Experimental evaluation of a downsized residential air distribution system: comfort and ventilation effectiveness.
Good air mixing not only improves thermal comfort but also enhances ventilation effectiveness by inducing uniform supply-air diffusion. In general, the performance of an air distribution system in terms of comfort and ventilation effectiveness is influenced by the supply air temperature, velocity, and flow rate, all of which are in part dictated by the HVAC system as well as the thermal load attributes. Any potential deficiencies associated with these design variables can be further exacerbated by an improper proximity of the supply and return outlets with respect to the thermal and geometrical characteristics of the indoor space. For high-performance houses, the factors influencing air distribution performance take on an even greater significance because of a reduced supply-air design flow rate resulting from downsized HVAC systems.
The downsizing results from implementation of energysaving measures that lead to a significant reduction in cooling and heating loads. For example, the central HVAC system of a standard-practice residential building is typically sized to provide a cooling capacity of 3.5 kW (1 ton) of refrigeration per 45-60 m2 (500-600 ft2) of the floor area. This rule of thumb translates into a nominal supply-airflow rate of about 165- 215 L/s (350-450 cfm) for the same range of area. In comparison, in an efficient home, the design cooling capacity and, consequently, the supply-airflow rate may be reduced by 20% to 50%, which can adversely affect the air distribution performance. This is a major concern in the heating mode for cases incorporating high sidewall diffusers that can cause significant stratification, particularly during the recovery from setback--a period in which a greater-than-normal temperature difference between the supply air and room can occur. In a recent theoretical study (Jalalzadeh-Azar et al. 2006), CFD simulations were carried out to evaluate the performance of an air-distribution system similar to that of the current project in both cooling and heating modes. The results of that study indicated that the indoor air was better mixed in the cooling mode but was significantly stratified during the heating process. The primary objective of that paper was to quantify the impact of indoor air distribution on the energy performance of a heat pump system and to address the implications regarding the comfort of the occupants in energy-efficient residential buildings.
This paper presents and analyzes the results of an experimental study on a specific air-distribution system in heating mode during recovery from setback. The system of interest incorporates a high sidewall supply air diffuser and a wallmounted, near-floor return air grille located directly below the diffuser. The tests discussed in this paper were conducted following the debugging of the experimental facility and the qualification tests (Jalalzadeh-Azar 2006). Although knowledge of the spatial temperature variation within a space is useful for assessment of comfort and HVAC system performance, it does not necessarily offer a clue on ventilation effectiveness. Therefore, in parallel with temperature measurements, tracer-gas analyses were also performed to evaluate this important indoor air quality parameter. The experimental approach implemented only a single gas analyzer and limited the measurements to a few test points with emphasis on the return-air gas concentration. The return air data were compared with the analytical solutions of an ideal case where the room air was assumed to be perfectly mixed. The results of this theoretical scenario provided a frame of reference for evaluating the air-mixing effectiveness of the system under study. Furthermore, use of a nondimensional parameter that normalized the variation of the return-air gas concentration relative to that of the supply airstream was useful in negating the effect of the supply-air concentration setpoint on the results.
The diffuser type examined in this paper is J&J 90 V, 20 X 10 cm (8 X 4 in.), with 0[degrees] deflection. Experiments were conducted in an empty test room in the absence of any internal heat sources. To simulate recovery from setback, a heating process followed a precooling phase. The heat losses through the test room envelope were relatively small; therefore, the possible effects of the exterior walls, roof, and windows have not been addressed in the study. Future work will investigate other types of diffusers and will address the effects of nonadiabatic boundaries as well.
Figure 1 illustrates the experimental facility consisting of a supply-air conditioning apparatus, a test room, an exhaust system, and instruments for measuring experimental variables at the principal points. The supply air apparatus incorporates a variable-speed fan, a tracer-gas ([SF.sub.6]) injection system, and a 6 kW electric heater. The supply airstream passes through a flow straightener in a well-insulated plenum before entering the room. A bypass line is provided to allow preconditioning of the supply airstream before introducing it to the test room. The mode of operation (i.e., bypass vs. normal) is set by an assembly of gate dampers (Figure 1).
