Click for next page ( 62


The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 61
Magnetic Recording Media Magnetic recording media, in the form of tapes and disks, are by far the most common machine-readable storage media in use today. It is estimated that the U.S. government alone at present uses over 15 million reels of half-inch computer tape. In 1985, more than 40 million video cassette recorders were manufactured worldwide, as well as some 200 million half-inch video cassettes. Today, over 1 billion computer flexible disks are produced annually. ARCHIVAL CRITERL~ The only other mass memory medium used on such a massive scale is, of course, photographic film, and it is therefore natural to make comparisons between the two. Photographic film is the result of more than 150 years of techni- cal development, and today it is a certifiable archival storage medium even though many early films {e.g., cellulose nitrate-base film) were patently unstable. Mag- netic tape has evolved over the past 50 years into a reliable, stable storage medium despite the problems of its early forebears {e.g., vinyl acetate-base film); however, it has not yet been awarded archival status. This discussion reviews the principal reports on magnetic tape stability pub- lished within the past 10 years. Generally, the conclusion derived is that a good- quality tape, stored in the proper environment ji.e., 65F or 18C, 40 percent RH) and accorded careful mechanical handling, is likely to remain usable for more than 20 years. The period of 10 to 20 years is of particular significance for all machine- readable records because it is also the useful life expectancy of the hardware itself. Today's electronic equipment {e.g., earth satellites, computers, television receiv- ers, and tape recorders) are not expected to remain in service for more than 10 to 20 years. Two important conclusions stem from this fact: first, the recording media may well outlast the hardware; and second, it will become necessary to recopy the tape record every 10 to 20 years on an ever-changing, probably incompatible, new machine with a new format. This operation, termed file conversion, carries with 61

OCR for page 61
62 PRESER VATION OF HIS TOPICAL RECORDS it, of course, the potential of ever-increasing data compaction. Concomitantly, the financial burden of file-converting the entire archival collection perhaps five or six times per century is likely to be out of the question except for relatively small collections that have great historical importance, sustain heavy use, or require rapid access. In machine-readable records, any realistic discussion of the archival proper- ties cannot be separated from questions on the longevity of their associated hard- ware or machines. The long-term stability of the recording medium is necessary, but it is not a sufficient criterion for its use. With human-readable records, on the other hand, the long-term stability of the medium is necessary and sufficient. For these reasons, it must be concluded that magnetic recording media and other machine-readable recording media {e.g., magneto-optic and optical disks) cannot be recommended for long-term Say, over 25 years) archival applications. Similar conclusions have been put forth by the NARS committee {National Archives and Records Service, 1984) and Mallinson {1985a). DEFINITIONS In machine-readable records i.e., magnetic computer, audio, and video tape, magnetic and optical disks, and phonograph records it is understood that the recorded information can be usefully recovered only by converting it to a human-readable form such as paper text, a photograph, or a video terminal display. In analog recordings, this conversion requires appropriate hardware, and in digital recordings it requires hardware, software, and documentation. On the other hand, the information in human-readable records is comprehensible simply by visual inspection of the record or a magnified image of the record. Only simple optical hardware, such as microscopes and projectors whose design principles need never change, are required to read the record completely. THE ELECTRONIC INFORMATION AGE Modern civilizations are now entering the so-called Information Age, wherein the vast majority of their information and records are stored, manipu- lated, and disseminated by electronic means such as computer networks, earth satellite relays, and television broadcasting. It seems that the archival community tends to forget that the principal motive for these technologies is their speed of access and that this speed is only achieved at an extremely high cost. The machines themselves {e.g., computers, satellites, and television receivers) are rarely expected to have a useful life in excess of 10 years. The machine-readable records are operated at ever-increasing information storage densities, not only to store more information but also to decrease access times. This is a trend that is surely inimical to long-term archival preservation. Since 1956 no less than eight differing, incompatible videotape formats of increasing storage density have emerged. In fact, coincidentally, since 1952 eight differing computer tape formats have been used. The sixteen tape formats are listed in Table 6-1. Each format typically mandates a different machine with its unique set of demodulators, decoders, and reformatters. This proliferation of incompatible systems is the root cause of the archivist's dilemma in adopting

