How Do You Design for Insert Molding — Metal Set into Plastic?

Last updated: 2026-09-01 · Reading time: about 10 min

Insert molding places a metal component — a nut, a pin, a contact, a tube — into the mold before every shot, and the melt forms around it, locking the two materials into one part.
The method buys strength and assembly savings that post-mold assembly cannot match, but it transfers the failure risk into the design: a loose insert flashes, a cold insert cracks the plastic around it, an unkeyed insert spins in service, and a poor plastic section over a metal core sinks and stresses. Design the interface deliberately — location, retention, preheat, and plastic around metal — and insert molding becomes one of the most reliable processes in the shop.

Key facts

<>

  • Insert molding fuses two problems into one design task: the plastic part obeys the wall rules, and the insert interface obeys mechanics — retention against pull-out and rotation, location to the tolerance the insert itself carries.
  • Thermal mismatch is the silent failure: metal and plastic shrink at very different rates as the part cools, so residual stress concentrates around every insert — the crack-initiation site that filled, stiff resins feel first.
  • Retention is cut, not hoped: knurls, grooves, hex flats and through-holes give the plastic something mechanical to grip; a smooth cylinder holds by shrink pressure alone and loosens with temperature and creep.
  • Placement is a process decision with a price: hand-loaded inserts cap the cycle and add operator dependence, while automated placement buys consistency at tooling cost — the part-cost math decides, not habit.
  • The insert must be verifiable in the mold: presence and seating detected before injection, because one shot over a missing insert can damage more than the part.

    • Shrinkage — The dimensional reduction from mold cavity to cooled part. Amorphous resins: 0.4–0.7% (uniform). Semi-crystalline: 1.0–2.5% (directional). Mold makers apply shrinkage factor when machining cavity. Actual shrinkage varies with wall thickness, gate location and packing pressure. See also related guide.

    What is insert molding — and when does it beat the alternatives?

    Insert molding
    A process in which a preformed component — typically metal — is loaded into the mold before injection so the melt forms directly around it, creating a permanent mechanical and chemical bond between insert and plastic in a single molding cycle.

    The economics decide by comparison, so run the comparison honestly. Against threaded bosses, molded-in inserts trade an assembly step for a placement step every shot — usually winning at volume, in service strength, and in sealed applications where screws loosen.
    Against post-mold heat-staked or ultrasonic inserts, insert molding gives stronger retention (the plastic forms into every knurl, no press-fit interface), no second process to schedule, and no risk of damaging the part at insertion — but it couples the insert supplier into the molding operation, because a dimension out of tolerance now stops the press.
    Against two-shot molding, inserts handle materials the second shot cannot provide — conductivity, threads in metal, stiffness — which is why the two methods coexist rather than compete. The honest failure mode to keep in view is volume: every part carries the insert-loading labor or automation cost, so a design that barely beats a screw at prototype volumes may lose badly at production volumes, or the reverse.
    Decide with the program’s actual volumes on the table, and revisit the decision if volumes shift — the change-order process converts a late redesign into a plan instead of a crisis.

    Which insert types carry which design duties?

    Insert type Typical duty Design focus
    Threaded bushings Reusable metal threads in a plastic housing. Knurl or hex outside for torque; blind end sealed against flash into threads.
    Pins and shafts Locating, pivoting, electrical contact. Grooves for pull-out; flat or D-shape against rotation; free end supported in the mold.
    Electrical contacts and terminals Current path through the housing. Position tolerance dominates; plating masked and protected; flash into contact faces rejected.
    Tubes and fittings Fluid path through the part. Openings sealed against flash; thin plastic over metal watched for sink and stress.
    Stamped metal carriers Terminals and shields at volume. Carrier holes give anchorage; automation feeds the strip; burr control decides seating.
    Bushings and bearings Wear surfaces aligned in the plastic. Concentricity to the part datum; plastic section uniform around the insert.

    Whatever the type, three interface duties repeat: retention (the insert must not pull out or spin in service), location (the insert arrives at the position the drawing tolerates), and sealing (plastic must not leak into threads, bores or contact faces where it does damage).
    Each duty is answered in the mold — grooves cut retention, locating details set position, shutoffs and seals control flash — and verified in the process: first articles measured on the trial press, retention and torque tested to the numbers the application needs, recorded per the acceptance discipline.
    An insert interface designed without one of the three duties shows its gap quickly: the nut spins at assembly, the pin wanders off position, or the thread arrives plastic-filled.

    How do you manage the stresses where metal meets plastic?

    1Design the plastic section around the metal deliberately. Too thin over a large insert starves and sinks; too thick shrinks harder against the metal and raises residual stress — the section is balanced per the same logic that governs the boss and rib stock, then confirmed on the flow study.
    2Preheat the inserts. A room-temperature insert quenches the melt against it, freezing stress and weak bonds into the interface; warming inserts to a moderate temperature before placement softens the thermal shock and lengthens the process window.
    3Give the plastic something to hold. Knurls, rings, undercuts and through-holes turn retention from shrink-fit luck into geometry; smooth inserts belong only in low-load, low-life applications.
    4Control the insert’s temperature history in service, not just in molding. Creep relaxes a pure shrink-fit over years; where the joint carries load across temperature swings, mechanical retention is the design, and the shrink fit is a bonus.
    5Test the interface as a joint. Pull-out force and torque on samples from the trial, at temperature extremes where the application demands it — numbers from the drawing’s duty, not from a feeling of firmness.

