Polybutylene terephthalate (PBT) is a semicrystalline thermoplastic polyester processed at melt temperatures of 240–260 °C and mold temperatures of 60–80 °C. Its dry unfilled density is 1.31 g/cm³ per ISO 1183-1, and the melting point is typically 223–225 °C. In injection-molded electrical connectors, relay housings, sensor carriers, and glass-fiber-reinforced structural components, the resin arrives with a certificate of analysis that lists lot number, moisture, melt volume-flow rate, and mechanical test values. The internal production record—drying conditions, machine settings, regrind ratio, colorant lot, cavity-pressure data, and test-coupon results—constitutes the PBT paper trail. When that trail breaks, the failure mode is not limited to a nonconformance report. The affected lot cannot be cleared for shipment, audit evidence cannot be reconstructed, and field returns may escalate to recall. The operational sequence addressed here is What Goes Wrong Without the PBT Paper Trail: Stops, Audits, Recalls.

Drying Record Gaps and the Hydrolytic Molecular Weight Crash

PBT hydrolyzes through ester-link cleavage when melt processing occurs above 0.02 % residual moisture. A desiccant dryer with a dew point of −30 °C or lower and a bed temperature of 120–130 °C is required for 3–4 h before molding. The dew-point value, bed temperature, and drying residence time are recorded as the first active control on the paper trail. If the drying log is missing, there is no evidence that moisture was reduced below the 0.02 % limit, and the hydrolytic chain-scission event cannot be excluded. The melt volume-flow rate may then rise under ISO 1133-1:2022, while tensile strength per ISO 527-2 and notched Charpy impact per ISO 179-1 fall below engineering margins. Exact property loss is grade-dependent and published data for a specific undocumented drying deviation is limited. On a production line, the absent log is nevertheless sufficient for a stop-order because the incoming moisture certificate cannot be matched to the material actually molded. Open containers held at relative humidity above 60 % require re-drying; amine-based additives should be avoided with hydrolytically sensitive PBT unless compatibility testing is documented.

In a multi-cavity hot-runner mold producing automotive underhood connector bodies, batch-to-batch variation in PBT viscosity is controlled by the recorded melt condition and fill time. A 80-ton electric injection molding machine running a hot-runner manifold at 250 °C may generate cavity-pressure curves that distinguish a resin-viscosity shift from a blocked gate or worn check ring. If the saved curves are absent, a lot with higher melt flow will still fill the cavity but may produce flash, warpage, or a fall in weld-line strength below the PPAP acceptance limit. The paper trail must tie the cavity-pressure signature to the material lot number and the certificate of analysis. Without that linkage, automotive supplier quality engineers stop the line under IATF 16949 because dimensional and material traceability cannot be demonstrated. Published data for the effect of undried PBT on cavity-pressure variation in hot-runner connector molds is limited; industrial practice, however, is quarantine until the lot can be re-verified by ISO 527-2 and ISO 179-1 testing.

What Goes Wrong When the Flammability File Loses Its Lot Number?

Flame-retarded PBT grades are classified by vertical burn testing under UL 94 and glow-wire testing under IEC 60695-2-12 or IEC 60695-11-20. The V-0 listing printed on a yellow card is valid only for the exact formulation, thickness, and colorant system submitted to the test laboratory. If the lot number of the flame-retardant masterbatch or the compounding line log is missing, the molded part cannot be linked to the listed formulation. That is a near-term audit finding, not a theoretical one: a customer auditor cannot confirm whether the supplied connector still contains the same brominated or halogen-free FR package, the same antimony trioxide ratio, or the same polymer base resin. Re-testing may show V-0 at 0.75 mm, but a different colorant lot can alter the rating to V-2, with afterflame time exceeding the V-0 limit of 10 s. The absence of the lot-level flammability file forces a stop-shipment and a third-party re-test at an ISO/IEC 17025 laboratory; if the component is already in field use, the issue escalates to a recall under the relevant safety regulatory framework. Comparative tracking index per IEC 60112 may also shift below 250 V when the additive lot changes, compromising electrical spacing assumptions.

