CFD Investigation of Vehicle’s Ventilation Systems and Analysis of ACH in Typical Airplanes, Cars, and Buses

: The simulation of the ventilation and the heating, ventilation, and air conditioning (HVAC) systems of vehicles could be used in the energy demand management of vehicles besides improving the air quality inside their cabins. Moreover, traveling by public transport during a pandemic is a concerning factor, and analysis of the vehicle’s cabin environments could demonstrate how to decrease the risk and create a safer journey for passengers. Therefore, this article presents airﬂow analysis, air changes per hour (ACH), and respiration aerosols’ trajectory inside three vehicles, including a typical car, bus, and airplane. In this regard, three vehicles’ cabin environment boundary conditions and the HVAC systems of the selected vehicles were determined, and three-dimensional numerical simulations were performed using computational ﬂuid dynamic (CFD) modeling. The analysis of the airﬂow patterns and aerosol trajectories in the selected vehicles demonstrate the critical impact of inﬂow, outﬂow, and passenger’s locations in the cabins. The CFD model results exhibited that the lowest risk could be in the airplane and the highest in the bus because of the location of airﬂows and outﬂows. The discrete CFD model analysis determined the ACH for a typical car of about 4.3, a typical bus of about 7.5, and in a typical airplane of about 8.5, which were all less than the standard protocol of infection prevention, 12 ACH. According to the results, opening windows in the cars could decrease the aerosol loads and improve the low ACH by the HVAC systems. However, for the buses, a new design for the outﬂow location or an increase in the number of outﬂows appeared necessary. In the case of airplanes, the airﬂow paths were suitable, and by increasing the airﬂow speed, the required ACH might be achieved. Finally, in the closed (recirculating) systems, the role of ﬁlters in decreasing the risk appeared critical.


Introduction
Many researchers have analyzed the efficiency of air conditioners (AC), particularly the vehicle producers, to reduce energy consumption and greenhouse gas emissions [1]. Marshall et al. analyzed the thermal management strategies for the vehicle, and their results determined the high impact of HVAC and ventilation on the energy demands, especially in electric vehicles [2]. Suárez et al. analyzed the ventilation and the HVAC system in a railway vehicle with a CFD simulation. They modeled different scenarios in summer and winter, and the results established the methodology for this type of analysis [3].
The experimental analysis of Kale et al. showed the complex airflow inside buses with open windows and determined that the speed of airflow could be about 0.1 of the bus speed [4]. Mathai [5]. In another study, Gajewski analyzed the indoor air quality in a bus with fresh air during its journeys, focusing on the the outcomes more reliable as the results were not limited to any specific model of vehicles.
Therefore, this study aimed to analyze the cabin airflow pattern for three vehicles, including an airplanes, buses, and cars, and to investigate ACH by CFD simulation to improve IAQ. Moreover, the second goal of the study was to show the application of CFD simulation in improving design elements of HVAC systems such as the vent location, the AC loads, and the airflow rate, besides the passenger placement. Although the medical analysis and transmission of the virus are beyond the scope of the present study, it could be useful in decreasing the health risk as it predicts the aerosol trajectory in the vehicles.
In this study, the main factors in analysis of the HVAC systems in the three vehicles including cars, buses, and airplanes were first investigated. Second, different case studies were evaluated, and the average values of the dimension, the airflow speed, and the airflow rate (based on 14 case studies) of the vehicles were determined. Third, the geometry and boundary conditions of the models were defined and validated according to the selected case studies. Fourth, the validated models were used for airflow analysis and ACH. Finally, the results were compared to each other and some suggestions to decrease the health risk in the vehicle cabins were provided.

Materials and Methods
The CFD solver model is used to analyze the cabin's airflow pattern, ACH, and the trajectory of aerosols for three vehicles, including airplanes, buses, and cars. The analysis flowchart is presented in Figure 1.   The correlations among the health risk and the indoor parameters were approved, and the monitoring of the relative humidity (RH), CO 2 , and T was suggested for improving the IAQ [37]. The main factors that affect IAQ in the vehicles included the humidity percentage, CO 2 , T, the airflow rate and direction, the type of the utilized air filter [38][39][40], the seat arrangement, and the displacement of ventilation [19,41]. The main approaches for airflow pattern analysis include experimental measurements and CFD simulations [42,43]. In general, to simulate and estimate the parameters, there are several methods, including mathematical, statistical, and numerical techniques [44][45][46][47][48][49]. Thus, we used the numerical technique in this study.

