TH ESIS (1‘35“?)1 TT ”1|ijlili’fllflifllfiflflifliifi”!!! IIH 1 19022 This is to certify that the thesis entitled CHANGES IN FIBER 8-DING CHARACTERISTICS AS A GARMENT AGES presented by Anne H. Colyer has been accepted towards fulfillment of the requirements for .__H._S.__degree in Mstice Maj r professor Date /L/L} /q7 0-7 639 MS U is an Affirmative Action/Equal Opportunity Institution LIBRARY Michigan State University PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE MA" 0 3 20m SEP 1 9 2002 ‘032404 11/00 cJCIRCJDuaDmpGS-p.“ Li 1,7 . .\I|\\'lii"|.\. III? Ni l?l CHANGES IN FIBER SHEDDING CHARACTERISTICS As A GARMENT AGES By Anne M. Colyer A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCES School of Criminal Justice 1997 ABSTRACT CHANGES IN FIBER SHEDDING CHARACTERISTICS As A GARMENT AGES By Anne M. Colyer Forensic scientists analyze trace evidence every day. Fibers are one of the most commonly discovered types of trace evidence found at crime scenes. A great deal of research has been done in the area of forensic fiber analysis; most of it focused on instrumental methods of fiber analysis. Significantly less research has been done on fiber transfer and shedding characteristics of textile types. The few articles that examine fiber shedding characteristics make no attempt to control for the age of the garment. Age- related wear may affect the number of fibers transferred in a contact. This study simulates the aging process by repeatedly washing several garments and then testing the number of fibers shed after a simulated transfer incident. The results Show that most garments shed less through time, though initial peaks in shedding may or may not be present, depending upon the composition of the donor fabric. To Bryan, for his faith in me and support of all my endeavors ACKNOWLEDGEMENTS So many people have contributed to the successful completion of this thesis and I am indebted to them all. First and foremost, I wish to thank Dr. Jay Siegel and the Midwestern Association of Forensic Sciences for the grant that supported this research. Without this help, I’d still be at the drawing board trying to find a thesis project! Dr. Siegel has been an excellent instructor, mentor and friend throughout my graduate education and I am grateful for his patience with me. I must also thank Dr. Christina DeJong and Dr. Christopher Smith for their contributions to my committee. It is too easy, when writing such a manuscript, to lose sight of how necessary it is to clearly explain background and methods to your audience. Their comments on the draft and during my defense helped me to clarify the document. Lee Brun-Conti was both a valuable committee member and sounding board for me during the infancy of this project. Both she and Amy Michaud helped me to clarify the focus of my research and simplify the methods section. Lee’s contributions to the defense were also excellent and helpful in finishing the final draft. Finally, I must acknowledge the personal support given me by my friends and family during this program. Irena, Jen, Heather, Amy, Bill, Liz, Mom, Dad, Lisa and, of course, Bryan, have all been wonderful cheerleaders for me. They helped me keep motivated to work on the drafts and stay with the program and I am deeply in their debt. All of my hard work has paid off and I hope they will enjoy celebrating the end of this long road at MSU with me! it TABLE OF CONTENTS LIST OF TABLES .................................................................................. vii CHAPTER 1 INTRODUCTION .................................................................................... 1 The Value of Fiber Evidence in Forensic Science ....................................... 1 Locard’s Exchange Principle and Fibers ................................................. 2 Previous Fibers Research .................................................................... 3 Relevance and Applications of this Study ................................................ 4 CHAPTER 2 LITERATURE REVIEW AND BACKGROUND OF STUDY ................................ 5 Occurrence of Fibers as Evidence at Crime Scenes ...................................... 5 Types of Analyses Performed on Fiber Evidence ........................................ 5 Fiber Transfer Research ..................................................................... 7 Fiber Persistence Research .................................................................. 8 Fiber Shedding Research ................................................................... 9 Garment Laundering Research ............................................................. 9 Hypothesis of Current Study ............................................................. 10 CHAPTER 3 METHODS AND RESULTS OF ANALYSIS ................................................. 11 Practical Considerations of the Study ................................................... 11 Garments Chosen for the Study .......................................................... 11 Simulated Contact Apparatus Utilized ................................................... 11 Simulated Aging of a Garment Through Repeated Washing ........................ 12 Data for Garment 1 ........................................................................ 13 Data for Garment 2 ......................................................................... 13 Data for Garment 3 ......................................................................... 14 Data for Garment 4 .......................................................................... 14 Data for Garment 5 .......................................................................... 14 Data for Garment 6 ......................................................................... 14 Data for Garment 7 ......................................................................... 15 Data for Garment 8 ......................................................................... 15 Data for Garment 9 ........................................................................ 15 Data for Garment 10 ....................................................................... 16 Data for Garment 11 ....................................................................... 16 Data for Garment 12 ....................................................................... 16 Data for Garment 13 (Recipient Fabric Control Garment) ............................ 17 CHAPTER 4 DISCUSSION AND CONCLUSIONS .......................................................... 