NONWOVENSJournalA Science and Technology PublicationVolume 11, No. 3 Fall, 2002

I N T E R N AT I O N A L

Role of Fiber Morphology In Thermal Bonding

Fiber Motion Near The Collector During Melt Blowing Part 2: Fly Formation A Comparison of Needlepunched Nonwoven Fabrics Made From Poly(trimethylene terephthalate) and Poly(ethylene terephthalate) Staple Fibers

Linear Low Density Polyethylene Resins For Breathable Microporous Films

Fiberglass Vs. Synthetic Air Filtration Media Patent Review ... Researchers Toolbox ... Technology Watch ... Directors Corner ... The Nonwoven WebSponsored By

The International Nonwovens Journal is brought to you from Associations from around the world. This critical technical publication is provided as a complimentary service to the membership of the Associations that provided the funding and hard work.

PUBLISHERINDA, ASSOCIATION OF THE NONWOVEN FABRICS INDUSTRY TED WIRTZ PRESIDENT P.O. BOX 1288, CARY, NC 27511 www.inda.org

SPONSORTAPPI, TECHNICAL ASSOCIATION OF THE PULP AND PAPER INDUSTRY WAYNE H. GROSS EXECUTIVE DIRECTOR/COO P.O. BOX 105113 ATLANTA, GA 30348-5113 www.tappi.org

NONWOVENSJournalA Science and Technology Publication

I N T E R N AT I O N A L

Vol. 11, No. 3

Fall, 2002

The International Nonwovens Journal Mission: To publish the best peer reviewed research journal with broad appeal to the global nonwovens community that stimulates and fosters the advancement of nonwoven technology.Publisher Ted Wirtz President INDA, Association of the Nonwoven Fabrics Industry Sponsors Wayne Gross Executive Director/COO TAPPI, Technical Association of the Pulp and Paper Industry Teruo Yoshimura Secretary General ANIC, Asia Nonwoven Fabrics Industry Conference Editors Rob Johnson 856-256-1040 rjnonwoven@aol.com D.K. Smith 480-924-0813 nonwoven@aol.com Association Editors Cosmo Camelio, INDA D.V. Parikh, TAPPI Teruo Yoshimura, ANIC Production Editor Michael Jacobsen INDA Director of Publications mjacobsen@inda.org

ORIGINAL PAPERS Role of Fiber Morphology In Thermal Bonding Original Paper by Subhash Chand, Gajanan S. Bhat, Joseph E. Spruiell and Sanjiv Malkan, University of Tennessee-Knoxville . . . . . . . . . . . . . . . . . . . . . . . 12 Fiber Motion Near The Collector During Melt Blowing: Part 2 Fly Formation Original Paper by Randall R. Bresee, University of Tennessee-Knoxville, and Uzair A. Qureshi, Jentex Corp.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21 A Comparison of Needlepunched Nonwoven Fabrics Made From Poly(trimethylene terephthalate) and Poly(ethylene terephthalate) Staple Fibers Original Paper by Dr. Ian G. Carson, Shell Coordination Centre s.a., Monnet Centre International Laboratory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28 Linear Low Density Polyethylene Resins For Breathable Microporous Films Original Paper by W.R. Hale, E.D. Crawford, K.K. Dohrer, B.T. Duckworth, Eastman Chemical Company . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 Fiberglass Vs. Synthetic Air Filtration Media Original Paper by Edward Vaughn and Gayetri Ramachandran, Clemson University . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41 DEPARTMENTSEditorial Researchers Toolbox Directors Corner Technology Watch 4 5 7 9 Nonwovens Web Nonwovens Patents Association News Pira Worldwide Abstracts Meetings 54 57 61 63 66

EDITORIAL ADVISORY BOARD Chuck Allen BBA Nonwovens Cosmo Camelio INDA Roy Broughton Auburn University Robin Dent Albany International Ed Engle Fibervisions Tushar Ghosh NCSU Bhuvenesh Goswami Clemson Dale Grove Owens Corning

Frank Harris HDK Industries Albert Hoyle Hoyle Associates Marshall Hutten Hollingsworth & Vose Hyun Lim E.I. duPont de Nemours Joe Malik AQF Technologies Alan Meierhoefer Dexter Nonwovens Michele Mlynar Rohm and Haas Graham Moore PIRA D.V. Parikh U.S.D.A.S.R.R.C.

Behnam Pourdeyhimi NCSU Art Sampson Polymer Group Inc. Robert Shambaugh Univ. of Oklahoma Ed Thomas BBA Nonwovens Albin Turbak Retired Larry Wadsworth Univ. of Tennessee J. Robert Wagner Consultant

EDITORIAL

Stealth ReadersBy Rob Johnson and DK Smith Technical Editors, International Nonwovens Journal

e know you are out there because our website people tells us that each issue of the INTERNATIONAL NONWOVENS JOURNAL receives more than 10,000 hits during the quarter after publication. Even more remarkable, the older issues of the INJ still each get up to 5,000 hits during the same period. Yes, we know you are out there ... and we would like to hear from you from time to time. The editors of the INJ currently have plenty of contact with several groups. There is frequent discussion with the authors of the technical papers and with members of our outstanding Editorial Advisory Board who peer review these papers every issue. We also receive significant feedback and input from the INDA Technical Advisory Board, whose Mission Statement now includes the line: Assure that the INJ remains an effective technical vehicle. What we would like, in addition to these important elements, is input from our readers. We need to know and understand what you are thinking so we can better serve you. We welcome comments on any aspect of the journal, even the stuff you dont like about it. Of course, the primary mission of the INJ is to publish peer reviewed research papers and, consequently, we consider this the most important aspect of the journal. Your suggestions on topics as well as comments on the papers pub-

W

lished to date are always welcome and can only serve to strengthen the journal. Do you agree with the authors results and conclusions? Perhaps you have additional insight to offer or comments that might spur further research. Well never know unless you tell us. The other key portion of the INJ consists of the various departments where our objective is to collect and disseminate useful information pertinent to technical professionals and others in the

nonwovens and related industries. The regular key departments include: Editorials Directors Corner Researchers Toolbox The Nonwovens Web Technology Watch Worldwide Abstracts Organization/University Focus Patent Review Association Page Here, again, we seek your comments and suggestions. Are these the correct subjects for departments to reflect your interests and needs? What do you like? What do you dislike? Are there topics for inclusion? Perhaps you have a suggested article that can be summarized in one of the departments. Perhaps you feel strongly about something and want to offer a guest editorial. Just let us know. You can reach us and forward your comments, suggestions and submissions to Rob Johnson at rjnonwoven@aol.com.

INJs Electronic Patht has been almost two years since we announced the online format of the INTERNATIONAL NONWOVENS JOURNAL that commenced with the Spring 2001 issue. It seems that we were ahead of the curve at the time and it is now fully apparent that this move was correct in that we see many journals and other publications that have followed us online. As we stated earlier, we now get more than 10,000 hits during the quarter after publication and we feel this compares favorably with the prior hardcopy press run of 5000 copies. Further, the online format has provided several additional advantages, including allowing INDA and TAPPI to make the decision to offer the INJ free to anyone in the world with Internet access. For another, being online offers the use of color, which increases the clarity of many tables, graphs and photos included in the journal. A good example of this value is a paper in the Winter 2001 issue, Use of Infrared Thermography To Improve The Melt Spinning And Processing of Polyester Fibers by Glenn Gibson and Mark Tincher, Eastman Chemical Company, Kingsport, TN. This paper obviously benefited from color, as much of the information would have been lost in black and white. RJ, DKS

I

4 INJ Fall 2002

INJ DEPARTMENTS

RESEARCHERS TOOLBOXDigital Cameras for Microscopy Because of their convenience, flexibility, and low-cost per photo, digital cameras have gained a great deal of popularity among the picture-taking public. Operation of the camera can be almost foolproof as well as flexible, giving surprisingly good results in a wide variety of conditions. These features, coupled with the ability to see the results immediately, as well as a cost per shot that makes multiple exposures almost mandatory, have made the transition to digital photography an irresistible force. The same movement is occurring within photomicroscopy, the union of the camera and the microscope. Although photomicroscopists are generally skilled photographers as well, the ease, convenience and cost are major strong drivers for the trend. Further, the ease of storage, retrieval, transferral and quantitative analysis of digital photomicrographs make this capability a significant research tool. With the capability of 3.3+ million pixel CCD resolution, superb digital images are almost guaranteed, even of very fine structures within a specimen. Olympus Optical Co. of Hamburg, Germany, has concentrated on the development of a line of digital microscope cameras. These cameras are fitted with a universal C-mount thread, allowing attachment to almost any microscope. The Olympus DP 12 is a compact digital camera with 3.34 million pixel resolution. The camera system is provided with a tilting 3.5 LCD monitor, which is integrated into the control pad. This allows adjustable observation at the ideal angle. Real-time display of large, easy-to-see images allow faster, more accurate focusing and framing. Date, time, shutter speed and file name are displayed and stored together with the image; up to 16 acquired images can be displayed on a connected computer screen at one time for on-screen image selection. Sharp focusing, even at low magnifications is made possible with an electronic focus indicator and a 2x digital zoom function. One touch, automatic and manual white balance modes are available for optimal color representation, and users can choose 1% spot and 30% exposure metering and automatic or manual exposure modes. Removable SmartMedia cards store up to 138 MB of images, which can be easily transferred to any PC. Optional software allows images to be downloaded directly from the camera to the PC. This company has just introduced a new, compact digital microscope camera, the Olympus ColorView II. This unit incorporates Firewire Technology, which is similar to a USB connection, but has a much higher data transmission rate. The camera is wired to the LCD screen or a computer, and transfers the photomicrographic image very rapidly. For information: Olympus Optical, D20097 Hamburg, Germany; 49+40/237730; www.olympus-europa.com . Coating and Laminating Equipment New capabilities for studying CCL processes (Coating, Combining and Laminating processes) in the laboratory and plant are emerging, as new small scale and production scale equipment is developed. The following describes some recent introductions. American Santex has introduced their Cavitec Modular Hot Melt Coating and Laminating system. This system provides for more than one application method in the same process line. The Cavilex base station can be equipped for three different processes: Cavimelt engraved roll coater; Caviroll roll coater; and the Cavislot slot die coater. All three systems provide considerable variability for suitable substrates

