Machines and factories make extremely precise things. How were we able to make the very first factories and machines that can make precise things without them? By hand? And I mean stuff like picrel
Machines and factories make extremely precise things.
pretty sure that was done with an EDM wire machine
stopped watching when that queer said "you totally should watch my other video".
we get it, daddy touched you badly.
This was the vid I thought of when I read OP's question.
Machine Thinking is the shit.
Thanks anon.
The very first manual machines were hand scrapped flat.
Go up to a Bridgeport wipe the oil off it then run your finger on the ways
See that cross hatch pattern? Some boomer’s dad scraped that by hand
That was Back when machining was an art. Now we just type some gcode into a machine and vape while machine goes burrr
Like people who know to machine? Everyone on the planet knows how to machine that’s why machine shops pay minimum wage to bring you in
Just like everyone under the age of 45 these days has a bachelor’s degree and the first time you meet someone who didn’t go to college you’re like whaaaaa???
Every post of yours I see, the less I believe you actually work in the machining industry.
>The very first manual machines were hand scrapped flat
They still are, traditional ways literally require scraping to function properly.
>That was Back when machining was an art. Now we just type some gcode into a machine
CNC machining is just as, if not more technical than manual machining if you're actually trying to be efficient about it. The principles of workholding, process design, and tooling selection are all the same.
>Everyone on the planet knows how to machine that’s why machine shops pay minimum wage to bring you in
Absolutely laughable.
I could take a dude from McDonald’s, stand him in front of a cnc machine and he will figure out parallels, he’ll look at the screen and be like oh x,hand z like a 3d printer!
Oh look handle jog, hmmm tool set measure button and a chart with 1,2,3,4 I bet those are tool lengths!
By the end of the day he’ll have a complete program, and probably figured out fixturing
Probably mid level fixturing programs
Literally anyone can do it that’s why they have dudes that shit in the river machining stuff in India for Americans now
>a dude from
>like a 3d printer!
First, very few people have used a 3d printer. But more importantly, even people who've used 3d printers a lot can't into g-code. They let the slicer to it for them. But most importantly, even if you can figure out g-code and and x-y-z and shit you are still going to break taps and ruin metal if not the mill itself because milling is also an art.
99% of 3d printer morons cant even home their machine properly.
they use a piece of paper and accidentally introduce a -0.2mm z offset.
But if ones standard is "it sticks to the bed" and "it fits if i redrill all the holes" the person doesnt even notice how shit of an operater he is
Most people under 40 have used a 3d printer.
The average McDonald’s employee has used a 3d printer at school, and likely has a non-stem bachelors degree
Like everyone under 40
I'm under 40 and have never used a 3d printer. I struggled through learning a tiny bit of cad/cam type stuff to set up a cnc router for engraving brass dog tags and have done a couple projects on a cricut printer...
I would rather just envision what I want to build in my head, then build it using materials i already have on hand, and cross any obstacles I encounter when i get there. For me it beats the hell out of sitting there fiddling with a computer drawing something up before I can even start the project.
>Most people under 40 have used a 3d printer.
> break taps
Hmm maybe I’ll read this manual that says canned cycles that’s in this pocket behind the controller on the machine
Face it machining isn’t a skill
Being a human inherently makes you a machinist, banging two rocks together… he was a machinist…
I use machine therefore I skilled
Nah try framing a house like those Mexicans who turn up to a job site with a stolen Milwaukee a couple 6 packs for lunch
>Face it machining isn’t a skill
Face it, you're not a machinist. You're a button pusher and draw shit on a computer. There are still machinists out there doing very real machining and they are skilled as hell.
I also turn knobs on bridgeports and lathes
Those machines were designed for children to work on production lines in the 1800s
Numbers on the dials because child labor back then wasn’t literate
Ok so you're a button pusher, computer drawer, and knob turner/polisher. I do agree with you 100% when you say you are unskilled. That being said not every machinist is unskilled like you.
