'Boning up' on 3D milling

[lloydsp]

I've got a complex oval bowl to make from a rare hardwood -- the Sunuke (heartless) version of Iso No Ke.  I cannot afford to go out and buy an oval lathe-turning apparatus for a job I might do twice in a decade, and nobody around me has one.  So...

Fortunately, the wood is very-conducive to milling and routing.  It's dense, almost without any visible grain (except for elegant colorations), and doesn't splinter.  In fact, it may be finished with only fine sanding and buffing, without even wax!

So, I've got to re-learn everything I already forgot about how to both draw and machine 3D object/s.  The wood is terribly expensive, so I'll be doing a couple of dummy runs in cheaper woods, before I'll commit to the Iso No Ke.

This is for the base of a very attractive desk ornament, and the very plain base must be as elegant-looking in its simplicity and contours as the ornament is complex-looking.

[Bubba]

Oh, wow. What is the overall size? It looks like 4th axis type of work..

[lloydsp]

I'm not machining the spinner, just a very elegant-looking dark-red/brown wooden base for it.  But I don't want it to be circular or retangular, for sake of being too plain.  I wish to make it in an oval shape with an ellipsoid top profile and a tiny round flat spot on which the spinner sits.

The ball is chromed steel.  The helical frame is aluminum, about 30mm in diameter by about 60mm tall, and rotating on a silent, invisible roller bearing in the bottom.  When turned by hand, the illusion is that of the ball being suspended in space while the helical frame rotates around it.

I have contrived and tested a means to turn the spinner by magnetic eddy currents, without any visible means of propulsion.  All the works/electronics are contained in the base, with no openings, switches, or visible ports.  It appears to be a hunk of solid wood with a glued-on felt bottom.  But when the spinner is placed on top, it turns by itself.

As designed, there's a hidden gravity-operated switch inside the housing, so it may be turned on and off by the operator.  On C-size alkaline batteries, it will run for about 220 hours continously before the batteries must be replaced.

Lloyd

[dave benson]

For doing the drawing, I'd use Freecad to produce the stl or stp file, I don't have a 4th axis
although this would be my first choice.

The 3d mops using a ballnose tool are slow and for a piece that small to pick up the subtle curvature
on the spiral arm's top and bottom faces would need a fairly small tool.

If you are good at working with wood, I'm not, I would leave a little bit of extra roughing clearance
 to support the spiral and finish by hand, If you require finish off the machine then It'll take a while.

To speed things up I'd use roughing passes with as large a tool to remove the bulk material as practical, the
work being so small, I think on flipping the work piece over that the spiral would
be unsupported  so would back off the aggressiveness of the cuts for that side.
This will help with chatter and the work piece moving around under the cutting forces.

If this was made out of metal, I would calculate my feeds and speeds run a test piece and adjust the
spindle speed up or down so as not to excite the spiral into oscillation.

Dave

[lloydsp]

I'll repeat:  I AM NOT MAKING THAT SPIRAL SPINNER.

I'm making a smooth, oval, elipsoidal-top wooden base upon which the spinner will sit.  The spinner is a commercial item.

The base will be in the approximate shape of a slice of an American football, with the slice taken parallel to the long axis of the ball, and not centered -- thinner than it would be if it were half of the ball.

Lloyd

[Bob La Londe]

Lloyd, you already know how to make cylindroid objects in ViaCAD and you already own a copy of ViaCAD. 

The only thing I might suggest is if exporting to STL make sure you set your surface deviation very low to eliminate the appearance of facets.  Then in CB remember that both the resolution and stepover settings impact resolution when 2.5D (3D) machining to the surface of a mesh. 

I can't picture what you are making, but it sounds like a simple bowl or vase with a flat top to set the spinner on.  I might be tempted to rough very close in CNC, and then finish by hand on my wood lathe.  I have made aluminum 3D parts that way, except using my metal lathe.  I've also roughed with a 4th axis, and still hand finished on the lathe. 

[Bob La Londe]

If it's as simple as the objects I am picturing you can probably use 2D tools to sketch the profile, and then lathe the object.  Now that I think about it, you can do that in CamBam directly.  I sometimes use those tools to make casting sprues/funnels in mold jobs. 

[Bob La Londe]

The other bug to be aware of is crossover distance.  You can get bad tool paths if cross over distance is set to less than stepover.  Since I am almost always 3D finishing an irregular shape I can still get full retraction on (most) crossovers to avoid crossover tool marks by setting max crossover to a very small fraction above stepover. 


Example: 
Stepover 0.03
Crossover 0.030001


Another way is to design a lead in and lead out shelf or area in your geometry so crossovers act normally, but are off the actual part.  This can be helpful for a 3D positive that may be quite tall to avoid alll that extra vertical rapid travel.   

[lloydsp]

Bob,
It's not round.  It has a tiny (1.5" dia.) round 'seat' on top to help properly locate the spinner above the magnetic drive.  The bottom is about a 2:1.3 oval, 6" long-axis.  About 1.5" high from desk surface to the seat.  The vertical profiles are both elipsoid, with different elipses in X and Y, due to the oval base.

It can't be lathed without an oval-turning gearbox, which is out of the question for a one-off.

I'm still 'hacking' with the drive electronics right now.  I found a couple of tricks to save power and alter the speed.  Parts come tomorrow to test those ideas.  So I won't even get to plotting the final base dimensions until I'm sure all-what hardware needs to fit inside.

Lloyd

[Bob La Londe]

Not knowing exactly what you are making you could do it in two pieces and glue them together.  I've found very little I can't machine, fabricate, cobble, weld, or glue if I really want to.  You can't violate the laws of physics, but withing those constraints there is a large subset of things you can do without buying more machinery.  I've been known to hand dress parts with a file to get them just right. 

[dh42]

Hello

Something like on the picture ?

++
David

[lloydsp]

Yes, David!  The primary difference is the height of that 'slice' will be taller, such that the 'flat' on top will diminish to a mere 30mm diameter (nearly circular) spot, while the long axis of the bottom will remain about 155mm.  The width, front-to-back will be about 60mm.

Bob, it very-well could be made in two pieces, but part of the goal is to make this look and feel like a solid block of wood, with no visible seams or joints of any kind.  There will be a set-in, removable bottom plate planed from a 'self slice' of the same piece of wood (sawn from the same face of the blank), so the grains nearly-exactly match on the plate and body, and the thin visible joint of that 'inset' bottom plate will be covered with a narrow ring of felt, as IF only a cushion, but making it appear as if the entire thing is one piece.  The bottom will be held in by tiny magnets, so it can be removed, but with no visible means of attachment.

The mechanism inside is totally silent, and will be turned on and off by an arc-shaped (bi-stable) mercury switch by tipping the unit side-to-side, so there are no visible means of controlling it.

It's a fussy, audacious attempt at making the illusion of perpetual motion perfect!

[lloydsp]

David,
May I ask what software you used to create that drawing?  And does it provide the surface meshes or other information necessary to do a waterline machining of the contour?

Thanks,
Lloyd

[dh42]

Hello

SolidWorks

And yes the soft can export to many formats like STL, STEP, etc ..

++
David

[dh42]

Re

not sure to understand the dims you give me (the height is missing ?)

have a look on the pictures.

++
David