Regeneration of Glass Mercury Rectifiers
Abstract Generated abstract
This paper describes a practical method for regenerating spent glass mercury rectifiers when replacement units were unavailable, with particular reference to rectifiers used in Kuzbass mine electric haulage. It identifies common failures, including loss of electrical contact, vacuum leaks, gas desorption, electrode deposits, and glass damage, and presents repair procedures involving cleaning, glass sealing, mercury purification, high-temperature evacuation, repeated arc ignition, and vacuum testing with accessible industrial equipment. The author reports that stable high vacuum formation is the most critical stage and gives operating parameters for heating, pumping, overload testing, and leak detection. Experience with regenerated rectifiers is said to show reliable operation and, in many cases, longer service life than new units, while reducing demand for new rectifiers and transport of fragile equipment.
Full Text
REGENERATION OF GLASS MERCURY RECTIFIERS
N. P. Kurin
Glass mercury rectifiers are widely used in engineering for converting alternating current into direct current. During the Patriotic War, as a result of the blockade of Leningrad, the production of new glass mercury rectifiers ceased, and mechanized electric haulage of coal in the mines of Kuzbass was threatened with disruption. Replenishing failed rectifiers of the ZVN 100 and ZVN 100/500 types by organizing the manufacture of new ones was, for a number of reasons, not possible; therefore, at the request of Kuzbassugol and the People’s Commissariat of the Coal Industry, we developed a method for regenerating spent glass mercury rectifiers of any make.
Faults of Spent Rectifiers
Without dwelling on a description of the types and designs of various mercury rectifiers, since this information is available in the literature [1], we shall point out that mercury rectifiers can operate only as long as the electrical connections are in good order and a sufficiently high vacuum is maintained in the rectifier.
According to our observations, the following rectifier faults arise during operation: melting of the solder that creates contact between the bases and the metallic electrode leads; loss of vacuum due to the presence or appearance of the finest channels and cracks—often invisible to the eye—in the leads of the molybdenum rods, as well as in the glass walls hidden under the electrode bases; loss of vacuum, with all parts of the rectifier in complete working order, due to desorption of adsorbed gases if they were poorly removed by the manufacturer; darkening of the horns of the main anodes, the excitation anodes, and the bulb body as a result of electrode sputtering and decomposition of organic impurities; perforated melt-throughs, cracks, fractures, and breakage of the glass parts of the rectifier.
All the listed damage is caused by the inevitability of certain processes (for example, slow sputtering of the electrodes), by improper operating conditions (overheating under overload and switching off the cooling), and by rough handling by the service personnel.
Regeneration Technique
The detection and elimination of faults in the electrical connections (for example, loss of contact between the bases and the molybdenum electrode leads) do not present substantial difficulties and are carried out by ordinary electrical-engineering methods.
Soldering work during restoration of glass parts must begin with thorough cleaning of all internal surfaces of the rectifier, since subsequently this will accelerate the evacuation process and create—
…of a stable high vacuum. For this purpose the inner walls of the bulb, the anode horns and the ignition branches, by washing with periodically replaced clean dry mercury, are thoroughly cleaned of films and powdery deposits capable of retaining considerable quantities of adsorbed gases; if this proves insufficient, wiping with a clean cloth, washing with a 5% solution of nitric acid and with distilled water are used, taking care that the liquid does not wet the electrodes. Newly installed graphite electrodes, in order to destroy organic contaminants deposited on their surface (for example, grease and sweat from the glassblower’s fingers), are first calcined in the flame of a soldering torch.
Restoration of glass parts (cracks, breakage of the horns of the main anodes, excitation anodes, the ignition branch, the cathode lead-in, and the lower cylindrical part of the bulb) is carried out by the method of sealing in the flame of illuminating gas or water gas [2]. The guarantee of the strength of the resulting seals is the cleanliness of the surfaces being treated and the use of one and the same grade (molybdenum) of glass. If, despite preliminary cleaning, small spots and points heated to white heat appear at the place of the joint being treated by the flame, this indicates the presence of contamination, and it must be removed by pulling off part of the glass with a glass rod or tweezers; otherwise, after cooling, cracks will arise in the contaminated joint.
If the indicated rules are observed, restoration of glass parts can be carried out without substantial difficulty by a glassblower of average qualification, provided he has at his disposal, in sufficient quantity, illuminating or water gas, and stationary and hand burners.
After the indicated repairs, clean mercury must be poured into the rectifier; when dirty mercury is used, the process of creating a stable high vacuum is hampered, owing to the slow release from it of dissolved gases and vapors.
