New Developments In The Assessment of Protective Fabrics Using Computational Models

Transcription

1 ORIGINAL PAPER New Developments In The Assessment of Protective Fabrics Using Computational Models By Roger W. Hill, Ph.D., Engineer, and James J. Barry, Ph.D., Principal Engineer, Creare Incorporated, Hanover, NH Abstract Advances continue in the use of computational fluid dynamics (CFD) for predicting the performance of fabrics, especially those used for chemical and steam/fire protective clothing. In this CFD modeling approach, solution of transport equations determines diffusive and convective transport of heat and gases/vapors; capillary transport of liquids; vapor and liquid sorption phenomena and phase change; and the variable properties of various fabric materials. Models employing the approach have been developed ranging from simplified representations of air flow over a 2-D clothed human arm or test facilities used for material characterization to large 3-D human body regions with multi-layered clothing ensembles. In this continuing research, cases considered include adsorption of chemical vapors by activated carbon fabrics, liquid wicking into fabric samples and evaporation into the surrounding gas, and flow around and through fabrics covering a human form. Applications of the models include analysis of chemical protective garment design for military and emergency response personnel, comparisons of thermally protective materials for steam or fire protection, evaluation of clothing test data, design of clothing fabric testing systems, and engineered fabrics. Keywords CFD, modeling, comfort, protective clothing, activated carbon, engineered fabrics Introduction Protective clothing provides military personnel, emergency responders, laboratory and hazardous materials workers, and others with the means to control their exposure to chemicals, biological materials, and heat sources. Depending on the specific application, the textile materials used in protective clothing must provide high performance in a number of areas, including impermeability to hazardous chemicals, breathability, light weight, low cost, and ruggedness. Nonwoven materials provide key components of an increasing number of these protective garments. Development of protective materials presently relies heavily on testing. Swatches of textile materials undergo laboratory tests to measure their properties, and the performance of partial or complete clothing products are measured in test chambers or field tests. The objective of the work reported here is development of computational models for predicting the performance of textile materials in protective applications. Such models complement testing by enabling property data from tests of textile swatches to be used in performance predictions of integrated multilayer garments under varying environmental conditions. Modeling can explore parameter ranges that are difficult or costly to reach experimentally. In addition, modeling provides detailed information throughout its computational domain that can enable improved understanding of the fluid flow, heat and mass transfer, and textile properties and can aid in the design process of advanced protective clothing systems (Barry et al, 2003). CFD provides the basis for the models. In CFD, the governing equations for mass, momentum, and heat transfer in fluids are solved over a two- or three-dimensional computational mesh. The calculations result in predictions of the flow velocity, temperature, pressure, and composition at each location in the mesh. Though commercial CFD software provides 22 INJ Winter 2004

2 many built-in capabilities, it does not offer all of the physical models required to address the complex multiphase processes that occur in textile fabrics. Detailed models for these processes have been developed, integrated into widely used commercial software, and used for a series of validation and applications calculations. Computational Tools The computational tools used in the present work are similar to those used in prior reported work (Barry and Hill, 2003) and consist of standard commercial CFD software (FLUENT, Fluent, Inc.) enhanced with specialized models for simulating heat and mass transfer through textile fabrics. Figure 1 illustrates