Composite Technologies for 2020: Proceedings of the Fourth Asian-Australasian Conference on Composite Materials ACCM 4

COMPOSITES TECHNOLOGIES FOR 2020

This page intentionally left blank

COMPOSITES TECHNOLOGIES FOR 2020

ACCM-4

Proceedings of the Fourth Asian-Australasian Conference on Composite Materials (ACCM-4) University of Sydney, Australia 6 - 9 July 2004

Edited by L. Ye, Y.-W. Mai and Z. Su

Organised by The Asian-Australasian Association for Composite Materials (AACM) The University of Sydney

x e

The University of Sydney

WOODHEAD PUBLISHING LIMITED Cambridge England

Published by Woodhead Publishing Limited, Abington Hall, Abington Cambridge CB1 6AH, England www.woodhead-publishing.com First published 2004, Woodhead Publishing Limited © 2004, Woodhead Publishing Limited The authors have asserted their moral rights. This book contains information obtained from authentic and highly regarded sources. Reprinted material is quoted with permission, and sources are indicated. Reasonable efforts have been made to publish reliable data and information, but the authors and the publisher cannot assume responsibility for the validity of all materials. The publisher makes no representation, express or implied, with regard to the accuracy of the information contained in this book and cannot accept any legal responsibility or liability for any errors or omissions. Neither the authors nor the publisher, nor anyone else associated with this publication, shall be liable for any loss, damage or liability directly or indirectly caused or alleged to be caused by this book. Neither this book nor any part may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, microfilming and recording, or by any information storage or retrieval system, without permission in writing from the publisher. The consent of Woodhead Publishing Limited does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific permission must be obtained in writing from Woodhead Publishing Limited for such copying. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation, without intent to infringe. British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library.

ISBN 1 85573 831 7 Printed by Antony Rowe Limited, Chippenham, Wilts, England

Contents Preface Conference Organisation ACCM-4 Sponsors

xix xxi xxiv

Part I: Bio/Eco-Composites All-plant Fiber Composites Zhang, M. Q., Rong, M. Z, Lu, X.

3

Compression Molding and Mechanical Properties of Composite Materials from Post Consumer Type Fiber Waste Hatta, S., Kimura, T., Gonno, H., Kadokura, K.

9

Various Lignocellulosics Fibre Reinforced Polyester Composites: The Study on Mechanical, Physical and Biological Properties Abdul Khalil, H. P. S., Issam, A.M.

15

Development of Eco-Cement Containing High Volumes of Waste Glass Sobolev, K., Iscioglu, G, Tiirker, P., Yeginobali, A., Ertiin, T.

21

Screwless Extrusion of Natural Fibre-Reinforced Thermoplastic Composites Galea, T., Mills, T., Halliwell, R., Jayaraman, K.

27

Mechanical Properties of "Green" Composites Made from Starch-Based Biodegradable Resin and Bamboo Powder Takagi, H., Takura, R., Ochi, S.

33

Part II: Characterisation Effect of Fibre-Orientation on Mechanical Properties of Polypropylene Composites Houshyar, S., Shanks, R. A., Hodzic, A.

41

Friction and Wear Properties of Potassium Titanate Whiskers Reinforced PTFE Composites Feng, X., Chen, D. H., Jiang, X. H., Sun, S, Lu, .X, Jin, Y.

46

Study on Jute Fiber Reinforced Polypropylene (PP) Composite Ma, S., Zhang, A. D., Ding, X., Wang, Y. M. Mechanical and Thermal Properties of Composites of Epoxy Resin Derived from Kraft Lignin Filled with Cellulose Particles Funabashi, M., Hirose, S., Hatakeyama, H.

52

57

vi

Contents

Effects of Microcracks and Surface Roughness on Thermal Oxidation of CarbonFiber Reinforced Polyimide Composite Kung,H.-K., Chen,H.-S.

62

Effects of Fillers on the Tensile Properties of Polyimide Composite Films at Room and Cryogenic Temperatures Zhang, Y. H., Fu, S. Y, Li, M., Li, Y., Li, L. F., Yan, Q.

68

Processing Effects on Electrical Conductivity and Mechanical Properties of Particulate Composite Mohafezatkar, F., Haddadi-asl, V., Nazokdast, H.

72

Study on the Hydrolysis-resistant Polyethylene Terephthalate (PET) Fibers Wang, Y. P., Wang, Y. M. Size Effect on the Compressive Strength of T300/924C Carbon Fiber-Epoxy Laminates in Considering Influence of an Anti-buckling Device Lee, J. W., Kong, C. D., Soutis, C. Environment Effect of Natural Sisal Fibre Reinforced Epoxy Composites Manufactured by Resin Transfer Molding Zhang, X. P., Yuan, Q., Ngatimin, W., Whitbourn, J., Ye, L. Effective Thermomechanical Properties of Interpenetrating-Structured Composites Tilbrook, M., Moon, R., Rutgers, L., Hoffman, M.

78

83

88

94

Mechanical and Thermal Properties of Phenolic Composites Reinforced with Hybrid of Spun and Continuous Carbon Fabrics Kang, T. J, Shin, S. J., Jung, K, Cho, Y. J.

100

Specific Properties vs. Microstructures for Syntactic Foam Wouterson, K, Boey, F., Hu, X., Wong, S.-C.

106

Functionally-Graded Structure and Properties in Human Teeth Low, I. M., Mahmood, U.

112

Experimental Investigation of Porosity in Carbon/Epoxy Composite Laminates Zhang, B.M., Liu, L, Wu, Z. J., Wang, D. F. Characterisation of a Reinforced PPS Thermoplastic Laminate For Forming Simulations Chen, Z. P., Phung, T., Paton, R., de Bruijn, P.

118 124

Characterisation of the Thermo-mechanical Behaviour of a Glass Reinforced Vinyl Ester Composite St John, N. A., Gardiner, C. P., Dunlop, LA. 131 Experimental Study on the Flexural Behavior Using Polyethylene Coated Bars Kim, Y. J.

137

Contents Moisture Absorption by Cyanate Ester Modified Epoxy Resin Matrices: Effect of Resin Structure Karad, S., Jones, F. New Epoxy Resins Based On Azomethine Groups For Potential Polymer Applications Issam, A. M., Abdul Khalil, H. P. S., Wan Rosli, W. D.

vii

143

149

Mechanical Properties of Rotational Moulded Polyethylene Composites Experiments and Theories Yan, W., Lin, R. J. T., Bhattacharyya, D.

154

Morphology and Mechanical Properties of HDPE Reinforced with PET Microfibres Seltzer, R., Fasce, L., Frontini, P., Rodriguez Pita, V. J., Pacheco, E. B. A. V., Dias, M. L.

163

Mechanistic Evaluation of Environments on Degradation of E-Glass/Vinylester Composites Karbhari, V. M., Chu, W., Wu, L. X.

169

High Value Composites from Recycled Polyolefins and Rubbers Fainleib, A., Grigoryeva, O., Tolstov, A., Starostenko, O.

175

Viscoelastic Behaviour, Thermal Properties and Morphology for New Composites from Recycled HDPE, EPDM, Ground Tyre Rubber (GTR) and Bitumen Grigoryeva, O., Tolstov, A., Starostenko, O., Fainleib, A., Lievana, E., Karger-Kocsis, J.

181

Effect of Coupled Long-Term Seawater Exposure and Bi-Axial Creep Loading (2:1) on Durability of Fiber-Reinforced Polymer-Matrix Composites Chen, X. H., Gokdag, E., Wang, S. S. 187

Part III: Composite Structures Deformation Analysis of Kinematically Constrained Thermoplastic Composite Plates in Forming Temperature Daghyani, H. R., Abadi, M. T., Fariborz, S.

195

Stability Analysis of Loaded Columns Made of Pultruded Composites Mahajerin, E.

201

Identification of Elastic Parameters for Cross-ply Laminated Plates and Shells Hosokawa, K., Matsumoto, K.

207

Parametric Instability Analysis and Experiment of Laminated Composite Shell Yeh,M.-K., Huang, H.-C.

212

viii

Contents

Bending Properties of Braided Composite Tubes Okano, M., Sugimoto, K, Nakai, A., Hamada, H.

218

Analytical Stress Analysis of Rotating Composite Beams Due to Material Discontinuities Tahani, M., Nosier, A., Rezaeepazhand, J., Zebarjad, S. M.

223

Thin-plate Splines for Thick Composite Plate Analysis Ferreira, A. J. M.

229

Part IV: Delamination Evaluation of Fatigue Delamination Behavior in Hybrid Composite Material using the Delamination Shape Parameters Song, S.-K, Kim, C.-W., Oh, D.-J. The Effect of Stitch Distribution and Stitch Pattern on Mode I Delamination Toughness of Stitched Laminated Composites Wood, M. D. K, Sun, X. N., Tong, L. Y., Katzos, A., Rispler, A. Dynamic Analysis for Delaminated Composites with Arbitrary Shaped Multiple Delaminations Based on Higher-Order Zig-Zag Theory Cho, M., Oh, J., Kim, J.S., Kim, G.-I.

237 243

249

Part V: Design and Optimisation Application of Stochastic Optimization to Reconstruction of Random Microstructures Pyrz, R-, Bochenek, B. 257 Optimal Design of Filament Wound Structures Based on the Semi-geodesic Path Algorithm Kim, C.-U., Kang, D.-H., Hong, C.S., Kim, C.-G.

264

Development of a Material Mixing Method for Topology Optimization of Multiple Material Structures Han, S. Y., Lee, S. K, Park, J. Y. 270 Structural Design of a 750kW Composite Wind Turbine Blade Jung, C. K, Park, S. K, Han, K. S.

276

Axiomatic Design of Composite Track Pin Park, D. C, Kim, S. S., Lee, S. M., Lee, D. G.

282

Characterization and Design Optimization of FRP Composite Modular System for Slab-on-Girder Bridges Cheng, L. J., Karbhari, V. M.

288

Contents

ix

Design of Composite-Antenna-Structures with High Electrical and Mechanical Performances You, C.S., Hwang, W.

294

Filament Wound Spherical Composite Pressure Vessel Design by an Energy Method KimB.-S., Joe, C.-R.

299

Part VI; Failure Analysis Numerical Study on Buckling of Z-pinned Composite Laminates Yan, W. Y., Liu, H.-Y., Mai, Y.-W.

307

A Three Dimensional Approach of Fatigue Crack Propagation for Aluminum Panels Repaired with Single-Sided Composite Laminates Hosseini-Toudeshky, H., Sadeghi, G., Daghyani, H. R.

313

Time-temperature-water Absorption Superposition Principle for Flexural Fatigue Strength of Unidirectional CFRP Laminates Ichimura, J., Sekine, N., Nakada, M., Miyano, Y.

319

Buckling of Composite Plates with Cutouts Rezaeepazhand, J., Darbari, A. M.

325

Fatigue Crack Propagation in Graded Composites Tilbrook,M., Rutgers, L., Moon, R., Hoffman, M.

331

Dynamic Response Behavior of Stiffened Delaminated Plates Considering Failure Bai,R.X., Chen,H.R,, Wang, M.

337

Tensile Behaviour of Polymer Coated Optical Fibres Law, S., Yan, C, Ye, L.

343

Statistical Model for Multiaxial Fatigue Behavior of Unidirectional Laminates Diao.X.X, Lessard,L.B.

349

Part VII: FEM/Simulation Simulation of Three-dimensional Flow in Compression Resin Transfer Molding by the Control Volume/Finite Element Method Shojaei, A., Boorboor, D., Ghaffarian, S. R.

357

Modeling of Two Dimensional Cellular Solids with Two Types of Imperfections Li,K., Gao,X.-L.

363

Finite Element Modeling of Fine Structure of Natural Plant Fibers for Statistical Characterization of Their Tensile Strengths Suzuki, K., Kimpara, I., Funami, K.

370

x

Contents

Multilayered and Selective Higher-Order-Deformable Sandwich Finite Element Modeling for Numerical Accuracy Improvement Suzuki, K., Kimpara, I.

376

Adhesion Measures of Elasto-plastic Thin Film via Buckle-driven Delamination Li, Q. Y., Yu, S. W.

382

Evaluation of Intra-ply Shearing Stiffness for a Plain Weave Fabric Prepreg

Yu,X.B., Ye,L, McGuckin, D.

388

Part VIII: Fracture Initial Fracture Behaviour of the Weft-Knitted Textile Composites Having WeltKnit Architectures Khondker, O. A., Fukui, T., Nakai, A., Hamada, H.

397

Mixed-mode Fracture of a CF/PEI Composite Material Choupani, N., Ye, L, Mai, Y.-W.

403

Effects of Molecular Structure on the Essential Work of Fracture of Amorphous Copolyester at Various Deformation Rates Chen,H.B., Wu, J. S., Karger-Kocsis, J.

409

How to Eliminate Buckling in the Essential Work of Fracture Measurement with Very Thin Plastic Films Chen, H. B., Liu, S. L, Wu, J. S.

416

Fracture Behaviour of Sandwich Laminates Reinforced by Short-Glass Fibres Khatibi, A. A.

422

Analytical and Numerical Simulations of Plastic Zone at Crack Tip in Anisotropic Solids Liu,H.-X., Ye,Z.-M., Liu, H.-Y.

428

Application of Essential Work of Fracture Methodology to Polymer Fracture Duan, K., Hu, X. Z, Mai, Y.-W.

433

Three-Dimensional Micromechanics Analysis of Strain Energy Release Rate Distribution along Delamination Crack Front in FRP Tanaka, H., Nakai, Y.

439

Influence of Fibre/Matrix Interphase on Crack Bridging Behaviour During Mode I Fracture in Glass Fibre Composites Feih, S., S0rensen, B. F. 445 Behavior of Brittle Reinforced Composites Fracture at Elevated Temperatures Mohamed, A. T.

452

Contents

xi

Part IX: Impact Non-woven Fabric Reinforced Cellular Textile Composites with Improved Energy Absorption Capacity Lam, S. W., Tao, X. M., Yu, T. X.

461

Energy Absorption Properties of Braided Composite Tubes Okano, M., Sugimoto, K., Nakai, A., Hamada, H.

466

Characterization of Damage Resistance and Damage Tolerance of Composite Materials Shen, Z, Yang, S. C, Fu, S. Y., Ye, L.

472

Simplified Prediction Method of Impact Response on Composite Laminates Kim, S. J., Hwang, I. H.

All

Indentation Responses and Damage in Kaolin/Cellulose-Fibre Epoxy Nanocomposites Vaihola, S., Vilaiphand, W., Lopez, A., Low, I. M.

482

Impact Performance of 3D Interlock Textile Composites Byun, J.-K, Urn, M.-K., Hwang, B.S., Song, S.-W.

488

The Ballistic Impact Behavior of Composites Reinforced by Biaxial Weft Knitted UHMWPE Fabrics Liang, Z.-Q., Qiu, G.-X., Yi, X.-S. Low Speed Impact Behavior of Aluminum Honeycomb Sandwich Panel Song,J.-L, Bae,S.-I., Han, M.S., Ham, K.-C.

494

500

Part X: Industrial Applications Thermomechanical Analysis of Water Aged Pultruded Composites Al-Assafi, S.

509

Secondary Bonding in the Construction of Large Marine Composite Structures Simpson, G. J., Burchill, P. J.

515

Surface Analysis of "Class A" Polymer Composite Substrates for the Automotive Industry Schubel, P. J., Harper, L. T., Turner, T. A., Warrior, N. A., Rudd, C. D., Kendall, K. N.

521

Injection Molding of Silk Composite from Industrial Fiber Waste Kimura, T., Suzuki, T., Hatta, S.

527

Test of Full Scale Integrally Stiffened Composite Spoiler Rispler, A.

533

xii

Contents

Study on the Polypropylene(PP) Fiber/Cement Mortar Workability Zhang, H., Zou, L. M., Ni, J. H., Wang, Y. M. Hybrid Composites for Engineering Application Ahmad, F., Latif, M. Ridzuan. A., Nisar, H.

539 545

Development of a Knowledge Warehouse for Intelligent Risk Mapping and Assessment System Savci, S., Kayis,B.

551

Opportunities for Nanocomposites in the Oil & Gas Industry Varley,R., Leong, K. H.

557

Part XI; Interface Interfacial Properties of Polypropylene Fibre-Matrix Composites Houshyar, S., Shanks, R. A., Hodzic, A. Surface Grafting of Nano-SiC with Glycidyl Methacrylate in Emulsion and Its Effect on the Tribological Performance of Epoxy Composites Rong, M. Z, Luo, Y., Zhang, M. Q., Wetzel, B., Friedrich, K.

565

571

Influence of Matrix Type and Processing Conditions on the Morphology of the Interface and the Interfacial Adhesion of PE/PE Composites Masoomi, M., Ghaffarian, S. R., Mohammadi, N.

577

Filler-Elastomer Interactions: Effect of Ozone Treatment on Adhesion Characteristics of Carbon Black/Rubber Composites Park, S.-J., Lee, H.-Y., Lee, J.-R., Min, B.-G.

583

Interface End Theory and Fragmentation Test Ji, X, Dai, Y., Ye, L, Mai, Y.-W.

588

Stress Singularity Analysis of Interface End and Specimen Design for Fiber Pullout Test Dai, Y, Ji, X., Ye, L, Mai, Y.-W. 594

Part XII; Joint Evaluation of Strength of SiC/SiC Composite Joint Using New Interface Potential Serizawa, H., Lewinsohn, C. A., Murakawa, H.

603

Static and Fatigue Analysis of Double-Bolted-Joints for Gr/Epoxy after Thermal Cyclic Loading Yip, M.-C, Li, R.-Y., Yang, C.-H.

609

Tensile Strength and Fatigue Properties of Z-Pinned Composite Lap Joints Chang, P., Mouritz, A. P., Cox, B. N.

615

Contents

xiii

The Effect of Thickness on Joint Property of Mechanical Joint with Washer and Torque Ochi, A., Sugimoto, K., Nakai, A., Hamada, H.

621

Part XIII: Metal Matrix Composites Micromechanical Modelling of Hybrid Metal Matrix Composites Babu, P. E. J., Savithri, S., Pillai, U. T. S., Pai, B. C.

629

Formation of Nanostructured Magnesium Composite Reinforced by in-situ TiC Lu, L, Gupta, M., Toy, K. W.

635

Preparation of Mg-based Hydrogen Storage Nano-composite by Reaction Ball Milling Hu, Y. Q., Yan, C, Zhang, H. F., Ye, L, Hu, Z. Q.

639

A Tin-based Composite Solder Reinforced by Nano-sized Particulates and its Soldering Ability Zhang, X. P., Shi, Y. W., Ye, L, Mai, Y.-W.

645

Suitability of Metal Composite Suspensions for Injection Moulding Ahmad, F.

651

Effect of Intermetallic Volume Fraction on the Mechanical Properties of Intermetallic/Metal Micro-laminated Composites

Kim, H. Y., Hong, S. H., Chung, D. S., Enoki, M.

Part XIV: Nanocomposites Fracture Behaviour of Nano-composite Ceramics Soh,A.K., Fang,D.-N., Dong, Z.-X.

665

Structure-property Relationships of Polymer Nanocomposites Filled with Mechanochemically Grafted Nanoparticles Ruan, W. K, Zhang, M. Q., Rong, M. Z, Friedrich, K.

671

A Numerical Model for Evaluating Elastic Properties of Carbon Nanotube Reinforced Composites Hu, N., Fukunaga, H, Kameyama, M.

677

Interfacial Bonding Strength between Carbon Nanotubes and Epoxy Resin Matrix: Experimental and Computational Studies Wang, B., Liang, Z. Y., Gou, J. H, Jiang, T. H., Zhang, C, Kramer, L.

683

Epoxy-clay Nanocomposites: Morphology, Moisture Absorption Behavior and Thermo-mechanical Properties Hu, C. G., Kim, J.-K., Ban, S.

689

xiv

Contents

Study on Fabrication and Properties of Nano-alumina Particles Reinforced Thermosetting Matrix Composites Cui, Y. H, Tao, J., Wo, D. Z. Investigating High Strain Rate Responses of Nylon 6/ Clay Nanocomposites Huang, J. C, Tsai, J.-L. Mechanical Properties of SiC>2/Epoxy Nanocomposites at Cryogenic Temperature Huang, C. J., Fu, S. Y., Zhang, Y. H, Li, L. F.

695 701 707

Mechanical Properties and Fracture Performance of Nanoclay-reinforced Polypropylene modified with Maleic Anhydride Wong, S.-C, Chen, L, Liu, T. X., He, C. B., Lu, X. H.

713

Nanoclay reinforced UV Curable High-barrier Coatings Uhl, F. M., Davuluri, S. P., Wong, S.-C, Webster, D. C.

719

A Comprehensive Study on Intercalation and Exfoliation of Epoxy/Clay Nanocomposites Liu, J., Wu,J.S. Hydrogen Bonding, Mechanical and Physical Property, and Surface Morphology of Waterborne Polyurethane / Clay Nanocomposite Ma, C.-C. M., Kuan, H.-C, Chuang, W.-P., Su, H.-Y. Thermal Mechanical and Electrical Properties of Multiwall Carbon Nanotube/Waterborne Polyurethane Nanocomposite Kuan, H.-C, Ma, C.-C. M.

725

731

736

Preparation and Properties of Toughened Novolac Type Phenolic /SiO2 Flame Retardant Nanocomposite Ma, C.-C. M., Tai.H., Chiang, C.-L, Kuan, H.-C, Yang, J.-C, Hsu, C.-W.

742

Synthesis, Thermal Properties and Flame Retardance of Novel Phenolic Resin/Silica Nanocomposites Chiang, C.-L, Ma, C.-C. M., Kuan, H.-C, Chang, H. R., Lu, S.-C.

748

On the Enhancement of the Creep Resistance of Polymer by Inorganic Nanoparticles Zhang, Z., Yang, J.-L, Friedrich, K.

754

The Stress Transfer in a Single-Walled Carbon Nanotube-Reinforced Epoxy Xiao, K. Q., Zhang, L. C.

760

A Study on Mechanical Properties of MWNT/PMMA Nanocomposites Kim, H.-C, Lee, S.-E., Kim, C.-G., Lee, J.-J.

766

Fabrication and Microstructure of Si3N4-TiCnano Composites Zhao, J., Huang, X. P., Ai, X, Lu, Z. J.

772

Contents

xv

Synthesis, Thermal and Wear Properties of Waterborne Polyurethane/Polysilicic Acid Nanocomposite Su, H.-Y., Ma, C.-C. M., Kuan, H.-C, Wang, C. P.

778

Preparation and Properties of Epoxy-Bridged Polyorganosiloxanes NanoComposite Lee, T.-M., Ma, C.-C. M., Hsu, C.-W., Chiang, C.-L.

784

Moisture Absorption and Hygrothermal Aging of Organo-montmorillonite Reinforced Polyamide 6/Polypropylene Nanocomposites Chow, W. S., Mohdlshak, Z. A., Karger-Kocsis, J.

790

Geopolymer Reinforced Polyethylene Nanocomposites Yuan, X. W., Easteal, A. J., Bhattacharyya, D.

796

Part XV; Processing Understanding the Thermoforming Issues of Carbon Fibre Reinforced Polyphenylene Sulphide [PPS] Composite Hou,M., Ye,L.

805

A Numerical Approach to Analyze the Curing Process of Railroad Composite Brake Shoe Shojaei, A., Abbasi, F.

811

Possibility of Fabricating Mixed cx/(3 Sialon Ceramics as Composite Materials Karunaratne, B. S. B.

817

High Quality and Low Cost Manufacture of Potassium Titanate Whiskers Lu, X. H, Liu, C, He, M., Yang, Z. H, Bao, N. Z, Feng, X.

823

What Darcy Really Meant - the Truth on Permeability Bechtold, G.

828

Study on Re-pull Force in Pultrusion Processes—I. Experimental Observations Smith, C, Johnstone, B., Lu, M., Ye, L, Mai, Y.-W.

834

Study on Re-pull Force in Pultrusion Processes—II. Theoretical Analysis Lu, M., Ye, L, Mai, Y.-W., Smith, C, Johnstone, B. Influence of Foaming Temperature and Time on the Hardness of Cellular Al-Si-Cu-Mg Alloys Hasan, MD. A., Kim, A., Lee, H.-J., Cho, S.-S.

840

846

Forming Characteristics of Aluminium and Glass-Reinforced Thermoplastic FibreMetal Laminates Mosse, L., Compston, P., Kalyanasundaram, S., Cardew-Hall, M., Cantwell, W. 852

xvi

Contents

Compaction of Single Layer Plain Weave Fabric Preform Chen, Z.-R., Ye, L, Kruckenberg, T. A Study on the Control Strategy to Minimize Voids in Resin Transfer Mold Filling Process Park, Y.-H., Lee, D. H., Lee, W. II, Rang, M. K.

858

864

Fabrication Process and Characterization of Conductive Composite for PEFC Bipolar Plates Heo, S. I., Yun, J. C, Yang, Y. C, Han, K. S. 870 SiO2/Sulfonated PEEK Doped with Dodecatunstophosphoric Acid Hybrid Materials — Preparation and Properties Wu, H.-L, Ma, C.-C. M.

876

Thermoplastic Composite Access Cover Manufactured by Co-Consolidation after Thermoforming Stiffers Wang,K.J., Yi,X.-S.

882

Dome Forming of Triaxial Non-Crimp Fabrics Kong, H., Mouritz, A. P., Paton, R.

888

In situ Microfibrillar Reinforced Composites of PET/PC Liang, G. G., Easteal, A. J.

894

Part XVI: Smart Composites EPR and Magnetic Susceptibility Studies on the Structure and Polaron Dynamics on V2O5-, MOO3- and CuO- Containing Glasses Das, B. B., Ambika, R., Ageetha, S., Vimala, P.

903

Increase of High Burst Pressure in CFRP Vessels Reinforced by SMA Fibers Ben, G., Sakata, K.

908

Monitoring the Strain in the CFRP Laminates and CFRP/Concrete Structures Ogi, K., Takao, Y.

914

Free Vibration of Perforated Aluminum Plates Reinforced with Bonded Composite Patches Rezaeepazhand, J., Sabouri, H.

920

Design Method for SMA Super Hybrid Composite Materials Zhang, B. M., Li, S. L, Wu, Z. J., Du, S. Y, Li, Q. F. Control of Crack Closure in Shape Memory Alloy TiNi Fiber embedded CFRP Composite Materials Shimamoto, A., Lee, C.-C.

926

931

Contents Influence of Stress Induced Birefringence on FBG Sensors Embedded in CFRP Laminates Mizutani, T, Takeda, N., Nishi, T, Tsuji, R., Okabe, Y.

xvii

937

Magnetoelectric Properties of Piezoelectric and Magnetostrictive Composites with 2-2 and 3-1 Connectivity Huang, H. T, Zhou, L. M.

943

Shape Memory Effect on Interfacial Strength of SMA-reinforced Composites Poon, C.-K., Zhou, L. M.

949

Part XVII: Structural Health Monitoring Modified Acoustic Emission Generated in a Full-Scale Aircraft Wing Subjected to Simulated Flight Loading Paget, C. A., Atherton, K, O'Brien, E.

957

In-situ Health Monitoring of Filament Wound Pressure Tanks using Embedded FBG Sensors Rang, D.-K, Kim, C.-U., Park, S.-W., Hong, C.S., Kim, C.-G.

963

An Approach towards Predicting the Evolution of Fire Damage for Marine Composites Mathys, Z., Gardiner, C. P., Burchill, P. J.

969

A Bayesian Artificial Neural Network Method to Characterise Laminar Defects using Dynamic Measurements Lam, H. F., Veidt, M., Kitipornchai, S. 975 A Damage Detection Technique of Composite Laminates with Embedded FBG Sensors Kim, W.S., Kim, S.-H, Lee, J.-J.

981

Damage Detection in Composites Using Fiber Bragg Grating Sensors as Ultrasonic Receivers Okabe, Y., Tamaue, H, Kuwahara, J., Takeda, N. 987 Inverse Analysis for Damage Identification in CFRP Laminates with Embedded FBG Sensors Yashiro, S., Okabe, T, Takeda, N.

993

Parameterised Modelling Technique & Its Application to Artificial Neural Networkbased Structural Health Monitoring Huang, N., Ye, L, Su, Z. 999 Information Fusion in Distributed Sensor Network for Structural Damage Detection Wang, X. M., Foliente, G, Su, Z, Ye, L.

1005

xviii

Contents

Remaining Life of FRP Rehabilitated Bridge Structures Lee, L. S., Atadero, B., Karbhari, V. M., Sikorsky, C.

1012

Delamination Monitoring of CFRP Laminates Using Electrical Potential Method Ueda, M., Todoroki, A., Shimamura, Y., Kobayashi, H.

1018

Damage Detection in Glare plate-like structures Rosalie, S. C, Chiu, W. K.

1025

Quantitative Nondestructive Evaluation in Composites Beam Using Piezoelectrics Choi, Y.-G., Su, Z, Chen, Z.-R., Ye, L. 1032

Part XVIII: Textile Composites Mechanical Properties of Textile Hybrid Composite Inoda, M., Sugimoto, K., Nakai, A., Hamada, H.

1041

Effects of Fabrication and Processing Techniques of Aramid/Nylon Weft-Knitted Thermoplastic Composites on Tensile Behaviour Khondker, O. A., Fukui, T., Nakai, A., Hamada, H.

1047

Modeling and Characterization of 3D Heterogeneous Tissue Scaffolds Fang, Z., Starly, B., Darling, A., Sun, W.

1052

Continuative Fabrication and Mechanical Properties of Multi-axial Warp Knitted Thermoplastic Composites Using Micro-braided Yarn Narita, T., Nakai, A., Hamada, H, Komiya, I., Fukui, E.

1058

Multi-Scale Analysis for Material Characterization of Textile Composites Liang, J., Wang, K. S., Du, S. Y.

1064

Study on Damage Development of Woven Fabric Composites with Spread Tow Kurashiki, T., Zako, M., Hayashi, Y., Verpoest, I.

1070

Measurement of Material Damping Properties of Triaxial Woven Fabric Composites in Low-Pressure Condition Zako, M., Kurashiki, T., Nakanishi, Y., Matsumoto, K.

1076

CTE Model of 3D Orthogonal Textile Reinforced Aluminum Matrix Composites Lee,S.-K., Byun, J.-H, Hong, S. H.

1081

Permeability of Sisal Textile Reinforced Composites by Resin Transfer Molding Li, Y., Mai, Y.-W., Ye, L. Index of Authors

1087 1093

Preface Over the past three decades, the terminology of composite materials has been well acknowledged by the technical community, and composite materials have been gaining exponential acceptance in a diversity of industries, serving as competitive candidates for traditional structural and functional materials to realise current and future trends imposed on high performance structures. Striking examples of breakthroughs based on utilisation of composite materials are increasingly found nowadays in transportation vehicles (aircraft, space shuttle and automobile), civil infrastructure (buildings, bridge and highway barriers), and sporting goods (Fl, golf club, sailboat) etc., owing to an improved understanding of their performance characteristics and application potentials, especially innovative, cost-effective manufacturing processes. As the equivalent of ICCM in the Asian-Australasian regions, the Asian-Australasian Association for Composite Materials (AACM) has been playing a vital leading role in the field of composites science and technology since its inception in 1997 in Australia. AACM aims to encourage the interchange of knowledge in all aspects of composite materials amongst both the scientific and engineering communities. The first three ACCM conferences were successfully held in Osaka-Japan, Kyongju-Korea, and Auckland- New Zealand, respectively, every two years since 1998. Following the excellent reputations and traditions of previous ACCMs, ACCM-4 is held in scenic Sydney, Australia, 6-9 July 2004. The theme of ACCM-4, Composites Technologies for 2020, provides a forum to present state-of-the-art achievements and recent advances in composites sciences and technologies, and discuss and identify key and emerging issues for future pursuits. By bringing together leading experts and promising innovators from the research institutions, end-use industries and academia, ACCM-4 intends to facilitate broadband knowledge sharing and identify opportunities for long-term cooperative research and development ventures. We are very pleased with many contributions from authors and overwhelming response from participants. Nearly 200 manuscripts were received from 25 countries and regions, and 65% of them were from areas of the Asian-Australasian region. The proceedings were published from camera-ready manuscripts prepared by the authors. Each paper was vigorously peer-reviewed by at least two independent referees from the local scientific committee and other selected referees. Some papers required multiple revisions. The selected papers, classified under 25 categories, are in a wide spectrum, ranging from general manufacturing and processing techniques to the latest and hottest topics such as nano-composites and eco-bio composites. Together they represent an authoritative documentation of current advances in the field of composite materials. We wish to thank the following for their contributions to the success of this Conference: Air Force Office of Scientific Research, and Asian Office of Aerospace Research and Development. We are also grateful to the Cooperative Research Centre-Advanced Composite Structures (CRC-ACS), the Sydney University Centre for Advanced Materials Technology (CAMT) and the DSTO-AED Centre of Expertise in Damage

xx

Preface

Mechanics, Australian Composite Structures Society (ACSS) and University of Sydney, for their financial and other forms of supports to this Conference. Finally, we would like to thank all members of the ACCM International Advisory and Scientific Committees in promoting this Conference. Special thanks are due to all the authors for the careful preparation of their manuscripts; and conscientious reviewers for maintaining the high scientific quality of the ACCM-4 proceedings.

Lin Ye Yiu-Wing Mai Zhongqing Su Sydney, Australia July 2004

Conference Organisation ACCM-4 Co-Chairmen L. Ye, University of Sydney S.Y. Du, Harbin Institute of Tech. T. Uenoya, Tech. Research Institute of Osaka Prefecture

AACM Steering Committee M. Zako (President), Osaka University L. Ye (Vice-President), University of Sydney J.K. Kim (General Secretariat), Hong Kong Uni. Sci. & Tech.

International Advisory Committee (AACM Council) D. Bhattacharyya, University of Auckland O.I. Byon, Nihon University H.R. Daghyani, Amirkabir University of Tech. Iran T. Fujii, Doshisha University T. Fukuda, Osaka City University K.S. Han, Pohang University of Sci. & Tech. C.S. Hong, KA1ST S.H. Hong, KAIST D.A.L. Juwono-Soenarso, University of Indonesia S.J. Kim, Seoul National University I. Kimpara, Kanazawa Institute of Tech. W.I. Lee, Seoul National University K.H. Leong, Petronas Research & Scientific Services C.C.M. Ma, National Tsing Hua University Y.W. Mai, University of Sydney Y. Miyano, Kanazawa Institute of Tech. M. Nasir, University of Sains Malaysia V.A. Phan, Institute of New Tech. Promotion of Vietnam S. Ramakrishina, National University of Singapore D.Z. Wo, Nanjing University of Aero. & Astro. M.C. Yip, National Tsing Hua University

International Scientific Committee A. Baker, Platforms Sciences Laboratory/DSTO B. Banks, University of Strathclyde T.-W. Chou, University of Delaware I. Crivelli-Visconti, Univerista' Degli Studi Di Napoli Federico II K. Friedrich, University of Kaiserslautern H. T. Hahn, University of California, Los Angeles Y. Hamada, Kyoto Institute of Technology

xxii S. V. Hoa, Concordia University H. Ishida, Case Western Reserve University T. Katayama, Doshisha University D. Kelly, University of New South Wales F. K. Ko, Drexel University W. I. Lee, Seoul National University A. C. Loos, Virginia Tech I. Marshall, Monash University K. Niihara, Osaka University T. K. O'Brien, NASA Langley Research Centre R. Byron Pipes, University of Akron A. Pousartip, University of British Columbia F. Rose, Platforms Sciences Laboratory/DSTO M. Scott, Royal Melbourne Institute of Technology P. Smith, University of Surrey C.-T. Sun, Purdue University N. Takano, Osaka University N. Takeda, University of Tokyo X. Tao, Hong Kong Polytechnic University J. S. Wu, Hong Kong University ofSci. & Tech. X.-S. Yi, Institute of Aerospace Materials/China M. Zhang, Zhongshan University Z. Zhang, University of Kaiserslautern L. Zhou, Hong Kong Polytechnic University

Local Scientific Committee A. Afaghi-Khatibi, Melbourne University M. Bannister, CRC-Advanced Composite Structures A. Beehag, CRC-Advanced Composite Structures W.K. Chiu, Monash University A Crosky, University of New South Wales S. Galea, Platforms Sciences Laboratory/DSTO Israel Herszberg, Royal Melbourne Institute of Tech. M. Hoffman, University of New South Wales M. Hou, CRC-Advanced Composite Structures X.-Z. Hu, University of Western Australia J. Li, University of Technology Sydney I. M. Low, Curtin University of Technology A. Mouriz, Royal Melbourne Institute of Technology Rowan Paton, CRC-Advanced Composite Structures L. Tong, University of Sydney C. Wang, Platforms Sciences Laboratory/DSTO X. Wang, Manufacturing and Infrastru. TechJCSIRO D.-Y. Wu, Manufacturing and Infrastru. TechJCSIRO

Conference Organisation

Conference Organisation Local Organising Committee Z. Chen, University of Sydney R. Connell, University of Sydney N. Huang, University of Sydney K.-Y. Kim, University of Sydney T. Krunkenberg, University of Sydney W. Liang, University of Sydney M. Lu, University of Sydney Y. Lu, University of Sydney Z. Su, University of Sydney C. Yan, University of Sydney

xxiii

ACCM-4 Sponsors We wish to thank the following for their contributions to the success of this conference

Air Force Office of Scientific Research (AFOSR) Asian Office of Aerospace Research and Development (AOARD)

Cooperative Research Centre-Advanced Composite Structures (CRC-ACS), Australia

Australian Composite Structures Society (ACSS)

The University of Sydney

\

AEROMECH @usyd

School of Aerospace, Mechanical and Mechatronic Engineering, University of Sydney

11//

I lie Limcrsih ol SxJruL\ Or*forAtbatvhi Mawnoh i., »a

Parti

Bio/Eco-Composites

This page intentionally left blank

All-Plant Fiber Composites Ming Qiu Zhang , Min Zhi Rong, Xun Lu Materials Science Institute, Key Laboratory for Polymeric Composite and Functional Materials of Ministry of Education, Zhongshan University, Guangzhou 510275, P. R. China

ABSTRACT Plasticization of fir sawdust was carried out in the present work to prepare natural resources based plastics. It was found that thermoplasticity and mechanical properties of the chemically modified wood flour changed with the substitution reaction conditions. By compounding sisal fibers and the plasticized fir sawdust, unidirectional laminates were manufactured similar to conventional thermoplastic composites. Such an all-plant fiber composite material is characterized by easy processing, environment friendly and low cost. Instead of chemical heterogeneity of conventional composites, physical heterogeneity of the current natural fiber composite is favorable for interfacial interaction.

