First Infinity Tournament: Redcap’s RECON

military-orders-iconYesterday I entered my first Infinity event. My home shop, Redcap’s Corner, hosted a beginner’s tournament using the unofficial RECON missions. This was a great driver to actually start playing. I’ve had a few models for two years now but other than a tutorial by my buddy Steve this summer and another by Alex this Thursday, I’d never gotten them on the table. Only four players showed up, but it was an excellent day of gaming. I’ve totally got it now: Though almost overwhelmingly complex to get started, and I think unnecessarily convoluted in rules presentation as well as jarringly inconsistent here and there in some details and mechanics, Infinity is an incredible game with a ton of strategy and narrative.

Though just beginner games, interesting things happened throughout, so this is a battle report for the day. There are a few more photos in my Flickr gallery.

Wounded but undaunted, my Father Knight leads the charge.

Wounded but undaunted, my Father Knight leads the charge.

Army

RECON is basically half size games. Lists are 150 points, with a single combat group and fireteam permitted, along with some other restrictions. The table is cut down to 2’x3′ and the ITS mission scenarios reduced to match, with all specialists (forward observers, medics, and hackers) able to activate all objectives.

I fight for the PanOceania technological hyperpower, specifically its Military Orders sectorial army. Sadly I chose that sectorial because it looks awesome. Coming from 40k as a Space Marine commander doggedly serving the gothic emperor, I have a special appreciation for religious nut jobs running around in power armor with swords while everybody else is packing oversized rifles, orbital bombardments, and so on. Only after I had a handful of models painted did I come to know that universal consensus is that the faction is really hard to play competitively, and certainly one of the hardest to start with… Most critically, the models are so expensive in game points and you field so few that you just can’t afford to make mistakes. As the kids say: Womp womp.

In any event, I brought the following army list:

screenshot_2016-10-30_07-31-30

I lead off choosing a Father Knight just because the physical sculptures are amazing. I really wanted to field both of mine (blister + Icestorm variant), but putting 2/3 of my points into two models seemed excessively foolish. However, I do think the unit is solid. It’s weird that a strong close combat fighter doesn’t have a Martial Arts skill, but they’re tough models with 2 wounds, 5 armor, and 9 bio-technological shield.

The Fusiliers are my cheapest options for the specialists needed to claim the scenario objectives. I didn’t field a hacker mostly to conserve points and a little bit to minimize the game mechanics involved in my squad. Taking the maximum of three permitted by the sectorial also got me to 7 models (plus the AUXBOT) generating 7 regular orders, which felt viable at 150 points. Anything less I felt was very risky in terms of actually being able to do anything, even before casualties started.

The Spec Sergeant with Auxiliary Bot essentially filled out points I had remaining. The bot wields a heavy flamethrower, which pulls at my heartstrings as a Salamanders-rules 40k player. But it’s also half again faster than standard infantry, which I thought might be handy for a late move on an HVT or similar action.

The last two models were my main strategic army choices. The Crusader has Airborne Deployment L4. The Spec Sergeant Forward Observer has Infiltration and TO Camo. Both give me a way to deploy upfield, obviating spending precious orders just moving models forward. With its boarding shotgun, the Crusader can hopefully run amok in my opponent’s backfield and take out weak order generating cheerleaders, reducing their army’s ability to act. Meanwhile, the Spec Sergeant hides in perfect camouflage until the moment comes to break for an objective it can activate as an FO specialist.

Everybody watching Thursday's tutorial game lead by Alex.

Everybody watching Thursday’s tutorial game lead by Alex.

Round 1

First up was Rae and his Yu Jing in the Exfiltrate mission. With both of us super unfamiliar with the rules, this was a sloppy, slow affair. We only made it through one of the standard three rounds. One notable moment though was actually in deployment, when an Oniwaban with a lethal monofilament close combat weapon failed to infiltrate and had to run up from his backfield. Unfortunately a Ninja did pop up right next to my Father Knight and couldn’t be kept out of close combat. The humble AUXBOT saved the game though in the final actions, just barely managing to tag the charging Oniwaban with its heavy flamethrower and burn the ninja to a crisp before it could grab a civilian.

Outcome: Draw, 0-0 with no civilians rescued and no HVTs claimed. Fortunately for my standings the other table came to a 0-0 draw as well. Most importantly though, coming out of this match I felt reasonably in the groove with the basic mechanics of the game.

Analysis: The AUXBOT was MVP, deployed to guard one of the enemy civilian objectives and doing exactly that to preserve the draw. One note was that the hidden Ninja attack on my Father Knight essentially forced me to reveal my TO Camo Spec Sergeant in my reactive phase in a desperate attempt to stop its charge. Didn’t matter here, but in general that would be unfortunate, wasting the hidden sergeant’s ability to make a surprise shot, late grab for objectives, and so on.

Ninjas, attack!

