Testing PBT Properly: DSC, HDT, Tensile and the Conditioning Step Everyone Skips

Polybutylene terephthalate is a semi-crystalline aromatic polyester that crystallizes rapidly from the melt and reaches a melting peak between 220 °C and 225 °C, with dry-state glass transition values reported between 45 °C and 60 °C depending on measurement technique, thermal history and moisture content. Injection-moulded PBT parts are seldom in thermodynamic equilibrium immediately after ejection; the outer skin cools faster than the core, mould temperature controls spherulite growth, and exposure to ambient humidity alters free volume and tensile response. A test programme that ignores this history produces DSC, HDT and tensile data that may be internally repeatable but not comparable across laboratories or across shifts.

The conditioning step in “Testing PBT Properly: DSC, HDT, Tensile and the Conditioning Step Everyone Skips” refers to more than moisture equilibrium. It includes the stabilisation of post-moulding crystallinity, relaxation of orientation and residual stress, and equilibration to a defined atmosphere before destructive testing. Standards such as ISO 291 and ASTM D618 exist precisely because polymeric test specimens do not have a fixed state at the moment of moulding.

How Should DSC Data Be Interpreted for Injection-Moulded PBT?

Thermal analysis of PBT should be carried out under ISO 11357-1 and ISO 11357-3 or ASTM D3418. Specimen mass is typically 5–10 mg, sealed or non-hermetically crimped aluminium pans are used, nitrogen purge is maintained at 20–50 mL/min, and a heating rate of 20 K/min is standard for routine melting-point and enthalpy measurements. The first heating scan records the as-received or as-moulded semi-crystalline state; a second heating scan after controlled cooling erases most fabrication history. For PBT, that distinction is not trivial because the first scan often contains a low-temperature melting population, a recrystallization exotherm and a dominant high-temperature melting endotherm. At slower heating rates the low-temperature population melts, recrystallizes and remelts, producing a double melting pattern that can be misinterpreted as bimodal crystal size distribution. Reporting only the peak temperature of the second heating scan without the cooling rate and thermal history loses the morphological information needed for root-cause analysis.

Crystallinity is calculated from the first heating melt enthalpy against a reference 100% crystalline enthalpy for PBT commonly cited between 140 J/g and 145 J/g. Laboratories should state which reference value is used; switching between 140 J/g and 145 J/g shifts the reported crystallinity by approximately 3–4% absolute on a 35% crystalline sample. The glass transition in semi-crystalline PBT may appear as a weak step between 45 °C and 60 °C, often overlapped with enthalpy recovery from physical ageing. If the sample is quenched or moulded into thin sections, a cold-crystallization exotherm can appear above Tg and must not be confused with degradation.

Instrument calibration also affects the reliability of PBT DSC results. Heat-flux instrumentation should be calibrated with indium and zinc over the range 150–250 °C, with temperature accuracy within ±0.5 K and enthalpy accuracy within ±2% based on indium. Cooling-rate control is equally important because PBT crystallizes rapidly; if the instrument cannot maintain a linear 20 K/min cooling ramp between 200 °C and 160 °C, the second-heating crystal population will differ. Fast DSC cooling rates of 100 K/min can suppress the high-temperature crystallization and freeze more amorphous material, but such data are not comparable with conventional 20 K/min results.

HDT Is Not a Material Property

For unfilled PBT, HDT/A at 1.80 MPa is typically reported in the 50–65 °C range, near the dry-state glass transition, while HDT/B at 0.45 MPa can exceed 150 °C because the lower stress permits greater contribution from the crystalline phase. Glass-fibre-reinforced PBT with 30 wt% short fibre can shift HDT/A above 200 °C, but this shift is a creep-deformation response of the fibre network and not a change in the melting point. The test follows ISO 75-2 or ASTM D648; the specimen is heated at 2 K/min in an oil bath while loaded in edgewise three-point bending. A reported HDT value without stress level, specimen thickness, thermal history and conditioning atmosphere is incomplete.

The conflict in HDT testing of PBT is that the property is often used as a heat-resistance ranking, but the standard itself does not measure creep rate, dimensional stability under load, or long-term heat ageing. For applications with sustained load above 80 °C, HDT/A values on unfilled PBT are not sufficient evidence of performance; additional creep modulus or DMTA data are required. The edgewise specimen is loaded in a three-point bending fixture with a span of 64 mm for ISO bars and heated in a stirred oil bath. The temperature sensor should be close to the specimen centre, and thermal uniformity must be stable enough to avoid premature deflection at an artificial hot spot. Because PBT is semi-crystalline, the deflection transition under 1.80 MPa is governed by the amorphous phase and can shift with moisture and ageing; under 0.45 MPa the test is more sensitive to crystal perfection and lamellar thickness. Reporting both HDT/A and HDT/B is therefore more informative than a single value. Comparing HDT/A values between a dry unfilled PBT and a conditioned glass-filled PBT without stating filler content and fibre orientation is invalid.

