Financial Reality Check

I've been doing some sums, and it looks like I'd need to sell about eighty to a hundred copies of each of my models on Shapeways in order to pay myself a reasonable hourly rate for the amount of labour put in to designing the things.

Frankly, that's looking wildly unlikely. I don't think there are enough Interwar Nerds in the world, with enough discretionary income, to make this exercise in 3d-Printed Interwar Nerdery a financial success.

I could increase my own markup on the models, but frankly, 3d printing is already expensive enough to be a niche luxury market. I'm just hopeful that the cost of higher-resolution printing comes down to affordable levels within the reasonably near future.

Ah well, for the moment I'm enjoying the process of building these digital models, so I guess until that changes I'll just keep going. It's not like it's costing me anything but time, after all.

3d Modeling: The War Years

I've leaped ahead into the dark days of World War II with my next model — the little Beaverette Beaverbug.

The Beaverette was a rush design, completed in only four days in 1940 when there was a real and credible fear of a Nazi invasion of Great Britain. This is the Beaverette Mk.III, used by the RAF for airfield defence, and called by them the Beaverbug. It's armed with a pair of Vickers "K" guns, but some later ones had the gun turret from the Boulton-Paul Defiant mounted in place of the production turret.

It's available, as usual, from Shapeways at http://shpws.me/LMlV.



Vickers in Paint

 I've painted my 3d-printed Vickers Medium Mk.III, and here it is with some 15mm WW1 Brits from Peter Pig.

The WSF material from Shapeways takes a bit of preparation before it's ready for painting.

The printing process is one of laser sintering in a bed of powdered resin, and when the model is cleaned up the excess powder is blown away with compressed air. That leaves quite a bit of powder still hiding in the various crevices and around detail elements, and that needs to be cleared away or else the paint will solidify it. The ideal tool for this would be a dentist's ultrasonic water pick, but since I don't have one of those, I use various steel modeling probes.

Also, the printed surface is very granular, with a texture like fine sponge. That gives a fine tooth for paint, but it's not ideal for small-scale modeling purposes. It can be smoothed down to a certain extent by burnishing, but it can't feasibly be eliminated entirely.

Hopefully, the cost of higher-resolution printing will come down within the foreseeable future. This plastic is OK, but it's not the best option, and I only entertain it because it's the only material Shapeways offer that I can afford in this scale..

White, strong and flexible frosting

At long last the first examples of Shapeways' "White Strong & Flexible" 3d printing has arrived. It's the cheapest medium available from them, and the only one that would be affordable for 15mm vehicles for the penurious likes of me.


Shapeways' own material render
(enhanced for contrast)
My opinion: it could be worse. It does show quite a decent degree of detail, but the surface is very granular, more so than I was expecting, and it shows nowhere near as much detail as their material render would suggest.

The surface can be polished, using steel burnishing tools, to mitigate this granularity somewhat. It doesn't get rid of it entirely though. Shapeways can do this themselves, but they do it by tumbling the item with little plastic balls, which (of course) pummel out not only the surface granularity, but also any small raised details, so it's no good for my purposes.

New turret and wheels for
Battlefront's 15mm Rolls Royce
armoured car
The final proof of the pudding will be in the painting, of course. I'll be interested to see how that turns out.

I suspect that the material would be fine for wargaming pieces, but it's certainly not fine enough for general scale modeling purposes.

More digital armaments

Port
The Carden Loyd Light Tank Mk.VIII from the mid-1920s. Now available from Shapeways in 1:100 scale and 1:56 scale.

This is the vehicle on which Vickers based their enduring line of light tanks, after they bought out Carden Loyd in the 1920s..
Starboard

It's tiny. At full size, it's less than three and a half metres long, which is a little bit smaller than my car — although it would have been a lot heavier than my car, and also much slower and less fuel-efficient. Though my car doesn't have a machine-gun mounted, which I sometimes regret when I'm in traffic

Yet another Vickers

I'm waiting on a book full of scale drawings of interwar military vehicles to arrive from bookdepository.com, and I thought while I wait, I'd whip up a quick Vickers Medium Mk.II. It would be simple and easy, I thought.

I was not entirely correct.

However, it's more or less done now; there are just a few more bits and bobs to add and then it'll be another one off to Shapeways.

It would be nice if I could actually afford to buy the stuff I'm designing. That would be really nice.

NOTE:
It's now available in 15mm (1:100) scale on Shapeways for a measly twenty yankee dollars.

Tiny Vickers

Now I've got some paint on to the 3d-printed 1:300 scale Vickers Medium Mk.III, and it's ready for its solitary lonesome duty on the tabletop battlefield.

Note that what looks like three tanks is actually only one, thanks to the magic of Photoshop trickery.

Now with added markings!

Leaving

It's really quite small.

My God, It's Real!

The raw resin
 I'm more excited than is truly proper for a Gentleman of a Certain Age, as the very first actual physical 3d-printed results of my labours has just arrived.

The model, primed for visibility
It's the two-piece 1:300 scale Vickers Medium Mk.III, in Shapeways' Frosted Ultra Detail resin.

I'm pretty impressed, I have to say. And it shows me a lot about what I need to be adjusting and exaggerating when designing in this very small scale, and what can just be left to be suggested.

The rivets, for example: they're visible (in places), but they don't get across the rivetiness of the tank. There could be fewer of them, but larger. There is visible striation on vertical surfaces, though it isn't visible from arm's length, and I doubt that it would be visible at all on a pewter casting.

I'll be interested to see how the Frosted Extreme Detail resin fares. It's more expensive, but in terms of the overall labour and expense involved in creating a master for moulding and casting, the difference isn't that great, and hopefully it will deal better with vertical surfaces.

My digital legions increase!

Since I'd already gone to all the effort of building the running gear for the Birch Gun, I thought I might as well make my next model the Vickers Medium Mk.C, from about 1925, which uses basically the same chassis.

I don't have scale drawings for this one, so I'm working by best guess from some very old, grainy photographs. This particular one is designed at 1:100 scale for 15mm; I'll probably do a redesigned version for 1:285 (6mm).

I don't believe it saw service with the British army, but Vickers supplied some to the Irish, and to  the Japanese who modified it and turned it into the Type 89.

Birch Gun for sale, one owner, low mileage

I don't know why the starboard running gear is bright white
while everything else is grey.


But not the port side, which was just mirrored from
the starboard.

Looking down into the fighting compartment

The view from behind.
I've built a 1:285 digital model of a Birch Gun from 1925, and put it up for sale in my brand new shop on Shapeways.

The limitations of the printing process means that things like the gun shield track guards and tracks are hugely over-thick in scale, but most of the model can just be resized upwards to form the core of a 15mm model.

On the plus side, this is the first model I've submitted to Shapeways that passed all their automated pre-acceptance tests first time, so I'm clearly getting a handle on how to design for the materials.

Digital Modelling (again)


I finally figured out (with the aid of the internet) how to fix the issue that was preventing me from importing the running-gear file, so I could put together this mock-up assembly of all the components.

For 3d printing purposes, the colour is really irrelevant. I thought I'd give it some though, because why not? In the process, I discovered that the turret and sub-turret files really need a bit of refinement. They were my first attempts in Blender, and it does rather show.

Now it's on to 3d printing the thing. I don't have local access to a printer, unfortunately. Even a relatively low-resolution printer would be useful for testing purposes, since the alternative would be to get Shapeways or one of those companies to do it, which would likely end up costing me an arm and a leg before I get a successful product.



Birthday Booty

I had a birthday party last night. The actual birthday was a week ago, but the weather was shitty, and we like to do our partyish socializing out in the back yard around a burning brazier, so we put it off until this weekend.

People brought me much wine, enough to ensure that my liver will suffer terribly.

Steve brought me an expansion to Kittens In A Blender, so we're no longer limited to only four players. Now we're limited to only six players, which is better.

Jo & Bastian brought me a whole bunch of ancient roleplaying miniatures, mostly Ral Partha, but there are some others in there as well. A few of them I already have examples of in my collection, but most are new to me (and I certainly never mind having more than one of a thing).



Now they can all go into a bath of Simple Green for a couple of days to strip off the old paint, ready for new paint.

All in all, I'd rate that birthday a Good Score.

Digital Modelling

Vickers Medium Mk.III c.1930
Back in the far distant past, I got pretty good at writing PovRay scene files in a text editor. Then a graphical front-end for PovRay appeared, called MoRay, and it was really good until it no longer worked with the new version of Windows (Vista), and the guy who originally wrote it lost interest in updating it. I believe he passed it on to the PovRay Foundation, but the last time I tried it, things hadn't really got any better.

Anyway, that was some years ago, and although I've dabbled from time to time I haven't really made the effort to come to grips with newer software.

As may have become apparent, I have an interest in the clunky rolling boilers that passed for AFVs in the inter-war period. Unfortunately, in my preferred gaming scales of 15mm (1:100) and 6mm (1:285 - 1:300), that period is very poorly served by manufacturers. However, now that 3d printing is beginning to mature, and reasonably high resolutions are now becoming available from companies like Shapeways, I thought I might see what I could do about the paucity of tanks from between the World Wars by building some digitally.

Machine-gun sub-turret
My software of choice is Blender, which is both very capable and free. I've tried it out before, when I was looking for a replacement for MoRay, but back then its UI was excruciatingly bad — it had clearly been designed by and for severely autistic engineering nerds, with little or no account taken of the needs or limitations of normal people, and with some truly unique assumptions about what was intuitive. Happily, it has been vastly improved since then. It's still not very straightforward to learn, but it's so much better that it might almost be a different program altogether.
Main turret

I decided to teach myself how to use Blender by building a Vickers Medium Mk.III, which (as far as I know) isn't manufactured by anybody anywhere in either 15mm or 6mm scales. So far, all I've got is the main turret and a front machine-gun turret, but it's a start.

I am learning a lot, but there's a lot more still to learn.




Cupola

Turret with imported OBJ cupola in place
Here's a bit of progress for the morning.

I separated out the commander's cupola to its own file, and added some detailing to it. Then I exported it from Blender as an OBJ file, and imported it back into the turret file.

I don't know why the imported object is so much brighter than the native Blender stuff; I guess they use different default material settings. I haven't yet delved into any of that; it's not really relevant for the purposes of this particular model, which is 3d printing. However, I'll have to take account of it eventually.

I realised that I'd actually created the  main turret and mg turret at about 1:50 scale, not 1:100, so I had to do some resizing so that all the components would be consistent when I come to put them all together.




This will be about as far as I get today, I think. I've got the main shape of the hull complete, now it's just a case of adding all the bits.


Interestingly, the files size for the MG turret is much larger than for the main turret. I'm wondering if that's because I used UVSpheres instead of IcoSpheres for the rivets? I don't really know.

So far, modelling has been pretty straightforward — it's all primitives really, messed about with to one degree or another with Boolean sculpting. The side frames will be a bit trickier, I think; I might have to investigate some more advanced construction methods.



A little bit more fiddling about before I go to sleep. So many rivets.




February 3rd

Very slow progress today — most of my time was spent flailing about just trying to make that little grab-handle on the front of the driver's cupola. It turned out to be ridiculously simple once I found out how to do it, but the Boolean unions I was originally trying were driving me up the wall.

Apart from that, now the driver's cupola is also covered in rivets, and we have some track guards. It's progress, if slow.



February 4th

I've made a start on the rather complicated running gear, which has been an opportunity to try out importing and extruding SVG curves. It's not an entirely straightforward task, but it's really not terribly difficult either.

I created the form for the framework in CorelDraw, exported it to SVG and it went into Blender without a hitch. For some reason, Blender ignores the scale in the SVG file, and it appears in Blender's stage very, very tiny — it would be easy to assume that the import had failed unless you're prepared for that. You have to zoom right in to find it, and resize it a lot to get it up to its proper size. Also, the origin point for the curves is set way off to the side and needs to be reset. I have no idea why either of these things happen, but it does.

Since the finished model is intended for moulding and casting from a 3d printed master, it meant that I could simplify the framework a bit and make it solid, rather than the sandwich construction the original vehicle had.

Next up, wheels and tracks. That should be interesting. Once I've got this whole side finished, I should be able to just duplicate and mirror the whole thing to get the port side.


Running wheels and return rollers added. The drive sprocket and front idler will be ore candidates for an SVG extrusion I think. They could be built up within Blender, but it would be a chore.

Oh yes, and I'll be needing a track tensioning screw as well.



Well, putting the tracks on was a lot easier than I expected. The Array modifier saved a lot of time and fiddling around, though being able to set an offset value for simple duplications would have been almost as quick, and a bit more flexible as regards individual link rotation — it would have been easier to emulate track sag, for example.

The view from the front...

...and from the rear.

These things start getting a bit complex.


February 8th

I've gone back to the hull for the moment, to add all the clutter on the engine deck.


The exhaust pipes are going to be an interesting learning experience. I'm not really sure how to go about them, though I know that bendy pipes is something Blender can do.

Apart from the pipes, pretty much all that's left to do is the storage boxes that run along the track guards amidships. That and trying to work out what's wrong with my running gear file that makes it unimportable, and how to fix it.


As it turned out, the exhaust pipes were easier to build than I had expected, since I could just build them up from a filled, bevelled bezier curve transformed to a mesh.


On to the last leg. This has proven to be rather an enjoyable project.

The storage boxes have been added, and that is about that for the hull. No doubt there are more bits of detail that could go on, but essentially, it's done.







Hills — the search for perfection continues

I've started another couple of hills, with the lessons of the first lot in mind. In the foreground is a long (about 800–900mm) rocky...