Ninjas, attack!

Round 2

Next was Alex and his Haqqislam in Seize the Antennas. Fortunately I saw his list Thursday when Tim borrowed it for a tutorial. So I knew to watch out for the sniper, missile bot, and the hacker lighting up my models to call down orbital bombardments.

Also very helpful was a short discussion we all had beforehand about the different tables. Bonsky (the TO) pointed out a slight asymmetry on this map: One deployment zone had a wall closing off an alleyway which was clear on the other side. I got to choose deployment zones, so I chose the side with the extra wall. Partly that would prevent any shenanigans with models appearing near my table edge, running behind the buildings, and seizing the antenna objective in the alley. More importantly, if Alex had the extra wall, it would potentially cripple the effectiveness of my Crusader combat jumping into his backfield by limiting firelanes and hampering movement.

Missile bot.

Missile bot.

That decision had huge impact. Alex quickly seized two of the three antennas and set up a bunch of models in defensive postures around the midfield. However, the Crusader came down right in the middle of his deployment zone and went on a rampage with its boarding shotgun. First it went forward to execute a specialist on the antenna—I was worried about Haqqislam doctor tricks bringing her back if left just unconscious—and then ran back behind the buildings to take out both baggage droids and successfully draw and rebuff fire from one of the rooftop teams. Cutting those three orders from Alex’s next turn put us on much more even footing, and taking out the specialist poised to run up the center toward the next antenna didn’t hurt.

Downed baggage bot.

Downed baggage bot.

Meanwhile, my Fusilier Forward Observers advanced as a fireteam duo toward the central antenna, carefully using the buildings to stay hidden from long range shooting teams on the rooftops. Then they dashed across an alleyway, dodged the incoming fire, and climbed up the objective building. After a brief but intense firefight with the Haqqislam specialist guarding it and the various models arrayed to shoot on that position, they managed to huddle down and seize that antenna.

From there, Alex’s lieutenant made a play for my home antenna, winning the firefight with my Father Knight leaping at the threat. That put me into loss of lieutenant, but I had just enough command tokens for the final turn to convert the resulting irregular orders to regular. That gave me just enough to get it done. Suddenly my Spec Sergeant in hiding revealed himself on a nearby rooftop, slid down a ladder, and ran up to recapture the antenna. At the same time, with the central antenna secured, the Fusiliers jumped from their position and ran through incoming fire from the rooftop enemies to secure the Haqqislam HVT, counterbalancing the loss of my own HVT.

Battle!

Battle!

Outcome: Victory for the Military Orders 8-5, with two and one antennas respectively claimed, and both HVTs secured by the opposing sides.

Analysis: My plan to use the Crusader to take out vulnerable order generating units in the enemy backfield worked perfectly. There was also an interesting decision at the very end to send the Fusiliers either toward the enemy HVT or to clear off a light infantry model on my HVT. I went for the former, which worked out. My rationale was that if the Fusiliers were going to take roughly similar fire along both courses, then clearly I should go for the objective that just required moving into place. Otherwise I’d risk failing rolls and not taking out the enemy model.

The real clutch of this game though was deploying the TO Camo Spec Sergeant very defensively, and holding him hidden until the absolute end. At multiple points I was really tempted to have him pop out for some surprise shooting. But it was much more valuable to still have him perfectly positioned for a last minute critical objective grab.

Checking to see if there's any way to get those pesky Fusiliers.

Checking to see if there’s any way to get those pesky Fusiliers.

Round 3

The last round was a minefield, literally. With a central tech-coffin and its data package as linchpin of the Smash and Grab scenario’s capture-the-flag, Tom had his Nomads deploy mines and crazy koalas (basically a hopping mine) all around it.

I tried to trigger all the mines and koalas at once by dropping my Crusader into the thick of them. But it dispersed and only triggered just enough to get killed, not enough to clear the area. So then my infiltrating Spec Sergeant revealed himself right next to the tech-coffin in slim hopes to quickly grab the data package. He didn’t weather everything that came at him, but did set off most of the remaining mines.

Nomad remote oversees the battlefield.

Nomad remote oversees the battlefield.

With that a Nomad Moran charged onto the objective building but fortunately failed to operate the tech-coffin and retrieve the data package. That gave one of my Fusiliers time to climb up on the building and take down the enemy, but not enough to grab the data package. Unfortunately, a Jaguar was able to use smoke grenades to make his way unhampered to the tech-coffin, operate it (this mission did not require specialists), and run away with the package before I could respond.

With the endgame approaching, the chase was on! I burned my command tokens to make most of my remaining squad run after the Jaguar in coordinated orders. Charging across the rooftops, rebuffing mines and repeated hacking, the Father Knight was able to gun him down from the back in an alley. A Fusilier then grabbed the data package and ran for safety while everybody else set up a kill box on the only path between him and the remaining enemies. The enemy lieutenant that came leaping down was thus easily dispatched, while my defenders also claimed the enemy HVT.

