Under-bonnet polymers: how temperature picks the plastic

The hardest environment on the car

An engine bay combines everything that destroys a polymer: sustained heat, thermal cycling, oil, coolant, fuel, road salt and vibration, for fifteen years. Yet intake manifolds, coolant components, connectors, covers and housings are all routinely plastic, because the weight and part-consolidation gains are large and because the right polymer genuinely survives.

Choosing that polymer is mostly a temperature question, and the way engineers answer it is more subtle than reading one number off a datasheet.

Two transitions, and which one is the ceiling

Every polymer has a glass transition Tg, where the amorphous chains gain enough mobility to move past each other and the material softens from glassy to rubbery. Semi-crystalline polymers additionally have a melting point Tm, where the ordered crystalline regions finally melt.

For an amorphous polymer — polycarbonate, PMMA, PEI, polystyrene, rigid PVC — there are no crystallites to hold anything together above Tg. So Tg is effectively the ceiling, and the useful service temperature sits below it.

For a semi-crystalline polymer — PA66, PA6, PP, PET, PEEK, HDPE — the picture is different and much more useful. Above Tg the amorphous fraction softens, but the crystallites are still solid and act as physical crosslinks, tying the structure together. The material loses stiffness but keeps working. That is why a semi-crystalline polymer can serve far above its glass transition, and it is the key to the whole under-bonnet material set.

The example that makes it click: PA66

PA66 is the classic under-bonnet workhorse, used for intake manifolds and countless housings. Our database gives it a glass transition of 60 °C and a melting point of 262 °C, at 3.2 GPa and 1.14 g/cm³.

Reported continuous-use temperatures for PA66 in air are roughly 120–140 °C, with glass-filled grades used in environments that exceed 150 °C for shorter excursions. Look at where that sits: it is more than twice the glass transition, and less than half the melting point.

Both halves of that sentence matter. It works above Tg because the 262 °C crystallites are holding the structure. It is limited well below Tm because what actually kills it is not melting — it is thermo-oxidative degradation, the polymer slowly reacting with oxygen and breaking its own chains, plus creep under sustained load at temperature. Melting is almost never the failure mode. Chemistry and time are.

Compare polycarbonate, with a Tg of 147 °C in our data — nearly two and a half times PA66's. Naively it looks like the better hot part. But it is amorphous, so 147 °C is a hard ceiling rather than a waypoint, and PC is also notoriously vulnerable to the hydrocarbons and stress-cracking agents that an engine bay is full of. PA66 wins the application despite losing the Tg comparison.

The ladder

In rough order of increasing capability and cost, and using our own Tg/Tm data alongside published continuous-use figures: polypropylene (Tg −10 °C, Tm 165 °C) handles ducting, splash shields and low-temperature housings; PA6 and PA66 (Tm 220 and 262 °C, continuous use roughly 120–140 °C) cover the bulk of structural under-bonnet parts; PPS reaches continuous service around 200–220 °C; and PEEK (Tg 143 °C, Tm 343 °C in our data) is quoted at roughly 250–260 °C continuous, at a price that keeps it to parts that genuinely need it.

PEEK is the clearest illustration of the semi-crystalline argument. Its glass transition of 143 °C is essentially the same as polycarbonate's 147 °C, yet it serves continuously at around 250 °C where PC cannot approach that. The difference is entirely the 343 °C crystalline phase holding the structure together above Tg.

Reinforcement shifts the whole ladder upward. Glass fibre raises stiffness retention above Tg substantially, which is why under-bonnet parts are usually a filled grade — PA66-GF30 rather than neat PA66 — and why a neat-resin datasheet understates what the actual part does.

What the datasheet will not tell you

Continuous-use temperature is a long-duration rating, not a peak. A part may briefly see far more than its rating during heat soak after shutdown, when airflow stops and under-bonnet temperature rises rather than falls. Peak and continuous are different specifications and both need checking.

Time matters as much as temperature. Thermo-oxidative degradation is cumulative, so a rating always implies a duration — a temperature that is fine for 1000 hours may not be fine for 15 years.

Load matters too. Creep accelerates sharply with temperature, so a bolted joint that is fine cold can relax and lose preload hot. Our viscoelasticity article works through exactly this: the effective modulus after sustained load can be a fraction of the datasheet value.

And chemistry is often the real limit. Coolant, oil and fuel exposure at temperature can matter more than the temperature alone, which is why material selection here always ends in fluid-compatibility testing rather than a table lookup.

A note on the numbers here

The Tg, Tm, modulus and density values are from this site's materials database, each with its source, test method and conditions — you can open any of them and check. They are typical values for neat polymers; filled and modified grades differ substantially, and a supplier datasheet always overrides a textbook figure.

The continuous-use temperature figures for PA66, PPS and PEEK come from published supplier and industry material-selection literature. They are strongly grade-dependent and are quoted here to illustrate the reasoning, not as specifications.