Making Tapered Castellations in Aluminium

All, my son was given an old - and very knackered - Schumacher R/C car to rebuild as an engineering project:



He's stripped it, and we've been making various parts like turnbuckles, a layshaft, spring perches, bushes etc., and bought a few parts too:



We are however stuck with the side plates for the ball differentials, two of which are cracked. A hardened washer fits on the non-drive side. They are apparently very rare and difficult to find in undamaged condition. They were available in aluminium as an extra cost option.





The problem is the castellated output ends. We turned up a test piece, and transferred the tapered tooth profiles onto the plain face with paint. Then used a razor saw to make vertical cuts, followed by a wood chisel to bend and chop the segments out. The output stubs are 8mm diameter, so filing to clean up is very difficult.



Obviously this gave an appalling result, but since it's aluminium, and the corresponding castellations on the output stubs fit (kind of), it will work. The differential halves are forced togeher with a screw and nut through the entire assembly, so the output stubs cannot fall off:



Question, is, how can we make a better job of these castellations? I can only think to buy a thin diamond wire, place it at the bottom of the cuts, and try to saw the cut-out bits across their bases. It would be a simple milling operation if the segments weren't tapered.

Any ideas? Thanks in advance.

Should be easy enough to mill with the spigot stood vertically, cut one side of a "tooth" rotate and cut the other. Off out now but will post later if you don't get a better answer

Have you got a milling machine & rotary table?

Tony

The existing parts don't look like they've got a taper on the teeth. Machining straight-sided teeth is a simple operation involving a rotary table to machine the sides of each tooth in turn. If a taper really is needed, simply tilt the head of the mill, or the rotary table. Aluminium seems an odd choice of material. Steel would be better wearing and less prone to fatigue.

Andrew

Sorry guys I didn't explain that very well.

By tapered, I mean that the gap at the middle is smaller than the gap at the edge, ie the sides of each castellation - in plan - are radial lines from the centre:



The gap towards the middle of the part is approx. 1.6 mm wide. So if I had a 1.6 mm milling cutter, and made a cut from one side of the part to the other (again in plan), I'd get straight sided cuts, not wedge shapes.

I don't have a rotary table.

Edited By Dr_GMJN on 11/08/2021 13:43:27

How about a jeweller's piercing saw?

Looking at your photo of the castellations, would it not be possible to use a fine saw and cut straight across the end (e.g from 10 o'clock to 4 o'clock hope that makes sense), and similarly with the other flanks of the 'teeth', then use a small drill to remove the bulk of the metal from between the sawcuts, and finish off to size with a needle file?

.

For parts on that scale, with only a few divisions … It should be possible to cobble-together some sort of dividing contraption.

MichaelG.

As Jason and others say it is simple enough with a rotary table to cut the parallel paths then rotate the table and offset to cut each side of each castellation to the taper
If you don't have a rotary table then you need to make a fixture with a central spigot to mount the part. The part will need a witness mark with calculated witness marks around a circle scribed on the fixture: clamp, cut, rotate, clamp etc.
One of the easiest ways of creating those fixture marks is to draw it in CAD and print, cut out and stick to the fixture

The primary cross-cuts are made on their centre lines but for the tapers the mill is moved 1/2 the cutter width off the centre line and only run circumference towards centre for 4 teeth then swapped to 1/2 cutter width the other side of centre line and repeat.

pgk

Given that the sides of the wedges are radial lines all that is needed is to offset the cutter by half its diameter, away from the wedge, while indexing round each tooth with the rotary table and the wedge shapes will appear. Opposing sides of the wedges need equal, but opposite, offsets. Each cut only goes to the centre, not across the diameter.

I expect Jason will knock out a quick Alibre drawing to show the method.

Andrew

Would a thin (say 0,75mm or 1/32"  ) slitting saw pass below the southern edge of the eastern merlon and above the northern edge of the western one in the above pic?

If so, you could use the sawcuts as flank markers for end milling out the crenels. If you have a swivel vice you could presumably calculate and set the angles. You might have to finish with a Swiss file, Dremel or suchlike.