and coating formulations, along with precise control of process variables. Additional details are available from: American Santex, Spartanburg, SC; 864574-7222: www.santex-group.com . For laboratory work, the Coatema Easycoater discontinuous lab unit offers an economical and easy-to-operate set-up for preparing small hand samples with constant coating weight and thickness. The coating head in this unit is a high-precision stainless steel Doctor Blade that can also be used as an Air Knife System. The coating head can be adjusted to various heights and angles with a precision screw and micrometer gauge. This unit also has a companion mixing set-up for preparation of 3-5 liters of coating formulation. The Easymixer is fitted with an explosion-proof motor and is scaled for splitting batches to cover a variety of formula modifications. For more: Coatema Coating Machinery GmbH, Spartanburg, SC; 846-582-1900. Reliant Machinery, the major UK manufacturer of flatbed laminators, has positioned its Reliant Powerbond Mark III series unit into their line of powder, film and adhesive web laminators. They claim flexibility, ease of operation, maximum production and improved quality. It has special features such as a heat tunnel that adjusts from zero to 50 millimeters for thick and thin materials, along with standard heat tunnels from 1.7 to 5.7 meters length in 1-meter increments. The heat tunnel can be fitted with 10-zone heat controls; it comes in standard width of one to three meters. The Mark III also includes their Synchro-Trak automatic belt tracking system, refrigerated cooling modules, microprocessor controls, and embedded diagnostics. In the U.S., Reliant is represented by Apparel Equipment, Philadelphia, PA; 215634-2626; www.reliant-machinery.com . Liquid Carbon Dioxide In Apparel Cleaning The use of liquified carbon dioxide has generated a considerable amount of interest over the past few years. The reason for this interest is the tremendous solvent power of liquid carbon dioxide. In this state, the material acts as both a liquid andINJ Fall 2002 5

RESEARCHERS TOOLBOXa gas, hence is able to more easily penetrate into materials and exert its strong solvent action. This unique physical state is achieved at an elevated pressure and temperature of the carbon dioxide. The use of the strong solvent power of this system has been exploited in the research laboratory to some extent. Also, the use of liquid CO2 in textile cleaning and scouring operations has been studied rather extensively at North Carolina State University and the University of North Carolina. Joseph M. DeSimone, a chemistry and chemical engineering professor at the University of North Carolina in Chapel Hill has done considerable research on this system. In 1995, Professor DeSimone founded the company Micell Technologies to market an apparel dry cleaning process based on this research. This latter company has been exploiting the technology through a series of dry cleaning establishments under the name of Hangers dry cleaners (16 stores in Southeast U.S.). In recent years almost all dry cleaning operations have been based on the use of hydrocarbon and chlorinated solvents, particularly perchloroethylene (so-called Perc). Such solvents have had real disadvantages to their use including flammability and potential for causing cancer. Hence, there has been an interest in replacing such solvents. Liquid carbon dioxide has none of these disadvantages in this application. To use this solvent, the dry cleaning equipment has to be pressurized, but this has been accomplished fairly easily. Expanding the use of this solvent has been accelerated by the introduction of special boosters into the solvent to facilitate the removal of some types of soil and spots. This has been achieved by additives, generally thought to be based on fluorine- or silicon-based surfactants. Such an improved solvent system based on liquid carbon dioxide has been introduced to the industry under the trade name Washpoint, by a joint development of ICI and Linde. These two companies joined in product development efforts in 2000, which has resulted in the proprietary Washpoint product. This product is now6 INJ Fall 2002

being introduced into the dry cleaning industry. Linde previously had entered the dry cleaning business through its merger with AGA a few years ago. This earlier effort was based on a solvent termed Dry Wash fluid, which had been developed by Raytheon Environmental Systems and Los

Alamos National Laboratories. The new Washpoint solvent system is compatible with the Micell system. It is apparent that these commercial activities will expand the use of this solvent system, and will very likely extend the use in the textiles and apparel industries, as well as increased laboratory use. INJ

Laboratory Techniciansmong the numerous unsung heroes of the R&D scene, laboratory technicians often comprise a group that is significant in number and contribution. Generally the workhorses that get the uninteresting and tedious assignments, their suggestions and contributions can often prove critical in bringing home the successful development project. One company within the nonwovens industry makes it a practice to include laboratory technicians as co-inventors when they honestly made a contribution to a new invention. A couple of the Techs within that Research Division had more patents to their credit than some long-time professionals, and rightly so! Too often, however, these unsung heroes are just that playing a significant role, making a contribution, but always on the sidelines. Recently, more attention has been paid to this group, and their gripes, hopes and views have been seriously noted and considered. Some scientific and engineering organizations have taken steps to recognize and highlight the expressions of this important group. Several professional societies have modified their bylaws and clearly established membership categories for qualified technicians. This often involves a clear statement that an individual with an associate degree with a certain minimum amount of applicable experience can join with other professionals in the society. A study of this situation by a professor at Stanford University has identified the three major Rs desired by laboratory technicians to clearly establish their status and recognize the value their work brings to the scientific community. These three major desires include the following: Respect technicians wants to be respected for the professionals they are, for the value of their experience and ability to contribute. Recognition Acknowledgment of their efforts and occasional public recognition of their contributions and accomplishments. Responsibility Commensurate with their skill and capability, the technician wants and needs opportunities to do more and learn more, to take on more responsibility and thus experience personal grow. The seasoned, long-time researcher who has worked with a variety of technicians can often recall one or more Techs that would truly be preferred on the R&D team over a lot of professionals. When the Three Rs desired by laboratory technicians are thoughtfully considered, they are not at all surprising. After all, respect, recognition and responsibility are consciously or unconsciously sought by all rational human beings. A discussion in these columns several issues ago dealt with appropriate and meaningful actions and recognitions that managers can arrange for deserving professional researchers. Those same items can often be very appropriate for the technician as well.

A

INJ DEPARTMENTSIt is important that the manager keep a checklist in each employees file that details who was contacted as a reference and what was learned. It is important to try to check each reference give. Little information may be gained, but the file must show that an honest and reasonable effort was made to get such information. If you do not try, you might be found negligent. For an employee with a previous conviction, the Equal Employment Opportunity Commission says that employers must consider three factors to justify use of a conviction record: The nature and gravity of the offense for which the applicant was convicted; The amount of time that has elapsed since the applicants conviction and/or completion of the sentence; The nature of the job in question as it relates to the nature of the offense committed. Further, if an employers finds out that an employee had problems with violence in the past and nothing is done about it, the employer could be found liable for Negligent Retention. Also, another related employer fault is gaining accep-

DIRECTORS CORNERBeware Negligent Hiring As has been discussed in this column in the past, the job of todays manager is becoming more and more complex; the task is truly filled with hazards, potential lawsuits and actions to be avoided. Here comes another one. A new employee is added to the payroll after the usual interview and reference check. Very shortly, it becomes obvious that the new hire is a discontented, rather unbalanced individual. As a reward for the baggage of associated problems, the person is assigned to a less desirable job. A dangerous and violent nature comes to the surface and a clash or incident occurs where a fellow employee or worse yet, a visitor is injured. What comes next? This could be an example of what is becoming known as a Negligent Hire, a legal term that describes a violation of the basic duty of a company or organization, the duty to exercise care in hiring. In such situations, an employer can be questioned as to whether they took all reasonable steps before the hiring decision to identify whether or not the problem employee had any past misconduct or unfit behavior on the job. The reference check may be cited as evidence of due care. However, everyone knows that most organizations are reluctant to give a former employee a less-than-average rating. So, what is a manager to do? Recent court rulings have found that managers who are contacted by any company doing a pre-employment check on a former employee must reveal any serious misconduct by that employee. Withholding such information can put them at risk for a lawsuit. Almost every state now has a law which is designed to address this problem. Invariably, former employers are legally obligated to mention any misconduct involving violence or acts that physically endangered other individuals. Failure to disclose such past misconduct by an employee subjects the previous employer to damages the employee might inflict in future workplaces. Most managers who hire new employees know that it is important to conduct some kind of applicant pre-screening or background check. This means checking references, talking with previous employees, and for certain jobs, conducting criminal or motor vehicle department checks. Also, managers should find out if applicants have ever been convicted of a crime. This question is usually on the written application. It is illegal to ask about arrests, but it is okay to ask about convictions.