OCR for page 61
MAGNETIC RECORDING MEDIA TABLE 6-1 Video and Computer Tape Formats Product Current Status Video tape formats since 1956 2-inch quadrupled 2-inch quadrupled, double density 1-inch helical, type A 1-inch helical, type B 1-inch helical, type C 3/~-inch helical, U-Matic /-inch helical, Beta-max /-inch helical, VHS /-inch computer tape formats since 1952 7-track NRZI, 100 BPI 7-track NRZI, 200 BPI 7-track NRZI, 556 BPI 7-track NRZI, 800 BPI 9-track NRZI, 800 BPI 9-track PE, 1,600 BPI 9-track GCR, 6,250 BPI 18-track NRZI, 19,000 BPI Obsolete Obsolete Obsolete Obsolete Obsolete Obsolete Obsolete KEY: NRZI = Non-Return to Zero Inhibit; PE = Phase Encoding; GCR = Group Code Recording. 63 machine-readable records. The speed of access and the electronic data processing abilities are indeed attractive, but it must be recognized that the records and their associated hardware will become obsolete within a couple of decades. Since the information and communication industries are most definitely not driven by long-term archival considerations, it seems futile to expect technology to resolve this problem. Advances in technology continue to cause the machine- readable problem, and obviously these advances will not solve the problem. ARCHIVAL PROPERTIES OF MAGNETIC RECORDING MEDIA Magnetic recording media are made up of three components: the substrate, the magnetic particles or grains, and the binder system. In rigid computer disks shard disks d, the substrate is an aluminum alloy. The magnetic particles are gamma-Fe2O3, and the binder system is usually one of the epoxy family. Because of the relatively high cost of storing data on rigid disks ( 10-3 cents per bit versus 10-6 cents per bit on taped, rigid disks are rarely considered for archival applications and, therefore, will not be discussed further. An additional factor against its archival use is the fact that the majority of today's large hard disk files "Winchester drivesJ cannot be physically separated from the head-disk assem- bly tHDAJ, a sealed unit. Magnetic tapes and flexible disks almost universally have a polyethylene terephthalate {PETJ film substrate; common trade names are Mylar {DuPontJ, Celanar {Celanese), and Estar {KodakJ. In a recent publication from the National Bureau of Standards it was concluded that, given storage at 20 to 25C {68 to 77FJ and 50 percent RH, PET films are expected to have a lifetime of 1,000 years (Brown et al., 1984J. Consequently, PET films will not be discussed further.

OCR for page 61
64 PRESERVATION OF HISTORICAL RECORDS The magnetic particles used in most half-inch-wide computer tape are gamma-Fe2O3. In the most recent half-inch computer tape format ~ 18-track NRZI, 19,000 BPI), CrO2 is the magnetic material used. Most of today's video tapes use cobalt-surface-modified gamma-Fe2O3. Flexible computer disks use, in the main, gamma-Fe2O3, with increasing adoption of cobalt-surface-modified gamma- Fe2O3. It is believed that all these magnetic materials are stable chemical entities under normal storage conditions. All are produced by high-temperature {above 200C) processes, which implies great stability around room temperature. The magnetic stability of these particles is well understood. Their coercive forces are all above 300 Oe {more than 200 times the earth's magnetic field), and they are, accordingly, unaffected by the stray fields {about 10 Oe) associated with most electronic equipment. Their Curie temperatures {the temperature above which they become nonmagnetic) are above 400C except in the case of CrO2, which is only 120C. These Curie temperatures are so far above the normal archival storage temperatures that no difficulty is anticipated. Other magnetic effects of concern include the print-through phenomenon, in which the magnetic fields from one layer of written tape can slightly magnetize the particles in the adjacent layers on the reel. The effect is known to be extremely small at room temperature but increases with temperature. However, at 65C in a 4-hour test the print-through signal level typically remains a factor of 500 below the normal signal levels Bertram and Eshel, 1979~. The binder systems in universal use in tapes today are of the polyester-ure- thane type. Because all tapes and flexible disks are intended to be operated with the writing and coding heads in as close physical contact as possible, the binder system has been chosen because of its extreme resistance to mechanical abrasion and its chemical stability. Most manufacturers of tape use slightly differing formu- lations, and no standards have yet been instituted. The Achilles' tree! of magnetic recording is the extremely close head-to- medium spacings required. Accordingly, most of the published reports deal directly or indirectly with the archival stability of the polyester-urethane binder systems. The particular area of concern is the hydrolysis of the binders. The basic reaction is (Brown et al., 1984; Cuddihy, 1980; Bertram and Cuddihy, 1982) Ester + Water ~ Polycarboxylic Acid + Alcohol The problem with hydrolosis of the binder system is that its mechanical properties degrade if it is allowed to progress too far. It is thought that the adhesion of the binder to the substrate is particularly vulnerable to hydyrolysis brown et al., 1984~. Earlier work showed that hydrolysis is a reversible reaction and sug- gested that not only can over-hydrolyzed tapes be recovered but that the system may equilibrate at a point where the hydrolysis reaction rate is zero {Cuddihy, 1980; Bertram and Cuddihy, 1982~. To attain equilibrium at a satisfactorily low level of hydrolysis, tapes should be stored at 20C {68F) and 40 percent RH {Bertram and Eshel, 1979; Bertram and Cuddihy, 1982~. Other satisfactory envi- ronments for limiting hydrolysis are shown in Figure 6-1. A tape wound on a hub relies entirely on the maintenance of the layer-to-layer pressures and friction to transmit torque to the outer layers. The layer-to-layer pressure has been found to vary considerably when the tape reel is exposed to temperatures and humidities that differ from those that existed when the reel was