PBT grades used in drug-delivery actuators, inhalation-device components, and some food-contact articles require biological reactivity data and extractables test results carried through the material lot. Under ISO 10993-5:2009 and USP Class VI, a molding supplier may be expected to retain the material lot number, the processing lot, and the test house raw data for cytotoxicity and systemic injection. A PBT grade that meets 21 CFR 177.1660 for poly(tetramethylene terephthalate) still requires documentation of the additive composition and the extraction conditions. If the paper trail breaks between the material supplier and the mold lot, a medical device manufacturer cannot verify that the component lot is equivalent to the biological safety file. Under 21 CFR 820, production and process controls, including device master records and device history records, make that missing link a formal audit observation. For implant-adjacent or respiratory devices, field containment may be initiated even before recall because the risk assessment cannot rule out a formulation change. The operational boundary is clear: without lot-level extractables documentation, the component is not acceptable for ISO 13485 release.

Tracing the Audit File in Color Concentrate and Additive Dispersion

Color concentrates supplied for PBT molding are not neutral to performance. A carbon-black masterbatch can lower volume resistivity; a pigment lot can shift the tracking index and aging behavior. The paper trail records the masterbatch lot number, the let-down ratio, the mixer screw speed, and the location of the feed port. In a compounding line with a twin-screw extruder of L/D 40:1 and side-feeding of glass fiber, the dispersion log is an audit attachment to the material file. If the log is missing, nonuniform pigment dispersion may be found only after parts are molded; melt streaks and inconsistent color measured by a spectrophotometer are then traced back to the unrecorded masterbatch addition. In regulated electrical equipment, a colorant change without documentation invalidates the material declaration and may expose the final parts to RoHS compliance re-evaluation under IEC 63000. Under ISO 9001:2015 clause 8.7, nonconforming outputs are segregated, but segregation without a root-cause record is not corrective action.

A twin-screw compounding line running 30 % glass-fiber-reinforced PBT records die pressure, screw torque, zone temperatures, vacuum level, and pelletizer speed. The same record must state the glass-fiber lot, the coupling-agent addition, and the sieve-pack pressure before the die. Without that log, the glass content cannot be assumed from the pellet density alone; ash content per ISO 3451-1 must be re-run, and the tensile strength of 130–140 MPa expected for a 30 % glass-fiber-reinforced grade cannot be certified. If the extrusion log is missing, an audit will treat the entire silo of reinforced PBT as unverified stock. The cost is not only re-testing; the compounded lot cannot be assigned to automotive or electrical customers until the fiber percentage and coupling-agent trace are restored. Published data for the exact effect of unrecorded side-feeder speed on dispersion in L/D 40:1 PBT compounding is limited, but the missing value itself is an audit nonconformance because the process definition cannot be proven.

When Regrind Ratios Are Not Recorded

Regrind is permitted in PBT molding only when the heat history is controlled and the ratio is documented. A typical supplier limit is 20–25 % regrind for noncritical parts; flame-retarded and medical grades may require virgin material only. Each pelletizing or grinding event is a thermal history cycle that can reduce molecular weight and shift the melt-volume rate. If the regrind ratio is not recorded, the lot genealogy collapses. A part molded with 30 % regrind from an unrecorded source may contain material that has already been dried, melted, pelletized, re-dried, and remelted, changing the notched Charpy impact and the UL 94 performance. The only defensible documentation is a fill-in log that records the virgin material lot, the regrind lot, the blend percentage, the granulator cleaning record, and the moisture analysis after blending.

At incoming inspection, PBT resin sampled from a bulk silo truck may differ from the certificate of analysis if the sample was drawn before settling or if the carrier compartment was not washed. The paper trail must connect the sample port, the silo load, the retained grit-screen specimen, and the incoming moisture result. When the incoming log shows 0.08 % moisture against a CoA of 0.02 %, the receiving lot is placed on hold, and downstream production stops until the drying record can be reconstructed from the silo and hopper data. If the retained sample is missing, the audit finding is irrecoverable because the discrepancy cannot be assigned to transport contamination or resin hydrolysis. In that state, the batch cannot enter the material handling stream; the traceability chain is closed at intake, and no compounding or molding record downstream can repair it.