the transport and phase change processes embodied in these specialized models. Each fabric layer is a porous medium composed of four components: 1) gas/vapor, 2) solid fiber, 3) bound liquid adsorbed into the fiber, and 4) free liquid in the interstitial spaces between fibers. Normally, air constitutes the primary component of the gas/vapor mixture with water or other vapor (e.g., vaporized chemical agent) comprising the remainder. The liquid either free or bound can be water or a liquid chemical agent. (The present models do not allow for both simultaneous liquid water and agent.) In the fabric, transport equations are derived for mass, momentum, and energy in the gas and liquid phases by volume-averaging techniques (Gibson, 1994, 1996). Definitions for intrinsic phase average, global phase average, and spatial average for porous media are those given by Whitaker (1977, 1998). A summary of the equations for the models may be found in Barry and Hill (2003). For modeling of clothing performance, the outer surface of the wearer s skin forms one boundary to the clothing model. To provide a suitable skin boundary condition, a simple model of the skin surface is included to impose the effects of sweating and the rejection of metabolic heat by the body. Several enhancements have been made to the fabric models. These enhancements include 1) addition of an activated carbon model, 2) modifications to the evaporation and condensation model to improve convergence, 3) an evaporation and condensation exchange model between fabric layers and surrounding gas regions, and 4) more accurate treatment of sweat evaporation and skin heat transfer. To simulate clothing performance, computational representations of clothed humans have been developed using geometry and grid generation tools. These physical representations are then used in conjunction with the fabric physics models to provide insight into the air, vapor, and heat flow in and around the clothing layers. Barry and Hill (2003) provide a summary of geometry models used to date in their work. These included two- and three-dimensional representations of clothed arms/legs and a preliminary model of a three-dimensional clothed torso. 23 INJ Winter 2004 Figure 1 PHYSICS OF TRANSPORT/PHASE CHANGE IN FABRIC Figure 2 THREE-DIMENSIONAL TORSO MODEL WITH TWO SHIRT LAYERS AND ONE PANTS LAYER. GRID DENSITY ON CLOTHING LAYERS IS SIMILAR TO THAT SHOWN ON PANTS. COLORS ARE APPLIED TO DIFFERENT SURFACES FOR DISPLAY In more recent work, a complete torso model has been developed having two layers of clothing above the waist and one below the waist (Figure 2.) A prototype of a fully clothed human has also been developed but not yet used for clothing model simulations (Figure 3). More complex models of the 3- D arm have also been developed having cuff closures around the wrist and a simple hand (Figure 4.) Each of these models is based on a laser-scan image of a human. A scan from the CAESAR (Civilian American and European Anthropometry Resource) database was used for the torso and full-body models (Subject 232). Clothing layers are creat-

3 Figure 3 CLOTHED 3-D FULL-BODY MODEL PRELIMINARY PROTOTYPE Cuff Closure Agent Spot #1 Agent Spot #2 Figure 4 THREE-DIMENSIONAL ARM MODEL WITH HAND AND GAP CLOSURE. ARM GEOMETRY FEATURES SINGLE CLOTHING LAYER AND CUFF CLOSURE WITH VARIABLE PERMEABILITY ed as new surfaces from scaled up cuts through the skin location defined by the scans. Computational Results Barry and Hill (2003) reported computational results and data comparisons for several example problems using the special features of the fabric models. More complex problems have now been considered, some using the 3-D clothed human models shown above. ACTIVATED CARBON WITH EVAPORATING DROPLET Activated carbon is utilized as a vapor adsorbent in military chemical protective garments and other applications such as hunting wear (for scent suppression). The physics of agent adsorption by activated carbon are complex but also central to the function of most garments now used for chemical protection. A model of a single evaporating droplet was developed to investigate the local transport and adsorption process in the immediate vicinity of the droplet and activated carbon layer. Model Description. The objective