1. Introduction

Although the psychrometer is one of the oldest and most common instruments used to measure the humidity of air, no theory adequately predicts its performance. Empirical and semiempirical formulas exist which describe, under limited conditions, the performance of psychrometers of particular dimensions and configurations.

In 1967, we developed and constructed a laboratory model of an adiabatic saturation psychrometer [1] the performance of which is specified by means of an equation. This instrument differed from other psychrometers in that it was designed to utilize a steady-flow adiabatic isobaric saturation process, whereas other psychrometers, even under steady-state conditions, are an open system undergoing a nonequilibrium process which cannot be described completely by classical thermodynamics. It was tested with various fluids and gases under conditions of zero vapor content. Because the results agreed with an equation derived from classical thermodynamics to within the limits of the experimental uncertainties associated with the conducted tests, it was concluded that the equation did indeed predict the behavior of this adiabatic saturation psychrome

ter.

This was particularly significant in the tests with vapor-gas systems other than water-air where other psychrometers give results which differ markedly from those derived from the postulates of classical thermodynamics.

The adiabatic saturation psychrometer has been developed further into a portable and self-contained instrument, intended for both laboratory and field use. In order to permit its employment at low ambient dry-bulb tempera

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

tures of meteorological interest without freezing of the wet-bulb water supply and wicking, provision was made for heating the test air to a fixed elevated temperature. It appeared to us that the heated-air adiabatic saturation psychrometer could be used to investigate more fully the formed in accordance with the derived equation. It had validity of our earlier conclusion that the instrument permade only under conditions of zero vapor content, the been suggested that because the original tests had been equation had been validated only under this unique condition and that the use of the relationship at other humidities could not be accepted with complete certainty.

We believed that the condition of zero vapor content was a unique condition only in that it represented the most severe condition under which to test the behavior of a psychrometer, and that the instrument would behave in accordance with the derived equation at all vapor con

The availability in our laboratory of a highly accurate humidity generator [2] made it feasible for us to perform an extensive series of tests over a wide range of

humidities.

The general design and operational features, as well as the test results, of the heated-air adiabatic saturation. psychrometer are the subjects of this paper.

2. Theory

When a quantity of liquid or solid water at pressure P and temperature Tw is evaporated into a vapor-gas mixture at pressure P, temperature T and mixing ratio r to bring the gas adiabatically to saturation at pressure P, temperature Tw and mixing ratio r, the sum of the enthalpies of the various phases are conserved. Thus the

by thermistor R, which in conjunction with a proportional heat controller (not shown) controls the voltage supplied to heater L, thereby regulating the temperature of the air entering the saturator tube to 414 °C. ± 4 °C.

Figure 3 is a circuit diagram of the two Wheatstone bridges, power supply and galvanometer circuit. The Wheatstone bridges and galvanometer measure the entrance (dry-bulb) and exit (wet-bulb) thermistor resistances. In addition, there are two temperature control circuits (not shown) operating from mechanical thermostats. One of the temperature control circuits regulates the air temperature surrounding the Wheatstone bridge and the other prevents the section of the instrument which contains water from falling to freezing temperature.

FIGURE 2. Schematic diagram of auxiliary components. B. Glass Dewar flask; G. brass cylinder; H. trap drain; J. critical nozzle flow controller; K. syringe pump; L. nickel-chrome heater coil; N. flow exit; 0. flow entrance; P. vacuum pump; Q. differential pressure gage; R. control thermistor; S. liquid filler tube; T. liquid trap; U. thermistor leads; V. cylinder cap; W. "O" ring; Y. compression seal; Z. "O" ring; AA. polystyrene foam packing; BB. rubber stopper; CC. insulated polytetrafluoroethylene tubing; DD, saturator tube connector; EE. cap seal assembly; FF. pressure signal tubing; HH. polytetrafluoroethylene liquid flow tubing.