[FIGURE 1 OMITTED]
The test room, measuring 7.16 X 4.30 X 2.74 m (23.5 X 14.1 X 9 ft), is precooled by circulating air between the room and the adjacent walk-in cooler (not shown). The envelope of the test room is well insulated. To achieve spatial temperature uniformity within the test room at the beginning of each experiment, a portable mixing fan is operated during the precooling process. To capture spatial variation of properties (temperature and gas concentration) within the test room, an instrumentation matrix was adopted as shown in Figure 2. The measurements in the room were made across the test planes A, 273.7 cm (108 in.) above the finished floor (AFF); B, 242.6 cm (95.5 in.) AFF; C, 168.4 cm (66.3 in.) AFF; D, 94.5 cm (37.2 in.) AFF; E, 20.3 cm (8 in.) AFF; and F, 0.3 cm (0.1 in.) AFF. Note that (1) test planes C, D, E, and F are equally spaced and (2) instrumentation planes B and E are in alignment with the centerlines of the supply air diffuser and the return air grille, respectively. For the air temperatures on test planes B and E along the instrumentation line adjacent to the east wall, the supply air and return air temperature measurements were used, respectively, in lieu of additional measurements. (The supply air and return air temperatures were represented by single-point measurements at the center of each respective plenum, and the uncertainties associated with the spatial temperature variations were taken into account (JalalzadehAzar 2006). No measurements were made near the ceiling or the floor at this east-wall array. Because of the room symmetry with respect to the supply air and return air outlets, no measurements were taken near the south wall either.
[FIGURE 2 OMITTED]
The operating variables measured in the tests included the supply airflow rate, temperature, and tracer gas concentration. The measuring devices are described as follows:
* The temperature variables were measured using T-type thermocouples, appropriate for the temperature range of interest, 20[degrees]C-60[degrees]C (60[degrees]F-140[degrees]F). The temperature data were collected at one-minute intervals.
* The gas concentration measurements were taken by an INNOVA AirTech Instruments 1303 Multipoint Sampler and Doser. However, because (1) the ports are activated sequentially and (2) sampling and measuring the gas concentration takes about two minutes for each channel, only two of the ports were used in this project to increase the sampling frequencies. One of the ports was allocated to the return airstream, and the other one was used to measure the gas concentration at a single point within the test room. This arrangement allowed sampling at four-minute time intervals and, consequently, provided more data on the transient Return-air gas concentration. As for the supply airstream, the gas concentration remained fairly constant because the supply-airflow rate and the amount of injected [SF.sub.6] were virtually kept constant throughout each test. In all tests, the flow rate of the tracer gas was maintained at 50 std. [cm.sup.3] per minute by an automatic valve and a flow-controlling device (PORTER Instrument CM-2 Module). Therefore, a few measurements of the supply-air gas concentration throughout a given test (particularly before and after the heating process) were quite sufficient for the tests involving a constant supply-airflow rate. To do so, a sampling port was borrowed to measure the gas concentration of the supply airstream in the bypass mode. The measurement was then adjusted to account for the difference in the flow rates between the normal-test mode and the bypass mode.
* The supply airflow rate was measured by a flow nozzle with a TSI mass flowmeter. The measured data were in SCFM. The accuracy of the flowmeter readings were confirmed through calculations based on the constant mass flow rate of the tracer gas and the gas concentration measurements in the supply airstream.
A Campbell Scientific CR10 was used for data acquisition, except for the tracer gas data. Up to 25 channels of this system were available for temperature measurements. The tracer gas data were collected independently by a different data acquisition system (DAS) from INNOVA AirTech Instruments. The two DAS systems were controlled by two separate computers that were synchronized at the beginning of each experimental test.
