Using the temperature pattern shown in the center of figure 19, tests were made to determine the effect of increasing the magnitude of the temperature difference between portions of the cross section of the air stream over the range from 4 to 13 deg F. Results shown in figure 20 indicate that there was no change in performance within this range. As in the case of the other mixers, which have two elements in series, it was necessary to determine at what distance, with respect to each other, the mixing elements would best perform. For this determination, the downstream measuring station was held in a fixed position and the location of the second mixing element with respect to the first was varied. The results illustrated in figure 21 show that effective mixing was gained when the elements were placed approximately two duct diameters apart, but that very little mixing would be lost by reducing the spacing down to one duct diameter. Tests using a spacing of two duct diameters between mixing elements showed that the overall distance from the first mixing element to the point of maximum observed uniformity was 4.75 duct diameters. Figure 22 shows how the effectiveness varied as the overall distance was changed. TEMPERATURE DIFFERENCE BETWEEN PORTIONS OF THE AIR STREAM BEFORE MIXING ( DEGREES F) FIGURE 20. Effect of air temperature difference at mixer inlet upon the effectiveness of the concentric louver mixing device. The effect of velocity upon the mixing process was determined by varying the average speed of the air stream over the range from 150 fpm to 600 fpm. The results plotted in figure 23 show that the effectiveness of the concentric louver mixer is relatively unaffected by the speed of the air stream. 4.3.1. Applications of the Concentric Louver Mixer The concentric louver mixer has been used successfully in several test apparatuses at the National Bureau of Standards in which uniformity of temperature of an air stream greatly facilitated the purpose of the investigation. In one case it was used to mix the air stream leaving a coil being tested in a psychrometric calorimeter apparatus for measurement of the heat transfer characteristics of cooling coils. After determining that the uniformity of temperature downstream from the mixer was good, only three thermocouples were used to obtain a representative sampling of the air stream for the psychrometric determinations. In the second case the mixer was used, in a study of the effect of thermal radiation upon temperature-measuring sensors, to obtain a condition of reasonably uniform air tempera The overall distance of 4.5 duct diameters from the inlet of the mixing device to the downstream FIGURE 21. Effectiveness of the concentric louver mixing device relative to the ture at all points within the cross section of the air stream prior to entering the heated portion of the duct. Figure 24 is a drawing of the heated wall apparatus showing the mixer, measuring stations, probe location, and heating elements. The study was made by locating the sensors in a duct whose wall was heated to a higher temperature than the air which surrounded the sensors. Heat was applied to the duct walls and, as the air passed downstream, an approximately symmetrical pattern of temperature from the duct wall to the sensors was achieved. At the point where the sensors were placed there existed a flow of air in which temperature fluctuations were very small. Error of measurement for temperature sensors of thermocouples, a thermistor and a resistance thermometer, due to the radiation from the heated walls, was then determined FIGURE 23. Relation of mixer inlet velocity to effectiveness for the concentric louver mixing device. The performance of three types of air temperature sensing devices encased in the probe was evaluated using this heated wall apparatus. The air mixer location can be seen at the left of the illustration. FIGURE 24. A sketch of an apparatus in which the concentric louver mixing device was used to produce a uniform temperature in an air stream. Due to the heat applied to the wall of the measuring section, the temperature of the air at measuring stations 2 and 3 was higher near the duct wall than at the center FIGURE 25. Temperature patterns at the three measuring stations in the heated wall apparatus, shown in figure 24. through an extensive investigation. Figure 25 shows temperature profiles at the three measuring stations when the duct wall at stations 2 and 3 was 50 deg F higher than the air stream at the center of the duct. Recently one of the mixers described in this paper, the concentric louver, was proposed as a recommended mixing device in the ASHRAE Standard 41-66 [5], Section on Temperature Measurement of the Standard Measurements Guide, Part I. The recommendation has been incorporated in the new issue of the standard. An appendix is included at the end of the paper which gives detailed drawings of the three mixing devices and discusses techniques helpful in construction of the mixers. 4.4. Comparison of Other Air Mixer Designs In the previous investigation on orifices and orifice-target combinations [1], it was determined that an orifice having a diamter ratio (ratio of the orifice throat diameter to the duct diameter) of approximately 0.33 was consistently more effective in reducing the thermal differences in an unmixed air stream than orifices having diameter ratios of 0.50 and 0.67. However, the mixing process was accompanied by a high pressure drop across the mixer. If the pressure drop can be tolerated, temperature nonuniformity could be reduced to 3 or 4 percent of its original value, at a distance of approximately 4.5 duct diameters downstream from the orifice. The study indicated that diameter ratio was the most important parameter affecting the mixer effectiveness, but that the interface area between cold and warm elements of the air stream and size of the nonisothermal elements also had a bearing on the mixing process. The performance of the orifice-target combinations showed no improvement over the plain orifice even when using a target as large as two-thirds of the duct diameter. During the course of the investigation of mixing devices, a number of designs (including the three reported herein) were evaluated using the flow visualization technique for observing the flow pattern, with smoke used as the indicator. Although observation of flow patterns is qualitative, good correlation between flow visualization and quantitative method of evaluation using temperature patterns was gained. This correlation indicates that flow visualization is a rather good but coarse method of evaluation. Listed below are a number of designs which revealed relatively poor mixing characteristics. 