INTRODUCTION Fiber reinforced polymer composites have been widely used in many fields mainly because of their high specific stiffness and strength. Recently, with increasing energy crisis and ecological problems, material scientists began to take interests in vegetable fibers serving as substitutes for man-made fibers in the area of composites [1]. In fact, however, these composites are not environment friendly enough due to the fact that the matrix resins are synthesized products from earth oil and mostly non-biodegradable. Considering the existing dilemma, the authors of the present work use modified plant fiber that can be processed like conventional polymers as matrix to make plant fiber reinforced plasticized plant fiber composite. The following benefits will be gained accordingly, (i) As natural fibers are slightly modified as a whole by chemical methods, cost effectiveness characterized by the renewable raw materials is maintained, (ii) Fully biodegradable ability is associated with the composite because both the reinforcer and matrix are biomaterials. (iii) Instead of chemical heterogeneity of conventional composites, physical heterogeneity of the all-plant fibers composite is favorable for interfacial interaction. Plant fiber consists of cellulose, hemicellulose, lignin and a small amount of extractives. Cellulose is the essential component and belongs to an isotactic /3-1, 4polyacetal of cellulose. High degree of crystallinity of cellulose as well as three dimensional reticulate structure of lignin make plant fibers far from thermoplastic materials. However, the crystalline structure of cellulose might be disrupted by substituting its hydroxyl group with some chemical reagent [2, 3]. This decrystallization process helps to improve thermoplasticity of cellulose since the substitution groups act as plasticizer. Etherification, esterification and graft-copolymerization are proved to be * Corresponding Author, Prof. M. Q. Zhang, Materials Science Institute, Zhongshan University, Guangzhou 510275, P. R. China. Fax: +86-20-84036576, e-mail: ceszmq(5),zsu.edu.cn.

4

All-plant Fiber Composites

effective to introduce plasticization into cellulose. Therefore, plant fibers can thus be converted into thermoplastics by using these techniques [4]. EXPERIMENTAL China fir sawdust was ground into a mesh size of 80-100 and then dried under vacuum at 80°C overnight. After being pre-swelled by ION NaOH for 1 hour, the powder was transferred into a flask containing (CFFj^NI and benzyl chloride. The reaction was carried out under vigorous stirring at 120°C for 4-10 hours to get benzylated products with various reaction extents. The products were purified through washing for 2-3 times with distilled water to remove inorganic salts, and with ethanol to remove residues of benzyl chloride and by-products, respectively. Finally the treated sawdust was dried again under vacuum at 80°C overnight for being used as composite matrix. As the modified fir exhibits rather high viscosity which can not be measured by a conventional melt indexer, a home-made tester was used to assess its flowability in terms of melt index (MI, defined as the weight of the material flowing out of the nozzle in gram within 10 minutes under given pressure and temperature) and flow temperature (defined as the temperature at which the material begins to flow out of the nozzle during heating under a given pressure). To produce plates of neat matrix and unidirectional sisal laminates, benzylated fir sawdust alone and its mixture with continuous sisal fiber at a desired proportion were compressed into sheets using a hot press at 130°C under lOMPa. The molding conditions were chosen based on the viscosity measurements. RESULTS AND DISCUSSION Benzylation of wood is a typical Williamson synthesis reaction, which involves nucleophilic substitution of an alkoxide or a phenoxide ion for a halide ion [5]. Since cellulose constitutes the majority of wood and lignin contains little hydroxyl, benzyl chloride has to react mainly with the hydroxyl of cellulose. Hon et al reported that wood species have little effect on the rate of benzylation [5], but the reaction turned out to be quite slow in the case of fir powder when the conditions suggested in ref.[5] were followed. Our experimental results show that the reaction is remarkably influenced by the concentration and amount of NaOH, amount of benzyl chloride, reaction temperature and time. Based on these findings, the extent of benzylation of fir can be adjusted by taking appropriate measures. Figure 1 illustrates that the chemical structure of benzylated fir is

3700

3300

2900

2500

2100

1700

1300

900

500

1

Wavenumber [cm ]

FIGURE 1 FTIR spectra of (1) unmodified fir sawdust and (2, 3) benzylated fir sawdust with different weight gains (curve 2: 47%, curve 3: 98%).

All-plant Fiber Composites

5

quite different from that of the untreated version. Hydroxyl absorption at about 3400cm"1 diminishes after benzylation as a result of etherification. Because the appearance of the bands at 1800-1950,1600, 736 and 695cm"1 is indicative of the mono-substituted benzene rings in benzyl groups, it can be concluded that the hydroxyl groups of cellulose have been substituted by benzyl groups. By comparing curves 2 and 3, it is seen that the characteristic peaks of benzyl group become stronger with a rise in the amount of the substituent. For an accurate analysis of the reaction and the resultant, degree of substitution originating from benzylation process should be known. Due to the chemical heterogeneity of wood, this parameter is factually hard to be measured. Instead, the extent of benzylation reaction can be evaluated by the percentage weight gain, which has explicit engineering meaning. As exhibited in Figure 2, an increase in the weight gain of the benzylated fir sawdust corresponds to an improvement of its thermal flowability. That is, the more hydroxyl groups of cellulose are substituted by benzyl groups, the higher the melt index and the lower flow temperature. The data in Figure 2, on one hand, demonstrate that the fir sawdust has been converted into a thermoplastic material after benzylation and acquired thermoforming ability. On the other hand, they reveal that the processing window of bezylated fir sawdust is a function of molding temperature and pressure.

20

40

60

Weight gain [%] FIGURE 2 Effect of weight gain on flow temperature and MI of benzylated fir sawdust (pressure=6MPa)

FIGURE 3 WAXD patterns of (1) unmodified fir sawdust and (2, 3) benzylated fir sawdust with different weight gains (curve 2: 47%. curve 3: 98%).

6

All-plant Fiber Composites

Figure 3 gives the WAXD patterns of untreated and treated fir sawdust. It is seen that the peak at 20=22.2° of the as-received sawdust (curve 1), which is derived from the reflection of (002) plane of cellulose I lattice, diminishes after benzylation (curve 2) and forms broader scattering when the extent of benzylation is significantly increased (curve 3). As the diffraction peak profiles shown in curves 2 and 3 resemble those of ball-milled cellulose which had been decrystallized [6], it can be concluded that the above-stated variation in (002) reflection is a result of decrystallization. Such a decrystallization process helps to improve thermoplasticity of cellulose by breaking hydrogen bonds between cellulose molecules. It is believed that the large benzyl groups introduced onto cellulose bring in more free volumes and account for the change in the supramolecular structure. It is worth noting that an increased molecular mobility of cellulose is hard to result in an increased flowability of fir sawdust if the network of lignin keeps intact. Therefore, partial removal and damage of lignin structure as reflected by the absence of the carbonyl band at 1760cm"1 and aromatic ether band at 1275cm"1 (Figure 1) is necessary to get rid of the fetter, which further improves the extent of substitution of cellulose. Prior to the discussion of sisal laminates, mechanical properties of the matrix, benzylated fir sawdust, should be known at first. As shown in Table 1, the molded sheet of modified fir sawdust with a weight gain of 72.8% has the highest performance except for the flexural modulus. Similar to the case of original wood material, crystallinity and molecular weight of cellulose determine the mechanical properties of its plasticized version. For highly substituted wood, more crystalline regions of cellulose are destroyed, while a higher reaction temperature or longer reaction time that is required to increase substitution degree would result in severer degradation than the slightly substituted one. These contradictory effects are responsible for the reduced mechanical properties coupled with increased thermoplasticity (cf. Table 1 and Figure 2). On the other hand, if the degree of substitution is so low that the modified fir sawdust can not be sufficiently melted, a sheet with lower mechanical properties but higher crystallinity retention would be yielded. Evidently, the benzylated fir sawdust with a weight gain of 72.6% seems to be well balanced. As a result, the benzylated products with weight gains of 70%-80% were selected as the matrix material for making sisal laminates in this work. Compared with cyanoethylated pine [4], the benzylated fir sawdust has much higher impact toughness but lower static strength and modulus. The difference should be attributed to the modification techniques rather than wood species. Tensile and flexural properties of unidirectional sisal reinforced benzylated fir sawdust composites are shown in Figure 4 as a function of sisal volume fraction, Vf. It is seen that moduli increase with sisal content in an approximately linear way, while the TABLE I Mechanical properties of thermoformed sheets of benzylated fir sawdust Weight Gain [%] 37.0 54.5 65.2 72.8 81.2 96.7 105.3

Tensile Strength [MPal 12.5 13.8 15.2 17.5 16.3 15.1 12.7

Young's Modulus [GPa] 1.92 1.98 2.21 2.39 2.30 2.24 2.12

Flexural Strength [MPa] 33.6 34.3 34.9 36.8 35.5 34.5 33.9

Flexural Modulus [GPa] 2.20 2.18 2.33 2.47 2.51 2.55 2.50

Impact Strength [kJ/m2] 3.9 4.4 4.7 5.6 5.1 4.5 3.7

All-plant Fiber Composites

7

strengths reach a maximum at Vf=30vol%. Although the latter phenomenon can be attributed to the competition between the effects of reinforcement and micro-crack initiation as a result of fiber incorporation, this drop in strengths appears too early. This is different from the results of thermosetting polymer based plant fiber composites. For example, a linear relationship between tensile strength and fiber loading were found up to Vf=60vol% in unidirectional jute/unsaturated polyester and sisal/epoxy systems [1, 7]. Obviously, wetting and absorption problems at the fiber/matrix interface should be responsible for this behavior, hi the case of thermosetting resins, the low molecular weight monomers before curing facilitate impregnation of the reinforcing fiber bundles to form intimate adhesion. For benzylated fir sawdust that has a quite high melt viscosity, it is somewhat difficult to impregnate sisal fibers under the current processing conditions. That is, the kinetic barrier leads to insufficient interfacial bonding in spite of the fact the fiber reinforcement and matrix are thermodynamically compatible. Consequently, the strengthening effect of sisal fibers can not brought into full play especially at a fiber content higher than 30vol%.

.

—D—Tensile strength —0—Young's modulus

20

20 a. O

8s

15 2 10

10

20

30

10

40

20

30

I

40

Sisal content [vol%]

Sisal content [vol%]

FIGURE 4 Tensile (a) and flexural properties (b) of unidirectional sisal/benzylated fir sawdust laminates as a function of sisal content

100

2

3

Strain [%] Figure 5 Typical stress-strain curves obtained from tensile tests of (1) benzylated fir sawdust, (2) sisal/benzylated fir sawdust laminates (V(=19.7vol%), (3) sisal/benzylated fir sawdust laminates (Vf=31vol%) and (4) sisal/benzylated fir sawdust laminates (Vf=40.4vol%).

It is interesting to examine the failure behavior reflected by the tensile stress-strain plots in Figure 5. Unlike unidirectional sisal/epoxy laminates that are characterized by

8

All-plant Fiber Composites

brittle failure [7], the current composites show long tails after the predominant damage. Such a post-failure crack propagation resistance improves the safety of use and must be related to the interfacial failure characteristics. That is, the localized shear deformation of the ductile matrix induced by the reinforcing fiber accounts for the phenomenon. It is thus expected that an increased interfacial interaction in the present laminates would result in a higher elongation to break. CONCLUSIONS (1) By means of benzylation, fir sawdust can be converted into a thermoplastic material. In comparison with cyanoethylated products, benzylated wood flour has higher thermoplasticity and toughness. (2) Melt viscosity and mechanical performance of benzylated fir sawdust is a function of the extent of benzylation, which provides posibilities for tailoring the structure-properties relationship. From technical point of view, the modified fir sawdust melt is thermally sensitive but not pressure sensitive. (3) Unidirectional sisal reinforced plasticized fir sawdust laminates can be manufactured by similar techniques available for conventional thermoplastic composites and exhibit moderate modulus and strength. To improve the mechanical properties, the flow behavior of the matrix should be greatly improved to ensure sufficient impregnation and interfacial adhesion. Another possible solution might lie in the production of self-reinforced composites, i.e. surface of the reinforcing plant fibers is appropriately plasticized and then the fibers are bound together without additional matrix resin under the joint action of temperature and pressure. In this way, uncoiling of the spirally arranged micro fibrils inside plant fibers, which consumes substantial energy, would occur and provide the composite with enhanced performance. A paper on this topic will be published in the near future. ACKNOWLEDGMENT The financial support by the National Natural Science Foundation of China (Grant: 50173032) is gratefully acknowledged. REFERENCES 1. 2. 3. 4.

5. 6. 7.

Bledzki, A. K. and J. Gassan 1999. "Composites Reinforced with Cellulose Based Fiber", Progr. Polym. Sci, 24:221-274. Morita, M. and I. Sakata 1986. "Chemical Conversion ofWood to Thermoplastic Materials", J. Appl. Polym. Set, 31:832-840. Hon, D. N. S. 1992. "Chemical Modification of Lignocellulosic Materials: Old Chemistry, New Approaches", Polym, News, 17:102-107. Lu, X., M. Q. Zhang, M. Z. Rong, G. Shi, G. C. Yang, and H. M. Zeng. 1999. "Natural Vegetable Fiber/Plasticized Natural Vegetable Fiber - A Candidate for Low Cost and Fully Biodegradable Composite", Adv. Compos. Lett, 8:231-236. Hon, D. N. S. and M. S. L. Josefina. 1989. "Thermoplastization of Wood. I. Benzylation of Wood", J. Polym. Sci, Part A: Polym. Chem., 27:2457-2482. Heritage, K. J., J. Mann, and L. Roldan-Gonzalez. 1963. "Crystallinity and the Structure of Celluloses", J. Polym. Sci, Part A: Polym. Chem., 1:671-695. Rong, M. Z., M. Q. Zhang, Y. Liu, G. C. Yang, and H. M. Zeng. 2001. "The Effect of Fber Treatment on the Mechanical Properties of Unidirectional Sisal-Reinforced Epoxy Composites", Compos. Sci. Techno!., 61:1437-1447.

Compression Molding and Mechanical Properties of Composite Materials from Post Consumer Type Fiber Waste

SeijiHatta* Kyoto Municipal Industrial Research Institute, Textile Technology Center, Japan Teruo Kimura, Hirohisa Gonno Kyoto Institute of Technology, Advanced Fibro Science, Japan Kenzo Kadokura Kadokura Trading Company Co.LTD., Japan

ABSTRACT The composite materials consisted with rag were molded by using the compression molding method and their mechanical properties were investigated. The rag was pre-treated by opener and card machine and mixed with Polypropylene fiber to make the web. After that the felt was made from the web by using the needle punch method. Polypropylene fiber was used for the matrix material, and the volume content of PP was varied in the experiments. The results suggest that the molding method described herein shows promise for contributing toward the material recycling of post consumer type fiber wastes as raw materials of composites. INTRODUCTION In recent years, increased emphasis has been placed on developing the recycling system of fiber waste with the goal ofprotecting the environment. However, relatively little investigation has been conducted regarding such a recycling system. There are two types of fiber wastes such as industrial fiber waste resulted from the process of manufacturing products and fiber waste named "rag" as a post consumer. It is estimated that the used fibers in Japan are released about two million tons per year. Although 10 per cent of waste fiber have been recycled or reused as the used cloths, the industrial wiping cloths, the shoddy and the felt, almost 90 % of waste fiber is destroyed by fire and buried underground. These conditions have raised the concerns about the necessity of the finding innovative usage of rag and new recycling technologies are strongly required. Meanwhile, to meet a wide variety of recent engineering requests, many research efforts related to the development of new materials have been carried out actively in many fields. In this paper, to establish the new material recycling system of the •rags, the compression molding method in which the rags were used as reinforcements of the composite materials was proposed.

* Corresponding Author, kamigyo-ku, Kyoto, 602-0898, Japan, +81-75-441-3165, [email protected]

10

Compression Molding and Mechanical Properties

The rag was used as a reinforcement of composite materials. Figure 1 shows the aspect of rag. Various kind of fibers were mixed in the rag, and the fibers used here are shown in Table 1. As can be seen in this Table, the rag used here is mainly consisted by the cellulose fibers. The measured mechanical properties and the size of fibers are shown in Table 2.The monofilament PP fiber wastes were used as a matrix material.

FIGURE 1 Aspect ofragused TABLE I Materials and component of rag used 2

2 PP30Wt% l3kg/m 2 ) PP50Wt% t5kg/m ) PP70Wt% t5kg/m )

Cellulose fibers (Cotton,Rayon,another) Polyesters Wool Another fiber PP

45— 54%

43— 4 5 %

15— 20%

12— 15% 5— 8% 2— 3% 2 7 - 33%

10— 15% 3— 4% 2— 3% 3 8 - 48%

10— 5— 2r^ 59-

12% 7% 3% 68%

TABLE II Properties of each fiber

Cotton Rayon Polyester PP

Tensile strength(MPa) Young's modulus(GPa) Fiber length(mm) Fineness(D) 1 0 -- 3 0 2 0 0 --350 9 . 3 - -12.7 1.5- - 2 . 3 1 5 0 -- 4 0 0 4 . 5 - - 8.3 1 5 -- 3 5 1.5- - 3 . 0 3 2 0 -- 4 5 0 13--35 3 . 1 - - 7.5 1.5- - 2 . 5 10--55 3 5 0 -- 5 3 0 4 . 5 - - 7.5 2 5 -- 5 0

PRE-MOLDING OF FELT The rag was pre-treated by opener and card machine and mixed with PP fiber to make the web. After that the felt was made from the web by using the needle punch method. Figures 2(a) and (b) show the aspects of molded felt and the cross section. As can be seen in the figure, the various fibers are mixed at randomly together with the PP fiber. The content of PP fiber was varied in the wide range in the experiment.

Compression Molding and Mechanical Properties

(a) Aspect of pre-molded felt

11

(b) Cross section of felt

FIGURE 2 Pre-molded felt

COMPRESSION MOLDING METHOD AND MECHANICAL TESTS For molding of composite materials, the compression molding method was performed. The molding process was as follows. Namely, the pre-molded felts were firstly heated in the furnace with forced convection air flow at 190 degree during 25 minutes in order to melt the PP fibers. After the heating process, the felts were compressed in the die of 190 degree with dimensions of 250 mm x 250 mm until 3 mm or 5 mm thickness. Then the natural cooling was done by keeping the proper pressure. The aspect of molded composite material is shown inFig.3.

FIGURE 3 Molded composite

The tensile, three-point bending and Izod impact tests were performed in accordance with JIS K 7054, JIS K 7055 and JIS K 7062, respectively. The aspects of cross section of composites were also observed by using SEM and micro scope. RESULTS AND DISCUSSION Figures 4(a), (b) and (c) show the aspects of cross section of composites for Wf=30%, 50% and 70%, respectively. Here Wf means the weight fraction of waste fibers in the composite. The good penetration of matrix material can be seen for Wf=30% and 50%. However, the many voids can be clearly seen for Wf=70% because of the lack of matrix material. The adhesion between fiber and PP matrix is not good as shown in these figures because of no-treatment of fiber surface and PP matrix in this experiment.

12

Compression Molding and Mechanical Properties

(a)Wf-=30%

(b)Wf=50% FIGURE 4 SEM observation of molded composites

Figure 5 shows the tensile stress-strain curves for molded composites. The stress-strain curve shows an inflection after the initial linear portion and the tensile strain is found to be maximum for matrix material. The tensile strength and modulus of composite were evaluated from the maximum tensile stress and the slope of initial linear portion of stress-strain curve, respectively.

0.05 Strain E t

0.1

FIGURE 5 Tensile stress-strain curves

C52.5 25

"

to"

I 15

% 1.5 o 6 1

I 10

3

'g 0.5

1 5 *

2

3

0

0 20 40 60 80 Weight fraction of waste fiber Wf (%) (a)Tensile strength

0 20 40 60 80 Weight fraction of waste fiber Wf (%) (b)Tensile modulus

FIGURE 6 Relationship between weightfractionof wastefiberand tensile properties

13

Compression Molding and Mechanical Properties

Figures 6(a) and (b) show the tensile strength and modulus as a function of Wf. As can be seen from the figures, the strength and modulus increase with increasing Wf. However, the notable increase of tensile strength can not be expected for the composite molded here. This may be caused by the lack of adhesion between matrix and waste fibers. Meanwhile, the fairly large value of modulus can be obtained at Wf=50%. It should be noted here that in the large range of Wf such as 70%, the strength and modulus take smaller values than those of matrix material. This may be caused bay the lack of matrix resin in the larger Wf in addition to the lack of adhesion between matrix and waste fibers. Figure 7 shows the bending stress-strain curves. The stress-strain curve shows an inflection after the initial portion as same as the result of tensile stress. The slope of initial linear portion for composites was severe in comparison with that of matrix material.

0

0.05 Strain E b

0.1

FIGURE 7 Bending stress-strain curves

Figures 8(a) and (b) show the bending strength and modulus as a function of Wf. Similar tendency to the tensile properties can be obtained for the results of bending tests. Namely, the strength and modulus take maximum values at Wf=50%, and decrease in the larger Wf. 153.5

1?60 OH

O

•§,50

i40

t

3

W2.5

l 10

60

:

|

•

•

11 1

1

M20

1

0.5

m o 0 20 40 60 80 Weight fraction of waste fiber Wf (%) (a) Bending strength

0 20 40 60 80 Weight fraction of waste fiber Wf (%) (b) Bending modulus

FIGURE 8 Relationship between Weight faction of wastefiberand bending properties

14

Compression Molding and Mechanical Properties

Figure 9 shows the Izod impact values as a function of Wf. It should be noted here that the impact value increases largely with increasing Wf. It is seen from the figure that the impact value becomes larger for larger Wf. For example, the impact value at Wf=70% is the value of about six times of that of matrix material. The said large impact value may be caused by the large energy absorption of flexible long fibers used here.

j? 40 •

^ 30 u |

20

o

a, 10 B 0 0

20

40

60

80

Weight fraction of waste fiber Wf (%) FIGURE 9 had impact value

CONCLUSIONS The composite materials based on the post consumer type waste fiber such as rag have been molded, and their mechanical properties were investigated. The tensile, bending and impact properties were improved with the increasing of waste fiber content in the range of Wf =£50%. Especially, the fairly larger values of modulus and Izod impact value can be obtained in comparison with those of matrix material. Although minor problem were encountered, the molding method of composite materials described in the present paper showed promise as a contribution towards the recycling of waste fibers such as rag.

Various Lignocellulosics Fibre Reinforced Polyester Composites: The Study on Mechanical, Physical and Biological Properties Abdul Khalil, H. P. S. and Issam, A. M. Bio-Resources, Paper and Coating Technology, School of Industrial Technology, Universiti Sains Malaysia, 11800 Minden, Penang, Malaysia.

ABSTRACT Non-woven lignocellulosic fibres (oil palm empty fruit bunches (EFB), bamboo fibres and pineapple leaf fibres) reinforced thermoset composites were fabricated at different fibre weight fractions, i.e. 10, 20, 30, 40 and 50%. The polyester resin (unsaturated) and MEKP as catalyst were used. The mechanical, physical and biological properties of composites were analyzed with and without fibre treatments (control untreated, acetylation, prepreg ABS and silane) in composites. In general, as increased the weight fraction in matrix, the composite properties increased. The mechanical properties followed the order: bamboo (highest) > pineapple > EFB (lowest). Fibres treatments with ABS exhibited the highest mechanical strength and lowest water absorption of composites followed by acetic anhydride, silane and untreated fibres composites. However, the value was still lower as compared to the glass fibre composites. INTRODUCTION Lignocellulosics or plant fibres have been used by mankind as structural materials since prehistoric times. More recently, interest in the use of materials derived from natural resources has increased dramatically. The use of lignocellulosic fibres (oil palm empty fruit bunches (EFB), bamboo fibres, pineapple leaf fibres etc) as a replacement for glass in the reinforcement of composites is currently generating much interest in the research community. Lignocellulosic fibres offer a number of advantages over glass in such applications because of low costs [1,2], a very high performance/weight ratio [2], light weight [3], easy processing [4], reactive surface chemistry [5] and the fact that they maybe burned at the end of their product life-cycle [6]. Since lignocellulosic fibres are hydrophilic surfaces and polyester matrix is generally hydrophobic, poor fibre-matrix dispersion and wetting of the fibres occur. This incompatibility leads to poor mechanical properties and higher water absorption. The uses of lignocellulosic fibre with or without [5-7] chemical modification and the coupling agent's treatments [7] for polymer composites application have been increasingly studied in recent times. In this study, non-woven lignocellulosics fibres such EFB, bamboo and pineapple leaf fibres reinforced thermoset composites were fabricated using unsaturated polyester resin as a matrix. The mechanical, physical and * Corresponding Author, Email: Email: akhalil(%usm.mv

16

Lignocellulosics Fibre Reinforced Polyester Composites

biological properties were evaluated with and without fibre treatments. The properties of the composites were then compared with glass fibre composites. EXPERIMENTAL Fibre Preparation and Treatments The empty fruit bunches fibres was supplied by Sabutek Ltd. (Malaysia); bamboo fibres and pineapple leaf fibres were supplied by Malaysian Palm Oil Board. Non woven fibre mats of various fibres were prepared using wet process. The fibres were dried in oven at 105°C overnight. The fibre was the extracted before fibre treatments. Silane [7], acetylation [8] and ABS treatments [9] were carried out using method as explained in detailed in the previous study. Formation of Composites The unsaturated polyester resin, type W-1905, used was a commercial product supplied by Euro-Pharma Sdn Bhd, Malaysia. The formulation used consists of 100 parts of resin by weight for 1.5 parts catalyst (methyl ethyl ketone peroxide). Extracted fibres (without modification) and modified fibre (with acetic, silane and ABS prepreg) mats were used to make laminates of non-woven hybrid (random) polyester composites. Composites of varying fibre weight fraction (Wf) of 10%, 20%, 30%, 40% and 50% were prepared using a Resin Transfer Moulding (RTM) machine as explained in detailed in the previous study [10]. Mechanical, Biological and Water Absorption Testing Flexural and impact properties were studied in the test. Flexural tests were carried out on an Instron model 5582 according to ASTM D 790 respectively. For water absorption test, the samples were cut to a size of 5 x 5 x 0.6 cm and immersed in deionised water at ambient temperature. Mechanical and water absorption testing were carried out according to method as explained in detailed in the previous study [10]. Biological Testing The test was performed for 12 months, following the BS standard EN ISO 846:1997. The samples were completely buried in natural soil at 90% water holding capacity (WHC) and 50% soil moisture content. The test was carried out according to the method explained in detailed in previous study [10]. RESULTS AND DISCUSSION Mechanical Properties The results for flexural strength and modulus tests were presented in Table I. It showed that, flexural strength (FS) and flexural modulus (FM) for all unmodified and modified fibres increased steadily as the percentage of fibre increased from 10% to 50%. In general, the flexural properties of treated and untreated fibres composites followed in the order; EFB (lowest) > pineapple leaf > bamboo (highest). As compared to various treatment, fibres treated with ABS showed higher flexural

Lignocellulosics Fibre Reinforced Polyester Composites

17

properties, followed by acetylated and silane treated fibres composites, as compared to unmodified fibres. TABLE I Flexural strength and modulus for different fibre reinforced polyester composites Silane

ABS

Unmodified

Flexural strength (MPa) 45.6 45.6 40.0 37.3 40.0 37.3 43.4 46.6 48.1 51.6 50.2 53.9

45.6 45.5 45.5 53.0 58.7 61.3

3220 3457 3520 3621 3876 3942

45.6 48.6 51.9 53.8 54.5 58.5

45.6 61.8 66.0 68.5 63.3 74.4

3220 3445 3756 3926 4220 4286

45.6 48.8 51.3 54.2 55.0 59.9

45.6 65.2 68.4 72.4 73.3 79.8

3220 3720 4025 4511 4620 4482

Fibre

Fibre/ matrix (Wf)%

Unmodified

EFB

0 10 20 30 40 50

45.6 34.5 34.5 40.2 44.5 46.5

0 10 20 30 40 50

45.6 44.2 47.2 48.9 49.5 53.2

45.6 53.0 56.6 58.7 59.4 63.8

0 10 20 30 40 50

45.6 43.4 45.6 48.2 48.9 53.2

45.6 54.3 57.0 60.3 61.1 66.5

Pineapple Leaf

Acetylated

Acetylated

Silane

ABS

(MPa) 3220 3664 3731 3838 4108 4179

3220 4287 4364 4489 4806 4888

3220 3962 4319 4515 4853 4929

3220 3703 4038 4220 4536 4607

3220 4479 4882 5104 5486 5572

3220 4278 4629 5188 5313 5154

3220 4092 4428 4962 5082 4430

3220 4836 5233 5865 6006 5826

Flexural modulus 3220 3872 3942 4055 4341 4415

Bamboo

Table II showed the results of impact strength for all types of fibres at different fibre treatments. It showed that impact strength of all unmodified and modified composites increased gradually as the fibre loading increased. For unmodified composites, EFB showed the highest value of impact strength at fibre loading more than 40%. Same phenomenon was observed with all treated fibres composites. TABLE II Impact strength for different fibre reinforced polyester composites Fibre

Fibre/matrix (Wf) (%)

Unmodified

Acetylated

Impact strength (kJm"2) 5.5 5.5 6.0 6.6 9.4 8.5 10.2 11.2 14.8 16.3 15.2 16.7

Silane

ABS

5.5 6.3 8.9 10.7 15.5 15.9

5.5 7.2 10.3 12.2 17.8 18.2

EFB

0 10 20 30 40 50

Pineapple leaf

0 10 20 30 40 50

5.5 9.2 10.8 11.7 13.2 13,0

5.5 10.7 12.5 13.6 15.3 15.7

5.5 9.9 11.6 12.6 14.3 14.7

5.5 12.2 14.2 15.5 17.4 17.8

Bamboo

0 10 20

5.5 8.2 9.1

5.5 9.8 10.92

5.5 9.0 10.0

5.5 11.4 12.7

18

Lignocellulosics Fibre Reinforced Polyester Composites 30 40 50

12.24 14.16 15.84

10.2 11.8 13.2

11.2 12.9 14.5

14.2 16.6 18.4

Water Absorption Results for water absorption showed at Figure 1-3. The results showed that after 20 days exposure in water, the rate of water uptake for all types of composites remain almost constant except for the unmodified composites. Comparison with different types of fibres, bamboo showed the highest moisture absorption, followed by pineapple leaf and EFB fibres. This was due to the hydrophilic nature of the lignocellulosic, as well as due to the higher cellulose content in bamboo. -•-Unmodified -•-Acetylated -±-Silane -S-S-ABS HI-Cast Resin -•-CSM

Time {days)

Time (days)

FIGURE 1: Water absorption of bamboo reinforced composites

FIGURE 2: Water absorption of pineapple leaf reinforced composites

Time (Days)

FIGURE 3: Water absorption of EFB reinforced composites

Biological Properties The effect of fibre treatment upon the resistance of fibres to microbiological attack was showed in Figure 4-12. Flexural properties of the composites deteriorated when samples were exposed in soil burial tests. Whereas, no significant changes in properties were observed when unreinforced resins samples were exposed, all of the composites exhibited deterioration to varying extent. Both ABS and silane treatment of fibres provided a significant degree of protection.

Lignocellulosics Fibre Reinforced Polyester Composites H 0 month

Unmodified Aealylated

Sllana

ABS

Cast resin

n 3 rnonth

Unmodified Acotylated

CSM

19 E3 6 month

SI la no

9 12 n

ABS

Cast resin

CSM

Types of treatment

Types of treatment

FIGURE 4: Effect of biological test on flexural strength of EFB reinforced composites

FIGURE 5: Effect of biological test on flexural strength of pineapple leaf reinforced composites

• 0 month G3 month B6 month • 12 month!

• 0 month Q 3 month E 6 month a 12 month

Unmodified Acotylated Types of treatment

Types of treatment

FIGURE 6: Effect of biological test on flexural strength of bamboo reinforced composites

Flexural mo dulus

g 5

=

• 3 month

i|

Unmodified Acetylalsd

B 6 month

iTTTT

• 0 month

1 Silane

FIGURE 7: Effect of biological test on flexural modulus of EFB reinforced composites

B12 month

E

["•Omil onth

a 3 month

B 6 month

B12 month

• . =

II 1

ABS

Cast resin

CSM

Unmodified Acstylatad

Types of treatment

Silane

ABS

Cast resin

CSM

Types of treatment

FIGURE 8: Effect of biological test on flexural mod. of pineapple leaf reinforced composites

FIGURE 9: Effect of biological test on flexural modulus of bamboo reinforced composites

Figure 10-12 showed the effect of biological tests on impact properties of the composites. All figure showed similar phenomenon that CSM composites exhibited the highest impact strength which was more than 50 kJm2.

-•-Asetilasl -•-Sllana -*-ABS -K-Castrasin -«-CSM 52

FIGURE 10: Effect of biological test on impact strength of EFB reinforced composites

1

«

FIGURE 11: Effect of biological test on impact strength of pineapple leaf reinforced composites

20

Lignocellulosics Fibre Reinforced Polyester Composites

Masa(BuIan)

FIGURE 12: Effect of biological test on impact strength of bamboo reinforced composites

CONCLUSIONS The conclusions from above study were summarized as followed: • In general, mechanical properties of unmodified and modified composites increased gradually as the weight fraction of the composites increased from 10%-50% • Mechanical properties of the unmodified and modified (acetylated, silane, ABS) composites followed the order: bamboo (highest) > pineapple leaf > EFB (lowest) • According to types of treatment, mechanical properties of the composites followed the order: ABS pre-preg (highest) > acetylated > silane > unmodified (lowest) • For all types of composites and fibre treatments, the rate of water absorption decreased in the order: unmodified (highest) > silane > acetylated > ABS > CSM > cast resin (lowest). • The results of biological studies showed that changes in impact and flexural strength for pineapple leaf, bamboo and EFB reinforced composites were as followed: CSM (highest) > ABS > acetylated > silane > unmodified > cast resin (lowest). ACKNOLEDGMENTS The authors would like to thank to Universiti Sains Malaysia, Penang, Yayasan FELDA Grant No. P/TEKIND/650214/Y 104 and Kementerian Sains, Teknologi dan Alam Sekitar (MOSTE) Project Number: 09-02-05-1071 RM8 EA001 that has made this work possible. REFERENCES 1. 2. 3.

