If you have a block of metal and one side has been scraped flat, how would you prepare a flat surface parallel to the first using only handwork methods? The method that I have seen uses a spring caliper to check for high spots. This seems very difficult and error prone and I wondered if there were another trick.
A granite plate with a dial indicator on a stand is very accurate. If you have scraped deep valleys, you can put a flat ground on top where you measure. But it should be enough to measure the thickness on the high spots. You just touched them, so you know where they are.
Surface plate and indicator will show you high and low spots, would be extremely tedious to file and scrape things down that way, though. Would be easier to set up a grinding spindle on a surface plate, ala Lautard's Machinist's Bedside Reader, spot grind, then lap afterwards. Or use a hand-powered planer or shaper. Hey, they're for "handwork"! Saw an early glass negative of a machine shop in the
1800's, every bench had a hand-operated shaper on it. You didn't say to what tolerance, though and how big the workpiece is. Thousandths would be pretty easy, tenths of thousandths would be tedious without precision grinding machinery. Are you going to further handicap things by not allowing a micrometer?
Thousandths is what I had in mind. Surface plate is allowed, but no mike and no shaper either. I don't quite understand what you have in mind regarding the indicator. The procedure as I envision it is basically identical to feeling for high spots with a spring caliper.
If you are only working to thousandths, then file the other side flat-ish (checked with surface plate). Then check for systematic slope with a dial gauge from a stand on the surface plate, against the scraped surface, with the filed surface against the surface plate. Slide the block under the dial gauge and note the slope in X and Y directions. File until the systematic slope is acceptable. Then, if you need better than thousandths, work on the flatness with the scraper. Continue checking as you go. Easy, but repetitive.
I've still got a set of three internal square standards to finish off using this method. I've scraped the outside faces flat and square to a close-but-not-finished 30 spots/inch^2 by comparing them with each other and the surface plate. Now I need to do the internal faces, using the external faces as reference... It would save about two months' of evenings/weekends if I spot ground (grinded?) them rather than scraped them.
I was going to suggest spotting the surface down with a die grinder, as we do at work. We have to flatten the bottoms of irregular insert steels using this technique. Takes perhaps 20min to take a surface from out by 0.1mm down to effectively complete contact (surfaces range from perhaps 100x100mm to 500x300mm). I've never had the opportunity to measure the flatness (not a trivial task) of these spotted surfaces, but you can tell by the way they slide across other flat surfaces that they are *very* flat.
Sometimes we have to change the angle of a (newly) hardened block to square up previously reamed dowel holes. Takes some time, but can certainly be done using this technique.
Why are you scraping the surface instead of grinding it? I'd assume some type of hobby project where you're doing it the way it used to be done?
I needed to generate standard quality squares without having a standard quality square to start from, so I couldn't use a surface grinder. I'm using normalised cast iron rather than hardened steel, so a die grinder isn't really appropriate (unless I used a polishing mop on it ;-)
If I were paying for my time at my chargeable rate, it would have been cheaper to go out and get a granite square with an NPL certificate but, yes, this is hobby stuff.
If I spot grind the internal faces (mount surface plate on table of surface grinder, then manually slide item on stationary plate under grinding wheel) then I can get the inside faces as good as the outside faces independent of the grinder. That is a worthwhile exercise.
Once I've got the internal squares, I'll use them for checking the ways of the mill that I'm refurbishing. That really would be un-economical if I were paying for it. I'll use a power scraper for some of the roughing and alignment work on that and finish it by hand. That should have been this winter's job. Unfortunately an abused HLV-BK followed me home and jumped the queue :-)
You need Lautard's Machinist's Bedside Readers, one of them tells how to make squares from scratch, also how to grind and lap blocks so that faces are square and parallel. Lots of other helps and hints, too.
I have Lautard, and Connelley and Smith and Moltrecht and Galyer&Shotbolt and Moore and Morgan and Thomas and Law and Bray and a few others. That's how I know what I'm doing.
Be careful. The Lautard article tells how to make a tiny swing-type surface grinder, and also a little bit about lapping. The only thing it says about parallelism is if you're lapping and it's getting out of parallel, hang a weight on the high side and keep on lapping.
Once again I'm going to repeat my advice: don't buy any books unless you've already read them and know they will work for you. How to read a book you don't have? Simple - check it out of your local library. They don't have it? Simple - request it via interlibrary loan.
There. I just saved you about $120 for all 3 Lautard books. Isn't Usenet great? :-)
I have all three (TMBR) and I must say that they are exactly what the title suggests. All are a good read and actually contain good information that is hard to come by. The price is steep but, in this case you DO get what you pay for. These books could form the basis of a university-level course in numerous diciplines.
This would be possible, if laborious, to achieve by hand scraping, with one surface plate and either two square gage blocks or one surface gage.
If you are thinking primitive, like all you have is a surface plate, then you will have to use it to originate the other gages. Obviously one gage cannot gage two degrees of freedom of two planes' parallelism, or three degrees if you also must constrain a specific parallel separation.
PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.