of the evaporating droplet model is to assess the effect of different flow rates on evaporation from the droplet and sorption of the vapors into an activated carbon layer. The model utilizes the geometry of a swatch test cell typically used in evaluation of military protective fabrics. For simplicity and computational efficiency, an axisymmetric representation was used to study the evaporation from a single 1 µl liquid droplet placed in the center of the fabric sample. The liquid is methyl salicylate (MeS), a typical simulant for mustard-type chemical agents. For simplicity, the droplet is assumed to have a hemispherical shape and does not wick into the fabric. The simulated multilayer has the following properties (approximating the typical military protective fabric): a 0.15 mm thick cover fabric having a flow resistance of 2.3x10 8 m -1, an activated carbon layer 0.7 mm thick having a flow resistance of 4.4x10 6 m -1, and a 0.15 mm thick backing layer having a flow resistance of 3.0x10 5 m -1 (Schreuder-Gibson et al., 2002). The carbon loading was established as 120 g/m 2 which corresponds to a dry solid volume fraction of 0.09 in the model. The equilibrium capacity of the carbon is 0.47 g MeS /g C. A Dubinin-Radushkevich relationship was used for capacity down to P vap /P sat of 10-6 with a Henry s Law approach at lower pressures (Wood, 1992; Carbon Filtration, 1999). The cover fabric and backing layer were assumed to be nonabsorbing for the MeS vapor. A diffusion parameter (D ls /d f2 ) of 7.7x10-4 s -1 was used for the activated carbon model vapor to solid sorption (Barry and Hill, 2003). Transient simulations were performed at two flow rates: 0.3 L/min and 1.0 L/min. (At 0.3 L/min, the pressure drop across the swatch is approximately 0.1 water gage). The flow was assumed to be laminar throughout the flow domain as well as isothermal at 30 o C. At each time step, the model solves the transport equations and computes the evaporation rate from the liquid droplet, vapor transport from the droplet, and vapor sorption into the activated carbon layer. Gas Flow. Figure 5 shows contours of the velocity magnitude for the two flow rates. The contours of the higher flow rate appear to penetrate deeper into the flow tube than they do for the lower flow rate. (Note the different scales used for the contour levels. The scale for each represents the full range of velocity magnitude that exists for each case.) The figure also depicts the region for the close-up views shown later. Figure 6 provides the velocity vectors for the 0.3 L/min case in the vicinity of the droplet. Vapor Distribution and Penetration. Figures 7 to 9 show contour plots of P vap /P sat for the two flow rates at 60, 120, and 180 minutes after the start of the transient, respectively. Note 24 INJ Winter 2004

4 Air Inlet Droplet Figure 5 VELOCITY MAGNITUDE FOR TWO FLOW RATES; A) 0.3 L/MIN.; B) 1.0 L/MIN Cover Fabric Closeup region Fabric Sample L/min flow rate. At 1.0 L/min the contours in the region above the sample are notably compressed compared to 0.3 L/min due to the increased strength of gas flow relative to diffusion in the higher flow case. A reduction in the radial spread of the contours is also noted for the higher flow case. The higher flow appears to confine the vapor transport through the fabric sample to a smaller cross sectional area. Before 120 minutes at 1.0 L/min, the carbon under the droplet has reached capacity allowing vapor to pass through the sample. These results are consistent with prior experimental data that show faster breakthrough at higher convective air flow rates. The higher convective flow narrows the plume of agent vapor, hence spreading the vapor over a smaller part of the activated carbon and enabling a quicker local overload of the adsorption capacity. It will also result in more vapor passing through without being adsorbed at higher flow rates. that the contours are shown as a log scale and are clipped such that no color is indicated for values of P vap /P sat less than 1x10-6. At 30 o C, the saturation pressure for MeS is approximately 21 Pa. Therefore, P vap /P sat of 1x10-6 corresponds to approximately 2.1x10-4 ppm. The vapor evaporates from the droplet and diffuses into the gas stream above the fabric sample. The vapors