Bolton, A. J. 1994. "Natural Fibres For Plastic Reinforcement," Mater Tech., 9:12 Sreekala, M. S., Thomas, S. and Neelakantan, N. R. 1997. Journal Polym Engin., 16:265 Matsuda, H. 1993. "Preparation and Utilisation of Esterified Woods Bearing Caboxyl Groups," Wood Sci Tech., 27:23 4. Bisanda, E. T. N. and Ansell, M. P. 1991. Comp Sci Tech., 41:165 5. Abdul Khalil, H. P. S. and Ismail, H. Polym Test., 9:42-56 6. Gassan, J. and Bledzki, A. K. 1996. "Modification Method on Natural Fibres and Their Influence on the Properties of the Composites," ANTEC, 2:225 7. Rozman, H. D., Abdul Khalil, H. P. S., Kumar, R. N., Abusamah, A and Kon, B. K. 1995. Int J Polym Mater., 32:247 8. Hill, C. A. S., Abdul Khalil, H. P. S and Hale, M. D. 1998. "A Study of The Potential of Acetylation to Improve The Properties of Plant Fibre," Industrial Crops and Products., 8:53 9. Abdul Khalil, H. P. S., Maulida and Nasir, M. 2000. Mechanical and Water Absorption Properties of Lignocellulosic-Based Hybrid Composites. In: The 3 rd Regional IMT-GT Conference, Medan, Indonesia: 101-110 10. Abdul Khalil, H. P. S., Rozman, H. D., Ahmad, M. N. and Ismail, H. 2000. "Acetylated Plant Fibre Reinforced Composites: A Study On Mechanical, Hygrothermal and Ageing Characteristic," Polymer Plastics Technology Engineering., 39:757

Development of Eco-Cement Containing High Volumes of Waste Glass Konstantin Sobolev* Division de Estudios de Postgrado, Facultad de Ingenieria Civil Universidad Autonoma de Nuevo Leon, Mexico Gunsel Iscioglu BEM Cement, TRNC Pelin Tilrker, Asim Yeginobali, Tomris Ertiin R&D Institute, Turkish Cement Manufacturers' Association, Turkey

ABSTRACT Waste glass is a serious environmental problem in many countries, mainly because of the inconsistency of the waste glass streams. Consequently the ability of glass industry to recycle waste glass is limited. Therefore, alternative technologies are needed to boost the recycling of waste glass beyond the present restraints of the glass industry. The application of waste glass as a finely ground mineral additive (FGMA) in cement is one of the promising directions for the recycling of waste glass. Based on the method of mechano-chemical activation, a new group of ECO- cements was developed. In ECO- cement, relatively large amounts (up to 70%) of portland cement clinker can be replaced with waste glass or another locally available mineral additives. This report examines the effect of waste glass materials (window glass, black-andwhite monitor glass, brown and green bottle glass) on the micro structure and strength of ECO-cement based materials. According to the research results, the developed ECO-cement (with 50% of waste glass) possessed compressive strength at a level similar to normal portland cement, in the range of 44.5 - 66.7 MPa. Best compressive strength values were demonstrated by the ECO-cement based on waste window and green bottle glass. SEM observations detected a visible densification around the glass grains, due to partial hydration of glass grains and formation of C-S-H. INTRODUCTION Glass, one of the earliest man-made materials, has been known for over 9000 years. Because of its availability, cost effectiveness and unique mechanical, chemical, thermal, and optical properties, glass has many useful applications. In the USA alone about 20 million tons of glass products are manufactured annually with a shipment value of about $29 billion [1].

* Corresponding Author, A.P.#17, Ciudad Universitaria, San Nicolas de los Garza, Nuevo Leon, Mexico, 66450; Fax: (+l)-925-663-0491; e-mail: [email protected]

22

Development of Eco-Cement

Theoretically, glass is a 100% recyclable material and it can be indefinitely recycled without any loss of quality. According to EPA official statistics [2-3], the municipal solid waste (MSW) stream in the USA contains about 5.5% of waste glass which yields 12.8 million tons per year. In 2000, only 23% of this volume was recycled [2-6]. Therefore, in spite of the apparent simplicity of glass recovery, its recycling rate is still insufficient (at an average MSW recovery level of 30%) [3]. Waste glass comes from different sources: glass containers (bottles and jars), construction glass (windows), and electrical equipment (lamps, monitors and TVs). Most (89%) of the waste glass comes from various containers [2-3]. Generally, recovered glass containers are recycled into new glass containers, others are used in newly emerging sectors such as fiberglass insulation, abrasives, light-weight aggregates, yet others are used for concrete and asphalt [2-25]. When waste glass is proposed as a constituent of cement (as a mineral additive) and concrete (as aggregate), concern about the strength reduction and potentially deleterious alkali-silica reaction (ASR) is often raised [18-25]. The usual precautions to avoid ASR (such as the application of low alkali cement and pozzolanic additives) were found to be effective when waste glass was used in concrete [22-25]. The application of waste glass as a finely ground mineral additive (FGMA) in cement is another promising direction for waste glass recycling [20, 26]. According to [20], FGMA glass with its high surface area, participates in the relatively quick pozzolanic reactions that eliminates the danger of a slower alkali-silica reaction at a later stage. It was demonstrated that the technology of High Performance (HP) Cement can be used for engineering ECO-cement with a high volume of mineral additives (HVMA) [26, 27]. Supersilica, a reactive silica-based complex admixture is added during the cement grinding process, promotes the mechano-chemical activation of cement and imparts high strength and extreme durability to the concrete or mortar made from such cement [27]. hi ECO-cement, relatively large amounts (up to 70%) of portland cement clinker can be replaced with inexpensive locally available mineral additives, including waste glass. It was proposed that the complex admixture and FGMA glass containing significant amounts of amorphous, reactive silicon dioxide can participate in simultaneous pozzolanic reactions [27]. RESEARCH SIGNIFICANCE The effect of FGMA glass (waste sital glass) on the strength behavior of HP cement- based materials has been reported in the literature [26], At the same time the recycling of different types of waste glass in ECO-cement was proposed [27]. Consequently, the evaluation of the effect of various groups of waste glass on the properties and microstructure of ECO-cement is important for the development and realization of this alternative way of waste glass recycling. EXPERIMENTAL PROGRAM Materials Used Four different waste glass materials (in a form of glass cullet) were used in the research: window glass (WG), black-and-white monitor glass (MG), brown and green bottle glass (BBG and GBG, respectively). The reference cement was portland cement CEM-I 42.5 (NPC) [28]. The ASTM Type I [29] clinker and reactive silica- based complex admixture Supersilica were used for the preparation of waste glass ECO-

Development of Eco-Cement

23

cement samples. The chemical composition of these materials was analyzed using XRD technique (Table I). It is noticeable that no lead was detected in the MG sample. TABLE I Chemical Analysis of Cement Components Composition SiO 2 A12O3 Fe 2 O 3 TiO 2

CaO MgO Na 2 O

K2O SO3 Cr 2 O 3 Loss of Ignition

Clinker

Window Glass

Monitor Glass

20.84 5.52 3.61 0.29 65.57 2.13 0.82 0.19 0.91 0.03 0.23

71.71 1.26 0.09 0.07 8.44 4.16 13.61 0.40 0.25 -

83.96 2.03 0.04 0.20 0.37 0.01 7.98 5.35 0.05 -

Bottle Glass Brown Green 71.19 2.38 0.29 0.15 10.38 1.70 13.16 0.70 0.04 -

71.12 1.71 0.24 0.07 10.02 3.01 13.17 0.19 0.25 0.23 -

Notations Used The following notations were used to distinguish the ECO-cement samples: WGC - for window glass (WG) cement; MGC - for monitor glass (MG) cement; BBGC - for brown bottle glass (BBG) cement; GBGC - for green bottle glass (GBG) cement. Mixture Proportioning The strength properties of five cement samples were investigated. These included cements based on high performance cement technology with different types of waste glass (WG, MG, BBG and GBG) and reference cement. 50% (by weight) of waste glass and 10% complex admixture were used to produce the waste glass containing cements. The composition and properties of investigated cements are presented in Table II. The mortars for compressive strength tests were prepared according to ASTM C109 [29]. Sand-to-cement ratio (S/C) of 2.75 was used for all mortars. These mortars were produced at a reduced water /cement ratio (W/C) adjusted to obtain a flow range of 140-190 mm. For SEM investigations, cement pastes of window glass (WG) cement with W/C of 0.5 were used. Preparation of Specimens Clinker was pre-ground in a ball mill for 60 minutes for subsequent use in the research program. Waste glass samples were washed to remove organic contaminants and crushed in the laboratory jaw crusher to a maximum size of 4 mm. Samples of glass cement were obtained by grinding a mixture composed of 35% pre-ground clinker, 5% gypsum, 10% complex admixture and 50% waste glass. The sample weight was 5 kg and the grinding media weight was 65 kg. Grinding time for all cement samples was 60 minutes. The investigated mortars were mixed following EN 196 [30]. The mortars were cast into three-gang (40x40x160 mm) prism molds, and compacted in accordance

24

Development of Eco-Cement

with EN 196 [30]. After the compaction procedure, the molds were placed in a humidity cabinet for 24 hours (keeping a relative humidity of 95% and a temperature of 20±l°C). Following this period, the specimens were removed from the molds and kept in water until the testing age. For SEM investigations, the cement pastes were prepared by the same mixing and molding method. Then, these pastes were cured for 28-day at 20+1 °C. After the curing period, the fractured specimens were dried, treated with acetone and coated with a thin layer of gold. Tests Performed The mortar samples were tested at the age of 2, 7 and 28 days for compression. Compressive strength tests were conducted using the portions of prisms broken in flexure [31]. The compressive strength results indicated are the average of the four values. SEM observations were performed using LEO Scanning Electron Microscope operated at an accelerating voltage of 15 kV. TEST RESULTS Compressive Strength of Mortars The compressive test results of glass cement mortars (following ASTM) are given in Table II. According to the test results, the best 28-day compressive strength value of 50.1 MPa was obtained from cement produced with window glass. The monitor glass, brown bottle and green bottle glass based cements reached a 28-day compressive strength in the range of 44.5 - 46.0 MPa, which is close to the strength of reference NPC (45.4 MPa). The compressive strength values at the 2-day age are almost the same for the investigated group of glass cements. Glass cement based on WG demonstrated the best compressive strength at the early ages. Similar behavior of waste glass ECOcements was observed at the 7- day age (Table II). Delay in the strength development of waste glass ECO-cements at early age: e.g. 50% at 2 days and 26% at 7 days can be explained by very low clinker content (35%) in these cements. Pozzolanic reaction of glass, as well as low W/C helps to offset this trend at later stages of hardening. TABLE II Compressive Strength of Investigated Mortars Composition Type NPC WGC MGC BBGC GBGC

Clinker

Gypsum

Glass

95 35 35 35 35

5 5 5 5 5

0 50 50 50 50

Compressive Strength, MPa Super Silica 0 10 10 10 10

W/C 0.45 0.30 0.30 0.30 0.30

2 days

7 days

28 days

26.5 16.4 12.1 13.2 11.6

36.1 31.0 25.0 25.0 25.5

45.4 50.1 44.5 45.3 46.0

SEM Observations It was observed that glass particles are well dispersed in the paste, resulting in a dense structure with low porosity (Figure 1). It is also noted that the glass grains are well connected to matrix and are coated with a thin layer of reticulated gel of C-S-H (Figure 1). Further, it was found that there is a visible densification around the glass

Development of Eco-Cement

25

grains, possibly due to partial hydration of glass grains, leading to an additional formation of C-S-H. SEM detected no sign of ASR in the investigated samples.

FIGURE 1 Microstructure of ECO Cement

CONCLUSIONS The developed ECO-cement containing 50% of waste glass possessed compressive strength properties at a level similar to normal portland cement, in the range of 44.5- 66.7 MPa. Best compressive strength values were demonstrated by the ECO-cement based on waste window and green bottle glass. It should be pointed out that the low-water demand property of ECO-cements results in high workability at low W/C. It helps to improve the strength of mortars based on these cements and also to offset the use of mineral additives in the cement composition. SEM observations detected a visible densification around the glass grains, due to partial hydration of glass grains and an additional formation of C-S-H. No sign of ASR was found in the investigated samples of waste glass cement. The research leads to the conclusion that the application of HP cement technology helps to recycle waste glass in ECO cement. Additional investigations may be necessary to improve the chemical activity of waste glass in the cement system. Further research is also required to explain and quantify the hydration mechanism and the microstructural development of ECO cement containing large volumes of waste glass, as well as to examine their resistance to a number of detrimental factors including the possible adverse effects of the alkali-silica reaction. ACKNOWLEDGEMENT The authors would like to acknowledge the receipt of samples of Supersilica from SCI Con Technologies. The author is grateful to the staff of Quality Control Laboratory of BEM Cement, Cement and Concrete Research Institute of TCMA, R&D Division of SISECAM Holding and Environmental & Educational Research Center of FAST at LAU for their help in conducting the experiments. The suggestions of Dr. U. Kersting and Dr. C. Podmore are highly appreciated. REFERENCES 1.

U.S. Department of Energy. 2002. "Glass Industry of the Future - Energy and Environmental Profile of the U.S. Glass Industry", pp. 1-23.

26 2. 3. 4. 5. 6. 7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19.

20. 21. 22. 23. 24.

25. 26. 27. 28. 29. 30.

31. 32.

Development of Eco-Cement U.S. Environmental Protection Agency. 1992. "Markets for Recovered Glass", pp. 1-15. (EPA530-SW-90-071A) U.S. Environmental Protection Agency. 2002. "Characterization of Municipal Solid Waste in the United States: 2000 Update", pp. 7-41. Geiger, G. 1994. "Environmental and Energy Issues in the Glass Industry", American Ceramic Society Bulletin 73(2), pp. 32-37. Stewart, G. 1986. "Cullet and Glass Container Manufacture", Resource Recycling, 2. "Americans Continue to Recycle More Than One in Three Glass Containers", October 27, 1999, Glass Packaging Institute, www.gpi.org/98rate.htm Apotheker, S. 1989. "Glass Processing: the Link between Collection and Manufacture", Resource Recycling, 7, p. 38. Rodriguez, D. 1995. "Application of Differential Grinding for Fine Cullet Production and Contaminant Removal", Ceramic Engineering Science Procedures, 2(16), pp. 96-100. Mayer, P. 2000. "Technology Meets the Challenge of Cullet Processing", Glass Industry, 2. Guter, E. 1996. "Quality Cullet Is Required for Fiberglass", Glass Industry, 1, pp. 13-35. Pascoe, R.D., Barley, R.W. and Child, P.R. 2001. "Autogenous Grinding of Glass Cullet in a Stirred Mill", in Recycling and Reuse of Glass Cullet, R.K. Dhir, M.C. Limbachiya and T.D. Dyer, eds. London: Thomas Telford, pp. 15-29. Clean Washington Center. 1995. "Evaluation of Crushed Recycled Glass as a Filtration Medium in Slow Rate Sand Filtration". Clean Washington Center. 1994. "Glass Feedstock Evaluation Project: Engineering Suitability Evaluation. Evaluation of Cullet as a Construction Aggregate". (GL-93-3) Day, D.E., and Schaffer, R., Glasphalt Paving Handbook, University of Missouri-Rolla, p. 53. Malisch, W.R., Day, D.E., and Wixson, B.G. 1975. "Use of Domestic Waste Glass for Urban Paving: Summary Report", U.S. Environmental Protection Agency. (EPA-670/2-75-053) "Glasphalt May Pave the Way for Worldwide Aviation in the 21 st Century", March 1, 2003, http://www.sciencedaily.eom/releases/1997/l 1/971110064723.htm Nash, P.T., Jayawickrama P., et al. 1995. "Use of Glass Cullet in Roadway Construction", FHWATCEQ. (0-1331) Shin, C.J., and Sonntag, V. 1994. "Using recycled glass as construction aggregate", Transportation Research Board, National Research Council. (No. 1437) Shao, Y., Lefort, T., Moras, S., and Rodriguez, D. 1998. "Waste Glass: A Possible Pozzolanic Material for Concrete", International Symposium on Sustainable Development of the Cement and Concrete Industry, CANMET/ACI, Ottawa, pp. 317-326. Dyer, T.D., and Dhir, R.K. 2001. "Chemical Reactions of Glass Cullet Used as Cement Component", Journal of Materials in Civil Engineering, 13(6), pp. 412-417. Naik, T.R., and Kraus, R.N. 1999. "Use of Glass Cullet as Aggregates in Flowable Concrete with Fly Ash", CBU. (CBU-1999-03) Meyer C. 2003. "Glass Concrete", Concrete International, Vol. 25, No. 6, pp. 55-58. National Research Council - Strategic Highway Research Program. 1993. "Eliminating or Minimizing Alkali-Silica Reactivity". (SHRP-C-343) Xie, Z., Xiang, W., and Xi, Y. 2003. "ASR Potentials of Glass Aggregates in Water-Glass Activated Fly Ash and Portland Cement Mortars", Journal of Materials in Civil Engineering, (15)1, pp. 67-74. "Making Concrete with Glass - Now Possible", May 13, 2002, Ref: 2002/90, http://www.dbce.csiro.au/news/viewpress.cfm/109 Sobolev, K., and Arikan, M. 2002. "High Volume Mineral Additive ECO- Cement", American Ceramic Society Bulletin, 81(1), pp. 39-43. Sobolev, K. 1999. "High Performance Cement: Solution for Next Millennium", Materials Technology, 14(4), pp. 191-193. American Society for Testing and Materials. 1999. "Standard Specification for Portland Cement", Annual Book of ASTM Standards, ASTM C150. European Committee for Standardization. 1994. "European Standard Specification for Cement", European Standard, EN 197-1. American Society for Testing and Materials. 1999. "Compressive Strength of Hydraulic Cement Mortars (Using Portions of Prisms Broken in Flexure)", Annual Book of ASTM Standards, ASTM C349. American Society for Testing and Materials. 1999. "Compressive Strength of Hydraulic Cement Mortars Using 2-in or 50-mm Cube Specimens", Annual Book of ASTM Standards, ASTM C109. European Committee for Standardization. 1994. "Test Method for Determining Compressive Strength of Cement Mortar", European Standard, EN 196.

Screwless Extrusion of Natural Fibre-Reinforced Thermoplastic Composites Timothy Galea, Tony Mills, Rex Halliwell and Krishnan Jayaraman Centre for Advanced Composite Materials, Department of Mechanical Engineering, The University of Auckland, Private Bag 92019, Auckland, New Zealand.

ABSTRACT Natural fibres possess good reinforcing capability when properly compounded with polymers. These fibres are relatively inexpensive, originate from renewable resources and possess favourable values of specific strength and specific modulus. Thermoplastic polymers possess shorter manufacturing cycle times and reprocessability despite problems with high viscosities and poor fibre wetting. The renewability of natural fibres and the recyclability of thermoplastic polymers provide an attractive eco-friendly quality to the resulting natural fibre-reinforced thermoplastic composite materials. Common methods for manufacturing natural fibre-reinforced thermoplastic composites, injection moulding and extrusion, often require compounding of the constituents in mixers leading to degradation of the fibres. Development of a screwless extruder, that minimises fibre degradation and employs a reliable and low technology process for compounding the constituents, is the main objective of this study.

INTRODUCTION Natural fibres come from renewable resources and are relatively inexpensive. These fibres are now well recognised to impart good reinforcing capability to composites. While their tensile strengths and moduli are generally inferior to those of polymeric fibres, they often exhibit significantly larger elongation giving them better damage tolerance [1,2]. A relatively large body of published literature [3-9] in the area of wood fibrereinforced virgin thermoplastic composites exists. These studies have examined the mechanical properties of the composites and the effects of various coupling agents on the interfacial bonding between the fibres and the polymer. The presence of a suitable coupling agent has been shown to be important for the achievement of significant gains in the mechanical properties of these composites in a recent review [10]. Common methods for manufacturing natural fibre-reinforced thermoplastic composites are injection moulding and extrusion. These techniques often require compounding of the constituents in mixers leading to the degradation of natural fibres. Hence, there is a need for a simple and reliable screwless extruder that minimises fibre degradation during compounding of the constituents.

* Corresponding Author: Department of Mechanical Engineering, The University of Auckland, Private Bag 92019, Auckland, New Zealand; Fax: + 64 9 373 7479; E-mail: [email protected]

28

Screwless Extrusion of Natural Fibre-Reinforced Thermoplastic Composites

A screwless extrusion technique for plastication and extrusion of thermoplastics, called elastic melt extrusion, has been developed by Maxwell and Scalora [11] based on the normal force effect. When a visco-elastic material is sheared between a stationary plate and a rotating plate, a normal force will be generated which will tend to push the two plates apart. Therefore, if an orifice is made in one of the plates, and both plates are constrained axially, the viscous material will be extruded though this orifice due to the normal force effect. The elastic melt extrusion technique results in much shorter material residence times than typical, and has a complete lack of contact between moving parts. It also subjects the material to a high degree of mixing compared with screw extrusion, the most common short fibre composite compounding method [12]. Since natural fibres will not be damaged through abrasion with moving parts, and shorter exposure times should reduce degradation due to high temperatures, it seems that the elastic melt extrusion process should result in a better fibre quality than screw extrusion. The purpose of this investigation is to assess the feasibility of the elastic melt extrusion process for compounding woodfibre-high density polyethylene and woodfibre-polypropylene composites. MATERIALS The woodfibres used in this study were Pinus Radiata fibres supplied by the New Zealand Forest Research Institute in Rotorua. They were light brown in colour with widths varying from 15 urn to 40 urn, lengths ranging from 1.5 mm to 5 mm, density of 400 kg/m3 and a nominal tensile strength and stiffness of 125-150 MPa and 2.5-4 GPa, respectively [13]. It should be noted that these values are very dependent on the source of woodfibres and their fibril angles. The high density polyethylene (HDPE) used was obtained from recycled milk bottles shredded into flake form. The flakes were roughly circular with a diameter of 5-10 mm. There were remnants of coloured bottle components present in the HDPE flakes, which caused the HDPE extrudate to be of a green colour rather than the clear extrudate that would occur from extrusion quality HDPE. The Polypropylene (PP) used was Cotene grade JE6100 in granular form with a diameter of 1-2 mm. The PP was reasonably free of impurities and a clear transparent extrudate was achieved. The requisite mass of woodfibres was put into an open top container followed by the required mass of HDPE or PP to achieve specific fibre mass fractions. The fibres and the polymer were mixed in the container by hand and efforts were made to separate the clumps of fibres so that the mixture would be as homogeneous as possible. ELASTIC MELT EXTRUDER The elastic melt extruder, shown in Figure 1, is powered by a 3-phase motor with variable speed control. The motor runs the internal drum of the extruder, which is in the shape of a cone, through an elliptical reduction gearbox (6.41:1). The machine is heated by four heating coils mounted in the front or orifice plate. These coils are controlled by a thermocouple attached to the exterior of the orifice plate and a variable temperature controller. The fibre-polymer mixture, dropped into the feed tube, is forced into the shearing zone between the front face of the cone and the front plate. During initial operation, heat is introduced to the conical section by conduction between the front face of the cone and the front plate through the molten material. Once the machine is running, most of the heat is provided through shear of the molten material lying between

Screwless Extrusion of Natural Fibre-Reinforced Thermoplastic Composites

29

the cone and the outer walls. The rear of the drum is thermally insulated from the front by a heat resistant gasket.

Polymer feed tube Shaft support bearings

A

Orifice plate ;one/drive assembly Extrudate

FIGURE 1 Schematic diagram of the elastic melt extruder

EXTRUSION OF COMPOSITES Preliminary studies were carried out on the elastic melt extruder using HDPE and PP to determine the effects of the drum speed, gap size (distance between the front face of the cone and the front plate) and the operating temperature on the extrusion rate or the material output from the extruder. The process variables that would provide maximum material output were then selected for the rest of this study, Table 1. The operation and output of the elastic melt extruder was sensitive to feed rates. An excessive rate of material addition created localised clumps of partially melted polymer and fibres in the melt, causing the drum to seize up. If the amount of material in the melt cavity was low, the output was significantly reduced causing material residence time to increase. Further, a decreased rate of material addition can cause the melt temperature to increase by up to 20°C. The feed rate was approximate for this study, Table 1. TABLE I Process variables used in elastic melt extrusion

Variable Drum rotational speed Gap size Temperature Feed rate

Unit rpm mm °C g/s

Value 15-20 10 + 0.5 180-200 0.2+0.1

Once the elastic melt extruder had reached the operatmg temperature, the fibrepolymer mixture was added to the hopper. Roughly a handful of fibre-polymer mixture was added for every 20-30cm of extrudate produced. The fill level was visually gauged by the amount of melt covering the drum. These rough approximations were sufficient for the small volume batches of composites produced in these initial studies, Table 2.

30

Screwless Extrusion of Natural Fibre-Reinforced Thermoplastic Composites TABLE II Materials produced using elastic melt extrusion

Material type Wood-PP Wood-HDPE

Fibre mass fractions as mixed before extrusion 30% 10% 15% 20% 25% 40%

0%

Y V

•

/

•

/

S

•

•

/

m

/

x

The woodfibre-HDPE and woodfibre-PP extradates contained a number of small voids. These voids may have been due to the moisture released by the woodfibres at the high temperatures of extrusion. Microscopic examination revealed that the fibres were aligned in the direction of extrusion. With higher fibre mass fractions, a degree of fibre clumping occurred at irregular intervals. These clumps were approximately 5 mm in diameter and did not have any polymer penetration. PRODUCTION OF COMPOSITE SPECIMENS The individual batches of the extradates were fed piece by piece into the loading hopper of an industrial granulator. Once a batch had been granulated, the collected material was bagged and labelled for injection moulding, and the granulator opened and cleaned to prevent cross contamination of samples. A BOY50M screw injection moulder was used to mould the tensile specimens. Machine variables were set to those typically used in the extrusion of HDPE or PP depending on the matrix of the composite material being moulded. TESTING OF COMPOSITE SPECIMENS The tensile properties of the woodfibre-thermoplastic composite specimens at room temperature and humidity (23°C and 50% RH) were determined in a computercontrolled Instron universal testing machine (Model 5567) using five replicates for each test by following the standard ASTM D 638M - 02. These specimens were conditioned at room temperature and humidity for 24 hours prior to the tests. The tensile strengths of the composites generally increase with increasing fibre content, with significant improvements over the tensile strengths of the base polymers occurring after 15% fibre mass fraction for woodfibre-HDPE composites and 20% fibre mass fraction for woodfibre-PP composites, Figure 2. The tensile moduli of the composite materials show significant improvements over the tensile moduli of the base polymers with increasing fibre content, Figure 3. SUMMARY A screwless extruder has been built based on a similar extruder intended for the extrusion of thermoplastics. Some of the advantages of this extruder include: complete lack of contact between moving parts, strong mixing action, shorter material residence times and the direct addition of natural fibres to the melt. The pressure developed during this extrusion process is low compared to that developed in conventional screw extruders. However, the laboratory prototype that has been built is smaller than the conventional screw extruders. Woodfibre-HDPE and woodfibre-PP composites have been successfully compounded by means of the elastic melt extruder. Mixing and dispersion of the

Screwless Extrusion of Natural Fibre-Reinforced Thermoplastic Composites

31

Wood-PP -

Wood-HDPE PP

—

15

20

25

30

-

HDPE 35

40

Fibre mass fraction (%)

FIGURE 2 Tensile strengths of woodfibre-PP and woodfibre-HDPE composites as a function of fibre content

15

20

25

30

Fibre mass fraction (%) FIGURE 3 Tensile moduli of woodfibre-PP and woodfibre-HDPE composites as a fimction of fibre content

45

32

Screwless Extrusion of Natural Fibre-Reinforced Thermoplastic Composites

constituents have been achieved in a short time due to intensive local shearing of the fibre-polymer mixture. Heat degradation of the fibres has also been minimised due to the heating of a small amount of feed material at any given time. Tensile properties of the composite specimens produced through granulation and injection moulding of the extrudates show reasonable improvements over the tensile properties of the base polymers. REFERENCES 1. 2.

3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

13.

Chand, N., R. K. Tiwary and P. K. Rohtagi. 1988. "Structure - properties of natural cellulosic fibres an annotated bibliography," Journal of Materials Science, 23: 381-387. Groom, L.H., S. M. Shaler and M. Mott. 1995. "The mechanical properties of individual lignocellulosic fibres," in Virgin and Recycled Wood Fiber and Polymers for Composites. Proceedings of the Third Woodfiber-Plastic Composites Conference, 1-3 May, 1995, Madison, USA, pp. 33-40. Woodhams, R.T., Q. Thomas and D. K. Rodgers. 1984. "Woodfibres as reinforcing fillers for polyolefins," Polymer Engineering and Science, 24: 1166-1171. Beshay, A.D., B. V. Kokta and C. Daneault. 1985. "Use of wood fibers in thermoplastic composites II: Polyethylene," Polymer Composites, 6(4): 261-271. Zadorecki, P. and A. J. Mitchell. 1989. "Future prospects for wood cellulose as reinforcement in organic polymer composites," Polymer Composites, 10(2): 69-77. Bataille, P., L. Ricard and S. Sapieha. 1989. "Effects of cellulose fibres in Polypropylene composites," Polymer Composites, 10(2): 103-108. Balatinecz, J J. and R. T. Woodhams. 1993. "Wood-plastic composites: Doing more with less," Journal of Forestry, 11: 22-26. Sain, M.M., B. V. Kokta and C. Imbert. 1994. "Structure-property relationships of wood fiber-filled Polypropylene composite," Polymer-Plastic Technology and Engineering, 33(1): 89-104. Bhattacharyya, D., M. Bowis and K. Jayaraman. 2003. "Thermoforming of woodfibre-polypropylene composite sheets," Composites Science and Technology, 63: 353-365. Lu, J.Z., Q. Wu and H. S. McNabb. 2000. "Chemical coupling in wood fiber and polymer composites: A review of coupling agents and treatments," Wood and Fiber Science, 32(1): 88-104. Maxwell, B. and A. J. Scalora. 1959. "The elastic melt extruder- works without screw," Plastics Engineering, 10:107-114, 202-210. Bledzki, A.K., V. E. Sperber and O. Faruk. 2002. "Natural and Wood Fibre Reinforcement in polymers," Rapra Review Reports, Volume 13 number 8, Rapra Technology, Shrewsbury, United Kingdom. Bowis, M.E. 1997. Thermoforming Woodfibre-Polypropylene Composite Sheets. PhD thesis, Department of Mechanical Engineering, University of Auckland, Auckland, New Zealand.

Mechanical Properties of "Green" Composites Made from Starch-Based Biodegradable Resin and Bamboo Powder Hitoshi Takagi* Department of Mechanical Engineering, Faculty of Engineering, The University of Tokushima, Japan Ryuki Takura and Shinji Ochi Department of Ecosystem Engineering, Graduate School of Engineering, The University of Tokushima, Japan

ABSTRACT Recently, bamboo has been reevaluated from an environment-friendly viewpoint. The reason for this is that bamboo is typical of the yearly-renewable bioresource and that it inherently has an advantage in its high growing speed. With the increase in the volume of the bamboo used in the near future, it is anticipated that the quantity of bamboo powder discharged in various cutting processes will increase accordingly. It is therefore important to develop an adequate technology utilizing the bamboo powder. For this purpose, we developed totally biodegradable "green" composites with bamboo powder and a starch-based resin. These "green" composites were fabricated by the hot-pressing using at 10 MPa and 130°C for 10 minutes. It is found that the "green" composites fabricated with bamboo powder of 0.5 mm in diameter had an acceptably high flexural strength and high flexural modulus. It is also shown that an alkali treatment applied to the bamboo powder affects the mechanical properties of the "green" composites with an increased flexural strength of about 20 percent. Furthermore, it was seen from biodegradation tests that the "green" composites could be easily biodegrade as they are buried in compost environment.

INTRODUCTION Since general-purpose plastics have several features such as lightweight, durable and, above all, easy to mold, the plastics are used abundantly in commercial goods in our daily life. However, the waste disposal of the used plastic becomes an urgent social problem with increasing both mass production and mass consumption of the plastics. Since general plastic products are non-biodegradable and are chemically stable in the environment, the increases in the volume of waste plastics become one of the biggest sources for a landfill shortage problem. hi these situations, as an approach to solve the waste problem, the use of biodegradable resin products is supposed to reduce the volume of plastic waste in landfills [1]. In addition, the biodegradable resin fully produced from natural crops, such as corn, * Corresponding Author, 2-1 Mmajijosanjima-cho Tokushima 770-8506, Japan, fax: +81-886-656-9082 and e-mail: [email protected]

34

Mechanical Properties of "Green" Composites

sugar beet and potato, has been recognized as a circulating material, because after full biodegradation, the biodegradable resin breaks down into water and carbon dioxide, and then these two substances are absorbed into crops again. Biodegradable plastic products should be rapidly replacing petroleum-based plastics in many industrial applications. The cost of the biodegradable resin, however, is higher than that of general-purpose plastics, and moreover, the biodegradable resins have relatively poor mechanical properties as compared with those of general-purpose plastics. Especially, since there exists no high-strength biodegradable material, the social demand for the development of high-strength biodegradable composite materials is continually expanding. hi recent years, a wide variety of research projects have been performed on the biodegradable composite materials, which are composed of a biodegradable resin and biodegradable reinforcement, such as natural plant fibers, and are often called as "green" composites [2-6]. The use of natural fibers as reinforcement for polymer composites contributes not only to strengthening of matrix but also reducing the total material cost of the composites, since many common plant fibers are much cheaper than biodegradable resins. Since the combination of natural plant fibers and thermoplastics is attractive from an ecological viewpoint, many researches tried to use a wide variety of natural plant fibers; bamboo [6-8], hemp [3], MAO [4], henequen [5], cotton [9], flax [9], and pineapple [10]. Bamboo fiber is recognized as one of the most attractive candidates for strengthening natural fiber [11]. It has been proposed, therefore, that bamboo has several advantages such as (1) the environmental load is small, since it is yearly renewable and its growing speed is fast, thus it is easy to regenerate after cutting, and (2) the bamboo fiber has relatively high strength as compared with other natural fibers such as jute, cotton, etc. The discharge volume of the bamboo powders (saw dust) is expected to increase year by year with increasing the amount of consumption of bamboo fibers and bamboo products. However, no studies have ever tried to utilize the bamboo powders discharged from bamboo factories as ingredient in "green" composites. hi this study, we tried to fabricate "green" composites made from starch-based biodegradable resin and bamboo powders. The mechanical properties of the composites are evaluated using tensile tests and fiexural tests. The effect of alkali treatment of the bamboo powder on the mechanical properties of "green" composites was also investigated. EXPERIMENTAL METHODS Raw Materials A water-dispersive biodegradable resin (Miyoshi Oil & Fat Co., Ltd., CP-300) was used as matrix material. This resin has several features; high-strength, emulsion type with average particle size of 6 jum and non-volatile components of 40% by weight. The bamboo powder is prepared from the saw dust of Moso-bamboo (Phyllostachys pubescens) discharged from a bamboo factory. Bamboo powder less than 0.5 mm in diameter was obtained by using a sieve with a mesh size of 0.5 mm. Alkali Treatment of the Bamboo Powder hi order to increase the adhesion strength of fiber/matrix interface, we applied alkali treatment to bamboo powder. Since it is the surface treatment method for using alkali treatment abundantly even in the inside for naturalfiber,in this study, the bamboo powder was treated with alkali by following procedures. First, aqueous NaOH solution of 5% by

Mechanical Properties of "Green" Composites

35

weight was prepared, and then bamboo powder was put in the alkali solution. The fraction of bamboo powder to the solution is fixed to 10% by weight. After stirring of the mixture at room temperature for 100 minutes, the bamboo powder was taken out, and then rinsed with distilled water. Bamboo powder was dried at 105°C for 2 hours in an oven. Sample Preparation Method The mixture of bamboo powder and biodegradable resin, which are weighed out into a predetermined composition, was well mixed in a beaker of 500 milliliter, then put into a rectangular metal frame, followed by drying at 105°C for 2 hours in an oven. Next, the metallic molds, which have the dimensions of 100x15 mm for flexural test and 100x10 mm for tensile test, were used to hot-press the dried sample. The samples were hot-pressed at 130°C and 10 MPa for 10 minutes. Then, the metallic mold was cooled to room temperature using an electric fan. The density of "green" composites with bamboo powder of 66.6% by weight was 1.18 g/cm3 and the water content was 3% by weight. Mechanical Testing Both tensile tests and flexural tests were carried out using an Instron Mechanical Tester (Model 5567) in order to evaluate the mechanical properties of "green" composites reinforced with bamboo powder. The flexural strength was measured using a three point flexural test, with a span length of 50 mm and a crosshead speed of 1.0 mm / min. Shapes of specimen for flexural test were 100 mm in length, 15 mm in width and 1.5 mm in thickness. The tensile test was also carried out at a crosshead speed of 1.0 mm / min. The specimen configuration for tensile test was 100 mm in length, 10 mm in width, 1.5 mm in thickness, with a gage length of 30 mm. Furthermore, 2 mm thick aluminum tabs were glued at the both ends of the tensile specimen to prevent from damage introduced during fixing. Biodegradation Testing Biodegradation tests were carried out by using a home-use garbage-processing machine (Hitachi, Ltd., BGD-150). After trial operation for several days in order to activate microorganisms, "green" composite sample was placed within the processing media (wood chips), and the biodegradation behavior of the composites was investigated. The sample was put into a nylon baggy net to avoid the difficulty in recovering of degraded sample scattered in the media, and then this net including sample was buried in the compost media. After biodegradation test, the appearance of the sample was examined by using stereoscopic microscope and optical microscope. EXPERIMENTAL RESULTS AND DISCUSSIONS Effect of the Bamboo Powder Content on the Bending Strength A representative flexural stress-deflection diagram for composites with different bamboo powder content is shown in Fig. 1. Data for a neat resin sample is also given in this graph for reference. It can be seen that flexural strength and flexural modulus increase with increasing the bamboo powder content, and that maximum deflection decreases with increasing the content. It is therefore clear that the strengthening of the biodegradable resin is accomplished by the addition of bamboo powder.

