Scott Stevens
COMMSCOPE INC.
The optical fiber industry has recently transitioned from 50-thousand-pound-per-square-inch fiber to 100-kpsi fiber strength as the standard in its cables. For local area networks, where fiber lengths are relatively short, some installers have been confused about the need for this transition.
Improvements in the glass-making process provide the impetus for the change. Because most major fiber manufacturers now achieve sufficient yields when proof-testing to 100 kpsi, a price premium is no longer charged.
The migration to 100-kpsi fiber is beneficial to the end user because the cable manufacturer can reduce the number of strength members used in the cable and lower its price. It is not, however, a response to reliability problems with 50-kpsi fiber.
Understanding fiber flaws
Flaws in fiber are microscopic imperfections in the glass. The size of the largest flaw will determine how much strength the fiber has. In its perfect state, fiber has a stronger tensile rating than steel. However, the introduction of large flaws will severely weaken the fiber. Much as a chain, which is only as strong as its weakest link, a length of fiber is only as strong as its largest flaw permits.
Fiber flaws can grow if enough stress--which must exceed a minimum threshold level--is placed on the fiber. Once the threshold is exceeded, the flaw grows at a rate that depends directly on the amount of stress and the size of the flaw. The larger the flaw, the faster it grows.
Whenever the stress applied to the fiber is equal to, or greater than, the strength of the largest flaw, the fiber will break at that point. If the stress applied is less than the weakest flaw but constant over time, all of the flaws above the threshold will grow in size. Eventually, the largest flaw--because it grows fastest--will weaken to a level below that of the applied stress, and the fiber will break. (Note, however, that the time required for such a failure could be hundreds of years.)
Understanding proof-testing
Proof-testing optical fibers is a method used to screen out the largest flaws. After manufacturing, fiber is passed through a test station that applies tension to the fiber at a preset value, usually 50 or 100 kpsi. Fiber with a large flaw, whose strength is below the preset tension value, will break at the point of the flaw. Fiber passing the test without breaking is given the rating of the preset tension.
The fiber passing through the proof-test station comes from a continuous reel, typically 2 to 20 kilometers long. If a break occurs during testing, the fiber is not discarded, nor is its strength downgraded. The fiber that has passed through the test station has already been screened, so it is rated at the preset tension value. The remaining fiber is restrung through the test station and the operation resumes. The only difference is that the glass manufacturer now has two short lengths of fiber rather than one long length. Both fibers are rated at the same proof level.
From a manufacturing process standpoint, there is no difference between 50- and 100-kpsi fiber. The glass screened to 50 kpsi is identical to that screened to 100 kpsi. Manufacturers do not start with different materials or change their procedures to create stronger or weaker glass; the production process is the same.
A fiber screened to 50 kpsi has a theoretical minimum strength of 50 kpsi. Its actual strength is unknown, because it depends on the largest flaw, which cannot be identified in the screening process. The fiber might have any strength greater than 50 kpsi--for example, 51, 200 or 719 kpsi.
The potential disadvantage of 50- versus 100-kpsi fiber is that it might contain a flaw whose strength is in the 50- to 100-kpsi range. However, the probability of this occurring is small. Fiber manufacturer Corning Inc. (Corning, NY), for example, has indicated that the probability of such a flaw is less than 0.01%. The same study shows that 99.9% of all 1-km lengths of fiber, tested to 50 or 100 kpsi, have a tensile strength greater than 300 kpsi.
Field-related problems
Well-designed fiber-optic cable is inherently robust. If installed properly, a loose-tube outdoor cable has an estimated life expectancy of more than 40 years. For indoor tight-buffered cable not exposed to temperature cycling, high moisture or ultraviolet radiation, the fiber should last as long as the building remains standing.
Installed fiber rated at 50 kpsi has not had a history of breaking because of weakness. In fact, 50-kpsi fibers have been used for nearly 20 years without problems attributed to their screening strength level.
Excluding catastrophic events, installed fiber-optic cable breaks are invariably of two types: The cable was not installed properly, or a problem occurs at the point where the jacket is stripped back.
Installation errors usually involve exceeding the specifications for pull tension or bend radius. Because manufacturers` cable designs limit fiber strain to a percentage of the proof-test level--for example, 80%--exceeding the pulling tension would have similar consequences for both 50- and 100-kpsi fiber. For instance, to reach a fiber strain equivalent to 50 kpsi in a cable designed with 50-kpsi fiber, a 600-pound-rated cable would have to be pulled at 720 pounds. If the cable were designed with 100-kpsi fiber, the same 720-pound pulling tension would strain the fiber to 100 kpsi. (Note that the standard practice of using breakaway lines during installation would prevent such abuse.)
If the bend radius is exceeded, the static fatigue placed on the fiber could lead to the growth of flaws at the point of stress. Unless the stress is relieved, the fiber, whether 50- or 100-kpsi, may weaken and eventually break. Fortunately, most fiber installations require periodic testing that can identify such problems before the cable fails.
To understand the added risk of using 50-kpsi fiber, consider the following worst-case scenario for a large outdoor armored cable. The probability that a flaw whose strength is between 50 and 100 kpsi will be contained within 0.5 meter--the arc length through a 90-degree bend of the fiber--is approximately 0.000005%.
