|
The pin must consume some amount of volume in the hole; if the pin and hole are perfectly mated at that level, no air in that volume can escape and you have to compress it. Have you ever played with one of those sets of gauge blocks made for super-precise manufacturing? You can kind of slide two of them together on their ultra-fat faces and they'll stick very firmly. It's sort of like magnets only totally not, a really disconcerting feeling. One theory about why they stick has to do with a virtual vacuum created between them holding them together, although I think there's also theories about interacting atomic forces due to the atoms in each surface being so close to one another. e. TooMuchAbstraction posted:Wouldn't you then run into an issue with the shaft vacuum-welding itself to the hole? My understanding of this is extremely rudimentary, but I believe that if the two surfaces of a metal are perfectly clean and in perfect contact, there is no functional difference between there being a hard vacuum around them or not, because either way, no gas can interfere with their joining. The problem with vacuum welding in space is that the very clean surfaces don't even have to be perfectly mated: small points of contact can spontaneously weld. Or something. Like I said I don't know much about it. Leperflesh fucked around with this message at 21:33 on Jan 29, 2018 |
#
?
Jan 29, 2018 21:29
|
|
|
|
| # ? May 24, 2024 01:54 |
|
|
Leperflesh posted:
Yes, every day. I still remember the first time when I started here forever ago and a guy showed me that, it's crazy how well they actually stick together.
|
|
#
?
Jan 29, 2018 22:07
|
|
|
Relavant: https://www.youtube.com/watch?v=sFrVdoOhu1Q He shows a 1 micron fit in the first minute.
|
|
#
?
Jan 30, 2018 14:37
|
|
|
M_Gargantua posted:Relavant: https://www.youtube.com/watch?v=sFrVdoOhu1Q At work so can't watch, but isn't a micron .039"? Edit: that doesn't sound right, but that's what the internet is saying. God drat metric! Edit again: Never mind, guess it's .00004"? A Proper Uppercut fucked around with this message at 18:52 on Jan 30, 2018 |
|
#
?
Jan 30, 2018 18:48
|
|
|
A Proper Uppercut posted:At work so can't watch, but isn't a micron .039"? No. 1 micron is ~39/1,000,000ths of an inch or 0.000039" e: You can always use google: 1 micron to inch
|
|
#
?
Jan 30, 2018 18:52
|
|
|
Sigourney Cheevos posted:No. 1 micron is ~39/1,000,000ths of an inch or 0.000039" Yea I knew it wasn't right, but sites like this say otherwise. I must be missing something.
|
|
#
?
Jan 30, 2018 19:25
|
|
|
A Proper Uppercut posted:Yea I knew it wasn't right, but sites like this say otherwise. I must be missing something. Nope, that site is correct, it successfully calculates 1 micron is .039 thou. And since a thou is a thousanth of an inch thats .000039"
|
|
#
?
Jan 30, 2018 19:30
|
|
|
M_Gargantua posted:Nope, that site is correct, it successfully calculates 1 micron is .039 thou. And since a thou is a thousanth of an inch thats .000039" Oh gently caress, I get it. I've never seen something written out as a fraction of a thousandth like that. When I work in tenths and millionths it's .0001" or .00001"
|
|
#
?
Jan 30, 2018 19:36
|
|
|
I was just wondering about it mathematically. I'd never thought about it that much but I guess the '10mm / 1cm' mark on a ruler isn't 1cm, as a mark cannot represent a point in space - a point/line has no thickness. So 1cm is either the right hand side of the mark or the left side. If we were to make the wall, say of a square hole, be exactly in line with the left side of the 1cm mark, so they formed a 'continuous wall', and the pin to go in the hole sized such that it is pushed up against this wall (but no further) then I guess they are both 1cm in size and should theoretically fit with no air gap. Of course, if they overlapped an amount of 0.00000000001cm then you would be asking the material to deform this much for the pin to go in - not much of an ask - so this should work fine. edit: But if they are both on the same line then they are touching an infinitesimal amount, and if the material was infinitely stiff they could not pass through one another ? Agggh Mudfly fucked around with this message at 22:59 on Jan 30, 2018 |
|
#
?
Jan 30, 2018 22:46
|
|
|
Mudfly posted:I was just wondering about it mathematically. I'd never thought about it that much but I guess the '10mm / 1cm' mark on a ruler isn't 1cm, as a mark cannot represent a point in space - a point/line has no thickness. So 1cm is either the right hand side of the mark or the left side. At that kind of tolerance, you'll be seeing different fits depending on the relative temperature of the two parts, as well. Also the cleanliness of the surfaces, and maybe even the rates of oxidation depending on the material. also at some point you're measuring the widths of atoms... oh yeah also take into account van der waals forces Leperflesh fucked around with this message at 23:06 on Jan 30, 2018 |
|
#
?
