What's new
What's new

Filing Flat

horomancer

Plastic
Joined
Oct 22, 2009
Location
Texas
To all the old machinist out there, how does one learn to file dead flat?
Back when i was in watchmaking school, learning to file was the first course of study. I never got over making a small hump in the middle of what ever plain I was working on despite countless hours of practice. I can get reasonably close, the deviance in my 'hump' being about 0.005 mm, but I know from my instructor and other students that a dead flat surface is achievable.
 
Use a curved file! I'm not kidding. A really handy thing for your toolbox is a fine flat mill file, which you've bent with a torch into a slight curve. You want the center to "droop" maybe 1/16" or so. I also usually bend up the tip to form a handle at the far end; that also helps to see at a glance that it's a bent file, and which side is down.

You use a flat file to bring the surface down close, watching the height of the edges. This leaves a crown in the center. Use the curved file to work the crown down to flat, checking with a straightedge, and establishing the surface finish.
 
Is this something that should have been posted yesterday?:confused: 0.1mm is pretty good with a hand file. But 0.005 of a mm tolerance with a hand file.:nutter By lapping it you could get that tolerance.
 
My father, who started his apprenticeship in Portsmouth Dockyard in the mid 1920s, told me that he once asked one of the old fitters how to file so beautifully flat.

The fitter explained: 'Hold the file like that and that. Now go away and practise for twenty years'.
 
Yes the years of practice answer. I hate that one since I know people that didn't spend years of practice and can file dead flat :: Envious::
I couldn't hit a .005 tolerance with a file ::maybe if I was REALLY trying and had some fine files...maybe:: but if I have an adjacent flat surface my mic shows roughly .005 difference from edge to center.
My files are curved slightly, but I'll try picking up a cheap Nicholson and making a modified one as suggested.
 
Swiss hand files are tapered toward the tip to aid filing flat. I don't think it's possible to file a flat surface with a flat file.

When I first started clockmaking I began the BHI distance learning course that obsesses about filing flat and square. I've since learned that antique clock parts are actually not flat or square, which makes reproducing them a lot easier.

Rich
Richard McCarty, Conservation and Restoration of Antique Clocks
 
Yea I just had to make an arbour for a small carriage clock. It should function well, but I'm not satisfied with my results. No one but me will know that the square section isn't quiet flat or truly square, but I know and I want to rectify that.
 
My former father in law who apprenticed as a machinist in Switzerland some time in the late 1940's or early 50's said that as his first exercise he was handed a rough hunk of metal and a file, and his job was to file it into a cube. Overall dimensions were unimportant, but it had to be a proper cube, flat, true and square. He would file and file, take it to the master, who would critique it and send it back, and he would file and file and file some more. By the time he passed the test, it was a whole lot smaller than it had started out, but he'd learned how to file something flat.
 
My father, who started his apprenticeship in Portsmouth Dockyard in the mid 1920s, told me that he once asked one of the old fitters how to file so beautifully flat.

The fitter explained: 'Hold the file like that and that. Now go away and practise for twenty years'.
Mate worked at RR, they gave him a steel ball & a steel cube. Then handed him a file with the words "make a cube from the ball & a ball from the cube"
 
My former father in law who apprenticed as a machinist in Switzerland some time in the late 1940's or early 50's said that as his first exercise he was handed a rough hunk of metal and a file, and his job was to file it into a cube. Overall dimensions were unimportant, but it had to be a proper cube, flat, true and square. He would file and file, take it to the master, who would critique it and send it back, and he would file and file and file some more. By the time he passed the test, it was a whole lot smaller than it had started out, but he'd learned how to file something flat.

That was the first job I ever did in engineering. I started out with a 2" chunk, finished 3 weeks later with a 1.5" cube. My fellow apprentice filed his lump down to about 1", the instructor took one look at it , threw it in the bin and said " Too small, go and get another piece" ! We looked enviously at the nearby shaper and surface grinder. The next two jobs were- Make one side of the cube into a triangle so you had a square and a triangle at either end. Drill and file triangular and square holes in a piece of flat plate about 1/2" thick and fit the cube and triangle into the plate, feeler tight.
I must have spent most of my first 3 months on various filing exercises, I hated it at the time but it's stood me in good stead , I can still file like "ringing a bell". My shoulders know it though. Regards Tyrone.
 
We looked enviously at the nearby shaper and surface grinder.

