The three variables that decide it
Most plastic joint debates reduce to three questions: how many times will the joint be opened, how thick is the boss wall, and how many parts per year will you build? Answer those and the choice is usually obvious.
- Service cycles: 1–2 cycles favours thread-forming; 5+ favours an insert.
- Wall thickness: below 0.8 mm around the thread, an insert may not fit; above 2 mm, an insert is easy.
- Volume: above roughly 100,000 per year, moulded-in inserts usually beat heat-set on cost.
Thread-forming: fast, cheap, finite
A thread-forming screw cold-flows plastic around the thread, producing a joint that is strong in pull-out and cheap in part cost. It requires no insert, no secondary operation and no additional component in the BOM.
Its limitation is fatigue. Each drive removes material and enlarges the formed thread slightly, so strip torque falls with every cycle. In unfilled resin, 3–5 cycles is realistic; in glass-filled resin, 1–2.
Insert: durable, but not free
An insert moves the thread into metal. Pull-out and torque-out become predictable, the joint can be opened dozens of times, and the plastic only has to retain the insert rather than carry the thread load.
The cost appears in three places: the insert itself, the installation operation (heat-set or moulding), and the larger boss required. A heat-set insert typically needs a boss diameter of about twice the insert diameter, which may not fit in a thin-wall design.
| Criterion | Thread-forming screw | Machine screw + insert |
|---|---|---|
| Part cost | Lowest | Higher (insert + installation) |
| Service cycles | 1 – 5 | 20 – 50+ |
| Boss diameter | 1.8 – 2.2 × screw Ø | 2.0 – 2.5 × insert Ø |
| Assembly steps | One (drive) | Two (install insert, drive screw) |
| Pull-out strength | Moderate | High and predictable |
| Best volume band | Any | Above ~100k/year for moulded-in |
The hybrid that often wins
For products that are assembled once but serviced occasionally, a thread-forming screw into a reinforced boss with a metal sleeve can be the best compromise. It costs less than a full insert, supports 8–10 cycles and fits in a wall thinner than a standard insert requires.
We supply both the screws and the inserts, so we have no particular interest in which one you choose. Send the housing section and the service requirement and we will propose the joint that meets both at the lowest total cost.