The following text may have been generated by Optical Character Recognition, with varying degrees of accuracy. Reader beware!

4 LOCALLY ANNEALING HARVEY!ZED PLATES. (SEE ILLUSTRATIONS. ) The following is the substance of a paper recently read before a meeting of the American Institute of Blectrical Engineers, by Mr. Herman Lemp, Jr.: One of the latest advances in the making of protective armor for battleships, or even forts, has been the in- troduction of what is known as the Harvey process. For those not conversant with the latter, I will briefly state that it consists in taking an ordinary low carbon | a yigen y, hee FIG. 2. FIG. 1. steel plate and introducing an additional percentage of carbon into the surface metal, thus changing the crust to the depth of about an inch, into a steel resembling tool steel. A plate thus treated is lastly water-hard- ened, similarly to an ordinary tool, and by experience has shown to offer, under equal conditions, more re- sistance to the impact of a projectile than any other armor known, and therefore we will take this as an illustration of hard faced armor. The extreme hard- ness of the surface of a Harvey plate, while exceedingly valuable in preventing projectiles from piercing it, has a disadvantage when it is required to be pierced by a drill and tap. j The methods heretofore used to produce these holes were principally two: 1. To protect the surface of the plates in patches or strips to prevent carburation wherever holes -were ex- pected to be drilled. 2. To make accurate drawings and patterns of each plate beforehand, to which all holes are drilled before plate is hardened. . Plan No. 1 was practiced in the United States until it had to be abandoned. In practice it was found that numerous alterations in construction, errors in either draughting-room or mills, made it necessary to pierce holes where no provisions for annealing were made. It also happened quite frequently that the method for prevention of carburation did not always work satis- factorily, and while white paint indicated well enough where soft metal was expected, very often hard metal was actually found. The second method, which is used in England, has the advantage over the first, that there is no difficulty in carrying it out, provided there are no alterations made, and no errors committed. The present method of con- struction seems to be, to make, beforehand, a complete model of the vessel to be built; then make accurate drawings and full-size patterns of every plate, giving the exact location of every hole. These are then bored, tapped and countersunk to a depth of approximately % of aninch and of a much larger diameter, filled with clay, and then the plate as a whole is heated and hard- ened with water. (See Fig. 1.) This method works apparently all right, although in spite of drawings and models, errors are made. There is no doubt, however, that it is a slow and very costly method of working, and would hardly ever be resorted to in the United States. If, by sending a current of large volume through any spot thus to be treated, the spot is brought to a temper ature of approximately 1,000° F., there can be no doubt that the temper has been withdrawn. Experi- ments carried out to that effect at once showed, how- ever, that upon taking off the heating current the heat was so rapidly con- ducted away by the sur- rounding metal masses as to cause the heated spot to become chilled just as effectually as if it had been plunged into cold water. No method of outside protec- tion of the heated spot would prevent this, and the gradual cooling of the spot had to be attained by differ- ent means; namely, a gradual and slow withdrawing of CRACKED HERE Fig. 8. | aa FIG. THE MARINE RECORD. the heating current. The method of introducing the annealing currents is best shown in Fig. 2. - C C are two copper contacts cooled by water. circulat- ing inside. The current enters the plate by one end, and leaves it by the other. Right under the contact the metal comes to a bright cherry heat, (shown in black) while the portion intervening and partly sur- rounding the contacts acquires a temperature of justa visible rd. Mine H H indicates where the influence of the Harvey treatment stops. The shaded portion in Figs. 2 and 3 shows the zone softened and ready to be . machined, while the dotted line shows how far the heat radiation would cause the metal to turn blue. When cooled, the annealed portion shows a chocolate color, while the place where the contacts have been resting is scaled and hard, and can not be touched by a tool to a depth of about 4% of an inch. These places can again be annealed later on, if required- The apparatus necessary to carry out this process consists of the generator, the annealer proper (trans- former) and the regulating apparatus. The annealer proper is a transformer similar to the well-known welding transformers. As the apparatus is to be oper- ated outdoors, on board of vessels dur- ing construction, it is important that the same may be handled with immunity from electric shocks, even when operated in mist and rain. 9 |. To meet these con- ditions a copper-clad type of transformer is used, one in which the secondary is com- posed of two copper castings, each hay- ing a rectangular groove, which two halves, when bolted together, form a closed’ rectangular frame in which the primary is held en- closed. The hollow space intervening between primary and secondary is, moreover, filled with a heavy oil, which acts both as the in- sulator and conduc- tor of heat from primary to second- ary. The secondary, by completely sur- rounding the .