Pittsburgh Paper - Soft Steel for Boiler-Plates

The technical papers of the last few years give numerous in stances of serious failures by cracking or rupture of soft steel boiler plates, marly of which have satisfactorily passed the rigid inspection and tests required by both Lloyds' Register and the Board of Trade in England. I speak of the inspecting department of Lloyds' and the Board of Trade, for we do not begin to have as systematic and intelligent work regularly done in this country as is done by these institutions in England. Tension-tests of this steel have given from 65,000 to 65,000 pounds per square inch tensile strength, with from 20 to 30 per cent. elongation in 8 inches. Chemical analyses have shown the steel to contain from 0.12 to 0.15 per cent. carbon, from 0.25 to 0.40 per cent. manganese, and from 0.03 to 0.05 per cent. phosphorus. All of these tests are considered normal, and, in the words of one of our members, " boiler-plates with these qualities ought not to fail." There are, however, several causes of failure in soft steel plates which, although the writer does not presume to ascribe to then], or either of them, any particular instance, are still dangerous, and ought to be guarded against much more carefully than in usual present practice. These causes are: 1. The hardening property of all steel. 2. Insufficient work from the ingot to the finished plates. 3. Temporary fiber in certain steels. 4. Internal strains caused by unequal work upon the material. Hardening.-The writer, in a paper read before the Institute in October, 1883," made the statement that soft .steel, no matter how low in carbon, would harden to a certain extent upon being heated red-hot and plunged into water, and that it hardened more when plunged into brine and less when quenched in oil, and gave the results of one experiment, merely as an illustration of many which had all confirmed the statement made. The illustration was a heat of open-hearth steel made by the . writer, of 0.15 per cent. carbon and 0.29 per cent. of manganese, which gave (he following results upon test-pieces from the same 1/4-thick plate.