As our experiments have shown, to remove the oxide film it is sufficient to pass the mercury in a thin stream through a paper filter with a punctured fine opening, and then, to extract metallic impurities, wash it with a 5% solution of nitric acid [3] and with distilled water. Mercury purified in this way must, before being poured into the rectifier, be heated to the boiling temperature; in this process most of the aqueous vapors and dissolved gases are released from it.
The most complicated, difficult, and lengthy part of the regeneration of glass mercury rectifiers is the creation in them of a stable high vacuum. As is known [1], a high vacuum ensures stable burning of the voltaic arc, its valve action, and reduces to a minimum the possibility of the appearance of so-called “back ignitions,” which have a disastrous effect on the preservation of the rectifier. Meanwhile, the constituent parts of air—nitrogen, oxygen, carbon dioxide, water vapor, and other impurities—not only fill the internal volume of the rectifier prepared for evacuation, but are adsorbed and firmly retained by the glass walls and graphite electrodes [3]. In addition, in the course of soldering work the glassblower may introduce into the rectifier organic impurities capable, upon heating, of decomposing with the release of gases and vapors.
The layout of the installation for evacuation is shown in Fig. 1.
Its main features are: 1) the possibility of heating the glass walls of the rectifier and its graphite electrodes to elevated temperatures in order to accelerate the desorption of adsorbed gases and vapors and the decomposition of organic impurities; 2) the possibility of creating a high vacuum.
Fig. 1. Diagram of an installation for regenerating glass mercury rectifiers.
As our experiments have shown, the gases and vapors filling the internal volume of the rectifier are evacuated rapidly, whereas the removal of gases and vapors adsorbed by the graphite electrodes and glass walls proceeds extremely slowly at room temperature. To accelerate this process we heated the entire rectifier to 185–200°C, and the graphite electrodes to a red heat.
Owing to the properties of molybdenum glass, heating of the glass walls of the rectifier can quite well be brought up to 450°C, which will accelerate still further the process of desorption of gases and vapors. However, with the electrode-heating procedure we adopted, we used this successfully only in those cases where gypsum \((t_{\text{decomp.}} = \sim 185^\circ\text{C})\) [4] and tin \((t_{\text{melt.}} = 232^\circ\text{C})\) [5] had not been used in mounting the electrode bases. Naturally, the maximum temperature for heating the cathode space was determined by the boiling point of mercury.
As is known, in factories, when new rectifiers are manufactured, heating is carried out in special installations, in particular by high-frequency currents. In order to avoid difficulties associated with the absence of the corresponding equipment, for electric heating we used the appropriate mercury-rectifier installations that are ordinarily employed by enterprises in operating bulbs of this type (for example, for the rectifier ZVN 100/500, ZVN 100 and 2V 20, respectively, the mercury-rectifier installations URV-33, URV-43 and URV-9) [1]. The cabinet (2) of these installations is insulated with sheet asbestos and is provided with an illuminating lamp (43), a thermometer (44), and a nichrome electric-heating coil (4).
After the repaired bulb (1) has been filled with mercury, placed in the cabinet of the rectifier installation (2), and connected to the electric circuit, it is sealed through the narrowed branch tube (3) at point (A) to the vacuum installation. The flexibility of the connection required when the rectifier is ignited is provided, if possible, by a short insert (4) of thick-walled vacuum rubber. Then the electric-heating coil (5) is switched on and the temperature in the closed cabinet (2) is brought to 185–200°C (or higher). Simultaneously, with the cocks (6), (7) open, the oil fore-vacuum pump (8) [3] is switched on, and after evacuation of the vacuum installation and the rectifier to \(0.1\text{--}0.01\) mm of mercury, which usually takes 20–30 minutes, a metal mercury pump of the Leytmgor type (9) [3] is started. For this purpose the electric-heating coil (10) is switched on, and the mercury vapor formed in the evaporator (11) is directed through the tube (12) by the umbrella (13) into the condensation space of the cooler (14), entraining along the way the gases and vapors entering from the trap (15) and the bulb (1). The drops of mercury condensed on the walls of the cooler (14) flow down between the tube (12) and the protective cylinder (16) and through the tube (17) again enter the pump evaporator. The gases and vapors entrained by the mercury vapor enter the 2–3-liter fore-vacuum balloon (18) and from there are evacuated to the atmosphere by the oil pump (8). The pressure in the system during evacuation is measured with a shortened mercury manometer (20) and a McLeod manometer (21) [3]. In the latter case, with cock (22) closed, cock (23) is slowly opened and the mercury from vessel (24) is transferred into manometer (21) until it reaches in capillary tube (25) the level shown in the diagram; the difference in height of the mercury columns \(h\) in capillaries (25) and (26) makes it possible, from the calibration curve or from the equation \(P = a h^2\), to determine the pressure in the system (for our manometer the value \(a = 1.85\)). After the value of \(h\) has been read, cock (7) is closed; with cock (23) closed, cock (22) is opened, and by evacuation the mercury is again transferred into vessel (24); then cock (22) is closed, and after 1–2 minutes, when pump (8) again creates a good
