jute fiber in MAPP solution (with 5% MAPP) led to increased tensile strength in the composites with 30% of jute. Tensile strength of composites increased from 24.4 MPa to over 31 MPa. 2.3.1.2 Physical surface treatment of natural fibers Alkalization (or mercerization) is a common method for physical treatment of natural fiber surfaces. In this method, lignocellulosic fibers are immersed in aqueous NaOH solution for a period of time. NaOH solution dissolves lignin, wax and oils from the fiber surface and leaves a clean and porous cellulosic surface. This treatment leads to higher specific mechanical properties since cellulose has much higher mechanical properties compared to lignin and also increases specific surface area leading to better interaction with the matrix. The main drawback of this approach is increased hydrophilic behavior of natural fibers 23
which happens due to higher concentration of hydroxyl groups on cellulose molecules compared to lignin. This is why this method is usually proposed along with other chemical modification methods such as addition of silanes or maleic anhydride [40]. Sgriccia et al. [48] studied the effect of alkali treatment on morphological and water absorption properties of several natural fibers. The fibers were submerged in 5% solution of sodium hydroxide for one hour at room temperature. They concluded, via SEM microscopy, that alkali treatment of hemp fiber led to hemicellulose and lignin removal from the fiber surface. As presented in Figure 2.8, surface of hemp fibers is much clearer after alkalization. The authors reported however that fiber alkalization increased their water uptake. After 700 hours of immersion in distilled water, epoxy composites based on untreated hemp absorbed around 16% water, while alkali treated hemp had a water uptake of around 22%.
(a)
(b)
Figure 0.8: SEM micrographs of (a) untreated and (b) alkali treated hemp fibers [48].
2.3.2 Modification of polymeric matrix Modification of the matrix is also a common method to enhance surface interaction between lignocellulosic fibers and polyolefins. This type of compatibilization is usually 24
achieved via two different approaches: (1) inclusion of a maleated polymer to the matrix and (2) chemical modification of the matrix by grafting active groups (especially maleic anhydride). In both methods the active group (maleic anhydride) reacts with the hydroxyl groups on the natural fiber surface, while the polymeric chains entangle with the matrix [40,52]. Efficiency of maleic anhydride grafted polymers (as compatibilizer) depends on acid number and molecular weight. Macromolecules with longer chains have better ability to entangle with the matrix. Acid number, on the other hand, represents the concentration of maleic anhydride groups in maleated polymers. Needless to say that molecules with higher acid number provide higher interaction with natural fibers. Keener et al. [52] studied the effects of different grades of MAPP as compatibilizer in PP/agro-fiber (jute and flax) composites. They reported that addition of different types of MAPP increased the mechanical properties of both PP/jute and PP/flax composites, while the increase was more significant for specific grades of MAPP. For instance, adding 3% of Epolene G-3003 and Epolene E-43 to PP/flax (70/30) increased tensile strength by 50% and 38%, respectively. It is reported that the acid number of Epolene E-43 is 5 times higher and its molecular weight is 80% lower compared to Epolene G-3003. They also supported their findings with morphological observations. It is clear in Figure 2.9 that adding MAPP resulted in increased surface interaction between natural fibers and PP as less fiber pull-out is seen and the natural fiber surfaces are covered with matrix molecules (Figure 2.9-b).
25
(a)
(b)
Figure 0.9: SEM micrographs of PP/agro-fiber composites (a) before and (b) after compatibilization with MAPP [52].
2.4 Matrices for natural fiber composites Selection of a suitable matrix for natural fibers is a vital step in the fabrication of composites. Shape, environmental tolerance, surface appearance and total durability of natural fiber composites are controlled by the matrix phase. Currently, 80% of the composite matrices are based on non-renewable petroleum resources which have caused concerns due to environmental issues [53]. Generally, two different approaches have attracted attention to decrease these environmental issues caused by the production of natural fiber composites. 2.4.1 Bio-based polymers Bio-based plastics are suggested as an important substitute to petroleum based plastics to reduce the dependence on petroleum and also decrease the environmental impacts coupled with use of petroleum resources. Several new polymers have been developed from renewable resources. Figure 2.10 presents current and emerging plastics as matrices for composites regarding their biodegradability [53]. It is clear that bio-based plastics are environmentally degradable in comparison with common petroleum based plastics. 26
Figure 0.10: Current and emerging matrices for natural fiber composites and their biodegradability [53].
Poly(lactic acid) (PLA) has been investigated as a matrix for fabrication of kenaf fiber reinforced composites by a number of researchers. PLA is a bio-based material that can be produced from lactic acid (a fermentable sugar) and polyhydroxyalkanoate (PHA). Huda et al. [54] studied the mechanical properties of kenaf fiber reinforced poly(lactic acid) laminated composites. They also investigated the effects of compatibilizing kenaf fibers with PLA through alkalization and silane-treatments. They reported that inclusion of both silane-treated and alkali-treated kenaf fibers led to increased PLA mechanical properties. Alkali treatment resulted in 50% improvement in impact strength of surface-treated 27
composites, while silane treatment resulted in more improvement in flexural performance. The flexural modulus and strength of surface-treated composites increased up to 69% and 50%, respectively. Starch is also a naturally occurring polymer which has attracted attention as a plastic matrix for natural fiber composites. Natural fiber composites based on biodegradable starch matrix and sisal fibers were made by Alvarez et al. [55-57]. The thermal, rheological and creep properties of the composites were studied extensively. 2.4.2 Use of waste plastics As mentioned in Chapter 1, recycling of plastic materials is a promising approach to deal with environmental impacts caused by waste plastics. Among all waste plastic materials, polyethylene and polypropylene have attracted a great deal of attention due to ease of separation from other polymers and processing in the form of PE/PP blends. Many researchers have also investigated the possibility of such blends to serve as the matrix phase for natural fiber composites. Clemons [19] studied the characteristics of composites based on PE/PP blends containing wood flour. Although he did not use recycled materials, the fact that the matrix was a blend of PE and PP simulates the recycling conditions. He studied the effects of maleated compatibilizers on surface interactions between wood flour and matrix and also between different phases in the matrix (PE and PP). He reported that inclusion of maleated ethylenepropylene-diene monomer (MA-EPDM) as compatibilizer between all three phases (PE, PP and wood flour) leads to increase in strength and deformability of wood plastic composites. Tensile strength of composite with PE/PP (75/25) as matrix and 30% wood flour increased from 23 to 25 MPa after inclusion of 10% MA-EPDM. However, tensile modulus of the composite decreased around 32% due to inclusion of the rubber phase. Najafi et al. [58] also fabricated composites based on virgin and recycled plastics (PE and PP) and 50% wood sawdust. They concluded that tensile modulus of composites containing recycled PE/PP (50/50) as matrix (6.5 GPa) was comparable with composites based on 28
virgin PE/PP (50/50) matrix (7.1 GPa). Tensile strength of composites with recycled PE/PP blend as matrix was less than 10 MPa compared to around 11 MPa for composites with the virgin matrix. Gao et al. [59] fabricated wood plastic composites based on PP/PE (80/20) blends containing 60% wood flour. In order to increase the compatibility between the thermoplastic matrix and natural fibers, they grafted maleic anhydride groups on PE and PP through twin-screw extrusion in presence of dicumyl peroxide (DCP) as initiator. The authors reported that increasing maleic anhydride concentration up to 1% resulted in higher mechanical properties such as flexural and impact strength. Flexural modulus of the composite with 0.5% maleic anhydride (5.1 MPa) was lower than the unmodified composite (5.8 MPa), while inclusion of higher concentrations of maleic anhydride resulted in increased modulus. Incorporation of 1.5% of maleic anhydride resulted in flexural modulus of 6.2 MPa. Dintcheva and La Mantia [60] performed one of few researches dealing with separation and recycling of true light fraction of plastic waste stream and studied the effects of inclusion of wood flour to such blends. They reported that inclusion of wood flour led to increased tensile modulus of PE/PP blend, while tensile strength, elongation at break and impact strength decreased. For instance, tensile modulus increased from around 600 MPa to around 900 MPa when wood content increased from 20 to 40%, while tensile modulus decreased from 12.5 MPa to 11.5 MPa. Dintcheva et al. [61] also studied the effects of different filler types (namely wood fiber, glass fiber and calcium carbonate) and processing equipment (discontinuous mixer, single and twin-screw extruder) on the mechanical properties of light fraction based composites. They reported that inclusion of 20% of all types of fillers led to similar effects on the mechanical properties of the composites as presented in Figure 2.11. They also reported that although mechanical properties of light fraction (with no filler) was not affected by the processing method used, wood fiber based composites were shown to have higher mechanical performance after injection molding in comparison with compression molding. 29
This might be due to increased fiber alignment caused by higher levels of shear stress during injection molding.
Figure 0.11: Elastic modulus (E), tensile strength (TS), elongation at break (EB) and impact strength (IS) of light fraction (LF) based composites with different fillers (WF: wood fiber, GF: glass fiber) [61].
2.5 Structural design of composite materials Composite materials are extensively used in many modern constructional applications to provide certain characteristics. Structural modification of composites, in terms of multilayered structures results in fabrication of more efficient designs which provide great potential in comprehensive functions. For multilayered structures, the flexural properties (such as stiffness) are dependent on layer configuration and are no longer determined via the simple rule of mixture. This behavior can be used in optimization of layer configuration, thickness and stacking sequence [62]. Today, multilayered structures are attracting increasing attention in several applications due to their efficiency and advantages regarding load distribution. These structures can have layers for special purposes such as damping, decrease density or protection from environmental effects. Key feature of such structures include inhomogeneous distribution of mechanical properties though the thickness. Such structures can be produced using a 30
wide range of materials such as metal alloys, ceramics, polymers, foam and wood (plywood). Tables 2.3 and 2.4 present some applications with positive/negative effects of such structures, respectively [63]. Table 0.3: Applications of multi-layered structural composites [63]. Branch of industry
Application
Rocket construction
Load-carrying structural elements, fuel tanks, aerial elements
Aircraft construction
Tail assembly, stabilizers, inner lining of cabins
Machine building
Transmission cases, gear wheels, machine elements
Automotive industry
Wheel rims, trunk covers, hoods, steering columns, inner lining of cabins
Medical equipment
Implants, artificial joints
Sports industry
Surfing, skis, clubs, canoes
Telecommunication
Parabolic aerials
Oil production
Elements of frames for offshore drilling rigs
Civil engineering
Facing materials
Energetics
Rotor blades of wind power stations
Industrial engineering Reservoirs, pipelines
Table 0.4: Advantages and disadvantages associated with application of multilayered structures. [63]. Advantages
Disadvantages
High rigidity characteristics relative to mass
Loss of strength due to aging of adhesive joints
Thermo-insulation
High technological requirements to the accuracy of production
Sound-proofing High fatigue characteristics
Necessity of modifying the methods of nondestructive testing of structures
High corrosion resistance
High sensitivity to impact loads Brittleness
Low tendency to stability loss
31
Decrease in the number of assembling operations due to development of more complex components
-
In case of reinforced plastics, fabrication of multilayered structures suffers from a lack of information about the effects of design parameters on their mechanical performance. The most important design parameters for multilayer composite structures include: fiber orientation, concentration of reinforcement phase and cross-sectional arrangement of each layer. Dyer et al. [64] studied the effects of cross-sectional design of different fiber reinforced composites for prosthodontic applications. Figure 2.12 presents the changes in elastic modulus and toughness with unidirectional R-glass fiber and different cross-sectional designs.
Figure 0.12: Effect of different cross-sectional designs on properties of composites. Lines depict corresponding fiber placement. Superscript groups are not statistically different for each test [64].
32
It is presented in Figure 2.12 that alteration in structure design (at equal fiber content) led to noticeable change in mechanical properties of the composites. The authors reported that flexural modulus increased when one group of glass fiber reinforcements was located in the compression side of the specimen (intrados) during flexion tests. They also reported that toughness increased when the fiber group was placed in the tension side of the specimens and also when fiber content increased. Salomi and coworkers [65] studied the flexural performance of double-layered structural composites under different load directions. The structure contained a layer of PP/E-glass fiber (40/60) coupled with a layer of linear low density polyethylene. Figure 2.13 shows load-deflection curves of the structures under flexural load from different directions: i.e. the composite layer is in the intrados or extrados. It is clear that flexural strength and deformability were higher when the composite layer is in intrados. Beam stiffness, on the other hand, was shown to be independent of load direction.
Figure 0.13: Three-point flexion test when the composite layer is in extrados or intrados sides (the intrados represents the position of layer placed under the load nose, while the extrados represent the position of other layer in a double-layer laminate) [65].
In addition, the authors studied the failure modes of specimens in bending tests. Figure 2.14 shows that the failure started in composite layer for both load directions. In cases where the 33
composite was in extrados, tension failure occurred (Figure 2.14-a), while buckling was responsible for composite failure in intrados layer (Figure 2.14-b).