Workplace Greenery Reduces Stresshis Department has frequently observed the importance and interest in workplace stress. With individuals spending a major portion of their lives in the workplace, it only makes sense to examine from time to time the factors that can alleviated or moderate the stress encountered there. Some stress-generating elements cannot be eliminated, of course, but that does not excuse consideration of anything that can help the situation. An interesting and surprising stress-reducing factor that has received consideration is the use of interior plants in the work environment. It has been shown by these recent studies that such greenery can be a helpful factor. Visual exposure to a plant setting has produced significant recovery from stress with five minutes, while enhancing productivity by 12%, according to a study by Texas A&M University and Washington State University (WSU). WSU research also confirmed that once exposed to plant settings, test participants demonstrated more positive emotions, such as happiness, friendliness, and assertiveness, as well as fewer negative emotions, such as sadness and fear. The researchers concluded that interior workplace plants signal stability and offer employees a touch of humanity while stimulating a more productive environment. Growing plants also consume and lower the sleeping-promoting carbon dioxide level within an enclosure, replacing it with more stimulating oxygen. These finding may surprising some research administrators, but most housewives can vouch for their authenticity. For more detailed information, go to www.plantsatwork.org.

T

INJ Fall 2002 7

DIRECTORS CORNERtance in legal circles and that is Negligent Supervision. With all of these potential worries, an employer does have some help. Alist of characteristics that experts in the field have identified as indicative of the possibility of workplace violence has been assembled (www.noworkviolence.com/articles/preventing_violence.htm ). Studies have shown that more than 35% of job applicants lie on their employment applications. The courts have not ruled that employers must verify what applicant have written on their job applications. However, if the employer does not question about prior convictions, there may not be a defense against Negligent Hiring. On the other side of the coin, there is a growing concern, especially with union organizations, that some employers may be digging too deeply into the employees past. This has been particularly true for organizations who employ outside contracting firms which may not do as thorough a job with their employees past as the company wants. There is such a thing as a company digging too deeply into an employees past, as evidenced by some current law suits. In some cases, the defendant employers are claiming they are only following government mandates. So, like a lot of elements in life, there has to be a balance in appropriate actions. Safety Items The following are a variety of safety ideas that may be applicable to lab, plant or office environments. These items have been collected from numerous sources. Our business is in a small community, and we are serviced by a very dedicated volunteer fire department. Every year we invite the officers to tour our facility so they are aware of the layout of the structure and its contents. We believe this is beneficial to both parties and may assist in a rescue or their ability to put the fire out. One recommendation involves asking the local polices bomb squad and the local Emergency Preparedness team to join fire department officers in a tour of the facility. A similar suggestion points out that in many states, a county Emergency Management Agency (EMA) is mandated.8 INJ Fall 2002

In such states, any facility possessing a certain level of dangerous chemicals or materials must report approximate amounts and locations of these substances to the local EMA. This allows the agency to help in many aspects, including planning evacuation routes, should the need arise. Contacts with and visits from the local EMA should be a must do item for all pertinent locations. I am a volunteer firefighter and EMTIntermediate, and my company understands that I might come in late or leave early if I get a call. Companies should encourage employees to volunteer and design their personnel policies to support employees commitment to contributing to the community. A useful suggestion is having the local fire department help train on-site emergency responders. We make a generous donation to them for their training services whenever provided. A cautionary note regarding the suggestion to have the local fire department help train on-site emergency responders. C.J. Palmer, an EMS and fire science educator who has actively practiced in the field for more than 25 years, says, I would make sure the instructors are competent in the subject matter and that they hold an instructor credential from an agency recognized in the field. Most of us in EMS know little about OSHA requirements, and I continuously run into well-meaning employers who are relying on interpretations from people who are not well-versed in regulatory affairs. Enough said! Hearing is an invaluable sense we tend to take for granted. Hearing loss has been found to take place at noise levels of 85 dBA and higher. An easy way to determine if you are in noise levels higher than 85 dBA is when you and another person/co-worker have to raise your voices in order to communicate when standing about three feet from each other. A very helpful Internet site for virtually all aspects of safety is that of the National Safety Council (NSC), located at www.nsc.org .The site offers a very excellent First Aid and CPR training program. This program was recently offered free of charge for a week to commemorate the

September 11 Terrorist Attack; normally it is offered for a nominal charge. This site also offers a mobile field reference to emergency medical information that can be loaded in a PDA (Personal Digital Assistant). This reference can be loaded with the latest medical information, and can be undated as new information is available. You can even add your own local protocols to the database. The site has a wide variety of features that contain a substantial amount of interesting and useful information. Use it!! Science Safety, the website of the Laboratory Safety Institute (LSI) is a useful place for a variety of helps on various aspects of laboratory safety. While especially directed toward laboratory safety in science education, much of the information is universal in nature. The Laboratory Safety Institute (192 Worcester Road, Natick, MA 01760; 508657-1900) is a non-profit center for health, safety and environmental affairs. LSI indicates that its mission is to make health and safety an integral and important part of science education, the work and lives of scientists and science educators. That such a need exists is highlighted by their statement that the accident rate in schools and colleges is 100 to 1,000 times greater than at Dow or DuPont. The Science Safety site offers a variety of products and services including mini-grants, audio-visual lending library, a variety of products, seminars and training sessions, custom training services, audits, information on regulatory compliance, an online library with graphics and PowerPoint files, a newsletter and numerous other facilities - (www.labsafety.org). By way of reducing the inventory of chemicals and potentially hazardous materials, the American Chemical Society has prepared a publication entitled Less is Better. This offers a variety of techniques to reduce such inventories without hindering the progress of research and development efforts. (American Chemical Society, 1155 16th Street, N.W. Washington, D.C. 20036; 202-872-4600). INJ

INJ DEPARTMENTSnent of smog, as it can be an important pollutant in addition to being blamed for a variety of reactions that increase the smog-forming potential of various other chemical pollutants. To the chemist, the easiest solution to the ozone problem would be to simply convert ozone (O3) into normal oxygen (O2), which constitutes about 1/5 of the atmosphere and is a vital and good component of air. The problem, of course, is how to achieve such conversion easily, inexpensively, and without any other attendant problems. A fascinating approach to achieving this conversion is being exploited in a small way by some drivers using their automobiles to a greater extent. This sounds a little incongruous, as automobiles are considered to be a major part of the problem of ground level pollution. With the right technology, however, they could become part of the solution. The key to this approach is a catalyst system that can achieve such a conversion at ambient or somewhat elevated temperatures, without any deleterious side effects. Such a catalyst system can be coated on the radiator of an automobile. As such a car is driven, a large volume of air is pulled through the radiator and the ground-level ozone contained therein is converted to normal oxygen. The catalyst system has been termed PremAir Technology by its producer, the American company Engelhard. It is keeping the identity of the catalyst a secret for the time being, but industry sources point out that the patent literature suggests that manganese oxides MnO2 and Mn2O 3 are involved. Depending upon factors like the speed of the car, the catalyst can convert 60 to 80% of the ozone flowing through the radiator into oxygen. Several car manufacturers are looking at this technology to give them a boost in their environmental image, as well as in meeting some governmental requirements that become mandatory in the future. Volvo has had PremAir Technology on some of its models for several months. Also, BMW is usingINJ Fall 2002 9

TECHNOLOGY WATCHGMP and the Converting Industry Several sectors within the nonwovens converting industry pay close attention to GMP, or Good Manufacturing Practices. These are requirements mandated by the Food and Drug Administration to the pharmaceutical, medical and related industries involved in manufacturing and marketing products that can impact human and animal health. Companies that manufacture various classes of medical devices also must meet GMP standards; such standards not only relate to proper manufacturing procedures, but also cover raw materials, records, distribution and other elements of production and use. Some of these requirements are designed to provide the means to check all aspects of a specific production lot and its use if a problem should arise in the eventual consumption or use of the product. The FDA has just begun a sizeable review of its GMP requirements for the pharmaceuticals industry; while the current review will focus initially on pharmaceutical products only, it is likely that any new aspects of GMP will find their way into GMP standards for other manufacturing operations, including personal, medical and sanitary products. The current review will cover manufacturing of veterinary and human drugs, including biological products and vaccines. The effort will strive to make manufacturing processes consistent and safer, according to FDA officials. Deputy FDA Commissioner Lester M. Crawford said in discussing the review, Any system can be improved upon, and with this riskbased, highly integrative Good Manufacturing Practices initiative, we intend to do just that. Three goals are cited for the initiative: 1. Focus more on processes that present actual risks to public health. 2. Establish quality standards that do not impede innovation or introduction of new technologies. 3. Enhance predictability in FDAs approach to quality and safety. Over the next couple of years it is anticipated that the FDA will gather detailed information from the pharmaceutical industry, and also from manufacturing experts, academia, government and consumer groups relating to these issues. It is also anticipated that any principles, practices and standards developed from this review will be adapted for modifying GMP requirements in other related industries. Also, there will undoubtedly be some international implications of such review of GMP requirements, as there is significant importation of many of these product types into the U.S.; also, there is a tendency for such U.S. standards to be utilized domestically in other countries. Hence, the potential impact of this review may be quite significant. Cleaning Up the Ozone Ozone is one of the major targets in efforts to clean up the air surrounding the earth. Ozone is a important compo-

