casting metal

[WillRockwell]

I'd like to learn to cast small metal objects, so I can make USB drives that look like this

[Narsil]

As an overview of metal casting:

-Low melt alloys: various white metals, usually tin based. There are lots of different alloys desi9gned specifically for casting, the most obvious application being military models. The temperatures involved make this a practical 'kitchen sink' project. Because of the low temperatures involved quite a few mould materials are practical with silicone being particularly suitable as it can take a lot of detail and copes well with moderate undercuts and you can get quite a few casts out of a single mould.

The limitation with this material is that the alloys tend to be quite soft and aren't really up to structural applications, but its a good place to start and get a feel of the casting process with relatively little outlay.

-Aluminium: Involves somewhat higher temperatures but a small backyard furnace is certainly doable, if a reasonably involved project. The moulds generally need to be a refractory material such as green sand, bonded sand, refractory plaster or ceramic shell.

Aluminium castings can take a lot of detail of done right and have good mechanical properties and its a reasonably cheap material.  

-Bronze: Bronze casting is getting into the realms of being quite an ambitious project but certainly doable on a small scale. The processes for bronze an aluminium are pretty similar, requiring a gas fired furnace, refractory crucibles and good safety precautions, even on a small scale. Small castings can safely be done with one person. Bronze is an excellent casting metal, producing good quality castings with high levels of detail possible and good mechanical properties, it also has the advantage f being easy to weld and chase. A huge range of different patinas and finishes are possible too.

-Iron: Not really suitable for a solo project unless you have a lot of experience, furnace design and operation is a skilled job and even a small furnace really requires at least two people to run. The minimum practical output is also fairly large, furnaces tapping less than about 50 pounds are rare and tend to to be very practical. People do make tiny iron furnaces but usually only as demonstration pieces.

Processes

-Lost wax casting: a wax positive is made either by sculpting or is cast from an original positive by a low temperature process and has wax runners and risers added. This then has a mould built up around it, usually either bonded sand, refractory plaster or ceramic slurry dip. The mould is then heated in a kiln, melting the wax out and leaving a corresponding cavity in the mould. In some cases  the mould itself may also need to be fired. This method produces excellent detail and virtual any shape can be produced virtually regardless of undercuts etc.

-Styrofoam burn out: Similar to lost wax except that the positive is carved of cast in styrofoam, including all sprue etc. A mould is built up round it and rather than melting it out in a kiln it is burnt out out by the molten metal as it is poured. Great care needs to be taken in the design of the sprue to ensure that there is adequate venting for the large quantities of gas produced (which are also quite toxic). This is a relatively quick and easy method of mould making, especially on a large scale but you have very little control over surface detail.

-Multi-part moulds: The mould is built up in sections which need to be carefully designed so as to lock together and still allow the pattern to be removed. By this method you can produce a mould off virtually any original object, regardless of the material which can be useful if the shape of the pattern requires more structural strength or dimensional accuracy than wax offers. The down side is that it is a skilled and labour intensive process. A complex mould in bonded sand can take days to make.

-Direct carved moulds: Rather than using a pattern the design is carved directly into blocks of suitable material such as cured bonded sand, graphite, cuttlefish bone etc. Very simple moulds can be made in one part (open face) but more sophisticated designs require a multi-part mould.

-Silicone: A flexible material which can be be painted on to a pattern and then peeled off. It's flexibility gives a moderate, but no unlimited tolerance of undercuts. Only suitable for low temperature alloys (white metal etc) and not all moulding silicones are suitable for metal casting. On piece and multi-part mould are possible and moulds can usually be re-used several times.

[Narsil]

Design considerations.

The process of mould-making requires a considerable amount of thought and forward planning.

