INTERNATIONAL HVAC+R & SANITARY TECHNOLOGY SYMPOSIUM March 31 -April 2, 2016 - istanbul
Organized by Turkish Society of HVAC and Sanitary Engineers
SPONSORS
� iSiB Partner Sponsor
TURKISH HVAC·R EXPORTERS Platinum Sponsors
Gold Sponsors
·-
Silver Sponsors
OAl«IN D�...�
P..".!.'.1.-.! � .... �m-1 -
i,.-
CfYl.tftT'IM HSK TRox·ncHHUt .......
LINTES -
v••IMANN wilo
Supporters
-@
.s, , .li!I..
�x �
ISKID @� 3 ��
m S:: .,_=
sestAo
[Abstract:0054][Comfort Cooling]
INVESTIGATION OF THE EFFECT OF AIR FLOW MALDISTRIBUTION ON EVAPORATOR THERMAL PERFORMANCE Ergin BAYRAK1,2, Alp Er Şevki KONUKMAN2 Friterm A.Ş Research and Development Department, Tuzla 34957, Istanbul, Turkey 2 Gebze Technical University, Department of Mechanical Engineering, Energy Systems Division, Gebze 41400, Kocaeli, Turkey Corresponding email:
[email protected] 1
SUMMARY Evaporators have some problems such as air velocity, temperature and humidity maldistribution depending on fan characteristics, limitations of heat exchanger and fouling. When designing an evaporator, the parameters taken from manufacturer is used and admitted that air flow where passes on evaporator surface is equal at each point of flow surface. Since this is not possible in practice, this issue directly affects the evaporation rate of internal two-phase fluid. In the literature, the study on airflow maldistribution (AMD) has not been documented very well due to difficulties of elaborate measurements. In the studies performed, this difficulty was handled by changing the path of internal fluid. In this study, effect of inlet air velocity maldistribution on evaporator thermal performance was investigated experimentally. Accordingly, an evaporator, which is consistent with the dimensional limits of laboratory air channel, was designed and identified the distribution and AMD of air velocity passing on each module by using an anemometer. Each circuit of evaporator was considered as an individual module and thermal capacities corresponding to each circuit was calculated by using FrtCoils software considering measurement outcomes. Eventually, non-uniform total capacity was determined and the difference with uniform case was detected. In order to confirm as experimental, the product was tested at calorimetric room at first and then tested at conditioning room ensured same airflow rate but more uniform flow. The capacity difference between experimental and analytical results was seemed to be very close, thus validated the effect of airflow maldistribution on performance. INTRODUCTION The finned tube heat exchangers are used as evaporator and condenser at most of the HVAC (Heating Ventilating Air Conditioning) systems. These devices have some problems due to the incorrect design and the incorrect usage. The most important problems are the non-uniform distribution of both internal and external fluids in the evaporator. In the literature, most of the researchers investigated the effects of internal fluid distribution instead of external flow media distribution. And, practitioners also tried to solve the problem of maldistribution of internal flow. Thereagainst, this study focus on maldistribution of external flow. The effect of maldistribution of external flow on the capacity of the evaporator first investigated computationally via commercial software FrtCoils and then compared with the results of the experiments performed in the Friterm Research and Development Laboratory. The large number of materials may cause in blockage in front of evaporator, which is taken inlet air flow, thereby this issue may induce decreasing air flow rate or non-uniform air flow maldistribution. Datta’s study [1] investigates blockage effect on overall performance of system in automotive cooling systems by creating various type blockage effects. As a result of performing 52 experiments, depending on severity of air maldistribution, cooling capacity decreased by 8.16% for an area blockage of %50 compared with the normal operating condition. Chen [2] has investigated the effect of airflow maldistribution on heat exchanger performance depending on oblique angel of inlet air velocity. Eventually, he has deducted an equation defining the decreasing of evaporator capacity related with oblique angle tightly. Jianying [3] has claimed that some important parameters may be affected some extend depending on AMD. In the course of the experimental analysis, the air velocities were measured at 56 points at inlet of evaporator and determined three different AMD and these degrees according to relative standard deviation formula and then each AMD were entitled as uniform, non-uniform and seriously non-uniform at %18, %49 and %93, respectively. Domanski [4] has considered the effects of air side and refrigerant side