DESCRIPTION OF METHOD
As a means to evaluate the mixing effectiveness of the air distribution system under various operating conditions, a theoretical case was considered as a reference scenario in which the indoor air was assumed to be perfectly mixed. The actual return air concentration data were compared with those of the theoretical case where the return-air gas concentration represented the concentration throughout the space at any given time. The solution to the theoretical scenario is determined based on the following continuity equation for the tracer gas:
[Q.sub.S.A].[[C.sub.[SF.sub.6],S.A.] - [C.sub.[SF.sub.6],R.A.]][congruent to][V.sub.Room][[d[C.sub.SF,Room]]/[dt]] (1)
where [C.sub.SF6,Room] is the gas concentration within the room and is uniform throughout the space. The time variable t denotes the time elapsed from the commencement of the test when the supply airstream is directed to the room by closing the bypass line and opening the supply air and return air dampers. The solution to this ordinary differential equation takes on the following form:
[C.sub.[SF.sub.6],R.A.] = [C.sub.[SF.sub.6],S.A.][1 - EXP( - [[Q.sub.S.A.]/[V.sub.Room]]t)] (2)
An important aspect of air distribution is the air change effectiveness defined as (ASHRAE 2005a):
[epsilon] = [[tau]/[t.sub.age]]
where the nominal time constant [tau] and the nominal age of the air [t.sub.age] are, respectively, determined from:
[tau] = [[V.sub.Room]/[Q.sub.S.A.]] (4)
[t.sub.[age].sup. * ] = [[integral].sub.0.sup.[infinity]][[[C.sub.[SF.sub.6],S.A.] - [C.sub.[SF.sub.6],R.A.]]/[[C.sub.[SF.sub.6],S.A.][C.sub.Room,Initial]]]dt = [[integral].sub.0.sup.[infinity]][xi]dt (5)
The initial room gas concentration, [C.sub.Room,Initial], is zero, as the growth (step-up) method is adopted for the tracer gas analysis in this study. The air change effectiveness is essentially a measure of the ventilation effectiveness of the air distribution system in removing the pollutant from the indoors. For perfectly mixed indoor air, the value of [epsilon] approaches unity. Because the duration of the heating process in the experiments was limited, Equation 5 was applied to the isothermal tests only. However, the integrant of Equation 5 signifies a normalized relative concentration, [xi], which is used here for an assessment of air distribution performance under different supply air operating conditions. At a given time during the heating process, the smaller the variable [xi], the more effective is the air mixing, given the positions of the S.A. and R.A. outlets. In a prolonged isothermal test, [xi][right arrow]0 as t[right arrow][infinity].
In this study, uncertainty analyses were performed to determine the elemental uncertainties of the tracer gas concentration measurements, the supply air and return air temperatures, and the supply airflow rate. Tables 1 and 2 list the experimental variables, their typical values, and the corresponding overall uncertainties. The overall uncertainty of a test variable x (such as temperature or gas concentration) at 95% confidence level was determined as
[U.sub.x] = [[[B.sub.x.sup.2] + [P.sub.x.sup.2]].sup.[1/2]] (6)
where [B.sub.x]. and [P.sub.x], respectively, represent the systematic and random uncertainties of the variable. The systematic uncertainty accounts for all elemental uncertainties, including those resulting from the instrument response time and the transient effects.
Table 1. Experimental Variables and Uncertainties Variable Symbol Range Gas [C.sub.SFB] 0-20 ppm concentration Flow rate Q 25-75 Std. L/s (50-150 scfm) Pressure [[DELTA] . 0-25 Pa (0-0.1 in. w.g.) differential P.sub.Room/Amb Supply air [T.sub.SA] 20[degrees]C-50[degrees]C temperature (60[degrees]F-120[degrees]F) Return air [T.sub.RA] 20[degrees]C-24[degrees]C temperature (60[degrees]F-75[degrees]F) Test room [T.sub.Room](x,y,z) 20[degrees]C-50[degrees][C(b) temperature (60[degrees]F-120[degrees]F) Variable Uncertainty Gas See Table 2 concentration Flow rate [+ or -]3% Pressure [+ or -]3 Pa differential ([+ or -]0.02 in.w.g.) Supply air [+ or -]0.5[degrees]C (a) temperature ([+ or -]1[degrees]F) Return air [+ or -]0.5[degrees]C (a) temperature [+ or -]1[degrees]F) Test room [+ or -]0.1[degrees]C to temperature [+ or -]0.5[degrees]C ([+ or -]0.2[degrees]F to [+ or -]1[degrees]F) (a) Based on a single measurement at or near the middle of plenum. (b) High temperatures occur near the supply air register. Table 2. Uncertainties of Gas Concentration Measurements Elemental Systematic Uncertainties Time Related Gas Range of Instrument Transient Synchronization Concentration Variable (Mfg. Effects Variable Specs.) (a),(b) Supply air 10-20 [+ or -]2% Negligible Negligible ppm Return air 0-10 ppm [+ or -]2% Varies Negligible Room air 0-10 ppm [+ or -]2% Varies Negligible Gas Random Concentration Uncertainty Variable Supply air [+ or -]3% Return air [+ or -]3% Room air [+ or -]3% (a) The uncertainty of the gas sample delivery time is transformed to gas concentration uncertainty based on the transient behavior at a given time. (b) The asymmetric uncertainty varies with the rate of gas concentration change--negative for increasing growth and positive for decay.
