Impact Spatter Bloodstain Patterns on Textiles

Introduction

Bloodstain pattern analysis (BPA) plays an important role in forensic science by analyzing the size, shape, number, and distribution of bloodstains left at a scene in order to provide scenarios consistent with the creation of these stains (1). The first systematic BPA studies are attributed to Eduard Piotrowski in 1895 (2). Since then, sporadic investigations of BPA were reported until the 1980s, when systematic studies of BPA began in earnest with the formation of the International Association of Bloodstain Pattern Analysts in 1983 (3).

Impact bloodstain patterns are formed when a source of blood is impacted with some object. Impact bloodstains are frequently found at violent scenes including accidents, beatings, or gunshots (4). Mouse and rat trap devices have been used in training and research to produce impact spatter (5)(6)(7)(8)(9) and are easily made from commercial mouse or rat traps with a wooden board attached to the spring. One problem we encountered with the wooden device in our laboratory was low reproducibility as the traps closed differently during each impact. To overcome this variability, a modified rattrap device was developed and is reported below.

Most research on BPA has been performed on nonabsorbent surfaces. Only 51 of 1041 papers in the OSAC BPA bibliography mention fabrics, textiles, clothing, or carpets in their titles (10) even though textiles are often present at crime scenes in the form of clothing, curtains, upholstery, and carpeting (11). Bloodstain pattern analysis on textiles is more difficult than on nonabsorbent materials due to the complexity of textile properties. There are thousands of different types of textiles in use today (for a small sampling, see ref. [12]). They may be designed to repel, absorb, or wick liquids (13); they may be clean, or more typically, dirty. Bloodstain patterns on textiles exhibit large diversity as they are sensitive to the fiber type, the yarn manufacturing process (14,15), fabric structure (16), finishing treatments, and even after-sale care (17).

Several researchers have conducted bloodstain pattern analysis research on textiles, primarily focused on passive bloodstains (13)(14)(15)(16)(17). They have rarely paid attention to the detailed textile structure and how structural differences may lead to altered bloodstain appearance. Among the vast array of textiles, knits are of great importance because they are flexible, durable, and easily constructed into different sizes and garments including hats, socks, T-shirts, and sweaters (18). These garments are often found at bloody scenes. Another aspect of BPA on textiles is that it is often difficult to distinguish between impact spatter and transfer where blood was deposited onto a surface and subsequently transferred to a textile (19).

Analysis of impact spatter on textiles has been identified as an area that needs further research (20). To begin to address impact spatter on textiles, we present our results for impact spatter on knit fabrics where the primary variable was restricted to different yarn sizes. To accomplish this, we modified a rattrap-based impact spatter device to improve the reproducibility of the spatter over our homemade rattrap device and used it to create impact spatter on the knits. The resulting stains were analyzed with ImageJ and are presented below.

Materials and Methods

In our initial studies, we built a rattrap device to create impact spatter. We found that each time we created a stain, the stain occurred in a different region, often missing a large portion of the target. This made it difficult if not impossible to compare one stain to another. Thus, we began by modifying our rattrap to produce more consistent results. We describe our modified device below and our results using it, but not an evaluation of the modified device itself as that is beyond the scope of this study.

Modified Impact Device

The modified impact device, based on the traditional rattrap, is shown in Fig. 1. The main parts of the device were made of two polypropylene boards to prevent blood absorption and to enable removal of residual blood to prevent interference in subsequent tests. The polypropylene boards and the small joint components were purchased from McMaster-Carr Supply Company (Douglasville, GA). The springs were disassembled from a Victor Metal Pedal rattrap and modified to fit the new impact device. To stabilize the modified rattrap, it was bolted to a 2cm-thick steel plate. This was found to be heavy enough to prevent unintended movement of the modified rattrap. To ensure the whole device was mounted horizontally, a bubble level was attached to the surface of the lower polypropylene board. The device was checked for level each time prior to releasing the spring.