maldistribution on the coil capacity. Experimental results have showed that maldistributed air also affects refrigerant distribution, which caused further coil capacity degradation. Another study belonging to the same author [5], it has been detected that the maximum capacity degradation depending on refrigerant and air side are as much as %30 and %8,7, respectively. Besides, [6] and [7] showed that nearly all of the capacity reduction due to nonuniformities in the velocity profile can be recuperated by simply redesigning the tube-to-tube connection sequence. Lee [8]’s numerical and experimental study performed by taking into account of airflow measurement on air cooled condenser by dividing to different segments of air cooled condenser has investigated the effects of different included angles between the air cooled condenser (V type) performance. Consequently, they have detected that changing the angles of coils has considerable effects on airflow distribution and therefore heat transfer performance can be increased %5,29 respectively. Kim [9] has
167
investigated the using of flow balance valve at inlet and outlet of evaporator with respect to AMD in order to recover of loss cooling capacity and COP resulting from non-uniform air flow distribution. For 40% air flow maldistribution, reducing capacity ratio was about %6. Within the scope of this study, an evaporator is designed by considering the test operation as unit cooler used with its fan and DX evaporator used by detaching the axial fan of unit cooler as it can show as image and specification at Figure 1 and Table 1, respectively. Then this product will test at calorimetric room and conditioning room, which have different air maldistribution in front of the evaporators. Thanks to these tests, we would learn not only if air maldistribution identified have an impact on heat transfer rate or not, but also FrtCoils program would verify because this program doesn’t take into account of changing of internal flow characteristic as depend on airflow maldistribution.
Figure 4. General view and dimensional characteristics of tested unit Table 1. Geometric parameters of tested unit Values
Geometric parameters Number of rows Number of tubes per row Transverse tube pitch Longitudinal tube pitch Tube length Tube diameter (inner/outer) Fin thickness Fin spacing Fin type Fin height Tube arrangement
4 22 35 mm 35 mm 745 mm 11.86 mm/ 12.5 mm 0.15 mm 7 mm Flat 770 mm inline
METHODS Two different experimental setups for the capacity measurement and a hot bulb anemometer to identify the air flow distribution formed at the entire face are used in the scope of this study in order to investigate whether the air flow maldistribution has significant effect on heat transfer rate or not. On the other hand, a commercial program entitled as FrtCoils is utilized in order to understand whether flow maldistribution on air side has got any effect on the characteristic of internal fluid. Figure 3 and 4 show the experimental setup and its general schematic perspective including each room together respectively used in this study, including calorimetric and conditioning room. It must be noted that experimental setup illustrated in Figure has been shown for one of the rooms but this diagram is valid for other one too. The refrigerant used in these experiments is R404A. These test setups of two rooms consist of a test unit, air handling unit for conditioning of air, which have humidifier, heaters and centrifugal fans, refrigeration line for regulating of temperature, pressure and flow rate of the refrigerant. The only difference of these room, as the air flow rate can be adjusted with the aid of wind tunnel of conditioning room and calculated of air capacity according to outcomes of sensors. It doesn’t matter for calorimetric room because unit coolers having constant air flow rate have already been used for test operations and the air capacity for this room is calculated from electrical loads.
168
Figure 5. P&I diagram of experimental set-up
Figure 6. Schematic view of experimental set-up The test processes have been carried out according to Eurovent Standard [10] meticulously. The test results have been taken when the capacity difference between air side capacity and refrigerant side capacity is less then %4. The test duration is about 5 hours. Air flow measurement was carried out at 80 points of each circuit via a hot bulb anemometer, which is only 3 mm diameter and placed at a manual traverse system as shown at Figure 5. The airflow distribution map occurring in front of evaporator surface at each circuit was created according to measurement outcomes. Moreover, the uncertainty value of this measurement was taken into account in the following section of study and specified at Table 2.