More common than improper cable installation, however, are problems related to fiber handling. Installers are required to strip cable jacket to access fibers so they can splice or connectorize the cable. During these procedures, fiber faces an increased risk of damage. It may, for instance, be inadvertently nicked or scratched from handling or from tools. This can create new flaws in the fiber. Fiber can also be damaged during routing and storing operations--for example, you can catch fibers between the lids of a splice tray.
These handling problems are not related to the proof-test level, however. One-hundred-kpsi fiber is no different than 50-kpsi in its susceptibility to nicks, scratches or mishandling.
Manufacturing differences
The difference between 1 km of 50-kpsi and 1 km of 100-kpsi fiber after delivery from the glass manufacturer is small. However, users specifying 100-kpsi fiber might be lulled into a false sense of security by claims about this fiber`s superiority. What such users should remember is that specifying a manufacturer that can properly design and manufacture cable to avoid stress-related problems is more important than specifying 50- or 100-kpsi fiber.
Proper cable design says that the more the cable decouples external forces from the fiber, the more effectively the cable will perform, both optically and mechanically. Manufacturers that design their cable using 100-kpsi fiber will make a different cable than those that base their designs on 50-kpsi fiber. A cable designed with 100-kpsi fiber will let the manufacturer decrease the strength members of the cable, because the fiber is stronger. This, in turn, will lead to a greater transfer of external forces to the fiber. For example, if a cable designed with 100-kpsi fiber is placed at a maximum long-term tensile rating of 25%, the fibers would experience a stress level of 25 kpsi. Fibers in a similarly placed cable designed with 50-kpsi fiber would see a stress level of only 12.5 kpsi.
The higher strain levels permitted for 100-kpsi cables, then, tend to negate the strength advantages conferred by 100 kpsi-fiber. This does not necessarily mean that problems exist with 100-kpsi cable designs, but it does illustrate how little impact 100-kpsi fiber has on cable reliability.
The manufacturing process may be as important to cable strength as design considerations. At each manufacturing step, optical fiber faces risk of damage because of handling. In addition, because processing lines operate under tension, the fiber may encounter significant stress levels, causing flaws to grow and the fiber to weaken.
Cable design and manufacture, in fact, have more impact on the longevity of fiber than whether it is originally proof-tested to 50 or 100 kpsi. Installers and users, therefore, should be more concerned about choosing a high-quality manufacturer of fiber-optic cable than they should about the strength of the fiber component of the cable.
Optical fibers have different size flaws that can grow larger under stress.
Proof-testing optical fiber involves running a continuous fiber through a test station, where it is subjected to stress of a preset value--usually 50 or 100 kpsi.
Commercial Wiring Standard Revised
The Telecommunications Industry Association, or TIA, (Arlington, VA) has announced the long-awaited Revision A of its TIA/Electronic Industries Association standard 568 for commercial wiring. This revision has involved the collaboration of hundreds of individuals who have met over a three-year period to pool their technical expertise in revising the 1991 version of the standard. The goal has been to incorporate the ever-changing technology of premises distribution.
The development of this standard and other related technical documents, such as Telecommunications Systems Bulletins, represents the joint effort of both manufacturers and users. In addition to the approximately 100 manufacturers who contributed to the development of the revision, valuable input was also received from members of the Building Industry Consulting Service International (Tampa, FL) through its representative, Donna Ballast.
Special mention also goes to Paul Kish, chair of the TR-41.8.1 Working Group, and Jackie Wag- ner, editor of the document, for their work in the revision of TIA/EIA-568. George Lawrence, chair of the TR-41.8 Premises Distribution Committee, should be also be recognized for his guidance of TR-41.8 and its working groups.
Many people have asked why this document took so long to be developed. It should be noted that the TIA is accredited by the American National Standards Institute (New York, NY) and that a prescribed ANSI procedure is in place for the development of standards. Key to this process is that the development of documents be conducted in a forum versus a consortium and that documents be approved by consensus.
When the proposed draft revision of EIA/TIA-568 was balloted for the first time in 1992, more than 70 substantial comments were received. Under the ANSI procedure, all comments had to be addressed. The committee worked to resolve all issues and reballoted the proposed draft in the summer of 1994. Once again, several comments were received and had to be incorporated into the final document.
The final text for the standard, along with supporting documentation, was submitted to the Standards and Technology Department of the TIA, which processed the proposed standard for submission to ANSI. Within ANSI, the document is submitted to the Board of Standards Review, and if all prescribed procedures have been followed, ANSI approves the document for publication.
It should be noted that the documentation, not the technical content, for a standard is reviewed. The TIA cannot publish a document until the official notice from ANSI is received.
The review of TIA/EIA-568A took longer than expected, but the TIA is just as eager as the users of the standard to have the document published in an expeditious manner.
Telecommunications Systems Bulletins, it should be noted, are not ANSI standards. These bulletins are informative documents that serve as technical references. A Telecommunications System Bulletin does not follow the same procedures as a standard for publication. For example, after TR-41.8 approved TSB-67 and TSB-72, the documents were reviewed by the Technical Standards Subcommittee of TIA, after which they were published.
The TIA values its ANSI accreditation and will continue to follow the prescribed procedures to develop its industry standards. It is anticipated, however, that the use of electronic media will allow the acceleration of the standards process, and the average turnaround time for a standard will be lowered to 12 to 15 months. The TIA`s goal is to remain the industry standard developer for premises distribution.
Scott Stevens is technical services manager at Commscope Inc. in Claremont, NC.