Mechanics and Reliability of Flexible Magnetic Media Second Edition Springer Science+Business Media, LLC Bharat Bhushan Mechanics and Reliability of Flexible Magnetic Media Second Edition With 326 Illustrations i Springer Bharat Bhushan, Ph.D., D.Sc., P.E. Ohio Eminent Scholar and the Howard D. Winbigler Professor Director, Computer Microtribology and Contamination Laboratory Department of Mechanical Engineering Ohio State University Columbus, OH 43210 USA Cover illustration: Tape path in an IBM 3480/3490 data-processing tape drive. Library of Congress Cataloging-in-Publication Data Bhushan, Bharat, 1949- Mechanics and reliability of flexible magnetic media / Bharat Bhushan.-2nd ed. p. cm. Includes bibliographical references and index. ISBN 978-1-4612-7069-0 ISBN 978-1-4612-1266-9 (eBook) DOl 10.1007/978-1-4612-1266-9 1. Magnetic tapes-Reliability. 2. Magnetic disks-Reliability. 3. Magnetic recorders and recording. I. Title. TK5984.B49 2000 621.34-dc21 99-045614 Printed on acid-free paper. © 2000, 1992 Springer Science+Business Media New York Originally published by Springer-VerJag Berlin Heidelberg New York in 2000 Softcover reprint of the hardcover 2nd edition 2000 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher Springer Science+Business Media, LLC, except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use of general descriptive names, trade names, trademarks, etc., in this publication, even if the former are not especially identified, is not to be taken as a sign that such names, as understood by the Trade Marks and Merchandise Marks Act, may accordingly be used freely by anyone. Production managed by Steven Pisano; manufacturing supervised by Joe Quatela. Typeset by Asco Typesetters, Hong Kong. 9 8 7 6 543 2 I To my wife (Sudha), my son (Ankur), and my daughter (Noopur) Preface According to some estimates, 95% of information today is stored on paper, 3% on microfiche, and only 2% on magnetic/optical and semiconductor storage devices. Semiconductor storage is almost exclusively used for dy namic random access memory (D-RAM) in computers, and constitutes a very small fraction of the total storage capacity. Magnetic storage devices include hard disk, flexible disk, and tape drives. Estimates for worldwide storage is 12,000 petabytes (12 million terabytes). It is estimated that mag netic tapes store about 95% of the information, and the balance is stored equally by magnetic hard disk and optical disk drives (250 petabytes each). For comparisons, the human brain has 1014 neurons and holds approxi mately 200 megabytes of information. For a world population of 6 billion people, the total human memory is therefore 1200 petabytes, which is about 10% of the electronically recorded information. Magnetic and optical storage industry for consumer and data recording applications is at present an industry grossing more than $80 billion per year. It is expected to grow at cumulative rate of about 10% per year. Revenue is as follows: for magnetic disks and drives, about $35 billion; for flexible disks and drives, about $4 billion ($1.5b/$2.5b); for data tape and tape drives, about $8 billion ($2b/$6b); for consumer video- and audiotape and tape drives, about $25 billion ($8b/$17b); for CD/DVD read-only disk and disk drives, about $7 billion ($lb/$6b); and for other optical products, less than a $1 billion. The magnetic-recording process is accomplished by relative motion be tween a magnetic head and a magnetic medium. Types of magnetic media for magnetic recording are flexible media (tapes and flexible disks) and rigid disks. Physical contact between head and medium occurs during starts and stops, and hydrodynamic air-film develops at operating speeds. Flying heights (mean separation between head and media) are of the order of 0.1 pm comparable to surface roughnesses of the mating members. The need for higher and higher recording densities requires that surfaces be as smooth as possible and flying heights be as low as possible. Smoother surfaces lead to increased wear. In the case of magnetic tapes, in order to have high bit vii V111 Preface density for a given size of spool, we like to use as thin a tape substrate as possible. Thinner tapes are prone to local or bulk viscoelastic defonnation during storage. This may lead to variations in head-tape separations result ing in problems in data reliability. Anisotropic mechanical properties and viscoelastic defonnations of the flexible disks may result in data-reliability problems. All magnetic media have to be sufficiently lubricated to minimize head and magnetic-medium wear. The lubrication is carried out either topi cally or in bulk. A