Cooling Load by Infiltration

Amount of infiltration to the vehicle depends on several conditions, like vehicle speed, position (Open/Closed) and leakage rate of doors and windows,

ventilation system operation (On/Off). Here ACH (Air Change per Hour Method) were used to calculate the infiltration-cooling load.

Here, in Table 3.31, the values for the vehicle situation is assumed as;

parked, doors and windows fully opened, ventilation on. And speed 60 Mph, doors and windows fully closed, A/C on.

Table 3.31. ACH values of a vehicle (Ott et. al., 2007)

Speed

(mph) Windows and Doors Ventilation System

Air Change Rate (1/h) Front

As a result of whole calculations, a proper A/C unit is selected to the vehicle.

After determining the A/C unit to be used in the vehicle, the unit was mounted on the test vehicle Photos of selected A/C unit, installation area on the roof of the vehicle and a view from inside the vehicle showing the inlet suction areas and evaporator outlets can be seen in Figure 3.14, 3.15 and 3.16 respectively.

Figure 3.14. Photo of the selected A/C unit

Figure 3.15. Installation area on the roof of the vehicle

Figure 3.16. A view from the test vehicle

3.2. Current Situation

In the current situation, as indicated previously, an air channel created by the boundaries of roof structure and luggage racks is employed in the vehicle. Also this air channel accommodates luggage rack connection brackets, heating water and refrigerant fluid pipes of A/C unit, electrical harnesses and some small brackets to mount electrical components. Since this structure of air channel creates a non-uniform air distribution, a non-non-uniform temperature distribution along the vehicle is observed.

Assemble of the roof top unit to the roof of the bus with the structure of the current air channel with a section view can be seen in Figure 3.17.

Figure 3.17. Current cross-section of the air channel and roof top A/C unit

In order to see the current situation of the vehicle and air conditioning channel in detail a set of tests was performed. During the tests, air velocity at the outlet of the nozzles, temperature distribution along the vehicle, temperature distribution in different levels of the passenger compartment (Head, Lumbar and ankle levels of the passenger) and internal noise level in different locations of the passenger cabin were measured.

3.2.1. Test Preparations

In order to determine the temperature distribution along the vehicle and in different levels of the passenger compartment, some of the passenger seats indicated

in Figure 3.18 and driver seat were equipped with thermo-couples as shown in Figure 3.19.

Figure 3.18. Seats equipped with thermo-couples

Figure 3.19. Thermo-Couple installation points of each seat (SAE, 2004)

Thermo-Couples were so positioned that to measure the temperature of the regions in head, lumbar and ankle region of the passenger. A photo of a sample seat thermo-couple installation and all the seats to be equipped with thermo-couples can be seen in Figure 3.20 and 3.21 respectively.

Figure 3.20. A photo of a sample seat thermo-couple installation

Figure 3.21. Seats to be equipped with thermo-couples

In order to see the effect of radiation heat transfer from sun light, a device measuring the effect of Sun light radiation was attached to the roof of the vehicle in a region near to the front. The installation photos of the radiation-measuring device can be seen in Figure 3.22.

Figure 3.22. Installation photos of the Sun light radiation-measuring device

After completing the installation of all the measurement sensors, data logger, which collects data from sensors and transfers to the computer, was installed to the vehicle. Data logger and the computer can be seen in Figure 3.23.

Figure 3.23. Data logger and the computer

The passenger compartment was allowed to soak in the sun for approximately 4 hours prior to initialization of airflow and data acquisition (Huang and Han, 2002).

The suitable time for soaking and testing was determined as follows;

Mansour et. al., (2007) developed a computer code according to ASHRAE, (1997) to estimate the cooling load variation (Figure 3.24).

Figure 3.24. Hourly cooling load distribution at different typical day times around year (Mansour et al, 2007)

As shown in Figure 3.24, the peak load occurred between 10.00 to 15.00 hours and this is commanding the AC system to operate at maximum capacity (Mansour et. al., 2007). The cooling load calculations conducted in the present study also showed that the maximum cooling load was achieved at 15:00 hour. So by referencing this information, soaking time is determined as between 10:00-14:00 and time to start testing was determined as 14:00.

