Developing a new product family quickly using @blender

My blog’s been pretty quiet of late as I got ready for @shapeways at the Maker Faire http://bit.ly/10z8u4V  And here’s why…

As you know I’m really into designing typographic cufflinks… but I’ve also been working on something a bit more accessible to a wider audience (who aren’t type geeks or cufflink wearers, which is a pretty small audience to be fair).

Introducing the “Neckbangle”  (I wasn’t sure what it was, it’s not a necklace, it might be a choker, it’s like a bangle hence the new product framily name is neckbangle).

Here’s the basic concept - I built it using Blender (which is awesome for rapid development).  It’s a contoured, spherical wrap around one-piece design.  Styled to sit at the base of the throat…

This took about a day to work out the specifics… it’s an asymmetrical slice through a hollow sphere - which gives the base structure a close-fitting, natural look which should fit most female necks.  It’s got a 76mm aperture at the back which allows it to hook on without too much trouble but not fall off.  In theory anyway…

Then I took a couple of hours to experiment with the form and create two other versions…  the ancient neckbangle (inspired by the metalwork styles of old Saxon / Roman jewellery)

And the Shoredtich neckbangle - a punk themed version:

Cutting holes in an asymmetric slice of a sphere ain’t easy… you really need to do it by eye, changing the slice angle by -1 degrees on the X axis, regular (ish) rotation increments around the Z axis and adjusting Y axis coordinates by instinct.  Blender is awesome for that - the controls for rotation around the object’s origin mean you can quite literally click away in each axis until it looks right on the little control box, no complex movements necessary.  Also, for each hole or spike, just <shift><d> duplicate the last one and move it again… easy really.

What’s next?  Well I’ve got a few prototypes in plastic and steel coming, so gauge the sizes with some real women… then it’s a matter of working out how to attach pendants and, of course, mount typography on it.  

I’ll keep you posted ;)

One for @bartv and the @shapeways team - new prototypes, more wine, learning the lessons of articulation, 0.2mm tolerances and materials

a whole load of typographic 3D printed cufflinks - finally put the vid on tumblr

Working with re-usable components for 3D print prototypes - @Blender3D tips

It’s been a long couple of weeks… I’m finally starting to think like a 3D print designer.  Here’s the story…

I’ve been puzzling over creating a mechanism to act as a clasp/lock to open and close chains.  A hoop / bar works fine, but I’m a natural tinkerer so I started thinking about making something that was more of a universal coupler - so it could connect two ends of a chain OR secure a pendant to a chain OR interchangable pendants / charms to chains and bracelets. Here’s my finished concept (waiting for @shapeways to deliver some test models back)

image

It’s a lock and key design, where the key can only pass through the lock at 90 degrees to its resting vertical position (below the H shape on 2 guide pins) - basically, once it’s connected in theory you can dance about the place or go base jumping (or whatever) and it can’t unlock.  If it works, you’re looking at the two part connector that fastens chains, pendants and all kinds of jewellery components to each other, with all the innards of the mechanism exposed.  

Then I started making my first one-print necklace… using a space invader motif:

image

It’s pretty rough at the moment, but a good working prototype to test material strength and how to compose an STL of many parts to work with Shapeways size and tolerance limits.

Then it hit me - why not take the bar off my type cufflinks and connect it to the space invader?  I realised once I had a component that worked, if I made it as a stand alone STL from Blender I could re-use it.

Make your clean STL by selecting the bits you want to keep {ctrl-J} then enter Edit mode {tab} > select all vertices {A} > Remove doubles {press w and select from the menu} then back to view mode where you select File > export > STL.

Import multiple STLs into a new file and assemble all the bits you want.  I took the Space Invaders, then the cufflink connector, made them into STLs and did just that…

image

Recognise the bits? (see posts on cufflinks below).  It must have taken me about 2 minutes.  That’s 5 new 3D printable items for sale on Shapeways in about 10 mins - efficiently re-using the items I’ve spent a week making.  All I have to do is get my connector working, and I’ll be able to make everything - the type cufflinks and the invaders into pendants in about 20 mins… which feels really efficient.