fundamental understanding of the tribology, mechanics, and reliability of the head-magnetic-medium interface is, therefore, very crucial for the future of the fast-growing magnetic-recording industry. Volumetric density of a drive is a function of areal density (product of the linear and track densities) and the total tape thickness. Today, the best linear and track densities in data processing drives are 86 kbpi and 1000 TPI (86 Mb/in2) in linear drives (DLT 7000) and 122 kbpi and 2800 TPI (340 Mb/in2) in helical drives (DDS3). The thicknesses of double-layer metal par ticle (MP) and metal-evaporated (ME) tapes are about 8 f.1.m (4.5-5.5 f.1.m sub strate, 2.6 pm non-magnetic underlayer, 0.4 f.1.m magnetic layer, and 0.5 f.1.m backcoat) to as high as 18 f.1.m and about 7 pm (6.3 f.1.m substrate, 10-25 nm non-magnetic underlayer, 65-170 nm thick dual magnetic layer, 8-10 nm DLC, 2 nm lubricant, 0.3-0.5 f.1.m backcoat and 8-10 nm lubricant), respectively. Therefore, for MP and ME tapes, the volumetric densities are 88 and 198 Gbytes/in.3 in helical drives, respectively. The physical spacing between the head and tape is on the order of 40-50 nm. Thin-film inductive type heads are commonly used. MP particles used today are about 100 nm by 20 nm with about 3 nm thick oxide coating for corrosion protection ('" half of volume in oxide). ME coating is dual layer with each layer about 65-85 nm thick. Dual layers are used for unifonn orientation of the columns. The target in the next five years is to achieve an areal density of about 2 Gb/in.2 (200 kbpi and 10,000 TPI) and a volumetric density of more than 2 terabytes/in.3 Based on the Wallace equation, physical spacing needs to be reduced for increased linear density in order for shorter flux reversals to in tersect the head. The linear density would be achieved by scaling the physi cal spacing with bit size to about 20 nm, which would require a smoother tape with a decrease in nns roughness from the current 6-8 nm nns to about 2 nm nns and with very low defect density. The particle length in MP tapes and ME film thickness would have to be reduced because bit length should be less than half of particle length or of one of the layers of ME tape. For higher TPI, the tape and its substrate have to be very dimensionally stable, and servo may be required. Increase in noise would be offset to maintain SNR by MR, GMR, or spin valve-type heads. For high volumetric density, the tape thickness would be required to be about 4-8 f.1.m and 4.5 f.1.m for MP and ME tapes, respectively. For high data rates, from a typical value of 3 Mb/s to a value of 20-100 Mb/s, higher linear velocities would be re quired, as compared to typical 4 and 14 m/s for linear and helical drives, Preface lX respectively. For high reliability, the air bearing surface (ABS) wear and pole-tip recession (PTR) growth should be on the order of 50 and 10 nm, respectively, over the head life (1000 h with ~ 20,000 km of tape over head) and the tape wear should be on the order of 10 nm over 10,000 file passes! MP and ME tapes are leading candidates. Higher tape smoothness, thinner tape substrates and magnetic layers, larger head-to-tape relative velocities, and increased sensitivity to head-to-tape interface instability place new de mands on the durability of the head-to-tape interface in ultra-high density re cording applications. The ultra-thin tapes should have improved mechanical properties as well as good slit edge quality and low damage with edge guid ance. Debris generation and particulate contamination must be minimized. To meet the new challenges, ultra-thin substrates with high smoothness, low defect density, good dimensional stability, and improved mechanical properties are being developed. The tribology and mechanics of magnetic storage devices are covered in a separate book published by Springer-Verlag in 1996. This book is a systematic compilation of the current knowledge of mechanics and the reliability of flexible magnetic media. The organization of the book is straightforward. Chapter 1 presents brief descriptions of the physics of magnetic recording, magnetic storage systems, and the manufacturing pro cesses of magnetic media. Chapter 2 presents a brief description of the manufacturing processes of poly(ethylene terephthalate) films commonly used as flexible media substrate, and the physical and chemical properties of the PET films and coated magnetic media. Chapter 3 then presents the vis coelastic properties of the PET films and coated magnetic media. Chapter 4 is a new chapter that discusses advanced ultra-thin substrates. Chapter 5 presents analytical predictions of