Hand-filing questions

[SagaciousKJB]

Well, I'm starting a two year Machine technology course at a local technical school. However, for the first six months or so, we'll be doing a lot of projects by hand.

The first project, the one I'm currently working on, is what they call a "screw plate”. It's just a 4.5 inch by 2.5 inch steel plate (about .325" thick), with various sized threaded holes to use for checking bolts.

My problem is a very basic one, but a big one. We're required to file the sides of the screw plate flat and square. We're using a flat-plate and layout paste (they call it prussian blue, I'm not too familiar with terminology) to check for square ness. I'm not too clear with the instructors as to how I'm supposed to interpret the blue markings that are made on my screw plate after I run it over the surface of the flat-plate..

Initially, my understanding was that the blue layout-paste on the flat-plate would mark the high-spots, showing me where to file. My instructor, Hans, then confuses me by telling me that the only parts of the blue dye I should file off are the “shiny” parts, and that I should leave the dark marks alone. I’m not too clear on what he says… There seems to be kind of a language barrier between the two of us. He speaks very good English, but I still find a hard-time understanding him a lot of the time.

Does anyone have any idea about what I’m talking about? I figured that I should probably ask some other people about this, as I’ve been trying to get the side of this screw plate square for two weeks now, and every time I get close, I start gaining low-spots. Every time I ask my instructor about this, though, I just become more unclear on what I’m doing.

I’m using mostly needle files, but I have a feeling that the ones I have are too coarse. I’m mostly using a No. 2 Three-Square, Flat, and Warding swiss file, all double-cut. I’m not too sure on the specifics of the steel we’re using, but it’s fairly soft.

[Mcgyver]

ok the flat-plate (surface plate) is a reference surface - its flat, to a high degree of accuracy. You are right that the blue is used to identify high spots - it is called Prussian blue or just blue and is basically oil paint without the drying agents.

when you bring the work into contact with the blued surface plate, Imagine a highly magnified side view: very flat surface plate, layer of blue and then the work. the highest spots on the work are going to extend down through the layer of blue and touch the surface plate, the surrounding area, or those that are not so high will not make it all the way to the surface plate. When you pull it away, the thinnest layer of blue (looking a bit shiny because of the steel underneath) is the high spot - it was in contact with the surface plate where as the darker areas were close to, but not quite in contact. the areas with nothing are lower still and didn't even make it to the top of the blue layer

it also has to with how thick you spread the blue on the surface plate. how you tell to what standard it has been done to is with a loop with a grid - an inspector would count how many flakes per square area there were.

you will be doing this for six months if you're starting out only taking off the shiny bits, imo remove material at more wholesale rates until you get close, then use the technique he's instructing you to do.

to get it square, you'll need more than a surface plate, but maybe that’s his lesson for the following six months

[SagaciousKJB]

By "loop", do you mean like a jewler's loop? Explaine that process a little more thorougly... Being in a class with other students, keeping a consistent amoutn of blue on the surface plate can also be a challenge.

Also, thanks for your explanation, it's a lot more clear as to what I need to do now. I'm not quitre sure what else I would need to check for squareness (maybe a machinist's square), but the process is to get four of the sides flat (according to the surface plate), and then surface grind the actual face of the screw plate.

[Mcgyver]

yeah by loop i mean loupe, I'm only partially illiterate, but frequently make mistakes like that trying do internet brain dumps.

I don't have one but was shown one by an old time tool and die maker. they had a grid and the inspector would count how many marks there were in each square. This was the quality control or standard to which things were scraped to – saying it was completely blued out I guess was subjective, based on how thick the blue was smeared so this was a way to quantify the quality of the scraping job.

[ctate2000]

Depends on how you apply the blue. If you apply to the part and rub on the plate you file the shiney parts. If you apply to the plate you will file the blue parts. Remember use the blue sparingly. Best way I have seen it done is using a balled up shop rag with a tear sized daub of blue imbeaded in its surface. Apply the blue to rag and pat it into the fibers so that it puts only the slightest amount onto the surface. Too much will make everything blue. Think about what you are trying to achieve and reason through the process. Also remember school should teach you the basic skills, but most importantly it should teach you to think. If you hand a job to 100 toolmakers you will get 100 different ways to complete it.

