By PolyTakt Team · 9/9/2026
If the pellets are wet — or over-baked — the process window on the press is fiction. This article treats drying as four fundamentals (temperature, dew point, airflow, time), explains why −40 °C is a useful floor rather than a law, and contrasts PBT heat history with PET/PC hydrolysis — including when vacuum drying and setback actually help.
Scientific moulding is very good at fill, pack, and cooling.
It is quieter about the hopper.
If the pellets are wet — or over-baked — the process window you measured on the press is fiction. Viscosity has already moved. Impact strength may already be gone. Splay is the cosmetic version of a chemical problem that started before the screw turned.
Drying is not “blowing hot air on plastic.” It is a process variable with its own window, its own failure modes, and its own equipment choices.
−40 °C and −40 °F are the same temperature. That coincidence made −40 a memorable shop-floor target for desiccant process air.
Treat it as a useful floor for many engineering resins, not a law of polymer physics.
The table follows one rule. If a family has a range, the wet end is the left number and the dry end is the right. The plant target sits in the middle or at the dry end — never wetter than the wet end, and never outside the band. If a family has a single number, that number is the spec. The dryer target and the wet alarm are the same; if process air goes wetter than that setting, the pellets are out of spec.
That is why PET and PBT look different. PET’s window is −40 to −50 °C, so the target is −45 °C (middle). −40 °C is already the wet limit — a bed-change swing to −38 °C is under-dried PET. PBT’s window is −30 to −40 °C, so the target is −40 °C (dry end). A swing toward −35 °C is still inside the band; past −30 °C is wet. PC is listed as a single −40 °C: there is no published band to sit in the middle of, so −40 °C is both the setpoint and the line. Wetter than −40 °C is wet PC.
PP, PE, and PS are a different job. They hold surface moisture, not chemically bound water. A desiccant plant at −40 °C for commodity polyolefins is usually wasted energy. Hot air that actually reaches the pellets is enough on a dry day; on a wet day you are fighting condensation, not hydrolysis.
Datasheet first. These bands are typical industry practice, not a substitute for the grade. PolyTakt generic materials store the plant target column.
| Resin family | Process-air window (wet → dry) | Plant target | Why it is picky |
|---|---|---|---|
| PET / rPET | −40 to −50 °C | −45 °C (middle) | Hydrolysis; very low moisture spec |
| PC | −40 °C (no band) | −40 °C | Hydrolysis at melt; splay. Wetter than −40 °C is wet. |
| PA6 / PA66 | −35 to −45 °C | −40 °C (middle) | Strongly hygroscopic; viscosity and splay |
| PBT / POM | −30 to −40 °C | −40 °C (dry end) | Moisture and heat-history limits |
| ABS | −30 to −40 °C | −40 °C (dry end) | Mostly cosmetics and splay |
| PP / PE / PS | not a dew-point spec | — | Surface moisture only |
Chasing −60 or −80 °C is rarely the next quality win. Diminishing drying returns, rising energy, and — on some grades — a false sense that “drier air” can replace correct temperature, airflow, and time.
A dry pellet is the product of four variables. If one is wrong, dew point on a badge does not save you.
PolyTakt’s Drying Time and Dryer Load calculators estimate hopper residence against throughput. They assume the dryer is actually delivering temperature, dew point, and airflow. Instrumentation and a moisture check still win the argument.
Non-hygroscopic resins (PP, PE, PS, many polyolefins) pick up water on the pellet surface. Heat and moving air strip it.
Hygroscopic resins (PA, PET, PBT, PC, many TPU/TPE grades) absorb water into the polymer. You need dehumidified air or another method that actually lowers the vapour pressure around the pellet — not shop air on a wet Monday.
Two different problems. Two different machines. Using a hot-air hopper on nylon because “it feels dry” is how splay and brittle tensile bars get blamed on pack pressure.
PET is the obvious drying villain: wet pellets, hydrolysis in the melt, molecular weight gone. Polycarbonate is the same chemistry with a slightly more forgiving solid-state story.
PBT (polybutylene terephthalate) is the other common polyester on the floor. It is hygroscopic. It usually dries around 120 °C. So shops copy the PET fear — chase a drier dew point, leave the hopper hot “just in case.”
