Designing ribs and bosses for PBT: shrinkage, packing and the tolerance stack requires treating the mould cavity as a shrinkage-compensated metrology system rather than a fixed geometric replica. Polybutylene terephthalate is a semicrystalline aromatic polyester processed by injection moulding at melt temperatures between 245 °C and 270 °C. Unfilled PBT moulding grades exhibit typical flow-direction shrinkage of 0.018–0.023 mm/mm and transverse shrinkage of 0.020–0.025 mm/mm when tested as ISO 294-4 plaques at a mould wall temperature of 80 °C. A glass-fibre content of 30% by mass reduces flow-direction shrinkage to 0.002–0.006 mm/mm while transverse shrinkage remains 0.007–0.012 mm/mm, producing an anisotropy that must be compensated separately for rib and boss features.

From a process standpoint, PBT crystallises rapidly during cooling; the pressure-volume-temperature path of the melt therefore contains a sharp density change as the solidifying layer thickens. Packing must be completed before the gate freezes because material cannot be added to compensate crystallisation shrinkage after gate seal. Cavity pressure transducers provide the primary measured variable for gate freeze time rather than nominal timer settings. In multi-cavity production tools, cavity-to-cavity shrinkage scatter in PBT is controlled more strongly by gate freeze time and melt cushion consistency than by barrel temperature alone. Published process windows for glass-reinforced PBT specify mould temperatures from 80 °C to 120 °C to slow crystallisation enough to allow adequate packing of rib roots.

Because PBT is semicrystalline, the volumetric solidification shrinkage is larger than in amorphous resins such as polycarbonate. The crystalline phase is denser than the amorphous phase, so a substantial fraction of total shrinkage occurs during the liquid-to-solid transition. Tensile modulus of unfilled PBT is typically 2.5–2.8 GPa under ISO 527-2; a 30% glass-fibre grade raises flexural modulus to approximately 9.0–11.0 GPa. The higher modulus means that shrinkage-induced internal stress produces less strain but higher stress in reinforced PBT, which can manifest as rib-root crack formation if the radius at the rib base is below 0.5 mm.

Shrinkage anisotropy and cavity pressure requirements in PBT

Shrinkage anisotropy in reinforced PBT arises from skin-core orientation differences and the alignment of glass fibres along the flow field. Flow-direction shrinkage is restrained by fibre orientation, while transverse shrinkage is controlled by matrix crystallisation and fibre-fibre packing. A 30% glass-fibre PBT grade can therefore show an anisotropy ratio of 2:1 to 4:1 between transverse and flow shrinkage. For a rectangular rib attached to a nominal wall of 3.0 mm, the local wall thickness around the rib base must be analysed as a separate shrinkage zone. If the tool is cut using a single isotropic shrinkage factor, the transverse dimension of the rib root will be undercompensated and the flow-length dimension will be overcompensated.

Cavity pressure traces recorded during injection of unreinforced PBT show that holding pressure in the range of 40–70 MPa is required to maintain mass flow into thin rib sections after the wall has filled. The required holding pressure is not a machine setpoint only; it is the cavity pressure at which the gate remains open and melt continues to flow into the rib root. A drop in cavity pressure below the solidification threshold, typically observed as a plateau in the cavity pressure trace, indicates that the rib is no longer being packed. Parts produced after this plateau show higher sink-mark depth and a wider spread in rib-root shrinkage than parts packed until gate seal.

Rib base width in unfilled PBT should be limited to 0.5–0.6 of the adjacent wall thickness t. This ratio prevents excessive local mass accumulation and reduces the depth of sink marks on the opposite surface. A rib height of more than 3 × t introduces fill hesitation, particularly with glass-reinforced PBT, because the thinner rib section begins to freeze before the main wall has reached full packing. Draft angle on the rib sidewalls should be 0.5° minimum for unfilled PBT and 1.0° minimum for 30% glass-fibre PBT. Ejection drag on polished surfaces is often observed when textured PBT ribs are drafted below 0.25°, and stress whitening at the rib base becomes visible after repeated ejection cycles.

Spacing between adjacent ribs in PBT is recommended at 2 × t or greater. When ribs are spaced closer than this, the interrupted wall behaves as a thicker average section and the mould cavity replicates a series of cold-wall zones. Differential cooling at the rib intersections produces local crystallinity gradients that can drive warpage. For a nominal wall of 2.5 mm, a rib base width of 1.25–1.50 mm and height of 7.5 mm maximum are consistent with published design guidelines for unfilled PBT.

What packing pressure actually influences in PBT rib features

The main effect of packing pressure in PBT rib features is not the elimination of part shrinkage as a single value, but the reduction of the shrinkage difference between the rib and the adjoining wall. When the rib base is too thick, the rib solidifies more slowly than the wall. The thicker local section continues to crystallise and shrink after the surrounding wall has already frozen, producing a sink mark that can be measured as a local thickness depression. Increasing holding pressure can reduce sink depth only if the gate and runner remain open. In a poorly gated rib, the pressure drop along the rib length prevents the cavity pressure from reaching the threshold required to force melt into the rib root.

Cavity pressure data from injection moulding trials show that effective packing of PBT requires a cavity pressure curve with a distinct holding plateau and no sharp pressure drop before gate seal. If the gate freezes in less than 2 s, the PBT melt has not received sufficient compensation for crystallisation shrinkage and the rib base will exhibit voiding. Hydraulic injection machines with screw diameters above 35 mm must be tuned so that transfer to holding occurs before the melt front reaches the end of the rib; otherwise the shot volume cannot be reduced into a controllable packing phase.