Edited By Mick B1 on 11/08/2021 14:50:19

Do they need to be tapered. I would guess from a power transmission point of view they will be well over engineered.

A suggestion might be to modify the existing stub axles to simplify things.

Mill the existing outboard pieces as per the bottom diagram to remove the red bits. It's four straight cuts (two "vertical, two "horizontal) with a small milling cutter.

Then make the new components to the top design. This time only two cuts straight across. Without dimensions, I don't know what size cutters you might need, but if the gaps between the teeth in the centre are 1.6mm then a 1mm cutter will probably suffice.

PS thanks for the question, solving your problem has made me think again about the design of  adog clutch I have made.

Edited By Peter Cook 6 on 11/08/2021 14:57:24

Can anyone recommend a rotary table then for an SX2P Mill?

Like I said I've not got one, but it would be a nice toy to play with and then it's a simple matter with a 1.5 mm diameter milling cutter.

Not specifically as I'm not familiar with the mill, but offer the following points for consideration:

• Get a bigger rotary table than you think you need, a lot of room can be taken up with clamps
• A vertical/horizontal mounting table is useful - I use vertical as much, or more, than horizontal
• My rotary table has a 1" parallel hole in the centre; to my mind that's more useful than a Morse taper, as it's simple to make location spigots and fixtures

To illustrate the points, here's a horizontal setup:

Note the central rod; simply stock bar with a 1" spigot turned on the end. The clamps under the gear only just fit on the table. And a vertical setup:

Andrew

Not having a rotary table is a bit constricting, but not insurmountable. In your position I would either make it from square stock and use the sides to do the indexing, against a vice stop and turn it down afterwards or make it out of round stock with a stub on the "back". Drill and/or bore a central hole in the aforementioned square stock, securing it with a grub screw or just superglue. Mount the square stock vertically in the mill vice and proceed to mill as described in previous posts.

A rotary table will be a useful addition to your workshop and I would suggest a 4"/100 mm one, preferably with division plates.

John

.

That's fine ... new toys are always fun ... ignore my suggestion

MichaelG.

You don't really need a rotary table for that, infact you could do it by just rotating the part once, either way a simple printed out paper disc and pointer would do.

First thing is to look at the part and asses how it can be machined, simply by rotating slightly you will see that the radial faces can be lined up with the X and Y planes.

A bit of CAD to draw out teh shape, I've assumed a 5mm hole and shown your 1.6mm gap. Which a 1.5mm dia cutter will just fit through so offset by half cutter dia and machine one face, probably go down in 0.5mm steps

This will leave your part looking like this

You can either leave it in the same position and make another cut in X coming in from the right and offset 0.75mm below the x axis then two cuts from top and bottom either side of teh Y axis. Or if tooling permits rotate 90deg and make a cut repeat twice. You will end up with 4 slots forming 4 of teh dog faces

Now rotate the work by the complimentry angle of the 37.05deg, offset below X axis and make a cut. 

Which and then do it three more times by rotating 90deg or cutting in teh other three axis and part will then look like this

Edited By JasonB on 11/08/2021 16:18:40

As a rule of thumb I would say go for a rotary table approx the siz eof the mills table front to back so maybe 5" diameter. My reason being that with limite dhead room on some of the bench mills you can soon run out of head room particularly if teh table is vertical.

This is how the SX2.7 is set up at the moment from rounding a couple of rod ends. 6mm dia cutter sticking out no more than 20mm from the collet. yet there is only 1" of vertical movement left in the head as the 150mm R/T (larger than table size) and 5" chuck stack up.

Also a little video of the progression of the cuts to form your dog clutch

Get a suitable size bit of square bar, machine spike on it to locate the bore of your castellated thing, then you can grip this in the machine vice and use it to index pretty accurately.

Or even easier make it from square bar, finally turn it round after you've done the teeth 

Edited By duncan webster on 11/08/2021 19:09:14

Could the mating parts be made from an epoxy putty, presssed into the existing parts after putting some grease on the original part?

Brian