stresses in wound magnetic tapes and flex ible disks. In Chapter 6, long-term reliability problems of magnetic tapes encountered during storage and use are discussed. Their descriptions, mech anisms, and methods of preventing them are presented. In Chapter 7, the long-term reliability problems of flexible disks encountered during storage and use are discussed. Typographical errors have been corrected throughout the book. I hope that readers find the second edition useful. I have tried, wherever possible, to discuss theories and types of experi mental measurements that can usefully be made in corroborating theories and/or in developing our understanding. Emphasis has been on the funda mental understanding of the subject matter before proceeding to a diversity of practical applications. I have presented ample experimental data, and relevant properties of materials and surfaces, to make this book useful to engineers working in the industry. The book is intended for three types of reader: the graduate student of magnetic recording, the research worker who is active or intends to become active in this field, and the practicing engineer who has encountered a reliability problem and hopes to solve it as ex peditiously as possible. Viscoelasticity theories presented in this book are very general and are applicable in other than magnetic-storage systems. x Preface I wish to thank my wife, Sudha, my son, Ankur, and my daughter, Noopur, who have been very forbearing during the years when I spent long days and nights in conducting the research and preparing this book. I would also like to thank many colleagues at the IBM Corporation who con tributed to some of the research reported in this book. I have immensely benefited from my association with F.W. Hahn, D. Connolly, I.e. Heinrich, D.l. Plazek (University of Pittsburgh), B.A Bartkus, D.l. Winarski, B.s. Sharma, R.L. Bradshaw, C.l. Heffelfinger (DuPont), B.S. Berry, T.L. Smith, 1.1. Gniewek, S.M. Vogel, W.B. Phillips, M.A Marchese, and AA Gaudet. Columbus, Ohio BHARAT BHUSHAN Contents Preface................................................................... VB 1. Introduction.......................................................... 1 1.1. Physics of Magnetic Recording .................................... . 1.1.1. Basic Principle ............................................. . 1.1.1.1. Magnetism.......................................... 1 1.1.1.2. Electromagnetic Induction. . . . . . . . . . . . . . . . . . . . . . . . . . . 9 1.1.1.3. Magnetic Recording................................. 9 1.1.2. Vertical Recording. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11 1.1.3. Signal-Processing Methods. . . . . . . .. . . .. . .. . .. . .. . . .. .. . . . . . . . 12 1.1.4. Design Considerations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15 1.1.4.1. Recording Density.................................. 15 1.1.4.2. Reproduced Signal Amplitude....................... 16 1.1.4.3. Signal-to-Noise Ratio. . . . .. .. . . .. . ... . .. . . . . . . .. . .. . 19 1.2. Magnetic Storage Systems. . . . . . . . . . . . . . . .. .. . . .. . . .. . . . .. . . . . .. . .. . 20 1.2.1. History of Magnetic Recording . . . .. . . . . .. . .. . . .. . . . . .. .. . .. . 20 1.2.1.1. Storage Hierarchy................................... 21 1.2.2. Examples of Modern Storage Systems Using Flexible Media .. 22 1.2.2.1. Tape Drives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24 1.2.2.2. Flexible-Disk Drives. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42 1.2.3. Head Materials. . .. . . . . . . . . . . . .. . .. .. . . .. . . .. . . . .. . . . . .. . . . . 51 1.2.3.1. Permalloys.......................................... 52 1.2.3.2. Mu-Metal and Hy-Mu 800B . . . . . . . . . . . . . . . . . . . . . . . . . 52 1.2.3.3. Sendust Alloys. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52 1.2.3.4. Alfenol Alloys.. .. .. .. .. .. .. .. .. .. .. . .. .. .. .. .. .. .. . 54 1.2.3.5. Amorphous Magnetic Alloys ........................ 54 1.2.3.6. Ferrites............................................. 55 1.2.3.7. Some Examples of Head Constructions. . . . . . . . . . . . . . . 57 1.2.4. Flexible Media Materials.. .. .. .. .. .. .. .. . .. .. .. .. .. .. .. .. . .. 59 1.2.4.1. Base Film........................................... 59 1.2.4.2. Magnetic Medium. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 1.2.4.3. Particulate Magnetic Coatings. . .. . . . . . . . ... . . . .. . .. . 64 1.2.4.4. Magnetic Thin Films. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 1.2.5. Functional Requirements . . . . .. .. . .. . .. . . . . . . . .. . . .. .. . .. . . . . 68 xi