To better control the blood flight direction, a triangular groove was cut into the lower board and a matched raised triangular region was machined into the upper polypropylene board. During preliminary experiments, we identified a suitable location to place a 0.2 mL blood drop and marked it with a permanent marker. In all subsequent experiments, 0.2 mL of blood was placed on this spot. For safety and easier operation, a metal latch was placed on the rear of the device. The tail of the latch was bent to hold the upper board in the center of the back edge allowing the upper board to be consistently released without twisting. Two laser modules were mounted inside the rattrap base facing forward to assist alignment of the impact position on the target. The plane of the two laser modules center axes was 5.0 mm below the plane of the impact zone.

Device Setup

The modified impact device (rattrap) was mounted on a laboratory jack to adjust the height. An iron ring stand was placed 30 cm away from the impact device to hold the sample, as depicted in Fig. 2. To confine all the impact blood spatter within a contained space, the experiments were conducted inside a clear glove box with one end removed for access. Two baffles made of transparent plastic sheet were attached to a small laboratory jack in front of the device to reduce blood spatter outside of the target zone. Any drops beyond this area were blocked by the baffles.

Paper

Fisherbrand P8-creped filter paper was purchased from Fisher Scientific (Pittsburgh, PA) and was used as a control surface.

Knit Fabrics

For the research discussed below, the textile materials were restricted to three 100% cotton single jersey knit fabrics, which are commonly used for clothing manufacture such as T-shirts. Two ring-spun singles cotton yarns of different yarn sizes were obtained from Cotton Incorporated, namely 492dtex (cotton count 12Ne) and 295dtex (20Ne). A third yarn sample was obtained by extracting yarns from a bleached cotton single jersey knit T-shirt fabric made of 197dtex (30Ne) ring-spun yarns. This fabric was a commercial fabric purchased from Test Fabrics, Inc. (West Pittston, PA). The twist levels of the yarns were measured by twist-untwist method, according to ASTM D1422/D1422M-13, "Standard Test Method for Twist in Single Spun Yarns by the Untwist-Retwist Method" (21), results are reported in Table 1. The twist multiplier TM was obtained by TM ¼ TPI= ffiffiffiffiffi ffi Ne p where TPI is the turns/inch as measured above and Ne is the cotton count. Ring-spun yarns with the same twist multiplier have the same helix angle on their surface and the same internal structure (22).

The two yarns supplied by Cotton Incorporated were converted into single jersey knit fabrics in the knitting laboratory of

the College of Textiles at North Carolina State University. These knit fabrics underwent scouring and bleaching to remove knitting oil and any other additives, which could alter the interaction of blood with the fabric. Scouring and bleaching were conducted in a Thies mini-soft fabric-dyeing machine. Soda ash (3 g/L), BASF Primasol â N-SA surfactant (3 g/L), hydrogen peroxide (2 g/L), and defoamer (1 g/L) were added to the machine in the process with a liquor ratio of 1:50 applied (14). The temperature was raised to 104°C (220°F) and kept for 30 min. The fabrics were then rinsed three times using cold water to remove residual chemicals.

All three knit fabrics were subsequently washed and dried following the direction of AATCC Monograph M7 "Standard Laboratory Practice for Home Laundering Fabrics Prior to Flammability Testing to Differentiate Between Durable and Non-durable Finishes" (23). The washer was filled with water at 60 AE 3°C, and the rinse temperature was set to 29 AE 3°C. 1993 AATCC standard reference detergent WOB (without brightener) was used while washing. After the washing process, the load was moved into a dryer and was dried for 45 min at 67 AE 6°C. After laundering, the fabrics were ironed to remove unwanted wrinkles. Then, all the fabrics were allowed to equilibrate for 24 h to allow conditioning to the laboratory environment.

The diameters of the yarn were measured at different positions along the knit loop, including the sinker loop, needle loop, and side legs, as shown in Fig. 3 and reported in Table 1.

