I'm looking at an old rackmount cabinet I bought to be a "techie mans 'curio' cabinet". (For the collection of all those toys I've collected. Firiona Vie rubber figure, Star Wars BustUps, etc.)
I'm thinking it would make an excellent case, and possibly partial chassis for a cartesian robot, with room for one of my old desktop computers to spare.
As for upgradability. My opinion is the thing should switch heads, as having all the possible heads hanging of the manipulator arm would quickly limit the working area. If the one that hangs the furthest to the right, can't go past point 260, and the pone furthest to the left can't pass 20, then the maximum width becomes 240, when the table originally measured 280 units. (The numbers listed are arbitrary values.)
As for speed...
If two heads can both work at the same time, on a single pass, but the one head has to be swapped out to make a second pass to do the same thing, than making a very dense object takes slightly longer than 2x, that extra time being the tool change. If the object tends to be less than very dense, on any given layer, it would tend to be less than twice as long per layer.
However, if the thing can work unsupervised, I can, theoretically in the future, tell it to fab me an MP3 player on my way out the door, (or on my way to bed,) and my time isn't compromised, so I am only inconvenienced by not having whatever it is NOW. Even multiple toolchanges per layer, (plastic, filler, circuit trace, (probably a conductive paint,) pigment, limited machining, etc,) wouldn't bother me, if I don't have to babysit the machine while it is working.
Oh, unless the head is given a fourth, and fifth, axis, painting such a thing would probably be handled by not one head, but five. One down, one each for the cardinal directions...maybe four, if each head covers a one hundred twenty degree arc.
Edit for page 2.
I see the need to "standardize" two things, and it may be solvable by another solution.
One, I think is. The dimensions are given in a real-world unit. SI, granted, (and why can you Brits tolerate using a measurement system invented in France, when we, your former colony, are still getting along happily with your old system, anyway?)
The second is material palette. Right now, items are being thrown around that use a single homogeneous material, with a possible second material to fill the voids. With just two materials, it doesn't matter if you're forming something out of PET or tin, the results are essentially the same.
However, later down the line, when one party keeps plastic as material 1, and another party adopts an alloy as material one, there will be less transferability of their designs as those designs will have to be translated to be formed on a capable machine, but one that belongs to a different material "school".
I see three ways to handle that. One, leave it alone, the problem will develop, but maybe automatic translators will as well.
Two, develop a table of what material gets what designation, so when the multiple material fabbers come around, everyone knows that this "color" is this material, and that "color" another. Call this the IP port solution. In such a solution, a certain number of ports should probably be reserved for local "exotics".
Three, make sure a palette is included with any file. One that spells out, for the machine, exactly what materials go where (Oh we used metal for the primary color, since the thing is mostly metal, with a bit of plastic and a small sliver of ceramic.) Call this the GIF approach.
Now, I haven't opened the 3D modeler yet, I just installed it. Maybe that capability already exists.
For other things? Who knows. Maybe someone will make a seven joint arm with a toolhand, (counting shoulder as three joints, and the wrist as three.) Perhaps that arm will use magnetic placement, and induction heating, to fuse metal powder, where everyone else is using a bernoulli stream and laser head, or a hacked wirefeed welder. Perhaps it'll produce plastics from binary chemicals, instead of melting pre-made plastic "wicks". So long as it can read the model, its dimensions, and its materials list, it won't matter.
I can print this document out any of three ways. I can use my current ink-jet, I can go dig out an old impact printer, or I can dump the file to a memory stick, walk into the library in front of me, (I'm using the schools more reliable connection, at the moment,) and print it out on one of the laser printers. Heck. My old drafting professor probably still has a plotter stashed away in his office. The result would be the same printed page, from any of four rather different technologies.
Edited 1 time(s). Last edit at 04/30/2007 07:16PM by Sean Roach.