All articles
Buyer GuideTroubleshootingFunctional PartsFDMMaterialsEquipment

Why 3D Printed Enclosures Cook Your Electronics

3D Prototyping Hub·
Why 3D Printed Enclosures Cook Your Electronics

The board ran for a week on the bench without a complaint. It went into the printed case on Friday, and by Sunday the Pi is throttling, the regulator is too hot to touch, and there is a visible dip in the lid above it. Nothing about the circuit changed. 3D printed enclosures fail electronics because a printed box is two things at once that nobody designs for: a thermos, and a sealed one.

This post contains affiliate links. If you purchase through these links, 3D Prototyping Hub may earn a small commission at no extra cost to you.

This post also contains Amazon affiliate links. As an Amazon Associate I earn from qualifying purchases.

Find Which Failure You Have First

Two different things go wrong and they need opposite fixes. Read the symptom before you reprint anything.

What you see What it means Section
Lid bowed, standoffs sunk, case distorted The plastic is above its service temperature 2
Case fine, board throttling or resetting Air temperature inside is fine for the box, too high for the silicon 1, 3
One component scorching, everything else cool No conduction path off that part 4
Fine in winter, failing in summer Ambient stacked on top of your internal rise 2
Failed outdoors after a season, case chalky UV, not heat 2
Screw bosses stripped on the third opening Warm plastic creeping around printed threads 5

1. The Box Is a Thermos, Not a Heatsink

Still air conducts heat at about 0.026 W/m·K. The common printing plastics sit around 0.2 — eight times better than air, and about a thousand times worse than the aluminium a commercial housing uses when it has to dissipate anything. So a printed enclosure does not move heat out; it holds it in and then slowly leaks it through the walls.

The practical consequence is that the component's rating is not the number to design to. A chip rated to 85°C sitting in a box whose internal air is 25°C above the room has lost 25°C of its margin before it does any work. A sealed enclosure of roughly a litre with 3-5W inside typically settles 15-25°C above ambient, and the air touching the hottest part is hotter than that average.

Measure it rather than estimating it. Run the thing under normal load for an hour, then put an infrared thermometer on the outside of the lid, on the wall next to the hottest part, and — with the lid off, which will read low but still rank them — on each component. You are looking for one of two patterns: everything uniformly warm, which is a ventilation problem, or one component far above the rest, which is a conduction problem. Most single-board computers will also report their own die temperature, and that number is worth more than any surface reading.

Does It Actually Need Drying?

Tell moisture from the faults that imitate it. Then dry it without wrecking it.

The file downloads on this page as soon as you submit. No waiting on an email.

2. The Material Sets the Ceiling, and Ambient Eats the Margin

The case itself fails when the plastic passes its softening range under load. That is a different and earlier failure than the electronics cooking, and it is the one you can see.

Filament Holds shape to roughly Where it belongs
PLA 50-55°C Nowhere with heat in it — it sags in a warm room under its own fasteners
PETG 70-75°C Indoor low-power housings; the sensible default
ABS 90-95°C Power supplies, drivers, lofts, vehicle interiors
ASA 90-95°C Everything above, plus sunlight
Polycarbonate 110-115°C The desktop ceiling, for genuinely hot builds

Three things about that table are worth more than the numbers in it.

Add the ambient to your internal rise, not to zero. A loft reaches 45-50°C in summer, a greenhouse more, and a parked car's interior routinely passes 60°C — the mechanism is laid out in why 3D prints deform in a hot car. A box that runs 20°C above ambient and lives in a loft is asking PETG to work at 70°C, which is exactly where it stops working.

Heat plus sustained load is creep, and creep is slower and quieter than melting. The lid stays bolted down, the standoffs stay loaded, and over weeks the plastic flows. A case that is 10°C below its stated limit will still slowly bow if it is loaded the whole time, which is why a bowed lid is usually the first symptom rather than a crack.

Outdoors, UV decides before heat does. ASA is the answer there, for the reasons set out in why 3D printed parts fail outdoors — and if you are reaching for polycarbonate, the practical printing trade-offs are in our guide to the best polycarbonate filament.

3. Vents Only Work If Convection Has Somewhere To Go

Most printed enclosures have vents that do nothing, because they were drawn as a pattern rather than as a path. Hot air rises and has to be replaced from below. That gives you three rules:

  • An inlet low and an outlet high, on opposite vertical faces. A grille on the base with a sealed lid is a decoration. So is a row of slots along the top with no way for air to get in.
  • Put the outlet directly above the hottest part. The plume goes straight up; meet it.
  • Open area matters more than hole count. Aim for a few square centimetres of genuine open area rather than forty 1mm holes. Small printed holes come out undersized, so a fine grille delivers measurably less open area than the CAD claims, and the walls between the holes block more again.

Angle the slots at roughly 45° and they print on a vertical wall without supports, shed dust downwards, and stop a dropped screwdriver going in. If the box will sit on a desk, give it feet — a case with its inlet on the base and its base on a table has no inlet at all.

4. Spread the Heat Before You Move the Air

When one component is far hotter than everything else, no vent helps, because the heat cannot get from the part into the air fast enough to be carried away. That is a conduction problem and it is solved with metal.

An aluminium heatsink on a thermal pad multiplies the surface area that part has to work with, and it is the highest-value few pounds in this entire article. Two refinements worth knowing:

  • A thermal pad is a compromise, not a bonus. It fills the air gap, and air is the enemy, but the pad itself conducts far worse than metal-to-metal contact. Use the thinnest one that actually touches both surfaces.
  • Give the heatsink air. A heatsink with 3mm of clearance to a plastic wall is doing a fraction of its job. Put it under the outlet vent.