Battle!

Battle!

Outcome: Victory for the Military Orders 6-2, after Tom scored the data package at the end of the second round but I scored at the end of round three when it was worth double for the endgame bonus, as well as claiming his HVT.

Analysis: Sacrificing the Crusader in attempt to trigger the mines wasn’t ridiculous, but I probably should have sent it against Tom’s backfield baggage bots just like I did with Alex. Partly I did not because I thought his units back there were more deadly than they are, partly there wasn’t enough space to ideally position the drop zone template. But mostly I wanted to keep the TO Camo Spec Sergeant hidden to pop up and grab the data package after the mines were cleared. However, in the same position, I would now probably have the Crusader go to reduce the enemy order count, while either a Fusilier ran up to trigger all the mines or the Spec Sergeant did—it being well placed to do so perfectly—and the Fusiliers came in to grab the objective.

Fusilier engaging a Nomad.

Fusilier engaging a Nomad.

Once the Fusilier was in place to grab the objective, I made a conscious decision to not grab the objective right then. My thinking was that I had no remaining orders to take it anywhere, the Fusilier would probably die exposed on the rooftop so I wouldn’t hold it to score anyway at the end of the round anyway, and it would be better to have Tom potentially fail the test to release the data package after he finally got in position. That logic again wasn’t terrible, but I’m not sure I would do it again. The Fusilier might have survived, and the Jaguar’s smoke grenades made it vastly easier than I expected to just walk up and grab the objective since I couldn’t see to respond at all.

At the very end I failed to put my kill box models into suppressive fire mode. It was a longshot anyway for anything to move through them to get at my data package Fusilier, but I just didn’t think about it due to inexperience.

The Father Knight was probably the MVP here, with his very high BTS 9 shrugging off several critical hacking attacks and enabling him to take down the escaping Jaguar, in addition to knocking down several enemies earlier in reactive shots.

Mines everywhere!

Mines everywhere!

Final Results

With two wins and a draw I came out with the best record and most points. I chose for my prize a small gift card and the Druze Hacker special model that came with pre-release orders of the N3 human Sphere book. I figure it’ll make a good HVT.

Druze Hacker pre-release model.

Druze Hacker pre-release model.

Observations

Father Knight especially but the Military Orders overall feel very 40k Space Marine-y: Expensive in points, very robust, pretty good at everything, not great at anything. Also echoing 40k Space Marines, the Military Orders units being religious and thus not retreating when your army is well depleted is probably a boon at this low point value, since you might hit that point pretty quickly.

Initially I was really worried about not having enough orders, but seven is probably reasonable for this format. With a single combat group in RECON the max is 10, and only Alex fielded that many in this event. I didn’t feel overly limited.

Ironically for Military Orders, I may have fielded the greatest number of offensively capable models. I’m not sure about Rae, but Alex and Tom both had two baggage remotes and another one or two remotes also without serious shooting capabilities. In some sense I spent those same points on Fusiliers. The tiniest bit more costly, but with reasonable attacks. Both games were really tight until the end. But both those opponents also hit a point a bit before that where I unexpectedly felt they had more or less run out of ways to come at me and the objectives even though they had a similar number of models on the table. In Tom’s case the active attackers were mostly dead, and in Alex’s most of the remaining models were long range shooters and supporting models in backfield positions. Meanwhile, I had just a few more models still actively running around midfield shooting and grabbing objectives even though the number of casualties weren’t very imbalanced. Despite being super worried about Military Orders not fielding enough models, I wound up running comparatively more of a swarm of relatively basic troopers. As another sign of that, note that I spent 0 SWC, i.e., spent no squad points on fancy equipment. Those extra bodies perhaps gave me a bit more robustness to make mistakes and more depth to have a strong end game. That gels with my experience in other miniatures games, especially at low points. There’s a lot to be said for simply having more bodies—the other army list I considered possibly better but less fun and cool looking was just a bunch of Order Sergeants and the Fusiliers. I was considering that primarily in terms of order generation, but simply having more models running around would also be a strength. In preparing a RECON list, you need to be really careful that you’re not bringing so many cool toys that it cuts into your basic ability to suffer some casualties but still fight and maneuver.

Haqqislam lieutenant just rudely pushes the paramedic aside.

Haqqislam lieutenant just rudely pushes the paramedic aside.

The only special ability among my models I didn’t use was the Paramedic, through poor placement on my part either getting her killed early or just not being in position to resuscitate anybody. I’ll have to work on that.

Everybody else though came through in at least one critical moment by leveraging their unique traits, so overall I was really happy with my army list selection. The AUXBOT preserved the draw in the first round by having a good weapon to reliably defend an objective. The Crusader and TO Camo Spec Sergeant enabled the second round win. The Father Knight earned his points in the final round by being tough enough to shrug off a bunch of physical and hacking attacks to chase down and eliminate the Jaguar carrying the data package.

