I think you're worrying a little too much about the stiffness
In what way? Do I have too many rails, too many carriages, rails too far apart? I am more concerned about the stiffness of the granite structure. The reason I was thinking about having the spindle facing away was to let me have bigger granite side pieces. I do worry about those parallelograming because of high accelerations of the X axis. Although, come to think of it, the X axis is quite low mass.
where do you plan on mounting the linear encoders?
That I am still not sure about. The X and Y axes are no problem I think, but for the Z axis, the easiest thing would be to mount it next to one of the motors. Assuming the X axis can't rotate about the Y axis much (because the rails are stiff enough), this shouldn't be a problem.
we took great pains to position the linear motor coil at the CG of the moving mass for each axis.
That's exactly what I'm trying to do by using pairs of motors. So that the centre of the force is at the centre of the gantry. I think that the Y and Z axes are well balanced, but I can't work out how to get the X axis balanced in the same way, without using a crazy number of rails.
I would reccomend "way covers" of some type, especially if you intend on cutting ferrous materials!
Yes, definitely. I had a really nice idea for the Y axis covers, to completely prevent any swarf getting to the motors, rails and encoders. I'll see if I can work out something similar for the other axes.
You will also need some sort of solid hardstops for each axis, as if the motors run-away, they are an impressive slide-hammer!
That's something that really scares me. I can just imagine the machine smashing itself to pieces the first time I turn it on. I'm going to have to carefully calculate the minimum stopping distances it can cope with, and put in some buffers and cut-out switches.
Keep an inch or two between the encoders and the magnet tracks, as if you are planning on renishaw encoders, the magnet tracks can sometimes trip the limit switches or the reference marks.
That's something I was wondering about. Would it be better to use magnetic or optical encoders? The magnetic ones are less prone to dust, but might be thrown by the powerful magnets.
Hugo
Create free flowing PCBs: http://www.liquidpcb.org