Fabric characteristics are also given in Table 1. The thicknesses of the fabrics were measured by AMES thickness tester following test option 1 in the ASTM D1777-96, Test Method for Thickness of Textile Materials (24). The thicknesses of the 492dtex (12Ne), 295dtex (20Ne), and 197dtex (30Ne) knitted fabrics were approximately twice the yarn diameters. The weight of the fabric in grams per square meter was tested according to ASTM D 3776, Standard Test Method for Mass Per Unit Area (weight) of Fabric (25). Cotton fabric was cut by a J.A. King universal manual sample cutter (Model No. SASD-688). Fabric count was measured according to ASTM D8007-15e1, Standard Test Method for Wale and Course Count of Weft Knitted Fabrics (26).

Porcine Blood

To avoid exposure to human blood, porcine blood was used in this research since it has been shown to have properties similar to human blood (27). EDTA anticoagulated porcine blood was purchased from Lee BioSolutions Inc. (Maryland Heights, MO) and was kept refrigerated at 2-8°C. Each time before use, the container of porcine blood was put on a Fisher Scientific Digital Bottle Roller and rolled for at least 60 min at 15 rpm until the plasma, blood cells, and other blood constituents mixed thoroughly without destruction of red blood cells.

Surface tension and hematocrit value of porcine blood were tested as received. The surface tension was measured through "Pendant Drop" method (28). Hematocrit was measured by centrifuge method; both are reported in Table 2. Prior to each experiment, the temperature and viscosity of porcine blood were measured to ensure the blood was fresh enough to be used. Temperature of the porcine blood was tested using a thermocouple and was ready for use when the blood temperature reached room temperature. Blood viscosity was tested with a Brookfield DV-E Viscometer with spindle SC4-18 according to ASTM D2196-15 Standard Test Methods for Rheological Properties of Non-Newtonian Materials by Rotational (Brookfield type) Viscometer (Test Method B) (29). Viscosity results are reported in Table 3.

Impact Bloodstain Test Method

The three jersey knit fabrics and filter paper were used as impact spatter targets in this research. Four schemes, as shown in Fig. 4, were used for each knit fabric in this experiment. First a sheet of filter paper was laid on an open 18 cm (seven-inch) diameter embroidery hoop. Next, a half-round of the selected knit fabric was placed on the left, right, upper, or lower half of the filter paper and the top of the embroidery hoop was attached and tightened. The wale direction of the knit was always arranged so that it would be vertical during the impact test while the machine direction of the filter paper was arranged to be horizontal. The embroidery hoop with the fabric and filter paper was then suspended from the ring stand at a distance of 30 cm from the modified rattrap. The plane of the embroidery hoop was mounted perpendicular to the travel direction of the blood (i.e., a 90°i mpact).

Then, 0.2 mL porcine blood of room temperature was taken by automatic transfer pipette (Gilson Inc., Middleton, WI) and was placed on the designated spot on the modified rattrap (Small drops will rapidly change temperature when in contact with another surface. In order to be consistent, the blood was only heated to room temperature since the apparatus could not be maintained at higher temperatures). When the upper board of the modified rattrap was released, it struck the blood and the resulting spatter hit the target to form a spatter pattern. Each fabric was mounted in each of the four positions (up, down, right, and left) with five repeats. This resulted in 20 sample specimen for each fabric and for the paper, that is, 240 total specimen. Each specimen was allowed to dry for 24 h prior to photographing. As discussed below, we found that the four positions of the fabric were statistically the same. This allowed grouping the 20 specimens from each fabric together when comparing the different fabrics and paper.

Analysis of Bloodstains

Each sample was photographed using a Nikon D90 camera with AF-S Nikkor 35 mm f/2D lens under CIE D75 lighting. The high-resolution digital photographs were analyzed using ImageJ software ImageJ software is freely available from https:// imagej.nih.gov/ij/download.html.). The "Color Threshold" tool was used to change the RGB value for all of the pictures. After a suitable RGB threshold value was determined and applied, the stained area was mapped, as shown in Fig. 5. Next, the "Particle Analysis" plug-in was applied for quantitative analysis of the impact spatter. The number of detected bloodstains on the impact spatter targets was counted, and their areas were measured by ImageJ.

