Designing for Change: A Parametric Note Holder in Fusion 360
14 Sept 2026
I make a lot of notes. To-do lists, things I need to print, things I need to fix, things I will definitely remember later and then immediately forget. The problem is that my handwriting isn't exactly helping. Notes end up everywhere, checkboxes slowly turn into random shapes, and sometimes even I have trouble figuring out what I wrote.
At the same time, I needed a small break from another project. I had been working on an enclosure for Flappy Banana, a little game I was building, and I was trying to make that design parametric as well. It wasn't going particularly well. Small changes, tolerances and dimensions kept causing problems elsewhere in the model. So instead of solving that problem, I did the sensible thing and started another project. This time, something smaller.
Starting with the problem#
The original idea came from the notes I was already using. My dad's company has these rectangular note blocks, and I had a pile of them lying around. They were actually a pretty good starting point. There is a logo in the top-left corner and some unused space next to it, which gave me the idea of using that space for a title. The rest of the note could then be used for the actual list.
But instead of drawing my usual questionable checkboxes by hand, I wanted removable stencil sheets. The holder itself should store the notes, hold a pen without making the design unnecessarily wider, and have several thin stencil sheets stored underneath.
For the rectangular version I made three stencil layouts: a two-column version with circles, a two-column version with squares, and a three-column version with squares for those days when apparently everything needs to become a task. After a few revisions, everything fitted and the stencil sheets worked surprisingly well.
Then my dad left it in his car.
The car test I didn't plan#
A few days after taking the holder to his office, my dad messaged me saying the note block didn't fit anymore. That made no sense. I knew it fitted because I had tested it myself. When I got the holder back, the explanation became pretty obvious: it had warped.
It was summer, the inside of the car had become very hot, and I had printed the holder in PLA. The dimensions I had spent all that time getting right weren't the dimensions anymore. So I printed the same design in PETG and gave it a rather unscientific durability test: back into the hot car. This time it survived.
That was probably the first indication that this little note holder was going to teach me more about designing for 3D printing than I expected.
Parametric, until you change something#
Then my girlfriend wanted one. The rectangular company notes were fine for my dad, but not particularly useful if you need to buy replacements somewhere else. So we went to a shop and bought regular square sticky notes instead.
No problem. The model was parametric. At least, I thought it was.
Changing the dimensions exposed parts of the design that weren't properly connected to the parameters. Some features changed correctly, others didn't. The stencil sheets in particular still depended on dimensions I had effectively designed around the original note block. So I went back into Fusion 360 and started fixing it properly.
For anyone unfamiliar with parametric design, the idea is that instead of designing everything around fixed measurements, you define important dimensions as parameters. The note might have a width and depth of 75 mm. The wall has its own thickness. There is a clearance around the notes, another clearance for the stencils, and other dimensions can be calculated from those values.
If the note changes, I shouldn't have to start moving faces and editing sketches all over the model. I should be able to change noteWidth and let the relationships and constraints in the model do the rest.
The change from 75 mm to around 150 mm isn't just a stress test either. That's roughly the difference between the square notes and the rectangular company note block my dad uses. The holder becomes wider, the stencil lines follow it, and the rest of the design adapts with it.
That's exactly what I wanted from the project: change a number, not the model.
Dimensions aren't tolerances#
This was also where I finally started treating dimensions and tolerances as different things. That sounds obvious now, but it was something I had been doing wrong in other designs.
If a note is 75 mm wide, its width is 75 mm. I shouldn't change that value to 75.5 mm because my printer needs some extra room. The physical object hasn't suddenly become wider. Instead, the design should know that the note is 75 mm and that the printer needs a certain amount of clearance around it. That means having things such as noteWidth, noteClearance and stencilClearance as separate parameters.
This was one of the things that had caused trouble with Flappy Banana. I was sometimes changing the actual dimensions to compensate for a print that didn't fit. It fixes that particular print, but slowly makes the CAD model lie about the object you're designing. This project was small enough that I could finally get that distinction into my head properly.
Doing the modelling myself#
This was also a little different from some of my other projects. For the Durian Lamp, I experimented with using AI to help create parts of the 3D design. That was interesting, and it can get you surprisingly far, but making revisions isn't always pleasant. You explain what you want, wait, get a result, discover something isn't quite right, change the prompt and try again.
Here I wanted to build the model myself in Fusion 360. That turned out to be much more satisfying for this kind of project. Every time I found a problem, I could go into the parameter dialog, change a value and immediately see what happened. More importantly, I understood why it happened.
The computer wasn't generating a new interpretation of my design. I had defined the relationships myself, and that made experimenting with small changes much faster.
From rectangular to square#
Once the model worked with the standard square notes, I could start having a bit more fun with the stencil sheets. For the square holder I kept the selection fairly simple: a regular square checkbox, a circle, a two-column circle layout and a heart. The heart was mostly for my girlfriend.
Naturally, the shape that looks the simplest became one of the annoying ones. Making a heart shape isn't difficult. Making a heart that works as a stencil with an actual pen is.
I made the stencil shapes in Affinity Designer, which has become my replacement for Illustrator. I used Illustrator for years, but I wasn't particularly interested in following Adobe further down the road of subscriptions and increasingly restrictive contracts. Affinity gives me the vector tools I need without much drama, which is exactly what I want from software.