169
Figure 7. Hot bulb anemometer used for air flow measurements
Figure 6. Each circuit performed measurement A mathematical method was adopted in order to describe the air flow maldistribution (AMD) similar as Jianying’s study [3]. This formula is the following: 𝜎= √
1 𝑛−1
𝑣𝑖− 𝑣̅ 2 ) 𝑣̅
∑𝑛𝑖=1(
(1)
Where 𝑣𝑖 is the air velocity measured, 𝑣̅ is the average value of the air velocities measured and 𝑛 is the number of test points. The uncertainty values for the air velocity measurements, temperature and calculated total capacity according to Stephanie Bell’s study [11] is presented below. Table 2. Uncertainties in experimental measurements Variable
Max. Uncertainty
Capacity [kW]
∓2.47%
Air velocity [m/s]
∓ (Reading valueൈ0.5+0.03)
Thermocouples (T type)
∓0,3°C
170
RESULTS
Table 3 shows that the values of AMD for both each circuit and entire coil. The first set of results belonging to conditioning room were obtained from the wind tunnel measurements which has no contraction part at inlet of air. In order to ensure more uniform flow, a contraction part for wind tunnel was designed and installed in conditioning room, therefore an improvement trend at outlet temperatures of refrigerant detected. The results without contraction; 6.6°C, 7.8°C, 7.7°C, 6.9 °C and with contraction; 6.6°C, 7.5°C, 7.3°C, 6.5°C. Depending on these results, it was supposed somewhat AMD improvement about by between %2-%5, that is close to %10. Determining of this improvement clearly wasn’t possible due to difficulties of measurements in contraction part. With reference of this detection, test operations and its outcomes were evaluated with respect to this situation. It was seen a deviation of measured velocities between two tests due to accuracy of probe but these values are in the limits of uncertainty range. Table 3. The value of AMD occurring at each circuits and entire coil Air Maldistribution Degrees Calorimetric Room
Conditioning Room Mean Velocity AMD (m/s)
Mean Velocity (m/s)
AMD
Circuit 1
4,6035
14,75%
3,234854167
16,93%
Circuit 2
4,285
18,43%
3,818677637
10,35%
Circuit 3 Circuit 4
3,79875
17,17%
4,105001488
12,20%
4,464375 4,28790625
15,95%
3,866578571 3,756277966
15,7% (∓1,60%)
Circuits
The entire coil
21,6% (∓1,87%)
15,02%
a) b) Image 1. Airflow distribution maps a) calorimetric room b) conditioning room During the test process, the most important parameters such as evaporation pressure, air inlet temperature and relative humidity were ensured quite stable as presented in Table 4. Also, the obtained heat transfer rates are very close and in the limits of uncertainty range, that are 7.567 kW and 7.439 kW.
Table 4. The average value of test parameters Test Room
Coil
mr (kg/h)
RH (%)
Q (kW)
with fan
21,6
212,1
0,600
10
36,6
7,567
Conditioning Room
no fan
about 10-13
208,35
0,600
9,99
22,88
7,439
171
Pevap. (bar)
Tinlet (°C)
AMD (%)
Calorimetric Room
Test Result (kW ) 7,8 7,6 7,4 7,2 7
7,733 7,622 7,255
7,38
Min
Max
Min
Calorimetric Room
Max Conditioning Room
Figure 8. The test results for each room The effect of air maldistribution on refrigerant side is seen clearly in Table 5. Owing to more uniform flow media at conditioning room, the outlet temperatures have approached to desired value, which is 6.5°C. Table 5. The refrigerant outlet temperature values measured via thermocouple 8,5
temperatures (°C)
8 7,5 7 6,5 6 5,5 5 circuit 1 circuit 2 circuit 3 circuit 4 Desired Outlet Temperatures Calorimetric Room Conditioning Room
Table 6 and 7 show the capacity outcomes calculated by means of FrtCoils software considering measured air velocities and uniform situation. Table 6. FrtCoils results according to measurement of conditioning room Circuits (from up to bottom)
AMD (%)
Measured velocity (m/s)
1
16,93
3,235
2
10,35
3,82
3
12,20
4,104
4
15,06
3,86
Uncertainty range (m/s) 3,42675 3,04325 4,041 3,599 4,3392 3,8688 4,083 3,637 Max. capacity Min. Capacity
172
FRT Capacity considering outlet temp. (kW) 1,781 1,689 1,856 1,767 1,931 1,836 1,925 1,832 7,493 7,124
FRT Capacity for uniform cond. (for 3,94 m/s) 1,9 1,9 1,9 1,9 7,6
Table 7. Frt Coils results according to measurement of calorimetric room Circuits (from up to bottom)
AMD (%)
Measured velocity (m/s)
1
14,75
4,6035
2
18,43
4,285
3
17,17
3,798
4
15,95
4,464
Uncertainty range (m/s) 4,863675 4,343325 4,52925 4,04075 4,0179 3,5781 4,7172 4,2108 Max. capacity Min. Capacity
FRT Capacity considering outlet temp. (kW) 2,059 1,957 1,901 1,833 1,829 1,732 1,955 2,002 7,744 7,524
