MakeUp Pilot 5.13.0 With Full Version Crack

MakeUp Pilot 5.13.0 With Full Version Crack

The book deals with all aspects of advanced composite materials: what they are a master or mandrel for lay-up cracked weldment. during tool fabrication. Diana Chan 2019 General English Australia 0.0.0 2021-11-11 16:30:00 yes yes 16:9 HDTV yes stereo G The Chase The Chase G It's game on as teams of complete. apply to and form all Sections of the Contract Documents and the Work. Salvaged materials shall not be chipped, cracked, split.

You for: MakeUp Pilot 5.13.0 With Full Version Crack

MakeUp Pilot 5.13.0 With Full Version Crack
MakeUp Pilot 5.13.0 With Full Version Crack
MakeUp Pilot 5.13.0 With Full Version Crack

MakeUp Pilot 5.13.0 With Full Version Crack - opinion

Источник: []

Beauty Pilot 2.13.0 Multilingual

Beauty Pilot 2.13.0 Multilingual
Beauty Pilot 2.13.0 Multilingual US OUTLOOK

Oregon inferno spreading for miles

Biden grapples with virus

PORTLAND, Ore. — Firefighters scrambled Friday to control a raging inferno in southeastern Oregon that’s spreading miles a day in windy conditions, one of numerous wildfires across the U.S. West that are straining resources. Crews had to flee the fire lines late Thursday after a dangerous “fire cloud” started to collapse. An initial review Friday showed the Bootleg Fire destroyed 67 homes and 117 outbuildings overnight in one county. Authorities were still counting the losses in a second county. The blaze has forced 2,000 people to evacuate and is threatening 5,000 buildings that include homes and smaller structures in a rural area just north of the California border, fire spokeswoman Holly Krake said. Active flames are surging along 200 miles of the fire’s perimeter, she said. The Bootleg Fire is now 377 square miles — larger than the area of New York City — and mostly uncontained.

President says US outbreak a ‘pandemic of the unvaccinated’ ZEKE MILLER

Associated Press

WASHINGTON — Two weeks after celebrating America’s near “independence” from the coronavirus, President Joe Biden is confronting the worrying reality of rising cases and deaths — and the limitations of his ability to combat the persistent vaccine hesitance responsible for the summer backslide. Cases of COVID-19 have tripled over the past three weeks, and hospitalizations and deaths are rising among unvaccinated

people. While the rates are still sharply down from their January highs, officials are concerned by the reversing trendlines and what they consider needless illness and death. And cases are expected to continue to rise in coming weeks. While the national emergency may have faded, officials say the outbreak is now a more localized crisis in communities where not enough people have rolled up their sleeves. “Look, the only pandemic we have is among the unvaccinated,” Biden said Friday, echoing comments made earlier in the day by Dr. Rochelle Walensky, director of the Centers for Disease Control and Prevention. The rising numbers are being driven by large pockets of infection

among the more than 90 million eligible Americans who have yet to get shots. Just four states with low vaccination rates made up 40% of new cases last week, and nearly half of them came from Florida alone. However, there is little appetite in the White House for a return to broad mandates for masks or other measures, as 161 million Americans are already fully vaccinated. Walensky said that in low-vaccination areas with rising cases, “local policymakers might consider whether masking at that point would be something ... helpful for their community.” With three highly effective vaccines authorized for use in the U.S., the Biden administration believes the most effective way to attack the virus is not trying to

slow the spread with mass masking and such — something the U.S. showed it was not very good at last year — but to continue to press the importance of vaccinations. It’s no easy fix. Many Americans remain resistant or unmotivated to get shots. Surgeon General Vivek Murthy added that while government can play an important role, “this has got to be an ‘all of the above’ strategy with everybody in,” including schools, employers, technology companies and individuals. “They’re killing people,” Biden said Friday of social media companies, speaking a day after Murthy warned that false information about vaccines spreading on platforms like Facebook posed a public health risk to the nation.

Источник: []

Structural Composite Materials

Citation preview

Structural Composite Materials F.C. Campbell

ASM International® Materials Park, Ohio 44073-0002

Copyright © 2010 by ASM International® All rights reserved No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the written permission of the copyright owner. First printing, November 2010 Great care is taken in the compilation and production of this book, but it should be made clear that NO WARRANTIES, EXPRESS OR IMPLIED, INCLUDING, WITHOUT LIMITATION, WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, ARE GIVEN IN CONNECTION WITH THIS PUBLICATION. Although this information is believed to be accurate by ASM, ASM cannot guarantee that favorable results will be obtained from the use of this publication alone. This publication is intended for use by persons having technical skill, at their sole discretion and risk. Since the conditions of product or material use are outside of ASM’s control, ASM assumes no liability or obligation in connection with any use of this information. No claim of any kind, whether as to products or information in this publication, and whether or not based on negligence, shall be greater in amount than the purchase price of this product or publication in respect of which damages are claimed. THE REMEDY HEREBY PROVIDED SHALL BE THE EXCLUSIVE AND SOLE REMEDY OF BUYER, AND IN NO EVENT SHALL EITHER PARTY BE LIABLE FOR SPECIAL, INDIRECT OR CONSEQUENTIAL DAMAGES WHETHER OR NOT CAUSED BY OR RESULTING FROM THE NEGLIGENCE OF SUCH PARTY. As with any material, evaluation of the material under end-use conditions prior to specification is essential. Therefore, specific testing under actual conditions is recommended. Nothing contained in this book shall be construed as a grant of any right of manufacture, sale, use, or reproduction, in connection with any method, process, apparatus, product, composition, or system, whether or not covered by letters patent, copyright, or trademark, and nothing contained in this book shall be construed as a defense against any alleged infringement of letters patent, copyright, or trademark, or as a defense against liability for such infringement. Comments, criticisms, and suggestions are invited, and should be forwarded to ASM International. Prepared under the direction of the ASM International Technical Book Committee (2009–2010), Michael J. Pfeifer, Chair. ASM International staff who worked on this project include Scott Henry, Senior Manager, Content Development & Publishing; Steven R. Lampman, Content Developer; Eileen De Guire, Senior Content Developer; Ed Kubel, Technical Editor; Ann Britton, Editorial Assistant; Bonnie Sanders, Manager of Production; Madrid Tramble, Senior Production Coordinator; Diane Whitelaw, Production Coordinator; and Patricia Conti, Production Coordinator. Library of Congress Control Number: 2010937090 ISBN-13: 978-1-61503-037-8 ISBN-10: 0-61503-037-9 SAN: 204-7586 ASM International® Materials Park, OH 44073-0002 Printed in the United States of America

This book is dedicated to my youngest granddaughter, Matilda, who is so little yet is so brave and strong.

“This page left intentionally blank.”

Contents Preface About the Author

xi xv

Chapter 1 Introduction to Composite Materials 1.1  Isotropic, Anisotropic, and Orthotropic Materials 1.2  Laminates 1.3  Fundamental Property Relationships 1.4  Composites versus Metallics 1.5  Advantages and Disadvantages of Composite Materials 1.6  Applications

1 4 7 8 10 14 18

Chapter 2 Fibers and Reinforcements 2.1  Fiber Terminology 2.2  Strength of Fibers 2.3  Glass Fibers 2.4  Aramid Fibers 2.5  Ultra-High Molecular Weight Polyethylene Fibers 2.6  Carbon and Graphite Fibers 2.7  Woven Fabrics 2.8  Reinforced Mats 2.9  Chopped Fibers 2.10  Prepreg Manufacturing

31 31 32 33 39 41 42 49 52 52 52

Chapter 3 Matrix Resin Systems 3.1  Thermosets 3.2  Polyester Resins 3.3  Epoxy Resins 3.4  Bismaleimide Resins 3.5  Cyanate Ester Resins 3.6  Polyimide Resins 3.7  Phenolic Resins 3.8  Toughened Thermosets 3.9  Thermoplastics 3.9.1  Thermoplastic Composite Matrices 3.9.2  Thermoplastic Composite Product Forms 3.10  Quality Control Methods 3.10.1  Chemical Testing 3.10.2  Rheological Testing 3.10.3  Thermal Analysis 3.10.4  Glass Transition Temperature 3.11  Summary

63 64 65 67 70 71 72 74 75 81 82 87 90 91 92 94 97 99

vi  /  Contents

Chapter 4 Fabrication Tooling 4.1  General Considerations 4.2  Thermal Management 4.3  Tool Fabrication

101 101 104 111

Chapter 5 Thermoset Composite Fabrication Processes 5.0  Lay-up Processes 5.1  Wet Lay-Up 5.2  Prepreg Lay-Up 5.2.1  Manual Lay-Up 5.2.2  Flat Ply Collation and Vacuum Forming 5.2.3  Roll or Tape Wrapping 5.2.4  Automated Methods 5.2.5  Vacuum Bagging 5.2.6  Curing 5.3  Low-Temperature Curing/Vacuum Bag Systems 5.4  Filament Winding 5.5  Liquid Molding 5.5.1  Preform Technology 5.5.2  Resin Injection 5.5.3  Priform Process 5.5.4  RTM Curing 5.5.5  RTM Tooling 5.5.6  RTM Defects 5.5.7  Vacuum-Assisted Resin Transfer Molding 5.6  Resin Film Infusion 5.7  Pultrusion

119 119 119 122 123 124 125 125 131 133 137 141 146 148 162 164 166 167 170 172 174 175

Chapter 6 Thermoplastic Composite Fabrication Processes 6.1  Thermoplastic Consolidation 6.2  Thermoforming 6.3  Thermoplastic Joining

183 183 186 192

Chapter 7 Processing Science of Polymer Matrix Composites 7.1  Kinetics 7.2  Viscosity 7.3  Heat Transfer 7.4  Resin Flow 7.4.1  Hydrostatic Resin Pressure Studies 7.4.2  Resin Flow Modeling 7.5  Voids and Porosity 7.5.1  Condensation-Curing Systems 7.6  Residual Curing Stresses 7.7  Cure Monitoring Techniques

201 202 206 207 209 214 217 219 226 226 232

Chapter 8 Adhesive Bonding 8.1  Theory of Adhesion 8.2  Surface Preparation 8.2.1  Composite Surface Preparation 8.2.2  Aluminum Surface Preparation 8.2.3  Titanium Surface Preparation 8.2.4  Aluminum and Titanium Primers 8.3  Epoxy Adhesives 8.3.1 Two-Part Room-Temperature Curing Epoxy Liquid and Paste Adhesives 8.3.2  Epoxy Film Adhesives

235 235 235 237 239 242 243 244 245 247

Contents  /  vii 

8.4  Bonding Procedures 8.4.1  Prekitting of Adherends 8.4.2  Prefit Evaluation 8.4.3  Adhesive Application 8.4.4  Bondline Thickness Control 8.4.5  Bonding

248 249 249 250 251 252

Chapter 9 Sandwich and Integral Cocured Structure 9.1  Sandwich Structure 9.2  Honeycomb Core Sandwich Structure 9.2.1  Honeycomb Processing 9.2.2  Cocured Honeycomb Assemblies 9.3  Foam Cores 9.3.1  Syntactic Core 9.4  Integrally Cocured Unitized Structure

255 255 255 264 267 271 272 273

Chapter 10 Discontinuous-Fiber Composites 10.1  Fiber Length and Orientation 10.2  Discontinuous-Fiber Composite Mechanics 10.3  Fabrication Methods 10.4  Spray-Up 10.5  Compression Molding 10.5.1  Thermoset Compression Molding 10.5.2  Thermoplastic Compression Molding 10.6  Structural Reaction Injection Molding 10.7  Injection Molding 10.7.1  Thermoplastic Injection Molding 10.7.2  Thermoset Injection Molding

285 285 287 289 289 290 290 295 296 297 298 304

Chapter 11 Machining and Assembly 11.1  Trimming and Machining Operations 11.2  General Assembly Considerations 11.3  Hole Preparation 11.3.1  Manual Drilling 11.3.2  Power Feed Drilling 11.3.3  Automated Drilling 11.3.4  Drill Bit Geometries 11.3.5  Reaming 11.3.6  Countersinking 11.4  Fastener Selection and Installation 11.4.1  Special Considerations for Composite Joints 11.4.2  Solid Rivets 11.4.3  Pin and Collar Fasteners 11.4.4  Bolts and Nuts 11.4.5  Blind Fasteners 11.4.6  Interference-Fit Fasteners 11.5  Sealing and Painting

307 307 309 311 311 314 315 316 317 317 318 320 322 323 323 326 328 329

Chapter 12 Nondestructive Inspection 12.1  Visual Inspection 12.2  Ultrasonic Inspection 12.3  Portable Equipment 12.4  Radiographic Inspection 12.5  Thermographic Inspection

333 333 335 341 342 345

viii  /  Contents

Chapter 13 Mechanical Property Test Methods 13.1  Specimen Preparation 13.2  Flexure Testing 13.3  Tension Testing 13.4  Compression Testing 13.5  Shear Testing 13.6  Open-Hole Tension and Compression 13.7  Bolt Bearing Strength 13.8  Flatwise Tension Test 13.9  Compression Strength After Impact 13.10  Fracture Toughness Testing 13.11  Adhesive Shear Testing 13.12  Adhesive Peel Testing 13.13  Honeycomb Flatwise Tension 13.14  Environmental Conditioning 13.15  Data Analysis

351 351 352 353 354 356 357 358 361 361 362 364 364 367 367 369

Chapter 14 Composite Mechanical Properties 14.1  Glass Fiber Composites 14.2  Aramid Fiber Composites 14.3  Carbon Fiber Composites 14.4  Fatigue 14.5  Delaminations and Impact Resistance 14.6  Effects of Defects 14.6.1  Voids and Porosity 14.6.2  Fiber Distortion 14.6.3  Fastener Hole Defects

373 374 376 379 383 388 393 393 397 398

Chapter 15 Environmental Degradation 15.1  Moisture Absorption 15.2  Fluids 15.3  Ultraviolet Radiation and Erosion 15.4  Lightning Strikes 15.5  Thermo-Oxidative Stability 15.6  Heat Damage 15.7  Flammability

401 401 411 411 412 415 416 417

Chapter 16 Structural Analysis 16.1  Lamina or Ply Fundamentals 16.2 Stress-Strain Relationships for a Single Ply Loaded Parallel to the Material Axes (θ = 0° or 90°) 16.3 Stress-Strain Relationships for a Single Ply Loaded Off-Axis to the Material Axes (θ ≠ 0° or 90°) 16.4  Laminates and Laminate Notations 16.5  Laminate Analysis—Classical Lamination Theory 16.6  Interlaminar Free-Edge Stresses 16.7  Failure Theories 16.8  Concluding Remarks

421 421

Chapter 17 Structural Joints—Bolted and Bonded 17.1  Mechanically Fastened Joints 17.2  Mechanically Fastened Joint Analysis 17.3  Single-Hole Bolted Composite Joints 17.4  Multirow Bolted Composite Joints 17.5  Adhesive Bonding

449 449 450 455 459 463

425 427 429 430 439 440 446

Contents  /  ix 

17.6  Bonded Joint Design 17.7  Adhesive Shear Stress-Strain 17.8  Bonded Joint Design Considerations 17.9  Stepped-Lap Adhesively Bonded Joints 17.10  Bonded-Bolted Joints

464 466 475 479 481

Chapter 18 Design and Certification Considerations 18.1  Material Selection 18.2  Fiber Selection 18.3  Product Form Selection 18.3.1  Discontinuous-Fiber Product Forms 18.3.2  Continuous-Fiber Product Forms 18.4  Matrix Selection 18.5  Fabrication Process Selection 18.5.1  Discontinuous-Fiber Processes 18.5.2  Continuous-Fiber Processes 18.6  Trade Studies 18.7  Building Block Approach 18.8  Design Allowables 18.9  Design Guidelines 18.10  Damage Tolerance Considerations 18.11  Environmental Sensitivity Considerations

489 489 490 491 492 493 494 496 496 497 498 499 501 503 508 512

Chapter 19 Repair 19.1  Fill Repairs 19.2  Injection Repairs 19.3  Bolted Repairs 19.4  Bonded Repairs 19.5  Metallic Details and Metal-Bonded Assemblies

517 517 517 520 523 533

Chapter 20 Metal Matrix Composites 20.1  Aluminum Matrix Composites 20.2  Discontinuous Composite Processing Methods 20.3  Stir Casting 20.4  Slurry Casting—Compocasting 20.5  Liquid Metal Infiltration 20.5.1  Squeeze Casting 20.5.2  Pressure Infiltration Casting 20.5.3  Pressureless Infiltration 20.6  Spray Deposition 20.7  Powder Metallurgy Methods 20.8  Secondary Processing of Discontinuous MMCs 20.9  Continuous-Fiber Aluminum MMCs 20.10  Continuous-Fiber Reinforced Titanium Matrix Composites 20.11  Continuous-Fiber TMC Processing Methods 20.12  TMC Consolidation Procedures 20.13  Secondary Fabrication of TMCs 20.14  Particle-Reinforced TMCs 20.15  Fiber Metal Laminates

537 540 542 542 544 545 545 545 546 546 548 549 550 554 557 560 562 566 567

Chapter 21 Ceramic Matrix Composites 21.1  Reinforcements 21.2  Matrix Materials 21.3  Interfacial Coatings 21.4  Fiber Architectures

573 575 578 580 580

x  /  Contents

21.5  21.6  21.7  21.8 

Fabrication Methods Powder Processing Slurry Infiltration and Consolidation Polymer Infiltration and Pyrolysis (PIP) 21.8.1  Space Shuttle C-C Process 21.8.2  Conventional PIP Processes 21.8.3  Sol-Gel Infiltration 21.9  Chemical Vapor Infiltration (CVI) 21.10  Directed Metal Oxidation (DMO) 21.11  Liquid Silicon Infiltration (LSI)

581 581 583 584 585 587 588 589 592 594

Appendix A Metric Conversion Factors Index

597 599

Preface Composite materials are pervasive throughout our world and include both natural and man-made composites. For example, in nature, wood is a composite consisting of wood fibers (cellulose) bound together by a matrix of lignin. Composite materials have been used by mankind for thousands of years; many of the sun-dried mud brick buildings of the earliest known civilization in Mesopotamia at Sumer were reinforced with straw as early as 4900 b.c. However, with the advent of high-strength man-made fibers and the tremendous advances in polymer chemistry during the twentieth century, in many instances composite materials now can be made that offer advantages comparable to those of competing materials. The advantages of these advanced composites are many, including lighter weight, the ability to tailor composites for optimum strength and stiffness, improved fatigue life, corrosion resistance, and, with good design practice, reduced ­assembly costs due to fewer detail parts and fasteners. The specific strength (strength/ density) and specific modulus (modulus/density) of high-strength fiber-reinforced composites, especially those with carbon fibers, are higher than those of comparable metal alloys. This translates into greater weight savings, resulting in improved performance, greater payloads, longer ranges (for vehicles), and fuel savings. This book is intended primarily for technical personnel who want to learn more about modern composite materials. It would be useful to designers, structural engineers, materials and process engineers, manufacturing engineers, and production personnel involved with composites. The book deals with all aspects of advanced composite materials: what they are, where they are used, how they are made, their properties, how they are designed and analyzed, and how they perform in service. It covers continuous- and discontinuous-­ fiber composites fabricated from polymer, metal, and ceramic matrices, with an emphasis on continuous-fiber polymer matrix composites. The book covers composite materials at the introductory to intermediate level. Throughout the book, practical aspects are emphasized more than theory. Because I spent 38 years in the industry, the information covers the current state-of-the-art in composite materials. The book starts with an overview of composite materials (Chapter 1) and how highly anisotropic composites differ from isotropic materials, such as metals. Some of the important advantages and disadvantages of composites are discussed. Chapter 1 wraps up with some of the applications for advanced composites. Chapter 2 examines the reinforcements and their product forms, with an emphasis on glass, aramid, and carbon fibers. Chapter 3 covers the main thermosetting and thermoplastic resin systems. Thermoset resin systems include polyesters, vinyl esters, epoxies, bismaleimides, cynate esters, polyimides, and phenolics. Thermoplastic composite matrices include polyetheretherketone, polyetherketoneketone, polyetherimide, and polypropylene. The principles of thermoset resin toughening are also presented, along with an introduction to the physiochemical tests that are used to characterize resins and cured laminates.