Undercuts:

It is necessary to make sure that you can remove the pattern from the mould once it's finished. For rigid mould materials this often requires making a mould in several parts which must be carefully planned so they they can be reassembled accurately and not lock onto the pattern. Silicone moulds have a degree of flexibility which mitigates this problem and lost wax and burn-out methods more or less eliminate it as virtually all mould can be made in one piece since teh pattern doen not have to be physically removed but is consumed during the process.

Metal flow:

In order to fill to mould properly consideration needs to be given to the way that metal flows through the mould and cools. Normally the mould is filled with molten metal from a single point and is is desirable to minimise the distance that it has to travel to fill the mould, especially through thin or narrow sections.

Runners are channels in the mould which give a path for the incoming metal these need to be large enough to allow a free flow of metal and avoid premature freezing. It is generally desirable to fill a mould from the bottom upwards so metal is not trying to flow back on itself.

Risers/Vents and channels in the mould which allow gas to escape, if this isn't done effectively then the resulting back pressure can cause voids and incomplete filling of the mould. Risers should generally be places at hight points in the form where gas is likely to accumulate.

Shrinkage

All metals shrink to a certain extend as they cool and solidify. It is therefore necessary to compensate for this. If dimensional accuracy is critical then the pattern should be made slightly oversized to compensate. Even if this isn't a problem shrinkage can cause voids in the casting. They way to avoid this is to create reservoirs to supply fresh metal into the mould as the surface freezes and to narrow the points at which the runners and risers meet the form (gates) Bearing in mind that the metal will cool from the outside in and large volumes will cool more slowly than smaller ones. Particular problem areas are any point where there is an abrupt change of section ie where a thick section meets a thin section.

[Otto Von Pifka]

optionally there are resins for casting that are very strong and can be colored and tinted to look almost exactly like that sub. saves the added trouble of dealing with heat and flow.

a friend bought a rubber mold set to cast chess pieces with, what a nightmare! the molds were of dubious design and the metals we bought from the same outfit were terrible for casting, the finished pieces were way too soft. the pot we used to melt with was a bullet casting pot, it seemed adequate. I eventually made a vibrating box to clamp the molds to, to help with the fill.
the vents in the molds were poorly designed and the sprue lead was too short to allow for shrinkage. the box was simply a wooden box upside down with rubber feet, a motor with a concentric weight on the shaft inside the box to vibrate it, and a block of wood sticking out the top to hold the molds to. I had made a speed controller out of an industrial ceiling fan controller, for a heavy duty die grinder thet only had one speed, it worked well with the box.

I have casted a few lead pieces in silicone, the trick seems to be to design the mold vertically so the weight of the lead forces it into the trickier areas. even the sprue can act as a sort of ballast to the casting and compensate for tunneling from shrinkage.

I even cast a few very low temp metal pieces in urethane to see if it would work. the alloys weren't safe for extended handling. the stuff is made for casting machining samples for quality control and prototyping. a pity since it cast fairly well and the pieces were every bit as hard as the new cast minifigs (the bismuth alloys). the molds bore up fairly well but the urethane did become more flexible while warm.

the mold is just a cavity so it didn't want to fill the tailfins well, so I just ran the metal extra hot (still cooler than regular lead by 100 degrees) and used a toothpick to drag the airbubbles out of the center fin and to scrape the gelling metal into the flat fins. was actually pretty easy to do. if I were going to make a serious mold to do the zeps, it would be a two part mold with the zep nose up and the sprue filling the mold from just above the tip of the nose but molded into the flat half to leave the nose details intact. the tailfins I would cast thicker to make the molds easier to use.

you should look into resins at first and silicone molds, much of what you learn making the molds will serve you if you step up to metals. smooth-on has some good tutorials and FAQ reads, molded in color would work nicely with a design like that. I myself am looking into making some molds to do scarab beetle cartouche/seals to hide the drives and also fossilized trilobites, even got a snappy name, "trilobytes" 
just need to get all these other projects done first. if I don't get pocky and darkshines their syringes before halloween, there won't be a flagstone low enough for me to hide under.