Protective ensembles are essential for protecting workers against dermal exposure to aerosols that contain potentially toxic particles. Currently, a standardized methodology that examines the barrier effectiveness of a full protective clothing ensemble to particulate hazards does not exist. Previous studies on particle penetration through protective clothing materials have most commonly been conducted with a standardized respiratory filter tester that does not reflect ensemble usage conditions,^[1,2] such as the TSI model 8130^[3] and, more recently, the TSI model 3160.^[4] However, in the workplace, especially in the outdoor environment, penetration is primarily driven by wind and body movement.^[5-8] In addition, the wearer behind the protective clothing acts as a blunt body, causing a change in airflow streamlines^[8,9] and, thus, the approaching particle trajectories^[5,8,10] would behave according to the streamlines, affecting particle penetration through the ensemble. To address this, Hill and colleagues^[10] designed a system to measure the barrier protection of woven garments worn by naval workers against particles driven by ambient wind conditions. They tested the penetration of particles between about 2 μm and 20 nm delivered from 9–37 m/sec to a fabric draped around a vertical tube with dimensions similar to a human arm. The system supported the test fabric with a screen and provided an annular region that approximated the air space that exists between loosely fit PPE and the skin. Using measurements at a single point normal to the incoming wind direction, the authors reported a maximum particle penetration size of about 1 μm with the penetration ratio increasing at higher ambient wind speeds. However, results may differ for indoor air exposures where wind speeds are much lower. In a survey of indoor workplaces, Baldwin and Maynard^[11] reported that 85% of 55 occupational environments had much lower background wind speeds (less than 0.3 m/sec). Their results include testing an entire suit system for leakage through seams, closures, areas of transition to other protective equipment, and from movement and activities. A system-level aerosol test (i.e., aerosol man-insimulant test [MIST]), similar to the chemical MIST^[12] is currently non-existent. Such a test would require 30 passive aerosol dosimeters (approximately 25 × 35 × 2 mm) placed at different locations throughout and under the protective clothing layer worn by a human subject.

Thus, a multidomain magnetic passive aerosol sampler (MPAS) was designed to enable an aerosol MIST.^[13,14] The MPAS generates a magnetic force to collect paramagnetic iron (II, III) oxide (Fe³O⁴) particles as the challenge aerosol. The magnetic force of the MPAS provides a much higher collection efficiency than a conventional passive sampler, which allows collecting a sufficient quantity of particles within the time frame of a MIST procedure.^[13] The multidomain arrangement allows magnetic force to dissipate within a small distance between the sampler opening and the substrate to avoid overestimating penetration levels caused by pulling particles from the outside of ensembles, and significantly improves the uniformity of magnetite aerosol particle deposition across the substrate. Particle deposition across the substrate was shown to have a consistent pattern which enables measuring reproducible particle numbers using a randomized counting method.^[14] The objectives of this study were to develop a benchtop particle penetration cell (i.e., MPAS holder and shroud), to calculate penetration ratios (active/passive), and to correlate penetration ratios with physical characteristics of four fabrics (Table 1). Fabrics in this study were selected by structural feature and are not necessarily typical of protective clothing.

Airflow differences without (Figure 4a) and with (Figure 4b) the shrouded unit show that no significant air recirculation was present at the exhaust region of the SP-Cell. Downstream turbulence was observed without the shroud; while a smoother smoke plume was observed with the shroud.

Numerical simulation showed that the screen of the SP-Cell created a pressure differential that oriented velocity vectors more directly toward the MPAS (Figure 5). As airflow approached the SP-Cell, velocity decreased and vectors upstream diverged. Airflow was generally smooth near the screen and symmetrical around the SP-Cell for a majority of air flows around the SP-Cell. Inside the SP-Cell, velocity vectors were oriented toward the outlet on the downstream side of the MPAS, and air velocity was about 500-fold lower than wind speed. From this, it was assumed that the SP-Cell with the shrouded exhaust unit would be able to prevent particle re-entrance to the MPAS after exiting the exhaust outlet.