xii  /  Preface

Chapters 4 through 11 describe the progression of composite fabrication steps. Chapter 4 covers the basics of cure tools. This is followed by a discussion of thermoset composite fabrication processes (Chapter 5). Important thermoset lay-up methods include wet lay-up, prepreg lay-up, automated tape laying, fiber placement, filament winding, and pultrusion. Vacuum bagging in preparation for cure is also discussed, along with the cure processes for both addition and condensation curing thermosets. Thermoset liquid molding covers preforming technology (weaving, knitting, stitching, and braiding) followed by the major liquid molding processes, namely, resin transfer molding, resin film infusion, and vacuum-assisted resin transfer molding. In Chapter 6, thermoplastic composite consolidation is covered, along with the different methods of thermoforming thermoplastics. Finally, the joining processes that are unique to thermoplastic composites are discussed. After these processing fundamentals are fully described, Chapter 7 deals with some of the detailed processing issues unique to thermoset and thermoplastic composites. The concept of cure modeling is ­introduced along with the importance of both lay-up and cure variables, hydrostatic resin pressure, chemical composition, resin and prepreg, debulking, and caul plates. Residual cure stresses and exothermic reactions are also covered, followed by a brief description of in-process cure monitoring. Adhesive bonding, sandwich, and integrally cocured structures are introduced in Chapters 8 and 9. The basics of adhesive bonding are covered, along with its advantages and disadvantages. The importance of joint design, surface preparation, and bonding procedures is discussed, along with honeycomb bonded assemblies, foam bonded assemblies, and integrally cocured assemblies. Large, one-piece composite airframe structures have demonstrated the potential for impressive reductions in part counts and assembly costs. The properties and fabrication technology for discontinuous-fiber polymer matrix composites are addressed in Chapter 10, with an emphasis on spray-up, compression molding, structural reaction injection molding, and injection molding. Assembly (Chapter 11) can represent a significant portion of the total manufacturing cost, as much as 50 percent of the total delivered cost. In this chapter, the emphasis is on mechanical joining, including the hole preparation procedures and fasteners used for structural assembly. Sealing and painting are also briefly discussed. Chapters 12 through 15 cover the test methods and properties for composite materials. Important nondestructive test methods (Chapter 12) include visual, ultrasonics, radiographic, and thermographic inspection methods. Mechanical property test methods (Chapter 13) include tests for both composite materials and adhesive systems. In Chapter 14, the strength and stiffness for both discontinuous and continuous reinforced composites are compared. Chapter 15 covers the important topic of environmental degradation, including moisture absorption, fluids exposure, ultraviolet radiation and erosion, lightning strikes, thermo-oxidative behavior, heat damage, and flammability. Chapters 16 through 19 cover the analysis, ­design, and repair of composites. Structural analysis (Chapter 16) starts with analysis at the lamina, or ply, level and then uses classical lamination theory to illustrate the analysis methods for more complex laminates. The concept of interlaminar free edge stresses is introduced. Four failure theories are discussed: the maximum stress criterion, the maximum strain criterion, the Azzi-Tsai-Hill maximum work theory, and the Tsai-Wu failure criterion. The important topic of analysis of composite joints, both bolted and bonded, is covered in Chapter 17. Chapter 18 deals with composite design and certification considerations, including materials and process selection, design trade studies, the building block approach to certification, design allowables, and design guidelines. Considerations for handling damage tolerance and environmental issues are also discussed. Repair of composites (Chapter 19) includes fill repairs, injection repairs, bolted repairs, and bonded repairs. Metal matrix composites (Chapter 20) offer a number of advantages compared to their base metals, such as higher specific strengths and moduli, higher elevated-temperature resistance, lower coefficients of thermal expansion, and, in some cases, better wear re-

Preface  /  xiii 

sistance. On the downside, they are more expensive than their base metals and have lower toughness. Because of their high costs, commercial applications for metal matrix composites are limited. As with metal matrix composites, there are few commercial applications for ceramic matrix composites (Chapter 21), also because of their high costs, as well as concerns for reliability. Carbon-carbon composites have been used in aerospace applications for thermal protection systems. However, metal and ceramic matrix composites remain an important material class, because they are considered enablers for future hypersonic flight vehicles. The reader is cautioned that the data ­presented in this book are not design allowables. The reader should consult approved design manuals for statistically derived design allowables. I would like to acknowledge the help and guidance of Ann Britton, Eileen De Guire, Steve Lampman, and Madrid Tramble, ASM International, and the staff at ASM for their valuable contributions. I would also like to thank my wife, Betty, for her continuing support. F.C. Campbell St. Louis, Missouri July 2010

“This page left intentionally blank.”

About the Author F.C. Campbell’s 38-year career at The Boeing Company (retired 2007) was closely divided equally between engineering and manufacturing. He worked in the engineering laboratories, manufacturing research and development, as well as engineering on four production aircraft programs, and in production operations. At the time of his retirement, he was a Senior Technical Fellow in the field of structural materials and manufacturing technology. He is knowledgeable about a large number of materials, fabrication, and assembly processes for airframe structural materials. Previously, he was director of manufacturing process improvement (1995–2000), and from 1987–1995, he was director of manufacturing research engineering. Earlier in his career, he worked in materials and process development with responsibility for composite related research and development programs. He has also worked on the F-15, F/A-18, AV-8B, and C-17 aircraft programs, conducted manufacturing research on composite and metallic materials, and worked as a laboratory engineer doing process development on both metal matrix and organic matrix composite materials.

“This page left intentionally blank.”

Structural Composite Materials F.C. Campbell

Copyright © 2010, ASM International® All rights reserved.

Chapter 1

Introduction to Composite Materials A composite material can be defined as a combination of two or more materials that results in better properties than those of the individual components used alone. In contrast to metallic alloys, each material retains its separate chemical, physical, and mechanical properties. The two constituents are a reinforcement and a matrix. The main advantages of composite materials are their high strength and stiffness, combined with low density, when compared with bulk materials, allowing for a weight reduction in the finished part. The reinforcing phase provides the strength and stiffness. In most cases, the reinforcement is harder, stronger, and stiffer than the matrix. The reinforcement is usually a fiber or a particulate. Particulate composites have dimensions that are approximately equal in all directions. They may be spherical, platelets, or any other regular or irregular geometry. Particulate composites tend to be much weaker and less stiff than continuousfiber composites, but they are usually much less expensive. Particulate reinforced composites usually contain less reinforcement (up to 40 to 50 volume percent) due to processing difficulties and brittleness. A fiber has a length that is much greater than its diameter. The length-to-diameter (l/d ) ratio is known as the aspect ratio and can vary greatly. Continuous fibers have long aspect ratios, while discontinuous fibers have short aspect ratios. Continuous-fiber composites normally have a preferred orientation, while discontinuous fibers generally have a random orientation. Examples of continuous reinforcements include unidirectional, woven cloth, and helical winding (Fig. 1.1a), while examples of discontinuous reinforcements are chopped fibers and random mat (Fig. 1.1b). Continuous-fiber composites are often made into laminates by stacking single

sheets of continuous fibers in different orientations to obtain the desired strength and stiffness properties with fiber volumes as high as 60 to 70 percent. Fibers produce high-strength composites because of their small diameter; they contain far fewer defects (normally surface defects) compared to the material produced in bulk. As a general rule, the smaller the diameter of the fiber, the higher its strength, but often the cost increases as the diameter becomes smaller. In addition, smaller-diameter high-strength fibers have greater flexibility and are more amenable to fabrication processes such as weaving or forming over radii. Typical fibers include glass, aramid, and carbon, which may be continuous or discontinuous. The continuous phase is the matrix, which is a polymer, metal, or ceramic. Polymers have low strength and stiffness, metals have intermediate strength and stiffness but high ductility, and ceramics have high strength and stiffness but are brittle. The matrix (continuous phase) performs several critical functions, including maintaining the fibers in the proper orientation and spacing and protecting them from abrasion and the environment. In polymer and metal matrix composites that form a strong bond between the fiber and the matrix, the matrix transmits loads from the matrix to the fibers through shear loading at the interface. In ceramic matrix composites, the objective is often to increase the toughness rather than the strength and stiffness; therefore, a low interfacial strength bond is desirable. The type and quantity of the reinforcement �determine the final properties. Figure 1.2 shows that the highest strength and modulus are obtained with continuous-fiber composites. There is a practical limit of about 70 volume percent reinforcement that can be added to form a composite. At higher percentages, there is too little matrix to support the fibers effectively. The theoretical

2╇ /╇ Structural Composite Materials

Fig. 1.1

Typical reinforcement types

strength of discontinuous-fiber composites can approach that of continuous-fiber composites if their aspect ratios are great enough and they are aligned, but it is difficult in practice to maintain good alignment with discontinuous fibers. Discontinuous-fiber composites are normally somewhat random in alignment, which dramatically reduces their strength and modulus. However, discontinuous-fiber composites are gen� erally much less costly than continuous-fiber composites. Therefore, continuous-fiber composites are used where higher strength and stiffness are required (but at a higher cost), and discontinuous-fiber composites are used where cost is the main driver and strength and stiffness are less important. Both the reinforcement type and the matrix affect processing. The major processing routes for polymer matrix composites are shown in Fig. 1.3. Two types of polymer matrices are shown: thermosets and thermoplastics. A thermoset starts as

a low-viscosity resin that reacts and cures during processing, forming an intractable solid. A thermoplastic is a high-viscosity resin that is processed by heating it above its melting temperature. Because a thermoset resin sets up and cures during processing, it cannot be reprocessed by reheating. By comparison, a thermoplastic can be reheated above its melting temperature for additional processing. There are processes for both classes of resins that are more amenable to discontinuous fibers and others that are more amenable to continuous fibers. In general, because metal and ceramic matrix composites require very high temperatures and sometimes high pressures for processing, they are normally much more expensive than polymer matrix composites. However, they have much better thermal stability, a requirement in applications where the composite is exposed to high temperatures. This book will deal with both continuous and discontinuous polymer, metal, and ceramic matrix

Chapter 1: Introduction to Composite Materials╇ /╇ 3╇

Fig. 1.2

Influence of reinforcement type and quantity on composite performance

Fig. 1.3

Major polymer matrix composite fabrication processes

4╇ /╇ Structural Composite Materials

composites, with an emphasis on continuousfiber, high-performance polymer composites.

1.1 Isotropic, Anisotropic, and Orthotropic Materials Materials can be classified as either isotropic or anisotropic. Isotropic materials have the same material properties in all directions, and normal loads create only normal strains. By comparison, anisotropic materials have different material properties in all directions at a point in the body. There are no material planes of symmetry, and normal loads create both normal strains and shear strains. A material is isotropic if the properties are independent of direction within the material. For example, consider the element of an isotropic material shown in Fig. 1.4. If the material is loaded along its 0°, 45°, and 90° directions, the modulus of elasticity (E) is the same in each direction (E0° = E45° = E90°). However, if the

Fig. 1.4

Element of isotropic material under stress

material is anisotropic (for example, the composite ply shown in Fig. 1.5), it has properties that vary with direction within the material. In this example, the moduli are different in each direction (E0° ≠ E45° ≠ E90°). While the modulus of elasticity is used in the example, the same dependence on direction can occur for other material properties, such as ultimate strength, Poisson’s ratio, and thermal expansion coefficient. Bulk materials, such as metals and polymers, are normally treated as isotropic materials, while composites are treated as anisotropic. However, even bulk materials such as metals can become anisotropic––for example, if they are highly cold worked to produce grain alignment in a certain direction. Consider the unidirectional fiber-reinforced composite ply (also known as a lamina) shown in Fig. 1.6. The coordinate system used to describe the ply is labeled the 1-2-3 axes. In this case, the 1-axis is defined to be parallel to the fibers (0°), the 2-axis is defined to lie within the plane of the plate and is perpendicular to the fibers (90°), and the 3-axis is defined to be normal

Chapter 1: Introduction to Composite Materials╇ /╇ 5╇

Fig. 1.5

Element of composite ply material under stress

Fig. 1.6

Ply angle definition

6╇ /╇ Structural Composite Materials

to the plane of the plate. The 1-2-3 coordinate system is referred to as the principal material coordinate system. If the plate is loaded parallel to the fibers (one- or zero-degree direction), the modulus of elasticity E11 approaches that of the fibers. If the plate is loaded perpendicular to the fibers in the two- or 90-degree direction, the modulus E22 is much lower, approaching that of the relatively less stiff matrix. Since E11 >> E22 and the modulus varies with direction within the material, the material is anisotropic. Composites are a subclass of anisotropic materials that are classified as orthotropic. Orthotropic materials have properties that are different in three mutually perpendicular directions. They have three mutually perpendicular axes of symmetry, and a load applied parallel to these axes produces only normal strains. However, loads that are not applied parallel to these axes produce both normal and shear strains. Therefore, orthotropic mechanical properties are a function of orientation.

Fig. 1.7

Shear coupling in a 45° ply. Source: Ref 1

Consider the unidirectional composite shown in the upper portion of Fig. 1.7, where the unidirectional fibers are oriented at an angle of 45 degrees with respect to the x-axis. In the small, isolated square element from the gage region, because the element is initially square (in this example), the fibers are parallel to diagonal AD of the element. In contrast, fibers are perpendicular to diagonal BC. This implies that the element is stiffer along diagonal AD than along diagonal BC. When a tensile stress is applied, the square element deforms. Because the stiffness is higher along diagonal AD than along diagonal BC, the length of diagonal AD is not increased as much as that of diagonal BC. Therefore, the initially square element deforms into the shape of a parallelogram. Because the element has been distorted into a parallelogram, a shear strain gxy is induced as a result of coupling between the axial strains exx and eyy. If the fibers are aligned parallel to the direction of applied stress, as in the lower portion of

Chapter 1: Introduction to Composite Materials╇ /╇ 7╇

Fig. 1.7, the coupling between exx and eyy does not occur. In this case, the application of a tensile stress produces elongation in the x-direction and contraction in the y-direction, and the distorted element remains rectangular. Therefore, the coupling effects exhibited by composites occur only if stress and strain are referenced to a non– principal material coordinate system. Thus, when the fibers are aligned parallel (0°) or perpendicular (90°) to the direction of applied stress, the lamina is known as a specially orthotropic lamina (θ = 0° or 90°). A lamina that is not aligned parallel or perpendicular to the direction of applied stress is called a general orthotropic lamina (θ ≠ 0° or 90°).

1.2 Laminates When there is a single ply or a lay-up in which all of the layers or plies are stacked in the same orientation, the lay-up is called a lamina. When the plies are stacked at various angles, the lay-up is called a laminate. Continuous-fiber compos-

Fig. 1.8

Lamina and laminate lay-ups

ites are normally laminated materials (Fig. 1.8) in which the individual layers, plies, or laminae are oriented in directions that will enhance the strength in the primary load direction. Unidirectional (0°) laminae are extremely strong and stiff in the 0° direction. However, they are very weak in the 90° direction because the load must be carried by the much weaker polymeric matrix. While a high-strength fiber can have a tensile strength of 500 ksi (3500 MPa) or more, a typical polymeric matrix normally has a tensile strength of only 5 to 10 ksi (35 to 70 MPa) (Fig. 1.9). The longitudinal tension and compression loads are carried by the fibers, while the matrix distributes the loads between the fibers in tension and stabilizes the fibers and prevents them from buckling in compression. The matrix is also the primary load carrier for interlaminar shear (i.e., shear between the layers) and transverse (90°) tension. The relative roles of the fiber and the matrix in detemining mechanical properties are summarized in Table 1.1. Because the fiber orientation directly impacts mechanical properties, it seems logical to orient

8╇ /╇ Structural Composite Materials

Fig. 1.9

Comparison of tensile properties of fiber, matrix, and composite

1.3 Fundamental Property Relationships

Table 1.1â•…Effect of fiber and matrix on mechanical properties Dominating composite constituent Mechanical property



Unidirectional 0º tension 0º compression Shear 90º tension

√ √ … …

… √ √ √

Laminate Tension Compression In-plane shear Interlaminar shear

√ √ √ …

… √ √ √

as many of the layers as possible in the main load-carrying direction. While this approach may work for some structures, it is usually necessary to balance the load-carrying capability in a number of different directions, such as the 0°, +45°, -45°, and 90° directions. Figure 1.10 shows a photomicrograph of a cross-plied continuous carbon fiber/epoxy laminate. A balanced laminate having equal numbers of plies in the 0°, +45°, –45°, and 90° degrees directions is called a quasi-isotropic laminate, because it carries equal loads in all four directions.

When a unidirectional continuous-fiber lamina or laminate (Fig. 1.11) is loaded in a diÂ� rection parallel to its fibers (0° or 11-direction), the longitudinal modulus E11 can be estimated from its constituent properties by using what is known as the rule of mixtures: E11 = EfVf + EmVm

(Eq 1.1)

where Ef is the fiber modulus, Vf is the fiber volume percentage, Em is the matrix modulus, and Vm is the matrix volume percentage. The longitudinal tensile strength s11 also can be estimated by the rule of mixtures: s11 = sVf + smVm

(Eq 1.2)

where sf and sm are the ultimate fiber and matrix strengths, respectively. Because the properties of the fiber dominate for all practical volume percentages, the values of the matrix can often be ignored; therefore: E11 ≈ EfVf s11 ≈ sVf

(Eq 1.3) (Eq 1.4)

Chapter 1: Introduction to Composite Materials╇ /╇ 9╇

Fig. 1.10

Cross section of a cross-plied carbon/epoxy laminate

Fig. 1.11

Unidirectional continuous-fiber lamina or laminate

10╇ /╇ Structural Composite Materials

Figure 1.12 shows the dominant role of the fibers in determining strength and stiffness. When loads are parallel to the fibers (0°), the ply is much stronger and stiffer than when loads are transverse (90°) to the fiber direction. There is a dramatic decrease in strength and stiffness resulting from only a few degrees of misalignment off of 0°. When the lamina shown in Fig. 1.11 is loaded in the transverse (90° or 22-direction), the fibers and the matrix function in series, with both carrying the same load. The transverse modulus of elasticity E22 is given as: 1/E22 = Vf /Ef + Vm/Em

(Eq 1.5)

Figure 1.13 shows the variation of modulus as a function of fiber volume percentage. When the fiber percentage is zero, the modulus is essentially the modulus of the polymer, which increases up to 100 percent (where it is the modulus of the fiber). At all other fiber volumes, the E22 or 90° modulus is lower than the E11 or zero degrees modulus, because it is dependent on the much weaker matrix. Other rule of mixture expressions for lamina properties include those for the Poisson’s ratio n12 and for the shear modulus G12:

Fig. 1.12

Influence of ply angle on strength and modulus

n12 = nfVf + nmVm 1/G12 = Vf /Gf + Vm/Gm

(Eq 1.6) (Eq 1.7)

These expressions are somewhat less useful than the previous ones, because the values for Poisson’s ratio (nf) and the shear modulus (Gf) of the fibers are usually not readily available. Physical properties, such as density (r), can also be expressed using rule of mixture relations: r12 = rfVf + rmVm

(Eq 1.8)

While these micromechanics equations are useful for a first estimation of lamina properties when no data are available, they generally do not yield sufficiently accurate values for design purposes. For design purposes, basic lamina and laminate properties should be determined using actual mechanical property testing.

1.4 Composites versus Metallics As previously discussed, the physical characteristics of composites and metals are significantly different. Table 1.2 compares some properties of composites and metals. Because composites are highly anisotropic, their in-plane strength and

Chapter 1: Introduction to Composite Materials╇ /╇ 11╇

Fig. 1.13

Variation of composite modulus of a unidirectional 0° lamina as a function of fiber volume fraction

Table 1.2â•… Composites versus metals comparison Condition

Comparative behavior relative to metals

Load-strain relationship Notch sensitivity â•… Static â•… Fatigue Transverse properties Mechanical property variability Fatigue strength Sensitivity to hydrothermal environment Sensitivity to corrosion Damage growth mechanism

More linear strain to failure Greater sensitivity Less sensitivity Weaker Higher Higher Greater Much less In-plane delamination instead of through thickness cracks

Source: Ref 2

stiffness are usually high and directionally variable, depending on the orientation of the reinforcing fibers. Properties that do not benefit from this reinforcement (at least for polymer matrix composites) are comparatively low in strength and stiffness—for example, the through-thethickness tensile strength where the relatively weak matrix is loaded rather than the highstrength fibers. Figure 1.14 shows the low through-the-thickness strength of a typical composite laminate compared with aluminum.

Metals typically have reasonable ductility, continuing to elongate or compress considerably when they reach a certain load (through yielding) without picking up more load and without failure. Two important benefits of this ductile yielding are that (1) it provides for local load relief by distributing excess load to an adjacent material or structure; therefore, ductile metals have a great capacity to provide relief from stress concentrations when statically loaded; and (2) it provides great energy-absorbing capability (indicated by the area under a stress-strain curve). As a result, when impacted, a metal structure typically deforms but does not actually fracture. In contrast, composites are relatively brittle. Figure 1.15 shows a comparison of typical tensile stress-strain curves for two materials. The brittleness of the composite is reflected in its poor ability to tolerate stress concentrations, as shown in Fig. 1.16. The characteristically brittle composite material has poor ability to resist impact damage without extensive internal matrix fracturing. The response of damaged composites to cyclic loading is also significantly different from that of metals. The ability of composites to withstand cyclic loading is far superior to that of metals, in contrast to the poor composite static strength when it has damage or defects. Figure 1.17

12╇ /╇ Structural Composite Materials

Fig. 1.14

Comparison of through-the-thickness tensile strength of a composite laminate with aluminum alloy sheet. Source: Ref 3

Fig. 1.15

Comparison of typical stress-strain curves for a composite laminate and aluminum alloy sheet. Source: Ref 3

Chapter 1: Introduction to Composite Materials╇ /╇ 13╇

Fig. 1.16

 ompared with aluminum alloy sheet, a composite laminate has poor tolerance of stress concentration because of its C brittle nature. Source: Ref 3

Fig. 1.17

Comparative notched fatigue strength of composite laminate and aluminum alloy sheet. Source: Ref 3

14╇ /╇ Structural Composite Materials

shows a comparison of the normalized notched specimen fatigue response of a common 7075T6 aluminum aircraft metal and a carbon/epoxy laminate. The fatigue strength of the composite is much higher relative to its static or residual strength. The static or residual strength requirement for structures is typically much higher than the fatigue requirement. Therefore, because the fatigue threshold of composites is a high percentage of their static or damaged residual strength, they are usually not fatigue critical. In metal structures, fatigue is typically a critical design consideration.

1.5 Advantages and Disadvantages of Composite Materials The advantages of composites are many, including lighter weight, the ability to tailor the layup for optimum strength and stiffness, improved fatigue life, corrosion resistance, and, with good design practice, reduced assembly costs due to fewer detail parts and fasteners.

Fig. 1.18

The specific strength (strength/density) and specific modulus (modulus/density) of highstrength fibers (especially carbon) are higher than those of other comparable aerospace metallic alloys (Fig. 1.18). This translates into greater weight savings resulting in improved performance, greater payloads, longer range, and fuel savings. Figure 1.19 compares the overall structural efficiency of carbon/epoxy, Ti-6Al-4V, and 7075-T6 aluminum. The chief engineer of aircraft structures for the U.S. Navy once told the author that he liked composites because “they don’t rot [corrode] and they don’t get tired [fatigue].” Corrosion of aluminum alloys is a major cost and a constant maintenance problem for both commercial and military aircraft. The corrosion resistance of composites can result in major savings in supportability costs. Carbon fiber composites cause galvanic corrosion of aluminum if the fibers are placed in direct contact with the metal surface, but bonding a glass fabric electrical insulation layer on all interfaces that contact aluminum eliminates this problem. The fatigue

Comparison of specific strength and modulus of high-strength composites and some aerospace alloys

Chapter 1: Introduction to Composite Materials╇ /╇ 15╇

Fig. 1.19

Relative structural efficiency of aerospace materials

resistance of composites compared to highstrength metals is shown in Fig. 1.20. As long as reasonable strain levels are used during design, fatigue of carbon fiber composites should not be a problem. Assembly costs can account for as much as 50 percent of the cost of an airframe. Composites offer the opportunity to significantly reduce the amount of assembly labor and the number of required fasteners. Detail parts can be combined into a single cured assembly either during initial cure or by secondary adhesive bonding. Disadvantages of composites include high raw material costs and usually high fabrication and assembly costs; adverse effects of both temperature and moisture; poor strength in the out-ofplane direction where the matrix carries the primary load (they should not be used where load paths are complex, such as with lugs and fittings); susceptibility to impact damage and delaminations or ply separations; and greater difficulty in repairing them compared to metallic structures. The major cost driver in fabrication for a composite part using conventional hand lay-up is the cost of laying up or collating the plies. This cost is generally 40 to 60 percent of the fabrication cost, depending on part complexity (Fig. 1.21). Assem-

bly cost is another major cost driver, accounting for about 50 percent of the total part cost. As previously stated, one of the potential advantages of composites is the ability to cure or bond a number of detail parts together to reduce assembly costs and the number of required fasteners. Temperature has an effect on composite mechanical properties. Typically, matrix-dominated mechanical properties decrease with increas� ing temperature. Fiber-dominated properties are somewhat affected by cold temperatures, but the effects are not as severe as those of elevated temperature on the matrix-dominated properties. Design parameters for carbon/epoxy are colddry tension and hot-wet compression (Fig. 1.22). An important design factor in the selection of a matrix resin for elevated-temperature applications is the cured glass transition temperature. The cured glass transition temperature (Tg) of a polymeric material is the temperature at which it changes from a rigid, glassy solid into a softer, semiflexible material. At this point, the polymer structure is still intact but the crosslinks are no longer locked in position. Therefore, the Tg determines the upper use temperature for a composite or an adhesive and is the temperature above which the material will exhibit significantly

16╇ /╇ Structural Composite Materials

Fig. 1.20

Fatigue properties of aerospace materials

reduced mechanical properties. Since most thermoset polymers will absorb moisture that severely depresses the Tg, the actual use temperature should be about 50 ºF (30 ºC) lower than the wet or saturated Tg. Upper Use Temperature = Wet Tg – 50 °F

(Eq 1.9)

In general, thermoset resins absorb more moisture than comparable thermoplastic resins. The cured glass transition temperature (Tg) can be determined by several methods that are outlined in Chapter 3, “Matrix Resin Systems.” The amount of absorbed moisture (Fig. 1.23) depends on the matrix material and the relative humidity. Elevated temperatures increase the

rate of moisture absorption. Absorbed moisture reduces the matrix-dominated mechanical€ properties and causes the matrix to swell, which relieves locked-in thermal strains from elevated-temperature curing. These strains can be large, and large panels fixed at their edges can buckle due to strains caused by swelling. During freeze-thaw cycles, absorbed moisture expands during freezing, which can crack the matrix, and it can turn into steam during thermal spikes. When the internal steam pressure exceeds the flatwise tensile (through-the-thickness) strength of the composite, the laminate will delaminate. Composites are susceptible to delaminations (ply separations) during fabrication, during assembly,

Chapter 1: Introduction to Composite Materials╇ /╇ 17╇

Fig. 1.21

Cost drivers for composite hand lay-up. NDI, nondestructive inspection

and in service. During fabrication, foreign materials such as prepreg backing paper can be inadvertently left in the lay-up. During assembly, improper part handling or incorrectly installed fasteners can cause delaminations. In service, low-velocity impact damage from dropped tools or forklifts running into aircraft can cause damage. The damage may appear as only a small indentation on the surface but it can propagate

through the laminates, forming a complex network of delaminations and matrix cracks, as shown in Fig. 1.24. Depending on the size of the delamination, it can reduce the static and fatigue strength and the compression buckling strength. If it is large enough, it can grow under fatigue loading. Typically, damage tolerance is a resin-dominated property. The selection of a toughened

18╇ /╇ Structural Composite Materials

Fig. 1.22

Effects of temperature and moisture on strength of carbon/epoxy. R.T., room temperature

resin can significantly improve the resistance to impact damage. In addition, S-2 glass and aramid fibers are extremely tough and damage tolerant. During the design phase, it is important to recognize the potential for delaminations and use sufficiently conservative design strains so that a damaged structure can be repaired.

1.6 Applications Applications include aerospace, transpor� tation, construction, marine goods, sporting goods, and more recently infrastructure, with construction and transportation being the largest. In general, high-performance but more costly continuous-carbon-fiber composites are used

Chapter 1: Introduction to Composite Materials╇ /╇ 19╇

Fig. 1.23

Absorption of moisture for polymer matrix composites. RH, relative humidity

where high strength and stiffness along with light weight are required, and much lower-cost fiberglass composites are used in less demanding applications where weight is not as critical. In military aircraft, low weight is “king” for performance and payload reasons, and composites often approach 20 to 40 percent of the airframe weight (Fig. 1.25). For decades, helicopters have incorporated glass fiber–reinforced rotor blades for improved fatigue resistance, and in recent years helicopter airframes have been built largely of carbon-fiber composites. Military aircraft applications, the first to use highperformance continuous-carbon-fiber composites,

drove the development of much of the technology now being used by other industries. Both small and large commercial aircraft rely on composites to decrease weight and increase fuel performance, the most striking example being the 50 percent composite airframe for the new Boeing 787 (Fig. 1.26). All future Airbus and Boeing aircraft will use large amounts of high-performance composites. Composites are also used extensively in both weight-critical reusable and expendable launch vehicles and satellite structures (Fig. 1.27). Weight savings due to the use of composite materials in aerospace applications generally range from 15 to 25 percent.

20╇ /╇ Structural Composite Materials

Fig. 1.24

Delaminations and matrix cracking in polymer matrix composite due to impact damage

The major automakers (Fig. 1.28) are increasingly turning to composites to help them meet performance and weight requirements, thus improving fuel efficiency. Cost is a major driver for commercial transportation, and composites offer lower weight and lower maintenance costs. Typical materials are fiberglass/ polyurethane made by liquid or compression molding and fiberglass/ polyester made by compression molding. Recreational vehicles have long used glass fibers, mostly for their durability and weight savings over metal. The product form is typically fiberglass sheet molding compound made by compression molding.

For high-performance Formula 1 racing cars, where cost is not an impediment, most of the chassis, including the monocoque, suspension, wings, and engine cover, is made from carbon fiber composites. Corrosion is a major headache and expense for the marine industry. Composites help minimize these problems, primarily because they do not corrode like metals or rot like wood. Hulls of boats ranging from small fishing boats to large racing yachts (Fig. 1.29) are routinely made of glass fibers and polyester or vinyl ester resins. Masts are frequently fabricated from carbon fiber composites. Fiberglass filament-wound SCUBA

Chapter 1: Introduction to Composite Materials╇ /╇ 21╇

Fig. 1.25

Typical fighter aircraft applications. Source: The Boeing Company

tanks are another example of composites improving the marine industry. Lighter tanks can hold more air yet require less maintenance than their metallic counterparts. Jet skis and boat trailers often contain glass composites to help minimize weight and reduce corrosion. More recently, the topside structures of many naval ships have been fabricated from composites. Using composites to improve the infrastructure (Fig. 1.30) of our roads and bridges is a relatively new, exciting application. Many of the world’s roads and bridges are badly corroded and in need of continual maintenance or replacement.

In the United States alone, it is estimated that more than 250,000 structures, such as bridges and parking garages, need repair, retrofit, or replacement. Composites offer much longer life with less maintenance due to their corrosion resistance. Typical processes/materials include wet lay-up repairs and corrosion-resistant fiberglass pultruded products. In construction (Fig. 1.31), pultruded fiberglass rebar is used to strengthen concrete, and glass fibers are used in some shingling materials. With the number of mature tall trees dwindling, the use of composites for electrical towers and

22╇ /╇ Structural Composite Materials

Fig. 1.26

Boeing 787 Dreamliner commercial airplane. Source: The Boeing Company

Fig. 1.27

Launch and spacecraft structures

Chapter 1: Introduction to Composite Materials╇ /╇ 23╇

Fig. 1.28

Transportation applications

light poles is greatly increasing. Typically, these are pultruded or filament-wound glass. Wind power is the world’s fastest-growing energy source. The blades for large wind turbines (Fig. 1.32) are normally made of composites to

improve electrical energy generation efficiency. These blades can be as long as 120 ft (37 m) and weigh up to 11,500 lb (5200 kg). In 2007, nearly 50,000 blades for 17,000 turbines were delivered, representing roughly 400 million pounds

24╇ /╇ Structural Composite Materials

Fig. 1.29

Marine applications

(approximately 180 million kg) of composites. The predominant material is continuous glass fibers manufactured by either lay-up or resin infusion.

Tennis racquets (Fig. 1.33) have been made of glass for years, and many golf club shafts are made of carbon. Processes include compression molding for tennis racquets and tape wrapping or

Chapter 1: Introduction to Composite Materials╇ /╇ 25╇

Fig. 1.30

Infrastructure applications

filament winding for golf shafts. Lighter, stronger skis and surfboards also are possible using composites. Another example of a composite application that takes a beating yet keeps on per-

forming is a snowboard, which typically involves the use of a sandwich construction (composite skins with a honeycomb core) for maximum specific stiffness.

26╇ /╇ Structural Composite Materials

With the number of mature tall trees dwindling, the use of composites for electrical towers and light poles is greatly increasing.

Fig. 1.31

Construction applications

Although metal and ceramic matrix com� posites are normally very expensive, they have found uses in specialized applications such as those shown in Fig. 1.34. Frequently, they

are used where high temperatures are involved. However, the much higher temperatures and pressures �required for the fabrication of metal and ceramic matrix composites lead

Chapter 1: Introduction to Composite Materials╇ /╇ 27╇

Fig. 1.32

Composite clean energy generation

to very high costs, which severely limits their application. Composites are not always the best solution. An example is the avionics rack for an advanced fighter aircraft shown in Fig. 1.35. This part was machined from a single block of aluminum in about 8.5 hours and assembled into the final component in five hours. Such a part made of composites would probably not be cost competitive. Advanced composites are a diversified and growing industry due to their distinct advantages over competing metallics, including lighter weight, higher performance, and corrosion resis-

tance. They are used in aerospace, automotive, marine, sporting goods, and, more recently, infrastructure applications. The major disadvantage of composites is their high cost. However, the proper selection of materials (fiber and matrix), product forms, and processes can have a major impact on the cost of the finished part. References

1. M.E. Tuttle, Structural Analysis of Polymeric Composite Materials, Marcel Dekker, Inc., 2004

28╇ /╇ Structural Composite Materials

Fig. 1.33

Sporting goods applications

Fig. 1.34

Metal and ceramic matrix composite applications

Chapter 1: Introduction to Composite Materials╇ /╇ 29╇

Fig. 1.35

 omposites are not always the best choice. This avionics rack machined from an aluminum alloy block would not be C cost-competitive if made of composites. Source: The Boeing Company

2. M.C.Y. Niu, Composite Airframe Structures, 2nd ed., Hong Kong Conmilit Press Limited, 2000 3. R.E. Horton and J.E. McCarty, Damage Tolerance of Composites, Engineered Materials Handbook, Vol 1, Composites, ASM International, 1987

Selected References •

High-Performance Composites Sourcebook 2009, Gardner Publications Inc S.K. Mazumdar, Composites Manufacturing: Materials, Product, and Process Engineering, CRC Press, 2002

“This page left intentionally blank.”

Copyright © 2010, ASM International® All rights reserved.

Structural Composite Materials F.C. Campbell

Chapter 2

Fibers and Reinforcements Reinforcements for composite materials can be particles, whiskers, or fibers. Particles have no preferred orientation and provide minimal improvements in mechanical properties. They are frequently used as fillers to reduce the cost of the material. Whiskers are single crystals that are extremely strong but are difficult to disperse uniformly in the matrix. They are small in both length and diameter compared to fibers. Fibers have a very long axis compared to particles and whiskers. They are usually circular or nearly circular and are significantly stronger in the long direction because they are normally made by either drawing or pulling during the manufacturing process. Drawing orients the molecules so that tension loads on the fibers pull more against the molecular chains themselves than against a mere entanglement of chains. Due to the strength and stiffness advantages of fibers, they are the predominant reinforcement for advanced composites. Fibers may be continuous or discontinuous, depending on the application and manufacturing process. This chapter covers the fibers used for organic matrix composites, with an emphasis on continuous fibers. Fibers used for metal- and ceramicmatrix composites are discussed in Chapters 20 and 21, “Metal Matrix Composites” and “Ceramic Matrix Composites,” respectively.

2.1 Fiber Terminology Before examining the various types of fibers used as composite reinforcements, the major terminology used for fiber technology will be reviewed. Fibers are produced and sold in many forms. •

Fiber—A general term for a material that has a long axis that is many times greater than its

diameter. The term aspect ratio, which refers to fiber length divided by diameter (l/d), is frequently used to describe short fiber lengths. Aspect ratios are normally greater than 100 for fibers. Filament—The smallest unit of a fibrous material. For spun fibers, this is the unit formed by a single hole in the spinning process. The term filament is synonymous with fiber. End—A term used primarily for glass fibers that refers to a group of filaments in long parallel lengths. Strand—Another term associated with glass fibers that refers to a bundle or group of untwisted filaments. Continuous strand rovings provide good overall processing characteristics through fast wet-out (penetration of resin into the strand), even tension, and abrasion resistance during processing. They can be cut cleanly, and they disperse evenly throughout the resin matrix during molding. Tow—Similar to a strand of glass fiber, tow is used for carbon and graphite fibers to describe the number of untwisted filaments produced at one time. Tow size is usually expressed as Xk; for example, a 12k tow contains 12,000 filaments. Roving—A number of strands or tows collected into a parallel bundle without twisting. Rovings can be chopped into short fiber Â�segments for sheet molding compound, bulk molding compound or injection molding. Yarn—A number of strands or tows collected into a parallel bundle with twisting. Twisting improves the handleability and makes processes such as weaving easier, but the twist also reduces the strength properties. Band—The thickness or width of several rovings, yarns, or tows as it is applied to a

32╇ /╇ Structural Composite Materials

mandrel or tool; a common term used in filament winding. Tape—A composite product form in which a large number of parallel filaments (such as tows) are held together with an organic matrix material (such as epoxy) commonly referred to as prepreg (preimpregnated with resin). The length of the tape in the direction of the fibers is much greater than the width, and the width is much greater than the thickness. Typical tape product forms are several hundred feet long, 6 to 60 in. (15 cm to 1.5 m) wide, and 0.005 to 0.010 in. (125 to 255 mm) thick. Woven Cloth—Another composite product form made by weaving yarns or tows in various patterns to provide reinforcement in two directions, usually zero and 90 degrees. Typical two-dimensional woven cloth is several hundred feet long, 24 to 60 in. (60 cm to 1.5 m) wide, and 0.010 to 0.015 in. (255 to 380) mm thick. Woven cloth is normally supplied either without resin (dry) or as prepreg with resin.

