- Email: [email protected]
Air flow simulation of HVAC system in compartment of a passenger coach
Accepted Manuscript Air flow simulation of HVAC system in compartment of a passenger coach Ali Aliahmadi, Morteza Abdolzadeh, Khosro Lari PII: DOI: Reference:
S1359-4311(16)33734-6 http://dx.doi.org/10.1016/j.applthermaleng.2017.05.086 ATE 10398
To appear in:
Applied Thermal Engineering
Received Date: Revised Date: Accepted Date:
28 November 2016 24 April 2017 16 May 2017
Please cite this article as: A. Aliahmadi, M. Abdolzadeh, K. Lari, Air flow simulation of HVAC system in compartment of a passenger coach, Applied Thermal Engineering (2017), doi: http://dx.doi.org/10.1016/ j.applthermaleng.2017.05.086
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Air Flow Simulation of HVAC System in Compartment of a Passenger Coach Ali. Aliahmadi1, Morteza Abdolzadeh2*, Khosro Lari1 1
Department of Mechanical Engineering, Graduate University of Advanced Technology, Kerman, Iran Energy Institute, Graduate University of Advanced Technology, Kerman, Iran
2
*
Corresponding author: Department of Mechanical Engineering, Graduate University of Advanced Technology, End of Haft Bagh Highway, Kerman, Iran, Email address: [email protected] [email protected]
Abstract: In this study, numerical simulation of airflow in the compartment of a passenger coach was carried out in summer. An experimental measurement was also conducted to verify the performance of the numerical model. The turbulent air flow in the compartment was supplied by the HVAC system of the coach. The numerical simulation in the compartment was performed in two cases: seated and slept manikins. The manikins were considered as heated manikins in all the simulations. The turbulent air flow was modeled using v2 f turbulent model. Two conditions in the compartment, including temperature and velocity comfort criteria were assessed in the compartment with and without the presence of the manikins. Moreover, some modifications were implemented in the compartment design to improve the comfort conditions around the manikin in the two cases. Results showed that due to the inappropriate design of the compartment, the thermal comfort conditions do not evenly distribute in the compartment and this issue makes the passengers uncomfortable. The outcome of the modifications showed that the air flow symmetrically enters to the compartment and provides better thermal conditions for the passengers seated/slept in the compartment. Keywords: Passenger coach, HVAC, Heated manikin, comfort
1
1. Introduction There are many passenger coaches in the railroads across the world which have weak air conditioning as the criteria for engineering design of the compartment have not been taken into account appropriately. This issue makes problems for even air distribution in the compartments of the passenger coaches and causes them to not provide an acceptable level of thermal comfort conditions and fresh air for passengers. Different parameters, including equipment as well as the location and geometry of inlet and outlet air registers in the compartments affect on the airflow distribution and heat transfer. These parameters can cause the passenger to feel uncomfortable in the compartment, because when the there is no fear air conditioning, the human body does not well dissipate its waste thermal energy to the surrounding air. The comfort air temperature in summer is between 23°C and 26°C and the comfort air velocity is between 0.2m/s and 0.8m/s [1]. These conditions should be sustained by the HVAC systems of passenger coaches. So far, many studies have been carried out to simulate the air flow distribution and assess the thermal comfort conditions in different enclosed spaces. For instance, Okland did a study on metro stations and carriages in London. They investigated the thermal comfort conditions as well as the air particulates in the studied systems. They realized that air temperature and humidity as well as particulate distribution, significantly affect the thermal comfort conditions of the passengers [2]. Berlitz and Matschke studied interior airflow simulation of a regional train. They simulated and measured the air flow properties in the train compartments [3]. Yongson et al. studied air flow analysis of an air conditioning room. They used k-ε turbulent model for turbulent air flow modeling. They investigated different locations of blower placement to analyze comfort of occupant in the room. They showed that the occupants will experience most comfort if the air conditioner blower is placed on the very corner of the side wall compared to the other locations [4].Khodakarami and Nasrolahi conducted a review about the thermal comfort of hospitals. They 2
concluded that it is important to find some solutions to reconcile the different thermal comfort conditions required by different occupants in hospitals [5]. Salmanzadeh et al. studied the buoyancy driven thermal plume near a heated manikin seated in a cubicle in turbulent flow using k-ε turbulent model. They showed that due to the thermal plume flow created by the temperature gradient adjacent to the body, a high air velocity is seen in the breathing zone of the manikin [6]. Gahever et al. studied thermal comfort of the surgical staff in the operating room. They summarized the technical standards and compared them to the thermal comfort standards and the data gathered on the site. They showed that significant discrepancies exist between the HVAC standards and the thermal comfort standard [7]. Yang et al. numerically studied transient natural ventilation driven by buoyancy. They showed that vent shape and the horizontal position of heat source have little effects on the ventilation [8]. Uścinowicz et al. studied thermal environmental conditions in polish operating rooms. They measured thermal conditions in 37 operating rooms and showed that operating staffs are different in thermal requirements [9]. Yang et al. numerically studied the airflow characteristics and thermal comfort conditions in buoyancydriven natural ventilation rooms. They investigated indoor thermal comfort during the transient ventilation. They showed that the initial indoor temperature may affect the steady state of natural ventilation. They also indicated that the vent shape and the horizontal position of heat source do not have a significant effect on the ventilation [10]. Arslanoglu and Yigit computationally and experimentally studied the effect of radiation heat flux from lighting lamps on human thermal comfort. They indicated that the upper part of the human body due to the lamp radiant heat flux is the most affected part of the body [11]. Wang et al. studied the thermal comfort characteristics under the vent with supplying air jets and cross-flows coupling in subway stations. They also studied relative warmth index to assess the thermal comfort conditions in the transitional
3
dynamic areas. They showed that the coupling airflows of supplying air jets and cross-flows can improve the transitional thermal comfort in air-conditioning systems [12]. Rullen et al. studied the thermal impacts of heat dissipation of electrical appliances on buildings by applying a dynamic thermal model of electrical appliances. Their results indicated the necessity of dynamic thermal modeling of electrical appliances for more accurate and efficient building energy simulation [13]. Ansaripour et al. studied air flow distribution and particle concentration emitted from a Laserjet printer in a ventilated room. They used v2-f turbulent model to predict the turbulent air characteristics. Their results showed that when the printer is located on the front side of the manikin, the particle concentration in the breathing zone is quite high in most of the used ventilation configurations [14]. Ahmed et al. designed and assessed a new local exhaust ventilation system in a room. They showed that the proposed system improved the thermal conditions as well as the particle concentration decrease up to 61% [15]. Abdolzadeh et al. computationally studied air flow in a standard room and investigated energy and exergy analyses of two heating systems including floor heating and skirt boarding systems. They showed that the thermal efficiency of the skirt boarding system with the same amount of heat flux given to the room is 8% higher than that of the floor heating system [16]. In the present study, airflow simulation in the compartment of a passenger coach was numerically carried out. The compartment was simulated in two cases: with the presence of seated/slept manikins and without manikins. This study also conducted an experiment to verify the results of the numerical modeling. The airflow properties were measured in different parts of the compartment in the experimental study and compared with the counterpart predicted properties. Moreover, after investigation of airflow characteristics in the above mentioned cases,
4
a few modifications were suggested and implemented on the compartment in order to improve the performance of thermal comfort in the HVAC system. 2. System description Fig. 1 shows the schematic of the passenger coach compartment studied in the present study. This compartment was selected from a passenger coach, which is currently running on one of the Iranian railway routes. The dimensions of the compartment are 1.986×1.868×2.48m. The main air channel comes from the air handling unit and is located under the floor of the compartment. This channel is divided into two parts and delivers the air to the compartment. The main air inlet in the compartment is located under the window and the second air inlet which sends a small portion of inlet air to the compartment, is located under the seats in the left side of the compartment. The outlet vents are located on the wall adjacent to the aisle. In order to approach a real condition, the compartment was studied in the presence of no passenger (Fig. 1), four seated passengers (Fig. 2a), and four slept passengers (Fig. 2b). The manikins were designed and placed in the compartment to simulate the passengers inside the compartment. The seated manikin’s height is 1.3m and the length of manikin slept on the bed is 1.8m.
