When a PBT GF30 injection-molded article is approved from a first article inspection (FAI) and subsequent production lots exhibit measurable warp, twist, or dimensional drift, the failure mode generally involves a convergence of processing variables rather than a singular material defect. The scenario described as "The First Article Was Flat and Production Is Not: Drift in a PBT GF30 Part" encompasses post-molding anisotropic shrinkage, differential cooling, and long-period crystallization relaxation that manifests within hours to weeks after ejection from the tool. The first article may be produced under conditions that are nominally identical to production, yet subtle differences in machine response, material lot characteristics, and mold thermal equilibrium generate dimensional outcomes outside the tolerance band established during qualification. Polybutylene terephthalate reinforced with 30 wt% glass fiber (PBT GF30) is a semi-crystalline engineering thermoplastic widely used in electrical connectors, automotive sensor housings, and precision mechanical components where flatness and hole-to-hole dimensional integrity are critical. Its rapid crystallization from the melt makes it an efficient molding material, but the same rapid crystallization creates sensitivity to process drift that is frequently underestimated until dimensional rejection occurs downstream.

What Mechanisms Initiate Post-Molding Dimensional Drift in PBT GF30?

Polybutylene terephthalate reinforced with 30% glass fiber by weight exhibits a melting temperature (Tm) of approximately 223°C to 225°C as determined by differential scanning calorimetry in accordance with ISO 11357-3 or ASTM D3418. The glass transition temperature (Tg) is observed in the range of 45°C to 50°C. These thermal transitions govern the solidification behavior during injection molding. PBT crystallizes rapidly from the melt; the addition of 30 wt% glass fiber further accelerates crystallization through heterogeneous nucleation at the fiber-matrix interface. This rapid crystallization, in combination with a mold temperature typically specified between 60°C and 90°C for dimensionally stable applications, produces a solidified skin layer at the cavity walls while the core remains molten during packing and early cooling. The first-article approval condition may inadvertently reference a specific thermal history—produced on a prototype mold with a particular cooling circuit configuration that happens to yield a flat part. Production tooling, after cooling channel balancing, tool steel replacement, or altered cycle parameters, imposes a different thermal footprint on the polymer melt. The resulting differential shrinkage between flow-direction and transverse-direction regions of the same part produces warpage that is not immediately apparent at ejection but develops over 48 h to 96 h as post-crystallization and moisture equilibration proceed. This post-molding dimensional evolution is underreported in engineering documentation because dimensional inspection frequently occurs within 24 h of molding, before asymptotic shrinkage is attained. Published data for this specific configuration—PBT GF30 parts that pass first-article flatness checks but fail production dimensional audits—is limited in the open literature, although the constituent crystallization kinetics and shrinkage behavior of glass-filled PBT are well-characterized in polymer processing studies. Post-ejection dimensional drift in PBT GF30 parts is governed by two competing processes: continued crystallization of amorphous regions that were frozen during rapid cooling, and physical aging of the constrained amorphous phase below Tg. Differential scanning calorimetry of molded PBT GF30 specimens frequently reveals a cold crystallization exotherm upon reheating from ambient temperature, indicating that cooling rates during production leave a metastable amorphous fraction that densifies over time. The density of PBT GF30 typically ranges from 1.50 g/cm³ to 1.55 g/cm³ per ISO 1183 or ASTM D792. An increase in crystalline fraction of 1% to 2% corresponds to a volumetric reduction that can produce a linear dimensional change of 0.1% to 0.3% depending on part geometry and the degree of mechanical constraint. This crystallization-driven shrinkage proceeds with a time constant dependent on storage temperature. At 23°C storage, equilibrium may not be reached for 7 to 14 days; at elevated storage temperatures approaching 70°C, the process accelerates substantially because the crystallization rate for PBT reaches its maximum value typically observed between 130°C and 150°C under quiescent conditions. The first article that appears flat may have been allowed to condition for a longer period prior to measurement, or may have been formed under conditions that minimized retained amorphous content. Production parts produced on a shorter cooling time or with a colder mold surface retain a greater amorphous fraction and consequently exhibit greater post-mold shrinkage. The cold crystallization peak temperature and enthalpy, measured per ISO 11357-7, provide quantifiable evidence of the metastable amorphous content; a molded part that displays a cold crystallization enthalpy of 2 J/g to 5 J/g has significant latent dimensional instability, while a part conditioned to equilibrium displays minimal cold crystallization signal.