V

Figure 2 is a schematic representation of all hydraulic and pneumatic components not shown in figure 1. A brass cylinder G surrounds the Dewar flask B shown in I figure 1. Flask B is held rigidly in place on the axis of G by means of polystyrene foam packing, AA. Cap V is sealed to saturator tube A by means of "O" ring Z within assembly EE, and to cylinder G by means of "O" ring W. In addition to providing the means for sealing, cap supports and locates saturator tube A coaxially within Dewar flask B. Insulated polytetrafluoroethylene tubing CC connects to the saturator tube. The thermistor leads, pressure tap, plastic water-feed tube and gas flow exit tube pass through compression seals in cap V. Test gas enters the instrument at O, is heated by the nickel-chrome heater coil L, flows through tubing CC and enters the saturator tube. After passing over the moist wicking it leaves the psychrometric section at Y and passes through trap T, where entrained water is separated from the gas stream. The trap is drained through tube H. The gas then passes through the critical nozzle flow controller J and on through vacuum pump P, and exits at N. Differential pressure gage Q measures the pressure difference between flask B and atmospheric pressure. Syringe pump K forces liquid into the instrument at a constant rate of flow. The syringe is refilled through tube S. The temperature of the air entering the saturator tube is sensed

2.5k2

FIGURE 3. Temperature measurement circuit.

D. Dry-bulb temperature bridge; W. wet-bulb temperature bridge; G. galvano. meter; RD. dry-bulb heliopot; Rw. wet-bulb heliopot; Setp. dry-bulb step switch; Setw. wet-bulb step switch; Sp. polarity reversing switch; Sg. wet-bulb or dry-bulb selector switch; S1. low sensitivity galvometer switch; S2. medium sensitivity galvanometer switch; S3. high sensitivity ga.vanometer switch.

Figure 4 is a top view photograph of the instrument assembled and ready for operation. The instrument is 23-in long, 16-in wide and 13-in high, and weighs 772 pounds.

Figure 5 is a photograph of the section where the water is drained and the syringe refilled, taken with the side panel removed and lying in front of the instrument.

The saturating element is a helix with a fiber glass surface. It was made by covering 7 ft of polytetrafluoroethylene tubing, having an i.d. of about 0.022 in and an o.d. from 0.042-0.050 in, with number 22-gage fiber glass spaghetti. The fiber glass spaghetti extends beyond the upstream end of the polytetrafluoroethylene tubing and is tied off with linen thread. The tubing was wound into a helix of 1/4-in o.d. and placed in boiling water for 1 hr. This had the effect of cleaning the fiber glass spaghetti and setting the polytetrafluoroethylene into a quasipermanent helical shape.

3.3 Water Heat-Exchanger

Eight feet of stainless steel tubing with an o.d. of 0.0355 in and an i.d. of 0.023 in, wound into a 34-in helix, serves as the water heat-exchanger. This exchanger surrounds the saturator tube and is joined at one end to the polytetrafluoroethylene tubing of the saturating element and at the other end to a polytetrafluoroethylene tube which is fed through a pressure seal in cap V to the water-feed pump. At the normally used water flow-rate of 10 cm3/h. there is a 4 min supply of water within the heat exchanger. The heat exchanger is surrounded by a Dewar flask with an i.d. of % in. and a straight section of 14 in. This flask reduces heat losses from the region around the heat exchanger.

In order to protect the glass elements and to seal the saturator tube so that the test gas flows as desired, a brass cylinder 17 in long, with an o.d. of 3 in and a wall thickness of 1/8 in, closed at one end, surrounds the Dewar flask. The space between the flask and the brass cylinder is filled with foamed polystyrene and a rubber stopper. permanently positioning the flask within the cylinder along its axis. At the open end of the cylinder is a flange which mates with cap V and seals to the cap by means of a 3-in i.d. "O" ring. The cap is held to the cylinder with

screws.

The cap has four openings within it. At its center is an opening through which the saturator tube protrudes. By means of an "O" ring seal, the cap is sealed to the outer surface of the saturator tube and positions this tube within the Dewar flask.

Two of the other openings in the cap are polytetrafluoroethylene packing-gland compression seals. One of the seals holds the pressure tap and water-feed tube, and the other serves as a pass-through for the two thermistor leads. The remaining opening in the cap serves as a flow exit for the test gas.

3.5. Air Heater

The test gas heater L is formed by winding 25 ft of bare nickel-chrome wire on a 1-ft long, 1/4-in diam, polytetrafluoroethylene rod. The rod has a few longitudinal grooves for the lead wire and to expose more of the wire to the air flow. The heater is contained within a 3-in polytetrafluoroethylene tube which is surrounded with rubber foam insulation. The small diameter bare wire has a low thermal lag.

3.6. Temperature Regulator

The inlet test gas temperature is controlled by a proportional electronic temperature controller. The sensor R which activates the controller is placed in the test gas flow stream just at the inlet to the saturator tube.