Air Inlet 1
Main air channel of air handling unit
Inlet air 2
Air outlets to aisle
5
Fig. 1:Compartment geometry
Right side
left side
(a) Fig. 2: The compartment with (a) seated manikins (b) slept manikins
(b)
3. Experimental measurement
In order to verify performance of the numerical modeling, an experiment was also conducted to measure the air flow properties. Temperature and velocity distributions in the studied compartment were measured when the air conditioning system was working while the coach was stoped. Ten points were chosen in the compartment to measure the air flow properties, Fig. 3a. The Testo (454) measuring instrument was used to measure the air flow properties. This instrument was held in the selected points shown in Fig. 3a and then the air velocity and the air temperature were measured (Figs. 3b and c). An infrared thermometer was also used to measure the walls’ temperature in the compartment, Fig.3d. These temperatures are later used as the boundary conditions of the computational model. It should be pointed out that the following conditions were held for the measurements: 1. The coach was not running.
6
2. Air conditioning system
as well as the compartment reached to the steady state
conditions 3. There was no passenger in the compartment. 4. The airflow properties reached to the steady state condition and their variations were very small. 5. The compartments on the right and left sides of the test compartment have the same temperature with the test compartment. . 6. The compartment door and beds were closed. 7. The lighting system was off. The airflow properties for each point were measured several times and then the average value of all the measurements was obtained.
Measuring the inlet air temperature
The measured ° temperature, Tin=17 C
Measuring device (Testo 454)
(a)
(b)
7
Air temperature probe
Air velocity probe
(c) (d) Fig.3: (a) The selected points for measuring the airflow properties (b) measuring the inlet air velocity and temperature (c) measuring the air velocity and temperature close to the ground, (d) Measuring the ceiling temperature
4.
Governing equations and numerical methods
The airflow in the compartment was simulated using the v2-f turbulence model [17, 18, 19, 20] as well as the energy equation. The model formulation has the following general form: _ _ _ _ uj , eff S t x j x j x j
(1)
Where represents the independent flow variables,
,eff
the effective diffusion coefficient, S
the source term, ρ the flow density and the bars denote the Reynolds averaging. In Table 1 the mathematical form of each transport equation of the v2-f model are summarized. p is the air _
pressure, µt the turbulent viscosity, S the rate of strain, f a part of the v '2 source term and Tl the turbulent time scale.
8
Table 1: Coefficients and source terms in Eq. (1)
Name of conservation equation
Effective diffusion coefficient ,eff
1
0
Independent flow variable _
Continuity X-Component of momentum
Source term
S 0
U
t
1 p u v w (t ) (t ) (t ) x x x y x z z
Y-Component of momentum
V
t
Z- Component of momentum
W
t
Energy Turbulent kinetic energy(k)
Tf
f t / T
0
K
t / k , t
Gk
Turbulent kinetic energy dissipation rate(ε) Wall normal turbulence fluctuation to kinetic energy( )
Ε
t / , t
(C' 1Gk C 2 ) / Tl
_ '2
t /
2 p k ( t / k , t ) S k x j x j
L2 1
fh
, Gk
t S 2 , S Sij Sij
( v /k ) , S
1 p
z
1 p
y
+
u v w (t ) (t ) (t ) x x y y z z
g (T f T )
v u w (t ) (t ) (t ) y y x y z z
(1 p ) f w p f h Gk / k
k G 1 (C1 1 k )( 2 / 3) , f w T max , CT k l Tl
_ k3/ 2 3/ 4 L CL max , C 1 / 4 , Pk 2C v '2 TSij2
t C , k Tl , , C' 1 1.44(1 0.04(1 p )
1
)
C 2 1.83, C1 1.7, C2 1.2, C 0.22, CL 0.161, C 90, CT 6, k ,t 1, ,t 1.22, ,t 1, p 3
The appropriate boundary conditions of turbulence variables near the walls are as follows: k v '2 0,
2
k y 2p
(2)
yp is the distance from the cell center adjacent to the wall. It should be mentioned that thev2-f model is similar to the standard k-epsilon model. Moreover, it incorporates also some near-wall turbulence anisotropy as well as non-local pressure-strain effects. It is a general turbulence model for low Reynolds-numbers which do not need to make use of wall functions because it is valid up to solid walls [18]. It has been shown that this model
9
adjacent to the wall predicts the turbulence characteristics with a higher accuracy compared to other turbulent models such as k-ε and k-w models [19]. However, the entire mentioned turbulent models are the same in the region far from the walls. In this study, the commercial CFD software, FLUENT (version 6.2), was used to predict the turbulent airflow. The v 2-f model used in this study is based on Davidson et al. [18] study. An unstructured grid (tetrahedral cell topology) was applied to the compartment as well as the surrounding air of the manikins, Figs 4a and b. To resolve the boundary layer around the manikins, fine meshes were created at the surface of the manikins. The manikins were considered as a heated body and the buoyancy movement due to the generated thermal plume was taken into account in the calculation. This was performed by taking the manikin as a constant temperature body which its temperature is higher than the surrounding air. The boundary conditions used in this study are given in Table 2. It should be mentioned that all the boundary conditions given to the computational model were based on the measurements conducted in this study. At the inlet register, the air flow rate, the air turbulent intensity, and the air temperature were taken 460m3/hr (Fig. 5), 10%, and, 17°C, respectively,. The temperatures of manikin body’s, ceiling, window’s wall, and window’s glass were taken 32°C, 27°C, 25°C, and 34°C, respectively. The sidewalls on the right and left sides of the compartment as well as the door’s wall were considered isolated. The pressure outlet boundary condition was given in the outlet registers of the compartment. Grid dependency analysis of the computational domain was carried out and a computational mesh containing about 2.896 ×106 cells showed a sufficient accuracy for continuing the analysis, Table 3.
10
(a)
(b)
Fig 4: Generated mesh (a) slept manikins (b) seated manikins
Inlet air from air handling system: Q=920m3/h, T in=17 °C
The outlet air to the next compartments, Q=460m3/h, Tout=17°C
Fig. 5: Given boundary conditions in the main inlet and outlet sections Table 2: The boundary conditions used in the present study Total inlet flow rate (m3/s)
Inlet Air Temperature (°C)
Manikin Body Temperature (°C)
Glass Temperature (°C)
Ceiling Temperature (°C)
Window’s wall Temperature (°C)
460
17
32
34
27
25
Outlet registers
Side walls +door wall
Patm
Isolated
11
Inlet Turbulent intensity (%)
10
Table 3: Checking the mesh dependency based on the flow properties Cell length (m)
Number of Cells
0.08 723910 0.06 965213 0.04 1447820 0.03 1930427 0.02 2895641* 0.015 3860854 * The selected grid in this study
5.
Wall shear stress (N/m2) 0.4235 0.2564 0.1432 0.1243 0.1123 0.1119
Air temperature at the center of compartment (°C) 19.93 20.02 20.1 20.17 20.21 20.235
Results and Discussion
5.1 Validation of the numerical model In this section, the performance of modeling was verified with the experimental data measured in this study. Fig. 6 shows the predicted properties of air flow in the selected points using the computational model when there is no manikin in the compartment. These data were later put close to the measured data to check out the modeling accuracy, Fig. 7. As shown in Fig. 7a, the velocity magnitude difference between the measured and the predicted values relative to the measured data all over the selected points is between 3% and 20%. However, the temperature distribution accuracy is good enough and has a higher accuracy compared to the velocity magnitude distribution, Fig. 7b.
)a(
12
)b(
)c(
)d( Fig. 6: The predicted velocity magnitude and temperature at the locations in the experiment (a) points 1,3, and 6 (b) 2,3, and 4 (c) 5,6, and 7 (d)8, 9, and 10
13
(a)
(b)
(c) (d) Fig. 7: Comparison of predicted and measured air flow properties in the selected points (a) velocity magnitude (b) Temperature (c) relative error of velocity (d) relative error of temperature
5.2 Flow analysis when manikins seated in the compartment In this section, the flow analysis is investigated when the manikins are seated in the compartment. Fig. 8 shows the air flow velocity magnitude distribution entered to the compartment from the main air vent. As shown in this figure, a portion of the airflow in the main channel turns to the inlet register of the compartment. The flow pattern in the inlet air channel is not good enough due to the sudden turn of the air flow and this causes the airflow to enter the compartment with a non uniform and high velocity in the very right of window’s wall. Fig. 8b 14
shows the air flow velocity magnitude distribution in the middle plane of the compartment. As shown in this figure, the air flow delivered to the compartment has the maximum velocity close to the ceiling and the minimum velocity in the center of the compartment and this states that most portion of the airflow entered to the compartment, leaves it from the exit registers without any efficient heat transfer with the confined air in the room.