Results and Discussions

Impact Bloodstain Pattern Appearance

Impact bloodstain patterns formed on the filter paper and on the three knit fabrics were examined using a Nikon SMZ1000 Zoom Stereo Microscope. Filter paper and knit fabrics are all fibrous materials, but the appearance of the patterns on the filter paper was different than those on the knit fabrics, as shown in Fig. 6. Filter paper is made by draining a slurry of short fibers in water through a screen. This results in a random distribution of fibers in the plane of the paper sheet. Paper fibers are also much shorter than those used to make fabric (30). As shown in Fig. 6a, the bloodstains on filter paper are more circular than on the knit fabrics. Cotton single jersey knit fabrics are made by first making the cotton fibers parallel, then twisting them to form yarns. These yarns are then knitted on a jersey knitting machine. As shown in Fig. 6b-d, the bloodstains on the knit fabric wicked into the yarns and followed the local helical path of the fibers, unlike the stain shown on the filter paper. Some blood drops landed on a single yarn and wicked along the fibers as shown in Fig. 6b. Some drops landed at the intersection where two yarns came together. In this case, the blood wicked along fibers within the different yarns and followed the local fiber directions, as shown in Fig. 6c. On occasion, blood drops would pass directly through the hole in the middle of a single loop, as shown in Fig. 6d. It is clear in these images that blood wicks through the space between the fibers within the yarns of the knit structure. No similar structure exists in paper so the blood spreads uniformly in paper.

Individual Drop Stain Area Distribution

All bloodstains with areas >0.06 mm 2 and <0.70 mm 2 were measured, and the distributions of the stain areas were determined. To be able to distinguish between the stains and the background texture within the fabric, the lower limit for stain area was set to 0.06 mm 2 (stain diameter of 0.28 mm). The upper limit on stain area was chosen to be 0.7 mm 2 (stain diameter 0.84 mm) since there were few stains with larger areas. Figure 7 shows the stain area distributions for the 20Ne fabric mounted in the lower position and on filter paper mounted in the upper position in the same impact event (see Fig. 3). Fabrics placed in the other positions gave similar distributions.

Statistical Significance

The areas of all stains were measured using ImageJ for all fabric specimens (20 specimens for each fabric, five specimens in each of four positions) and paper test specimens (60 total specimens, 15 in each position) and the distribution of stain areas determined using Excel. The number of stains with areas between 0.06 and 0.08 mm 2 was counted for each specimen, and the results were compared using single factor ANOVA in Excel. We found no significant dependence on fabric or paper placement (i.e., up, down, left, and right as shown in Fig. 3) at the p = 0.05 level. We also found no significant difference in stain areas for the 30Ne knit fabric and paper. On the other hand, the stain areas were significantly lower for the 20Ne and 12Ne fabrics. Furthermore, the stain areas for 12Ne fabric were significantly lower than for 20Ne knit fabrics at p = 0.01.

Average Stain Areas

The average area of individual stains on different substrates was calculated by dividing the total area of bloodstains as determined in ImageJ by the number of bloodstains observed; the results are shown in Fig. 8. The average area of the individual stains was larger on filter paper than on the knit fabrics, and average area became larger as the yarn size decreased (i.e., as the yarn count increased). The 12Ne yarn has the largest diameter; thus, this knit fabric had the smallest average stain area while the 30Ne yarn has the smallest diameter and the knit fabric had the largest stain area.