fore-vacuum, valve (7) is opened, and pumping is continued until the next pressure measurement. For evacuation, the communication connecting the Langmuir pump and the rectifier should, if possible, be made short and of wide glass tubes, with a diameter of not less than 10–15 mm. The lubrication of ground joints (27) and (28) and of the valves must be done with high-vacuum cement \[3\]. For cooling trap (15), a mixture of solid carbon dioxide and ethyl alcohol or liquid air should be used. However, as our experiments have shown, such cooling of the trap may be dispensed with if ground joints (27) and valve (6) are lubricated with good high-vacuum cement, since in this case, owing to the low elasticity of the vapor, the probability of organic vapors entering bulb (1) is reduced to a minimum. Cooling of the condenser of mercury pump (14) is carried out with tap water; if the water supply cannot provide continuous feeding, then in this case cooling may be effected by circulating water between two glass bottles (28) and (29) of 15–20 liters capacity.
As soon as the pressure in the bulb drops to 0.001–0.008 mm of mercury, heating is stopped, the cabinet doors are opened, fan (45) is started, and after the rectifier has cooled approximately to room temperature, it is connected, by switching on disconnectors (30) and (31), through autotransformer (32), excitation transformer (33), choke coils (34) and (35), to the alternating-current source (3). By the usual breaking of the mercury in ignition branch (36), with careful rocking of the bulb, and by closing contact (37), the arc is ignited on the excitation anodes (38) and the main anodes (39), and through wire rheostats (40), (41), or liquid rheostat (42), a direct current of 25–50% of the rated value characteristic of the given type of rectifier is drawn (for example, for rectifiers ZVN 100 and ZVN 100/500 this is 25–50 amperes; for ZVN 30 and ZV 30—7.5–15; for 2V 6—3 amperes).
The gases and vapors desorbed from the electrodes are continuously pumped out. If the rate of desorption, which depends on the temperature of the electrodes and consequently also on the current strength, is greater than the pumping rate, then the pressure in the bulb rises and reaches limits at which the arc is extinguished; by evacuation during repeated heating of the rectifier in cabinet (2) to 185–200°C, the pressure in the bulb again drops to the above-indicated limit, and then, after cooling, the arc is reignited. As a result of several ignitions and pumpings, the arc begins to burn steadily at the indicated current strength, after which the current strength is gradually increased by means of the rheostats and brought up to the rated value.
During operation of the voltaic arc, the mercury evaporating in the cathode spot rushes at high speed into condensation vessel (1) of the rectifier and partly carries with it the gases and vapors released from the heated electrodes and glass walls. The latter gradually accumulate in the upper part of the vessel, and in order to remove them, after every 1–2 hours of burning, it is advisable to switch off the voltaic arc and pump out the indicated gases. After the arc is switched off, the pressure curve in the bulb, as a function of time, passes through a maximum. This can be explained by: 1) the gradual cessation of the suction action of the jet of mercury vapors rushing from the cathode spot into the condensation chamber, as a result of which the gases accumulated in the dome of the bulb spread throughout its entire volume; and 2) the release of gases and vapors adsorbed by the very fine droplets of mercury rain filling the entire volume of the bulb during arc burning. As Fig. 2 shows, as pumping proceeds with alternating ignitions, the magnitude of these maxima gradually decreases.
For more intensive heating of the electrodes and the release from them of the most strongly adsorbed gases and vapors, after the next 1–2-hour
of hourly operation at the rated current (for example, for ZVN 100 and ZVN 100—500—100 amperes), the rectifier is overloaded by 50% for 5–10 minutes. Usually, after an hour of continuous operation at the rated current, the electrodes acquire a constant temperature characteristic of red heat; overloading the rectifier by 50% raises the temperature to light-red heat. After pumping down to a pressure of \(1\cdot10^{-4}—4\cdot10^{-4}\) mm of mercury column, the rectifier is sealed off with a hand soldering torch at branch (3). During evacuation, the production log should record the time, temperature, magnitude and voltage of the rectified current, the current in the excitation anodes (if present), and the pressure in the rectifier according to the McLeod manometer scale.
