Polybutylene terephthalate (PBT) is a semicrystalline aromatic polyester produced by the transesterification of dimethyl terephthalate with 1,4-butanediol. In reinforced injection-molding grades, short glass fibers are incorporated at loadings such as 30% by weight (PBT GF30) and 15% by weight (PBT GF15). The 30% glass-filled grade typically provides a flexural modulus of 8.0–10.0 GPa under ISO 178:2019, while the 15% glass-filled grade typically falls to 5.0–6.5 GPa. A downshift from PBT GF30 to GF15 without losing stiffness is therefore not a direct material substitution; it is a system-level exercise in which fiber length retention, fiber orientation, interfacial coupling, mold design, and load-bearing geometry are manipulated. The lower density of GF15, typically 1.41 g/cm³ versus 1.53 g/cm³ for GF30 under ISO 1183-1:2019, is one of the primary motivations for downshifting in weight-sensitive automotive and electrical applications.

What Limits Stiffness Retention When Glass Content Drops from 30% to 15%?

The modulus of a short-glass PBT compound is controlled by the volume fraction, aspect ratio, and orientation state of the fiber. Glass fiber has an elastic modulus near 72 GPa, while unfilled PBT has a tensile modulus near 2.5 GPa. At 30% by weight, the fiber volume fraction is approximately 0.18; at 15% by weight, it is approximately 0.08. In a random-orientation short-fiber composite, the effective modulus can be estimated from the Halpin-Tsai equation. Using a fiber aspect ratio of 25 after molding and an orientation efficiency factor of 0.5, the calculated tensile modulus drops from roughly 9.5 GPa to 5.5 GPa when the weight fraction is halved. The structure of PBT itself—rapid crystallization, high nucleation density in the presence of glass, and a glass transition temperature near 45–60 °C—does not compensate for this filler-volume reduction.

In a twin-screw compounding line using a 40:1 L/D co-rotating extruder, glass fiber is side-fed downstream of the polymer melt seal to limit breakage. Short-glass PBT compounds typically retain a number-average fiber length of 250–350 µm after compounding. After injection molding, fiber length decreases further: GF30 may retain 220–300 µm; GF15 often falls to 200–280 µm. The viscosity reduction in GF15 increases shear transmission to the fiber surface during plastication, so screw design and back pressure become critical. A low-shear screw with a compression ratio of 1.8–2.2, a free-flow non-return valve opening of at least 5.0 mm, and a back pressure of 3.0–5.0 MPa have been used to preserve fiber length. Published data for the exact shape of the fiber length distribution shift in downshifted GF15 is limited; however, scanning electron microscopy of ashed specimens shows that retaining number-average fiber length above 300 µm narrows the flexural modulus gap by 8–15% compared with severely degraded GF15 at 180 µm or below.

Glass fiber diameter also influences stiffness retention in low-glass compounds. Fibers with a diameter of 10 µm provide a higher specific surface area and can increase tensile modulus relative to 13 µm fibers at the same weight fraction, but they are more prone to breakage during compounding. The roving tex, typically 2400 tex for continuous glass supply, is less critical in short-glass compounds than the chopper length and feed location. A chopper length of 3.0–4.5 mm before side feed, combined with a distributive mixing zone of 8–12 L/D elements, can produce a number-average fiber length in GF15 above 280 µm after extrusion.

Fiber Orientation at the Gate and Its Effect on Flexural Modulus

Injection-molded short-glass parts develop a layered microstructure: a thin frozen skin with little orientation, a highly oriented subsurface layer, and a core with more randomized or transverse fiber orientation. The proportion of the oriented subsurface layer increases as wall thickness decreases. For a nominal wall of 1.5 mm or less, the flow-direction flexural modulus of a GF15 part can approach 7.2–7.8 GPa, while the cross-flow modulus remains 4.8–5.5 GPa. A film gate or edge gate with a land length of 0.5–1.0 mm and a gate cross-section of 60–80% of the part wall thickness promotes planar fiber alignment. Hot runner valve gates with a nozzle tip diameter of 0.8–1.2 mm generate high shear and can degrade fiber length; sequential valve gate opening may be required to shift weld lines away from load-bearing ribs. Mold-filling simulation using a Folgar-Tucker orientation model has predicted that a strongly aligned GF15 flow-direction modulus can close the gap with GF30 by 1.8–2.2 GPa under simulated tensile loading.