Fiber Orientation Mapping and Shrinkage Anisotropy

Glass fiber orientation within a molded PBT GF30 part determines the local shrinkage coefficient. Along the fiber axis, shrinkage is constrained by the low thermal expansion and essentially zero shrinkage of E-glass fibers, which exhibit a coefficient of linear thermal expansion of approximately 5 × 10⁻⁶ K⁻¹. Perpendicular to the fiber axis, the polymer matrix dominates shrinkage, yielding values that are 2 to 3 times greater than the longitudinal values. The anisotropic molding shrinkage of PBT GF30 is frequently reported as 0.2% to 0.5% in the flow direction and 0.5% to 1.1% in the transverse direction when molded and measured per ISO 294-4 protocols. Fiber orientation itself is a function of shear rate history, mold geometry, and processing parameters. A first-article part produced on a lab-scale machine with a specific injection velocity profile develops a characteristic skin-core orientation structure. The skin layer, typically 0.1 mm to 0.3 mm thick depending on mold temperature and injection rate, contains fibers aligned predominantly in the flow direction due to high shear at the frozen-layer interface. The core region contains fibers with more random or transverse orientation resulting from extensional flow at the melt front and reduced shear in the core. Production transfer to a higher-clamp-force machine with different screw recovery characteristics and a modified injection velocity profile changes the skin-to-core thickness ratio and the degree of core fiber alignment. This shift in orientation structure can produce a part that is flat after initial cooling but warps as the temperature cycles through service or through a paint or coating operation. Under thermal cycling between −40°C and 85°C, a typical automotive validation protocol per ISO 16750-4, differential thermal expansion between flow-oriented and transverse-oriented regions generates internal stress and resultant curvature. The first article might coincidentally possess a symmetric fiber orientation about the part's neutral axis; production parts with asymmetric skin-core fiber distribution will inevitably develop curvature upon any temperature change, even if they measure flat at ambient temperature immediately after molding. Microscopic verification of fiber orientation is typically performed by polished cross-section inspection under reflected light microscopy, with specimen preparation guidance per ISO 20753, or by computed tomography for volumetric orientation tensor evaluation. Quantifying the orientation tensor using the Advani-Tucker formulation provides a scalar measure of alignment that can be correlated with measured shrinkage anisotropy.

Property	Test Method	Flow Direction	Transverse Direction
Molding shrinkage	ISO 294-4	0.2–0.5%	0.5–1.1%
CLTE (23–55°C)	ISO 11359-2	25–35 × 10⁻⁶ K⁻¹	70–100 × 10⁻⁶ K⁻¹
Density	ISO 1183 / ASTM D792	1.50–1.55 g/cm³
Water absorption (24 h, 23°C)	ISO 62	0.07–0.10 wt%

