Printed threads, and when to use a metal insert instead
Where printed features stop being adequate.
A printed thread earns its place in a fastener that will be assembled once or twice and left alone, and stops being adequate the moment that same fastener needs to be removed and reinserted repeatedly, because a plastic thread's peaks wear down a little with every cycle in a way a metal thread simply does not.
Why a printed thread wears out fast
A thread cut or printed into plastic works, the first few times, almost as well as a thread cut into metal, guiding a screw in along a helical path and gripping it once seated. What plastic cannot do as well is survive the small amount of shearing and crushing that happens at the tip of every thread crest each time a screw is driven in or backed out, since plastic yields and deforms permanently under a load that steel would simply spring back from unharmed. A handful of insertions is often enough to visibly round over a printed thread's crests, and once that happens the screw stops gripping cleanly and starts spinning in a hole that no longer has a real thread left to bite into.
Layer lines make this worse for a printed thread specifically, since a thread's crest is often only as strong as the bond between the two or three layers that happen to make it up, and a crest sitting right at a weak layer boundary can shear off entirely rather than merely rounding over gradually, stripping in a single overzealous turn of a screwdriver rather than wearing down slowly across several gentler ones. A machined plastic thread avoids this particular failure since it is cut from solid material with no internal seams to shear along, but even a machined thread in plastic still eventually rounds over from repeated crushing in a way a metal thread resists far longer.
The softwood-screw comparison
Driving a wood screw directly into a block of soft pine holds well enough the first time, but backing that same screw out and driving it in again a few more times noticeably enlarges and softens the hole, until eventually the screw simply spins without gripping anything at all, the wood fibres around it crushed past the point where they can hold a thread's shape. A metal wall anchor set into that same hole first, with its own hard internal thread for the screw to engage instead of the wood directly, can be driven in and taken out repeatedly without ever degrading, because the screw is now engaging a material that does not crush or round over the way soft wood, or soft plastic, does under the same repeated load. A printed thread and a raw wood hole fail for exactly the same underlying reason, and a metal insert solves both problems the same way, by giving the screw a harder, more durable material to grip instead of the part's own printed body.
What a heat-set insert actually does
A metal threaded insert, typically a small knurled or barbed brass sleeve, is pressed into an oversized printed hole and set in place using a heated tool that softens the surrounding plastic just enough for the insert's outer ridges to sink in and lock, leaving a genuine, reusable metal thread once the plastic cools and re-solidifies around it. The insert's outer surface is deliberately roughened or barbed specifically to resist being pulled or twisted back out once it is embedded, spreading the load over a much larger bonded area than a screw threaded directly into plastic ever manages on its own, and it is this larger, mechanically locked contact area, not simply the insert being made of metal, that gives it most of its extra pull-out strength.
The one number worth remembering
A screw threaded directly into a printed hole often shows visible thread wear and a noticeably looser grip after only a handful of insertion cycles, while the same fastener engaging a properly set metal insert can be removed and reinserted many dozens of times with essentially no change in how firmly it grips, since the metal thread the screw is actually engaging never deforms the way the surrounding plastic would have on its own. That gap, a handful of cycles against many dozens, is really the entire argument for an insert compressed into a single comparison, since almost nothing else about the two joints differs once the insert is properly set.
Why this matters in practice
Deciding between a printed thread and a metal insert comes down almost entirely to how many times the joint will actually be opened and closed, a panel fastened once during final assembly can reasonably use a printed thread, while any access panel, battery cover or serviceable fastener expected to be removed repeatedly over the life of the device deserves an insert from the outset rather than a printed thread nursed along until it eventually strips. Adding an insert costs a slightly larger printed hole and a few extra seconds with a heated tool during assembly, a small price against having to redesign, reprint or repair a part later once its printed threads have quietly worn past the point of holding anything securely.
Where this stops being true
A printed thread genuinely reserved for a single assembly, never intended to be removed again once fitted, loses none of its usefulness to this argument, since the wear that ruins a printed thread only accumulates across repeated cycles and a fastener installed once and left in place indefinitely never gives that wear the chance to build up. The judgement that matters is an honest one about how the joint will actually be used over its life, not a blanket rule that a metal insert always improves on a plastic thread regardless of how many times the fastener will ever really be turned. A very small or very thin-walled part can also struggle to fit an insert at all, since the insert needs enough surrounding plastic to actually anchor into, and forcing one into a wall too thin to contain it can crack the part outright rather than strengthening it, at which point a self-tapping screw driven carefully into a purpose-sized pilot hole, accepted as a single-use joint from the outset, is often the more honest choice than an insert the part was never really large enough to hold.