by Graham Howe

A FRIEND emailed me as he was doing some fly-cutting and asked if I new how he could get a good finish. He was machining some hot rolled steel which I personally like as it does not have too many stresses and is much cheaper! The finish he was getting using a single point HSS fly-cutter was very uneven and full of ridges and he said: "I'm afraid that from my side, I find the going a little frustrating right now and it all has to do with the damn poor finish I am getting while fly cutting. It feels as if the job is in control of me instead of the other way round.

“It is almost impossible to aim for a size and get a good finish at the same time. Just when it's going well and you think you’re in control on the final cut another few flying saucers appear so you think you'll take yet another very light  final - final cut when the same thing happens..... the result is that the job just keeps getting smaller and smaller....... have nothing left just now. I know that it does not affect the workings of the job at all but I would really like to get a super finish as well......  just looks so much more professional."

Time to  think

I searched through a couple of books and the Internet but it seems finish and small milling machines do not altogether mix well. I decided it was time to think things through as I, too, much prefer a machined finish that looks good. In the past I have just done this by surface grinding but it should be possible to get a fly-cutter to behave.

I was thinking along the lines of how cutting with a single point tool in the lathe works with very little side rake and how end-mill side helix edges manage to do a reasonable job. When it comes to the milling machine, assuming it is trammed correctly, you see the tell-tale cross-hatch pattern which is, I think, quite nice looking but running your finger nail across the machined surface reveals all is not smooth.

What I wanted was to eliminate the cross-hatch evidence and yet know the surface was perfectly flat. I have seen several suggestions that one sets the mill out slightly of tram to ensure there is no cutting on the reverse side but this will inevitably produce a dished effect, even if it is slight. The main problem is that once trammed correctly I just hate to undo my previous diligence and patience - if it is set up correctly then leave it be!

Single cutting tool

Back to sorting this problem out. I decided to keep a single cutting tool simply because it is so difficult to set multiple cutters in the cutter-holder and ensure that each cutting tip is exactly at the same height. The cutter-holder I use is a home-made affair which can take up to four cutters, each of 1/4in. diameter.

Fortunately, I had a store of round HSS tool steel which, I think had been, in a previous life, from an engraving cutter. What I do know is they are very hard and once an edge is ground it seems to keep sharp - some of the modern HSS tooling is pretty poor and frankly soft and a nuisance to use especially in the case of fly-cutting since the edge is constantly hammered by intermittent cutting.

That said, for the tool I have come up with the amount of metal removed is very little and so even the cheap HSS or hardened silver steel may serve equally well, but I have not tried using it. The other thing I decided was that the cutter would only be used as a finishing tool.

Incidentally, for ensuring an even table feed it is probably essential to have an auto-feed which you can choose the best speed. Alternatively, if feeding by hand then you need a great deal of patience and ability to keep the feed speed constant.

One way I used to do this in times now long gone was to turn the hand-wheel with one hand and use the other hand to add a little drag on the hand-wheel perimeter. You still need a lot of patience!

The basic cutter design turned out to be much easier than I first anticipated.  Although I finished the main cutting edge with my Stent TC&G, it could just as well be done free-hand

and later honed on an oil stone.  The objective is to shape a single cutter such that it effectively does the cutting work gradually by having a leading edge that starts the cut and then continues cutting along an edge that is angled until it gets to the trailing edge. 

What this means is that, unusually, the cutter has a long edge in gradual contact to emulate the same effect as the side edge of an end milling cutter (i.e. the trailing edge tip provides the final finishing cut). 

The cutter is mounted in the tool holder so that the bottom face of the cutter is basically square to the work surface with just the slightest relief given to the leading edge. The corners of both the leading and trailing edges also have a small, honed radius. The top rake and front clearance (imagine for this discussion it is a lathe tool) are about 5 degrees but this is not critical, side rake does not matter and so could be left as neither positive or negative.

Critical angle

The only crucial angle is the small relief given to the long edge doing the cutting.  Imagine the cutting edge to be a knife blade edge but angled so that the back end is lower than the leading front end.

Note: For the picture above the cutter has been repositioned after completing the full length.

I think the clearance from front tip to back tip should be about one a degree at most but when I made mine I just ensured the leading edge was clear of the work surface when the trailing edge tip touched the work surface or such that I could see a light gap at the leading tip corner. The other thing to mention is the radial positioning of the cutter in it's holder which needs to ensure that the long cutting edge, discussed above, is not on a direct radial line from the spindle centre but instead angled at about 30 degrees.

In operation, the work-piece will have been pre-machined in all its glory displaying grooves and ridges and since the amount of metal being removed by the finishing-cutter is so little (circa 4 thou) I oiled the work surface with a mixture of neat oil and WD40, the latter just to thin it out). There is absolutely no requirement to use suds as cooling is not necessary (the spindle was running at 500 rpm) and the thin curly chips produced are absolutely cold. For this job the oil just needs to lubricant the cutting edge so that it does not rub or dig-in to the work-surface.

The finishing-cutter leading edge (lowest corner) is touched to the work-piece such that it removes just a few thou of metal. As the feed proceeds, the leading edge corner removes the bulk of this few thou of metal and as things progress the tapered cutting edge removes just the remainder of peaks and valleys until it runs out at the trailing corner.

At this point another few thou has been removed but the surface is now perfectly flat and very smooth. The resulting chips produced are like tiny hairs but when magnified (enlarge picture above) it can be seen that they are in fact small curly metal chips. I have not experimented with different lubricants or spindle speeds as the method above seems to give the desired results. The feed rate was steady and about 2 ipm.

In the picture above you can just make out the curly chips produced, these are not much thicker than a hair.

Solution found

Results. The finish was looking good, no sign of cross-hatching and a regular appearance to the radial cutting. On subjecting it to the finger-nail test it passed with flying colours - not a single ridge or surface imperfection could be felt. As finished in the mill surface grinding was really unnecessary as it looked professional and perfectly flat. I may have just been lucky, but my first attempt at designing and using this cutter met all my criteria and so no further experiments have been made.   At last I can reply to my friend and tell him a solution has been found and await his next challenging email!