When Packing Pressure Drops Before Gate Freeze

Gate freeze time is the duration required for the molten polymer at the gate to solidify to the extent that further pressure transmission into the cavity becomes impossible. For PBT GF30 molded with a melt temperature of approximately 250°C and a mold temperature of 80°C, gate freeze time typically ranges from 3 s to 8 s for a nominal gate diameter of 1.5 mm to 2.0 mm in a cold runner system, depending on gate thickness and mold thermal conductivity. If the pack pressure hold time is set below this threshold, the cavity pressure at the moment of gate freeze is insufficient to compensate for the volumetric shrinkage that accompanies cooling and crystallization. The result is a part that exhibits lower density, higher void content, and greater shrinkage than a properly packed part. The first article may be produced with a conservative packing time that fully satisfies gate freeze requirements. During production scale-up, cycle time optimization frequently reduces packing time to meet throughput targets. A reduction from 8 s to 4 s may result in incomplete packing that is not immediately visible because the part may still be nominally flat at demolding. However, the reduced polymer density in the core region, combined with continued crystallization, produces delayed shrinkage drift that manifests as concavity on flat surfaces and altered hole-to-hole dimensions measured days after molding. Cavity pressure instrumentation provides a definitive method for detecting premature gate freeze. Piezoelectric cavity pressure sensors installed flush with the cavity surface, in accordance with ISO 17212, record the pressure-time profile during molding. A properly packed PBT GF30 part maintains cavity pressure above 10 MPa to 20 MPa until gate freeze, followed by a controlled exponential decay as the part cools under residual pressure. A prematurely frozen gate produces an immediate pressure drop toward atmospheric pressure at the end of packing, followed by a secondary pressure rise during crystallization-driven contraction. This characteristic pressure signature should be monitored on every cavity during production to detect process drift before dimensional rejection occurs. Moisture absorption in PBT GF30 is relatively low compared to polyamides, with equilibrium water uptake of approximately 0.07% to 0.10% by weight after 24 h immersion at 23°C per ISO 62. This low hygroscopicity means that moisture-induced dimensional change is typically smaller than crystallization-driven shrinkage; however, it is not negligible when dimensional tolerance bands are specified at ±0.05 mm or tighter. Parts equilibrated in a 50% relative humidity environment at 23°C will exhibit slightly different dimensions than parts measured in a dry-as-molded state. Dimensional inspection protocols must specify conditioning time and environment. ISO 291 defines standard atmospheres for conditioning and testing of plastics; the most commonly cited condition is 23°C / 50% RH, class 2. For PBT GF30 parts, a minimum conditioning period of 48 h is recommended before dimensional inspection to allow partial relaxation of molded-in stress and moisture equilibration. The first article's flatness may have been measured after extended conditioning, while production parts measured within 4 h of ejection will contain residual differential shrinkage that will resolve later, producing apparent drift that is actually measurement artifact. Molded-in stress in PBT GF30 parts can be evaluated by solvent stress cracking tests or by observing birefringence patterns under polarized light. The presence of high residual stress in production parts indicates that the process window has shifted from the first-article condition. Annealing of PBT GF30 parts at 120°C for 2 h to 4 h in circulating air ovens is practiced to stabilize dimensions; however, annealing induces additional crystallinity and must be accounted for in the dimensional specification. The annealed part will exhibit dimensions different from the as-molded part, and this difference must be captured in the FAI documentation. Flatness measurement itself requires specification of the reference datum system; coordinate measuring machine (CMM) inspection per ISO 1101 with a defined number of measurement points is recommended over manual feeler-gauge methods, which are inadequate for detecting the low-amplitude, long-wavelength curvature characteristic of differential shrinkage.