Additional fiber and textile terminology will be introduced as different fiber types, processes, and product forms are discussed. The physical and mechanical properties of some commercially important fibers are given in Table 2.1, and a graph of specific strengths versus specific moduli is given in Fig. 2.1. Several types of fibers are used for polymeric composites, with glass, aramid (for example, Kevlar), and carbon being the most common. Boron fiber was the original high-performance fiber before

carbon was developed. It is a large-diameter fiber made by pulling a fine tungsten wire through a long, slender reactor, where it is coated with boron using chemical vapor deposition. It is very expensive because it is made one fiber at a time rather than thousands of fibers at a time. Due to its large diameter and high modulus, it has outstanding compression properties. On the negative side, it does not conform well to complicated shapes and is very difficult to machine. Other high-temperature ceramic fibers, such as silicon carbide (Nicalon), aluminum oxide, and alumina boria silica (Nextel), are frequently used in ceramic-based composites but rarely in polymerbased composites. The stress-strain curves in Fig. 2.2 show that high-strength fibers are usually linearly elastic up to the point of failure. Figure 2.3 shows some relative fiber costs versus performance data. Carbon and graphite fibers are the most expensive, followed by aramid, S-2 glass, and E-glass. Because E-glass is a very affordable high-performance fiber, it is the most widely used and dominates in commercial composite applications.

2.2 Strength of Fibers Fibers generally exhibit much higher strengths than the bulk form of the same material. The probability of a flaw per unit length present in a sample is an inverse function of the volume of the material. Since fibers have a very low volume

Table 2.1â•…Properties of some commercially important high-strength fibers Type of fiber

Tensile strength, ksi

Tensile modulus, msi

Elongation at failure, %

Density, g/cm2

Coefficient of thermal expansion, 10−6 °C

Fiber diameter, µm

4.9–6.0 2.9 0.5

╇ 5–20 ╇ 5–10 ╇ 9

Glass E-glass S-2 glass Quartz

500 650 490

10.0 12.6 10.0

4.7 5.6 5.0

2.58 2.48 2.15

Organic Kevlar 29 Kevlar 49 Kevlar 149 Spectra 1000

525 550 500 450

12.0 19.0 27.0 25.0

4.0 2.8 2.0 0.7

1.44 1.44 1.47 0.97

–2.0 –2.0 –2.0 …

12 12 12 27

PAN-based carbon Standard modulus Intermediate modulus High modulus

500–700 600–900 600–80

╇ 32–35 ╇ 40–43 ╇ 50–65

1.5–2.2 1.3–2.0 0.7–1.0

1.80 1.80 1.90

–0.4 –0.6 –0.75

╇ 6–8 ╇ 5–6 ╇ 5–8

Pitch-based carbon Low modulus High modulus Ultra-high modulus

200–450 275–400 350

╇ 25–35 ╇ 55–90 100–140

0.9 0.5 0.3

1.9 2.0 2.2

… –0.9 –1.6

11 11 10

Note: Representative only. For specific properties, contact the fiber manufactures. PAN, polyacrylonitrile

Chapter 2: Fibers and Reinforcements╇ /╇ 33╇

Fig. 2.1

Specific strength and modulus of some commercially important fibers. Source: Ref 1

per unit length, they are much stronger on average than the bulk material, which has a high volume per unit length. On the other hand, because a bulk material has a much higher content of weakening flaws, it exhibits much lower variability in strength. Thus, the smaller the fiber diameter and the shorter its length, the higher the average and maximum strength but the greater the variability. The effect of fiber diameter on the strength of glass fibers is shown in Fig. 2.4. Therefore, fibers have higher strength than their bulk counterparts, but they have greater scatter in their strength. The variability in the strength of fibers is a function of the flaws they contain and, in particular, the flaws they contain on the surface. Flaws can be minimized by careful manufacturing processes and the application of coatings to protect them from mechanical and environmental damage. Precursors used in fiber manufacturing processes must be of high purity and free of inclusions. Many fiber manufacturing processes involve drawing or spinning operations that impose very high degrees of orientation parallel to the fiber axis, thus producing a more favorable orientation in the crystalline or atomic structure. In addition,

some processes involve very high cooling rates that produce ultrafine-grained structures, which are not achievable in most bulk materials.

2.3 Glass Fibers Due to their low cost, high tensile strength, high impact resistance, and good chemical resistance, glass fibers are used extensively in commercial composite applications. However, their properties do not match those of carbon fibers in high-performance composite applications. Compared to carbon fibers, they have a relatively low modulus and inferior fatigue properties. Although there are many types of glass fibers, the three most commonly used in composites are E-glass, S-2 glass, and quartz. E-glass is the most common and least expensive, providing a good combination of tensile strength 500 ksi (3.5 GPa) and modulus 10.0 msi (70 GPa). S-glass, which has a tensile strength of 650 ksi (4.5 GPa) and a modulus of 12.6 msi (87 GPa), is more expensive, but is 40 percent stronger than E-glass and retains a greater percentage of its strength at elevated temperatures. Quartz fiber is a rather

34╇ /╇ Structural Composite Materials

Fig. 2.2

Comparison of stress-strain curves of high-strength fibers. Source: Ref 1

expensive ultrapure silica glass that is a lowdielectric fiber and is used primarily in demanding electrical applications. High-strength glass fibers were first developed in the 1930s and today represent a worldwide structural composite reinforcements market of roughly four to five million tons (3630 to 4540 kg) per year. Glass is an amorphous material that consists of a silica (SiO2) backbone with various oxide components to give specific compositions and properties. Glass fibers are made from silica sand, limestone, boric acid, and minor amounts of other ingredients such as clay, coal, and fluorspar. Silica, which melts at 3128 °F (1720 °C), is also the basic element in quartz, a naturally occurring rock. However, quartz is crystalline with a rigid, highly ordered atomic structure and is

99 percent or more silica. If silica is heated above its melting temperature and slowly cooled, it crystallizes and becomes quartz. On the other hand, glass is produced by altering the temperature and cool-down rates. If pure silica is heated above 3128 °F (1720 °C) and then quickly cooled, crystallization can be prevented and the process yields a glass with an amorphous, or randomly ordered, atomic structure. Although the process is refined and improved continuously, today’s glass fiber manufacturers combine this high-heat/quick-cool strategy with other steps in a process that is basically the same as that developed in the 1930s. High-strength glass fibers are made by blending raw materials, melting them in a three-stage furnace, extruding the molten glass through bushings in the bottom of the forehearth, cooling

Chapter 2: Fibers and Reinforcements╇ /╇ 35╇

Fig. 2.3

Relative fiber cost and performance of some high-strength fibers. Source: Ref 1

Fig. 2.4

Effect of fiber diameter on strength of glass fibers. Source: Ref 2

36╇ /╇ Structural Composite Materials

the filaments with water, and applying a chemical size. The filaments are gathered and wound into a package. The production process can be broken down into five basic steps: batching, melting, fiberization, coating, and drying/packaging. Batching. Although a viable commercial glass fiber can be made of silica alone, other ingredients are added to reduce the working temperature and impart other properties that are useful in specific applications. For example, originally targeted for electrical applications, E-glass with a composition including silica, alumina (Al2O3), calcium oxide (CaO), and magnesium oxide (MgO) was developed as a more alkali-resistant alternative to the original soda-lime glass. Later, boron oxide (B2O3) was added to increase the difference between the temperatures at which the E-glass melted and then formed a crystalline structure to prevent clogging of the nozzles used in fiberization. S-glass fibers, developed for higher strength, are based on a silica-aluminamagnesium oxide formulation but contain higher percentages of silica for applications in which tensile strength is the most important property. The maximum use temperature of glass fibers ranges from 930 ºF (500 °C) for E-glass up to 1920 ºF (1050 °C) for quartz. Batching consists of weighing the raw materials in exact quantities and thoroughly mixing them. Batching is automated using computerized weighing units and enclosed material transport systems. Each ingredient is transported by pneumatic conveyors to its designated multistory storage bin (silo), which is capable of holding 70 to 260 ft³ (2 to 7.5 m3) of material. Directly beneath each bin is an automated weighing and feeding system that transfers the precise amount of each ingredient to a pneumatic blender in the batch-house basement. Melting. From the batch house, another pneumatic conveyor sends the mixture to a high-temperature, approximately equals 2550 ºF (1400 °C), natural gas–fired furnace for melting. The furnace is typically divided into three sections with channels that aid glass flow. The first section receives the batch, where melting occurs and uniformity is increased, including the removal of bubbles. The molten glass then flows into the refiner, where its temperature is reduced to 2500 ºF (1370 °C). The final section is the forehearth, beneath which is located a series of four to seven bushings that are used to extrude the molten glass into fibers. Large furnaces have several channels, each with its own forehearth. Control of oxygen flow rates is crucial because

furnaces that use the latest technology burn nearly pure oxygen instead of air; pure oxygen helps the natural gas fuel to burn cleaner and hotter, melting glass more efficiently. It also lowers operating costs by using less energy and reduces nitrogen oxide (NOx) emissions by 75 percent and carbon dioxide (CO2) emissions by 40 percent. Two processes used to make high-strength glass fibers are the marble process and the direct melt process. In the marble process, the glass ingredients are shaped into marbles, sorted by quality, and then remelted into fiber strands. Alternatively, molten glass is introduced directly to form fiber strands. In the marble process, molten glass is sheared and rolled into marbles roughly 0.6 inches in diameter, which are cooled, packaged, and transported to a fiber-manufacturing facility where they are remelted for fiberization. The marbles allow visual inspection of the glass for impurities, resulting in a more consistent product. The direct melt process transfers molten glass in the furnace directly to fiber-forming equipment. Because direct melting eliminates the intermediate steps and the cost of forming marbles, it has become the more widely used method. Fiberization. Glass fiber formation, or fiberization, involves a combination of extrusion and attenuation. After being heated to approximately 2200 ºF (1200 °C), the molten glass flows or is extruded from the forehearth through an electrically heated platinum-rhodium alloy bushing or spinneret containing a large number (200 to 8000) holes in its base to form filaments, which are immediately quenched with water or an air spray to yield an amorphous structure (Fig. 2.5). Bushing plates are heated electrically, and the temperature is precisely controlled to maintain a constant glass viscosity. Attenuation is the process of mechanically drawing the extruded streams of molten glass into fibrous elements called filaments. A high-speed winder catches the molten streams and, because it revolves at a circumferential speed of approximately two miles per minute (much faster than the molten glass exiting the bushings), tension is applied, drawing them into thin filaments. Nozzle diameter determines filament diameter, and the number of nozzles equals the number of ends. Fiber diameter (typically around 5 to 20 mm) is controlled by hole size, draw speed, temperature, melt viscosity, and cooling rate. In typical glass fiber terminology, a number of individual strands (or ends) are usually incorporated into a roving to provide a convenient form for subsequent processing. Rovings are preferred

Chapter 2: Fibers and Reinforcements╇ /╇ 37╇

Fig. 2.5

Manufacturing processes for fiberglass fibers

for most reinforcements because they have higher mechanical properties than twisted yarns. Rovings are wound onto individual spools (Fig. 2.6) containing 20 to 50 lb (9 to 23 kg) of fiber. If the material is to be used for weaving, it is usually twisted into a yarn to provide integrity during the weaving operations. The strands are specified by their yield (yd/lb) or denier (the weight in grams of 9000 m of fiber). Another textile term frequently encountered is tex, which is the weight in grams of 1000 m of fiber. Coating. Because glass fibers are monolithic, linearly elastic brittle materials, their high strength depends on the absence of flaws and defects. Flaws are nanometer-size submicroscopic inclusions and cracks. Tensile strength depends on the internal stresses at the surface, which are different than in the interior due to the very high cooling rates on solidification. Although this surface layer is only about one nanometer thick, it is on the order of the flaw size that controls the strength of glass fibers. Virgin glass filaments are very susceptible to degradation from exposure to both air and mechanical abrasion. The tensile strength of as-drawn fibers can be reduced by more than 20 percent after contact with air during drawing under normal ambient conditions due to absorption of atmospheric moisture into

microscopic flaws, which reduces the fracture energy. Therefore, a sizing is applied immediately after manufacturing to prevent scratches from forming on the surface during spooling and from mechanical damage from weaving, braiding, and other textile processes. Sizings are extremely thin coatings that account for only about one to two percent by weight. The sizing (usually a starch and a lubricant) can be removed by either solvents or heat scouring after all mechanical operations are completed. After the sizing is removed, it is replaced with a surface finish that greatly improves the fiberto-matrix bond. For example, organosilane coupling agents have one end group that is compatible with the silane structure of the glass and another end group that is compatible with the organic matrix. The silane molecule on hydration in water can be represented by the following simplified formula:

R ... Si(OH)3

The silane bonds with the oxide film at the surface of the inorganic fiber glass, while the organic functional group R is incorporated into the organic matrix during its cure. Coupling agents are critical to the performance of the glass-reinforced

38╇ /╇ Structural Composite Materials

Fig. 2.6

Typical glass fiber product forms

composites. Improvements of over 100 percent have been demonstrated in composite tensile, flexural, and compression strength. The coupling agent also helps to protect the glass fiber from attack by water. Some sizings also function as coupling agents, and therefore remain on the fiber throughout the manufacturing process. The organic functional group R must be a group that is chemically compatible with the matrix resin. For example, for an epoxy resin, an epoxy silane may be used. Many different types of sizings/ finishes are available, and it is important that the one selected is compatible with both the fiber and the matrix. A partial list of some of the coupling agents for various resins include vinyl silane (methacrylate silane) used for polyester resins; Volan (methacrylate chromic chloride) used for polyester and epoxy resins; amino silane used for epoxy, phenolic, and melamine

resins; and epoxy silane used for epoxy and phenolic resins. Antistatic agents and lubricants can also be used to improve handling and processing characteristics such as hardness and softness. If the glass fiber will be chopped into short lengths for fiber spraying, hardness is a desirable property because it improves the ability to be chopped. On the other hand, if the fiber will be used in a lay-up operation where drapablility and forming are important, then softness is a desirable property. Drying/Packaging. Drawn, sized filaments are collected together into a bundle, forming a glass strand consisting of 51 to 1624 filaments. The strand is wound on a drum into a forming package. Forming packages, still wet from water cooling and sizing, are dried in an oven, after which they are ready to be palletized and shipped

Chapter 2: Fibers and Reinforcements╇ /╇ 39╇

or further processed into chopped fiber, roving, or yarn. Roving is a collection of strands with little or no twist. For example, an assembled roving might be made from 10 to 15 strands wound together into a multi-end roving package. Yarn is made from one or more strands, which are twisted to protect the integrity of the yarn during subsequent processing operations such as weaving. While glass has a high tensile strength comparable to that of some high-strength carbon fibers, its modulus and fatigue properties are inferior to those of carbon. In addition, glass fibers are susceptible to static fatigue in which their strength decreases with time under a static tensile load. While glass fibers do not absorb water into their bulk, water molecules can attach to their surface, forming a very thin, softened layer. This adsorbed layer can have an important effect on fiber strength because water and other surface active substances lead to surface microcracks. Humid environments reduce the strength of glass fibers under sustained loading, as the moisture adsorbed onto the surface of the flaw reduces the surface energy, thus promoting slow crack growth to a critical size.

2.4 Aramid Fibers Aramid fibers are organic fibers with stiffness and strength intermediate between those of glass and carbon. Dupont’s Kevlar fiber is the most prevalent. These aromatic polyamides are part of the nylon family. Aramid fibers are based on the amide linkage formed by the reaction between carboxolic acid and the amine group. When the linkage occurs between straight-chained satu-

Fig. 2.7

Chemical structure of para-aramid. Source: Ref 3

rated molecules, aliphatic amides known commercially as nylon are formed. When the linkage occurs between unsaturated benzene rings, aromatic amides called aramids are formed (Fig. 2.7). The aromatic ring structure contributes high thermal stability, while the para configuration leads to stiff, rigid molecules that contribute high strength and high modulus. Kevlar is made by reacting para-phenylene diamine with terephthaloyl chloride in an organic solvent to form polyparaphenylene terephthalamide (aramid). This is a condensation reaction that is followed by extrusion, stretching, and drawing. The polymer is washed and then dissolved in sulfuric acid. At this point, the polymer is a partially oriented liquid crystal. The polymer solution is then extruded through small holes (spinnerets). The fibers are oriented in solution and during spinning and are further oriented as they pass through the spinneret. The �fibers are then washed, dried, and wound up. The structure of aramid fibers (Fig. 2.8) consists of highly crystalline, aligned polymer chains separated into distinct zones or fibrils. The aromatic rings provide thermal stability and result in a crystalline, rigid, rod-like structure held together by strong covalent bonds. However, the bonds between the chains are relatively weak hydrogen bonds, which result in defibrillation failures in tension and kink band formation in compression (Fig. 2.9). The compression strainto-failure is only about 25 percent of the tensile value. As a result of this compression behavior, the use of aramid fibers in applications that are subject to high-strain compressive or flexural loads is limited. However, the compressive

40╇ /╇ Structural Composite Materials

Fig. 2.8

Conventional polymer and liquid crystal–like aramid polymers. Source: Ref 3

Fig. 2.9

Aramid fiber failure modes. Source: Ref 3

buckling characteristics have led to the development of crashworthy structures that rely on the fail-safe behavior of aramid composites under sustained high compressive loading. Due to their extreme toughness, aramid fibers are often used for ballistic protection. A major advantage of aramid fibers is their ability to absorb large amounts of energy during fracturing, which results from their high strain-to-failure values, their ability to undergo plastic deformation in compression, and their ability to defibril-

late during tensile fracture. The Kevlar fibrillar structure and compressive behavior contribute to composites that are less notch sensitive and that fail in a ductile, nonbrittle, or noncatastrophic manner, as opposed to glass and carbon. Unlike glass and carbon, or graphite, fibers, aramid fibers are not surface treated because no acceptable surface treatment for aramid fibers has been developed to date. Sizings are used when the fiber will be woven, made into rope, or used in ballistic applications. Aramid fibers are

Chapter 2: Fibers and Reinforcements╇ /╇ 41╇

lightweight and have a combination of good tensile strength and modulus, excellent toughness, and outstanding ballistic and impact resistance. However, due to the lack of adhesion to the Â�matrix, they exhibit relatively poor transverse tension, longitudinal compression, and interlaminar shear strengths. Like carbon and graphite, aramid fibers exhibit a negative coefficient of thermal expansion. The three most prevalent aramid fibers are Kevlar 29 (low modulus, high toughness), Kevlar 49 (intermediate modulus), and Kevlar 149 (high modulus). The property differences in the different grades of Kevlar are due to changes in process conditions that promote additional crystallinity in the intermediate- and high-modulus fibers. The tensile modulus of Kevlar fibers is a function of molecular orientation. Kevlar 29 has a modulus of 12 msi (83 GPa) that is slightly higher than that of E-glass 10 msi (70 GPa). Heat treatment under tension increases crystalline orientation, and the resulting fiber (Kevlar 49) has a modulus of 19 msi (130 GPa). Kevlar 149 has an even higher modulus approximately 27 msi (185 GPa); and is available on special order. Normal bundle size ranges from 134 to 10,000 filaments per bundle. Like carbon fibers, aramid fibers are available in tows or yarns of various weights that can be converted to woven cloth or chopped fiber mat. However, because the fibers are extremely tough, they are more difficult to cut, which causes some handling problems. Because aramid yarns and rovings are relatively flexible and nonbrittle, they can be processed in most conventional textile operations, such as twisting, weaving, knitting, carding, and felting. Yarns and rovings are used in filament winding, prepreg tape, and pultrusion processes. Applications include missile cases, pressure vessels, sporting goods, cables, and tension members. Aramid paper used in honeycomb sandwich constructions is made of Nomex aramid fiber. Nomex is chemically related to Kevlar, but its strength and modulus are considerably lower, more like those of conventional textile fibers. The aromatic chemical structure of aramid fibers imparts a high degree of thermal stability. They decompose in air at 800 °F (425 °C) and are inherently flame resistant, with a limiting oxygen index of 0.29. They can be used at a temperature range of -330 to +390 °F (-200 to +200 °C), but they are not generally used for extended periods at temperatures above 300 °F (150 °C) because of oxidation. Aramid fibers are resistant

to most solvents except strong acids and bases, and they have a strong tendency to absorb moisture. At 60 percent relative humidity, Kevlar 49 absorbs about four percent moisture and Kevlar 149 absorbs around 1.5 percent moisture due to its higher crystallinity. The effect of moisture on tensile strength at ambient temperatures is not significant. Aramid fibers are prone to significant short-term creep even at modest temperatures, but long-term creep is negligible. Aramid fibers are also prone to stress rupture under prolonged static loading, but they are much less sensitive to this mode of failure than glass fibers. Aramid fibers are degraded in strength by prolonged exposure to ultraviolet radiation, and while this can be a serious concern with cables that have exposed fibers, it is not a significant problem for aramid composites because the fibers are protected by the resin matrix.