After all the preparations were completed, the vehicle was parked under direct Sun light radiation with a magnitude of 890 W/m², and 38 °C of average ambient temperature, between 10:00 and 14:00 times, 4 hours. The vehicle so

positioned that the right side of vehicle was subjected to direct Sun light, which is the worst case in the cooling load calculations as determined in the current study. During the heat up stage the doors and the windows of the vehicle was kept closed. Photos of the test vehicle during heat up stage can be seen in Figure 3.25.

Figure 3.25. Photos of the test vehicle during heat up stage 3.2.2. A/C Performance Test and the Measurements

After the 4-hour heat up stage, average inside temperature of the vehicle was approximately 45 °C, even in some regions up to 50 °C. Because of the high rate of Sun light radiation, especially surface temperature of some plastic parts, which is dark colored and subjected to direct sun light, measured as 47 °C with a laser type thermometer. However, since the inside roof panels have isolation, light color and are not subjected to direct sun light, the temperature of the inside roof panels were observed as lower than the black parts like dashboard and steering wheel. Photos of surface temperature measurements of dashboard, steering wheel and inside roof panels can be seen in Figure 3.26, 3.27 and 3.28 respectively.

Figure 3.26. Measurement of surface temperature of the dashboard of the vehicle with laser thermometer

Figure 3.27. Measurement of surface temperature of the steering wheel of the vehicle with laser thermometer

Figure 3.28. Measurement of surface temperature of the inside roof panels of the vehicle with laser thermometer

The vehicle, whose inside material surface and inside air temperatures raised, was driven in Adana-Gaziantep highway with a constant speed of 90 km/h so that, the right side of the vehicle subjected to sun light during all the test. During the test, temperature distribution of the test vehicle and sun light radiation were measured and recorded by means of 41 thermo-couples (2 for outside ambient temperature) and radiation measurement device installed to the vehicle and data logger. At the same time, the exit velocities of the nozzles were measured with an anemometer and the results were tabulated. The figure showing the measurement of exit velocity of the nozzles can be seen in Figure 3.29.

Figure 3.29. Measurement of exit velocity of the nozzles 3.2.3. Test Results

Air velocities at the outlet of the nozzles of each passenger measured from the current vehicle is shown in Figure 3.30.

AIR VELOCITIES AT NOZZLES

LEFT 5,27 5,44 5,47 5,49 6,50 7,03 5,65 5,49 5,46 5,36 4,74 4,62 3,43 3,21 RIGHT 4,15 4,28 4,43 5,33 6,04 6,54 6,06 5,56 4,62 4,42 4,27 4,12 4,04 3,66

Drive

Figure 3.30. Air velocities at the nozzles measured from the current vehicle

As can be seen obviously from Figure 3.30, there are great amount of differences between air velocities at the outlet of each passengers. In the regions under the A/C unit the nozzle exit velocities are very high, while in the front and rear portions of the vehicle, where are subjected to high level of solar radiation and excessive heat transferred from the engine room, is very low. As a result of this uniform air distribution, temperature distribution along the vehicle is also non-uniform. This can also be observed from the results of the temperature distribution test.

At the end of 100-minute test, the temperature distribution in the level of head, lumbar and feet ankle of the passenger can be seen in Table 3.32.

Table 3.32. Temperature distribution along and different levels of the passenger

When Table 3.32 was investigated, because of the high rate of solar radiation from the front windshield of the vehicle, it was observed that, the starting temperature of the driver head region is about 51 °C. Since the rear glass of the vehicle is smaller than the front windshield, the temperature of the rear region is lower than the front but still it is higher than the middle portion of the vehicle. Also it was seen that, since the vehicle is stationary during the heat up stage, the air heated

up was lifted up and the effect of radiation of side glasses near to the head region, the temperatures of head regions in all locations are higher than the lumbar and ankle regions.

The graphic representing the temperature distribution along the vehicle in different levels at the 60^th minute of the test can be seen in Figure 3.31.

L6R6

Temperature Distribution at 60th min.

