Rotary Broaching for Asymmetric Hole?

I've an 'ole to make in a valve 'andle boss, like this:-

I've a dozen to make, so I don't want to be filing out the radius and corners. People have suggested milling out a 12 mm wide slot in the bar the bosses are made from before parting, and then welding in a segment of a matching disc with a 14,4mm hole in it - but I've done no more than a couple of hours welding in 73 years.

So I'm wondering if it'd be easier to buy/make a rotary broach?

...and could it make such a hole?

...or does it require balanced material removal all around, as in regular polygons?

I think you may find it a difficult shape to broach, for one reason making the broach and also if the angular timing of the shape is important and if its only 12 to do think how your filing skills would improve!

Looking at a company that makes rotary broaches web site, it would seem that it may be possible. The problem may be removing enough metal before broaching to get the broach to cut easily. When broaching a square or any shape that can be scribed in a circle requires a pre drilled hole slightly bigger than the final across flats size. It would be easier to just make a simple broach to form the hole in a press than lathe.

I was thinking of a slot drill a bit over 12mm, fed over by about 1,2mm to elongate, and touch the rad inside. Still means there'll be more to take out where the rad corners out to the parallel faces than there is in the square corners.

A multistep broach looks a lot more complicated to make.

The angle can be set when welding the boss onto the handgrip part.

If a rectangular hole 12 x 13.2 was acceptable then rotary broaching in, say, a 13.5 diameter drilled hole could work. Theoretically the hole would have to be offset slightly to be on centre but hardly likely to matter for so small amount in a manual lever.

Drill a suitable round hole then take the rest out by EDM, spark eroder. I had an almost identical job in stainless worked like a charm. Noel.

And yet you are happy doing the bunch of fillet welds...

Is it me or are the dimensions wrong in the drawing? A 12 x 12 square isn't cut by an R7.2 arc as shown.

In his first post Mick mentions using bushes made from a rod with a slot milled in it. That's what I would do, and braze them.

Maybe the handle is designed with an asymmetrical socket with one curved side because a slotted bush of the required shape is easy to make and to fit into a round hole. The designer intended it to be fabricated with a bush? I could be completely wrong - maybe asymmetry is to stop the handle of a safety critical valve being fitted at the wrong angle, in which case a weak brazed bush is dodgy.

Dave

Why is a brazed bush going to be weak??

I don't think I'm gonna be doing it - but I think it'll be easier to lash up some sort of fixture to set the angle so that a welder of normal skill can add the handgrip, than to precision-weld in a segment with an inside radius. The angle needs to be maintained within a degree or three, because in some applications the valves are ganged together in some way I haven't seen, and need to operate together.

The dimensions are basically a clearance fit on the milled and turned shank of the valve shaft. I don't know whether brazing would be strong enough - the welded-radius suggestion came from a very experienced locomotive fitter, but I'd rather make it in one piece as that removes any question, and looks like the original handle I'm wanting to copy.

Arranging spark-eroding would certainly be expensive and probably be more exhausting than filing/shaping out a round hole.

Thanks for the suggestions, fellas, but I'm wondering whether anybody actually knows, so likely I'm gonna hafta try it...

Edited By Mick B1 on 01/06/2023 22:28:15

Failed SOD's safety test, you can put it on back to front and get it 90 degrees out of position !

Alternative to the attractive spark eroded hole.

Drill the boss 14.4 diameter thru from the back then end mil 25 diameter from the front but not right thru, say 15 deep (assuming boss is about 25 thick). Make a plug 25 dia x 15 thick with keyway 12 wide. Braze in place. Minor changes to the spindle could make incorrect assembly impossible as long as the spindles have not already been made. Bush would be robust I believe.

It's all about how much time you have qnd what the jobs worth? Having built a spark eroder it was a good way to prove it's worth ! Noel.

Not that much work to do one with a file. Even your dozen will take less time than finding an alternative.

Drill to 12mm then use a good sized file such as 10" round file to rough out the radiused end then a smaller half-round file to finish it off in the corners. A 10" three-sided file will make quick work of the other two square corners, finishing off with the small half-round to get the very corner point sharp.

Edited By Hopper on 01/06/2023 23:46:00

Plunge into the centre with a 12mm milling cutter, move the table by 3.22mm to remove the majority of the material in the radiused area.

Using a 2mm diameter milling cutter to square off the hole, travelling 10mm across the bottom and 10.686mm along the sides should give you something like in the picture.

Finish the corners with a file.

Martin.

Edited By blowlamp on 02/06/2023 00:40:59

EDM wirecut is getting to be very common these days, and the price is about as low as it will get too.

You can supply them with the 12 blanks , with or without a centre hole, won't make much difference.

The outside shape can be cut and then set up and the inside shape cut as well. They can get the cutting very accurate for alot more money, but done within a couple hundredths should not be too expensive.

I am making some bearing rings, for a bearing size that is no longer available. They can cut the inner race blank and outer race blank from the same stock. I am leaving .5mm diameter inside and outside to clean up, as I do not want any stress from the wircutting in the parts. The other advantage is that the material can already be heat treated before being cut too.

Wire cutting , with 3d printing , is making alot of changes to how things are being made. Especially the electron beam printing capabilities.

All this aside, if it is actuating a valve, I am sure that the parts could be made as section with a slot through it, and the other piece with the radius on it, and then brazed in place , with a simple fixture to keep the distance correct. If the heat from brazing or soldering is too much, it could be pinned and glued in place with a high temp glue, of which there are many brands now, that can sustain a temp of over 250c.

The outer bump or cam part , could also be made from a segment that is then cut into the shape, and then also glued and pinned in place.

There are many ways to make something, if it is not stressed to the maximum or is not having to always operate at a very high temp.

It depends how you read the drawing, if just going by th e12, 6 and 7.2 dimensions you get one shape. If you assume the other two parallel sides are 12mm you get another shape. I would want to find out which is correct before making the hole as methods may need to be altered.

If the valve shank was "milled & turned" as Mick says then then that raises questions as neither could be turned without the tool taking the corners off the "square" part

Well, if I'm remembering the conventions I once used as a Tool Designer, that 7,2 rad is thrown from the CL of the boss,and there's only one way to read it.

You're quite correct that the milled flats on the mating shaft leave a narrow line of turned diameter on the 'corners', and that might leave theoretical scope for milling the hole (as in eg. Blowlamps drawing), but with the boss being 22mm thick, trying it with a 2mm dia. endmill doesn't look an attractive proposition.

I've shaped out keyways and square corners before using the carriage handwheel, but with a dozen to do I'm imagining the blisters already...

By the same token you could say the corners of the "square" extend to the two ctr lines on the drawing which they won't if the 7.2 is from where the two lines cross

Yeah, but most drafting sheet blanks have 'Do Not Scale' pre-printed on 'em, so the dimensions and conventions are supposed to take precedence over mere appearance!

Perhaps I should've put an 'NTS' by the rad dimension, because it clearly is. But hey, I'm a volunteer, and I do these drawings so I don't have to keep measuring every part the railway ask me to copy.

And I'm no nearer an answer about whether to try rotary broaching.

Edited By Mick B1 on 03/06/2023 16:43:36

At 18mm deep you may be beyond rotary broaching, looking at the Hemmingway offering it does 8mm major dia x 7mm deep so even if scaled up to 12mm dia you may not get much deeper than 10mm. I suppose as the hole gets deeper there is less room (angle) for the tool to wobble.