TECHNOLOGY WATCHthe system on cars sold into certain states in the U.S. While the heat from the car radiator doesnt hurt, the reaction does not require the elevated temperatures necessary for precious metal catalysts as used in conventional catalytic converters. The technology is not considered to be a complete solution to the ozone problem, as it can process the ozone in only a small fraction of the earths atmosphere. However, to concept of using the automobile to do some cleaning of the air is certainly novel. Engelhard is looking for other applications for the catalyst system in addition to the use in automobiles. Use in air conditioner condensers and other architectural applications may be feasible and advantageous. In view of the extensive use of nonwovens in air filtration applications, it is not a wild stretch of the imagination to think of a modification of this system to engineer nonwoven fabrics that not only rid air of its particulate contaminants, but also chemical contaminants that are not now amenable to carbon filter media. Nonwoven fabrics that have chemically modified fiber surfaces are being exploited in blood filtration by selective chemical actions; why not a similar approach to cleaning up the IAQ (Indoor Air Quality) problem. Also, for anyone who has a laser printer close by, the odor of ozone may be apparent from time to time. Some people claim a little ozone can be helpful, but basically it is a poisonous gas, so elimination of an excess by such an active ventilation system might be a good idea. Recycling PVC Plastic Polyvinyl chloride plastic has been under the gun from numerous environmental groups, who perceive the material to be a real environmental problem. Recently, some favorable publicity was gained by the PVC industry by some significant success achieved in recycling waste PVC. Earlier this year, the Vinyloop10 INJ Fall 2002

recycling process went into commercial operation at a plant in Italy. This first industrial unit was started up at a plant of Solvay, a major chemical company headquartered in Brussels, Belgium, that is a major producer of PVC resin. The new operation is designed to recycle 10,000 tons per year of waste PVC plastic, most of it insulation material coming from electrical cable, 80% of which is of post-consumer origin. The plant is being operated by Vinyloop Ferrara SpA, which is a joint venture of four European PVC producers: SolVin Italia (a Solvay company), Adriaplast, Tecnometal, and Vulcaflex. The venture has received financial support from Vinyl 2010, the European PVC Industry committee devoted to the voluntary recycling effort. A second Vinyloop recycling plant is

being designed to recycle PVC-coated tarpaulins and fabrics produced in Europe; Ferrari S.A. of France is a major producer of such coated textile fabric and provided considerable assistance in developing the process. This second unit is scheduled to begin operation in 2004; other recycling units are being considered for Europe, Canada and Japan. Obviously, recycling is becoming a major factor wherever a product raw material is used in large volume. Digital Printing of Nonwovens The company Leggett & Platt has roots going back several years into the nonwovens industry. For many years, the Nashville company was noted as a major producer and marketer of highloft fabrics, needlepunch fabrics, waddings and other specialty nonwo-

PDAs to PocketPCsll PDAs (Personal Digital Assistants) are not created equal. If all you want to do is store names and phone numbers, any electronic organizer will fill your needs. However, if you often find yourself away from your computer, whether out of the room or out of the country, you may want to consider one of the beefier handheld offerings that are now becoming popular. The latest development in the digital assistant world is the introduction of the PocketPC a device that is kind of a cross between a laptop computer and a simple digital organizer. Compaqs iPaq was the first such device to really hit the market a few months ago, but recently companies like Toshiba, Sony, and others have rolled out their own version of the PocketPC. The thing that sets these handhelds apart from the Palm Pilot of five years ago is that they run much of the office software that people are already familiar with. Most of them can run Windows CE, a lightweight version of Microsoft's popular desktop operating system. The majority of them also run stripped-down versions of MS Word, Excel, Outlook, and Windows Media Player. Several of the current models also support Java. If you get a model that is equipped for a wireless network (optional in most cases), you can also check your email and surf the Web on your palmtop. Their familiar interface and inter-operability with desktop computers have made the new generation of handheld computers very popular as a practical office tool. Imagine that instead of recording laboratory data by hand and then recopying it to your desktop computer, you simply enter it into an Excel spreadsheet on your Pocket PC. Once the data is stored, it can either be transferred via a wireless network connection, or through its cradle, which connects it to a desktop computer. For more information on the various commercial models: www.compaq.com/products/handhelds/pocketpc/H3870.html; www.pda.toshiba.com; http://products.hp-at-home.com/products/

A

TECHNOLOGY WATCHvens. L&P marketing activities were well-known in the furniture, bedding, home furnishings, wipes and other product areas. In the past few years, it has been involved in several acquisition and mergers. The present L&P is actually Leggett & Platt Digital Technologies and it has focused on digital printing of a wide variety of substrates. Also, a major business for the company is the development and marketing of digital printing equipment, accessories and substrates for digital printing. Much of this current business, both equipment and substrates, is in the graphics industry, specifically for soft signage, banners, flags, pennants, point-of-purchase displays and similar items. In the digital printing equipment area, L&P Digital offers many models of industrial printing machines in a wide and super-wide format (98 to 138). These units utilize piezo dropon-demand inkjet printing heads, with as many as 8 heads on a unit for bidirectional printing. These units can handle roll-to-roll operations, as well as some models designed for discontinuous operation on rigid substrates, up to 3-inches in thickness. These digital printing machines are capable of processing a variety of substrates, including fabric, coated papers, textiles of a variety of types, film, vinyl (film and sheet), canvas, mesh and other types of soft/flexible and rigid specialty substrates. The digital printing can involves complex patterns, photos, color in an amazing variety, as well as selected textures. The company, harkening back to its roots, recently introduced a line of nonwoven products for the graphics industry. This line included 100% polyester and 100% polypropylene fabrics, along with blends in their VirtuNonWoven fabric line. These current fabrics are relatively light weight and are translucent for optimal signage use. Within the product line, VirtuMesh is a durable, bright white polyester mesh at 8 osy. The VirtuPoly Cotton and VirtuPoly Cotton Plus fabrics are made of polyester/cotton blends, and are for applications requiring a softer hand. This is certainly a specialized application for nonwovens, but it well illustrates the synergistic combination of nonwovens and advanced technology. (Leggett & Platt Digital Technologies, Jacksonville, FL; 904-249-1131; www.lp-digital.com). As has been mentioned in this column in the past, other nonwoven producers are taking an active interest in the application of digital printing to nonwoven substrates. INJ

INJ Fall 2002 11

ORIGINAL PAPER/PEER-REVIEWED

Role of Fiber Morphology In Thermal BondingBy Subhash Chand, Gajanan S. Bhat*, Joseph E. Spruiell and Sanjiv Malkan, The University of Tennessee, Knoxville, Tennessee USApopular method of bonding used in nonwovens. The main advantages of thermal bonding are low raw material and energy costs, product versatility, small space requirements, cleanliness of the process, better product quality characteristics, and increased production rates. Of the several types of thermal bonding such as area-bonding, point-bonding, air oven bonding, ultrasonic bonding and radiant bonding, point bonding is the most widely used technique [2]. Nonwoven fabric properties are determined by the characteristics of bond points and in particular by the stress-strain relationship of the bridging fibers. During point bonding, the bond points and the bridging fibers develop distinct properties, different from those of the virgin fibers, depending on the process variables employed. The changes in fiber properties have been hinted at by several authors [2-7] but have not been investigated. So far most of the research work [6-15] has been done to study the effects of bonding conditions on fabric properties. Some work [12, 16-17] has been done on the effects of fiber properties on final fabric properties. However, the role of fiber morphology and morphological changes taking place in the fibers due to applied heat and pressure in thermal bonding has been almost untouched. This has been mainly due to the fact that it is almost impossible to characterize the bond points and the fibers surrounding the bonds without the use of some innovative techniques. Point bonding is used for a wide range of fibers, from those with less developed morphology as in spunbonding to those with fairly well developed morphology as in staple fibers. Thus it becomes very important to investigate the effects of fiber morphology on bonding conditions and web properties. In this study, polypropylene fibers with a wide range of crystallinity and orientation, but with the same diameter, were produced. The fibers were then used in studies of their bonding behavior and web forming characteristics. Spunbond studies were also done in a similar way in order to see the generality of the observations made in the staple fiber study. It was reported earlier that fiber morphology has a definite Abstract The role of fiber morphology in a thermal point bonding operation was investigated. Primary objectives were to understand the changes taking place in fiber structure due to applied heat and pressure, and the role of fiber morphology in determining optimum process conditions and properties of the webs. To study fibers with varying morphology, i.e., from partially drawn as in spunbonding to fully drawn as in staple fiber nonwovens, fibers with a wide range of crystallinity and orientation were spun and characterized, from two polypropylene resins. Thermally bonded carded webs were produced, using these fibers, and characterized in order to understand thermal bonding behavior of fibers with different morphology. The fibers with different morphology differed significantly in their bonding behavior. The fibers with higher molecular orientation and crystallinity tended to form a weak and brittle bond due to lack of polymer flow and fibrillation of the fibers in the bonded regions. In general, fibers with lower molecular orientation and lower crystallinity yielded stronger and tougher webs. Fibers with relatively less developed morphology also exhibited lower optimum bonding temperature. Morphological changes in fibers were observed during the thermal bonding process, in bonded as well as unbonded regions of the web. As a final step to see how the observations from staple-fiber study translate to one of the relevant processes during scale-up, spunbond studies were also conducted in a similar way. Introduction The basic idea for thermal bonding was first introduced by Reed [1] in 1942. Since then, there have been a number of developments in this field. Thermal bonding is now the most Subhash Chand, currently with Nylstar, Inc., Ridgeway, VA Sanjiv Malkan, currently with Synfil Technologies, Knoxville, TN Gajanan Bhat, Corresponding Author12 INJ Fall 2002

role on the structure and properties of the thermal bonded nonwoven webs [18-20]. A summary of results from this comprehensive investigation is reported here. Experimental Methods Processing Fiber grade polypropylene, which had a melt flow rate of 17 dg/min, supplied by Montell USA Inc. was used for the production of fibers. Fibers were produced using a Fourne extruder and spinning setup and a conventional two-stage drawing machine. Extrusion temperature was kept constant at 230C. Polymer throughput rate and take-up speed were varied together in order to achieve the same final diameter for all the fibers. Out of six fiber samples produced, three were as-spun with no drawing and three were drawn after spinning. Drawing was done at 140C. The processing conditions used to prepare the fiber samples are summarized in Table 1. Continuous fibers were chopped into staple fibers of length 40 mm for carding. Staple fibers, with an appropriate level of water (10%) and LUROL PP-8049 spin-finish (0.4%) supplied by Goulston Inc., were carded on a Saco-Lowell carding machine to produce webs with a nominal basis weight of 40 g/m2. As the fibers did not have any crimp, it was important to have sufficient finish on the fibers, and to control the humidity of the room for successful carding. Carded webs were then bonded at several different bonding temperatures and at a speed of 5 m/min using a Kuster point-bonding calender having 15% bonding area. Speed was kept low due to difficulties in handling of small carded webs. Nip pressure was kept constant at 350 pli for all the samples. Spunbond studies were carried out using 35 MFR EXXON PP on the modified Reicofil-I line at the University of Tennessee, Knoxville. A schematic of the process variables is shown in Figure 1. Melt temperature and cooling air temperature were the main variables. Airflow rate was adjusted to achieve the same fiber diameter for all the three sets. Webs were bonded at four different bonding temperatures for each set of fibers. Other process parameters such as bonding speed and calender pressure were kept constant. Filament samples before bonding were also collected for analysis.