Chemical coupling at the glass-matrix interface is usually achieved with an amino-silane size applied to the glass fiber. The effect of coupling on flexural modulus is less than the effect of fiber orientation, but it can shift the effective modulus of GF15 by 0.4–0.7 GPa. In PBT, carboxyl end groups participate in transesterification and hydrolysis; a resin with low carboxyl end group content reduces chain scission during compounding and extends the processing window. A coupled GF15 compound may show tensile strength of 95–110 MPa under ISO 527-2:2012, while a non-coupled GF15 of the same glass content may fall to 80–90 MPa. Nucleating agents such as sodium stearate at 0.1–0.3 wt% raise crystallization temperature and reduce post-mold shrinkage, but they do not materially change flexural modulus. Avoid combination with amine-based flame retardants or mold release agents because these can accelerate hydrolytic degradation at melt temperatures above 270 °C.

When Rib Design and Wall Thickness Restore Load-Bearing Behavior

For structural housings and brackets, the performance criterion is usually deflection under a specified load, not the coupon modulus. Bending stiffness of a plate scales with the third power of wall thickness. Increasing a base wall from 2.0 mm to 2.4 mm raises geometric bending stiffness by approximately 73% at constant modulus; this can compensate for a 20–25% lower flexural modulus in GF15. A rib root thickness of 0.6–0.8 times the adjacent wall, a rib height of 3–5 times the wall, and a draft angle of 0.5°–1.0° minimize sink marks while increasing the moment of inertia. Mold-flow analysis must account for the higher linear mold shrinkage of GF15 in the flow direction, typically 0.4–0.6% compared with 0.2–0.4% for GF30 under ISO 294-4:2018. This anisotropy difference changes the dimensional tolerance stack in clip latches and snap fits.

Processing Window Conflicts Between GF30 and GF15 Compounds

Drying of both compounds is mandatory: 120–130 °C for 4–6 h in a desiccant dryer to a moisture content below 0.02%, measured by ISO 15512:2019. Hydrolytic degradation at melt temperatures above 280 °C will reduce molecular weight and cause a fall of 0.3–0.5 GPa in flexural modulus before visible splay appears. GF30 typically requires a rear zone of 230–245 °C, middle zone of 245–260 °C, front zone of 250–265 °C, and nozzle of 255–265 °C. GF15 is lower in viscosity and should be run with the rear zone reduced by 5–10 °C, screw speed capped at 60–100 rpm, and cushion held at 3–6 mm to prevent fiber orientation relaxation and drool. The mold temperature for GF15 should be 80–100 °C for general applications; raising to 110–120 °C improves crystallinity and heat deflection temperature under ISO 75-2:2013 method B by 3–5 °C, at the cost of 8–12% longer cooling time.

Capillary rheometry data at 250 °C and a shear rate of 1000 s⁻¹ typically show apparent melt viscosity near 250–350 Pa·s for GF30 and 120–180 Pa·s for GF15 according to ISO 11443:2014. This viscosity reduction allows GF15 to fill thin walls at lower filling pressure. In a spiral flow mold with a 2 mm thick channel, GF15 flow length may be 25–35% longer than GF30 at the same injection pressure. The lower filling pressure reduces core pin deflection and permits smaller clamp tonnage; however, the same viscosity drop can cause jetting in large gates and fiber orientation loss in thick sections. A valve-gated hot runner with a tip diameter of 0.8–1.0 mm and a shear rate below 100,000 s⁻¹ is used to avoid brittle weld lines.