Jan 30, 2018 23:03
|
|
|
Mudfly posted:I was just wondering about it mathematically. I'd never thought about it that much but I guess the '10mm / 1cm' mark on a ruler isn't 1cm, as a mark cannot represent a point in space - a point/line has no thickness. So 1cm is either the right hand side of the mark or the left side. We've got one machine that can hold almost +/- 1 micron. Realistically we double that for quoting purposes as gauge correlation has proven to be a huge sticking point. The machine has a coolant chiller, hydraulics chiller, and 3 micron filter. We use a dedicated laser micrometer as well as some fancy-dancy lobing gages. Measuring form is tough and we use a special granite surface plate with ruby indicator set to make sure there's no weirdness. Things get really wonky during final inspection and unless the customer is setup in a comparable way you get some really frustrating correlation conversations. All in all it's impressive as we got about 1,000 pcs per shift on the last run that needed it. As Leper said, temperature is an issue. You can totally squeeze it tight in your hand and watch it grow on the laser mic. We've never had issues with cleanliness or the likes affecting but I think most people aren't equipped to measure it properly. It's always cool to see people push the boundary, but sometimes I think some designs compensate for basic flaws by just tightening tolerances.
|
|
#
?
Jan 31, 2018 01:01
|
|
|
I run into that problem of our customer not measuring things things the same way we are and it drives me up a wall. We have been building gages for a major gun manufacturer for them to check their machined parts on. Their engineers stick .0001" tolerances all over the place, and more than once I get a call from their QC guy saying something is out of tolerance, and then I tell him how we checked it, and then *somehow* the part is fine now. We're a 8 man shop and they have 1800 employees. You'd think someone there would have an idea how to check their own stuff.
|
|
#
?
Jan 31, 2018 02:15
|
|
|
Anyone know what would be the main hurdles to a DIY surface grinder? In Australia they start at 4k for either a new Chinese model or a second hand something else. Occasionally around here you can get old mill tables quite cheaply which got me thinking... The Chinese surface grinders use what look like Hiwin rails, so I don't think the x/y system would be too hard to make, especially since I have access to a surface plate and can scrape everything. The spindle on the other hand.... Could you make something cheaply with plain bearings? Does anyone have any good readings that could point me in the direction of research for this sort of project? I also read somewhere the 3 phase motors were precision balanced.
|
|
#
?
Feb 5, 2018 09:08
|
|
|
Mudfly posted:Anyone know what would be the main hurdles to a DIY surface grinder? This is probably of interest. No idea how well it will work in the end, but it's at least worth looking at. You should also check out the CNC thread. CBJamo fucked around with this message at 12:12 on Feb 5, 2018 |
|
#
?
Feb 5, 2018 12:00
|
|
|
Mudfly posted:Anyone know what would be the main hurdles to a DIY surface grinder? We've got a big surface grinder and a little baby one. The little one is extraordinarily simple with a bronze bushing for the spindle. The slide is just a dovetailed way with a cable system that you actuate with a lever. All manual. I think a person could replicate that with linear rails. I'm not sure generic Hiwin rails are the answer, you'll need a good preload and possibly a gothic arch or crossed X style of roller bearing. Surface grinder tables move way quicker than a mill table. Our larger unit uses a high speed hydraulic cylinder and it is smooooooth. Maybe you could replicate that with a linear motor? The hydraulics are definitely non-standard. Beyond that flatness and rigidity is huge. It's a whole different world when you build a grinder compared to a generic milling machine. If you'd like pictures of our lemme know.
|
|
#
?
Feb 5, 2018 17:01
|
|
|
I'm definitely interested in seeing that baby surface grinder. I'd also think you could put together a real small surface grinder with off-the-shelf hobbyist-oriented CNC parts... if you're fine with hobby machine-level precision and travel speed. It's that ~0.0001" precision goal that'll really get ya here, the grinder retrofitting part itself seems fairly simple in comparison. I also found an example of a CNCed mini mill converted into a surface grinder here, apparently he just stuck an angle grinder to the spindle, made an adapter so it can mount surface grinding wheels, hooked it up to a variac to drop the RPMs way down. No comment on how well it actually works. https://www.youtube.com/watch?v=QIg1qC3wl-E I'd also be amazed if people don't throw rotary tools on gantry-style CNC routers from time to time to do some impromptu surface grinding on small parts. Ambrose Burnside fucked around with this message at 20:59 on Feb 5, 2018 |
|
#
?