WEBS WERE BONDED AT FOUR DIFFERENT BONDING TEMPERATURES FOR EACH OF THE SETS

Figure 1 SCHEMATIC OF SPUNBOND PROCESS VARIABLESCharacterization of the Fibers and the Webs Fiber diameter and birefringence were measured using an optical microscope. Thermal analysis of the fibers and the webs was done using the Mettler thermal analysis system consisting of TC11 controller, DSC25 and TMA40 modules. The scans were done at a heating rate of 10C/min in air. Crystallinity was calculated from the DSC scans assuming that the heat of fusion of 100% crystalline polypropylene is 190 J/g. X-ray diffraction photographs for fibers were obtained using a flat plate camera and a Phillips x-ray generator. The x-ray wavelength was 1.542 A0 in all the x-ray studies. Crystal size was calculated using the Scherrer equation from the measured full width half maximum intensity of reflection peaks in the equatorial scans [21]. Duco Cement was used as a glue for sample preparation for equatorial scans. Use of Duco Cement was helpful in sample preparation from bonded regions (only) and from very short fibers taken from unbonded regions of the web. Bonded and unbonded regions of the web were carefully separated from the web using a sharp pair of scissors and analyzed for molecular orientation, crystallinity and crystal size. Tensile properties of the fibers and the fabrics were mea-

Table 1 PROCESS CONDITIONS FOR PRODUCTION OF FIBER SAMPLES. Polymer Throughput Rate (G/Hole/Min) 0.28 0.41 0.55 0.42 0.72 0.96 Nominal Spinning Speed (M/Min) 1000 1500 2000 1000 1000 1000

Sample Id As-spun 1 As-spun 2 As-spun 3 Drawn 1 Drawn 2 Drawn 3

Draw Ratio Undrawn Undrawn Undrawn 1.5 2.5 3.5

Denier 2.7 2.5 2.5 2.4 2.7 2.4

INJ Fall 2002 13

Figure 3 THERMOMECHANICAL RESPONSES OF STAPLE FIBERScm/min). A total of twenty tests were done for each sample. SEM images of the fabrics and the tested samples were taken using a Hitachi S-3000N electron microscope. Backscattered images with 30 Pa gas were taken in order to minimize the problems due to static charge generation. Results And Discussion Staple Fiber Studies Fiber diameter, crystallinity, and their mechanical properties are given in Table 2. Thermomechanical responses (TMA) of the staple fibers are shown in Figure 3. The six fiber samples covered a very wide range of morphology and properties. Fiber diameter was kept the same in all the cases so that the differences due to change in diameter could be minimized and the role of fiber micro-morphology in thermal bonding could be analyzed. As can be expected, there was an increase in crystallinity of the fibers with increase in spinning speed and draw ratio. This is an expected trend and the tensile data, i.e., increase in tenacity and decrease in elongation, is also consistent with the development of structure. TMA (Figure 3) data also supported the morphological differences between the fibers. Fibers with less developed morphology deformed easily, compared to the well drawn fibers that

Figure 2 SCHEMATIC OF SINGLE-BOND STRIP TENSILE TESTsured using a United Tensile Tester with test conditions described in the ASTM D3822-91 for filaments and ASTM D1117-80 for nonwoven fabrics [22]. However, for fiber samples, a gauge length of 2 (5.08 cm) and an extension rate of 10/min (25.4 cm/min) were used. For webs, a gauge length of 5 (12.7 cm), width of 1 (2.54 cm), and extension rate of 5/min (12.7 cm/min) were used in both machine direction and cross direction. A Single-Bond Strip Tensile Test was developed in order to estimate the bond strength and the degree of load sharing between the fibers during tensile deformation. A schematic of this test is shown in Figure 2. A tiny strip of size 80 mm X 5 mm was cut from the web. The strip was cut in the middle in the width direction from two sides to leave only one bond uncut in the middle of the strip, as shown in the figure. The strip was then subjected to a conventional tensile test. The test was conducted on the United Tensile Tester with a gauge length of 1 (2.54 cm) and extension rate of 0.5/min (1.27

Table 2 FIBER STRUCTURE AND PROPERTIES Sample Id As-spun 1 As-spun 2 As-spun 3 Drawn 1 Drawn 2 Drawn 3 Diameter M 20.8 19.5 19.7 19.9 20.7 19.5 Crystallinity (%) 36.7 41.3 45.0 48.9 53.7 56.4 Tenacity GPD 2.9 4.8 6.4 6.4 7.4 8.5 Breaking Extension (%) 290 280 190 160 60 25

14 INJ Fall 2002

Figure 4 WEB TENSILE STRENGTH VS. BONDING TEMPERATURE FOR STAPLE FIBERS

Figure 6 WEB BREAKING EXTENSION VS. BONDING TEMPERATURE FOR STAPLE FIBERS

Figure 5 FIBER TO WEB STRENGTH REALIZATION FOR STAPLE FIBERSshowed higher thermal stability. Tensile strength values of the webs produced from different fibers and bonded over a wide range of bonding temperature are shown in Figure 4. It was observed that web strength decreased with increase in fiber molecular orientation and crystallinity. Fibers with relatively less developed morphology yielded stronger webs compared to fibers with more developed morphology. Fiber to web strength realization (ratio of fiber strength to web strength) for different fibers is shown in Figure 5. Fiber to web strength realization decreased sharply with increase in fiber molecular orientation and crystallinity. Higher strength realization for the fibers having lower molecular orientation and crystallinity may be partly attributed to higher breaking extension of the fibers. Higher breaking extension of the fibers leads to greater degree of load sharing between the fibers during the deformation of the web. Optimum bonding temperature for drawn fibers was found to be higher than that for the as-spun fibers. Further, optimizing the bonding temperature did not help much in the case of highly drawn fibers, as can be seen from web strength versus bonding temperature relationship. Web breaking extension as shown in Figure 6 exhibited a trend similar to tensile strength. Wei et al. [14] and Bechter et al.

Figure 7 OPTICAL MICROGRAPHS OF THE BONDS AFTER THE TENSILE TEST: TOP = AS SPUN FIBERS; BOTTOM = DRAWN FIBERS[16] have also studied the effect of fiber draw-ratio on polypropylene nonwoven fabric properties and reported that fibers with lower draw-ratio resulted in fabrics with higher tensile strength. Fracture mechanism of the webs was studied using both optical and scanning electron microscopy (SEM). Optical micrographs of the bonds after the tensile test are shown in Figure 7, at optimum bonding temperatures, for as-spun and drawn fibers. The bonds did not rupture during web failure inINJ Fall 2002 15

Figure 8 SEM IMAGE SHOWING DISINTEGRATION OF BONDthe case of webs produced from as-spun fibers, for bonding temperature at and above the optimum. Whereas in the case of drawn fibers, web failure involved rupture of the bonds at all the bonding temperatures studied. It was observed that bonds were very weak and brittle in the case of drawn fibers. It is further evident from the image of elongated bond in Figure 7 that bonds were very ductile and strong in the case of as-spun fibers. Disintegration of the bonds during web failure in the case of drawn fibers is shown in Figure 8. Fibers are pulled out from the bond one by one during disintegration. A similar kind of disintegration of the bonds occurred in the case of as-spun fibers at low bonding temperatures. In the case of as-spun fibers, drop in web strength above optimum bonding temperature may be attributed to very severe thermomechanical damage to the fibers in the bond vicinity at higher temperatures. Figures 9 and 10 show SEM images of bond points of webs for as-spun-1 and drawn-1 fibers, respectively. It is evident from the figures that the bond is not well formed and there is less polymer flow and fibrillation of the fibers in bonded regions of the web in the case of drawn fibers. Insufficient polymer-flow and fibrillation of the fibers appear to be the main factors responsible for the weak and brittle nature of the bonds in the case of drawn fibers. No fibrillation was observed in the case of as-spun fibers. Fibrillation of the

Figure 9 SEM IMAGE OF A BOND FOR AS-SPUN 1 FIBERS

Figure 10 SEM IMAGE OF A BOND FOR DRAWN 1 FIBERSfibers is further clear from the SEM image in Figure 11. In the case of drawn fibers, polymer flow could be improved by increase in bonding temperature. However, web failure occurred due to rupture of the bonds even at higher bonding temperatures. A good correlation was observed between the bondability of the fibers and the TMA failure temperature of the fibers. The higher the TMA failure temperature, the higher the temperature required to obtain a good bond.