19 Trends Noted in Fiber Shedding Characteristics ....................................... 19 The Effects of Fiber Composition on Shedding ........................................ 20 An Investigation of Shedding Mechanisms ............................................. 21 Consideration of Natural Versus Simulated Aging of a Garment ................... 22 Applications of the Study. Methodology ............................................... 23 Applications of the Study: Toward an Index of Shedding ........................... 24 Suggestions for Future Research .......................................................... 24 Summary .................................................................................... 25 APPENDIX A GARMENT DESCRIPTIONS .................................................................... 27 APPENDIX B GRAPH OF GARMENT 4 SHEDDING RESULTS ........................................... 29 GRAPH OF GARMENT 5 SHEDDING RESULTS ........................................... 30 GRAPH OF GARMENT 6 SHEDDING RESULTS ........................................... 31 GRAPH OF GARMENT 7 SHEDDIN G RESULTS ........................................... 32 GRAPH OF GARMENT 8 SHEDDING RESULTS ........................................... 33 GRAPH OF GARMENT 9 SHEDDING RESULTS ........................................... 34 GRAPH OF GARMENT 10 SHEDDING RESULTS ......................................... 35 GRAPH OF GARMENT 11 SHEDDING RESULTS ......................................... 36 GRAPH OF GARMENT 12 SHEDDING RESULTS ......................................... 37 GRAPH OF GARMENT 13, FELT A SHEDDING RESULTS .............................. 38 GRAPH OF GARMENT 13, FELT B SHEDDING RESULTS .............................. 39 GRAPH OF GARMENT 13, FELT A AND B SHEDDING RESULTS .................... 40 BIBLIOGRAPHY .................................................................................. 41 vi LIST OF TABLES Table l—Fiber Count Data ......................................................................... 18 vii Chapter 1 INTRODUCTION This study addresses the question of whether or not the age of a garment will affect its shedding characteristics. Several garments are repeatedly washed, dried and subjected to a simulated contact in an attempt to elucidate trends based upon their fiber composition. Several trends are noted and conclusions are made about the mechanisms of shedding responsible for these trends. Finally, information about how the simulated aging process affects shedding is given. The Value of Fiber Evidence in Forensic Science Fibers are the quintessential class evidence. It is not possible to individualize them to a particular source to the exclusion of all other possible sources except in the extremely rare case of a tear match. The main forensic use of fiber evidence is to associate a suspect to a crime scene or to a victim through the characterization of fibers from clothing, bedding, vehicle interiors or carpeting (Gaudette 1988). It is possible to perform some rough statistical analyses of the commonality (or rarity) of certain fibers and the relative odds of finding a particular fiber type at a crime scene by chance (Cook and Wilson 1986; Cook et al. 1993; Deadman 1984; Jackson and Cook 1986). Such evidence can often be highly significant if numerous matches based on several fiber types can Show that a suspect most likely had contact with a victim or crime scene. Fibers are traditionally classified by subdividing them according to their type (Grieve 1994). The first consideration is whether the fiber is naturally-occurring or synthetic. Natural fibers are further differentiated by their source: animal, vegetable or mineral. Synthetic fibers can be separated into groups of synthetic polymers, natural polymers and “other” fibers. Forensic fiber examiners take advantage of the varying chemical and physical properties of these various fiber types to yield information about associating fibers with crime scenes. Once it is determined that fiber evidence is relevant to the case, fiber analysis begins. First, fibers are removed from garments by any of several methods; brushing, shaking, vacuuming or the use of adhesive tape (Grieve 1994). The latter technique, applying adhesive tape to a piece of evidence, lifting the loose fibers and labeling the taping (or mounting the harvested fibers on a microscope slide), is the most popular method of fiber retrieval. The fibers are characterized according to their origin, subtype, color, shade, indications of the type of textile material of origin and the manufacturer of the textile, where possible (Gaudette 1988). Once these data are known about fibers, the appropriate kind of further analysis can be undertaken. Locard’s Exchange Principle and Fibers Edmond Locard (1877-1966), one of the founders of modern forensic science, theorized that a cross-transfer of evidence would occur in every encounter between people and/or objects. In the case of a crime, the perpetrator will leave some form of physical evidence on the victim and the victim’s presence will be similarly recorded on the perpetrator. Locard’s Exchange Principle, as it has come to be known, is the premise upon which the transfer of many types of evidence (especially fibers) rely. Fiber evidence can be easily lost or transferred to other surfaces. It is important to begin evidence collection as soon as possible after a crime is committed. Studies show that soon after fibers are transferred, they are lost or re-distributed, thus eliminating several pieces of potentially important evidence (Grieve et al. 1989; Grieve 1990; Jackson and Lowrie 1994; Robertson 1992; Siegel 1996). The transfer of fiber evidence is affected by several factors: characteristics of the donor and recipient textiles, the nature of contact (prolonged, forceful, etc.) between the two textiles, and the time interval between contact and specimen collection (Siegel 1996). Unfortunately, there have been few studies of fiber transfer characteristics and persistence rates that help to clarify these factors. Previous Fibers Research The lion’s share of the sparse literature on fiber persistence consists of anecdotal studies which cite unusual forensic cases to illustrate the application of a particular technique or instrument to fiber evidence (Deadman 1984; Church 1991; White and Tebbet 1992; Koons 1996;). Rigorous, exhaustive scientific studies are rarely undertaken in this highly applied field. Fiber transfer and persistence studies (Pounds and Smalldon 1975; Robertson et al. 1981; Parybyk