Shrinkage
A contraction or shrinking of materials into a less compass as they change from a hot to a cold state. Some hard irons shrink or contract 3/8 of an inch in 12 inches, while soft irons of choice grades will sometimes not exceed 1/16. Medium grades of good quality generally shrink about 1/10 of an inch in 12 inches when used in heavy castings; but in light castings the same brand will as a rule show 1/8 of an inch in 12 inches. Bronze shrinks about 3/16 of an inch in 12 inches. General brass work, according to mixture, will shrink from 3/16 to 1/4 of an inch in 12 inches. Copper shrinks 3/16, tin about 1/8, silver 1/8, lead 5/16, zinc and bismuth each 5/16 of an inch in 12 inches. These can only be approximate measures of shrinkage; the exact amount which takes place in any particular casting must necessarily be determined by its general outline and bulk, and to some extent by the temperature of the metal used for casting with, dull metal always favoring the least contraction. Very little if any shrinkage would seem to occur in heavy castings of limited compass, but what in this instance seems to be lack of shrinkage is in all probability to be attributed to distention of the mould under extreme pressure. The metal remaining fluid a longer time in this class of moulds than is ordinarily the case, gives ample opportunity for the pressure to act upon the mould surfaces to severely try them. Bottoms of cylinders and pipes would appear to shrink more than the tops; but it is a mistake to say they do. The smaller diameter at that point is because the material of which the core is composed cannot effectively resist the extra pressure to which the lower portions of all deep moulds are subjected. When 12 feet added to the depth will make a difference in pressure equal to about 37 pounds per square inch, it is plain that unless extraordinary measures are adopted to resist this added pressure, cylinders will always appear to shrink more at the bottom than top. The commonly accepted theory that castings shrink less vertically than in any other direction is undoubtedly wrong also, and similar reasons may be advanced to refute this as in the case above, especially when the upper flask or covering-plate is connected with a system of coring, the bottom surfaces of which extend a considerable distance down. These extra strains, if not resisted absolutely both at the top and bottom, gives an increased length to the casting, which in many instances exceeds the legitimate shrinkage. Gear-wheel rims shrink less when the arms and hubs are cast on than when made as a separate casting, and it may always be expected that the heaviest wheels will shrink the least. The same may be said in reference to other classes of castings, which are composed of cross-ribs and plates internally; these will invariably shrink less than open castings, such as plain frames, etc.