UNITED STATES DEPARTMENT OF COMMERCE Maurice H. Stans, Secretary NATIONAL BUREAU OF STANDARDS Lewis M. Branscomb, Director Performance of Louvered Devices as Air Mixers T. K. Faison, Jr., J. C. Davis, and P. R. Achenbach As part of a study of evaluating methods for reducing thermal gradients within the cross section of an air stream, three louvered mixing devices were investigated. Each of these devices was found to be capable of reducing the cross-sectional nonuniformity of air temperature to a few percent of the entering value. The three devices covered in this report contain combinations of louvers (directing vanes) and baffles as mixing elements. Two of the devices were designed at the National Bureau of Standards; the third was a modification of a previous design. The three mixers (the louvered strip, the concentric louvers, and the louvered-baffle) required 4.75, 3.8, and 3.0 duct diameters, respectively, to reach a mixing effectiveness level of 97 percent. The mixing effectiveness of the louvered strip and concentric louver models was independent of the approach velocity, whereas the effectiveness of the louver-baffle model was somewhat dependent on the approach velocity. The pressure drops accompanying air flow through the mixers, expressed as multiples of the velocity head of the entering air, were approximately 7, 5, and 38 for the louvered strip, concentric louver, and louver-baffle mixers, respectively. Key words: Effectiveness; forced mixing; mixing device; pressure drop; temperature; uniformity. 1. Introduction The lack of uniformity of temperature within the cross section of air streams has long posed a problem to investigators trying to make accurate measurements of air-stream temperature. Several mixing devices have been developed and many rather complicated designs have been devised to promote uniformity of temperature, although the literature contains little information of a quantitative nature on their effectiveness. Each of the methods seems to have its own particular drawback-too complicated to use, the pressure drop too high, the length for mixing too long, etc. A recent study [1] at the National Bureau of Standards showed that thermal gradients within the cross section of an air stream can be effectively diminished through forced mixing. With the removal of the undesirable condition of temperature nonuniformity, the average temperature of the air stream can be more ac curately obtained and air stream temperaturedependent properties, such as enthalpy, can be determined with greater precision. Several types of air mixing devices were evaluated in the laboratory but evalutations of only three types are presented here; louvered strip, louvered-baffle combination, and concentric louvers. The performance of each is given for various initial conditions of nonuniformity. The names given these mixers tend to describe the geometric configuration of each. The authors have previously described the apparatus in which the devices were evaluated and discussed the performance of the orifice as a mixer for removing thermal gradients in an air stream [1, 2]. The three types presented here and the orifices previously reported [1] were studied in detail. Several other designs, which were not as effective, were evaluated only qualitatively. 2. Apparatus and Test Procedure The test appartus [2] was designed to produce in a 24-in-diam duct a condition (nonuniformity of temperature in the air stream) that might be commonly encountered in heating and air conditioning systems and laboratory applications. A schematic of the apparatus is shown in figure 1. Figure 2 is a photograph of the apparatus. A detailed description of the apparatus, the instrumentation, and procedure is given in reference [2]. Briefly, the apparatus functioned as follows: Air preconditioned and maintained at a constant temperature was drawn into the inlet blower at a controlled rate. Downstream from the blower, just 1 Figures in brackets indicate the literature references at the end of this paper. FIGURE 1. A sketch of the mixing apparatus used to produce the test condition and house the mixing devices to be evaluated. after an enlarging transition, the duct cross-sectional area was divided into four equal 1-ft squares (sect. A-A, fig. 1). The dividing partitions were made to extend from beginning of the 24-in square conditioning portion of the apparatus on through a portion of the round duct and terminated at a position a short distance upstream from the section shown as B-B in figure 1. The partitions in the circular duct divided the duct space or air stream into four equal pie-shaped quadrants. The portions of the apparatus immediately upstream and downstream of section A-A in figure 1 contained electrical resistance heaters in each of the spaces to produce and maintain selected conditions of temperature. Immediately following the conditioning section of the apparatus, a measuring station was positioned for determining the temperature pattern of the air prior to its entering the mixing device to be tested. From the first measuring station, the air passed through the mixing device and along