Trials and Tribulations of 3D Designing

Last week, I outlined my fruitless search for a modular, hex-based terrain system for tabletop RPGs. As a result, I started working on my own solution. I’m not a designer of any description by trade, and CAD programs give me a huge headache, so this is something that I’ve no doubt made more complicated for myself than it needed to be.

Today, I’ll run you through some of the processes I used and give you the opportunity to yell at me. “By all the Gods, why did he do that?” is undoubtedly going to be a common phrase coming from those of you reading this who have experience in this type of thing.

But stick with me, because I’m still on track to release the full basic set of tiles next week. Just a few tweaks to go…

Remember this from last week? These are the seven basic tile shapes that I’m working on, with additional heights to come as well.

In the beginning, there was Microsoft 3D Builder

I say that I’m not a designer, which is true, but it doesn’t mean I haven’t tinkered about in various programs in the past. Initially, I was only interested in splitting models into various smaller pieces to fit on my print bed (The Monoprice Select Mini is not exactly a giant machine). I quickly found that Microsoft 3D Builder was perfect for this job. It’s free, and should already be installed on your PC if you’re running Windows 10.

Opening the program takes mere seconds, and you’re presented with a couple of basic options to open an existing file, or create a new scene. Really simple and straightfoward. If you open a model, highlight it, and click Edit -> Split, it will take you no time at all to play around with the plane and break your model into multiple pieces (very handy for printing more complicated models that require excessive supports).

A model in the process of being split in Microsoft 3D Builder

And it was good…

I’ve created some basic designs in Microsoft 3D Builder for Horizons – not that I’ve revealed them yet. I must get back to that soon – so I figured this would be a good place to start for my terrain. Building creations from scratch is pretty simple in such a basic program, but of course your options are a bit limited when compared to the industry heavy hitters like Fusion 360. My initial design included a series of hooks and slots to tie the various pieces together.

v1 had more than one catastrophic design issue…

It printed just fine, but when it came to test fitting, it was… a dismal failure. I honestly don’t even know why I thought it would work. There were a couple of major issues:

  1. My hooks weren’t aligned in a way that would allow them to match the slots in various orientations. D’oh!
  2. When I did manage to align a few pieces, they were far too difficult to latch together, as the hook on one piece was getting in the way of the rotation necessary to seat the hook of the second piece.

Version 2 included some on-the-spot brute-force changes (that is – I snapped off a heap of hooks to make a symmetrical pattern and turned the ex-hook locations into slots with a pair of trusty pliers). This worked a little better, as it was easier to line up the pieces, but it was still not going to work on a larger scale with multiple pieces.

No photographic or screenshot evidence exists of Version 2, as it very quickly became Version 3. I snapped off the rest of the hooks and designed a separate clip piece in a simple U-shape. This could possibly work… My crappy, hand-built (hand-broken?) Version 3 quickly got mocked up in 3D Builder and became Version 4.

v3 – An ugly attempt, but the first one that showed some degree of promise. If you look closely you can see where I snapped off the previous hook system, and gouged out extra depth for the slots.


v4 design – no integrated clips, hooks or slots at all. So clean!

Version 4 looked so clean and simple. Time for a test print and see how it works.

v4 – bottom view on the left, top view on the right. We might finally be on to something. If only we could make those clips flush with the base…

Version 4 seems to have some potential. The clips hold it all together really well from the bottom, and when viewed from the top, they seem to fit nicely. Trouble is, of course, that the clips aren’t sitting flush with the base which makes the whole assembly a little uneven. That shouldn’t be too hard to address – let’s go back to our notches for Version 5.

v5 – it works!

And here, at Version 5, we have some thing that actually works. And better yet, they stack really nicely, too, locking in to each other vertically as well as horizontally (I didn’t write about the changes I made at each revision in this regard for the sake of brevity).

With the base hex sorted, it was time to work on some of the different shapes I needed. And this is where I found a critical limitation of 3D Builder – edge alignment.

… but Fusion 360 is better.

When I tried to make two (or three, or more) individual hexes line up, I found that 3D Builder had some issues. Namely, while the edges would “snap” together when they were brought near each other it seemed that there was some lack of precision. After much trial and error, I gave up on 3D Builder and delved deeper into Fusion 360.

Don’t be put off by the price ($310/year) – there’s a free hobbyist subscription, with all the power of the full version. What you are allowed to be put off by is the learning curve. Fusion 360 is horribly powerful and detailed, which fairly logically makes it a bit trickier to pick up than 3D Builder.

Look at all those options… oh, my.

Eventually though, I managed to pick up the 2% of the program’s capabilities that I needed, and rebuilt my basic model from scratch. From there, I just needed to duplicate the single hex, align the edges of each piece, and remove some of the now-excess central “legs” (to save time and material when printing).

The underside of models A, B and C (from right to left. I do enjoy being a bit backwards). You can see the edges have been removed where the individual hexes join.

Version 6 turned out to do everything I needed and wanted. Now, I just needed to adjust some of the heights to allow stacking, and consistent levels, and we’re good to go!

Next week, with just the tiniest bit of luck, the weekly post will contain a link to download the basic set of terrain in varying heights. Make sure you come back then!


    1. Post

      From corner-to-corner at the widest point, a single tile is 38mm.
      The widest point of the surface is 32.9mm
      The ledge is 2.25mm.

Leave a Reply

Your email address will not be published. Required fields are marked *