>Hmm maybe I’ll read this manual that says canned cycles that’s in this pocket behind the controller on the machine
>still breaks taps left and right because that's 10% of the equation
>Face it machining isn’t a skill
This attitude is why you get paid minimum wage as a "machinist", you don't actually know anything about the field.
They give you the formula, and they give you the chart for form and cut taps, and the tools when your order them from Kenna or micro 100 says to leave room for the throat…
Idk what to tell you , or why you drank to look aid thinking this is some hard and difficult job
The heaviest thing we lift all day is a Kurt vise what’s a dx6 weight like 25 pounds?
And you only have to do that on jobs where you need to slap an indexed or angle plate in
Tramming it isn’t a big deal either
McDonald’s job
Also cam programming is hard
>clicks line
> climb
> type in the information on the sticker on the packaging your tool came with
>but work holding is an art!
Raise the part 1/4” in the air and make the vise hold that
But fixturing!!!
Bolt it to a flat piece of iron
>read this manual
Americans can't read. Factory works can't read. Nobody is going to pay you to sit in from of a CNC mill and read the manual.
This homosexual is a troll. He's doing the same shit on the EMT thread. Disregard.
you're just a bunch of consoomers and troony worshippers in the EMT threads so you deserve everything you get
>
An operator is not a machinist. Set the same person in a repair machining shop using a mix of tools and equipment including lathes far too large to make economic sense as CNC machines for that business and he'd take months to be useful. Repair machinists need to know fun stuff like how to properly weld up a worn shaft for remachining, weld and line bore holes on heavy equipment and much more. They need a working understanding of press fits, boring, making sleeves, installing same etc etc etc.
>ways still require hand finishing
Unbolt the way covers and you’ll just see linear rails now adays
Sometimes box
That’s fine that’s more maintenance and millwright knowledge anyhow
>manual machines
>linear rails
Box ways also need one side scraped you fricking moron.
The same thing can be said for basic knee mill or lathe us, and neither of those makes you a machinist. Now tell me how long it takes to give them a print and nothing else, and make 100 of them without the cycle time being 5x what it should be, along with 3 extra, unneeded operations.
> way covers
‘> thinks manual machines
You're the one that brought up manual machines first and that's what I was replying to.
>The very first manual machines were hand scrapped flat.
Sure bud, box ways are ground these days. Cnc controlled grinders
They sometimes hand scrape off brand or older machines for repair but it’s dying trade then again so is machining.
I called the death of production arc welding with laser and fsw but nobody believed it there either
Cnc machine programming is a dying trade, machine operators won’t exist.
The only jobs left in machining are fanuc robot programmers who can also setup cnc machines and industrial maintenance
> but tech leads to more jobs
You lay-off 30 operators and 10 programmers
Hire 1 setup dude and robot programmer and 2 maintenance crew
That’s a net loss in jobs
>Sure bud, box ways are ground these days. Cnc controlled grinders
Scraping ways, and scraping for flatness are two different things. You cannot have two, raw, ground surfaces slide across each other like that.
You very blatantly don't have a clue what you're talking about, you're just regurgitating half-info you've heard actual machinists say. I'm sure you're about to tell me AI is going to replace CNC programmers any day now, when it can barely do a proper drill cycle.
ChatGPT free version can already write gcode with a bit of error
>> but tech leads to more jobs
>You lay-off 30 operators and 10 programmers
>Hire 1 setup dude and robot programmer and 2 maintenance crew
>That’s a net loss in jobs
You forgot about the +4 Chinese workers making the tech.
Thank you. I love you for sending the video
I just finished the video. Not exactly what I was looking for but I found this video by the same guy
?si=KIfizjE6Lk-Pksy9
“Origins of precision”
This seems closer to my og question and by the same guy. Why not send this instead?
This is an excellent documentary.
