(Condensed from Van Nostrand's Magazine) BY MAKTIN BALCKE The following remarks do not refer to riveting for the purpose of merely uniting two parts of machinery or two sheets of iron, but they apply to rivetings which require a higher degree of strength and solidity, as, for instance, for boilers and working parts of machines It is a general rule for all constructions, especially for those in iron, to distribute the strain which has to be withstood by a certain part of a machine, as evenly as possible over the solid mass of the said part This rule is also very important in the use and arrangement of rivets The simplest and safest way to carry out this rule is to calculate directly the areas of the working sections, and to see that the strain which acts on any part of a section, does not exceed certain limits generally conceded to the respective materials This is the way also to avoid the use of empirical formulae, the most important coefficients of which are always dictated by the personal opinions and notions of their authors The force necessary to terr a wroughtiron bar of a certain section, is so nearly equal to that required for cutting or shearing the bar, that both may be considered as equal in calculations, for practical purposes The limit of elasticity of soft wrought iron, as generally used for rivets, is at a pressure of about 18,000 lbs on the square inch With boilers the strain of tension per square inch of section of the material, ought not to reach 9,000 lbs; because continued heating and long use weaken the material considerably In the construction of stationary boilers, one square inch of section, taken through the riveting, ought generally not to be strained above 12,000 lbs But if a riveted part of a machine has to sustain a strain acting alternately in two different and opposite directions, this strain should never exceed 8,000 lbs per square inch of section If a quite uniform distribution of the strain over all the sections cannot practically be obtained, at least the tension of the sections which are exposed to the highest strains ought to be kept within the above mentioned limits The shape of the head of a rivet is dependent on the kind of strain to which the rivet is subjected This strain can have the tendency of tearing or of shearing the rivet, or of both simultaneously If a rivet has to withstand a tearing strain, the hight of its head must be such that the cylindrical surface which would make its appearance when the head of the rivet would be stripped off, is equal to the area of a crosssection through the rivetthat is, the hight of the head has to be one half of the radius, or one fourth of the diameter of the rivet Practical experiments on the strength of rivets have come to the same result, and have besides shown very distinctly that the rivet holes should never have sharp edges, and that the head of a rivet ought to be connected with the shaft by a conical part Whenever this part is omitted, and when, consequently, the rivets have sharp corners below their heads and the rivet holes sharp edges, the rivets break close to the head, when subjected to a strain of tension and when the heads are strong enough not to be stripped off When, on the contrary, the rivets have a conical connecting part between their heads and shafts, they extend considerably before they break, and the rupture finally occurs in the middle of the shafts All experiments have given this result without exception Rivets subjected to a shearing strain only, would theoretically not require any head at all But it is good also in this case to make the heads of the rivets as high as above determined, because generally a close contact af the riveted parts is desirable, and because the rivets, being set in redhot, have to resist the strain of tension produced by their contraction in cooling If the heads of rivets have to be countersunk, their best shape is that of a truncated cone, the angle at the point of which cone would be of 75 The sectional area of the shaft of a rivet, expressed in square inches, is found by dividing the actual and total strain on the rivet, by the strain practically admissible on the square inch of the respective material We will now examine the riveting of simple round boilers The shearing strain in pounds on every rivet in the length rows is equal to one half the diameter of the boiler in inches, multiplied by the rivet distance in inches, multiplied by the steam pressure in pounds less 15 pounds atmospheric pressure The strain upon every rivet distance in rows round the boiler expressed in pounds is equal to one fourth the diameter of the boiler in inches, multiplied by the distance bet ween any two rivets in the same row round the boiler taken in inches, multiplied by the steam pressure per square inch in pounds, less 15 pounds atmospheric pressure Now, to obtain an even distribution of the total pressure in the boiler over all its sections, the sectional area of a rivet has to be equal to the sectional area of the plate between two rivet holes, and equal also to the double area of a section through the plate, from a rivet hole to the edge That is, the sectional area of the rivet in square inches must equal the distance between any two rivets in a row round the boiler in inches minus the diameter of the rivet, multiplied by the thickness of the boiler plate in inches ; or, conversely, the distance in inches between any two rivets in a row round the boiler must equal the sectional area of the rivet in square inches, divided by the thickness of the boiler plate in inches, plus the diameter of the rivet From this we conclude that the rivet distance is dependent on the diameter of the rivets, and, reciprocally, the diameter on the distance To determine these, it is necessary to take into consideration the possibility of making and keeping the boiler tight, which possibility depends principally on the relation between the thickness of the plate and the rivet distance Let us consider a special case to explain this more fully We suppose a simple cylindrical boiler to have a diameter 42 inches ; the thickness of the plate, 03 inches ; the excess of the steam pressure over the atmospheric pressure, 42 lbs Under these conditions the strain of tension per square inch of plate section, taken parallel to the axis of the boiler, is j 21 x 42 =2,940 lbs In taking the areas of the rivet sections equal to those of the plate sections contained between two rivet holes, according to the above rule, and in calculating the following items for three different rivet diameters, for the sake of comparison, we find The rivet diameter being f in, f in, $ in Area of rivet section (sq in), 0307, 0442, 0601 Distance between rivets (inches), 1 648, 222, 2878 Shearing strain on a rivet (lbs), 1,453, 1,959, 2,538 Strain per square inch on a section through the plate, or through the rivets in the length rows of the boiler (lbs), 4,730, 4,430, 4,220 [The shearing strain on rivets and the strain per square inch of section in the rivet rows round the boiler, are one half of those in the length rows] The strength of the riveting compared to the strength of the simple plate is 062, 066, 0'70, for the three different rivet diameters The advantages and disadvantages of the one or other of the chosen rivet diameters are clearly shownby these figures The fin rivets produce a very small comparative strength of the riveting (062) The Jin riveting has a great comparative strength ; but the distance between the rivets (2878 in) is too large in proportion to the thickness of the plate, to allow of a good and safe tightening of the joints The fin rivets, not showing either of the two mentioned disadvantages in a considerable degree, are evdently the best in this special case

This article was originally published with the title "On Rivets and Riveting" in Scientific American 21, 8, 115 (August 1869)

doi:10.1038/scientificamerican08211869-115