Strains in Glass Produced by Applied Color Labels *

Abstract

The fact that "applied color lettering" can weaken a bottle has been recognized from the beginning of the practice, and the fact that it normally does not but may even increase the strength slightly is also well known. Weakness arising fr?m the "color" causes fractures with origins on the outer'surface of the bottle wall and within the colored area. They are thus readily distinguished from those fractures which arise from weaknesses of the inside surface of the wall.The author investigated this subject with commercial and experimental colors on pressure bottles of various manufacturers. A special thermal-shock test was used, which tests the strength of the barrel of the bottle without producing large strains on the base. By correlation of such shock tests with the strain patterns observed in ring sections examined under the polarizing microscope and with thermal-expansion data, it has been possible to set up a suitable procedure for insuring the proper "fit" of a ceramic color to a glass. With minor exceptions, the problem is considered to be similar to those of applying enamels to metals, glazes to pottery, and in sealing metals to glass, which have already been discussed in the literature.By an analysis of the respective thermal-expansion curves for the glass and the ceramic colors, the author suggests the reason for the experimentally found fact that the ceramic color must be of slightly loww coefficient of expansion than the glass in order to minimize the strain between the color and the glass when the color has been fired on to the glass and the two cooled down to room temperature. The suggested reason is that the fluxes used in the manufacture of ceramic colors are of necessity softer than the glass and, therefore, their annealing range is below that of the glass. From the shapes of the respective total expansion curves, it is shown that, when the color passes through its lower critical annealing temperature, its curve falls below that of the glass. If the coefficient of expansion for the color is less than that of the glass, the lower rate of contraction of the color balances out this initial high contraction so that at room temperature no residual strain is left.A brief discussion of the interface between the glass and the ceramic color is also included, with particular reference to the case of a flux containing lithia. The smalldiameter, highly mobile lithium ion migrates into the glass, creating an interface tension cord, visible under the polarizing microscope. Present information shows that this is permissible, provided the other factors mentioned have been properly adjusted.