INTERNAL STRAINS IN TREATED TUBE.
354
Diameters were established along■ the bore in two planes, and at intervals of J" each. The tube was measured on these diameters, then out apart at the middle of its length and remeasured. For identification the halves were marked A-B and A,-!», and the measured diameters successively numbered from 1 to 27, beginning at the inner ends of the two sections. After cutting apart, a slice long was taken from the inner ends of each piece, which were designated as A-B 1 and Aj-B! 1. Using section A-B, additional slices were detached, remeasuring a number of the diameters in the remaining part of the section after groups of four slices were detached, and thus proceeding until seventeen slices in all were removed. The detached slices were remeasured at the bore, surface diameters established on the ends and rings detached, one at the bore and one from the outside. Rings from four of the slices were taken out with the metal in the state it came from the forging. Ten slices were annealed at temperatures ranging■ from 207° F. to 1,100° F. prior to detaching the individual rings, and three slices are reserved for experiments on local heating at the bore. The remainder of section A-B, comprising bore diameters 18 to 27, was used for determining the longitudinal strains released upon turning off the exterior metal, leaving a thin cylinder of metal next the bore. Longitudinal gauged lengths were established in the bore on four lines.
From observations made upon cutting off slices from the ends of the tube it appears that the metal tends to expand and assume a bell-shaped form at' the ends of the section, and when slices are successively detached the new ends thus formed each in turn display this action. At a distance of i" from the end the mean expansion in diameter was ".00224, which diminished to ".00092 at a distance of 1" from the end. On more remote diameters the effect was lost or of insufficient magnitude to present decisive indications. The stresses corresponding to the strains above stated are 13,200 and 5,400 pounds per square inch, respectively, which were strains of compression.
A comparison of the bore diameters of the detached slices with the same diameters when in the tube, at a distance from its ends, beyond the bell-shaped zone, makes it appear that the detached slices are slightly larger than when forming an integral part of the tube. The conditions under which the measurements were taken precluded the accuracy sometimes attained in experiments of this class. From all of the data developed it appears, however, that the slices were at least ".0005 larger at the bore than when forming a part of the tube, which on this diameter corresponds to a stress of 3,000 pounds per square inch. That is, the metal at the surface of the tube is apparently under a compressive stress 3,000 pounds per square inch higher than when in the detached slice. The detached rings showed the release of compressive stresses at the bore, from the state of the metal in the slice, amounting to 25,625 pounds per square inch, the means of four rings, while rings from the exterior of the same slices gave the mean tensile stress 6,725 pounds per square inch, the mean strains released being in rings 1, ".00445 and in rings 2, ".001S2. The rings from the annealed slices gave the following results: •