Smoothing the Small Stuff: Learning to Anneal 3D Prints

There is a category of problem that shows up once everything else is technically working: The part fits. The mechanism moves. You install it, use it once or twice, and then notice something you cannot quite ignore. It doesn't feel right.

That was the case with the locking rod touchpoint on our Tunnel Tray. It did exactly what it was supposed to do. It just felt a little too rough, a little too unfinished, like it belonged one prototype earlier in the process.

So we went looking for a way to fix something that was not broken.

When Function Isn’t the Whole Story

The touchpoint in question is simple. It is where your hand meets the mechanism. You grab it, move it, and move on with your day. There is nothing complicated about it, which is exactly why it matters. When a part like that feels off, it sticks.

We had printed the component in-house. It was durable. It held up fine under repeated use. The geometry was correct. From a purely functional standpoint, it checked every box. 3D printed surfaces have a certain honesty to them, however. Layer lines, slight texture, a sortof dryness to the finish. In some applications, that is perfectly acceptable. In a touchpoint, it starts to feel like friction where there should be none.

We could have left it alone. It worked. Most people would not complain. It just didn't feel finished.

Looking for a Better Finish

There are a few ways to approach surface quality on printed parts. You can redesign the geometry to hide imperfections. You can sand and polish by hand. You can change materials and print settings until the surface improves. All of those options cost time in different ways. What we were looking for was something repeatable. Something that did not turn each part into a small, manual finishing project. Something closer to a process than a workaround.

That is where acetone annealing came into the picture. The concept is straightforward. Certain plastics react to acetone vapor. When exposed carefully, the outer layer softens just enough to smooth itself out. The surface tension does the work for you, pulling those visible layer lines into a more continuous finish. In theory, you end up with a part that looks molded instead of printed. In practice, it required a bit of experimentation.

The Setup (and a Bit of Guesswork)

Our version of the process was not complicated. A plastic container. Paper towels. A small amount of acetone. The paper towels get saturated and placed inside the container. The printed part sits above them, not directly in contact, while the acetone evaporates and fills the space. The container gets sealed, and then you wait.

At first, nothing seems to change. The part looks exactly the same as when it went in. Then, gradually, the surface begins to shift. Gloss replaces texture. Edges soften slightly. The whole thing starts to look less like it came off a printer and more like it was meant to be handled.Timing is important. Too little exposure and nothing really changes. Too much and you risk losing definition or structural integrity. Finding that middle ground took a few attempts.

We learned quickly that this is less of a switch and more of a dial.

What Changed (and What Didn’t)

Once we found a workable timing window, the difference was immediate. The same part, same geometry, same print settings, but a completely different surface. The texture that used to catch your fingers was gone. In its place was a smooth, slightly glossy finish that felt intentional. More importantly, it felt consistent. That touchpoint is something you use every time you open or close the tray. It is a small interaction, but it repeats. When it feels clean, the whole mechanism feels more refined, even though nothing else has changed.

We did not redesign the part. We did not change materials. We simply gave the surface a different ending. There are limits to it, of course. Annealing will soften sharp edges and can blur fine details if pushed too far. It is not a universal solution for every printed component. Structural parts, tight tolerances, or anything with fine mechanical interfaces need a lighter hand or a different approach entirely. For something like this, a human-facing surface with simple geometry, it fit.

A Process Worth Keeping

What made this stick for us was the repeatability. Once we understood the timing and setup, the process became predictable. Print the part, prep the container, run the cycle, and let it settle. No sanding. iterative hand-finishing, or variation from one piece to the next beyond what we could control. It turned a one-off improvement into something we could repeatedly use. That is usually the threshold for whether something makes it into the shop long-term. If it depends on patience and a steady hand every single time, it tends to fall off. If it can be dialed in and repeated, it stays. This one stayed.

The Broader Lesson

There is a tendency to treat surface finish as cosmetic, especially when working with parts that are buried inside a larger system. This was a useful reminder that the parts people touch carry a different weight. You can have a well-engineered mechanism underneath, but the interface is what people remember. That is the part that communicates quality, whether intentionally or not. In this case, the fix was relatively simple. A container, some paper towels, and a solvent working quitely in the background. Nothing about it felt particularly advanced, but the result shifted the experience of using the tray in a noticeable way. It also added a new tool to the process, which tends to be the more valuable outcome over time.

Where This Goes Next

We are not suddenly annealing everything that comes off the printer, but we are paying closer attention to where it makes sense. Touchpoints, visible components, anything that benefits from a smoother interaction. Those are the candidates. The rest can stay as-is, with all the honesty that comes with a printed surface.Like most things in the shop, it is another dial to adjust when something feels just slightly off.

If you have been experimenting with finishing methods on your own prints, we are curious what has worked and what has not. There is usually more than one way to get a part from functional to finished.