and Lokan 1986; Jackson and Lowrie 1994) are beginning to be published with more regularity, but more are needed. Basic research on shedding characteristics has been undertaken also (Salter et al. 1987; Coxon et al. 1992). Some articles concerned with the effects of washing on fiber transfer capabilities have been published (Robertson and Olaniyan 1986; Bresee and Annis 1991). Relevance and Applications of this Study One aspect of fiber evidence unexplored in the current literature is the effect that aging has on a garment. Since virtually all garments are washed or dry cleaned by their owners, certainly their fiber shedding characteristics are subject to change throughout the “lifetime” of the garment. Do certain fiber types and blends show a greater propensity to shed as they age? Could this affect their abundance at a crime scene? The hypothesis to be tested by this study is that through time (simulated by repeated washings), garments will shed differentially based upon their fiber composition. This is simulated in controlled laboratory conditions with several garments undergoing mock transfers, tape lifting, washing, drying and tape lifiing. By examining a variety of garment fiber types, a realistic picture of patterns of fiber shedding through time emerges. Chapter 2 LITERATURE REVIEW AND BACKGROUND OF STUDY Occurrence of Fibers as Evidence at Crime Scenes Most fiber evidence arrives in the forensic laboratory on articles of clothing. The goal is to match fibers found on one participant’s clothing with the garments worn by the other participant in the crime and vice versa. Such cross-transfers, particularly if they are numerous and varied, can prove to be effective evidence in criminal cases as demonstrated by the Wayne Williams murder trial (Deadman 1984). Fibers may also be transferred onto other surfaces at a crime scene. They can be recovered from hair, skin, carpeting, vehicle interiors, shoe and tire treads, vehicle bumpers, as well as such unlikely places as animal burrows and nests. In short, any object within the scope of a crime scene is a possible fiber evidence recipient. Types of Analyses Performed on Fiber Evidence The types of analyses that can be performed upon fiber evidence are as many and varied as the fiber types themselves. Simple visual examination and basic physical and chemical tests can be used to rule out obvious non-matches. Analysis of the cross- sectional shape of fibers is particularly useful since many manufacturing techniques vary this aspect of fibers in order to produce some desired characteristic (Grieve 1994). Measurement of a fiber’s refractive index, both parallel and perpendicular to its long axis, provides additional information for fiber comparison. Comparison of pyrograms obtained by pyrolysis gas chromatography of fibers from a scene is commonly performed in forensic cases. Since fibers are by definition small pieces of evidence, several types of microsc0pes are useful in comparing potential donors with crime scene evidence. Features such as the diameter, shape, surface features, amount of processing, delustrant particles, color and fluorescence are examined by using a comparison microscope. Polarized light microscopy can be used to broadly categorize synthetic fibers, considering optical properties such as birefringence and the Sign of elongation of the fiber (McCrone et al. 1978). Combining a hot stage apparatus with a polarized light microscope can yield additional information about the constituents of a fiber. The changes in interference colors seen as fiber samples are slowly heated can be tracked and compared. This method can be useful in reconstructing traffic accident scenes where fiber fragments are embedded into melted plastic materials (Grieve 1992). The fiber dye or combination of dyes used on textiles can be used to differentiate fibers in a forensic case. Microspectrophotometry, using either visible or ultra-violet light can quantitatively define a fiber’s color for comparison with known samples (Grieve 1994). Thin layer chromatography is effective in separating the constituent dyes from a fiber sample. This can also be accomplished using high performance liquid chromatography or one of the newest techniques in fiber dye analysis, capillary electrophoresis. Finally, there are several techniques that take advantage of the infrared optical activity of fiber samples. F ourier-transform infrared spectroscopy can reliably determine if a detergent residue is present on a sample, if two samples contain varying amounts of the same co-polymer, and which, if any, monomers are present (Grieve 1994). Raman spectroscopy can identify differences in crystallinity and molecular orientation of individual fibers. Polarized infrared spectroscopy takes advantage of the optical properties of substituent chemical groups and their molecular orientation in the fiber (Tungol et al. 1991). Fiber Transfer Research One of the earliest, and most significant, scientific studies of fiber transfer characteristics was undertaken by Pounds and Smalldon (1975a; 1975b; 1975c). They performed a basic fiber transfer experiment and concluded that fiber shedding characteristics are dependent upon the type of donor and recipient materials, the contact pressure and the number of contact passes (1975a). Robertson and Lloyd (1983) found that fibers deposited on a particular area of a garment gradually move to other parts of the garment. They warn that these secondary transfers may mislead fiber examiners who do not consider the time since contact when analyzing a fiber transfer. Fiber diameter affects shedding characteristics. Cordiner et al. (1985) studied the transfer characteristics of wool fibers of varying diameters and found that fine, rather than coarse, wool fibers are more likely to transfer upon contact. The differential transfer rates of the constituents of blended fabrics were studied by Parybyk and Lokan (1986). They determined that the ratio of fibers found on a recipient garment cannot be used to determine the ratio of fibers used in the construction of the donor garment. Due to the differential shedding characteristics of the donor garment’s constituent fiber types, predictions based upon tapings of recipient garments may be misleading. Recently, fiber transfer research has sought to elucidate more reliable information about fiber transfer in forensic settings. To study the question of fiber contamination within a crime lab, Moore et al. (1986) studied the transfer characteristics of wool