Fig. 8: Airflow velocity magnitude distribution compartment
(a) at inlet air channel (b) middle surface in the
Fig. 9 shows the velocity magnitude and temperature distributions across a vertical plane placed in the middle of two manikins seated in front of each other in the compartment. As shown in this figure, the head of Manikin 1 experiences the maximum air velocity among all the manikins in the compartment. This velocity is also higher than the maximum air velocity for the comfort velocity [1]. This states that the air distribution in the left side (seats of Manikins 1 and 3) of the compartment make the passengers uncomfortable. Fig. 10 shows the velocity magnitude distribution along the line AB (shown in Fig. 9a). As shown in this figure, the air velocity in the front of Manikin 2 is eight times lower than that of Manikin 1. The air temperature distribution (Fig. 12) also confirms that Manikins 1 and 3 have lower temperatures compared to the other 15
manikins. This issue is mostly due to the non symmetrical distribution of entering air in the compartment.
A
B
(a)
(b) Fig. 9: Distributions of temperature and velocity magnitude in (a) surface crossed manikins 1 and 2 (b) surface crossed manikins 3 and 4
16
V(m/s)
1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
Z(m) Fig. 10: Airflow velocity magnitude variation form Manikin 1 to Manikin 2.
(a)
(b)
Fig.11: Distributions of velocity magnitude (a) temperature (b) in surface crossed manikins 1 and 3 and surface crossed manikins 2 and 4
Manikin Head
On the seat
Fig. 12: Temperature distributions across vertical lines drawn in front of the manikins
17
Fig. 11 shows the airflow velocity magnitude and temperature across a vertical plane which cross the manikins seated on the same side. This view also confirms that Manikin 1 has the maximum air flow velocity and the minimum air temperature over its head. Fig. 12 shows the temperature distributions along all the lines drawn in the front of the manikins. As shown in this figure, Manikin 1 and Manikin 2 on their head have the lowest and highest temperatures, respectively, among all the manikins. This figure also states that the mean temperature of line 1 is lower than the other lines as the air temperature in the lower body part of Manikin 3 is lower than the other manikins. This is mainly due to the good circulation of cool air around this manikin. The temperature in the vicinity of manikins’ (#1 and #3) head is between 21.5 and 21.8°C. This states that the airflow’s temperature around these two manikins is lower than the comfort temperature [1]. Manikin 4 also has a cooler air temperature up to 0.9m of its height in the vicinity of its body compared to Manikins 1 and 2. However, Manikin 2 has the highest mean air temperature around its body due to the poor air circulation in this region. It should be mentioned that the air temperature around all the manikins is not in the range of comfort temperature (23-26°C). As shown in Fig. 11, the left side of the compartment is cooler than the right side. To clarify this fact better, four lines were selected in the front of manikins seated on the same side and seated in front of each other. These lines are shown in Fig. 13a. The temperature distributions along these lines are shown in Figs 13b and 13c. As shown in these figures, the temperature differences between the end of lines #1, 2, 3, and 4 are 0.5, 0.3, 0.7, and 0.7 °C, respectively. It is clear that the mean temperature of the line 1 is higher than that of line 2.
18
(a) 23
23
22.8
From Manikin1 to Manikin3 (line 1)
22.8
22.6
From Manikin2 to manikin4 (line2)
22.6
From Manikin1 to manikin2 (line 3) From manikin3 to manikin4 (line4)
22.4 22.4
T(C°)
22.2 T(C°)
22.2
22
21.8
22
21.6
21.8
21.4
21.6
21.2
21.4
21 0.4
0.6
0.8
Z(m)1
1.2
1.4
1.6
0.3
0.5
0.7
(b)
0.9
Z(m)
1.1
1.3
(c
Fig. 13: Temperature distributions (b, c) across different lines in the front the manikins (a)
5.3 Flow analysis when manikins slept in the compartment In this section, the flow analysis is investigated when the compartment is in the sleeping mode. In this case, all the beds are opened and the manikins lied on them. Figs. 14a and b show the velocity magnitude and temperature distributions across the vertical plane which cut the manikins’ upper and lower body parts. As shown in these figures, Manikin 3 slept on the top bed in the left side of the compartment, has higher air velocity and lower air temperature around its
19
1.5
upper part compared to the other manikins. It shows that the manikins slept on the beds in the left side of the compartment have a better thermal comfort conditions compared to those seated in the right side of the compartment. Fig. 15 shows the air velocity magnitude and the temperature distributions across the plane which cut the manikins heads. Fig. 16 shows the air velocity magnitude and temperature distributions across line A-B. As shown in this figure, the maximum air velocity is seen in the middle of the top beds and the minimum air speed is at the head of Manikin 3. This figure also states that the location with the highest air velocity has the lowest air temperature and the location with the lowest air velocity has the highest air temperature. To check out what manikin has the best thermal condition, the air temperature distribution on the lines shown in Fig. 17a, are found and then the mean temperature on these lines were obtained. As shown in Fig. 17b, manikins in the left side of the compartment have a lower temperature in their vicinity compared to the manikins in the right side. This figure also states that the mean air temperature around the manikin sleeping on the top bed in each side is lower than that of the manikin sleeping on the bottom bed. Moreover, the manikins slept on the bottom beds have a lower temperature in their lower part compared to their upper part. However, this trend is reversed for the manikins slept on the top beds.
20
(a)
(b) Fig. 14: Temperature and velocity magnitude distributions in surfaces crossed (a) manikins’ upper body part (b) manikins’ lower body part
21
Temperature(°C)
Fig. 15: Temperature and velocity magnitude distributions in surfaces crossed manikins’ head
(a)
(b)
Fig. 16: Velocity magnitude (a) and temperature (b) distributions across line AB (Fig 14)
22
(a)
Manikin Legs
Manikin Head
(b) Fig. 17: Temperature distribution (b) across lines drawn over the lied manikins (a)
5.4 Modifications for improving the airflow performance in the compartment As stated in the above sections, the non symmetrical entering of the air flow to the compartment causes the flow to lean on the left side of the compartment and this makes the left side to experience higher air velocity and lower air temperature compared to the right side of the compartment. To solve this issue, it is required to individually send the airflow via a channel into each compartment. This causes the inlet air flow entering symmetrically to the compartment and makes similar the left and right sides of the compartment. Two changes were applied in the compartment design, including the addition of a direct channel to the compartment in order to symmetrically distribute the air flow in the compartment and smooth the airflow pattern in the 23
upside part of the window (Fig. 19). In following, the effects of these two changes were tested and assessed on the studied air flow properties in the compartment. Fig. 20 shows the air flow velocity magnitude in the case of applying the two modifications stated in above. As shown in this figure, the new inlet cool air causes the flow to symmetrically distribute in the left and right sides of the compartment. Fig. 21 confirms that the symmetrical condition is established in all the planes crossed the manikins seated in the front of each other. Fig. 22 shows the velocity magnitude and temperature distributions on a line drawn in the front of the manikins to check out the airflow characteristics. It is shown that the manikins seated in a same seat have almost the same temperature difference compared to the non symmetrical case. However, in the symmetrical condition the manikins seated in front of each others have the same air velocity and the thermal condition around their body and this makes identical the left and right sides of the compartment.
