An interesting way to make a multi tooth form cutter

[BCRider]

https://www.youtube.com/watch?v=oEzVEs306s4

I've long thought that a good way to make a multi tooth form cutter to mill some manner of shape needing a saw blade would be to index it and cut the form as shown here and then to cut the fronts so the relief is built in as shown. But I never thought about the idea of doing it with a jig like this that makes use of indexing holes.

While this one used a flat disc with holes I can see it working with a round bar tool where indexing divots are first drilled into the sides in the shank portion. Then set screws into the spot drilled starter do the indexing for the turning and other subsequent steps.

This quick sketch shows some sort of tool for something being held in the green holder in the lathe and the tool itself is spot drilled for three positions.

The jig would be used both in the lathe for the initial cuts as well as in the mill to make the gullet cuts. Making it from stock which can be held in a 5C collet seems like a good idea as well. Gives me more options for how to hold it in the mill.

I've often thought about making the clearance cuts this way for multi tooth cutters but never really thought on how best to ensure even indexing of the lobes. Clickspring's indexing pin and holes inspired me.

[nc5a]

Nice find. Clickspring makes things look so quick and easy. It's not until you try some of his machining operations that you learn otherwise but I still enjoy watching him do his thing.

[Doc Nickel]

One thing you gotta remember about Clickspring is he works tiny. Things like those shiny screws and whatnot, are truly itty-bitty. That cutter he just showed is about the size of a quarter, and the gear he made with it about the size of a dime.

That both makes it easier- because he's not having to remove huge amounts of metal- and at the same time harder simply due to the scale and thus the accuracy needed.

Dude knows his beans, no matter what.

Doc.

[Dan Dubeau]

I'm always impressed at his level of finishing. That's an area of mine that is severely lacking. The things I machine just never need that level of finishing, so I never bother. Not even sure I could if I wanted too.

Using his fixturing would be pretty straight forward with the button cutter method no?

[BCRider]

Button cutter method?

[Dan Dubeau]

Button cutter method?

The method of making the tooth form cutter by using 2 hardened steel "buttons" a certain diameter, and certain distance apart (depending on gear tooth needed) to cut an approximate involute profile. It's described in detail in Ivan Law's book on gear cutting, and been discussed a number of times here over the years.

[nickel-city-fab]

IIRC the method that he used is described briefly in Ivan Law's book, but it still takes quite a bit of skill and imagination to properly pull it off like Clickspring does. I seem to recall also that he teaches engineering, so that certainly helps.

That video was beautiful, in that it describes things clearly with nothing extra. I am sure the fantastic finishes he gets are due to having a very rigid setup compared to the size of the parts, and also to spending some time with the honing stones on his tools.

[Ringo]

I just browsed Law's book, I didnt see that cutter mandrel
Can someone help me out?
I understand Clickspring turning the eccentric, but, I cant see how the cutter back clearance reflects to the correct amount of eccentric.
That looks like the single button method he used, just a different way to hold the 'button' or tool

[J Tiers]

He offsets the tool he is making such that the back relief is a circular arc, but the rest of the cutter does not intersect it, so he can just spin the whole thing. I don't think he could do it with 8 edges on the tool, which may be why he makes it with 4 only. I'd have to look at it more carefully to prove that, though, so he could just consider that 4 is enough.

The usual "relieving system" just moves the cutting tool in and out with a cam, so that the tool being made spins on its center, and the cutting tool moves in and out to form the back relief. That can be done with as many cutting edges as are considered to be needed.

[J Tiers]

Actually, it can be done with any number, but the total relief gets smaller with the number. So he probably figured 4 was enough. For a cutter, the relief need not be that large, it just needs to not rub. That may depend on the exact shape of the cut profile. As long as there is "draft" on the profile, then theoretically, "any" amount of relief is OK so long as there is definite relief.

There is probably an optimum radius of relief for any size cutter.

Here is one with 8 edges (I do not show any profile, but it could be anything



And the geometry of the relief. The size of the hole and its location is arbitrary as long as it is outside the center hole of the tool being made. Here it just indicates the actual center of rotation. An actual hole at that location could be used to index, I suppose, if the center hole of the cutter is offset on an arbor similar to clickspring's. Probably a different location would be better.
The radius of the relief is not a number that is needed, the 0.625 is sufficient if 8 holes are drilled to use as index holes, and the offset is correct.



Proof of actual relief

[Paul Alciatore]

Basically he was using the button (circular cutter) method. He just did it one side at a time in stead of both sides at once. So he only needed one round lathe tool. And that tool was at an angle, so it was converted to an eclipse. That may be more suited for cutting cycloidal gears. I have wondered about using a round button or cutter like his for making gears with involute teeth. There are errors with the two button method because the buttons are round and the gear teeth are not. I think a carefully calculated ellipse could be a closer fit to the involute curve.

