Performance evaluation of personalized ventilation personalized exhaust (PV PE) system in air conditioned healthcare settings 4

Chapter 4: Preliminary Studies 4.1 Introduction The objective of the preliminary studies is to evaluate and verify the feasibility of the novelPersonalised Ventilation - Personalised Exhaust (PV-PE) devices. There are quite a few variations influencing the design of PV-PE system, such as the position of PE, the pressure of PE, background ventilation type, PV ATDs type etc. Before building the actual experimental facilities, the preliminary research involving a few parametric variation studies was envisaged to provide ideas to make the experimentalset-up more optimal.The preliminary research comprised two pilot studies. Pilot study I was aimed to evaluate the feasibility of PV-PE system using Computational Fluid Dynamics (CFD) simulation. It was found that the PV-PE system was able to divert the PV fresh air profile and exhaust the exhaled contaminated air before it mixed with the room air. In Pilot study II, three different types of PE devices were developed and compared using a CFD model. The results show that the top-PE and shoulder-PE could have a better performance in practice. 4.2 Pilot study I - Feasibility of the novel PV- PE system 4.2.1 Research Methodology CFD simulation was used to evaluate parametric variations and to find the feasibility of the novel PV- PE system. 4.2.2 Geometry and Grid Two human beings sitting at two sides of a desk in a consultation room were simulated as numerical manikins using CFD. The dimension of the simulated consultation room is m (length) x m (width) x 2.6 m (height). The two manikins were sitting face to face with a m distance between each other. One represented an Infected Manikin (Infected Person) that acts as a source of contaminated exhaled air and the other manikin was assumed to be a Healthy Manikin (Healthy Person) that inhales the contaminated air and acts as a sink. The geometry of the computer simulated thermal manikin used in the pilot study is obtained from physical thermal manikin by using a three-dimensional laser scanning technique (Gao& Niu, 2006). It is a real and accurate representation of a nude seated female occupant with the surface area of 1.57 m2. A pair of newly conceptualized localized exhaust devices, termed the Chair Personalised Exhaust (PE) devices, was placed at the upper part of the chair and just behind the human head, with a dimension of 0.08 m x 0.08 m. Room geometry was divided into two sections: lower part of the room enclosing the two manikins, with 2,863,045 unstructured tetrahedral cells around the manikin, and the remaining upper room volume with 1,496,069 cells. There are two layers of uniform boundary layer cells placed around the manikin’s body with size of 0.0022 m. Enhanced wall treatment was applied in the simulation. As shown in Figure 4.1, the value Y+ was approximately in the range of 0.5-1.8 for most part across the body surface and around for small part at face area. This is acceptable since in practice, an y + up to 4-5 is considered acceptable as it is still inside the viscous sub-layer. Figure 4.1 Contours of wall Y+ 4.2.3 Turbulence model and boundary conditions The indoor air flow patterns was simulated using standard k-epsilon model, which has been used in previous numerical research (Gao & Niu, 2004; Pantelic, 2010) to simulate indoor air flow based on the assumption that this model is capable of simulating convective heat transfer of buoyancy-driven air flow as long as a reasonable value of Y+ is achieved. Energy equation was activated. Continuity equation and momentum equations were solved to obtain velocity distribution. Species transportation equation was activated to study the personalized air distribution and exhaled air spread. The SIMPLE algorithm is used to couple the pressure and velocity fields. Considering the convection and diffusion, second order upwind was used to solve the momentum and energy equations. PRESTO was used for pressure. The room air was defined as incompressible air since the speed of room air flow was insignificant compared to the speed of sound of the fluid medium. As shown in Figure4.2, a four way ceiling supply air diffuser, demonstrated to be able to predict the air flow pattern accurately (Cheong et al, 2001), was modelled for the MV inlet since it is the most commonly used diffuser for MV. In addition, Under-Floor Air Distribution (UFAD) system was also considered. The UFAD outlet with a diameter of 200 mm is modelled