Iterate and improve

You have a real lemon squeezer in your hands now. You designed it, sliced it, printed it, washed it, and squeezed a lemon over it. Some of that went well. Some of it probably did not. This chapter is about what you do next, and it is the most important lesson in the whole book.

Here is the mindset shift: your first squeezer is not a grade. It is data. We will call it v1, short for version one, and the only thing v1 owes you is information. Did the juice flow? Did seeds slip through? Did it fit the lunch box? Every answer, good or bad, points at a specific thing to change. Back in the introduction we named the loop that makers run: design, print, test, learn, repeat. You have now done the first three steps. This chapter is the "learn" and "repeat" part, done on purpose.

The reason this matters is that nobody, anywhere, prints a perfect part the first time. Professional product designers go through dozens of versions. The difference between someone who makes useful things and someone who gives up is not talent. It is that the maker treats v1 as the start of a conversation, not the end of one.

Read your own test results

Open up the success checks you ran in Chapter 15. You wrote down, for each one, whether your squeezer passed, half-passed, or failed. Do not throw any of that away. Each weak spot is a clue, and each clue has a likely cause, and each cause has a design change you can try.

That chain (problem, then likely cause, then change) is the heart of iterating. It keeps you from flailing. Instead of "it is bad, let me change everything," you say "seeds got through, the slots are too wide, I will narrow them," and you change one thing.

Here is the centerpiece of this chapter: a table that maps the problems you are most likely to have hit straight to a fix you can try. The "design change" column tells you what to do; many of these are a single number in your model, which we will get to in a moment.

Notice that the table does not say "buy a better printer" or "the design is hopeless." Every fix is small and specific. That is on purpose. You are looking for the smallest change that addresses the actual complaint.

Don't be confused. There are two very different kinds of "fix," and mixing them up will waste your filament. Iterating the design means changing the shape of the object in your CAD model: a juicing problem or a fit problem (cone too short, slots too wide, bowl too small) gets fixed in OpenSCAD or Tinkercad. Re-tuning the printer means changing how the machine lays down plastic: a print-quality problem (stringing, warped corners, a rough bottom layer) gets fixed in the slicer or on the machine, and we covered those in Chapter 7. If your reamer is too short to juice a lemon, no slicer setting on earth will save you. Reach for CAD. If your reamer is the right shape but covered in plastic hairs, no CAD edit will help. Reach for the slicer.

The payoff of a parametric design

Here is where the work you did earlier pays off. Back in Chapter 10 and Chapter 13, you built the squeezer as a parametric model. That word means the shape is driven by a handful of named numbers at the top of the file, and the rest of the geometry is calculated from them. Change a number, and the whole part redraws itself to match.

This turns most of the fixes in that table into a one-line edit. Look at how directly the variables map to the problems:

• Low juice yield? The cone is built from your fruit size and ridge settings. Bump rib_count up for more ridges, or adjust the cone height so it reaches deeper. More, sharper ridges tear more pulp.
• Seeds getting through? The strainer is built from strainer_gap, the width of each slot. Lower that number (say from 2.0 mm toward 1.4 mm) and reprint. The slots get narrower everywhere at once.
• Walls too thin or too flimsy? wall sets the thickness of the bowl wall. Raise it from, for example, 2.0 mm to 2.8 mm and the whole bowl gets sturdier, no other edits needed.
• Wrong fruit entirely? This is the one that feels like magic. The whole model hangs off lemon_d, the lemon's diameter. Change that one number to a lime's diameter, hit reprint, and you have a lime squeezer. Change it to an orange's, and the bowl, cone, and rim all scale up together.

That last point is worth sitting with. You did not design "a lemon squeezer." You designed a family of squeezers, and lemon_d picks which member of the family you print. That is the real reward for the extra effort of writing the design as code instead of dragging shapes around by hand.

If you built your version in Tinkercad instead, the same fixes are still available, you just make them by hand rather than by typing a number. To narrow the strainer slots, select the slot-cutting shapes and make them thinner, then re-space them. To deepen the cone, stretch its height. To resize for a lime, scale the whole group down. It is more clicking and more eyeballing than editing one variable, but the logic is identical: find the feature that is wrong, change it, leave everything else alone.