are also carried by the flow through the sample where they are adsorbed in the activated carbon layer. The region of red contour levels under the droplet can be seen to become larger with time as the vapors are adsorbed and eventually reach the capacity of the carbon. Breakthrough concentrations of P vap /P sat greater than 1x10-6 are reached by the three-hour point for the 0.3 Activated Carbon Layer Backing Layer Figure 6 VELOCITY VECTORS IN THE VICINITY OF THE FABRIC SAMPLE AND DROPLET CLOTHED 3-D ARM WITH ACTIVATED CARBON Simulations were performed using the 3-D arm with hand and cuff closures (Figure 4) to investigate the effects of the closures at the wrist and elbow and the influence of an activated carbon layer in the garment. Closures are applied by assigning clothing properties to the cells spanning the gap between the outer clothing layer and the arm surface at both ends of the arm. Lack of closures was simply applied by assigning gas properties to these cells. All simulations use a steady 5 mph wind blowing in a direction parallel to the axis of the lower arm such that the species would be carried up the length of the arm. The incoming fluid is modeled as dry air at 300 K. Clothing fabric was modeled as a single layer with flow resistance of 2.3x10 8 m -1 and activated carbon properties similar to those used above. Again the properties of MeS were used for the agent. Without Activated Carbon. Figure 10 provides contour plots of P vap /P sat over the surface of the hand and arm for conditions without activated carbon in the clothing material. Note that the contours are presented on a log scale with the lowest contour representing P vap /P sat of 1x10-6. The top view in each figure shows the MeS concentration at the surface of the arm, and the bottom view shows the MeS concentration at the surface of the clothing. With no closure (Figure 10, the MeS vapor directly enters the gap between the clothing and the arm, and the concentration next to the arm is greater than it is 25 INJ Winter 2004

5 Figure 7 CONTOURS OF P VAP /P SAT AT 60 MINUTES: A) 0.3 L/MIN; B)1.0 L/MIN Figure 8 CONTOURS OF P VAP /P SAT AT 120 MINUTES: A) 0.3 L/MIN; B) 1.0 L/MIN Figure 9 CONTOURS OF P VAP /P SAT AT 180 MINUTES: A) 0.3 L/MIN; B) 1.0 L/MIN on the outer clothing. With a closure providing a flow restriction at the wrist, the vapor tends to bypass the air gap and flow along the outer length of the clothing (Figure 10. The concentration at the arm surface closely approximates the level at the outer clothing surface due to the lack of flow in the air gap. With Activated Carbon. Figure 11 presents similar contours of P vap /P sat for conditions for a fabric containing activated carbon. The condition without closures (Figure 11 appears quite similar to the no-carbon case presented above without closures. The MeS vapor is forced into the gap between the clothing and the arm by the incoming 5 mph flow and the car- 26 INJ Winter 2004

6 Arm Arm Wind Direction Wind Dire ction Clothing Su rf ace Vi ew Clothin g Surf ace Vi ew Figure 10 CONTOURS OF P VAP /P SAT WITHOUT ACTIVATED CARBON: A) WITHOUT CLOSURES; B) WITH CLOSURES Arm Surf ace Vi ew Arm Wind Direction Win d Directi on Clothing Cl othin g Figure 11 CONTOURS OF P VAP /P SAT WITH ACTIVATED CARBON: A) WITHOUT CLOSURES; B) WITH CLOSURES bon appears to have little effect on reducing the concentration at the arm surface. Inclusion of a closure has a dramatic effect on reducing the MeS concentration at the arm surface as seen in Figure 11b. The region covered by the clothing is clearly visible as a dark blue region in the upper view of Figure 15b where the activated carbon significantly reduced the presence of MeS in the gap between the clothing and the arm. HEAT AND SWEAT LOSS FROM CLOTHED TORSO Using the 3-D clothed torso model (Figure 2), the effects of multiple clothing layers and closures on the heat and sweat transfer from the body were evaluated. All simulations were performed with a 5 mph wind (300 K and relative humidity of 70%) blowing toward the front of the torso. The computations were steady-state and used a standard k-ε turbulence model. The clothing layers and gaps between the clothing and the torso were defined as laminar zones. The shirts and pants were modeled with the