in the simulation. Both the axial-velocity and tangentialvelocity were assigned in boundary conditions in order to represent the swirling flow from a floor-mounted circular diffuser. Detailed boundary conditions used in this study are listed in Table 4.1. Figure 4.2 Models for the four way air diffuser Table 4.1: Detailed boundary conditions in pilot simulation study Turbulence model Standard k–epsilon model Number of cells used in the Fluent model Mixing ventilation inlet 3,763,604 cells Velocity inlet; T = 23 °C; I = 10% UFAD ventilation inlet Room air outlet Velocity inlet; flow rate=20l/s ; I = 20%; D=0.2 Pressure outlet; Gauge pressure=0 Pa Room wall Adiabatic wall Manikin body T =34°C Mouth L = 8.4 l/min; I = 0.5%; D = 0.02 m RMP ATD Velocity inlet; I =10%; D =0.12 m VDG ATD Velocity inlet; I =10%; D =0.04 m Personalized exhaust outlet Pressure outlet 4.2.4 Parametric Variation Studies Two sets of CFD simulation were carried out. The first set was conducted to evaluate the influence on PEE by adding a Personalized Exhaust (PE) working together with PV. In this simulation, both the manikins keep inhaling at a velocity of 8.4 l/min. Two different kinds of PV air terminal devices (ATD) were simulated: Round Moveable Panel (RMP) and Vertical Desk Grill (VDG). Two different background air supply systems were simulated: ceiling supply Mixing Ventilation (MV) system (Figure 4.3) and Under-Floor Air Distribution (UFAD) system (Figure 4.4). Detailed conditions used in this study are listed in Table 4.2. Figure 4.3 Configuration of the simulated office with mixing ventilation (1-MV four- way inlet; inlet d=200 mm; 2-MV outlet 500x500 mm; 3-RMPd=120 mm; 4-PE 80x80 mm; 5-VDG 220x20 mm; 6-Numerical manikin) Figure 4.4 Configuration of the simulated office with UFAD ventilation (1UFAD inlet d=200 mm; 2- UFAD outlet 500x500 mm; 3-RMP d=120 mm; 4-PE 80x80 mm; 5-VDG 220x20 mm; 6-Numerical manikin) Table 4.2: Detailed ventilation combinations studied in Set simulation Backgr ound air supply Mixing Ventilation Workst ation Configu ration RMP +PE PV flow rate (l/s) Chair PE device Gauge Pressur e (Pa) -10; -30; -50 12 UFAD ventilation VDG +PE 16 -10; -10; -30; -30; -50 -50 12 -10; -10; -30; -30; -50 -50 RMP+PE 16 -10; -30; -50 -10; -30; -50 12 VDG+PE 16 -10; -10; -30; -30; -50 -50 12 -10; -10; -30; -30; -50 -50 16 -10; -30; -50 In order to evaluate the ability of personalized exhaust (PE) system in preventing the spread of contaminated air exhaled by infected people, another set of CFD simulation was conducted. The dimensions of the simulated room were the same. One manikin (Healthy Manikin) keeps inhaling at a velocity of 8.4 l/min through mouth (20 × 10 mm) and the other manikin (Infected manikin) keeps exhaling at the same velocity through mouth (20 × 10 mm). Tracer gas was introduced in the exhaled air. Based on Set I simulation results, the combination of MV and Vertical Desk Grill (VDG) has the lowest PEE. Set II simulations chose this worst case combination as the ventilation combination to further studies. Detailed conditions used in this study are listed in Table 4.3. Table 4.3: Detailed ventilation combinations studied in set simulations Ventilat ion Mixing Ventilation +VDG combin ation PV flow rate (l/s) Chair PE device Gauge Pressur e (Pa) -10 -30 12 -50 -80 -10 -30 16 -50 -80 -10 -30 -50 -80 4.2.5 Evaluation Index Two indices, i.e. Personalised Exposure Effectiveness (PEE) and Intake Fraction (iF) were used to assess the performance of personalized exhaust system which has been discribed in Chapter 3. 4.2.6 Results and Discussion - Ability to change the PV air direction and profile The first objective of the pilot study was to examine the advantages and feasibilities of the novel PV-PE system, which are mainly focused on the ability to pull the PV fresh air by the chair integrated PE devices. CFD simulation performed in Set I studies compared the PV air streamlines in each PV-background combination with PE and without PE. Figure 4.5 to Figure 4.8 show that there are great differences between each case. The streamlines in Figures 4.5 through 4.8 use the “Unique option” which gives each streamline a different color along its whole length, and can be used to track individual streamlines through the domain. The purpose of the figures is to show how the PE diverts the PV air. So only streamline from PV is shown. The streamlines from UFAD or MV inlet are not shown in these figures. When PV ATD (either RMP or VDG) was coupled with a PE device, almost 100% PV air went through the seated