The National Bureau of Standards has undertaken a study to determine methods or processes by which nonuniformity of temperature in an air stream in a duct can be corrected by forced mixing. Since the enthalpy of an air water-vapor mixture is related to its temperature and moisture content, the flow of energy represented by the moving air stream can best be determined if the measured values of dry-bulb temperature and dewpoint or wet-bulb temperature are truly representative of the entire air stream. Under conditions where nonuniformity of velocity, temperature, and humidity exist, realistic averages of temperature and humidity are very difficult to obtain. During laboratory tests where moving air is involved, the measurement difficulties due to poor mixing can sometimes be compensated for by taking many temperature measurements throughout the cross section of the air stream. But even when a large number of measurements are taken, the observed values must be evaluated and weighted, with respect to both temperature and velocity. For nonuniform distributions of temperature and velocity, a method of weighting should be used for proper representation of the average temperature of the total mass of air that is passing a measuring station or stations.

A much better method of determining the aver

age air temperature is to produce a high degree of uniformity of the air temperature through use of a mixing device. Determination of the average temperature could then be accomplished by making only a few measurements. If the mixing is good enough, one observation will suffice.

A number of mixing devices are under investigation in this study. This paper, the second in a series of three on the apparatus and mixing devices, presents performance data for the square-edged orifice, and the combination of orifice and target (circular baffle), two methods of mixing which have been in use in laboratories for a number of years.

These methods mix the air principally by a jet action. There has been considerable work done on jet mixing, but most other investigators [1, 2, 3, 4] have studied the mixing of the jet with the surrounding air under unbounded conditions whereas this study is concerned with mixing in a duct. Theoretical and experimental work has been done with both heated and unheated jets. In previous investigations, the jets have been axially symmetrical, whereas the present work deals with an unbalanced distribution of temperature within the air stream.

1 Figures in brackets indicate the literature references at the end of this paper.

1

2.1. Temperature Measurements A test apparatus [5] was designed and built for producing controlled temperature and humidity conditions in an air stream to facilitate measurement of the effectiveness of air mixing devices. A schematic layout of the apparatus is shown in figure 1. A stream of air, maintained at a constant and substantially uniform temperature, was drawn into the apparatus with an inlet blower. The air stream was conditioned to a desired pattern of nonuniformity of temperature in that portion of the apparatus shown in figure 1 as Section A-A. The air stream was not conditioned with respect to humidity for the tests covered in this paper. Temperature measurements were made in a 24-in diam duct at stations upstream and downstream from a mixing device. These measurements provided information about the temperature pattern upstream from the mixer and the resulting distribution downstream. Conditions of nonuniformity at the upstream measuring station were controlled and reproduced from test to test. A mixing device, in this case an orifice, was placed just downstream from the location shown in figure 1 as Section B-B and the air stream of nonuniform temperature was forced through the mixer.

At selected distances downstream from the mixer, measurements were made to determine the temperature distribution as an indicator of the effectiveness of the mixing device in producing

uniformity of temperature. By observing the temperature distribution in the duct at each of several stations downstream from the mixer, the progress of the mixing process could be evaluated and the station of optimum mixing could be determined.

A statistical analysis was made of the upstream and downstream temperature values to determine the standard deviations at the respective stations. For each of five sets of temperature determinations recorded in a test, standard deviation values were computed from which mixing effectiveness was calculated. The five effectiveness values were averaged to obtain the effectiveness for a particular test. A value of effectiveness for a mixing device was calculated by subtracting from unity the ratio of downstream to upstream standard deviation. The functions, standard deviation and effectiveness, are shown below:

3.D. - - -
-(-3)+]"

This latter method based on the range is simpler to use but the results can be much more affected by variation of a single observation.

Tests were performed under varied physical arrangements such as different orifice sizes and different longitudinal positions of the plane of measurement downstream of the orifice. Tests were also performed under the following varied stream conditions: different temperature patterns, flow rates, and magnitudes of temperature difference. Steady state conditions were maintained throughout the system, a necessary condition for the above formulas to give a valid indication of effectiveness.

The conditioning section of the 24-in square duct upstream of the mixing device was divided into four quadrants for producing distinct and reproducible temperature distributions. The temperature patterns were obtained by means of different electrical energy inputs into the air streams in the quadrants. Most of the tests were performed with a temperature difference between warm and cool quadrants of approximately 3 deg F, with three of the quadrants kept at the higher temperature and the fourth at the lower temperature.

The temperature distribution in the cross-sectional area of the air stream at each station was measured with 24 calibrated copper-constantan thermocouples, fixed in similar patterns at each station. Each thermocouple was soldered to a small washer which, because of its mass, provided damping of small temperature fluctuations. Location of the thermocouples at the upstream measuring station is shown in figure 2. The tubes which served to aline the air stream as it issued from the quadrants toward the upstream measuring station can be seen in the background of figure 2. The enlargement shows a typical assembly of the thermocouple and its holder.

The measuring station upstream from the mixing device was always located at the same place in the test duct and was used to determine the temperature distribution of the air as it left the four quadrants of the duct. The downstream measuring station was movable from test to test; thus, temperature distributions within the cross section of the air stream could be taken at different distances downstream from the mixing device.

FIGURE 2. View of the upstream measuring station and detail of the thermocouples and holders.

Periodically the thermocouples at each measuring station were checked to determine the magnitude of deviation between any two thermocouples at a station. For this purpose the thermocouples were immersed in a container of water under isothermal conditions and the greatest deviation between any 2 of the 24 thermocouples at either station was determined. These differences never exceeded 0.7 μV (equivalent to about 0.03 °F) and are considered adequate for this work. The indicating instrument was a precision manual potentiometer, providing resolution for comparison or difference readings to the nearest 0.1 μV.

2.2. Static Pressure Measurements Static pressure measurements were taken at 4in intervals along the length of the duct to determine the variation in pressure from the upstream measuring station to the most remote location of the downstream measuring station. Figure 3 shows a characteristic static pressure profile for a squareedged orifice which illustrates the static pressure regain and the location that has proved to be the approximate location of optimum mixing.

It was found in this study that no appreciable increase in mixing effectiveness was observed beyond the station where maximum regain occurred. Static pressure regain for the orifice has been studied and described in other publications [6] and the data gathered in this study were in good agreement with previous work.