However, the MVPs for the whole day were my Forward Observer Fusiliers, quietly but significantly making the central pushes without which the second and third round victories wouldn’t have happened. Both times they charged up and onto a central objective, won firefights to clear it off, claimed it, and then moved on to also secure enemy HVTs. That’s pretty good for a pair of 12 point models! They proved more resilient and potent than expected. I’m also really happy with their paint job, so I’m pretty pleased with these guys at the moment.

Fusiliers saving the day.

Fusiliers saving the day.

RECON

Although pretty solid for a fan effort, the RECON packet itself could stand some cleanup. There are some clear presentation errors, such as the Smash and Grab mission seemingly copying some text from the full ITS missions about specialists but apparently not actually requiring specialists to operate the tech-coffin. It’s unclear if that’s an oversight or intentional. There are also at least a few substantive changes needed that we made on the fly. Dispersements should be cut down to match the smaller table. We played it as 8″, but I would probably do 9″ in the future so that you would definitely disperse out of any zone of control effects you intended to apply. The second player should also only be able to use a command token to cut one of their opponent’s orders in the first turn, and maybe none at all.

More subjectively, for beginners the missions probably either need to be chosen carefully or some of them tweaked down in difficulty. The Exfiltrate mission in particular seems just too difficult to start off with. Either the scoring needs to be tweaked (e.g., shift the emphasis between grabbing civilians versus bringing them back toward the former) and/or the exclusion & saturation zone needs to be diminished.

Welcome to the Sphere!

That said, RECON was a great format for my first event. I learned quite a bit about Infinity and am eager to play it more often from here on. If you’re in the Philadelphia area, Redcap’s has a small crowd of regular players on Sunday afternoons, and people sometimes play on Thursday nights. There is a small Philadelphia Infinity Facebook group to hook up with people for those or other meetups.

Again, there are a few more photos from this event in my Flickr gallery.

See you out there!

3D Printing Tutorial: Modeling & Making a GR-75 Cargo Pod

Jedi, Sith, starfighters, walkers, whatever. No one could argue that the true hero of the Star Wars universe is the homely GR-75 medium transport. Cheap, poorly built, cargo literally simply floating in the void, well past due for retirement, clean lines marred by all manner of greeblies, hull panels invariably burned through… I love it.

A GR-75 evacuating Rebels from Hoth, escorted by X-Wings.

A GR-75 evacuating Rebels from Hoth, escorted by X-Wings.

container-tokenNot unrelated, wargaming is all about supply containers and cargo bits. “Buck up, soldier, we’re shipping you across the galaxy to fight horrible monsters over… some oil drums… and maybe a weapons crate.” So in crafting missions for our Molokh Gambit X-Wing Miniatures narrative campaign event, of course there’s going to be a Supply Depot scenario.

Just like the shuttle mission, that means we need cool tokens for the mission. A couple good 3D printable options already exist. There’s a great 3D rendition of the art on the Den of Thieves scenario tokens that come with the Millenium Falcon (remixed here to be printable w/ no supports). At a larger size is the Class A container from the X-Wing and TIE Fighter video games, which has also been made into a BFF-1 Bulk Freighter.

Neither of those worked for me though. I wanted something that could be printed without needing to cut supports or glue halves, was small like the container token to minimize impact on the basic game dynamics, and ideally had fewer Imperial associations. So I modeled the cargo pods on the underside of FFG’s GR-75 model.

This tutorial is a walkthrough of that, in hopes that newcomers to 3D modeling and printing might learn from seeing the process of constructing this simple artifact. These designs have also been uploaded to Thingiverse as a free download. In addition, I have a previous article up more generally introducing 3D printing for miniatures.

It's a hunk of falling-apart junk, so of course the Rebels would decide the transport could do double duty as an assault ship...

It’s a hunk of falling-apart junk, so of course the Rebels would decide the transport could do double duty as an assault ship…

Shape

Fantasy Flight’s GR-75 model is really fun. There’s a ton of detail, and just enough color to really draw the eye onto them. Most captivating is the underside, with a whole mess of tiny cargo pods sprinkled with bright colors. It’s exquisite.

Fantasy Flight's GR75 model.

Fantasy Flight’s GR75 model.

Pods filling the underside cargo bay of the medium transport.

Pods filling the underside cargo bay of the medium transport.

Fortunately for my task here, these pods have a distinctive but really simple design:

  • The core shape is basically a central box with two large angled wings;
  • The wings are themselves top and bottom triangles with a thin central box;
  • There are recesses on the top and bottom of the wings;
  • The front and back faces have pads with slightly angled corner cutouts;
  • The top is textured with two boxes and two circles.