One really important Blender tip is to us the P button (in Blender 2.6 & later) if you need to break apart a joined mesh.  It’s a bit of a pain, because it breaks up all the linked surfaces, so a space invader breaks down into about 10 parts, but using the scene navigator window you can quickly work out what’s what and re-join them into usable parts.  This was very useful when I accidentally made my necklace into one object, exported it as an STL then found it was too big to be printed - so I had to go back and make a couple of rotations to segments to get it into the right bounding box footprint to coloured plastic.  It took about an hour of fiddling… I guess the lesson there is simple:

SAVE MULTIPLE VERSIONS OF YOUR COMPONENTS BEFORE YOU JOIN AND ASSEMBLE THEM.

I had been doing it… the one time I got too excited, boom!!!  

Opening my Spore prototypes video… great to see the materials working so well, especially laser sintered sandstone

Spore prototypes arrive: 2 piece interlocking design (thanks @shapeways)

The Spore is my first experiment with interlocked printed components.  It’s a 12 point hollowed-out conical star inside a 2mm thick spherical shell with 24 holes cut in it.  It was a material test as much as anything - I wanted to get a feel for alumide, transparent plastic and sandstone.

As you can see, it’s worked out quite well.  The really interesting thing is how sturdy and usable the sandstone is (the white one) and how good the definition of the alumide is (the grey one).

The next step is working out how to link them so I can 3D print a whole necklace… stay tuned ;)

Here’s my first prototype typographic cufflinks… fresh from Shapeways via the UPS guy…

First typography cufflink prototypes arrive - and they rock!!!

This morning the UPS guy brought me my first prototype cufflinks from those nice people at @shapeways…  and they’re awesome.

They’ve done what I set out to achieve, the type looks like it’s floating on your wrist, the supporting arm structure means you tend not to see the horizontal bar holding them up, so the type isn’t distorted from the character shapes.  They read properly, the resolution of the print gives a smooth outline and they look good.

The supporting arms are just under 2mm deep, so not very strong but fortunately the fact each character as at least 2 or 3 positioned behind it means they distribute the pressure on the letter quite effectively, making the plastic ones feel sturdy with only a little flex in the bar when you put them on, hard to notice.  They’re definitely strong enough for the purpose. 

The silver ones are rock solid.

The arm and button design works effectively too, in fact, I’m delighted at how well the 3D model actually translated into reality, it’s quite hard to guess how small dimensions will look from an image on a monitor, but they actually work better than I’d imagined.

The next step is to create a full range.  Now the engineering is sorted out it’s time to get creative and explore the font options.  FS Lola Extrabold Italic works very well, time to find some serifs and see if trad fonts can cut it.  Bold cuts work best, italics have great 3D potential because of the surfacing, but I don’t want to limit the options too much until I’ve done more testing.

More material tests & and some articulation

Whilst I’m waiting to road test the typographic cufflinks, I’m testing a two piece interlocked model called “Spore”.  It’s based on a fungal spore, consisting of a hollow spherical shell with a spiky multi-pointed star sitting loose inside it.  The design is intended to allow the spikes to move in and out of the shell with gravity.  It’s a complex, slightly frivolous idea which could create funky beads for a necklace - but the idea wasn’t really just a design concept, it had some strategic thinking behind it… 

Firstly, I wanted to test out the tolerances / clearances required to print something that was a two piece interlocking design.  This is the basis of chains and machinery, so a natural next step to move beyond the parameters of thickness (walls and supporting wires) and into realm of printer clearances, i.e. how much space do you need between interlocking component 1 & 2 to be printable.  The answer is around 2mm to be safe… which is what I’ve done here.

Secondly, I wanted to experiment with how much of each item I could carve out (using boolean difference calculations in Blender).  The outer shell is a 2mm thick sphere, the inner star is a 2mm thick hollow structure as well.  I then took out more material around seams / joins to leave a consistent 2mm of surface everywhere, but to make as many holes as possible to reduce the volume of the material getting printed, and thus reduce cost.