(a)
(b)
Figure 0.14: Fractured areas of specimens from flexion test with composite layer placed in the (a) extrados side and (b) intrados side [65].
Smith and Partridge [62] studied the flexural properties of planar multilayered structures based on two dissimilar materials, namely high strength titanium alloy (Ti-834) and titanium metal matrix alloy (Ti-MMC). They studied the development and evolution of several types of two-material multilayer systems and Figure 2.15 presents the effects of changing layer configuration and thickness on the flexural modulus of such beams. As shown in Figure 2.15, the layer position strongly influenced the stiffness of symmetric and asymmetric systems.
34
Figure 0.15: Flexural modulus of two and three layered systems with dissimilar materials (different colors in these structures represent the different materials: 834 and MMC for white and black layers, respectively) [62].
For symmetric and asymmetric multilayered systems with a given volume fraction of each phase (50/50), the flexural modulus is shown to vary significantly with altering layer configuration (Figure 2.16). Maximum flexural modulus was observed when the stiffer layers (Ti-MMC) were placed as the top and bottom skins. On the other hand, the minimum stiffness was achieved when Ti-834 (less stiff) were in the top and bottom skins. For a given volume fraction of layers, all other proposed configurations where shown to have intermediate stiffness values. This proves that the flexural moduli of multilayered structures are controlled by the skin layers. They also concluded that layer position strongly changes other mechanical properties such as strength, impact toughness and damping capacity.
35
Figure 0.16: Flexural modulus of symmetric and asymmetric systems with similar volume fraction of phases (MMC in black and 834 in white) [62].
2.6 Thesis objectives and organization Despite the extensive efforts towards improvement of PE/PP blends for plastic recycling purposes, only a limited number of research works were found in the literature dealing with true plastic waste materials. Therefore, the main objective of this research work is to adopt the current enhancing techniques i.e. polymer compatibilization, inclusion of wood fiber residue and structural design, on municipal plastic waste streams to improve their mechanical and morphological performance. Chapter 1 presented a general overview of plastic recycling with the aim of introducing the mechanical techniques as the most appropriate method in recycling municipal plastic waste streams. In addition, the light fraction (PE/PP blend) was introduced as the major component of municipal plastic wastes. Then, the technical challenges in mechanical 36
recycling of light fraction were studied and compatibilization techniques were proposed to compensate the low compatibility of the different components in such waste streams. In addition, a literature review on plastic recycling was presented along with each topic in this chapter. In Chapter 2, wood plastic composites were introduced as a fascinating application of recycled plastic materials. Inclusion of wood fibers into recycled polymer matrices was shown to give rise to low costs and high mechanical performance products with particular emphasis on improving the stiffness of recycled plastics. To improve the performance of these products, the compatibility of wood fibers and polymer matrices was discussed. In addition, structural design was proposed to improve the mechanical performance of composite materials. A literature review was also performed on these topics. Based on this first part, the main objectives of this research work include: 1) Reprocessing of municipal waste light fraction plastics to fabricate homogenous thermoplastic resins using mechanical recycling. 2) Fabrication and characterization of wood plastic composites from the light fraction of municipal plastic wastes and wood processing residues. 3) Compatibilization of PE and PP as the main components of the light fraction along with impact modification of these materials. 4) Enhancing the compatibility of light fraction resins and wood fiber residues via incorporation of a coupling agent package including MAPE and MAPP. 5) Fabrication of multilayered structural composites, including symmetric and asymmetric structures, based on recycled materials. 6) Studying the design parameters (stacking sequence, layer thicknesses and wood contents, load direction and structure symmetry) on the mechanical performance of three-layered wood plastic composites. 37
7) Investigating the validity of the classical beam theory in predicting the flexural modulus of multilayered composites based on recycled materials. Chapter 3 presents the experimental efforts regarding the first four objectives of this research work. Wood-plastic composites are thoroughly investigated in terms of the effects of compatibilizing technique and wood content on their mechanical and morphological characteristics. The remaining three objectives are achieved through the experimental works presented in Chapter 4. The compatibilized composites of Chapter 3 are used to fabricate three-layered structures in order to optimize the structural performance of these products. A detailed discussion on the effects of design parameters is provided in this chapter with comparison of flexural modulus predictions by the classical beam theory. Finally, the last chapter includes a general conclusion based on the results obtained and suggests a number of potential investigations for future works which might be beneficial for further developments of these materials.
38
Chapter 3. Mechanical and Morphological Properties of Wood Plastic Composites Based on Municipal Plastic Waste
Résumé La production et la caractérisation de composites bois-plastique (WPC: wood plastic composites) à partir de la fraction légère des déchets plastiques municipaux (post-consommation) et de résidus de la transformation du bois (sciure) ont été étudiées ici. Le polyéthylène (PE) et le polypropylène (PP) ont été identifiés par analyse de la composition comme les composants principaux de la matrice. Afin d’améliorer la compatibilité et l’adhésion entre les différentes phases, un copolymère d’éthylène-octène (EOC: ethylene-octene copolymer) a été utilisé pour développer la compatibilité entre les différentes phases polymères tout en agissant comme modificateur d’impact, alors que l’ajout de PE et PP maléatés (MAPE: maleated polyethylene; MAPP maleated polypropylene) a permis de jouer le rôle d’agent couplant entre la matrice polymère et la farine de bois. Les effets combinés de tous les composants ont mené à la production de composites présentant des propriétés morphologiques (dispersion et adhésion) et mécaniques (traction, torsion, flexion et impact) intéressantes après l’optimisation de l’ensemble des additifs (mélanges d’agents couplants).
Mots-clés: Recyclage; Composites Bois-Plastique; Mélanges de Polyéthylène/Polypropylène; Interface Fibre/Matrice; Morphologie; Propriétés Mécaniques.
39
Abstract Production and characterization of wood plastic composites (WPC) from the light fraction of municipal plastic wastes (post-consumer) and wood processing residues (sawdust) were investigated. Composition analysis revealed the presence of polyethylene (PE) and polypropylene (PP) as the two main components of the matrix. In order to improve compatibility and adhesion between all the phases, an ethylene-octene copolymer (EOC) was used to compatibilize the polymer phases and was also acting as an impact modifier, while the addition of maleated polyethylene (MAPE) and maleated polypropylene (MAPP) were acting as coupling agents between the polymer matrix and the wood flour. The combined effect of all the components was found to produce composites with interesting morphological (dispersion and adhesion) and mechanical properties (tension, torsion, flexion and impact) after optimization of the additive package (blend of coupling agents).
Keywords:
Recycling;
Wood
Plastic
Composite;
Polyethylene/Polypropylene
Fiber/Matrix Interface; Morphology; Mechanical Properties.
40
Blends;
3.1 Introduction Over the years, specific properties of polymers gave rise to the production of a wide variety of low-cost and high-performance products. The popularity of these products, however, resulted in an increase in production of plastic wastes. Proper management of such wastes is a crucial matter due to the environmental concerns associated with the durability of polymers. In the past, life cycle of polymeric materials only included one use. After their service life, these products were discarded as wastes to landfills which ultimately led to legislation pressures caused by severe environmental impacts. Inclusion of different recycling processes to the traditional life cycle of polymeric materials is the current strategy to alleviate the problems related to landfill disposal. However, only small amounts of post-consumer plastics are recycled.
There are four accepted techniques to recover plastic wastes including energy recovery (incineration), feedstock recycling (pyrolysis, hydrogenation and gasification), chemical (monomer recovery) and mechanical recycling [66]. Of all these options, mechanical recycling seems to be the most beneficial in the terms of energy efficiency and greenhouse gases emission. Mechanical recycling is referred to processes including physical means (grinding, washing, separating, drying, re-melting and compounding) producing recyclates with desired or at least acceptable end-use properties. The origin of the plastic waste streams is a crucial concept in determining the recycling challenges. The most difficult recyclates are post-consumer plastics, especially of municipal solid waste stream, since a mixture of several commodity polymers and contaminants are obtained. This waste stream is typically composed of polyethylene (PE), polypropylene (PP), polyvinyl chloride (PVC), polystyrene (PS) and polyethylene terephthalate (PET) [67]. Mechanical recycling of such a heterogeneous waste needs some preliminary separation steps to improve resin quality and processability. But a complete sorting of these materials is economically nonviable and even in some cases impossible [5]. However, the easiest sorting method proposed for this waste stream is based on density segregation (floatation in sink-float tanks of water). In this technique, the materials are separated in two fractions, in which the light 41
fraction, which floats on water surface, mainly includes polyethylene and polypropylene with the possibility of some impurities. Although polyethylene and polypropylene have many similarities in their structures and properties, their blends show a severe deterioration of morphological and mechanical properties. This behavior is related to immiscibility and incompatibility between both components [16]. Incorporation of a proper compatibilizer into such a thermodynamically immiscible blend lowers the separation tendencies of the components and leads to [67]: lower interfacial tension, finer morphology and better phase dispersion, stabilization of the generated morphology, and better adhesion between the polymer domains which facilitates stress transfer and improves mechanical performance of the final products. Commonly used compatibilizers include graft or block copolymers with chemically similar segments or even identical with the blend components. Compatibilization is therefore achieved through physical compatibility of the blocks with each phase leading to increased interfacial adhesion between the components. Some of the most frequently used compatibilizers for PE/PP blends include: styrenic block copolymers (SBC) [24], ethylene-propylene elastomers (EPR) including ethylene-propylene copolymer (EPM) [25-28] and ethylene-propylene-diene copolymer (EPDM) [19,29,30], ethylene-vinyl acetate (EVA) [24,31] and ethylene-octene copolymers (EOC) [32,33]. In addition, incorporation of wood flour into such waste plastic matrix can lead to the development of a new value-added group of composites from recycled origin. Despite the obvious advantages of natural fiber composites [53], there are some limitations in their use for structural applications. The main problem arises from the inherent incompatibility with most polymer matrices due to the high cellulose content of natural fibers. The presence of polar hydroxyl (-OH) groups on cellulose leads to hydrophilicity of such fibers which is in contrast with hydrophobic matrices [68]. Low compatibility between both phases results in poor interaction and therefore poor ability for the matrix to transfer the applied loads to the reinforcing fibers. To solve the problem, maleated polyolefins are commonly used as coupling agents to establish good interaction between each phase in the composite [52].
42
So far, studies have been performed on the production of natural fiber composites based on PE/PP blends [19,52,58,60,61,69,70]. However, very few focused on municipal plastic wastes as the matrix [60,61]. Therefore, the main objective of this work was to produce WPC based on wood sawdust and PE/PP blends from a post-consumer origin to produce composites based on almost 100% recycled raw material. A strategy was developed to optimize the mechanical properties of such composites based on a blend of coupling agents and compatibilizer addition with respect to wood content [71]. A complete morphological and mechanical characterization of the composites is also presented. 3.2 Experimental 3.2.1 Materials Flakes of post-consumer plastics were received from RECYC RPM (Quebec, Canada). These materials were collected from municipal solid waste. Using density segregation, i.e. flotation in water, polyethylene and polypropylene were separated as the light fraction. The material composition was simply determined via differential scanning calorimetry (DSC) (Figure 3.1). DSC was performed using a Perkin Elmer apparatus model DSC7. The sample was heated from 60 to 380˚C with a heating rate of 10˚C/min under nitrogen atmosphere.
Figure 0.1: DSC curve of the municipal plastic waste light fraction used. 43
For additional insight, Fourier transform infrared (FTIR) spectroscopy of the light fraction is presented in Figure 3.2. The FTIR spectra were recorded using a Nicolet Magna 850 spectrometer (Thermo Scientific, Madison, WI) equipped with a liquid-nitrogen-cooled narrowband MCT detector with Golden-Gate (diamond IRE) ATR accessories (Specac Ltd., London, U.K.). The spectrum was obtained from the acquisition of 128 scans at 4 cm-1 resolution, from 4000 to 750 cm-1 using a Happ-Genzel apodization. The FTIR spectrogram indicates the presence of polyethylene and polypropylene as the main components of the light fraction in post-consumer plastic blends. In addition, C=O signals are due to the possible oxidation of plastic materials or the presence of contaminants such as paper, carbonates, etc.
Figure 0.2: FTIR spectrogram of the municipal plastic waste light fraction used.
From both DSC and FTIR investigations, the analysis revealed the presence of only two components (PE and PP) in the light fraction of recycled plastics (contaminant materials are assumed to be less than 1% which is about the detection limit of each method). The wood flour used was a blend of sawdust from different species (sieved between 125 μm and 1 mm, with an average of 330 µm) and was kindly supplied by the Department of Wood and Forest Sciences of Université Laval.