Feb 5, 2018 18:43
|
|
|
Just as a comment, threw one of those cup grinder wheels on my bridgeport with a homebrew arbor to do some light clearancing once...
|
|
#
?
Feb 5, 2018 19:07
|
|
|
Slungblade pics of that would be great. Somehow I get the feeling your little surface grinder will be quite larger compared to my machines. It looks really easy to just make A surface grinder, I just wonder how hard it is to make one that gets something flat within tenths, and what the big requirements are, especially with regard to the rotating parts. I was sort of thinking you could geometrically eliminate runout with a good dresser, but you would need to eliminate play either free or under load with preloaded spindle bearings. Not sure if I'm on the right track.
|
|
#
?
Feb 5, 2018 19:43
|
|
|
Mudfly posted:Slungblade pics of that would be great. Somehow I get the feeling your little surface grinder will be quite larger compared to my machines. I'm headed to that building later, I'll snap some pics. You can eliminate wheel runout by dressing but not wheel imbalance. Once the wheel is under load the runout becomes irrelevant as the spindle will want to climb up on the part. It will literally climb, reach peak stress, dig in, and repeat. This can be audible, visual, measurable, or irrelevant. It's one of those things you learn the first day on the machine. Surface grinders take light light light cuts. My big machine is beefy as gently caress and can only take a few tenths cut. I'll second using a cup wheel on a bridgeport. The R8 taper isn't beefy enough so we cut notches in the wheel to reduce the load. But your ways will be all  regardless how good you think your guarding is. Grinding swarf gets everywhere. Everywhere. My rule for buying lathes is if it comes with a tool post grinder I pass, regardless how nice it is.
|
|
#
?
Feb 5, 2018 20:17
|
|
|
All I was doing was cutting a thou or two taper trim on some cast iron engine parts... really pretty light cuts. I should really rehab that bridgeport as is.
|
|
#
?
Feb 5, 2018 20:52
|
|
|
First we have baby surface grinder. It's a Harig Super 612. I don't think they make these anymore but they make some really nice little workheads.  We've got a tiny little magnet plate, golf ball for size. It has a coolant system but is a messy fucker. Machine doesn't see much use unless the big one is tied up. Big machine! Rawr!   Golfball for scale. It's sitting on an electromagnet.  Our newest acquisition for the fleet. That'll be #10 of that size, plus we've got a few bigger ones too. Cincinnati centerless grinder. We're currently rebuilding it, redoing the electrical, etc.
|
|
#
?
Feb 5, 2018 20:58
|
|
|
Yooper posted:First we have baby surface grinder. It's a Harig Super 612. I don't think they make these anymore but they make some really nice little workheads. Those Harigs are nice little surface grinders. We have an old Reid of similar size, and our "big" one is a Kent (8x16 magnet, automatic feed and all that stuff)
|
|
#
?
Feb 5, 2018 23:44
|
|
|
Do surface grinders normally move the table forward and back or the grinding wheel? I can imagine that moving the table would be throwing a lot of mass around, but moving the wheel might introduce an unacceptable few tenths of misalignment?
|
|
#
?
Feb 6, 2018 01:13
|
|
|
They move the table. The issues from mass getting tossed around gets mitigated by the very tactile nature of the feed wheels, and the only adjustment directly affecting the position of the wheel is height/z axis.
|
|
#
?
Feb 6, 2018 01:26
|
|
|
It's impressive to watch the machine grind a big slide or plate. The whole table slings from side to side. You set the x feed hydraulically as well. Once you define the boundary of the work all you need to do is turn the dial, which I believe is one rotation for 1 tenth. (0.0001") The big pain in the rear end is dressing as once you move you need to find your size once again. Oh and when you forget to turn the magnet on. That's a mistake you only make once. Next time I run it I'll try and record a video.
|
|
#
?
Feb 6, 2018 01:44
|
|
|
Also the table on a typical machine might weigh like 100 lbs while the rest of the machine is like 1500 or more. It ain't going anywhere.
|
|
#
?
