In manufacturing, dimensions are free, but tolerances cost money.
Try grinding plano optical flats @ within tens nanometers [ Bilionths of a Meter ]
Ground & polished optics @ Huges Research Labs for a brief time.
Other wise known as Newton rings or color, the air space rings between flatsurfaces.
It has been noted the people that grind & polish the super large telescopes wind up commiting Autocide before or after the mirror suface is completed.
If not for watchmaking modern electronics would not have happened.
The growing of bules for the silicon wafers = same process for synthtic jewels.
The company that makes Hairsprings in CH makes the heating elements in aicraft windows. Can you find them sandwiched in there somewhere?
Contest years ago, who can ( Country ) can make the smallest drill bit.
CH sent one from the USA back with a hole drill through it.
Wow, I would think just the natural vibrations all around would not allow flatness to be ground down to that many decimal places. Must be lots of proprietary technology around that process.
I'd want to get a look at that drill bit with the hole drilled through it - sounds a little like Robin Hood splitting the arrow already in the bulls-eye. There are other ways to make small holes than drilling.
I briefly worked for a tool-making division of Kyocera and we were making 0.050mm diameter drills. I'm confident we could have made them smaller, but that was the smallest in our catalog. The coolest tiny cutting tool we made was a 0.080mm diameter 4-fluted ball endmill.
Hughes research lab had some nice lasers too.
I ground & pollished this recangular block of Non-Oxidising copper, it had a very very small hole in the center. To keep the desired piece true there were blocks arround it so the edges would not round. They use ruby red lasers to align the chemical lasers.
On another note for those interested in early quartz watches, Huges made some but not sure how many. I am sure they are RED LED displays though.
I haven't checked ebay to see if any have shown up.
I'd want to get a look at that drill bit with the hole drilled through it - sounds a little like Robin Hood splitting the arrow already in the bulls-eye.
So I had a look through some photos, and I thought I might have something that isn't a marketing video, or hearsay/guesses, but actual drawings with dimensions and tolerances...here is what I knew in the back of my mind I had somewhere in my archives...
This is a photo I took of a manufacturing drawing for the main plate of a watch movement. Here is a closer view where you can see some tolerances for some of the hole diameters:
So the tolerances you see on some hole diameters there are 0 to +4, -2 to +2, some are looser so -2 to +5, and even a few at -2 to +7. These would be in 100ths of a mm.
I've not yet come up with anything for tolerances on things like pivot diameters, but I'll keep looking.
One thing to keep in mind with tolerances is that in a production situation, there are different types of tolerances. In a situation where you are making a one off part, the only tolerance that is important is the final tolerance to determine if the part is good or not.
In a production process that is continuous, meaning that you are machining one part after another for an extended period of time, there will be final tolerances, but also another complete set of tolerances - process control tolerances. These are tolerances that are tighter than the final tolerances, and in a situation where the machine tool is for example turning part after part, the operator would take samples at set intervals, check the critical dimensions, and plot the dimensional changes over time. This is often done via computerized systems with gauges that are connected to a computer and software that will plot subgroups of checks, and plot the process in graphical form. This allows the operator of the machine to view the process to follow the trends for things like cutting tool wear, and when for example an offset needs to be put into the controller to account for tool, wear, or when the tooling may need changing.
My engineering days were spent in the automotive world working for a tier 1 supplier, so this sort of thing was the norm for machining processes of any kind. I've been through a lot of watch factories, so Blancpain, VC, PP, AP, Chopard, JLC, as well as some independent shops.
Of all the factories I've been through, the only one where I saw quality assurance data posted up on bulletin boards in the same manner I would expect to see in an automotive plant, was at Patek. They were a little sensitive (to say the least) about me taking photos at all, and I had to ask permission for each one I took, and those charts were off limits. However, looking at it and understanding the numbers, I was a bit shocked at the rate of defects - in an automotive world these processes would be seen as being out of control.
Cheers, Al
So I had a look through some photos, and I thought I might have something that isn't a marketing video, or hearsay/guesses, but actual drawings with dimensions and tolerances...here is what I knew in the back of my mind I had somewhere in my archives...
This is a photo I took of a manufacturing drawing for the main plate of a watch movement. Here is a closer view where you can see some tolerances for some of the hole diameters:
So the tolerances you see on some hole diameters there are 0 to +4, -2 to +2, some are looser so -2 to +5, and even a few at -2 to +7. These would be in 100ths of a mm.
I've not yet come up with anything for tolerances on things like pivot diameters, but I'll keep looking.
One thing to keep in mind with tolerances is that in a production situation, there are different types of tolerances. In a situation where you are making a one off part, the only tolerance that is important is the final tolerance to determine if the part is good or not.
In a production process that is continuous, meaning that you are machining one part after another for an extended period of time, there will be final tolerances, but also another complete set of tolerances - process control tolerances. These are tolerances that are tighter than the final tolerances, and in a situation where the machine tool is for example turning part after part, the operator would take samples at set intervals, check the critical dimensions, and plot the dimensional changes over time. This is often done via computerized systems with gauges that are connected to a computer and software that will plot subgroups of checks, and plot the process in graphical form. This allows the operator of the machine to view the process to follow the trends for things like cutting tool wear, and when for example an offset needs to be put into the controller to account for tool, wear, or when the tooling may need changing.
My engineering days were spent in the automotive world working for a tier 1 supplier, so this sort of thing was the norm for machining processes of any kind. I've been through a lot of watch factories, so Blancpain, VC, PP, AP, Chopard, JLC, as well as some independent shops.
Of all the factories I've been through, the only one where I saw quality assurance data posted up on bulletin boards in the same manner I would expect to see in an automotive plant, was at Patek. They were a little sensitive (to say the least) about me taking photos at all, and I had to ask permission for each one I took, and those charts were off limits. However, looking at it and understanding the numbers, I was a bit shocked at the rate of defects - in an automotive world these processes would be seen as being out of control.
Cheers, Al
Wow, this is so cool that you can share this spec!
The format is a bit unfamiliar to me, but I'm going to study it for awhile. It's definitely not GD&T per ASME Y14.5 with the X and Y coordinates on the table, but it's not immediately obvious to me what the X' column is for, and I'm wondering if that column all the way on the right "R-[diameter]-S" is some sort way they're expressing true position tolerancing?
Based on the tolerances the holes are most likely reamed, but not honed.
Your observation regarding what would be acceptable for Statistical Process Control (SPC) limits in automotive compared to watchmaking actually doesn't surprise me. Nobody is going to die if a watch movement jams (except maybe James Bond?) I teach Tolerance Analysis in the medical device industry and one of the primary points I try to drive home for the young engineers is to always consider severity and detection in designing for manufacturability. An undesirable fit in an assembly, if undetected (and 100% inspection is NOT an effective detection method!), will result in some consequence that may be insignificant, or it may potentially result in patient harm.
I think they prefer “sales rep” to “salesman”, but yeah it’s a great industry to make your career in whether in engineering or sales. I was in the Army Reserves for 22 years so no desire to get reacquainted with the DOD. Where I work now I design catheter delivery systems to place stents and valves into the heart. Instead of open-heart surgery I give the docs a tool that they can run up the plumbing starting at a 3/8 inch incision in the inner thigh. Patients are typically back home next day. My current project is for treating pediatric congenital heart defects, so this is the most rewarding work I’ve ever done - it really feels like a privilege sometimes to get to work on solving these technical problems.
Really cool - what was the set-up for this, and at what distance? I'm imagining something like the way they set-up the splitting a bullet with the blade on "Forged In Fire"
You might be familiar then with Dr. Gundry's research.
A buddy works for Medtronics in R & D.
Self taught electronics whiz, name on a few patents with them.
When he started, wound up teaching the folks with sheep skins how to design ac to dc filters. For a clean dc voltage supply.
Shot with my Olympic recurve set-up at 18m, which is the regulation indoor distance. Same set-up you see here (but this is at 90m):
We actually try to avoid this, because it's costly ($$), and because the arrow that hits the arrow that's already in the target will deflect out of the middle and cost you points.
So in tournaments we shoot 3 arrows per end, and then go to score and collect them. We use "3-spot" target faces, meaning that each target has 3 smaller targets printed on them, and you shoot 1 arrow in each of the 3 targets to avoid this sort of thing happening. This one was shot in practice when I was shooting multiple arrows at one target.
Help me understand, I have no experience in archery, that picture you shared of where the one arrow split the other - that was something that just happened whiteout intent? I assumed there was some pre-designed, installed, and validated set-up with the goal of doing that. How frequently does that happen even with the process safeguards you mentioned that they put in place for competitions?