occupant towards the PE device, compared with only a small portion reaching the seated human being without PE device. After hitting the surface of the manikin in the face region, a fraction of the air flow may move up towards the ceiling exhaust grill. This fraction depends largely on the gauge pressure of the PE. (a) (b) (c) (a) No PE device (b) Gauge pressure at -10Pa (c) Gauge pressure at -50Pa Figure 4.5 Streamlines under the combination of VDG and MV (a) (b) (c) (a) No PE device (b) Gauge pressure at -30Pa (c) Gauge pressure at -50Pa Figure 4.6 Streamlines under the combination of RMP and MV (a) No PE device (b) Gauge pressure at -10Pa (c) Gauge pressure at -30 Pa (d) Gauge pressure at -50 Pa (e) Gauge pressure at -80 Pa Figure 4.12 Concentrations of exhaled air at the mouth of Healthy Manikin with a PV flow rate of l/s The inhaled tracer gas concentrations at the mouth of Healthy Manikin when PV air flow rate was set at 12 l/s are shown in Figure 4.13 (a) No PE device (b) Gauge pressure at -10 Pa (c) Gauge pressure at -30 Pa (d) Gauge pressure at -50 Pa (e) Gauge pressure at -80 Pa Figure 4.13 Concentrations of exhaled air at the mouth of Healthy Manikin with a PV flow rate of 12 l/s The inhaled tracer gas concentrations at the mouth of Healthy Manikin when PV air flow rate was set at 16 l/s are shown in Figure 4.14 (a) No PE device (b) Gauge pressure at -10 Pa (c) Gauge pressure at -30 Pa (d) Gauge pressure at -50 Pa (e) Gauge pressure at -80 Pa Figure 4.14 Concentrations of exhaled air at the mouth of Healthy Manikin with a PV flow rate of 16 l/s Figure4.15 illustrates the iF for all the cases in Table 4.3 for at the mouth of the Healthy Manikin. Generally, iF reduces with the increase of PV supply rate, which indicates that PV has the potential to protect people from inhaling pollutant air. Furthermore, for PE with gauge pressure of -80 Pascal, iF of all the three supply flow rates was about one order of magnitude lower than without PE. In terms of protecting the healthy manikin from pollutants exhaled by the polluting manikin, adding a PE device is much more effective than increasing the PV flow rate. This is because PE devices are able to exhaust the exhaled air directly before it mixes with the room air, as shown in Figure4.16. MV+ VDG 1.60E-­‐06   1.40E-­‐06   Intake fraction 1.20E-­‐06   1.00E-­‐06   8l/s   8.00E-­‐07   12l/s   6.00E-­‐07   16l/s   4.00E-­‐07   2.00E-­‐07   0.00E+00   No PE -10 -30 -50 -80 PE Gauge Pressure (Pascal) Figure 4.15 Comparison of iF at different Gauge pressure of PE Figure 4.16 Air streamlines showing exhaled air from Infected Manikin 4.3 Pilot study II - Evaluation of different Personalized Exhaust devices The first pilot study has examined the feasibility of Personalized Ventilation Personalized Exhaust (PV-PE) system and supports the idea of supplying more fresh air to a person as well as to exhaust the exhaled contaminated air directly around the Infected Person before it mixes with the room air in hospitals and healthcare centres. Thus, the location and design parameter of the PE devices can be further explored since it plays a major role in the distribution of air around the human body. In pilot study II, three different PE devices were developed, simulated and compared: a chair-PE, the same as used in Pilot study I; a top-PE, which is a round device above the human head; a shoulder-PE, which are two local exhaust devices installed at the chair, just above the shoulder level. The PV air terminal devices chosen in this study are VDG and Desk-top PV (DPV). The performance of the PV-PE system in regard to occupants’ inhaled air quality and the transmission of exhaled aerosols between two occupants was studied and investigated numerically by computational fluid dynamics. 4.3.1 Methodology Figure 4.17 illustrates the configuration of the simulated consultation room in a health centre, the same as in pilot study I. The origin of the coordinate system was selected at the centre of the room volume. The Infected Manikin was to simulate an Infected Person who is sitting below the ceiling supply diffuser. The other Healthy Manikin was to simulate a Healthy Person who is sitting below the ceiling return grill. DPV was put on the desktop at a distance of 550 mm from the manikin’s mouth and the Vertical Desk Grill was located 390 mm from the manikin’s mouth. Three kinds of newly conceptualized localized exhaust device were simulated. Top-PE was located 50 mm above the manikin’s head, Chair-PE and Shoulder-PE had a dimension of 80 mm x 80 