The uncertainty of an experimental result was determined by propagating the uncertainties of the individual variables through the data reduction equations (Coleman and Steele 1999). The asymmetric uncertainty of the normalized relative gas concentration ([xi] in Equation 5) was estimated from:
[U.sub.[xi].sup.( + )] = [[[([B.sub.[xi].sup.( + )]).sup.2] + [P.sub.[xi].sup.( + )]].sup.[1/2]] (7a)
[U.sub.[xi].sup.( - )] = [[[([B.sub.[xi].sup.( - )]).sup.2] + [P.sub.[xi].sup.2]].sup.[1/2]] (7b)
The asymmetric bias terms in Equations 7a and 7b were determined from
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (8)
where [beta] is the correlated systematic uncertainty. The random uncertainty, [P.sub.[xi]], was estimated as
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII] (9)
RESULTS AND DISCUSSIONS
The test results for two levels of supply airflow rates, low and high, are presented and discussed here. In the nonisothermal tests, the heating process commenced following a precooling period (to simulate recovery from setback) and ended as soon as the designated thermostat node (Figure 2) registered a temperature of about 4.5[degrees]C (8[degrees]F) higher than the room's initial temperature. For these tests, the supply air temperature was set between 38[degrees]C and 44[degrees]C (100[degrees]F and 110[degrees]F). In addition to evaluating the spatial temperature variations within the test room (Figure 2), tracer gas analyses were also conducted in these tests. One of the tests presented here was designed to conduct only a tracer gas analysis under an isothermal condition without any heating. In this test, unheated supply air was introduced to the test room to investigate the diffusion of the tracer gas in the absence of thermal stratification.
As discussed before, only two channels of the available multi-port gas analyzer were activated, one for the return air and the other for a test point within the room, to allow more frequent sampling. This was imperative for accurately capturing the transient behavior of the return-air gas concentration, especially in the high-flow test in which the heating process was considerably more rapid. However, as will be seen later, the gas concentration measurements at these two points did not provide conclusive results. To overcome this limitation, the concentration measurements at the return air were compared with those of a theoretical case where the indoor air was perfectly mixed (Equation 2). In the analyses, a nondimensional parameter was also utilized to negate the effect of variation of the supply-air gas concentration setpoint from one test to another. These measures proved to be a valuable tool in enhancing the usefulness of the limited tracer-gas measurements.
The results of these tests are illustrated in Figures 3 through 9. The low-flow tests were designed to evaluate the air distribution in high-performance buildings or in residential spaces, such as interior zones where the design airflow rate is relatively small.
Figure 3 depicts the time-dependent supply air and return air temperatures for a low-flow test in which the actual supply airflow rate was maintained at 41 [+ or -] 0.2 L/s or (87 [+ or -] 2.6 acfm)--corresponding to the standard-condition flow rate of 31 [+ or -] 1 L/s (65 [+ or -] 2 scfm), given the operating temperature and the local atmospheric pressure of 82 kPa (11.9 psia). The average air velocity at the supply air register was about 3 m/s (10 ft/s). A snapshot of the horizontal and vertical variations of the indoor air temperature at the end of the heating process is captured by Figure 4. This figure indicates that (1) in the occupancy zone below plane C, the horizontal temperature variation is insignificant and (2), excluding the vicinity of the supply air outlet (east wall), the highest vertical temperature difference is observed near the west wall facing the supply air register. Based on the correlation provided by ASHRAE (2005b), the vertical temperature difference between head (1.1 m above the finished floor) and ankles (100 mm AFF) observed near the west wall translates into dissatisfaction of up to 15% of the seated occupants. In other areas, including the center of the room, the vertical temperature difference is less than 3[degrees]C (5[degrees]F), resulting in a dissatisfaction level of no more than 6%. (It should be pointed out that the temperatures at the head and ankles levels were not directly measured but estimated based on the measurements at the adjacent test planes.)
