Threaded Inserts for Plastics, Wood and Soft Metals: Which Family the Evidence Supports, and How Deep to Engage

First, decide whether you need an insert at all

A tapped hole works when the parent material can carry the thread itself: strong, sufficiently thick metal with enough engagement for the load ³³. Where that condition fails — plastics, softer metals, thin sections — the thread wears, strips or pulls out, and an insert gives you a separate threaded interface that does not depend on the parent material's strength ³³. The same property makes inserts the repair route for a stripped connection, and the practical choice for fastening points that come apart repeatedly ³³. That benefit is not free. An insert adds a component and an installation step to every joint, so the case for one rests on durability, serviceability or repair rather than on cost ³³. If your joint is a strong metal part that is assembled once, direct tapping is the simpler route ³³. Before you pick a family, write down the inputs that actually drive the choice: parent material, fastener thread size and pitch, load type (pull-out, torque, shear, vibration), wall thickness around the insert, blind or through hole, installation method, service conditions and production volume ³⁵. These eight factors are what the family decision is made against, and skipping any of them is how a correctly chosen insert still fails in service ³⁵.

Which insert families the sources name for your substrate

The families below are the ones the sources actually name per substrate. Note the asymmetry: plastics have a broad named set, soft metals a narrow one, and wood has no insert family named at all in this evidence — so a wood joint cannot be specified from these sources ^{19, 9, 22}. Where a substrate is absent from the table, the sources do not cover it.

Molded-in or post-mold: rule out families your process cannot install

The installation step is the hard constraint. Molded-in inserts are placed during molding; heat-set, ultrasonic, press-fit and self-tapping are all post-mold processes ^{19, 22}. If the part is not injection molded, molded-in is off the table regardless of how well it suits the load. If the part is molded but you need to add or replace a fastening point after the fact, only the post-mold families remain available ¹⁹. Volume does not separate the families as cleanly as you might expect. Molded-in, heat-set and ultrasonic inserts are all listed as options for high-volume molded parts ¹⁹, and heat-set is also described as common in thermoplastic and 3D-printed parts ²². So a high-volume molded part does not by itself force molded-in — you are choosing between three named options, and the sources do not rank them ¹⁹. What the sources do not give you is the equipment. They name the process families and agree that installation method is a selection factor ^{22, 33, 35}, but they do not specify a soldering iron or thermal press for heat-set, an ultrasonic welder and horn, or an arbor press for press-fit, and they give no process parameters ^{22, 33, 35}. Confirm with your supplier or contract manufacturer that the required equipment exists in the cell that will run the part before you commit the family to the drawing.

Set the engagement depth from the substrate's softness

Engagement depth is where the substrate's softness converts directly into a drawing dimension. The rule of thumb is a multiplier times the nominal diameter, counting full threads in contact, with tap chamfer and drill point allowances added on top for blind holes ¹⁸. Softer material needs proportionally more thread in the hole, because the thread flanks in the parent material carry the load and a soft material distributes that load over fewer effective threads ¹⁸. The multipliers run from 3.0 x d for unreinforced thermoplastics down to 1.0 x d for medium-strength steel ¹⁸. The mechanism behind the spread is hardness: aluminium sits at roughly 60-80 HV against 250-370 HV for structural steel grades, so a torque spec written for steel-into-steel will strip aluminium threads ⁶. For a plastic boss, that means the depth you need is three times the nominal diameter of full thread — an M4 insert in unreinforced thermoplastic needs 12 mm of engaged thread, not the 4 mm a steel part would need ¹⁸. Check that the boss and wall thickness can actually accommodate that depth before you release the drawing; the sources flag thin walls and undersized bosses as a cause of cracking, deformation and pull-out ³⁵. One caution on aluminium: the two sources do not agree on the multiplier, so do not treat either figure as settled. See the comparison below.