and cotton fibers from bench to bench at a crime laboratory. They concluded that the wool fibers were more likely to fall to the floor while the cotton fibers stayed airborne, allowing for more cotton fiber contamination of lab benches. Grieve et al. (1989) continued the simulations of real life transfers in their study of acrylic fiber transfers. They advised fiber examiners to be cognizant that while fibers do indeed undergo secondary transfers, there are many cases where the recipient garment or object will retain fibers from the primary transfer for extended periods of time. Lowrie and Jackson (1991) emphasize the importance of considering the effectiveness of fiber recovery efforts in making conclusions based upon fiber transfer evidence. Finally, Roux et al. (1996) studied fiber transfer among automobile seats, noting that differences in garment type, seat covers, drivers and driving time all affected transfer characteristics. Fiber Persistence Research Once again, the Pounds and Smalldon research (1975b, 1975c) on fiber persistence provided some of the earliest scientific data regarding the adhesion of foreign fibers on various recipients. They concluded that fiber persistence depends more upon the characteristics of the recipient textile than those of the donor (1975b). They concluded that mechanical interactions between the two garments more significantly affect fiber persistence rates than other factors such as electrostatic potential (1975c). Similar to the trend in fiber transfer research, forensic scientists have examined fiber persistence with the goal of providing more realistic simulations and producing beneficial studies for fiber analysts. Robertson et al. (1982) studied the persistence of acrylic, wool and polyester/viscose fibers on a large range of recipient garments. They concluded that fiber persistence was decreased by factors such as contact pressure, fiber size, deposition of the fiber on areas of the recipient subject to further contacts (such as through arm or leg movements) and simply wearing the recipient garment. The persistence of fibers in human head hair was studied by Ashcroft et a1. (1988). They found that fibers consistently remain available for collection even afier a suspect has washed his or her hair. Fiber Shedding Research The literature on fiber shedding characteristics is sparse. Cook and Wilson (1986) established the possibility of finding common fiber types on a garment by chance. This type of research is referred to as a target fiber study. Cook and Wilson determined that finding any more than a few fibers purely by chance on a garment is unlikely in a forensic context. Differential shedding of fibers from blended fabrics was examined by Salter et al. (1987), concluding (as did Parybyk and Lokan in 1986) that the ratio of fibers shed is not indicative of the ratio of fiber types used in the garment. Garment Laundering Research Only two studies of the effect of laundering on fiber shedding characteristics have been reported. Robertson and Olaniyan (1986) compared various washing methods and their effect on fiber transfers of acrylic and nylon garments. They found that none of the washing methods significantly affected either textile type’s propensity to shed fibers in simulated contacts. Bresee and Annis (1991) concluded that the use of fabric softener significantly increased the fiber shedding characteristics of most garments. 10 Hypothesis of Current Study None of these studies has addressed the issue of whether or not the age of a garment affects its propensity to shed fibers in a contact situation. It is assumed that people wash their clothing on a regular basis. It is similarly assumed that through time, these garments will shed differentially based upon their composition. This can be simulated in controlled laboratory conditions and data collected using washing cycles in commercial clothes washers and dryers to simulate age. This study will produce a replicable set of procedures for testing garment Shedding characteristics and a data set that can be used to generalize about a garment’s propensity to shed based upon its age. Chapter 3 METHODS AND RESULTS OF ANALYSIS Practical Considerations of the Study In keeping with the current literature trend of providing studies and data that will be of realistic use to fiber examiners, this study is designed to provide easy replicability and useful information to the forensic community. The apparatus for making simulated contacts is modeled after that of Coxon et al. (1992). All materials used in the study are readily available. The laboratory equipment used is basic in nature and available at most laboratories. Garments Chosen for the Study Thirteen garments were bought for this study. They represent a wide range of fabrics and blends, to provide a realistic set of garment types encountered in forensic casework. Each garment has at least one 20 x 36 cm area where the simulated contact apparatus can be applied without crossing a sewn seam. The garments are all uniformly woven, with no irregular patches that could affect shedding characteristics. A full description of each garment’s color and fabric composition (according to the label on each item) is in Appendix A. Simulated Contact Apparatus Utilized The simulated contact apparatus consists of a 19 x 20 cm wood cigar box, weighted with 50 steel hex nuts. The entire apparatus weighs 1140.97 g, representing a ll 12 contact pressure of 30 kg per square meter, determined to represent a “realistic contact pressure” by Coxon et al. (1992: 152). Red or white felt is clamped around the box and fishing line is tied to the front of the box. The garment is clamped to a large piece of plywood into which small grooves are cut. These are to guide the experimenter and ensure that each successive drag measured exactly 36 cm. The box is dragged across the surface of the garment over a time period of 5 seconds, mimicking the Coxon et al. 1992 study. After each box is dragged across the garment, clear, colorless adhesive tape ( 2‘/z inches wide) is applied to the felt surface and any overhanging remnants removed. The remaining square of tape is removed from the felt and applied to the surface of a labeled acetate overhead projector film. These are put onto a lightbox and the colored fibers are easily counted. Simulated Aging of a Garment Through Repeated Washing After the garments are tape lifted from the first dragging, they were laundered in a standard laundromat washing machine with liquid laundry