Stability of cylindrical shells with unreinforced holes

Abstract

Listen

Translate article

and 8 had no holes. Shells with an internal surface of radius 100 mm were made from sheets of tin plate (E = 9.2 �9 101~ N/m2; an = 2.14" 108 N/m2; a t = 2.6.108 N/m2; crr = 3.16.108 N/m 2) of thickness 5 = 0.54 mm. Flat rectangular b~anks of 320  635 mm, the width of which corresponded to the shell length l = 320 mm, were prepared for the cylinders and the holes were cut. An overlap of 5 mm was allowed for the longitudinal seam, which was made along a generatrix equidistant from the holes. The flexible blank was shaped on a drum to cylindrical form. The longitudinal seam was soldered on a special mandrel with a clamp, the excess solder was removed, and the shell was held with the clamp. The edge of the shell 4 (Fig. 1) was fixed to the base 1, which had two annu= lar grooves. Since the inner diameter of the base and the clamp was the same, self-centering of the shell and the base in assembly was possible. The shell was rolled past the annular grooves onto the base and then the spring circular clips 3 were put in place and secured by the collar 2. At the center of the base a conical depression was drilled, to position the spherical bearing through which the shelI is axially loaded. This method of fixing the shell edges and axially loading the shell ensures a uniform distribution of axial forces over the shell perimeter, as verified by tensometry. The shells were divided into six groups. The zero group consisted of eight shells without holes. Shells in groups t-V had four holes: square with unrounded corners, square with rounded corners, and round. A single group consisted of shells of the same linear dimension but holes of different shapes. Each i-th shell in the various groups had holes of the same shape; for example, all the third shells had round holes. All the shells were prepared in the same conditions on the same equipment. The initial flexure was not measured.