As has been stated elsewhere, there should be twice as much thermit steel poured for a weld as is necessary to fill the space between the joined pieces and to provide for shoulder around the joint. The first of the thermit pour that reaches the inside of the joint expends most of its heat in raising the temperature above redness. It passes up the riser, leaving the interior so hot that the last of the pour settles easily around the half-molten joint and is fluid enough to make a homogeneous casting.

The amount of the thermit powder used in a weld is eighteen times the unoccupied space in the mold after the joint is adjusted and ready to weld. The thermit is estimated in ounces, the space in cubic inches (page 124). This number, eighteen, provides twice as much thermit steel as is needed for the weld, the rest, as already stated, going into the riser. However, P. Redington1 and H. L. Des Anges2 advise that three or even four times the amount of steel is needed to get the best results. This may be due to imperfect preheating.

The Reaction

The reaction is rapid and violent. There is no explosion, but the crucible sends up a shower of sparks much like the kind of fireworks called a " flower-pot." So to prevent this and to conserve the heat, a loose metal top is slipped over the crucible as soon as the fuse is lit.

The workman should use smoked or colored glasses to protect his eyes.

The reaction takes not longer than thirty seconds. The crucible should not be tapped for at least ten seconds thereafter, because the reaction has left an intimate mixture of slag in steel in the crucible, and a little time is allowed for the slag to float to the surface.3 I believe that outside of insufficient perheating, one of the common causes of failure of thermit welds is premature tapping.1 No steel is strong if it is premeated with slag.

1 Foundry, April, 1905.

2 Foundry, August, 1905

3 Ibid., R. Webb, July, 1905.

After Pouring

After pouring, you have an ordinary steel casting, with this exception - that the heat of the joint will be conducted by the body of the part much faster than is good for a steel casting. If you are welding a fractured locomotive frame, and you want to assure yourself that the joint will be as tough as the frame, you had best give the joint several hours' annealing by such means as are at hand.

Annealing is not so necessary in a thermit joint as it is in the oxy-acetylene and other welds. Thermit steel shows a low-carbon content. Rapid cooling will not temper it highly. But no chilled steel is as tough as the annealed product. Tests in practice seem to show that after-heating gives an even-grained and tougher joint.2

The results of after-heating may be attained in a lesser degree by keeping the mold in place until the joint is cooled. Cooling may take several hours with the mold on if the pieces are large.

Nickel Addition

Nickel thermit is an allied substance to thermit proper. It is a mixture of nickel oxid and aluminum, and the reaction sets free the nickel in the metallic state.

3NiO+2Al = 3Ni+Al203.

If the operator wants a higher tensile strength without diminishing his elastic limit, he introduces a can of nickel thermit into his ladle of molten iron, as already described. Or the nickel thermit is fired in a hand-ladle, using a small quantity, and pouring in the remainder of the package gradually as the reaction progresses. The entire contents of the hand-ladle are poured into the big ladle, which should be one-third full of molten iron. The big ladle is then poured full of iron and a can of titanium thermit is poled in to cause a thorough mixing of the iron and nickel.

One per cent, of nickel is sufficient to increase the strength of ordinary iron about one-third. Two per cent, of nickel thermit gives a little more than 1 per cent, metallic nickel.

1 Foundry, Jas. F. Weber, July, 1905.

2 Journal U.S. Artillery, Gustav Reiniger, July-August, 1907.

Metallic nickel is also added to thermit, using 5 ounces nickel to each 100 pounds thermit if you wish to make a 1 per cent, alloy.

Titanium Addition

Titanium thermit is another "alumino-thermic" substance having the reaction

3Ti02+4Al = 2Al203+3Ti.

It is introduced into the ladle in foundry practice for the purpose of purifying the iron. About 1 per cent, is recommended. As its office is to reduce the sulphur and nitrogen, most of it re-appears in the slag. Its effect is to greatly increase the strength, presumably by making the metal close-grained and homogeneous.