By the l890s, ordinary steel was replaced by nickel steel. (Krooth 89).
Cannon Manufacture
Hollow casting. In the early-seventeenth century, muzzle-loading iron cannon were cast as single hollow blocks. Making the mold was tricky. In essence, there were two clay molds, a hollow one for the exterior and a solid one (core) for the interior. (Hall, 11ff; Fisher).
The hollow mold was built up over a pattern made of wood, rope, clay and a friable material like horse dung; the pattern defined the desired shape of the cannon interior. The pattern was coated with a release material, such as an ash-fat mixture or a wax, so the actual mold material wouldn't stick. This mold material was also clay-based, and might be reinforced with rope or animal hair. The mold was reinforced with metal straps, the pattern was carefully removed from its interior, and the mold was baked. The core mold of course was simpler to make.
The complete mold was lowered into a pit, muzzle up. Note that the interior (core) mold had to be held centered inside the larger mold by a metal spacer ( cruzeta; chaplets), which would become part of the gun. In general, this didn't work out perfectly, the core would shift so the bore wouldn't be quite straight. (WeirML, 132).
The pit was filled with earth so as to hold the mold upright, a "feeding head" (riser) was attached, and the molten metal was poured into it. The latter had to have the right fluidity to properly fill the mold. Once the metal had cooled and solidified, the mold was broken so the casting could be removed. That meant that no two cannon could be identical.
The cannon was then finished off; the most important finishing steps were cutting off the riser, and reaming out the cast bore so that it had a smoother surface. Diego Prado y Tovar (1603) described a machine, possibly driven by animal power, for accomplishing this, However, the drilling was vertical, with the cannon suspended and slowly lowered over the drill. Note that the machine was merely finishing a hollow casting. Indeed, hollow casting is plainly described in Mieth, A New Description of Artillery(Frankfurt 1684); chapter V discusses the cruzeta (Rainer Prem transl.). Bores were cast to the diameter of the shot and drilled out to the added diameter of the windage. (Hoskins 43).
Clay molds are criticized in 1634: TBW, Chapter 38: "Clay had a very low porosity, which meant that air bubbles in the molten iron were often unable to escape when the guns were cast and, instead, formed dangerous cavities and weak points in the finished guns." The gun barrels of the USE ironclads used in the Baltic War were fabricated by sandcasting. "Sand was far more porous, which made for much stronger, tougher artillery pieces." Historically, sand molds were introduced in Britain about 1750. (Lavery 84).
Another problem was intrinsic to the vertical casting method; since the bottom (breech) was under greater pressure than the top, and also better insulated, it would have been the last to solidify, and therefore tin would have migrated downward. The muzzles were thus only 3–5 % tin, resulting in brittleness, which was compensated for by flaring the muzzle.(Guilmartin).
There were other modest improvements over the eighteenth century. In Britain, these included providing full-size drawings to the gun founders (1716) and using copper rather than wood cores.
We may deduce the improvement in tolerances by examining the weight variation of the pieces. "In 1665, guns from a single batch of 9ft demi-cannon varied from 44 to 62cwt, those of 8.5 feet from 43 to 47, and culverins of 10ft varied from 40 to 46 cwt." (83). In contrast, the 32-pounders surveyed in 1803 -6 were 55–57 cwt. (84).
Solid Casting. Over the period 1715 -45, Johann Maritz developed a new fabrication method. The cannon was cast solid, breech down, and then the bore was drilled out horizontally. The casting itself was much as in prior times, except that the core mold was no longer required. Boring itself, using an animal- or water-powered machine, took several days. (Kimpton). One curious aspect of the process is that it was the cannon that was rotated, the bit remaining stationary. (Alder 42). Solid casting was adopted in Britain in 1776 (Lavery 84).
Hot Blast. In the 1830s, American gunfounders attempted to cast iron by the more economical "hot blast" method, resulting in a disastrous loss of strength. At West Point foundry, 68.5 % of those cast by cold blast (1826–1834) were deemed "first class," compared to 4.02 % of those produced (1835 -39) by hot blast. (Hazlett 36, 42).
Rodman Guns. These were hollow cast, with a trick; the core was itself hollow, in fact, two concentric tubes, and was cooled with pumped water while the molten iron was poured in around it. The metal would thus cool inside out, pre-stressing it in a desirable way. (Wikipedia/Rodman_Gun).
Built-Up Construction. The 1855 Griffen "Ordnance Rifle," a 10-pounder cannon with a 3 inch rifled bore, was built up by welding wrought iron bands together around a mandrel, boring, and rifling. (Kinard 192), or by building up a mandrel with welded iron rods and then winding several bars in spiral fashion about it (Hazlett 121). Note its similarities to the ancient bombard, in that it was "built up" from wrought iron! However, it is important to note that instead of forging the iron with a hammer-as was done with the 1844 "Peacemaker," which burst and killed two cabinet members-Griffen forged his iron rods in a rolling mill.
There was also the British Armstrong gun. This went through several permutations. In one, wrought iron bars were twisted into spirals and welded on their edges to form the barrel. (Tennent 106). In some cases the twisted coils were themselves shrunk onto an inner tube of mild steel. (Morgan xvi).
Wrought iron's advantage was that it was four times stronger than cast iron, and thus able to help resist the higher internal stresses (the result of the reduced windage) of a rifled gun. Saving manufacturing cost and time, Parrott shrunk a wrought iron reinforcing hoop onto the breech of a rifled barrel cast in the usual way. However, "large Parrott rifles had the worst record of any Union cannon for premature bursting. Of 110 large caliber Union cannon that cracked or burst in action during the war, 83 were Parrotts. " After the first 1864 assault on Wilmington, Admiral Porter declared that the guns were "calculated to kill more of our men than those of the enemy." (Bell 8).
Around the end of the nineteenth century, the British and Japanese made use of wire wound construction. The "A" tube was wrapped multiple times with a high tensile strength wire and then the "B" tube was shrunk over this. (DiGiulian). The ten-inch guns of the new time line's USE Constitutionare "wire-wound" ( 1634: TBWChap. 38), presumably over a cast tube, but I don't know if a "B" tube was added.
In the early-twentieth century, heavy naval guns were built-up in hoop-over-tube fashion. The inner tube was placed breech end down in a cold pit, supported by a short mandrel. Heated hoops were placed one by one over the tube and cooled with a water spray, shrinking them onto the tube. (NAVORD. 136).
Spun Cast Monoblocs. In the 1920s, this was superseded by monobloc construction, made possible by the development of centrifugal spun casting. Despite the name, it typically involved concentric assembly of two or three tubes. Autofrettage was used to permanently deform the tubes in a desirable way. In autofrettage, the tube was pressurized hydraulically, just enough so that the outer limit was at its elastic limit, and then slowly relaxed. This increases the diameter of the bore and there is a permanent strain in the tube which varies from the inside diameter to the outside one.