The individual stain size is influenced by the absorbency and surface nature of the contact materials. Both filter paper and cotton knit fabrics are cellulose based and are hydrophilic but have different surface properties. As illustrated in Fig. 9, fabric made of coarser yarns has bigger hollows between loop structures and possess more texture (sometimes referred to as "roughness"). This leads to a larger volume of air within the yarn per unit area of the fabric. As blood wicks into the yarn, it displaces air within the yarn. Thus, the more air within a yarn, the shorter the wicking distance required to contain the entire drop. This limits the wicking area for the bloodstain. As the yarn diameter becomes smaller, the volume of air within a yarn decreases per unit fabric area. As blood displaces air within the yarns, it must occupy a larger area to contain the same volume of blood. Thus, fabrics with smaller yarn diameters (higher count) have larger stain diameters.

Number of Bloodstains and Total Area of Bloodstains

The number of bloodstains detected and the total area of these bloodstains on knit fabrics are reduced compared to those on filter paper. By comparing the number of stains and total area of the stains observed on the three knit fabrics to those observed on filter paper, the percent reduction could be obtained as follows:

number of stains on knit fabric number of stains on filter paper % area reduction ¼ 100% 1 À area of stains on knit fabric area of stains on filter paper

The results are shown in Fig. 10. Clearly, the percent reduction is related to the size of the yarns within the knit fabrics. The lack of a significant difference in sample placement for number of stains (left-right-top-bottom) indicates that the same number of blood drops must have been deposited onto the filter paper and all three knit fabrics.

For this set of experiments, a single impact spatter drop intersects only one to three yarns since the maximum drop diameter is less than three yarn diameters, even for the thinnest yarn. The blood drops spread out in these yarns as the drops wick into the yarns as sketched in Fig. 9. Therefore, the cumulative area of the impact stains should be proportional to the yarn thickness, or equivalently, the fabric thickness, and the number of observed stains decreases, as shown in Fig. 11.

In the analysis above, a stain was only counted if it exceeded a certain size (0.06 mm 2 ). The stain area decreased as the yarn or fabric thickness increased and, therefore, the number of observed stains also decreased as the thickness increased since some of the stains were no longer larger than the minimum stain size counted. This is clearly seen in Fig. 10 where the % number and % area reductions closely parallel each other.

Conclusions

During this research, a modified rattrap device was designed, constructed, and used in impact bloodstain pattern creation prior to analysis. The device was made of polypropylene and steel. This avoids potential health hazards due to blood absorption. It is also easy to clean, which is important to avoid unknown effects of dried blood. A triangular groove and baffles were added and are used to direct the flight of blood drops.

Filter paper along with 492dtex (12Ne), 295dtex (20Ne), and 197dtex (30Ne) 100% cotton knit fabrics was used in this research. The filter paper and the cotton fabrics are all highly absorbent for blood. Four schemes were used in target arrangement. The upper, lower, left, or right half of a paper substrate was covered with a semicircle of fabric, and they were mounted together in an embroidery hoop. No statistically significant differences (p = 0.05) in the number of stains were found between the different fabric and paper positions.

The average stain size was calculated by dividing the total area of all stains by the number of bloodstains on the same fabric. The average stain size decreased with increasing yarn size (decreasing Ne). The observed bloodstains were also more abundant on filter paper and on the 30Ne (197dtex) knit fabric compared to the patterns on the 20Ne (295dtex) and 12Ne (495dtex) knit fabrics. The number of observed bloodstains on the knit fabric made with the 30Ne (197dtex) yarn was statistically the same as on paper. The average number and total area of patterns all decreased with the increasing yarn diameter and fabric thickness.

The authors attribute the observed differences to the substrate structure and fabric (yarn) thickness. Knit fabrics are made of porous yarns of different diameters, while paper is made from short randomly oriented fibers. Wicking of the blood drops into the knit fabrics followed the fibers that are twisted into yarns which form the knit loops. Thus, it was observed that the blood drops on the knit fabrics became highly distorted. Wicking into the paper occurred randomly since the fibers are arranged randomly and the drop shapes remained circular.

These findings and others (14,15) highlight the importance of including the yarn and fabric structural characteristics in the experimental design and analysis of bloodstains when performing BPA research on textiles.