mm. Table 4.4 shows the details of the simulated conditions in this study. PV flow rate was set at l/s or l/s. Figure 4.17: Configuration of the simulated consultation room (1-Mixing Ventilation four-way inlet; 2-Mixing Ventilation outlet 500x500 mm; 3-DPV 96x80 mm; 4-VDG 220x20 mm; 5-Top-PE d=120 mm; 6-Chair-PE 80x80 mm;7Shoulder-PE 80x80 mm) According to Holmgren et al. (2010), the size distribution of exhaled particles peaks at around 0.07 µm, and an additional broad and strong peak was found between 0.2 and 0.5 µm. Gao (2008) and He et al. (2011) found that the concentration profile of exhaled particles with diameter smaller than 0.8µm was similar to gas. Thus, in this study, the exhaled particles were simulated as gas and the exhaled air from the Infected Manikin was marked as contaminated air. The experimental results of Pantelic (2009) demonstrated that droplets concentrations measured in the breathing zone of a thermal manikin were similar with and without a normal breathing process. Therefore, the respiratory process of the Infected Manikin was simplified to constant exhalation with a flow rate of 8.4 l/min. The Healthy Manikinwas simulated as keeping inhalation at the same velocity through mouth (20 × 10 mm). Although the actual expiratory process is through the nose, the experimental results of Rim and Novoselac (2009) demonstrated that breathing of a sedentary manikin has small impact on the airflow field in the breathing zone as well as the occupant’s thermal plume. Table 4.4: Detailed ventilation combinations studied in Set simulation PV air terminal devices Vertical Desk Grill (VDG) Personali zed Exhaust device Top-PE Chair-PE PV flow rate (l/s) PE device Gauge Pressure (Pascal) -10; -20; -30 -10; -20; -30 -10; -20; -30 -10; -20; -30 -10; -20; -30 Desktop PV Devices(DPV) ShoulderPE Top-PE Chair-PE ShoulderPE -10; -20; -30 -10; -20; -30 -10; -20; -30 -10; -20; -30 -10; -20; -30 -10; -20; -30 -10; -20; -30 CFD simulation was conducted using Fluent 6.3. Considering the complexity of combined buoyant flow, PV air flow and background air flow for the PV-PE study, it was difficult to find one turbulence transport model capable of resolving every aspect of the flows. Nielsen (1998) suggested having a compromise to select one turbulent model for the room airflow prediction. Similar to the pilot study I, standard k-ɛ model was adopted for the Pilot Study II. Energy equation was activated. Continuity equation and momentum equations were solved to obtain velocity distribution. The convection and diffusion term for all variables were the same as described in pilot study I. The boundary conditions applied in the simulation are shown in Table 4.5. The supply air temperature was 26° C for mixing ventilation system and 23° C for personalized ventilation. The exhaled air was set to be 34°C. Table 4.5: Detailed boundary conditions in simulation Turbulence model Number of cells used in FLUENT model Mixing ventilation inlet Standard k–epsilon model 4,605,081 cells Velocity inlet; T = 23 °C; I = 10% Room air exhaust Personalized exhaust Room wall, floor and ceiling Manikin body Mouth DPV air terminal device VDG air terminal device Chair Pressure outlet; Gauge pressure=0 Pa Pressure outlet Adiabatic wall T =34 °C L = 8.4 l/min; I = 0.5%; D = 0.013 m Velocity inlet; I =10%; D =0.087 m Velocity inlet; I =10%; D =0.04 m Adiabatic wall 4.3.2 Results and Discussion The purpose of PV is to achieve the highest possible quality of the air inhaled by occupants by providing clean air at the breathing zone. In order to evaluate the improvement of general inhaled air quality after adding different types of PE devices, Personalized Exposure Effectiveness (PEE) was used as an evaluation index. Figure 4.18: Comparison of PEE under different types of PE device tested. a) DPV supplies PV air at l/s, b) DPV supplies PV air at l/s, c) VDG supplies PV air at l/s, d) VDG supplies PV air at l/s Figure 4.18 compares the PEE obtained with the addition of tested PE devices under different PV air terminal devices. The performance of the PE devices was different and it changed with PE gauge pressure. The results show that for all the four cases, the PEE decreases with the decrease of PE gauge pressure for chair-PE device, especially when DPV was used. The largest reduction amount is about 27% when DPV was supplying PV air at a flow rate of l/s. PEE remains unchanged if a Top-PE device is added and it increases when a Shoulder-PE device is added. The largest increased amount is around 20% when VDG was supplying PV air