Loss of Kinitic Energy
No airflow
(a)
(b)
Fig. 18: The issues of air distribution in the compartment (a) inlet air channel (b) non curved plate over the window 24
A curved path
A Direct channel from air conditioning system
Fig. 19: Modifications for improving the air distribution in the compartment
Symmetrical air distribution
(a)
(b)
Fig. 20: Air flow distribution after implementing the proposed modifications (a) curved plate over the window (b) direct inlet channel from the air handling system
25
(a)
(b) Fig. 21: Distributions of temperature and velocity magnitude in (a) surface crossed manikins 1 and 2 (b) surface crossed manikins 3 and 4
26
(a)
(b) Fig. 22: Velocity magnitude (a) and Temperature (b) distributions across lines 1 and 3 drawn in the front of the manikins
6. Conclusions In this research, a computational study was carried out to predict the airflow properties of HVAC system in the compartment of a passenger coach in the summer. This study also conducted an experiment in the studied compartment to verify the numerical results. The turbulent air flow was considered in the compartment and it was simulated by using v2-f turbulent model. The compartment was simulated in two cases: with seated and slept manikins. Temperature and velocity magnitude distributions were obtained in two cases, including the compartment with 27
manikins and without of manikins. The results showed that due to the position of inlet air channel in the compartment, the airflow entered from the HVAC system mostly inclines on the left side of the compartment and this causes the airflow to deviate from the thermal comfort conditions in some parts of the compartment and makes the seated/slept passengers uncomfortable. The results also showed that performing some simple modifications in the compartment design can improve the HVAC system performance and also make the air flow distribution symmetrical in the compartment.
References [1]. ASHRAE, 1992, Standard 55-Thermal Environmental Conditions for Human Occupancy. Atlanta, ASHRAE, Inc. [2]. Oakland innovation and information services Ltd; 18 Cambridge science park; Milton road ,Cambridge, cb40fh, thermal comfort on trains, 2002 [3] Tim Berlitz, Gerd Matschke. interior airflow simulation in railway rolling stock. Research and Technology Centre of the Deutsche Bahn AG, Division for aerodynamics and airconditioning, 2004 [4] Yongson Ooi,, Anjum Badruddin Irfan, ,Z.A. Zainala, P.A. Aswatha Narayana , Airflow analysis in an air conditioning room, Building and Environment 42(3), 1531–1537, 2007 [5]. Khodakarami, N. Nasrollahi,Thermal comfort in hospitals–A literature review”, Renewable and Sustainable Energy Reviews, 16, 4071–4077, 2012
[6]. Salmanzadeh, M.; Zahedi, G.; Ahmadi, G.; Marr, D.R.; and Glauser, M. Computational modeling of effects of thermal plume adjacent to the body on the indoor airflow and particle transport. Journal of Aerosol Science, 53, 29–39, 2012.
[7]. Gaever R. Van, Jacobs V.A., Diltoer M., Peeters L., Vanlanduit S., Thermal comfort of the surgical staff in the operating room, Building and Environment, 81, 37-41, 2014, 28
[8]. Xiufeng Yang. Ke Zhong, Hui Zhu,Yanming Kang, Experimental investigation on transient natural ventilation driven by thermal buoyancy, Building and Environment, 77, 29–39, 2014 [9] Piotr Uścinowicz, , Marta Chludzińska, Anna Bogdan, Thermal environment conditions in Polish operating rooms, Building and Environment, 94, 296–304, 2015, [10] Xiufeng Yang, Ke Zhong , Yanming Kang, Tianyin Tao, Numerical investigation on the airflow characteristics and thermal comfort in buoyancy-driven natural ventilation rooms, Energy and Buildings, 109, 255–266, 2015 [11] Nurullah Arslanoglu, Abdulvahap Yigit, Experimental and theoretical investigation of the effect of radiation heat flux on human thermal comfort, Energy and Buildings, 113, 23-29, 2016 [12]. Lihui Wang, Zhiping Du, Jianshun Zhang, Rui Chen, Shaolong Yu, Minglu Qu, Study on the thermal comfort characteristics under the vent with supplying air jets and cross-flows coupling in subway stations, Energy and Buildings, 131, 113-122, 2016 [13] Marie Ruellan, Herie Park, Rachid Bennacer, Residential building energy demand and thermal comfort: Thermal dynamics of electrical appliances and their impact, Energy and Buildings, Volume 130, 46-54, 2016, [14]. Mehrzad Ansaripour, Morteza Abdolzadeh, Saleh Sargazizadeh, Computational modeling of particle transport and distribution emitted from a Laserjet printer in a ventilated room with different ventilation configurations, Applied Thermal Engineering, 103, 920-933, 2016
[15] Ahmed Qasim Ahmed, Shian Gao, Ali Khaleel Kareem, Energy saving and indoor thermal comfort evaluation using a novel local exhaust ventilation system for office rooms, Applied Thermal Engineering, 110, 821-834, 2017 [16] Morteza Abdolzadeh, Mohammad Saleh Sargazizadeh, Mohammad Hadi Dehghan,
Quantitative and Qualitative Energy Assessments of Floor Heating and Skirting Board Heating Systems in a Room, Journal of Energy Engineering, 143(1), http://dx.doi.org/10.1061/(ASCE)EY.1943-7897.0000385, 2017 [17]. Lien F., Kalitzin G., Computations of transonic flow with the v2-f turbulence model. International Journal of Heat and Fluid Flow 22, 53–61. [18] Davidson L., Nielsen, P.V., Sveningsson A.,. Modification of the V2F model for computing the flow in a 3D wall jet. Turbulence Heat and Mass Transfer , 4, 577–584 ,2003.
29
[19]. Abdolzadeh . M,Mehrabian M.A, Soltani, A., Numerical study to predict the particle deposition under the influence of operating forces on a tilted surface in the turbulent flow, Advanced Powder Technology, 22(3), , 405-415, 2011 [20].Abdolzadeh M., Mehrabian M.A., Combined effect of thermophoretic force and other influencing parameters on the particle deposition rate on a tilted rough surface, International Journal of Thermal Sciences 50 (6), 954-964, 2011
Table captions: Table 1: Coefficients and source terms in Eq. (1) Table 2: The boundary conditions used in the present study Table 3: Checking the mesh dependency based on the flow properties Figure captions Fig. 1: Compartment geometry Fig. 2: The compartment with (a) seated manikins (b) slept manikins Fig.3: (a) The selected points for measuring the airflow properties (b) measuring the inlet air velocity and temperature (c) measuring the I air velocity and temperature close to the ground, (d) Measuring the ceiling temperature Fig 4: Generated mesh (a) slept manikins (b) seated manikins Fig. 5: Given boundary conditions in the main inlet and outlet sections Fig. 6: The predicted velocity and temperature at the locations in the experiment (a) points 1,3,6 (b) 2,3, and 4 (c) 5,6, and 7 (d)8, 9, and 10 Fig. 7: Comparison of predicted and measured air flow properties in the selected points (a) velocity (b) Temperature (c) relative error of velocity (d) relative error of temperature Fig. 8: Airflow velocity distribution (a) at inlet air channel (b) middle surface in the compartment
30
Fig. 9: Distributions of temperature and velocity in (a) Surface crossed manikins 1 and 2 (b) surface crossed manikins 3 and 4 Fig. 10: Airflow velocity variation form Manikin 1 to Manikin 2. Fig.11: Distributions of velocity (a) temperature (b) in surface crossed manikins 1 and 3 and surface crossed manikins 2 and 4 Fig. 12: Temperature distributions across vertical lines drawn in front of the manikins Fig. 13: Temperature distributions (b, c) across different lines in the front the manikins (a) Fig. 14: Temperature and velocity distributions in surfaces crossed (a) manikins’ upper body part (b) manikins’ lower body part Fig. 15: Temperature and velocity distributions in surfaces crossed manikins’ head Fig. 16: Velocity (a) and temperature (b) distributions across line AB (Fig 14) Fig. 17: Temperature distribution (b) across lines drawn over the lied manikins (a) Fig. 18: The issues of air distribution in the compartment (a) inlet air channel (b) non curved plate over the window Fig. 19: Modifications for improving the air distribution in the compartment Fig. 20: Air flow distribution after implementing the proposed modifications (a) curved plate over the window (b) direct inlet channel from the air handling system Fig. 21: Distributions of temperature and velocity in (a) surface crossed manikins 1 and 2 (b) surface crossed manikins 3 and 4 Fig. 22: Velocity (a) and Temperature (b) distributions across lines 1 and 3 drawn in the front of the manikins
31
Highlights
Computational simulation of airflow inside compartment of a passenger coach The turbulent air flow in the compartment was modeled using v2 f turbulent model. The thermal comfort conditions in the compartment were not evenly distributed
32
S1359-4311(16)33734-6 http://dx.doi.org/10.1016/j.applthermaleng.2017.05.086 ATE 10398
To appear in:
Applied Thermal Engineering
Received Date: Revised Date: Accepted Date:
28 November 2016 24 April 2017 16 May 2017
Please cite this article as: A. Aliahmadi, M. Abdolzadeh, K. Lari, Air flow simulation of HVAC system in compartment of a passenger coach, Applied Thermal Engineering (2017), doi: http://dx.doi.org/10.1016/ j.applthermaleng.2017.05.086
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Air Flow Simulation of HVAC System in Compartment of a Passenger Coach Ali. Aliahmadi1, Morteza Abdolzadeh2*, Khosro Lari1 1
