Monday, July 28, 2014

Moar Technical Appendix: Guns, An Early IT Revolution, and Gyproc

I don't know why I do this to myself. It's about as much fun as ripping out drywall.

Actually, I lie. But only as an excuse to get that video in.

So, Reader Alex writes, etc.:

Anyway, I doubt that the French group's APCR design for the Canon de 75 mm mle 1897 modifié 1933 was developed for the M4 tank gun. The M4 was itself a fairly formidable development of the Soixante-Quinze by the standards of 1942, following the fairly obvious and easy development path of extending the barrel and the propellant charge to get a 50 m/s improvement in muzzle velocity. The next step, insofar as we can construct it, was the M1 76mm, 57 calibre barrel length, or 51 calibres when cut down for installation in M4 Shermans as the M1A1. The wiki announces that the M1 had "comparable" anti-armour performance to the 17 pounder, then admits that it had inferior penetration but made up for it with "greater accuracy," adding that in spite of its slightly shorter barrel, the 17 pounder has a higher muzzle velocity, and had a larger propellant charge. Which is to say, that the M1's propellant was not an improvement on the M4s.  

Now here is a cut and paste of something I've probably said on this blog already, but perhaps with editing aimed more at battlecruisers or some such.

Volatile chemicals were used in WWII for a number of purposes, of which the most important were demolition explosives, weapon bursters and artillery propellants. As a class, explosives, bursters and propellants are self-reducing chemical compounds containing a high energy molecular bond. When the bond is broken by an initial catalyzing event such as an electrical charge, percussion or ignition flame, the oxidant and reactant decompose into gasses, producing considerable energy, although less than, say, gasoline, the point of being "explosive" is that the reduction takes place independent of external air supply. The very energetic gas bubble expands rapidly in a way that may be loosely described as an “explosion.” Technically, this is often not accurate. A distinction is made on the basis of the shock wave propagation velocity between what are known as “deflagrants” and true explosives. The distinction is that in open spaces the velocity of the shockwave is too low to carry actually burning gas, with the result that deflagrants burn, while explosives explode. Various deflagrants are more or less stable in their response to external catalysis,  and compounds containing trinitrotoluene can “sweat” this constituent nitroglycerine. Sweated nitroglycerine can and does explode, and this concern may appear as an explanation for the American decision not to embrace dual-base explosives and propellants (deflagrants). 
   Most of the explosives used by the British armed forces in WWII were still trinitrotoluene (TNT), and ammonium nitrate, the choice of the mining engineer. Production of these on continuous flow basis had been well established in WWI, and although there was substantial increase in productive capacity, development of industrial technique was left to Canadian and American engineers, who developed the batch production method until it was even more efficient than the 20 year old continuous flow method. 

The need for new explosives was recognised first for military demolitions. Various workers identified RDX, and an arsenal explosion in Italy in 1938 signalled that military production of RDX was underway there. In reality, the Italians were just more careless --or, more likely, underfunded compared with their rivals. A small-scale plant was producing 75lb/day of RDX at the Waltham Gunpowder Factory associated with the Royal Arsenal Woolwich by 1933, while work went on to improve the synthesis method. Production was still at the level of only a few hundreds of tons of RDX a year in the autumn of 1941, notwithstanding the opening of one unit at a dedicated factory at Bridgwater in August 1941. Even with the opening of the second unit in 1942, dependence on Canadian and American supply remained heavy, and as with progress in conventional explosives production, it was the North American manufacturers who developed high volume RDX production techniques, and incidentally of PETN and HMX.

Summary: RDX production was not high enough, even taking into account North American production, to meet all interest in it. It went into depth charges and torpedoes rather than air-dropped bombs, for example, although the Germans eventually used it to charge the V-1s and V-2s. 