OCR for page 61
MAGNETIC RECORDING MEDIA 100 90 80 70 60 I 50 > OCR for page 61
66 PRESERVATION OF HISTORICAL RECORDS private communication, 1985~. After longer periods {say, 5 years) of storage, it is advisable to rewind the tape gently whenever it is withdrawn from the archive. The published literature contains a host of other sound suggestions for the archival care of magnetic recording media. The review by Geller { 1983) is particu- larly detailed and is highly recommended as a guide for action. Of the many excellent practices suggested, it is perhaps well to note those that concern the response of magnetic tape to changes in humidity. The coefficient of linear expan- sion of magnetic tape is almost the same for each percent change in relative humidity as it is for each degree Celsius change in temperature. Whereas the temperature may equilibrate in several hours, similar equilibria/ion of humidity in a tightly wound tape reel may take several days. It follows, then, that if an archival tape is to be subjected to a change in humidity (e.g., when brought to the computer room environment), it ideally should be allowed to equilibrate for sev- eral days even before being rewound or retensioned. Given proper care, a safe conclusion can be drawn that a good-quality tape is an archivally reliable storage medium for periods of 10 to 20 years as a minimum, with much longer periods a distinct but as-yet unproved possibility. The essential point to be remembered is that today's recording media are most likely to outlive the period of utility of the other components of magnetic storage systems. TRENDS In the future the inexorable trend to higher storage densities will lead to magnetic tapes and disks that use very thin metallic layers iMallinson, 1985a, 1985b). Projections are that not only will such tapes approach within a factor of two the storage densities of optical and magneto-optical disks, but also the physi- cal properties of the metallic storage layers will be very similar. Thus, in magnetic disks, a 200-A-thick layer of Co-Ni may be used, compared with magneto-optical disks that have a 150-A-thick layer of Co-Fe-Tb and optical disks with a 150-A- thick layer of a Te alloy. The archival properties of such metallic thin-film media, be they magnetic or optical, are not at present known; current estimates are for lifetimes of 10 to 30 years. Again, the salient point is that the recording medium may well outlast the hardware. ARCHIVAL PROPERTIES OF SOFTWARE AND DOCUMENTATION In a certain class of machine-readable records, namely those employing com- puter or digital technologies, a further archival problem arises. The mere recovery of the digital data is not possible without some software. The proper operating system {called software to distinguish it from hardware) must be available at the time that data are to be recovered. Sadly for the archivist, software today is chang- ing more rapidly than hardware; for example, Western Electric's UNIX operating system has been offered in about 30 versions over the past decade. Offsetting this serious problem are, of course, some potentially attractive reasons for using digital recording techniques; compatibility with the computer environment and the abil- ity to perform perfect error detection and correction are prime examples. Given the proper operating system for reading out the digital data, a further requirement arises. Appropriate documentation must be at hand that will provide

OCR for page 61
MAGNETIC RECORDING MEDIA 67 Contro~e3-environment tape storage area. Archival care of magnetic recor~ingme~a requires attention to come-ons of temperature and humi~tyas well as carefu~peno~c rewinding.