44
An ethylene-octene copolymer (Engage 8180, Dupont-Dow Chemical) was used to compatibilize the PE/PP blend. This material is an olefin block copolymer (OBC) with 28 wt.% of octane content, melt flow index (MFI) of 0.5 dg/ min (190°C, 2.16 kg weight- ASTM D1238) and density of 0.865 g/cm3. In addition, maleic anhydride-grafted polyethylene (MAPE) and polypropylene (MAPP) were used to enhance the compatibility of wood flour with the polymer blend. Both coupling agents were purchased from Westlake Chemical Corporation. The MAPE was Epolene C-26 with an average molecular weight of Mw = 65000 g/mol, acid number of 8 (mg KOH/g) and melt flow index of 8 dg/ 10 min (190˚C, 2.16 kg weight, ASTM D1238), while the MAPP was Epolene E-43 with an average molecular weight of Mw = 9100 g/mol and acid number of 45 (mg KOH/g). The density of MAPE and MAPP was 0.908 and 0.936 g/cm 3, respectively. 3.2.2 Processing First, flakes of the light fraction recycled polymer were homogenized using a co-rotating twinscrew extruder, (Leistritz ZSE-27, L/D=40) with a screw speed of 110 rpm and a temperature profile for the 10 zones of: 170/185/200/215/230/230/230/215/210/205˚C. The material was then pelletized to be used subsequently for WPC production. Then, wood flour was dried overnight in an oven at 80°C before processing. The homogenized pellets of recycled plastic along with 5 wt.% of ethylene-octene copolymer (EOC) were fed into the main feeder (first zone) of the extruder using the same extruder with a screw speed of 115 rpm and a temperature profile of: 170/180/185/185/185/185 /185/190/190/195˚C. Wood flour (between 0 and 40 wt.%) and 5 wt.% (based on total wood content) MAPE/MAPP blend (80 wt.% MAPE + 20 wt.% MAPP) were introduced in the fourth zone of the extruder using a side-stuffer. The MAPE/MAPP: 80/20 formulation was selected based on previous study [71]. Finally, the compounds were pelletized and dried (overnight at 80°C) before being compression molded at 190°C in a laboratory Carver press. The compounds were first preheated for 3 minutes and pressed for 7 minutes in a mold having dimensions of 250×250×3 mm3 under a load of 3 tons.
45
3.2.3 Morphological observation Scanning electron micrographs (SEM) were used to investigate the morphology and interfacial adhesion of the components. A JEOL model JSM-840A was used to take SEM micrographs at different magnifications. Samples were fractured in liquid nitrogen, coated with a thin layer of gold/palladium alloy and then examined at 15 kV. 3.2.4 Mechanical testing Tension, flexion, torsion and impact tests were performed to evaluate the mechanical properties of the samples produced. All test specimens were cut from the 3 mm thick compression molded plates. For tensile characterizations, dog bone samples were cut according to ASTM D638 type V. The tests were performed at a strain rate of 10 mm/min on an Instron model 5565 with a 500 N load cell at room temperature (23°C). Five replicates were tested for each sample to get an average and standard deviation for Young’s modulus (E), tensile strength (σ), tensile elongation at break (εb) and ductility (D) (calculated as the area under the stress-strain curve). Flexural tests (60 mm span) were conducted on specimens with dimensions of 75×12.7×3 mm3 which were cut according to ASTM D790. An Instron model 5565 (load cell of 50 N) with a cross-head speed of 10 mm/min was used to measure the flexural modulus (Em) of the samples. The tests were performed at room temperature (23°C) with five replicates for each sample. Torsion modulus (Etm) was measured using an ARES Rheometer with dynamic frequency sweeps between 0.05 and 315 rad/s in the linear viscoelastic regime (0.05% strain). The modulus was determined at a frequency of 1.2 rad/s for comparison purposes. Three rectangular specimens (75×11×3 mm3) were tested for each sample and the average value was reported for modulus with its standard deviation. Charpy impact tests were conducted on a Tinius Olsen (model Impact 104) impact tester according to ASTM D6110-10. The samples (120×12.7×3 mm3) were notched with an automatic notcher Dynisco model ASN 120m. Ten specimens were tested for each sample and the average value of impact strength (Is) with standard deviation is reported.
46
3.2.5 Density and Hardness Measurements Density was determined using a gas pycnometer (ULTRAPYC 1200e) from Quantachrome Instruments. In addition, hardness (shore D) data were measured by a PTC Instruments Model 307L (ASTM D2240). These tests were performed on five replicates to report the average value and the standard deviation. 3.3 Results and discussion 3.3.1 Blend morphology Morphology was investigated using SEM. Figure 3.3 presents typical fractured surfaces of PE/PP blends with and without the additives. Based on these micrographs, a two-phase morphology was observed for the recycled plastic because of poor miscibility between polyethylene and polypropylene. Due to large interfacial tension [72], large domain size in SEM micrographs was observed before adding EOC (Figure 3.3-a). Adding the compatibilizer (5 wt.% of ethyleneoctene copolymer) resulted in reduced interfacial tension [67] and promoted adhesion between the blend components. Therefore, the compatibilized blend clearly revealed a finer phase dispersion and smaller domain size in SEM micrographs (Figure 3.3-b).
Figure 0.3: Typical SEM micrographs of the recycled light fraction plastics: (a) without compatibilizer and (b) with 5 wt.% of EOC. The arrows indicate typical domain sizes in SEM micrographs.
47
Figure 3.4 shows SEM for composites with 40% wood flour. The wood particles are well dispersed in the matrix as presented Figures 3.4-a, b. However, due to the low compatibility, wood particles were easily pulled out of the composites without coupling agent (Figure 3.4-a, c). Observation of clean surfaces of wood particles along with presence of gaps between the matrix and reinforcement (Figure 3.4-c) infer lack of compatibility. Adding 5 wt.% of MAPE /MAPP: 80/20 blend improved the interaction between wood flour and the polymer blend matrix. Fewer voids can be seen on the fractured surface of the compatibilized blends indicating enhanced compatibility between the phases (Figure 3.4-b, d). In addition, it is clear in micrographs with higher magnification that coupling agent addition resulted in elimination of gaps between matrix and wood flour (Figure 3.4-d). Figure 3.4-d also shows that wood particles are broken at the surface (no fiber pull-out) indicating good load transfer from the matrix to the wood particles (improved adhesion).
Figure 0.4: SEM micrograph of composites with 40 wt.% wood flour: (a,c) without coupling agents and (b,d) with additives (5 wt.% of EOC and 5 wt.% (MAPE/MAPP : 80/20)) at different magnifications.
48
3.3.2 Mechanical characterizations Mechanical properties of the composites with different wood concentrations are reported in Tables 3.1 and 3.2 for uncompatibilized and compatibilized samples, respectively. As presented in the Experimental part, tensile, flexural, torsion and Charpy impact tests were performed to characterize and compare the mechanical performance of these composites. Table 0.1: Mechanical properties of the composites without compatibilizer and coupling agent. (E: Tensile modulus, σ: Tensile strength, εb: Elongation at break, D: Ductility, Ef: Flexural modulus, Etm: Torsion Modulus, Is: Impact strength). Wood flour (%)
Tensile properties E (MPa)
σ (MPa)
εb (%)
D (MJ/m3)
0
265 (15) 296 (27) 305 (8) 344 (25) 412 (61)
13.1 (1.2) 12.4 (0.4) 11.6 (0.5) 11.1 (0.8) 10.0 (1.0)
10.0 (1.6) 7.6 (0.7) 5.9 (0.4) 5.3 (0.5) 4.6 (0.4)
1.10 (0.11) 0.67 (0.07) 0.45 (0.06) 0.42 (0.05) 0.36 (0.03)
10 20 30 40
Ef (MPa)
Etm (MPa)
Is (J/m)
1254 (120) 1350 (123) 1739 (96) 2065 (46) 2488 (163)
346 (12) 384 (24) 437 (27) 472 (26) 487 (17)
46.2 (4.2) 43.7 (4.1) 36.4 (2.6) 32.7 (1.7) 29.8 (2.2)
*Numbers in parenthesis denote standard deviations.
49
Table 0.2: Mechanical properties of the composites with compatibilizer and coupling agent. (E: Tensile modulus, σ: Tensile strength, εb: Elongation at break, D: Ductility, Ef: Flexural modulus,
Wood flour (%)
Etm: Torsion Modulus, Is: Impact strength).
0 10 20 30 40
Tensile properties E (MPa)
σ (MPa)
247 (17) 286 (28) 295 (24) 334 (59) 394 (33)
12.7 (0.4) 12.8 (0.4) 13.1 (0.8) 14.3 (0.4) 15.3 (0.7)
εb (%) 15.1 (0.9) 8.9 (0.7) 8.5 (1.0) 6.4 (0.4) 5.7 (0.4)
D (MJ/m3) 1.54 (0.17) 1.17 (0.21) 0.84 (0.11) 0.64 (0.05) 0.59 (0.06)
Ef (MPa)
Etm (MPa)
950 (99) 1296 (69) 1421 (103) 1736 (193) 1889 (115)
326 (11) 369 (23) 411 (11) 425 (20) 461 (19)
Is (J/m) 65.6 (2.8) 62.6 (4.0) 53.0 (4.5) 43.9 (2.1) 38.0 (2.8)
*Numbers in parenthesis denote standard deviations.
3.3.2.1 Recycled polymeric matrix As shown in Tables 3.1 and 3.2, a slight decrease in tensile modulus (from 265 to 247 MPa) is observed after incorporation of EOC. This was expected due to the elastomeric nature of this compatibilizer. But elongation at break and ductility of the blends showed significant increase (up to 50%) for the compatibilized blend in comparison with their uncompatibilized counterparts, which is in agreement with literature data for virgin blend of polyethylene and polypropylene [33] and confirms the morphological observations of Figures 3.3. Flexural and torsion moduli have a similar trend as tensile modulus. After adding 5 wt.% of EOC into the blends, the flexural and torsion modulus decreased by 32 and 6%, respectively. Inclusion of EOC is also shown to enhance impact properties of the recycled blend. Impact strength of the matrix increased from 46.2 to 65.6 J/m (42% increase) suggesting that EOC has positive effect as both impact modifier and compatibilizer. It is suggested in literature that the compatibilizing performance of EOC is achieved through alternating semi-crystalline ethylene/α-olefin and amorphous ethylene/α-olefin segments of the
50
copolymer structure [73]. While the crystalline blocks have good compatibility with PE, the amorphous ones seem to be more compatible with PP in the recycled plastic blend [33,73]. 3.3.2.2 Uncompatibilized composites Inclusion of sawdust residues as a reinforcing phase into the recycled matrix increased the rigidity of the plastic blend (Table 3.1). Enhancement in tensile, flexural and torsion modulus is attributed to the higher modulus and stiffness of wood fibers [53], in comparison with the corresponding values of the neat polymer blend. Nevertheless, introducing wood flour reduced tensile strength and elongation at break, as well as ductility of the samples without compatibilizer (Table 3.1). Reduction in tensile strength indicates low compatibility (poor stress transfer) between wood flour and matrix which was evidenced previously in SEM observations by the presence of voids (Figure 3.4). Decreasing ductility and elongation at break is due to the inherent rigidity and low elasticity of natural fibers in general [53]. An analysis of the stressstrain curves in tensile tests indicates that all samples broke at the yield point before exhibiting any cold drawing, even for the matrix without wood. This behavior was also reported in the literature for blends of virgin polyethylene and polypropylene [28]. Inclusion of wood flour also reduced the impact performance of the recycled plastic blends. Wood particles are creating stress concentration points in the composite which can act as crack initiators under the high deformation rates of impact tests. In this case, increasing wood content decreased impact strength by up to 36% after adding 40% of wood flour. 3.3.2.3 Compatibilized composites As mentioned before, EOC was chosen among several different compatibilizers for increasing the compatibility between PE and PP in the matrix. Furthermore, to promote compatibility between wood and the polymer matrix, an optimized coupling agent formulation (MAPE/MAPP: 80/20) was used to investigate the effect of increasing wood content on the mechanical performance of recycled polymers [71]. Effects of adding these compatibilizers are summarized in Table 3.2. Tensile modulus of the composites remains almost constant after incorporation of the compatibilizers. Interestingly, it is observed that tensile strength of samples with compatibilizer 51
had an opposite trend compared to samples without coupling agents: increasing wood content led to increased tensile strength of the composites with coupling agent. The difference in these trends justifies the effectiveness of the compatibilizing procedure used. Tensile elongation at break showed an increase in the compatibilized composites (Table 3.2) in comparison with the corresponding values for uncompatibilized ones (Table 3.1). It was shown that elongation at break of sample without wood flour increased 50%, while the sample with 40% wood showed an increase of only 24% after coupling agent addition. This behavior is expected to occur as a result of two opposite mechanisms: A) EOC increases the elongation at break due to better compatibility between the phases in matrix, and B) MAPP/MAPE decreases the elongation at break through better interaction between the matrix and wood particles (wood particles have much lower elongation at break). Flexural and torsion modulus showed a similar trend as tensile modulus. Although the impact strength of the composites improved after incorporation of compatibilizers, a reduction in impact strength of the composites is still observed with increasing wood content. Impact strength of sample with 40% wood increased from 29.8 to 38.0 J/m after adding compatibilizers. This behavior is due to elastomeric nature of EOC and improved homogeneity in the compounds. It is shown that impact strength of the composites with 10% (62.6 J/m) and 20% (53.0 J/m) wood flour in the compatibilized composites is still higher than that of the neat matrix (46.2 J/m). Mechanical properties of composites are proven to be higher compared to those of the matrix and are comparable with composites from virgin materials in structural applications. Considering the fact that both major phases of our composites are from recycled origin, remarkable environmental and economic benefits are expected. 3.3.3 Density and hardness results Table 3.3 shows that increasing wood content for both compatibilized and uncompatibilized composites increased density and hardness of the composites because of higher density and hardness values for wood flour [74]. Incorporation of the additives into the compatibilized composites slightly reduced their density and hardness. This behavior is considered as the result of lower density and hardness of the additives used. 52
Table 0.3: Hardness and density results for compatibilized and uncompatibilized composites.