Head Lumbar Ankle

Figure 3.31. Temperature distribution at the 60^th minute of the test

Under normal conditions, it is targeted to reach an average inside temperature of 25 °C after a 60-minute test. (Temsa, 2007). When we investigate the test results without considering the differences between the regions, it can be said that the target was achieved, because the average temperature inside the vehicle after 60 minutes was determined as approximately 25 °C. On the other hand, from the temperature distribution point of view, there are serious differences between regions. For example; while the maximum temperature was observed in the driver head region as 33,1 °C, minimum temperature was recorded as 20 °C in L3 and R3 ankle regions, where the locations just under the A/C unit. At the end of 60^th minute, although the temperatures in the regions near to the A/C unit (L2, L3, R2, R3 and R4) were under 23 °C. In front and rear regions of the vehicle, temperatures changed between 27-30

°C because of the non-uniform air distribution. In that situation, while the passengers in the middle regions, where is near to the A/C unit, complaint about the low temperatures especially in some ankle regions about 20 °C, the driver and passengers in the front/rear regions of the vehicle complaints about the insufficient performance of the A/C unit.

As the test continued, it was observed that temperature values in all points (Except the driver head), could reach the target value of 25 °C at the end of 100^th minute. However, in that situation, the average inside temperature of the vehicle reached to 22 °C and temperature of some regions were 19 °C.

Vertical temperature difference can be defined as the air temperature difference between the head and ankle levels of a person. Generally it is not a satisfied situation, which the air temperature around head level is high, while around ankle level is low. (Fanger, 1970). 3°C of vertical temperature difference, for a seated person can be taken as the comfort limit according to standard ISO 7730, (2006). In Table 3.33, vertical temperature difference along the vehicle can be seen.

Table 3.33. Vertical temperature difference along the vehicle

Vertical Temperature Difference (°C) Head-Lumbar Lumbar-Ankle Head-Ankle

3,1 2,0 5,1

In the current situation, in the most of the regions, the vertical temperature difference can be observed over 3 °C, which creates an uncomfortable media to the passengers.

On the other hand, another important source of uncomfortable media is the temperature difference along the vehicle. Maximum/minimum and difference values of the temperatures in different levels of the passenger compartment can be seen in Table 3.34.

Table 3.34. Temperature differences along the vehicle

Level Max difference along the vehicle is very high.

As a result, since the period of time to be reached the comfort level of the passengers in the front and rear portions of the vehicle is longer than the middle region, passengers in the middle region of the vehicle complaint about the low temperatures. Also during the operation of A/C unit, since the compressor, which is driven by the engine, is working, fuel consumption of the vehicle will increase.

Improving air conditioning performance and occupant thermal comfort requires an understanding of the fluid motion prevailing in the compartment for any given ventilation setting and passenger loading (Jalil and Alwan, 2007). So, as a result of all these tests, measurements and comments, since the main source of the non-uniform temperature distribution is non-uniform air distribution, it is obvious that the vehicle’s A/C system must be improved from uniform air distribution point of view.

Another comfort parameter of a passenger is the internal noise level. And A/C evaporator fans play an important role for this situation. In the current situation, the time needed to reach the set temperature is very long, and during this period of time, fan speed level is very high, generally maximum. Consequently the level of internal noise created by the evaporator fans is in a very high level. Internal noise measurements were performed in the current vehicles, front, middle and rear portions of the vehicle with a decibel meter to determine this effect.

Some Photos, representing the measurement locations (Front, middle and rear) of the test vehicle can be seen in Figure 3.32, 3.33, 3.3. And internal noise values measured in these locations with different fan speeds and without A/C can be seen in Table 3.35.

Figure 3.32. Internal noise measurement point, front

Figure 3.33. Internal noise measurement point, middle

Figure 3.34. Internal noise measurement point, rear

Table 3.35. Internal noise values in different locations with different fan speed levels

Fan Level Internal Noise (Decibel) Front Middle Rear No A/C 51,7 55,2 57,8 Level 1 51,8 56,0 57,9 Level 2 52,3 57,7 58,1 Level 3 54,1 61,2 58,5 Level 4 60,6 68,9 61,4 Level 5 65,7 74,5 66,3 Level 6 65,6 74,6 66,5

As can be seen from the noise values in Table 3.35, A/C unit fan noise plays an important role for the internal noise issue, especially the high values of internal noise in the middle region of the vehicle, where is the underneath of the A/C unit shows this effect obviously.

Since inside of the vehicle cannot reach set value fast, the A/C control unit sets the fan speed to higher levels and consequently internal noise level is higher. So improving the A/C performance will also improve the internal noise level by employing lower fan levels.