the duct to a second measuring station. The second station could be repositioned along the length of the duct to determine the uniformity of the air stream at different locations downstream from the mixing device. The air was then passed from the apparatus through the exhaust blower. The two blowers, the supply and exhaust, were dampered to control both the flow rate and the static pressure within the apparatus. Temperature patterns shown in figure 3 were used to evaluate the performance of the mixing devices. These patterns indicate distinct differences in temperature within the cross section of air streams prior to entering a device under test. The cases shown in the center and left part of the figure might represent situations where two streams of fluid merge together and form patterns of nonuniformity. The case shown at the right in the figure illustrates a situation where air flows for a distance in a duct located in an ambient temperature that is higher or lower than the initially uniform temperature in the duct, thus causing the temperature near the wall of the duct to be different from that at the center of the stream. These patterns are fairly common in practice and should give meaningful information on the ability of a mixing device to promote mixing and thus produce a homogeneous stream. FIGURE 3. Patterns of temperature distribution within the cross section of the test stream. 2.1. Instrumentation Temperature measurements were made at each of two stations. Measuring station No. 1 (see fig. 1) was located just downstream from the termination of the quadrant partitions, where the nonuniform temperature pattern of the air was measured as it existed prior to passing through the mixing device. The second measuring station was used to make measurements of the temperature distribution at selected points along the duct length as the mixing progressed. Both stations were constructed having the same physical pattern as shown in figure 1, with the only difference being that the upstream station was stationary and the downstream station was movable. At each station an array consisting of 24 copper-constantan thermocouples constructed of 30 gage (A.W.G.) wire was used. Thermocouples were calibrated at the National Bureau of Standards and were manually read on a precision-type potentiometer capable of direct reading to 1.0 V (0.05 deg F for the wire used) with interpolation to 0.1 μV. A zone box [3] was used in the thermocouple circuit to permit the use of a common ice reference junction and all switching was accomplished in the copper portion of the thermocouple circuit. From time to time comparison tests to determine the magnitude of variations amongst the 24 thermocouples at each station were made by immersing all of the thermocouples in a Dewar containing room-temperature water. The difference between any two of the 24 thermocouples was always less than 0.7 μV (0.03 deg F). 2.2. Conditioning and Control Air flow rate in the test apparatus was controlled by varying adjustable dampers at the inlet of the first blower and at the exhaust downstream from the second blower. By proper adjustment of these two dampers, the pressure inside the apparatus was maintained at a positive level to prevent leakage of unconditioned air into the system from the room. Flow rates through the four quadrants were approximately equal. The volume rate of air flow was determined by using pitot-static tubes to measure the velocity head at 18 positions in the cross-sectional area of the rectangular inlet duct. The duct had a straight length of 8 ft ahead of the station of measurement. The pressures from the pitot-static tubes were measured with a Hook gage, which could be read to the nearest 0.001 in of vertical water column. Using potentiometer-type thermostat controllers and manually operated base heaters, air temperature was maintained constant at preselected temperature levels in the apparatus. The major increase of the air temperature was obtained from base heaters. Final control was obtained by thermostatic operation of trimmer heaters of smaller capacity. Each of the four quadrants individually controlled to provide flexibility in selection of temperature patterns. Electrical resistance heaters positioned in a circular pattern around the periphery of the duct near the duct wall were used to produce concentric temperature patterns. was 3. Description of Mixing Devices In this study, three types of mixers were investigated: louvered strip, a louver-baffle combination, and concentric louvers. These mixers were designed with the objective of obtaining a satisfactory combination of mixing effectiveness, pressure drop, and space requirements for their use. The design and construction of the three types of mixers are described in the following sections. 3.1. Louvered Strip In designing the louvered strip mixer, the area enclosed by a circular band 24 in in diameter was divided into 6 horizontal strips, each 4 in high, as shown in figure 4, left view. The areas bounded by the strips were fitted with vertical louvers for deflecting the air from its normal path. Louvers were designed so that |