36

Mechanical Properties of "Green" Composites

Figure 2 shows the relationship between both flexural strength and flexural modulus and bamboo powder content. It can be seen that both flexural strength and flexural modulus linearly increase with increasing the bamboo powder content. At the content of 66.6% by weight, both flexural strength and flexural modulus show peak values of 36.7 MPa and 4.3 GPa, respectively. These strength values are higher than those of bamboo powder particle boards reported elsewhere (flexural strength of 27.0 MPa and flexural modulus of 4.0 GPa) [12]. In addition, Shibusawa reported that the flexural strength of bamboo particle board produced using small pieces of bamboo was 39 MPa and that this value was higher than the regulated flexural strength of 19.6 MPa in JIS (Japan Industrial Standards) [13]. Thus, it is shown that the "green" composites have also enough strength comparable to the regulation value of JIS. Effect of the Bamboo Powder Content on the Tensile Strength The representative tensile stress-strain diagram for the composite with different bamboo powder content and that for neat resin are show in Fig. 3. A similar dependence of bamboo powder content on mechanical properties has been found in this figure. The relationship between both tensile strength and Young's modulus and bamboo powder content is shown in a Fig. 4. It can be seen that both tensile strength and Young's modulus increase with increasing the bamboo powder content. The "green" composites with bamboo powder content of 66.6 % by weight have a maximum tensile strength of 25.8 MPa and a maximum Young's modulus of 2.9 GPa. Effect of Alkali Treatment The chemical surface treatment of natural fibers is of the great importance and is considered as a low-cost method, therefore a considerable number of studies have been made on alkali treatment [14-16]. Figure 5 shows the result offlexuraltest for the "green" composites fabricated using bamboo powder with and without alkali treatment. It can be seen that alkali-treated composites shows highflexuralstrength showing the improvement about 20% and large maximum deflection. The reason for these increases in mechanical properties is that adhesion strength between bamboo powder and resin increases by alkali treatment of bamboo powder and that the increases in surface roughness derived from alkali treatment enhance the stress transfer between bamboo fiber and resin. The decrease in lignin on the surface of bamboo powder directly contributes to increases in strength, because alkali treatment has an action of removing hydrophobic lignin existing on the surface vicinity of natural fiber as reported by Mohanty et al. [14]. Biodegradation Behavior The results of biodegradation tests shows that sample buried in composting media for seven days macroscopically breaks down into a lot of small pieces. Especially, the superficial resin of the sample is preferentially decomposed, and the surface morphology changes from flat surface to rugged one. Similar experimental results were reported by Luo and Netravali for poly(hydroxybutyrate-co-hydroxyvalerate) resin composted in soil [17]. It is thus concluded from the biodegradation results that the "green" composites are biodegradable. CONCLUSIONS Environment friendly biodegradable "green" composites were made from bamboo powder and a starch-based biodegradable resin. Their tensile andflexuralproperties were

Mechanical Properties of "Green" Composites

37

evaluated, and the effect of alkali treatment on their strength was also investigated. As the result, the "green" composites with bamboo powder of 66.6% by weight has flexural strength of 36.7 MPa and flexural modulus of 4.3 GPa. These mechanical properties exceed the regulation of 18 MPa in Japan Industrial Standards (JIS) for a high strength particleboard. These composites, therefore, have a potential to apply to commercial products. The effectiveness of the alkali treatment for bamboo powder was confirmed experimentally, showing the improvement about 20% in flexural strength in comparison with the strength for the composites fabricated with untreated bamboo powder. REFERENCES 1.

2. 3.

4.

5.

6.

7. 8.

9.

10.

11.

12.

13. 14.

15.

16.

17.

Cutter, C. N., J. L. Willett, and G. R. Siragusa. 2001. "Improved Antimicrobial Activity of Nisin-Incorporated Polymer Films by Formulation Change and Addition of Food Grade Chelator," Letters in Applied Microbiology, 33(4): 325-328. Netravali, A. N. and S. Chabba. 2003. "Composites Get Greener," Materials Today, 6(4): 22-29. Takagi, H. and S. Ochi. 2002. "Characterization of High-Strength 'Green' Composites Using Manila Hemp Fibers and Starch-Based Resin," presented at the Third Japan-Canada Joint Conference on New Applications of Advanced Composites (JCJC-JJI), May 14-15, 2002. Takagi, H., C. W. Cindy, and A. N. Netlavali. 2002. "Tensile Properties of Starch-Based 'Green' Composites Reinforced with Randomly Oriented Discontinuous MAO Fibers," presented at the International Workshop on "Green" Composites, November 19-20, 2002. Chabba, S. and A. N. Netravali. 2002. "Characterization o f Green" Composites Using Henequen Fibers and Modified Soy Protein," presented at the International Workshop on "Green" Composites, November 19-20, 2002. Takagi, H. and R. Takura. 2003. "The Manufacture and Mechanical Properties of Composite Boards Made from Starch-Based Biodegradable Plastic and Bamboo Powder," Zairyo, 52: 357-361 (in Japanese). Chen, X., G. Qipeng and M. Yongli. 1998. "Bamboo Fiber-Reinforced Polypropylene Composites: A Study of the Mechanical Properties," Journal of Applied Polymer Science, 69: 1891-1899. Okubo, K. and T. Fujii. 2002. "Eco-Composites Using Bamboo and Other Natural Fibers and Their Mechanical Properties," presented at the International Workshop on "Green"' Composites, November 19-20, 2002. Jiang, L. and G. Hinrichsen. 1999. "Flax and Cotton Fiber Reinforced Biodegradable Polyester Amide Composites, 1 Manufacture of Composites and Characterization of Their Mechanical Properties," Die Angewandte Makromolekulare Chemie, 268(1): 13-17. Luo, S. and A. N. Netlavali. 1999. "Mechanical and Thermal Properties of Environmental-Friendly 'Green' Composites Made from Pineapple Leaf Fibers and Poly(hydroxybutyrate-co-vallerate) Resin," Polymer Composites, 20(3): 367-378. Ochi, S., H. Takagi, and R. Nflri. 2002. "Mechanical Properties of Heat-Treated Natural Fibers," presented at the International Conference on High Performance Structures and Composites, March 11-13,2002. Fujimoto, Y., Y. Nakashima, J. Kawabe, Y. Mataki, and S. Kumon. 1998. "Manufacturing of Particleboard from Bamboo Particles - Influence of Particle Size on Properties of Bamboo Particleboard," Mokuzaikogyo, 53(5): 212-217 (in Japanese). Shibusawa, T. 1998. "New Material -Bamboo-," Ringyogijutsu, 672: 23-26 (in Japanese). Mohanty, A. K., M. A. Khan, S. Sahoo, and G. Hinrichsen. 2000. "Effect of Chemical Modification on the Performance of Biodegradable Jute Yara-Biopol® Composites," Journal of Materials Science, 35(10): 2589-2595. Ray, D. B. K. Sarkar, and N. R. Base. 2002. "Impact Fatigue Behaviour of Vinyleester Resin Matrix Composites Reinforced with Alkali Treated Jute Fibers," Composites Part A: applied science and manufacturing, 33(2): 233-241. Valadez-Gonzales, A., J. M. Cervantes-Uc, R. Olayo, and P. J. Herrera-Franco. 1999. "Effest of Fiber Surface Treatment on the Fiber-Matrix Bond Strength of Natural Fiber Reinforced Composites," Composites PartB: engineering, 30(3): 309-320. Luo, S. and A. N. Netlavali. 2003. "A Study of Physical and Mechanical Properties of Poly(hydroxybutyrate-co-hydroxyvalerate) during Composting," Polymer Degradation and Stability, 80(1): 59-66.

Mechanical Properties of "Green" Composites

38

IS

BP66.6%-BDP33.3% BP50.0%-BDP50.0% BP33.3%-BDP66.6% BDP(CP-300)

50

1

• Bending strength A Bending modulus

PH

S 40 § 30

}

1 20

i

bo

I 10 5 10 Deflection (mm)

-

•

f o

' .

]

i

I

x

25

1

1

50

75

(BDP:CP-300)

Content of bamboo powder (mass%)

FIGURE 1 Typical stress-deflection curves as a function of content of bamboo powder.

FIGURE 2 The relationship between flexural strength and flexural modulus and content of bamboo powder.

1

BP66.6%-BDP33.3% BP50.0%-BDP50.0% BP33.3%-BDP66.6% •--BDP100%(CP-300)

30

-

30

t

20

o Tensile strength ^ Young's modulus

20

I

0 10

10

\

'£•-;• 0.01

• • ''

0.02 Strain E

:

f

I

1

T

.-

f

:

8

-4

o

2

i

1 0.03

0

0

25

50

75

0

CBDP:CP-300)

Content of bamboo powder (mass%

FIGURE 3 Typical stress-strain curves as a function of content of bamboo powder.

FIGURE 4 The relationship between tensile strength and Young's modulus and content of bamboo powder.

1 PQ

2

4 6 8 Deflection (mm)

10

FIGURE 5 Effect of alkali treatment on the flexural behavior of "green" composites.

Part II

Characterisation

This page intentionally left blank

Effect of Fibre-Orientation on Mechanical Properties of Polypropylene Composites S. Houshyar, R. A. Shanks*, A. Hodzic** Applied Chemistry, RMIT University, GPO Box 2476V, Melbourne, 3001, Australia

ABSTRACT The mechanical and structural properties of a composite consisting of polypropylene fibres (PP) in a random poly(propylene-co-ethylene) (PPE) has been prepared and its properties evaluated. The mechanical properties of PPE laminates were largely determined by the presence of a complex fibre orientation distribution in the composite. The results showed that all-PP composites demonstrated enhanced stiffness and creep resistance with decrease in the orientation angle, 9, between the fibre axis and the load direction; the friction coefficient decreased linearly as 9 increased from 0° to 90°. Composites with zero angle (unidirectional composites) between fibre axis and applied load (9 = 0°) possess the highest stiffness, because the fibre efficiency is inherently strong for this system as all the fibres are able to contribute to the composite stiffness and to carry the load. An increase in 9 leads to a decrease in composite stiffness, which indicates that by increasing 9, a lower proportion of the applied load is transferred to the fibres and thus it is not completely distributed among the fibres. The composite with 8 = 90° showed the highest relative creep, whereas the composite with 9 = 0° displayed the lowest creep. In general, the relative creep increased steadily with increasing 9; also the increment in 9 produced a decrease in the creep modulus, and it was observed up to 90°. Composites with woven fabric showed the best properties, after the unidirectional composites, due to the interlaced structure of the woven fabric. The bundles of plain fibre cloth restrict displacement in each other and result in high stiffness. In addition the mechanical properties of the composite with random fibres are somewhat between the composite with 0°< 9 < 90°, due to alignment of fibers along any direction in the composite.

INTRODUCTION There are some important parameters, which affect the properties of thermoplastic composites. These are: (a) the fibre-matrix bond (b) the type and volume of fibre, (c) the distribution and orientation of fibres within the matrix, (d) the ability to obtain isotropic and orthotropic behavior if required (e) ease of handling of the reinforcement and a (f) suitable method of manufacture [1,2,3]. A crucial parameter for design of composite material is the fibre orientation, as it controls the mechanical response. In order to obtain the favoured materials properties for a particular application, it is important to know the function of performance ' * Correspondence Author, Applied Chemistry, RMIT University, GPO Box 2476V, Melbourne, 3001, Australia, Tel and fax: +613 9925 2122, robert.shanks (ftjrmit.edu.au ** Current address: School of Engineering, James Cook University, Townsville, Australia

42

Effect of Fibre-Orientation on Mechanical Properties

change with the fibre orientation under given loading conditions. The mechanical properties of polymer composite systems as a function of fibre orientation have been studied in the past [4,5]. Krenchel [6] presented an analytical approach, which suggested that the mechanical properties of a laminated composite depend on the individual mechanical properties of its constituents. The mechanical properties of these thermoplastic composites can be modified by changing fibre orientation. EXPERIMENTAL The materials employed in this investigation were random poly(propylene-coethylene) matrix (PPE) (density, p = 0.905 gem"3 , MFI = 0.8 dg/min, melting temperature= 147.5 °C, ~ 5 % ethylene) and PP fibres, woven and non-woven mat (fibre diameters = 50 (j.m, fibre tensile strength = 250-350 MPa, tensile modulus = 4.7 GPa and length = 2-3 cm, fabric modulus = 5.2 GPa with plain weave). The fibres were washed with acetone to remove any processing lubricants. The fibres were obtained from Melded Fabrics Pty Ltd, the PPE from Basell Australia Pty Ltd. According to DSC results [7], 155 -160 °C was selected as the molding temperature range. A heated press was used in two stages. In the first stage, long PP fibres, woven or non-woven mat were distributed according to Table 1, on top of a PPE film and placed between two Teflon sheets, then pressed at 155 -160 °C for 5-7 minutes. TABLE I. Designations of the samples with different orientation angles

Sample Number of Plies Ply one

Cl

C2

C3

C4

C5

C6

C7

C8

One& three Woven

One& three Random

One& three

Three

Three

Three

Three

0°

0°

45°

0°

45°

Ply two

Woven

Random

0°

90°

0°

45°

90°

Ply three

Woven

One& three Nonwoven Nonwoven Nonwoven

Random

0°

0°

45°

0°

45°

After initial pressing, 11-14 kPa pressure was applied for 8-10 minutes. In the second stage, three layers of the composite were laminated together under the same conditions as in the first stage, to provide a more uniform composite. The PP fibre concentration in the composite was 50 %v/v. The mechanical properties were measured with a Perkin-Elmer DMA7e in extension and three-point bend mode at 25°C. The static force was scanned from 100 mN to 8000 mN at 100 mN/min and 0.0 mN to 6400 mN at 400mN/min for three-point bend and extension modes, respectively. The samples were cut from sheets, with dimensions 1x12x5 mm and 1x10x5 mm for three-point bend and extension modes, respectively. Creep recovery was recorded with a Perkin-Elmer DMA7e in extension mode at 25 °C with constant time (~15min) for creep and recovery. To investigate the effect of different PP fibre orientation in one layer composites, creep tests were carried out with a constant static force of 2000 mN at room temperature and duration time of 25 minutes. A small residual load (~ 200 mN) was left on each specimen in the unload condition to maintain the integrity of the loading assembly. For composites with three plies, creep tests were carried out with the same parameters except for the values of static force of 3000 mN and residual load of 400 mN. The samples were cut from the sheets with dimensions 1x10x5 mm. In order to test the reliability of experimental

Effect of Fibre-Orientation on Mechanical Properties

43

results, at least five specimens for each type of composite were used for creep recovery behavior tests. RESULT AND DISCUSSION PP-PPE composite with one ply The effect of fibre orientation on the mechanical properties of PP-PPE composites was analysed by DMA. Mechanical properties such as tensile modulus of the composites were measured and the results are shown in Figure l(a,b). Figure l(b) shows that the composite with zero angle between fibre axis and applied load (9 = 0°) posses the highest stiffhess. The fibre efficiency is inherently strong for this system as all the fibres are able to contribute to the composite stiffness and to carry the load. An increase in 6 leads to a decrease in composite stiffness, which indicates that by increasing 8, a lower proportion of the applied load is transferred to the fibres and it is not completely distributed among the fibres. In other words, a decrease in the fibre efficiency results in a decrease of the composite stiffness. In the composite where 9 = 90°, the fibres cannot act as a reinforcement in the matrix and fibre efficiency is minimal. The fibres cannot support the matrix, which flows and deforms under the stress. 1400-1 12001000-

— re o. 5 ~~ UJ

800-

6004002000-

Different type of fibre orientation

FIGURE 1. Tensile modulus of the one layer composite system different type of fibre orientation

On the other hand, the composite reinforced with the woven fabric (Figure 1 (a)) shows an inherently strong stiffness, due to the well-proportioned structure of the woven fabric. The bundle of plain fibre cloth restricts displacement of each fibre and result in a high stiffness. Also, the results provide information on the composites with non-woven and random mat. Although the composite with non-woven mat provides better stiffness than the composite with random fibres due to the high entanglement of fibres in some parts, they have the same properties in both directions. In addition, the modulus of the composite with random fibres is somewhat between the composites with 0° < 9 < 90°, due to alignment of fibres along any direction in the composite.

Effect of Fibre-Orientation on Mechanical Properties

44 PP-PPE composite with 3 plies

Figure 2 shows the tensile modulus as a function of fibre orientation in the composite with three plies. There is a decrease in the tensile properties of the composite with increasing 9 of each ply. In both data sets, one and three plies, it can be assumed that the composite stiffness clearly expresses a dramatic increase with decreasing 9. This is due to the fact that the reinforcement imparted by the fibres allows stress transfer from the matrix to the fibres. The load transmittance is a function of fibre orientation and magnitude of the fibre-matrix interfacial bond. As the angle (6) increases, the stiffness of the composite will decrease and lead to breakage of the composite at lower stress [5,6]. According to these results, different fibre orientations can be used in each ply depending on the direction of the applied load on the composite. D longitudinal E3 Transverse

Different type of fibre orientation

FIGURE 2. Tensile modulus of the three layers composite

CREEP BEHAVIOUR The effect of PP fibre orientation on the creep response of the composite is shown in Figure 3. The load transmittance is a function of fibre orientation and fibre/matrix interfacial bond. The composite with 6 = 90° showed the highest relative creep value, whereas the composite with 9 = 0° displayed the lowest value. In general, the relative creep increased steadily with increasing 9; also the increment in 9 produces a decrease in the creep modulus, and is observed up to 90°. As 6 increased, fibres were not able to contribute to the stiffness of the composite, which resulted in a decrease of the composite stiffness at lower stress. The results clearly indicate that the creep resistance is dependant on fibre orientation and the quality of the fibre-matrix interface.

Effect of Fibre-Orientation on Mechanical Properties

45

Time (min)

FIGURE 3. Creep response of the composite with different orientation

CONCLUSION The effect of fibre orientation on the creep and mechanical properties of PPE has been analyzed by static and dynamic mechanical analysis [DMA], The results clearly indicate that there is a decrease in composite modulus by increasing 6, between the fibre axis and applied load direction, due to a decrease in the contribution of the fibres in carrying the applied load. The study proves that the composite with woven fabric has the best properties due to the stiffness of the reinforcement. ACKNOWLEDGEMENTS Financial support from International Postgraduate Scholarship (IPRS) is acknowledged. REFERENCES 1. N. Barkoula, J. Karger-Kocsis. 2002," Effects of fibre content and relative fibre-orientation on the solid particle erosion of GF/PP composites " Wear, 252, 80-87. 2. Y. Liang, S. Li, R. Zhang, S. Li. 1996." Effect of fiber orientation on a graphite fiber composite in single pendulum scratching," Wear, 198, 122-128. 3. Karger-Kocsis, J. 1995. Polypropylene: Struct. Blends Composites, Vol. 1, Chapman & Hall, UK, pp. 1-50. 4. M.A. Lopez-Manchado, M. Arroyo. 2000. "Thermal and dynamic mechanical properties of polypropylene and short organic fiber composites," Polymer, 41, 7761. 5. J. Rosenthal. 1992. "A model for determining fibre reinforcement efficiencies and fibre orientation in polymer composites, " Polymer Composites, 13(6), 462-466. 6. William, D.J. 1994. Materials Science and Engineering, an Introduction, John Wiley and sons, pp. 106. 7. S. Houshyar , R. A. Shaiiks. 2003. "Morphology, thermal and mechanical properties of polypropylene fibre-matrix composites," Macromolecular Materials and Engineering, 288, 599.

Friction and Wear Properties of Potassium Titanate Whiskers Reinforced PTFE Composites Xin FENG *, Donghui CHEN, Xiaohong JIANG, Shenghua SUN, Xiaohua LU Nanjing University of Technology, P. R. China Yuansheng JIN National Key Laboratory of Tribology, Tsing-Hua University, P. R. China

ABSTRACT Potassium titanate whiskers (PTW) is a promising reinforcer with good abrasion proof properties. Tribological performance of PTW reinforced PTFE composites (PTW-PTFE) was investigated. The results showed that addition of PTW caused an increase in wear resistance of PTFE. The wear of PTW-PTFE composites was only about one tenth of PTFE. Compare to PTFE, both of the limit loading and the limit sliding speed of PTW-PTFE increase 10% and 60% respectively. The friction coefficient and hardness of PTW-PTFE were slightly changed. The friction coefficient of PTW-PTFE was much steadier than that of PTFE. Both PTFE and PTW-PTFE had lower wear mass loss but higher wear traces surface at 200 °C than that at 25 °C. The PTW with larger size had better wear resistance. SEM showed that PTW were effective in impeding large-scale fragmentation of PTFE, thereby reducing the wear.

INTRODUCTION Polytetrafiuoroethylene (PTFE) is a thermoplastic with a variety of high performance like extremely low friction coefficient and high temperature stability. However, PTFE exhibits poor wear, abrasion resistance and severe creep deformation, leading to early failure and leakage problems in sealing material [1, 2]. Potassium titanate whiskers (PTW, K2O-6TiO2) is a promising reinforcer for the composites, due to its unique properties such as outstanding mechanical performance, excellent heat resistance and good abrasion proof properties. Its tensile strength is higher than carbon fibers. PTW is tiny which brings about a micro-reinforcing effect in PTW composites. It is suitable to produce ware with complex shape, high precision and high degree finish of surface [3, 4]. hi this paper, tribological performance of PTW-PTFE of various filler contents at different load, sliding speed was investigated. EXPERIMENTAL Materials The PTFE powder (30um diameter) was provided by Shanghai Tianyuan resin * Corresponding Author, College of Chemical engineering, Nanjing University of Technology, Nanjing 210009, P.R.China, Tel:0086-25-83588063, Fax:0086-25-83588063, e-mail: [email protected]

Friction and Wear Properties of Potassium Titanate Whiskers

47

group (China). PTW, with an average diameter 1.86/xm and length 10.47 jtm, was supplied by Shengyang Jinjian Co. in China. PTW was modified with surface modifier [4]. Preparation of PTW- PTFE composites sample is same as Glass Fibers-PTFE composites. Friction and Wear Testing Friction and wear tests were carried out in an Optimol SRV( Germany) on the dry friction condition. Before testing, specimens ( $24mm> 0.80.60.40.20.0 50N

200N

10wt%PTW-PTFE

FIGURE 3 Effect of PTW size on Wear properties

SEM Analysis of Worn Surfaces Figure 4 (a) and (b) showed micrographs of worn surfaces of PTFE and 20wt%PTW-PTFE composite respectively under 50N load, at 0.04m-s"1 sliding speed and 25 °C, which revealed that worn surface of PTFE was with severe crack, however worn surface of 20wt%PTW-PTFE was very smooth. It indicated that addition of PTW prevented PTFE from large-scale fragmentation generated and grown, thereby reducing fatigue wear, and then increasing the abrasive resistance.

(a) PTFE (b) 20wt%PTW-PTFE FIGURE 4 SEM photographs of worn surfaces of PTFE and 20wt%PTW-PTFE

CONCLUSIONS 1. Filler PTW is effective in impeding large-scale fragmentation of PTFE, thereby addition of PTW causes PTFE composites an increase in wear resistance, meanwhile makes the friction coefficient steadier, while the friction coefficient and hardness is slightly affected. 2. The wear of PTW-PTFE composites was only about one tenth of PTFE.

Friction and Wear Properties of Potassium Titanate Whiskers

51

Compare to PTFE, both of the limit loading and the limit sliding speed of PTW-PTFE composites increase 10% and 60% respectively. It indicated that PTW-PTFE composites with higher PV limit value can work under more rigorous condition. 3. There is the highest wear resistance for PTW-PTFE composites when content of PTW is about 5wt%. 4. Both PTFE and PTW-PTFE have lower wear mass loss but higher wear traces surface at 200 °C than that at 25°C.It is notable that both wear mass loss and wear traces surface should be as criterion whether the material is out of work or not. 5. The size of PTW also influences the tribological properties. The PTW with larger size has better wear resistance. ACKNOWLEDGMENTS Authors appreciate the Outstanding Youth Fund of National Natural Science Foundation of P. R. China (29925616) and National High-tech Research Development Program(863 Program: 2003AA333010) and The Tribology Science Fund of National Tribology Laboratory (SKLT02-2) REFERENCES 1.

Khedkar, J., I.Negulescu, and E. I. Meletis. 2002. "Sliding wear behavior of PTFE composites, " Wear, 252 (5-6): 361-369.

2.

Bijwe, J., S. Neje, J. Indumathi, and M. Fahim. 2002. "Friction and wear performance evaluation of carbon fibre reinforced PTFE composite, " Journal of Reinforced Plastics and Composites, 21(13): 1221-1240.

3.

Feng X.i J. Z. Lii,> X. H. Lu. et al. 1999, "Application of potassium titanate whisker in composite," ACTA MATERIAE COMPOSITAESINICA, 16(4): 1-7

4.

Lii, J. Z., and X. H. Lu. 2001. "Elastic Interlayer Toughening of Potassium Titanate Whiskers-Nylon66 Composites and Their Fractal Research," J. Appl. Polym. Set, 82(2):368-374.

5.

He, C. X., L. P. Shi, and H. P. Shen. 2000. "Friction and Wear Properties of Nanocrystalline A12O3 Filled-PTFE Composite," Tribology, 20(2):153-155.

6.

Wu, R.J. 1998. The Surface and Interface of Polymer. Beijing: Science Press

Study on Jute Fiber Reinforced Polypropylene (PP) Composite Ma Sheng and Wang Yimin State Key Laboratory for Chemical Fibers and Polymer Materials Donghua University, Shanghai, 200051, P. R. China Zhang Anding and Ding Xin College of textile, Donghua University, Shanghai, 200051, P. R. China

ABSTRACT Jute fiber reinforced PP composites are studied in this article. The composites are made of jute fiber as reinforcement for thermoplastic polypropylene. The effect of different fiber content and different fiber length on composite properties, such as the tensile strength, the flexural strength and impact strength are studied. The process conditions for making the composite and the invalidation theory of the composite are also discussed. The results indicate that the comprehensive properties of the composites with 12mm fiber length and 10% fiber weight content are improved obviously. INTRODUCTION Composite materials made of cellulose-based fibers such as jute fiber demonstrate remarkable environmental and economical advantages. This is because the cellulosic fibers have many advantages, such as low cost, low density, high specific strength and modulus, limited requirements on processing equipments, no health problems, ease of fiber surface modification, and availability as renewable natural resources. And this composite combines good mechanical properties with a low specific mass and offers an alternative material to glass-fiber reinforced plastics in some technical applications. However, their high level of moisture absorption, poor wettability by non-polar plastics, and insufficient adhesion between untreated fibers and the polymer matrix frequently exhibit unsatisfactory mechanical properties. Despite this intrinsical disadvantages, the study of jute fiber-reinforced plastic composites is continually increasing. For example, some scientists and experts of India have given lots of studies about jute fiber reinforced thermoplastic composites. Through these earlier papers, it is summarized that mechanical performance of a fiber-reinforced plastic composite primarily depends on three factors: (a) strength and modulus of the reinforced materials, (b) strength and chemical stability of the matrix, and (c) effectiveness of the bond between the fibers and the polymer matrix in transferring stress across the interface. Most of the composite properties are strongly depended on microstructure parameters such as fiber length, fiber content and

* Corresponding author, Donghua University, Tel: 86-21-62379785, Fax: 86-21-62379309, E-mail address: [email protected]

Jute Fiber Reinforced Polypropylene Composite

53

alignment and packing arrangement of fibers. In randomly oriented short fiber composites, the fiber length and fiber content play an important role in determining their mechanical performance. The optimal length of the jute fiber depends on the bonding between the fiber and the resin. And similarly, the volume fraction of the fiber lies on the interfacial bonding exists between the fiber and the matrix. Several types of thermoplastics like polyethylene, polypropylene and polystyrene have been reported as matrices for natural fiber composites. These polymers may have different agglutination towards the fiber due to the difference in their chemical structure. As a consequence, the reinforcement effect of the fibers in these matrices may vary widely. It is well known that some polymers are susceptible to reinforcement; others are not with respect to particular filler. For fiber filled composites it has been found that the higher the flow limit of the matrix the lower the critical length of the fiber. This paper selected polypropylene (PP) as the matrix. However, the studies reported in the literature on the use of jute fiber as reinforcement in thermoplastic polymer like PP have been scanty. This can be attributed to the invalidation behavior for jute fiber reinforced PP composite is less well understood than for glass fiber reinforced plastic composite because of the lack of systematic and detailed information available. Base on this reason, this article deals with the effects on the mechanical properties of jute fiber reinforced PP composite with different fiber length and fiber content. By comparison, the parameter and process optimization of the composite is introduced in this paper. EXPERIMENTAL Materials Jute fiber was obtained from Zhejiang Jute Co., Ltd, China. And PP was gained from Jin Shan Petrochemical Corporation Ltd, Shanghai, China. The properties of jute fiber and the matrix PP are presented in Table 1. Jute fiber was cut to 3mm, 5mm and 10mm in length respectively. And the average value of the diameter of the jute fiber is 54um. The melt flow index (MFI) of PP is 5g/10min. TABLE I Properties of jute fiber and PP

,, . . , Materials

_ ., Density , 4

Tensile ,, strength

Tensile , . modulus

Elongation ., , at break

Moisture . regain

rpastern

-

/MPa

/GPa

/%

/%

r

J/°

1.41 0.91

424 28.5

14.13 0.867

2.54 10.34

12.6 —

4.24 —

/g cm

,.

I/O

Jute fiber PP

Sample Preparation A screw mixing technique was used to make the jute fiber reinforced PP composite. The composite containing 10, 20 and 30% by weight of fiber was prepared using fiber of length in 3mm, 5mm and 10mm. The temperature of the screw head is controlled no more than 180°C because of the degradability of the jute fiber. The tensile specimen, the flexional specimen and the impact specimen were prepared by injection

54

Jute Fiber Reinforced Polypropylene Composite

molding according to the standard of ASTM D63814, ASTM D790 and ASTM D256. Testing of Composite Tensile testing and flexural testing of thermoplastic composite was carried out using a Universal testing machine of Tianshui Sansi of China at a crosshead speed of 5mm/min and a gauge length of 100mm and 95mm respectively. The tensile modulus and elongation at break of the composite were calculated from the stress-displacement curve. At least five specimens were tested for each set of samples. Impact testing was carried out using a pendulum RESILIMPACTOR of Italy. RESULTS AND DISCUSSIONS Effect of Fiber Content Table II Effect of jute content on mechanical properties of jute/PP composite

B D Materials PP C 29.88 29.82 29.78 Tensile strength/MPa 28.5 1.875 1.634 1.157 Tensile modulus/GPa 0.867 34.68 39.2 40.7 Flexural strength/MPa 30.16 1.184 1.768 2.5 Flexural modulus/GPa 0.776 4.95 4.75 6.57 4.78 Impact strength/J • m"2 The weight content of the jute fiber in the composite is 10, 20 and 30%; and corresponding volume content is about 5.6, 13.1, and 20.5%. The effect of jute content on mechanical properties of jute fiber reinforced PP composite is presented in Table II. From the results it is observed that all mechanical properties except impact strength improve with the adding jute fiber to the resin. From T Table II, the most remarkable observations are: (a) The tensile strength of the composite increases by about 4.5% compare to PP. However, the improvement quantity is slightly falling with increase of jute content. And it was easy to find that the improvement is the most remarkable when the fiber weight content is 10%. (b) The flexural strength of jute fiber reinforced PP composite improves greatly compare to PP, and the increase of the flexural strength with increase of jute fiber. The flexural strength increases by 35% compare to PP. (c) Similarly, both the tensile modulus and the flexural modulus improve more or less. The improvement quantity is increasing with increase of jute fiber for the tensile modulus and contrarily for the flexural modulus. This can be found clearly from figure 1.

Jute Fiber Reinforced Polypropylene Composite

2

4

6

55

8 10 Elongation/%

FIGURE 1 The stress-elongation curve of the composites with different fiber content Note: Apresents for PP, and B, C and D present for 10%, 20%, and 30% jute/PP composite respectively

In contrast to the tensile and flexural properties, the work of impact test is found to be pessimistic. The impact strength of the composite has a slightly decline compare to PP. The reason for low impact strength of jute/PP composite is due to the high bonding of the fiber with polypropylene resin, which resulted in the fracture of fiber at the crack without any fiber pull-out.

Effect of Fiber Length In this paper, the jute fiber was cut 3mm, 5mm and 10mm respectively. Generally speaking, the fiber reinforced thermoplastics composite showed an increasing trend in their mechanical properties with the fiber length. However, the flexural properties of the composite demonstrate an unusual change. Table III shows all the variation in tensile and flexural and impact properties of the thermoplastics composite. Table III Effect of jute length on mechanical properties of jute/PP composite

Materials PP Tensile strength/MPa 28.5 Tensile modulus/GPa 0.867 Flexural strength /MPa 30.16 Flexural modulus /GPa 0.776 Impact strength /J • m"2 6.57

B 29.66 1.833 35.8 1.334 5.78

C 29.88 1.875 34.68 1.184 4.78

D 29.99 1.976 37.7 1.365 4.41

From Table III, the jute/PP composite shows an enhancement in their tensile strength and modulus by increasing the fiber length from 3mm to 10mm, whereas the flexural strength and modulus go through a decrease with the fiber length from 3mm to 5mm and followed a significant increase with the fiber length from 5mm to 10mm. This decreasing trend in flexural properties of the composite may be due to the jute fiber didn't completely disperse and the fiber-to-fiber contact occurred. In terms of

56

Jute Fiber Reinforced Polypropylene Composite

impact strength, the change rule is approximate with effect of fiber content.

Elongation/%

FIGURE 2 The stress-elongation curve of the composites with different fiber length Note: This time A presents for PP, and B, C and D present for 3mm, 5mm, and 10mm jute/PP composite respectively

Figure 2 shows the stress-elongation curve of the composite with the change fiber length. CONCLUSIONS In this paper, the mechanical properties of short jute fiber reinforced PP composite have been investigated as a function of fiber content and fiber length. The various parameters lead to the various properties of the composite. As a consequence, the reinforcement effect of the fiber to the resin varies widely. By comparison, the composite with 10mm fiber length and 10% fiber weight content is improved most obviously.

REFERENCES [1] A.C.Karmaker, J.A.Youngquist. 1996. "Injection Molding Polypropylene Reinforced with Short Jute Fibers," Journal ofApplied Polymer Science, 62(8): 1142-1151. [2] Karmaker A. C, Schneider J. P. 1996. "Mechanical performance of short jute fiber reinforced polypropylene," Journal of Materials Science Letters, 1(2): 201-202. [3] A.K.Rana, A.Mandal, B.C.Mitra, R.Jacobson. 1998. "Short Jute Fiber-Reinforced Polypropylene Composites: Effect of Compatibiliar," Journal of Applied Polymer Science, 69: 329-338. [4] Rana A.K., Mitra B.C, Banerjee A.N. 1999. "Short jute fiber-reinforced polypropylene composites: dynamic mechanical study," Journal of Applied Polymer Science, 71(4): 531-539. [5] Jochen Gassan, Andrzej, K.Bledzki. 1999. "Influence of Fiber Surface Treatment on the Creep Behavior of Jute Fiber-Reinforced Polypropylene," Journal of Thermoplastic Composite Materials, 12: 388-398. [6] Saha AK, Das S, Bhatta D, Mitra BC. 1999, "Study of jute fiber reinforced polyester composites by dynamic mechanical analysis," Journal of Applied Polymer Science, 71(9): 1505-1513. [7] Ghosh P, Bose NR, Mitra BC, Das S. 1997. "Dynamic mechanical analysis of FRP composites based on different fiber reinforcements and epoxy resin as the matrix material," Journal of Applied Polymer Science, 64(12): 2467-2472.