Department of Mechanical Engineering, Graduate University of Advanced Technology, Kerman, Iran Energy Institute, Graduate University of Advanced Technology, Kerman, Iran
2
*
Corresponding author: Department of Mechanical Engineering, Graduate University of Advanced Technology, End of Haft Bagh Highway, Kerman, Iran, Email address: [email protected] [email protected]
Abstract: In this study, numerical simulation of airflow in the compartment of a passenger coach was carried out in summer. An experimental measurement was also conducted to verify the performance of the numerical model. The turbulent air flow in the compartment was supplied by the HVAC system of the coach. The numerical simulation in the compartment was performed in two cases: seated and slept manikins. The manikins were considered as heated manikins in all the simulations. The turbulent air flow was modeled using v2 f turbulent model. Two conditions in the compartment, including temperature and velocity comfort criteria were assessed in the compartment with and without the presence of the manikins. Moreover, some modifications were implemented in the compartment design to improve the comfort conditions around the manikin in the two cases. Results showed that due to the inappropriate design of the compartment, the thermal comfort conditions do not evenly distribute in the compartment and this issue makes the passengers uncomfortable. The outcome of the modifications showed that the air flow symmetrically enters to the compartment and provides better thermal conditions for the passengers seated/slept in the compartment. Keywords: Passenger coach, HVAC, Heated manikin, comfort
1
1. Introduction There are many passenger coaches in the railroads across the world which have weak air conditioning as the criteria for engineering design of the compartment have not been taken into account appropriately. This issue makes problems for even air distribution in the compartments of the passenger coaches and causes them to not provide an acceptable level of thermal comfort conditions and fresh air for passengers. Different parameters, including equipment as well as the location and geometry of inlet and outlet air registers in the compartments affect on the airflow distribution and heat transfer. These parameters can cause the passenger to feel uncomfortable in the compartment, because when the there is no fear air conditioning, the human body does not well dissipate its waste thermal energy to the surrounding air. The comfort air temperature in summer is between 23°C and 26°C and the comfort air velocity is between 0.2m/s and 0.8m/s [1]. These conditions should be sustained by the HVAC systems of passenger coaches. So far, many studies have been carried out to simulate the air flow distribution and assess the thermal comfort conditions in different enclosed spaces. For instance, Okland did a study on metro stations and carriages in London. They investigated the thermal comfort conditions as well as the air particulates in the studied systems. They realized that air temperature and humidity as well as particulate distribution, significantly affect the thermal comfort conditions of the passengers [2]. Berlitz and Matschke studied interior airflow simulation of a regional train. They simulated and measured the air flow properties in the train compartments [3]. Yongson et al. studied air flow analysis of an air conditioning room. They used k-ε turbulent model for turbulent air flow modeling. They investigated different locations of blower placement to analyze comfort of occupant in the room. They showed that the occupants will experience most comfort if the air conditioner blower is placed on the very corner of the side wall compared to the other locations [4].Khodakarami and Nasrolahi conducted a review about the thermal comfort of hospitals. They 2
concluded that it is important to find some solutions to reconcile the different thermal comfort conditions required by different occupants in hospitals [5]. Salmanzadeh et al. studied the buoyancy driven thermal plume near a heated manikin seated in a cubicle in turbulent flow using k-ε turbulent model. They showed that due to the thermal plume flow created by the temperature gradient adjacent to the body, a high air velocity is seen in the breathing zone of the manikin [6]. Gahever et al. studied thermal comfort of the surgical staff in the operating room. They summarized the technical standards and compared them to the thermal comfort standards and the data gathered on the site. They showed that significant discrepancies exist between the HVAC standards and the thermal comfort standard [7]. Yang et al. numerically studied transient natural ventilation driven by buoyancy. They showed that vent shape and the horizontal position of heat source have little effects on the ventilation [8]. Uścinowicz et al. studied thermal environmental conditions in polish operating rooms. They measured thermal conditions in 37 operating rooms and showed that operating staffs are different in thermal requirements [9]. Yang et al. numerically studied the airflow characteristics and thermal comfort conditions in buoyancydriven natural ventilation rooms. They investigated indoor thermal comfort during the transient ventilation. They showed that the initial indoor temperature may affect the steady state of natural ventilation. They also indicated that the vent shape and the horizontal position of heat source do not have a significant effect on the ventilation [10]. Arslanoglu and Yigit computationally and experimentally studied the effect of radiation heat flux from lighting lamps on human thermal comfort. They indicated that the upper part of the human body due to the lamp radiant heat flux is the most affected part of the body [11]. Wang et al. studied the thermal comfort characteristics under the vent with supplying air jets and cross-flows coupling in subway stations. They also studied relative warmth index to assess the thermal comfort conditions in the transitional
3
dynamic areas. They showed that the coupling airflows of supplying air jets and cross-flows can improve the transitional thermal comfort in air-conditioning systems [12]. Rullen et al. studied the thermal impacts of heat dissipation of electrical appliances on buildings by applying a dynamic thermal model of electrical appliances. Their results indicated the necessity of dynamic thermal modeling of electrical appliances for more accurate and efficient building energy simulation [13]. Ansaripour et al. studied air flow distribution and particle concentration emitted from a Laserjet printer in a ventilated room. They used v2-f turbulent model to predict the turbulent air characteristics. Their results showed that when the printer is located on the front side of the manikin, the particle concentration in the breathing zone is quite high in most of the used ventilation configurations [14]. Ahmed et al. designed and assessed a new local exhaust ventilation system in a room. They showed that the proposed system improved the thermal conditions as well as the particle concentration decrease up to 61% [15]. Abdolzadeh et al. computationally studied air flow in a standard room and investigated energy and exergy analyses of two heating systems including floor heating and skirt boarding systems. They showed that the thermal efficiency of the skirt boarding system with the same amount of heat flux given to the room is 8% higher than that of the floor heating system [16]. In the present study, airflow simulation in the compartment of a passenger coach was numerically carried out. The compartment was simulated in two cases: with the presence of seated/slept manikins and without manikins. This study also conducted an experiment to verify the results of the numerical modeling. The airflow properties were measured in different parts of the compartment in the experimental study and compared with the counterpart predicted properties. Moreover, after investigation of airflow characteristics in the above mentioned cases,
4
a few modifications were suggested and implemented on the compartment in order to improve the performance of thermal comfort in the HVAC system. 2. System description Fig. 1 shows the schematic of the passenger coach compartment studied in the present study. This compartment was selected from a passenger coach, which is currently running on one of the Iranian railway routes. The dimensions of the compartment are 1.986×1.868×2.48m. The main air channel comes from the air handling unit and is located under the floor of the compartment. This channel is divided into two parts and delivers the air to the compartment. The main air inlet in the compartment is located under the window and the second air inlet which sends a small portion of inlet air to the compartment, is located under the seats in the left side of the compartment. The outlet vents are located on the wall adjacent to the aisle. In order to approach a real condition, the compartment was studied in the presence of no passenger (Fig. 1), four seated passengers (Fig. 2a), and four slept passengers (Fig. 2b). The manikins were designed and placed in the compartment to simulate the passengers inside the compartment. The seated manikin’s height is 1.3m and the length of manikin slept on the bed is 1.8m.
