Interchangeable parts revolutionised the way things are made | Hacker News Reader

Interchangeable parts revolutionised the way things are made

bbc.co.uk

32 pointskitd14 comments

Fun Fact: One of the first mass assembly operations that used an early form of interchangeable parts was the Venetian Arsenale, which could produce up to one ship per day. The ships were floated down a canal in an assembly like manner and fitted with pieces made by artisans to standardized specification, allowing the same piece to be fitted to different ships as well as allowing for pipelining. As ships were also repaired in the Arsenale, this could be viewed as an early form of interchangeable parts as replacement pieces were fitted from stock.

Also, Galileo visited the arsenale and was fascinated by the many interesting questions of mechanics raised. For instance, why do the oars in the center of the ship most account for the ships movement? Galileo experimented with new oar designs as well as questions of naval architecture and artillery design, eventually moving to Venice and working as a consultant for the Arsenale. It was the world's first mass munitions factory.

Some links for further reading:

https://italoamericano.org/story/2016-7-7/arsenale

http://echo.mpiwg-berlin.mpg.de/content/shipbuilding/venice_...

https://leanfactories.com/venetian-arsenal-venice-ship-build...

The more I dive into the history of various technologies (and even professions), the more it becomes apparent that militaries are responsible for a vast proportion of historical advancements, even in the most unexpected areas.

Edit: Also responsible for a vast supply of historical datasets.

Edit 2: IMHO, the most fascinating is all the tie-ins between artillery, spycraft, astronomy, aerospace/rockets, geodosy & cartography (and invention of the chronograph), tidal monitoring & weather keeping, nuclear weapons, gps, and probably a dozen more topics I'm forgetting off the top of my head.

Individually, each of these is a mammoth of a topic on its own. But the interplay in their development (as influenced by big-picture military strategy) is just mind-boggling to me. It is by far, the most interesting, persistent, and deep rabbit-hole I've encountered.

(Hint: Ballistic missiles are essentially long range artillery. Hint 2: They require gravitational variation to be taken into account for targeting.)

I think this is absolutely correct. Huge advances are driven by war, for obvious reasons -- survival is at stake. John Napier discovered the logarithm due to his work on artillery. A lot of Archimedes's innovations were driven by the desire to build weapons. The demand for accurate clocks was driven by the Navy, giving rise to many innovations. And of course in the 20th Century, integrated circuit design was driven by a desire to build computers that could be embedded in ballistic missiles and withstand large amounts of heat. Fairchild semiconductors' first big contracts were primarily for defense and NASA.

I am rather concerned that this failed to leave out Eli Whitney[0], whose contributions for interchangeable gun parts were taught to me in university history class.

> By January 1801, Whitney had failed to produce a single one of the promised weapons, and was called to Washington to justify his use of Treasury funds before a group that included outgoing president John Adams and Jefferson, now the president-elect. As the story goes, Whitney put on a display for the group, assembling muskets before their eyes by choosing (seemingly at random) from a supply of parts he brought with him. The performance earned Whitney widespread renown and renewed federal support. It was later proven, however, that Whitney’s demonstration was a fake, and that he had marked the parts beforehand and they were not exactly interchangeable. Still, Whitney received credit for what Jefferson claimed was the dawn of the machine age. [0]: https://www.history.com/topics/inventions/interchangeable-pa...

I believe 1911 (handgun) slides and frames still have to be hand fit to one another. Though there is a company claiming tight enough tolerances that theirs don’t. I have two identical 1911, but have not tried it.

I wonder how original Colt 1911 were produced for military orders? I can’t see hand fitting those.

The AR-15 is fully interchangeable, but I have read of takedown pins on Colts not fitting more modern lowers receivers. Also wonder if the gas system has to be tweaked? The FN-FAL has an adjustable gas block, but that may be due more for ammo than parts compatiblity.

Sounds like an urban legend. Colt and John Browning shipping something that needs to be hand fit doesn't pass the smell test. Major world powers don't adopt a gun that can't have parts interchanged as their standard sidearm in the early 1900s. For a specialty weapon of 30yr earlier maybe but only the cheapest of "boring old small arms" were still being hand fit around that time and even then it was rare in industrialized countries.