Wood Content (% wt.)
Uncompatibilized composites
Compatibilized composites
Density (kg/m3)
Hardness (Shore D)
Density (kg/m3)
Hardness (Shore D)
0
948 (1)
65.4 (0.7)
936 (1)
65.0 (0.2)
10
972 (1)
67.0 (0.3)
964 (1)
66.4 (0.7)
20
993 (1)
69.3 (0.6)
989 (1)
66.9 (0.5)
30
1046 (1)
69.8 (0.3)
1032 (1)
68.1 (0.3)
40
1102 (1)
73.2 (0.5)
1071 (1)
69.6 (0.6)
3.4 Conclusions Wood plastic composites from municipal plastic waste (light fraction) and wood sawdust were produced to obtain a new value-added material. Morphological and mechanical properties obtained suggested the necessity of using an appropriate compatibilizer/coupling agent package to enhance the mechanical performance of the composites due to the presence of at least three components: PE, PP and wood. From the results obtained, it was shown that adding EOC and maleated polyolefins led to noticeable improvement in adhesion between the different phases in the compounds. Morphological observations supported the effectiveness of adding EOC to the recycled blend through reduction in domain sizes. Improvement in blend compatibility resulted in an increase in elongation at break up to 50% in comparison with the neat matrix. In addition, incorporation of 5 wt.% of MAPE/MAPP (80/20) enhanced the interfacial adhesion between the wood particles and the matrix of the composites.
53
Tensile, flexural and torsion modulus of the recycled plastic blends showed an increase after adding wood flour in both compatibilized and uncompatibilized composites. Because of inherent incompatibility of hydrophilic natural fibers with hydrophobic polymers, inclusion of wood flour reduced the tensile strength of uncompatibilized composites. But, incorporation of the compatibilizers suppressed this effect and tensile strength of the composite with 40% wood was enhanced up to 50% for the compatibilized composites. Compatibilization was also shown to have beneficial effect on impact properties. Impact strength of the matrix improved up to 42% after introducing the compatibilizers. Finally, hardness and density of the uncompatibilized composites were shown to be higher in comparison with their compatibilized counterparts. Incorporation of wood flour along with compatibilizers led to enhanced mechanical properties of the composites in comparison with the polymer matrix (from municipal plastic waste). The composites produced covered a wide range of mechanical properties with respect to the concentration of each phases, but final formulations must be adjusted based on the intended application. Acknowledgments The authors acknowledge the financial and technical support of the Natural Sciences and Research Council of Canada (NSERC) and FPInnovations, as well as Centre de Recherche sur le bois and Centre Québécois sur les Matériaux Fonctionnels (CQMF). The technical help of Mr. Yann Giroux was also much appreciated.
54
Chapter 4. Design Analysis of Three-Layered Structural Composites Based on Post-Consumer Recycled Plastics and Wood Residues
Résumé Des composites structuraux à trois couches ont été produits à partir de déchets plastiques municipaux et de résidus de farine de bois afin d’étudier les effets des paramètres de design sur les performances en flexion et à l’impact. Les paramètres étudiés incluent la teneur en bois, l’épaisseur des couches individuelles de composite, ainsi que la séquence et la configuration d’empilement des différentes couches (structures symétriques et asymétriques). Les résultats obtenus montrent que la couche constituant le cœur a une influence plus faible sur les propriétés en flexion que les couches de peau. Selon la configuration d’empilement choisie, différentes propriétés en flexion peuvent être obtenues pour un même ensemble de couches composites. La théorie classique des poutres a été utilisée afin de prédire avec précision le module en flexion de ces matériaux. Enfin, les performances de ces structures lors de tests d’impact ont été montrées comme étant indépendantes de la séquence d’empilement et de l’épaisseur des couches pour les différentes configurations.
Mots-clés: Structures multicouches; Propriétés mécaniques; Recyclage; Bois.
55
Abstract Three-layered structural composites were produced from municipal plastic wastes and wood flour residues to investigate the effects of design parameters on their flexural and impact performance. The studied parameters include wood content, thickness of individual composite layers, as well as stacking sequence and configuration (symmetric and asymmetric structures). The results indicate that the core layer has a lower influence on the flexural properties of structural beams in comparison with the skins. But depending on beam configuration (staking sequence), different flexural characteristics can be obtained using the same composite layers. The classical beam theory was used to predict the flexural modulus with high precision. In addition, performance of the beams under impact tests was shown to be independent from their stacking sequences and layer thicknesses for each configuration.
Keywords: Layered structures; Mechanical properties; Recycling; Wood.
56
4.1 Introduction Wood- plastic composites (WPC) have attracted an increasing amount of interest as an emerging area in polymer and wood industries. The main idea for inclusion of natural fibers of vegetal (lignocellulosic) origin into thermoplastic matrices is to reduce the cost of the final product and produce environmentally friendly materials. Moreover, natural fibers offer other specific properties which make them comparable to their conventional counterparts. The main advantages of WPC include light weight, high stiffness, biodegradability, less abrasiveness and a wide range of availability [68]. Owing to these characteristics, WPC have made a significant inroad in residential markets and construction industries including decking, fencing, flooring, landscaping, railings, window framing, and roof tiles [75]. Nevertheless, poor resistance to moisture and low compatibility of natural fibers with most polymer matrices hindered WPC’s from achieving their full potential in commercial markets. Substantial efforts have been made to compensate these drawbacks [76,52]. In spite of the large amount of research and development on WPC, the use of recycled materials in these products has not been fully exploited, principally because of low homogeneity and contamination effects. These effects lead to low compatibility among the different components in the composite and therefore lower mechanical properties of the final products. But several methods have been proposed in the literature to overcome such defects [19,58,60,61,69,70,77]. Recent works by the present authors [71,78] studied the compatibilizing effects of several modifiers on the properties of recycled composites. These composites were produced based on wood residues (sawdust) and PE/PP blends from a post-consumer origin to make a composite based on virtually 100% recycled materials. To further improve the mechanical properties of polymer composites, development of multilayered structures is considered as an interesting approach to functionalize these products for several applications. In this technique, distinct composite layers are bonded together to form an efficient load bearing assembly. These structures are designed to use the benefits of each separate layer and therefore optimize the structural properties of the whole assembly. Lamination designs offer many advantages over conventional monolithic structures, particularly when flexural loadings are predominant [63,79]. 57
Flexural properties of multilayered structures depend largely on the configuration of the layers, i.e. stacking sequence and layer thickness relative to the mid-plane [62]. This configuration dependence of bending properties in multilayered structures can be exploited in lamination design to optimize the structural properties of composites with a particular emphasis put on the economics of the final products. In this work, lamination design of three-layered structural composites was studied to investigate the effects of configuration parameters such as stacking sequence, layer thickness and wood content on the flexural and impact properties of the structures produced. As a specific case, postconsumer recycled plastics are used as the matrix and wood residues (sawdust) are used as reinforcement. Several symmetric and asymmetric configurations were studied to optimize the mechanical properties for structural applications. Finally, flexural modulus results are compared with predictions obtained from the classical beam theory. 4.2 Materials and methods 4.2.1 Materials Municipal plastic wastes (mainly polyethylene and polypropylene) were used as the matrix phase of the composite layers [71,78]. This post-consumer material (RECYC RPM, Quebec, Canada) was received in flake form. The wood flour used was a blend of sawdust from different species (sieved between 125 μm and 1 mm, with a 330 µm average sieve size) and was kindly supplied by the Department of Wood and Forest Sciences, Université Laval (Quebec, Canada). A formulation of ethylene-octene copolymer (Engage 8180, Dupont-Dow Chemical), maleic anhydride-grafted polyethylene (Epolene C-26, Westlake Chemical Corporation) and maleic anhydride-grafted polypropylene (Epolene E-43, Westlake Chemical Corporation) was used to enhance the compatibility between PE, PP and wood flour in the composite structures. Composites with different wood contents (0, 10, 20, 30 and 40% wt.) were used to fabricate the three-layered structures. A complete characterization and detailed processing conditions of these composites has been addressed elsewhere [78].
58
4.2.2 Processing Single layered composites, used for the production of three-layered structures, were first produced by compression molding, with different thicknesses (2, 3 and 4 mm) and wood contents (0-40 wt%). Three plates of these composites were then chosen and transferred to a new mold according to the desired configurations. Sample coding defines the stacking sequence which is the relative locations of each layer through the thickness. The sandwich structures were subsequently compression molded at 190°C in a laboratory Carver press. The plates were first preheated for 4 minutes and pressed for 8 minutes in a mold having dimensions of 250×250×9 mm3 under a load of 3 tons. The pressure applied in compression molding resulted in structure consolidation and interlaminar adhesion of the layers. For design considerations, two different types of three-layered structures were fabricated: symmetric and asymmetric laminates. In the symmetric configuration, the structure of the bottom layers was a mirror image of the structure above the mid-plane. In this case, 3 mm thick layers were used. In the asymmetric configuration, depending on the thickness of each layer, two types of three-layered laminates were produced: 1) laminates with 3 mm layers thickness but of different wood content, and 2) laminates of different layer thicknesses (2, 3 or 4 mm). However, the overall thickness of all laminates was constant at 9 mm (see Figure 4.1). For both symmetric and asymmetric structural composites, several configurations were produced to study the influence of stacking sequence, layer thickness and wood content on the mechanical properties of the beams. 4.2.3 Sample coding A sample coding system is used to identify the structure of the laminates. Stacking sequence is presented in the format of A(x)-B(y)-C(z), where A, B and C represent the wood content (% wt.) of each layer, while x, y and z represent the nominal layer thicknesses (mm) used in the structure. For the sake of conciseness, the layer thickness was not mentioned for structures having three layers of 3 mm in thickness. For instance, sample 10-0-10 represents a three-layered structure with skins having 10% wood and a core of neat polymer. For asymmetric configurations however, the load direction in bending should be considered. In this case, the first 59
layer in sample coding represents the layer which is in extrados of the beam in flexion. This means that sample 10-20-30 has a 30% wood content under the load nose in flexion test, and 10% wood in the extrados. However, by flipping over the beam (changing the load direction) the sample coding changes to 30-20-10. For asymmetric beams with different layer thicknesses, sample 20(2)-0(4)-40(3) represents a beam composed of a 2 mm thick 20% wood in extrados, a neat matrix 4 mm thick in the core, and a 3 mm thick layer of 40% wood in intrados (under the load). 4.2.4 Microscopy Scanning electron micrographs (SEM) were used to investigate the morphology and interlaminar adhesion of the layers. A JEOL model JSM-840A was used to take SEM micrographs at different magnifications. Fractured surface of Charpy impact test was used for SEM microscopy. The samples were previously coated with a thin layer of gold/palladium alloy and then examined at 15 kV. For more insight into the failure mechanism, the fractured specimens were also studied using a Stereo-Microscope (Olympus, SZ-PT). In addition, these optical micrographs were used to measure the exact thickness of each layer in a specimen. 4.2.5 Mechanical testing Three-point flexural tests were conducted according to ASTM D790. An Instron model 5565 with a 500 N load cell was used to measure the flexural properties. A support span to overall thickness ratio of 16:1 was chosen to eliminate shear effects in flexural modulus measurements. Sample width of 13 mm and crosshead speed of 3.8 mm/min were chosen, according to the standard. The tests were performed at room temperature (23°C) with five replicates for each sample. For asymmetric composites, flexural tests were performed on both sides to study the effect of load direction on flexural behavior. The values of flexural modulus (Ef), flexural strength (σmax) and strain at maximum stress (εmax) were obtained. Charpy impact tests were also performed with a Tinius Olsen (model Impact 104) impact tester, following the ASTM D6110 standard. The samples (120×12.7×9 mm3) were notched edgewise,
60
with an automatic notcher Dynisco model ASN 120m. Ten specimens were tested for each sample to report the average value and standard deviation. 4.2.6 Theory Once a multi-layered structure is subjected to a load in three points bending, the stress is shared differently among the layers. Thus flexural load varies through thickness from compression in the side under the load nose to tension in the opposite side of the beam. Consequently, there is a neutral surface (no stress) in the beam separating these two sides as illustrated in Figure 4.1. For symmetric structures, the neutral axis is on the mid-plane of the beam cross-section. However, for asymmetric structures, the position of the neutral axis should be calculated according to the beam geometry. Considering the assumptions of the classical beam theory, this model can be applied to predict the flexural modulus of symmetric and asymmetric structural composites. For small strains and short periods of time, the model neglects the shear strain in the beam. In addition, it considers that the plane cross-section of the beam remains planar and normal to the longitudinal direction of the beam during bending. The materials in the beam are assumed to follow Hooke’s law and have similar tension and compression modulus. Finally, for multilayered structures, perfect adhesion between layers also supposed [80].
