"PA6 absorbs 9% water" — the number everyone quotes is read wrong

In every second discussion about polyamide 6 the same figure comes up: "PA6 picks up 9% water." On that basis the grade gets struck off the project before the first trial part. The problem is that 9% is not working water absorption — it is saturation under full immersion, the worst state physically possible. In a real shop and on a real part the numbers are different, and the quantity that actually matters to the design engineer is not a percentage at all. This article explains where the 9% comes from, why one TDS holds three different "water absorption" values, and how not to lose a perfectly workable grade such as Examid® PA6 GF30 because of it.

Every figure below is given either with a reference to the test method or with the note "typical for the class — verify against the grade TDS". We do not assign specific curves to Material Wizard grades where they must be taken from the batch certificate. This is a piece about how to read a number, not about our price list. A broader choice of "dry" families is collected in the engineering polyamides hub; the focus here is narrower: what the notorious 9% in PA6 actually means.

Where the 9% comes from and why it is not a working figure

The 9% figure is real. It is the mass of water that unfilled polyamide 6 will take up if kept immersed in water until its mass stops rising. That is the saturated state — a physical ceiling, not an operating point. Aliphatic polyamides are hygroscopic because of the chain chemistry itself: water molecules are attracted to the amide groups. How many attach depends on how much water is available around them. Under full immersion practically all of it is available; in air at 50% humidity, several times less.

So "9%" answers a question most parts never ask. A connector in an instrument housing, a gear in a reducer, a clip under the bonnet all work in air of varying humidity, not under water. What matters for them is not the saturation ceiling but equilibrium with the surrounding air. For PA6 that equilibrium is about 2.5–3.0% by mass at 23 °C and 50% relative humidity. The difference between 9% and 3% is neither an error nor marketing: it is the difference between two different questions answered by one and the same grade.

Three different "water absorptions" in one data sheet

When an engineer opens a TDS and sees the line "Water absorption", he almost always sees not one number but several — and that is exactly where the confusion starts. Under ISO 62 (the base method for determining water absorption of plastics) a producer may report at least three different values, each matching its own scenario.

The first number is water absorption after 24 hours of immersion. The specimen is placed in water for a day and weighed. For PA6 this is roughly 1.5–1.9%. It is a short check, convenient for quality control, but it says nothing about equilibrium or about service.

The second is saturation under immersion, the very same 9–9.5%. The specimen is kept in water until its mass stabilises. This is the physical ceiling, and it is precisely this value that gets misquoted as "working water absorption".

The third, and the most important one for the design engineer, is equilibrium moisture in the standard atmosphere of 23 °C / 50% RH, about 2.5–3.0%. The catch is that reaching this equilibrium naturally on a thick specimen can take many months. Hence a separate standard exists — ISO 1110, accelerated conditioning of polyamides: specimens are held at 70 °C and 62% humidity until they have taken up at least 95% of their equilibrium moisture, and the resulting value is close to equilibrium at 23/50. When a data sheet says "conditioned", it almost always means this state, not immersion.

The conclusion is simple: before comparing the "water absorption" of two grades, make sure both numbers were measured in the same state. Otherwise you are comparing one grade's day of immersion with another's saturation — and making the wrong choice.

"Dry as moulded" versus "conditioned": the comparison trap

The same split of states haunts not only the water absorption line but the whole mechanical section. A serious TDS reports mechanical properties in at least two states: dry as moulded (DAM — straight out of the mould, almost without moisture) and conditioned (conditioned to equilibrium at 23/50). For polyamide the difference is fundamental, because water acts as a plasticiser: it increases chain mobility, and the material becomes less stiff and more ductile.

For glass fiber reinforced PA6 GF30 it looks like this: tensile modulus in the dry state is roughly 9,500–11,000 MPa, while on a conditioned specimen it falls to approximately 6,000–6,500 MPa (typical for the class — verify against the TDS). In other words, moisture can take away about a third of the declared stiffness. If the designer took the "dry" number from the data sheet while the part will run in humid air, he has overestimated the stiffness of the assembly by tens of percent.

Hence the key practical rule: grades may only be compared in the same state. If DAM is taken for one grade and conditioned for another, the comparison is void — and that is exactly how conclusions such as "PA6 is useless" are born. The right approach is to take the conditioned row for both candidates, because it is closer to real service, or at least to keep both honestly in DAM and remember that the part will be worse.

What water actually changes: dimensions and modulus, not "percentages"

The percentage of absorbed water breaks nothing on its own. What breaks things are its two consequences, and those are what must go into the calculation.