Figure 11 SEM IMAGE OF A BOND FOR DRAWN 1 FIBERS AT HIGHER MAGNIFICATION (500X)

Table 3 RESULTS OF SINGLE-BOND STRIP TENSILE TEST Sample Id Breaking Nature Of Load (G) Bond Failure As-spun 1 260 No failure As-spun 2 212 No failure As-spun 3 154 No failure Drawn 1 96 Semi-ductile Drawn 2 74 Brittle Drawn 3 73 Very brittle

16 INJ Fall 2002

orientation, crystallinity, crystallite size and other morphological aspects. Fiber diameter was within the desired range for all the three sets. As in the case of staple fiber studies, fiber diameter was intentionally kept the same so that the differences due to change in diameter could be minimized and the role of fiber micromorphology in thermal bonding could be analyzed. Set I fibers had the most developed morphology followed by Set II and Set III, respectively. Diffused peaks in WAXD patterns of Set III fibers indicate the significant presence of smectic phase in Set III fibers. Formation of smectic phase is favored at higher melt temperature [23], as was the case for Set III. This is probably due to the fact that higher melt temperatures lead to lower stress in the spinline. This allows greater supercooling to occur before crystallization begins. When this temperature drops below about 700C, smectic phase rather than a-phase is formed [19]. Fiber Figure 12 birefringence and breaking extension of spunWAXD PATTERNS OF SPUNBOND FIBERS bond fibers did not go hand in hand. The difA single-bond strip tensile test was done in order to esti- ferences in phase structure may be responsible for lower mate the bond strength and the degree of load sharing breaking extension of Set III fibers, in spite of their lower between the fibers. The results of this test are shown in Table birefringence. 3. In this test also, no failure of the bonds was observed in the Differences in the web properties for different sets were case of as-spun fibers. Whereas, in the case of drawn fibers, marginal in the case of spunbond webs owing to small differbond failure was observed at breaking loads much less than ences in their fiber properties. Tensile strength and breaking that in the case of as-spun fibers. Therefore, it may be con- extension of the spunbond webs from different sets of the cluded that bonds were much stronger in the case of as-spun fibers bonded over a wide range of bonding temperature are fibers as compared to drawn fibers. The bonds became more shown in Figures 13 and 14, respectively. Optimum bonding brittle and weak with increase in draw ratio of the fibers. temperature was the lowest for Set III fibers followed by Set Difference in breaking loads between as-spun fibers, as there II and Set I, respectively. Better bondability of Set III fibers was no failure of bonds, was attributed to the difference in the may be due to their smaller crystal size, paracrystalline strucdegree of load sharing between the fibers. The degree of load ture and less molecular orientation, which provide better sharing between the fibers was directly related to breaking polymer flow at lower temperatures. A good correlation was extension of the fibers. The higher the breaking extension, the observed between the TMA failure temperature and the optimum bonding temperature of the fibers. Fibers with lower higher the degree of load sharing. TMA failure temperature, such as Set III, had lower optimum bonding temperature than the fibers with higher TMA failure Spunbond Studies The morphological characteristics and mechanical proper- temperature, such as Set I. A similar kind of correlation ties of spunbond fibers for the three sets are listed in Table 4. between the TMA failure temperature and the bonding temWAXD photographs are shown in Figure 12. The results perature has been reported by Zhang et al. [20]. Improved show that the three sets differed in terms of their molecular- bondability of the fibers from Set I to Set III could also be

Table 4 STRUCTURE AND PROPERTIES OF SPUNBOND FIBERS Sample Id Set I Set II Set III Diameter (M) 19.3 19.3 18.8 Birefringence Crystallinity (%) Crystal Size (X x 10-3) % (Ao) 21.8 45.4 110 21.2 46.5 50 18.8 47.3 35 Tenacity (G/Denier) 3.1 2.7 2.4 Elongation (%) 300 280 225

INJ Fall 2002 17

Two competing factors in this case may be speculated to be the bondability and the mechanical properties of the fibers. At lower bonding temperatures, bondability of the fibers seemed to dominate the web properties, and at higher bonding temperatures, mechanical properties of the fibers were dominant. In general, bonding behavior of spunbond fibers was similar to that of asspun staple fibers. Morphological Changes During Thermal Bonding Morphological changes in the fibers were studied at medium bonding temperature, which was 145C in the case of staple fiber studies, and 135C in the case of spunbond studies. Noteworthy changes in fiber structure were observed in both the cases. The effects were less prominent in the case of spunbond studies as compared to staple fiber studies due to relatively shorter residence time in spunbonding. The changes in molecular orientation of the fibers during the thermal bonding process are shown in Table 5. Birefringence of the fibers increased as a result of annealing under constrained length during calendering. Increase was more for the fibers with comparatively less developed morphology before bonding. The changes in crystallinity of the fibers during thermal bonding are shown in Table 6. A significant increase in crystallinty was observed from virgin fibers in bonded as well as unbonded regions of the web, in the case of staple fiber studies. Such a substantial increase may be due to much higher residence time in the case of staple fiber studies, which allowed sufficient recrystallization to occur. No significant changes in crystallinity were observed in spunbonding. However, crystal size increased during thermal bonding in both staple fiber as well as spunbond studies, as shown in Table 7. Here it needs to be noted that crystal size data for smectic phase are only reasonable approximations. Increase in crystal size was even more prominant for spunbond fibers. Crystals in the case of spunbond fibers grew bigger and fewer. Such a rearrangement of crystalline structure in spunbond fibers was also indicated by WAXD equatorial scans shown in Figure 16. Change in location and width of reflection peaks from virgin fibers to bonded and unbonded regions of the web suggested transformation of smectic phase to the more stable a-monoclinic phase during the thermal bonding process. Conclusions Fiber morphology plays a very important role in determining the optimum bonding conditions and the mechanical properties of the web. Fibers with relatively less developed morphology yielded stronger and tougher webs as compared to fibers with more developed morphology. The fibers with high molecular orientation and crystallinity tended to form a weak and brittle bond mainly due to

Figure 13 TENSILE STRENGTH VS. BONDING TEMPERATURE FOR SPUNBOND WEBS

Figure 14 BREAKING EXTENSION VS. BONDING TEMPERATURE FOR SPUNBOND WEBS

Figure 15 MAXIMUM FIBER STRENGTH REALIZATION FOR SPUNBOND FIBERSseen in terms of increase in fiber to web strength realization from Set I to Set III, as shown in Figure 15. However, as can be seen from Figures 13 and 14, the trend in web properties for different sets reversed from lower to higher temperature.18 INJ Fall 2002

lack of polymer flow and the presence of fibrillation of the fibers in the bonded regions. Fiber breaking extension was found to be equally important, if not more, as fiber strength, in governing the web properties. Higher breaking extension of the fibers leads to a greater degree of load sharing between the fibers during deformation, thus improving the mechanical properties of the web. Fibers with less developed morphology showed lower optimum bonding temperature. A good correlation was observed between the thermomechanical stability of the fibers as measured by TMA and the bondability of those fibers. Optimizing the bonding temperature did not help much in improving the web properties in the case of highly drawn fibers, i.e. fibers with very high molecular orientation and crystallinity. In general, findings with spunbond studies are also similar to that in staple fibers. In addition, it was observed that crystalline structure and crystal size do affect thermomechanical stability and, thus, bondability of the fibers. Less perfect and less stable structure, such as smectic phase with smaller crys* In bonded regions, molecular orientation was estimated in terms of tals in the case of Set III, led to lower thermomechanical stachange in bond- dimensions when heated up to 160 C. bility and, thus, better bondability of the fibers. In general, bonding behavior of spunbond fibers was found similar to that of as-spun staple fibers. It was observed that fibers do undergo Table 6 some structural changes in bonded CHANGE IN CRYSTALLINITY (%) DURING THERMAL BONDING. as well as unbonded regions of the Sample Id Crystallinity Crystallinity Crystallinity web during the thermal bonding Of Virgin In Unbonded In Bonded Region process. The extent of change in Fibers (%) Region (%) (%) fiber structure would depend upon As-spun 1 36.7 41.9 50.1 the structure of original fibers and As-spun 2 41.3 47.8 55.1 the process variables employed.