Figure 0.1: Load distribution in a three-layered composite under three point bending test.
61
In this case, the flexural modulus of a three-layered structure (Et) is calculated as [81]: (4.1) where t (=t1+t2+t3) is the total thickness of the beam with a rectangular cross-section, E and I respectively represent flexural modulus and moment of inertia (about the neutral axis of the cross-sectional areas) of each layer, as presented in Figure 4.1. The values for I are given by: (4.2) (4.3) (4.4) where y0 represents the position of the neutral axis. For an asymmetric three-layered structure, y0 is calculated by: (4.5)
where t1, t2 and t3 are thicknesses of each layer, with respect to Figure 4.1. Exact thicknesses values were measured by optical microscopy. The measured values of thicknesses are different from the nominal values, owing to shrinkage or loss of materials in the compression molding step. In this work, the values of flexural modulus, obtained by Equation (4.1) are compared with experimental data in the following sections. For example, Smith and Partridge [62], studied the elementary bending theory for structures with two dissimilar materials (namely a high-strength titanium alloy and titanium metal matrix alloy) to determine the flexural stiffness envelopes of multi-layered systems. The evolution of symmetric and asymmetric binary multilayer systems was studied to address the effects of configuration parameters on their flexural stiffness. Using elementary bending theory, they showed that layer position changes the flexural modulus in different multi-layered structures. However, they concluded that it is possible to organize the layers into different structures to
62
achieve the same stiffness. This indicates that for a given volume fraction of layers, it is possible to arrange the layers to get an optimum flexural stiffness [62]. This is discussed next. 4.3 Results and discussion 4.3.1 Microscopy First, morphology and interfacial interaction between layers were investigated using scanning electron microscopy (SEM). Figure 4.2 presents micrographs of the fractured surfaces from notched impact specimens at different magnifications. SEM observations indicate a perfect interlaminar contact between the layers in the structures. They also show a uniform structure and no distinct boundary between the layers at the interfaces. In addition, no delamination was observed in the fractured surfaces after experiencing the impact tests which also indicates good adhesion. The principal defect of laminated composites is their low interlaminar strength which results in structure failure via separation along the interfaces of the layers (delamination) [82-84]. Using composite layers with the same matrix results in miscibility of the materials at the interface, which contributes to high interlaminar properties as evidenced by SEM observations. Processing conditions such as temperature, pressure and time of compression molding also play an important role in the quality of interlaminar bonding. Depending on the configuration and properties of each layer, different flexural failure modes are proposed in the literature for multilayered structures. The main failing modes to consider are shear failure in the core, interlaminar debonding and skin failures (compression buckling in the loaded face or tensile failure in the other face) [85].
63
Figure 0.2: SEM micrographs of: a) 40-0-40 and b) 0-20-10 samples. The arrows indicate the position of the interface between two composite layers.
64
According to optical micrographs (Figure 4.3), the fracture of all the samples (regardless of layer configuration) exhibited the same failure mechanism in three-point bending. Failure mode in such structures starts with tensile failure of the composite material in the extrados layers. The crack then propagates through the thickness and results in failure of the beam as shown in Figure 4.3. Such failure mode in three-layered structures infers the necessity of enhancing the tensile properties of the extrados layer to delay beam failure under flexural loading. In agreement with SEM observations, the optical micrographs of all the studied laminates indicate good bonding between each layer; i.e. no delamination (cracked interface) was observed between the layers in fractured specimens.
Figure 0.3: Optical micrographs of fractured samples for: a) 0(3)-40(4)-20(2) and b) 20(4)-0(3)40(2).
4.3.2 Mechanical characterizations Based on the configurations of the studied structures, mechanical characteristics are presented as follows. First, flexural properties are presented for three-layer structures to report their flexural modulus (Ee), strength (σmax), and strain at maximum stress (εmax). In the case of asymmetric structures, the beams are tested on two different faces to assess the flexural properties depending on load direction. Furthermore, for all three-layered structures, the experimental modulus (Ee) is compared with the theoretical values (Et) calculated by Equation (4.1). For this, values of 65
flexural modulus of single layers were taken from a previous study [78]. Finally, to compare the impact properties of the laminates, Charpy impact strengths (Is) are presented for all structural composites. 4.3.2.1 Symmetric structural composites Different symmetric structures were fabricated with the same cores or skin layers to assess the relative effect of each layer on the properties of the sandwich structure. The properties of symmetric structures are reported in Table 4.1. For Charpy impact tests, inclusion of composite layers with higher wood content reduced the impact strength. For instance, the impact strength of sample 10-0-10 was 50.7 J/m compared to 42.4 J/m for sample 40-30-40 which is a decrease of 20%. Because of good interlaminar bonding, these materials behave essentially as monolithic composites. Therefore, in edgewise impact tests the crack in one layer can propagate readily into adjacent layers and result in the failure of the whole assembly. Increasing the overall wood content of the laminates resulted in increased flexural modulus of the beams, as presented in Table 4.1. Nevertheless, increasing wood content in the skins was more effective in comparison with the core. For instance, increasing wood content in skins from 30% (30-0-30) to 40% (40-0-40), which represents only 6.7% increase in overall wood content, increased the flexural modulus of the structures from 1628 to 1900 MPa (17% improvement). However, increasing the wood content in the core, from 40-0-40 to 40-30-40 (10% increase in overall wood content) resulted in only a slight improvement in modulus (5% improvement). This behavior demonstrates that skin properties are more important to determine the flexural modulus of three-layered composites. In addition, model deviation (Dev.) in Table 4.1 indicates good agreement between theoretical and experimental modulus values for symmetric three-layered structures. This confirms that the assumptions of the classical beam theory are valid in this case. One of the most crucial assumptions is the perfect interlaminar adhesion and this was assessed by SEM observations as presented in Figure 4.2. Increasing wood content in the composite layers reduced the strain at maximum stress of threelayered structures. Strain at maximum stress (εmax) decreased from 4.8% for 10-0-10 to 2% for 40-30-40. However, similar to the flexural modulus, the effects of skins properties were more pronounced in comparison with the core layer. As an example, increasing wood flour content in 66
skins from 30-0-30 to 40-0-40 structure reduced εmax from 2.9% to 2.2%. However, inclusion of wood flour in the core had less effect (εmax is 2.7% for 30-20-30). Unexpectedly, the values of flexural strength did not exhibit a significant trend by increasing wood content in symmetric laminates, since there was a minimum for flexural strength of 20-020 structures. This behavior may be explained by the existence of two parallel mechanisms controlling the flexural strength of such materials. In this case, increasing the wood content in skin layers increases the flexural modulus of the structures. On the other hand, this also reduces the extension of each composite layer, especially for the skin layer placed in extrados for flexural tests. The low extension of extrados skin reduces its ability to sustain the tensile load and consequently leads to premature failure of the beam under flexural loads. Therefore, increasing the wood content in both skins may not improve the flexural strength. In this case, asymmetric designs may be a suitable alternative to compensate this defect. Table 0.1: Flexural and impact properties of symmetric three-layered structural composites. Sample code
t1 (mm)
t2 (mm)
t3 (mm)
10-0-10
2.8
2.7
2.5
20-0-20
2.9
2.6
2.5
30-0-30
3.0
2.8
2.9
40-0-40
2.9
2.7
3.1
20-10-20
2.8
2.7
2.6
30-20-30
3.0
2.9
2.9
40-30-40
2.9
2.9
2.9
Is (J/m)
σmax (MPa)
εmax (%)
Ee (MPa)
50.7 (2.6) 50.0 (2.0) 47.8 (1.8) 44.3 (3.2) 47.2 (3.7) 45.4 (2.7) 42.4 (1.6)
20.2 (0.2) 18.9 (0.4) 20.9 (0.2) 21.0 (0.8) 20.2 (0.2) 21.5 (0.6) 22.4 (0.8)
4.8 (0.2) 3.8 (0.1) 2.9 (0.1) 2.2 (0.1) 3.4 (0.2) 2.7 (0.2) 2.0 (0.1)
1217 (75) 1366 (56) 1628 (102) 1900 (148) 1459 (28) 1691 (45) 1958 (167)
Et (MPa)
Dev. (%)
1282
5
1403
3
1710
5
1861
-2
1416
-3
1725
2
1883
-4
4.3.2.2 Asymmetric structural composites with equal layer thickness In this section, the relative positions of composite layers are altered in the structure to examine the performance of different configurations in three-layered structural composites. Three 67
different configurations can be considered for each set of layers (for example layers with 10, 20 and 30% wood) according to Table 4.2. Regarding the asymmetric design, the laminates were examined from two different directions in flexion and the results are presented in the same rows of Table 4.2. Impact strength (Is) of composites in each set did not exhibit significant changes with to the position of each layer. However, a slight decrease in impact strength was observed from set 1 (49.7 J/m) to set 3 (45.5 J/m), which is due to the increase in overall wood content. The results in Table 4.2 show a direct relationship between wood content in skin layers and flexural modulus; higher wood contents in the skins produced higher flexural modulus values for three-layered structures. For instance, flexural modulus of sample 0-20-10 was 1073 MPa, while the value increased to 1355 MPa (26% increase) for the same set of layers with a 10-0-20 configuration. Similar to symmetric structures, the classical beam theory resulted in good predictions for the flexural modulus of asymmetric structures (Table 4.2). The model deviation in these structures was less than 10% which confirms again the validity of the assumptions made in the model. Interestingly, the results of Table 4.2 (both experimental and theoretical) suggested that the flexural moduli were independent of load direction. Flexural modulus of sample 10-0-20 was 1355 MPa compared to 1379 MPa for sample 20-0-10. This observation shows that both skins (intrados and extrados layers) have equal impact on flexural modulus of composite structures, probably because of small differences in composite’s properties between 10 and 20% wood. Contrary to flexural modulus, asymmetric beams exhibit different strain at maximum stress (εmax) depending on the direction of applied load. Hence, a beam with lower wood content in extrados has higher εmax, owing to the higher extension of this layer. The maximum value of εmax (5.2%) in set 1 belongs to beam 0-10-20 with the higher wood concentration (20%) in intrados. It is also observed that variation of wood content in the core layer does not affect significantly the extension of the structures. For instance, a beam with a 10-0-20 structure has the same εmax (3.9%) as a 10-30-20 beam (3.7%) with a different core layer. Similar result was observed for other configurations with the same skins/core layer configuration (Table 4.2). 68
Furthermore, the effect of skin layers was studied by comparing the results of εmax in Tables 4.1 and 4.2. For instance, a beam with a 20-10-20 structure has a εmax of 3.4%, while samples 0-1020 and 20-10-0 have εmax of 5.2 and 4.3%, respectively. Higher deformability of asymmetric samples is ascribed to lower overall concentration of wood in skins leading to higher elasticity (elongation at break). It is also shown that intrados also plays a slight role in controlling beam deformability regarding the higher values of εmax for 20-10-0 (4.3%) in comparison with the symmetric 20-10-20 structure (3.4%). In the case of flexural strength (σmax), different results were obtained depending on load direction. For all the studied beams, the flexural strength of the side with lower wood content in extrados was higher. For instance, flexural strength of sample 20-10-0 was 18.8 MPa compared to 22.3 MPa (19% increase) for 0-10-20. This shows that the layer in the extrados should have high deformability, while the intrados layer must have high modulus. This conclusion is confirmed by comparison of the results from Tables 4.1 and 4.2. For instance, flexural strength of sample 0-10-20 (22.3 MPa) is higher than sample 20-10-20 (20.2 MPa), while sample 20-10-0 shows the lowest flexural strength (18.8 MPa). Comparison between beams with similar skins and different cores exhibits similar flexural strength. For instance, the flexural strength of sample 10-0-20 in set 1 (19.9 MPa) is similar to the value of sample 10-30-20 in set 2 (19.4 MPa). This behavior shows that the skins are more predominant in flexural strength. According to Figure 4.1, in three points bending test, the core layer experiences the lower amount of load in comparison with the skin layers and therefore the flexural properties of this layer are less important. Salomi et al. [65] studied the flexural properties of two-layered laminates consisting of a linear low-density polyethylene (LLDPE) layer and a composite layer of isotactic polypropylene (PP)/E-glass fabric. Even in this case, their results indicated that the flexural strength (σmax) and strain (εmax) were higher when the composite layer was placed in the intrados. However, they indicated that the stiffness is not significantly affected by the relative position of both layers. This behavior may be attributed to the same weight of intrados and extrados layers in determination of flexural modulus in asymmetric beams (two layers only).