Easy. The one detail I opted to skip is that the front and back faces have slight angles to their top and bottom halves. This feature would be a hassle for me to model, the loss of it wouldn’t really impair the look, and it’s probably largely a production requirement. Vertical faces of plastic injection molded parts have to be drafted (swept back) so that the piece doesn’t get stuck in the die or the metal scrape across and mar the surface as the two halves of the mold are pulled apart and the component ejected. That’s probably the primary driver of why these faces are angled. This isn’t a concern in 3D printing however, as the part is formed in place with no encasing mold.

Cargo pod deconstruction.

Cargo pod deconstruction.

Scale

With that shape deconstruction, the next question is “How big?” I decided to upscale my cargo pods to 1/270, the nominal scale of X-Wing’s large and small ships. The huge ships, like the GR-75, are done at a mix of scales to tradeoff between manufacturability, cost, gameplay, and look. As you might guess, this is a whole huge topic area online that dives way deep into the “correctness” of FFG’s models versus previously published material, with many lists around cataloging the variances.

In this case, there are no concrete canonical measurements for the GR-75. Length, width, etc., are not published in its starwars.com data bank entry, and they’ve never been given in authoritative roleplaying books, video games, and so on. A well established baseline though, derived from comparison to other ships and as reported by Wookieepedia, is that the “real world” transport is 90 meters long.

It fills me with no little joy that so many people have produced such serious documentation for so many Star Wars ships. #seriousbusiness!

It fills me with no little joy that so many people have produced such serious documentation for so many Star Wars ships. #seriousbusiness!

FFG’s model is 225 millimeters long. Given the 90m length, that means it’s at exactly 1/400 scale (225mm/90,000mm = 1/400). If the model were done at the same 1/270 scale of the smaller ships, it would be 50% longer, 333mm. No doubt FFG decided that would cost too much to manufacture, even if the gameplay was reasonable—there’s only so big you can make a ship and still have it fly around on a kitchen table.

Moving on to the pods, I started by taking precise measurements of the overall extents as well as the vertical & horizontal central boxes. The pods were a little tricky to measure in place, and they no doubt vary a bit due to minute differences in cooling and plastic shrinking rates as they were cast, but some basic numbers are easily attained. Sampling a bunch of them, I came up with a range for each dimension. Those were roughly averaged and then upscaled by the trivial formula (observed * 400) / 270. That output I adjusted ad hoc to produce nice numbers. This isn’t rocket science, and it’s a lot easier to work with a 17mm wide piece than 16.59mm.

Measuring the pods.

Measuring the pods.

Working notes of observed measurements and calculations.

Working notes of observed measurements and calculations.

FFG’s 1/400 scale pods are about 11mm x 8mm x 5mm. Converting that to 1/270 scale, we get pods that are 17mm x 12mm x 8mm, with a 5mm central horizontal box, and a 2.5mm middle vertical box. In real life these would be 4.48m wide, 3.2m deep, and 2m tall. That seems inefficiently small for an interplanetary modular freight system, but not ridiculous. Those dimensions are just slightly bigger than the cargo space of a 15′ box truck (a meter is 3.28 feet). Regardless of real world issues, as a sanity check, that lines up with drawings and other representations of the GR-75 and how it gets loaded.

A GR-75 being loaded.

A GR-75 being loaded.

Modeling

For a small project like this for 3D printing (as opposed to, say, laser cutting or injection molding), I’m a huge fan of TinkerCAD. It’s browser based, so unlike most CAD packages it works well on my Linux laptop. The interface is intuitive, and the feature set plenty for simple pieces. I was able to do a lot with it very quickly, and highly recommend the app. It’s free for non-commercial use, so you can try it out easily.

The first step in modeling the cargo pod was checking the overall dimensions, in case they needed to be fudged for gameplay or aesthetics. To start I made a base to exactly match the core set Container token. I wanted to keep that as the normative playing piece, with pods merely decoration. After making a box from the dimensions calculated above, I could see that it would be a reasonable size to put a couple on the base. In addition, I imported an X-Wing sized TIE Fighter model to see how the pod would look against a ship. It looks a bit small, but I think that’s an artifact of almost everyone picturing jet fighters and star fighters as much smaller than they are. This is especially true for the TIE Fighter, even setting aside FFG’s arguably upsized take on it.

Roughing out the dimensions with an appropriately sized box.

Roughing out the dimensions with an appropriately sized box.

From there I blocked out the basic shape, creating three boxes capturing the overall dimensions of the center box and wings. The same was done for a wing, and then the top and bottom of it replaced with a wedge.The sizes for all of these boxes, notably implicitly defining the angle of the wings, were taken directly from the scale calculations above. Of course, I only needed to model a single wing in detail. As they’re the same left and right, when it was all done I could just duplicate and mirror that side.

Breaking the overall shape up into the central box and side wings.

Breaking the overall shape up into the central box and side wings.