[FIGURE 3 OMITTED]
[FIGURE 4 OMITTED]
Figure 5 presents the supply-airflow rate and the gas concentration measurements of the indoor air at the center on plane C (168 cm AFF) and at the return air outlet. The supplyair. gas concentration is maintained at about 13.7[+ or -]0.41 ppm throughout the test. The difference between the readings at these two locations is well pronounced and persists throughout the entire heating process--reflecting a poor mixing effectiveness from the ventilation standpoint. This notion is also supported by Figure 6, which compares the measured returnair gas concentrations and the resulting normalized relative quantities [xi] (Equation 5) with those of the scenario in which the air is theoretically well mixed. In Figure 6, the time is normalized ([THETA] = t/[tau]) with respect to the time constant ([tau]), which is 34.5 minutes for this low-flow test.
[FIGURE 5 OMITTED]
[FIGURE 6 OMITTED]
In contrast, the isothermal test conducted with a low volumetric flow rate of 39 [+ or -] 1.2 L/s (83[+ or -]2.5 cfm) points to the presence of only a small spatial variation in the gas concentration in the test room (Figure 7). The good agreement found between the experimental results and those of the ideal case (Figures 8 and 9) further supports this observation. Therefore, the comparison of these two low-flow and high-flow tests reflects the effect of thermal stratification on limiting the ventilation effectiveness, as implied by the impeded diffusion of the tracer gas in the presence of stratification. Note that in Figure 9
[FIGURE 7 OMITTED]
[FIGURE 8 OMITTED]
[FIGURE 9 OMITTED]
[psi] = [[integral].sub.0.sup.t][xi]dt,
and the age of the air for the ideal case approaches the time constant as the time increases, yielding a ventilation effectiveness of unity (i.e., [epsilon] =1). The limited results of [xi] based on the measurements closely track those of the well-mixed air scenario.
The results of a high-flow test are presented in Figures 10, 11, and 12. The actual supply airflow rate was maintained at 87[+ or -]2.6 L/s or (184 [+ or -] 5.5 cfm), corresponding to the standardcondition flow rate of 65 [+ or -]2 L/s (138 [+ or -] 4 scfm), given the operating temperature and the local atmospheric pressure. The average air velocity at the supply air register was estimated to be about 6.3 m/s (20.7 ft/s). The flow rate of this test is more in line with that of the constant-air-volume HVAC systems in standardpractice buildings. Comparison of Figure 10 (high-flow test) with Figure 3 (low-flow test) indicates that at the higher flow rate, the return air temperature is higher at the end of the heating process when the thermostat setpoint is satisfied. Despite the higher supply air temperature, the vertical temperature variations in this test (Figure 11) are less significant than those in the low-flow test (Figure 4). However, because of the higher temperatures in the occupancy zone near the west wall, the horizontal temperature variation is more pronounced compared to that of the low-flow test. With respect to comfort, the maximum vertical temperature difference in the seating zone was reduced to about 3[degrees]C (5[degrees]F), limiting the maximum level of dissatisfaction (stemming from this comfort variable alone) to less than 6%.
[FIGURE 10 OMITTED]
[FIGURE 11 OMITTED]
Interestingly, the gas concentration trends for the return air and the test point on the centerline (at 168 cm above the finished floor) virtually collapse as seen in Figure 10. This by no means is indicative of perfectly mixed indoor air, according to Figure 12, which compares the experimental results with those of the reference case. However, it is evident that, at the higher flow rate, the actual return air concentration and its normalized relative quantity are in betters agreement with those of the ideal case.
[FIGURE 12 OMITTED]
Because the test room boundaries were well insulated, the effect of the envelope losses on the stratification levels observed in these tests was rather limited. The average heat loss from the test room was estimated to be about 40 to 60 W (150- 200 Btu/h) at the end of the heating process, translating into less than 6% of the supply air heating capacity for the high-flow test and less than 12% for the low-flow test. (The heat loss was approximated based on the room temperature drop during the no-flow period. This estimate does not include the effects of infiltration and the internal thermal mass associated with the walls, ceiling, and floor during the heating process.) In situations where significant envelope heat losses occur, the stratification can be influenced by the boundary conditions (not addressed in this study).