OCR for page 61
68 PRESER VATION OF HIS TOPICAL RECORDS the necessary information on the digital codes used, the organization or format of the record, and several other minor but critical details. Operating systems are usually resident on computer tapes or floppy disks, thus compounding an already difficult archival problem. The documentation may be in machine-readable or human-readable form but, given the human species' well-known tendency to procrastinate, the needed data may well be incomplete or missing {National Acad- emy of Sciences, 1982~. ARCHIVAL PROPERTIES OF HARDWARE The fact that most electronic hardware is expected to function for no more than 10 to 20 years raises very serious problems for long-term {more than 20 years) archival preservation. Even if the operating systems and documentation problems somehow are dealt with, what is the archivist to do when the machine manufac- turer declares the hardware obsolete or simply goes out of business? Will there be an IBM or a Sony in the year 2200? If they still exist, will they maintain a 1980- 1990 vintage machine? Moreover, it must be realized that no archival organization can hope realistically to maintain such hardware itself. Integrated circuits, thin film heads, and laser diodes cannot be repaired today, nor can they be readily fabricated, except in multimillion-dollar factories. The inescapable conclusion is that, if a long-term archive preserves records in machine-readable form, it will be committed eternally to file conversion {i.e., rerecording the old obsolete versions into the new current format) approximately every 10 to 20 years. Not only would such an operation be enormously expensive, but also, in an archive where by definition no records can be disposed of it is a task that grows exponentially with time. Precisely such file conversions take place all the time, of course, in today's computer facilities, but the critical differ- ence is that records in such facilities are continually being retired perhaps to be sent to an archive! CONCLUSIONS The committee's conclusions in the area of magnetic media are as follows: 1. Magnetic recording media today are of sufficient stability that only short- term {10 to 20 years) storage is practical. 2. Operation of short-term magnetic tape archives in accordance with the rec- ommended storage practice detailed by Geller (1983) is possible. 3. Magnetic recording media and other machine-readable recording media can- not be recommended for long-term Over 20 years) storage because of the difficul- ties in maintaining software, hardware, and documentation; provision for repeated file conversion can overcome this limitation. REFERENCES Bertram, H. N., and E. F. Cuddihy. 1982. Kinetics of the humid aging of magnetic recording tape. IEEE Trans. Magn., 18(5, September):993-999. Bertram, H. N., and A. Eshel. 1979. Recording Media Archival Attributes (Magnetic). U.S. Air Force Systems Command, RADC F 30602:78:C-0181.

OCR for page 61
MAGNETIC RECORDING MEDIA 69 Brown, D. W., R. F. Lowry, and L. E. Smith. 1984. Predictions of Long-Term Stability of Polyester- Based Recording Media. National Bureau of Standards, NBSIR 84-2988, December. See also Kinetics of hydrolytic aging of polyester urethane elastomers, Macromolecules, 13:248-252 {1980~; Hydrolytic degradation of polyester polyurethanes containing carbodiimides, Macro- molecules, 15:453-485 jl982J; Equilibrium acid concentrations in hydrolyzed polyesters and polyester-polyurethane elastomers, J. Appl. Polym. Sci., 28:3779-3792 jl983~; end Hydrolysis of crosslinked polyester polyurethanes, Div. Polym., Mater. Sci. Eng., 51:155-161 (1984~. Cuddiby, E. F. 1980. Aging of magnetic recording tape. IEEE Trans. Magn., 16(4, July]:558-568. Geller, S. B. 1983. Care and Handling of Computer Magnetic Storage Media. National Bureau of Standards, NBS SP 500-101, June. Mallinson, J. C. 1985a. The next decade in magnetic recording. IEEE Trans. Magn., 21 (3, May]: 1217-1220. Mallinson, J. C. 1985b. Archiving human and machine readable records for the millenia. Society of Photographic Scientists and Engineers Second International Symposium: The Stability and Preservation of Photographic Images. Ottawa, Canada. August 1985. National Academy of Sciences. 1982. Data Management and Computation, Vol. 1: Issues and Rec- ommendations. Washington, D.C.: National Academy Press. National Archives and Records Service. 1984. Advisory Committee on Preservation White Paper: Strategic Technology Considerations Relative to the Preservation and Storage of Human and Machine Readable Records. July. Unpublished.

OCR for page 61
Optical disk a recent development in data storage technology. The magniped view shows digitized data encoded on the disk.