Deconstructing the wing into the middle box and two halves.

Deconstructing the wing into the middle box and two halves.

Replacing the top and bottom halves with wedges.

Replacing the top and bottom halves with wedges.

With the basic shape arranged, I made a checkpoint copy to keep in the background in case I decided to start over from this point, then started on the front pads. The center is simply a box slightly smaller than the center box itself. To make the side pad, I grouped the three components of the wing, duplicated the amalgam, shifted it forward, and scaled while maintaining aspect ratio. Both of these pads were sized and positioned to create a 0.5mm border around and between the pads.

The pads also stick out from the body 0.5mm, but they’re modeled 0.75mm deep and extend into the latter. A major issue in CAD system implementation and use is numerical precision. In particular, when shapes are combined to make a single solid, if two faces meet but are positioned apart by an infinitessimal decimal difference, the system may not realize they’re completely joined. It doesn’t typically matter much for this kind of casual modeling, the final printed piece would usually look the same either way. However, it could affect the efficiency of the print. It also quickly becomes an issue if you use a combined shape with such a gap as a hole to make a cut: You’ll be left with an extremely thin slice in the middle of the hole. I find this issue arises often in TinkerCAD, so whenever possible I extend parts into each other a bit to make absolutely clear that they should be a single solid when joined.

Another detail is that the 0.5mm visible pad depth is not an arbitrary value. Much below that and many home printers, including mine, would not have the resolution to be able to reliably produce the outline. Further, even if it could, the detail would be lost with all but a very careful paint job. The pads are also hanging in space, creating an overhang that might need external support to be printed. However, a 0.5mm overhang is easily managed by many or most slicers and printers. There’s little enough material that the overhanging region can be largely supported from the body.

Front pads made by copying the center box and wing, then scaling down and positioning.

Front pads made by copying the center box and wing, then scaling down and positioning.

A minute detail from the original piece, I then made very thin wedges to cut angles out of the corners of both pads. This is quickly done by making one wedge, then using the align and mirror tools to flip it into each corner. Those are then grouped together, switched to a hole, and grouped with the pad.

Wedges used as holes to cut corner angles into the front pads.

Wedges used as holes to cut corner angles into the front pads.

The recess on the wings could be made by scaling down the body similarly as the pads. However, a small tweak can be made to afford a critical difference. The front pad on the wing has a small border all around it. However, we can model the recess to only be on the top and bottom aspects. This is actually how it is on the FFG model: The side face of the wing body is aligned with the side extremity of the front and back ribs.

Modeling the recess this way is a significant boon to 3D printing the piece. By not recessing the side of the wing body, the side of the piece presents a completely flat plane. It’s also big enough to stand stably on the print platform. The piece can thus be rotated 90 degrees and printed on its side.

This is great news, because it entails no supports will be needed. If the piece were printed in its natural position, sitting upright, the wings would create an overhang requiring support. The issue is that the wings grow out to toward the sides at too shallow of an angle. The printer can’t print in thin air. Most start needing a support structure at anything shallower than a 45 degree overhang, and the wing rise is much less than that. Support structures are a pain to cut off or dissolve, assuming your printer even has the latter capability, so they’re best avoided whenever possible. They’re especially difficult to work on small pieces, and in this case the detail of the wing recess would be lost cutting out a same-material scaffold as required on a single-head fused deposition printer (the most common kind) on such a small piece.

Similarly, printing the piece on its front or back would risk losing the detail of the pads and their angle cuts. Although very small, the pads would create slight overhangs that might not be rendered as precisely. Even more likely, physical effects and common settings for the initial print layer(s) against the build plate, such as fatter and thicker material deposition, would probably cause the small recessed border around the pads and the detail of the angle cuts to be flattened, absorbed, or otherwise lost.

In contrast, by putting the piece on its flat side, no detail will be lost and the wing angles grow vertically very comfortably within printer tolerances. Combined with the point above about the front pad overhangs being kept well within tolerances, this means the piece can be printed without supports, and a whole mess of hassle and loss of detail avoided. As a bonus this mimics the original FFG piece as well, though it would otherwise be a tradeoff of printability versus accuracy well worth making anyway.

To preserve that flat side and create a solid print foundation with no supports needed, the wing body can’t be simply scaled down. Its body has to extend to the side extremity. So, rather than scaling, the original 3 piece wing construction was duplicated and ungrouped. The top and bottom wedges were then simply moved vertically toward the center, creating a border recess parallel to the top and bottom lines of the front and back ribs extending all the way to the side edge. The center box was then just downsized to this thinner space.

Similar to the pad, creating the wing recess by lowering the wedges and downsizing the middle box.

Similar to the pad, creating the wing recess by lowering the wedges and downsizing the middle box.