Table 5 CHANGE IN MOLECULAR ORIENTATION DURING THERMAL BONDING Sample Id Birefringence Birefringence Of Virgin Fibers In Unbonded 3 Region (X x 103) (X x 10 ) As-spun 1 19.0 23.3 As-spun 2 20.4 23.4 As-spun 3 17.8 25.0 Drawn 1 23.8 26.6 Drawn 2 29.4 29.6 Drawn 3 31.4 30.6 Set I 21.8 21.6 Set II 21.2 22.3 Set III 18.8 22.4o

As-spun 3 Drawn 1 Drawn 2 Drawn 3 Set I Set II Set III

45.0 48.9 53.7 56.4 45.4 46.5 47.3

48.3 52.6 54.2 56.9 45.0 44.8 45.8

58.8 53.5 54.3 56.1 48.6 46.3 47.1

Table 7 CHANGE IN CRYSTALLIZE DURING THERMAL BONDING. Sample Id Crystal Size Crystal Size Crystal Size For Virgin For Unbonded For Bonded Region Region (A) (A) Fibers (A) As-spun 1 140 160 185 As-spun 2 185 215 245 As-spun 3 150 170 180 Drawn 1 140 165 190 Drawn 2 155 160 170 Drawn 3 135 145 160 Set I 110 145 170 Set II 50 130 145 Set III 35 90 160

Acknowledgements This project was funded from Nonwovens Cooperative Research Center, NCSU, Raleigh, NC. Authors would like to thank Montell USA Inc. and ExxonMobil Corp. for providing the polymers. Support from TANDEC for providing the Spunbond equipment time is also appreciated. References 1. Reed R., U.S. Patent 2277049, assigned to Kendall Company, 1942. 2. Dharmadhikary R. K., Gilmore T. F., Davis H. A. and Batra S. K. Thermal Bonding of Nonwoven fabrics, Textile Progress, 1995(26), No. 2, pp. 1-37. 3. Warner S. B. Thermal Bonding of Polypropylene Fibers, Text. Res. J., 1989(59), pp. 151-159. 4. Kwok W. K., Crane J. P., Gorrafa A. and Iyengar Y. Polyester Staple fibers forINJ Fall 2002 19

Thermally Bonded Nonwovens, Nonwovens Industry, June 1988, pp. 30-33. 5. Gibson P. E. and McGill R. L. Thermally Bondable Polyester Fiber: the Effect of Calender Temperature, TAPPI J., 1987, No. 12, pp. 82-86. 6. Drelich A. Thermal Bonding with Fusible Fibers, Nonwovens Industry, Sept 1985, pp. 12-26. 7. Muller D. H. How to Improve the Thermal Bonding of Heavy Webs, INDA J. Nonwovens Res., 1989(1), No. 1, pp. 35-43. 8. De Angelis V., DiGiaoacchino T. and Olivieri P. Hot Calendered Polypropylene Nonwoven fabrics, Proceedings of 2nd International Conference on Polypropylene Fibers and Textiles, Plastics, and Rubber Institute, University of York, England, 1979, pp. 52.1-52.13. 9. Bechter D., Kurz G., Maag E. and Schutz J. Thermal Bonding of Nonwovens, Textil-Praxis, 1991 (46), pp. 12361240. 10. Malkan S. R., Wadsworth L. C. and Devis C. Parametric Studies of the Reicofil Spunbonding Process, Third TANDEC Conference, Knoxville, 1993. 11. Malkan S. R., Wadsworth L. C. and Devis C. Parametric studies of the Reicofil Spunbonding Process, International Nonwovens Journal, 1992, No.2, pp. 42-70. 12. Wei K. Y., Vigo T. L. and Goswami B. C. StructureProperty Relationships of Thermally bonded Polypropylene Nonwovens, J. Appl. Polym. Sci., 1985(30), No.4, pp. 15231534. 13. Phillipp P. Thermal Bonding with Copolyetster Melt Adhesive Fibers, Nonwovens World, Nov 1986, pp. 81-85. 14. Beyreuther R. and Malcomess H. J. Spunbonded Nonwovens-Linking Innovative Polymer, Technological and Textile Research, Melliand Textilberichte, 1993(74), No. 4, pp. E133-135. 15. Winchester S. C. and Whitwell J. C. Studies of Nonwovens-I: A Multivariable Approach, Text. Res. J., 1970(40), No.5, pp. 458-471. 16. Bechter D., Roth A., Schaut G., Ceballos R., Kleinmann K. and Schafer K. Thermal Bonding of Nonwovens, Melliand Textilberichte, 1997, No. 3, pp. E39-40. 17. Wyatt N. E. and Goswami B. C. StructureProperty Relationships in Thermally Bonded Nonwoven Fabrics, J. Coated Fabrics, 1984(14), pp. 100-123. 18. Zhang D., Ph.D. Dissertation, The University of Tennessee, Knoxville, December 1995. 19. Lu, F. M. and Spruiell, J. E., J. Appl. Polym. Sci., 34, 1541 (1987). 20. Dong Zhang, G. S. Bhat, Sanjiv Malkan and Larry Wadsworth, Evolution Of Structure And Properties In A Spunbonding Process, Textile Research Journal, 68(1), 2735 (1998). 21. Cullity B. D., Elements of X-ray Diffraction, Addison-Wesley Publishing Company Inc., Massachusetts, 1978, p. 284. 22. Storer R. A., ASTM, Easton, MD, USA, 1986.

23. Ahmed M., Polypropylene Fibers, Science and Technology, Elsevier Science Publishing Company, New York, 1982, p. 194. INJ

20 INJ Fall 2002

ORIGINAL PAPER/PEER-REVIEWED

Fiber Motion Near The Collector During Melt Blowing: Part 2 Fly FormationBy Randall R. Bresee, The University of Tennessee, Knoxville, Tennessee USA and Uzair A. Qureshi, Jentex Corporation, Buford, Georgia USAAbstract On-line and off-line measurements were obtained to gain an understanding of fly production during multi-hole melt blowing at commercial speed. These measurements allowed us to describe the effects of common processing parameters on fly production and develop a model for fly formation that begins to account for experimental measurements. Introduction In a previous paper [1], we reported results of experiments conducted to obtain a general understanding of fiber motions near the collector of the basic multi-hole melt blowing (MB) process operating at commercial speed. In the current paper, we address the problem of fly formation. Fly particles are fibers that have been broken and released from the fiber stream during MB. The phenomenon of fly formation has practical importance to web producers and knowledge of fly formation is important for understanding the MB process. Fly is undesirable and its formation is sometimes used to identify a processing limit during commercial MB. That is, preliminary processing conditions are determined, primary air pressure is increased until fly is produced and then air pressure is decreased until little fly is produced. In this paper, we will report numerous experimental measurements related to fly formation during multi-hole MB operating at commercial speed. Measurements include fly particle mass, fly particle length, total fiber length in fly particles, fiber bundle size in webs, air speed in the direction normal to the collector surface, air speed in the direction of collector motion and the direction of fiber flow near the collector. While obtaining these measurements, we varied primary air pressure, die-to-collector distance (DCD), collector speed and collector vacuum. These measurements were used to formulate a conceptual model of fly production based on aerodynamic drag and fiber entanglement. Experimental Procedures We processed PP-3546G polypropylene resin (1259 MFR) supplied by ExxonMobil Chemical Company on three different multi-hole MB lines in TANDEC at the University of Tennessee. These were a 180-hole (15 cm) horizontal line having a 47 cm diameter rotating drum collector, an Accurate Products 600-hole (51 cm) horizontal line having a 55 cm diameter rotating drum collector, and a Reifenhauser 601hole (61 cm) vertical bicomponent fiber line having a flat endless belt collector. Commercial speed processing conditions generally were used. A high-speed camera and pulsed laser were used to acquire images of fibers on-line. Procedures used to obtain fiber velocity from these images have been reported previously [2]. Air speed measurements were obtained using processing conditions similar to those used for fiber measurements but with no resin throughput. Air speed was measured on-line using a Pitot tube and anemometer. Fiber bundle size in webs was measured off-line using WebPro [3]. Fly particles were captured during MB using wire screens and analyzed off-line . Results and Discussion Figure 1 provides optical images of fly particles collected while processing polypropylene with a die temperature of 232O C, air temperature of 243O C, resin throughput rate of 0.42 ghm, primary air pressure of 2.5 psi and DCDs of 76, 30 and 15 cm using a 55 cm rotating drum collector. This figure qualitatively shows that the size of fly particles varied over a large range. Figure 1 also shows that fly particles produced with a particular set of processing conditions exhibited similar sizes. Finally, Figure 1 shows that DCD significantly influenced the size of fly particles. To obtain quantitative information about fly, we collected fly particles while varying processing conditions and measured the mass and length of individual particles and theINJ Fall 2002 21

diameter of fibers in particles. From this data, we computed the total length of fiber contained in individual fly particles. Measurements for individual particles collected with each processing condition were Figure 1 averaged and are summaFLY PARTICLES COLLECTED WITH 76 CM (LEFT), 30 CM (MIDDLE) AND 15 rized in Figure 2. CM (RIGHT) DCD; EACH IMAGE AREA = 9.0 CM X 6.7 CM (BAR = 3.0 CM) Figure 2 shows that primary air pressure, DCD Figure 2 and collector speed influenced the structure of fly. Increasing EXPERIMENTAL FLY DATA FOR VARIOUS primary air pressure 20% increased particle mass, particle PROCESSING CONDITIONS length and total fiber length in particles, although the increases were relatively small. Increasing DCD reduced particle mass, particle length and total fiber length in particles. Increasing collector speed increased particle mass, particle length and total fiber length in particles. We are aware of no phenomenological model for fly formation in the published literature. In the following pages, we will propose a basic model for fly formation based on aerodynamic drag and fiber entanglement and will show that this model begins to account for the experimental data in Figure 2. Mechanism of Fly Formation We believe that fly formation is controlled primarily by aerodynamic drag and fiber entanglement. That is, fly particles are released when (i) a drag force exists that is strong enough to break fibers and (ii) fiber entanglement is insufficient to retain broken fibers within the forming web. Drag Force Fibers must be broken to release fly particles from the fiber stream during MB. We previously showed that only two regions of the basic MB process are likely to produce a large drag force on fibers [1]. These regions are located near the die and near the collector where differences between air and fiber speeds are large. Consequently, these two regions are most favorable for producing fly whereas most of the region between the die and collector is less favorable for fly production because drag forces are smaller. Figure 2 showed that fly production is greatly influenced by two collector parameters - DCD and collector speed. Figure 2 also showed that individual fly particles contained as much as 150 m of fiber length. These observations suggest that fly is most likely released near the collector rather than near the die. Consequently, we will focus our discussion on fly formation near the collector although we recognize the possibility that fly also may be produced near the die. In a previous discussion of the basic MB process, we remarked that aerodynamic drag forces acting on fibers suddenly increase near the collector since fiber speed decreases to zero during laydown but air continues to flow at relatively high speed [1]. Recognizing this phenomenon allows us to