Mechanical and Thermal Properties of Composites of Epoxy Resin Derived from Kraft Lignin Filled with Cellulose Particles Masahiro Funabashi*, Shigeo Hirose National Institute of Advanced Industrial Science and Technology, Japan Hyoe Hatakeyama Fukui University of Technology, Japan

ABSTRACT A mixture of ester-carboxylic acid derivatives (KL polyacid, KLP A) was obtained by reaction of an ethylene glycol solution of Kraft lignin (KL) and succinic anhydride in the presence of a catalytic amount of dimethylbenzylamine at 100 °C. Cellulose particles were added to the above mixture with various mixing ratios from 0 to 60 wt%. A prepolymer of lignin-based epoxy resin was prepared by a reaction of the mixture of KLPA and cellulose particles with ethylene glycol diglycidyl ether (EGDGE) at 100 °C. The above mixtures were molded into sheets at 130 °C for 5 hours. The mechanical properties of the samples were investigated by tensile tests using plate type specimens. The elastic modulus and tensile strength were determined by the tensile tests. Thermal properties of composite samples were measured by differential scanning calorimetry (DSC) and thermogravimetry (TG). Glass transition temperatures of samples were determined by DSC. Thermal decomposition temperatures and mass residues were determined by TG. The values of strength and modulus determined by tensile tests were maximum at cellulose content of 60 wt% for composite samples with cellulose particles of 25 mm diameter. Thermal properties such as peak temperatures of thermal decomposition and glass transition temperature determined by TG and DSC were not affected by cellulose contents.

INTRODUCTION Polymer composites consisting of polymer matrices and fillers are used in various industrial fields, such as the automobile, construction, and aerospace industries, etc., since they have high specific modulus, high specific strength and are lightweight, easy to process and corrosion resistant. Polymer composites are difficult to dispose of, because they are highly durable. Large amounts of bio-waste are produced in the agricultural and industrial fields. The above problem can be solved by the production of biodegradable polymer composites using plant components from bio-waste. Biodegradable polymers can be prepared by the combination of polymers derived from plant components and fillers from solid parts of plants [1-6]. Plant components such as saccharides and lignin * Corresponding Author, National Institute of Advanced Industrial Science and Technology, AIST Tsukuba Central 5, Tsukuba, Ibaraki 305-8565, Japan, Tel. +81-(0)29-861-4584, Fax. +81-(0)29-861-6250, [email protected]

58

Properties of Composites of Epoxy Resin

have been recognized as important raw materials for polymer processing, since they are produced abundantly in nature. Polyurethane composites consisting of the polyurethanes, which were derived from saccharides and lignin, and solid plant materials were also studied at our laboratory [1-6]. Biodegradable polymer composites consisting of the poly-lactones, such as poly(lactic acid) and poly(caprolactone), and cellulose fillers were also studied in our laboratory [7-8]. In this study, the polymer composites were prepared by combining epoxy resin derived from Kraft lignin (KL) and cellulose particles. The mechanical and thermal properties of the above composite samples were investigated. EXPERIMENTAL Materials Kraft lignin (KL) was used as a lignin in this study. KL was dried in a vacuum oven at 60°C for 2 days. Cellulose particles, Avicel (Asahi Chemical Industry CO., LTD) were used in this study. Avicel (PH-M25) was used as fillers of composite samples, where the average diameter of particles was 25 |^m. Sample preparation An ethylene glycol solution of KL was reacted with succinic anhydride in the presence of a catalytic amount of dimethylbenzylamine at 100 °C. A mixture of ester-carboxylic acid derivatives (KLPA) was obtained by the above reaction. Cellulose particles were added to the above mixture with various mixing ratios from 0 to 60 wt% and were mixed well by a mechanical mixer at 70 °C for 2 hours. A prepolymer of lignin-based epoxy resin was prepared by a reaction of KLPA with ethylene glycol diglycidyl ether (EGDGE) at 100 °C for 20 min. The above mixtures were poured into mold to form sheets at 130 °C for 5 hours. The sheets were removed from the mold after cooling and were cut into plate shaped specimens for density measurements and mechanical tests. Density measurements The apparent density (/?a) was determined as the ratio of sample weight to the apparent volume of the plate shaped specimens. Tensile tests Mechanical properties of samples were investigated by tensile tests using a Shimadzu AG-10TB. The sizes of plate type specimens were 15 mm wide, 3 mm thick and 150 mm long. Span was 100 mm. Test speed was 5 mm min"1. The tensile strength and tensile modulus were determined from the stress-strain curves of tensile tests. Tensile strength (Oi) was defined as the maximum stress of the stress-strain curves. Elastic modulus (Et) was defined as the gradient of the linear part of stress-strain curve. The specific strength ((Tts) and the specific modulus (E^) were calculated as Oi and Et divided by apparent density of samples, /?a. Differential scanning calorimetry (DSC) DSC was carried out using a Seiko DSC 220 at a heating rate of 10 °C min"1 in the

Properties of Composites of Epoxy Resin

59

temperature range from -60 to 70 °C. The initial glass transition temperature (Tig) of samples was determined from DSC curves according to the method reported by Hatakeyama [9]. Thermogravimetry (TG) TG was carried out in a nitrogen flow at flow rate 300 ml min 1 using a Seiko TG 220 at a heating rate of 10 °C min"1 in the temperature range from 30 to 550°C. The peak temperatures (DTn and DT^) were observed from derivative thermogravimetry (DTG) curves. Mass residue of samples at 500°C (MR500) was determined from TG curves.

RESULTS AND DISCUSSION Density measurements The relationship between content of cellulose particles and density of samples is shown in Figure 1. Density of samples increases with increasing cellulose content when content reached 40 wt%, and then density slightly decreases with increasing content. In the range of cellulose content from 0 to 30 wt%, the increase of cellulose makes the density increase. At the same time, the viscosity of mixture of epoxy pre-polymer and cellulose particles increases with increasing cellulose content. It was thought that when content reached 30 wt%, the air bubbles could not escape during the sample preparation. The presence of void in the samples caused the density decrease at content more than 40 wt%.

0

20

40

60

80

content /wt% FIGURE 1 Relationship between content of cellulose and density of samples; cellulose particles, PH-M25

Mechanical tests The mechanical properties of samples were investigated by tensile tests using plate-shape specimen. The strength and modulus of samples are shown in Figure 2. Both values of strength and modulus of samples increase with increasing cellulose content, and when content exceeds 40 wt%, both values rapidly increase. The reinforcement effect of cellulose particles can be clearly observed as shown in Figure 2. These results indicated

60

Properties of Composites of Epoxy Resin

that KL epoxy resin used in this study could be used as a good adhesive for polymer composites.

1500

V) _3 3 •D

O

E

20

80

40 60 content /wt%

FIGURE 2 Relationships between content of cellulose and, strength and modulus of samples by tensile tests; cellulose particles, PH-M25

Thermal properties measurements The thermal properties of samples were investigated by TG and DSC analyses. TG curves and derivative TG (DTG) curves were obtained by TG analysis. Relationship between cellulose content and mass residues at 500 °C (M?5Oo) determined by TG analysis are shown in Figure 3. MRsoo increases with increasing cellulose content. The above results suggest that cellulose molecular changed to component in the residue after heating up to 500 °C. Peak temperatures of DTG curves were also determined by TG. DTG curve of KL epoxy sample without cellulose particles showed one peak at 376 °C. The DTG curves of KL epoxy samples with cellulose particles of PH-M25 showed two peaks at almost the same temperatures, that is, the higher temperature DTA2 was ca. 377 °C and the lower temperature D7d2 ca. 315 °C. The height of peak at DTdi increases with increasing cellulose content. In contrast, the height of peak at T)T&2 decreases with increasing content.

20

40

60

80

content /wt% FIGURE 3 Relationship between content of cellulose particles and mass residues at 500 °C in DTG curves; cellulose particles, PH-M25

Properties of Composites of Epoxy Resin

61

The initial glass transition temperatures (7jg) and heat capacity gap at Tig were determined by DSC analysis. 7ig of all the composite samples with cellulose particles of PH-M25 including pure epoxy sample were almost the same ca. -17 °C. The heat capacity gaps at Tig (ACP) of composites samples decreased linearly with increasing cellulose content and the extrapolated value of ACP at 100 wt% content was estimated as 0.

CONCLUSIONS Elastomeric polymer composites of lignin-based epoxy resin filled with cellulose particles were prepared by a reaction of an ethylene glycol solution of Kraft lignin and succinic anhydride with ethylene glycol diglycidyl ether in the presence of a catalytic amount of dimethylbenzylamine. Mechanical and thermal properties of the composite samples were investigated by the tensile tests, TG and DSC. Tensile strength and tensile modulus increase with increasing content of cellulose particles and reach the maximum values at cellulose content 60 wt%. Results of thermal analyses by TG and DSC indicated that there is no interaction between epoxy polymer and cellulose particles detected by the thermal analyses.

REFERENCES 1.

2.

3.

4.

5.

6.

7. 8.

9.

D. Kamakura, H. Hatakeyama, H. Kasahara, S. Hirose and T. Hatakeyama, Thermal and mechanical properties of polyurethane composites containing wood particles, the proceedings of 10th International Symposium on Wood and Pulping Chemistry, 1999, 3,442 H. Hatakeyama, D. Kamakura, S. Hirose and T. Hatakeyama, Biodegradable polyurethane composites containing coffee bean parchments, "Recent Advanced in Environmentally Compatible Polymers", J. F. Kennedy, G. O. Phillips, P. A. Williams and H. Hatakeyama, Eds., Woodhead Publishing Ltd., 2001, p.191 M. Funabashi, S. Hirose and H. Hatakeyama, Thermal and mechanical properties of polyurethane composites containing residue from palm oil production, the proceedings of 5th Pacific Rim Bio-Based Composites Symposium, Canberra, 2000, 591 M. Funabashi, S. Hirose, M. Sibaja, M. Moya and H. Hatakeyama, Thermal and mechanical properties of polyurethane composites consisting of pineapple molasses and banana fibers, the proceedings of USM-JIRCAS Joint International Symposium "Lignocellulose- Material of the Millennium: Technology and Application", Penang, 2001, 203 M. Funabashi, S. Hirose, T. Hatakeyama and H. Hatakeyama, Effect of filler shape on mechanical properties of rigid polyurethane composites containing plant particles, Macromolecular Symposia, 2003, 197,231-241 M. Funabashi, S. Hirose and H. Hatakeyama, Mechanical properties of polyurethane composites using fibers of oil palm empty fruit bunches, the proceedings of 3rd Asian-Australasian Conference on Composite Materials, Auckland, 2002, 345 M. Funabashi and M. Kunioka, Composites consisting of poly(e-caprolactone) and cellulose fibers directly molded during polymerization by yttrium triflate, Green Chemistry, 2003, 5,591 M. Funabashi and M. Kunioka, Biodegradable Composites of Poly(lactic acid) with Cellulose Fibers Polymerized by Aluminum Triflate, the proceedings of 1st IUPAC International Conference on Bio-based Polymers, Saitama, 2003, P2-1 T. Hatakeyama and F. X. Quinn, Thermal analysis: fundamentals and applications to polymer science 2nd edition, John Wiley & Sons, 1999, pp.59

Effects of Microcracks and Surface Roughness on Thermal Oxidation of Carbon-Fiber Reinforced Polyimide Composite Huang-Kuang Kung and Hung-Shyong Chen Graduate Institute of Mechatronic Engineering Cheng Shiu University, Taiwan, R.O.C.

ABSTRACT In high temperature applications, the oxidation of polymer composites will lead to micro structure change and mechanical property degradation and both of them have harmful effects on service life. In the literature, some researchers have suggested the fiber-matrix interface model to explain the discrepancy while others believed it had something to do with the surface preparation. In this research, a microstructure-based oxidation model with microcrack and surface roughness is proposed to predict the high-temperature oxidation of fiber-reinforced polymer composite. Consequently, predictions concerning high-temperature oxidation are favorably supported by experimental data.

INTRODUCTION Many studies on the high-temperature thermal oxidation of polyimide-matrix composites have focused on polyimide-matrix composites. Bowles [1] and Nowak [2] evaluated both the effects of different fiber reinforcements on the stability of oxidation and the properties of various fiber-reinforced PMR-15 composites. Scola et al. [3,4] studied changes in mechanical properties and the associated degradation mechanisms of a series of graphite-fiber/PMR-15 composites during isothermal aging at 316 °C (600 °F). Bowles [1], Nam and Seferis [5-7] took the principal directions of the material into account in investigating the anisotropic nature of thermal oxidation in a composite system. Furthermore, oxidation may induce chemical and microstructural changes, cause loss of material, and crack material [8]. Nam and Seferis[7] studied a unidirectional bismaleimide/carbon fiber composite, and showed that the oxidation rate of a composite may be determined from oxidation rates in the principle material directions. Alston [9] and Scola et al [3,4] found that the weight loss of a composite exceeds that of the neat resin. The findings in [3-4,9] contradict the expectation that the oxidation of a composite should be lower than that of neat resin owing to the shielding of high thermo-resistance fibers. In the literature, some researchers have suggested a fiber-matrix interface model to explain the discrepancy[8,10]. However, others believe that it is associated with the preparation of the surface [5-7].

* Corresponding Author, 840, Cheng-Ching Rd., Neausong, Kaohsiung 833, TAIWAN, R.O.C, Tel: +886-7-7310555; Fax: +886-7-7337100, E-mail address: [email protected]

Effects of Microcracks and Surface Roughness on Thermal Oxidation

63

THEORETICAL DEVELOPMENT Microstructure-based oxidation models are applied to evaluate the anisotropic oxidation of the fibers and the matrix and considering aspects of the micro structure. The oxidation correction factors may be included to account for the additional oxidation area, leading to a transverse oxidation weight loss, Aw'r) of,

where A^ and A^J are the actual and apparent oxidation area of matrix in the transverse direction; AfT) and AfTJ are the actual and apparent oxidation area of fiber in the transverse direction ; A ^ and A ^ are the microcrack-related oxidation area for matrix and fiber in the transverse direction, respectively. The ratio of actual matrix oxidation area, A^, to apparent matrix oxidation area, A(JJ, for the matrix in the transverse composite direction can be easily shown to be, bL + R L bL

'V n

«£1

(2)

=

where R is the radius of the fiber; R^x is the surface roughness in the transverse direction, and b is half the mean distance between fibers. For the oxidation of the fiber, neglecting any loose fibers on the surface, a similar procedure yields, n A

f

f

ji

-Af-^L—2

V

where AfT) is the actual fiber oxidation area and AjJ is the apparent fiber oxidation area. The microcrack-induced oxidation area may relate to crack density, crack length and crack depth. Then, the microcrack-induced oxidation area for matrix and fiber in the transverse direction can be written as, Amd

( O

_

?

_

(T) „ (T)r

(7-)

(7)

(A\

(T) A

*

where c(dT)is the microcrack density in the transverse direction ; c, ( r | is the average microcrack length in the transverse direction', c[T) is the average microcrack depth in the transverse direction ; r^' is the geometry-related correction factor for matrix in the transverse direction ; rf > is the geometry-related correction factor for fiber in the transverse direction. Including the effects of surface roughness and microcracks, the isothermal composite weight change per unit apparent oxidation area in the axial direction of a

64

Effects of Microcracks and Surface Roughness on Thermal Oxidation

composite may be expressed using the modified rule of mixtures as, A(A)

A™

A\A) fa

A{A)

(6)

fa

where A(A) and AmAJ are the actual and apparent oxidation area of matrix in the axial direction ; A(A) and A(fA) are the actual and apparent oxidation area of fiber in the axial direction ; , 4 ^ and Affare microcrack-related oxidation area for matrix and fiber in the axial direction, respectively. Following the same procedure as used for the transverse oxidation model, the ratio between the actual matrix oxidation area and the apparent matrix oxidation area in the axial oxidation of the composite is,

(7) (7)

where S^ is the surface roughness of the composite in the axial direction, and A(nA) and A^fJ are the actual and apparent matrix oxidation areas in the axial direction, respectively. Using a similar approach as taken to the fibers, 71

A

Afa

2

— =l

(8)

ER2 4

where A{A) and A(fA) are the actual and apparent fiber oxidation areas in the axial direction, respectively. Similarly, the microcrack-induced oxidation area for matrix and fiber in the axial direction can be written as, AA) A(A)

AA) fd

~AAT A fa

~;

d

(A) = 2C

"

C

I

-m

~*

(A) (A) C r

"

(A)

f

where c(dA)is microcrack density in the axial direction I c'^'is average microcrack length in the axial direction ; c[A) is average microcrack depth in the axial direction ; r^A) is the geometry-related correction factor for matrix in the axial direction ; r{fA) is the geometry -related correction factor for fiber in the axial direction. EXPERIMENTS Composite panels of carbon fiber/PEEK are cut by a water-cooled diamond-plated saw. The specimens had nominal dimensions of 25 mm long, 25mm

Effects of Microcracks and Surface Roughness on Thermal Oxidation

65

wide and 3mm thick. Silicon carbide CC-Cw emery papers with 80, 100, 120, 240, 400, 600, 800 and 1200 grit respectively, and lum alumina paste were used to polish the surface of the composite. All surfaces of specimens were checked using an optical microscope to ensure that they had been thoroughly polished. Average measurements were obtained using three of specimens to eliminate experimental error. Theoretically, the surface roughness should be the same in all directions on the surface since specimens are isotropic. However, the reinforced fibers make the surface roughness of composite specimens anisotropic. The surface roughness can be defined at the point at which the profile curve is perpendicular to the movement of stylus. The surface roughness of the composite was measured using Mitutoyo Surftest-4 and all the values of Ra, R2 and Rmil were recorded. The cut-off value was 0.8 mm and the traversed length was 4.8 mm, six times the cut-off value, for all measurements. The pre-polished specimens (polished with 80, 100, 120, 240, 400, 600, 800 and 1200 grit and lum alumina paste) were placed on the work stage to measure their surface roughness. RESULTS AND DISCUSSIONS The surface roughness parameters of the Carbon Fiber / PEEK composite are determined for specimens polished with different grits of emery papers. The finest surfaces of the specimens are polished with 1 um alumina paste. Figure 1 plots the surface roughness of the Carbon Fiber/ PEEK composite vs. emery grit size in the axial and transverse directions. The transverse surface roughness consistently exceeds the axial surface at any given grit level. The G3O-5OO/PMR-15 oxidation data in a parallel study [10] are used to verify the surface-roughness oxidation model. All specimens were hand-polished down to the 600 grit level. The corresponding surface roughness R^ and R^ at 600 grit are obtained from Fig. 1 as 3.34 /urn and 2.61 jum, respectively. The development of microcracks of PMR15/G30-500 subjected to 316°Cfor 1000 hours in the axial surface can be observed in Figure 2. The density, length and depth of microcracks are increasing with the exposure time. Figure 3 shows the microcrack density evolution of PMR15/G3O-5OO subjected to 316°Cin the axial direction. As can be seen in the figure, up to 1000 hour, the microcrack density still increasing rapidly. Figure 4 shows the statistic measurement of the average microcrack length and depth of PMR15/G30-500 composite. The average length of microcrack evolves increasingly with time in the first stage and then levels off at 400 hour. The restriction of growing of microcrack length may be due to the specimen width limited in 3 mm for current experimental data. The value of geometry-related correction factors r,(nA) and rjA) are assumed to be unity. Based on the data of Figures 1, 3 and 4, the predictions of weight loss flux AmlcA) of PMR15/G30-500 with the effect of surface roughness and microcrack-related oxidation at 316°C are presented in Figure 5. Compared with experimental data, the prediction of the rule of mixture is the least accurate. As shown in Figure 5, predictions with the effect of surface roughness only, and the effect of surface roughness and microcrack have better agreement with experimental results.

66

Effects of Microcracks and Surface Roughness on Thermal Oxidation

Emery grain size (jim)

FIGURE 1 Surface roughness of Carbon Fiber/ PEEK composite vs. emery grain size.

FIGURE 2 Microcrack observation of PMR15/G3O-5OO subjected to 316°C for 1000 hours.

FIGURE 3 Microcrack density of PMR15/G30-500 subjected to 316 " C on the axial surface.

FIGURE 4 Average microcrack length and depth of PMR15/G30-500 subjected to 316 "C .

Effects of Microcracks and Surface Roughness on Thermal Oxidation

67

Rule of Mixture ~~

Prediction w/ Roughness Only Prediction w/ Roughness & Damage

FIGURE 5 Predictions and experimental data of weight loss flux Am*"' of

PMR15/G30-500.

CONCLUSIONS 1. The fibers are more resistant than the matrix to abrasion. Thus, the polished specimens of the composite might exhibit surface roughness during surface treatment. 2. The transverse surface roughness is consistently larger than the axial surface roughness at a given grit level, because of the fiber-reinforced geometry in the axial and transverse directions. 3. The longer the oxidation time, the more microcracks on the oxidation surface. Therefore, discrepancies between experimental data and predictions are expected, due to severe cracking on the surface. 4. Experimental data confirm the predictions of the high-temperature oxidation by the proposed surface roughness and microcrack model over those of the rule of mixtures. REFERENCES 1.

Bowles, K. J., 1992."Effect of Fiber Reinforcements on Thermo-Oxidative Stability and Mechanical Properties of Polymer Matrix Composites", SAMPE Quarterly, April, pp. 2-12. 2. Bowles, K. J. and G. Nowak, 1988."Thermo-Oxidative Stability Studies of Celion- 6000/PMR-15 Unidirectional Composites, PMR-15, and Celion-6000 Fiber", J. Composite Materials., 22, pp. 966-985. 3. Scola, D. A. and J. H. Vontell, 1991."Mechanical Properties and Mechanism of the Degradation Process of 316 C Isothermally Aged Graphite Fiber/PMR-15 Composites", Polymer Engineering Science, 31, pp. 6-13. 4. Scola, D. A. and M. Wai, 1994."The Thermo-Oxidative Stability of Fluorinated Polyimides and Polyimide/Graphite Composites at 371 °C", J. Applied Polymer Science, 52, pp. 421-429. 5. Nam, J. D. and J. C. Seferis, 1992. "Anisotropic Thermo-Oxidative Stability of Carbon Fiber Reinforced Polymeric Composites", SAMPE Quarterly, October, pp. 10-18. 6. Nam, J. D. and J. C. Seferis, 1991. "A Composite Methodology for Multistage Degradation of Polymers", Journal of Polymer Science, Part B, 29, pp. 601-608. 7. Nam, J. D. and J. C. Seferis, 1992. "Generalized Composite Degradation Kinetics for Polymeric Systems Under Isothermal and Nonisothermal Conditions", J. of Polymer Science, Part B, 30, pp. 455-463. 8. Skontorp, A. and S. S. Wang, 1995. "High-Temperature Anisotropic Thermal Oxidation of Carbon Fiber Reinforced Polyimide Composites: Theory and Experiment", The Tenth International Conference on Composite Materials (ICCM-10), Whistler, British Columbia, Canada, 14-18. 9. Alston, W. B., 198O."Resin/fiber thermo-oxidative interactions in PMR polyimide/graphite composites", Polymer Composites, 1, 66-70. 10. Skontorp, A., 1995. Ph.D. Dissertation, Department of Mechanical Engineering, University of Houston, Houston, TX, USA.

Effects of Fillers on the Tensile Properties of Polyimide Composite Films at Room and Cryogenic Temperatures Y. H. Zhang1'2, S.Y. Fu1*, M. Li1'2, Y. Li1'2, L.F. Li1, Q.Yan1 1 Cryogenic Materials Laboratory, Technical Institute of Physics and Chemistry Chinese Academy of Sciences, Beijing 100080, China 2Graduate School, Chinese Academy of Sciences, Beijing 100039, China

ABSTRACT Nano- and micro-filler reinforced polyimide composite films with a high thermal conductivity and a low thermal expansion while still remaining high modulus and strength are desirable in cryogenic applications. Polyimide composite films were prepared by incorporation of fillers such as clay and silica particles into polyimide matrix. The silica particles were made by sol-gel process. The tensile properties of polyimide composite films were studied at room and cryogenic temperatures (77K) taking into account the effects of filler contents for involved fillers. SEM study was carried out on the fracture surfaces of the polyimide composite films. The results for the dependence of the tensile properties of polyimide composite films at room and cryogenic temperatures were discussed on filler contents for the involved fillers. INTRODUCTION Polyimide (PI) is one of the most promising thermally stable polymers with good mechanical properties. With the rapid developments in space, superconducting magnet and electronic technologies, the cryogenic properties of PI films have recently drawn much attention. So far, remarkable progress has been made in syntheses of PI/MMT, Pl/mica and Pi/silica nanocomposites, and properties of PI nanocomposite films at room temperature have been extensively studied [1-6]. Nonetheless, the conclusions obtained at room temperature cannot simply be transferred to the cryogenic case. Moreover, data for these properties at cryogenic temperature cannot be derived from those obtained at room temperature. Thus, it is important and necessary to study the mechanical and thermal properties at cryogenic temperature of PI nanocomposites for cryogenic engineering applications. However, up to now little report has been presented of the cryogenic properties of the PI hybrid nanocomposite films. Lately, a series of polyimide hybrid films such as PI/MMT, Pi/mica and Pi/silica etc have been synthesized in our laboratory. In this paper, the tensile properties of polyimide composite firms were studied at room and cryogenic temperatures (77K) taking into account the effects of filler contents for the involved fillers. The results for the dependence of the tensile properties of polyimide nanocomposite films at room and cryogenic temperatures were discussed on filler contents.

* Corresponding author, Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, P.O. Box 2711, Beijing 100080, P.R. China, Tel: +86-10-62659040/80669735; Fax: +86-10-62564049, Email: [email protected]/ [email protected]

Effects of Fillers on Tensile Properties of Polyimide Composite Films

69

EXPERIMENTAL WORK The hybrid nanocomposite films were synthesized in our laboratory. The polyimide was prepared from pyromellitic dianhydride (PMDA), 4, 4'—diamine—diphenyl ether (ODA) and the solvent of N, N—dimethylacetamide (DMAc). The PI/MMT composite films were prepared via intercalation method, and the Pi/silica hybrid films were synthesised by sol-gel process. The sizes of film specimens for the tensile tests were respectively 10 mm>

1

20


,mail.chosun,ac.kr

84

Compressive Strength of T300/924C Carbon Fiber-Epoxy Laminates

scaling in 0° fiber-dominated laminates should be regarded as an artifact of the test procedure and failure mode. For investigation of the size effect on the compressive strength of a carbon fiberepoxy laminate, specimens with four different gauge sections are tested statically in uniaxial compression. In order to avoid buckling of the relatively large specimens, the anti-buckling device was used. In this case, it was found that the compressive strength with the device was slightly greater than that without the device due to surface friction between the specimen and the device by pre-torque in bolts of the device. For investigation of influence by the anti-buckling device, the finite element method was used. SPECIMENS AND COMPRESSIVE TESTING The material used was Toray T300 carbon fiber in a Ciba-Geigy 924C epoxy resin (T300/924C). The pre-impregnated tapes (pre-preg) were laid up by hand into a [45/45/0/90]3s quasi-isotropic lay-up and cured according to the manufacturer's recommended procedure. The quality of the module laminates was examined by using ultrasonic C-scanning. Several specimens were cut from the 3 mm thick panels and glass fiber-epoxy reinforcement tabs were bonded giving gauge section of 30 x 30, 50 x 50, 70 x 70 and 90 mm x 90 mm. The geometry of the 30 mm long by 30 mm wide specimen is based on the Airbus Industry test method (AITM-1.008) [7]. Static compressive tests were carried out on a screw-driven Zwick 1488 universal testing machine with a load capacity of 200kN; a crosshead displacement rate of 1 mm/min was used. Load introduction to the specimen was mainly by end loading using modified ICSTM fixture [8]. For all 50mm x 50mm, 70mm x 70mm and 90mm x 90mm specimens an anti-buckling device similar to that used by Soutis [9] was employed to prevent column buckling. It contains a window at the device center, allowing damage around the hole to occur but restraining the specimen from general bending. In order to minimize frictional effect, the clearance between the antibuckling device and the specimen face was less than 100 urn. In case of 30mm x 30mm specimen, the clearance can be successfully kept, but in case of larger specimens than the 30mm x 30mm, the clearance cannot be sustained due to contact by some amount of pre-torque of bolts of the anti-buckling device to prevent buckling from the specimen. TEST STRENGTH RESULTS Failure of specimens was sudden and occurred mainly within the specimen gauge length. Post-failure examination suggests that in-plane fiber micro buckling in the 0° plies is the critical damage mechanism, which causes the catastrophic fracture. Longitudinal splits, fiber/matrix de-bonding and delamination between neighboring plies do not occur gradually but take place suddenly and concurrently with the final failure. This is supported by failure strain measurements. The average failure strain measured by the two back-to-back strain gauges at the point of failure was in the region of 1 %, which similar to the strain of the 0° unidirectional material, hi general, test results for all sizes were good and reproducible. The scatter in axial stiffness and strength for all specimen configurations was less than 5%, see Table 1; the results quoted in Table 1 are based on the average of five specimens tested for each different size.

Compressive Strength of T300/924C Carbon Fiber-Epoxy Laminates

85

TABLE I Average compressive strength of specimens

Compressive Strength of Unnotched Specimen Modified ICSTM

Test Fixture 30x30

30x30

50x50

70x70

90x90

Anti-buckling Device

No

Yes

Yes

Yes

Yes

Average Failure Strength

575

690

736

750

711

Coefficient Variation (%)

3.34

4.92

4.28

4.86

2.77

Specimen Size(mm)

The scatter in strength is probably due to imperfections introduced during manufacturing of the laminates resulting is fabrication defects and non-uniform laminate thickness. Imperfections in specimen geometry can produce misalignment of the specimen in the testing fixture that causes bending and reduction in the measured compressive strength. The average elastic modulus (measured at 0.25% applied strain) for all specimens was 66GPa, which is in a good agreement with that estimated by the laminate plate theory (62 GPa). However, the average strength of the specimen with the 30mm x 30mm gage section was at least 20% lower than that of the bigger size specimens. For this configuration (30mm x 30mm gage section), no anti-buckling device was used and failure occurred prematurely due to global Euler buckling. In the typical stress-strain curve of such a specimen, the strain readings of the two back-to back strain gauges are almost the same up to 0.6% applied strain and start to deviate at higher applied loads indicating out-of-plane deflection that leads to premature failure of the laminate and lower overall strength. However the compressive strengths of 50mm x50mm, 50mm x 70mm and 90mm x 90mm gauge length specimens using the anti-buckling device are 6-7% higher values than compressive strength of 30mm x 30mm gauge length specimen. It can be considered that the reason would be caused by frictional contact due to pre-torque of connecting bolts of the anti-buckling device to guide failure at specimen center as well as to avoid the buckling. From the limited strength data presented in Table 1, there is no evidence of a size effect which is not the case for specimens loaded in tension or flexure. Under compressive load, such a size effect does not seem to exist probably because of the different failure mechanism. The mechanism of failure under compressive loading is fiber micro buckling that initiates in regions of maximum fiber misalignment or waviness introduced during the fabrication process [10]. The fiber waviness is sensitive to thickness change of the specimen rather than the two-dimensional area charge introduced in this study. Therefore, an investigation on thickness effects should be performed. According to a compressive strength prediction model of Slaughter, Fleck and Budiansky [11] based on the microbuckling theory, the compressive strength of a composite, in the state of the axial normal stress, the loading directional shear stress and the transverse normal stress, was estimated. In the microscopic measurement of the quasi-isotropic [45/-45/0/90]3s laminate made of T300/924C, it was confirmed that the fiber waviness angle of the 0°-ply is average 2.5°, and the post failure kink band inclination angle is 25°. When the in-plane shear yielding strength of the used matrix is 63 MPa, and the compressive strength of the [45/-45/0/90]3s laminate predicted by the microbuckling theory is 690.3 MPa. This predicted result is in good agreement with the measured strength of the 30mm x 30mm size specimen, 690 MPa, which is

86

Compressive Strength of T300/924C Carbon Fiber-Epoxy Laminates

obtained from using the anti-buckling device without the pre-torque of bolts of the device. However in case of the measured compressive strengths of 50mm x 50mm, 70mm x 70mm, and 90mm x 90mm size specimens using the anti-buckling device as shown in Table 1, the measured values are 6-7% higher than the predicted values and the compressive strength of the 30mm x 30 mm size specimen. It was concerned as the reduction effect of the applied load on the anti-buckling device due to the surface friction and the deformation constrained boundary condition between the device and the specimen by the applied pre-torque of four bolts to guide failure at the center of specimen within the window of the device. The anti-buckling device has the 20 mm width by 20 mm length window and the thick steel plate of the device is closely contacted by pre-torque of four 1/4 inches bolts. When the applied torque Tis 10.16 N-m which is almost same as the nominal torque, the frictional pressure load Px on the contact surfaces can be expressed as the following equation. When the applied torque T is 10.16 N-m which is almost same as the nominal torque, the frictional pressure load Px on the contact surfaces can be expressed as the following equation. Px=juTm/(0.2d0A)

(1)

where T is the applied torque of the bolt, it is assumed as T = 0.2 Ft d0 by Faupel et al. [12]. The Fi is the axial force due to the torque, do is the bolt nominal diameter, m is the total number of bolts, A is the frictional surface area, and the ft is the frictional coefficient. In this study, d0 = 1/4 inch, m = 4, and /u = 0.245 that is measured at experimental tests. The contact surface boundary condition is assumed that the thickness directional displacements are constrained because of very stiff thick steel plate of the anti-buckling device. Using the finite element method the 1/2 model is analyzed, and the used finite element is the 2-D composite element [13]. The analysis results show equivalent stress distribution Sxxeq / So in width direction due to friction of the anti-buckling device, the equivalent stress decrease with increasing the frictional coefficient. In case of // = 0.245, it is noted that the equivalent stress at the center of the specimen is 7% higher than the compressive stress without the antibuckling device. In other word, it means that the compressive strength of the specimen with the anti-buckling device is 7% higher than that without the antibuckling device due to frictional effect and deformation constraints. Therefore it would be concerned that the realistic compressive strengths of 50mm x 50mm, 70mm x 70mm, and 90mm x 90mm size specimens using the anti-buckling device should be corrected to 684.5Mpa, 697.5Mpa, and 663.2 Mpa respectively which are 7% less than the measured values. CONCLUSION The effect of gauge section was investigated on the static compressive strength of a currently used carbon fiber-epoxy system; the lay-up was a quasi-isotropic and the specimen thickness was kept constant. The anti-buckling device was used on the unnotched specimens, and it was showed that the compressive strength with the anti-buckling device was slightly greater than that without the anti-buckling device due to surface friction between the specimen and the device by pre-torque in connecting bolts. In the analysis result on influence of the anti-buckling device using the finite element method, it was found that the compressive strength of the laminate with the anti-buckling device loaded by bolts was about 7% higher than real compressive strength and the corrected

Compressive Strength of T300/924C Carbon Fiber-Epoxy Laminates

87

compressive strength is in good agreement with the predicted strength based on the fiber micobuckling theory. From these limited data, it appears that the compressive strength does not depend on specimen size. Failure is matrix dominated and occurs due to fiber micro buckling in the 0° plies. The initiation of this failure mode is greatly dependent on initial fiber waviness, which highlights the importance of manufacturing processes. REFERENCE 1.

