Polybutylene Terephthalate (PBT) in Contact with Fuels, Oils and Coolants: Verifying Compatibility Before a Trial

Polybutylene terephthalate is a semicrystalline engineering polyester with a melt temperature of approximately 223–225 °C for unfilled grades and a dry glass transition temperature near 45–60 °C. The compound is selected for fuel sender flanges, coolant valves, oil-control pistons, and under-hood connector bodies because of its short crystallization time, high stiffness, and resistance to nonpolar aliphatic hydrocarbons. The response to aggressive fuels, oils, and coolants is not a single material constant. It is governed by molar mass retention, carboxylic acid end-group concentration, molded-in stress, fiber orientation, and additive package. Polybutylene Terephthalate (PBT) in Contact with Fuels, Oils and Coolants: Verifying Compatibility Before a Trial demands that the exact production compound be immersed in the actual service fluids, at the maximum expected temperature, for enough time to detect both progressive degradation and cliff-edge failure.

When Fuel Blends Shift from Reference Solvents to Oxygenated Aggressors

PBT absorbs only trace amounts of nonpolar aliphatic hydrocarbons. The aromatic fraction in commercial gasoline, principally toluene and xylene, plasticizes the amorphous phase and increases volume swell. ASTM Reference Fuel C is a mixture of 50 % toluene and 50 % isooctane by volume and is used to represent an aggressive aromatic gasoline without oxygenates. Swelling in a glass-fiber-reinforced PBT is anisotropic because the fibers constrain the matrix in the flow direction; a volume change of 1.2 % may produce a linear dimensional change of 0.20 % in the cross-flow direction but 0.05 % in the fiber direction. This introduces sealing stress at weld lines and insert interfaces. Methanol and ethanol add alcoholysis risk. At temperatures above 60 °C, methanol can attack the ester linkage and extract low-molar-mass oligomers, causing surface etching and slow crack development. Published data for long-term PBT immersion in methanol-containing fuel blends above 10 vol% alcohol is limited, so a trial should not rely on reference fuel alone.

Specimens are cut or molded as Type 1A tensile bars according to ISO 527-2:2012 or Type I bars according to ASTM D638-14. The immersion chambers are sealed stainless steel or fluoropolymer-lined vessels to avoid corrosion and plasticizer contamination. Headspace is limited to 10–15 % of vessel volume to reduce evaporative loss and oxygen availability. Exposure conditions for initial screening typically include 500 h at 60 °C in ASTM Reference Fuel C, followed by 1000 h at 40 °C in E10 for oxygenated service. At each interval, mass change is measured to ISO 175:2010, tensile strength retention to ISO 527-2:2012, and melt volume-flow rate shift to ISO 1133-1:2022. A mass increase above 1.5 % in fuel often signals plasticization sufficient to alter snap-fit retention and insert pull-out force.

Permeation is separate from immersion. For fuel system components, wall thickness and fiber orientation affect permeability. Published data for PBT in fuel sender flanges under SAE J2260 conditions is limited; if permeation is a gate criterion, a gravimetric cup test at 40 °C with E10 should be run in parallel with immersion.

What Limits Long-Term Exposure to Hot Coolant?

PBT undergoes hydrolysis by chain scission at the terephthalate ester linkage. The reaction is autocatalytic because carboxylic acid end groups reduce local pH and accelerate further chain scission. In hot ethylene glycol-water coolant, the glycol penetrates the amorphous phase, lowers the glass transition temperature, and increases the diffusion rate of water to the ester bond. Published data for unstabilized PBT in pure water at 100 °C show significant tensile strength loss within 500–1000 h; stabilizer packages and end-group capping shift the threshold but do not eliminate it. The apparent activation energy for hydrolysis varies with end-group concentration and coolant pH, so extrapolation from 60 °C to 100 °C is not linear and requires at least three time points.

Coolant is not a single fluid. Organic acid technology coolants buffer pH between 8.0 and 10.5, while silicated or hybrid coolants introduce silicate particulates that can abrade surfaces and create sites for capillary ingress. Chloride or sulfate contamination above 50 ppm can accelerate polyester hydrolysis. The screening medium should be the actual production coolant, mixed at the service concentration, typically 50 vol% ethylene glycol in deionized water. Immersion at 100 °C for 168 h provides a first-pass indication, but a full pre-trial verification should include 1000 h at 80 °C to detect slower hydrolysis. Dimensional change is measured to ISO 175:2010, water absorption to ISO 62:2008, and tensile strength retention to ISO 527-2:2012. Melt volume-flow rate at 250 °C and 2.16 kg according to ISO 1133-1:2022 frequently detects chain scission before a drop in tensile strength becomes statistically significant.