Gate diameter for ribbed PBT parts should not be less than 0.5 of the rib base width. If a tunnel gate of 0.8 mm is used on a 1.5 mm rib base, the gate freezes before the rib root packs and the effective shrinkage of the rib becomes decoupled from the wall. This decoupling is a common cause of sink marks that cannot be corrected by raising holding pressure alone.

Coring, gussets and the risk of flow-induced displacement

Boss side walls in PBT are specified as 0.5–0.6 × t to avoid sink marks at the base. A cored hole diameter of 0.6–0.8 of the boss outer diameter creates a uniform wall around the core and prevents non-uniform shrinkage. When the boss is taller than 2 × outer diameter, gussets with a thickness of 0.4–0.6 × t are required to resist core-pin deflection and to maintain the hole position under cavity pressure. The core pin should be cooled independently on deep bosses because the surrounding PBT wall can remain molten for a prolonged period, increasing the risk of core pin seizure.

Weld lines are a structural risk in PBT bosses located downstream of flow obstructions. Glass-fibre reinforced PBT forms a weld line with a tensile strength that may be 35–60% of the base material strength when tested under ISO 527-2. The reduction is caused by fibre orientation perpendicular to the weld plane and incomplete molecular re-entanglement across the flow front. Moving the gate closer to the boss or placing a flow leader between the gate and the boss reduces the weld-line area and improves packing at the boss base.

When a 30% glass-fibre PBT grade imposes new tolerance limits

Switching from unfilled PBT to a 30% glass-fibre grade changes the tolerance stack because the shrinkage data change from quasi-isotropic to anisotropic. For a nominal length of 100 mm, flow-direction shrinkage of 0.002–0.006 mm/mm produces a residual linear variation of 0.20 mm if the mould is cut at the midpoint 0.004 mm/mm. The same part dimension in the transverse direction, with shrinkage of 0.007–0.012 mm/mm, produces a residual variation of 0.25 mm when cut at the midpoint 0.0095 mm/mm. In complex housings with ribs and bosses oriented at different flow angles, the local shrinkage can fall between these two orthogonal limits, and the tolerance band must be widened unless flow-path simulation is used to generate a locally variable shrinkage map.

The difference between flow and transverse shrinkage of 0.005 mm/mm over 100 mm corresponds to 0.5 mm of differential contraction. This does not automatically produce 0.5 mm of warpage, but it defines the upper bound of the differential contraction term in the flatness tolerance stack. Rib and boss features that connect flow-parallel and flow-transverse walls are particularly sensitive because they act as local restraints and can move the neutral axis of the section.

The tolerance stack for a PBT rib or boss is assembled from cavity machining tolerance, mould steel thermal expansion, anisotropic mould shrinkage, post-mould moisture movement and warpage. Mould steel of the P20 type has a linear thermal expansion coefficient of approximately 12 × 10⁻⁶ K⁻¹. A cavity heated from 20 °C to 80 °C expands by 0.00072 mm/mm, which adds 0.072 mm to a 100 mm nominal dimension before shrinkage compensation. This term is removed from the measured steel dimension if the mould is measured at operating temperature, but many tool shops measure at ambient temperature and must correct the recorded values.

For a one-dimensional stack in the transverse direction, the arithmetic worst-case residual shrinkage uncertainty is U = L × ΔS / 2, where L is the nominal length and ΔS is the published shrinkage range. A root-sum-square combination is permissible only if the individual terms are statistically independent. In PBT the shrinkage terms are not independent, because crystallinity, fibre orientation and gate freeze time couple flow and transverse shrinkage to the same thermal history. The arithmetic stack is therefore preferred for prototype tooling acceptance.

Post-mould moisture uptake in PBT is measured by ISO 62. Conditioning before metrology at 23 °C and 50% RH for at least 24 h reduces the dimensional drift that occurs when dry-as-moulded parts absorb water during storage. The dimensional change from moisture uptake is not isotropic in glass-reinforced PBT, and commercial tolerance analyses often add a moisture movement allowance of 0.03–0.08 mm for dimensions in the 50–100 mm range when parts are not measured in the dry state.

Dimensional acceptance of PBT parts is typically evaluated against DIN 16742:2013 general tolerance series for moulded parts. Rib-to-wall thickness ratios and boss positions are functional dimensions that often require tighter mould-making tolerances than the outer envelope. The tolerance stack for a boss position should therefore separate form tolerances from position tolerances; if both are assigned to the same cavity feature, the resulting stack overstates the true process capability.

Boss design under near-zero draft conditions

Near-zero draft bosses are sometimes required for bearing seats or press-fit inserts. For PBT, draft angles below 0.25° create ejection forces high enough to score the boss surface on textured steel. With 30% glass-fibre PBT, the risk is higher because the surface can contain exposed glass fibres that increase friction against the core. Mould release agents reduce ejection force but can introduce interference with painting or bonding operations. In such cases, the boss should be fitted with a polished collapsible core or a sleeve-ejector system rather than relying on a fixed core pin. Production tools with polished core pins and ejection sleeves can operate with boss drafts below 0.5° by separating the ejection force from the cored surface, provided the core temperature is held below 60 °C to limit heat deflection.

If the near-zero draft boss is gated at the base, the sleeve must not obstruct the packing path. A gate placed tangentially to the boss allows melt to rotate around the core and decreases core displacement during fill. The resulting boss will exhibit less roundness deviation than a boss filled directly through a central gate because the pressure field around the core is more symmetrical.