AirGradient Open Source Air Quality Monitors
We design professional, accurate and long-lasting air quality monitors that are open-source and open-hardware so that you have full control on how you want to use the monitor.
Learn MoreThe banner image above is the mold for the smaller bottom part of our new portable monitor, AirGradient Go. Yes, it’s all for that very small blue part in the middle of the exploded view! Every plate, pin, spacer, and screw is pulled apart so you can see how the tool is built. I think this view really shows what exactly makes injection molding both fascinating and unforgiving.
In case the blue part is too tiny in above view, here is the actual plastic part that this mold produces.

In this post, I want to explain the basic structure and terminology of an injection mold for readers who have never worked with one. If you want the project-management view (quotes, DFM, T0/T1 samples, working with a mold maker), I covered that in an earlier article and I will reference it a few times here.
Imagine injection molding as a loop. Plastic pellets are melted, pushed under pressure into a shaped cavity, cooled until solid, and then the tool opens so the part can be ejected. Repeat that cycle thousands - or millions - of times.
That last part - “repeat” - is why molds look so overbuilt. A mold is not just a negative shape. It is a precision machine that must align perfectly, transfer huge forces, manage heat, and release the part reliably on every cycle.

Most molds have two main halves. The A-side is the fixed half mounted to the stationary side of the injection machine. It’s on the left side of the exploded view above. The B-side is the moving half that opens and closes every cycle.
Inside those halves are the two surfaces that form our part of the AirGradient Go. The cavity typically forms the outside (the cosmetic surfaces you touch and see). The core typically forms the inside features (ribs, screw bosses, and all the hidden geometry).
Where A and B meet is the parting line - the seam you sometimes see on plastic parts. Parting line placement affects appearance, flash risk, and how complicated the tool needs to be.
If you look at the AirGradient Go exploded view, you can already read some of this structure even without labels. The larger grey plates at the outer ends are part of the A-side/B-side stack that bolts into the injection machine. The more intricate blue-grey block in the center is where the core/cavity plates and inserts live - the region that actually creates the bottom shell’s shape.
A common beginner assumption is that a mold is just two chunky blocks of steel. In reality, it is a stack of plates, each doing a specific job.
There are plates that clamp the tool into the machine. There are plates where the part geometry is machined. There are support plates whose job is simply to keep everything stiff and flat under pressure. There are spacer blocks that create room for mechanisms to move. There are plates that carry the ejector system.

In the AirGradient Go tool, you can see that “layering” very clearly: big grey plates acting as the structural backbone, the central blue-grey assembly where the part-forming surfaces and inserts sit, and the “cloud” of hardware around it that makes the whole machine repeatable.
In the AirGradient Go exploded view, the most eye-catching components are the many long, thin rods scattered around the assembly. Those are primarily ejector pins.
When the tool opens, the part usually sticks to the B-side (core side). That is normal - the plastic shrinks onto the core as it cools. Ejector pins push the part off the core so it can fall out cleanly.
This is one of the places where “simple part” becomes “complicated mold” very quickly. Every pin leaves a tiny circular witness mark. You cannot eliminate them - you can only choose where they are acceptable. That turns ejection into both engineering and industrial design. You try to hide pin marks on the inside, on ribs, or in areas the user never notices. And you need enough pins to eject without bending the part, creating stress marks, or damaging delicate features. For the AirGradient Go we placed the pins on the inside of the enclosure. So you should not be able to see them.

Ejection also needs to reset perfectly. That is why molds include return systems so the ejectors retract before the next shot. In an exploded view like ours, that whole ejector system shows up as a set of plates plus a surprising number of pins, spacers, and fasteners - all in service of reliably releasing a single plastic part.
This mold weighs more than 100kg, but it still needs to close with precision. That is handled by leader pins and bushings (guiding elements that bring the halves together) and often additional alignment features such as interlocks.
If alignment is off by even a small amount, you get problems like flash (thin plastic that leaks into gaps), uneven wear, and inconsistent cosmetic edges. This is why so many parts in the exploded view are not “part-shaped” at all. They are the mold’s mechanical system - the pieces that make sure the A-side and B-side return to exactly the same position, every time.
In the AirGradient Go mold you can spot this logic in the way the main grey plates and the central blue-grey block are tied together with guide elements and a lot of carefully placed hardware. It looks like overkill until you remember it has to work reliably for a very long time.
Somewhere in the A-side (left side on the exploded view) there is a green sprue bushing, the entry point where the injection machine nozzle pushes molten plastic into the tool. From there, plastic flows through runners and then through a gate, which is the final narrow passage into the part.
Gate placement is one of those decisions that affects almost everything. It influences aesthetics (a small gate vestige), strength, how well the cavity fills, and where you might get weld lines or flow marks. In our “lessons learned” post, we talk about why reviewing the complete mold 3D is so important - gating is one of the things you want to see clearly before steel is cut.

For AirGradient Go’s bottom shell, the goal is straightforward: fill consistently, avoid cosmetic trouble on the outside surfaces, and keep the part stable so it fits the rest of the enclosure precisely.
Cooling channels are often the most important feature that is hardest to appreciate from a simple render.
Cooling determines cycle time and has a huge influence on warpage and sink marks. If one area cools slower than another, plastic shrinks unevenly and the part bends. That can turn a great CAD design into a part that will never sit flat, never seal properly, or require constant “fighting” during assembly.
This is also why we emphasize careful review and iteration in the earlier write-up. Cooling and flow issues are much easier to correct in a 3D model than in hardened steel.
A mold is not “done” when it arrives. The first shots are a validation process.
In our previous article we describe T0 as the first samples used to check fit, warpage, stress marks, sink marks, and whether everything behaves mechanically. Then T1 is typically the next step where you validate the production-ready result, including final cosmetics such as texture.
For AirGradient Go, this matters because the bottom shell is not just “a cover.” It needs to be dimensionally stable, assemble cleanly, and survive real-world use. Those requirements show up in the mold as pin placement, support strategy, alignment design, and careful control of how the plastic cools.
So far the mold manufacturing is on track, and we do expect the first parts early next month. Then hopefully the parts turn out well, that the PCB and the sensors fit well and that we can doing some stress tests of the enclosure before ordering a much larger batch for the first production units.
If you’re curious about AirGradient Go and want to be the first to hear when we open preorders, you can sign up with this form. We’ll send a short update once preorders go live - and we’ll keep it to the essentials.

We design professional, accurate and long-lasting air quality monitors that are open-source and open-hardware so that you have full control on how you want to use the monitor.
Learn MoreCurious about upcoming webinars, company updates, and the latest air quality trends? Sign up for our weekly newsletter and get the inside scoop delivered straight to your inbox.
Join our Newsletter