I would also recommend viewing this Australian who has managed to make a lathe out of wood in his quest to faithfully recreate the Antikythera mechanism:
>“Origins of precision”
vouching
>PrepHole now just follows whatever shit is posted on youtube that day
the demand for precision has pushed scientists to figure out how to more precisely measure things
a driveshaft could rotate 60 rpm in early industrialization and the fat israelite owners would be happy selling 100 widgets a day
now the israelites expect 300 widgets a second and you are running an assembly machine faster than the human eye can watch
>make thing
>pretty good tolerance
>labcoat jerk says it's off if you measure in nanometers
>make better thing
>says it's off if you measure in picometers
No matter how precise you get, some asshat will come around with a finer measurement and insist it's not enough.
>can barely do a proper drill cycle
Good enough for my workplace about 70% of the time.
Some butthole will always tell you to hurry up or that you’re not doing good enough.
But when you hand them the print and tell them to show you how to do it better they can’t or they say it’s not their job to
Yes, by hand. How the frick else do you think? You want to start with making a flat surface and then developing some kind of measurement system. After that you can make a lathe and then use your newly defined system to make screws that allow for precise, repeatable movements along a thread. You can use those to upgrade your lathe and then you can use your new more precise lathe to start making high accuracy measurement tools which you can use to build other high accuracy production tools like mills and drills, taps etc etc.
How did we make accurate measurements and tools in the first place that weren’t wibbly wobbly like non straight lines
Hand scraped ways. Screws and micrometers. Three plate lapping. Vernier scales.
You're all the way up to 1970s level with just this much.
Now computers can do continuous compensation for everything that goes wrong so it's even better when need be
This.
3 plate lapping is the cornerstone technology that gives you a very accurate flat surface reference. Then with a lot of hand fitting you can make very precise stuff, but not repeatable (think of 18th-19th century watchmaking).
The biggest next milestone for the advance of industry was to set an international standard for the inch and for the meter and the invention of the technical marvel that are the Johanson Blocks to calibrate your tools, from that you can take the drawing of a part and make a perfectly fitting and replaceable part, unlocking the possibility for precision in mass production.
Pic is the Harrison H1 Marine Chronograph, an example of pre industrial precision instrument.
>gauge blocks
why the frick do you need that after all the scales, callipers and dials
Gauge blocks are how you check those things. Even cheap shit, china sets are insanely more accurate than almost anything else in the shop. A micrometer might be +-.0001", gauge blocks are +-.00001" or even less for the crazy, lab grade ones. They're still the best way to check tight tolerance slots and pockets.
They're very important for checking flatness and parallel tolerances to narrower ranges. Which will make sense to you if you've ever had to make a part off a drawing with GD&T notes.
>3 plate lapping
mind = blown
you can pull a thread taught to create a "straight" line. you can dangle a weight on a thread to find plumb. you can use a surface of water to define a flat area. you can scrape two rocks against each other until they flatten each other out. you can pull a thread around a pin to mark a circle. you can make a triangle with side lengths of 3 units, 4 units, and 5 units to create a true right angle.
>surface of water to define a flat area
wouldnt it be round
>20 nm deviation over 1 m span
That's flat enough for most practical purposes. Just the dust in the air alone falling onto the surface would make bigger bumps, as well as vibrations from air movement and random truck passing by a few miles away. Even standing nearby would probably make enough gravitational pull to distort the surface by comparable amount.
There really was no need for anything that precise until after machines were already made. That allowed us as humans to create more complex machines that needed more precise tolerances.
To machine something like that, you don't even really need machines with tight tolerances since it's made out of two separate pieces. You could do that by hand
Your question does not merit human spoonfeeding. You are free to study all the relevant items and will become wiser for doing so.
ChatGPT will have the patience to spoonfeed OP so best ask there as it's ideal for responding to those too lazy to be genuinely curious. Industrial revolution and predecessor history abounds. Get thee some.
>ChatGPT will have the patience to spoonfeed OP
Anyone who recommends ChatGPT is a moron who doesn't actually understand how it works.