at l/s. In order to compare the performance of different kinds of PE devices on the Healthy Manikin’s exposure to exhaled contaminated air, Intake Fraction was adopted. Figure 4.19 illustrates the iF of all the PV-PE configurations.The results of this investigation showed that for the same flow rates studied (4 l/s or l/s) all the types of the tested PE devices were able to reduce the iF, which indicates that PE has the potential to protect people from inhaling contaminated air. Considering the energy usage, draught discomfort, and the potential noisy level, a higher PE gauge pressure is always preferred without the compromise of ability to protect Healthy Person (the same decreased amount of iF). Based on this, for DPV-PE system at the same flow rate of PV air, top-PE performs best, followed by chair-PE, and lastly, the shoulder-PE device. For VDG-PE systems, in order to reduce the same amount of iF, the highest gauge pressure that could be applied was achieved by the top-PE device, followed by the shoulder-PE device, and lastly the chair-PE device. Figure 4.19 Comparison of iF under different tested PE devices when. a) DPV is used, b) VDG is used Generally, for both VDG and DPV, top-PE device has much better performance in terms of reducing iF for the Healthy Manikin. For the same PE gauge pressure, the iF is one to two orders of magnitude lower when top-PE was applied compared with chair-PE device and shoulder-PE device. The minimum iF of 1.2e-6 is achieved by the top-PE with a gauge pressure of -30 Pascal under VDG supplying PV air at l/s. 4.4 Conclusions from Preliminary Studies The main conclusions of the first pilot study are as follows: Firstly, using PE could not increase the overall percentage of personalized air in inhaled air with MV based systems. Instead, it has the potential to decrease PEE when high PV flow rate is used. For UFAD based systems, both the combinations of RMPPE and VDG-PE are seen to increase PEE by more than 20%. Secondly, Intake Fraction by the Healthy Manikin can be one order of magnitude smaller by adding a PE device for both Infected Manikin and Healthy Manikin. Intake Fraction is also influenced by the PV supply flow rate. The higher the flow rate is the lower is the Intake Fraction. Adding a PE device is much more efficient in terms of protecting the Healthy Manikin from contaminated air than by increasing the PV air supply flow rate. Lastly, the pressure of PE is an important factor. Both the reduction of PEE with MV and Intake Fraction (iF) depend largely on PE pressure. A lower pressure is recommended in order to get a better inhaled air quality. In the pilot two study, the performance of three PE devices for a PV-PE system was studied in regard to a Healthy Person’s inhaled air quality and exposure to contaminated exhaled air from an Infected Manikin. The following main conclusions can be drawn: The inhaled air quality provided by the PV-PE system is influenced by the type of PE devices. PEE has a potential to increase, unchanged or decrease depending on the location of PE device. Shoulder-PE and top-PE devices are recommended since they will not compromise the ability of supplying outdoor conditioned air to the occupants. Intake Fraction is seen to decrease with addition of any type of PE device. In terms of achieving the lower Intake Fraction with higher gauge pressure, top-PE device has the best performance in all the simulation cases. In general, it may be expected that one of the tested air terminal devices, namely topPE, will perform well in practice since they will not affect occupants’ general inhaled air quality and will protect a Healthy Person from contaminated exhaled air effectively. 4.5 Discussions about further experimental study Based on the results of the preliminary studies, the PV-PE system is observed to be feasible and applicable. The top-PE and shoulder-PE is worth exploring and to be built-up for further experimental research. Since the size of simulated consultation room and the locations of supply and exhaust diffusers are chosen based on the general design of a consultation room in healthcare centres and hospitals, the chamber to be used for experiments was carefully chosen to get a similar room size and ventilation systems. The Indoor Environmental Chamber with a dimension of 6.6 m (L) × 3.7 m (W) × 2.7 m (H) was selected for further experimental research.   [...]... Figure 4. 15 Comparison of iF at different Gauge pressure of PE Figure 4. 16 Air streamlines showing exhaled air from Infected Manikin 4. 