Governing Equations
The main equations in the CFD analysis were Equations (1)- (3). Equations of motion can be used to predict the particle or aerosol trajectory in a discrete phase. The trajectory of a particle such as droplets could be predicted through particle force balance Equations (4) and (5). Continuity: Momentum: Energy: where: ρ: fluid density; p: pressure; T: temperature; u: fluid phase velocity u p : particle velocity; ρ p : particle density; Fx: additional forces; C D : drag coefficient µ: molecular viscosity; Re: relative Reynolds number F D u − u p : drag force per unit particle mass

Case Studies and Boundary Conditions
The three-dimensional models of cars, buses, and airplanes were developed according to similar case studies and the available information. To define the boundary conditions in our CFD models, we evaluated the boundary conditions in different case studies, and we selected the average values of dimension, airflow, and temperature for a typical car, bus, and airplane. The dimensions and airflow rates and speeds in five cars, four buses, and four airplanes, and the selected boundary conditions are presented in Table 1.
To define the boundary conditions in the second part of the analysis, discrete phase modeling, the average diameter and speed of the of the aerosol particles in a human sneeze was selected. The human respirational aerosol diameter was between 0 to 100 µm, on average 5 µm, and the aerosol speed in respiration was 0.12-1 m/s; and for a cough and a sneeze was between 4.2 to 11.2 m/s, on average 10 m/s with a duration of 0.5 s [50][51][52][53]. Therefore, the aerosol diameter of 5 µm with a speed of 10 m/s and duration of 0.5s was used for the boundary condition.

CFD Model Set-Up
The model details, including the geometry of the vehicles and locations of inlets/outlets, are shown in Figures 2-8. The models represent the entire cabins of the three vehicles with some passengers inside. We tried to generate the geometry of the vehicles' cabins accurately. The passengers' human bodies were designed with simplified details to decrease the uncertainty and error in the mesh, as suggested in previous studies [33,65].
To decrease the simulation error, we followed the suggested methodology and considerations in the previous case studies. The base mesh and set-up of the discrete model was done according to the recent studies on aerosol transmission [33,[66][67][68]. To solve the Navier-Stokes equations in the simulation model, the k−ε turbulence method was used as the performance has been verified in similar case studies [68][69][70][71][72]. For simulation, we used a detailed 3D-CFD model in the steady condition, and for the CFD analysis, we used the ANSYS Fluent package.                  Mesh sensitivity analysis is presented in Figure 9, and details of the final selected mesh is presented in Table 2. Mesh sensitivity analysis is presented in Figure 9, and details of the final selected mesh is presented in Table 2. Mesh sensitivity analysis is presented in Figure 9, and details of the final selected mesh is presented in Table 2.  The details of the solution set-up for the three models are presented in Table 3.   The details of the solution set-up for the three models are presented in Table 3. For the verification of the numerical analysis, the models were validated initially based on the velocity profile of the measurement data in the previous studies of similar case studies, and then the simulations were completed for the selected boundary conditions. The results of the validation are shown in Figure 10. The used case studies for the validation were as follows:

•
The airflow model of the car was validated based on the experimental studies of Khatoon and Kim [54] and Danca et al. [57].

•
The bus CFD model was validated based on the studies of Yang et al. [50] and Zhu et al. [19]. • The airplane CFD model was validated based on the experimental analysis by Shehadi et al. [7] and simulation of the Boeing 737 airplane by Talaat et al. [33].

Main Assumptions and Hypothesis in the CFD Modelling
The main assumptions in this study were as follows: • The movement of passengers and flight attendants in the bus and airplane were not considered in the analysis; • It was assumed that passengers were sitting upright with facing front; • The validation of the models by similar case studies would not affect the main results; • The average diameter of the respiration aerosol would not affect the main results; • The models were based on the HVAC system, with full fresh inflow, and the opening windows or recirculation system were not considered in this study.
• The airflow model of the car was validated based on the experimental studies of Khatoon and Kim [54] and Danca et al. [57].
• The bus CFD model was validated based on the studies of Yang et al. [50] and Zhu et al. [19].
• The airplane CFD model was validated based on the experimental analysis by Shehadi et al. [7] and simulation of the Boeing 737 airplane by Talaat et al. [33].