2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Zwen, C , Smith, W. S. and Wardle, M. W., 1979. "Test method for fiber tensile strength, composite flexural modulus and properties of fabric reinforced laminates". Composite Materials: Testing & Design, 5th Conf. ASTM STP674 (American Society for Testing and Materials, Philadelphia), 228-262. Herring, H. W., 1966. "Selected mechanical physical properties of boron filaments", NASA TN D3202, Bullock, R. E. 1974. "Strength of composite materials in flexure and in tension", J. Composite Mater., 8,, 200-206. Berg, K. R. and Ramsey, J., 1972. "Metal aircraft structural elements reinforced with graphite filamentary composites", NASA CR-112162, August Wisnom, M. R., 1991. "The effect of specimen size on the bending strength of unidirectional carbon fiber-epoxy", Composite Structures, 18, 47-63. Lavoie, J. A., Soutis, C. and Morton, J., 2000. "Apparent strength scaling in continuous fiber composite laminates", Composite Science & Technology, 60, 283-299. Soutis, C , Fleck N.A. and Smith, P.A., 1991. "Failure prediction technique for compression loaded in carbon fiber-epoxy laminate with open hole " Journal of Composite Materials, 25, 1476-1498 Airbus Industrie Test Method, 1994. AITM 1.0008. Issue 2 Haberle J. G., 1991. "Strength and failure mechanics of unidirectional carbon fiber-reinforced plastics under axial compression", PhD Thesis, University of London, Soutis, C , 1989. "Compressive failure of notched carbon fiber-epoxy panels", PhD Thesis, University of Cambridge Slaughter, S., Fleck, N. A., Budiansky, B., 1992. "Micrbuckling of fiber composite: the role of multi-axial loading and creep", J. of Engineering Mater. Tech., 115(5), 308-313. Faupel, J. H., and Fisher, A. E., 1980. "Engineering Design: A Synthesis of Stress Analysis and Materials Engineering", 2nd Edition, John Wiley & Son, EMRC, 1992. NISAII-Users Manual, version5.2,

Environment Effect of Natural Sisal Fibre Reinforced Epoxy Composites Manufactured by Resin Transfer Molding 1

X. P. Zhang1*, Q. Yuan2, W. Ngatimin1, J. Whitbourn1 and L. Ye1 Center for Advanced Materials Technology, School of Aerospace, Mechanical and Mechatronic Engineering, the University of Sydney, NSW 2006, Australia 2 Division of Manufacturing and Infrastructure Technology, CSIRO, Heighett, Vic. 3190, Australia

ABSTRACT Moisture/water absorption in fibre reinforced composites and the subsequent influence on properties of the composites are very important issues, in particular for natural fibre reinforced epoxy composites. In this work, the moisture absorption feature in a natural sisal fibre reinforced epoxy composite fabricated using Resin Transfer Molding (RTM) technology was studied, and the degradation of mechanical properties of both composites with treated and untreated fibre was characterized. The results show that increasing the content of water (moisture) in the composite results in an apparent reduction in tensile strength and impact toughness, and the deterioration in microstructure. The fibre surface treatment is an efficient method to improve the water/moisture resistance of the composite. These results may provide a sound theoretical basis for natural fibre composite design and engineering applications. INTRODUCTION In most environments, the moisture (or water) can penetrate into composites, which may result in obvious changes in the constituents and acute deteriorations in their interaction. These changes often compromise the load-carrying properties of the composites with the level of degradation increasing as the absorbed moisture content increases. Moreover, from micro-bonding point of view, a serious effect caused by moisture-absorption is the debonding between the fibre and matrix, which leads to a reduction in load transfer potential and a fall in mechanical properties of the composites. Reduction in stiffness, strength and toughness, as well as changes in thermo-mechanical behavior can often be related to the amount of moisture absorbed by the composites. For these reasons, recent years many studies have been focusing on characterization of moisture absorption (or water uptake) and the subsequent influence on properties of the composites. Most of these studies were performed for carbon and glass fibre reinforced epoxy resin composites destined for use by the aeronautical and aerospace industry [1-4]. However, little is known about natural fibre reinforced epoxy matrix composites which have been employed in some civil applications such as in roofing, building materials, and even automotive parts. Using natural fibres as a replacement for glass or carbon fibres in the reinforcement of composites has attracted increasingly the attention from materials researchers and engineers. These natural fibres include wood, bamboo, sisal, lignocellulos, flax, Correspondence author. Tel: 61-2-93517146; fax: 61-2-93517060; Email: [email protected]

Natural Sisal Fibre Reinforced Epoxy Composites

89

pineapple leaf, coir etc. Use of these natural fibres in fabricating composites broadens the applications for these agricultural by-products. Due to the potentially biodegradable feature, these natural fibres reinforced composites are also called eco-composites or environmentally benign composites. Among these natural fibres, sisal fibre is one of the strongest naturalfibres,just after flax. Sisal is extracted from the leaves of the sisal plant. The sisal leaf contains mechanical, ribbon and xylem fibres. Of these, the mechanical fibres are the most useful part [5]. In this study, a natural sisal fibre reinforced epoxy matrix composite was fabricated using resin transfer molding (RTM) technology which has become increasingly interested for the composites manufacturing industry over the last decade or more. The composite samples were then subjected to an elevated temperature moisture environment to investigate the effects of the environment on composite mechanical properties and microstructure. The moisture absorption feature in the composite was characterized for both treated and untreated fibres. The mechanical properties of composites with different moisture contents were estimated, and the fracture mechanism of both dry and wet composites was studied. COMPOSITE FABRICATION AND EVALUATION Materials, RTM System and Composite Fabrication In this work, 0-90° plain woven sisal fibre was used in fabrication of the composite. The used epoxy system is commercial materials, i.e., Araldite F resin and HY951 hardener. The mixture ratio of resin and hardener is 1:9 (wt%). The volume fraction of sisal fibre was about 0.50. The RTM system used in this work is a HYPAJECT portable injection machine which consists of a control unit, a vacuum pump, connection hoses and a steel mold. In order to investigate the influences of fibre surface conditions on water absorption feature and the bonding quality between the fibre and the resin, as well as, subsequently, the mechanical properties of the composite, a chemical solution, containing 0.05% 3-aminopropyltrethoxysilane, was used to modify the fibre surface. During the chemical treatment, the woven sisal fibre mats were immersed in the solution for 2 minutes followed by a cleaning in acetone for 2 minutes. The sisal fibre mats were dried in an oven at 80 °C for about 20 hours. The injection mold had a geometry of 410 x 205 x 5 (length x width x height, mm), which may fabricate a composite panel with the maximum size about 400 x 200 x 5 mm3. The post curing process was carried out in an oven at 80°C for three days, and the Differential Scanning Calorimetry (DSC 2920, TAInstruments) tests were conducted to evaluate the residual active chemical group in the composite panels and confirm all composite panels fully cured. Estimation of Environment Effects To characterize the effects of environmental exposure on the natural fibre reinforced epoxy composites, the specimens were immersed in a stainless steel container full of water, which was placed an oven at 70 °C for a period of one to two weeks. Over this time the specimens were taken out periodically to weight so as to determine the amount of water absorbed by the specimens. The percentage uptake of water by weight, Rw, can be defined as follows: W-W ^ l 0 0 %

(l)

90

Natural Sisal Fibre Reinforced Epoxy Composites

where W is the measured specimen weight corresponding to different dwell times, and Wo the original dry specimen weight. Both W and Wa were measured by an electronic microbalance (model: BP-210D, Sartorius) with an accuracy of 0.01 mg. During the first three days, 2 measurements were conducted due to the rapid moisture uptake of the composite, and then followed by one reading per day for the remainder of the first week as well as for the second week (in this work, readings were taken only till the 9th day due to the approximate saturation of moisture absorption). After moisture measurements each day, three tensile samples and three impact samples were taken out for mechanical property tests. Mechanical Property Estimation Tests Mechanical property tests include: tensile strength and impact resistance of the composite. The dog-bone specimens (the total length of specimens is 150 mm; gauge part length and width are 50 and 10 mm, respectively; and the specimen thickness 5 mm) were used to evaluate the tensile strength of the composite. The tensile tests were conducted using an Instron 5567 test machine with a loading speed of 0.5mm/min, referring to ASTM D638-96 [6]. The impact test specimens had a dimension of 60 x 13 x 5 (length X width x thickness, mm) with a small V-notch (2 mm in depth) cut on one edge of the specimen. Impact tests were conducted on an Izod Impact tester according to ASTM D256-02 [7]. The fractographies of the composite samples were alanyzed using an optical microscope and an SEM.

RESULTS AND DISCUSSION Mechanical Properties and Microstructures of the Composites The fabricated composite panels were inspected by optical microscopy. The surface examinations show that the composite panels had very fine quality. The microstructure observation of the polished cross-section area shows that no air bubbles existed both on surface and interior, and there existed very satisfactory bonding between the fibre and the resin, typically see Figure 1. The results of tensile strength and energy absorption for both composites with and without fibre surface treatment are shown in Table 1. Clearly, the fibre surface treatment can increase the mechanical properties of the composite. >jd:i-_.i-ibv, iii

N:i;iir,tE fritr in

2 dirci-iim

Y-J:>-.nm

FIGURE 1 Microstructure of the fabricated sisal fibre reinforced epoxy composite

Natural Sisal Fibre Reinforced Epoxy Composites

91

TABLE I Mechanical properties of the composite at different conditions Mechanical properties Tensile strength (MPa) Surface treated No treatment Absorbed energy (J) Surface treated No treatment

Dry

Day 1

Day 2

Day 3

Day 4

Day 5

Day 6

Day 7

Day 8

39.2 34.7

36.1 30.2

33.4 27.9

31.5 25.3

29.4 23.4

27.2 22.6

25.4 21.3

24.5 19.5

23.1 18.1

1.72 1.54

1.52 1.34

1.32 1.23

1.20 1.12

1.06 0.96

0.96 0.87

0.84 0.75

0.75 0.67

0.71 0.62

Water Absorption and its Influence on Property of the Composite The percentage uptake of water in the sisal fibre reinforced epoxy composite, for either untreated and treated fibres, is plotted against dwell time as shown in Figure 2. Clearly, the rate of water uptake is fairly rapid in the early stage, this is, during the first few days. Thereafter, percentage uptake of water reaches nearly saturation. But the composite without fibre surface treatment shows a longer time to reach the saturation condition of water uptake. Moreover, the composite with fibre surface treated has a lower water absorption rate compared with that of untreated one. This means that the untreated sisal fibre composite can absorb more water than treated one. The results of tensile strength and impact energy absorption for the composites with and without fibre treatment are also shown in Table 1. It can be seen that the composite exposed to the elevated temperature and moisture environment has a lower tensile strength and impact resistance compared to the original dry one, and an increase in water uptake leads to decrease in both tensile strength and impact resistance. However, the water/moisture absorption effect can be reduced by the fibre surface treatment. The composite with the treated fibres has a better resistance to water uptake, thus a relatively less degradation of mechanical properties compared to the untreated one. The main reason is that the surface treatment of sisal fibres has changed the chemical function group from hydrophilic to hydrophobic, and improved the bonding between sisal fibres and polymer matrices, and thus reduced the moisture absorption in the composite. As known, the interface plays a predominant role on the load transfer between fibre and matrix. Good interfacial bonding is crucial for the fulfillment of the advantages of both reinforcing fibre and matrix. 2.00 -fiber surface untreated -fiber surface treated

1.60 0)

1c

1.20

o

0.80

0.40

0.00 0

1

2

3

4

5

6

7

8

9

10

Dwell time (day)

FIGURE 2 Results of water absorption rate versus dwell time (at 70 DC)

92

Natural Sisal Fibre Reinforced Epoxy Composites

Figure 3 shows optical microscope images of a broken specimen after fracture test, which was immersed in 70 °C hot water for 9 days. The failure of the fibre bundle shows a fracture with less ductility; and no obvious fibre pull-out, which generally presents the relatively satisfactory bonding quality between the fibre and matrix, was observed. The moisture absorption resulted in serious swelling of the fibres, as shown in Figure 3(b).

(a) Cross-section of the broken specimen

(b) A: Breakage of a fibre bundle

v- • (c) B: Debonding along a poor interface

(d) C: Debonding along a well-bonded interface

FIGURE 3 Microstructure observation of a broken specimen immersed in 70 °C water for 9 days (without fibre surface treatment)

The studies showed that the major content of natural sisal fibre is cellulose [8, 9] which is a hydrophilic glucan polymer consisting of a linear chain of 1, 4-p-bonded anhydroglucose unit [10]. This large numbers of hydroxyl groups result in the hydrophilic properties of sisal fibre, that is, sisal fibre is prone to absorbing moisture (or water). This results in a very poor interface between sisal fibre and the hydrophobic polymer matrices and very poor moisture absorption resistance [11]. For the composite samples exposed in an elevated temperature moisture environment, therefore, the main reasons for the degradation of mechanical properties

Natural Sisal Fibre Reinforced Epoxy Composites

93

and microstructure are as follows: (1) the swelling of the fibres due to the moisture absorption; (2) the moisture absorption deteriorates the bonding quality between the fibre and matrix, leading to a poor interfacial property between sisal fibres and polymer matrices caused by the hydrophilic properties of the sisal fibre and hydrophobic characteristics of organic polymers. The interfacial adhesion and resistance to moisture absorption can be improved by fibre surface treatment with appropriate chemical solutions. Thus the mechanical properties of the composite can also be improved.

CONCLUSIONS 1. There exists an obvious water uptake in natural sisal fibre reinforced epoxy composite. The rate of water uptake is fairly rapid in the early stage of dwelling, before it reaches a saturation condition. 2. The moisture absorption in sisal fibre reinforced epoxy composite results in serious property deterioration of the composite. 3. The main mechanisms of the water uptake induced property deterioration are the swelling of the fibres in moisture environment, and the degradation of the bonding quality between the fibre and matrix. 4. Appropriate fibre surface treatments could improve the bonding quality between the fibre and matrix, and decrease the moisture absorption, thus improve the mechanical properties of the composites.

REFERENCES 1.

Dewimille, B. and Bunsell, A. R. 1982. "The modeling of hydrothermal aging in glass fibre reinforced epoxy composites." J. Phys. D: Appl. Phys, 12: 2079-2091. 2. Vaddadi, P., Nakamura, T. and Singh, S. P. 2003. "Transient hydrothermal stresses in fiber reinforced composites: a heterogeneous characterization approach," Composites Part A: Applied Science and Manufacturing, 34: 719-730. 3. Vaddadi, P., "Nakamura, T. and Singh, R. P. 2003. "Inverse analysis for transient moisture diffusion through fiber-reinforced composites," Acta Materialia, 51: 177-193. 4. Kootsookos, A and Mouritz, A. P. 2004. "Seawater durability of glass- and carbon-polymer composites," Composites Science and Technology, 64 (in press, and web on-line). 5. Bisanda, E. T. N. and Ansell, M. P. 1992. "Properties of sisal-CNSL composites." Journal of Materials Science, 27: 1690-1700. 6. ASTM D638-96. Standard Test Method for Tensile Properties of Plastics. ASTM, Philadelphia. 7. ASTM D256-02. Standard Test Methods for Determining the Izod Pendulum Impact Resistance of Plastics. ASTM, ASTM, Philadelphia. 8. Joseph, K. and Thomas, S. 1996. "Effect of chemical treatment on the tensile properties of short sisal fibre-reinforced polyethylene composites." Polymer, 37: 5139-5149. 9. Paul, A., Joseph K. and Thomas, S. 1997. "Effect of surface treatments on the electrical properties of low-density polyethylene composites reinforced with short sisal fibers," Composites Science and Technology, 57: 67-79. 10. Li, H., Pawel, Z. and Per, F. 1987. "Cellulose fiber-polyester composites with reduced water sensitivity (1) - Chemical treatment and mechanical properties," Polymer Composites, 8: 199-207. 11. Li, Y., Mai, Y.W. and Ye, L. 2000. "Sisal fibre and its composites: a review of recent developments," Composites Science and Technology, 60: 2037-2055.

Effective Thermomechanical Properties of Interpenetrating-Structured Composites Matthew Tilbrook*, Robert Moon, Lyndal Rutgers and Mark Hoffman. School of Materials Science & Engineering, University of New South Wales, Australia

ABSTRACT Composites exhibiting an interpenetrating structure have received recent attention for their potential improved toughness properties. Numerous models for prediction of effective properties of composite materials have been developed, however these have tended to focus on traditional composite structures. Alumina/epoxy and alumina/aluminium composites with interpenetrating structures were produced via a multi-step infiltration process, and their mechanical properties investigated via the impulse excitation technique. Effective properties of these specimens and other interpenetrating-structured composites appear to be predicted most successfully using the effective medium approximation (EMA). The EMA does not specify internal geometry, but rather considers inclusions of each constituent phase situated within an effective medium, which is appropriate for the non-periodic structures of the composites investigated.

INTRODUCTION Composites exhibiting an interpenetrating structure, in which both phases are continuously connected networks, have received significant interest recently for applications requiring high toughness, temperature resistance or multifunctionality [1], Most models developed for effective property predictions have tended to focus on traditional composite structures: particle/matrix, fibre/matrix and laminates; though there are models which do not assume a particular internal geometry [2], That interpenetrating structures do not usually display a periodic regularity suggests that the non-geometryspecific models may be more valid. Thermoelastic properties of particular composites with interpenetrating structures have variously been predicted with the HashinShtrikmann bounds [3,4], the effective medium approximation (EMA) [5,6] and the Tuchinskii unit-cell model [7,8]. The effective medium approximation was found to be the most appropriate of these for predicting the elastic properties of alumina-epoxy composites [9]. The extreme contrast in elastic properties of these two phases (Table 1) has two implications: amplified variation between predictions from different models; and possible aberrant behaviour due to high strain mismatch at phase interfaces. In this paper, it is shown that the effective medium approximation is appropriate for a range of composites with interpenetrating structures. The results of mechanical property measurements, on alumina-epoxy and alumina-aluminium composites and porous alumina specimens, are presented. These results and those of other researchers for similarly structured composites are shown to be simulated well using the EMA. * Correspondence Author, School of Materials Science & Engineering, University of New South Wales, Sydney NSW 2052 Australia. Email: [email protected] Fax: (02) 9385 5956.

Thermomechanical Properties of Interpenetrating-Structured Composites

95

EFFECTIVE MEDIUM APPROXIMATION The effective medium approximation, or self-consistent approach, was derived from the work of Eshelby [10], Hill [11] and others. It is based on the assumption that regions of each phase within the composite may be treated as being embedded in an effective medium. This results in a set of equations relating constituent phase properties to effective properties via weight functions, dependent on volume fraction and assumed inclusion shapes.

{(Kn-K*)-TK) = 0

((Gn-G*)-TG} = 0

(1)

The strain relation tensors TKW and TQ W are determined from a matrix A which is defined in terms of Eshelby's inclusion shape tensor P and shape functions f(v|/), where \|/ is the shape parameter describing the ellipsoid aspect ratio (prolate for V|/ < 1, spherical for v|/ = 1, and oblate for \|/ > 1). These tensors and functions are given elsewhere [5,9]. The EMA does not specifically assume continuity of phases, which can lead to problems: porous structures with >50% porosity are predicted to have zero stiffness, for example. This issue must be considered when interpreting EMA predictions.

A

200 um _

*

150

B •« 100

I

EMA - porous (Si)

Expt - porous 0.1 0.2 0.3 0.4 Composition [Phase 2 Volume fraction]

0.7

0.75

0.8

0.S5

0.9

0.95

Volume Fraction of Alumina

FIGURE 3 Effective composite properties, (a) Young's Modulus for porous alumina specimens, before (o) and after (•) infiltration with epoxy, and EMA with \j/ = 8. (b) Alumina-aluminium - Young's modulus and thermal expansion, experiment (•) and refs [13] (•) and [14 ] (»,°), and EMA with ip = 10.

Figure 3(a) displays Young's modulus values for several alumina preforms at different stages of processing which are concordant with EMA predictions. These were specifically measured to enable calculation of stresses during processing. Results for alumina-aluminium obtained by the authors and others [13,14] are displayed in Figure 3(b) along with EMA predictions, demonstrating good correlation for Young's modulus and also for thermal expansion coefficient, as discussed by Hoffman et al [6]. FURTHER APPLICATIONS To demonstrate the general applicability of the Effective Medium Approximation method to composites with interpenetrating structures, EMA predictions have been obtained for several composite systems investigated experimentally by other researchers. Tuchinskii compared theoretical predictions, from his model for co-continuous structured composites, to experimental results for Fe-Pb, W-Cu and Ti-Mg composites [7]. As shown in Figure 4(a), the EMA predictions are very close to the measured values. The Tuchinskii model also provides good predictions, however these are given as upper and lower bounds, and hence are less precise. Both Tuchinskii and Jedamzik et al. showed that Tuchinskii's model could be used for tungsten-copper composites [7,8]. Figure 4(b) shows that the EMA and Tuchinskii's model provide very similar good predictions across a wide range of volume fractions. Wegner and Gibson have conducted a significant investigation into mechanical properties of interpenetrating-phase composites [13,16,17]. They produced and characterised steel-bronze and steel-resin composites, and simulated their mechanical response using a three-dimensional finite-element model. Whilst their computational

98

Thermomechanical Properties of Interpenetrating-Structured Composites

model provided good predictions, it would be very time-consuming to set up, and cannot be applied across the full range of compositions due to the geometric assumptions involved. As shown in Figure 5, the EMA provides sufficiently accurate predictions, which can be improved by considering porosity in some specimens, as discussed by Wegner and Gibson, or by varying the shape factor, \\i. The steel-resin composites are of particular interest, due to the extreme disparity in constituent properties, providing a useful comparison with results obtained for aluminaepoxy composites. The sharp drop-off in stiffness with increase in compliant phase volume fraction is even more significant in the steel-resin system, which could be due to differences in stress transfer at phase boundary interfaces. A good fit was obtained with a shape factor \\i = 10. This suggests that the optimum choice of shape factor may depend on the disparity in constituent material properties as well as the geometry of the composite structure. As the appropriate choice of shape factor for a particular composite system is not clear however, experimental measurement of properties is recommended.

EMA Predictions Jedamzik ' Results

0.2 0.4 0.6 0.8 Volume Fraction (Second Phase)

0.2

0.4 0.6 0.8 Volume Fraction of Copper

FIGURE 4 Comparison of EMA predictions with experimental values and Tuchinskii model predictions for Young's modulus of (a) iron-lead and titanium-magnesium composites [7] and (b) tungsten-copper composites [7,8] with interpenetrating structures.

0.2

0.4

0.6

0.8

Volume Fraction of Steel

FIGURE 5 Comparison of EMA predictions with experimental measurements [15] of Young's modulus of steel-resin (o) and steel-bronze (•) composites with interpenetrating structures.

Thermomechanical Properties of Interpenetrating-Structured Composites

99

CONCLUSIONS The mechanical properties of composites with interpenetrating structures have been investigated. The measured elastic properties of alumina-epoxy and alumina-aluminium composites produced by infiltration were more precisely simulated with the effective medium approximation (EMA), than with other models. The EMA was also applied successfully to a range of composite systems investigated by other researchers, indicating a general suitability for use with interpenetrating structured composites. Investigation of composites with significantly disparate constituent properties allows improved differentiation between predictions from different models, which highlighted the general applicability of the EMA in preference to other models. For composites with significantly differing constituent properties, EMA predictions appear to be reasonably accurate, and are preferable to the model of Tuchinskii in this regard. The advantages of the EMA over the unit-cell model of Tuchinskii, or more involved models such as those of Wegner and Gibson, is that internal phase geometry is not specified and that exact values, rather than a range, are predicted. Nevertheless, experimental determination of properties for validation of models is recommended. REFERENCES 1. 2. 3. 4.

5. 6. 7. 8. 9. 10. 11. 12. 13. 14.

15. 16. 17.

Wegner L.D., LJ. Gibson, 2001. "The fracture toughness behaviour of interpenetrating phase composites"Int. J. Mech. Sci. 43:1771-1791. Hashin Z., 1983. "Analysis of Composite Materials - A Survey," J. Appl. Mech. 50:481-505. Hashin, Z., S. Shtrikman, 1963. "A variational approach to the theory of the elastic behaviour of multiphase materials,"./ Mech. Phys. Solids 11:127-140. Torquato S, Yeong CLY, Rintoul MD, Milius DL, Aksay IA. Elastic Properties and Structure of Interpenetrating Boron Carbide/Aluminum Multiphase Composites. J Am Ceram Soc 1999; 82(5):1263-1268. Kreher W., W. Pompe, 1989. "Internal Stresses in Heterogeneous Solids," Akademie Verlag, Berlin. Hoffman M., S. Skirl, W. Pompe and J. Rodel, 1999. "Thermal Residual Strains and Stresses in A12O3/A1 Composites with Interpenetrating Networks," Acta Mater. 47(2):565-577. Tuchinskii, L.I., 1983. "Elastic Constants of Pseudoalloys with a Skeletal Structure," Porosh Metall 7(247):85-92 (Russian). Translated in: Powder Metallurgy and Metal Ceramics, Plenum. Jedamzik R., A. Neubrand, J. Rodel, 2000. "Characterisation of electrochemically processed graded tungsten/copper composites," Mat. Sci. Forum 308-311:782-787. Tilbrook M.T., RJ. Moon and M. Hoffman, 2003. "On the Mechanical Properties of Alumina-Epoxy Composites with an Interpenetrating Network Structure," Mat. Sci. Engng. A (submitted). Eshelby J.D., 1957. "The determination of the elastic field of an ellipsoidal inclusion, and related problems," Proc. Roy. Soc. A 349:376-396. Hill, R., 1965. "A self-consistent mechanics of composite materials," J. Mech. Phys. Sol. 13:213-222. Cichocki FR Cichocki, KP Trumble, J Rodel, "Tailored Porosity Gradients via Colloidal Infiltration of Compression Molded Sponges," J. Am. Ceram. Soc. 81[6] (1998) 1661-64. Wegner, L.D., LJ. Gibson, 2000. "The mechanical behaviour of interpenetrating phase composites I: modelling," Int. J. Mech. Sci. 42:925-942. H Prielipp, M Knechtel, N Claussen, SK Streiffer, H Miillejans, M Riihle, J Rodel, 1995. "Strength and fracture toughness of aluminum/alumina composites with interpenetrating networks," Mat. Sci. Eng.A 197:19-. Neubrand, A., T.-J. Chung, J. Rodel, E.D. Steffler and T. Fett, 2002. "Residual stresses in functionally graded plates," J. Mater. Res., 17 (11): 2912-2920. Wegner, L.D., LJ. Gibson, 2000. "The mechanical behaviour of interpenetrating phase composites II: a case study of a three-dimensionally printed material," Int. J. Mech. Sci. 42:943-964. Wegner, L.D., LJ. Gibson, 2001. "The mechanical behaviour of interpenetrating phase composites III: resin-impregnated porous stainless steel," Int. J. Mech. Sci. 43:1061-72.

Mechanical and Thermal Properties of Phenolic Composites Reinforced with Hybrid of Spun and Continuous Carbon Fabrics Tae Jin Kang*, Seung Jun Shin, Kyungho Jung and Young Jun Cho School of Materials Science and Engineering, Seoul National University, Korea

ABSTRACT The mechanical and thermal properties of continuous carbon fabric/spun carbon fabric interply hybrid composite materials have been studied. The hybrid composites with continuous carbon fabric of high tensile, flexural strength and spun carbon fabric of better interlaminar shear strength and lower thermal conductivity are investigated in terms of mechanical properties as well as thermal properties. Through hybridization, tensile strength and modulus of the spun reinforced composites were increased by about 28% and 20%, respectively. The hybrid composite also shows better interlaminar shear strength than continuous carbon reinforced composites. The thermal conductivity of the hybrid composite is lower approximately only 4~8% along the in-plane direction than that of the continuous carbon reinforced composite. The transverse thermal conductivity of the hybrid composite decreases with increasing continuous carbon fiber volume fraction. We predicted the thermal conductivity of textile composites using thermal-electrical analogy. The predicted thermal conductivities showed good agreement with experimental results. The erosion rate and insulation index were calculated through torch test. The spun reinforced composite has a higher insulation index than the continuous carbon reinforced composite and hybrid composites over the entire range of the back-face temperature of the specimen. The different stacking sequence has influence on the insulation index and erosion rate of hybrid composites.

INTRODUCTION Carbon/phenolic composites show excellent ablation resistance so that they have been widely applied to the thermal protection system for reentry vehicles or rocket engine components [1]. One of the most critical variables to govern the ablation performance of the composite is reinforcing materials [2], For solid rocket motor applications, one of the key requirements of the composite is low thermal conductivity to minimize the thickness of pyrolyzed carbon layer and temperature rise at the back-face of the composite [3]. The rayon-based carbon fibers were commonly used in the past due to the low thermal conductivity. In these days, however, it's not easy to get rayon-based carbon fibers because the manufacturing process causes serious environmental problems. So the PAN-based carbon fibers have gradually replaced the rayon-based carbon fibers. But PAN-based carbon fibers show high thermal conductivity in comparison with rayon-based carbon fibers. Corresponding Author, San 56-1, Shillim-dong, Kwanak-gu, 151-742, Korea, +82-2-885-1748, [email protected]

Mechanical and Thermal Properties of Phenolic Composites

101

One of the methods to reduce thermal conductivity of carbon/phenolic composites is using spun carbon yarn. Continuous carbon fibers show high thermal conductivity in the direction of fiber axis. There is feasibility of reducing thermal conductivity of composite using spun carbon fabric as reinforcement. Moreover, the protruded fibers on the staple yarn in spun composites play an important role in suppressing delamination due to the effect of fiber bridging as supplementary reinforcement [4-6]. We produced spun carbon yarn from staple Oxi-PAN fiber and fabricated carbon fabric using spun carbon yarn. We manufactured phenolic composites reinforced with spun carbon fabric, continuous carbon fabric and both of them. The mechanical and thermal properties of composite materials have been studied. The hybrid composites with continuous carbon fabric of high tensile, flexural strength and spun carbon fabric of better mterlaminar shear strength and lower thermal conductivity are investigated in terms of mechanical properties as well as thermal properties. EXPERIMENTAL Preparation of Composites The continuous and spun carbon fabric was prepared from stabilized PAN fiber supplied by Zoltek Co.. Spun stabilized PAN yarns were manufactured through woolen spinning and then woven into eight harness satin fabric. The maximum treatment temperature was maintained at 1100 °C, which is relatively lower value than conventional ones for the purpose of lowering the thermal conductivity of carbon fabric. Nitrogen gas was purged throughout carbonization process. The continuous carbon fabric was fabricated using low temperature heat-treated carbon filament tows. The spun carbon fabric was made from the spun Oxi-PAN fabric by carbonization. The resol-type phenolic resin KC-98, supplied from Kang Nam Chemical Co., was used as matrix for composite materials. The phenolic composites were fabricated at 150 °C for 2 hours. The debulking process at 105 °C for 30 minutes was done to remove possible entrapped air and voids in the resulting composite. Five kinds of composites are manufactured. Those are continuous fabric reinforced composite, spun fabric reinforced composite and three kinds of continuous/spun fabric interply hybrid composites. Characterization of Composites The tensile, flexural and short-beam shear tests were performed using MTS Sintech 10/GL based upon ASTM D3039, D790 and D2344, respectively. The thermal conductivities were measured by employing a comparative steady-state method using a tailor-made apparatus based on ASTM E-1225-87 [7]. Fig.l shows the schematic illustration of the apparatus for thermal conductivity measurement. The erosion rate and insulation index were calculated through torch test based on ASTME-285-80[8]. PREDICTION OF THERMAL CONDUCTIVITY We predicted the thermal conductivity of laminated composite using thermal-electrical analogy. We assumed that the cross-section of the fiber bundle was horse-track shape. We adopted several assumptions. First, we neglect the thermal contact resistance between fiber and matrix. Second, the heat flows rectilinearly and the heat flow

102

Mechanical and Thermal Properties of Phenolic Composites Pressure

Insulating/ Materials^

FIGURE 1 Schematic illustration of the apparatus for thermal conductivity Measurement

are parallel one another. Finally, the thermal conductivity of unit cell represents the whole composite structure. Figure 2 shows the idealized unit cell of eight harness satin fabric. The unit cell is divided into four kinds of elements for the analysis of thermal conductivity in transverse direction. On the other hand, six kinds of elements are defined for the analysis of in-plane direction thermal conductivity. Most elements are divided again into several regions. We derived the thermal conductivity of textile composites in transverse direction and in-plane direction. We can predict the thermal conductivity of textile composites in transverse and in-plane direction using Equations (1) and (2).

01 / /

//

A4W

// // •' to

/r is also exceeded considerably. Thus, more time is available during cooling to below the Tg (wet) for the network to equilibrate through molecular conformational reorganization before the network becomes frozen. The resulting network structure can be considered to be in equilibrium with a lower unoccupied volume (in the form of free volume and micro voids). Thus, these samples will absorb less moisture than those spiked at the Tmax. This is observed for all of these three resin blends. TABLE II Thermomechanical data for AroCy LlO/Epon 828, AroCy L10/DEN 431, and AroCy L10/MY 720 resin samples, after 4,000 hours conditioning at 96% R.H./50 ° C , and 17 thermal spikes

Spike Temp

AroCy LI0/DEN 431 TgE' Tg

(°C)

AroCy LlO/Epon 828 TgE' Tg (°C) (°C)

(°C)

(°C)

AroCy LI 0/MY 720 TgE' Tg,/ Tg2 Tg (°C)

As-cured Control 120 140 160 180

187 176 169 166 165 165

201 185 177 163 162 158

185 162 150 142 143 137

259 240 240 236/198 230/188 227/168

(°C) 172 156 144 136 130 128

237 199 194 177 166 144

The Tg of the cyanate ester/epoxy blend increased from 187°C to 259 °C when the epoxy resin was changed from the di-functional (Epon 828) to tetrafunctional (MY 720). This is an expected result as a higher cross-link density will be obtained for tetra-functional epoxy resin. Many investigations [9-10] have shown that, increasing the glass transition temperature and crosslink density results in more unoccupied volume in the resin network, which will result in higher moisture absorption. From Figure I it can be seen that the isothermal diffusion coefficient and moisture content increased with increase in epoxy functionality which supports the hypothesis

146

Moisture Absorption by Cyanate Ester Modified Epoxy Resin Matrices

Figure 1 Effect of epoxy functionality on glass transition temperature (Tg) of as-cured samples, moisture content (Mt), and isothermal diffusion coefficient (Dc) for resin samples conditioned for 4000 hoursat96%R.H./50°C.

The Tg for AroCy LlO/Epon 828 difunctional blended resin falls continuously with moisture absorption and thermal spiking to 140°C. Spiking to 160°C and 180°C caused no further reduction in Tg. Above the maximum moisture enhancement spike temperature of 140°C, Tg is unaffected. This concluded that spiking above 140 °C does not result in thermal degradation of resin samples. As can be seen from Figure I, the diffusion coefficient appeared to increase with the increase in the functionality of the epoxy resin in the blend. In Figure II it can be seen that the main tan5 peak for AroCy L10/MY 720 (tetra-functional blend) resin samples is split into two when thermally spiked to 140°C. The secondary peak, which is defined asTg 2 , occurred at 198°C. This was found to drop further at higher spiketemperatures. As shown in Figure II, Tg2 fell by 30 °C when spike temperature was increased from 140°C to 180°C while Tgj fell by 9 ° C . The large decrease in Tg2 for the 180 °C-spiked samples without a concomitant increase in moisture content suggests that thermally induced hydrolytic degradation is responsible for the reduction in Tg2 and not conventional plasticisation of the polymer matrix.

Moisture Absorption by Cyanate Ester Modified Epoxy Resin Matrices

147

0.35

0.00 40

90

140

190

240

Temperature (°C) • as-cured • control A 140 o180

Figure 2 DMTA tan 8 traces for AroCy L10/MY 720 as-cured, control samples, and samples spiked to 140 and 180 ° C conditioned at 96% R.H./50 ° C .

In a recent report [11] we showed that the epoxy resin concentration in the blend significantly influenced the likelihood of hydrolysis. FTIR data indicated that the aryl triazine crosslinks were reduced in concentration when thermally spiked at 160 and 180°C. The commercial version of MY 720, MY 721 has been shown to contain a range of synthesis by-products [12]. Therefore, when the network forms by reaction with the cyanate ester, it will have a complex structure, with potential for differential plasticisation and hydrolysis [13]. The efficiency of the network formation through triazine ring crosslinks will be compromised. It is therefore concluded that the tetra functional epoxy cured cyanate ester is more susceptible to hydration at temperatures above 140°C. CONCLUSIONS The concentration of moisture which is absorbed and the moisture diffusion coefficient were shown to increase with the functionality of the epoxy resin employed in the cyanate ester/epoxy blend. The cured resin from the tetra functional epoxy (MY 720) has a more polar network than that from the di-functional (Epon 828) epoxy and hence a greater affinity towards water, which led to a higher moisture concentration on approaching saturation isothermally and under thermal spiking. Splitting of the tan 5 peak after thermal spiking was observed for the resin blend containing the tetrafunctional epoxy. From the DMTA data, it can be concluded that degradation occurred as a result of thermal spiking at a temperature > 140°C and was postulated to be associated with the incomplete reaction of the functional groups during cure. Since more unreacted functional groups will be present in tetra-functional blend, there is a higher potential for hydrolytic degradation at the higher temperatures used for thermal spiking.