69
4.3.2.3 Asymmetric structural composites with different thickness of layers To study the effects of layer thickness on the properties of three-layered structures, asymmetric structural composites were produced. Permutation of three composite layers (0, 20 and 40% wood) with three thicknesses (2, 3 and 4 mm) in three-layered structures resulted in 36 different configurations, as shown in Table 4.3. Owing to the presence of layers with similar wood contents in all the studied configurations, the values of impact strength remained almost constant as presented in Table 4.3. The average value of impact strength for these materials was 49.8 J/m with a standard deviation of only 1.6 J/m. Again here, two different sides of the beam exhibit similar values of flexural modulus as presented in Table 4.3. This behavior was confirmed by the classical beam theory. The overall trend of flexural properties in Table 4.3 shows that wood content in each layer had a more important effect in determining the flexural modulus in comparison to layer thicknesses. For instance, the 0-40-20 configuration can have different layer thicknesses; i.e. 0(2)-40(3)-20(4) and 0(4)-40(3)-20(2), but showed similar flexural moduli: 1264 MPa and 1096 MPa, respectively. On the other hand, sample 20(2)-0(4)-40(3), which has the same layers and thicknesses as 0(4)40(3)-20(2) but different stacking configuration, showed a much higher flexural modulus of 1548 MPa. Even in the case of asymmetric structures with different layer thicknesses, σmax and εmax were different when the load direction was changed. According to the results of Table 4.3, flexural strength and strain at maximum stress were higher when the face with higher wood content was placed in the intrados of the beam. Figure 4.4 presents pictures of sample 0(4)-20(3)-40(2) which was tested on two different sides. In Figure 4.4-a, the layer with 40% wood and 2 mm thickness was placed in the extrados, while Figure 4.4-b presents the reverse case. This first configuration resulted in a specimen failure within 174 seconds (3.3% deformation). However, when the direction of the load was changed (layer with 0% wood and 4 mm thickness in extrados) the beam was more resilient and failed after 252 seconds (4.8% deformation).
70
Figure 0.4: Three-point bending of a beam with two different directions of flexural load: a) 40(2)-20(3)-0(4) for 3.3% deformation and b) 0(4)-20(3)-40(2) for 4.8% deformation.
Figure 4.5 is presented to understand changes in the flexural load-deflection behavior for selected laminates produced in this work. The maximum value of flexural strength was observed for beam 0(4)-20(3)-40(2) with 24.5 MPa (εmax = 4.8%). However, when tested on the other face, σmax was only 19.0 MPa and εmax was 3.3%. In comparison with 0(4)-20(3)-40(2), symmetric beams with higher (40-30-40) and lower (20-0-20) wood content were found to have lower values in terms of flexural strength and strain, as shown in Figure 4.5. Consequently, designing a multilayered structure gives rise to optimization of the mechanical properties of each layer. For instance, in Figure 4.5, a lower amount of wood is used in 0(4)-20(3)-40(2), but higher strength and strain is achieved compared to 40-30-40 beam. This is a noticeable feature as reinforcement distribution inside the whole structure can be optimized, which is especially important for composites using expensive, abrasive, hazardous or dense fibers. In addition, one surface of the beam can contain no fiber which may be interesting for outdoor applications. In this case, there will be less water absorption or biological degradation of the composite when a face without natural fibers is in contact with the environment.
71
Figure 0.5: Typical flexural stress-strain curves for different configurations of three-layered structural composites.
4.4 Conclusion In this work, flexural and impact behavior of three-layered structural composites based on recycled plastics (polyethylene and polypropylene blend) and wood flour (sawdust residues) was investigated. Symmetric and asymmetric structural composites (with equal and different layer thicknesses) were produced and analyzed. For the processing conditions selected, it was shown that good interlaminar adhesion between the layers occurred which was evidenced by morphological observations through SEM and optical microscopy. For symmetric beams, increase in overall wood content led to lower impact strength, while their flexural modulus increased substantially. Skin layers proved to be more effective in controlling flexural modulus than core ones. On the other hand, increasing wood content, led to lower deformability of the composites, while flexural strength remained almost constant for the range 72
of conditions studied. It was also observed that in all cases, failure started in the extrados layer and then propagated through the sample thickness. For asymmetric structures with equal layer thicknesses, flexural modulus remained constant after changing the load direction. Flexural deformability and strength of these structures were higher when the load was applied to the skin face with higher wood content. For asymmetric structures with different layer thicknesses, the effect of wood content in each layer was shown to be more significant than their thicknesses. The maximum value for flexural strength (24.5 MPa) of all the structures studied was obtained for sample 0(4)-20(3)-40(2). Finally, the classical beam theory was shown to provide good predictions for the flexural modulus of the structural composites produced in this work. Model deviation was always smaller than 10% showing that the assumptions made in the model (especially perfect interlaminar adhesion) are satisfied in our samples. It should also be mentioned that the model covers a wide range of structures (over 60 different configurations here) with different layer thicknesses and wood contents. Thus, the experimental results show that the classical beam theory is a reliable model to predict the flexural modulus of three-layered structural composites produced from recycled materials (polymers and wood). Based on the results obtained, a wide range of mechanical properties can be covered via structural design of composites. Nevertheless, selection of the best structure (design) relies on the final application. For example, sample 40-30-40 has the highest flexural modulus (1958 MPa), while samples 0(4)-20(3)-40(2) and 0-10-20 have the highest flexural strength (24.5 MPa) and maximum strain (5.2%). For impact strength, sample 0(4)-40(2)-20(3) performed best with 52.2 J/m.
Acknowledgements The authors acknowledge the financial support of the Natural Sciences and Research Council of Canada (NSERC), and the technical support of FPInnovations, as well as Centre de Recherche sur le Bois (CRB) and Centre Québécois sur les Matériaux Fonctionnels (CQMF). The technical help of Mr. Yann Giroux was also much appreciated for the experimental work. 73
2.7
1122
-4
10-20-0
3.0
2.7
2.4
1339
1
20-0-10
2.7
3.0
2.5
18.8 (1.1) 17.8 (0.4) 18.4 (0.5)
4.3 (0.1) 3.7 (0.2) 3.6 (0.3)
1216 (83) 1154 (37) 1379 (105)
18.4 (0.4) 18.5 (0.3) 19.9 (0.4)
2.7 (0.1) 3.8 (0.1) 3.2 (0.1)
1499 (62) 1391 (31) 1512 (55)
19.6 (1.0) 19.1 (0.6) 20.2 (0.3)
2.4 (0.2) 2.7 (0.3) 2.0 (0.1)
1708 (27) 1577 (51) 1928 (95)
Dev. (%)
2.7
Et (MPa)
2.6
Ee (MPa)
20-10-0
εmax (%)
t3 (mm)
8
σmax (MPa)
t2 (mm)
1165
Sample code
t1 (mm)
Dev. (%)
Et (MPa)
Ee (MPa)
εmax (%)
σmax (MPa)
Is (J/m)
t3 (mm)
t2 (mm)
t1 (mm)
Sample code
Table 0.2: Flexural and impact properties of asymmetric three-layered structural composites with equal layer thickness
1162
4
1129
2
1336
3
1496
1
1375
1
1557
-3
1640
4
1586
-1
1794
7
(Set 1: 0%, 10%, 20%) 0-10-20
2.5
2.6
2.6
0-20-10
2.7
2.9
2.8
10-0-20
2.6
2.9
2.8
50.6 (4.9) 48.5 (1.3) 50.0 (2.4)
22.3 (1.1) 18.1 (1.3) 19.9 (0.3)
5.2 (0.3) 4.1 (0.3) 3.9 (0.1)
1261 (101) 1073 (95) 1355 (78)
(Set 2: 10%, 20%, 30%) 10-20-30
2.6
2.8
3.0
10-30-20
2.5
2.9
2.9
20-10-30
2.5
2.8
3.1
49.9 (1.7) 47.3 (3.0) 48.7 (2.4)
20.7 (1.0) 19.4 (0.5) 20.8 (0.1)
3.7 (0.2) 3.7 (0.3) 3.1 (0.1)
1483 (61) 1325 (34) 1494 (31)
1496
-1
30-20-10
2.9
3.0
2.5
1375
-3
20-30-10
3.0
3.0
2.6
1558
-4
30-10-20
3.0
3.0
2.5
(Set 3: 20%, 30%, 40%)
74
20-30-40
2.8
3.0
2.8
20-40-30
2.9
3.2
2.7
30-20-40
2.8
3.0
3.0
45.9 (2.9) 46.3 (1.9) 44.2 (2.7)
21.4 (0.6) 20.9 (0.5) 20.6 (0.3)
2.7 (0.3) 2.9 (0.1) 2.5 (0.1)
1648 (139) 1545 (74) 1904 (126)
1640
1
40-30-20
3.0
2.8
2.8
1584
-2
30-40-20
3.0
3.1
2.8
1794
6
40-20-30
3.0
3.0
2.7
Table 0.3: Flexural and impact properties of asymmetric three-layered structural composites with different thickness of layers Sample code
t1 (mm)
t2 (mm)
t3 (mm)
0(2)-40(3)-20(4)
2.1
3.1
3.6
0(4)-40(3)-20(2)
3.8
3.0
2.1
0(3)-40(2)-20(4)
3.1
2.1
4.2
0(4)-40(2)-20(3)
3.8
2.1
2.9
0(2)-40(4)-20(3)
1.8
3.7
2.9
0(3)-40(4)-20(2)
3.0
4.1
2.2
20(2)-0(3)-40(4)
2.0
3.1
4.2
20(4)-0(3)-40(2)
4.0
3.2
2.0
20(3)-0(2)-40(4)
2.6
1.9
3.8
20(4)-0(2)-40(3)
4.0
2.2
3.3
20(2)-0(4)-40(3)
2.2
4.0
3.1
20(3)-0(4)-40(2)
2.7
3.9
1.8
0(2)-20(3)-40(4)
2.1
3.3
3.8
0(4)-20(3)-40(2)
4.0
3.3
1.9
0(3)-20(2)-40(4)
2.9
2.2
4.0
0(4)-20(2)-40(3)
4.0
2.0
3.1
0(2)-20(4)-40(3)
2.2
4.0
2.9
0(3)-20(4)-40(2)
2.9
4.1
2.0
Is σmax (J/m) (MPa) 48.2 21.0 (2.4) (0.8) 50.6 18.9 (3.8) (1.4) 49.3 21.1 (4.3) (0.9) 52.2 20.2 (3.7) (0.6) 50.2 21.6 (2.8) (0.4) 49.6 21.1 (3.1) (0.5) 47.5 20.7 (2.9) (0.5) 51.0 21.4 (3.8) (0.4) 51.0 21.8 (3.7) (0.4) 47.9 21.2 (2.8) (0.5) 48.4 21.6 (2.8) (0.9) 50.0 20.7 (3.1) (0.6) 51.6 21.6 (2.8) (0.3) 51.7 24.5 (4.2) (0.8) 50.5 22.9 (2.8) (0.4) 46.4 21.8 (2.3) (1.7) 49.2 22.3 (2.9) (0.1) 51.0 22.2 (3.2) (0.4)
εmax (%) 4.9 (0.3) 4.2 (0.1) 4.3 (0.1) 4.1 (0.3) 4.4 (0.2) 4.4 (0.2) 3.1 (0.3) 3.3 (0.1) 3.2 (0.2) 3.3 (0.1) 3.5 (0.7) 3.4 (0.4) 4.2 (0.2) 4.8 (0.5) 4.7 (0.2) 3.8 (0.7) 4.2 (0.1) 4.5 (0.1)
Ee Et Dev. (MPa) (MPa) (%) 1264 1247 1 (74) 1096 1175 -7 (130) 1235 1186 4 (83) 1133 1159 -2 (92) 1313 1275 3 (89) 1166 1211 -4 (11) 1442 1563 -8 (103) 1574 1565 1 (74) 1569 1608 -2 (116) 1568 1621 -3 (107) 1548 1570 -1 (104) 1415 1550 -10 (89) 1311 1379 -5 (49) 1405 1288 8 (25) 1358 1334 2 (5) 1360 1305 4 (53) 1292 1368 -6 (92) 1256 1316 -5 (48)
Sample code
t1 (mm)
t2 (mm)
20(4)-40(3)-0(2)
4.0
3.3
20(2)-40(3)-0(4)
2.1
3.1
20(4)-40(2)-0(3)
4.0
2.0
20(3)-40(2)-0(4)
3.2
1.9
20(3)-40(4)-0(2)
2.9
4.0
20(2)-40(4)-0(3)
2.1
4.2
40(4)-0(3)-20(2)
4.0
3.2
40(2)-0(3)-20(4)
2.0
3.2
40(4)-0(2)-20(3)
3.8
1.9
40(3)-0(2)-20(4)
3.2
2.3
40(3)-0(4)-20(2)
3.1
4.0
40(2)-0(4)-20(3)
1.8
3.8
40(4)-20(3)-0(2)
4.0
3.3
40(2)-20(3)-0(4)
2.0
3.3
40(4)-20(2)-0(3)
4.1
1.9
40(3)-20(2)-0(4)
3.1
2.0
40(3)-20(4)-0(2)
3.2
3.9
40(2)-20(4)-0(3)
2.0
4.0
t3 σmax (mm) (MPa) 17.0 2.0 (0.8) 16.6 4.0 (0.2) 17.9 3.1 (0.3) 18.6 3.8 (0.4) 19.7 1.8 (0.6) 19.0 2.9 (0.2) 19.8 2.0 (0.5) 19.5 4.0 (0.3) 19.8 2.5 (0.2) 19.1 3.8 (0.4) 20.0 2.1 (0.7) 19.4 2.8 (0.8) 19.2 2.1 (0.3) 19.0 4.0 (0.7) 17.5 2.8 (0.7) 18.9 3.8 (0.3) 17.3 1.8 (0.2) 16.7 2.8 (0.4)
εmax (%) 3.2 (0.1) 3.7 (0.4) 3.3 (0.4) 4.5 (0.3) 3.7 (0.2) 4.0 (0.1) 3.0 (0.1) 3.0 (0.1) 2.7 (0.1) 2.6 (0.1) 2.8 (0.2) 3.2 (0.3) 3.0 (0.1) 3.3 (0.2) 2.8 (0.2) 3.2 (0.1) 2.6 (0.1) 2.8 (0.1)
Ee Et (MPa) (MPa) 1196 1269 (67) 1203 1176 (50) 1239 1181 (54) 1191 1158 (85) 1296 1284 (130) 1263 1217 (45) 1546 1568 (44) 1597 1565 (15) 1616 1606 (117) 1615 1619 (103) 1518 1566 (50) 1555 1552 (81) 1505 1382 (87) 1404 1291 (9) 1354 1334 (48) 1423 1307 (28) 1331 1396 (93) 1209 1319 (39)
Dev. (%) -6 2 5 3 1 4 -1 2 1 -1 -3 1 8 8 1 8 -4 -9
75
76
