US2527298A - Baffled time delay mercury switch - Google Patents

Description

Oct. 24, 1950 J. w. CAMPBELL BAFFLED TIME DELAY MERCURY SWITCH Filed Sept. 25, 1947 I INVENTOR. Jamar W F [1/72 Patented Oct. 24, 1950 UNITED STATES PATENT F BAFFLED TIME DELAY MERCURY SWITCH James W. Campbell, Rockville Centre; N.-?Y. Application september 25, 1947, Serial No. 776,128

.5 Claims. 01. res-42 :The invention relates to delay action'mercury contact switches.

More particularly, this invention relates to the means for controlling the delay in the mercury flow of mercurycontact switches. J The invention consists in'='a mercury fiow con.- trol'device in which the surface tension'of the mercury is neutralized, that is prosophy (adhesion of the fluid) is greater than the synaphy (cohesion of the fluid). and capillary "action u'tilizedinthe spaces through which the mercury passes,"so as to increase the delayfactor of the same quantity of mercury usually contained in the standarddelay mercury-switches, which type seldom" attains a delay of "more than seconds.

The improved device is of a materialwhi'ch is ire-amalgamated or "wetted with mercury. This wettingof the surfaces enables the mercury to 'flow readily in a spread out form over the wetted surfaces and move by capillary action flow. As a practical embodiment, the invention contemplates a plurality or'pile of copperdisks, pie-amalgamated save for their circumferential surfaces, and each having a conduit passing transversely through the disk, the axis of which {is substantially parallel to the axis of the disk, the conduit of one disk being staggered to the conduit'of the'adjacent disk. Thedi'sks, in suitable number, are held together-by suitable means, in a pi1e,to-have their ra'dial surfaces contiguous to each other. The electrodes are also preamalgamated for positive make andbreak. The invention will be more fully described hereinafter, an embodiment thereof f'shown in the drawing, andthe invention will -be finally pointed out in the claims.

In the accompanying drawings,

Fig. l is a longitudinalsectional view of the improvedmercurysw-itch including a hermetically sealed glass tube containing-the improved control means, the mercury "charge, and the electrodes; v v

Fig. 2 is a 'per'spectivevi'ew 0f the improved control means separate'fromfthe glass tube;

Fig. 3 is aperspective view of a single disk;

Fig. 4 is a central section of the control means;

"Fig. 5 is an end view of the same;

Fig. 6 is a spacedassembly of the improved control means. j

Fig. 7 is 'a diagrammatic drawing of apile-of disks, without shims, showing "the ."capillary action, and

Fig. 8 shows an electrode with mercury'adhering=thereto.

Similar characters of reference indicate correspondingparts throughout the various views.

The improveddelay mercury switch consists of t a hermetically sealed glass tube II in which there is placed the improved control means or unit-l2. An electrode connection His "electri- 2 1 cally-connected withsaid unit l2 and passes out of the tube and anothed electrode I3 is-spaced from said'unit l2-and passes out ofthe tube.

The improved unit l2, consists of aplurality of disks [1 having-an outside diameter-substantially equal to'the inside diameter of theglass tube l I. Each disk I! hasa central bore 23;and an'ofi centre transverseor axially aligned'conduit l9 extending from face to face of each disk [1. The disks I! may be separated from each other by shims I8, in the form of a thin washer non-soluble by mercury, the outer diameter of which is less than that of the disks. The shims l8, when used, have a centralbore I 8a ofsub- 'stantially the same diameter as the bore 23 of the disks. The disks I! with the intermediate shims. iii, if present between the disks I'L'are assembled by means of a headed bolt l6,"with"a washer 2|, at each end of the'group or pile of disks I1, and with a nut l6a engaging the threaded end IBb of the bolt It. The disksare so disposed that the successive conduits l9 are not in alignment with each other, preferably with each conduit IS in a successivequadrant of the next disk. I

The -disks H are made of pure-copper (for practical purpose 99% purity is suitable). This has a solubility in mercury of one tenth 'of'one percent (reference: Mellors Chemistry). The single plates may be separately amalgamated with 'mercury'by letting each stand in a poolof mercury for "a suitable length of time. Or, each disk maybe provided with a surface him of domestic solder, and then introduced into the mercury, whereby they will become immediately amalgamated. The peripheraledge 22 in each case must remain unamalgamated. This may be accomplishedby staining the same with Wright Stain, or'applying an aluminium silicate and silicate of soda'mixture providing the temperature whenthe tube is sealed is not excessive. This operates to prevent the mercuryrfrom running along the peripheral edges of thedisks and between the rims of the disks and glass tube, which would short cut its intended run through thecontrol means or unit l2. The circumferential surface of the bolt I6 is preferably covered with the foregoing stain, or the like. If the bolt is of wrought iron, this'covering is not necessary. Electrolytic amalgamation may also be used.

