Rotary Broach Tool Design

[excalibus]

Hiya people,

The other, after reading some information on rotary broach tooling and seeing a rotary broach in action, I was having my lunch at work and thought I would try my hand at designing a rotary broach tool.

In the initital rough sketch i had set out the design with a fixed MT2 Arbor (my lathe when it arrives next weekend has a MT2 on the tailstock). I had initially designed it with the arbout thread on the 1 degree angle which would put the whole tool off skew - bad idea I thought lol...

Last night I sat down and decided to redo the design after seeing that a local SIEG reseller has MT2 Arbors threaded to take a drill chuck for sale on ebay for a very decent price.

So using that as the basis for the tooling, I designed a piece that will screw onto the arbour, has a rear thrust bearing, 2 centering bearings and the tool holder held in place with a face plate.

The centre hole where the Rotary Broach cutter will be placed is drilled at a 91/89 degree angle perpendicular to the face.

I am pleased to share the drawing with you guys and feel free to comment and make suggestions on how to improve it

A couple of notes on the design, the bearing retainer plate will contain 6 or 8 screws to hold it onto the tooling body though the design does not show that clearly as I did not do a end on drawing of the face plate.

The design contains NO dimensions. it is an intial design with dimensions to be set by the available bearings, screws and the arbor (I have not purchased the arbor off ebay and so do not really know the dimensions of its backplate or the depth of its threaded face)

Also just for the record, the name Scallipus - is me... It is a name I use for a lot of gaming forums as there is another Excalibus out there on those forums lol and is the name I signed the designs with.

 

[polygon]

Nice sketch, keep us up to date on your progress.

 

[rkepler]

The trick with these is that with the tool being non-square to the work the length of the tool will move the end from center, and center is where the tool has to be. So you need to include some provision to move the tool vertically and horizontally to center it on the work.

Another thing to keep in mind is that rotary broaching is a high pressure operation. Larger tools are going to be pretty hard on the screw & nut in the smaller lathe's tailstock quill. For something like the smaller Atlas (I've got a 12x36) I would be tempted to make a quill with a lot of thread and very little travel to try and spread that force out some. For onsey-twosey it wouldn't matter much but if you were making a decent run of parts it will start to add up.

 

[Ken I]

Here's one in ACAD.dwg - zipped that I posted the other day.

Ken

View attachment WOBBLY.zip

 

[excalibus]

thanks for that drawing ken...

I am wondering, seeing the tooltip has to be centred etc, would it be better to make it so the rotary broach holder is attached to the tool post instead of the tailstock? this would allow centring on the workpiece a lot easier I imagine... especially if setup to be used with a QCTP...

Anyone have any thoughts on that?

 

[Ken I]

Excalibus, you will notice that there are elongated holes to allow the tool to "float".

This is pretty much the design I've used to produce hundreds of thousands of holes - years of N.C. lathe production of VW Golf (Rabbit, Jetta etc) McPherson strut rods (7mm AF Allen key hole).

To set up, drill the hole and chamfer the lead in - loosen the two securing screws - bring it into the predrilled hole - float it into place under slow rotation and tighten the bolts while maintaining force - job done.

When you feed in from a tailstock you will notice a "cogging" sensation this is the broach advancing during the "slack" portion of the broaching (on say a hexagon) - the torque then generates the thrust as it turns to a corner - during the high thrust portion you aren't actually advancing the tool. Thus the broaching seems to take a lot less force than you might imagine. You are sneaking in during the slack periods and the torque is generating the thrust in between - the thrust against the tailstock is however very real - just you don't have to apply it. - Sorry its rather arkward to explain.

Even so the thrust is a fraction of simply trying to force the broach straight in.

Again for this reason - since the tailstock is non-reversible (ie the force generated can't unwind the handle) the tailstock works better than the carriage which will tend to push back - unless you are using power feed.

The drilled hole needs to be just a whisker larger than the broach - the difference in force between a whisker bigger and a whisker smaller is very large.

For standard Hex sizes such as 6mm 8mm etc. use a size drill ie 6 or 8 but grind one cutting edge slightly longer - this causes the drill to cut oversize like 8.05 - helps a lot.

I used to broach the 7mm holes in EN8 at 900rpm @ 0.05mm feed - FYI.

The broach needs to have slightly more clearance than the incline angle - (typically only ¼° more) - otherwise it will snap off - but don't use too much clearance as the broaching can start to spiral on you - particularly if your thrust bearing is going off.

This broaching principal is also used to create the external "knurled" finishes of longitudinal grooves (the circulars are turned in) such as you see on camera focus rings etc. External broaches are fairly expensive to make - typically via wire cut EDM.

Hope this helps.

This process is sometimes called a "wobbly broach" - the name's a misnomer as it doesn't really wobble (although relative to the work it does wobble) - originally a Westinghouse patent I believe.

Ken

 

[excalibus]

Ken,
Thank you very much for that very informative description of how the broaching tool actually works, you have indeed taught me something I know the angle of the broach itself has to have clearance behind the actual tool tip and I figured this was to allow for the waste metal that is removed.

I can see why the additional size on the hole would benefit as there would be less material to remove in the 'corners' which would mean less force placed on the tailstock. Other than this, the actual force on the tailstock is the same as the force that the broach tool generates as it advances due to the torque forces placed on the tool.
Assuming each force in one direction has an equal force in the opposite direction...