Feb 6, 2018 01:46
|
|
|
A heads up because I think most of y'all would be interested in it, the last few pages of the OSHA thread are filled with detailed effort posts about tooling and process for drilling oil wells. Discussion starts here: https://forums.somethingawful.com/showthread.php?threadid=3763899&pagenumber=631&perpage=40#post480945232 Yooper posted:Grinding swarf gets everywhere. Everywhere. My rule for buying lathes is if it comes with a tool post grinder I pass, regardless how nice it is. QFT. The single most expensive machine in any serious grinding shop is the coolant filter. shame on an IGA fucked around with this message at 01:56 on Feb 6, 2018 |
|
#
?
Feb 6, 2018 01:49
|
|
|
Finally got a chance to finish this thing up! Spin up the second quasi-hemisphere, and a few other bits to get started.  Some complications, mostly because I'm stubborn and wanted to use one of those shiny 3/4"-10 stainless setscrews for the main thread- my only matching tap was a taper tap, so thru holes only really, and def. no shallow blind holes since you don't reach full thread depth for 7+ turns. Also the lathe doesn't want to cut less than 11 tpi, so single pointing 'em was out. Decide to make a threaded bushing to press into a matching bore.  Finish the donor screw first- remove threads for .400", leave .375" of nice thread, step drill thru to 1/2", leaving about .06" wall at the narrow end. Plenty of meat for this one. Pick out a slug of brass for the female thread, it should run really nicely on the rolled stainless threads without galling up. Bushing is ~.900" OD, .405" long, with a generous bottom-side OD chamfer to make install easy, and the topside OD left sharp. Final installed thread depth will be ~.025" longer than the male stud, to ensure things seat on the mating faces instead of bottoming out on the screw. Once the thread assembly is happy and all is deburred, move on to finishing the shells. Put 'em back on the same 1/2" threaded mandrel and finish all OD work you may have forgotten (grooves, etc) before the next step!  Drill out 1/2"-13 threads to about .62" deep to form the bottoms of the storage cavity at either end. Happily rediscover a reground ball-ish end drill from a previous project, so they even look real nice down in there  Bore .400" deep for a snug slip fit to the screw, we're gonna zap it in with loctite 609 so it wants a ~half thou to not squeeze out, ideally.  Final step- skim a thou off the face to confirm all is square and the mating surface is unbruised, then pull it, degrease both parts with acetone, loctite and press home.  While that's curing, finish up #2. Matching drill op, but the bushing dia. is a bit larger on this guy.  .400" deep by ~.901" dia, should still have at least ~.13" of meat at the thin corner down there, if my eyeball facets on the outside aren't too far off print.  Nip a quick chamfer with the compound on the drilled/bored bottom corner, it'll try and hide a sharp wire edge and you know someone's gonna stick their finger in there. Blue sharpie makes it easy to confirm it's gone, what with all the reflections. Leave the top edge of the bore sharp. Confirm bushing fit, and pull/loctite/press when satisfied.  Set it aside for ~15 mins to chooch, and go get some coffee. Mmmm, coffee. Okay now go pop it back in the lathe. Because we left the bushing standing proud of the mating face by ~.005" with some sharp corners on purpose, and that's because we want to skim a thou or two off the assembly as a whole. So get things running true, and do it.    Looks good, but does it screw??   Hell yeah, it sure does.   It closes up pretty drat good, the seam is "there" but blends in nicely, and the coarse threads and stainless/brass combo means it shoudn't get stuck closed, even if you crank the wrong way. Just gotta make sure the faces are wiped clean when it goes together so they don't scratch. It's certainly not the way to make hundreds of them, but for a nice one-off G-job it should do the trick. e: shame on an IGA posted:Well, might as well share the spreadsheet from my apprenticeship when I had to manually cut hemispheres all the time Yeah this works drat well, I prefer to blue up after stepping to see progress and use single cut files though, as I hate getting much emery/grit near a (good) lathe unless I can help it. Lathespin.gif fucked around with this message at 12:24 on Feb 10, 2018 |
|
#
?
Feb 10, 2018 12:02
|
|
|
That's a beauty
|
|
#
?
Feb 10, 2018 17:42
|
|
|
Lathespin.gif posted:Okay now go pop it back in the lathe. Because we left the bushing standing proud of the mating face by ~.005" with some sharp corners on purpose, and that's because we want to skim a thou or two off the assembly as a whole. So get things running true, and do it. That is very pretty and loctite should minimize galvanic corrosion. Nice job!
|
|
#
?
Feb 10, 2018 22:02
|
|
|
I’m super impressed, the brass threads and stainless screw, that’s some quality work there, WAY beyond what I was hoping for. And the cuts around the perimeter in each face really help to hide the center seam. A+, top notch work.
|
|
#
?
