US1605860A

US1605860A - Thermocouple

Info

Publication number: US1605860A
Application number: US78739A
Authority: US
Inventors: Walter O Snelling
Original assignee: Individual
Current assignee: Individual
Priority date: 1926-01-02
Filing date: 1926-01-02
Publication date: 1926-11-02
Anticipated expiration: 1943-11-02

Description

Nm@ 2 9 E926@ E@ W. o, SNELLIN@ THERMOCOUPLE Filed Jan, 2, 1926 FIG. i

FIG. 2

Patented Nov.. 2, 1926..

UNITED STATES PATENT, FFICE.

WALTER O. SNELLING, OF ALLENTOWN, `IPEN'NSYL'V'AN'IA.

THERMOCOIT PLE.

l .Application led January My invention relates to improvements in the manufacture of thermocouples and thermopiles, andmore articularly relates to an improved method or uniting or welding together the wires or rods of a thermocouple or of a series of thermocouples to form a force 'to permit its use as a. substitute for an electric battery, and one of the objects of my invention isto provide means by which thermoelectric elements or thermocouples of low electrical resistance and of uniform thermoelectric capacity may be rapidly and cheaply made.

The two wires or rods forming a thermocouple. must necessarily be of different naturein order to ,produce a difference of electrical pressurey when the junction is heated, anda difference in melting point is a common characteristic of all pairs of thermoelectric elements which are used in the preparation of thermocouples and thermopiles. This difference in melting point' makes the rapid welding of the thermocouples somewhat diiiicult, and up to this time has pre- 'vented the automatic mechanica-l preparation of therinocouples of satisfactory character, although the welding of single thermoelectric elements by manual operations can be very conveniently and satisfactorily performed.

I have discovered that by the very simple method of using two wires or rods of somewhat diierent length at my thermoelectric junction, so that during the welding operation one wire or rod will project into the heating zone somewhat further `than the other, it is possible to make thermoelectric junctions much more rapidly than by the methods at present known, and that the method has the additional advantage of perlmitting the automatic mechanical forming of the thermojunction welds, and of producing welded junctions of exceptional mechanical and electrical uniformity.

In the drawings, Figure lis an illustration of a pair of twisted wires -icrmiiug a K thermoelectric junction, before welding -is begun. Figure 2 shows the same pair of twisted wires after being brought into a 2, 1926. Serial N0. 78,739. l

heating zone sufficiently to fuse the end of the projecting length of one of the wires. Figure 3 shows the same pair of twisted wires yafter further heating, the globule of molten metal Aformed by the fusion of the Erojecting length of the longer wire having een drawn by capillarity into the spaces between the two twisted wires, and welding them together.

Figure 4 is an illustration of a portion of acontinuous length of thermoelectric elements forming a thermopile, showing the progressive condition of the junctions as they ,are4 mechanically moved past theI heating zones used to produce fusion.

By modifying the excess length of the pro` jecting wire or rod, as compared with the length ofthe shorter wire or rod, I can controllably modify the temperature at which the weld is made, and also the amount of metal present in the welded portion of the junction. By using a projecting length of the longer wire of from five to ten times the diameter of the wire employed and using as myY longer wire themetal of lower melting point, I obtain a globule of molten metal which on solidifying consistspredominantly of the metal or alloy of the longer wire, which under some conditions is advantageous. By using the metal of higher melting poi-nt as the longer wire, the excess heat in the molten .globule of metal formed in the fusion step tends to melt the second wire to a somewhat greater extent than in the first case described, and under these conditions gives a weld in which the two metals 'are quite uniformly mixed in the welded zone.

In the preferred embodiment of my invention, I employ a projecting length of my longer wire equal to six to eight times the diameter of the wire, and I pass my thermocouples very rapidly through the heating zone. Heating may be conveniently produced by an electric arc, in which case the thermocouple may be used as one'of the electrodes,r or the arcm'ay be formed between two independent electrodes, the thermocouples vbeing passed through the arc without coming incontact with either electrode. Instead of using an electrode arc as a source of heat, I'may emplo an Oxy-,hydrogen or an Oxy-acetylene torc or. any other suitable source of heat capable of fusing the wires i more heat from the heating zone than does.

the portion of the thermocouple where the two wires are together due to its projecting further into the heating zone, and as the loss of heat by conductivity to the unheated portion of the wires is substantially the same for the two wires, this leads to the projecting length of wire fusing before the other wire has reached a melting temperature.

Heat is rapidly carried by conduction from the longer wire to the twisted junction of the two wires however, and by the time the molten globule formed by the fusion of the projecting wire is drawn by capillarity into the twisted portion of the two wires, the shorter wire is sutiiciently close to its melting point so that the large excess of heat in the molten globule of metal formed by the fusion of the longer wire is suiiicient to cause a very etlicient weld possessing very desirable charac teristics of mechanical and electrical uniformity.