detergent (in amounts recommended on the detergent bottle). A generic liquid detergent with no bleaching agents or fabric softeners was used in the study. Garments 1 through 7 were then put in a clothes dryer and run for two cycles, approximately 30 minutes. Garments 8 through 13 were air dried in accordance with the cleaning suggestions printed on the inside label of each garment. The clothes were then re-dragged, tape lified and the washing cycle begins again. The raw data of fiber counts generated are presented in Table 1. A graph of the fiber counts for Garments 4 through 13 was generated and linear regression used to find a l3 best-fit line for each. These are presented in Appendix B. Garments 1, 2 and 3 did not generate any shedding, so no graphs were made. In addition, an estimation is made of the relative percentages of short (0-Smm), medium (5-10mm) and long (over 10mm) fibers for each sheet. This was quantified as a three numbered ratio. A ratio of 40:35:25 means the garment shed 40% short fibers, 35% medium fibers and 25% long fibers. Trends noted for each garment are presented in the following paragraphs. These data are used to understand the mechanisms of shedding involved in this study. Data for Garment l Garment 1, a pale green blouse made of 100% polyester, was washed and dried in laundromat machines 10 times. The tapings showed no fibers shed at all after any of the washes. These results are to be expected because tightly woven polyester fibers are long and retain their stability. Data for Garment 2 Garment 2, a navy blue shirt made of 55% cotton and 45% polyester, was subjected to 10 wash and dry cycles in the laundromat. It shed one fiber after the first washing. This fiber was long (12mm) and very thick (relative to the other fibers analyzed in this project) and appears to have originated from a frayed edge inside seam of the garment. Obviously, the mean of fibers shed was low (0.09). These results are consistent with expectations because the weave of the garment creates a “lab coat” texture which is resistant to snagging and fiber shedding. 14 Data for Garment 3 Garment 3 is a goldenrod blazer made of 100% cotton. It did not shed a single fiber after any of the washes. It is also woven into the “lab coat” texture, so these results are as anticipated. Data for Garment 4 Garment 4 is a bright pink shirt made of 40% rayon, 33% polyester and 27% cotton. It shed fibers that ranged from 1 to 7mm in length. Garment 4 shed a range from 1 to 127 fibers, with a mean of 22.8. These figures do not accurately represent the trend noted in fiber shedding for this garment. Afier purchase, the garment shed 127 fibers. The fiber count after the first wash was 15. The subsequent fiber counts were similarly low, never reaching a number higher than 23 fibers shed. It shed mostly short fibers at the beginning (70:20: 10) and showed a gradual change to 85: 1 5:0 after the final wash. Data for Garment 5 Garment 5 is a blue shirt made of 65% linen and 35% cotton. It shed a range between 74 and 34 fibers, with an average of 46. These fibers were all between 1 and 3mm in length. Unlike Garment 4, Garment 5 showed a simple linear decline in fibers shed after each wash. All of the fibers were in the short (O-Smm) category. Data for Garment 6 Garment 6, a charcoal gray 50% cotton/50% polyester T-shirt, showed a shedding pattern similar to that of Garment 5. It shed a range between 176 and 53 fibers, with a mean of 101.7. The fibers ranged from 2 to 10mm long. The decline was linear with no initial spike of fibers shed, as seen in Garment 4. This garment shed half short and half 15 medium length fibers (50:50:0) at the beginning and eventually shed more short fibers (80:20:0) after the last wash. Data for Garment 7 Garment 7 is a bright pink 100% acrylic skirt. The fibers shed were between 2 and 5mm long. Garment 7 shed between 4 and 60 fibers, with a mean of 20.9. Its shedding pattern was unlike any seen in the previous garments. After purchase it shed 3O fibers. An increase to 60 fibers shed after the first wash was noted and then a linear decline similar to that of the other garments was noted. Garment 7 shed all short (0- 5mm) fibers. Data for Garment 8 Garment 8 is a cream skirt made of 40% rayon, 25% nylon, 20% wool, 10% angora and 5% cashmere. Tape lifts of it produced a range between 437 and 73 fibers with a mean of 176.3. Most of the fibers recovered were about 20mm long, but some were as short as 2mm. These data followed the trend seen in Garment 7 of an initial modest amount, followed by a marked increase after the first wash and then a gradual decline. It shed mostly small fibers. (80: 15:5) at the beginning, which only increased in frequency (95:5:0) by the end of the study. Data for Garment 9 Garment 9 is a charcoal gray skirt made of 80% lambswool and 20% nylon. It shed between 948 and 260 fibers with a mean of 632. The fiber lengths varied from 2 to 15mm in length. It followed the same pattern as Garment 8 first shedding 394 fibers. After the first wash, the fiber count soared to 926 and then more-or-less declined from this point. Garment 9 shed mostly small fibers at the beginning of the study (60:20:20) l6 and followed the general trend seen in other garments of increasing this number of small fibers to 75:15:O at the end. Data for Garment 10 Garment 10 is an off white sweater made of 55% ramie, 15% angora, 15% acrylic and 15% nylon. It shed fibers that ranged from 2 to 20mm in length. Garment 10 shed a range between 253 and 4 fibers, with a mean of 72.7. It followed the pattern noted for Garments 8 and 9 also, shedding most fibers after the first wash, followed by a fairly marked decline in fiber counts. Garment 10 shed exclusively long fibers at the beginning of the study (0:0:100). This ratio changed to 40:40:20 in a linear fashion by the last wash. Data for Garment l 1 Garment 11, a bright red 100% acrylic sweater, shed 1 to 3mm long fibers. It shed a range between 149 and 20 fibers, with a mean of 63.4. It generally showed a gradual descent in fiber numbers after each consecutive wash. All of the fibers shed were in the short (0-5mm) category. Data for Garment 12 Garment 12 is a chartreuse 100% acrylic (or chenille) sweater. It shed between 163 and 27 fibers per wash with a mean of 65. The fibers recovered were between 3 and 7mm long. The greatest number of fibers was shed afier the first wash, with first a marked, then a gradual decline thereafter. Garment 12 is the only garment that shed a consistent ratio of 50:50:0 throughout all of its washings. 