You could conceivably grind that way, traverse is only possible (at least, accurate and low-friction) when you have just about sparked out one area, and ultimate precision is determined by the rods, not the bearing surface - in fact it could flatten its own surface -  but it would be unsuitable for milling.<< Let me put it this way: I don't know what will work in practice, because I've never seen these things done. But, like a lot of HSMs, I'm sure, I've given some thought to better or easier ways to accomplish some of the functions of traditional machine tools. Someone probably will come up with some better ways of doing things from all of this thinking, and I'm the last one who would discourage it. However, it would be a good idea to try some of them, with models or simple examples, before speculating too much about it. I never got to try my laser-guided way grinder, but I did try making a concrete lathe bed. You learn a lot more by trying these things. I hope some of these ideas lead to simple experiments. Ed Huntress

Well, you took the concrete lathe a few steps farther than I did. One way you could seal those bearings is with epoxy. But there are other ways to do it. My design was _base_d on the idea that I'd use a commercial cartridge-type spindle, with a variable-speed Fuji DC motor and direct drive. I wanted high speed for small parts and extreme smoothness for fine finish. The spindle would cost more than my South Bend, but that was what the project idea was about. Anyway, I see that $400 for 3 cu. ft. figure again, and I have to remark that you're looking at some higher-class material than you need

When designing any type of machine or mechanical device, kinematic principles serve extremely well. The goal of the designer is to try to mimic these idealized motions as simply as possible. For instance, 2 points define a straight line in geometry. This is mimicked by the use of 2 ball bearings and a straight bar. Ground round stock is extremely straight because 3 surfaces cannot be rotated against each other and remain in full contact unless they are all circular. I think you're over-estimating the straightness of ground shafting. Even Thompson shaft, which is intended for use as linear ways and is held to very close tolerances on diameter cannot be counted on to be any straighter than about .002 inches/foot. This isn't to say it's not possible to produce exceptionally straight shafting. I have a Rathbone chucker with round ways (but not rolling bearings) that is very accurate, but not particularly rigid, despite being quite heavy. But standard ground shafting just isn't all that straight. You'd also need to be very careful to exclude chips fron the ways...a dovetail or flat way is self-clearing, a rolling bearing will want to climb over small obstructions rather than pushing them out of the way. The ball bearings on a round way are not as bad as a roller on a flat way in that regard, they will tend to push some stuff to the sides. And mounting round ways is not trivial, a moment connection at the ends of the shaft will tend to distort the way unless everything is carefully aligned. A simple support will sacrifice rigidity. I'm not trying to pooh-pooh the idea of rolling linear bearings, I've been designing machinery with them for over twenty years, and there's a 4 stack of IKO and THK catalogs on the floor behind me as I type. But they come with their own batch of problems that can't be ignored. Ned Simmons

Thanks, Ned. I, in my own stupid uneducated way, keep letting good sense interfere with judgment.  It is still my humble opinion that the reason machine tools are built as they are, and from gray iron, is because it is the best possible design and material for the job, considering all aspects of machine tool building (that includes economics).  To build a machine circumventing conventional design is simply changing the nature of problems encountered with machine tools. .  If one is building because he enjoys building, go for anything and everything.  If one is building because he is poor, it is my humble belief that when he is finished, he will be poorer and still need a machine when finished with a clever design.  Why re-invent the wheel in the name of economy? Harold I recieved the following email that is of interest from Richard Danzey. Thanks Richard! Steve, I've been following the precision hobby machine tool thread and must say you make a reasonable case for some expermental efforts. I'm a klutz and a dummy.  I just enjoy reading about such things.  Can't really work to high standards.  Lack proper patience, too. Please look up Stewart Platforms or maybe Stuart, I don't remember now, and see if you think there might be some contribution possible to this discussion from this direction.  They seem to be a relatively immature technology, but promising, in my estimation, and I'd like to hear your thoughts on them. I can't post to RCM, hence this email.  But put any response you might have on the forum, please.  Regards... danz <<<<< After a brief search I found numerous _link_s to work on the Stewart Platform. This appears to be Kinematic Design carried to a high degree of refinement. It is being used at NASA, Lawrence Livermore and a number of university engineering schools. Ingersol built a commercial model they called Hexapod. It was first reported in a paper by D. Stewart in 1965. He developed it for flight simulator platforms, but was aware of its machine tool potential. Stewart did not patent the idea, and it has been in the public domain since. Basically, the Stewart Platform is a radical departure from standard machine design. It mimicks all 6 degrees of kinematic motion. It's accuracy is superior to any standard milling table design. One university has even calculated the effects of heat distortion from the cutting tools and written algorithms for compensating. The limiting mechanical accuracy is in the actuator motors, but compensation for this has been calculated as well. The Stewart Platform only uses 5 percent as much _meta_l (steel is recommended) in its construction as a typical 3 axis milling table, is very much cheaper to make, is more accurate, plus the fact that it's a 6 axis milling table, something that traditional methods have never achieved. Here's a _link_ if you want to take a look: http://www.i-way.co.uk/~storrs/lme/LMEHexapodMachine.html An open mind is your most powerful tool. Steve