[Geof]

Felt pen ink works pretty near as good as blue. But given that your instructor is traditional enough to have you doing the hand filing if he caught you using felt pen he would probably have kittens.

[Mcgyver]

wax on wax off. Is the teaching this traditional skills still standard practice these days? it wouldn't have surprised me hear learning to file or scrape had been replaced with some computer app training. knowing these traditional skills is a good thing, makes one a more versatile and better metal worker imo, but I also get the impression that there are lots of shops that wouldn't know what to do with a file - then again that may be more production environment vs mro

[One of Many]

I have used hi-spot bluing ink, felt marker, white lead, red lead, dry spray-on moly or graphite and plain old grease. The simpler the better with less mess. Anything that is thin enough not to influence the mating surface contact points to proof out a load bearing contact percentage.

White and red lead is not as available as it was in the past. Not many of the old timers left to pass on these hand fitting skills either.

A file would not be the best tool to use since it covers too wide of an area. Hand scrapers will keep the area to be removed localized with much better control.

DC

[RotarySMP]

For a good surface finish you need a "Mill file" with a single set of teeth grooves, rather than the diamond pattern teeth of a double cut. You should chalk you file often, and keep it clean with a file card (short bristle wire brush) otherwise it will load up and ruin the surface finish. The last key to good surface finish is draw filling.

Draw filling will tend to make a concave surface so you have to be very careful on the ends. If you have to file a larger area, a half round file will normally have a slight bend along its length, so you can use its flat side to lower a high spot (convex centre).

[HuFlungDung]

I would agree with One of Many, that a scraper is what you need to use to create a flat. The best part of a file is the end, once you sharpen and hone the edge. It should be ground square to the filecut faces, with a slight bit of convexity, so that the corners do not scratch when scraping. I also usually sand the file cut teeth flat for about 1/4" back from the end, so that they surface of the file is smooth leading up to the end edge. When the edge is very sharp, you can easily push off light layers of metal without an extreme effort. At this stage of making a flat, you should be past the 'extreme effort stage' anyways Unfortunately, a regular file does not hold a good scraping edge for very long, so you'll get lots of practise sharpening the scraping edge.

Hold the scraper at about a 45 degree angle to the surface. Push it ahead to scrape exactly where you want. Lift the scraper on the return stroke, otherwise you just dullen it.

I don't believe that a regular two handed grip on a file will ever produce a flat. You can get better accuracy by applying pressure to the back of the file immediately over the area where you wish to remove material. Use the other hand to pull or push the file, but the stroke should still be very short. You must maintain the fingertip pressure over the area you want to remove the high spots from. For this, you will want to use a thin mill file, one that is flexible enough to bend, so that your fingertip pressure has the desired effect in making the file cut mostly underneath your fingers.

[Mcgyver]

HuFlung, I happened to have some pics of a file like you described on photobucket - it's a lot nicer than the commercial scraper and I 100% agree with your comments.

Remember though, he's using needle files so its easy to localize the removal of material. I assumed (yeah i know the joke) that they were doing it this way as it might be easier to start a beginner off with a needle file rather than a scraper, but admit were it me i'd for sure be using the scraper. sounds like with hold Hans on the job, they're lucky they're being allowed to use tools, much less take one of Han's files and smithy and grind it into a scraper

[NC Cams]

Wide (as opposed to Sharpie pointed) magic markers (permanent ink) are less messy than prussian blue. The low spots retain the color and the shiney spots are "high".

Another way to lap is to use progressively finer grades of wet/dry sand paper (220 to 280 to 320 to 400 to 600 etc) on a piece of thick (1/2" or 3/4" glass). Keep the paper wet (water is fine, add a couple drops of dish washing detergent as a wetting agent).

Don't just lap in 1 direction. By changing directions, you can see the low spots by the sand marks that did NOT disappear after you changed direction.

We have generated near mirror finishes with this method if you get finer and finer - it is the same method used when you do metallurgical grain structure analysis so it is not some sage wisdome deal.

Can be tough to lap a milling table this way but the project you're working on will work just fine this way....

EDIT: files tend to cut in one direction. Thus you do NOT want to slide them back and forth under pressure. IT is pretty easy to "see" which direction the file wants to go to cut.

So, push in proper direction to cut, lift and move file back, re apply to surface and repeat. Hans will probably have a royal fit if he sees you sliding it back and forth as if it were sanding block.