In any case, I don't have customers to please. So as long as I'm paying attention to all the safety issues, I don't have to achieve perfection day 1.
I've got two or three 50 millionth indicators, and used them a lot for setting of bearing packs on machine centers. It all started out with a scrape job I did on a Heald internal grinder, and we were using a pair of .0001" indicators to check our work. The Heald used a V and a flat in the bed, and it looked like Norm and I had it within .00025" in about 48". But I knew that the relationship between the flat and the v had to be very close or the machine would wear out quickly. Yet I didn't know how to check this for sure (I had an idea, and it turned out I was correct for a change). Well I knew one guy that knew how to do this check, but he was also a pain. We go see him and asked him just what was good enough. He came over and had a look see, and comes back with some indicators and a home built gauge.
The indicator was a .000050" Federal, and he did the checks for us. He said "it would be good enough this time!" I'd never seen a .000050" indicator before, and he said it was what you need to do it right. Turned out he was pulling my leg and the grinder was near perfect. Thru the years Norm and I worked together scraping machines and other things, and when we wanted to put the other on the jack up we'd give the other a 50 millionth indicator to use!
The .000020th indicators are used to calibrate surface plates. They usually have a ruby tip, and don't come cheap. I gave the ones I had away, as I hoped to never work that close again
After that, dialing in a barrel sounds like child's play for you.
Thanks for sharing!
I guess I'm 3/4 nuts, because I enjoy working close. I actually wore out .0001" indicators, and destroyed more than one as well. That's how I came into the first Interrapid. I had a new Federal setup on a VTL adjusting the backlash out of the screw & nut. My boss came up and said to let him see where I was at. He pulled the wrong lever and made the indicator about as thick as a dime! He laughed and said he'd get me another one asap. About a week later he hands me a new Interrapid. I was ruined! I now own four or five of them.
I think a lot of folks tend to make a mountain out of a mole hill. If you see movement in the dial you can always take it out till there virtually no movement. Then do a check with a finer indicator. Working close is scarey to a lot of folks, but it's just a mental frame of mind. You learn to remove as many variables as possible, and the rest will fall in place. When I cut surface plates the spec was .000025" error per foot. Sounds frightening, but with the correct piece of equipment it's not all that bad. Getting one flat was easy, but getting one area to repeat with another area was often a pain in the rear. I've gotten surface plates so flat and smooth that you couldn't check them with a laser, and would have to go back over them with a piece of Scotchbrite to break the glaze.
Many years back I was handed the job of scraping in a Brown & Sharpe universal gauge grinder. The table was about 90" long, but the width of the V & flat were very narrow. You could not use a standard width strait edge to master it off of. Nothing was easy except for the quality of the cast iron they used. The bed was not ever planed right from the factory, and the pad the wheel head mounted on was a little over .03" out of parallel. Norm and I scraped on that thing till we were blue in the face. I actually burnt up a Bix electric scraper! When we finally got it to where we thought it was kinda close (about .0004" in 90") the guy that ran it dropped by to see how we were comming along (we'd been cutting on this thing for about a month). Told me that it had to ground round within .000010"!!! I was ready for Malox!! Norm went upstairs to the restroom and puked. Then this guy I knew from Gleason came buy, and said he knew about that grinder. They assembled it for B&S when they were down. He said the bed was not planed right (I knew that the wheel head mount wasn't already), and told me to measure the center line of the V to the flat before I went any further (we were about 85% done). The flat was .120" too high!!! The machine is junk as far as I'm concerned, and it also has to grind within 10 millions of being round. It would take months to bring the flat down to where it needed to be. Norm is sure we're as good as unemployed, and goes back upstairs for another upchuck. I call a high level meeting and dropped the hammer on them. Boss wants to seal up the bed and send it out to Viking Engineering to replane the bed. His boss won't let him, and the tool room needs the grinder. So we put it together, and gave it back to them. They loved it (??).
But in six months it needed to be rescraped. If they'd replaned the bed and then scraped it back in the machine would have ran for twenty years without being touched.
I might add that I lost about 15lb. on the scrape job!
You ever take any pictures of your work? I would be interested to see some before and afters or just a layout showing the process of rescraping a machine.
I tried to post a couple pictures of my front shooting rests a couple days ago, but so far can't get nothing to work right. Can anybody here help an old guy thru the procedure?
Machine scrapeing is a black art! No two guys do it the same, and after working with many guys over time I can look at the pattern and almost tell you who did it! It's hard work, and not very rewarding. I learned to scrape from old Dutchmen and Germans. These guys were extremely picky about work quality, and would often refer to another's work as "chicken scratches!" Later I worked with a couple others that were pretty good, and learned from them. I think my greatest advancement in the use of the idiot stick was when I started mastering surface plates and strait edges. I learned to develope patterns and how to cheat. It very hard to find two guys that can scrape together, as they both must cut about the same amount of metal in a pass (about .000030"). Norm and I were very similar, but our patterns were not quite the same. My dots were football shaped, and Norm never quite got that part. Not a big deal unless your cutting plates. I also read the dots like the old men did, and Norm never quite figured that part out (takes many years to learn). But Norm was the only guy I would scrape with, and I was the only guy he trusted. I used to make Norm extremely nervous when I drug out the electric scrapers! (he later relented) I remember ontime we had to rescrape a Devlieg Jig Mill that somebody had really screwed up. The "W" axis was out over .0065". I was laying on the electric scraper pretty hard, and the cast iron was comming off smoking. I knew I was taking about .00075" a pass, and was cutting it in layers to get it down to less than .001". The boy was on Malox by the time I made my third pass <g>!! The table had to be cut under one tenth parrallelism, and I had a long way to go. I had it parallel by the middle of the second day, and was done in four days. Right in the middle of the first day I jumped on the boss's scooter and went after my cigar box of stones (recycled thread grinding wheels). He looked at me with a lot skepticism, and I just laughed. The job actually took about a half day longer because somebody rubbed their fingers over the area I was scraping (the old men would have boxed his ears or at least smacked his fingers with a scraper). When they do this it leaves a glaze, and the metal dosn't cut the same. As for photgraphs of my work; there are none that can be published. TACOM has some in their files. And even with a photograph you couldn't tell much, but the idicator and autocollumator never lie! I do have an old book at the house that really goes in to detail, but the equipment is old and kinda crude
I never had to worry about loosing a job much with the ability to scrape in a set of ways on a machine. Nobody wanted to learn to do it as it was hard work and kinda dirty. Yet I enjoyed it.
I have been told a Billion dollars is one-thousand million dollars. A Trillion dollars? I won't even go there...
Nonetheless, I do believe I can wrap my tiny little brain around .0005". But .000025" error per foot?
What is a millon times a huge WOW?
surface plates and strait edges come in three ranges Lab Grade, Inspection Grade, and Tool Room Grade. The Government regulates all of them even if they don't own it. Most folks do surface plates to Inspection Grade (about .000050" per foot), and they pay for it in the end. Once you get it good enough to be an Inspection Grade plate it usually is only two or three more passes to make it a Lab Grade. A Lab Grade plate will last much longer than the others (longer usable life span). It actually takes three plates to cut a Lab Grade plate (you master off three plates). These are known as "masters", and are never ever used for anything but mastering another plate. I kept mine locked up in sealed boxes. A Master Plate will often be in the single digit range. So why the .000025" spec? Take a strait line on a surface plate and multiply it out to five feet long. Now you can have an error of .000125" in the total length. A good six foot by nine foot table will usually be under .0002" in the lines running from corner to corner. The one I used to master off of had about .0001" error in it's lines. But nobody ever used that piece of granit but me.
We cut granit with scraped in cast iron plates. These are very flat, and come in all sorts of sizes. The actual cutting material is diamond dust in several grades. I loved doing this, and the funny thing was that everybody thought I was killing myself! I really wasn't working all that hard.
Your gauging equipment must have a fraction of error in it when compaired to the parts being measured on it. If your working with in a thousandth of an inch, then you need to be thinking around two and a half tenths max. The better the job you do, the easier it is for the guy using the piece of equipment, and the faster and better he works. A Devlieg jig mill has a max compound error of .00025" in all it travel. A SIP Diaoptic is about two thirds of that. I used to check a rotary table that had a max error of two tenths of an arc second in it's moves. How you fixed it I never knew because it always checked out within specs. You just have to learn to work close, and some folks never get there