3.7. Temperature Measuring Circuit The temperature measuring circuit is shown in figure 3. It consists of two separate Wheatstone bridges D and W, each connected to a thermistor encased in polyethylene tubing. One of the thermistors is held in place at the entrance end of the saturator tube by means of a polytetrafluoroethylene disk containing many holes, while the other thermistor is secured along the axis at the exit of the saturator tube by means of a small diameter wire. The bridge circuit is powered by 0.4 V from the power supply and the voltage is continuously supplied to each bridge circuit, including thermistor, whenever the psychrometer is in operation, in order to maintain constant self-heating of thermistors. A selector switch Sg allows the reflecting galvanometer G to be connected to either of the bridge circuits as desired. A switch Sp provides for a reversal of voltage polarity to the bridges in order to obtain an electrical zero in the balancing of the bridges. The polarity switch also removes the voltage from the bridge circuits whenever it is placed in its center position. There are three buttons, S1, S2, S3, which connect the galvanometer into the circuit, each having resistance circuits which provide for three different galvanometer sensitivities. The power supply is supplied with 18 V from a transformer.

The bridge circuits, power supply and galvanometer are contained in a separate enclosure within the instrument cabinet which is temperature regulated. Temperature in the enclosure is maintained by means of a 44-W heater and a miniature thermoswitch preset to 40 °C. A miniature blower within the enclosure operates continuously, and a panel light on the top of the psychrometer indicates whenever the heater is on.

3.8. Flow System

Flow is drawn through the psychrometer by means of a moisture-resistant vacuum pump, P. Upstream of the vacuum pump is a nozzle assembly J which limits the flow to approximately four ambient liters per/minute. Upstream of the nozzle assembly is a water trap T which separates liquid from the exit gas. The level of liquid in the trap is determined visually, and the trap is drained by means of a plastic tube H connected to the trap.

3.9. Water-Feed System

When operating, water is pumped into the psychrometer at the rate of 10 cm3 per hour by means of a syringe pump K. A motor drives the plunger of a 100 cm3 glass syringe. An "O" ring fits into a groove on the piston to provide a leak-free seal in the syringe. Connected to the syringe is a two-way automatic valve which allows for filling of the syringe without disconnecting it or disassembly. A plastic filler tube S remain connected to the twoway valve at one tap. Attached to the other tap of the

two-way valve is a hypodermic needle to which, in turn, is connected the tubing HH that goes through the pressure seal in the cap V of the brass case G to the heat exchanger M in Dewar flask B.

In the region around the syringe K and liquid trap T are located three 15-W miniature light bulbs connected in parallel. These are connected, in turn, through an adjustable liquid-bellows type thermoswitch, directly to the power cord to the instrument, and constitute the freezeprotection circuit.

3.10. Pressure Measurements

A plastic tube connects the pressure tap in the psychrometer to a differential pressure gage Q, having a range of 0 to 20-in of water. The gage is mounted at the top of the instrument. The gage pressure, when subtracted from the ambient pressure (independently determined), gives the pressure P in the psychrometer. If the pressure of the test gas is not at atmospheric pressure but is known, another plastic tube can be connected to the reference port of the differential pressure gage Q and connected to the test gas source. The psychrometer pressure P is then the source pressure less the gage pressure.

x = 0.9847 Set + 0.001 RD for the inlet thermistor and

x = 0.9846 Setw+0.001 Rw for the outlet thermistor. Set D and Setw are indicated switch positions from 1 to 5 and represent nominal bridge resistances of 5,000 N per unit. RD and Rw are helipot readings from 0 to 1,000 and represent nominal bridge resistances of 5 per unit. The values 0.9847 and 0.9846 are the ratios of the mean of the step resistances to the maximum variable ratio of 2.5 to 1 and therefore the resistance of the therresistance in the corresponding bridge. The bridge has a mistors is approximately 2,000 x. The thermistors were simultaneously calibrated at 34 different temperatures against a calibrated platinum resistance thermometer. Two of the calibration points were eliminated from both calibrations and one other from the outlet thermistor calibration for statistical reasons. Fits were then obtained as follows:

T = 67.0965232.43244 In x + 2.109868 (ln x)2 °C (3)

Tw=61.16210 — 32.65236 In x + 2.230575 (ln x) 2 °C. (4)

The residual standard deviation for TD was 0.01 °C and for Tw it was 0.007 °C.