In that process I hit upon a really pressing 3D modelling issue.  The rotation angles of holes on a spherical surface isn’t like lining-up angles on a 2D one (duh, yes I know it’s obvious) but working out the maths is very complex.  I put the equator / pole holes at 90 degrees to each other, equidistant to make the 6 cardinal points of the star.  The other 8 points of the star fit in the middle of cardinals, making each into a 3 hole cardinal triangle with a hole at each corner and one in the middle of the surface.  The middle hole, bisecting each 90 corner angle sits at 52.5 degrees of rotation x 45 degrees.  I kinda know why, but that was reached by trial and error.  Anyway it worked.

Finally, I wanted to use this to test different materials, especially transparent materials.  I made extra holes in the surface of the spherical shell so you could see the insides (what’s the point of creating a complex 3d printed shape) and also to test out new materials to me - i.e. Almuide and standstone.  

They’ll be arriving in a few weeks, I’m now working on a ceramic 3D type concept… so that’s 3 new collections of print experiments this month.  The next step will be creating commercial executions which will mean getting the production cost down and the final executions sorted.  The aim is to, by April, be hard into selling this stuff…

Typographic cufflink progress… beefing-up the structure

The thing about 3D printing is… the materials affect the design specs.  This might seem obvious, it does to me when I write it down, but coming from a graphic design background you’re seldom ask to make design choices based on physical limits.  Okay, if your into print design or fabrics, you’re aware of a degree of physical limits, but never to the point whereby you need to rethink the choice of font - and that’s the difference when you’re working at the limits of what a 3D printer can handle.

Last week, the nice people at Shapeways.com checked over my initial prototypes and let me know the supporting struts for my letters were too thin.  Even the ones that weren’t theoretically too thin (but at the limit of how thin they could be) they wouldn’t stand the post-print process of polishing, which naturally reduces thicknesses down by a few fractions of a millimetre.  

So it was back to the drawing board.  Actually, it was a good chance to work with dimensions in Blender a bit more precisely (up to now I’ve been using SVGs created in Illustrator and then resizing in blender by eye).  So I rebuilt the whole cufflink from scratch.  This time, I ensured none of my struts were less that 1.2mm - which, in theory, means even if I lose 0.2 - 0.4mm in the post-print finishing process on some of the materials, they’ll still retain enough tensile strength.  Here’s the new design…

Of course there’s a few ugly join angles on the structure, but I’m hoping it will be printable.  Otherwise, I’ll need to have another rethink…  on that front I had to rethink the font.  Fatter characters are best, so I’ve worked with my favourite fat font FS Lola Bold Oblique from Fontsmith.  It’s good… next I’ll be trying out heavy cuts of Helvetica, Gill and a great Fontsmith Serif called FS Clerkenwell.  

I guess this is the whole point of prototyping… it also enabled me to fine tune the thickness of the whole structure, experiment with an embossed UP logo on the button end, and generally get to grips with the organic modelling UX of Blender, which remains awesome - the new 2.65 release is working very well on my tiny little MacBook Air 11” - although I’m using a second big screen monitor and a Logitech 3 Button mouse with wheel to give me the fine control needed to rotate and slide stuff around.

New project: Typographic cufflinks…

Type cufflinks are nothing new, but I wanted to address the typography angle and the ergonomics.

One major problem with type cufflinks (and typographic jewellery in general) is in the mounting of the type on the housing structure.  The thing is, type floats in space (on a screen or on a page) it doesn’t link together.  That’s great for graphic designers but in the 3D world you’ve got to attach the type somehow.  I learned from my initial experiments that you can do this by connecting the letters to each other by extruding the characters, but that means the type changes it’s typographic identity, it works ok with italic, chunky fonts but really spoils the look of traditional serif and san serif characters.

Unfortunately, connected the dot to the ‘i’ is still a necessity.  I could have done this with a central bar rather than a full width connector to give it a proper 3D impression, but given the small sizes of such a connector, I’m concerned the little dots would break off to easily.  It will need more testing.