-A' second set of electrodes must'be placed in the lower pole' of the chamber if a repeatable actionis desired. I

The assemble'd'unit' may be secured'in the tube with a tight fitting non-wettabler ("in mercury) sleeve, "or if an aluminium silicate and silicate of soda mixture is used on the edge 22 the unit may be introducedintdthe glass chamber with a relatively tight fit. The tube is then sealed after a vacuum is 'created and is readyior-service by mounting on a known mechanical tilting device as commonly employed.

The electrodes 13 and M are also amalgamated with mercury, in a similar manner as for the disks, with the exception of the portion extending along the glass chamber as in Fig. 1. Electrode 14 may be secured to the proximal plate of the unit l2 if :runs down to contact the first disk of the control unit l2 and enters the conduit l9, slowly oozes through the capillary'space created between the disks, and then finds the conduit IQ of the second disk in its opposite lower quadrant, and continues on into the second capillary space between the next disks, and so on. At the exit end orifice of the last conduit IS the mercury'accumulates as a drop where it adheres in a rounded form of a meniscus created by the cohesion of the surface tension. As this drop gradually increases in size and weight, the weight becomes greater than the surface tension adhesion and the drop falls into the lower part of the tube l l, accumulations forming a pool of mercury. When sufficient mercury has passed from the upper pool to the lower tube to free the spaced electrode l3, then the circuit is broken. I e

The pre-amalgamated electrode, as copper treated with mercury, as described, is important because it being wetted, the remaining pool of mercury frees itself from the electrode abruptly, after the adhesion of the mercury to the electrode has been overcome by the weight of a cone (Fig. 8) of mercury acting through its cohesion when abrupt severance of the cone from the electrode takes place, the cone of mercury merging into the remaining pool of mercury. In the embodiment, a break action takes place. In another embodiment, where a make action takes place, the wetted or amalgamated electrode is also. im-

portant, since as the pool of mercury approaches .closely the electrode, a joining cone of mercury 'will bind the electrode and the pool of mercury, :and even though the mercury pool 'may be vi- :brated as is frequently the case, the adhesion of the mercury to the electrode Will begreater than the cohesion of the mercury of the pool. As adhesion and cohesion in fiuids is under variables in distinction to solids, Frankenstein in 1835 prefers prosophy (adhesion of fluid) and synaphy (cohesion of fluid); This phenomena ofprosophy greater than synaphy is utilized in the movement of the mercury through the contiguoussurfaces of the disks, and in the use of the improved electrode surface.

The retaining of the features described 'depends on the selection of the material which allows this phenomenon to take place without dissolving the equipment away. We believe that copper, as pure as possible, containing little or no alloy achieves this, as its solubility in' mercury 3- so eroded as to render the equipment of no further use. However, rarer metals, rhodium, molybdenum and tungsten, as also nickel and cobal, are not readily soluble in mercury, and should be considered with iron and platinum as metals which might be usable.

For finer accuracy and lack of change in the timing delay, the mercury used in the pool can be treated by placing it in contact with pure copper for a period of time so as to substantially saturate it with copper before it is introducd into the tube for sealing, which would assure the mercury of the pool when passing through the conduits and capillary surfaces from dissolving the amalgam film or the body of the disks. For most purposes this will not be necessary. It would be within the invention to install the control disks in unamalgamated form into the tube and then operate it for a while, until the surfaces have been amalgamated, in which case the mercury would take up some copper. But, this is not practical for commercial purposes, which require testing before sale.

The more disks used in the capillary unitand the less space between plates, the greater will be the'delay of the mercury flow. The tension of the bolt maybe used to change the delay. The shim may be omitted for longer delays. In this case the plates would be separated by a microscopic roughness while yet affording capillary passage to the mercury. The estimated space created thusly is probably not greaterthan one ten thousandth of an inch, in view of the fact that a, three thousandth shim gives only a four hour delay and with direct contact I have achieved 56 hours.