3 Pilot study II - Evaluation of different Personalized Exhaust devices The first pilot study has examined the feasibility of Personalized Ventilation Personalized Exhaust (PV- PE) system and supports the idea of supplying more fresh air to a person as well as to exhaust. .. PV air in front of the face instead of reaching the breathing zone For UFAD systems, PEE concentration distributions increase when air flow rates increase as well For the same air flow rate, RMP has a better performance than VDG in terms of PEE PEE shows an obvious increasing trend by adding a PE device For both RMP and VDG, PE device enlarged the concentration of personalized air in the breathing... in the simulation are shown in Table 4. 5 The supply air temperature was 26° C for mixing ventilation system and 23° C for personalized ventilation The exhaled air was set to be 34 C Table 4. 5: Detailed boundary conditions in simulation Turbulence model Number of cells used in FLUENT model Mixing ventilation inlet Standard k–epsilon model 4, 605,081 cells Velocity inlet; T = 23 °C; I = 10% Room air exhaust. .. quality of the air inhaled by occupants by providing clean air at the breathing zone In order to evaluate the improvement of general inhaled air quality after adding different types of PE devices, Personalized Exposure Effectiveness (PEE) was used as an evaluation index Figure 4. 18: Comparison of PEE under different types of PE device tested a) DPV supplies PV air at 4 l/s, b) DPV supplies PV air at... Figures 4. 9 and 4. 10 display the changes of Personalized Exposure Effectiveness (PEE) with the change of gauge pressure of PE From the two figures, it can be deduced that the percentage of personalized air in inhaled air has a different trend with mixing ventilation and with UFAD ventilation In the following analysis, the profile of PEE will be discussed separately together with different background ventilation. .. was supplying PV air at a flow rate of 4 l/s PEE remains unchanged if a Top-PE device is added and it increases when a Shoulder-PE device is added The largest increased amount is around 20% when VDG was supplying PV air at 4 l/s In order to compare the performance of different kinds of PE devices on the Healthy Manikin’s exposure to exhaled contaminated air, Intake Fraction was adopted Figure 4. 19 illustrates... and shoulder-PE device The minimum iF of 1.2e-6 is achieved by the top-PE with a gauge pressure of -30 Pascal under VDG supplying PV air at 4 l/s 4. 4 Conclusions from Preliminary Studies The main conclusions of the first pilot study are as follows: Firstly, using PE could not increase the overall percentage of personalized air in inhaled air with MV based systems Instead, it has the potential to decrease... in terms of protecting the Healthy Manikin from contaminated air than by increasing the PV air supply flow rate Lastly, the pressure of PE is an important factor Both the reduction of PEE with MV and Intake Fraction (iF) depend largely on PE pressure A lower pressure is recommended in order to get a better inhaled air quality In the pilot two study, the performance of three PE devices for a PV-PE system. .. ability of supplying outdoor conditioned air to the occupants Intake Fraction is seen to decrease with addition of any type of PE device In terms of achieving the lower Intake Fraction with higher gauge pressure, top-PE device has the best performance in all the simulation cases In general, it may be expected that one of the tested air terminal devices, namely topPE, will perform well in practice since... studied in regard to a Healthy Person’s inhaled air quality and exposure to contaminated exhaled air from an Infected Manikin The following main conclusions can be drawn: The inhaled air quality provided by the PV-PE system is influenced by the type of PE devices PEE has a potential to increase, unchanged or decrease depending on the location of PE device Shoulder-PE and top-PE devices are recommended since . the influence on PEE by adding a Personalized Exhaust (PE) working together with PV. In this simulation, both the manikins keep inhaling at a velocity of 8 .4 l/min. Two different kinds of PV air. -10; -30; -50 In order to evaluate the ability of personalized exhaust (PE) system in preventing the spread of contaminated air exhaled by infected people, another set of CFD simulation. concentrations of exhaled air at the mouth of the Infected Manikin for all the cases are the same as shown in Figure 4. 11 Figure 4. 11 Concentrations of exhaled air at the mouth of the Infected Manikin