Virtual Ways?

A while back I read a reference by Don Lancaster (guru's lair;
formatting link

about "Virtual Ways" for a milling machine that didn't have real ways.
The reference is as follows:
"Described in the August 15, 1994 issue of Design News. Take a base plate
and place six ball screws or linear steppers uniformly around a circle placed
on it. Now lean each ball screw in 45 degrees or so and attach them to a
smaller upper head plate. The upper head plate can easily assume all of the
normal lathe or mill motions. But there are no precision sliding contacts
anywhere! All bearing are plain old round ones. All forces are pure tension
or compression with zero side loadings."
The description is clear enough, but I've got two problems. I don't see
how there is no side loading. Agreed that MOST of the force would run down
the ball screws, but to say "ZERO side loadings" seems a stretch.
Second, how would this become a practicle machine? If it can assume all the
lathe and mill motions, then this would be a great machine for the home/hobby
shop tight on space.
I've google'd on "virtual ways" and some other terms, but can't
seem to get any other infomation or a picture of what such a
machine would look like, let alone any information on actually
building one.
So, what am I missing here? Thoughts and references are welcome.
Reply to
Kevin C.
Loading thread data ...
Zero side loads is correct. You cannot have side loads on a beam supported at both ends by ball and socket joints. (neglecting gravity and friction)
Reply to
Ian Stirling
Rigidity would be a big issue, and I don't know how you're going to make a set of six rigid lead-screw assemblies without some pretty precise machineing on sliding contacts.
None the less it's a good idea. Check out the Polytec PI
formatting link
website for an animated picture.
Reply to
Tim Wescott
formatting link
The machine has been built. It's called the Hexapod, and Ingersoll Milling fiddled with it for years. I don't know the current status but I hear someone has it under "continuing development."
The biggest problem has been software. That's one hell of a lot of number-crunching, and some aspect of it -- perhaps compensation for elasticity -- has given the designers fits.
I saw an early version at IMTS years ago. Although I was still a big tension-structure/Bucky Fuller fan at the time, and I love to analyze space frame structures (hey, everybody needs a hobby), I've always thought this one was a loser. It always looked to me as if Young's Modulus was going to defeat any advantages the basic design may have.
Oh, as someone else pointed out, there are no side loads. Think tetrahedron/geodesic dome: it's all tension and compression. This is only slightly compromised by the platform at the top, which introduces very slight side loads because the elements don't intersect at a point.
Reply to
Ed Huntress
Also commonly known as a "Stewart Platform"
A google search with keywords "Stewart Platform" along with "Hexapod" will turn up quite a bit of information.
formatting link

Have fun............
Reply to
Try googling "hexapod" and you will get scads of references. Fascinating concept. Think like the thing that moves an aircraft flight simulator platform. Hope that helps, -AL
Reply to
It is called a Stewart platform. It is simple in concept, but calculating the lengths of the actuators for any arbitrary position is not easy, especially if you are trying to follow a line or circle.
Reply to
Charles Erskine (Charles Erskine) wrote in news:
LinuxEMC has Hexapod support, don't know how good it is though..
Reply to
Hey Ed and ??,
I'm pretty sure there has been a working demo of something in the R&D scale of things at NAMES for a couple of years now. I don't recall it making any chips though, but then again lots of stuff there is just a static demo, so don't take my word for it.
Take care.
Brian Lawson, Bothwell, Ontario. XXXXXXXXXXXXXXXXX
Reply to
Brian Lawson
Uhhhh Steve???
Not sure about NAMES.
But I did happen upon this site this evening featuring a very nice looking 5 axis cnc unit designed to retrofit onto a bridgeport or other milling machine:
formatting link

A very cool video of it in action can be accessed via this hotlink :
formatting link

Reply to
Sounds like you're describing a flight simulator motion base like this one:
formatting link
Best Regards, Keith Marshall
"I'm not grown up enough to be so old!"
formatting link
Reply to
Keith Marshall
Try looking for Stewart Platform or hexapod. These things have been used for full-motion flight simulators for 40+ years, and as machining centers for at least 25.
Not all bearings are "plain old round ones". It needs 12 ball joints, one at each end of each extendable strut.
Reply to
Jon Elson
Try searching for "kinematic machines", also "Stewart machine". and "Hexapod". It is a clever idea, which yields a very precise, very rigid machine (with the possibility of dynamic compensation) without a lot of weight or precision ground surfaces. However, it takes a lot of realtime computer power, and strong rapid acting servos, to pull it off.
Often, all 12 actuators have to move at the same time, each at a precise rate, in order to accomplish the desired tool motion. Of course the price of computer power continues to decline rapidly, while the price of precision machined cast iron does not.
So there's reason to believe that this sort of concept, which substitutes computer and electric power for massive high precision cast iron structures, will eventually be the cheapest (and perhaps best) way to make very versatile and very precise CNC machine tools that are also light and handy, even portable.
Reply to
Gary Coffman

Site Timeline

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.