17 Data for Garment 13 (Recipient Fabric Control Garment) Garment 13 is a red 80% lambswool, 20% nylon sweater. This was used as a control for recipient fabric variations. It was washed, air-dried, then dragged twice with a different piece of felt each pass. One pass, Felt A, was not changed throughout all 10 washes. Felt B, conversely, was changed each time. Felt A collected a range of fiber lengths between 3 and 15mm. It collected a range between 436 and 96 fibers, with a mean of 23 1 .6. Felt B collected a similar range of fiber lengths and a count ranging between 462 and 128 fibers, with a mean of 275. A statistical comparison of means showed no significant difference between the two means. These results indicate that prior uses and tapings of a recipient textile have little or no effect on results of a study such as this. Felts A and B collected a fiber size ratio of 40:40:20 initially. These gradually changed to 60:35:5 for Felt A and 50:45:5 for Felt B. This indicates that surface conditions of the recipient textile affected the Size of fibers recovered from the simulated contacts. 18 Table l—FIBER COUNT DATA Wash Wash Wash Wash Wash Wash Wash Wash Wash Wash Wash 0 l 2 3 4 5 6 7 8 9 10 # l 0 0 0 0 0 0 0 0 0 0 0 # 2 0 1 0 0 0 0 0 0 0 0 0 # 3 0 0 0 0 0 0 0 0 0 0 0 # 4 127 15 19 23 9 8 14 7 15 13 l # 5 74 58 39 47 49 56 42 29 37 41 34 # 6 176 143 145 127 139 94 77 53 64 48 53 # 7 30 60 39 21 18 9 ll 17 12 9 4 # 8 166 437 208 186 200 194 128 89 75 83 73 # 9 394 926 948 834 744 689 649 510 555 443 260 # 10 85 253 198 91 72 43 28 9 7 4 10 # 11 149 94 106 87 64 49 42 35 27 24 20 # 12 34 163 97 64 79 61 56 59 43 32 27 13A 164 436 367 321 285 268 198 167 141 105 96 13B 147 462 411 387 351 307 263 239 188 143 128 Chapter 4 DISCUSSION AND CONCLUSIONS Trends Noted in Fiber Shedding Characteristics Three general trends were noted in examination of the charts of shedding patterns for the garments in this study. Garments either did not shed any fibers at all, exhibited a gradual decrease in the number of fibers shed, or shed a large number of fibers after the first wash with a gradual decline thereafter. It is hypothesized that several features affect a garment’s shedding characteristics, including weave, fiber composition, fiber length, age and recipient textile features. Garments l, 2 and 3 shed no fibers throughout the study. One large string was found after the first wash of Garment 2, but this most likely represents a fiber lost from a frayed seam since no other fibers were found in any wash of this garment. These items were made of tightly woven cotton, polyester or a mix of both. This explains the lack of shedding initially and after several washings. Polyester fibers are created as long strands and when tightly woven with other polyester or even cotton fibers, these anchor firmly into the cloth matrix and resist shedding. Cotton fibers are shorter, but when tightly woven will similarly maintain cohesion. A second trend noted was a gradual decrease in the number of fibers shed through time. Garments 4, 5, 6, and 11 conformed to this trend. All of these items exhibited looser weaves than Garments l, 2 and 3. Three of these were blends of cotton with other 19 20 fibers. One was a 100% acrylic sweater. The combination of fiber types in the blend of the garment is assumed to be responsible for the trend. Garments 7, 8, 9, 10, 12, and 13 all showed a marked increase in fibers shed after the first wash and then a decrease. This decrease was either gradual and linear, as seen in the charts of Garments 9 and 13 (see Appendix B) or initially extreme, followed by a smoother, more gradual decrease beginning around wash 4 or 5. The initial spikes of long fibers shed represent the watershed expulsion of whole, loose fibers from the garment, followed by the release of more tightly-bound fibers which would naturally decrease in number after subsequent washes. The Effects of Fiber Composition on Shedding Based upon these three trends noted, one can investigate the fiber blends of the garments and hypothesize relationships between fiber composition of a garment and its shedding characteristics. From an investigation of the first three non-shedding garments, it is shown that most polyester garments do not shed fibers. Polyester does not fragment or break in the washing or contact process. Blends of polyester may or may not shed (see Garments 3 and 6). Garment 3, a 50/50 blend of cotton and polyester did not shed, while Garment 6 did. Since the fiber composition of these garments is identical, some other mechanism must be responsible for the differences in shedding noted. Garment 3 is a tightly woven jacket while Garment 6 is a more loosely woven T-shirt. This indicates that it is the weave of the garment, then, that may be a strong determinant of whether or not it will shed fibers. This conclusion is supported by comparing the charts of the cotton garments. Garments 2 and 21 3 are tightly woven and do not Shed, whereas the Garments 4, 5, and 6 are more loosely woven and do shed in the linearly decreasing fashion previously noted. By looking at the characteristics of the garments that shed a gradually declining number of fibers through time, it is evident that weave is a significant predictor of fiber shedding. The earlier comparison of cotton garments supports this conclusion. Acrylic garments also show this trend. The tightly woven Garment 11 shed in the gradual decline trend, while the loosely woven Garments 7 and 12 Shed a spike of initially high numbers, then a marked decline followed by a gradual decrease. These data corroborate the hypothesis that tightly woven textiles shed less throughout the study. One last trend noted that relates fiber composition of a garment to shedding characteristics is that wool and angora blends are all placed in the spike category. These natural fibers are generally short and pull away from the weave easily, as demonstrated by the initial spike in the chart. The washings pull these short fibers out of the weave relatively easily, so the shedding is continued with only a gradual decline noted in most cases. An Investigation of Shedding Mechanisms Many authors have postulated mechanisms of fiber shedding in forensic contexts. The intrinsic characteristics of the fiber such as thickness, shape, length and garment weave affect shedding characteristics (Siegel, in press). The seminal work of Pounds and Smalldon (1975a; 1975b; 1975c) did away with the commonly accepted notion that only