2.5 Ultra-High Molecular Weight Polyethylene Fibers Gel-spun polyethylene fibers are ultra-strong, high-modulus fibers based on the standard polyethylene molecule. Gel-spun polyethylene fibers are produced from polyethylene having a very high molecular weight (UHMWPE). This material is chemically identical to normal high-density polyethylene, but the molecular weight is higher than that of the commonly used polyethylene grades. The chemical nature of polyethylene remains in the gel-spun fiber, which can be both an advantage and a limitation; abrasion and flex life are very high, but the melting point is sometimes too low for use in many applications. UHMWPE fibers are commercially produced under the trade names Dyneema in the Netherlands and Japan and Spectra in the United States. In the gel-spinning process, the molecules are dissolved in a solvent and spun through a spinneret. In solution, the molecules become disentangled and remain in that state after the solution is spun and cooled to produce filaments. Because of its low degree of entanglement, the gelspun material can be drawn to a very high extent (super-drawn). As the fiber is superdrawn, a very high level of macromolecular orientation is attained (Fig. 2.10), resulting in a fiber with a very high strength and modulus. Gel-spun fibers are characterized by a high degree of chain extension, parallel orientation greater than 95 percent, and a high level of crystallinity (up to 85 percent).

42╇ /╇ Structural Composite Materials

cause the fibers cannot be drawn to their full extent in commercial production. However, sliding of the polymer chains encouraged by the weak interchain bonding also makes a significant contribution.

2.6 Carbon and Graphite Fibers

Fig. 2.10

 acromolecular orientation of ultra-high molecuM lar weight (UHMWPE) and normal polyethylene (PE). Source: Ref 4

Owing to their low density and good mechanical properties, their performance on a weight basis is extremely high. At a density of 0.97 g/cm3, they are even lighter than aramid fibers; they are so light that they will float on water. They have high impact resistance and attractive electrical properties (in this case, low dielectric constant and low loss tangent), and they have exceptional chemical resistance and low moisture absorption characteristics. Due to the relatively low melting point of polyethylene, UHMW-polyethylene fibers are limited to a maximum use temperature of 200 ºF (95 °C). For example, Spectra melts at a temperature of around 300 °F (150 °C). On a specific basis, polyethylene fibers have tensile properties exceeding those of most other fibers, including aramid. Like aramid fibers, polyethylene fibers exhibit low compression strength and fail under compression through kink band formation. Also, they do not form a strong bond to the matrix, which results in poor transverse tension and compression strengths. Because they are thermoplastic, polyethylene fibers are subject to creep under continuous loading, even at modest temperatures, which limits their use in applications involving high, prolonged static loading. Creep occurs in part be-

Carbon and graphite fibers are the most prevalent fiber forms used in high-performance composite structures. They can be produced with a wide range of properties. They generally exhibit superior tensile and compressive strength, have high moduli, have excellent fatigue properties, and do not corrode. Although the terms are often used interchangeably, graphite fibers are subjected to heat treatments above 3000 ºF (1650 °C), have three-dimensional ordering of their atoms, have carbon contents greater than 99 percent, and have elastic moduli (E) greater than 50 msi (345 GPa). Carbon fibers have lower carbon contents (93 to 95 percent) and are heat treated at lower temperatures. The graphite structure consists of carbon atoms arranged in a lamellar structure of hexagonal layers (Fig. 2.11). The high strengths and moduli of carbon and graphite fibers are a result of the strong covalent bonds (≈525 kJ/mol) along the basal planes (in this case, the ABABAB stacking sequence). Carbon fibers actually consist of graphite and nongraphitic carbonaceous material. The graphitic phase is in the form of crystallites with discrete dimensions that can be oriented difÂ� ferently from each other, with high-stiffness carbon fibers containing a large portion of graphite aligned in the fiber direction. As the graphitization heat treatment temperature increases, the planes become oriented more parallel to the fiber axis. Carbon and graphite fibers are extremely anisotropic. Because the covalently bonded basal planes are held together by rather weak van der Waals forces (≈10 kJ/mol), the transverse strength and modulus of the fiber are much less than the longitudinal values. For example, the longitudinal modulus can be as high as 145 msi (1000 GPa), while the transverse modulus may only be approximately 5 msi (35 GPa). Carbon and graphite fibers can be made from rayon, polyacrylonitrile (PAN), and petroleumbased pitch precursors. Although PAN fibers are more expensive than rayon fibers, PAN is used extensively for structural carbon fibers because the carbon yield is almost double that of rayon fibers. The pitch process produces fibers that

Chapter 2: Fibers and Reinforcements╇ /╇ 43╇

Fig. 2.11

Structure of graphite crystal

have lower strength than those produced from PAN, but it can produce high-modulus fibers 50 to 145 msi (345 to 1000 GPa). The generalized manufacturing processes for PAN and pitchbased carbon fibers are shown in Fig. 2.12. Rayon-Based Carbon Fibers. Rayon-based fibers originate from wood pulp in which the cellulose is extracted and the tows are produced by wet spinning. The fibers are then heat treated at 400 to 800 °F (205 to 410 °C) to start the conversion process to carbon. Unlike PAN- and pitch-based fibers, rayon-based fibers do not require a stretching and heat-setting process prior to carbonization; they can be carbonized without melting. During carbonization and graphitization, there is a gradual ordering of the structure. However, the alignment is rather poor compared to that of PAN- and pitch-based fibers. To improve the strength and modulus of rayon-based fibers, it is necessary to hot stretch the graphitized fibers up to 50 percent at 5000 to 5500 °F (2760 to 3040 °C), which increases the orientation of the basal planes; they rotate toward the fiber axis, and the porosity within the fibers is reduced. The disadvantages of rayon-based carbon fibers are (1) the decreasing availability of a suitable rayon precursor; (2) a low yield, only 15 to 30 percent; and (3) problems with hot stretching the fiber at the end of the cycle. However, rayon-based carbon fibers work well in carbon-carbon and carbon-phenolic composites, where fibers produced with low moduli minimize microcracking of the matrix during fabrication.

PAN-Based Carbon Fibers. Polyacrylonitrilebased fibers exhibit a high degree of orientation in the raw precursor form and have much higher yields of approximately 50 percent than rayonbased fibers. The production process of PANbased carbon fibers can be divided into five steps: 1. Spinning and stretching the PAN copolymer to form a fiber 2. Stabilization and oxidation in air at 390 to 570 ºF (200 to 300 °C) under tension 3. Carbonization in an inert atmosphere at 1800 to 2900 ºF (980 to 1595 °C) 4. Graphitization in an inert atmosphere at 3600 to 5500Â�ºF (1980 to 3040 °C) 5. Surface treatment and sizing Polyacrylonitrile is formed by the polymerization of the acrylonitrile monomer, an olefin derived from the substitution of the nitrile group for a hydrogen. Because PAN decomposes before melting, it is necessary to make a solution using a solvent such as dimethyl formamide to be able to spin the material into a fiber. The spinning operation can be done either dry, where the solvent evaporates in a spinning chamber, or wet, where the fiber is placed in a coagulating bath solution. In dry spinning, the rate of solvent removal is greater than the rate of diffusion within the fiber, and the surface of the filament hardens faster than the interior, resulting in its collapse and the formation of a dogbone cross section. In wet spinning, the fiber dries uniformly and the cross section is circular. Of the two processes,

44╇ /╇ Structural Composite Materials

Fig. 2.12

Polyacrylonitrile (PAN) and pitch-based carbon fiber manufacturing processes

only wet-spun PAN is currently used for carbonfiber precursors. The spun fiber is composed of a fibrillar, or ribbon-like, network, which acquires a preferred orientation parallel to the fiber axis, provided that the fiber is stretched either while it is still in the coagulating bath or subsequently in boiling water. Stretching results in an elongation of 500 to 1300 percent, and is an essential step in obtaining high-strength fiber. Heat setting or oxidation crosslinks the PAN and stabilizes the structure so that it will not melt during the carbonization process. The heat-setting process converts the thermoplastic PAN into a nonplastic cyclic or ladder compound capable of withstanding the high temperatures used during carbonization. The PAN fibers are heated to 390 to 570 ºF (200 to 300 °C) in air for one to two hours while the fibers are under tension. Sufficient tension, causing elongation of 300 to 500 percent, is used to unfold the tightly folded chain molecules. The oxidizing environment causes the unfolded chains to crosslink with oxygen molecules, replacing hydrogen molecules on adjacent chains. Hydrogen combines with excess

oxygen and is evolved as water vapor. Oxidation causes the formation of C=C bonds and the incorporation of hydroxyl (-OH) and carbonyl (-CO) groups in the structure, which promotes crosslinking and thermal stability. The product at this stage is often referred to as oxy-PAN. Carbonization is conducted in a nitrogen atmosphere at 1800 to 2900 ºF (980 to 1595 °C) and converts the PAN to carbon. During carbonization, the fibers shrink in diameter and lose approximately 50 percent of their weight. As the PAN is slowly heated to the carbonization temperature, approximately equals 40 °F/min (22 °C/min), considerable amounts of volatile by-products are released, including water, carbon dioxide, carbon monoxide, ammonia (NH3), hydrogen cyanide (HCN), methane (CH4), and other hydrocarbons. The carbon yield is 50 to 55 percent. The circular morphology of the fiber is maintained, and the final diameter varies from 5 to 8 mm, which is approximately half that of the precursor PAN fiber. The removal of nitrogen occurs gradually over a range of temperatures: Nitrogen evolution starts at 1110 °F

Chapter 2: Fibers and Reinforcements╇ /╇ 45╇

(600╯°C), reaching a maximum at 1650╯°F (900╯°C), approximately six percent remains at 1830 °F (1000 °C), and only 0.3 percent remains at 2380 °F (1305 °C). Carbonization results in a carbon network of hexagonal ribbons, known as turbostatic graphite, that tends to align parallel to the fiber axis (Fig. 2.13). The crystal structure is very small, which contributes to its high strength. If a true graphite fiber is desired, the fiber is graphitized at temperatures between 3600 and 5500 °F (1980 and 3040 °C), which produces a more crystalline structure and a higher elastic modulus. The final carbon content is greater than 99 percent. This treatment completes the conversion of the remaining carbonaceous material to graphite. The graphite tends to align the basal planes in the direction of the fiber. However, during the process, the crystallite size can increase. The increase in graphite content results in increased stiffness, while the increase in crystallite size lowers the strength. Carbon-fiber costs reflect the cost of heat treatment (Fig. 2.14). Higher-modulus fibers re-

Fig. 2.13

Carbon fiber structure. Source: Ref 5

quire higher heat-treating temperatures to produce the greater amounts of aligned graphite. Graphitization improves the degree of alignment of one ribbon to the next along the fiber axis. As the final graphitization temperature increases, the elastic modulus increases, while the strength reaches a maximum and then decreases in the manner shown in Fig. 2.15. When carbon fibers were first introduced in the 1960s, it was soon realized that the asmanufactured carbon fiber did not bond well to epoxy resins. The importance of the effect of a proper surface treatment on fiber-to-matrix adhesion is shown in Fig. 2.16. Untreated carbon fibers exhibit almost no adhesion to the epoxy matrix. While poor adhesion does not affect 0° tensile strength, it does adversely affect matrixdependent properties such as 90° tensile strength, 0° compressive strength, and interlaminar and in-plane shear strength. Therefore, in the next to last production step, the carbon fibers are subjected to an electrolytic oxidation that removes weak surface layers, etches the fibers,

46╇ /╇ Structural Composite Materials

Fig. 2.14

Carbon fiber cost increases with the heat treatment used to obtain a higher modulus. Source: Ref 1

Fig. 2.15

Effect of heat treatment on the strength and modulus of PAN-based carbon fiber. PAN, polyacrylonitrile. Source: Ref 6

Chapter 2: Fibers and Reinforcements╇ /╇ 47╇

Fig. 2.16

Effect of surface treatments on fiber-to-matrix bonds: (a) good bond; (b) poor bond. Source: Ref 7

and generates reactive, or polar, groups. The surface treatment attaches carboxyl, carbonyl, and hydroxyl groups to the fiber surface, which can bond to the polymeric matrix. The fiber emerges from the heat-treating furnace and passes around a positively charged roller (anode) and into an aqueous electrolytic cell. Various electrolytes, such as sodium hydroxide (NaOH), are used to conduct current and to create surface groups. The fiber is washed and dried before it enters the size-application bath. The surface generated by this treatment greatly improves adhesion to thermosets and to some thermoplastic resins. The last step is the application of a protective material called sizing to the carbon fiber. If the fiber will be woven, sizes (usually uncatalyzed epoxy) are applied to the fiber to protect the fiber surface from mechanical abrasion. Size, sizing, and finish are all names for coatings applied to carbon fibers to make them easier to handle. The size protects the small filaments from damage by holding them together and reducing friction. Only a small amount of sizing is used, typically 0.5 to 1.5 percent. It should be noted that sizes for carbon fibers are not coupling agents, as are silane compounds used with glass fibers. Pitch-Based Carbon Fibers. Pitch-based carbon and graphite fibers are made by heating coal

tar pitch for up to 40 hours at 800 °F (425 °C), forming a highly viscous liquid having a high degree of molecular order known as a mesophase. The mesophase is spun through a small orifice that aligns the molecules along the fiber axis. Pitch-based fibers are then processed following the same basic steps used in PAN-based fiber manufacturing, namely, carbonization, graphitization, and surface treatment. Pitch is a by-product of the distillation of coal, crude oil, and asphalt. Its carbon yield can exceed 60 percent, which is appreciably higher than the yield of PAN approximately 50 percent. The composition of pitch includes four generic fractions in variable proportions: (1) saturates, which are aliphatic compounds having a low molecular weight similar to that of wax; (2) napthene aromatics, which are low molecular weight compounds having a saturated ring; (3) polar aromatics, which are medium molecular weight compounds with some heterocyclic molecules; and (4) asphaltenes, which have a high molecular weight and a high degree of aromaticity. The higher the ratio of asphaltene, the higher the softening point, thermal stability, and carbon yield. Pitch-based fibers can be divided into two groups: isotropic pitch fibers, which have low mechanical properties but are relatively low in cost, and mesophase pitch fibers, which have

48╇ /╇ Structural Composite Materials

very high moduli but are more expensive. Lowcost carbon fibers are produced from an isotropic pitch with a low softening point. The precursor is melt spun, thermoset at a relatively low temperature, and carbonized. The resulting fibers generally have low strength and a low modulus. Carbon fibers from mesophase pitch have medium strength and a high modulus. The processing of mesophase pitch fibers is similar to that of PAN fibers, except that the stretching step during heat treatment is not necessary. The processing steps can be summarized as follows: 1. Polymerization of the isotropic pitch to produce mesophase pitch 2. Spinning the mesophase pitch to obtain a green fiber 3. Thermosetting the green fiber 4. Carbonization and graphitization to obtain a high-modulus carbon or graphite fiber The pitch is heated to approximately 750 °F (400 °C) and is transformed from an isotropic to a mesophase, or liquid crystal, structure consisting of large polyaromatic molecules with oriented layers in parallel stacking. The mesophase pitch is melt spun in a mono- or multifilament spinneret heated to 570 to 850 °F (300 to 455 °C) and pressurized with inert gas. It is drawn at a speed greater than 400 ft/min (122 m/min) with a draw ratio of approximately 1000 to l to a diameter of 10 to 15 mm (0.4 to 0.6 mil). The draw ratio is an important factor in the control of the orientation of the fiber structure; the higher the draw ratio, the greater the orientation and uniformity. At this stage the fiber is thermoplastic, and a thermosetting operation is needed to avoid relaxation of the structure and to prevent the filaments from fusing together. This thermosetting operation is carried out in an oxygen atmosphere or in an oxidizing liquid at approximately 570 °F (300 °C), causing oxidation crosslinking and stabilization of the filament. Temperature control during this thermosetting step is critical, because a temperature that is too high will relax the material and eliminate its oriented structure. The thermoset fibers are then carbonized at temperatures of up to 1830 °F (1000 °C). This is done slowly to prevent rapid gas evolution and the formation of bubbles and other flaws. Carbonization is followed by high-temperature graphitization at 5000 to 5500 °F (2760 to 3040 °C). Pitch-based graphite fibers have a higher modulus and lower strength than PAN-based carbon fibers. In addition, pitch-based fibers tend to have more flaws such as pits, scratches,

striations, and flutes. These flaws are detrimental to tensile properties but do not necessarily affect the modulus and thermal conductivity. As with PAN-based fibers, as the heat treatment temperature during carbonization/graphitization is increased, the fiber strength increases to a maximum and then decreases, while the modulus continues to increase. The higher temperatures used in the graphitization process for graphite fibers result in more orientation of the graphite cystallites parallel to the fiber axis. The better the alignment of the crystallites, the higher the modulus of the fiber. However, high crystallinity also causes the fiber to be weak in shear, which results in lower compressive strength. Therefore, high-crystalline graphite fibers do not exhibit balanced tensile and compressive mechanical properties. Pitch-based, high-modulus graphite fibers having a modulus between 50 and 145 msi (345 and 1000 GPa) are often used in space structures requiring high rigidity. In addition to their high modulus and low thermal expansion, pitch-based graphite fibers have high values of thermal conductivity—for example, 900 to 1000 W/mK compared to only 10 to 20 W/mK for PAN-based carbon fibers. The large crystallites in graphite fibers are structurally close to the perfect graphite crystal and well aligned along the fiber axis, offering few scattering sites for phonons. This means that these fibers have high thermal conductivity along the fiber axis. Fibers with the highest degree of orientation, such as the pitch-based fibers, have the highest thermal conductivity. Their conductivity along the axis is higher than that of even the best metal conductor. Polyacrylonitrile-based fibers have much lower thermal conductivity because of their more pronounced isotropic structure. These high thermal conductivities are used to remove and dissipate heat in space-based structures. High-modulus graphite fibers can also be manufactured using the PAN process. However, the highest modulus attainable is around 85 msi (585 GPa). The strength of carbon and graphite fibers depends on the type of precursor used, the processing conditions during manufacturing such as fiber tension and temperatures, and the presence of flaws and defects. Flaws in the carbon fiber microstructure include internal pits and inclusions, external gouges, scratches, and stuck filament residues, as well as undesirable characteristics such as striations and flutes. These flaws can have a considerable impact on fiber tensile strength but have little, if any, effect on modulus,

Chapter 2: Fibers and Reinforcements╇ /╇ 49╇

conductivity, and thermal expansion. Both carbon and graphite fibers usually have a slightly negative coefficient of thermal expansion, which becomes more negative as the modulus of elasticity increases. One consequence of using high and ultrahigh-modulus carbon fibers is the increased possibility of matrix microcracking during processing or environmental exposure due to the larger mismatch in the coefficients of thermal expansion between the fibers and the matrix. Carbon fibers are available from a number of domestic and foreign producers having a wide range of strength and moduli. Polyacrylonitrilebased carbon fibers having strengths ranging from 500 to 1000 ksi (3.5 to 7 GPa) and moduli ranging from 30 to 45 msi (205 to 310 GPa) with elongations of up to two percent are commercially available. Standard-modulus PAN fibers have good properties with lower cost, while higher-modulus PAN fibers are higher in cost because high processing temperatures are required. Heating the fibers to 1800 °F (980 °C) yields PAN fibers containing 94 percent carbon and six percent nitrogen, while heating to 2300 °F (1260 °C) removes the nitrogen and raises the carbon content to around 99.7 percent. Higher processing temperatures increase the tensile modulus by refining the crystalline structure and the three-dimensional nature of the structure. Carbon fiber diameters usually range from 0.3 to 0.4 mil (7.6 to 10 mm). Carbon fibers are provided in untwisted bundles of fibers called tows. Tow sizes can range from as little as 1000 fibers/tow up to more than 200,000 fibers/tow. A typical designation of “12k tow” indicates that the tow contains 12,000 fibers. Normally, as the tow size decreases, the strength and cost increase. Due to its cost, the small 1k tow size is normally not used unless the property advantages outweigh the cost disadvantages. For aerospace structures, normal tow sizes are 3k, 6k, and 12k, with 3k and 6k being the most prevalent for woven cloth and 12k for unidirectional tape. It should be noted that there are very large tow sizes greater than 200k, which are primarily used in commercial applications and are normally broken down after manufacturing into smaller tow sizes such as 48k for subsequent handing and processing. The cost of carbon fibers depends on the manufacturing process, the type of precursor used, the final mechanical properties, and the tow size. Costs can vary from less than $10 per pound for large-tow commercial fibers to several hundred or even several thousand dollars per pound for small-tow, ultra-

high-modulus, pitch-based fibers. The maximum use temperature of carbon and graphite fibers in an oxidizing atmosphere is 930 ºF (500 °C). The ideal engineering material would have high strength, high stiffness, high toughness, and low weight. Carbon fibers combined with polymer matrices meet these criteria more closely than any other material. Carbon fibers are elastic to failure at normal temperatures, creep resistant and not susceptible to failure; are chemically inert except in strong oxidizing environments or in contact with certain molten metals; and have excellent damping characteristics. Some disadvantages of carbon fibers are brittleness and low impact resistance, low strains-to-failure, lower compressive strengths than tensile strengths, and higher in cost than glass fibers.