Main Assumptions and Hypothesis in the CFD Modelling
The main assumptions in this study were as follows: • The movement of passengers and flight attendants in the bus and airplane were not considered in the analysis; • It was assumed that passengers were sitting upright with facing front; • The validation of the models by similar case studies would not affect the main results; • The average diameter of the respiration aerosol would not affect the main results; • The models were based on the HVAC system, with full fresh inflow, and the opening windows or recirculation system were not considered in this study.

Cabin Airflow Analysis in the Selected Vehicles
In this section, the airflow path, the trajectory of contaminated aerosols, and the ACH in the selected vehicles were simulated. The simulations were based on the use of an HVAC system and were completed according to the defined boundary conditions. The cabin airflow patterns in the selected vehicles are shown in Figures 11-17.

Cabin Airflow Analysis in the Selected Vehicles
In this section, the airflow path, the trajectory of contaminated aerosols, and the ACH in the selected vehicles were simulated. The simulations were based on the use of an HVAC system and were completed according to the defined boundary conditions. The cabin airflow patterns in the selected vehicles are shown in Figures 11-17.
The analysis of the selected vehicles showed the critical impact of inflows' and outflows' locations on the airflow's patterns. Moreover, the results revealed the importance of passenger locations in the cabins. Therefore, it seems that the lowest infection risk based on the airflow paths could be in the airplane, and the highest risk in the bus could be due to the locations of inflows/outflows and the number of passengers.         The analysis of the selected vehicles showed the critical impact of inflows' and outflows' locations on the airflow's patterns. Moreover, the results revealed the importance of passenger locations in the cabins. Therefore, it seems that the lowest infection risk based on the airflow paths could be in the airplane, and the highest risk in the bus could be due to the locations of inflows/outflows and the number of passengers.

Trajectory of Contaminated Aerosols and ACH Inside Vehicle Cabins
In this section, to check the ACH in the three vehicles, the discrete model analysis

Trajectory of Contaminated Aerosols and ACH Inside Vehicle Cabins
In this section, to check the ACH in the three vehicles, the discrete model analysis was applied. The trajectory of contaminated aerosols in the vehicles are shown in The analysis of the selected vehicles showed the critical impact of inflows' and outflows' locations on the airflow's patterns. Moreover, the results revealed the importance of passenger locations in the cabins. Therefore, it seems that the lowest infection risk based on the airflow paths could be in the airplane, and the highest risk in the bus could be due to the locations of inflows/outflows and the number of passengers.

Trajectory of Contaminated Aerosols and ACH Inside Vehicle Cabins
In this section, to check the ACH in the three vehicles, the discrete model analysis was applied. The trajectory of contaminated aerosols in the vehicles are shown in Figures  18-20          According to the discrete models, the respiration particles in the car could be cir lated in all cabin environments with the HVAC system. In the bus, since the outflow w placed at the end, the respiration aerosols from the front would pass the entire cabin a According to the discrete models, the respiration particles in the car could be circulated in all cabin environments with the HVAC system. In the bus, since the outflow was placed at the end, the respiration aerosols from the front would pass the entire cabin and reach the end. In the airplane, there were separate outflows under each row, causing restriction of the airflow circulation. Thus, the contaminated respiration particles could have a movement about one-two raw forwards and backwards. In addition, the movement of the contaminated respiration aerosols after just 60 s demonstrated the high risk of an infection event during a few seconds without personal protective equipment (PPE) or during eating/drinking inside the cabins. The estimation of the ACH for the three vehicles is presented in Figure 24. reach the end. In the airplane, there were separate outflows under each row, causing restriction of the airflow circulation. Thus, the contaminated respiration particles could have a movement about one-two raw forwards and backwards. In addition, the movement of the contaminated respiration aerosols after just 60 s demonstrated the high risk of an infection event during a few seconds without personal protective equipment (PPE) or during eating/drinking inside the cabins. The estimation of the ACH for the three vehicles is presented in Figure 24. The graph shows the percentages of the remaining particles over time. The periods for which more than 99% of the particles escaped from the cabins were about 7, 8, and 14 min for the airplane, the bus, and the car, respectively. Therefore, the CFD model analysis determined the ACH for a typical car at about 4.3 (60 min/14 min), a typical bus at about 7.5 (60 min/8 min), and in a typical airplane at about 8.5 (60 min/7.1 min), all less than the minimum protocol of infection prevention at 12 ACH.