    Which insert-molding checks belong in the review?

  • Insert drawing toleranced for molding — the insert’s own variation known and accommodated, because the mold cannot hold what the insert does not.
  • Retention features specified — knurl type, groove count and positions stated, not left to the insert supplier’s default.
  • Location and shutoffs designed — position held against the drawing, flash paths sealed away from threads and contacts.
  • Placement method chosen with eyes open — manual versus automated priced into the cycle and the rate, with misload detection included.
  • Stress reviewed at the mismatch — section balance and preheat planned; crack-sensitive geometries flagged on the flow study.
  • Interface tests defined — pull-out, torque and position checks with limits written, run at the trial and archived.

    • What does insert molding demand from the molding operation?

    It demands discipline the plain process never needed.
    Every shot begins with an operator or machine placing parts into moving steel — so the placement interface must be ergonomic or automated, the mold must sense presence where a miss would damage it, and the cycle must absorb the placement time honestly rather than rushing it into errors. Inserts bring their own logistics: cleaning, sorting, orientation and supply are now production inputs, and a batch of insert burrs or oil becomes a molding defect by way of seating.
    The payoff for meeting these demands is real — parts that arrive assembled, threads that outlive the housing, contacts aligned to positions machining would struggle to hold — but the payoff arrives only when the operation treats every insert as a component with a specification, not as a fitting the crew will “work in.” Programs that respect this run for years without a single insert-related stop; the ones that treat placement as trivia meet it at the first flash complaint.

    “The insert is a component, the interface is a design, and the placement is a process step — miss any one of the three and the molding machine will find it for you, one shot at a time.”
    — Arthur, Tooling Engineer at aCatMould

    Frequently asked questions

    Why did the plastic crack around our molded-in inserts?

    Almost always the thermal mismatch: the metal barely shrinks while the plastic pulls hard around it, and the residual stress sits at the interface waiting for impact, assembly loads or cold weather.
    The fixes stack — preheat the inserts, balance the plastic section, radius the stress concentrations, and reconsider very stiff or filled resins at tight radii. Cracking that appears days or weeks later is the same stress relaxing out; the design, not the lot, is usually the suspect.

    How much plastic do I need around a metal insert?

    Enough to transmit the load and enough to avoid sink and starvation — a balance found per geometry and resin, not a single universal number. The flow study reports both risks on your actual section: thin over metal shows fill trouble, thick shows shrink stress. What must never happen is plastic over a sharp insert edge at minimal thickness — that corner concentrates stress and initiates the cracks the previous question describes.

    Can inserts be loaded automatically?

    Most can, and at volume most should be — robotics or feed systems place bushings, pins and stamped carriers faster and more consistently than hands. The prerequisites are insert-friendly features: consistent stock, orientation-friendly geometry, generous placement access in the mold, and sensing that stops the press on a missing or unseated insert. The decision is priced per the part-cost arithmetic — automation cost against operator cost against scrap risk, over the program volume.

    What tolerance can we hold on insert position?

    Sum the stack honestly: the mold locates the insert, the insert’s own variation adds in, and the plastic’s shrink moves the plastic around it. Position tolerances tighter than the stack will hold are the most common insert-molding disappointment — the cure is locating from the insert’s datum where the function is, dimensioning the function rather than the whole insert, and confirming the achieved positions with measurements at the tolerance review.

    Do inserts need cleaning before molding?

    Yes — oils, drawing compounds and swarf interfere with bonding and seating, and contaminated inserts contaminate the interface no process setting can rescue. The level of cleaning matches the application: a degrease suffices for most mechanical inserts; electrical contacts need handling and packaging that protect plating as well as cleanliness. Make cleanliness a specification with a check, not an assumption.

    Is insert molding still worth it if volumes grow or shrink?

    It changes character more than it loses validity. At low volumes, manual placement keeps the tooling investment modest and the per-part cost honest. At high volumes, automation shifts the economics decisively. The trap is judging the method at one volume and never revisiting: the volume assumptions behind the placement decision belong in the process file, so the program re-prices the choice when volumes move instead of inheriting a decision made for a different business.

    Key takeaways

    1. Insert molding succeeds at the interface: retention by geometry, location by the mold, sealing by design — each duty answered before steel is cut.
    2. Thermal mismatch is the standing risk; preheat, balanced plastic sections and radius discipline keep the residual stress survivable.
    3. Placement is a process decision — manual or automated, sensed and priced — and the insert itself is a component with a specification, not a fitting.
    4. The interface is verified as a joint: pull-out, torque and position tested at trial and archived with the process file.
    5. Revisit the placement economics when volumes move; the honest method is decided per program, not inherited.

    If your housing needs metal where plastic cannot go, send the insert drawing and the housing CAD. You will get the interface reviewed — retention, location, stress and placement — with the risks ranked, from a named tooling engineer, before the mold commits to a guess.

    Sources
    ISO 20457:2018 — Plastics moulded parts: Tolerances and acceptance conditions, International Organization for Standardization
    Plastics Industry Association — industry resources on insert molding practice
    Autodesk Moldflow — injection molding simulation, reference for residual stress around inserts
    MatWeb — material property database, reference for thermal expansion of metals and resins

    Sources: Plastics Technology · Plastics Technology Online · Society of Plastics Engineers (SPE) · ThomasNet