2.7 Woven Fabrics Two dimensional woven products (Fig. 2.17) are usually offered as a 0º, 90º construction. However, bias weaves (45º, 45º) can be made by twisting the basic 0º, 90º construction. Weaves are made on a loom by interlacing two orthogonal (mutually perpendicular) sets of yarns (warp and fill). The warp direction is parallel to the length of the roll, while the fill, weft, or woof direction is perpendicular to the length of the roll. Textile looms (Fig. 2.18) produce woven cloth by separation of the warp yarns and insertion of the fill yarns. Most weaves contain similar numbers of fibers and use the same material in both the warp and fill directions. However, hybrid weaves (Fig. 2.19), such as carbon and glass, and weaves dominated by warp yarns can also be produced. These hybrid weaves can be used to obtain specific properties, such as mixing carbon with aramid to take advantage of the toughness of aramid, or to reduce costs, such as mixing glass with carbon fibers. Woven broadgoods may be purchased either as a dry preform or preimpregnated with a B-staged resin. In most applications, multiple layers of two-dimensional weaves are laminated together. As with tape laminates, layers are oriented to tailor the strength and stiffness. Weaves may be classified by the pattern of interlacing, as shown in Fig. 2.20. Probably the two most prevalent weaves used in high-performance composites are the plain and satin weaves, which are compared in Fig. 2.21. The simplest pattern is the plain weave, in which every warp and fill yarn goes alternatively over and then under

50╇ /╇ Structural Composite Materials

Fig. 2.17

Two-dimensional dry woven carbon cloth

successive warp and fill yarns, respectively. Plain weaves have more interlaces per unit area than any other type of weave, and therefore the tightest basic fabric design, and are the most resistant to in-plane shear movement. Therefore, plain weaves resist distortion during handling but may be difficult to form on complex contours. They are also more difficult to wet-out during impregnation. Another disadvantage of the plain weave is the frequent exchanges of position from top to bottom made by each yarn. This waviness, or yarn crimp, reduces the strength and stiffness of the composite. The basket weave is a variation of the plain weave in which two (or more) warp and two (or more) fill yarns are woven together. An arrangement of two warps crossing two fills is designated as a 2â•›×â•›2 basket, but the arrangement of fibers need not be symmetrical; it is possible to have 8â•›×â•›2, 5â•›×â•›4, and other variations. The basket weave has less crimp than the plain weave and is therefore somewhat stronger. Satin weaves are characterized by a minimum of interlacing and, therefore, have less resistance to in-plane shear movement and have the best drapablility. Plain weaves are often used for less curved parts, while harness satin weaves are used for more highly contoured parts. In the four-harness satin weave, the warp yarns skip over three fill yarns and then under one fill yarn.

In the five-harness satin weave, the warp yarns skip over four fill yarns and then under one fill yarn. In the eight-harness satin weave, the warp yarns skip over seven fill yarns and then under one fill yarn. Due to less fiber crimp, satin weave fabrics are stronger than plain weave fabrics. They also provide smooth part surface finishes at a minimum per ply thickness. The eight-harness satin weave has the best drapablility of this group. However, five-harness weaves normally use a 6k carbon tow, which is less expensive than the 3k carbon tow used for eight-harness satin weaves. The trend in industry has been to move toward greater use of the less expensive fiveharness satin weaves. Twill weaves are occasionally used because they have better drapablility than the plain weaves and are known for their extremely good wet-out during impregnation. In this weave, one or more warps alternately weave over and under two or more fills in a regular repeating manner, producing the visual effect of a straight or broken “rib” to the fabric. The leno and mock leno weaves are rarely used for structural composites. The leno weave is also a form of the plain weave in which the adjacent warp fibers are twisted around consecutive fill fibers to form a spiral pair, effectively locking each fill in place. This construction produces an extremely open fabric with a low fiber

Chapter 2: Fibers and Reinforcements╇ /╇ 51╇

Fig. 2.18

Weaving glass cloth on a textile loom

content. The leno weave is frequently used to tie the edges of dry fabric together so that it will not unravel during handling. The mock leno (also a version of the plain weave) has occasional warp fibers at regular intervals but several fibers apart, which deviate from the alternate under-over interlacing and instead, interlace every two or more

fibers. This happens with similar frequency in the fill direction, and the overall result is a fabric having increased thickness, a rougher surface, and more porosity. Woven fabrics often contain tracer yarns. For example, yellow aramid tracers are often woven on two-inch (50 mm) spacings in carbon cloth

52╇ /╇ Structural Composite Materials

of glass reinforcements. These products form a drapable reinforcement that combines the bidirectional fiber orientation of woven roving with the multidirectional fiber orientation of chopped strand mat. This saves time in hand lay-up since two layers can be placed in the mold in a single operation. Other combinations are available for surface finish improvement as well as for multilayer reinforcement.

2.9 Chopped Fibers Fig. 2.19

Examples of hybrid weaves

along the warp direction to help fabricators identify warp and fill direction during composite part lay-up. The selection of a weave involves manufacturing considerations as well as final mechanical properties. The weave type affects dimensional stability and the conformability (or drape) of the fabric over complex surfaces. For example, satin weaves exhibit good conformability. Unfortunately, good conformability and resistance to shear are mutually exclusive. Thus, while woven fabrics are frequently the material of choice for complex geometries, the designer must be aware that specified material directions may be impossible to maintain on compound contours and other complex shapes; that is, initially orthogonal yarns may not remain orthogonal in the finished product. Commonly used weave styles for highperformance composites are shown in Table 2.2.

2.8 Reinforced Mats Reinforced mats (Fig. 2.6) are made of either chopped strands or continuous strands layed down in a swirl pattern. Mats are generally held together by resinous binders. They are used for medium-strength parts having uniform cross sections. Both chopped and continuous-strand reinforcing mats are available in weights varying from 0.75 to 4.5 oz/ft2 (240 to 1430 g/m2) and in various widths. Surfacing mats, or veils, are thin, lightweight materials used in conjunction with reinforcing mats and fabrics to provide good surface finish. They are effective in blocking out the fiber pattern of the underlying mat or fabric. Combination mats, consisting of one ply of woven roving chemically bonded to chopped strand mat, are available from several processors

Chopped fibers (Fig. 2.6) produce higher strength in compression and in injection-molded parts. Chopped fibers are usually available in lengths ranging from 0.125 to 2 in. (3.2 to 50 mm), although shorter milled fibers and longer fibers are available. They are blended with resins and other additives to prepare molding compounds for compression or injection moldings, encapsulation, and other processes. Choppedglass reinforcement is available with many surface treatments to ensure optimum compatibility with most thermosetting and thermoplastic resin systems. The shorter chopped reinforcements are best suited for blending with thermoplastic resin systems for injection molding. Longer chopped reinforcements are blended with thermosetting resins for compression and transfer molding. Milled fibers combine reinforcing properties with processing ease in encapsulation or injection molding. Milled fibers are 0.03125 to 0.125 in. (0.8 to 3.2 mm) lengths of fibrous glass. They are used to reinforce thermoplastic parts where strength requirements are low to moderate and for reinforcing fillers in adhesives.

2.10 Prepreg Manufacturing Prepreg is an important product form in which either unidirectional fibers or woven cloth is impregnated with a controlled amount of resin. The resin is staged or advanced (B-staged) to the point where the resin in the prepreg is a tacky semisolid, which allows the layers to be layed up to form a laminate that can be cured. A resin goes through several stages in the manufacturing process. Resins are normally made by batch manufacturing, in which the ingredients are placed in mixers (Fig. 2.22) and slowly heated to the A-stage condition, or initial mixed state, where the resin has a very low viscosity, which allows

Chapter 2: Fibers and Reinforcements╇ /╇ 53╇

Fig. 2.20

Common two-dimensional weave styles. Source: Ref 8

flow and impregnation of the fibers. Because the resins and curing agents used in composite matrices can be quite reactive, careful temperature control during mixing is critical to prevent an exothermic reaction, which could result in an

explosion or a fire. Some components may require premixing before being added to the main mix. After mixing, the resin is usually placed in plastic bags and frozen until it is needed for prepregging or until it must be shipped for processes

54╇ /╇ Structural Composite Materials

Fig. 2.21

Plain and satin weave cloths are most often used in high-performance composites. Source: Ref 1

Table 2.2â•… Common weave styles used in highperformance composites


Style 120 Style 7791 Style 120 Style 285 8-harness satin 5-harness satin 5-harness satin Plain

Type of fiber

Construction yarns/in. warp × fill

Fiber areal weight, g/m2

Approximate cured per ply thickness(a), in.

E-glass E-glass Kevlar 49 Kevlar 49 3K carbon 6K carbon 1K carbon 3K carbon

60 × 58 57 × 54 34 × 34 17 × 17 24 × 23 11 × 11 24 × 24 11 × 11

107 303 61 17 370 370 125 193

0.005 0.010 0.004 0.010 0.014 0.014 0.005 0.007

(a) Actual cured per ply thickness depends on resin system, resin content, and processing conditions.

such as wet filament winding, liquid molding, or pultrusion. Prepreg is the most prevalent product form used in advanced composite manufacturing. It usually consists of a single layer of fibers embedded in a B-staged resin (Fig. 2.23). During prepregging, the resin advances to a B-stage condition in which it is a semisolid at room temperature, and remelts and flows during the cure cycle. The B-staged resin normally contains some tack, or stickiness, to allow it to adhere to itself and tooling details during the lay-up operation. Because the resin is in a state of continual advancement (i.e., reaction), the degree of advancement and the resultant tack and flow be-

havior will change unless it is kept refrigerated when not in use. Variables that define a prepreg are fiber type, fiber form (such as unidirectional or woven), resin type, fiber areal weight (FAW), prepreg resin content, and cured per ply thickness. Fiber areal weight is simply the weight of the fibers in a given area and is usually specified in g/m2. Prepreg resin content specifies the percentage of resin by weight in the prepreg; this is not necessarily the resin content of the cured part. Some resins are prepregged with excess resin (e.g., 42 percent), which will be bled off (removed) during cure to yield a final cured resin content of 28 to 30 percent by weight. Other resins, called net-resin-content prepregs, will be prepregged to almost the same resin content as the final cured part; therefore, no bleeding of excess resin is required for this product form. Cured per ply thickness is specified as the thickness of each ply in inches. It should be noted that the final cured per ply thickness is strongly dependent on part configuration (especially part thickness) and the user’s processing conditions. Prepreg is usually supplied as narrow unidirectional tape, roving (or towpreg), wide unidirectional tape, and woven cloth called broadgoods (Fig. 2.24). Prepreg rovings or tows are bundles of fibers that are used primarily for filament winding or fiber placement. As the name

Chapter 2: Fibers and Reinforcements╇ /╇ 55╇

Fig. 2.22

Resin-mixing schematics. Sources: Photo, Cytec Engineered Materials; figure, Ref 9

implies, a single bundle of fibers is impregnated with the resin during prepregging. The cross section of both product forms is a flat rectangle with a width between 0.100 and 0.250 in (2.5 to 6.4 mm). The material is supplied in long lengths, up to 20,000 ft (6100 m), on a single spool. Unidirectional tape prepreg is a combination of mul-

tiple tows aligned in parallel that are impregnated with resin. Typical FAWs range from as low as 30 to as high as 300 g/m2, with typical values being 95, 145, and 190 g/m2, which correspond to cured per ply thicknesses of 0.0035, 0.005, and 0.0075 in. (0.09, 0.13, and 0.19 mm), respectively. Widths range from 6 to 60 in. (15 cm to

56╇ /╇ Structural Composite Materials

Fig. 2.23

Unidirectional and woven cloth prepreg. Source: Ref 10

Fig. 2.24

Unidirectional prepreg tape, towpreg, and broadgoods (unidirectional tape and woven cloth). Source: The Boeing Company

Chapter 2: Fibers and Reinforcements╇ /╇ 57╇

1.5 m). Automated tape-laying machines usually use 6 or 12 in. (15 or 30 cm) wide material, while the wider 48 to 60 in. (1219 to 1524 mm) broadgoods are machine cut into ply shapes and used for hand lay-up. Fabric prepregs consist of a woven fabric impregnated with a resin. Because fabric prepregs are primarily used for hand lay-up, the material is usually supplied as wide rolls (up to 60 in. (1.5 m) wide) to minimize the number of splices required in a part. Fabric prepregs usually have higher FAWs than unidirectional tape and thicker cured per ply thickness such as 0.014 in./ply. Prepregs normally produce the highest laminate strengths when they are layed up and autoclave cured. Some typical cured carbon/epoxy properties for unidirectional tape materials are shown in Table 2.3, and woven cloth laminate properties are given in Table 2.4. Prepregging can be accomplished by hot melt impregnation, resin filming, and solvent impregnation. In the original hot melt process (Fig. 2.25), fibers are fed from a creel, collimated, impregnated with the melted resin, and immediately cooled prior to spooling on the roll. The newer resin filming process consists of two operations. The resin is first filmed to a controlled thickness on backing paper, as shown in Fig. 2.26. The spooled film can then be either taken directly to the prepregging operation or frozen for future use. The majority of prepreg is Table 2.3â•…Typical properties of 0° tape carbon/ epoxy laminates AS-4

Fiber type AS-4 IM-6



Resin type 3501-6 3501-6


0º tensile strength, ksi 0º tensile modulus, msi 0º compression strength, ksi Interlaminar shear, ksi Compression after impact, ksi

275 20.3 240 18.5 22

310 20.5 240 17.5 22

330 20.4 221 18.6 33

350 23.0 240 18.0 22


Source: Ref 11

currently made by the filming technique, because it allows better control of resin content and FAW. Typical filming weights are in the range of 0.06 to 0.25 oz/ft2 (20 to 80 g/m2) with speeds of up to 40 ft/min (12 m/min). When the resin film is ready for impregnation, it is conducted on a separate machine (Fig. 2.27) in which the fiber web is protected on both surfaces with a backing paper. Impregnation is achieved by the application of heat and nip-roll pressure as the fibers, resin film, and upper and lower backing papers are pulled through the line. After the material passes through the second set of nip rollers, it is immediately chilled to raise the resin’s viscosity and produce the semisolid prepreg. At the exit, the upper paper sheet is removed and discarded, the edges are trimmed straight with slitter blades, and the finished prepreg is rolled up on the spool. This process runs at the rate of about 8 to 20 ft/min (2.5 to 6 m/ min) in widths of up to 60 in. (1.5 m). A third method, solvent impregnation (Fig. 2.28), is used almost exclusively for towpreg, woven fabrics, and high-temperature resins (e.g., polyimides) that are not amenable to hot melt prepregging and must be dissolved in solvents. A disadvantage of this process is that residual solvent may remain in the prepreg and cause a volatile evolution problem during cure. Therefore, the recent trend has been to use hot melt or resin film methods for both unidirectional and fabric prepregs. The solution process is operated with a treater line. The fabric web is drawn off a reel into a dip tank containing the resin solution (acetone is a common solvent for epoxies) and pulled through a controlled set of nip rollers to set the resin content. Typical speeds are 10 to 15 ft/min (3 to 4.5 m/min) for 60 in. (1.5 m) wide material. The web moves to a hot-air oven that serves to both evaporate the bulk of the solvent and advance the resin for tack control. At the end of the oven, the material is spooled up with a layer of plastic film applied to one side to

Table 2.4â•…Typical properties of woven carbon/epoxy cloth laminates Style weave

Fiber type Tow size Areal weight, g/m2 Tensile strength, ksi Tensile modulus, msi Interlaminar shear, ksi Cured per ply â•… thickness, mils Source: Ref 11

A193-P plain

A280-5H 5-harness satin

A370-5H 5-harness satin

A280-5H 8-harness satin

1360-5H 5-harness satin

AS-4 3K 193 100 10.0 9.5 6.9

AS-4 3K 280 100 10.5 10.0 10.0

AS-4 6K 370 100 10.5 10.0 13.2

AS-4 3K 370 100 10.0 10.0 13.2

IM-6 12K 360 146 14.0 10.2 12.8

58╇ /╇ Structural Composite Materials

Fig. 2.25

Hot melt resin impregnation process. Source: Ref 9

provide a separator. Some products require compaction at this stage to close the weave. Others may be distorted during the passage through the treater line and require reworking in a tenter, which pulls and straightens the fabric to restore alignment. References

1. R. Bohlmann, M. Renieri, G. Renieri, and R. Miller, Training course notes given to Thales in the Netherlands entitled “Advanced Materials and Design for Integrated Topside Structures,” 15–19 April 2002 2. A.G. Metcalf and K.G. Schmitz, ASTM Proc., Vol 64, 1075, 8, 1974

3. K.K. Chang, Aramid Fibers, ASM Handbook, Vol 21, Composites, ASM International, 2001 4. J.L.J. Van Dingenen, Gel-Spun HighPerformance Polyethylene Fibres, HighPerformance Fibres, CRC Press, 2000 5. S.C. Bennett and D.J. Johnson, Structural Heterogeneity in Carbon Fibers, Proceedings of the 5th Carbon and Graphite Conference, Vol 1, Society for Chemical Industries, 1978, p 377–386 6. W. Watt, Proc. R. Soc., Vol A319 (No. 5), 1970 7. L.T. Drzal, Interfaces and Interphases, ASM Handbook, Vol 21, Composites, ASM International, 2001 8. SP Systems “Guide to Composites”

Chapter 2: Fibers and Reinforcements╇ /╇ 59╇

Fig. 2.26

Resin filming process. Source: Ref 9

9. C. Smith and M. Gray, ICI Fiberite Impregnated Materials and Processes—An Overview, unpublished white paper 10. Hexcel Product Literature, “Prepreg Technology,” 1997 11. Hexcel Product Literature, “Graphite Fibers and Prepregs,” 1988

Selected References •

S. Backer, Textiles: Structures and Processes, The Encyclopedia of Materials Science and Engineering, Pergamon Press, 1986 K.K. Chawla, Fibrous Materials, Cambridge University Press, 1998

Fabrics and Preforms, ASM Handbook, Vol 21, Composites, ASM International, 2001, p 59–68 T.G. Gutowski, Cost, Automation, and Design, Advanced Composites Manufacturing, John Wiley & Sons, Inc., 1987 P. Morgan, Carbon Fibers and Their Composites, Taylor & Francis, 2005 Pitch Fibers Take the Heat (Out), High Perform. Compos., September/December 2001 T.L. Price, G. Dalley, P.C. McCullough, and L. Choquette, “Handbook: Manufacturing Advanced Composite Components for Airframes,” Report DOT/FAA/AR-96/75, Office of Aviation Research, April 1997

60╇ /╇ Structural Composite Materials

Fig. 2.27

Process for producing hot melt tape from resin film. Source: Ref 9

Fig. 2.28

Solution impregnation line for producing prepreg

Chapter 2: Fibers and Reinforcements╇ /╇ 61╇

S. Rebouillat, Aramids, High-Performance Fibres, CRC Press, 2000 D.A. Schultz, Advances in UHM Carbon Fibers, SAMPE J., March/April 1987 A.B. Strong, Fundamentals of Composite Manufacturing: Materials, Methods, and Applications, SME, 1989

A.B. Strong, Practical Aspects of Carbon Fiber Surface Treatment and Sizing J.-S. Tsai, Carbonizing Furnace Effects on Carbon Fiber Properties, SAMPE J., May/ June 1994 P.J. Walsh, Carbon Fibers, ASM Handbook, Vol 21, Composites, ASM International, 2001

“This page left intentionally blank.”

