A Mercury-Balance Pressure Regulator A , J. BAILEY, University of Washington, Seattle, Wash.
0
F THE numerous pressure regulator designs reported,
the large majority involve elaborate and costly apparatus which permits neither rapid assembly and easy adjustment nor portability and application t o different operations involving reduced pressure. Broadly, pressure regulators can be classified as intermittent or continuous in operation. Most of the elaborate electrical assemblies maintain the pressure at a n average value by intermittent pumping, or intermittent leaking with constant pumping Basically, all these intermittent devices would appear t o produce pressures which oscillate above and below the average value, the degree of oscillation depending upon the ballasting effect of the entire system and the sensitivity of the compensating device. Continuous regulators, on the other hand, adjust the amount of leakage so that the sum of gas from the vacuum system plus the controlled leakage equals the capacity of the pump. The precision of pressure regulation then depends on the precision of the proportioning valve-i. e., the sensitivity of the regulator-which is a compound factor determined by the smallest change \vhich causes compensation, the rapidity of compensation, and the magnification of the compensation. Ideally, a continuous regulator should react swiftly and to a n exaggerated degree to a minute departure from established pressure. The designs involving direct control of leakage by me1 cury balance appeared t o offer a precise method of obtaining continuous regulation with all three of these characteristics. The precise types devised by Schierholtz ( 4 ) and Thelin (6) used the translation of mercury in a manometer to actuate a lever which controlled the magnitude of the leak. The mercury control of Emerson (3) employed mercury entrained in the stream of leakage to obtain the final degree of throttling. The designs of Swayze (6) and Watts, Riddick, and Shea (7) are examples of intermittent operation of pump and leak, respectively. The apparatus of Caldwell and Barham ( 2 ) employed a glass valve operated diiectly by the mercury in a manometer. Numerous other devices, including many of considerable complexity, have been reported.
screw and pivot were supported by the cold-rolled iron rod, H , which was attached to a ring stand by a right-angle clamp. To this rod was attached a rubber stopper, I , which served as the valve seat. The stopper was preboiled in dilute alkali and the glass seat was ground flat with a fine pocket whetstone.
Operation The regulator in Figure 1 was operated by connecting to the vacuum system with the spring relaxed. When evacuated to the proper degree, the spring was tightened until the valve barely opened. While the actual opening was not usually easily visible, the effect was noted on a separate manometer. Once adjusted, a decrease in pressure in the system caused the mercury in the manometer to flow into the reservoir, unweighting the manometer-beam and opening the valve. An increase in pressure caused opposite effects.
Eltvolion
-Apparatus
5O
Apparatus patterned originally after those of Schierholtz and Thelin was improved through four successive models. The use of a mercury column to balance the excess pumpage by leakage was found to be precise, rapidly responsive t o disturbance of equilibrium, and capable of indefinite magnification of compensation. Two modifications of the final device are presented here. They employ the same principle as those of Schierholtz and Thelin, but have the mechanical beam eliminated and incorporated in the manometer; one has a loop type of manometer which permits wide range and easy adjustment. Both require only elementary glassblowing technique to construct. The construction is shown in Figures 1 and 2.
ClnhrneIerS
FIGURE 1. PRESSURE REGULATOR WITH CLOSEDMERCURY MANOMETER
The regulator in Figure 2 was similar in operation, except that the stopcock was left in the open position while the proper degree o! vacuum was established roughly by adjusting the spring tension. The stopcock was closed and the final adjustment made by adjusting the spring tension. A decrease in pressure then caused the mercury to flow out of the closed end of the manometer, unweighting the manometer-beam and allowing the spring tension to open the valve.