Just to make sure I do fully understand...

The broach tool itself should taper back around 1.25% (1% + 1/4%) to allow clearance for the metal being cut.
The tool holder should allow the broach tool to wobble a little bit until it is aligned on the rotating part and then locked into place.
The Tailstock must be able to at least withstand the same forces placed on the tool or it will be damaged...


Just making sure I fully understand you is all ken Thanks again for your help it is appreciated...

 

[Ken I]

Excalibus - there is no waste material - it is pushed ahead of the broach and remains in the bottom of the hole - you drill the hole deeper than the broached hex by 2-3mm to accomodate this.

If you are broaching a hexagon with a slightly oversize hole the swarf is going to separate into 6 strands - with an undersize hole you are trying to turn a "tube" insideout - much greater force required.

To remove the "swarf" you re-enter the pilot drill after broaching - generally you simply leave the material in the bottom of the hole.

Don't worry about the forces on the tailstock quill - unless you find you are having to apply undue force with the handwheel.

The tool "floats" by releasing the clamping screws - the face of the broach is ultimately going to be on the same centreline as the work (the ass end of the broach obviously isn't).
Because the length of the broach may vary from broach to broach - you would have to readjust the tool holder. Once on centre the holder should be securly bolted down to the shank.

Clearance - you obviously must match the tool inclination otherwise its going to bind and via repeated flexing will be snapped off (this is what would happen if you tried to use a straight allen key as a broaching tool) - beyond that you just need a bit more (¼°) just "to be sure".
Too much clearance and you increace the propensity for the broached hole to spiral on you.
The original toolholder I used was supplied by Index in Germany who recomended dismantling and greasing of the thrust bearing every day - the hell with that - I ran it and relied on the coolant - until it broke - bearing failure starts to induce spiraling in the broached hole - normally we would pick this up because the go gauge would enter the hole easilly enough - but wouldn't reach the bottom - time to replace the bearings (maybe 10 - 20 000 parts).

The following post's tips are spot on. The undercut allows the chips to break out but can be ommitted (difficult on small holes).

Regards,
Ken

 

[polygon]

Some hole prep tips:

 

[Ken I]

This from elsewhere on the forum.

Lms actually sells some rotary Broaching bits and holders. Unfortunately they are out of stock right now. Here is the link. http://www.littlemachineshop.com/products/product_view.php?ProductID=3772&category=

The LMS broach has no "float" adjustment and is therefore made for one length of broach only - certainly simplifies it and mild resharpening shouldn't throw it off (at least not before the broach is indersize).

I still prefer the "float" feature.

Ken

 

[polygon]

Ken, that's a fantastic explanation of rotary broaching.
I'd also like to point out that the same tool works in a mill. This is a 'no float' holder.
[ame]http://www.youtube.com/watch?v=uTsPXFy7Xx0[/ame]

 

[Ken I]

Polygon, you're obviously "in the game" - I presume you supply those wondrous external polygon milling attatchments as well ?

When you use it in a mill, you can see how the term "wobbly broach" came about.

When used in a lathe you can't see that action - but its still there relative to the work.

I lost count of the number of perplexed visitors who asked me "how did you turn that hexagon hole ?"

Ken

 

[excalibus]

thanks for all the input it is really appreciated

 

[GRAYHIL]

Am I missing something?
If the wobbler is held in the tool post at centre height (ie make a square holder) but with the angle of 1degree at the side not the bottom then the crosslide can be used to set the toolbit to centre regardless of the tool length.
Doing it this way will eliminate the need for the cross slots and bolts and the housing and holder can be made as one?
Graham

 

[GRAYHIL]

More thoughts.
Make the wobbler with no angle, and set the topslide at 1degree when holding it to do the hexagon.
Graham

 

[GRAYHIL]

Something like this
Graham

 

[Ken I]

Graham, yes you can make it straight - but then you have to mount the assembly on your toolpost at 1° off true from the machine centreline. (Alternatively "swing" your toolpost 1°)

This may sound simple but 1° is a bit subtle to "eyeball" although you can obviously clock for the correct runout for 1°

But you are correct in that this will allow you to use your cross slide and toolpost height adjustment to get it on centre.

Caution on setting the angle - get it too shallow and you lose the action, get it too steep and you'll break the broach.
Lesson - clock the 1° accurately.

Set up from the cross slide is always going to be a lot more fiddly than from the tailstock.

When working from the cross slide, lock the carriage and use the compound slide to advance the tool - the non-revesibility of the screw thread is easier to generate the force than the carriage handwheel (but either will work).

If you do make such a holder put a 1° angle on one face of the shank - that way you can mount it against a known square toolpost position - alternatively mount the 1° side facing out for clocking - just a thought.

Ken

 

[GRAYHIL]

Ken
Feed in from the carriage will be at 90deg to chuck with topslide at 1 degree offset and a a parallel holder.
Feed in on the topslide will be one degree out unless the holder is machined with a one degree offset and the topslide set parallel as you suggest.
I think I see what you mean!
Graham

 

[Ken I]

Graham,

The face of the broach is on the machine axis centreline.

The broach's axis of rotation is 1° off the machine centreline.

The direction of feed is paralell to the machine centreline.

So you cant swing the compound slide 1° and then use that to feed with (although you can still use the carriage).

View attachment DUMP.zip

 

[GRAYHIL]

Ken
Agreed
Graham