The anisotropy of shrinkage in glass-filled PBT is the primary source of warpage in downshifted parts. GF30 typically exhibits parallel shrinkage of 0.2–0.4% and perpendicular shrinkage of 0.9–1.2%, while GF15 can show parallel shrinkage of 0.4–0.6% and perpendicular shrinkage of 0.8–1.3%. Because the flow-direction shrinkage increases when glass content falls, the molded part shrinks less anisotropically and may have lower total out-of-plane warpage. This is often the main reason for downshifting. The mold designer can use the lower anisotropy to reduce the number of correcting tool modifications, but the part must be checked for sink marks and dimensional compliance under ISO 294-4:2018. Gas counterpressure or conformal cooling has no direct effect on glass content but can stabilize crystallization and reduce asymmetric shrinkage.

Failure Modes in Downshifted PBT Components Under Thermal and Chemical Load

In underhood environments, PBT GF15 is more sensitive to creep and fatigue than GF30 because the lower fiber volume fraction reduces load transfer under sustained stress. Creep modulus measured under ISO 899-1:2018 shows a greater time-dependent decline for GF15. At 80 °C and 20 MPa tensile stress, GF30 may retain 70–80% of its initial tensile modulus after 1000 h; GF15 may retain 55–70% under the same conditions. This difference means that a simple short-term flexural modulus comparison under ISO 178:2019 is insufficient. Chemical exposure to hot engine oil, coolants, or road salt can hydrolyze the PBT matrix preferentially at the fiber interface. If the part operates in a condensing environment at 85 °C and 85% RH, moisture uptake increases the loss tangent near the glass transition; retention of stiffness in GF15 requires a hydrolysis-stabilized PBT grade, typically containing a polymeric carbodiimide or an epoxy-based chain extender at 0.5–1.5 wt%.

Differential scanning calorimetry at 10 °C/min under ISO 11357-1:2016 shows that GF15 crystallizes at 190–200 °C on cooling, while GF30 may crystallize at 195–205 °C due to higher nucleation density. The lower crystallization temperature of GF15 can produce a slightly lower crystallinity in thick sections, reducing modulus by 0.2–0.4 GPa if mold temperature is below 80 °C. Raising mold temperature to 110 °C restores crystallinity and stabilizes dimensions.

In a 32-pin automotive connector housing originally molded from PBT GF30, a downshift to GF15 was evaluated to reduce warpage and improve pin retention. The part wall thickness is 1.2 mm, flow length 120 mm, and the tool uses a valve-gated hot runner. The GF15 material reduced injection pressure from approximately 85 MPa to 62 MPa at the same melt temperature and allowed the use of a 1,000 kN clamp force machine instead of 1,500 kN. To recover bending stiffness, two longitudinal ribs with a root width of 0.7 mm and a height of 2.5 mm were added along the housing back wall. The gate was repositioned to align the major fiber orientation axis with the pin row. After conditioning at 23 °C and 50% RH according to ISO 291:2008, the ribbed GF15 housing exhibited deflection of 0.12 mm under a 40 N insertion load, while the unmodified GF30 housing deflected 0.14 mm in the same fixture. This result is specific to the tested geometry and load case; it does not establish general material equivalence under ISO 527-2:2012 tensile modulus or ISO 179-1:2010 impact resistance.

A laser printer paper-guide bracket molded in PBT GF30 was similarly evaluated with GF15. The bracket is a cantilever beam with a base width of 18 mm and a wall thickness of 2.2 mm. Direct material substitution produced a deflection increase of approximately 25% under a 5 N end load. The design was corrected by increasing the wall thickness to 2.6 mm and adding a gusset with a root thickness of 1.5 mm. In the revised geometry, the GF15 bracket deflection at 23 °C was 0.42 mm, compared with 0.45 mm for the original GF30 bracket. Mold-filling analysis showed that the thicker wall did not create a sink mark exceeding 0.01 mm because the glass fiber remained oriented along the flow path. The change eliminated 120 g of material per 1,000 parts due to the lower density of GF15 and reduced cycle time by 4 s because of the lower melt viscosity and faster plasticating.