Thermal Imbalance Between Cavity Surfaces Constitutes a Root Cause

Differential cooling across the thickness of a PBT GF30 part is a frequently overlooked contributor to dimensional drift. If the core side of the mold is maintained at 80°C and the cavity side at 60°C, the asymmetric cooling produces a non-uniform solidified skin layer. The hotter mold surface delays skin formation, allowing a thicker oriented layer on that side and greater crystalline content near the surface. Upon ejection, the bimaterial-like structure of the solidified part relaxes toward a curved geometry. This warpage may not be apparent at ejection because the part is still above Tg and retains viscoelastic compliance, but it develops fully as the part cools to ambient temperature. Mold designers routinely specify conformal cooling circuits for dimensionally critical PBT GF30 parts, with a maximum temperature differential between core and cavity of 5°C recommended for flat parts with thin walls. In production, this differential is affected by coolant flow rate, coolant inlet temperature, flow line scaling, and local hot spots near gates or thick sections. Thermography of ejected parts using a calibrated infrared camera provides a spatial temperature map that can identify regions of non-uniform cooling. Published data from molding process studies indicates that a 10°C differential between cavity surfaces can produce warpage of 0.5 mm to 1.5 mm across a 100 mm span in glass-reinforced semi-crystalline materials, depending on part thickness and fiber content. This behavior must be validated for each specific part geometry and tool configuration. Equipment specifications for dimensional stability molding of PBT GF30 typically include mold temperature controllers capable of maintaining setpoint within ±2°C, coolant flow meters with ±0.5 L/min accuracy, and cavity pressure sensors with a sampling rate of at least 1 kHz to capture pressure decay profiles. On production lines, the use of turbulent flow conditions with Reynolds number greater than 4000 in cooling channels ensures adequate heat transfer coefficients, and sequential valve gating prevents packing irregularities in multi-gate parts. Material lot-to-lot variation in PBT GF30 compounds contributes to dimensional drift through changes in melt viscosity, fiber length distribution, and crystallinity. Melt flow rate (MFR) is typically specified at 250°C with a 2.16 kg load per ISO 1133-1 or ASTM D1238; PBT GF30 grades generally exhibit MFR values in the range of 10 g/10 min to 30 g/10 min, though the specific range depends on the manufacturer's formulation. A change of 5 g/10 min between material lots, while within the supplier's specification, is sufficient to alter the filling pattern, packing efficiency, and resultant fiber orientation in a thin-walled part. Production parts molded with a higher-MFR lot fill more easily but pack differently; the lower melt viscosity may also permit greater fiber migration toward the surface, changing the skin-layer thickness. Intrinsic viscosity (IV) of the PBT base resin, measured per ISO 1628-5 in a phenol/tetrachloroethane solvent mixture at 25°C, is a molecular weight indicator that correlates with mechanical integrity and crystallization rate. Typical IV values for injection-molding grades of PBT range from 0.80 dL/g to 1.20 dL/g. A decrease in IV from 1.10 dL/g to 0.90 dL/g reflects hydrolytic degradation of the polyester backbone, which reduces molecular weight, accelerates crystallization due to increased chain mobility, and increases shrinkage. Moisture exposure during material handling is the primary cause of IV degradation. PBT must be dried to a moisture content below 0.02% (200 ppm) prior to molding. The recommended drying conditions are 120°C for 4 h in a desiccant dryer with a dew point of −40°C or lower. If production drying bins are set to 80°C to reduce energy cost, or if dried material is left exposed to ambient air at 50% RH for more than 30 min, sufficient moisture is absorbed to begin hydrolysis at melt temperature, producing measurable IV reduction and downstream dimensional drift. Glass fiber length distribution is altered by screw shear during compounding and molding. A production twin-screw extruder with a 40:1 L/D ratio and specific energy input of 0.15 kWh/kg to 0.25 kWh/kg produces a particular fiber length distribution that affects shrinkage. After injection molding, the mean fiber length typically decreases from the pellet value due to additional screw shear; mean fiber lengths of 200 μm to 300 μm in pellets may reduce to 150 μm to 200 μm in molded parts, depending on screw design and back pressure. Fiber length influences both the constraining effect on shrinkage and the mechanical interlock at the fiber-matrix interface. A subtle reduction in mean fiber length, even within the supplier's published tolerance, reduces the dimensional constraint provided by the fiber network, resulting in slightly greater matrix-dominated shrinkage in transverse directions.

Verification Parameter	Standard / Method	Acceptance Criterion for Dimensional Stability
Melt flow rate	ISO 1133-1 (250°C, 2.16 kg)	Within ±2 g/10 min of FAI lot value
Intrinsic viscosity	ISO 1628-5	≥ 0.90 dL/g
Moisture before molding	Karl Fischer titration	≤ 0.02 wt% (200 ppm)
Mold temperature differential	IR thermography / cavity sensor	≤ 5°C across part surfaces
Packing time relative to gate freeze	Cavity pressure decay profile	≥ 1.2 × gate freeze time
Dimensional inspection conditioning	ISO 291 Class 2	48 h minimum at 23°C / 50% RH
Molding shrinkage verification	ISO 294-4	Flow: 0.2–0.5% / Transverse: 0.5–1.1%