Turning question

[Art Ransom]

First commercial job will be 5.25" by 60" red oak posts and this could lead to 100 per month. One end has an ball tip. If I was turning manually I would leave a 1" spigot untill I had finished all turning and sanding. Then would turn end till I had a 3/8" spigot and then hand finish the tip. How would I do this on CNC?
The post has 4 transions from square to round, several 3/4" flats, several 1/4" v's and 2 convex 1/2" coves plus ball tip. Since I don't have a tool changer (yet) I was thinking if using a 1/2" ball and side cutting or a 1/2" up cut and side cutting taking taking .1" deep cuts. I would then move piece to manual lathe and finish the parts that the bit get to.
Suggestions please, rotation speed etc. I am just starting in the CNC world so I know darn little.
I am running Mach 3 with a PC 3.5 HP router

[ger21]

Since I really haven't seen many (any?) posts from cnc wood lathe users, it looks like you pretty much on your own as far as the learning curve goes.

Id play around with some scrap to determine rotation speed and depth of cuts. You should be abler to cut much deeper with a spiral bit than a ball end bit, but rotation speed will play a part in that.

As for the ball end, how about programming in 2 parts. First, do the rest of the column. Then, mount a steady rest and cut the ball complete. If, you can mount a steady rest fast enough to be productive that way.

If the part will need a lot of sanding to clean up anyway, I might think about roughing with a large straight bit, and then finish cut the whole thing with a V bit. Then put it on the manual lathe for sanding while the next one is being CNC'd. Just a thought. You're in for a bit of trial and error I think.

[Drawdog]

I am doing research which will lead to a design for a CNC woodworking lathe and from a woodworking book, a 5.25 in diameter hardwood should be turned at 1200 - 1300 RPM. Now that's for manual turning and certainly if you're using a live spindle it might call for some adjustment. Hapfo (a german company) makes a cnc lathe that can turn ball with a blade/chisel/tool that pivots to provide a left facing edge and a right facing edge, can do 90 degree edges, V's , curves all with one setup. www.hapfo.co.uk. I'd be interested in knowing (for the sake of research) what you end up doing. Thanks!

[Art Ransom]

Commercial CNC mills/lathes to handle stock the size I am working with weigh many tons because the are spinning the stock at a high RPM. Mine has a max RPM of 50 so the cutting tool has to rotate ie router. Mine weighs 600 lbs and cost less than 6K. If mine will finish a 24" by 10' column in 8 hours I will be very happy. When you reduce the price of a machine from $200,000 plus to $6,000 there are sacrifices to be made.

[Art Ransom]

geting close see http://turningaround.org/4_axis_mill.htm

[HuFlungDung]

Visualize a piece of tubing sitting on top of a sphere. Note that it makes perfect contact all round the inner edge of the tubing.

Now if you substitute some sort of trepanning type tool for the tubing, you can generate spheres by slowly turning the stock beneath the trepanning tool. Perhaps a rotabroach cutter would work , such is commonly used with the magnetic base drills to drilll larger diameter holes.

Caution: you would not be safe running a large diameter tool at high rpms such as produced by a router. The tool or pieces from it could shatter and fly with deadly consequences. A shield around the tool would be mandatory.

Anyway back to the subject, it is possible to use any size trepanning tool with a radius less than the radius of the sphere you wish to cut. What you need to do is determine the chordal distance that your chosen tool will span on a circle of a given diameter part. A relatively simple layout on a cad system will then make it possible for you to determine the angle and reposition of the tool head for each cut.

By cleverly choosing a trepanning tool with a radius exactly equal to the desired output sphere and positioned at 45 degrees to the work axis, you can generate a perfect hemisphere in one rotation of the work. Now obviously, the part cannot have a tailcenter engaged for a cut like this. So you would perhaps calculate a slightly smaller diameter trepanning tool that would leave a little spud on the part. You will need to calculate a new angle and entry position for the trepanning tool to create an undercut on the other half of the sphere, to whatever depth.

Taking a WAG, I suppose that if a trepanning tool has a diameter equal to the radius of the part, then it could theoretically cut a sphere in 3 passes, each one with the workhead indexed 60 degrees from the previous one. A small oversize allowance on the diameter of the trepanning tool would do a better blend in of the surface.

Some of the advantages of this method would be that you do not need a high rotational speed on the workpiece to achieve cutting action. A simple chain drive worm reduction would be all that is required on the lathe spindle. The surface speed of the tool would be constant wherever it was cutting.

I have not done any of the above to know the gritty problems that may arise, but I imagine a person would need to take lots of care in creating an indexable fixture to hold the spindle at accurate angles. The pivot center of this fixture would need to be accurately positioned beneath the workpiece at the exact XZ position of the spherical surface. Supposedly you can then simply rotate the work several times as you feed the tool in to depth as you knock the corners off the stock.