Chapter 5. Conclusions and Recommendations 5.1 General conclusion Plastic recycling has emerged as an issue of great importance in recent decades. Despite the substantial efforts devoted to this matter, it should be considered that the efficient largescale implementation of plastic recycling represents a goal which is still far from being fully achieved by the recycling industry. The major obstacle arises from the low compatibility of different components in plastic recyclates. The compatibility of virgin PE and PP (as the major contributors of plastic waste stream light fraction) is substantially studied in the literature. However, the recycled blends of PE and PP still suffer from the lack of experimental analysis. The aim of this research work was to study the blends of recycled PE and PP in light fraction of municipal plastic waste streams. Mechanical and morphological analysis revealed a severe deterioration in recycled PE/PP blend properties. Therefore, three different strategies were followed to improve the mechanical performance of this blend including enhancing the compatibility of the blend components, inclusion of wood fiber and structural design of the composites (sandwich structure). Inclusion of 5 wt.% of EOC as a compatibilizer reduced the interfacial tension which resulted in smaller PE and PP domains in SEM micrographs. Such an improvement in morphological characteristics resulted in enhanced mechanical performance. Consequently, in compatibilized blend the values of elongation at break, impact strength and ductility were improved by up to 50%, 42% and 40%, respectively. Incorporation of wood flour improved the stiffness of recycled PE/PP blend. Increasing the wood content in composites enhanced tensile, flexural and torsion moduli, but also decreased ductility, elongation at break, as well as tensile and impact strengths. In comparison with the net matrix, inclusion of wood fiber in a composite with 40% wood content reduced the tensile and impact strength by up to 24% and 35%, respectively. 77
SEM observations revealed the low compatibility between wood flour and polymeric matrix. The wood particles were easily pulled out in the fractured surface of composites without coupling agent. Adding 5 wt.% of MAPE/MAPP: 80/20 blend improved the interfacial adhesion between wood flour and the polymer blend matrix. In addition, incorporation of MAPE/MAPP blend suppressed the low compatibility of wood fiber and polymeric matrix which resulted in an increase in tensile strength with increasing wood content. Compatibilization was also shown to have beneficial effect on impact properties of the composites. Therefore the compatibilized composite with 40% wood exhibited enhanced elongation at break (24%), ductility (64%), tensile strength (53%) and impact strength (27%), compared to their uncompatibilized counterparts. In addition, hardness and density of the uncompatibilized composites were shown to be higher in comparison with their compatibilized counterparts. In the second part of this research, structural design of composite materials was investigated to enhance their structural performance. As a special case, three-layered structural composites were fabricated in both symmetric and asymmetric configurations to study the effects of different design parameters including stacking sequence, layer thickness and wood content. SEM micrographs reveal a good interlaminar adhesion between the layers owing to the proper processing conditions and similarity of matrices in different layers. For symmetric beams, increase in overall wood content led to lower impact strength (up to 20% decrease from sample 10-0-10 to sample 40-30-40), while their flexural modulus increased substantially (up to 61% increase for sample 40-30-40, in comparison with sample 10-0-10). For these structures, skin layers were shown to be more effective in controlling flexural modulus than the core. On the other hand, increasing wood content, led to lower deformability of the composites (up to 58% decrease for sample 40-30-40, in comparison with sample 10-0-10), while flexural strength remained almost constant for the range of conditions studied. It was also 78
observed that in all cases, failure started in the extrados layer and then propagated through the sample thickness. For asymmetric structures with equal layer thicknesses, flexural modulus remained constant after changing the load direction. Flexural deformability and strength of these structures were higher when the load was applied to the skin face with higher wood content. For asymmetric structures with different layer thicknesses, the effect of wood content in each layer was shown to be more significant than their thicknesses. The maximum value for flexural strength (24.5 MPa) of all the structures studied was obtained for sample 0(4)20(3)-40(2). Finally, the classical beam theory was shown to provide good predictions for the flexural modulus of the structural composites produced in this work. Model deviation was always smaller than 10% showing that the assumptions made in the model (especially perfect interlaminar adhesion) are fully satisfied in our samples. It should also be mentioned that the model covers a wide range of structures (over 60 different configurations here) with different layer thicknesses and wood contents. Thus, the experimental results show that the classical beam theory is a reliable model to predict the flexural modulus of three-layered structural composites produced from recycled materials (polymers and wood). Based on the results obtained, these multi-layered structures represent a wide range of applications due to the asymmetric load distribution in these composites. The main applications of these products are in construction markets including decking, fencing, flooring, landscaping, railings, window framing, and roof tiles. 5.2 Recommendations for future works As mentioned earlier, experimental analysis on true post-consumer plastic material is still limited to a few number of research works in the literature. Although positive effects of compatibilizers on compatibility of virgin PE/PP blends have been extensively reported, improvement of blend characteristics of recycled origin remained a challenge. Lower 79
performance of compatibilizers in recycled blends demonstrates the role of interfering elements which suppress the compatibilization. Therefore it is essential to investigate these defects in recycled blends. Consequently, it is recommended to study the compatibilizing effects of different types of compatibilizers in both virgin and recycled blends to quantify the interfering effects of impurities present inside recycled materials. In addition, investigating the viscoelastic behavior of polymer blends is essential in collecting valuable information on flow mechanism of the blend components which has an effect on phase morphology and ultimate mechanical properties of these materials. Therefore, a rheological analysis of PE and PP blends with both virgin and recycled materials is recommended to determine the effect of contamination and degradation on the flow properties of these materials. In the case of multilayered structures, additional design parameters including different number of layers or inclusion of different fibers can give more insight regarding the optimum performance of such structures. As a special case, fiber orientation and size should be included. Finally, to decrease the weight and obtain higher specific properties, foamed core structures are recommended for these materials. Design analysis of these sandwich structures should be an interesting subject for future works in relation with strength over weight ratio optimization.
80
References [1] Plastics - The Facts 2012 - An analysis of European plastics production, demand and waste data for 2011, a study carried out by EuPR, EuPC, PlasticsEurope and EPRO: www.plasticsrecyclers.eu. [2] USEPA, 2010. Municipal solid waste generation, recycling, and disposal in the United States: facts and figures for 2010: www.epa.gov. [3] Al-Salem, S.M., Lettieri, P. and Baeyens, J., Recycling and recovery routes of plastic solid waste (PSW): A review, Waste Manag., 29, 2625-2643 (2009). [4] Vilaplana, F. and Karlsson, S., Quality concepts for the improved use of recycled polymeric materials: A review, Macromol. Mater. Eng., 293, 274-297 (2008). [5] Karlsson, S., Recycled polyolefins. Material properties and means for quality determination, Adv. Polym. Sci., 169, 201-230 (2004). [6] Finnveden, G., Johansson, J., Lind, P. and Moberg, A., Life cycle assessment of energy from solid waste-part 1: General methodology and results, J. Cleaner Prod., 13, 213229 (2005). [7] Williams, E.A. and Williams, P.T., The pyrolysis of individual plastics and plastic mixture in a fixed bed reactor. J. Chem. Tech. Biotechnol., 70, 9-20 (1997). [8] Mastellone, M.L., Thermal treatments of plastic wastes by means of fluidized bed reactors. Ph.D. Thesis, Department of Chemical Engineering, Second University of Naples, Italy (1999). [9] Aznar, M.P., Caballero, M.A., Sancho J.A. and Francs, E., Plastic waste elimination by co-gasification with coal and biomass in fluidized bed with air in pilot plant. Fuel Process. Technol., 87, 409-420 (2006).
81
[10] Maio, F.D., Rem, P., Hu, B., Serranti, S. and Bonifazi, G., The W2Plastics project: Exploring the limits of polymer separation, Open Waste Manag. J., 3, 90-98 (2010). [11] Bakker, E.J., Rema, P.C. and Fraunholcz, N., Upgrading mixed polyolefin waste with magnetic density separation, Waste Manage., 29, 1712-1717 (2009). [12] Mikulionok, I.O., Pretreatment of recycled polymer raw material, Russ. J. Appl. Chem., 84, 1105-1113 (2011). [13] Utracki, L.A., Commercial Polymer Blends. Cornwall: Chapman and Hall (1998). [14] Bertin, S. and Robin, J.J., Study and characterization of virgin and recycled LDPE/PP blends, Eur. Polym. J., 38, 2255-2264 (2002). [15] Pospisil, J., Nespurek, S., Pfaendner, R. and Zweifel, H., Material recycling of plastics waste for demanding applications: Upgrading by restabilization and compatibilization, Trends Polym. Sci., 5, 294-300 (1997). [16] Teh, J.W., Rudin, A. and Keung, J.C., A review of polyethylene-polypropylene blends and their compatibilization, Adv. Polym. Tech., 13, 1-23 (1994). [17] Zhao, L. and Choi, P., A review of the miscibility of polyethylene blends, Mater. Manuf. Proc., 21, 135-142 (2006). [18] Lednicky, F., Coufalova, E., Hromadkova, J., Delong, A. and Kolarik, V., Lowvoltage TEM imaging of polymer blends, Polymer, 41, 4909-4914 (2000). [19] Clemons, C., Elastomer modified polypropylene-polyethylene blends as matrices for wood flour-plastic composites, Composites Part A., 41, 1559-1569 (2010). [20] Koning, C., Duin, M.V., Pagnoulle, C. and Jerome, R., Strategies for compatibilization of polymer blends, Prog. Polym. Sci., 23, 707-757 (1998). [21] Scobbo, J.J., Compatibilization strategies for upgrading recycled plastics performance: Use of reactive rubbery polymers, Macromol. Symp., 57, 345-352 (1992). 82
[22] Hettema, R., Tol, J.V. and Janssen, L.P.B., In situ reactive blending of polyethylene and polypropylene in co-rotating and counter-rotating extruders. Polym. Eng. Sci., 39, 1628-1641 (1999). [23] Cheung, P., Suwanda, D. and Balke, S.T. The reactive extrusion of polyethylene/polypropylene blends, Polym. Eng. Sci., 30, 1063-1072 (1990). [24] Souza, A.M.C. and Demarquette, N.R., Influence of coalescence and interfacial tension on the morphology of PP/HDPE compatibilized blends, Polymer, 43, 39593967 (2002). [25] Dorazio, L., Greco, R., Mancarella, C., Martuscelli, E., Ragosta, G. and Silvestre, C., Effect of the addition of ethylene-propylene random copolymers on the properties of high-density polyethylene/isotactic polypropylene blends: Part 1- Morphology and impact behavior of molded samples, Polym. Eng. Sci., 22, 536-544 (1982). [26] Chiu, W.Y. and Fang, S.J., Mechanical properties and morphology of crosslinked PP/PE blends and PP/PE/propylene-ethylene copolymer blends, J. Appl. Polym. Sci., 30, 1473-1489 (1985). [27] Zhu, W., Zhang, X., Feng, Z. and Huang, B.J., Effect of ethylene-propylene copolymer with residual PE crystallinity on mechanical properties and morphology of PP/HDPE blends, J. Appl. Polym. Sci., 58, 551-557 (1995). [28] Radonjic, G. and Gubeljak, N., The use of ethylene/propylene copolymers as compatibilizers for recycled polyolefin blends, Macromol. Mater. Eng., 287, 122-132 (2002). [29] Vranjes, N. and Rek, V., Effect of EPDM on morphology, mechanical properties, crystallization behavior and viscoelastic properties of iPP+HDPE blends, Macromol. Symp., 258, 90-100 (2007).