loosely adhering fibers would be shed from a garment. They found a linear, dependent relationship between the number of fibers shed and the contact pressure applied to the garment. They concluded that there are three types of fiber exchanges: fragmented 22 fibers that are easily transferred, loosely held complete fibers that are easily transferred and those that are fragmented by the contact process. One trend noted in this study is that acrylics generally shed short to medium length fibers in a consistent ratio regardless of the number of washing cycles the garment undergoes. Therefore, it appears that no whole fibers are being shed since no outlying long fibers are found. Breakage during contact appears to be the mechanism responsible for acrylic fiber shedding. Natural fibers, such as wool and angora, shed fiber ratios that begin with high numbers of long, whole fibers, then exhibit a spike of these followed by a gradual increase in short fibers noted. This supports the Pounds and Smalldon model of garments shedding loosely bound whole fibers first and then giving way to breakage of the more tightly bound fibers. Finally, the different results of Felt A and Felt B on Garment 13 must be considered. Both of these began with a 40:40:20 ratio of fibers shed. After the last wash, Felt A shed a ratio of 60:35z5 and Felt B shed a ratio of 50:45:5. Felt A trapped more short fibers, relatively, than Felt B which trapped more medium length fibers. The general trend is the same, but the constantly re-used Felt A appears to have broken more fibers than Felt B. Consideration of Natural Versus Simulated Aging of a Garment Previous studies of garment shedding involved the use of donated clothing, often older garments that had been worn an unknown number of times. In the research design of this study, this factor was eliminated by the use of new garments. The washing process simulates the aging of a garment in an individual’s wardrobe. 23 A limitation of this study is the necessity to simulate the normal aging process that a garment undergoes. Throughout the course of the normal “life” of a garment, it is subjected to several contacts of varying strength, duration and placement. It is impossible to simulate these random events that may dramatically alter its propensity to shed fibers. The natural affect of light, temperature and weather on garments is similarly impossible to mimic in the laboratory. The simulated aging only involves the washing, air- or machine-drying of the garment, the simulated contact and then repeat washing. No provision for other interactions is made. These garments were never actually put on by anyone, so they never participated in any contacts with other surfaces or garments. They were kept in plastic bags once dry, to reduce the chance of cross contamination of the various garments, so they were not subjected to normal levels of light. This should cause overestimation of the shedding propensity of the garments under study. There is no opportunity for fiber loss due to contact, or for garment degradation due to environmental factors. If the goal of the study is to determine how aging affects a garment’s shedding characteristics, however, a controlled study is necessary. Therefore this study fulfills that need even if it does not mimic the full range of contacts that a garment undergoes. Applications of the Study: Methodology One of the primary goals of the study is to provide the forensic community with an easily-constructed simulated contact apparatus. This is to encourage further research into fiber shedding characteristics of other fiber types and weaves. By using the 1992 Coxon et al. study as a guideline, the simulated contact apparatus can be assembled from materials available at most hardware stores. Any fiber examiner or forensic science 24 research could find and purchase all the materials needed for a similar study in a short time. Applications of the Study: Toward an Index of Shedding Although fibers are class evidence, they may provide crucial information otherwise unavailable in forensic cases. Databases of fiber types, such as carpet and clothing, are of great value to forensic scientists. As more and more methods of fiber analysis are developed, simpler methods of categorizing fiber types seem to be falling to the wayside. This research helps lay the path for a determination of an index of shedding for various fiber types. By assigning a numerical index to fiber types and weaves (based upon how many fibers that textile type typically sheds under controlled conditions), databases of fiber shedding indexes could be established. These could be used to corroborate witness testimony about the duration and nature of contact between two participants in a crime. If a particular fabric and weave are known to shed easily (possessing a high shedding index) and relatively few are found at a crime scene, the analyst would conclude that the contact must have been neither lengthy nor forceful. Suggestions for Future Research Since this study has illuminated several garment shedding trends, studies of textile weaves are the most logical follow-up to this research. Performing a similar study of several 100% cotton garments differing only in their weave types would provide interesting information on cotton shedding characteristics. Most research indicates that it is the weave of the garment that affects the mechanical forces responsible for Shedding. 25 Therefore, one would expect such a study to demonstrate a linear relationship between the strength of fiber weave and the number of fibers shed. Another valuable avenue of inquiry would be investigating the shedding characteristics of several blended fabrics. Most research suggests that the shedding characteristics of blends such as cotton/polyester garments change according to the relative amounts of each fiber type. No one has studied the stability of this trend through time. Will the ratio of cotton to polyester fibers shed remain constant over the “life” of a garment? Lastly, a target fiber study of a group of garments washed together would be worth analysis. One could address the question of cross-contamination of garments involved in a crime by garments with which they were washed. If two shirts are washed together, will fibers of the unwom item be present on the one worn during the criminal act? If so, would they be present in a significant quantity to mislead the fiber examiner into concluding that the contaminating garment was worn during the crime? Summary In