148

Moisture Absorption by Cyanate Ester Modified Epoxy Resin Matrices

ACKNOWLEDGEMENTS The authors acknowledge the University of Sheffield for a scholarship and ORS award to Dr. Sunil Karad. We also acknowledge BAe Systems Ltd. for additional financial support. We thank Advanced Composites Group for the supply of resins. We acknowledge valuable discussions with Mr. E. Shahidi (ACG Ltd.). REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

Browning, C. E., 1978. Polymer Engineering and Science, 18 (1):16. Jiming, Z., L. James. 1995. Composites Science and Technology, 53:57. Adams, R. D., M.M. Singh. 1996. Composites Science and Technology, 56:977. Bauer, J., M. Bauer. 1988. Ada Polymerica, 39:4. Lin, R.H. 2000. Journal of Polymer Science: Chem Edition, 38:934 Guo, B.C., W.W. Fu, D.M. Jia, Q.H. Qui, L. L. Wang. 2002. Polymer Composites, 10:237. Xiang Z. D., F.R. Jones. 1997. Composites Science and Technology, 57:451. Apicella A., R. Tessieri, C. De Cataldis. 1984. Journal of Membrane Science, 18:211. Jeffrey K., R.A. Pethric. 1994. European Polymer Journal, 30:153. Enns J.B., J.K. Gillham. 1983. Journal of Applied Polymer Science, 28:2831. Karad S.K., D. Attwood, F.R. Jones. 2002. Composites -.Part A, 33:1665. Gumen V.R., F.R. Jones. 2002 In Proceedings of 9th International Conference on Fibre Reinforced Composites A.G. Gibson , editor. 151. Gumen V.R., F.R. Jones, D. Attwood. 2001. Polymer, 42:5717.

New Epoxy Resins Based on Azomethine Groups for Potential Polymer Applications A. M.ISSAM*, H. P. S. ABDUL KHALIL and W. D. WAN ROSLI School of Industrial Technology. Universiti Sains Malaysia

ABSTRACT A series of new epoxy resins containing azomethine groups were synthesized by condensation reaction. The structures were characterized and confirmed by FTIR, 1H— NMR, 13C-NMR, UV and elemental analysis. Thermal stability and degradation behavior of these epoxy resins were examined by thermogravimetric analysis (TGA) and differential scanning calorimetry (DSC). The results of thermal analysis showed that, all resins possess high thermal stability. The epoxy resins based on p hydroxybenzaldehyde exhibited high thermal stability as compared to 4-hydoxy-3methoxybenzaldehyde. The resins produced showed good properties and can be used as matrix in polymer composites. INTRODUCTION Epoxy resins are among the most important thermosetting polymers in wide use as a matrix for fiber-based composites, structural adhesives, surface coatings, etc. [1,2]. Most of the commercially available epoxy resins are oligomers of DGEBA [1-3]. The epoxy resins are characterized by the presence of the oxirane group which is able to react with compounds possessing active hydrogen atoms, including amines, amides, or mercaptans. Various glycidyl esters, glycidyl amine derivatives and thioethers have been synthesized using this approach [3,4]. The synthesis, characterization, and polymerization of epoxy resins of various glycidyl ethers and esters containing azomethine groups have already been reported [5-7]. Owing to the relatively high thermal stability given by the presence of azomethine linkages [8-11], heat-resistant epoxides were prepared by reacting hydroxy and/or carboxy substituted azomethines or bis-azomethines with epichlorohydrin in the presence of a quaternary ammonium bromide as catalyst. The present paper deals with the synthesis of some new epoxy resin with azomethine linkages included in the main chain. The products, obtained by the direct reaction between DGEBA epoxy resin and various azomethine bisphenols, were characterized by both spectral and thermoanalytical techniques, the results being related to the chemical structure of the synthesized polymers.

* Corresponding Author, School of Industrial Technology, Universiti Sains Malaysia, 11800 Penang, [email protected]

150

New Epoxy Resins Based on Azomethine Groups

EXPERIMENTAL Materials P-phenylenediamine and ethylene diamine (Fluka Co.) were used without further purification. Vanillin, 4-hydroxy-3—methoxybenzaldehyde, p-hydroxybenzaldehyde and epichlorohydrine (Aldrich Co.) were used without further purification. Preparation of Azomethine Bisphenols Aldehyde (0.1 mol) was added dropwise to a solution of diamine (0.05 mol) in absolute ethanol. The mixture was refluxed for 6 h with stirring in 500 ml flask to allow complete reaction, followed by precipitation, filtration and washing several times with diethylether. The precipitate was finally dried for 24 h in a vacuum oven at 70 °C. Final purification was carried out by re-crystallization from 1-butanol and then dried for 24 h in a vacuum oven at 75 °C. Preparation of the Epoxy Resins Containing Azomethine Groups Preparation of the epoxy resins with azomethine groups in the main chain of the polymer was carried out in bulk using DGEBA epoxy resin and the azomethine compounds, synthesized as previously described. The reagents, taken in the molar ratio DGEBA/azomethine of 2:1, were first heated at 100 CC for 1 hr, then nbutylamine, used as a selective catalyst for the ring opening of the epoxide compound [12], was added. The mixture was stirred at 100 °C for 2-4 hr, and at 130 °C for 1 h to complete the polymerization process. The product, obtained as a solid glassy resin, was purified by dissolution in acetone, filtered several times and precipitated in toluene. Finally, the product was dried in vacuum at 80 °C for 10 h. The yields ranged from 75 to 80%. Instrumentation The FTIR spectra of the newly synthesized epoxy resins were recorded on Perkin Elmer 2000. ^ - N M R spectra were obtained using Brucker 300 MHz NMR spectrometer in CDCI3 as the solvent and TMS as the internal reference. The glass transition temperature ( Tg) was obtained by differential scanning calorimetry (DSC) by means of a Perkin Elmer DSC7 Series at a heating rate of 20 °C min"1 in a nitrogen atmosphere. The epoxy equivalent was evaluated by dissolution of the sample in pyridine (HC1) solution) and titration with aqueous NaOH solution in the presence of phenolphthalein, as previously described [13].

RESULTS AND DISCUSSION The general reaction yielding the epoxy resins containing azomethine linkages is given in Scheme 1.

New Epoxy Resins Based on Azomethine Groups

151

•0—CH 2 —CH—CH 2

2 CH,—CH—CH,—0-

amme

Scheme 1 The chemical nature of the A radical and elemental analysis of the synthesized epoxy resins are listed in Table 1. The ring opening of the epoxide compound (Scheme 1) is followed by the appearance of the secondary alcohol group [14-15]. The degree of the selectivity of the reaction depends on the active hydrogen compound used, on the catalyst and on reaction temperature. The use of the selective amine catalyst and reaction temperatures higher than 90 °C determines a 100-fold increase in the epoxide-phenol reaction [13]. The experimental conditions used and the experimental data obtained confirm the linear structure of the obtained epoxy resins containing azomethine. TABLE I Elemental analysis of the synthesized epoxy resins (1-4) Sample 1

Aldehyde

O=C^

V-OH /OCH3

2

3

O=C—/

\-OH

O=C—^

V-OH

A

-(CH2)2-

-(CH2)2O=C-/

^OH

Values in brackets are calculated

H (%)

N (%)

(11.23)

(11.23)

(11.23)

11.24

11.24

11.24

(11.23)

(11.23)

(11.23)

11.24

11.24

11.24

(11.23)

(11.23)

(11.23)

11.24

11.24

11.24

(11.23)

(11.23)

(11.23)

11.24

11.24

11.24

-n-n-

/OCH3 4

C (%)

152

New Epoxy Resins Based on Azomethine Groups

The FTIR spectra of the epoxy resins containing azomethine linkages showed the presence of the characteristic absorption bands at 900, 1200, and 1250 cm"1, attributed to the epoxy group. The 575 - 585 cm"1 and 1120 cm"1 bands correspond to the vibration of the ether group (-CH 2 -O-C 6 H 4 -), while the bands within the 1615 - 1635 cm"1 range assigned to the -CH=N- bond. The bands indicating the presence of the aromatic ring are placed at 3100 and 1500 cm"1, respectively. The 'H-NMR spectra of the synthesized epoxy resins showed a singlet at 1.7 ppm, specific to the methylenic protons, a multiplet situated in the 2.6-3.8 ppm interval for the protons of the epoxy group, a multiplet observed in the 6.5-7.3 ppm interval for the aromatic protons and a singlet at 8.69 ppm, assigned to the azomethine protons. Figure 1 shows a typical 'H-NMR spectrum recorded for sample 1. The UV spectra of the epoxides showed absorption bands placed in the 330-360 nm interval (characteristic to the epoxy groups). Compounds with methoxy groups in the backbone (2,4) showed a little blue shift as compared to the compounds (1,3).

FIGURE 1 'H - NMR spectrum of Sample 1

The DSC curves recorded with repeated heating-cooling cycles allowed the evaluation of Ts values of the synthesized epoxy resin. The Tg values are situated in the 35-60 °C temperature range. It is obvious that they depend on the structure of the epoxides, Tg increasing with increasing polymer molecular weight [16]. One might expect the polymers with azomethine segments in the main chain to have high thermal stability. The thermal behaviour of the synthesized epoxides was evaluated by dynamic TG experiments in nitrogen and air. Epoxides containing azomethine groups in the main chain showed an apparent •thermal stability higher than that of the DGEBA epoxy resin. The polymers 1 to 4 suffer a degradation starting from about 200 °C in air whereas about 250 °C in nitrogen. An increased decomposition rate being observed in the 300-450 °C temperature range, when the weight losses reach about 60-70%. The very close similarity of the thermograms suggests that the heat stability of the synthesized epoxy resin is not significantly influenced by the structure of the azomethines introduced in the main chain of the DGEBA resin. Considering that the compounds incorporating azomethine groups (i.e. mesogenic units) and flexible spacers in the main chain could

New Epoxy Resins Based on Azomethine Groups

153

possess both heat resistance and liquid crystalline properties [17-18], further investigations will concentrate on this aspect. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.

May, C. A. (ed.). 1988. Epoxy Resins : Chemistry and Technology, 2nd edn, Marcel Dekker, New York. Goulding, T. M. 1994. in Handbook of Adhesive Technology, (edited by A. Pizzi and K. L. Mattel), p.531. Marcel Dekker, New York. Lee, H. and Neville, K. 1967. Handbook of Epoxy Resins, Mc-Graw Hill, New York. Wright, C. D. and Muggee, J. M. 1986. Structural Adhesives: Chemistry and Technology (edited by S. R. Hartshorn), p. 113. Plenum Press, New York. Mikroyannidis, J. A. 1989. Makromol. Chem. 190, 1867. Mikroyannidis, J. A. 1990. J. Appl. Polym. Sci. 41, 2625. Mikroyannidis, J. A. 1991. Polym. Int. 25, 91. D'Alelio, G. F. D., Strazik, W. F., Feigel, D.M., and Schoenig, R. K. 1968. J. Macromol. SciChem. A2, 1457. Preston, J. 1982. Angew. Makromol. Chem. 109/110, 1. Kricheldorf, H. R. and Awe, J. 1989. Makromol. Chem. 190, 2579. Aharoni,S.M. 1988. Macromolecules 21, 1941. Cascaval, C. N., Mustata, F., and Rosu, D. 1993. Angew. Makromol. Chem. 209, 157. Alvey, F. B. 1969. J. Appl. Polym. Sci. 13, 1473. Shechter, L. and Wynstra, 1956. J. Ind. Engng. Chem. 48, 86. Enikolopyan, N. S., Markevitch, M. A., Sakhonenko, L. S., Rogovina, S. Z., and Oshmyan, V. G. 1982. J. Polym. Sci., Polym. Chem. Edn 20, 1231. Turi, E. A. 1981. Thermal Characterization of Polymeric Materials. Academic Press, New York. Sek, D. 1984. Eur. Polym. J. 20, 923. Tang, J. C. and Chang, T. C. 1994. Eur. Polym. J. 30, 1059.

Mechanical Properties of Rotational Moulded Polyethylene Composites - Experiments and Theories W. Yan, R.J.T. Lin* and D. Bhattacharyya Centre for Advanced Composite Materials Department of Mechanical Engineering The University of Auckland, Private Bag 92019, Auckland, New Zealand

ABSTRACT The dominant usage of linear polyethylene as the raw material for rotational moulding, a fast growing industrial method, has shown insufficient mechanical properties for certain applications where the strength and stiffness of a product are of importance. Worldwide rotational moulders have an urgent need for stronger, stiffer materials to be available; therefore, the introduction of reinforcements into rotomoulding process is attracting increasing attention. However, the incorporation of reinforcements in rotational moulding process has often shown difficulty of achieving a uniform distribution within the matrix material, which results in unsatisfactory mechanical properties. This paper describes an investigation using various particles of different volume fractions as the reinforcements, and verifying the various mechanical properties of rotomoulded products using different mathematical models. The results show that the Halpin-Tsai-Nielsen equation and the Nicolais-Narkis equation are well suited to predict the tensile modulus and tensile strength respectively for the rotomoulded particulate reinforced structure. A very good agreement can be achieved between the experimental results and the predictive models for the composites with particle volume fractions of 20% or less.

INTRODUCTION Rotational moulding is a method using plastic in powder or liquid form to produce hollow plastic products in a rotating mould, hi principle, rotational moulding comprises four stages: material charging, heating, cooling and de-moulding. The rapid development of this process as a mainstream plastics manufacturing method is attributed to its many advantages, such as the absence of residual stresses (no external pressure needed), low manufacturing and material costs, and the capability of manufacturing large and high quality products. However, some limitations also exist such as high mould cost, long processing time, limited mouldable material selection and mould shape complexity [1]. In order to improve the efficiency of rotational moulding, research has been carried out in various aspects, such as using special geometrical features for mould

Corresponding author, Centre for Advanced Composite Materials (CACM), Department of Mechanical Engineering, School of Engineering, University of Auckland, Private Bag 92019, Auckland, New Zealand; Phone: +64-9-3737599 Ext. 84543; Fax: +64-9-3737479; E-mail: ii.lin(fl),auckland.ac.nz

Rotational Moulded Polyethylene Composites

155 TM

design to provide stiffness for thin wall products [1,2], using ROTOLOG wireless temperature acquisition system for facilitating process control [2], identifying critical factors and enhancing heat transfer properties of moulds for the reduction of cycle time in rotational moulding [3,4]. However, incorporating reinforcements into the rotationally moulded components has created new challenges [5,6]. The first critical step for producing rotomoulded composite structures with better mechanical properties is to achieve uniformity of the reinforcement distribution. However, because of the high aspect ratio of most fibrous reinforcement, poor results have been reported when attempting to produce long glass fibre reinforced rotomoulded parts [6]. Other research has shown that the incorporation of short glass fibres or flax fibre can obtain some improvement of the mechanical properties by modifying the rotomoulding process, such as introduction of inner air pressure or double shot mould charging process [7-10]. However, the distribution of the fibres is still not completely satisfying. On the other hand, due to the low aspect ratio of particulate reinforcements, good distribution has been achieved in rotational moulding [11,12]. Some improvements in tensile modulus and impact strength can be obtained but there are some problems with the deterioration of the tensile strength [6-8,11-12]. With the addition of rigid particles to polymers or other matrices, a number of effects on mechanical properties, such as an increase in stiffness, reductions in tensile strength and the coefficient of thermal expansion, improvements of creep resistance and fracture toughness can be observed [13]. The degree of change in the mechanical properties of the composites can be affected by the size, shape, aspect ratio and the distribution of the reinforcing particles. In the case of non-spherical particles, the degree of orientation with respect to the applied stress is also important. Previously developed theories for modelling mechanical properties of particulate reinforced composites manufactured by different processes have shown good agreement with the experimental results obtained from composite parts [14-19]. However, for the pressure-free rotational moulding process, verification of the mechanical properties using these models has not been attempted yet. All these models are based on Einstein's equation for the viscosity of suspension with rigid spherical inclusions [20] to evaluate the reinforcing effect of particles in polymeric materials. However, they have their inherent weaknesses when predicting the mechanical properties of the composite products resulting in no single theory being suitable for analysing the overall composite material performance. In this study, for modelling the mechanical properties of rotomoulded products, several theories have been discussed and the predictions from these mathematical models have been compared with the experimental results obtained from composites rotomoulded with different kinds of particulate reinforcements. EXPERIMENTAL DETAILS The polymer matrix used in this study was Linear Polyethylene (LMDPE) Cotene 9042 with Melt Flow Index of 4.0g/10 min. This material is generally suitable for rotational moulding of large tanks and products that require a high degree of rigidity. The particulate reinforcements used are shown in Table I. The moulded parts had a wall thickness of 3.2 mm with the reinforcement volume fraction of up to 20 %. The sizes of the particles were in the same order of magnitude as that of the plastic powder. One laboratory designed rock-and-roll machine was used for rotational moulding experiments. The aluminium mould used had the dimensions of 100mm x 100mm x 220mm, and the wall thickness of 3mm. During the moulding process, the oven

156

Rotational Moulded Polyethylene Composites

temperature and internal air temperature were constantly monitored and recorded. The rotational speed ratio of the two axes was set at 10 to 3 with the main longitudinal axis rotating at 25 rpm. The oven temperature was set at 220°C and the heating process was stopped when the internal air temperature had reached 185°C. Forced air-cooling was used until the internal temperature dropped down to 40°C for demoulding. TABLE I Particulate reinforcement

A12O3

SiC

Irregular

Spherical glass beads with coupling agent (SGB+CA) Spherical

Irregular

Irregular

90-150

240

90-150

105-149

105-149

1

1.35

1

1.25

1.21

Spherical glass beads (SGB)

Recycled glass particles (RGP)

Shape Major dimension

Spherical

Aspect ratio

Reinforcements

The test specimens were cut out of the walls of rotomoulded boxes (94mm x 94mm x 214mm) and standard tests (stiffness - 0.1% secant modulus, BS2782; tensile strength ASTM D638-00 and impact strength, ASTM D6110-97) were conducted to determine the mechanical properties of the moulded composite specimens at room temperature. THEORETICAL MODELLING For polymer manufacturing processes such as injection moulding, compression moulding etc, there have been several theoretical models developed to predict the tensile moduli and tensile strengths of the particulate reinforced polymers, which are summarised in Table II. In general, it has been pointed out that, based on the assumption of a composite material failing when fracture is initiated from a stress concentration around a reinforcing particle, the theories for predicting the strength of a particulate reinforced system are less developed than those for predicting the moduli [20,21]. When applying these models for predicting the composite strength and modulus, the values of some important constants or parameters have to be decided. The first important parameter is the maximum packing factor of particulate reinforcement, <j>m, which can be obtained either theoretically or experimentally once the particle distribution condition is determined. In this study, random close packing is considered for deciding this factor. The other important parameter is the Einstein Coefficient, KE, which can be calculated by knowing the Poisson's ratio of the matrix material and the relative Einstein Coefficient ratio, KE/2.5, where 2.5 is the KE value of a material with a Poisson's ratio of 0.5 [20]. TABLE II Mathematical models for tensile modulus and tensile strength Name Guth model (ref[17])

Models For spherical particles

Ec=Ep(l + KEVf+U.Wf2) For non-spherical particles

Ec = Ep(l + 0.67aVf

+\.62a2V/)

Nomenclature Ec = Tensile modulus of the reinforced polymer Ep = Tensile modulus of the matrix KE = Einstein coefficient Vf= Reinforcement volume

Rotational Moulded Polyethylene Composites

Halpin-TsaiNielsen Model (ref [20])

157

1 + ABVf c

where: B -

\-B V l w,-*% w < ' V 0=1,2)

(5) where r\ is the resin shear viscosity and 8 is the resin rich layer thickness. Thus, the contact work can be written as:

(6) that Q(l) is the surface of ith ply. The work done in the edge of all plies can be given by:

-y2 (7) where tm,tnl and t are the normal, tangential and transverse tractions, respectively. The potential function due to previous defined constrains can be written as:

^ - t [ |»« Q

A

(8) where, T{>) ,p(i) and /l(Oare Lagrange multipliers, d^n is the component of strain rate tensor and a^is the component of unit vector in fiber direction of r'th layer. Integrating Eq. (8) over thickness yields: .n

(11)

This particular problem is chosen because interlaminar stresses arise only near the material discontinuities. Due to identical ply orientations in region B no free-edge effects exist at x=L. Also because the beam has length 2L and it has symmetric boundary conditions at x=0 no interlaminar stresses arise at this point.

Analytical Stress Analysis of Rotating Composite Beams

227

The material properties of the layers are taken to be those of a T300/5208 lamina [8]: £,=132 GPa, £ 2 =£ 3 =10.8 GPa , Gl2 = G13 = 5.65 GPa, G23= 3.38 GPa vl2 = v13 = 0.24 , v23=0.59, p = 1540kg/m3 (12) where the subscripts 1,2, and 3 indicate the on-axis (i.e., principal) material coordinates, hi what follows, the interlaminar stresses are determined by using Hooke's law with six numerical layers in each physical lamina (see [7]). The in-plane stress ax at z=h/4 in 0° plies in regions A and B is shown in Figure 2a. Also the in-plane stress <JX at z=h/4 in 90° ply in region A and 0° ply in region B is presented in Figure 2b. It is seen that stress distributions are discontinues near x=a. It is seen that there are close agreements between the present solutions and those obtained by the finite element method using ANSYS software [9].

4 3.5 3

5

2

D

"1.5

1.5

05

1 0.5

1 0.5 0

0.2

0.4

0.6

0

0.8

0.2

0.4

0.6

x/L

x/L

(a)

(b)

0.8

FIGURE 2 Distributions of in-plane normal stress ax at z=h/4 (a) in 0° plies in regions A and B and (b) in 90° ply in region A and 0° ply in region B.

0.1 0.05 0 2-0.05

s X-o.

[ :

/

~ ~ — ^

-0.1 O-

Present FEM

i

-0.15

s B

-0.2 -

-0.2 -0.25

i

0-

0.2

< • < < • • •

0.4

0.6

0.8

-0.3,

x/L

(a) FIGURE 3 Distributions of interlaminar stresses (a) az at the middle plane and (b) aa at z=h/4 along the 0790° interface in region A and along the 070° interface in region B.

228

Analytical Stress Analysis of Rotating Composite Beams

Distribution of interlaminar normal stress <JZ at the middle plane is shown in Figure 3a. The results show that there are sharp interlaminar stress gradients near the material discontinuity and decays away from the region of discontinuity as expected. The current solutions are seen to match the finite element solutions reasonably well except in the region close to the material discontinuity where the stresses have a steep gradient. Also Figure 3b illustrates the distribution of interlaminar shear stress <JXZ at z=h/4 along the 0°/90° interface in region A and along the 0°/0° interface in region B. It is noted that <JXZ at the free edge (x=L) meet the stress free boundary condition with a good approximation, even though this condition has not been enforced a priori. CONCLUSIONS A layerwise laminated beam theory are developed by using a layerwise laminated plate theory and it is used to predict interlaminar stresses in the vicinity of material discontinuities in rotating composite beams with general laminations. Governing equations of motion are obtained by using Hamilton's principle. The results obtained from this theory are compared with those obtained by a finite element method. The results indicate that there are severe interlaminar stresses in regions near the sudden transition of material properties (material discontinuities). These stresses may initiate heterogeneous damage in the forms of delamination and transverse cracking and may cause the damage to propagate to a substantial region of the beam, resulting in a significant loss of strength and stiffness. To this end, these stresses must be considered in design of such structures. APPENDIX The coefficients appearing in Eqs. (8) are defined as: [Aii] = [An]-[B23]T[D22T'[B2i],

[Bn] = [Bti]-[Dl2][D22T'[B2i]

[B,6 ] = [5 36 ] - [D26 ] [D22 r ' [5 23 ] , [ A , ] = [ A , ] - [Dl2 ] [D22 ]"' [Dl2 ] [A.] = [Di6l-[Dl2][D22r[D26],

[D66] = [D66]-[D26][D22y[D26]

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.

Pipes, R. B. and N. J. Pagano. 1970. "Interlamiiiar Stresses in Composite Laminates under Uniform Axial Extension," J. Compos. Mater., 4: 538-548. Hsu, P. W. and C. T. Herakovich. 1977. "Edge Effects in Angle-Ply Composite Laminates," J. Compos. Mater., 11: 422-428. Wang, A. S. D. and F. W. Crossman. 1977. "Some New Results on Edge Effect in Symmetric Composite Laminates," J. Compos. Mater., 11: 92-106. Pagano, N. J. 1978. "Free Edge Stress Fields in Composite Laminates," Int. J. Solids Structures, 14: 401-406. Wang, S. S. and I. Choi. 1982. "Boundary-Layer Effects in Composite Laminates: Part 2- Free-Edge Stress Solutions and Basic Characteristics," J. Appl. Meek, 49: 549-560. Bhat, N. V. and P. A. Lagace. 1994. "An Analytical Method for the Evaluation of Interlaminar Stresses Due to Material Discontinuities," J. Compos. Mater., 28(3): 190-210. Tahani, M. and A. Nosier. 2003. "Three-Dimensional Interlaminar Stress Analysis at Free Edges of General Cross-Ply Composite Laminates," Materials & Design, 24(2): 121-130. Herakovich, C. T. 1998. Mechanics of Fibrous Composites, John Wiley, New York. ANSYS, Release 5.4 UP19970828, SAS IP, Inc., 1997.

Thin-Plate Splines for Thick Composite Plate Analysis Antonio J. M. Ferreira* Departamento de Engenharia Mecanica e Gestao Industrial, Faculdade de Engenharia da Universidade do Porto, Portugal

ABSTRACT Composite laminated plates are a typical and relevant application of composite materials in various industries. Such structures are heterogeneous in the thickness direction and orthotropic in each lamina plane. Most of the methods for plate analysis are based on the finite element method. In this paper it is proposed to interpolate the system of partial differential equations by radial basis functions, in particular by thin-plate splines. The thin-plate splines (TPS) method represents an alternative to multiquadrics (MQ) radial basis functions. The TPS method does not rely on a shape-parameter like the MQ method, being this a more stable approach. Composite structures can be analysed by various sets of shear deformation theories: the classical Love-Kirchhoff plate theory, the first-order shear-deformation theory, the third-order shear deformation theory and layerwise theories. In this paper we apply the TPS method together with the third-order shear deformation theory for the analysis of moderately thick laminated plates. The methodology proves to be stable and accurate. INTRODUCTION Collocation with radial basis functions is a recent meshfree collocation method with global basis functions. The multiquadric method for the solution of partial differential equations (PDEs) was first introduced by Kansa in the early 1990s and showed exponential convergence for interpolation problems. In this paper we formulate and discuss the use of thin-plate splines radial basis functions in the solution of moderately thick laminated composite plates using a third-order shear deformation theory. Composite materials are a very important class of engineering materials with great properties and applications in a variety of complex structures, such as those typically found in space, automobile and civil applications. The correct design of such structures requires adequate stress analysis and in particular numerical tools. The proper modeling of composite laminates remains an open field of discussion and developments as a consequence of their complex behavior. It is very important to accurately determine the transverse shear stresses since they are the reason why delamination mechanisms are active. A proper laminate theory that accounts for shear deformation and an adequate numerical tool are needed in order to capture all such effects. Several laminate theories have been proposed in the literature. The simplest one is the classical laminate theory (CLT) [1] which is based on the Kirchhoff-Love theory of plates. The CLT neglects shear deformations and can lead to inaccurate results for moderately thick composite laminated or sandwich plates. The first-order shear deformation theory (FSDT) for laminated composite plates was developed as an * Corresponding Author, Departamento de Engenharia Mecanica e Gestao Industrial, FEUP, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal, Fax +351 22 9537352, email: [email protected]

Thin-plate Splines for Thick Composite Plate Analysis

230

extension of the theory of Mindlin [2] for isotropic plates which considers shear deformations and gives satisfactory results for a wide range of structures, in particular for moderately thick laminates [3-5]. The FSDT efficiency is dependent on a good choice of shear correction factors. This is not trivial and is dependent on the lamination scheme and on the deformation. Whitney [4] and Ferreira [5,6] have developed procedures for such calculation, based on cylindrical bending. The higherorder shear deformation theories (HSDT) have been developed to obviate the limitations of FSDT [7]. Following the research by the author with FSDT and MQ in the analysis of laminated composite plates [8], this paper addresses the application of thin-plate radial basis functions and HSDT to the analysis of moderately thick composite laminated plates. THIRD-ORDER THEORY OF PLATES The third-order theory of Reddy [7] is based on the same assumptions than the classical and first-order plate theories, except that the assumption of straightness and normality of a transverse normal after deformation is relaxed by expanding the displacements (u,v,w) as cubic functions of the thickness coordinate. The displacement field is then obtained as dw u(x,y,z)=u\x,y)+z(j)x '—f (1) ~dx dw (2) ' V" ) ^ 1" ) • V~ >J I ' " Yy 11,2 ~dy~ w(x,y,z)=w°(x,y) (3) where u and v are the inplane displacements at any point (x,y,z), u° and v° denote the inplane displacement of the point (x,y,0) on the midplane, w is the deflection, (j>x and <j)y are the rotations of the normals to the midplane about the y and x axes, respectively. The strain-displacement relationships are given as: ^xx

e

yy

y(0) /xy

city tiyy

'xy

du0 dx dv^ dy dun _ dvn

t XZ

fy?

(0) / xz

eO)

yy y(3) 1 xy

+ z2

/yz

ri 2)

y(l) /xy

dx dx

dx Cyy

~?>y

dx

dy

dy

dx

y(2) 1 xz

(4)

y(2) 1 yz

dx

-+ •

dy

+ z3

+ Z

%

(5)

dy

/xy

dy

dx

dw0 x

dx dwn

i

(6)

dy + dy = 3c,. Neglecting az for each orthotropic material layer, the stresswhere c, = strain relations in the fiber local coordinate system can be expressed as

Thin-plate Splines for Thick Composite Plate Analysis Qn

2,2

0

0

Qn

222

0

0

0

0

233

0

0

0

0

244

0

0

0

0

231

(7) Qs.

Here subscripts 1 and 2 are, respectively, the fiber and the normal to fiber inplane directions, 3 is the direction normal to the plate, and the reduced stiffness components, Qy are given by ii

=

~

ten

-

>

1 — 1/1/

I — V

V

=G 23 ;g 55 =G 3I ;v 21 = v ] 2 - ^

Qn =

(8)

in with £ , , £ 2 ) v12, G12, G23 and G31 are materials properties of the lamina. By performing a coordinate transformation, the stress-strain relations in the global x-y-z coordinate system can be obtained as X

Gy

Txy Tyz

Tzx

Qn

2,2

2,6

0

0 0

By

0

Yxy

245 255

Yyz

Qn

222

226

0

= 2,6

226

266

0

0

0

0

0

0

0

244

£

x

(9)

Yzx

The third-order theory of Reddy [7] satisfies zero transverse shear stresses on the bounding planes. The equations of equilibrium of the third-order theory are derived from the principle of virtual displacements as: dx dx

dy

3h2 { 8x2

^-2>o;'

\., d • + •

dx

(10)

dy2

dxdy dy

•-O

=0

(11)

with (12) In previous equations Q^R^M^^y, are the shear forces and bending moments, respectively. The Euler-Lagrange equations can be written in terms of the displacements by substituting strains and stress resultants into this equation. For example, equation (10) is given as

232

3h2

Thin-plate Splines for Thick Composite Plate Analysis

n

[dxdy2

dx2dy J

9h4

i

\dy2dx

' dx2dy ' "dy2dx2

hi this equation, stiffness components are obtained as A

a = tMl'J

= 4 ' 5 ;^ = T £ ^ W + . -zt3J,7 = 1,2,3,4,5 L ~z5JJ = 1.2.3JH, = ^£ei;W + 1 -^>-J = 1,2,3

(15)

hi (14) and (15), zk and zt+l are the lower and upper z coordinates of the A:'* layer. RADIAL BASIS FUNCTIONS The radial basis function method relies on the Euclidian distance between nodes and in some cases on a shape parameter, user-defined and object of various discussions. The influence of such parameters not only defines the RBF but also may provide illconditioned problems with inadequate solutions. To obviate the use of such parameter it is proposed in this paper the use of thin-plate splines. An RBF depends only on the distance to a center point x ; and is of the form g (Ik - x;. |h. Consider a set of nodes xux2,...,xN

EQCJ!",

The radial basis functions centered at xy. are defined as R\j=l,...,N

(16)

where x - x ; is the Euclidian norm. The thin-plate splines RBFs are given as = ||x - x.|2m logfx - Xjf ;m = 1,2,... (17) An important feature of the RBF method is that is does not require a grid. The only geometric properties needed in an RBF approximation are the pairwise distances between points. Working with higher dimensional problems is not difficult as distances are easy to compute in any number of space dimensions, hi this paper it is proposed to use Kama's unsymmetric collocation method [11]. Consider a boundary-valued problem with a domain Q c R" and a linear elliptic partial differential equation of the form Lu(x)=s(x)cR";Bu(x)]BQ =f(x)eR" (18) where dQ represents the boundary of the problem. We use points along the boundary (Xj,j = l,...,NB) and in the interior (xy., j = NB +1,...,N). Let the RBF interpolant to the solution w(x) be gj(x)

Thin-plate Splines for Thick Composite Plate Analysis

233

Collocation with the boundary data at the boundary points and with PDE at the interior points leads to equations

sB(x,c) =

(20)

= O(xi),i=NB

X -X;

(21)

where -^(x,), (x;) are the prescribed values at the boundary nodes and the function values at the interior nodes, respectively. This corresponds to a system of equations with an unsymmetric coefficient matrix, structured in matrix form as

(22)

NUMERICAL EXAMPLE A square laminate of side a and thickness h is composed of four equally thick layers oriented at (0°/90°/90°/0°). It is simply supported on all edges and subjected to a sinusoidal vertical pressure of the form:

. (nx\ . (ny\ pz =Psm — sin —

V a ) \ a ) being the origin of the coordinate system located at the lower left corner of the middle plane. The material properties are given as Ei = 25.0£2;G12 = Gn = 0.5E2;G23 = 0.2E2;v12 = 0.25 The numerical results are presented in table I, in a normalized form, as _ \02w(a/2,a/2,0)h3E2_ axx{al2,al2,hl2)h2 _ _ ayy(a/2,a/2,h/4)h2 ~<Jxx

=•

Pa' Pa2 Pa2 Tzx{0,a 12,first _ layer)h _ Txy(a,a,h/2)h Pa '* Pa2 In this table a laminated composite plate is analyzed with N = 11x11,15x15 and 21 x21 points. It can be seen that the present methodology is very accurate for the analysis of composite laminates. Both the transverse displacement, normal and transverse stresses are accurately predicted. Transverse shear stresses are calculated by the equilibrium equations. CONCLUSIONS In this paper the thin-plate splines radial basis function method and a third-order shear deformation theory were applied to the structural analysis of isotropic and symmetric laminated composite thick beams and plates. Results were compared with existing solutions showing excellent performance. Results obtained using an unsymmetric collocation strategy give very good agreement with available theories or previous results. This method, based on radial basis functions has very large potential for the

234

Thin-plate Splines for Thick Composite Plate Analysis

solution of structural problems, as a real meshless method, insensible to spatial dimension. TABLE I Composite plate a/h 100

20

10

4

N

w

11 15 21 Exact [12] 11 15 21 Exact [12] 11 15 21 Exact [12] 11 15 21 Exact [12]

0.5250 0.4575 0.4409 0.4347 0.5116 0.5029 0.5084 0.517 0.7050 0.7115 0.7133 0.743 1.8288 1.8509 1.8686 1.954

°* 0.6421 0.5650 0.5463 0.539 0.5465 0.5364 0.5423 0.543 0.5376 0.5452 0.5433 0.559 0.6289 0.6601 0.6614 0.720

a

y

0.3684 0.3434 0.3373 0.271 0.3663 0.3623 0.3657 0.309 0.4333 0.4368 0.4374 0.403 0.6175 0.6212 0.6261 0.666

0.4701 0.4295 0.4011 0.339 0.2149 0.2056 0.5906 0.328 0.1341 0.3553 0.3551 0.301 0.4422 0.1796 0.2208 0.270

0.0304 0.0238 0.0219 0.0214 0.0207 0.0221 0.0214 0.0230 0.0218 0.0238 0.5252 0.0276 0.0225 0.0315 0.0372 0.0467

REFERENCES I.