The other thing he was doing was to offset the cutter blank and then cutting circular arcs four times around it's outer edge. Then he cut a bit over half of each arc off with the milling cutter. That created the needed clearance with the offset arcs and the rake angle of the cutting edge with the milling cutter. This method has been used and explained before. It only works for a small number of teeth on the cutter you are making. If you tried to make a cutter with, say, 12 teeth, it would not work so well. Five or six teeth may be the practical limit. The four tooth cutter he was making is a good number for this technique.

I'm always impressed at his level of finishing. That's an area of mine that is severely lacking. The things I machine just never need that level of finishing, so I never bother. Not even sure I could if I wanted too.

Using his fixturing would be pretty straight forward with the button cutter method no?

[Paul Alciatore]

J Tiers said,

For a cutter, the relief need not be that large, it just needs to not rub. That may depend on the exact shape of the cut profile. As long as there is "draft" on the profile, then theoretically, "any" amount of relief is OK so long as there is definite relief.

That is not completely true. The angle of the relief is connected to the rate that a cutting tool can be feed into the stock. Generally speaking, lathe tools with relief angles around 5 degrees or even more will have enough relief for almost any feed rate that it makes sense to use. So we do not worry about it so much with them. We just grind them and happily use them to make chips.

But when using this method for making a multi-tooth cutter, as you increase the number of teeth, that relief angle decreases due to the geometry. With enough teeth, it could easily drop below one degree and the cutter may start to rub if it is fed into the work quickly enough. What happens is the angle of the ramp that the cutter is cutting into the stock becomes greater than the relief angle and the two will rub. The relief angle must be greater than the ramp angle of the cut. So, when the relief angle starts to become relatively small, you will have to slow down the feed rate.

Sure you could make a 20 or 30 tooth cutter with this technique, but it would be difficult to use.

[J Tiers]

Well, you can take a look at a gear cutter, or a hob sometime..... or look at one of the cutter grinding booklets put out by Norton and others years ago. The gear cutters have a curved relief, and can have a rather low angle of relief directly behind the edge.. Many cutters with ground relief have only 3 to 4 degrees relief, others may have 10, it just depends on the material being cut as to the optimum relief.

You will find that is not a practical limitation.

Yes, the amount of relief varies with the displacement of the circular relief center from the cutter center.

They are not really visible angles. At the radial face of the cutter, the relief angle is really the difference between the tangent line of the cutter OD at the radial face and the tangent line of the relief circle at the same point. There is no flat surface as with a cutter where the relief is a flat ground with the wheel. These sorts of cutters are ground on the radial face.

For instance, in my example cutter above, that difference is 7 1/2 degrees, even though there are 8 cutting faces. The relief is then 7.5 degrees, which is a medium amount of relief, as such things go.

The connection to feed rate is obviously that you cannot feed faster than the relief "falls away". That limit is where you would get a rub. But that is only a theoretical limit, and theory falls down here.

In actual practice, the feed rate is far less, and is expressed in thou per tooth, or "chip load". A rate of 5 thou per tooth is pretty large for many cutters, and takes a goodly amount of power. To illustrate the reason why the feed rate is low, and generally will not come close to "following" the relief, look at the example.

The illustrated cutter has a relief that puts its surface at 0.031" below the OD at the point where the cutaway for the next tooth starts (about halfway between tooth faces). To get into trouble with that would require a chipload of 1/16" per tooth, which is way beyond the capability of the tool or the machine it would be used on. The corresponding rate of advance would be a half inch per turn, and at an SFM of 100ft/minute, feed rate would be away up at 80 inches per minute on the 2.5" diameter cutter. Not gonna happen.

No, the true limit on these cutters is going to be the chipload and allowable SFM, you are not going to get in trouble at any reasonable, meaningful number of teeth And, of course, as the number of teeth goes up, even if the allowable chipload per tooth decreases somewhat, the allowable feed likely goes up as well.

The method of making a multi-tooth formed cutter looks like it will remain practical up to at least 10 or 12 teeth. Above that, you do not have a formed cutter at this size, you are starting to have a saw.😁

[Ringo]

I still don't get it.
Cutting the one-button or two-button form ,,,,,,I get that part.
I don't see the mandrel offsets of the elliptical portions.
there is a threaded hold down bolt, and a corresponding pin locations, on the lathe, this cuts the ellipse
but,
how did the pin locations align with mill setup to cut on the apex of the ellipse?

[Ringo]

I had to draw this with my morning coffee, I can see it, but, not the way clickspring did it.
he never relocated the hold down bolt, how he did that?
I understand my picture, but it moves the hold down bolt each facet
how he maintain the bolt to lathe center?