OP, I had your exact same chicken-or-egg question, and I ended up reading Foundations of Mechanical Accuracy by Wayne Moore. That should give you some idea of the work involved
I get what you mean. What makes machines precise is how the compound error is calculated. While your current capabilities might not let you machine a part to .0001 tolerance, a lever mounted on a very stiff pivot can reduce the tolerance significantly, allowing you to machine a more precise part. Here is an example: you might have a tiny, quick but imprecise motor. It runs at tens of thousands of RPMs. With a controller, you can get it to stop within a few hundred rotations. Add gear reduction (rotating levers) and that ±200 rotations can mean a difference of only .0001 in your application. The trade-off is that it takes two minutes for a full travel instead of two seconds. With a varied knowledge of systems, you can integrate many different techniques to achieve a particular goal (increased precision for instance). As an example, if you do not have the means to make a precise vall screw or rack and pinion mechanism, you could use hydraulics. Those are often avoided because they are complex, expensive and impractical in small contained systems, but they are very precise: it is easy to reason how much fluid you need to inject for a cylinder of radius r to move a distance d. Furthermore, this cylinder could be connected to a lever that reduces the amplitude, increasing the precision even more. Thus an impractical hydraulic system could be the basis of producing a very precise part for a completely different system altogether.
?si=5xTFD_GXgCe3SfzI
>By hand?
the modern metal working lathe was literally invented to replace live long trained craftsmen by monkeys that can turn a crank
This is made by a edm, and except big compagny like GF or Makino, no one know how to make it, i never see any diy project about it
Oldest trick in the edm book right there.
That was made from 2 separate stocks. One was cut on the left side of the contour and the other was cut on the right side. Finally put together and ground to a nice finish it looks like its mad from a single stock.
Titans of cnc and even the guy that DIYd an edm machine did one
In the most basic sense, precision can be applied by getting simply a straight line.
Take two sticks and a string between them, hold one stick where it is and move the other around it, keeping the string straight.
You now have a perfect circle.
Scale this up to better materials, drawn wire and metal rods. It’s even more precise.
You can use your straight line and perfect circles to make almost any two dimensional shape with the right math. Using a standard measurement unit of the string you can scale pieces in relation to eachother.
And with these two basic principles, a straight line and a perfect circle, you can make some pretty impressive tools, and with those tools you can make precise things, to make more precise tools.
Start with shitty tool. Spend absurd amounts of time using it to build a better tool. Repeat. If you need a demonstration of the absurd levels of accuracy that can obtained with basic tooling and a huge number of manhours (and maybe a bit of autism), look up videos of people hand-lapping reflector telescope mirrors in their garage.
With gears and pulley systems. Just like in ancient times.
the two kind of main things to think about are
number one, getting something perfectly flat. you can look up the three plate method where two plates ground upon each other may become spherical, three plates would cancel out and could only be flat.
the second is the idea is that errors are not proportional to size, that is, one can make a part at large scale (by hand) with a relatively small error as opposed to a small part with a relatively large error. using this for example on a lathe one might produce a large scale feed screw and then apply that at an angle to divide down the effective pitch, reducing proportionately both the size and error together.
10th of a mm ought to be enough for anybody right
>make rough thing
>sand
>use to make more precise thing
>sand
>use to make more precise thing
>sand
>use to make more precise thing
etc.
something like that could only be made by an EDM wire machine.
consider that at first, things were not precise and then they became precise over time.
Big heavy machine has a part that spins very fast and at least 3 other parts that spin very predictably. Not a complicated concept.
What improved was metrology equipment and computer control.
Turns out that if you remove the meatbag and leave the machine under the control of an abstract set of gears, it runs with pretty predictable performance.
In general, the first machine tools were lathes. Even with poor construction, the lathe allows for perfectly circular cross sections, allowing rods and shafts and whatnot to be manufactured. These become parts for the next machine, which can use circular motion to convert to linear motion. A few cycles of this (making even better lathes, then even better mills) and you can get some impressive precision. Things take off once you have individual electric motors, rather than some belt or chain running off a central steam engine.
Starting point is a flat surface. You need to grind 3 surfaces against each other. If you just grind two surfaces, one of them becomes concave and the other convex.