Train Journey: 3D Printing a Train

When an off-the-shelf train just isn't enough of a challenge, what can you do? Design and print your own, of course!

This is an instructable about how I thought it would be easy to design and build a new train concept, within a few days. Spoiler alert: It's harder than I thought. However it can be done, so read ahead to get tips and advice on trying this yourself.

Firstly, a short background behind this project: Due to coronavirus, I've been working from home and have been unable to go on holiday or do anything fun in the outside world. To keep my mind occupied during this copious free time, I dug through my spare parts box and came up with an idea for a homemade train set.

1) 3D printer + filament

2) Fusion 360 and a computer to run it on

3) any old small motor

4) any battery that'll turn your motor

5) small bearings + screws

Measure your motor, measure your battery, and figure out a track design. My initial idea was to have a perfectly round track, lying down on the table. The train would drive around, sideways-on to the table. I quickly realised that it would fall off if it went too slowly, so I turned it more into a rollercoaster-style train, with "clamp" wheels on the underside.

The motor is possibly from an old VHS player, but I found it in a pile in my garage so it could be from anything. I had an issue with mine - I broke it when trying to push a part onto it. This resulted in a minor redesign with a back-up motor.

The battery is an old big NiMh rechargeable - this was actually useless in the end, so I switched it for a 9v battery. I didn't want to go down the route of LiPo and complex circuitry, I wanted to be able to just plug the motor straight into the battery. Simple and effective.

The design was all drawn up on Fusion 360. It's a powerful bit of software, with the added bonus that it's free for us makers.

I also paid lots of attention to how it would print. This meant flat surfaces where possible, and minimal overhang. Of the 38 parts I printed, only 1 needed supports and only 1 needed print bed adhesion. Taking the time to think about these aspects when you're designing will save a lot of failed prints and post-print tidying.

If you're not sure how well things will work, make the most of the best part of 3D printing - iterative design. Design something, print it, test it, and if it doesn't work then just make the changes and try again. If you were designing something for injection moulding, this might take months. With a 3D printer you can test multiple parts in an afternoon.

As mentioned in the previous step, print while designing. You can pick up errors earlier this way, letting you go back to fix them before it requires a complete redesign.

Make sure you pay attention to all the different 3D printing options. One setting I particularly like is adaptive layer height. If you don't need perfect quality, this is a great setting to reduce print times whilst ensuring high enough quality where it's needed.

Another point to watch for is the strength of the design. Print a narrow item vertically and it will break apart where layers meet. Print it horizontally and it will be much stronger.

Support can be used, but if you can design your part to not need it, you'll save time and money. If you're not sure you need it, use either the Draft Analysis on Fusion 360, or use layer view on your printer slicer.

Also ensure your print will stick to the bed. I use a Prusa i3 Mk3 with a heated print bed, and a spray-on glue called Dimafix, which is only sticky at high temperatures. It works 99% of the time. However, I got my first print spaghetti whilst printing this, due to a part having a very small bed contact area. I fixed this using a brim around the print, just helping to hold it on.

If a train doesn't have a track, it's just a big car.

To ensure it didn't move anywhere, I screwed the track together and hot-glued it to the table. Hopefully when I'm finished, the glue should just lift off the table with minimal effort. I hope it does, because I promised the wife that it will...

Note that hot glue is hot enough to warp the PLA I printed these with. Luckily I noticed this on one of the support legs and just held it in place until it solidified again. After all, a hot glue gun is just a really basic handheld 3D printer.

I also taped the inside of the track to ensure the drive wheel had good traction. PLA slides fairly well on PLA, but rubber is obviously far grippier.

To hold the train to the track at low speeds and when stationary, I needed to design these rollercoaster-style retention wheels.

I used an elastic band to hold them tight, which was enough to hold the train on, but also allowed me to unclip the train from the track without having to disassemble the track itself.

Design note: Ensure the elastic band will survive. After using it a bit, the arms have almost cut through the elastic band. A simple way to fix this is to give them rounded edges.

As I mentioned, the original motor broke (because I pushed too hard on the spindle, whoops) and the original battery was far too heavy. I swapped for a different motor and a smaller battery.

I'd already printed the original battery holder as part of the design, but luckily the new battery fitted anyway. It left lots of spare room. Let's call it cargo space?

I designed in loads of small bearings to ensure minimal friction. One problem I had was getting the drive wheel to stop wobbling. This was achieved with two bearings and a sleeve between, reducing the effect of play in the bearings.

All the other wheels consist purely of bearings.

With any luck, it'll work first time. If it does, buy a lottery ticket.

No project works perfectly immediately.

Firstly my original battery was far too big. I picked it because it's the first one I saw in the garage. If it worked, it would have kept the train running for days on end, but this is not what was required. It was too heavy and the train wouldn't even stay on the rails.

Once the battery had been changed out, I had issues with traction, solved by tightening the retention arms and adding tape to the track.

But once it started working, my doggo was amazed. And so was I. I'd just designed and built a trainset from scratch. Feels far more productive than spending lockdown playing videogames.

Try it yourself, whether it's a train or something else, it'll feel good. The total cost of this is probably about £1/$1/€1 for printer filament and hot glue, with all other parts reusable.

No project is ever finished. Here are some improvements I'd make if I had the time:

1. Aerodynamic adjustments: In free space, a teardrop is the most aerodynamic shape. I made the tail of this pointy, now I just need to improve the front end. Other alterations could include a fairing around the retention wheels, and smoothing the entire model with sandpaper.
2. Speed increases: Aerodynamics don't really apply at the speed this goes, so improving the speed through a better motor or a gearbox is high on the list.
3. Electronics: WiFi control is perfect in every scenario.
4. Stronger track supports: If the speed is increased, the existing track supports will not survive long.

Do you have any other suggestions? Please let me know in the comments.

I have included the Fusion 360 design, feel free to use it and evolve it, but the real joy is to start from scratch. Good luck with your own design and let us know how you get on.