Shapeways -

For the past week and a half I've been spending most of my "free" time making up a Polish TOPAS-2AP model for my Shapeways "shop", along with various related vehicles, and I thought I might post some words about the process.  Since I have no intention of starting collecting a Polish 7th Landing Division force (touch wood!), these aren't meant for me, so they are really meant for the gaming community at large.  Maybe somebody out there is desperate to field a 7DD force, but is just stymied by the lack of available models in 1:300th scale...

OT-62A as rendered by Shapeways

 I started out by creating a plain unarmed OT-62A, as shown on the right.  Now this, of course, is a Czechoslovakian model, so I should be able to find a use for it in my collection.  Well, at least I would be able, if I hadn't already got dozens of Heroics & Ros BTR-50PUs standing in for OT-62s, that is...  Once I had this model done, it wasn't so hard to modify it to produce a bunch of variants.  For example, there is the standard OT-62 (often called an OT-62B) that features a small turret on the right cupola (from the viewpoint of the vehicle's driver) mounting a machine gun, and, optionally, an 82 mm recoilless gun on the side of the turret.  Not that these weapons can be printed out in plastic - they are too thin.  But the turret to hold them should be fine.  Then was the OT-62R3 command variant, with a generator on the back, just like the BTR-60PU.  Next was a DTP-62 maintenance vehicle featuring more stuff on the back deck. 

WPT-TOPAS in sketchup

Then I made three Polish versions - a TOPAS, which is simply an OT-62A with a higher engine intake on the back deck so it's more difficult to flood in choppy marine conditions, a WPT-TOPAS, which is essentially a DTP-62 with a large (machine) gun shield over the right cupola, and finally a TOPAS-2AP as shown in the first picture, and which has a centre-mounted turret mounting a high-angle 14.5 mm machine gun (again, my model just has a hole for a length of 0.3 mm diameter wire to be inserted to represent the actual gun). The software I use to make these models is Google's Sketchup, because it's freeware.  I've never used anything like AutoCAD, just Sketchup, which I first used before having my house built in Tai Tapu.  Which alas, I have never lived in, because it's in New Zealand, and I am still in Japan...   Note that Sketchup works with polygons - curved surfaces have to approximated by curved lines (see the gas cylinder heads in the WPT picture to the left).  However, you are free to make a circle be composed of pretty well much as many segments as you want, so this isn't a practical limitation in terms of modelling a curve (although curves can take a lot of time to get to mesh correctly with the rest of your model).  Sketchup was initially written for architectural-sized designs in mind - occasionally it has problems connecting things when the dimensions get really really small, and it also fails to tell you what your dimensions actually are when you get under 1 mm in size, which can be a serious problem when you work at 1:300th scale! 

Netfabb in operation

Accordingly, I make up all my models in silico at 1:1 scale, and only shrink them by a factor of 300 as the antepenultimate step before converting them into an "stl" file, which is the format the Shapeways printers require. (The ultimate step before uploading to Shapeways is checking the stl file for errors in another piece of freeware called Netfabb, because there will inevitably be some.  For example, the model isn't a single "shell" because there is a plane hidden inside it somewhere.  Or it isn't a closed shell, because there is an opening in it somewhere.  Many of these errors can actually actually automatically fixed by Netfabb, like the ones shown on the left picked out in yellow.)

I'm not a very efficient CAD worker.  Every time I finish a model I invariably think to myself "it's got C2 symmetry - I'd have saved myself a third of that time by just drawing half of it and mirroring the image" - and promptly forget that whenever I come to make up my next model!  And there are no doubt all sorts of software shortcuts and tricks and things that I have no idea about that would make life much easier for me in terms of draughting the designs. So it takes me something like 16 hours of work  - or more - to come up with a new design from scratch like my OT-62A; any subsequent modifications are obviously much easier; they can usually be done in an evening.

Once one of my models is on my Shapeways "shop" anyone can order it for themselves.  It isn't an ideal set-up.  For one thing, Shapeways currently has a ridiculous postage policy - it's  20 dollars (US) minimum - in addition to the handling fees.  So this completely negates the primary advantage of on-demand 3-d printing - producing stuff only as required.  My father thinks I should be buying my own printer (he models as well, but he doesn't work in anything smaller than railway's 00 scale: 1:76.2 in other words) - but I can't justify the several thousands of dollars this would currently entail...  The handling fees are actually bundled into the amount of plastic printed out, and not charged "separately".  The more plastic ordered, the less the handling fee, at least, per unit volume of plastic.  This is actually quite a reasonable way of doing things, but is very frustrating for someone working at 1:300th scale, because the charges really kick in once your print volume gets down to 1 cubic centimetre or so in size.  Just what you are looking at with a 1:300 vehicle, in other words...

So if you order a single OT-62 of mine, the amount you will have to pay Shapeways, in US dollars, not counting the postage, is currently 2 dollars 93 cents, and that's for the very cheapest plastic.  Far too expensive for the average gamer, who is likely to need lots of them...  And if you order, for example, 4 of these in one go, you will be paying four times that amount.  However, if I take my model, multiply it four times in Sketchup, and connect them with a sprue, to make a single item of four OT-62s, the amount you pay Shapeways is only $6.89.  Barely twice as much, not four, because you are ordering just "one" item, not four, so you pay less handling fees.  Now seven bucks for four simple APCs is still not exactly cheap, but it is much more reasonable than 12 dollars!  But to get the price down like this means I have to make up yet another model for such a multiple set on a sprue - both in Sketchup, and in my "shop", and that will take yet another half an hour of my time...   

What's more, the printer operators often arbitrarily reject models because they are "too small" - even if they have been printed out successfully before, which is very frustrating (you will get informed about this only after your order has been placed, and the rest of the order has gone through successfully, and I will get informed about it afterwards too, in which case I will have a fruitless argument explaining why they they can't read their own guidelines). Of course, occasionally it actually is my fault, and the model has structural details that are too delicate for their printing standards to cope with (or to be more exact; their post-printing handling procedures to cope with).  I'm getting better at avoiding these mistakes, but it's a long road...

Having said that, 3-d printing is clearly going to come on leaps and bounds in the immediate future.  The current print resolutions aren't all that good, at least for the cheap plastics (the more expensive ones at Shapeways cost up to 3 times as much as the prices I have quoted above!), so traditional moulded metal currently has a clear advantage there.  But moulded metal is only commercially viable for large enough production runs to justify the cost of the physical sculpting and the mould.  OK for bog-standard items, but not for the kind of model we all only need one or two of.

Of course, in the longer term, we won't actually be using figures anyway.  We'll all be using holograms.  Think how much the cost of touch-screens has fallen recently.  Soon we will have gaming-table sized touch-screens in our houses, and they will make great gaming boards.  And since holographic TV is already being prototyped, it's just a matter of time before we will be virtually pushing our troops across the table rather than doing it in meat-space...