Next I added simple shapes above the top of the center box to give it texture, just like the FFG model. Another bit of minute detail is that the one circle appears slightly smaller relative to the other on the original model, and I kept that here. Like the pads, these details are again sized keeping in mind the capabilities of many home printers. Mine has a 0.5mm nozzle, which is fairly common. That means it more or less produces a 0.5mm wide path. It’ll try, but it can’t realistically and reliably print features below 1mm resolution, essentially a wall made by a path going out and back. So that’s my usual size threshold in making small details like this, the smallest circles and boxes are 1mm. That turns out to be fine though for throwing in vague details like these.

Once that was done I duplicated the front pads and full-size wing rib and shifted the copies to the rear. From there I grouped all the pieces of the wing, duplicated them, shifted them to the other side, and mirrored them.

Adding bits on top for texture.

Adding bits on top for texture.

Copying the pads and full size wing rib from the front to the back.

Copying the pads and full size wing rib from the front to the back.

Duplicating the wing to the other side.

Duplicating the wing to the other side.

The modeling of the piece is now essentially done. All that remains is to create a hole on the bottom for a flight stand peg to hold the pod off the base, and group everything together to form a single solid.

The hole is simply a cylinder center aligned on the central box, switched to a hole, and grouped in. One small note here about TinkerCAD is that it doesn’t have many convenience functions. For example, there’s no functionality to choose a reference for an align action or to lock one piece down for aligning against (the button to lock pieces unfortunately—oddly—also prevents it from being used to align). So, if you select two displaced objects and center them, they’ll both move. Assuming the moving piece is smaller, the pattern is therefore to first align the piece that can be moved against the far extremity of the piece you don’t want to move, and then center them. Since the moving piece is now within the extent of the larger piece, the latter won’t move. In the next picture, this means aligning the cylinder against the right edge of the body, and then centering it. If the moving piece is larger then you need to position it manually or craft a more elaborate temporary construction.

Centering a cylinder to become a hole for the stand peg.

Centering a cylinder to become a hole for the stand peg.

The hole grouped into the bottom of the center box.

The hole grouped into the bottom of the center box.

That’s it! The cargo pod is now done.

The completed GR-75 cargo pod.

The completed GR-75 cargo pod.

Printing

The next step is making a test print of the pod, base, and a simple flight stand peg. My printer is a Lulzbot Mini, which has been phenomenal to work with. Though I’m not an expert on current market offerings, it seems to strike a very good balance between cost, ease of use, reliability, print quality, print size, and time. An important observation is that it seems there are many printers now offering high enough resolution to print miniatures of acceptable quality. Some are even very low cost, down to ~$300. But it seems one of the biggest tradeoffs made in achieving that is that they print much more slowly. That’s probably fine to download and print parts, but in developing new parts from scratch could seriously slow down design iterations.

In any event, one of the cost tradeoffs made in the Mini is that it does not have an SD card reader and/or internal high level controller, it must have a computer connected to drive the print. So I use Lulzbot’s CURA distribution to slice objects into G-Code printer commands, and send that file to a Raspberry Pi running an Octoprint server which actually executes the job and can be monitored remotely.

Slicing the test print.

Slicing the test print.

Another logistical point is that I currently have my printer set up by a window, with fans to draw and push air out of the room and ventilate my workspace. Although there doesn’t seem to have been much study yet of health effects of 3D printers, ultimately fused deposition modeling is melting plastic to form a part. That typically creates toxic fumes, so I’ve been erring toward an abundance of caution as I use it heavily.

My current printer setup.

My current printer setup.

I’ve been exclusively using eSun HIPS filament for miniatures work, to great success. The resulting product has great resolution, is hard and strong, is easily cut and sanded, and takes paint great. A kilogram spool runs $24–40 on Amazon and can be shipped same day Prime. A part like the cargo pod and its base only uses literally a couple grams of material, so you can print an awful lot of miniatures bits per spool.

As referenced above, caveat upgrades, the Lulzbot Mini has a 0.5mm nozzle. After a fair bit of experimentation tuning settings for this kind of project, most of my prints are at 0.18mm layer height; 0.54mm bottom/top thickness; same initial layer thickness; 100% initial layer width; and 20% infill at 0.5mm shell or 10%, 15%, or 20% at 1mm shell. Those settings have struck a good balance between speed, quality, and strength, with very acceptable dimensional accuracy (mostly a challenge in the initial layers).

In the end, the first test print came out excellently! The part came out very cleanly and definitely looks like the cargo pods. The only changes made were to raise the top details from 0.25mm to 0.5mm tall so they’d not be lost in paint, and to increase the flight peg diameter to 3mm. I’d expected to have to do that, but was hoping in vain that the thinner 2mm would magically work. It looks good and would be perfectly fine for at-home play, but I need these pieces to work in a public event setting where they’re not being babied, so I had to accept the 2mm pole would be too easily snapped. I also ran the test with a separate flight stand peg so that in case it was useless, the base would still be useful somewhere. However, the final design integrates the peg and base to make a solid connection with no gluing required.

