Back in the dark ages when the GPO (who remembers them?) maintained their own vehicles after a crankshaft had been reground to its maximum undersize is would be metal sprayed and the ground to standard size to repeat the cycle of reconditioning. I am sure there will come a time soon (if it isn't already here) when 3D printers can spray metal to make parts. It is only a matter of time.
Peter
This is part of my professional area of interest.
The repair process you refer to for crankshafts dates back (so far as I have found documentary evidence for so far) to the 1920s where metal spraying was in use for crankshafts on ships' engines. I have a photograph of the process in operation in a copy of Newnes' Complete Engineer from around that time.
The metal spraying process either uses a powder or a wire and heats it using a flame, arc or plasma in a high velocity gas stream which projects it onto the surface. Conditions are set so that it is semi-solid when it hits (think Blu-tac) and the splats (as they are correctly termed) build up layer by layer. There tends to be a thin oxide film formed on the surface of each splat so the bonding between them is very weak. This means they work really well as coatings, such as for corrosion protection or wear resistance where they are mainly in compression but they are poor under other stresses such as bending or tensile so they don't have much more strength than a biscuit.
There is a more recent development known as Cold Spray which prevents an oxide from forming and it does get nearer to a metallic part in strength but it uses huge quantities of helium and costs about £10k each time you fire it up, so it won't be coming near anything other than advanced parts for space/aerospace anytime soon.
There are alternative technologies such as laser fusion. There are two types of this technology, powder bed and blown powder. Powder bed is becoming reasonably common. You break the whole part down into layer by layer slices and then the idea is that you put a layer of powder in a tank, then raster a laser over for that slice, lower the tank by one increment, place another layer of powder over and do it again. This works well but is limited by the size of the tank, currently about 250mm diameter x about the same depth. You need to have enough powder to fill the whole tank which is up to £20k, although you can sieve it out and re-use most of it so it only costs a bit more than the volume of the parts. Most of the time goes in laying up the layer rather than rastering the laser so it is much more efficient to make as many parts at a time as you can fit in the tank. The other option is blown powder which is projected through a nozzle onto the surface where you want to fuse it, with a laser beam simultaneously impinged. The combination melts the powder on where it then rapidly cools, a bit like building up an arc weld on a surface but much finer scale. With the right multi-axis table you can make pretty much anything in free space. The biggest part we have done is the front casing of a gas turbine nearly a metre across. The kit needs the laser, the robot, the multi-axis stage and the software to control it all so all-in it is around £0.5M. Because the laser fused processes melt the metal it is a lot stronger than the oxide-coated sprayed structures but it is still a cast microstructure rather than forged so tends to be more brittle than a forged part. You can improve on it a lot with good heat treatment.
The final approach is to put the powder in a binder and then after forming to burn out the binder and sinter the parts. This is effectively how ceramics are made - everything from teacups to toilets. You can do this by using a wax binder and extruding into moulds (which can be 3D printed) known as Metal Injection Moulding or MIM, or by running a more conventional 3D printer of the resin type. As has been mentioned above, the resins are light cured (commonly UV) which does present a problem when you add a lot of metal powder to the resin as the light no longer goes through and it also tends to settle in the machine, making the layers uneven. Both these issues are being worked on. There are already systems on the market which work to an extent but it's worth a look at Photocentric's website to see what the state of the art is. We are working with them at them moment to develop this. Once you have your printed parts you need to remove the binder and then sinter the particles together. This works well but for metal it has to be done in an atmosphere which prevents the surface of the particles from oxidising so a specialist furnace is normally required.
All of the above is possible now and some of the processes will eventually become domestically accessible. Initiatives such as Barclays' Eagle Labs are likely to be one of the first routes to access. If you haven't come across these, they are pretty good. They tend to have advanced facilities available to use 'by the hour' so if you want to do some decent 3D printing to make your part additively, or even use 3D CNC milling to make your part subtractively from solid then both are available.
Alec