The diffusion-limited regime can be separated from the reaction-limited regime by plotting mass gain against retained tensile strength. In a 4 mm molded plaque, water reaches the core within approximately 24–48 h at 100 °C; longer exposure data therefore reflect hydrolytic chain scission rather than moisture uptake alone. For a 6 mm wall, the diffusion lag is roughly 2–4 times longer. This thickness dependence explains why thin-wall connector bodies can fail earlier than thick flanges in the same coolant circuit.

Oil Additive Packages and Stress Cracking in Molded Components

Unfilled and glass-reinforced PBT grades generally show low mass uptake in mineral oils. Absorption values below 0.5 % after 1000 h at 125 °C are typical for refined mineral oils. However, low mass uptake does not guarantee compatibility. Synthetic esters, extreme-pressure additives, zinc dialkyldithiophosphate decomposition products, and amine-based antioxidants can migrate into the amorphous phase and reduce local yield stress. Cracking occurs when the environmental stress exceeds the reduced threshold at molded-in strain concentrations. Weld lines, sharp radii, and fiber-rich surfaces are the most common initiation sites. The bent-strip method of ISO 22088-3:2017 at 1.0 % and 1.5 % strain is appropriate for ranking oils and PBT grades before molding trial parts. A component trial should be delayed if a parallel ESC coupon exhibits any crack deeper than 50 µm after 168 h at 80 °C in the actual oil.

Before immersion, the granulate is dried to a moisture content below 0.02 % using a desiccant dryer with a dew point of -30 °C or lower and an air inlet temperature of 120 °C for 4 h. Residual moisture above 0.03 % during melt processing hydrolyzes the resin and reduces the baseline mechanical properties that later become the reference for chemical ageing. Trials on injection molding machines should use a melt temperature of 250–270 °C for unfilled PBT and 260–290 °C for glass-fiber-reinforced compounds, with mold temperature controlled between 60 °C and 90 °C. Production-scale experience with hot-runner tools shows that dead spots, undersized gate drops, and excessive residence time above 300 °C generate yellowing, black specks, and a melt viscosity reduction that can be mistaken for chemical attack. A pre-trial melt volume-flow rate comparison on the same granulate processed in a cold-runner tool and the hot-runner production tool is therefore recorded before any fluid exposure begins.

A Pre-Trial Screening Matrix for Fuels, Oils, and Coolants

Table 1 condenses minimum screening conditions. The matrix is an initial gate, not a replacement for component-level functional testing under an OEM specification.

The conditions in Table 1 should be run in parallel with a fluid-free aged control at the same temperature. Without the control, thermal oxidation and post-molding crystallization are incorrectly assigned to the fluid. The control specimens are sealed in nitrogen-flushed foil laminate to exclude oxygen and moisture. The exposed specimens are normalized to the control value, and only then is a pass/fail decision made for trial entry.

Before the component trial, the molded plaques are checked against the following limits.

Retention Curves Do Not Decay Linearly

PBT does not lose mechanical strength as a linear function of immersion time. Chain scission reduces molar mass, but tensile strength remains near the initial value until the average chain length approaches the critical molar mass for entanglement. At that point, the load-bearing network deteriorates rapidly. This creates a property cliff edge: a part may pass a 500 h fuel immersion and fail abruptly at 750 h. The time-to-inflection is governed by water ingress rate, end-group concentration, temperature, and wall thickness. Because published data for specific PBT formulations in oxygenated fuels and OAT coolants is limited, the screening plan includes three sampling intervals rather than a single endpoint. The early marker is melt volume-flow rate: an increase above 25 % relative to the unexposed control at 250 °C and 2.16 kg indicates chain scission even when tensile strength retention remains above 90 %.

Surface cracking is rated under magnification of at least 20×. Crack length, crack density, and whitening are recorded separately because cracking is a surface phenomenon while tensile retention is a bulk measurement. A glass-fiber-reinforced PBT fuel flange that retains 85 % of baseline tensile strength may still fail a helium leak test if surface cracks connect the sealing bead to the fiber-rich weld line. This is the operational boundary that compatibility immersion cannot by itself close; it must be paired with a short geometry-specific pressure cycle using the actual component.

Approval to proceed to a component trial requires no crack deeper than 50 µm at 20×, mass change below the dimensional tolerance stack limit, and retained tensile strength of at least 80 % of the fluid-free aged control. If the production fuel contains methanol above 10 vol% or the coolant contains silicate or abrasive particulates, the immersion is repeated with that production fluid because published data for this specific configuration is limited. Machined specimens are not substituted for molded plaques, because the resin-rich skin removed by machining is a primary barrier to fluid ingress and its absence artificially accelerates crack initiation.