Feb 11, 2018 02:45
|
|
|
Nice work, very pretty! I like the thorough documentation and good photos, it's always a treat for challenging projects n one-offs where someone has gotten creative n clever and tried something a little different to make it all work out properly. Unrelated: any die cutters here, or anyone with knowledge/experience with steel rule dies? I have a product that I'm developing retail packaging for and it's looking like any of the generic standardized boxes/hang tags/etc I can find are very poor fits, and anything custom-printed is prohibitively expensive and/or has a minimum order quantity I can't justify. My designs are really simple, just simple polygons with big-radius bends and 4 small holes punched through at specific points and a die area of like 7x2" at the most. I'm not sure if I wanna buy them from a pro or try taking a crack at them myself. I've read a couple people say that steel rule dies are quite cheap by tooling standards and one of those things that often isn't cost-effective/sensible to attempt yourself, but I have access to a laser cutter that will eliminate a lot of the labour required. I'll get quotes on Monday but I always value The Thread's Collective Reckons soooo Ambrose Burnside fucked around with this message at 06:12 on Feb 11, 2018 |
|
#
?
Feb 11, 2018 02:49
|
|
|
I've been on two projects at work with dies like that. One used flat clear plastic with an embedded blade and the other was etched steel that wrapped around a magnetic drum. For both cases we outsourced them because they were so cheap: ~$100 for the plastic and under $200 for the steel. We're cutting polyurethane film and they last around 100k cuts. We run the flat ones through a roller press with another piece of plastic to back up the cut. They have foam to help keep our production team from getting cut but it still happens. Depending on your volume, you may be interested in partnering with a cnc pattern maker or cnc router with vacuum and a drag knife. bred fucked around with this message at 04:23 on Feb 11, 2018 |
|
#
?
Feb 11, 2018 04:20
|
|
|
I've got friends with those hobby-level Cricut machines, no good for actual production but I'll probably bribe one of em to let me prototype a bunch of designs all at once so I don't have to fold up sloppy library printout scissor-cut patterns all afternoon like a schmuck. For actual production... yeah, I oughta get some quotes from a couple non-die-cutting sources if I can find em, can't hurt. I like the idea of owning the dies and making packaging as needed for pennies, but that'll become a huge pain in the rear end if I suddenly need to fill triple-digit orders or whatever. Cross that bridge when I come to it I guess.
|
|
#
?
Feb 12, 2018 04:33
|
|
|
Also, be ready to iterate your die design, especially if your packaging has any folding or complex tabs. I feel like we go through 5-10 rounds of design changes before we understand what areas are important and how to control the dimensions on our boxes. With anything, perfection is possible but don't plan on it.
|
|
#
?
Feb 12, 2018 08:05
|
|
|
Oh yeah, no bespoke boxes or anything at first, I don't want to still be prototyping stuff a month from now. For the first packaging spread I'm sticking to custom backer/insert cards i'll twist-tie the product to, they can stand on their own or get thrown into standard-size kraft jewellery boxes depending on the sales context and price-point.
Ambrose Burnside fucked around with this message at 20:53 on Feb 12, 2018 |
|
#
?
Feb 12, 2018 09:14
|
|
|
Ambrose Burnside posted:I've got friends with those hobby-level Cricut machines, no good for actual production but I'll probably bribe one of em to let me prototype a bunch of designs all at once so I don't have to fold up sloppy library printout scissor-cut patterns all afternoon like a schmuck. For actual production... yeah, I oughta get some quotes from a couple non-die-cutting sources if I can find em, can't hurt. I have a Cricut machine. I don't think it would do cardboard, even for small run prototypes.
|
|
#
?
Feb 12, 2018 14:00
|
|
|
I'd go with cardstock, not actual cardboard. I have a friend who uses hers for cardstock fairly frequently, albeit iirc not an actual Cricut and with a heavier blade intended for thicker materials. She's cut brass shim stock with it somehow so I'm not too worried about that end of things.
|
|
#
?
Feb 12, 2018 20:53
|
|
|
I just finished a big project. A birthday / Christmas gift; they asked for a kitchen island. I've been blacksmithing for a few years, and this is my first attempt at doing a big finished piece. I'm pretty happy with the results! I posted a few more pictures on imgur : https://imgur.com/gallery/4C5I4
|
|
#
?
Feb 12, 2018 23:14
|
|
|
|
| # ? May 24, 2024 01:54 |
|
|
That is super rad, nicely done.
|
|
#
?
Feb 12, 2018 23:30
|
|