Variants worth trying as you grow

Once your basic squeezer works, you do not have to stop. The same project opens up into a small workshop's worth of follow-on builds, each one a real step up in skill:

• A hinged press or lever version. Instead of pressing the lemon down with your hand, you put it in a hinged cup and pull a handle. The lever multiplies your force, so juicing is easier and a weak grip stops being a problem. This adds a moving joint, which is a genuinely new challenge.
• A snap-on lid. A lid turns your squeezer into something you can toss in a bag with juice still in it. Designing the snap fit (a lip that clicks over a groove) teaches you about tolerance, the small gap you leave so two printed parts actually fit together.
• A two-piece nesting set. Make the strainer a separate piece that lifts out and nests inside the bowl for storage. This shrinks the packed size and makes cleaning easier, since you can scrub each piece on its own.
• Resized for other citrus. Limes, oranges, even a grapefruit if your printer bed is big enough. As covered above, this is mostly just changing the lemon diameter and reprinting. A whole produce drawer's worth of squeezers from one design file.

You do not need to build all of these. Pick the one that fixes your biggest annoyance with v1, or the one that just sounds fun. Both are good reasons.

Habits that make iterating work

Three small habits separate productive iterating from spinning your wheels.

Keep a build log. This is just a text file or a notebook page where you write, for each version, what you changed and what happened. "v2: lowered strainer_gap to 1.5 mm. Seeds stopped getting through, but pulp now clogs the slots." That one line saves you from re-discovering the same thing in three weeks. Your future self has a terrible memory; the log is how you talk to them.

Version your files. Save your model and your exported STL with the version in the name: squeezer-v1.scad, squeezer-v2.scad, and so on. Never overwrite a version that worked. If v3 turns out worse than v2, you want to be able to go straight back, not reconstruct it from memory. Storage is free; a working design you deleted is gone.

Change one thing at a time. This is the same discipline from Chapter 7, and it is just as important here. If you narrow the slots and sharpen the cone and raise the walls all in one go, and the result is better, you have no idea which change did it, or whether one of them quietly made something worse. Change one variable, print, test, write it down. Slower per step, faster overall, because you actually learn.

Don't be confused. "Change one thing at a time" is about understanding cause, not about being stingy. You are allowed to make many versions. You are just making them one change apart, so each version teaches you exactly one thing. Ten focused versions beat one version where you changed everything and learned nothing.

Knowing when to stop

There is always one more tweak. The spout could be a hair sharper. The grip could be a touch grippier. If you wait for perfect, you will never own a finished squeezer, you will own an endless project. So how do you know when you are done?

You go back to your project brief from Chapter 11. The brief is the list of success checks you wrote before you had any feelings about the part. It is your honest, ahead-of-time definition of "good enough." When your squeezer passes those checks, juices a lemon, holds back the seeds, pours without a mess, fits the box, you are done. Not done forever, but done with this project.

The goal was never a perfect squeezer. The goal was a squeezer you actually use. "Good enough and in my kitchen drawer" beats "perfect and still on the screen" every single time. Let the brief, not your perfectionism, tell you when to put the tools down. You can always start a v4 later, when a real annoyance gives you a real reason to.

And that is the whole maker loop, run once with your own hands. You designed, you printed, you tested, you read the results, and you know exactly what you would change. That skill, far more than this one lemon squeezer, is what you came here for. Point it at anything.

Takeaways

• Your v1 is data, not a verdict. Every flaw points at a specific design change.
• Use the problem-to-cause-to-change table: name the problem, guess the cause, change one feature, test again.
• A parametric design turns most fixes into a one-number edit: rib_count for grip, strainer_gap for finer straining, wall for sturdiness, lemon_d for a whole different fruit. Tinkercad users make the same edits by hand.
• Fix shape problems in CAD; fix print-quality problems in the slicer. Do not confuse the two.
• Keep a build log, version your files, and change one thing at a time so you know what helped.
• Stop when the part passes your Chapter 11 brief. Good enough and used beats perfect and unfinished.

👉 You have run the full loop and you know how to keep improving. Next, a plain-language Glossary so every term in this book has a one-line definition you can come back to.