That is the wrong fear, which is why this article keeps naming PBT. It is the resin that separates two clocks.
Clock 1 is moisture. PBT’s process-air window is −30 to −40 °C. The plant target is −40 °C — the dry end of that band, same number as in the materials library. A hopper leak that moves you to −35 °C is still inside the window; past −30 °C is wet. That is not the PET story, where −40 °C is already the wet end. POM sits on a similar band. ABS is the same −30 to −40 °C window with the same dry-end target. None of those families need the PET-style −45 °C middle.
Clock 2 is heat history. PBT grades often carry heat stabilisers, flame retardants, and impact modifiers that do not enjoy sitting at 120–130 °C all afternoon. Extended or repeated drying can deplete that package: colour shift, brittleness, and a material that no longer matches the datasheet even though the moisture analyser is happy. Other heat-stabilised or flame-retardant engineering grades can fail the same way. PBT is the shop-floor case that shows up often enough to teach the rule.
A practical floor method for many PBT jobs:
That is dryer setback: keep the sponge, pause the oven. Confirm the holding temperature with the grade owner; 95 °C is an illustration, not a universal SOP.
PC is usually more tolerant of repeated solid-state drying. You can often re-dry PC that sat in an opened bag. You cannot un-hydrolyse PC that was melted wet. Water at 280 °C+ cuts chains. Re-drying the next bag does not repair the last shot.
Under-drying PET or PC is catastrophic chemistry. Over-holding PBT at full temperature is a quieter, equally real scrap mode. The correct fear depends on the resin.
Equipment exists because resins and cells have different constraints: moisture type, energy, space, changeover speed, and how long material can sit at temperature.
Ambient air, heated, through the hopper, exhausted to the room. Cheap. Fine for surface moisture. On a humid day it can make hygroscopic pellets wetter.
Two beds of molecular sieve. One dries process air; the other regenerates at high temperature; they swap. This is the classic way shops hit around −40 °C. The catch is dew-point swing at changeover, valve wear, and the energy spike of regenerating a full tower.
A continuous wheel moving through process, regeneration, and cooling zones. The usual reason to buy one: a flatter dew-point line and less energy than heating an entire spare tower. Same physics as twin-tower — dehumidified air as the sponge — with steadier delivery.
Preheat the pellets, then pull a deep vacuum. Lowering pressure lowers the boiling point of water, so moisture flashes out of the pellet in tens of minutes instead of several hours. There is no desiccant bed to regenerate and typically far less mass sitting at drying temperature.
That last point is why vacuum drying shows up in the PBT case. A conventional hopper sized for 3–4 hours of residence is a thermal-history warehouse if the tool crashes. A vacuum system sized for a short cycle holds much less inventory at heat. Startup and material changes are faster for the same reason. The same inventory argument applies to other additive-sensitive grades; PBT is the example, not a vacuum-only resin.
Trade-offs are real: higher capital, batch rather than simple mass-flow, and the need for level control so the press never sees a gap. Follow the OEM. Energy-saving percentages on brochures are not a process spec.
Drying does not pause just because the cycle time went to infinity.
On a desiccant hopper, material keeps cooking at setpoint until someone setbacks, empties, or scraps it. Hours of PBT at 120 °C is how additive depletion shows up on Monday’s tensile bars.
On a vacuum dryer, controllers typically watch consumption. If weight stops dropping, preheat and vacuum heat should go to a holding or sleep state. Material sitting in a sealed, unheated vacuum chamber is not accumulating the same heat history as a full-temperature desiccant bed — but it is still subject to the grade’s maximum dry time and your SOP. Do not treat “indefinitely” as a licence to ignore the datasheet.
If you do have to dump the system after a long crash, hopper mass is the scrap ticket. A 4-hour desiccant inventory is a different write-off from a ~30–45 minute vacuum inventory. Size is not a moral argument for either technology; it is a number you should know before the tool fails.
Modern desiccant dryers can setback too. Vacuum is not the only way to stop cooking PBT — it is the way that also shrinks the mass at risk and the wait at start-up.
If fill and pack look textbook and the parts are still splayed or brittle, look upstream of the feed throat.