In Figure l, a represents the projecting length of the longer wire of a pair of twisted thermoelectric elements, before heating the junction to produce welding. In Figure 2, b represents the globule formed by the fusion of the projecting end of the longer wire.`y In Figure 3, c represents the completely welded thermocouple, after the longer wire has melted back to the point where the two wires are twisted together, and the globule has passed by capillarity into the spaces between the twisted portion of the two wires. In Fi re 4, a represents the projecting lengths of the longer wire on each side of a continuous length of thermocouple elements` every other wire of the zig-zag being formed of one of the thermoelectric metals or alloys, and the alternate wire being formed of the second thermoelectric metal or alloy. b represents the projecting wires, after fusion of the projecting ends has started, but before it is complete, and c represents the finished thermoelectric junction, after the welding is complete and the zigzag has passed out of the heating zone.

In commercial manufacture it is convenient to provide two rolls or reels of the two dissimilar metals, placed on opposite sides of the machine making the thermocouples. Lengths of wire are alternately extended from each of the rolls, and the length so projected is twisted onto the free end of the continuous zig-zag reviously made. The wire last extended is then cut ntl', a length of the second wire is then entende twisted onto the free end of the wire cut oif, again giving a free end, to t free end, after extending a further quant of wire from the first roll, the wire tended twisted. By the means or hy any equivalent means, a A produced in which every alternate 'o one metal or alloy, and. .the interni-enlace Lacasse wires are of another metal or alloy. The junctions ofthis zig-zag are mechanically welded, preferably by employing two heating zones on opposite sides of the zig-zag, the junctions being passed through the heating zones until suitable fusion is effected, the temperature of the heating zones being such as to produce the desired degree of fusion during the period in which the junctions remain within the heated zone.

By employing a projecting length of one of the two wires forming each thermoelectric junction, it has been found that the welding which is produced in passing through the heated zone can by suitable control of the temperature used be made superior in mechanical strength and electrical Characteristics to the thermoelectric junctions previously made by hand operations, while the mechanical preparation of thermoelectric elements in continuous lengths makes it ossible to provide thermopiles of very satisfac-l tory character, and free from the occasional mechanical defects which have prevented hand-assembled thermopiles from achieving commercial success. As from fifty to as many as live hundred individual thermocouples must be employed in the construction of a thermopile for use as a vsubstitute for dry batteries in a radio set for example, and as a single defective weld will destroy the usefulness of the entire thermopile, the importance of having every individual weld show low resistance, good mechanical strength, and proper thermoelectric properties, will be evident. It has hitherto been found impossible to consistently produce the requisite uniformity by hand manufacture, while by mechanical twisting of the thermoelectric wires and automatic fusion of the junctions. the commercial preparation of thermopiles comprising a very large number of individual thermocouples has been rendered both simple and inexpensive, and the character of the resulting product has been found superior to the thermocouple junctions formed by manual operations.

It will be evident that many modifications may be made without departing from the es# sential features of the disclosure as herein made, and accordingly no limitations should be imposed in connection with my invention except as indicated in the appended claims.

I claim:

l. ln the manufacture of thermocouples the process which comprises hrineing together the free ends of two wires or different composition so that the end of one wire projects iiejfond the end of the other, and

fusing the projectii'igportion of wire until the molten metal formed enveiops the free end of the shorter wire.

Q. ln the manufacture of thermocouples he process which co ses twisting to- ,ti a. a :3 -z' gethci the i oi. no wires el. tuner- Yreoaeeo ent composition so that the end of one wire projects beyond the end of the other, and

passing the projecting end of wire through a heating zone until it is fused suiciently to form a globule uniting the two wires.

3. ln the manufacture of thermooouples the process which comprises bringing together the -free ends of two Wires of dierent composition so that the end of one wire projects beyond the end of the other, and striking an arc between the projecting wire and another electrode, until the two wires are welded together.

4. ln the manufacture of thermopiles the process which comprises bringing together alternate lengths ofwire of different composition, so as to leave a projecting length of one wire in .each pair, and fusing such projecting lengths of wire to weld together the associated wires of different composition.

5. In the manufacture of thermopiles the process which comprises feeding a length of wire, twisting together the free end of the wire thus fed and the cut end of a wire of different composition previously fed, ,cutting the length of wire last fed to form a new free end, feeding a fresh length of wire, twisting together the free end of the wire thus fed and the cut end of the wire previouslyl fed, and repeating these steps indefinitely to form a continuous length or zigzag of alternately disposed wires of different composition.

6. ln the manufacture of thermopiles the process which comprises feeding a length of wire, twisting together the free end of the wire thus fed and the cut end of a wire of different composition previously fed, cutting; the length of wire last' fed to form a new free end, feeding a fresh length of wire, twisting together the free end of the wire thus fed and the cut end of the wire previously fed, and repeating these steps indefnitely to form a continuous length or zig-zag of alternately disposedl wires of different composition, and fusing together the ltwisted ends of each pair of wires.

Zz ln the manufacture of thermopiles the process which comprises feeding a length of wire, twisting together the free end of the wire thus fed and the cut end of a wire of different composition previously fed, cutting the length of wire last fed to form a new free end, feeding afresh length of wire, twisting together the free end of the wire thus fed and the cut end of the wire previously fed, and repeating these steps indefinitely to forni a continuous length or zig-zag of alternately disposedwires of different composition, and passing this strueture through two oppositely disposed heating zones to fuse together' the twisted ends of the different pairs of wires.

In testimony whereof, l have hereunto subscribed my name this th day of December,1925.

VALTER O. SNELLING.