If passive still is not enough, then a fan. A slow Noctua 40mm fan moving a modest volume continuously beats a cheap one screaming, and you will actually leave it switched on. Blow in rather than out so the box runs at slight positive pressure and dust goes where you choose, and accept the trade: a fan costs you noise, ingress protection and a mechanical part that will fail before the board does.

5. Mount the Board So the Case Is Not Carrying the Heat

Printed standoffs are the most common quiet failure. They conduct nothing, they sit directly under the hottest components, and under a bolted-down board at 50°C they creep until the board is no longer flat and the lid no longer seats.

Use metal standoffs with M3 screws instead. They do not soften, and the fasteners move a small but real amount of heat out of the board into the case wall where there is more area to lose it from. Put a heat-set brass insert wherever a screw enters plastic that gets warm and will be undone again — a warm boss is where printed threads fail first and fastest, and the full comparison of inserts, tapped holes and captive nuts is in why 3D printed threads strip.

The same logic extends to the hottest part itself: if it can be bolted through the case wall to an external metal plate, the plastic stops being in the thermal path at all and becomes just a cover.

What the Printer Has To Be Able To Do

PETG prints on anything. ABS, ASA and polycarbonate do not — they need a nozzle well past 250°C, a bed at 90-110°C and a chamber that holds its own heat, or the case delaminates up the corners as it cools and you have a weak box that also looks bad. No slicer profile substitutes for a chamber.

If warm-running housings are the work you keep doing, the machine is the constraint rather than the filament. Flashforge's enclosed CoreXY machines pair a high-temperature hotend with a chamber built around it, and ELEGOO's FDM range and Anycubic's FDM machines and spares both cover the enclosed, all-metal end of the market as well as the hotend parts if you are upgrading what you already own.

When Not To Print the Enclosure At All

There is a line, and on this topic it is about ratings rather than skill. Do not print a housing for mains voltage: that needs a declared UL 94 V-0 flame rating at a stated wall thickness with a certificate behind it, and ordinary filament has none of those — a printed case around a mains supply is also exactly the geometry that turns a small fault into a spreading one, which is the mechanism behind why 3D printers catch fire. Do not print a box that has to carry a real IP rating and real heat at the same time, because passively you can have one or the other. Do not print a case that is supposed to act as the heatsink; that part wants to be metal.

Those are sourcing questions rather than design questions, and a shop can answer them with a flame-retardant grade, an ESD-safe material, or a machined aluminium housing with a printed fascia. Our guide to 3D printing services for electronics enclosures covers what to send and what to ask for, and quotes by process and location come from the provider directory.

For everything below that line the method is short: measure the box under load, pick the material for ambient plus your internal rise rather than for the component's rating, give convection a low inlet and a high outlet on vertical faces, put metal on the hot part before you put a fan in the wall, and stand the board on metal rather than on plastic. The enclosure that works is the one that was designed as part of the thermal path instead of a lid over it.

Hero photograph by Jakub Żerdzicki on Unsplash.

Does It Actually Need Drying?

Tell moisture from the faults that imitate it. Then dry it without wrecking it.

The file downloads on this page as soon as you submit. No waiting on an email.

Recommended Resources

Disclosure: Some links below may be affiliate links. We only recommend services we have personally evaluated or that are used by providers in our directory. Clicking earns us a small commission at no cost to you.

Infrared thermometer
The first purchase, before any redesign. You cannot fix a thermal problem you have not located, and a surface reading on the lid and on each hot component tells you in thirty seconds whether you have a hot box or one hot part.
PETG filament, 1.75mm
The sensible default for an indoor housing that runs warm. Holds shape to roughly 70-75°C, prints on any machine without a heated chamber, and does not go brittle the way PLA does in a warm corner.
ABS filament, 1.75mm
A real step up in service temperature — usable to around 90°C — for cases around power supplies, motor drivers and anything that lives in a loft or a vehicle. Needs an enclosed printer to come out straight.
ASA filament, 1.75mm
ABS temperature performance plus genuine UV stability. This is the one for an outdoor box, a solar controller housing or anything mounted where sunlight lands on it.
Polycarbonate filament, 1.75mm
The highest service temperature available on a desktop machine, usefully past 110°C. Demands a hot nozzle, an enclosed chamber and dry filament, so check the printer before buying the spool.
Aluminium heatsink kit with thermal pads
Plastic conducts almost nothing, so the heat has to be spread by metal before the air can take it away. A few pounds of finned aluminium on the regulator or the SoC does more than any amount of vent redesign.
Noctua 40mm fan
When convection is not enough, a slow quality 40mm fan moving air through the box beats a loud cheap one running flat out, and it is quiet enough to leave running in a room where people are working.
Heat-set threaded inserts, M3-M5
Brass inserts give a lid that is opened repeatedly a metal thread, and a warm enclosure is exactly where printed threads creep and give up first.
Metric screw and standoff hardware, M3-M5
Metal standoffs and screws under a hot board carry a little heat out through the fasteners and do not soften, which printed standoffs under sustained load eventually do.

Frequently Asked Questions

Ready to find a 3D printing service provider?

Browse 2,000+ verified providers across the United States and submit your quote request free.

Browse Providers