22 INJ Fall 2002

Measurement Region

Figure 3 MEASUREMENT REGION NEAR A FLAT BELT COLLECTOR

Figure 5 AIR SPEED IN THE DIRECTION NORMAL TO THE COLLECTORslow fibers more rapidly as they traveled closer to the collector surface. This conclusion is consistent with fiber speed measurements that showed fiber speed decreased as far as 9 cm from the collector but decreased more rapidly within 3 cm of the collector [1]. In contrast to air traveling near the centerline, air 7.5-15 cm from the centerline traveled faster at locations closer to the collector surface. Faster moving air would be expected to increase the speed of some fibers approaching the collector in this region. This may seem to contradict the general concept that fiber speed must decrease to zero during laydown. However, we need to recognize that fibers near the collector of a commercial MB process are entangled with numerous other fibers to form an extensive network. Fibers near the airflow centerline that slow as they approach the collector help slow fibers traveling far from the airflow centerline. It is important to note, however, that Figure 5 provides evidence that a drag force exists far from the airflow centerline that may accelerate and break fibers. This suggests that fly is most likely produced in laydown regions far from the airflow centerline rather than laydown regions near the centerline. The interior of MB webs generally result from fiber laydown in the vicinity of the airflow centerline whereas laydown far from the centerline produces the collector-side and die-side of webs. Figure 5 provides evidence that aerodynamic drag may reduce the speed of fibers forming the web interior at a different rate than fibers forming the collectorside and die-side of webs. This leads us to expect that the interior of a MB web may exhibit a slightly different structure than the collector-side and die-side of the web. However, experimental measurements of web structure that could test this hypothesis have not been reported. Figure 2 provided experimental evidence that fly formation was influenced by collector speeds of 10-35 m/min. To learn more about this, we acquired air speed measurements similar to those of Figure 5 but using three collector belt speeds (0, 21 and 61 m/min) at each measurement location. These measurements are provided in Figure 6. This figure clearly showsINJ Fall 2002 23

-15.0 -7.5

0

7.5

15.0

Figure 4 MEASUREMENT LOCATIONS NEAR THE COLLECTORqualitatively explain experimental observations in Figure 2 that show fly formation apparently was reduced when primary air pressure was decreased or DCD was increased. That is, fiber speed decreases to zero during laydown for any processing condition so the aerodynamic drag force available to break fibers near the collector is determined mostly by the speed of air in the laydown region of the collector. Decreasing primary air pressure at the die or increasing DCD reduces the drag force near the collector since the speed of air arriving at the collector is reduced. Consequently, we expect less fiber breakage to occur and less fly to be produced when primary air pressure is decreased or DCD is increased. To learn more about drag force near the collector, we measured the distribution of airflow over a collector surface. The speed of air traveling in the direction normal to a flat collector belt was measured near the airflow centerline as well as plus and minus 7.5 and plus and minus 15.0 cm from the centerline and 1.5, 4.0, 6.6, 9.1 and 11.6 cm from the collector surface. The general measurement region is identified schematically in Figure 3 and specific measurement locations are denoted by vertical arrows in Figure 4. Figure 5 provides air speed measurements in the direction normal to the collector surface. Near the airflow centerline, air speed decreased as the collector surface was approached. Slowing was observed as far as 11.6 cm from the collector although air slowed more rapidly as it traveled closer to the collector. This effect would be expected to slow fibers near the airflow centerline as far as 11.6 cm from the collector and

Figure 6 AIR SPEED IN THE DIRECTION NORMAL TO THE COLLECTOR FOR THREE COLLECTOR BELT SPEEDS (SEE FIG. 5 LEGEND FOR DISTANCES FROM COLLECTOR SURFACE)that collector belt speed had little influence on the speed of air traveling in the direction normal to the collector belt at distances as close as 1.5 cm from the belt surface. We also evaluated the influence of collector belt speed on the speed of air traveling parallel to the direction of belt movement at various distances from the collector surface. Horizontal arrows in Figure 4 denote our specific measurement locations. Measurements were recorded only at the airflow centerline and 15 cm from the centerline to save time. Figure 7 provides measurements obtained at the airflow centerline whereas Figure 8 provides measurements obtained 15 cm from the centerline. Figures 7-8 show that collector belt speed had little influence on the speed of air traveling in the direction of belt movement at distances as close as 1.5 cm to the belt surface. Overall, Figures 6-8 lead us to conclude that the influence of collector speed on fly formation reported in Figure 2 did not occur as a result of collector motion affecting air speed. Figures 7-8 also show that air flowing in the direction of collector motion traveled fastest at locations far (15 cm) from the airflow centerline. This implies that some fibers may be swept during laydown toward the direction of belt movement by large drag forces. Since belt motion proceeds in the MD, Figures 7-8 support our previous claim [4] that fiber orientation is markedly changed during laydown from CD to MD. In addition, fast moving air in the MD would be expected to increase the speed of some fibers which, in turn, increases the probability of fiber breakage and fly formation. Next, we attempted to learn more about the influence of a vacuum applied to the collector laydown area on fly formation. To help understand this, we acquired air speed measurements that were similar to Figure 5 but while using a vacuum and combined these measurements to produce Figure 9. This figure shows that a vacuum applied to the collector significantly influenced the speed of air traveling in the direction normal to the collector belt. The vacuum influenced air speed24 INJ Fall 2002

Figure 7 AIR SPEED IN THE DIRECTION OF COLLECTOR BELT MOVEMENT AT THE AIRFLOW CENTERLINE

Figure 8 AIR SPEED IN THE DIRECTION OF COLLECTOR BELT MOVEMENT 15 CM FROM THE AIRFLOW CENTERLINE Figure 9 AIR SPEED MEASUREMENTS NORMAL TO THE COLLECTOR BELT

Figure 10 AIR SPEED IN THE DIRECTION OF COLLECTOR BELT MOVEMENT AT THE AIRFLOW CENTERLINE (DIAMOND, SQUARE AND TRIANGLE DENOTE 1.5, 4.0 AND 6.6 CM FROM COLLECTOR SURFACE, RESPECTIVELY)as far as 6.6 cm from the collector surface, although air traveling closer to the collector was influenced more. It is important to note that the vacuum increased air speed near the airflow centerline but decreased air speed in areas far (5-15 cm) from the airflow centerline. Since practical MB experience has demonstrated that fly is reduced when a vacuum is applied to the collector, Figure 9 suggests that fly is most likely released from regions located far from the airflow centerline and near the collector surface (where air speed was reduced most by the vacuum). That is, the vacuum ought to reduce aerodynamic drag and thus fiber breakage most significantly far from the airflow centerline and near the collector surface. Figure 9 also suggests that fiber laydown with a vacuum is different than laydown using the same MB equipment but without a vacuum since the distribution of air speeds in the laydown area are different. For example, Figure 9 shows that the vacuum increased air speed near the airflow centerline where fiber laydown forms the interior of webs but decreased air speed far from the centerline where web surfaces are formed. This may produce a slightly different fiber diameter distribution in the web than a web produced with the same average air speed at the collector but without using a vacuum. However, experimental measurements of web structure that could verify this expectation have not been reported. Next, we compared the speed of air traveling in the direction of belt motion with and without a vacuum. Figure 10 provides measurements at the airflow centerline and Figure 11 provides measurements 15 cm from the centerline. These figures show that the vacuum significantly reduced the speed of air traveling in the direction of belt movement. The greatest speed reduction occurred far (15 cm) from the airflow centerline and close to the collector surface. Since practical MB experience has demonstrated that fly is reduced when a

Figure 11 AIR SPEED IN THE DIRECTION OF COLLECTOR BELT MOVEMENT 15 CM FROM THE AIRFLOW CENTERLINE (DIAMOND, SQUARE AND TRIANGLE DENOTE 1.5, 4.0 AND 6.6 CM FROM COLLECTOR SURFACE, RESPECTIVELY)vacuum is applied to the collector, Figures 10-11 suggest that fly is most likely produced far from the airflow centerline and near the collector surface. Overall, drag force considerations resulting from measurements of air speed in the direction normal to the collector (Figure 5) and in the direction of collector motion (Figures 78) indicate that fly is most likely produced in collector regions located far from the airflow centerline and near the collector surface. Drag force considerations associated with the influence of a vacuum on air speed in the direction normal to the collector surface (Figure 9) and in the direction of collector motion (Figures 10-11) also support this conclusion . Fiber Entanglement Fibers in commercially produced MB webs are entangled with numerous other fibers so we now consider the role that fiber entanglement may play in fly formation. We have previously discussed why fiber entanglement occurs during MB [2,5], the influence of entanglement on fiber speed uniformity during MB [1], the influence of entanglement on fiber orientation in MB webs [1] and how DCD affects fiber entanglement [1]. We believe that entanglement plays two general roles in fly formation. First, entanglement causes the mechanical load on a fiber to be shared with other fibers so fiber breakage and thus fly formation ought to be reduced when entanglement increases. Second, entanglement inhibits the release of fibers that are already broken and thus ought to reduce fly formation. Since fiber entanglement increases when DCD increases [1], it seems likely that both increased fiber entanglement and decreased aerodynamic drag reduce fly formation when DCD is increased. Figure 2 clearly showed that fly formation was influenced by collector speed. We now consid