Reissner, E. and Stavsky, Y., 1961, "Bending and stretching of certain types of heterogeneous aelotropic elastic plates", J. Appl. Mech., 28 :402-412. 2 . Mindlin, R. D., 1951, "Influence of rotary inertia and shear in flexural motions of isotropic elastic plates", J. Appl. Mech., 18: 31-38. 3 . Yang, P. C. and Norris, C. H. and Stavsky, Y., 1966, "Elastic wave propagation in heterogeneous plates", Int. J. Solids and Structures, 2:665-684. 4 . Whitney, J. M., 1973, "Shear correction factors for orthotropic laminates under static load", J. Appl. Mech., 40:302-304. 5 . Ferreira, A. J. M. and Barbosa, J. T., 2000, "Buckling behaviour of composite shells", Composite Structures, 50:93-98. 6 . Ferreira, A. J. M. and Camanho, P. P. and Marques, A. T. and Fernandes, A. A., 2001, "Modelling of concrete beams reinforced with FRP re-bars", Composite Structures, 53:107-116. 7. Reddy, J. N., 1984, "A simple higher-order theory for laminated composite plates", J. Appl. Mech., 51:745-752. 8 . Ferreira, A. J. M., 2003, "A formulation of the multiquadric radial basis function method for the analysis of laminated composite plates", Composite Structures, 59(3):385-392. 9 . Reddy, J. N., 1997, Mechanics of laminated composite plates, CRC Press, New York 10 . Hardy, R. L., 1971, "Multiquadric equations of topography and other irregular surfaces", Geophys. Res., 176:1905-1915. II. Kansa, E. J., 1990, "Multiquadrics- A scattered data approximation scheme with applications to computational fluid dynamics. I: Surface approximations and partial derivative estimates", Comput. Math. Appl., 19(8/9): 127-145. 12 . Pagano, N. J., 1970, "Exact solutions for rectangular bidirectional composites and sandwich plates", J. Comp. Mater, 4, 20-34.

Part IV

Delamination

This page intentionally left blank

Evaluation of Fatigue Delamination Behavior in Hybrid Composite Material Using the Delamination Shape Parameters Sam-Hong Song Department of Mechanical Engineering, Korea University 1, 5ga, Anam-dong, Sungbuk-gu, Seoul 136-701, Korea Cheol-Woong Kim Mechanical System, fnduk Institute of Technology San 76, Wolgye-dong, Nowon-gu, Seoul 139-749, Korea Dong-Joon Oh Department of Mechanical Education, Andong National University 388, Songchun-dong, Andong, Kyoungbuk 769-749, Korea

ABSTRACT The applicability of the hybrid composite materials such as Al/GFRP laminates is restricted due to the frequent delamination of different materials at interlaminar. The previous researches showed that the major parameter to control the delamination of Al/GFRP laminates was a crack (a). On the other hand, it was also shown that a delamination width (b) could strongly effect on the delamination behavior. Therefore, the aim of this research is to define the delamination behavior using the above correlation. We obtained result as follow. The delamination aspect ratio (b/a) that was suggested from the increasing a-b relationship was decreased with increasing of a/W. The suggested the delamination area rate ((^DVAXDW) which represented the growth rate of delamination area (Ap), was going higher while a/Wv/as increasing.

INTRODUCTION Hybrid composite materials such as Al/GFRP laminates show a superior fatigue behavior to general metallic materials [1-4]. In spite of it, the reason why the applicable fields were restricted is the delamination caused between the Al layer and fiber/epoxy one. This delamination greatly decreases the fiber bridging effect. Therefore, the fatigue behavior of Al/GFRP laminates based on the delamination had been evaluated. The delamination was not determined just by the crack even though the first parameter to control the delamination was the crack. The second parameter to control the delamination was delamination width (b). However, it is difficult to fine the quantitative study of delamination using the a-b relationship. Therefore, the aim of this study is to evaluate the delamination behavior by a-b relationship. The details are as follows. 1) Analysis of crack (a) - delamination width (b) relationship. 2) Estimation of delamination aspect ratio (b/a) * Correspondence Author, Address : San76, Wolgye-dong, Nowon-gu, Seoul 139-749, Korea Fax : +82-2-921-8532, E-mail: woong25(a!korea.ac.kr

238

Evaluation of Fatigue Delamination Behavior 4-O10.5

2.1 Cross-section of A-A1 Drilled holes ••s

!

80 150

(a) Geometries of AI/GFRP laminates specimen

Bending f I J Moment! E 3

f

Saw-Cut

i_^..0^^=^-——J E B ^ S ^ ^ p l

Q Attending H Moment

Aluminum alloy UD Glass/epoxy: [0]2 (b) Cross-section of AI/GFRP laminates specimen

F I G U R E 1 Geometries and cross-section o f AI/GFRP laminates (unit: mm)

delamination area rate {(AD)N/(AD)Aii) relationship. 3) The effect of the b/a on the delamination shape factor (fs) and delamination growth rate (dAr/da). Through the above facts, the new parameters required for the delamination evaluation of the AI/GFRP laminate was proposed and the applied results were discussed. FABRICATION OF SPECIMEN AND EXPERIMENTAL METHOD Fabrication of AI/GFRP Laminates Specimen AI/GFRP laminates were manufactured as the 2/1 type that the unidirectional glass fiber/epoxy was inserted between two A15052 alloy sheets. During the curing, the postheating procedure was used to reflect the Differential Scanning Calorimeter (DSC) results of GFRP prepreg and it made the specimens more stable chemically. The geometries of AI/GFRP laminates were shown in Figure 1. Pre-crack was made at the low edge of the specimens by a wheel cutter and four holes were drilled at the fixing area of specimens. Fatigue Test Method The fatigue tests were performed by the bending & torsion fatigue testing machine (TB-10B, Shimadzu Co.) whose maximum moment was 98 N-m. The cyclic bending moment of 3.92 N-m was applied and the fatigue cracks were measured at 100 magnifications by the traveling microscope. C-scan (Mi-SCOPE exla, Hitachi Co.) was used to obtain the delamination images between the Al alloys sheet and glass-fiber/epoxy one and those of corresponding cyclic delamination were recorded. RESULTS AND DISCUSSION Relationship Between Crack Length (a) and Delamination Width (b) Unlike the monolithic Al alloy, AI/GFRP laminates did not show the sudden change of the crack growth with increasing of cycles [3]. During the second half of loading, this crack growth was the same as that of the first half of loading even though increasing of the crack length caused decreasing of ligament. Consequently, linear a-N relationship was obtained as Figure 2 (i). This phenomenon resulted from the stress bridging effect of fiber laminates. However, if the crack propagated the delamination field between the Al alloy layer and the glass-fiber/epoxy one was initiated and grown. When the major axes of delamination expansion direction were determined as x- andy- direction, it was know that the delamination length of x-axis direction was equal to the crack length. In the meantime, the delamination width (b) corresponding to ^-direction was broadened normal to the crack growth direction. From these facts, a lot of information about delamination behavior could be obtained using the a-b relationship. The triangle model (c=l) was selected for this study. If the triangle model was applied, the contour (c) and delamination area (AD) could be obtained easily by the a-b. The calculation of the b was carried out by the C-scan

Evaluation of Fatigue Delamination Behavior

239

c = 1, Triangle*

Bendin^Iiner

w r

yield ®yield ^^fiber

~

^f,design

>

^f,design

fiber

_ >°'liner ^ °'yield J

&

°fiber

fiber

>°'liner > °'yield

where Wmm , W, af4esign, aflber, ayidi, aliner are the possible maximum weight, the weight of the design point, the fiber directional strength considering the safety factor, the maximum fiber directional stress of the design point, the yield strength of the liner, and the maximum von Mises stress of the liner of the design point, respectively. Four design variables(the number of helical layers, the number of hoop layers, the winding angle of the cylinder part and the thickness of the liner) were set up and used in this optimal design because the goal of this application is the verification of the suggested algorithm and program. Each of them was separated to discrete values as many as 2 bits, 4

Optimal Design of Filament Wound Structures

269

bits, 5 bits and 4 bits in order to be applied to the genetic algorithm. From our previous researches, we determined the required parameters for genetic algorithm as follows; the population size was set as 100, the maximum number of generations as 100, the probability of crossover as 0.7, the probability of mutation as 0.1 and the tournament size for the genetic algorithm as 10. The initial seed value was generated randomly, and a total often optimal designs were performed. Table 1 shows the design results. Results of seven kinds were drawn. Among them, the best case, which satisfies the given design requirements, is a tank with a weight of 4.44 kg. TABLE I Design results Cases

Helical layer

Hoop layer

Winding angle

Liner

Weight

1

2

9

33.5°

1.9 mm

4.44 kg

2

3

10

34.5°

1.7 mm

4.47 kg

3

2

9

31.5°

1.9 mm

4.45 kg

4

3

10

28.0°

1.7 mm

4.50 kg

5

2

10

33.0°

1.9 mm

4.51kg

6

2

9

33.5°

1.9 mm

4.44 kg

7

2

9

33.5°

1.9 mm

4.44 kg

8

3

10

33.0°

1.7 mm

4.48 kg

9

3

9

34.0°

1.8 mm

4.58 kg

10

2

9

31.5°

1.9 mm

4.45 kg

CONCLUSION In this research, possible winding patterns considering windability and slippage were calculated using the semi-geodesic path algorithm. In addition, progressive failure analyses were performed to predict the behavior of filament wound structures. In particular, suitable element types and failure criteria for filament wound structures were studied, hi addition, on the basis of the semi-geodesic path algorithm and the finite element analysis method, an optimal design algorithm was suggested using the genetic algorithm. Finally, the developed design code was applied to a symmetric composite pressure vessel for verification. REFERENCES 1. 2. 3. 4. 5.

6.

J. Scholliers. 1992. "Robotic Filament Winding of Asymmetric Composite Parts," Ph. D Thesis (K.U.Leuven) M. Lossie. 1990, "Production Oriented Design of Filament Wound Composites," Ph. D Thesis (K.U.Leuven) P. A. Lowery. 1990. "Continued Fractions and the Derivation of Uniform-Coverage Filament Winding Patterns," SAMPE Journal, 26:57-64. Y. S. N. Reddy, C. M. Dakshina Moorthy, J. N. Reddy. 1995. "Non-linear Progressive Failure Analysis of Laminated Composite Plates," Int. J. Non-Linear Mechanics, 30(5):629-649. J. S. Park, C. S. Hong, C. G. Kim, C. U. Kim. 2002. "Analysis of Filament Wound Composite Structures Considering the Change of Winding Angles through the Thickness Direction," Composite Structures, 55:33-71. D. E. Goldberg. 1989. Genetic Algorithm in Search, Optimization, and Machine Learning, Addison-Wesley Publishing Company

Development of a Material Mixing Method for Topology Optimization of Multiple Material Structures Seog Young Han*, Soo Kyoung Lee, Jae Yong Park School of Mechanical Engineering, Hanyang University, 17 Haendang-dong, Sungdong-Gu, Seoul, 133-791, Korea

ABSTRACT This paper suggests a material mixing method to mix several materials in a structure. This method is based on ESO(Evolutionary Structural Optimization), which has been used to optimize topology of only one material structure. In this study, two criterions for material transformation and element removal are implemented for mixing several materials in a structure. Optimal topology for a multiple material structure can be obtained through repetitive application of the two criterions at each iteration. Two practical design examples of a short cantilever are presented to illustrate validity of the suggested material mixing method. It is found that the suggested method works very well and a multiple material structure has more stiffness than one material structure has under the same mass.

INTRODUCTION Topology optimization is very useful in making a conceptional design when considering weight and cost in the early design stage. An important development in topology optimization was made by Bendsoe and Kikuchi[l] who proposed the homogenization method, in which a material with an infinite number of microscale voids is introduced and the optimization problem is defined by seeking the optimal porosity of a porous medium using an optimality criterion. Mlejnek et al.[2] has accomplished shape and topology optimization using a simple energy method and a special type of function, that is, Kreisselmeier-Steinhauser function[3] for calculating effective properties. Recently, a simple method for shape and topology optimization, called Evolutionary Structural Optimization(ESO), has been proposed by Xie and Steven [4] and Chu[5], which is based on the concept of gradually removing redundant elements of the low stressed part of the material from a structure to achieve an optimal design. hi the area of MEMS, topology optimization techniques have also been actively applied. Especially, one of the thermal actuators, microgripper, has been designed by a bimorph structure[6] consisted of two materials with the different thermal strains, so that the improvement of its performance has been achieved by controlling the direction and magnitude of overall displacement effectively. hi this study, a material mixing method was suggested in order to obtain an optimal topology of a multiple material structure with the maximum stiffness under a static load based on ESO.

Correponding Author, School of Mechanical Engineering, Hanyang University, 17 Haendang-dong, Sungdong-Gu, Seoul, 133-791, Korea, FAX : 082-02-2298-46341, Email: [email protected]

Material Mixing Method for Topology Optimization

271

MATERIAL MIXING METHOD The explanation of the material mixing method for a bimaterial structure is given here in brief. Let the larger and the smaller stiffness and density of the two materials set material 1 and material 2, respectively. First of all, consider a design region made of material 1. Apply both the transformation and the removal lines for material transformation and element removal, respectively. As the first step, the elements having lower level of strain energy than the transformation line are transformed into material 2. As the second step, the elements having lower level of strain energy than the removal line which is established as the lower level of strain energy than the transformation line, are removed. An optimal topology can be obtained by iteration of this procedure at each iteration. Transformation and Removal Line TRANSFORMATION LINE The transformation line defined as eq. (1) is obtained from a journal paper [7] in this study. And the efficiency factor, a of strain energy in each element after stress analysis is defined as eq. (2). If a of a certain element is smaller than the transformation line, that is, eq. (3) is satisfied, the element is transformed from material 1 to material 2. TL = amm+Aax(

vol

5=1)"

(1)

V l

° ini,ial

« e = J ^r

(2)

a"

flexural demand

z\ 300 S

[Factored shear "

Test •• YEA (nonlinear) •-— M-C- analysis • Beam analysis J

500

10

300 -

S

200 .

200 3 a

tensile in composite

-2000

-1000

Test Girderbot Slab top

20 30 40 Midspan Displacement (mm)

•i * -I 100

, , , 1 1000 2000 3000 4000 Microstrain FEA M-C Analysis Beam analysis v Girderbot Gircierbol * Girderbot - Slab top -"— Slab top • Slab top 0

FIGURE 7 Global flexural test results of system assembly

At final failure, the load-displacement response of the system displayed essentially bi-linear response with a "softening point" (Figure 8), beyond which both cracking in the slab and slippage at deck-girder interface began to develop substantially. A quarter-span finite element model with idealized rigid connections for the deck-slab and deck-girder interfaces was used and results were seen to correlate closely to the test results (Figures 6, 7, and 8). c

1

Midspan DispU cement (in) 2 3 4

>t: Phase5 4000 _—•—Te • -Tc t: Ptase3 * FCA (Strain Criteria)

A

5

1000

•

4

£3000 600 e

/

J* Factored / flexural demanc

•J2000 -

1

1000

°c

400-a

A

t Factored shear demand , ! 1. .L-J— L-.!...!.-1.-J 40 60 8( 100 120 Midspan Displacement (mm)

20

200 ~ 140

Quarter span finite element model of two-girder system: Number of elements Number of nodes Number of D.O.F.'s

: 18363 :24023 : 90078

FIGURE 8 Final failure of system global flexural and finite element model

SUMMARY The characterization results for the structural behavior of the steel-free FRP/concrete modular slab-on-girder bridge system are presented in this paper. Good performance was obtained at both the full-scale component and system levels. Finite element analysis method and moment-curvature analysis method were found to be capable of predicting the structural response of the bridge components and the assembled system.

FRP Composite Modular System for Slab-on-Girder Bridges

293

ACKNOWLEDGMENTS The research presented in this paper was supported by the Federal Highway Administrations, State of California Governor's Office, and the California Department of Transportation, support of which is gratefully acknowledged.

REFERENCES 1. 2. 3.

4.

5. 6. 7.

Dunker K.R. and Rabbat, B.G. 1995. "Assessing infrastructure deficiencies: the case of highway bridges," Journal of Infrastructure Systems, 1(2): 100-119. Yost, J.R. and Schmeckpeper, E.R. 2001. "Strength and serviceability of FRP grid reinforced bridge deck," Journal of Bridge Engineering, 6(6):605-612. Reising, R.M.W., Shahrooz, B.M., Hunt, V.J., Lenett, M.S., Christopher S., Neumann, A.R., Helmicki, A.J., Miller, R.A., Kondury, S., and Morton, S. 2001. "Performance of five-span steel bridge with fiber-reinforced polymer composite deck panels," Transportation Research Record 1770, Paper No. 01-0337, pp.113-123. Salim, H.A., Davalos J.F., Qiao, and Barbero, E.J. ,1995. "Experimental and analytical evaluation of laminated composite box beams," 4tfh International SAMPE Symposium, May 8-11, pp.532539. Deskovic, N., Triantafillou, T.C., and Meier, U. 1995. "Innovative design of FRP combined with concrete: short-term behavior," Journal of Structural Engineering, 121(7): 1069-1078. Salim H.A. and Davalos J.F. 1999. "FRP composite short-span bridges: analysis, design and testing," Journal of Advanced Materials, 31(l):18-26. ABAQUS/Standard User's Manual, Version 5.8, 1998, Version 6.3, 2002.

Design of Composite-Antenna-Structures with High Electrical and Mechanical Performances Chi-Sang You and Woonbong Hwang Department of Mechanical Engineering, Pohang University of Science and Technology, Korea

ABSTRACT hi this paper we developed load-bearing outer surface that provides antenna performances, and it is termed Composite-Antenna-Structures(CAS). CAS is composite sandwich structure in which antenna part is inserted between honeycomb core and lower facesheet. Design procedure is presented including structure design, material selection and design of antenna elements in order to obtain high electrical and mechanical performances. Optimized honeycomb thickness is selected, in which maximum gain is obtained by resonance condition. Measured electrical performances show that CAS has wide bandwidth over 10% and higher gain by 3.5 dBi than initially designed antenna, and no doubt it has excellent mechanical performances by composite laminates and honeycomb cores. The CAS concept can be extended to give a useful guide for manufacturers of structural body panels as well as antenna designers, promising innovative future communication technology.

INTRODUCTION "Structural surface becomes an antenna." Structures, materials and antenna designers have recently joined forces to develop a new high payoff technology called smart skin or CLAS, for "conformal load-bearing antenna structure" [1-4]. The embedding of radio frequency (RP) antennas in load-bearing skin is a new approach to the integration of antennas into structural body panels. It emerged from the need to improve structural efficiency and antenna performances. It demands integrated product development from disparate technologies including structures, electronics, materials and manufacturing in order to generate a realistic design. The classic of division between structures and antennas is bridged in CLAS and the technical challenge is to satisfy structural and electrical requirements that often conflict. The present study aims to design electrically and structurally effective antenna structures for the next generation of structural surface technology [5,6]. This is termed Composite-Antenna-Structures (CAS). Design procedure is focused on high gain and wide bandwidth in the electrical part, and high strength, stiffness and environmental resistance in the mechanical part. Direct-feeding stacked patch antenna is used for the antenna performances and composite sandwich structure for structural performances. Measured electrical performances are presented.

Corresponding author, Department of Mechanical Engineering, Pohang University of Science and Technology, San 31, Pohang, Kyungbuk, 790-784, Republic of Korea, Tel: +82-54-279-2174, Fax: +82-54-279-5899, E-mail: [email protected]

Composite-Antenna-Structures

295

STRUCTURE AND MATERIALS The fundamental design concept for the CAS panel is an organic composite multi-layer sandwich panel in which microstrip antenna elements are inserted. As shown in Fig. 1 direct-feeding stacked patch antenna elements are used for the antenna performances. The basic panel layers are: two facesheets, two honeycomb cores and antenna elements with dielectrics. The facesheets carry a significant portion of the in-plane loads, contribute to overall panel buckling resistance, and provide low velocity impact and environmental resistance, however outer facesheet that is placed above upper patch provides signal loss by its high electrical loss (tan8). For the facesheet Glass/Epoxy[0/90]2s (UGN200, SK chemicals) of 1 mm is used. The honeycomb cores transmit shear load between layers induced from bending loads in the panel, support the outer facesheet against compression wrinkling, provide impact resistance and increase the overall panel buckling resistance. The thickness of honeycomb core contributes to the overall rigidity that they should be selected between panel thickness requirement and structural rigidity. Between outer facesheet and upper patch Nomex honeycomb core is inserted for mechanically high flexural strength. The thickness of this honeycomb core (h) will be selected for the most efficient antenna performances. For wideband performance, two radiating patches are used. Upper patch is placed on the Duroid 5880 (Rogers corp.) of 0.254 mm, and lower patch and feedline on the Duroid 5880 of 0.762 mm with ground plane on the other face. Upper patch and lower patch are separated by air gap that is provided by Nomex honeycomb of 2.54 mm. Though Duroid 5880 has good electrical properties of low dielectric constant and electrical loss, it doesn't contribute to structural performances. Consequently, CAS is composite sandwich structure in which antenna part is inserted between honeycomb core and lower facesheet. In order not to lower antenna efficiency by outer facesheet that is structural material with high electrical loss, the thickness of honeycomb core must be adjusted in the design procedure.

Glass/Epoxy[0/90]2s (1 mm) o= 4, tanS=0.03

Honeycomb (h mm) er=1.1

Upper Patch

Duroid 5880 (0.254 mm) er= 2.2, tanb=0.0009

Honeycomb (2.54 mm) zr=1.1

Lower Patch Feedline

Duroid 5880 (0.762 mm) er= 2.2, tan&=0.0009

Ground Glass/Epoxy[0/90]2s (1 mm) er= 4, fan8=0.03

FIGURE 1 Configuration of Composite-Antenna-Structures

Composite-Antenna-Structures

296 DESIGN PROCEDURE

In our design procedure, antenna performances are aimed for Ku-band satellite communication, frequency range between 11.7 and 12.75 GHz. At first, antenna elements are designed. The sizes of rectangular patches are 7.6 and 8 mm for upper and lower patches respectively. And then facesheets and honeycomb core are added. Fig. 2 and 3 show reflection coefficient and radiation pattern variations respectively for honeycomb thickness h by computer simulation and the performances of initially designed antenna are also presented by h = infinite. Gain in the broadside direction reaches maximum at h = 10 mm, at which it is approximately half a wavelength between lower patch and outer facesheet at central frequency 12.2 GHz, and reflection coefficient is also satisfied in the desired frequency range with bandwidth of 1.5 GHz.

1

CO

Nci

-10' >'

1

—

Arranging equation (14) for/,

Since F is the tension on a single fiber, F = as-

df-tf

(16)

<JS : Fiber Tensile Strength df : Fiber Robin Width tf : Fiber Robin Thickness

Combining equations (16) and (15), the total length,/, of reinforcing fiber robin is An r • P as

• df

• tf

Thus, the total volume, Vr, becomes Vr=l-df-tf=

4 n r3 • P

(18)

If the Fiber Volume Fraction of Filament Wound Pressure Vessel is vf, the total volume of the composite part is

Filament Wound Spherical Composite Pressure Vessel 4 ri p Vc=L-= « -

303

(19)

For composite thickness, t, and radius, r, the volume of the composite is Vc = An r2 • t

(20)

Equating equations (19) and (20), and arranging for t t

P-r =

—

(21)

This equation can easily be used by on-site engineers to figure out the thickness of the composite layer in near spherical composite pressure vessels. Even if it is not used as the final design value, it can be at least a good estimation to start with.

DESIGN EXAMPLE hi equation (21), substituting as = 4902 (MPa ), Vf = 0.5 P = 62 (MPa ) 0.5, r = Ro = 495 .7 (mm )

The composite thickness, t, becomes 62x495.7 t =

= 12.6 (mm) 4902x0.5

Since the outer radius of PE Liner is 497.5 mm and the composite thickness is 12.6 mm, the outer diameter of the final vessel becomes 1016.6mm which exceeds the required max, diameter of 994.2mm. Therefore, the vessel needs to have small portion of cylinder shape in the middle part of the vessel maintaining near spherical shape. Now, when the outer diameter of the vessel is 994.2mm and the Skirt's thickness is 5.4mm, the outer diameter of the cylinder is 983.4mm. Since the diameter of the vessel is reduced, the thickness of the composite could become thinner slightly. However, adopting the thickness of 12.6mm to this, the vessel design would be safer on the sidewall. The thickness of the PE liner is 6.7mm, and the inner diameter of the PE in the cylinder area is 944.8mm. If the length of the cylinder is L, the vessel's internal volume is

K, =„ ( »*£>.. I + ± ,

1

— 02 CD

U

a: i

n

I

5

i

1

10

- 1

15

20

Position across gradient, x [mm]

FIGURE 4 Material Gradient: Measured compositional distribution for a graded composite sample, along with stiffness and toughness distributions calculated from experiments on homogeneous specimens.

FINITE ELEMENT SIMULATION A finite element model has been developed to simulate the process of fatigue crack propagation in graded specimens. The commercial FEA software package ANSYS (Version 6.1, ANSYS Inc, Canonsburg, Pennsylvania) was utilised, with a number of modifications specific to graded materials. Full details of this model, and various aspects of its validation, are given by Tilbrook et al. [10] however several key aspects warrant mention here: 1) The spatial distributions of effective properties, in particular stiffness and fatigue resistance, were calculated from the compositional distribution and the experimental results for homogeneous composites described above. 2) Application of material stiffness gradient. As ANSYS does not allow graded elements, this was achieved by defining temperature-dependent properties, and assigning nodal temperatures individually, leading to a spatial temperature and property gradient. 3) Calculation of fracture parameters. Mode I and II stress intensity factors were determined from crack-opening and crack-sliding displacement values respectively for nodes near the crack-tip. Crack propagation was assumed to occur for: where Kc(x, Aa) was estimated to include crack-extension toughening effects. 4) Determination of crack propagation direction. The local symmetry criterion (Kn = 0) was used. A test kink was extended from the crack-tip and its angle varied to minimise Kii. The size of the test kink was 50 microns, which corresponded to

0.5

ter

O [0

0-0 0

1

2

2

3

4

5

6

7

8

LogNf

FIGURE 3 Normalized fatigue life curve for a unidirectional lamina under transverse tensile loading conditions.

FIGURE 4 Normalized fatigue life curve for a [0/90]s laminate under in-plane shear loading conditions.

TABLE I. Model parameters obtained from material characterization. Loading Mode (i) Trans. Tension (22) In-plane Shear (12)

Rsi [MPa] 52.56 136.33

OC;

16. 1 3 18. 6 3

P.

[MPa] 54.31 139.42

Ai

Bj

ai

bi

1.011 7 0 .124 5

0.093 7 0.178 5

9 .628 7 0 .160 0

7 .968 1 9.110 0

Multiaxial Fatigue Behavior of Unidirectional Laminates

353

1.2

6

i

£ 1.0 •a 2 $ 8

£ 0.8

1

,

0 ^

E ra

o-S0'6 a: 0.4 0.2

a22=9.6287 b22=0.1255

0.0 0.0

0.2

0.4

0.6

0.8

1.0

0.2

Normalized Cycles

0.4

0.6

0.8

1.0

Normalized Cycles

FIGURE 6 Normalized residual strength curve for a [0/90]s laminate under in-plane shear loadin

FIGURE 5 Normalized residual strength curve for a unidirectional lamina under transverse tensile loading.

STATISTICAL EVALUATION With all the model parameters obtained from materials characterization in Table 1, the statistical model can be applied to calculate the S-N curve of a lamina under a 30° off-axis loading. The calculated residual strength of the laminate subjected to 30° off-axis fatigue loading with maximum of 70% static strength is shown in Fig. 7. Similar calculations can be made for other applied stress levels. These results are normalized and shown in Fig. 8. If the normalized residual strength of unidirectional laminates with off-axis angle 9 is of the same form as Eq. (2): R e (n,q e ,K)-CT e

•[-

log(n)-log(0.25) f

(11)

vlog(Nffi)-log(0.25)j

where Rse, Re(n,ae,K) and OQ are the off-axis static strength, residual strength and applied stress, the value of parameters ae and be (9=30°) are determined to be 3.272 and 1.346, respectively. Therefore, if the fatigue life of an off-axis laminate under different applied stress levels is known, the normalized residual strength of the laminate can be determined based on Eq. (11). The fatigue life of unidirectional laminate with 30° off-axis angles is calculated from Eq. (6), shown in Fig. 9, which shows a good agreement with experiments [4].

E" 1S0 CD

Maximum Applied Stress - FVediction 95% - 1 0

1

2

3

4

5

6

LogN

FIGURE 7 Residual strength of [30] 16 offaxis laminate under an applied stress with maximum of 70% static strength.

0.2

0.4

0.6

0.8

Normalized Cycles

FIGURE 8 Normalized residual strength of [30] i6 off-axis laminate.

354

Multiaxial Fatigue Behavior of Unidirectional Laminates 1.0 I 0.8 0.6 0.4 •

n0.8

0.2 •

A 0.7 O 0.6

X0.5 0.0 i 0.0 0.2

§ ^l jF at i 0.4

i 0.6

i 0.8

Td 1.0

Normalized Cycles FIGURE 9 Calculated S-N curve and deviations for [30]]6 off-axis laminate.

SUMMARY 1) The residual strength and fatigue life of an off-axis laminate can be determined from the statistical model if the fatigue behavior of unidirectional lamina under longitudinal, transverse and in-plane shear loading are known through material characterization. 2) The normalized residual strength of an off-axis laminate against normalized fatigue cycles for different applied stress levels converges to a master curve, which can be used to determined the residual strength of the laminate under any applied stress levels if the corresponding fatigue life is known. 3) The fatigue life calculated for the unidirectional laminate with off-axis angle 30° agrees well with experimental data, indicating the usability of the statistical model. REFERENCES 1.

Garud, Y. S., 1981. "Multiaxial Fatigue: A Survey of the State of the Art", Journal of Testing and Evaluation, JTEVA, 9:165-178 2. Found, M. S., 1985. "A Review of the Multiaxial Fatigue Testing of Fiber Reinforced Plastics", Multiaxial Fatigue, ASTM STP 853, K. J. Miller and M. W. Brown, Eds., American Society for Testing and Materials, Philadelphia, pp.381-395 3. Shokrieh M. M. and L. B. Lessard, 1997 "Multiaxial Fatigue Behaviour of Unidirectional Plies Based on Uniaxial Fatigue Experimental: Part I. Modeling", International Journal of Fatigue, 4. Shokrieh, M. M. and L. B. Lessard, 1997, "Multiaxial Fatigue Behaviour of Unidirectional Plies Based on Uniaxial Fatigue Experimental: Part II. Experimental Evaluation", International Journal of Fatigue, 5. Sims, D. F. and V. H. Brogdon, 1977. "Fatigue Behavior of Composites Under Different Loading Modes", in Fatigue of Filamentary Composite Materials, ASTM STP-636, ASTM, Philadelphia, pp. 185-205 6. Harris, B., H. Feiter, R. Adam, R. F. Dickson and G. Fernando, 1990. "Fatigue Behaviour of Carbon Fibre Reinforced Plastics", Composites, 21:232-242 7. Diao, X.X, Lin Ye and Yiu-Wing Mai, 1995. "Statistical Prediction of Fatigue Failure of Fibre Reinforced Composite Materials", Applied Composite Materials, 2:153-173 8. Weibull, W. 1951. "A Statistical Distribution Function of Wide Applicability", Journal of Applied Mechanics, 18:293-297 9. Melsa, J. L. and A. P. Sage, 1973. An Introduction to Probability and Stochastic Processes, Prentice-Hall, Inc. Englewood Cliffs, New Jersey, 10. Adam, T., G. Fernando, R. F. Dickson, H. Reiter and B. Harris, 1989. "Fatigue Life Prediction for Hybrid Composites", International Journal of Fatigue, 11:233-237 11. Adam, T., R. F. Dickson, G. Fernando, B. Harris and H. Reiter, 1986. "The Fatigue Behaviour of Kevlar/Carbon Hybrid Composites", IMechE Conference Publications (Institution of Mechanical Engineers), 2:329-335 12. Fernando, G., T. Adam, B. Harris and H. Reiter, 1994. "Life Prediction for Fatigue of T800/5234 Carbon-Fibre Composites: I Constant-Amplitude Loading", International Journal of Fatigue, ' 16:523-532.

Part VII

FEM/Simulation

This page intentionally left blank

Simulation of Three-dimensional Flow in Compression Resin Transfer Molding by the Control Volume/Finite Element Method Akbar Shojaei* Chemical and Petroleum Engineering Department, Sharif University of Technology, Tehran, Iran Davood Boorboor and S. Reza Ghaffarian Polymer Engineering Department, Amirkabir University of Technology, Tehran, Iran

ABSTRACT In compression resin transfer molding (CRTM), resin injection and mold closing occur during the mold filling stage. In this paper, numerical simulation of threedimensional flow in compression resin transfer molding (CRTM) is presented. Numerical method is based on the control volume/finite element method (CVFEM) and the numerical algorithm to update the flow front at each time step is based on quasi-steady state approach. Numerical scheme presented in this article can be used to predict the flow progression, pressure distribution, mold clamping force in full threedimensional mold. Numerical example provided in the paper demonstrates the effectiveness of the developed numerical simulation in analyzing the CRTM process.

INTRODUCTION Due to advantages of Resin transfer molding (RTM), it increasingly becomes an important processing technology to fabricate net-shape polymer-matrix composites, ranging from nonstructural components with simple shapes, to structural parts with complex geometries. Despite various advantages of the RTM process, there are still some problems in the fabrication of part with large dimensions or high fiber content, including long mold filling time and presence of considerable void content in the final product. In order to overcome the above mentioned problems, various methodologies have been presented to modify the conventional RTM process. One important improvement in RTM is to combine the compression into the resin transfer molding. This process is called compression resin transfer molding (CRTM) and consists of resin injection and mold clamping or compression during the filling stage. Numerical simulation is known as effective tool to analyze filling stage in CRTM. Recently, numerical simulation of the CRTM process has been addressed in literature [1-4]. But, all of these simulations are restricted to two-dimensional flow in the CRTM mold. In some practical applications, two-dimensional flow assumption is no longer valid, and then three-dimensional simulation is needed to estimate the processing parameters accurately. Corresponding author. Chemical and Petroleum Engineering Department, Sharif University of Technology, Tehran 11365-9465, Iran. Fax: +98-21-6022853. Email: [email protected]

358

Three-dimensional Flow in Compression Resin Transfer Molding

The objective of the present article is to propose a numerical method for the full three-dimensional simulation of filling stage in the CRTM process. First, the current mathematical model of CRTM process, continuity equation, is extended to threedimensional domain. Then the control volume/finite element method (CV/FEM) is used to solve the governing equation. Various capabilities are provided in the computer code, permitting prediction of the flow front progression in threedimensional domain, pressure distribution and clamping force. Finally, application of three-dimensional simulation is demonstrated by providing a numerical case study. THEORETICAL MODELING During filling stage in the CRTM process, the liquid resin flows through the fibrous reinforcement. This can be assumed as flow through porous media. Mathematical models governing filling stage for pure injection process, namely RTM, have been well documented in literature [5]. Resin flow for both RTM and CRTM can be modeled by Darcy's law, but the continuity equation for CRTM needs to be modified for considering the mold closing effect. Resin flow through porous media described by Darcy's law is given as follows in three-dimensional domain: dP - dx xz dP yz (1) dy zz dP

Eq. 1 can be rewritten in tensor notation as:

=-~[K].\

(2)

where v is the superficial velocity vector with three components v, u and w in x, y and z directions, respectively, [K] represents the permeability tensor, Ky are its components, P is the pressure and u. is the resin viscosity. The general continuity equation for a deformable medium is given as follows: V.v=—L-^

dV dr

(3)

where dV is the infinite small elemental volume at time t and the term d(dv)/dt represents the rate of deformation of this elemental volume. This deformation rate depends on the mold closing speed and fiber preform deformation behavior. Eq. 3 had been first derived in soil mechanics [6,7], then it was rederived by Pham et al. [2] for CRTM in which the fiber deformation may be large. Substituting Darcy's law into Eq. 3 leads to single equation for pressure as: dV dt

(4)

Three-dimensional Flow in Compression Resin Transfer Molding

359

Eq. 4 is a general form of governing equation for three-dimensional flow in the CRTM process. By solving Eq. 4 with the aid of appropriate boundary conditions, one can obtain the pressure field in saturated zone and consequently estimate the flow front progression during filling stage. The total force exerted on the mold is composed of two components, including the resin pressure in saturated region and fiber bed stress. The total force may be estimated by the following model [2,4]: PdS

(5)

where Ftoal is total force exerted on the mold, S the surface area of the mold, P the resin pressure in saturated zone and af the fiber bed stress. NUMERICAL METHOD In the present formulation, control volume/finite element method (CV/FEM) is used to discretize the governing equation, i.e. Eq. 4. The volume of mold cavity is first divided into eight-node three-dimensional elements and then control volumes are constructed around a node as shown in Fig. 1.

^" -« '/:••

/

/-

J. /FIGURE 1 An eight node element and its subcontrol volumes used in this study

In order to track the flow front progression, a scalar parameter, /, called nodal fill fraction, showing the status of each control volume, is used. The fill fraction for each control volume represents the ratio of occupied volume by the resin to its total pore volume. During the filling stage, each control volume may have three different statuses:/= 1 for main region,/= 0 for empty region and 0—(* v ~ < • \—