properties of a cotton fabric having a flow resistance ranging from 3.2x10 8 m -1 (0% relative humidity) to 6.25x10 8 m -1 (100% relative humidity). A constant surface sweat flux of 3.0x10-5 kg/m 2 -s (corresponding to moderate workload) is applied at the skin surface along with a convective boundary condition to simulate metabolic heat rejection to the skin. Thermal radiation is included (gray, discrete ordinates approach) with a very large absorption coefficient for the clothing layers to approximate opaque behavior. Effect of Closure for One Shirt. Two simulations were performed with a single-shirt model: With closures (snug fit) at sleeves, neck, and waist Without closures (loose fit) at those same locations Figure 12a shows the temperature distribution over the torso surface for the case with closures. The views shown are rear and front on the left and right, respectively. The region at 27 INJ Winter 2004

7 Figure 12 TEMPERATURE DISTRIBUTION AT SKIN SURFACE FOR ONE-SHIRT TORSO MODEL: A) WITH CLOSURES; B) WITHOUT CLOSURES Figure 13 SWEAT AT SKIN SURFACE FOR ONE-SHIRT TORSO MODEL WITH CLOSURES: A) ACCUMULATION RATE; B) EVAPORATION RATE the neck is cool since it is not covered by clothing. Local warm spots are visible down the small of the back, at the sternum, and under the arms. The temperature contours shown in Figure 12b correspond to the simulation without closure. These temperatures are clearly lower than those in Figure 12a. Note the local warm spot in the right upper chest region for the no-closure case. This warm spot is caused by a narrowing of the clothing-skin gap in the region reducing circulation of air in that location which is a result of natural asymmetries in the original body scan data and the shirt model. A local warm spot still exists under the arms for the no-closure case though it is not visible in the view shown. Figure 13 shows contour plots of the sweat accumulation and evaporation rate at the skin surface under conditions of closure. (For a constant sweat flux applied across the skin surface, one plot can be inferred from the other.) Relatively high rates of sweat accumulation exist under the arms, in the small of the back, and in the region of the sternum. Regions of high sweat evaporation rates are at the neck and at various localized regions across the torso surface. Under conditions of closure, the transport under the clothing layers is primarily by diffusion alone. Consequently, the localized regions of high evaporation rates correspond to locations in which the distance between the torso and clothing surface are smallest. Figure 14 provides the contours of sweat evaporation at the torso surface for the no-closure case. Over almost the entire surface, the sweat evaporation is equal to the sweating rate of 3.0x10-5 kg/m 2 -s. Only in regions in the center of the back, the center and right upper chest, and under the arms (not visible) is sweat evaporation less than the sweating rate. In these localized regions, the remaining portion of the sweat flux does not evaporate but is accumulating as a liquid. Effect of Second Shirt. Figure 15 shows the effect of adding a second shirt on the temperature and sweat evaporation. At the junction of the shirt(s) and pants, a clear transition is evident. In comparison to the temperatures for the one-shirt case of Figure 12a, the region of high temperatures under the arms also appears to be larger for the two-shirt case. Relative to the one shirt case shown in Figure 13b, the sweat evaporation is reduced by a factor of two or more over most of the upper 28 INJ Winter 2004

8 computed from the heat transfer values shown and the areas of each region (neck: 36.2 cm 2, upper torso: cm 2, lower torso: cm 2 ). Implementation of closures and the addition of a second shirt have a clear effect of decreasing the overall heat transfer. Adding these effects also shifts more of the heat transfer to thermal radiation in the clothed regions due to the resultant increase in torso surface temperatures. The effect of decreased evaporation heat transfer is also evident with reduced ventilation and a second shirt. Figure 14 SWEAT EVAPORATION AT SKIN SURFACE FOR ONE-SHIRT TORSO MODEL WITHOUT CLOSURES torso. As for the one-shirt case with closure, mass transport from the