Fig. 2. Dependence of the pressure in a glass rectifier on evacuation time. On the sections of the curves shown by dotted lines, the electrodes were being heated.
The electrodes that were repaired before evacuation are fitted with bases. For this purpose, brass bases are placed on the corresponding leads, and the intervening space is filled with gypsum paste (water content 34–40% by weight). After the gypsum seal has hardened, the copper leads from the electrodes are soldered with tin to the bases; as flux, a mixture of rosin and glycerin is used (~45% glycerin). The rectifier is tested on the same installation.
As our experiments have shown, quite reliable results, in the sense of verifying the quality of the regenerated rectifier, are obtained if the rated current is passed through the rectifier for 2 hours and it is then overloaded by 50% for 10 minutes; in this case the rectifier must not go out. A well-evacuated rectifier, when the voltaic arc is ignited, should produce a greenish or bluish glow in the cathode space and, when mercury is poured from the cathode into the condensation chamber, should produce not a soft sound, but a characteristic sharp, metallic sound.
Considerable assistance in checking the tightness of glass seals and walls and in detecting the smallest channels, often invisible to the eye, is provided by the glow of gases in evacuated parts under the influence of the current from a small Ruhmkorff coil [3]. One end of the coil is connected to one of the rectifier electrodes, and the other is led along the part being examined.
surface; in this case the usual pinkish glow at the leak site intensifies to the character of a spark. However, by the indicated method it is impossible to detect a leak at the molybdenum lead-ins of the electrodes, as well as in places close to them, since the discharge from the Ruhmkorff coil does not pass through the glass but tends toward the metallic lead-ins. In these cases the electrode lead-ins of the previously evacuated and sealed-off bulb are successively filled with mercury by pouring it over from the cathode space; if there is a leak, air bubbles periodically bubble through the mercury.
The duration of creating a stable high vacuum is an individual property of each rectifier, if the same pumping installation is used. Under normal continuous operation, this value for rectifiers from 2V-6 to 3VN100/500 usually lies within 6–24 hours. Usually the greatest pumping duration is observed for those rectifiers whose glass walls have a dark deposit, whose repaired electrodes have been handled with sweaty fingers and have been weakly annealed in the flame of a soldering burner, and whose poured-in mercury has been poorly dried. In all these cases, the release of adsorbed gases and vapors proceeds more slowly and the total pumping time increases.
Experience with the glass mercury rectifiers regenerated by us in the mines of Kuzbass showed that these rectifiers possess high qualities and, in terms of service life, as a rule surpass new rectifiers (3000–4000 hours instead of ~1000 hours).
At various times the following took part in practical regeneration work: assistant P. I. Tumashev, assistant G. K. Shavrov, assistant L. L. Skripova, student A. T. Kuleshova, assistant A. A. Semenov, preparator A. F. Kolesnikova, glassblowers I. V. Chizhkov and G. I. Vorobev.
Conclusion
The method described makes it possible quickly to organize the regeneration of spent glass mercury rectifiers at various mines and other enterprises using rectified current on a large scale, on the basis of existing and accessible equipment.
Regeneration of rectifiers will sharply reduce the need for new rectifiers and will free railway transport from the carriage of bulky and fragile freight.
LITERATURE
-
A. G. Belyavsky—Mercury Rectifiers of Alternating Current. GEI. 1933.
V. P. Vologdin—Rectifiers. ONTI. 1936.
I. L. Kaganov—Electron and Ion Current Converters. ONTI. 1937.
I. I. Livshits—Underground Electric-Locomotive Haulage. ONTI. 1934.
S. A. Volotkovsky—Underground Electric-Locomotive Traction. GONTI. 1939.
I. A. Proskurin, V. M. Efimov, V. K. Stasyuk—Underground Electric-Locomotive Haulage. SONTI. 1941. -
A. E. Arbuzov—Manual for Independent Study of the Glassblower’s Art. ONTI. 1935.
-
M. Ya. Lapirov-Skoblo—High Vacuum. GNTI. 1931.
L. Du Noüy—High-Vacuum Technique. GTTI. 1933.
S. V. Ptitsyn—Vacuum Technique. Printing House of Lengidromez. 1934.
K. V. Chmutov—Technique of Physicochemical Research. ONTI. 1937.
N. A. Kaptsov—Physical Phenomena in Vacuum and Rarefied Gases. ONTI. 1937. -
Technical Encyclopedia. Vol. 5, p. 612. 1930.
-
Collection of Physical Constants. ONTI, p. 55. 1937.