summary, this research has shown that garments do shed differentially through time. Three trends were noted. Some garments did not shed any fibers. Some shed initially high numbers followed by a gradual decline. A third group shed relatively few fibers at first, followed by a high ntunber after the first wash and then a gradual decline. Fiber type and garment weave appear to greatly affect shedding characteristics. Polyesters and tightly woven polyester-cotton blends do not shed. Tightly woven cotton does not shed while looser cotton weaves and blends do. Acrylics always shed, but the 26 amount is related to the weave. Finally, natural fibers such as wool and angora shed many fibers, regardless of weave. These trends support the Pounds and Smalldon hypotheses of shedding mechanisms, specifically that loosely-held, whole fibers are shed readily in the first few washings. After this, tightly-woven fibers are fragmented with increasing frequency. This manifests itself in the general trend seen of increasing amounts of small fibers noted afier several washes. APPENDICES Appendix A GARMENT DESCRIPTIONS : ll 1 ll: .. E] C .. 1—pale green blouse 2—navy blue shirt 3—goldenrod blazer 4—pink shirt 5—blue shirt 6—charcoal gray t-shirt 7—pink skirt 8—cream skirt 9—grey skirt lO—off-white sweater ll—red sweater 12—chartreuse sweater l3——red sweater polyester 55% cotton, 45% polyester cotton 40% rayon, 33% polyester, 27% cotton 65% linen, 35% cotton 50% cotton, 50% polyester acrylic 40% rayon, 25% nylon, 20% wool, 10% angora, 5% cashmere 80% lambswool, 20% nylon 55% ramie, 15% angora, 15% acrylic, 15% nylon acrylic acrylic 80% lambswool, 20% nylon 27 Appendix B GRAPHS OF FIBER COUNTS FOR GARMENTS 4 THROUGH 13 28 Fiber Count Qua—:63 h .30 AND 30 mo we no no -No 2333 lol moaofl inane. can“: _ 29 Fiber Count mo wo mo mo no we no no 02.36.: m 5338 ._o ._N _ rel momma... Isattasnae _ 0 __ 3 Fiber Count Moo 3o 30 30 30 30 mo mo no no 02.32.. a fiancee AM If: mama, __ I Enema 839.: V I_ II. 3 Fiber Count 02.32: a «0 oo mo no. wo. no no 2338 no ._~ :1. 833--.. .l. . 5326333. 2 3 Fiber Count moo nmo noo woo woo woo woo noo noo mo $2.36.... a Season no AN _ IT mama ,. , [.532 662mm: _ 3 3 Fiber Count @232: o n ooo ooo moo woo ooo . moo 3o . woo woo I noo. It, Eon—.3 ._o ._N 3 tot - Women; s _ Jase nae: 4 3 Fiber Count woo woo woo noo noo mo -mo $2.32... .3 See-.8 Iol. - - - .Weamfl ,_ diagnose: _ 5 3 Fiber Count 0935:. ._ ._ So go A 30 1 3o mo we be no) 23:3 do in IT. ‘mmzma “ ll. :32 magi 6 3 Fiber Count 3o .50 30 ._No So we ao be no 02.32: .3 53:3 3 ._N +.-mm1ma . [momma Ammamm: ._ 7 3 Fiber Count woo Amo boo wwo woo woo Moo 30 .50 mo 62.32: .3. mo: > 5253 .5 AM If. mama IIH rimmxlmmawfl V _ 00 3 Floor Count moo Ame ooo wmo woo Mme moo .30 .80 mo 9938:. 3. mo: m 23:8 3 AN +mmammd-_i XII. rammfimlmqmé Fiber Count moo Amo boo wmo woo woo ,. moo Go 30 mo 92.32: .3" ooauuzoo: oq no.3 > 2:. m Emu—.3 3 A» 1+. z .933. mil: - Wong» 0 4 BIBLIOGRAPHY Bibliography Ashcroft CM, S Evans and IR Tebbett (1988) The persistence of fibres 1n head hair. mm 28: 289- 293. Bartick EG (1987) Considerations for fiber sampling with infrared microscopy. In Ihe D- '1 _..1n-u-,.u'_r° 1..$.. 1" 0 Ho. Wu. 0 00‘ -PBROUSh, ed., ASTM: Philadelphia, PA. Bresee RR and PA Annis (1991) Fiber transfer and the influence of fabric softener. JoumalnfEnnensieScienccs 36 (6): 1699-1713. Carroll GR (1992) Forensic fibre microscopy. In W. J Robertson, ed., Ellis Horwood: New York, NY. Cook R and C Wilson (1986) The significance of finding extraneous fibres 1n contact cases MW 33: 267-273 Cordiner SJ, P Stringer and PD Wilson (1985) Fibre diameter and the transfer of wool fibres WWW 25: 425-426 Coxon A, M Grieve and J Dunlop (1992) A method of assessing the fibre shedding potential of fabrics. WWW 32:151-158. Deadman H (1984) Fiber evidence and the Wayne Williams tn'al. Wm Bulletin; Part l—p 13, Part 2—p 141. Gaudette BD (1988) The forensic aspects of textile fiber examination. In Forensic W, Vol. 2. R Saferstein, ed., Prentice Hall. Englewood Cliffs, NJ. Grieve MC (1990) Fibers and their examination in forensic science. In We; Progress, Vol. 4. A Maehly and RL Williams, eds., Springer Verlag: Berlin, Germany. Grieve MC (1994) Fibers and forensic science—New ideas, developments, and techniques. WW 6: 59-79. Grieve MC, J Dunlop and PS Haddock (1988) Transfer experiments with acrylic fibres. WM 40: 267- 277 Jackson G and R Cook (1986) The significance of fibers found on car seats. Forensic Ssiencelntemational 32: 275- 281 JacksonG and CN Lowrie (1994) Secondary transfer of fibers. W International 64: 93-114. 41 42 Lowrie CN and G Jackson (1991) Recovery of transferred fibres. EQIensieSeienee International 50:111-119. Moore JE, G Jackson and M Firth (1986) Movement of fibres between working areas as a result of routine examination of garments. WM 26: 433-440. Parybyk AE and RJ Lokan (1986) A study of the numerical distribution of fibres transferred from blended fabrics. leumalefjhefierensieSeleneeSeeiefl 26: 61- 68. Pounds CA and KW Smalldon (1975a) The transfer of fibres between clothing materials during simulated contacts and their persistence during wear: Part 1- Fibre transference loumaloftheflremieSeiemSoem 15: 17-27. Pounds CA and KW Smalldon (1975b) The transfer of fibres between clothing materials during simulated contacts and their persistence during wear: Part II- Fibre persistence MW 15:29-37 Pounds CA and KW Smalldon (1975c) The transfer of fibres between clothing materials during simulated contacts and their persistence during wear: Part III-A preliminary investigation of the mechanisms involved. Wm Seieneefigeiety 15: 197- 207. Robertson J, Ed. (1992) WWW. Ellis Horwood: Chichester, England. Robertson J, CBM Kidd and HMP Parkinson (1982) The persistence of textile fibres transferred during simulated contacts. JenmalmhefierensieSeieneeSeeiety 22: 353- 360. Robertson J and AK Lloyd (1984) Observations on redistribution of textile fibres. Jenmal otthefiorensieSnienceSQciem 24: 3- 7. Robertson J and D Olaniyan (1986) Effect of garment cleaning on the recovery and redistribution of transferred fibers. lenmalnffienensieSeienees 31 (1): 73-78. Roux C, J Chable and P Margot (1996) Fibre transfer experiments onto car seats. Seienee andiustiee 36 (3): 143-151. Salter MT, R Cook and AR Jackson (1987) Differential shedding from blended fabrics. WM 33: 155-164 Siegel JA (1996) Forensic Textile Fiber Analysis. In EerensieSeienees, CH Wecht, ed., Matthew Bender: New York, NY. "7111111111111111111111155