83
[30] Slusarski, L., Bielinski, D., Wlochowicz, A. and Slusarczyk, C., Structure and properties of polyolefin blends, Polym. Int., 36, 261-268 (1995). [31] Blom, H.P., Teh, J.W. and Rudin, A., iPP/HDPE blends. II. Modification with EPDM and EVA, J. Appl. Polym. Sci., 60, 1405-1417 (1996). [32] Dias, P., Lin, Y.J., Poon, B., Chen, H.Y., Hiltner, A. and Baer, E., Adhesion of statistical and blocky ethylene–octene copolymers to polypropylene Polymer, 49, 2937-2946 (2008). [33] Lin, Y., Yakovleva, V., Chen, H., Hiltner, A. and Baer, E., Comparison of olefin copolymers as compatibilizers for polypropylene and high-density polyethylene, J. Appl. Polym. Sci., 113, 1945-1952 (2009). [34] Oksman, K. and Clemons, C., Mechanical properties and morphology of impact modified polypropylene-wood flour composites, J. Appl. Polym. Sci., 67, 1503-1513 (1998). [35] Denac, M., Musil, V. and Smit, I., Polypropylene/talc/SEBS (SEBS-g-MA) composites. Part 2. Mechanical properties, Composites Part A, 36, 1282-1290 (2005). [36] Jarukumjorn, K. and Suppakarn, N., Effect of glass fiber hybridization on properties of sisal fiber-polypropylene composites, Composites Part B, 40, 623-627 (2009). [37] Joshi, S.V., Drzal, L.T., Mohanty, A.K. and Arora, S., Are natural fiber composites environmentally superior to glass fiber reinforced composites?, Composites Part A, 35, 371-376 (2004). [38] Cheung, H.Y., Ho, M.P., Lau, K.T., Cardona, F. and Hui, D., Natural fiber-reinforced composites
for
bioengineering
and
Composites Part B, 40, 655-663 (2009).
84
environmental
engineering
applications,
[39] Ramezani K., A., Leduc, S. and Rodrigue, D., Effect of hybridization and compatibilization on the mechanical properties of recycled polypropylene-hemp composites, J. Appl. Polym. Sci., 124, 2494-2500 (2012). [40] Bledzki, A.K. and Gassan, J., Composites reinforced with cellulose based fibers, Prog. Polym. Sci., 24, 221-274 (1999). [41] Li, X., Tabil, L.G. and Panigrahi, S., Chemical treatments of natural fiber for use in natural fiber-reinforced composites: A review, J. Polym. Environ., 15, 25-33 (2007). [42] Nevell, T.P. and Zeronian, S.H., Cellulose Chemistry and Applications, New York: Wiley (1985). [43] Facca, A.G., Kortschot, M.T. and Yan, N., Predicting the elastic modulus of natural fiber reinforced thermoplastics, Composites Part A, 37, 1660-1671 (2006). [44] Abdelmouleh, M., Boufi, S., Belgacem, M.N. and Dufresne, A., Short natural-fiber reinforced polyethylene and natural rubber composites: Effect of silane coupling agents and fibers loading, Compos. Sci. Technol., 67, 1627-1639 (2007). [45] Shinoj, S., Panigrahi, S. and Visvanathan, R., Water absorption pattern and dimensional stability of oil palm fibre-linear low density polyethylene composites, J. Appl. Polym. Sci., 117, 1064-1075 (2010). [46] Dhakal, H.N., Zhang, Z.Y. and Richardson, M.O.W., Effect of water absorption on the mechanical properties of hemp fibre reinforced unsaturated polyester composites, Compos. Sci. Technol., 67, 1674-1683 (2007). [47] Rachini, A., Mougin, G., Delalande, S., Charmeau, J.Y., Barres and C., Fleury, E., Hemp fibers/poly propylene composites by reactive compounding: improvement of physical properties promoted by selective coupling chemistry, Polym. Degrad. Stabil., 97, 1988-1995 (2012).
85
[48] Sgriccia, N., Hawley, M.C. and Misra, M., Characterization of natural fibre surfaces and natural fibre composites, Composites Part A, 39, 1632-1637 (2008). [49] Rong, M.Z., Zhang, M.Q., Liu, Y., Yang, G.C. and Zeng, H.M., The effect of fiber treatment on the mechanical properties of unidirectional sisal-reinforced epoxy composites, Comp. Sci. Technol., 61, 1437-1447 (2001). [50] Felix, J.M. and Gatenholm, P., The nature of adhesion in composites of modified cellulose fibers and polypropylene, J. Appl. Polym. Sci., 42, 609-620 (1991). [51] Mohanty, S. and Nayak, S.K., Effect of MAPP as a coupling agent on the performance of jute-PP composites, J. Reinf. Plast. Compos., 23, 625-637 (2004). [52] Keener, T.J., Stuart, R.K. and Brown, T.K., Maleated coupling agents for natural fiber composites, Composites Part A, 35, 357-362 (2004). [53] Faruk, O., Bledzki, A.K., Fink, H.P. and Sain, M., Biocomposites reinforced with natural fibers: 2000-2010, Prog. Polym. Sci., 37, 1552-1596 (2012). [54] Huda, M.S., Drzal, L.T., Mohanty A.K. and Misra, M., Effect of fiber surfacetreatments on the properties of laminated biocomposites from poly(lactic acid) (PLA) and kenaf fibers, Comp. Sci. Technol., 68, 424-432 (2008). [55] Alvarez, V.A. and Vazquez, A., Thermal degradation of cellulose derivatives/starch blends and sisal fiber biocomposites, Polym. Degrad. Stabil., 84, 13-21 (2004). [56] Alvarez, V.A., Terenzi, A., Kenny, J.M. and Vazquez, A., Melt rheological behavior of starch-based matrix composites reinforced with short sisal fibers, Polym. Eng. Sci., 44, 1907-1914 (2004). [57] Alvarez, V.A., Kenny, J.M. and Vazquez, A., Creep behavior of biocomposites based on sisal fiber reinforced cellulose derivatives/starch blends, Polym. Compos., 25, 280288 (2004). 86
[58] Najafi, S.K., Hamidnia, E. and Tajvidi, M., Mechanical properties of composites from sawdust and recycled plastics, J. Appl. Polym. Sci., 100, 3641-3645 (2006). [59] Gao, H., Song, Y.M., Wang, Q.W., Han, Z., Zhang, M.L., Rheological and mechanical properties of wood fiber-PP/PE blend composites, J. Forest. Res., 19, 315-318 (2008). [60] Dintcheva, N.T. and La Mantia, F.P., Recycling of the “light fraction” from municipal post-consumer plastics: Effect of adding wood fibers, Polym. Adv. Technol., 10, 607614 (1999). [61] Dintcheva, N.T., La Mantia, F.P., Trotta, F., Luda, M.P., Camino, G., Paci, M., Di Maio, L. and Acierno, D., Effects of filler type and processing apparatus on the properties of the recycled “light fraction” from municipal post-consumer plastics, Polym. Adv. Technol., 12, 552-560 (2001). [62] Smith, D.J. and Partridge, P.G., Flexural stiffness envelopes for planar multilayered systems containing two dissimilar materials, Proc. Instn. Mech. Engrs., 213, 1-19 (1999). [63] Altenbach, H., Theories for laminated and sandwich plates. Mech. Compos. Mater., 34, 243-252 (1998). [64] Dyer, S.R., Lassila, L.V.J., Jokinen, M. and Vallittu, P.K., Effect of cross-sectional design on the modulus of elasticity and toughness of fiber-reinforced composite materials, J. Prosthet. Dent., 94, 219-226 (2005). [65] Salomi, A., Greco, A., Pacifico, T., Rametta, R. and Maffezzoli, A., Processing and properties of a polymer/composite double/layer laminate, Adv. Polym. Technol., DOI: 10.1002/adv.20267. [66] Scheirs J., Polymer Recycling. Chichester: Wiley (1998). [67] Utracki, L.A., Polymer Blends Handbook. Dordrecht: KIuwer Academic Publishers, p. 1117-1165 (2003). 87
[68] Dittenber, D. B. and GangaRao, H. V. S., Critical review of recent publications on use of natural composites in infrastructure, Composites Part A, 43, 1419-1429 (2012). [69] Tasdemır, M., Biltekin H. and Caneba, G.T., Preparation and characterization of LDPE and PP—Wood fiber composites, J. Appl. Polym. Sci., 112, 3095-3102 (2009). [70] Cui, Y., Lee, S., Noruziaan, B., Cheung, M. and Tao, J., Fabrication and interfacial modification of wood/recycled plastic composite materials, Composites Part A, 39, 655-661 (2008). [71] Kazemi Y., Cloutier A., Rodrigue D., Natural fiber composites based on postconsumer polyolefins and wood fiber residues: Effect of coupling agent addition. In: PPS Americas Conference. Niagara Falls, May, p. 410-411 (2012). [72] Jose, S., Aprem, A.S., Francis, B., Chandy, M.C., Werner, P., Alstaedt, V. and Thomas, S., Phase morphology, crystallization behavior and mechanical properties of isotactic polypropylene/high density polyethylene blends, Eur. Polym. J., 40, 21052115 (2004). [73] Arriola, D.J.,
Carnahan, E.M., Hustad, P.D., Kuhlman, R.L. and Wenzel, T.T.,
Catalytic production of olefin block copolymers via chain shuttling polymerization, Science, 312, 714-719 (2006). [74] Maiti, S.N. and Sinch, K., Influence of wood flour on the mechanical properties of polyethylene, J. Appl. Polym. Sci., 32, 4285-4289 (1986). [75] Clemons, C. Wood-plastic composites in the United States: The interfacing of two industries, Forest. Prod. J., 52, 10-19 (2002). [76] Fabiyi, J.S. and McDonald, A.G., Effect of wood species on property and weathering performance of wood plastic composites, Composites Part A, 41, 1434-1440 (2010).
88
[77] Clemons, C, Sabo, R.C., Kaland, M.L. and Hirth, K.C., Effects of silane on the properties of wood-plastic composites with polyethylene-polypropylene blends as matrices, J. Appl. Polym. Sci., 119, 1398-1409 (2011). [78] Kazemi, Y., Cloutier, A. and Rodrigue, D., Mechanical and morphological properties of wood plastic composites based on municipal plastic waste, Submitted to Polym. Compos. (2012). [79] Zenkert, D., An Introduction to Sandwich Construction, Cradley Heath: Engineering Materials Advisory Services (1995). [80] Gere, J.M. and Timoshenko, S.P., Mechanics of Materials. Boston: PWS Publishing Company (1997). [81] Backstrom, D. and Nilsson, A.C., Modeling the vibration of sandwich beams using frequency-dependent parameters, J. Sound Vib., 300, 589-611 (2007). [82] Miller, A.G., Lovell, D.T. and Seferis, J.C., The evolution of an aerospace material: Influence of design, manufacturing and in-service performance, Compos. Struct., 27, 193-206 (1994). [83] Bolotin, V.V., Delaminations in composite structures: Its origin, buckling, growth and stability, Composites Part B, 27, 129-145 (1996). [84] Tay T.E., Characterization and analysis of delamination fracture in composites: An overview of developments from 1990 to 2001, Appl. Mech. Rev., 56, 1-31 (2003). [85] Daniel, I.M., Gdoutos, E.E. and
Rajapakse, Y.D.S., Major Accomplishments in
Composite Materials and Sandwich Structures. New York: Springer (2009).
89
90
Appendix A Technical information of the ethylene-octene copolymer (EOC) used
91
92
Appendix B Technical information of the maleic anhydride grafted polypropylene (MAPP) used
93
94
Appendix C Technical information of the maleic anhydride grafted polyethylene (MAPE) used
95
96