Pre-RFQ checks on boss, hole and installation control

These are the checks that decide whether the family you picked can physically be installed in the part you have. The sources agree on what matters but publish no numbers for any of it, so each item below is something you verify with the insert supplier or your molder rather than read off a table ^{22, 35}.

• Confirm the boss diameter and wall thickness can carry the insert without cracking or deforming — the sources call out thin unsupported walls as a failure mode but give no minimum ^{22, 35}.
• Confirm the hole size matches the insert manufacturer's specified dimension — an oversized hole reduces retention, an undersized hole raises installation force and can damage the insert or part ³⁵.
• Confirm whether the hole is blind or through, and that there is installation space for the tooling ³⁵.
• For heat-set, confirm the heat and insertion force are controlled to the insert and parent material — too much heat deforms the thermoplastic, too little prevents material flow around the insert ^{35, 22}.
• Confirm the insert will be held straight during installation — an angled insert leaves the internal thread misaligned and affects mating parts ^{35, 22}.
• Confirm the installation torque is controlled, particularly for self-tapping inserts, where excessive torque damages the insert or surrounding material ¹⁹.

What performance evidence to demand instead of a catalogue rating

Do not accept a generic pull-out or torque-out figure for the insert family. The sources are consistent that insert and joint performance has to be validated on the actual material and the actual joint, not taken from a data sheet or a similar part ^{22, 27}. For plastic applications that means testing torque and pull-out strength with the real resin, at the real boss geometry, after the real installation process ²². The condition that makes borrowed data most dangerous is the joint environment. Where the joint sees vibration, thermal cycling, or mixed materials such as steel into aluminium, torque data must be validated on the actual joint rather than borrowed from a similar part ²⁷. A coating or insert that tests well on a loose sample can behave differently once it is paired with the mating component, washer or seating material ²⁷. Two further cautions the sources raise. An insert does not make the joint immune to over-tightening — fastener torque still has to be specified against the insert, parent material, fastener and component geometry ¹⁹. And in vibration-prone assemblies the insert alone may not prevent the fastener loosening; preload or a locking feature may still be needed ¹⁹. Ask your supplier for validation on your material and joint, and ask what test they would run to demonstrate it.

Temperature, corrosion and standards: what to ask because the sources do not state it for inserts

The environmental and conformance figures in this evidence are stated for fasteners and coatings, not for threaded inserts in plastics, wood or soft metals, so none of them transfers directly to your insert selection ^{13, 10, 32}. The corrosion grades and coating behaviour described — stainless grade selection by environment, zinc flake under thermal exposure, zinc nickel where chemical resistance and corrosion protection are both needed — are properties of the fastener or coating system, and the exact outcome depends on passivation and topcoat chemistry ^{10, 32}. On standards, the picture is the same. The standards cited in this evidence cover coatings, hydrogen-embrittlement testing and general quality systems, and no source cites a threaded-insert dimensional standard, thread class or insert-specific inspection criterion ^{7, 8, 16, 24, 36, 37}. If your drawing needs a thread class or an inspection criterion for the installed insert, that has to come from the insert supplier's specification, not from a fastener standard. So put the questions to the supplier directly: what is the maximum continuous service temperature for this insert in this resin, what chemicals will it see in service, and what dimensional or thread-class standard does the insert itself meet. The sources do not answer any of these for inserts, and a figure borrowed from a fastener or coating page would be the wrong number on your drawing.

The aluminium engagement figure the sources do not reconcile

Two sources give different minimum engagement values for aluminium, and neither is presented as subordinate to the other. Treat the figure as a rule set to be confirmed against your specific alloy and joint rather than a single authoritative number, and resolve it before you commit a depth to the drawing ^{18, 6}.

FAQ

Related pages

• Nuts product page
• Rivets product page
• Screw Family Selection by Base Material: Wood, Machine and Self-Tapping Threads, Engagement and Coating
• Thread Pitch and Engagement Length for Steel, Aluminium and Plastic Joints: What the Evidence Settles
• Steel Screws in Aluminium: How Much Thread Engagement Prevents Stripping?