Gunsmithing with a 3D printer
- Thread starter Odio
- Start date
replicator? Melting plastic into shapes laid down by what's effectively a CAM system, while extremely cool, isn't exactly reconstituting a cup of hot earl grey from pure energy.
the resolution on 3d printers is... well..
Great for prototyping parts, not up to the sort of tolerances you'd want for a gun, especially on the curves.
Great for prototyping parts, not up to the sort of tolerances you'd want for a gun, especially on the curves.
Yet.
even at "maximum awesomeness", it's still a digital system and as such is incapable of creating a true curved surface, merely a series of really small steps. You could get it close (probably 'close enough'), but it'd still pay to throw it on a lathe or more 'traditional' machining process to get it perfect.
pZ II Birdy
Veteran XX
Same with laser sintering, but yeah... It beats doing the initial hog out
Edit: reading through the link it looks like he undersized the holes then opened them up manually to get a nice finish.
Edit: reading through the link it looks like he undersized the holes then opened them up manually to get a nice finish.
Last edited:
Someone needs to build one with a lathe and a waldoe arm.
Esteban_Villa
Veteran XV
Togo is a lathe operator.
in other news britan bans 3d printers as soon as they realize you can make a half decent thrust dagger in one.
Its no longer metal detection but "Object orientation" in most airports. I had a mint in one pocket going through the DFW airport and they snagged me on it.
That's the one I was thinking of
The sintered ones are very strong and shit can be hardened after machining, but still not quite suitable to withstand being used as a firearm. I certainly wouldn't want to fire one anyway.
Er, why couldn't it make a curve?
I could see 3d printing the upper and lower ar15 receiver and misc parts
but not the barrel and bolt assemblies, not for a while
but not the barrel and bolt assemblies, not for a while
If I remember correctly, mathematically and computationally a true curve has an infinite number of data points. A computer simply cannot calculate a true curve as it MUST subdivide the curve into a set number of data points. Otherwise the algorithm trying to describe the curve would just calculate off into infinity.
So the computer chops the curve up into pieces (data points, pixels, etc) and calculates where each point is in space. If you have a low number of data points, to the human eye you end up with a sad looking curve made up of a small number of facets. But this curve can be calculated very quickly.
If you have a large number of data points, to the human eye the curve appears to be smoother, something closer to a true curve. But calculating all those data points costs you in time and processing. So you have to come up with a comfortable medium.
In 3D space you calculate the position of polygons. The more polygons you use to describe an object, say, a sphere, the smoother and more "round" the sphere appears. The fewer number of polygons you calculate to describe that sphere the more "rough" or faceted that sphere appears.
Zoom in on a high-polygon count sphere and it gets more "rough" or faceted again.
This is why graphics cards put so much attention on how many polys they can calculate per second. If it cannot calculate many polys per second, you get a Minecraft character walking across the screen. If it can calculate billions of polys per second you get a much more complex, detailed and realistic looking character walking across the screen.
So in the case of this 3D printer... what is its maximum resolution? What is the smallest dot of materiel it can deposit on your model? This will define how well it can print out a "curve". Can it print out individual atoms? Awesome, you'll get great "curves". But it will take forever to print and, zoom in enough, you'll still see a rough surface.
Last edited: