Land Dreadnaughts

I've long had a fondness for the ridiculous multi-turreted land-dreadnaughts that came briefly into fashion in the late 1920s and 1930s. The Soviets were the only ones to actually try using them in action, as far as I know, with their T-35 and T-28. The British A1E1 Independent never made it out of the prototype stage. The Germans built six Neubaufahrzeug of slightly different conformation,, but the nearest they came to seeing action was being sent to Norway in 1940, largely as a propaganda exercise.

I've just printed two types of Neubaufahrzeug in 1:100 scale (15mm) on my 3d printer, and this is the first one I've finished painting.


Neubaufahrzeug Type A

Neubaufahrzeug Type B
I got this model from TigerAce1945 on Thingiverse, from his Test Print Factory collection.

The Neubaufahrzeug was Germany's first attempt at a heavy tank after Hitler came to power. It wasn't considered a success, and only five were made.

It printed pretty well. I sliced it in Cura, using its experimental "tree" supports, which came away very easily.

The antenna on the Type A is a bit rough, but that's not completely unexpected — the loop could have done with more closely-spaced supports than Cura's automatic generation gave me; it would probably be best to import the turret STL into Blender and rebuild it with dedicated support structures in place. The struts are quite delicate pipe structures printed at a fairly steep angle, so the stair-stepping is a bit visible, even printed at 0.08mm layer height; I don't know that there's a lot I could do about that, considering the limitations of my printer. I could try printing it a 0.04mm, but it would be a long print, and I'm not sure it would be worth the effort.

The track-guards are another tricky printing area. The whole guard slopes slightly down towards the rear of the hull at maybe 3°, and so rather than a smooth length of plastic, I got a series of longish plateaus as each layer was printed.

I'll do as much cleaning up with sandpaper and scrapers as I can, and see how it goes. There's not a lot of fine surface detail to worry about, so a spray with sandable filling primer will probably take care of any remaining striations.

Zvezda Sturmtiger (15mm)

This is another 1:100 scale Zvezda snap-together kit, the SturmMörser Tiger, or Sturmtiger.

In the photographs I looked at of the Sturmtiger, none of them appeared to have any unit markings or vehicle numbers, which made life a lot easier for me.

This is my first attempt at the 1944 "ambush pattern" camouflage scheme. I'm not entirely dissatisfied with the result, but I feel it could be better — I just can't quite put my finger on just what's wrong.

Zvezda Tiger II (15mm)

I picked up a couple of these Zvezda Art of Tactic snap-together kits the other day, not because I have any immediate use for them, but because I had a sudden fancy to play around with German late-war three-colour camouflage patterns.

It's a nice little kit, and pretty straightforward to put together. The track detail is the only thing that really lets it down, but I can live with that in a wargaming model.

Operation Unfathomable

This just arrived for me from DriveThruRPG, the softcover POD book by Jason Sholtis of his long-running Swords & Wizardry FRPG campaign.

I got it primarily because I really like Jason Sholtis' artwork. I seldom used pre-packaged adventure settings or modules; I never really feel they're very satisfactory. However, this book is full of the sort of weird shit that does appeal to me, and I will, no doubt, loot and pillage it until it's been completely strip-mined and left crouching by the side of the road, huddled and despairing.

Two-page title spread
I bought the PDF when it was first released, and liked it so much that when they sent out an email with a discount code for the hard copy for PDF buyers, I thought "I'll have that", and so I did.

I do not regret it one whit.

PLA Dorchester in the raw

Model design by M. Bergman on Thingiverse
Excellent 1:100 and 1:200 scale models, highly recommended.
 Suspecting that my issues with stringing were largely to do with using damp PLA filament, I put the reel in a dehydrator for about six hours, and then in the hot water cupboard overnight.

Then I put this model on to print, a 1:100 scale (15mm) Dorchester ACV. It was about an eight hour print job at 0.08mm layer height, and I took the opportunity to try out Cura's new and experimental "tree" supports — they seem to have done their job, and they were easier to remove and left less of a mess than the usual zig-zag supports.

Stringing wasn't entirely absent, but it didn't start to appear until towards the very end of the job, when the reel had been exposed to our winter air for eight hours, so it was pretty minimal. It's something that I'm going to have to deal with though, as 20 or 30 hour print jobs aren't all that unusual. I may have to build an enclosed filament feed that I can load up with a dessicant or something, if I'm going to try any really long prints in the winter-time.

Anyway, now the reel is back in the hot water cupboard to dry out again.

Supports removed, looking not too bad.

Printing fault in the rear left wheel,
doesn't seem to have affected anywhere else.

Painted 15mm Dutchmen

Here are two prints of the 1:100 scale (15mm) Vickers Commercial Light Tank of 1936 I showed before in its semi-raw state, now all painted up. Each has been printed in a slightly different manner, though the fundamental print settings are largely the same.

The one on the left was printed up on its tracks, and the supports necessitated by that have made a real mess of the running gear. I will probably go back and turn all that crap into mud and churned up vegetation at some stage; at the moment it's just painted more or less like it.

The one on the right had its running gear printed as separate components, lying face up on the print bed, and the results are very much cleaner, though I did have to remove a great deal of "string" from the print. I'm learning a lot about just what and what isn't worth worrying about for these little FDM printed wargaming vehicles, and one thing that's not worth worrying about is detail on the inside (the hull side) of the bogies and tracks. That may be important for fine scale modelling, but for my purposes it's pointless complication that will likely never be noticed.

Even after this short experience of FDM printing, I'm becoming ever more eager to get myself an SLA resin printer. That's a way more expensive printing method though, both in terms of the printer itself as well as the consumables, and it's going to be a long way off. Still, I can dream.

New (Old) Monsters!

 Today, this arrived for me from DriveThruRPG , a softcover copy of the D&D Creature Catalog, first published in 1993 in this form, and now re-released as a print-on-demand volume.

It was intended as a companion volume to the Rules Cyclopedia, published in 1991, and both of them gathered together material scattered through a multitude of BECMI volumes into these two books.

The Creature Catalog presents a couple of hundred new (at the time) monsters, and it includes an index of all the critters for the system up to that point, detailing which could be found in which volume.

This reprint is based on scans of a printed book, rather than the original artwork. As a result, the tones in the greyscale images are noticeably compressed, and they tend to look rather muddy. The artwork is not, for the most part, particularly good, though I've seen much worse in other published RPG material. The images are good enough to do their job, which is to provide some reinforcement to the printed descriptions. The text is quite clear and legible, and the paper is acceptably white and opaque — it's not as luxuriously heavy as the original AD&D manuals, but then again you so seldom get that sort of quality in RPG material these days (or ever, really).

I have an original hardcover copy of the Cyclopedia, and although I've never actually played D&D using it, conceptually speaking it's my favourite edition of D&D ever. I like the idea of having the whole system collected into and playable from a single volume (which is one reason why I'm so fond of OSRIC, the AD&D retroclone) and I also really like the "change it if you don't like it" ethos of the system, which was quite the opposite of the rather restrictive and regimented AD&D. To be fair to AD&D, a big part of why it was published in the first place was to bring some consistency to D&D games throughout the world, and it did do that, though whether that was a good thing or not I leave to your own personal opinions.

RPG Irony

The most dangerous possible thing for any party of PCs is a climb of any kind, no matter how easy.

Unless you actually want them to slip and fall to get your scenario to move along. Then they all suddenly climb like fucking Spiderman.

1:100 Dutchman (FDM print)

This is the Vickers Commercial Light Tank of 1936, known colloquially as the "Dutchman" because the major purchaser was the Dutch East Indies Company. It was produced in several variants.

I printed this model on my Ender-3 FDM printer in PLA (0.08mm layer height). It's 1:100 scale for use in 15mm gaming, and it's about 35mm long — I've included a 15mm Peter Pig WW1 British officer for scale. I've given it an overall base coat of Khaki Green No.3, but I haven't gone any further with the decoration as yet. I included a separate turret with an open hatch. in the .STL, and eventually I'll find myself a commander to stick in it.

I'm pretty happy with it so far, but I'm wrestling with stringing on my prints which makes for a lot of cleanup work. I suspect the culprit is moisture build-up in my filament; I've got the roll in a dehydrator at this moment, and I'll try another print when it comes out. I've seen various ideas for easy-peasy PLA stringing removal on YouTube, but none of them are really very satisfactory, especially on a model like this which is 99% rivets, so it's just the arduous one-at-a-time removal for me.

I shall have to get a reel of black filament, which would make painting little models like this a lot easier. All I have at the moment is bright blue, which stands out like dogs' balls if you miss a spot.

You can have one* (or ten) for your very own: it's available for sale from my page at Shapeways. Unlike most of my models there, I haven't made this one available in their white nylon, but only in SFD resin. I could never get a printable model in the nylon without some major redesigning.

* Though not, as yet, with the open hatch turret. If there's a demand for it, I'll upload that version too.

P40 Warhawk (1:100 FDM 3d print)

 I found this .STL of a P-40 Warhawk a long time ago (I don't recall exactly where) but only now do I have the ability to print it myself. It's a very simple model, with no external detailing at all — no exhausts, no canopy ribbing, no panel lines or control surfaces. It's like an old air force recognition model, in which the silhouette is the important thing and all those other details are superfluous.

Meshmixer model visualization
I've painted it up in colours for the Desert Air Force in North Africa during WWII. At least, I think I have — I haven't done a great deal of research into exact markings. It's intended purely as a wargaming model, so for me, near enough is good enough as long as it's recognizable as what it's intended to be.

I've added a nail head to its belly so that it will stick to my magnetic flight stands.

Note: Painting RAF roundels freehand is a gigantic pain in the arse, but I haven't found any masking method for this scale that isn't just as much of a pain in the arse.

It didn't seem right, since it's a P-40, that it didn't have a shark-mouth painted on. So I went back and did that.

Demon Idol — OSL painting

 Here's my low-rez FDM 3d printed PHB Demon Idol, all painted up.

It turned out all right, if I do say so myself.

I haven't really tried OSL (object source lighting) painting before, so I wasn't quite sure how to approach it, but I think I managed a decent result for a first try.

I ironed most of the larger surfaces with a soldering iron before I started painting, to minimise as far as possible the layering from the 3d printing process. It worked OK for a model like this, but it would be a bit coarse a process for a more delicate model.

Unfortunately, PLA doesn't respond to acetone vapour the way that ABS does, so there's no really decent smoothing shortcut. Some people paint the surfaces with epoxy or self-levelling polyurethane.

PLA Colour Leaching

I've just discovered that the PLA filament I'm using is not colour-fast.

This very simple 1:144 scale model of a P40 Warhawk was printed in blue PLA (0.08mm layer height), and then sprayed with white primer. The blue from the PLA has leached through a bit and stained the primer a pale blue.

Hopefully when the primer has cured properly, it will seal and stabilize the underlying colour. I wouldn't want all my top-coats going blue too.

One good thing: the sandable primer has completely obliterated any remaining layer corrugations (though they were minimal in any case). There's no fine detail on this model to be affected by it, so that's a pure win.

Layer Height Experiment

The Z-axis stepper on the Ender 3 works in 0.04mm increments, and I thought I'd see what would happen if I printed something at exactly that layer height. This is the result.

It's not a success as a model (though it is identifiably a 1:285 scale Vickers Medium Mk.II), but it has taught me a few things.

  • The Ender 3 will indeed print at a layer height of 0.04mm, but
  • It doesn't much like it — there's a great deal of stringing. That might have been partly because of the nozzle heat, but I suspect it's mostly because the nozzle was squishing into layers already laid down.
  • All of the sloping planes are very smooth, but
  • The cleanup required afterwards is excessive.
  • The print time, even for something as tiny as this, is going to be very, very long. I did two of these simultaneously, and it took about four hours.

I printed two at once, as I mentioned — one (shown above) printed lying horizontally on the raft, the other (to the right) standing vertically on the hull rear.

As you can see, the one printed standing up is appallingly bad, and I haven't even bothered cleaning it up any further than just stripping away the supports as far as I can.

I gave this model another go, printed at 0.08mm, t see how that one fared. I got rid of its gun, and just added a locating dimple where it should go — I'll bore it out and use a cut off brass pin instead.

I doubt, somehow, that I'm going to be printing many 1:285—1:300 scale vehicles, and those I do will probably have to be simplified to the level of Axis & Allies tokens.

It's quite usable as a wargaming piece though. Indeed, I've had much worse metal models in the past. The biggest problem with the printed micros is the stringing of this blue filament; the short piece of white filament that came with the printer was much better, but I have no idea who made it or where to get it from. I haven't yet found out exactly what combination of temperature, printing speed, retraction etc. this particular blue filament likes — assuming it likes any of them.
Another one, printed at 0.08mm layer height

Just for comparison, here are some Vickers Mediums (based on the same digital model and in the same scale, but a different version of the tank) printed for me by Shapeways on their SLA resin printers. This is their second-to-finest material.

Benz-Mgebrov 3d Printing Test #2

1/100 scale Benz-Mgebrov armoured car.
Approximately 55mm long and 16mm to the top of the hull-drum.
Here's the same model as in my last post, arranged differently on the print-bed, and exactly as it came off the printer without any clean-up at all.

As I expected, I got much cleaner detail on the wheel hubs by standing them vertically rather than printing them lying face-up. There's still quite perceptible banding on the sloped surfaces, but by setting Cura to print the top surfaces in lines rather than concentric rings, and by enabling ironing, I've ended up with a much smoother top to the drum of the hull.

It's a pity that I couldn't also iron the sloping panels, but even so they're not too bad. This was printed at a 0.12mm layer height, and I could go as low as about 0.8, which would make those surfaces cleaner, but would also increase printing times considerably.

The hull front has lifted a bit up off the bed as the PLA has warped slightly. That doesn't matter much for this model, but it is an issue that will need to be considered for the future. The Cura slicing profile I used printed the model solid, rather than using an infill, which I don't think is either necessary or desirable for this sort of thing, as it increases the thermal stresses on the model substantially. It might not be an issue if the hull was printed vertically instead of horizontally, but I suspect that all it achieved in this case was to increase the printing time a bit. Printing it vertically would probably give me cleaner surfaces on the sloped panels too, as it has on the turret, but I'd lose the benefit of ironing on the drum top.

There's a bit of stringing; I could probably do with tinkering with my retraction settings a bit, but for the moment I'm fairly satisfied with the results. It's not yet perfect, but I'm coming closer.

WSF vs PLA FDM print comparison

Painted WSF on the left, primed PLA on the right.

These are 1/100 scale models of the WW1 Russian Benz-Mgebrov armoured car, a model that I've had printed before, and with all those curved and sloping panels I thought it would make a good comparison test subject. It's about 55mm (2") long.

The one on the left was printed by Shapeways in what they then called White Strong and Flexible, a sintered nylon material. (They've recently changed the names of their materials, though not the materials themselves). I forget precisely how much it cost, somewhere about twelve bucks I think, plus postage.

The one on the right I printed this afternoon, in PLA. It took about an hour and a half, and cost me maybe twenty cents.

Setting aside the issue of cost for the moment, they each have their advantages and disadvantages as models:

  • The WSF model from Shapeways was printed in a single piece; all I had to do was paint it, whereas the PLA model was printed as a set of components that had to be cleaned up and assembled.
  • The detail on the PLA print is much more crisp than on the WSF model. The rivets are actually discrete nubs, where on the WSF print they're really just fairly amorphous bumps, and pretty much had to be painted in to be visible at all.
  • The horizontal surfaces on the PLA print show very marked striations. The print lines are much more marked than on the WSF print. On the other hand, the vertical surfaces are much cleaner and smoother.

I printed the model at an angle to the bed in an attempt to get all those sloping surfaces printing level, or at least, as much as possible. That may have been an error, as horizontal surfaces all showed a lot of banding, and besides which, all the supports required increased the print time quite substantially, not to mention cleanup time. The wheels were printed lying flat, and maybe they would have been better printed standing up — I could live with having to remove supports from around their bottom halves if it meant that the hub detail was rendered more cleanly.

I'm slowly learning a thing or two about what to expect. Now I need to learn a bit more about how to ameliorate the limitations of FDM printing.

Going Lower and Lower

I've added a mirror to the print bed for absolute flatness, with some hair-spray to aid in adhesion
My next experimental print is being run at 0.3mm layer height, just to see what that's like. There's a very perceptible difference in quality between 0.2 and 0.3; the layers (from a 0.4mm nozzle) aren't squished together as much, so the layering is much more obvious, like a stack of wires. Which is, after all, pretty much what it is. I suspect that part of the issue is the colour of the PLA filament; the pure white tends to blend in with itself visually, while this medium blue shows much more tonal contrast.

A lot of what we see here are actually support structures; Cura (the free open-source slicer I use) is VERY VERY ENTHUSIASTIC about where and how many supports it adds, and as far as I know, they're not editable. There are other slicers available (e.g. Solidify 3d), but they tend to be the complete opposite of free, and therefore beyond my reach.

I don't think I'd go any lower than this with the current nozzle; I expect I'd likely start getting gaps and layer separation.

At this very low resolution, printing proceeds (relatively) quickly, and it would be quite useful for rapid(ish) prototyping before committing to a long, high-resolution print. Also, it would be perfectly good for creating armatures for later finishing with epoxy putty and the like.

It's odd to think that this was pretty much the expected quality for any FDM prints, even from quite expensive printers, just a few years ago.

Here's the finished print, an AD&D PHB Demon Idol, about 70mm tall.

The top two show front and back views with supports still in place, the bottom two having removed them.

For the most part the supports come away quite easily, but there are a few tricky places — the mouth was especially tough, and at one point on a leg, the support stripped away a length of wall as it came away. I assume this is because of the relatively low adhesion between layers due to the 0.3mm layer height.