1. Screens. Layers of flat wire screen, having approximately 50 percent free area, placed in series over the area of the duct, through which the air could flow, showed very little effect upon the air stream with respect to mixing. (Velocity profiles taken before and after a number of layers of screen showed a distinct improvement in the uniformity of the velocity within the cross-sectional plane of the duct.) It should be mentioned that for a temperature distribution having many small variations, the screen might be more effective, as the flow through the screen is in the form of many small jets. For the distribution as described in this report, one quadrant different from the other three, for example, screens provide little displacement of flow from one part of the stream to another; thus, the general effectiveness for screens was very limited. 2. Baffles. a. Semicircular baffles were placed in series approximately one duct diameter apart in such a way that one-half of the area of the round duct was blocked off by the first baffle. One duct diameter downstream, the second baffle was placed in a position corresponding to the space not occupied by the first baffle; thus, the air would pass through the restricted opening of the first baffle and strike the second baffle before passing through the second opening. This method proved unsatisfactory (but better than screens). It was evident that the patterns were not being fully broken up. b. Single slats one-third of a duct diameter in width were placed along the diameter of the duct and used as baffles. The baffles were installed in series, the first horizontal and the second vertical, approximately one duct diameter apart. The mixing was comparable to that of the semicircular baffles but was not considered good. c. Quartered baffles were used and showed some improvement over the above two types. These baffles consisted of openings in the first and third quadrants through which the air could pass, while the second and fourth quadrants were blocked off with metal plates. Approximately one duct diameter downstream, another baffle was fixed with openings in the second and fourth quadrants and the first and third blocked off. Still the mixing was not considered good, but an improvement. 3. Fans. a. A multi-blade fan was placed in the air stream and allowed to rotate as the air passed through it. This method did not produce mixing, as the air was not displaced but traveled essentially along its original path. b. The same multi-blade fan was held in a stationary position and the air was forced through the openings between blades and again little mixing occurred. The resulting action was that of swirling as the air moved downstream. The pattern remained essentially the same; there was simply an angular displacement as the air passed downstream. c. Powered fans were not tested because of the difficulty of mounting a moving mechanism in the duct and because of the introduction of unwanted heat. Table 4 is a simplified presentation of comparative performance of the three mixers described in this paper plus the orifice as described in a previous publication [1]. From these studies on mixing devices it was shown that a poor distribution of temperature could be effectively improved by either the louvered strip, the louver-baffle, the concentric louver, or the 0.33-diameter ratio orifice. Each possesses good features and each has shortcomings. Consequently tradeoffs must be considered when selecting a mixer for a particular application. The concentric louver, the louverbaffle, and the 0.33-diameter ratio orifice were relatively unaffected by temperature pattern, whereas the louvered strips showed some sensitivity to temperature pattern. As shown in table 4, the resistance to flow for the louvered strip and the concentric louver when described in multiples of velocity head was 7 and 5, respectively, as compared to 38, 43, and 166 for the louver-baffle, 0.50-, and 0.33-diameter ratio orifices. The overall distance required to attain an effectiveness level of 95 percent ranged from 2.3 to 4.2 duct diameters. A 97 percent effectiveness was obtained by all of the mixers with the exception of the 0.50-diameter ratio orifice. Through reference to table 4, it can be seen that the change in effectiveness from 95 to 97 percent was accomplished in 0.7 to 0.8 duct diameter for the louver-baffle, the concentric louver, and the 0.33-diameter ratio orifice. The louvered strip mixer was less responsive to the change and required a distance of 1.25 duct diameters for the 2.0 percent increase in effectiveness. Overall, the louver-baffle and the concentric louver mixers were comparable in their mixing effectiveness, with the tradeoffs being pressure drop and distance required for mixing. The mixing distance of the louver-baffle is approximately 0.8 of that required for the concentric louver, but the pressure drop is some seven times as high for the louver-baffle when compared to the concentric louver. For all of the mixing devices, the magnitude of the temperature differences in the cross section of the air stream does not seem to affect the mixing effectiveness a great deal. As a word of caution, careful consideration should be given in determining the average temperature of the mixed air stream when large temperature differences are encountered prior to mixing. With large initial temperature differences, even with a high effectiveness, the temperature differences in the mixed portion of the air stream could be significant in some applications. 6. References [1] Faison, T. K., Jr., Davis, J. C., and Achenbach, P. R., Performance of square-edged orifices and orifice-target combinations air mixers, Nat. Bur. Stand. (U.S.), Bldg. Sci. Ser. 12 (Nov. 1967). [2] Faison, T. K., Jr., Davis, J. C., and Achenbach, P. R., A test apparatus for the study of forced air-mixing devices, J. Res. Nat. Bur. Stand. (U.S.), 70C (Eng. and Instr.), No. 1 (Jan.-Mar. 1966). |