Air Inlet 1
Main air channel of air handling unit
Inlet air 2
Air outlets to aisle
5
Fig. 1:Compartment geometry
Right side
left side
(a) Fig. 2: The compartment with (a) seated manikins (b) slept manikins
(b)
3. Experimental measurement
In order to verify performance of the numerical modeling, an experiment was also conducted to measure the air flow properties. Temperature and velocity distributions in the studied compartment were measured when the air conditioning system was working while the coach was stoped. Ten points were chosen in the compartment to measure the air flow properties, Fig. 3a. The Testo (454) measuring instrument was used to measure the air flow properties. This instrument was held in the selected points shown in Fig. 3a and then the air velocity and the air temperature were measured (Figs. 3b and c). An infrared thermometer was also used to measure the walls’ temperature in the compartment, Fig.3d. These temperatures are later used as the boundary conditions of the computational model. It should be pointed out that the following conditions were held for the measurements: 1. The coach was not running.
6
2. Air conditioning system
as well as the compartment reached to the steady state
conditions 3. There was no passenger in the compartment. 4. The airflow properties reached to the steady state condition and their variations were very small. 5. The compartments on the right and left sides of the test compartment have the same temperature with the test compartment. . 6. The compartment door and beds were closed. 7. The lighting system was off. The airflow properties for each point were measured several times and then the average value of all the measurements was obtained.
Measuring the inlet air temperature
The measured ° temperature, Tin=17 C
Measuring device (Testo 454)
(a)
(b)
7
Air temperature probe
Air velocity probe
(c) (d) Fig.3: (a) The selected points for measuring the airflow properties (b) measuring the inlet air velocity and temperature (c) measuring the air velocity and temperature close to the ground, (d) Measuring the ceiling temperature
4.
Governing equations and numerical methods
The airflow in the compartment was simulated using the v2-f turbulence model [17, 18, 19, 20] as well as the energy equation. The model formulation has the following general form: _ _ _ _ uj , eff S t x j x j x j
(1)
Where represents the independent flow variables,
,eff
the effective diffusion coefficient, S
the source term, ρ the flow density and the bars denote the Reynolds averaging. In Table 1 the mathematical form of each transport equation of the v2-f model are summarized. p is the air _
pressure, µt the turbulent viscosity, S the rate of strain, f a part of the v '2 source term and Tl the turbulent time scale.
8
Table 1: Coefficients and source terms in Eq. (1)
Name of conservation equation
Effective diffusion coefficient ,eff
1
0
Independent flow variable _
Continuity X-Component of momentum
Source term
S 0
U
t
1 p u v w (t ) (t ) (t ) x x x y x z z
Y-Component of momentum
V
t
Z- Component of momentum
W
t
Energy Turbulent kinetic energy(k)
Tf
f t / T
0
K
t / k , t
Gk
Turbulent kinetic energy dissipation rate(ε) Wall normal turbulence fluctuation to kinetic energy( )
Ε
t / , t
(C' 1Gk C 2 ) / Tl
_ '2
t /
2 p k ( t / k , t ) S k x j x j
L2 1
fh
, Gk
t S 2 , S Sij Sij
( v /k ) , S
1 p
z
1 p
y
+
u v w (t ) (t ) (t ) x x y y z z
g (T f T )
v u w (t ) (t ) (t ) y y x y z z
(1 p ) f w p f h Gk / k
k G 1 (C1 1 k )( 2 / 3) , f w T max , CT k l Tl
_ k3/ 2 3/ 4 L CL max , C 1 / 4 , Pk 2C v '2 TSij2
t C , k Tl , , C' 1 1.44(1 0.04(1 p )
1
)
C 2 1.83, C1 1.7, C2 1.2, C 0.22, CL 0.161, C 90, CT 6, k ,t 1, ,t 1.22, ,t 1, p 3
The appropriate boundary conditions of turbulence variables near the walls are as follows: k v '2 0,
2
k y 2p
(2)
yp is the distance from the cell center adjacent to the wall. It should be mentioned that thev2-f model is similar to the standard k-epsilon model. Moreover, it incorporates also some near-wall turbulence anisotropy as well as non-local pressure-strain effects. It is a general turbulence model for low Reynolds-numbers which do not need to make use of wall functions because it is valid up to solid walls [18]. It has been shown that this model
9
adjacent to the wall predicts the turbulence characteristics with a higher accuracy compared to other turbulent models such as k-ε and k-w models [19]. However, the entire mentioned turbulent models are the same in the region far from the walls. In this study, the commercial CFD software, FLUENT (version 6.2), was used to predict the turbulent airflow. The v 2-f model used in this study is based on Davidson et al. [18] study. An unstructured grid (tetrahedral cell topology) was applied to the compartment as well as the surrounding air of the manikins, Figs 4a and b. To resolve the boundary layer around the manikins, fine meshes were created at the surface of the manikins. The manikins were considered as a heated body and the buoyancy movement due to the generated thermal plume was taken into account in the calculation. This was performed by taking the manikin as a constant temperature body which its temperature is higher than the surrounding air. The boundary conditions used in this study are given in Table 2. It should be mentioned that all the boundary conditions given to the computational model were based on the measurements conducted in this study. At the inlet register, the air flow rate, the air turbulent intensity, and the air temperature were taken 460m3/hr (Fig. 5), 10%, and, 17°C, respectively,. The temperatures of manikin body’s, ceiling, window’s wall, and window’s glass were taken 32°C, 27°C, 25°C, and 34°C, respectively. The sidewalls on the right and left sides of the compartment as well as the door’s wall were considered isolated. The pressure outlet boundary condition was given in the outlet registers of the compartment. Grid dependency analysis of the computational domain was carried out and a computational mesh containing about 2.896 ×106 cells showed a sufficient accuracy for continuing the analysis, Table 3.
10
(a)
(b)
Fig 4: Generated mesh (a) slept manikins (b) seated manikins
Inlet air from air handling system: Q=920m3/h, T in=17 °C
The outlet air to the next compartments, Q=460m3/h, Tout=17°C
Fig. 5: Given boundary conditions in the main inlet and outlet sections Table 2: The boundary conditions used in the present study Total inlet flow rate (m3/s)
Inlet Air Temperature (°C)
Manikin Body Temperature (°C)
Glass Temperature (°C)
Ceiling Temperature (°C)
Window’s wall Temperature (°C)
460
17
32
34
27
25
Outlet registers
Side walls +door wall
Patm
Isolated
11
Inlet Turbulent intensity (%)
10
Table 3: Checking the mesh dependency based on the flow properties Cell length (m)
Number of Cells
0.08 723910 0.06 965213 0.04 1447820 0.03 1930427 0.02 2895641* 0.015 3860854 * The selected grid in this study
5.
Wall shear stress (N/m2) 0.4235 0.2564 0.1432 0.1243 0.1123 0.1119
Air temperature at the center of compartment (°C) 19.93 20.02 20.1 20.17 20.21 20.235
Results and Discussion
5.1 Validation of the numerical model In this section, the performance of modeling was verified with the experimental data measured in this study. Fig. 6 shows the predicted properties of air flow in the selected points using the computational model when there is no manikin in the compartment. These data were later put close to the measured data to check out the modeling accuracy, Fig. 7. As shown in Fig. 7a, the velocity magnitude difference between the measured and the predicted values relative to the measured data all over the selected points is between 3% and 20%. However, the temperature distribution accuracy is good enough and has a higher accuracy compared to the velocity magnitude distribution, Fig. 7b.
)a(
12
)b(
)c(
)d( Fig. 6: The predicted velocity magnitude and temperature at the locations in the experiment (a) points 1,3, and 6 (b) 2,3, and 4 (c) 5,6, and 7 (d)8, 9, and 10
13
(a)
(b)
(c) (d) Fig. 7: Comparison of predicted and measured air flow properties in the selected points (a) velocity magnitude (b) Temperature (c) relative error of velocity (d) relative error of temperature
5.2 Flow analysis when manikins seated in the compartment In this section, the flow analysis is investigated when the manikins are seated in the compartment. Fig. 8 shows the air flow velocity magnitude distribution entered to the compartment from the main air vent. As shown in this figure, a portion of the airflow in the main channel turns to the inlet register of the compartment. The flow pattern in the inlet air channel is not good enough due to the sudden turn of the air flow and this causes the airflow to enter the compartment with a non uniform and high velocity in the very right of window’s wall. Fig. 8b 14
shows the air flow velocity magnitude distribution in the middle plane of the compartment. As shown in this figure, the air flow delivered to the compartment has the maximum velocity close to the ceiling and the minimum velocity in the center of the compartment and this states that most portion of the airflow entered to the compartment, leaves it from the exit registers without any efficient heat transfer with the confined air in the room.