The first heart versions looked perfectly fine as vectors, but tracing them didn't produce a particularly nice heart. A pen has thickness, you hold it at an angle, and it has to physically travel around the inside edge of the stencil. Sharp transitions that look good on screen don't necessarily feel natural when you're drawing through them. So the heart went through several revisions before the shape on paper started looking like the shape on screen.
That became another recurring theme of the project: CAD geometry and usable geometry aren't always the same thing.
Which stencil did I put in there?#
Once my girlfriend started using the holder, she found another problem I hadn't really considered. When the stencil sheets were stored inside the holder, you couldn't tell which one was which.
Fair point.
So I added a small symbol to every tab. A square stencil gets a square, a circle gets a circle, a two-column version gets two symbols and the heart gets a heart. It's a tiny detail, but it's one of my favorite changes because it didn't come from staring at the model. It came from someone actually using the thing.
Because the stencil system was parametric, I could also make the number of storage slots configurable instead of maintaining separate models.
Want three stencil sheets? Set stencilCount to 3. Want five? Set it to 5. The height of the holder and the storage area can be derived from that value rather than maintaining a different model for every version.
The tiny bump that took far too long#
There was still another problem: the stencil sheets could slide out. This was especially noticeable with the smaller square holder. Pick it up the wrong way and the sheets could simply escape from the front.
I tried several ideas before settling on a tiny bump on the holder combined with a corresponding recess in each stencil tab. Push the sheet in and it moves over the bump before settling into place. It's a tiny feature, so naturally it took far too many test prints.
I would change the bump by something like 0.10 mm, print another version and sometimes barely see any difference at all. Eventually I realized I was thinking about the CAD dimensions without thinking enough about how the printer actually creates them. My prints were using a 0.16 mm layer height. Asking for certain vertical changes smaller than that doesn't magically give the printer another physical layer.
So I stopped defining features like the bump purely as arbitrary millimeter values. Instead, the layer height itself became a parameter. If I'm printing at 0.16 mm and I want a feature to be three layers high, its height can simply be derived from:
layerHeight × 3
Now the design is describing something the printer can actually produce.
There's that saying about doing the same thing over and over again while expecting a different result. Looking at this pile, I may have spent parts of this project conducting my own practical research into it.
At least eventually I started changing the right parameters.
Did I actually make it parametric?#
At some point I wanted to know whether I had actually built a parametric design or just spent a lot of time making one note holder unnecessarily complicated. So I started making different versions.
There was the rectangular version for my dad's company notes, a square version for standard sticky notes, versions for myself, a personalized one for my girlfriend, and eventually versions for colleagues. Names and logos could be changed without rebuilding the holder. Different stencil combinations could be exported. Note dimensions could change. Even the number of stencil sheets could change.
The colleague versions became another small experiment in personalization. Instead of putting a huge name across the holder, I went with a smaller recessed name on the back. That worked without causing problems, so I pushed it a little further and added the company logo across the front of the stencil drawers as well.
And... nothing broke.
That was actually one of the more satisfying moments of the project. Earlier, every new idea felt like it might trigger another round of redesigning and test printing. Now I could add a name, add a logo, change a dimension or make another variation without worrying about the whole thing falling apart. I was finally becoming confident in the model itself.
I wasn't redesigning the note holder anymore. I was configuring it.
What I actually learned#
I'm not a 3D modelling expert, and I'm sure there are plenty of things in this Fusion 360 file that somebody with more experience would do differently. Feedback is always more than welcome. But I have started developing my own simple way of deciding whether a parametric design is going in the right direction: how much work does a change cost me?
If changing the note size means editing five sketches, moving some faces and fixing three broken features, I've probably designed something wrong. If I can open the parameter dialog, change one value and everything follows, I'm probably heading in the right direction.
The same thinking changed how I organize the Fusion 360 file itself. The holder and stencil sheets are separate components, so their features and timelines stay where they belong. Things that are likely to change between versions, such as names, text and logos, are deliberately added near the end of the timeline. If I want to make a version for somebody else, I don't want to dig through the entire history of the model just to find their name.
That's probably the biggest thing this project taught me. Parametric modelling isn't just about replacing 75 mm with a variable called noteWidth. It's about deciding which parts of a design are facts, which are tolerances, which depend on the printer, which belong together, and which things you're likely to want to change later.
The finished note holders aren't particularly complicated objects. The pile of failed prints suggests I did my best to change that. But now I can change the note size, clearances, layer height, number of stencils, names and logos without rebuilding everything around them.
And, perhaps more importantly, my checkboxes finally look like checkboxes.
I'm still learning 3D design, so if you spot something I could have done better, I'm always interested in hearing it. For now, my benchmark is pretty simple:
If the next change only requires changing a parameter, I know I'm moving in the right direction.
Frequently Asked Questions
- What is a parametric 3D model?
A parametric model uses named dimensions and relationships instead of relying on fixed measurements throughout the design. In this project, changing parameters such as the note width, clearance or number of stencil sheets automatically updates the related parts of the model.
- Why did you use PETG instead of PLA?
The first PLA version warped after being left in a hot car. I reprinted the holder in PETG and repeated the test, and the PETG version survived the heat without the same deformation.
- Why is layer height a parameter in the model?
Some small features, such as the bumps that keep the stencil sheets in place, need to match what the printer can physically produce. Defining their height as a multiple of the print layer height made those features much more predictable.