- pri- mary, affords an ex- cellent mechanical protection, and pre- vents electric as well magnetic leakage. The primary is a copper ribbon in- sulated with asbestos, and the ratio of conversion is 100to1. The transformer has two trunnions fastened to its sides, in a line a little above the center of gravity, which trunnions swing in bearings, part of a yoke which straddles the whole. The yoke in its turn has a hook which may be secured to the latter at any place of the arch, thus allowing the transformer to be sus- pended, like a compass in gimbals, in any position desired. It goes without saying that the copper castings, which compose the secondary, are cut through at one place in the circuit. On either side of the cut two short plat- forms form the base for copper contacts of various shapes and sizes, by means of which the currents are made to enter and leave the plate to be annealed. These copper contacts are of forged copper, hollowed out to receive circulation for cooling purposes, and terminate in narrow tips rounded at the end. The weight of the whole annealer, being approxi- mately 1,000 pounds, is sufficient to give proper contact 7 i cy, \ HI he st 3. Fig. 4. pressure for all work on a horizontal plate. Whe clined surfaces, vertical or otherwise, are to be wor! upon, the transformer is suspended so that its weig shall not interfere with the contact pressure, which obtained for work by bracing the contacts directly w wooden wedges against any object near by. Ont outside of a hull it is proposed to use a pair of electr the iron hull, and form a support for the annealer. The remarkable thing is the great amount of curren that is being carried by : the copper contacts into the plate. The contact surface is seldom more than about % inch square, and yet 10,000 amperes are made to flow through it continually. This is equiv- alent to 40,000 amperes per square inch, a density which is only possible on account of the thorough cooling by the water cir- culation. The regulating appara- tus is, in most cases, sim- ply a rheostat in series with the field of the gener- ator. When more an- nealers than one are run | E simultaneously from one generator, as stated before a reactive coil is interposed between the two (Fig. 4) The annealing operation is carried out as follows The transformer is placed in position, the contacts touching the plate either side of the place marked to b annealed, and the primary current brought up by ~ means of a rheostat to from 75 amperes to 90 amperes for about two minutes, according to the size of spot to be annealed, which will bring the metal to a dull red heat, a temperature at which a pine stick catches fire when held in contact with the plate. If no reactive coilis used, the current is now diminished by turning the rheostat one point every minute. If the reactive coil is used, the core is now raised by winch, the coil put in circuit by opening a short-circuiting switch, and then is allowed to descend on its own account. The operation generally takes seven minutes, all told. Ac- companying Figs. 5, 6 and 7 show the proper apparatus - in operation on the Massachusetts and Oregon. No difficulty has been experienced from the beginning. The annealing of individual spots was, however, only the stepping stone to a more important work of a simi- lar nature, work which was about to be given up, owing to what were considered insurmountable difficulties. In the construction of a modern man-of-war, there are many armor plates which act as shields tothe guns, and have to be perforated to allow the gun muzzle to pass through to be either raised or lowered. Some of these shields are circular or oval, with narrow edges around the ports, in the case of Harveyized plate. ‘To perforate these _ shields after carburation, and be- fore being water-hardened, was the only possible way, since previous experiments had shown that prevention of carburation could not be relied on, and no pro- cess was known to anneal the plate locally after hardening. The harden- ing of a plate once perfor- ated, showed itself to be, however, almost impos- sible; in fact, a matter of chance. In most cases the plate cracked in two, as shown in Fig. 8, or the whole plate became dis- torted in such a way that it could not be used. As individual spots for holes could be annealed, there could be no doubt that a series of annealed spots could be likewise obtained by the electric process, following a line along which a cutting tool was expected to be run. The first attempt, there- fore, and made in England, was to anneal a number of spots in proximity to each other in such a way Fig. 6.