The effective face velocity (V_f) driven by ambient wind speed showed an increase in pressure drop with V_f and substantially different pressure drops across each fabric and between fabric model (Figure 6).

Particle penetration through a representative fabric model (I) using the SP-Cell at the minimum and maximum tested face velocities is shown (Figure 7a). This figure does not indicate a “Most Penetrating Particle Size” (MPPS). The fractional penetration value (S/C) is computed by dividing the sample (S) penetration by control (C) penetration values for each particle size. Comparison between the passive (SP-Cell) and active sampling shows that penetration increases with increasing face velocity for both methods, and penetration for the active method is substantially greater for all conditions (Figure 7b). The active to passive ratio (A/P) varied between 1.5 and 5.5 and generally decreased with an overall increase in penetration, suggesting that it likely overestimated particle penetration by forcing particles through that which would have passed by naturally.

The difference between methods may be partially explained by differences in penetration as a function of the fabric’s physical parameters (Figure 8). Although the results are not direct comparison of each parameter (i.e., all four qualities change in each fabric) it appears that particle penetration through the SP-Cell (passive method) increased with larger fiber diameter (all fabrics), pore area (except for Fabric IV), and air permeability (except for Fabric IV), but decreased with fabric thickness (except for Fabric III). Most of these results are expected. For example, it is expected that the greater surface area provided by smaller fibers, and the flow restriction by smaller pores would reduce particle penetration. Additionally, it would be expected that penetration would increase with a higher air permeability, since permeability is positively associated with increases in air velocity and air pressure. The results of Fabric IV having the lowest penetration, but highest air permeation, may be explained in that it is the thickest, has the smallest fiber diameter, and the second smallest pore area, providing a long tortuous path for particles to pass. The fabric is used as insulation for fire fighter ensembles, and is designed to provide breathability and prevent particle penetration. In contrast Fabric III, which had the second lowest penetration overall, did not have the second lowest penetration as a function of pore size. This is likely because it was a four-layered structure, with the two inner layers reinforced with yarn, providing supplemental protection from particles, and effectively reduced its penetration with respect to pore area.

Although not entirely what we expected, in that a dominant MPPS does not result for all fabrics at all face velocities, these results may reasonably represent penetration that occurs through protective clothing worn by workers. For example, 1–.5 m/sec winds likely predominantly pass around the human body, which acts as a blunt body. A small fraction of air may penetrate through the worn clothing, potentially transporting particles similar to these findings (e.g., for fabrics with similar weave, porosity, and pressure drop).

This study found that particle penetration measured by the SP-Cell (passive method) increased with increasing face velocity, while penetration obtained by the active method was nearly six times greater than that measured by the passive method. Generally, the ratio of active to passive penetration increased with decreasing permeability, and decreased with decreasing pore area for all four fabrics. Since particles are not filtered by worn ensembles as they are for respirators, but can penetrate by wind, these results suggest that the active method overestimated penetration. Fabric IV, used as insulation for fire fighter ensembles, was the thickest fabric with the smallest fiber diameter, which showed a much greater ratio between active and passive methods, likely because of the effects of its greater surface area and lower residence time on particle deposition. Smoke tests in the wind tunnel visually showed that the penetration cell with a shrouded exhaust unit was able to minimize possible eddy effects downstream of the sampler. The results suggest that the shrouded exhaust unit could reduce the potential for particle re-entrainment from the exhaust of the SP-Cell, and thus, avoid overestimations of particle penetration. Limitations of this study include the use of two different lab test aerosols that are not commonly exposed to workers. The quantification of particles by this method was time consuming and a challenge to conduct. Additional work should explore a rapid and easy method for counting particles. A follow-up study would apply the MPAS to an aerosol MIST test by first evaluating its performance of measuring particle penetration through full-ensemble garments worn by a manikin in an aerosol wind tunnel of adequate size. This would be paired with measurements made using an SMPS on the inside and outside of a swatch wrapped around a screened tube housed in various sections of the manikin.^[17]