Tensile testing of PBT requires careful specimen selection because edge-gated ASTM D638 Type I bars and ISO 527-2 type 1A bars build different flow-induced orientation patterns. Specimens should be moulded to ISO 294 or the relevant ASTM mould standard, then conditioned before measurement. The tensile modulus is determined in the 0.05–0.25% strain interval at 1 mm/min; yield and break are recorded at 5 mm/min for unfilled PBT and many rigid grades. Use an extensometer meeting ISO 9513 class 1 or ASTM E83 class B-2. Unfilled PBT typically shows tensile strength of 50–60 MPa, modulus of 2.5–3.0 GPa, and elongation at break that can exceed 50% depending on molecular weight and moisture. Glass-fibre-reinforced PBT shows tensile strength near 120–140 MPa and modulus of 9–11 GPa, but the failure strain drops below 3–5%. These values are not generic; they shift with dampness, mould temperature and glass fibre sizing.

Moisture acts as a plasticiser in the PBT amorphous phase. A specimen sealed immediately after ejection can produce a higher tensile strength and modulus than a conditioned specimen because the external amorphous layer has not equilibrated with ambient water. The reverse can occur if the part is moulded with residual moisture or left in a humid environment; hydrolysis at moulding temperatures is an additional risk. In production-scale moulding trials on desiccant-dried PBT, the absence of a conditioning step is a common cause of inter-laboratory tensile disagreements and poor repeatability on yield strain.

Before any mechanical test, the moulding history requires drying. PBT hydrolyses at melt temperature if moisture exceeds 0.02 wt%. Desiccant drying at 120 °C for 4 h to a dew point of −40 °C is commonly specified; overheating above 140 °C can cause discoloration and molecular weight loss. The dry state is not a conditioning state for mechanical testing. A specimen can be moulded from dry granulate and still require standard atmosphere conditioning before tensile or HDT measurement.

When the Conditioning Step Is Omitted

Conditioning under ISO 291, class 23/50, or ASTM D618 Procedure A is not optional when comparative PBT data are required. The standard atmosphere is 23 °C and 50 % RH; specimens are held until constant mass or for a defined minimum interval, commonly 40 h for the ASTM procedure. For 4 mm multi-purpose ISO bars, equilibrium moisture uptake at 50 % RH is generally below 0.3 wt%, much lower than polyamides, but even this quantity influences the amorphous phase and the glass transition. If moulded bars are bagged immediately in polyethylene, the moisture state remains near the dry-as-moulded value, and subsequent tensile and HDT data represent a condition unlikely to exist in service or in another laboratory.

The skipped step also includes thermal stabilisation. After ejection from a 70–100 °C mould, PBT can continue to develop fine crystallinity over the following days at room temperature. A bar measured at 4 h and a bar measured at 7 days may therefore show different HDT/B and different first-scan DSC enthalpy, especially if the mould was operated at the low end of its recommended range. Conditioning to constant mass simultaneously provides a defined moisture level and a defined room-temperature ageing window. For production troubleshooting, the minimum data set is not just HDT or tensile strength; it is first-scan DSC, HDT/A and HDT/B, and tensile modulus, yield stress and elongation at break measured on specimens conditioned to the same protocol.

For glass-reinforced PBT, fibre orientation causes anisotropic tensile properties. ISO 527-2 type 1A bars moulded in an end-gated cavity have higher strength in the flow direction. Compared with injection-moulded plates machined into ASTM D638 Type I specimens, the orientation pattern differs and the test results are not interchangeable. This is especially relevant for HDT where a skin layer with longitudinal fibre orientation supports the flexural load; a specimen machined from a plaque may not reproduce the part-level fibre architecture.

When DSC cannot resolve the Tg clearly, dynamic mechanical thermal analysis at 1 Hz is an alternative because the storage modulus drop and tan δ peak are more pronounced in semi-crystalline PBT. However, DMTA data must be labelled with frequency and heating rate; tan δ values are systematically higher than DSC Tg values. The two methods do not produce identical glass transition values.

Failure to apply conditioning creates artificial outliers that are often attributed to resin lot variation. On a 1300 kN injection-moulding machine running 30 wt% glass-reinforced PBT at melt temperature 250–270 °C and mould temperature 80–100 °C, immediate tensile measurements may be biased by residual through-thickness temperature gradients and by water content below the equilibrium level of the testing laboratory. Published data for dry-as-moulded versus conditioned PBT with full resin provenance is limited, so the reproducible approach is to report the exact conditioning state in the test record and to use the standard atmosphere for pass-fail comparisons.