    Propellants also posed problems for chemists. Propellant gas eats gun barrels. The life of the metal at the breech, where the containment lasts longest, may be no more than 10 seconds, although this will correspond to a very large number of firings, from thousands in the case of a shoulder arm to at least more than 100 for a long naval rifle. Nor is the waste energy confined to the barrel. Incandescent gas is expelled from the gun barrel, producing an intense muzzle flash, which dazzles the weapon’s crew and identifies it to enemies during night fighting while contributing to muzzle wear through a back overpressure. The more energetically efficient the propellant, the greater the flash problems. Less energetic propellants, on the other hand, produced excessive amounts of smoke. 

All artillery propellants used in 1939–45 were classic carbon-nitrogen compounds. The simplest still in use was good old gun cotton, or nitrocellulose (cellulose nitrated with nitric acid) gelled with acetone.

Nitrocellulose was also an important ingredient in so-called double-base compounds, which in their first iteration contained a mixture of nitrocellulose and nitroglycerine to give greater energy density. The original modern brand name propellant was referred to as “Cordite,” and the name was perpetuated without distinction when pure nitrocellulose was succeeded by double-base, triple-base and even multiple base compounds. 

For example, Admiralty standard Cordite MC (65% nitrocellulose, 30% nitroglycerine, 5% cracked mineral jelly stabiliser) had a gas temperature of 3215K and a calorie content of 1025 calories/gramme(?).  MC was virtually a better-stabilised descendant of the WWI propellant, but by 1939 several triple-base propellants had become available. Cordite SC was 49.5% nitrocellulose, 41.5% nitroglycerine and 9% symmetrical diphenyl diethyl urea (Centralite), with a gas temperature of 970K and calorific value of 3090. The obvious major virtue of SC was that its lower gas temperature reduced barrel wear, but it was equally important that it could be extruded in high pressure presses, allowing its grains to be accurately sized and shaped cords and tubes. SC cords improved gun accuracy, while Cordite HSC or HSCK (potassium cryolite replacing Centralite) tubes gave higher muzzle velocity. Unfortunately, all of these propellants were doped with small amounts of chalk to counteract residual acidity, and the resultant calcium gave SC and HSC notoriously bright flashes. A flashless propellant was necessary, at least for special purposes. 
    During the interwar years intensive research at Woolwich led to the development of a new propellant combining nitrocellulose and nitroglycerine with nitroguanidine (CH4N4O2) and the gelatinising agent nitrodiethyleneglycol replacing acetone. The advantage of the nitroguanidine compositions was that they produced their impetus energy at much lower temperatures (as low as 2100C compared with 2840C for pure cordite), greatly reducing barrel erosion, while the new solvent permitted extrusion shaping. The resultant so-called triple base propellant NF (55% nitroguanidine, 16.5% nitrocellulose, 21% nitroglycerine, 7.5% Centralite, 0.3% cryolite) was awkwardly bulky, and never available in adequate quantities because calcium carbide is required in the synthesis of nitroguanidine, and its production demanded excessive amounts of electrical power. The sole Commonwealth production facility in Niagara Falls never produced more than 19% of the originally projected output. NF was thus never as widely available as hoped, and this was exacerbated by a decision to supply the US Navy, which after Guadalcanal decided that it needed low-flash cordite now now now now. 

Extrusion plant was also provided. This is another interesting case in which there were largescale manufacturing plant exports from Britain to the United States. I do not know if Michael Postan is aware of this, but am fairly sure that Corelli Barnett is not. Since not enough plant was available, inadequate quantities were delivered, preventing the U.S. Army from securing a supply of SC. 

This is where the rubber hits the road. Late war British tanks, using either HSK or some other material (that is, perhaps doping HSK with RDX), had approximately an 11% advantage in potential muzzle energy compared with tanks still using conventional nitrocellulose at the expense of greatly increased barrel erosion. 