Precision of Regulation The precision obtained with the regulator shown in Figure 1 was about 0.1 mm. of mercury; t h a t of the type in Figure 2 was not readable with the naked eye. Many factors influence precision. The precision obtained with these regulators was the algebraic sum of these variables under given conditions and was by no means t h e ultimate precision of n-hich the device was capable. Spring tension affected sensitivity greatly. A spring having a small load-deflection ratio was more sensitive than
Three sizes of tubing were used: 1-mm. capillary, 8 mm. in outside diameter, and 19 mm. in outside diameter, as indicated by A , B, and C, respectively, in Figures 1 and 2 (lettered identically). The glass manometer was supported in two places: by the spring, D, and the pivot, E. The spring contained 90 turns of 26-gaqe piano wire coiled on a 4-mm. rod and was connected by a swivel, F , t o the thumbscrew, G, which adjusted the tension in the spring. Pivot E was a glass tube at right angles to the manometer, which served both as the vacuum connection and tht. qhaft abnut which the manometer rotated. The thumb283
284
INDUSTRIAL AND ENGINEERING CHEMISTRY
Vol. 15, No. 4
sure by means of the thumbscrew, G, having 32 threads per inch, was easily effected. Both regulators controlled pressure with uniform facility over the entire pressure range. Both regulators were susceptible to “bouncing” under certain conditions-i. e., an oscillation of the mercury column and concomitant movement of the manometer-beam, causing intermittent admission of air. It was found to be due t o parallelism of valve seat and face, and was eliminated by an imperceptible bending of the valve seat support, so that valve seat and face were not quite parallel and hence the valve could not close completely. The bouncing apparently was due to complete closure of the valve. I n order to open the valve, pushing against atmospheric pressure required overcompensation of the O Centimeters 5 mercury-balance, and then rapid overcompensation in the opposite direction resulted FIGURE 2. PRESSURE REGULATOR WITH LOOPTYPEOF MANOMETER owing to the sensitivity of the device, again causing sealing of the valve. This bouncing phenomenon has been noted in other regulating devices (1) using a mercury column and is due t o a stiffer spring. The diameter of the tubing in the closed similar causes. limb of the manometer controlled the weight of mercury Generally speaking, the Figure 2 type of regulator was shifted for a given pressure change, hence larger tubing in more flexible and sensitive] but no stopcock under vacuum the closed limb of the manometer greatly increased sensitivity; can be considered completely dependable under long periods because of location the size of the open limb was unimpoitant. of operation. Either regulator may be made capable of A bubble of air was always trapped in the closed end of the increased sensitivity by increasing the diameter of the manometer in the Figure 2 type of regulator; a larger bubble tubing in the closed manometer limb, lengthening the beam, resulted in increased sensitivity. A bubble trapped above using a more elastic spring, or trapping a larger volume of the mercury in the Figure 1 type of regulator increased the air in the closed end of the manometer. Where great sensensitivity and increased the pressure range over which the sitivity is not required, a rubber cushion can be used to redevice was operative. While the pivot bearing on these place the spring suspension, but all attempts t o use rubber regulators was the glass tube, better practice would be to have necessitated continual adjustment, owing to the fatigue sheath this tube with a machined, split metal bushing, since of the rubber cushion. glass tubing is rarely round and is apt to stick in the bearing. It would likewise be better practice to connect to the vacuum Literature Cited connection, E,with a loop of thin-walled rubber tubing, prevented from collapsing by Raschig rings inserted a t intervals, (1) Bailey, A,, Science, 79, 277 (1934); 86, 525 (1937). and with the other end of the loop fastened permanently to (2) Caldwell, M. J., and Barham, H. N., IND. ENG.CHEM.,ANAL. ED.,14,485 (1942). the support bar, H , so that torque on the regulator by chang(3) Emerson, R. L., and Woodward, R. B., Ibid.,9, 347 (1937). ing the position of the suction tubing would not change (4) Schierholtz, 0. J., Ibid.. 7, 284 (1935). the adjustment of the regulator. (6) Swayze, C. I., Science, 76, 196 (1932). Both regulators were independent of variations in at(6) Thelin, J. H., IND. ENG.CHEM.,ANAL.ED.,13, 908 (1941). mospheric pressure. Adjustment to a predetermined pres(7) Watts, C. E., Riddick, J. -4., and Shea, F., Ibid., 14, 506 (1942).
-
Electrolytic Preparation of Quinhydrone ROBERT E. ELY’, Schenley Distilleries, Inc., San Francisco, Calif.
AN
ELECTROLYTE used extensively in p H measurements is quinhydrone, which consists of bright platinum immersed in a saturated solution of quinhydrone. I n view of the fact that quinhydrone is an equimolecular compound of hydroquinone and quinone, it was thought possible that hydroquinone, a relatively cheap product used in photography, might be oxidized to quinone by the use of an electric current. The resulting product would be free of foreign materials. The apparatus consists of an outer cup of nonporous material 20 cm. (8 inches) high and 20 cm. (8inches) in diameter and an 1
Present address, 109 West Moneta St., Bakersfield, Calif.
inner cup of porous material 17.5 cm. (7inches) high and 7.5 cm. (3 inches) in diameter. A carbon electrode, obtained from a dry battery such as those used in residential doorbell circuits, is placed in the outer cup (anode) and another is placed in the inner cup (cathode). TABLE I. PREPARATION OF QUINHYDRONE ’ Expt. No.
Electrolysis Time
Amperage
Hydroquinone
Hours
Amperes
Urams
Yield %
