[10,000 amp. per sq. in.] of sectional area of the object to be welded in the immediate neighbourhood of the joint. The same process is recommended for the heating of wires or rods, the wire being drawn over the two ends of an interrupted electric circuit in such a manner that it makes contact, and so forms a resistance between them. Since single wires may be heated in this fashion, there is no doubt that whole bundles or packets might be similarly raised to a temperature suitable for rolling, and the reheating furnace may possibly be even replaced under certain circumstances by an electrical heating plant.

Local Softening of Quenched Steel.-The melting of castiron for foundry work by the resistance heating method has not advanced to the practical stage, even though isolated reports of a not unfavourable character have been published concerning the Taussig smelting furnace to which allusion has been made above. One of the latest applications of this method of heating is to be found in the local softening of armour plates which have been superficially hardened by Harveyising; this may be required if extra rivet holes should be required at the last moment to secure the plates to the hull of the ship, or if the plates have to be subjected to some special treatment. In order to effect this two parallel copper contact pieces 5 sq. cm. [sq. in.] in cross-section and 25 mm. [1 in.] apart are caused to lead a sufficient current to the part that is to be heated. A dull red heat will shortly be observed in the iron between the contacts. From this moment the current strength is gradually reduced, so that after about ten minutes the heated part will have been cooled below the critical hardening temperature, after which no hardening is possible on the removal of the contact.

[A detailed account* of a plant supplied by the Thomson Welding Company for the local annealing of the Harveyised armour plates of the "Oregon" (U.S. Navy) has recently been published. A 55 H.P. engine is used to drive a 40 kilowatt alternating dynamo, wound for an outturn of 135 amperes × 300 volts, when run at 1000 revolutions per minute. The exciter (a D-type, shunt-wound generator of 110 volts at 2000 revolutions) is drawn by a pulley on the armature shaft, and is connected with the field magnet coils of the alternator through a regulating rheostat with German silver coils. The main alternator current is led to the primary coil of a "shell" transformer (i.e., one with the coils enclosed within a laminated iron shell); the secondary coil consists of a single U-shaped turn of cast copper, to the ends of which copper contact-pieces may be attached. These are of various sizes and shapes, according to the work to be done, and they are water-cooled (within) to prevent the heat from the annealing operation penetrating to the coils of the transformer. The transformer is so suspended

* Iron Age, August 29, 1895.

on trunnions that it may be turned at any angle, and thus brought to bear (if need be) upon any part of a plate that is already in position. In the actual process of annealing, the rheostat is so adjusted that the alternator current shall at first be of minimum strength, affording a current of about 3500 amperes and 4 volts from the transformer. The proper contact-pieces for the work in hand are attached to the latter, the distance between the pieces being about 1 in., if the hole that is ultimately to be drilled be required in. in diameter. The contact-pieces are then brought down upon the plate so that the spot to be annealed lies centrally between them and thus forms a part of the secondary circuit of the transformer. A slight humming sound is heard, and the steel begins to be heated; the resistance of the rheostat must then be gradually reduced so that a (maximum) current of 6000 amperes may flow through the secondary circuit, including the portion of the plate between the contact-pieces. The steel becomes visibly red hot locally, and bulges somewhat owing to expansion caused by the irregularity of the heating. After about three minutes from starting it is sufficiently hot between the poles to char or ignite a pine stick held in contact with it. The current is then gradually reduced again to the minimum (3500 amps.), so that at least ten or twelve minutes are occupied in reducing the temperature below the point (a dull red heat) at which the metal could harden by chilling from contact with the mass of cool steel around. The current may then be broken, and the metal cooled more rapidly. The annealed portion will be elliptical in shape, the major axis of the ellipse being about 4 in., and the minor axis 2 in., under the above-named conditions.-TRANSLATOR.]

The Benardos Electric Welding Process. Another method of heating objects for fusion or soldering consists in bringing them into contact with the poles of an electric arc. Metals were melted in this manner by Siemens more than sixteen years ago, as indicated on p. 131. Benardos has ingeniously adapted this idea to the soldering, perforating, riveting, and repairing of metals. The work forms the negative pole, and a carbon rod held in some form of movable handle is used as the positive pole (Fig. 183). While an arc plays between the carbon and the part of the metal to be heated, the latter may be locally raised to its melting point, so that the most varied soldering and melting work may thus be done. There are, however, certain difficulties which militate against the general use of the process. In the first place, the temperature of the arc is far too high for most metallurgical purposes, and it is scarcely capable of regulation. Again the transfer of

* German Patent 38,011.

[The work is now made the positive pole, see next paragraph. — TRANSLATOR.]

carbon from the carbon electrode to the molten iron of the object being heated is unavoidable, for the formation of an electric arc is of the nature of electrolysis with a gaseous electrolyte. The solidified metal of the part that has been soldered is, therefore, nearly always harder and more brittle than the rest of the material. Hence it is possible that the process will have to be restricted to the repairing of cast iron objects.

+

dos' method of electrical heating.

[The Benardos process is in constant use at the Halesowen Works of Messrs. Lloyd & Lloyd, to whom the translator is indebted for the following information, as well as for the opportunity of personally inspecting the work- Fig. 183.-Benaring of the installation. Three 150 - volt dynamos are employed to charge 1440 Benardos accumulators arranged in 24 groups of 60 each. These cells are composed of lead plates so perforated as to expose a large surface of the metal, and are made without "paste"; they thus have to be "formed" in the same manner as the original Planté battery. This construction of cell is obviously necessary, on account of the intermittent nature of the welding operations, and of the sudden demand for large volumes of current to be applied for a short time. Should the renewal of any lead plates become necessary in the cells, the fresh electrodes may be introduced directly, and the process of "forming" is then accomplished without any special treatment. All the leads from the dynamos, batteries, and welders are brought into the switch room, and are so arranged that they may be joined up in any required manner; the instruments necessary to the measurement of any, or every, circuit being placed in the same room.