First print: Success!

First print: Success!

Conclusion

After that was just making a few bases with different numbers of flight pegs, and the project was all wrapped up. I now have a sweet 3D cargo supply container token that matches my beloved GR-75 in 1/270 scale, and have started printing out piles of them.

Again, these designs have been uploaded to Thingiverse as a free download. Drop a line if you make use of them, or have any questions about this walkthrough!

Final design mockup of a cluster of cargo pods; prints are made as one base piece (red) and a bunch of pods (orange) that fit onto the pegs.

Final design mockup of a cluster of cargo pods; prints are made as one base piece (red) and a bunch of pods (orange) that fit onto the pegs.

Print layout for a stand of 4.

Print layout for a stand of 4.

Printing an initial batch of tokens.

Printing an initial batch of tokens.

Cargo pods and their mothership under attack!

Cargo pods and their mothership under attack!

Supply depot.

Supply depot.

Extreme closeup. These parts have not been cut, trimmed, sanded, glued, or otherwise cleaned up in any way yet, this is straight off the printer just popping pods onto pegs.

Extreme closeup. These parts have not been cut, trimmed, sanded, glued, or otherwise cleaned up in any way yet, this is straight off the printer just popping pods onto pegs.

Molokh Gambit: Production Begins!

Very very late last night I started physical production for our Molokh Gambit 1-day X-Wing narrative campaign. Behold, a fleet of CSS-1 Corellian Star Shuttles appears!

CSS-1s and X-Wing escort.

CSS-1s and X-Wing escort.

By this evening I’ll have four of these sweet babies to hand over to Matt for painting. In the process I learned that the Senator’s Shuttle token from the X-Wing Miniatures core set is probably the most out-of-scale “model” in the game. I barely remember it from the prequels at all, but Wookieepedia tells me that the CSS-1 Corellian Star Shuttle is actually 80 meters long and can carry 200 passengers. So more of a cruiser than a shuttle. To be in scale with the other 1/270 standard ship models, the shuttle would have to be almost three times the size seen above. Even against the 1/400 Epic ships it’d have to be double this size. However, I think this version looks great and the size gives it better and simpler playability for 100pt missions as we’ll be using it.

Modeling

I did not create this model, I found a design Losik published on Thingiverse. Err404 also has a great version that embeds the model directly into the shuttle token. I thought hard about using that, but we’re planning to build a bunch of models for the campaign so I went with the flying version so it would fit in better. In the end all I did to the model was recut the base to have a blank top, created a thicker flight pole with matching slots on base and ship, and added a tab & slot to the two ship halves to guide better gluing.

Exploded view.

Exploded view.

I also realized that despite instructions, the front half could be printed laying down rather than on its end. Although this model is thick enough it would almost certainly print fine, and you can put more on the build plate by going vertical, I try to avoid tall structures. Laying it down also let me easily add the gluing tab.

I’ve posted these re-cut models to Thingiverse for free download.

Print layout.

Print layout.

Printing

A somewhat interesting realization was that the front half isn’t actually perfectly flat though. I could see this looking at the shading on the model in various viewers as it moved around, but you can see it very clearly by examining the somewhat degenerate first layer toolpath. The nose of the shuttle lifts up a bit, creating a weird shape for the initial layer. However, it’s literally just the first layer. Given that, I’d assume this is unintentional. If I had to guess I’d say it’s possible (I have no idea) that the model was originally ripped from a game and this is some artifact from the video-oriented modeling and mesh construction. The print is unaffected though and it would have been an effort for me to correct using my software at hand, so I did not bother to fix it.

Somewhat degenerate first layer; the bottom shape should be the hull front half.

Somewhat degenerate first layer; the bottom shape should be the hull front half.

A more interesting general note about printing is that originally I had a taller flight pole. Most of it printed fine. However, toward the end when it was the last component being printed on the layer, the printer hot end cycling around continuously made the top of the piece too hot, keeping the deposited plastic melted a bit and letting the motion swirl and gnarl it. You could fix this with a sacrificial tower pulling the hot end away long enough for the piece to cool. I did the same thing here though by just cutting the flight pole down to the same height as the engine stack. As long as you print them together, the printer will be pulled away from the flight pole long enough, as it goes to print the engines, that the pole will cool and retain its shape.

Printing in progress.

Printing in progress.

The Gambit!

That’s a good start to physical production for the event, with three months still to go. Colin and I had a long discussion on our way to and from NOVA earlier this month about basic squad construction rules and logistics, the outcomes of which are on the event webpage. Matt, Colin, and I have also started passing around a draft dungeon master’s guidebook with detailed proposals for the campaign mechanics that I’m excited about.

More to come; hope you can join us!

14124333_1273061072745446_607619454505042502_o