Fig. 8: Airflow velocity magnitude distribution compartment
(a) at inlet air channel (b) middle surface in the
Fig. 9 shows the velocity magnitude and temperature distributions across a vertical plane placed in the middle of two manikins seated in front of each other in the compartment. As shown in this figure, the head of Manikin 1 experiences the maximum air velocity among all the manikins in the compartment. This velocity is also higher than the maximum air velocity for the comfort velocity [1]. This states that the air distribution in the left side (seats of Manikins 1 and 3) of the compartment make the passengers uncomfortable. Fig. 10 shows the velocity magnitude distribution along the line AB (shown in Fig. 9a). As shown in this figure, the air velocity in the front of Manikin 2 is eight times lower than that of Manikin 1. The air temperature distribution (Fig. 12) also confirms that Manikins 1 and 3 have lower temperatures compared to the other 15
manikins. This issue is mostly due to the non symmetrical distribution of entering air in the compartment.
A
B
(a)
(b) Fig. 9: Distributions of temperature and velocity magnitude in (a) surface crossed manikins 1 and 2 (b) surface crossed manikins 3 and 4
16
V(m/s)
1.8 1.6 1.4 1.2 1 0.8 0.6 0.4 0.2 0 0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
Z(m) Fig. 10: Airflow velocity magnitude variation form Manikin 1 to Manikin 2.
(a)
(b)
Fig.11: Distributions of velocity magnitude (a) temperature (b) in surface crossed manikins 1 and 3 and surface crossed manikins 2 and 4
Manikin Head
On the seat
Fig. 12: Temperature distributions across vertical lines drawn in front of the manikins
17
Fig. 11 shows the airflow velocity magnitude and temperature across a vertical plane which cross the manikins seated on the same side. This view also confirms that Manikin 1 has the maximum air flow velocity and the minimum air temperature over its head. Fig. 12 shows the temperature distributions along all the lines drawn in the front of the manikins. As shown in this figure, Manikin 1 and Manikin 2 on their head have the lowest and highest temperatures, respectively, among all the manikins. This figure also states that the mean temperature of line 1 is lower than the other lines as the air temperature in the lower body part of Manikin 3 is lower than the other manikins. This is mainly due to the good circulation of cool air around this manikin. The temperature in the vicinity of manikins’ (#1 and #3) head is between 21.5 and 21.8°C. This states that the airflow’s temperature around these two manikins is lower than the comfort temperature [1]. Manikin 4 also has a cooler air temperature up to 0.9m of its height in the vicinity of its body compared to Manikins 1 and 2. However, Manikin 2 has the highest mean air temperature around its body due to the poor air circulation in this region. It should be mentioned that the air temperature around all the manikins is not in the range of comfort temperature (23-26°C). As shown in Fig. 11, the left side of the compartment is cooler than the right side. To clarify this fact better, four lines were selected in the front of manikins seated on the same side and seated in front of each other. These lines are shown in Fig. 13a. The temperature distributions along these lines are shown in Figs 13b and 13c. As shown in these figures, the temperature differences between the end of lines #1, 2, 3, and 4 are 0.5, 0.3, 0.7, and 0.7 °C, respectively. It is clear that the mean temperature of the line 1 is higher than that of line 2.
18
(a) 23
23
22.8
From Manikin1 to Manikin3 (line 1)
22.8
22.6
From Manikin2 to manikin4 (line2)
22.6
From Manikin1 to manikin2 (line 3) From manikin3 to manikin4 (line4)
22.4 22.4
T(C°)
22.2 T(C°)
22.2
22
21.8
22
21.6
21.8
21.4
21.6
21.2
21.4
21 0.4
0.6
0.8
Z(m)1
1.2
1.4
1.6
0.3
0.5
0.7
(b)
0.9
Z(m)
1.1
1.3
(c
Fig. 13: Temperature distributions (b, c) across different lines in the front the manikins (a)
5.3 Flow analysis when manikins slept in the compartment In this section, the flow analysis is investigated when the compartment is in the sleeping mode. In this case, all the beds are opened and the manikins lied on them. Figs. 14a and b show the velocity magnitude and temperature distributions across the vertical plane which cut the manikins’ upper and lower body parts. As shown in these figures, Manikin 3 slept on the top bed in the left side of the compartment, has higher air velocity and lower air temperature around its
19
1.5
upper part compared to the other manikins. It shows that the manikins slept on the beds in the left side of the compartment have a better thermal comfort conditions compared to those seated in the right side of the compartment. Fig. 15 shows the air velocity magnitude and the temperature distributions across the plane which cut the manikins heads. Fig. 16 shows the air velocity magnitude and temperature distributions across line A-B. As shown in this figure, the maximum air velocity is seen in the middle of the top beds and the minimum air speed is at the head of Manikin 3. This figure also states that the location with the highest air velocity has the lowest air temperature and the location with the lowest air velocity has the highest air temperature. To check out what manikin has the best thermal condition, the air temperature distribution on the lines shown in Fig. 17a, are found and then the mean temperature on these lines were obtained. As shown in Fig. 17b, manikins in the left side of the compartment have a lower temperature in their vicinity compared to the manikins in the right side. This figure also states that the mean air temperature around the manikin sleeping on the top bed in each side is lower than that of the manikin sleeping on the bottom bed. Moreover, the manikins slept on the bottom beds have a lower temperature in their lower part compared to their upper part. However, this trend is reversed for the manikins slept on the top beds.
20
(a)
(b) Fig. 14: Temperature and velocity magnitude distributions in surfaces crossed (a) manikins’ upper body part (b) manikins’ lower body part
21
Temperature(°C)
Fig. 15: Temperature and velocity magnitude distributions in surfaces crossed manikins’ head
(a)
(b)
Fig. 16: Velocity magnitude (a) and temperature (b) distributions across line AB (Fig 14)
22
(a)
Manikin Legs
Manikin Head
(b) Fig. 17: Temperature distribution (b) across lines drawn over the lied manikins (a)
5.4 Modifications for improving the airflow performance in the compartment As stated in the above sections, the non symmetrical entering of the air flow to the compartment causes the flow to lean on the left side of the compartment and this makes the left side to experience higher air velocity and lower air temperature compared to the right side of the compartment. To solve this issue, it is required to individually send the airflow via a channel into each compartment. This causes the inlet air flow entering symmetrically to the compartment and makes similar the left and right sides of the compartment. Two changes were applied in the compartment design, including the addition of a direct channel to the compartment in order to symmetrically distribute the air flow in the compartment and smooth the airflow pattern in the 23
upside part of the window (Fig. 19). In following, the effects of these two changes were tested and assessed on the studied air flow properties in the compartment. Fig. 20 shows the air flow velocity magnitude in the case of applying the two modifications stated in above. As shown in this figure, the new inlet cool air causes the flow to symmetrically distribute in the left and right sides of the compartment. Fig. 21 confirms that the symmetrical condition is established in all the planes crossed the manikins seated in the front of each other. Fig. 22 shows the velocity magnitude and temperature distributions on a line drawn in the front of the manikins to check out the airflow characteristics. It is shown that the manikins seated in a same seat have almost the same temperature difference compared to the non symmetrical case. However, in the symmetrical condition the manikins seated in front of each others have the same air velocity and the thermal condition around their body and this makes identical the left and right sides of the compartment.