torso surface to the outer surfaces of the clothing is dominated by diffusion due to the very small velocities in these stagnant zones. The second shirt layer is not only a transport barrier itself, but it creates another stagnant layer over the first shirt that reduces the ability of the sweat vapor to be removed from the body. A variety of alternative conditions could be considered with this model by selectively applying closures to flow through the first layer only or the second layer only. Heat Transfer. Integral heat transfer was analyzed by separating the torso into three regions: 1) upper torso (including the arms) beneath the shirt(s), 2) lower torso beneath the pants, and 3) neck. For each case, the contributions to overall heat transfer by each of the three modes (conduction/convection, radiation, and evaporation) were computed. Tables 1 to 3 provide a summary of the heat transfer from each of these regions for the three cases considered. Average fluxes were Conclusion In the development of protective clothing and other textiles, modeling offers a powerful companion to experiments and testing. Detailed models for vapor and liquid phase transport within textile fabrics have been developed and integrated with CFD software. Validation of the models with experimental data has been successful for moisture absorption, permeability, and wicking (Barry and Hill, 2003). Applications of the software thus far have included analysis of chemical penetration of garments due to wind, investigation of flow in swatch testing equipment, effects of seal leakage on protective clothing performance, evaporation and adsorption of vapors with activated carbon, and heat and sweat loss from clothed partial bodies. Future applications could involve more extensive assessment of thermal comfort/stress on wearers of protective clothing, effects of layering on protective performance, and sensitivity to textile permeability and wicking properties. New opportunities for validation and application of the modeling tools are sought. Acknowledgements This work was supported by the U.S. Army Soldier Biological Chemical Command under contract DAAD C References Barry, J.J., Hill, R.W.; Computational Modeling of Protective Figure 15 SKIN SURFACE CONDITIONS FOR TWO-SHIRT TORSO MODEL WITH CLOSURES: A) TEMPERATURE; B) SWEAT EVAPORATION RATE Shirt/pants junction Shirt/pants juncti on 29 INJ Winter 2004

9 Table 1 HEAT TRANSFER FROM TORSO: ONE SHIRT AND NO CLOSURES Table 2 HEAT TRANSFER FROM TORSO: ONE SHIRT WITH CLOSURES Table 3 HEAT TRANSFER FROM TORSO: TWO SHIRTS WITH CLOSURES Clothing; International Nonwovens Journal, Vol. 12, 2003, No. 3, pp Barry, J. J., Hill, R. W., Brasser, P., Sobera, M., Kleijn, C. and Gibson, P., Computational Fluid Dynamics Modeling of Fabric Systems for Intelligent Garment Design, MRS Bulletin, Vol. 28, No. 8, 2003, pp Carbon Filtration for Reducing Emissions from Chemical Agent Incineration, National Academy Press, Washington D.C., (CAESAR data samples) Gibson, P. W., Multiphase Heat and Mass Transfer Through Hygroscopic Porous Media with Applications to Clothing Materials, Technical Report Natick/TR-97/005, U.S. Army Natick Research, Development and Engineering Center, Natick, MA, Gibson, P. W., Kendrick, C., Rivin, D., Sicuranza, L., Charmchi, M., An Automated Water Vapor Diffusion Test Method for Fabrics, Laminates, and Films, Journal of Coated Fabrics, Vol. 24, pp , Gibson, P. W., Governing Equations for Multiphase Heat and Mass Transfer in Hygroscopic Porous Media with Applications to Clothing Materials, Technical Report Natick/TR-95/004 (U.S. Army Natick Research, Development, and Engineering Center, Natick, MA, 1994). Schreuder-Gibson HL, Gibson P, Rivin D. Direct incorporation of granular carbon into melt blown thermoplastic polyurethane an approach for lightweight chemical protective fabrics. Presented at the 12th International Nonwovens Conference, November 20, 2002, Knoxville, TN. Whitaker, S., in Advances in Heat Transfer, Vol. 13, edited by J. Hartnett, (Academic Press, New York, 1977) p Whitaker, S., in Advances in Heat Transfer, Vol. 31, edited by J. Hartnett, (Academic Press, New York, 1998) p. 1. Wood, G.O., Activated Carbon Adsorption Capacities for Vapor, Carbon, Vol. 30, 1992, pp ). INJ 30 INJ Winter 2004