Which material for a part that gets hot
This is the commonest cause of a failed part we see: perfect on delivery, warped three weeks later.
Technical ·
A supplier who says yes to everything will cost you time and money. Here are the limits we state before taking an order, and why none of them is solved by paying for a better material.
This is the fundamental limit, the one everything else follows from. An FDM part is a stack of layers welded to one another: it is strong in the plane of the layers, distinctly less so perpendicular to them.
In practice a part can be twice as strong in one direction as in the other. That is why print orientation is a technical decision rather than a production detail: we orient the part according to the load it will take, which assumes you tell us what that load is.
It is also why we refuse safety parts. A bicycle stem, a lifting component, anything that injures when it breaks: the failure mode of an FDM part is abrupt and hard to predict. No material corrects that.
A printed part can be watertight as it leaves the machine, with enough walls and sufficient infill. The problem is that it rarely stays that way: the micro-interfaces between layers eventually let something through under thermal cycling and pressure.
For a planter cover holding watering runoff, that has no consequence. For a pressurised circuit it is disqualifying, and we do not take those orders.
A material can be certified food-safe as a pellet without the printed part being so. Two reasons: the nozzle leaves metallic traces, and above all the print striations form grooves that washing never fully cleans. Those are bacteria traps.
A cookie cutter used for a few minutes and then washed poses no real problem. A box holding food for days does. So we never claim food contact, even when the material datasheet would allow it.
We hold ±0.3 mm as standard, and ±0.15 mm on dimensions identified in advance. Below that the part has to be machined, which we do not offer.
Material shrinkage explains much of that spread: plastic contracts as it cools, and more so over large dimensions. A 300 mm part does not come out with the same precision as a 30 mm one. That is predictable, so it can be compensated, but not eliminated.
If a tight fit is critical, the right method is a test sample before the run: measure, compensate, then start.
Layer lines are always visible. A 0.1 mm layer makes them discreet, a 0.3 mm layer makes them pronounced, but they do not disappear. Downward-facing surfaces additionally show support marks.
We offer no cosmetic finishing — no sanding, no primer, no paint. What we can do is orient the part so the visible faces come out as clean as possible. Tell us which face will be looked at.
With those limits stated, one field remains where nothing competes: the one-off or the short run, with complex geometry, no tooling cost and delivery in days.
An injection mould costs several thousand euros and takes weeks. Below a few thousand pieces, printing wins. Above that, injection wins, and we will tell you so rather than take the order.
A question about your own part? Describe it to us. Firm quote within 24 working hours, from a single piece.
Request a quoteThis is the commonest cause of a failed part we see: perfect on delivery, warped three weeks later.
Four materials cover ninety per cent of enquiries. Telling them apart takes no technical knowledge: five questions are.
A €500 washing machine scrapped over a €3 programme knob that cannot be found because the model is eight years old.
We check every file before printing, and six problems come round in a loop.