The temperature gradient observed near the west wall (facing the supply air diffuser) is expected to diminish for cases in which (1) supply air diffusers with deflected blades are used and/or (2) the supply airflow is in the longitudinal direction of the room, providing a greater distance to accommodate the throw.
In this study, the performance of an indoor air-distribution system incorporating a high sidewall diffuser (with no deflection) was experimentally examined during a period of recovery from a setback, in the absence of any internal heat sources. The study encompassed assessment of the spatial temperature variations within the test room as well as evaluation of the ventilation effectiveness of the system under various operating conditions. To evaluate the air mixing and ventilation effectiveness, tracer-gas tests were conducted in parallel with the temperature measurements. The key findings of this study are summarized as follows.
At the high supply airflow rate comparable to those of standard-building HVAC systems, the stratification in the occupancy zone was rather small and had fairly minor adverse effect on the comfort level, despite the relatively high supply air temperature and the diffuser with no deflection. This finding, which was also supported by a tracer gas analysis, is consistent with the commonly made HVAC design assumption that the indoor air is fairly well mixed. From the comfort standpoint, the level of dissatisfaction associated with the vertical temperature difference in the seating zone was estimated to be less than 6% at this flow rate.
At the lower supply airflow rate and velocity, a greater degree of stratification was observed, translating into dissatisfaction of 6% to 15% of seated people. The temperature measurements did not suggest formation of a drastically different temperature distribution pattern at the lower flow rate. However, comparison of the tracer-gas results with the exact solutions for theoretically well-mixed indoor air, pointed to a lesser mixing effectiveness. This incongruity stemmed, in part, from the limitation of the temperature measurements in depicting the complete temperature distribution pattern in the entire space.
The approach adopted for the tracer gas tests and analysis of the ensuing results was demonstrated to be an effective tool for evaluating the mixing effectiveness during the recovery from setback.
The author thanks Ed Hancock for his efforts in instrumentation and data collection. The work of Doug Powell in installation of the experimental facility is greatly appreciated as well. Finally, this study would not have been possible without the support and guidance from Ren Anderson.
B = systematic uncertainty
[C.sub.SF,Room] = tracer-gas concentration in the test room, ppm
P = random uncertainty
= volumetric flow rate, L/s (cfm)
[V.sub.Room] = volume of the test room, L ([ft.sup.3])
t = time, minute (at t = 0, data collection/reporting begins)
[t.sub.age] = age of air, minute (Equation 5)
[beta] = correlated systematic uncertainty
[epsilon] = air change effectiveness (Equation 3), dimensionless
[xi] = integrant of Equation 3, labeled as normalized relative concentration
[tau] = time constant, minute (Equation 4)
[theta] = dimensionless parameter Equal to [t.sup.*] / [tau]
[MATHEMETICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]
ASHRAE. 2005a. 2005 ASHRAE Handbook--Fundamentals, Chapter 27. Atlanta: American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.
ASHRAE. 2005b. 2005 ASHRAE Handbook--Fundamentals, Chapter 8. Atlanta: American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.
Coleman, H.W., and W.G. Steele. 1999. Experimentation and Uncertainty Analysis for Engineers, 2d ed. New York: John Wiley & Sons.
Jalalzadeh-Azar, A. 2006. Experimental evaluation of indoor air distribution in high-performance residential buildings-- Part I: General descriptions and qualification tests. NREL Report No. TP-550-40392. Golden, CO: National Renewable Energy Laboratory.
Jalalzadeh-Azar, A.A., R. Anderson, and K. Gawlik. 2006. Role of indoor air distribution in performance of heat pump systems in high-performance residential buildings. IMECE2006-15079. 2006 ASME International Mechanical Engineering Congress and Exposition, November 5-10, Chicago, IL.
Ali A. Jalalzadeh-Azar, PhD, PE Member ASHRAE
Ali A. Jalalzadeh-Azar is a senior engineer at National Renewable Energy Laboratory, Golden, CO.
|Printer friendly Cite/link Email Feedback|
|Author:||Jalalzadeh-Azar, Ali A.|
|Article Type:||Technical report|
|Date:||Jul 1, 2007|
|Previous Article:||Examination of thermal adaptive effect of postural and positional adjustment of a seated human body exposed to spot airflow.|
|Next Article:||Field testing of a supervisory optimal controller for an air-conditioning system.|