Skip to main content

#BuildBrighton - Printing in 3D

I've been pretty dismissive of 3D printers, chiefly because I view them as complex, fiddly, erratic, temperamental and inelegant machines.
If you want a lumpy thimble made very, very... very slowly they are perfect.
If you want anything even moderately useful, then you probably need a machine bigger than you can afford, and it will still be slow...
and probably breakdown...
or print 80% of the thimble then suddenly have a funny turn and stop, or bung up or something.
They are a bit like "classic" cars, where "classic" means poorly designed, 1970's British monstrosity.

However the #BuildBrighton folk organised a 3D group, so  I thought I'd rock up and see if they couldn't dispel my negativity now I actually have a part I need to make  (a collet for the "gun" on my Dalek.)

The #BuildBrighton 3D printer group started with a talk from Alex about OpenSCAD, which is a simple scripting language that allows geometric primitives to be combined (union) or subtracted (difference or intersection).


The OpenSCAD tool provides a swift rendering of the object, and ultimately a union is created of the whole model which is exported.
Because it's a scripting language it means models can be built programmatically, with all the usual loops and logic.

for (i=[10,20,30,40]) {
translate ([i,i*2,i*3])
rotate([0,20*1,10*i])
sphere(20)
}

OpenSCAD can import dxf files, however these are dxf-r12 - which was a CAD standard format for the last version of AutoCAD that ran on DOS, this should allow seamless transport of data/models/shapes... but r12 isn't supported by much outside of Inkscape these days.
OpenSCAD delivers something close to the final STL (Stereo lithogram) object, whereas Google Sketchup requires a deal more work.
Blender provides a GUI alternative to OpenSCAD, but may have a steeper learning curve.
My experience of teaching Vectorworks indicates that moving 3D objects around in space is a pain in the bum, and a few lines of code may be much easier.
Pleasant3D provides  way of checking and editing the final STL file.

There are a huge number of models already built, in a truly Object Oriented way, ready for re-use :http://www.thingiverse.com


Next step is to dump out the object and chop or SLICE it into layers that the printer can print, a G code file.
Slicing software is pretty printer specific, but a tool like SLIC3R provides support for a wide range (RepRap (Prusa Mendel, MendelMax, Huxley, Tantillus...), Ultimaker, Makerbot, Lulzbot AO-100, TAZ, MakerGear M2, Rostock, Mach3, Bukobot and lots more. And even DLP printers.) SLIC3R also provides an option to add supports for geometry within geometry.

Stephen Cropp then talked about the types of plastics that can be used in the machines and the temparatures, PLA, XT (a type of polycarbonate) and nylon (619 grade)-  some of these need heated beds, some are very brittle, some need extra heat to melt.
Care of the plastic is important, if they are stored badly, the moisture in the plastic turns to steam.
Storing properly means keeping it dry - sealed container with dessicant.
Some new materials are super flexible which means there is an opportunity to mix rigid and flexible materials.
The materials come in different colours, which may be more successful than painting or dying.

Discussions then rolled into how to move things forward in the group, and the structure it should take - open sessions, practical printing or machine building, from basic simple printers to fancy machines - Reading Hackspace are apparently building a laser scintering 3D printer, which fuses metal powders.

So, have I changed my view of 3D printing?
mmmm, well I like the fripperies, of building gemotries, and the idea of the underpinning maths, I even came away liking the look of the Delta printer, which seems far better engineered, and elegant.

I guess I'll have to see, whether 3D grows on me.
I'll have a bash at printing something next week on the #BuildBrighton 3D printer... perhaps that will get me hooked.





Comments

Popular posts from this blog

The weakness of machines

All of the robots I have built have failed in fights, and usually at the axles/wheels, it's like a running joke. So why should it be any different with this featherweight? Meathammer was designed back in 2005/06 when using 4 drill motors was a bit of a novelty... it was made more novel still by mounting the epicyclic gear cages directly to the chassis, which unfortunately sends all the drive shocks back through the gearbox. The original version ran for a number of fights at modelworld 2006? and even won some, especially the sumo, which favoured its very low profile. I've just rebuilt the gearboxes with all metal gears, and tested it, and it seems it should last for at least one fight… but as I put it in the car after testing, I noted one of the wheels had nearly fallen off. Mark that bolt for Lockthreading! All this got me thinking if I could make something less likely to lose a wheel, break a gearbox etc I have a huge pile of bits and bobs that were destined for heavie

robot rebuilds

For two years in a row I've missed out joining in with Ian Watts robot bash at the Uni. So I'm making sure I've got something running before it swings around again. So I welded a couple of new gearbox cages into the 4 wheel push bot, and dug out a set of Electronize speed controllers. I fitted the titanium plate that has been cluttering the basement for a decade and have been doing some testing with a set of LiON batteries that used to belong to a cordless strimmer, which fell apart from overuse. I've run it up and the battery doesn't really have enough life to power 4 motors, but is useful for testing. I've stuck some old NiMH units on a charge cycle to see if they have enough life for a bout or two. The only things I actually need to "buy" are a 2.4GHz radio set, as the old 40MHz stuff is against the rules these days, so I've ordered a Hobby King 4 channel set for £20 off ebay. Strangely yesterday Mentorn emailed a bunch of robot folk to

Fake Plastic Trees...

So, for a bit of light relief I bought a few of Justin's LED Christmas tree kits. I think the original idea was to have a relatively easy kit to put together on Saturday workshops in the run up to Christmas, but they seem to have become quite popular with all manner of folk. Anyway, the kit is based around a couple of well made pcbs, a bag full of components, and the simple instruction that +ve is UP. It took about 10 mins to solder all the bits together, and with a huge amount of satisfaction over the end result. Justin and his Christmas tree kits can be found here: http://cyber-lane.com/ The Build Brighton Christmas workshops will be running each Saturday from 30th November to 21st December from 10am – 2pm. Entrance is completely free. Kits are priced between £2 and £5. #BuildBrighton can be found here: http://www.buildbrighton.com/blog/