how to make aluminium rod

[Hornblower]

How to make accurate cubes out of aluminium, wood and stone ? (0.1mm accuracy required). Again using 16th century tech at most. I have no idea about how that could be done. Crude cubes can be done using hand/eye precision, but how to make accurate ones ?

Wood is the easiest to cut and finish with simple hand tools, and your accuracy requirement could be met if the woodworker is persistent enough. Stone can be ground flat by rubbing two rough pieces against one another with sand between them, or with something like emery for really hard stone. The measurements could be checked with a ruler and a framing square. I think metal would be a beastly job without modern milling machinery. It has been my experience that pieces of metal yield to abrasives much more slowly than do pieces of stone or glass.

[swampyankee]

How to make accurate cubes out of aluminium, wood and stone ? (0.1mm accuracy required). Again using 16th-17th century tech at most. I have no idea about how that could be done. Crude cubes can be done using hand/eye precision, but how to make accurate ones ?

It sounds like you're trying to make Jo blocks with 16th Century methods.

I don't think that's possible (but I'd be very happy if you could prove me wrong), as they did not have the measuring technology for repeatably doing so. Jo blocks were traditionally made of high-hardness tool steel, ground and lapped to final dimensions, although they're now made in ceramics (iirc, zirconia). Don't even bother with aluminum. A good set will only set you back a couple of grand.

[profloater]

How to make accurate cubes out of aluminium, wood and stone ? (0.1mm accuracy required). Again using 16th-17th century tech at most. I have no idea about how that could be done. Crude cubes can be done using hand/eye precision, but how to make accurate ones ?

You call 0.1 mm accurate? That's 4 thou, not exactly a specular finish. You could do that with a file and gauge and some time to practice. Ok everything starts with reference. You need a straight edge, and an angle, but that level of accuracy could be obtained centuries ago if they needed it for some reason. It would not be for aluminium of course. Brass or bronze makes more sense. A simple calliper will measure to four thou. Press fits use one thou per inch, that kind of thing. Whether they had need, for a lock perhaps , is another question.

[roboticmhd]

You call 0.1 mm accurate? That's 4 thou, not exactly a specular finish. You could do that with a file and gauge and some time to practice. Ok everything starts with reference. You need a straight edge, and an angle, but that level of accuracy could be obtained centuries ago if they needed it for some reason. It would not be for aluminium of course. Brass or bronze makes more sense. A simple calliper will measure to four thou. Press fits use one thou per inch, that kind of thing. Whether they had need, for a lock perhaps , is another question.

0.1mm is accurate enough for me I guess, of course better accuracy is always welcome, 1 micrometer would be great (those advanced guys can do as good as nanometers I guess). Can you provide more detail about how to get accurate cube(or cuboid) using things you listed: straight edge, angle, ... ? (be it bronze or brass or wood or stone)

(Note: this might sound funny but if you profloater or any of you guys can help me with this, I would like to have conversation over skype, my name on skype is "mehdi.mehdizade" without quotes, please add me and let me know)

[NearABE]

unlike copper it is hard to join aluminium by mechanical pressure, you mentioned several pieces, it's because of the stable oxide on the surfaces.
If they had aluminium back then, which they did not, I think they might have worked it by hitting it. I have in mind the process of impact back extrusion (modern) which combines heating and working by being very fast. If you hit aluminium with a hammer on an anvil you can get some success but if you have an alloy rather than pure aluminium, you will get work hardening which then will crack and air will get the the crack so it won't join up again even if you heat it. Pure aluminium is much more malleable. You could combine alternate heating and hitting to make a thick rod. However I think I would go with casting first. If you made a long trough in wood with a suitable groove, melt the aluminium and cunningly hold back the oxide scum on top with a tap beneath the surface you should be able to easily run a 2m long rod if your groove is level and insulating so that it cools slowly. Then an anvil and hammer would allow some shape correction to make an aluminium arrow for example. It is quite easy to melt aluminium in a steel pan but take care as advised earlier.

Impact forging works with iron. The impurities have a different crystal structure from iron. Rust flakes off because of the mismatch between iron and iron oxide. Working the hot steel helps the impurities move through the material.
Aluminum oxide has a crystal structure that matches and aligns with pure aluminum. The phases do not have any reason to work out of the material. If you fold aluminum oxide into aluminum it will mostly stay in the aluminum. Aluminum oxide also does not float as well as slag floats out of pig iron.
Aluminum is extremely reactive with oxygen. Your aluminum can does not burst into flame because the 2 or 3 atomic layers on the surface already did that. The crystal lattices match so there is not gap between the aluminum grain and the aluminum oxide grain. If you liquefy aluminum the oxygen has no problem rapidly consuming the metal.
Carbon bonds oxygen stronger than iron bonds oxygen and much stronger than copper binds oxygen. So you can put iron in contact with carbon monoxide coming from charcoal. The oxygen atoms will tend to leave the iron metal and become carbon dioxide. Liquid aluminum would strip the oxygen off carbon monoxide. You end up with graphite and various aluminum oxides.

[profloater]

Yes but you can melt and pour aluminium, the oxide forms on the free surface of course.

[roboticmhd]

how to make accurate cube(or cuboid) using things you listed: straight edge, angle, ... ? (be it bronze or brass or wood or stone)

[Hornblower]

For starters, using methods available to ancient craftsmen, I would grind three suitable pieces of stone really flat, using the method described here.
http://www.jeffbaldwin.org/flats.htm
I would stop with the fine grit, because we are not looking for optical standards of flatness. Using one of these flats as a reference, we could check a metal straightedge for being really straight, and file or grind it as necessary to straighten the edge. Then we could make a T square or framing square from a pair of straightedges, and check for an accurate right angle by drawing a perpendicular line and then flipping the square over to see if it matches. When we are satisfied with the square, we could use it to check the accuracy of a cuboid.

[profloater]

Yes a flat surface is the start, and these are flat to better than one thou. The OP seems to find that hard to work with but it's easy to detect a one thou gap in hard materials. Skilled filing or sanding can work to that tolerance using a straight edge and then when you use engineers blue, a grease, you can work to finer tolerance. Even a simple tool like a calliper can find errors of a thou and allows correction with abrasion, lapping and so on. If you look at the modern micro sculpture movement, craftspeople are using their own steady hand and tiny tools to make sculpture too small to see without aids. Skilled workers are impressive.

[profloater]

image.jpegfor example here is a picture from a current exhibition.

[billslugg]

I would get a one inch diameter cardboard tube, suspend a wooden dowel in the center, fill the annulus with refractory material and tamp it down with another wooden rod. Let it cure then burn it out in a campfire. While still hot, stand vertically in a pile of sand and pour liquid aluminum into it.
Chemically pure aluminum does not work harden, but any alloy will.

[profloater]

I would get a one inch diameter cardboard tube, suspend a wooden dowel in the center, fill the annulus with refractory material and tamp it down with another wooden rod. Let it cure then burn it out in a campfire. While still hot, stand vertically in a pile of sand and pour liquid aluminum into it.
Chemically pure aluminum does not work harden, but any alloy will.

Yes that would work as a practical method, but I wonder more and more why the op wants to take aluminium back to when it was not available. Once you have extracted it from bauxite it becomes one of the most recycled metals. The history of metallurgy is about alloys that lowered melting point and added strength or ductility, real alchemy. Aluminium is such a modern material, like magnesium and titanium. Made possible by electricity.

[Jeff Root]

I suspect that roboticmhd just wants to use techiques that he
is already more-or-less familar with, and can do himself. Anything
"modern" is likely be beyond his means, or at least so expensive
that it would make the basic idea of using the resources already
available to him pointless. He has resources and wants to make
use of them. He doesn't want to start a commercial operation
fabricating aluminum parts.

A hundred and fifty years ago, it was the norm to re-use whatever
resources you had, and more often than not, you did it yourself.
Now we can usually give stuff to others to recycle. Aluminum
is unusual in that some of its cost can often be recovered by the
aluminum's possesor. Gold and silver have a much longer history
of recycling that way.

-- Jeff, in Minneapolis