Notice that I am not trying to reverse Postan's argument that "advanced American machine tools with lots of controls and widgets mean America is more advanced on the road to the world of there being computers and stuff like that" argument. On the contrary, not only Germany but even Japan were forging ahead on the same track as the British. Much of the information that I have presented here is not even classified. I am citing modern literature, but could just as easily point to a textbook reviewed in a 1938 number of The Engineer if I could just get at it. The Japanese did not, however, have solventless-forming methods, and substituted hydrocellulose and potassium sulphate for nitroguanidine, gaining the same virtues of flashless affect with the disadvantage of greater charge. 
So why did the Americans (and British tanks firing American-supplied ammunition) not have access to HSK (doped HSK?) propellant? Because it was not made in sufficient quantities in the United States. At some point it was probably deemed possible to go ahead with an antitank gun using projected new ammunition. I am guessing that this was the origin of the 76mm M1, although it is likely that the actual design story was more convoluted than that. We can at least understand why the American 6 pounder was slightly inferior to the British original, and why the 17 pounder was not put into production there. The 75mm M4 were manufactured by centrifugal casting followed by autofrettage, whereas British (and other) arsenals produced them from monoblocs, that is, from machining out a single forging and used autofretagged loose liners. Per Green et al. (82ff), Watertown Arsenal was making 100 M4s a month by the middle of 1942, and two Pennsylvania metalworking firms had received commercial contracts, but it was only with Oldsmobile's involvement that production reached the necessary targets. Three contractors, notably including Goodyear, received contracts for the "three inch" tank gun in 1942, but only an A-1-d priority  for machine tools,(priority inflation! "All my factory requirements are in the top 1% of all priorities") meaning that production did not go ahead. Another project for a 76 mm high velocity gun that did not go into production during the war is also noted.

This information is based on Ordnance papers. I suspect that the 76mm that did not go into production is actually the M1, and that there has been some confusion here. There is also some defensive obscurity in respect to manufacturing methods. Without centrifugal casting, the United States would not have been able to achieve the production miracle that it actually accomplished. There were simply not the machine tools to produce monobloc three inch guns, of whatever design. Modern centrifugal casting is certainly a good way of making high pressure gun barrels. It is, not, however, under the conditions in which Watertown did it, going to produce a comparably resistant gun barrel. 

When we see the family of improved ammunitions becoming available for the 76mm gun at the time of the Korean War, we of course are also seeing the flowering of wartime industrial investment. High pressure extrusion presses and direct-drive hydraulic lathes and, well, David Noble, Forces of Production, the rise of "numerical computing," the whole IT revolution thing. 

I say again: can't find the supposed productivity increasing effects of the "IT revolution?" You're looking seven decades too late. The spreadsheet is a huge improvement on the Cardineer rotary file, but the Cardineer is a  huge improvement on a ledger book. 

Why locate the changes in the postwar era instead of the modern? I have some ideas, but I'm working them out. As part of that exercise,  I am drawing your attention to "Gyproc." The Economist characterised "Gyproc" as an industry dating only to 1933. The Wiki article will give a different impression, as usual, a number of manufacturers, mostly American, occupy a shady terrain between the origins of "gyproc" in gypsum plaster and the transition to the modern, standardised drywall product.  I was curious, as recently as when I titled this post, as to whether there was a synergy between calcium carbide and gyproc production. This does not appear to be the case, leaving this as entire nonsequitur.

Shorter: In 1941-45, the United States invested in massive chemical and machining plant to bring its tank guns (amongst other things) up to the world standard. At the same time, drywall was just coming into use in construction. In 1945--75, the United States economy boomed dramatically. Construction boomed along with it, with costs brought down by the widespread use of drywall. At the same time, the precision control required by the new generation of machine tools led to the development of "numerical computing." I think that it is at least defensible to argue, in advance of actually doing [gasp] research, that there is a link here, between the development of a new chemical industry (drywall) and the expansion of the explosives and machine tool industries during the war. If not, I have dropped another day into an unplanned technical appendix to come up, at the end, with the ringing announcement that a rising tide lifts all boats.

No comments:

Post a Comment