3.3. Design Of Air Channel By Means Of CFD 3.3.1. Determination Of Channel Geometry

Tests conducted with the vehicle having current air channel showed us that a new air channel providing uniform air distribution along the vehicle must be designed and applied to the vehicle.

With the conducted study, instead of using a channel created by the boundaries of roof structure and luggage racks, accommodating luggage rack connection brackets, heating water and refrigerant fluid pipes of A/C unit, electrical harnesses and some small brackets to mount electrical components. A separated air channel, which is transferring only air and no other components like pipes, harnesses, hoses, or brackets and providing uniform air distribution along the coach, was designed by means of CFD.

So, base shape and dimensions of this channel must be determined by investigating the current air channel inside.

Geometrical examinations were performed and it is determined that the most suitable cross-section geometry is circle instead of a rectangle or a different polygon.

Because, Circle cross-section air channels provide less pressure drop, uniform streamlines and opportunity of using standard parts. Manufacturing and availability of standard part in market of a circle cross-section is good and since it has no corner, circle cross-section provides homogeneous flow inside it (Carrier, 2004). So, employing a circle cross-section channel will be better from lots of point of views

Considering the space inside the luggage rack of the coach, hoses, brackets and electrical components, it is determined that the maximum outer diameter of the channel can be 160 mm.

By taking account the position and dimension of outlets of A/C unit evaporator and air nozzles, geometry of inlet and outlet channels were determined as shown in Figure 3.35.

Figure 3.35. Section view of designed air channel in current luggage rack

After decided the cross-section dimensions of the channel, length of the channel was also determined for the 12-m vehicle as 10,2 m. Channel position on the vehicle can be seen in Figure 3.36.

Figure 3.36. Position of the air channel on the vehicle

The next step in designing the channel is to determine the inlet and outlet channel locations. For this purpose, positions of evaporator fan exit of the A/C unit and seat layout were considered and this first air channel was designed as shown in following figures (From Figure 3.37 to 3.40).

Figure 3.37. Cross-section view of the base channel

Figure 3.38. Air inlet region detail of the base channel

Figure 3.39. Front and rear end details of the base channel

Figure 3.40. Isometric 3D view of the base channel

3.3.2. CFD Analysis Studies

The computer has become a very valuable resource, and the continual increase in its performance gives great promises for the future. One of the areas that the computer has had a large impact in is fluid dynamics. Great progress has been made over the past several decades to incorporate computers into the prediction of fluid flow. The flow field around a complete aircraft configuration can now be calculated with the help of computers. The acronym CFD (Computational Fluid Dynamics) represents numerical solutions to fluid problems by solving some form of the governing equations of fluid motion. Most complex engineering problems involve analysis from both CFD and experimental testing. The use of both of these approaches allows a more complete and detailed view of the flow field, enhancing the final solution. At the present growth rate of technology, research in both of these areas will continue to remain important and necessary in both the design and analysis of engineering problems. (Neel, 1997)

For laminar flows, solutions to the Navier-Stokes equations are considered to be as accurate as numerical computations can be. But unfortunately, most problems where viscous effects are important are classified as turbulent flows. The unsteady Navier-Stokes equations have the ability to resolve all the small-scale structure of turbulent flow. The problem is encountered with the enormous number of grid points required to capture all the physics of the flow. Therefore, it is more common to solve the Reynolds-Averaged Navier-Stokes (RANS) equations. These equations are the highest level of approximation. The equations solve for the mean flow field, which in turn requires a turbulence model for closure. In the world of CFD, there is no one turbulence model that is general enough for all flow conditions. Instead, each problem will need to be studied and have an appropriate model attached to it. (Neel, 1997).

3.3.2.1. Production of Turbulence Model Determination Experimental Setup

Solving CFD problem usually consists of four main components: geometry and grid generation, setting-up a physical model, solving it and post-processing the computed data. The way geometry and grid are generated, the set problem is computed and the way acquired data is presented is very well known. Precise theory is available. Unfortunately, that is not true for setting-up a physical model for turbulence flows. The problem is that one tries to model very complex phenomena with a model as simple as possible. (Sodja, 2007).

The most important stage of a CFD analysis study is to determine the most suitable turbulence model for your flow domain geometry.

In the present study, in order to determine the most suitable turbulence model, an experimental study was carried out. In the previous section general geometry of

In the present study, in order to determine the most suitable turbulence model, an experimental study was carried out. In the previous section general geometry of