A lot of designs avoid joining the individual characters by extruding them out from a back plate, but then you’ve got a back plate, and I think the letters should have a face from all angles rather than only the front perspective.  Also the back plate approach affects the light cast on the type and it diminishes the elegance of the letter forms.

These experimental designs have taken the approach of mounting each individual letters on individual supporting arms.  That way you minimise the amount of material connected to the letters, making the type appear to float.  But this approach has it’s flaws.  To get the supports thin enough to leave the letter forms more or less intact, you need multiple supports - otherwise the letters will break off the cufflink structure or bend out of alignment.  

I’ve opted for a mix of supports, some based on circular designs (to distribute pressure along the letter evenly, if knocked) and secondary straight supports where a circular 2 pronged approach would be visible outside the edges of the face of the character, for example, on an ‘s’ you can’t attach a circular support at all, because you see it through the curves of the ‘s’.  With an uppercase ‘G’ you can only hide a circular support behind the thick curve of the letter, you need a single bar to connect the inner end of the G or you’ll see too much of the support… etc.

Obviously for strength, these supports are connected to a central bar, which is visible from the front of the word - but less so than a back plate, and distance from the characters is increased by the use of the supports, so it recedes into the background nicely.  

The second issue I have with cufflinks is the link structure itself.  Round bars enable the face of the cufflink to rotate in the buttonhole, so it needs to be flat to keep everything oriented correctly.  But I can’t be too wide or it won’t fit the buttonhole properly.  Adopting the strut approach, I tried to make a strong form of connected struts which will, as the cuff expands, give the lateral stability to prevent the word rotating in the button hole.  

This structure couldn’t be solid because I’ve carved out the Unique Plastique UP logo mark in the button end, and I wanted to keep the letters clear.  It also spells ‘up’ as a handy reminder to put the cufflink on straight.

The net result is a clearer, floating typeface of distinct, unadulterated letter forms, and a mounting structure of interconnecting struts that has an architectural vibe to it, functional and designed to be strong without being bigger and weightier than the word it’s supporting.

The font is Helvetica, the cut is bold because the thicker characters are essential to give them visual impact and create a strong enough base for the mounting structure… I’m going to try some other faces and words over the next week and post them.

Next step… do some material tests to see how it prints, plus experiment with the sizes… right now it’s about 50mm deeper than most cufflinks, which might make the links too long, on the other hand, it might make them easier to insert and less likely to break under the stress of repeated use… or fit people with big wrists.  I have an idea on how to fix that if it does, more on that as the project develops ;)

Back at the plastique… at last

Phew.  What a long enforced break from the 3D print project as my studio was repaired (along with my house) for 5 months after flooding last year.  It wasn’t a plumbing flood, it was a full-on flash flood last summer caused through heavy rainfall.  Anyway, it meant moving out and I didn’t get back in until just before Xmas.  Back in business now…

Phase 1 of the Unique Plastique experiment is over.  I’ve explored the potential of using typography and simple 2D linework to create 3D objects, which has worked well.  It’s tested the materials, got a feel for tolerances and the way fonts work in 3D.  Now it’s time to move on to the next phase - vases, chains and more complex models.

Day #77 - after a bit of a break I’ve finally got my first large size typographic wall hanger back.  It’s a birthday present for my old business partner Brad.  It’s designed to use interconnecting type to make the characters sit as naturally as possible without supports - using the merging of characters to support the text and make it ‘float’.  

The typeface is Aksidenz Grotesk bold, which adapts well to 3D printing.  It’s a bit thin in the z axis (only 5mm) which isn’t super strong - it could have used an additional 5mm to be really robust and function as a coathanger, but the price was a bit high.  

It also has a graffiti style cock (an in-joke) in relief which has worked quite well, but has a couple of flaws in it. Same with the frame, there’s a couple of nicks which came from the SVG not being super smooth (a vertex out of alignment).  So it’s a great prototype for giving insights into how to fix it next time, but overall the text works well and hanging against a coloured background works well as a piece of typographic art.

Top