Fellow student did just that at A&P school when we were tasked with taking a piece of roughsawn timber & told to plane it perfectly square. He took his home that day & ran it through his dads jointer/planer...
If I'd have had one at the time I'd have done the same thing ;)
 
Guess there is nothing left for me but shoulder abuse and tears until I get it right. I'm mortified that it will be all for naught, two years of schooling didn't do it so maybe it's just one of those things I have to making a work around for.
 
I too had to learn to file for squareness and flatness using test blocks, however I recall the final flatness of the main surfaces was achieved by scraping.

Yes, I recall wondering why we couldn't just use the surface grinder - but now a file is my friend, whereas untrained people would do almost anything rather than use a file. The good thing is that their files (and hacksaws) are only worn on the middle couple of inches, always an unused portion at each end!

BTW, I talked to a Taiwanese man who works for a huge company which makes most of the world's golf clubs. He assured me the Japanese tool and die makers also do the same sort of basic training, i.e. hand made projects.
 
Horomancer;

There's no magic secret about filing technique. Your body isn't a milling machine. It's physically impossible for you to stroke a file while holding it in a plane flat to 0.001". It doesn't matter how many years you've been doing it.

The big lesson from all those apprentice projects is learning the discipline to constantly measure your own work, and make little corrections to bring it to the final shape and dimension. The process of hand filing a surface to flat is all about checking it with a straightedge, finding a bump, then using a curved file (or a scraper, or whatever) to trim the bump down. Repeat, repeat, repeat until you can't find any more bumps. Then it's flat.

With years of experience, you just get faster at finding the bumps and trimming them the right amount the first time. And not going too far.

The old guys will try to make you believe that it's all due to their massive wisdom and skill. But it's not magic.
 
One thing that I picked up over the years is to use files of appropriate size for the job. Within reason, as flatness is approached a smaller file will be controlled more by the job than the filer whose part will be more influence than control. You might start a job with a 14 inch bastard & finish with a 6 inch smooth. Also warding files can be flexed a little during the stroke to 'hollow' the centre of a flat workpiece. And of course the technique we were forbidden as apprentices, draw-filing, generally produces a hollow centre if taken too far. Still can't understand that veto, as a 'free agent' I often find draw-filing useful. It's amazing what can be accomplished quite quickly with a file when machine tools are not available.
 
Horomancer;

There's no magic secret about filing technique. Your body isn't a milling machine. It's physically impossible for you to stroke a file while holding it in a plane flat to 0.001". It doesn't matter how many years you've been doing it.

The big lesson from all those apprentice projects is learning the discipline to constantly measure your own work, and make little corrections to bring it to the final shape and dimension. The process of hand filing a surface to flat is all about checking it with a straightedge, finding a bump, then using a curved file (or a scraper, or whatever) to trim the bump down. Repeat, repeat, repeat until you can't find any more bumps. Then it's flat.

With years of experience, you just get faster at finding the bumps and trimming them the right amount the first time. And not going too far.

The old guys will try to make you believe that it's all due to their massive wisdom and skill. But it's not magic.

Hi Bruce, There's a bit of a contradiction in your last two sentences. Filing's like most games, the more you "practice" the better you get. Apart from wear and tear on your joints you do actually seem to get better as you get older. It's not magic but it takes a while to get good at it. I suppose it depends how hard you want to learn.
Maybe some guys never get it. I'd love to play the guitar and I've tried to learn but I'm no nearer now than I was years ago, I just don't seem to have what it takes. Other guys just seem to be able to pick it up really quick. Regards Tyrone.
 
I was just going to say this whole guru on a mountaintop filing thing is stupid, but Bruce put it a lot better. I can file as well as parts need to be filed, but for removing any significant quantity of metal I use a machine designed for the job. As the footnote on the print says, "file to fit, weld to suit".

CH

(and don't give me any nonsense about freehand sharpening being as good as that done with a fixture)
 
Guy Lautard's Third Bedside Reader offers this: Put the file in the vise and hold the work.

If it's at all like scraping it seems to me it would require a lot of bluing to get it right.
 
My current work around is to place the file on the bench and draw the work across it like a sharpening stone. This does the trick, but it can throw the angle off a few degrees. I know there are better ways now to do the job, but I lack the equipment more than naught.
 
just a point of interest but how would you check for flatness???

i have a picture in my head of running a straight edge across the surface while looking for the gap/light/horizon between the work surface and the straight edge and to true a block to 90' you would do the same thing with a square right?
 








 
Back
Top