How are items manufactured with such precision, when it requires a machine with precise parts to manufacture the items (which presumably requires another machine to produce it’s parts)? Does everything eventually trace back to handmade parts or tools?

29 views

How are items manufactured with such precision, when it requires a machine with precise parts to manufacture the items (which presumably requires another machine to produce it’s parts)? Does everything eventually trace back to handmade parts or tools?

In: 9

As someone who has slight experience with this I can tell you you’re kinda right. We build systems that can allow us to observe what we made though and that’s how we really determine accuracy. Of course all these systems rely on the best of the best from the previous generations and TONS of testing, but we’re at a point where those original tools and equations we use have become so good that they’ve allowed us to make better versions essentially. It’s hard to explain without going through a specific piece of tech and it’s iterations.

Yes, the original precision tools were made using hand tools.

The Antikythera mechanism was built sometime around 100BC, but has gears and bearings that have very tight tolerances. We’re not entirely sure how it was built, but we know some tools that were available at the time, like simple files and hand engravers. We also know they had a form of hand powered lathe to make round cuts easier.

In fact, it’s arguably easier to be very precise with hand tools than modern machine tools. This is because you can remove small amounts of material, check the result, and then remove more only if needed.

On the other hand, it also takes orders of magnitude more effort to actually complete an operation, which makes objects made this way very expensive.

Edit: There’s a YouTube channel you might find interesting: Clickspring

Chris is making a replica of the Antikythera mechanism, and along the way, he’s exploring how it might actually have been built using only tools and technology from 100BC.

[This is an excellent video on the topic.](https://www.youtube.com/watch?v=gNRnrn5DE58)

Yes, essentially. There are a few fundamentals like getting something flat by rubbing three axis planes together which will create a flat surface, and from that theoretically everything is built off of that. Then you need systems of measurements, use a natural constant like a stick or something if you don’t need to communicate this to someone else. Then evolve into codified units….

Eventually there was even a very profitable European company making precisely measured blocks that could be used to ensure the quality of machining. These had to be smuggled out of Europe during the war to help boost US manufacturing.

If this is a subject of much interest, you might check your local library for The Perfectionists by Simon Winchester.

If you are interested in more modern, there’s a YouTube video by Machine Thinking that covers the move into true precision.

https://youtu.be/gNRnrn5DE58

I program and run mills and lathes. How precise these machines are usually comes down to manufacturing tolerances, which can be simplified to fit of parts, and the repeatability/rigidity of the parts making the machines.

Modern precision machinery traces back to a screw-duplicating lathe that was made around 1780, I believe. (Source: a series of videos by Jacob Bronowski on the history of science and technology, made for PBS.)

Early screw-cutting lathes were turned by hand. They have a screw that moves the cutting bit as the lathe turns. By using different gear ratios so that the screw turns at a different speed from the shaft being cut, different threads per inch can be cut.

The first such lathe, and the first screw, was made by hand, with files, etc. Once you have that very carefully made first screw, you can duplicate it indefinitely. You can also make screws with different thread counts, and nuts.

Once you have a reliable supply of screws and nuts, you can build precision measuring equipment. Say you have a 1/4-20 screw, with 20 threads per inch. One turn of the screw is 1/20 of an inch, or 50 thousandths of an inch. Put a pointer on top of the screw, and a wheel with 50 divisions evenly spaced, and you can now measure the thickness of a piece of metal to a thousandth of an inch. (You need to make the nut part of a c-clamp-like piece of metal, and have some way to set zero, but you get the idea.) (No doubt Metric screws were produced in the same way in France, as soon as the metric system was adopted.)

I have copies of old books from the 1800s on how to scrape and file metal surfaces with great precision, so there was still a lot of hand filing going on, through maybe WWI, but once you have some means of precision measurement, as standard screws provide, you are on the way to the modern world.