Copyright © 2010, ASM International® All rights reserved.

Structural Composite Materials F.C. Campbell

Chapter 3

Matrix Resin Systems BINDING THE FIBERS together in an orderly array and protecting them from the environment is the role of the matrix. The matrix transfers loads to the fibers and is critical in compression loading in preventing premature failure due to fiber microbuckling. The matrix also provides the composite with toughness, damage tolerance, and impact and abrasion resistance. The properties of the matrix also determine the maximum usage temperature, resistance to moisture and fluids, and thermal and oxidative stability. Polymeric matrices for advanced composites are classified as either thermosets or thermoplastics. Thermosets are low molecular weight, low viscosity monomers (≈2000 centipoise) that are converted during curing into three-dimensional crosslinked structures that are infusible and insoluble. Crosslinking (Fig. 3.1) results from chemical reactions that are driven by heat generated either by the chemical reactions themselves

Fig. 3.1

for example, exothermic heat of reaction, or by externally supplied heat. As curing progresses, the reactions accelerate and the available volume within the molecular arrangement decreases, resulting in less mobility of the molecules and an increase in viscosity. After the resin gels and forms a rubbery solid, it cannot be remelted. Further heating causes additional crosslinking until the resin is fully cured. This progression through cure is shown in Fig. 3.2. Since cure is a thermally driven event requiring chemical reactions, thermosets are characterized as having rather long processing times. In contrast, thermoplastics are not chemically crosslinked with heat and therefore do not require long cure cycles. They are high molecular weight polymers that can be melted, consolidated, and then cooled. Since thermoplastics do not crosslink, they may be subsequently reheated for forming or joining operations; however, due to their inherently high

Comparison of thermoset and thermoplastic polymer structures

64╇ /╇ Structural Composite Materials

Fig. 3.2

S tages of cure for thermoset resin. (a) Polymer and curing agent prior to reaction. (b) Curing initiated with size of molecules increasing. (c) Gelation with full network formed. (d) Full cured and crosslinked. Source: Ref 1

viscosity and high melting points, high temperatures and pressures are normally required for processing.

Table 3.1â•…Relative characteristics of thermoset resin matrices

3.1 Thermosets

Vinyl Esters



Thermoset composite matrices include polyesters, vinyl esters, epoxies, bismaleimides, cyanate esters, polyimides, and phenolics (Table 3.1). Epoxies currently are the dominant resins used for low and moderate temperatures (up to 275 °F or 135 °C). Bismaleimides are used primarily in the temperature range of 275–350 °F (135–175 °C). For very-high-temperature applications (up to 550–600╯°F or 290–315╯°C), polyimides are typically the material of choice. Polyesters and vinyl esters, which can be used at approximately the same temperatures as epoxies, are used extensively for commercial applications but rarely for high-performance composite matrices because of their lower mechanical properties and somewhat poorer environmental resistance. Cyanate esters are a relatively new class of resins that were designed to compete with both epoxies and bismaleimides. These newer resins offer some advantages in lower moisture absorption


Cyanate Esters

Polymides Phenolics

Used extensively in commercial applications. Relatively inexpensive, with processing flexibility. Used for continuous and discontinuous composites. Similar to polyesters, but are tougher and have better moisture resistance. High-performance matrix systems for primary continuous-fiber composites. Can be used at temperatures up to 250–275 °F. Give better high-temperature performance than polyesters and vinyl esters. High-temperature resin matrices for use in the temperature range of 275–350 °F with epoxylike processing. Requires elevated-temperature postcure. High-temperature resin matrices for use in the temperature range of 275–350 °F with epoxylike processing. Requires elevated-temperature postcure. Very-high-temperature resin systems for use at 550–600 °F. Very difficult to process. High-temperature resin systems with good smoke and fire resistance. Used extensively for aircraft interiors. Can be difficult to process.

and have attractive electrical properties but at a significantly higher price. Phenolics are hightemperature systems that offer outstanding smoke and fire resistance and are frequently used for aircraft interior components. Due to their high

Chapter 3: Matrix Resin Systems╇ /╇ 65╇

char yield, they can also be used as ablators and as precursors for carbon-carbon (C-C) components. Polyesters, epoxies, bismaleimides, and cyanate esters are all classified as additioncuring polymers, while polyimides and phenolics are condensation-curing systems. The most distinct difference between the two types of reactions is that condensation reactions give off water or alcohol, while addition-curing reactions do not give off a by-product. This difference is illustrated in Fig. 3.3, which compares a reaction for an addition-curing epoxy with one for a condensation-curing phenolic. The evolution of water and/or alcohols during curing presents a volatile management problem. If these by-products are not removed prior to the resin gelling or setting up, voids and porosity in the

cured matrix will occur. Thus, condensationcuring systems are much more difficult to process than addition-curing systems.

3.2 Polyester Resins Polyesters are used extensively in commercial applications but are limited for use in highperformance composites. Although lower in cost than epoxies, polyesters generally have lower temperature capability, lower mechanical properties and inferior weathering resistance, and they exhibit more shrinkage during cure. Polyesters cure by addition reactions in which unsaturated carbon-carbon double bonds (C=C) are the locations where crosslinking occurs. A typical

Condensation Reaction to Form Crosslink

Fig. 3.3

Comparison of addition and condensation reactions

66╇ /╇ Structural Composite Materials

polyester consists of at least three ingredients: a polyester; a crosslinking agent such as styrene; and an initiator, usually a peroxide such as methyl ethyl ketone peroxide or benzoyl peroxide. Styrene acts as the crosslinking agent and also lowers the viscosity to improve processability. Styrene is not the only curing agent (crosslinker); others include vinyl toluene, chlorostyrene (which imparts flame retardance), methyl methacrylate (improved weatherability), and diallyl phtalate, which has low viscosity and is often used for prepregs. The properties of the resultant polyester are strongly dependent on the crosslinking or curing agent used. One of the main advantages of polyesters is that they can be formulated to cure at either room or elevated temperature, allowing great versatility in their processing. The basic chemical structure of a typical polyester (Fig. 3.4) contains ester groups and unsaturated or double bond reactive groups (C=C). Polyesters are usually viscous liquids consisting of a solution of polyester in a monomer, usually styrene. Styrene in amounts up to 50 percent reduces the viscosity of the solution, thereby im-

Fig. 3.4

Typical polyester and vinyl ester structures. Source: Ref 2

proving processability, and reacts with the polyester chain to form a rigid crosslinked structure. Since polyesters have a limited pot life and will set or gel at room temperature over a long period of time, small quantities of inhibitors, such as hydroquinone, can be added to slow the reaction rate and extend the out-time. A solution of polyester and styrene alone polymerizes too slowly for practical purposes; therefore, small amounts of accelerators or catalysts are always added to speed up the reaction. Catalysts are added to the resin just prior to use to initiate the polymerization reaction. The catalyst does not actually take part in the chemical reaction but simply activates the process. Accelerators, such as cobalt napthenate, diethyl aniline, and dimethyl aniline, can also be added to speed up the reaction. A wide variety of monomers and curing agents are available that yield a broad range of physical and mechanical properties. For example, the bulky benzene ring improves rigidity and thermal stability. Vinyl esters are very similar to polyesters but only have reactive groups at the ends of the molecular chain (Fig. 3.4). Since this results in

Chapter 3: Matrix Resin Systems╇ /╇ 67╇

lower crosslink densities, vinyl esters are normally tougher than the more highly crosslinked polyesters. In addition, since the ester group is susceptible to hydrolysis by water and since vinyl esters have fewer ester groups than polyesters, they are more resistant to degradation from water and moisture.

3.3 Epoxy Resins Epoxies are the most common matrix material for high-performance composites and adhesives. They have an excellent combination of strength, adhesion, low shrinkage, and processing versatility. Commercial epoxy matrices and adhesives can be as simple as one epoxy and one curing agent; however, most contain a major epoxy, one to three minor epoxies, and one or two curing agents. The minor epoxies are added to provide viscosity control, impart higher elevated temperature properties, provide lower moisture absorption or to improve toughness. Two main major epoxies are used in the aerospace industry. The first is diglycidyl ether of Bisphenol A (DGEBA), which is used extensively in filament winding, pultrusion, and some adhesives. The second is tetraglycidyl methylene dianiline (TGMDA), also known as tetraglycidyl-4,40-diaminodiphenylmethane (TGGDM), which is the major epoxy used for a large number of the commercial composite matrix systems. The epoxy group, or oxirane ring, is the site of crosslinking:

The cure of epoxy resins is based on the oxirane ring opening and crosslinking with the curing agent. The epoxide group has unfavorable bond angles, which makes it chemically reactive with a variety of substances that can easily open the ring to form a highly crosslinked structure.

Fig. 3.6

The crosslinking may occur through the epoxy groups or the resulting hydroxy groups. The threemember epoxy group is usually present as either a glycidyl ether or a glycidyl amine or as part of an aliphatic ring. Glycidyl ethers and amines are normally used for composites, while cycloaliphatics are used extensively in electrical applications or as a minor epoxy in composite matrix systems. With amine curing agents, each hydrogen is reactive and can open one epoxide ring to form a covalent bond (Fig. 3.5). When the amine nitrogen contains two hydrogens, each reacts with a different epoxide ring. The reaction between epoxide and amine produces a carbon-nitrogen (C-N) bond. It should be emphasized that the properties of a cured epoxy are strongly dependent on the specific curing agent used. Like polyesters, they may be formulated to cure at either room or elevated temperatures. The most widely used epoxy type is DGEBA (Fig. 3.6). Since it is a difunctional epoxy (such as two epoxy end groups that can react) that can be either a liquid or a solid; and is available as a liquid at several viscosities, it is frequently used for filament winding and pultrusion. If the repeat unit (n) is between 0.1 and 0.2, it is a liquid with a viscosity in the range of 6000 to 16,000 cps.

Diglycidyl ether of Bisphenol A (DGEBA)

Fig. 3.5

Amine crosslinking reaction with epoxies

68╇ /╇ Structural Composite Materials

The effect of the repeat unit on viscosity is shown in Table 3.2. As n approaches two, it becomes a solid. For n values greater than two, it is not useful as a matrix material because the crosslink density becomes too low. Glycidyl amines contain higher functionality (such as three or four reactive epoxy end groups) because the resins are based on aromatic amines. The most important glycidyl amine is TGMDA or TGGDM. This resin, shown in Fig. 3.7, is the base resin used in the majority of composite epoxy matrix systems. Its high functionality (four) provides highly crosslinked structures that exhibit high strength, rigidity, and elevated temperature resistance. In some adhesive systems, where toughness (such as peel strength) is an important property, suppliers will mix difunctional DGEBA with tetrafunctional TGMDA to help provide more flexibility in the cured adhesive. If a lower crosslink density is desired, there are also trifunctional epoxies available, such as the triglycidyl derivative of p-aminophenol (TGAP), shown in Fig. 3.8. Minor epoxies are frequently added to improve processability (viscosity), elevated-temperature performance, or other properties of the cured resin system. Typical minor epoxies include amine-based phenols, novolacs, cycloaliphatics, and others. Table 3.2â•…Effect repeat unit on weight per epoxide (WPE) and viscosity for DGEBA resins Repeat unit n

0 0.07 0.14 2.3 4.8 9.4 11.5 30

Weight per epoxide(a)

Viscosity or melting point

╇ 170–178 ╇ 180–190 ╇ 190–200 ╇ 450–550 ╇ 850–1000 1500–2500 1800–4000 4000–6000

╇╇ 4–6000 cps ╇╇ 7–10000 ops ╇ 10–16000 cps ╇ 65–80 °C ╇ 95–105 °C 115–130 °C 140–155 °C 115–165 °C

(a) Epoxy prepolymers are characterized by the epoxide content or weight per epoxide (WPE), which is the weight of resin containing one mole of epoxide groups.

Fig. 3.7

Tetraglycidyl methylene dianiline (TGMDA)

The composition of a typical epoxy matrix system is: Component

Tetraglycidyl methylenedianiline (TGMDA) Alicyclic diepoxy carboxylate Epoxy cresol novolac 4,4’ Diaminodiphenyl sulfone (DDS) Boron trifluoride amine complex (BF3)

Total percent (wt)

56.4 9.0 8.5 25.0 1.1

In this case, TGMDA is the major epoxy, while the alicyclic diepoxy carboxylate and epoxy cresol novolac are the two minor epoxies. The curing agent is DDS, and BF3 is a catalyst. Diluents are sometimes added to epoxy resin systems to reduce viscosity, improve shelf and pot life, lower the exotherm, and reduce shrinkage. They are normally used in small amounts (three to five percent) because higher concentrations degrade the mechanical and thermal properties of the cured system. Typical diluents include butyl glycidyl ether, cresyl glycidyl ether, phenyl glycidyl ether, and aliphatic alcohol glycidyl ethers. There are a wide variety of curing agents that can be used with epoxies; the most common ones for adhesives and composite matrices include (1) aliphatic amines, (2) aromatic amines, (3) miscellaneous and catalytic curing agents, and (4) anhydrides. A number of these different types of curing agents are shown in Fig. 3.9. Aliphatic amines are very reactive, producing enough exothermic heat given off by the reaction to cure at room or slightly elevated temperature. In fact, if they are mixed in large mass, the exotherm can be large enough to cause a fire. However, since these are room-temperature curing systems, their elevated-temperature properties are lower than those of the elevatedtemperature cured aromatic amine systems. Aliphatic amine systems form the basis for many room-temperature curing adhesive systems. Their elevated-temperature performance can sometimes be improved by initially curing at room

Chapter 3: Matrix Resin Systems╇ /╇ 69╇

temperature followed by a second cure cycle at elevated temperature to increase crosslink density. Aromatic amines require elevated temperatures, usually 250 to 350 °F (120 to 175 °C), to obtain full cure. These systems are widely used for curing matrix resins, filament winding resins, and high temperature adhesives. Aromatic amines produce structures with greater strength, lower shrinkage,

Fig. 3.8

Triglycidyl derivative of p-aminophenol (TGAP)

Fig. 3.9

Curing agents for epoxy resins

and better temperature capability but less toughness than aliphatic amines. Aromatic amine curing agents are usually solids at room temperature that must be melted, although some low melting point eutectic liquids are available. Diaminodiphenyl sulfone (DDS) is by far the most common curing agent used in epoxy composite matrices and in a high percentage of the high-temperature adhesive systems. It should be noted that methylene dianiline, the curing agent used in the high-temperature polyimide PMR-15, and occasionally used with epoxies, is a suspected carcinogen that can either be inhaled or absorbed through the skin. Catalytic curing agents, such as BF3, promote epoxy-to-epoxy or epoxy-to-hydroxyl reactions. They do not serve as crosslinking agents; however, they produce very tightly crosslinked structures and are characterized by long shelf lives.

70╇ /╇ Structural Composite Materials

A typical catalytic curing agent is boron trifluoride mono ethyl amine (BF3-MEA), which is a Lewis acid. It is normally used along with another curing agent (for example, DDS) in small amounts (1 to 5 phr) to reduce the flow and improve the processability of the composite matrix. It is a latent curing agent with a long pot life that requires a temperature of 200 °F (95 °C) or higher to initiate cure. However, once cure is initiated, it proceeds at a very rapid rate. Dicyandiamide (dicy) is also a very important latent curing agent used in both prepregs and adhesives. It is a solid powder that must be thoroughly mixed into the resin to provide uniform cures. Anhydride curing agents require high temperatures and long duration cures to achieve full cure. They are characterized by their long pot lives and low exotherms. They yield good hightemperature properties, chemical resistance, and electrical properties. They can be blended with epoxies to reduce viscosity. Anhydride curing agents generally require the addition of a catalyst to proceed at a rapid rate. One anhydride group reacts with one epoxy group during cure. However, anhydride curing agents are susceptible to moisture pickup, which can inhibit the cure reaction. Nadic methyl anhydride is the most frequently used anhydride curing agent. The curing of an epoxy resin system consists of low molecular weight resins and curing agents reacting under heat at room or elevated temperature to yield high crosslinked structures. The important points to remember are: •

Commercial matrix resins and adhesives are usually a blend of two or more epoxies combined with one or two curing agents. The principal epoxy in the majority of epoxy matrix systems and high-temperature adhesives is TGMDA. Frequently two, or sometimes three, minor epoxies are added to control viscosity or influence the final cured properties, such as modulus or toughness. The major curing agent used in matrix resins and many adhesives is DDS. Catalytic curing agents can be added to reduce flow and accelerate the cure. Epoxy resins for both composite matrices and adhesives are truly engineered systems to yield the best combination of processability and final properties. Higher cure temperatures and long cure times give the highest glass transition temperatures, Tg. When these are combined with high functionality (for example, four reactive end groups), the highest possible crosslink densi-

ties are achieved, which yield strong, stiff, but somewhat brittle structures. The resin is frequently toughened by a number of different means, but this often results in lower usage temperatures. The use of flexibilizing units (either the epoxy or the curing agent) gives higher elongation and impact strength at the expense of Tg, tensile and compression strength, and modulus. However, recent advances in epoxy chemistry and formulation have allowed much tougher resin systems with acceptable elevated-temperature performance. Epoxy matrices and adhesives––in fact, almost all thermoset resins––will absorb moisture from the atmosphere that degrades their elevated-temperature matrix-dependent properties (Fig. 3.10). However, the moisture problem is well understood and can be accounted for in the structural design process. The effect of moisture absorption on composite properties is covered in more detail in Chapter 15, “Environmental Degradation.”