In Fig; 7, the capillary spaces are shown much larger than the disks, to enable action of the creeping, of the mercury under capillary forces to be shown. Mercury has a surface tension of 470 dynes per centimeter in air and a specific gravity of 13.5 in contrast to water with a surface tension of 75 dynes per centimeter in air and a specific gravity of 1. It is these two features plus the solubility of the walls or edges of the orifice or passage-way in the interposedbarrier on which the mercury flow depends. I have found experimentally using a non-soluble barrier (not wetted) that a three thirty seconds of an inch circular orifice using 8 cubic centimeters of mercury in a half inch diameter tube created a five second delay with no mercury left in the upper chamber. Under simi lar, condition, a /1000 inch or about a of an inch orifice gave an '18 second delay with -l.3 cc. of -mercury leftin'the upper chamber. Further, /64 of an inch gave 50 seconds delay with ice. left, but required about a 40 degree angle of tilt to start the'flow, since at a- 30 degree angle there was no mercury flow. Further, of an inch, gave about 55 seconds delay with 5 cc. left. This was at a 45-50 degree tilt, as there was no mercury flow at 30 degrees. However, with the improved device using 5 cc. of mercury and 6 disks and /1000 of an inch clearance with a, 30 degree angle, a 3 minute delay was effected with no mercury left'in the upper chamber. A /1000 of anxinch clearance gave a 4 hour delay with a 30 degree angle with nomercury left. An estimated /1o,ooo of an inch gave 56 hours with a, 30 degree angle with all mercury used.

In the case of the insoluble non-wetted or. unamalgamated orifice of the delay switches heretofore proposed, the difn'culty arises in the relationship of the surface tension to theflweight of the mercury. As the mercury contacts the baflle,

' the surface tension must have just been to exceed the force of the weight of remaining mercury. Further, as was shown a 36 degree angle of tilt produced no flow. In this case the lessened angle of tilt reduced the head or the act 0 weight of the mercury above orifice to the extent that the suriace tension vvit -holding were greater at the start and no flow of y was obtained. In contrast to this, the impre device, an estimated /l0,000 of an inch space le no mercury at a 30 degree angle. This was fected because the pre-arnalgamated disk 33 down the surface tension forces by allowing tne adhesion created by the pro-amalgamated dislz: to break down the rounded meniscus and convert it to a concave surface oi the mercury due to tn greater adhesive forces. A very positive capillary flow results which cannot fail.

While I have herein shown and described my invention in the particular form, it is obvious that structural variations may be employed without departing from the spirit of the invention, and therefore, do not desire to be limited to such precise forms beyond the limitations as may be imposed by the following claims.

I claim:

1. In a mercury switch, a control unit for the flow of mercury, consisting of a plurality of contiguous disks of copper, coated with a film of mercury amalgam, with narrow spaces between the disks as results from their contiguity, and each disk having a transverse conduit for guiding the mercury to a space between the disks, and for guiding the mercury from said space to the next space.

2. In a mercury switch, a control unit for the flow of mercury, consisting of a plurality of contiguous disks of copper, coated with a film of mercury amalgam, with narrow spaces between the disks as results from their contiguity, and each dish having a transverse conduit for guiding the between the dist-gs, and for y to a guiding the mercury orn said space to the neat the circumferential oi the disks being mercury amalgam.

In mercury switch, a control unit for the flow of mercury, consisting or -a plurality of contiguous disks copper, coated with a film of mertween the ;s as results from their ccntiguity, and each disk having a transverse conduit for guiding the mercury to a space oetween disks, and for guiding the merefrom said to the next space, the circumferential rim or" the disks eing free from mercury amalgam and covered with stain or a layer of a. mixture of aluminium silicate silicate of soda.

l. In a mercury switch, a control unit for the flow of mercury, consisting of a plurality of con.- tigucus dlslrs of copper, coated with a him or mercury amalgam, with narrow between the disks as results from th contiguity, and each having a transverse conduit for guiding the mercury to a space between the and for guiding the mercury from said space to the next space, successive conduits being disposed out of alignment with each other.

5. In a mercury switch, a control unit for the gravity flow of mercury, consisting of a member having opposed contiguous surfaces forming a narrow space as results from their contiguity, to induce a capillary action flow of the mercury, restricted means supplying mercury to said space, restricted means for discharging mercury from said space, and a coating of mercury amalgam on the surfaces over which the mercury passes.

JAMES W. CAMPBELL.

REFERENCES CITED The following references are of record in the file of this patent:

UNITED STATES PATENTS Number Name Date 2,010,398 Keen Aug. 6, 1935 2,182,216 Staley Dec. 5, 1939 2,206,436 Spencer July 2, 1940 2,229,441 Carlson Jan. 21, 1941

Images