FRT Capacity for uniform cond. (for 3,94 m/s) 1,9 1,9 1,9 1,9 7,6
Table 8 and Figure 9 demonstrate the percentage of capacity deterioration depending on FrtCoils and experiments separately. Table 8. The comparison of FrtCoils and experimental results Max. capacity deterioration (%) Test Room Calorim etric Room Conditio ning Room
Test Result (kW) Min
7,38
Max
7,733
Min Max
7,255
FrtCoils result for uniform flow (kW)
FrtCoils result for nonuniform flow (kW) 7,524
7,6
Frt Nonuniform resultFrt design capacity
Test Nonuniform resultFrt design capacity
-1,00%
-2,89%
-6,26%
-4,54%
7,744 7,124
7,6
7,622
7,493
CapacityDeterioration%
1,00%
0,00% -1,00%
Min
Max
Min
Calorimetric Room
Max Conditioning Room
-2,00%
-3,00% -4,00% -5,00% -6,00% -7,00%
Max. capacity deterioration (%) Frt Nonuniform result-Frt design capacity Max. capacity deterioration (%) Test Nonuniform result-Frt design capacity Figure 9. The percentage of capacity deterioration
173
DISCUSSION In this study, the impact of air flow distribution on evaporator with specific design has been investigated. FrtCoils software and experimental results have verified that although it has been ensured a little improvement at outlet of refrigerant circuits by more uniform air flow, no significant change at heat transfer rate was observed. Hence, the AMD up to about 19% doesn’t have significant impact on the heat transfer rate, which is consistent with Jianying’s study [3]. Therefore, unit cooler investigating in this case study can be used easily without considering air flow effects. Furthermore, the heat transfer rate outcomes taken with FrtCoils and experiments are quite close because of the fact that the AMD identified doesn’t affect the thermal characteristic of internal flow. Further research effort should be considered not only higher AMD but also the blockage types at same AMD, which may occur at practical application. Moreover, the air velocity measurement requires extra experimental effort, so that it should be studied to develop new air flow simulation methods such as CFD tools for typical configurations so as to facilitate the air velocity determination. ACKNOWLEDGEMENT Special thanks to Friterm A.Ş. and Dr. Hüseyin Onbaşıoğlu for financial support and assistance to this study. REFERENCES 1.
Datta, S.P., Das, P.K., Mukhopadhyay, S. 2014. Obstructed airflow through the condenser of an automotive air conditioner, effects on the condenser and the overall performance of the system, Applied Thermal Engineering, Vol. 70, pp 925-934.
2.
Chen, N., Xu L., Feng, H. D., Yang C. G. 2005. Performance investigation of a finned tube evaporator under the oblique frontal air velocity distribution, Applied Thermal Engineering, Vol. 25, pp 113–125.
3.
Jianying, G., Tieyu G., Xiuling, Y., Dong, H. 2008. Effects of air flow maldistribution on refrigeration system dynamics of an air source heat pump chiller under frosting conditions, Energy Conversion and Management, Vol 49, pp 1645–1651.
4.
Lee, J., Domanski, P.A. 1997. Impact of air and refrigerant maldistributions on the performance of finned-tube evaporators with R-22 and R-407C, Building Environment Division National Institute of Standards of Technology Final Report, U.S. Department of Commerce, Gaithersburg, Maryland.
5.
Choi, J.M., Payne, W.V., and Domanski, P.A. 2003. Effects of Non-Uniform Refrigerant and Air Flow Distributions on Finned- Tube Evaporator Performance, International Congress of Refrigeration 2003, Washington, D.C.
6.
Domanski, P.A., Yashar D., Kaufman, K.A., Michalski, R.S. 2004. Optimized design of finned-tube evaporators using learnable evolution methods, International Journal HVAC&R Research, Vol. 10 (2), pp 201-212.
7.
Yashar D., Lee, S., Domanski, P.A. 2015. Rooftop air-conditioning unit performance improvement using refrigerant circuitry optimization, Applied Thermal Engineering, Vol. 83, pp 81-87.
8.
Lee, T.S., Wu, W.C., Y.K. Chuah, Y.K. 2010. An improvement of airflow and heat transfer performance of multi-coil condensers by different coil configurations, Int. J. Refrigeration, Vol. 33 (7), pp 1370-1376.
9.
Kim, J., Braun, J.E., Groll, E.A. 2008. Analysis of Refrigerant Flow Distribution in Evaporators, International Refrigeration and Air Conditioning Conference, Paper 966, Purdue University, USA.
10. EUROVENT. 2010. Eurovent Rating Standard for Direct Expansion Forced Convection Unit Air Cooler for Refrigeration. Eurovent Certification Company, Paris, France.
174