3.4 Bismaleimide Resins Bismaleimides (BMIs) were developed to bridge the temperature gap between epoxies and polyimides with dry Tg in the range of 430 to 600 °F (220 to 315 °C) and use temperatures of 300 to 450 °F (150 to 230 °C). They process very similarly to epoxies by curing through addition reactions at 350 to 375 °F (175 to 190 °C). To obtain their high-temperature properties, they are given freestanding post cures at 450 to 475 °F (230 to 245 °C) to complete the polymerization reactions. Bismaleimide composites can be processed by autoclave curing, filament winding, and resin transfer molding. The tack and drape of most BMIs are quite good due to the liquid component of the reactants. Since BMIs process at the same temperature (such as 350 °F or 175 °C) and pressures (in this case 100 psig) as epoxies, conventional nylon bagging films, bleeder and breather materials, and other expendables can be used. In contrast, the higher-temperature traditional polyimide materials usually require higher temperature cures (600 to 700 °F or 315 to 370 °C) and higher pressures (in this case 200 psig or higher), resulting in more expensive and difficult tooling and bagging materials. Bismaleimide chemistry is quite varied, with many potential paths to producing matrix materials. Both BMIs and polyimides contain the imide group shown in Fig. 3.11. Bismaleimide

Chapter 3: Matrix Resin Systems╇ /╇ 71╇

Fig. 3.10

Fig. 3.11

Effect of moisture absorption on hot-wet matrix mechanical properties

Bismaleimide chemical structure

monomers are synthesized by reacting a primary diamine with maleic anhydride. The most prevalent BMI base monomer for matrices and adhesives is 4,49-bismaleimidedodiphenylmethane. Commercial BMIs are usually one of five forms: (1) BMIs or mixtures of different BMIs, (2) blends of BMIs and BMI-diamines, (3) BMI and olefinic monomers and/or oligomeric blends, (4) BMI and epoxy blends, or (5) BMI and 0,09dicyanobisphenol A mixtures. Although early BMI materials were characterized as being hard to process (low tack and short out-times) and possessing low toughness (brittle), the BMI materials being produced today have much better tack and long out-times. In addition, some BMIs (such as Cytec 5250-4) are almost as tough as some toughened epoxies. Bismaleimides can also readily be processed by using liquid molding processes such as resin transfer molding (RTM).

One potential usage problem with BMIs, and with any polymer containing the imide end group, is a phenomenon known as imide corrosion. This is a form of hydrolysis that results in degradation of the polymer itself. It was originally observed in the aerospace industry with carbon fiber/BMI composites galvanically coupled to aluminum in a sump environment (in this case, a stagnant mixture of salt water and jet fuel). If the aluminum corrodes, the composite, because it is electrically coupled to the aluminum through the carbon fibers, becomes the cathode. Water reduction in the presence of oxygen occurs at the cathode, leading to the formation of hydroxyl ions that attack the imide-carbonyl linkage of the BMI. The mechanism is shown schematically in Fig. 3.12. No corrosion occurs with nonconductive fibers such as glass or aramid, nor does corrosion occur if the metal is galvanically similar to carbon, such as titanium or stainless steel. Studies have shown that increases in temperature and bare exposed carbon edges accelerate the deterioration. Electrically isolating the carbon from the aluminum will prevent the problem, which can be accomplished by curing a layer of fiberglass on the composite faying surface and then sealing the edges with a polysulfide sealant.

3.5 Cyanate Ester Resins Cyanate esters are often used in applications requiring low dielectric loss properties, such as

72╇ /╇ Structural Composite Materials

Fig. 3.12

Galvanic corrosion mechanism for imide linkage. Source: Ref 3

antennas and radomes. They are potential substitutes for both epoxies and BMIs with dry Tgs ranging from 375 to 550╯°F (190 to 290╯°C). However, due to a rather limited market and an inherently expensive monomer, they are expensive materials. The prepreg is also susceptible to moisture pickup that can produce carbon dioxide during cure. Their adhesion, or bondability, is inferior to that of epoxies. The cured laminates exhibit lower moisture absorption than epoxies or BMIs and are inherently flame resistant. Cyanate esters are bisphenol derivatives containing a ring-forming cyanate functional group. This family of thermosetting monomers and their prepolymers are esters of bisphenols and cyanic acid that cyclotrimerize to substituted triazine rings upon heating (Fig. 3.13). During cure, three-dimensional networks of oxygenlinked triazine rings and bisphenol units crosslink by addition reactions. The high aromatic content of the triazine and benzene rings provides a high Tg. The single-atom oxygen linkages function somewhat like ball joints to dissipate localized stresses. Moderate crosslink densities contribute to toughness. In addition, both rubber and thermoplastic toughening mechanisms have successfully been used with cynate esters to further enhance toughness. The attractive electrical properties, low dielectric constant, and dissipation factor are a result of the balanced dipoles and the absence of strong hydrogen bonding. This lack of polarity, along with the symmetry of the triazine rings, makes cyanate esters more resistant to water absorption than most epoxies and BMIs. The low moisture absorption (in the range of 0.6 to 2.5 percent) creates less outgassing, a critical factor in the dimensional stability required for space structures.

3.6 Polyimide Resins Polyimides are high-temperature matrix materials intended for usage temperatures as high as 500 to 600╯°F (260 to 315╯°C). Polyimides can be thermoplastics or thermosets. They are normally condensation-curing systems. The condensation systems give off water or water and alcohol, which causes a severe volatile management problem during cure. If the volatiles are not removed prior to resin gellation, they become entrapped as voids and porosity that lower the matrix-dependent mechanical properties. In addition, polyimides are usually formulated with high-temperature solvents, such as dimethylÂ� formamide (DMF), dimethylactamide (DMAC), N-methylpyrrolidone (NMP), or dimethylsufoxide (DMSO), which must also be removed either prior to or during the cure cycle. Polyimides are much more difficult to process than epoxies or BMIs. They require high processing temperatures (such as 600 to 700 °F or 315 to 370 °C), long cure cycles, and higher pressures (such as 200 psig). Volatiles and voids are always potential problems when processing polyimides. Even the so-called addition-curing systems can exhibit volatile problems as the low molecular weight monomers are usually dissolved in solvents during manufacturing. The best known of the addition-curing polyimide materials is PMR-15. It’s a polymeric monomer reactants with a molecular weight of 15,000. In PMR-15, three types of monomers (Fig. 3.14) are mixed together along with a solvent, usually methyl or ethyl alcohol. However, one of the monomers, 4,49 methylene dianaline (MDA), is a suspected carcinogen that can be absorbed through the skin or inhaled if used as a

Chapter 3: Matrix Resin Systems╇ /╇ 73╇

Fig. 3.13

Cure of cyanate resins

Fig. 3.14

Components of PMR-15

74╇ /╇ Structural Composite Materials

spray. Several non-MDA polyimides have been formulated and sold; however, their elevatedtemperature performance is not as good as that of the original formulation. Even though PMR-15 is classified as an addition-curing reaction, it undergoes condensation reactions early in the cure cycle during the imidization stage that creates a volatile management problem, and the impregnation solvent must be removed before the resin gels; otherwise, the solvent will result in voids and porosity. PMR-15 has a usage temperature of 550 to 600 °F (290 to 315 °C) for 1000 to 10,000 hours, depending on the specific use temperature. The main disadvantages of PMR-15 are the potential for voids, poor tack and drape, inadequate resin flow for fabricating thick and complicated structures, a tendency toward microcracking, and the health and safety concerns regarding MDA. Considerable effort has been expended over the last 25 years on the development of hightemperature polymers that have good thermaloxidative behavior in the 500 to 600 °F (260 to 315 °C) range yet are easy to process. Much of this effort has either been led or funded by the National Aeronautics and Space Administration (NASA), most recently to support the High Speed Civil Transport program conducted during the mid-1990s. The goal was to develop a resin system capable of withstanding a temperature of 350 °F (175 °C) for 60,000 hours. After screening of the available materials, the most promising resin developed was PETI-5, a phenylethynyl terminated imide. A matrix resin, an adhesive, a RTM grade, and a resin film infusion grade of PETI-5 were developed during the program. As with other high-temperature resin systems that use high- boiling-point solvents for manufacturing––NMP in this case––management of volatiles during cure is a major consideration.

Fig. 3.15

Typical phenolic condensation reaction

However, successful demonstration parts were fabricated using all of these product forms.

3.7 Phenolic Resins Phenolics are normally very brittle and exhibit large shrinkage during cure. Their primary use is for aircraft interior structures because of their low flammability and low smoke production. They are also used for high-temperature heat shields, due to their excellent ablative resistance, and as the starting material for C-C composites because of their high char yield during graphitization. Phenolics are made by a condensation reaction with phenol and formaldehyde that gives off water as a by-product. A typical phenolic reaction is shown in Fig. 3.15. Phenolics are usually classified as either resoles or novolacs. If the phenol-formaldehyde reaction is carried out with an excess of formaldehyde and a base catalyst, the result is a low molecular weight liquid resole. If the reaction is carried out with an excess of phenol and an acid catalyst, the result is a solid novolac. Resoles are normally used for phenolic prepregs. Phenolics are one route to the production of high-temperature-resistant C-C composites. The phenolic is charred, or pyrolized, to produce a carbon matrix. Since the charring process produces a porous structure due to the vaporization of the organics in the phenolic, the process has to be repeated several times. Before each subsequent pyrolization, the porous structure is impregnated with either pitch, phenolic resin, or directly with carbon by chemical vapor deposition. This is a slow process that must be done with great care to prevent delaminations and severe matrix cracking. Quite frequently, three-

Chapter 3: Matrix Resin Systems╇ /╇ 75╇

dimensional reinforcements will be used to resist ply delaminations. Carbon-carbon composites can also be made directly using chemical vapor deposition. A carbon preform is impregnated with a methane gas. Since there is a tendency for the deposited carbon to seal off internal voids and porosity, intermittent machining operations are required to remove the surface layer to allow the carbon deposit access to the internal structure. Carbon-carbon composites are dealt with in more detail in Chapter 21, “Ceramic Matrix Composites.”

3.8 Toughened Thermosets The toughness limitation of the thermoset matrix is a direct result of the rigid, highly crosslinked, glassy polymer structures that form during cure. These rigid structures have both advantages and disadvantages. The main advantages are high-temperature capability and the ability of the rigid matrix to stabilize the reinforcing fibers during compression loading. The greatest disadvantage is their susceptibility to delaminations when impacted. Of particular concern are low-velocity impacts causing internal delaminations that cannot be detected during visual walkaround inspections. Since the early 1980s, a large amount of effort has been expended to develop resin systems that are tougher and less susceptible to impact damage. Two candidates emerged in the mid-1980s:

Fig. 3.16

Design details that cause out-of-plane loads

(1) damage-tolerant thermoplastic composites and (2) toughened thermoset composites. Although these systems are radically different in chemical structure, their resultant properties are somewhat similar. Both have improved resistance to low-velocity impact damage, resulting in greater load-carrying capability after being impacted. Also, compared to the stiff thermoset systems, both have a somewhat lower resin modulus and therefore exhibit lower compression strengths. Although it is certainly not true in every case, the tougher systems generally have less heat resistance than the rigid glassy thermosets. Due to the inhomogeneous nature of these systems, where the properties vary in many directions, toughness is more complicated for composites than for homogeneous metallic materials. In composite structures, in-plane loading is controlled primarily by the reinforcing fibers, while out-of-plane loading is dominated by the properties of the resin matrix. Therefore, composite structures are intentionally designed so that the load paths are stiff and are primarily inplane. However, out-of-plane loading can occur. Out-of-plane loads develop during in-plane compressive buckling but, more important, out-ofplane loads are induced by a variety of design features. Five of the more common design details that can cause out-of-plane loading are shown in Fig. 3.16. Even under normal in-plane loading conditions, interlaminar shear and normal stresses develop at these locations. These indirect loads can act either alone or in combination with direct

76╇ /╇ Structural Composite Materials

out-of-plane loads, such as fuel or air pressures. If the out-of-plane loads become large enough, interply delaminations can form and propagate. Fortunately, current design criteria are conservative enough that even if a small delamination is present, either from a manufacturing defect or from service abuse, it will normally not propagate. Crosslinks are the chemical ties between two or more polymer chains that give the cured polymer its strength, rigidity, and thermal resistance. As shown conceptually in Fig. 3.17, the higher the crosslink density (in this case, the number of crosslinks per unit volume) and the shorter the polymer chain lengths between the crosslinks, the more constrained the chain motion, and thus the stiffer and more thermally resistant the molecular structure. Further stiffening can be imparted by the use of stiffener backbones in the main chains. However, this rigidity results in brittleness, low strainto-failure, and poor impact and postimpact properties. Highly crosslinked rigid structures usually exhibit good thermal stability due to their ability to restrict the relative motion of the polymer chains by the chemical bonds holding them together. Since the crosslink bonds are primary covalent bonds, they retain a large portion of their strength as the temperature is increased. When the crosslink density is high and the bond lengths are short, acceptable properties are maintained up to near the glass transition temperature of the resin. Above the glass transition temperature, the rigid solid polymer converts to a softer rubber-like material as the main backbone chains themselves soften. Therefore, highly crosslinked polymers possess moderate to high strength and stiffness and excellent temperature resistance. However, being rigid and glassy structures, they are somewhat brittle and susceptible to impact damage. The molecular structure of a thermosetting polymer determines how it processes and its re-

Fig. 3.17

Effect of crosslink density on rigidity

sultant properties. Examples of properties that are a function of molecular structure include glass transition temperature, moisture absorption, strength, modulus, elongation, and toughness. By altering the molecular structure, it is possible to alter these performance properties. The molecular structure is controlled by its backbone structure (in this case the main polymer chains) and its network structure (in this case the number and types of crosslinks). The chemical structure of the main monomer defines the backbone structure and to some extent the network structure. The network structure is also influenced by the type of curing agent, or hardener, used in the curing reaction. Resin formulators have expended a significant amount of effort to develop new molecular structures that result in polymers with improved toughness. The improved postimpact compression strength of the toughened systems is a direct result of the amount of damage inflicted during the impact event. As shown in Fig. 3.18, the toughened systems (Hexcel’s IM-7/8551-7 in this example) experience much less internal damage when impacted. Therefore, when the specimens are subsequently loaded to failure, the toughened systems have a larger cross-sectional area of undamaged material to help support the compression load. To impart greater toughness to the crosslinked thermoset polymers, a number of different approaches have evolved. Some are used by themselves, while others are often combined to further enhance toughness. Four toughening approaches will be discussed: (1) network alteration, (2) rubber elastomer second phase toughening, (3) thermoplastic elastomer toughening, and (4) interlayering. Network Alteration. Since the brittleness of thermosetting polymers is a direct consequence of their high crosslink densities, one method of

Chapter 3: Matrix Resin Systems╇ /╇ 77╇

Fig. 3.18

Damage comparison during low-velocity impact

toughening a thermoset polymer is to lower the crosslink density. This approach, when taken to the extreme case in which there are no more crosslinks, results in a thermoplastic polymer. Since amorphous thermoplastic polymers are inherently tough, the more the crosslink density is reduced, the tougher the resulting polymer. However, the decrease in crosslink density is also accompanied by a decrease in desirable properties such as the glass transition temperature and the resin modulus. There are three well-known methods for reducing the crosslink density of thermoset polymers (Fig. 3.19). The first is to alter the main monomer backbone chain by increasing the molecular weight between crosslinks with longchain monomers. The resulting decrease in glass transition temperature can be somewhat offset by constructing a long-chain monomer with rigid bulky side groups, such as benzene rings. The decreased mobility of the polymer chains will somewhat compensate for the loss in glass transition temperature due to the lower crosslink density. A second method is to lower the monomer functionality. Most highly crosslinked thermosetting polymers have a functionality of four, which means that there are four reactive end groups that can react and crosslink during cure. If a portion of the polymer mix contains a mono-

mer with a functionality of two, there are fewer available sites for crosslinking during cure and the toughness will therefore be improved due to the lower crosslink density. However, the heat resistance is again affected, as evidenced by the lower glass transition temperatures of difunctional monomers. A third method is to incorporate flexible subgroups into the main chain backbone of either the resin or the curing agent. Although these subgroups are depicted as “springs” in Fig. 3.19, in actuality more flexible aliphatic segments are used in preference to the more rigid and bulky aromatic groups that contain the large benzene ring. As with the other approaches described above, there is a trade-off in glass transition temperature. This can be minimized by using some stiff segments along with the flexible ones. Rubber Elastomer Second Phase Toughening. When a crack begins in a brittle glassy solid, it requires very little energy to propagate. In fiber-reinforced composites, the fibers will prevent in-plane crack growth. However, if the crack is interlaminar or between the plies, the fibers are of no help in preventing crack propagation. One way of reducing crack propagation is to use second-phase elastomers. Discrete rubber particles help to blunt crack growth by promoting greater plastic flow at the crack tip, as shown in Fig. 3.20.

78╇ /╇ Structural Composite Materials

Fig. 3.19

Network alteration toughening mechanisms

The size of the elastomeric domains is a critical factor in determining the microdeformation processes that control toughening. Rubber particles, usually round, having domain diameters of 100 to 1000 Å, initiate shear yielding, while larger domains (10,000 to 20,000 Å) are generally believed to lead to crazing. Very small domains are

used to enhance the shear deformation processes. However, if the crosslink density permits the use of larger domain sizes, bimodal distributions (such as a mixture of large and small domains) of the elastomer will result in both crazing and shear. Since these mechanisms complement each other, the toughening effect can be nearly doubled.

Chapter 3: Matrix Resin Systems╇ /╇ 79╇

Fig. 3.20

Elastomeric toughening

A finely dispersed phase of elastomer-rich domains can be obtained by using either (1) preformed rubber particles or (2) a rubber elastomer system that is initially soluble in the liquid resin but then phase separates (i.e., precipitates) during cure. Preformed particles are advantageous because they can be used as additives or fillers. Also, their domain size is easier to control than when a phase separation process is used. Unfortunately, commercially available preformed rubber particles are suitable only for toughening some of the lower-temperature thermosets. When elastomer domains are to be precipitated during cure, it is necessary to use a rubber with suitable solubility. The rubber must be initially soluble in the resin and then phase separate into the desired small domains during the cure process. An elastomer that is too soluble will remain in solution too far into cure, and significant quantities will be trapped in solution when the resin gels. The trapped elastomer will then act as a plasticizer and subsequently will lower the glass transition temperature and the hot-wet properties. On the other hand, if the elastomer is not soluble enough, a stable solution with the resin will not be possible and a fine dispersion of particles will not be obtained. If just the right amount of compatibility exists, then the elastomer will initially remain in solution, and once the resin begins to cure, small particles will precipitate uniformly

into the resin matrix. Early and complete phase separation is necessary to maintain good hot-wet matrix performance. In epoxy-based composites and adhesives, reactive liquid polymers, such as carboxyl terminated butadiene-acrylonitrile (CTBN) rubber, are often used to provide the desired solution and compatibility characteristics. Functional carboxyl groups, such as CTBN rubbers, are usually blended and prereacted with one of the epoxy monomers to provide the desired solubility. In addition, the carboxyl groups react with some of the epoxy groups to form light crosslinks that increase the cohesive strength of the elastomeric domains. Good bonding between the elastomer domains and the continuous resin phase is important. If the bond is poor, the elastomer can debond from the resin during cooling and form voids. The elastomer itself must have good rubbery characteristics. Specifically, the elastomer-rich domain must have a Tg lower than about –100 °F (–75╯°C) so that when a crack travels rapidly through the material, the domains still act as elastomers. If the Tg is too high (in this case, too close to room temperature), then at high deformation rates it will behave in a glassy manner and the desired toughening effect will not be achieved. Another requirement is that the elastomer must be thermally and thermally-oxidatively stable. If unstable rubbers are used, they are likely

80╇ /╇ Structural Composite Materials

Источник: []
6.63 Mb

Beauty Pilot - software to smooth skin.

It allows you to bring out the beauty in women's portraits. Our sophisticated program is so easy to use that it needs no explanation regarding its user interface. Only a gentle touch is required to erase a skin imperfection on a photo: moles, warts, pimples, heat-spots, etc.
Place the cursor on the spot you want to remove;
Using the mouse wheel or cursor keys set the size of the cursor larger than the size of the spot you are about to remove;
Press and move the cursor circle a little over the imperfection.




Источник: []