Edit: adjustable gas systems are in fact there to tune for ammo variations and fouling.

I don’t think the original 1911 were hand fit, but all of the modern, upper tier 1911 claim to be hand fit. Why the purpose behind this; reliability?

If you're really lucky hand fit translates to "our QA department is stocked with measuring tools and they actually use them to confirm parts meet spec before being assembled, maybe if you're lucky they write down how over/under something is so that the assembly guys can pick combos that work well". (If you're paying big bucks this is hopefully the case.)

More likely it means someone cycled the thing to make sure it worked before slapping their "qa by <unintelligible scribble>" sticker on the product or packaging

It's mostly just marketing mumbo jumbo to make it seem like a fancy hand crafted product. Back in "the day" the marketing people would have advertised that things were made by precision machines. Today labor is expensive and machines are cheap so they're advertising that they put a lot of labor into it in part to justify the price (and the reverse would have been true in the early 1900s, they'd emphasize the machine contribution).

Firearms are tight like a retired hooker compared to the kind of tolerances any reasonably modern manufacturing operation of that sort (mostly machining and stamping/forming) is capable of working in (they need to be loose so all those steel on steel parts will move like they're supposed to). Loose fitment tolerances translate to less need for super precise parts to begin with. Say you have a slide that wants .010"+.002"/-0001" of clearance. Sure you can make that part to .0005 but insisting on that level of consistency just adds unnecessary cost to do so when you have an entire .0003" space to shoot for.

For ultra high precision stuff tighter tolerances and more consistency are used but that's mostly limited to the pressure bearing parts of the action that touch the case/bullet (the goal is to get that cartridge to fit in there the same way each and every time).

Some of the most reliable firearms out there use very loose tolerances. For a 1911 I feel very safe saying that the modern high end ones do it more because they can than because it improves performance in any measurable way. High end guns almost by definition will never see enough use to get to the point where manufacturing it more precisely could ever effect reliability.

I remember a piano repairman telling my parents (when I was way too little to understand) that replacement parts had to be within a certain window of serial numbers to fit. It wasn't because of deliberate changes in manufacturing, but had to do with wear and drift in the templates (or molds or presses or whatever). He described it as a "living" process, and that it worked because it cascaded in a complementary way; a small deviation in the shape of one key was "absorbed" and compensated for by the next key. It also prioritized consistent tuning (result) over persistent shape (method).

Put another way, the manufacturing process had built-in methods for self-correcting systematic error. They just weren't guaranteed to be consistent in the long term. (I'm sure there's got to be a more technical description for this, but it's beyond my knowledge).

I know semi-modern (opto-mechanical) survey equipment is manufactured on similar principles of self-correcting deviation (although these also tend to be dependent on strict method of operation for it to work itself out).

It wouldn't be a stretch to think a firearm as old as the 1911 was manufactured in a similar fashion as some point. Especially if there was a way to keep the long term deviation centered within spec. But that's just my guess.

Reminds me of this article, which I think was also discussed on HN

https://www.ribbonfarm.com/2012/03/08/halls-law-the-nineteen...

I never even considered that early machinery was custom made and had no readily available parts. I guess for any repairs, you had to go to the original maker, who would create a replacement part that would only fit on your machine.

Seems like a really simple concept, just like the drill press for that matter, but it took so long for people to realise the potential.

Also true of the most basic components that we take for granted now, e.g. screws. The concept was known to the ancients, but they weren't mass produced as fasteners until the 1760s [0]

[0] https://en.wikipedia.org/wiki/Screw

I remember watching a documentary about a country in central Africa where, less than 15 years ago, blacksmiths were making screws and nuts by hand. They needed to be kept together or the manufacturing variations would mean that last week's screws might not fit this week's nuts.

These guys were fast, but nowhere remotely as fast as thread-rolling machinery of course.

The Packard Merlin is an interesting story. The famous Rolls Royce Merlin was widely used in allied fighters and bombers during ww 2, but many parts where manually fitted. When licensed to the American auto manufacturers to mass produces, they tightened up the specifications and changed the things like the bearings.

https://en.wikipedia.org/wiki/Packard_V-1650_Merlin

Reply on news.ycombinator.com