The idol came away from its base, again due to low layer adhesion, and the PLA it's printed in proved to be immune to all the plastic cements I tried. Eventually I just hot-glued it back together.

First Print

After one or two minor hiccups, here's the very first fruit of my 3d printing loins.

It's a ruined house, a terrain piece for 6mm wargaming.

It's not very impressive, and it's not very ambitious, but I selected it as my first print for several reasons:

  1. Being a ruin, it might still be usable if the print crapped out
  2. It's small, so shouldn't take too long to print (it still took an hour and a half)
  3. It doesn't require any supports
  4. It wouldn't take up much filament, and I don't have much filament just yet
  5. It only took a few minutes to build in Blender and spit out an STL file.

All in all, I'm pretty happy with it as a first go. I suspect I could drop the quality settings a bit and still get a perfectly usable model, and a lot quicker too — I shall have to experiment to see where the sweet spots are.

There's a texture on the vertical surfaces — a pattern of vertical lines — that I'd like to avoid, if I could, though it probably wouldn't be all that noticeable from arm's length. Just how to avoid them, or even if it's possible, I don't yet know. The layer lines themselves are barely noticeable and would most likely disappear entirely under a coat of paint.


That first model was printed at 0.1mm layer height. I did another at 0.2mm, with minimal visual difference but printed in half the time. I'll try another couple at 0.3 and 0.4 and see at what point the printing artifacts become too obvious to bear... I suspect, for a model like this, I could go quite high and still get something that will look OK on the wargames table under a coat of paint.

The 3d Printing Adventure Begins

Fresh out of the box

I had a holiday planned for August that, due to circumstance, I am now unable to indulge in. So I took some of the money I'd intended to spend on that, and spent it instead on my very first 3d printer, a Creality Ender 3.

It's an FDM printer, and if it can be got running smoothly, the quality of prints possible from it are quite remarkable — almost as good as output from an SLA resin printer. Not quite, but almost. There are inherent limitations to FDM printing, but for the price (about two hundred bucks for the printer, plus the cost of filament to squirt through it) it's a good place to start I think.

That's a big "if" though, since I have absolutely no idea what I'm doing. First step — okay, second step, after unboxing the thing — is to watch a whole lot of "how to build your Ender 3 printer" videos on YouTube.

A few hours later....

Well, I've got it assembled, I think. I suspect that if one had built one of these things before, it wouldn't take much more than half an hour to get it together. I took quite a bit longer, a couple of hours at least.

The build quality is very good. Everything seems to be solid and square, and everything seems to me, from my position of blissful ignorance, to run freely.

The instruction sheet provided requires very close inspection as you go along; there are no text instructions (though, considering some of the translations I've seen from China, perhaps that's a mercy) and the illustrations, though decent enough, are not entirely without ambiguity in places. In most places, really.

The next step will be to give it some electricity and be ready to cut the power on the instant if anything goes bang or smokes excessively. The plugs are all labelled, and to be completely truthful there is a plugging diagram included, but it's one of the more opaque of the assembly drawings.

There is a very small roll of PLA filament included, but I shouldn't think it would be enough to do much more than prime the extruder and maybe print some tramming figures. I'll have to get some more immediately; if I had thought ahead, I should have ordered some at the same time as I ordered the printer. Stupid me.

Desert Covenanter

 This is my 3d-printed 15mm (1:100) Covenanter model, printed in white nylon by Shapeways.

There were not many Covenanters sent out to North Africa, but there were some — I've seen one in the background of a photograph of a tank park in Egypt, and they're briefly mentioned in one of the regimental diaries.

The Covenanter was a much-maligned tank, but in truth, by the time it was replaced by the Crusader (in which many of the same mistakes were made all over again), most of its issues had been fixed, and it was discarded more because the turret ring was incapable of supporting any weapon larger than the two-pounder rather than for any other issues. Much was made in the official literature about inadequate cooling, but it is notable that this issue is raised in none of the reports by the men who had to actually operate it.

It was retained as a primary training vehicle for some considerable time, at least up until late 1943.

I have a troop of Covenanters painted up now (plus another in Home Service livery). I've experimented with a variety of methods to paint this sintered nylon 3d printing material, and I'm fairly happy with them as wargaming pieces. As I've said before on more than one occasion though, it's not adequate for fine modelling purposes.

Land Dreadnaughts

I've long had a fondness for the ridiculous multi-turreted land-dreadnaughts that came briefly into fashion in the late 1920s and 19...