Conclusions
The analysis exhibited the critical impact of the inflow, the outflow, and the passengers' locations in the cabins. According to the results, the ACH in the three selected vehicles was less than the standard infection prevention protocols, with the lowest in the car at 4.3 and the highest in the airplane at 8.5. The analysis showed that the aerosols in the car could be circulated in all cabin environments with the HVAC system. In the bus, since the outflow was placed at the end, the respiration aerosols from the front would pass the entire cabin to reach the end. The circulation of the aerosols after just 60 s in the entire vehicle's cabin showed a high health risk even during a one-minute stay without PPE.
The analysis showed that the car's opening windows could decrease the contamination loads and improve the low ACH by the HVAC systems. However, for the buses, a new design for the outflow location or an increase in the number of outflows seems necessary. In the case of the airplane, the separate outflows under each row caused a restriction of the airflow circulation. Thus, the possible contaminated respiration aerosols could involve about one row forward and backwards. Since the airflow patterns were suitable in the airplane, it seems that by increasing the airflow rate, the required ACH might be achieved.
In conclusion, the CFD simulations showed that the lowest health risk could be in the airplane and the highest in the bus due to the location of airflows and outflows. Although the simulations in this study were performed for the selected vehicles with specific dimensions, since the boundary conditions are based on several case studies, the main achievements could be valid for similar vehicles.
For future studies, it seems that in the closed AC systems (recirculating), the role of filters in the health risk seems critical and recommended. Furthermore, the current simulation results could improve the indoor air quality in vehicle cabins, besides minimizing The graph shows the percentages of the remaining particles over time. The periods for which more than 99% of the particles escaped from the cabins were about 7, 8, and 14 min for the airplane, the bus, and the car, respectively. Therefore, the CFD model analysis determined the ACH for a typical car at about 4.3 (60 min/14 min), a typical bus at about 7.5 (60 min/8 min), and in a typical airplane at about 8.5 (60 min/7.1 min), all less than the minimum protocol of infection prevention at 12 ACH.

Conclusions
The analysis exhibited the critical impact of the inflow, the outflow, and the passengers' locations in the cabins. According to the results, the ACH in the three selected vehicles was less than the standard infection prevention protocols, with the lowest in the car at 4.3 and the highest in the airplane at 8.5. The analysis showed that the aerosols in the car could be circulated in all cabin environments with the HVAC system. In the bus, since the outflow was placed at the end, the respiration aerosols from the front would pass the entire cabin to reach the end. The circulation of the aerosols after just 60 s in the entire vehicle's cabin showed a high health risk even during a one-minute stay without PPE.
The analysis showed that the car's opening windows could decrease the contamination loads and improve the low ACH by the HVAC systems. However, for the buses, a new design for the outflow location or an increase in the number of outflows seems necessary. In the case of the airplane, the separate outflows under each row caused a restriction of the airflow circulation. Thus, the possible contaminated respiration aerosols could involve about one row forward and backwards. Since the airflow patterns were suitable in the airplane, it seems that by increasing the airflow rate, the required ACH might be achieved.
In conclusion, the CFD simulations showed that the lowest health risk could be in the airplane and the highest in the bus due to the location of airflows and outflows. Although the simulations in this study were performed for the selected vehicles with specific dimensions, since the boundary conditions are based on several case studies, the main achievements could be valid for similar vehicles.
For future studies, it seems that in the closed AC systems (recirculating), the role of filters in the health risk seems critical and recommended. Furthermore, the current simulation results could improve the indoor air quality in vehicle cabins, besides minimizing