Loss of Kinitic Energy
No airflow
(a)
(b)
Fig. 18: The issues of air distribution in the compartment (a) inlet air channel (b) non curved plate over the window 24
A curved path
A Direct channel from air conditioning system
Fig. 19: Modifications for improving the air distribution in the compartment
Symmetrical air distribution
(a)
(b)
Fig. 20: Air flow distribution after implementing the proposed modifications (a) curved plate over the window (b) direct inlet channel from the air handling system
25
(a)
(b) Fig. 21: Distributions of temperature and velocity magnitude in (a) surface crossed manikins 1 and 2 (b) surface crossed manikins 3 and 4
26
(a)
(b) Fig. 22: Velocity magnitude (a) and Temperature (b) distributions across lines 1 and 3 drawn in the front of the manikins
6. Conclusions In this research, a computational study was carried out to predict the airflow properties of HVAC system in the compartment of a passenger coach in the summer. This study also conducted an experiment in the studied compartment to verify the numerical results. The turbulent air flow was considered in the compartment and it was simulated by using v2-f turbulent model. The compartment was simulated in two cases: with seated and slept manikins. Temperature and velocity magnitude distributions were obtained in two cases, including the compartment with 27
manikins and without of manikins. The results showed that due to the position of inlet air channel in the compartment, the airflow entered from the HVAC system mostly inclines on the left side of the compartment and this causes the airflow to deviate from the thermal comfort conditions in some parts of the compartment and makes the seated/slept passengers uncomfortable. The results also showed that performing some simple modifications in the compartment design can improve the HVAC system performance and also make the air flow distribution symmetrical in the compartment.
References [1]. ASHRAE, 1992, Standard 55-Thermal Environmental Conditions for Human Occupancy. Atlanta, ASHRAE, Inc. [2]. Oakland innovation and information services Ltd; 18 Cambridge science park; Milton road ,Cambridge, cb40fh, thermal comfort on trains, 2002 [3] Tim Berlitz, Gerd Matschke. interior airflow simulation in railway rolling stock. Research and Technology Centre of the Deutsche Bahn AG, Division for aerodynamics and airconditioning, 2004 [4] Yongson Ooi,, Anjum Badruddin Irfan, ,Z.A. Zainala, P.A. Aswatha Narayana , Airflow analysis in an air conditioning room, Building and Environment 42(3), 1531–1537, 2007 [5]. Khodakarami, N. Nasrollahi,Thermal comfort in hospitals–A literature review”, Renewable and Sustainable Energy Reviews, 16, 4071–4077, 2012
[6]. Salmanzadeh, M.; Zahedi, G.; Ahmadi, G.; Marr, D.R.; and Glauser, M. Computational modeling of effects of thermal plume adjacent to the body on the indoor airflow and particle transport. Journal of Aerosol Science, 53, 29–39, 2012.
[7]. Gaever R. Van, Jacobs V.A., Diltoer M., Peeters L., Vanlanduit S., Thermal comfort of the surgical staff in the operating room, Building and Environment, 81, 37-41, 2014, 28
[8]. Xiufeng Yang. Ke Zhong, Hui Zhu,Yanming Kang, Experimental investigation on transient natural ventilation driven by thermal buoyancy, Building and Environment, 77, 29–39, 2014 [9] Piotr Uścinowicz, , Marta Chludzińska, Anna Bogdan, Thermal environment conditions in Polish operating rooms, Building and Environment, 94, 296–304, 2015, [10] Xiufeng Yang, Ke Zhong , Yanming Kang, Tianyin Tao, Numerical investigation on the airflow characteristics and thermal comfort in buoyancy-driven natural ventilation rooms, Energy and Buildings, 109, 255–266, 2015 [11] Nurullah Arslanoglu, Abdulvahap Yigit, Experimental and theoretical investigation of the effect of radiation heat flux on human thermal comfort, Energy and Buildings, 113, 23-29, 2016 [12]. Lihui Wang, Zhiping Du, Jianshun Zhang, Rui Chen, Shaolong Yu, Minglu Qu, Study on the thermal comfort characteristics under the vent with supplying air jets and cross-flows coupling in subway stations, Energy and Buildings, 131, 113-122, 2016 [13] Marie Ruellan, Herie Park, Rachid Bennacer, Residential building energy demand and thermal comfort: Thermal dynamics of electrical appliances and their impact, Energy and Buildings, Volume 130, 46-54, 2016, [14]. Mehrzad Ansaripour, Morteza Abdolzadeh, Saleh Sargazizadeh, Computational modeling of particle transport and distribution emitted from a Laserjet printer in a ventilated room with different ventilation configurations, Applied Thermal Engineering, 103, 920-933, 2016
[15] Ahmed Qasim Ahmed, Shian Gao, Ali Khaleel Kareem, Energy saving and indoor thermal comfort evaluation using a novel local exhaust ventilation system for office rooms, Applied Thermal Engineering, 110, 821-834, 2017 [16] Morteza Abdolzadeh, Mohammad Saleh Sargazizadeh, Mohammad Hadi Dehghan,
Quantitative and Qualitative Energy Assessments of Floor Heating and Skirting Board Heating Systems in a Room, Journal of Energy Engineering, 143(1), http://dx.doi.org/10.1061/(ASCE)EY.1943-7897.0000385, 2017 [17]. Lien F., Kalitzin G., Computations of transonic flow with the v2-f turbulence model. International Journal of Heat and Fluid Flow 22, 53–61. [18] Davidson L., Nielsen, P.V., Sveningsson A.,. Modification of the V2F model for computing the flow in a 3D wall jet. Turbulence Heat and Mass Transfer , 4, 577–584 ,2003.
29
[19]. Abdolzadeh . M,Mehrabian M.A, Soltani, A., Numerical study to predict the particle deposition under the influence of operating forces on a tilted surface in the turbulent flow, Advanced Powder Technology, 22(3), , 405-415, 2011 [20].Abdolzadeh M., Mehrabian M.A., Combined effect of thermophoretic force and other influencing parameters on the particle deposition rate on a tilted rough surface, International Journal of Thermal Sciences 50 (6), 954-964, 2011
Table captions: Table 1: Coefficients and source terms in Eq. (1) Table 2: The boundary conditions used in the present study Table 3: Checking the mesh dependency based on the flow properties Figure captions Fig. 1: Compartment geometry Fig. 2: The compartment with (a) seated manikins (b) slept manikins Fig.3: (a) The selected points for measuring the airflow properties (b) measuring the inlet air velocity and temperature (c) measuring the I air velocity and temperature close to the ground, (d) Measuring the ceiling temperature Fig 4: Generated mesh (a) slept manikins (b) seated manikins Fig. 5: Given boundary conditions in the main inlet and outlet sections Fig. 6: The predicted velocity and temperature at the locations in the experiment (a) points 1,3,6 (b) 2,3, and 4 (c) 5,6, and 7 (d)8, 9, and 10 Fig. 7: Comparison of predicted and measured air flow properties in the selected points (a) velocity (b) Temperature (c) relative error of velocity (d) relative error of temperature Fig. 8: Airflow velocity distribution (a) at inlet air channel (b) middle surface in the compartment
30
Fig. 9: Distributions of temperature and velocity in (a) Surface crossed manikins 1 and 2 (b) surface crossed manikins 3 and 4 Fig. 10: Airflow velocity variation form Manikin 1 to Manikin 2. Fig.11: Distributions of velocity (a) temperature (b) in surface crossed manikins 1 and 3 and surface crossed manikins 2 and 4 Fig. 12: Temperature distributions across vertical lines drawn in front of the manikins Fig. 13: Temperature distributions (b, c) across different lines in the front the manikins (a) Fig. 14: Temperature and velocity distributions in surfaces crossed (a) manikins’ upper body part (b) manikins’ lower body part Fig. 15: Temperature and velocity distributions in surfaces crossed manikins’ head Fig. 16: Velocity (a) and temperature (b) distributions across line AB (Fig 14) Fig. 17: Temperature distribution (b) across lines drawn over the lied manikins (a) Fig. 18: The issues of air distribution in the compartment (a) inlet air channel (b) non curved plate over the window Fig. 19: Modifications for improving the air distribution in the compartment Fig. 20: Air flow distribution after implementing the proposed modifications (a) curved plate over the window (b) direct inlet channel from the air handling system Fig. 21: Distributions of temperature and velocity in (a) surface crossed manikins 1 and 2 (b) surface crossed manikins 3 and 4 Fig. 22: Velocity (a) and Temperature (b) distributions across lines 1 and 3 drawn in the front of the manikins
31
Highlights
Computational simulation of airflow inside compartment of a passenger coach The turbulent air flow in the compartment was modeled using v2 f turbulent model. The thermal comfort conditions in the compartment were not evenly distributed
32