PDA

View Full Version : Rotary broaching



Michael G
27th Jan 2017, 08:28 PM
Rotary broaching is one of those processes that is interesting but buying the genuine Swiss or US made gear makes it horribly expensive, especially for a trial. In MEW 241 there was an article on doing it 'simply' which caught my eye and was the basis for today's experiment.
The end goal was to make a socket spanner like the one pictured - I have them in metric and in imperial but there are a couple of BSW fasteners on the J&S that I would really like a properly sized tool for.
366350
The way rotary broaching works is (in theory) simple. The broach is mounted slightly off square to a work piece and then pressure applied to force it into a preexisting hole. As the cutter also rotates effectively the pressure is applied to the cutting edge one bit at a time so the cutter works it's way in. The literature that I had previously read suggested that a practical limit for a hole made this way was 1.5 to 2D, but the hole could be blind.
This trial was semi-successful, in that it worked but not as well as I would have hoped.

The hex that I was trying to cut was 3/4" - A bit on the large side but I thought if I could do that everything else was easy. The cutter itself was made from some W1 steel, set in a dividing head with a down angle of 1 to 2 degrees (that is, so that the cutter narrows down it's length). The end is concave so that there is an edge to cut. The other part is a wobble shaft - a piece of rod for the drill chuck and with a centre drilled offset (this gives the wobbling action). The offset is calculated as the cutter length x sin 1 degree.
366355 366357
In between these two goes a greased ball bearing.
I started out with a piece of plastic (PE I think) and that worked quite nicely. Lathe speed was around 350rpm. Being the first time I'd done this I was a bit worried by some crunching noises deep into the 'cut', but it turned out this was just the cutter breaking through.
366356 366352
So far, so good. Next I tried doing the same thing with a bit of Al. Mounted up in the lathe as before, it didn't do anything - except push the ball into the back of the cutter.
366354
The issue was probably that I was trying to cut too deeply (turning a round into a square), so while it penetrated a mm or so, the material being pushed in front of the cutter was just too much. If you compare to a multi tooth push broach, they take off a lot less material per tooth. How much force was a good question and as I had a slightly formed blank and a cutter that would never be used for this again, I tried it in my 10t hydraulic press. I had to stop several times to lubricate, clear the tool and the press groaned at me in a slightly worrying way, but I did manage to get a hex made.
366351
One of my problems was the tool - I'd heated it with the gas torch and quenched it but not hot enough or long enough. Here is the broach after finishing the hole. Even though at the time I thought I had a uniform red colour across the cutting edges, you can see that a couple of the points have just collapsed. One or two are alright and I suspect they are the ones that were directly in front of the torch. in hindsight, it should have been soaked at temperature for a while. I was trying to avoid hardening all over and through but for something this size the MAPP torch just does not get it hot enough.
366353
In terms of making the wrench I wanted though, this is a dead end as I just can't get the force needed to get that sized hex. I may have to look at EDM perhaps and sink a hex in or maybe get a couple of WW sockets and do a bit of fabrication.

So in terms of lessons learnt -

Rotary broaching will work with a simple set up but is limited in size because of the forces involved
I need to heat treat tools properly if I expect them to do the job
My small dividing head is not concentric enough to produce a really symmetric tool
I may have to look at making multitooth broaches to produce larger shapes like these (at least for through holes)


Michael

Ueee
27th Jan 2017, 10:13 PM
I assume you mean W1 Michael? No a great choice for torch hardening as it will need a soak at the correct temp to get the best from the steel. Easier to use O-1/silver steel or something like 5160 or 8660.
I have noticed often that larger hex's and squares use a larger drill than the finished size. At the end of the day it is the corners of the tool doing the job not the middle of the flats. The larger hole means less material in the corners to have to remove.

Michael G
27th Jan 2017, 10:20 PM
Yes, W1 (now corrected). It was just what I had to hand, so yes there may have been materials out there that would be better.
Drilling the corners may be worth doing but very soon you drill out enough that filing then becomes an option...

Michael

Ueee
27th Jan 2017, 10:35 PM
I don't mean drilling out the corners but over drilling the center. So for your 3/4" socket drill 20 or 21mm.

Michael G
28th Jan 2017, 06:53 AM
I did drill out at 19.05; corner to corner the hex is 21.99. For a 1/4" square broach from memory the hole was 1/32" over the nominal for that reason. However, a hex is that much closer to a circle. If I drilled at 20 the land at each corner would only be 2.4mm and at 21 it is down to a mm. To go onto a well sized bit of hex stock that may be manageable but for a fastener that has copped 60 years of abuse it may be asking a little much.

Michael

franco
28th Jan 2017, 11:16 AM
Michael,

Should you be interested in an alternative method to achieve the same result, in the December/January 92 MEW there was an article on the Watts type drill, with a description of its function. The author describes his making of the three flute (square hole) drill with the calculations involved, and making the guide template for the drill.The square hole drill shape is based on the Reuleaux triangle.

A similar 5 flute drill can be made based on the shape of the Reuleaux pentagon for making hexagonal holes. I presume the calculations for the five flute drill would be similar to those given in the article for the three flute drill. Holes drilled with the square hole drill will have slightly rounded corners, but polygons with more than four sides will have sharp corners I believe.

The chuck holding the Watts drill must be able to float radially. There was a later article on making a chuck mounting which allowed this, which I did not copy. I do still have a copy of the article describing the making the square hole drill though, should you contemplate further experiments!

For a one-off item it would not be worth the effort, but it is an interesting concept.

I believe these are still available commercially, but they were not cheap last time I checked a few years ago.

Regards,

Frank.