In the smithy, the metal to be welded may be connected up to the conductor from the switch room, but more usually the anvil upon which the work rests is so connected. The other electrode is L-shaped, and consists commonly of a stout carbon rod, about an inch in diameter, which forms the short limb of the L, and is gripped at the end of a long holder held in the hand, and joined up to the opposite conductor from the batteries. It is especially to be noted that the carbon is connected to the negative, and the work to the positive battery lead, and not vice-versa as described in the accounts of the original Benardos process. Hence there is no longer the danger of a partial carburisation of the metal by the anode carbon, to which allusion has been made in the text. The process may, of course, be applied to the welding of straight bars; but it is most useful and economical when employed for such purposes as the welding of flanges on to pipes that are required to stand high pressures, or for the

burning out of surface flaws in steel castings, or the like, and making good the cavities thus produced by melting metal of similar character into them. There are obviously hundreds of other applications, and among them may be noted the burning of holes in tubes, &c., or the cutting of metal in any desired

manner.

*

In use, the object to be welded is placed on the anvil and connected to the positive lead; the welder, holding a coloured glass screen in the one hand to protect his eyes from the glare of the arc, and grasping the carbon holder in the other, touches the carbon on the work and then immediately and very rapidly withdraws it, so that an arc of suitable length is formed. The length of the arc used depends mainly upon the thickness of the metal under treatment. If very short the heat is not only too concentrated, but too intense, and a thin plate might be fused through almost instantaneously; and, even if a good joint were made with such an arc, there would be a possibility that the great (and purely local) contraction of the metal on cooling would then cause it to crack. There is no difficulty in obtaining an arc 6 ins. long, or more, with the aid of a powerful current, and the heat would then be diffused over a larger area of metal, but the cost would be needlessly high. For the work usually undertaken, the arc is preferably from 2 to 23 ins. in length. Within a few seconds from the start the portion of the metal upon which the arc is playing is brought into a plastic semi-fused condition that is most suitable for the welding operation. If an ordinary bar were to be welded the metal would be raised to the required temperature, and hammered as usual. In other cases (e.g., in uniting a flange to a tube) the parts to be joined may be cut, so that when placed in position there is a V-shaped groove between them, a small block of metal (toin. cube) is then placed at a convenient point in the groove, the arc is started and caused to play upon this portion, so that the area surrounding the metal block is raised to a welding heat, and the block itself is just melted into a plastic condition, the arc is then broken and the joint completed with a former. Another block is then placed [The greatest caution must be observed in working with these large electric arcs. They should never be looked at, even from a considerable distance, without using a screen consisting of several thicknesses of deep ruby and orange glass. The writer has on several occasions, when working with an electric furnace, suffered slightly from the effects even of an inadvertent glance at the arc. In more serious cases that have come within his knowledge, the most intense pain and irritation of the eye were felt, combined with a copious flow of tears. These effects are not commonly felt at the time, but only after the lapse of several hours; the pain may then, however, be almost unendurable for hours. If, for any reason, it be neces sary to approach the arc closely for any length of time, the screen should be of sufficient size to protect the whole face, as the action of the arc is comparable with that of direct sunlight, causing sun-burn and peeling of the skin.-TRANSLATOR.]

*

next to this, and the arc re-formed, and the joint made in the same way. These operations are alternated, so that the weld is extended about 1 to 2 in. each time until the whole is complete. In boring or cutting a pipe, or other object, the arc is caused to play sharply upon one point until a complete perforation is made, and this is extended in either direction by altering the position of the arc. In filling up cavities the metal is built up little by little, after the manner explained above in the case of the long weld.

At Messrs. Lloyd's works the current pressure used averages 110 to 120 volts, while the volume is varied by the introduction of resistances, according to the work in hand. Speaking generally, the current employed for welding metal in. thick may conveniently be about 200 amperes, for metal gin. thick about 250 amperes, and for in. thick 300 amperes. For steel castings, it might range from 200 to 500 amperes. It is evident that the volume of current may vary within wide limits, and somewhat smaller currents may be used for a longer time, but generally with less economy. Thus, in the welding of to in. metal, the current would range from 23,000 to 35,000 watts, which is equivalent to an absorption of 31 to 46 electrical H. P.-TRANSLATOR.] The Slavianoff Process.-The process of Slavianoff, which depends upon the same principle as that of Benardos, can also be applied to the repair of cast goods for like reasons.

This inventor has substituted a metal electrode for the carbon pole employed under the other system. It is evident that such an electrode must fuse, and will then be liable to run into the faulty places of the latter; it should, therefore, be composed of the same material as the metal to be repaired. It is true that there is no danger in this case that an excess of carbon shall be taken up by the iron surfaces to be united, but the operation itself requires more care and skill than that of Benardos on the part of the workman, for if the metal melt off too rapidly from the positive electrode, the arc will be subject to constant fluctuation. On the one hand, there is always danger of short circuits being formed, but, on the other hand, the breaking of the arc has equally to be avoided. [A current of about 4,800 to 5,200 amps. per sq. in. of sectional area of bar is used at a pressure of 50 to 70 volts. -TRANSLATOR.]

The Lagrange and Hoho System.-Lagrange and Hoho also cause the object to form the negative pole of an electric arc, as in the processes last described; but this arc is produced (and the whole object is immersed) in an atmosphere of hydrogen produced by the electrolysis of an aqueous solution. A leaden vessel, or a lead-lined wood or stoneware tank, or else leaden plates of the greatest possible area, placed in a suitable receiver, serve as anode, whilst the objects to be heated form the cathode and present a relatively small surface (Figs. 184 and 185). The