The first is dimensional change. Absorbed water pushes the chains apart and the part swells. For polyamides the linear swelling coefficient is roughly 0.2–0.3% for every 1% of moisture taken up (typical for the class). For unfilled PA6 moving from the dry state to equilibrium at 50% RH this gives a linear change of the order of 0.4–0.8% — for a gear or a connector with tolerances in hundredths of a millimetre that is a lot. This is why the "dry" choice is made not where percentages are feared, but where the part has to hold its dimensions.

The second is the loss of stiffness and strength discussed above: roughly 20–35% of modulus and strength lost on transition to the saturated state. Both consequences describe a direction of causality that is stable across the whole class of aliphatic polyamides; the specific magnitude depends on the grade, the filler and the geometry, so it is taken from the TDS rather than from thin air.

Values are typical for the class; verify against the TDS of the specific grade and the test state.

When PA6 GF30 wins, and when it is time to move to "dry" grades

The right selection logic starts not with the water absorption percentage but with the question: in what state will the part work, and what is its tolerance.

• PA6 GF30 is enough when the part tolerance is set for the conditioned state (not for the "dry" data sheet), the environment is air of moderate humidity rather than prolonged immersion, and what matters most is stiffness and price rather than absolute dimensional stability.
• PA6 GF30 is enough when the part is large and glass fiber reinforced: the fiber skeleton damps the swelling, and the economics of the production run weigh more than tenths of a percent of moisture content.

• It is time to move to a "dry" grade when the part is precision and works in water or condensate permanently — here even GF30 may fall outside tolerance.
• It is time to move when stability of electrical or mechanical properties in humid heat is critical — then look towards Examid® PA12 CF30 (water absorption several times lower than PA6) or Examid® PA610 CF30 (a bio-based polyamide — a compromise between dryness and price).

The logic here is not "PA6 is bad, PA12 is good" but "different chemistry — a different price for dryness". PA12 and PA610 pay for low water absorption with a longer carbon chain and higher feedstock cost. They are worth choosing when the calculation on conditioned PA6 GF30 genuinely fails on tolerance, not because someone was frightened by the number 9%. We covered the chemistry of the long-chain bio-based polyamide separately — see Polyamide 610 (PA610): a bio-based polyamide from castor oil.

Expert review: 5 questions worth asking

1. Why does one TDS give PA6 water absorption both as 9% and as 3% — is that a producer's error? No, these are two different states of the same material. 9–9.5% is saturation under full immersion (the physical ceiling); 2.5–3.0% is equilibrium with air at 23 °C / 50% RH. The producer gives both because they answer different scenarios: immersion and service in air. The error appears when the reader takes the saturation ceiling for the working figure.

2. If the part never contacts water, can water absorption be ignored entirely? No. Polyamide takes moisture from air, not only from liquid water. Even in a dry room at 30–50% relative humidity PA6 will in time reach equilibrium at about 2–3% and change its dimensions and stiffness accordingly. It can be ignored only for a hermetically encapsulated part; for everything else you calculate equilibrium with air.

3. PA6 GF30 takes less water than unfilled PA6 — where exactly does that difference come from? Glass fiber is not hygroscopic, so in GF30 only the polymer part absorbs water — about 70% of the mass. Water absorption by mass therefore falls roughly in proportion to the resin fraction. In addition the rigid fiber skeleton mechanically restrains swelling, so the dimensional effect of moisture is suppressed more strongly than the water percentage alone.

4. Why can't two grades simply be compared by the "water absorption" line in their data sheets? Because the number depends on the method and state: 24-hour immersion, saturation or equilibrium at 23/50 are different quantities for the same material. If one condition was taken for one grade and another for the second, the comparison is void. First bring both grades to the same state, then compare.

5. When is a switch from PA6 GF30 to PA12 or PA610 genuinely justified, and when is it over-insurance? It is justified when the calculation of the part on conditioned PA6 GF30 fails on tolerance, or on property stability in humid heat or prolonged immersion. It is over-insurance when the decision is taken because of the number 9% without checking the working state: in many applications GF30 with a tolerance set for conditioned holds its geometry, and the more expensive "dry" grades are taken only against a real need, not out of fear.

Summary

The 9% is real, but it is the saturation ceiling under immersion, not working water absorption. The part works in air, where PA6 equilibrium is about 2.5–3.0%, and the quantity critical for the designer is not a percentage at all but two consequences: dimensional change and the modulus drop on a conditioned specimen. Compare grades only in the same state, set tolerances for the working humidity — and PA6 GF30 remains a workable choice where it is hastily struck off because of one attractive number from the first line of the data sheet.

Examid® PA6 GF30 — The baseline glass fiber reinforced grade for series mouldingCompare on the conditioned row · the GF skeleton damps swelling · set tolerance for the working stateRequest supply terms →

Examid® PA12 CF30 — Long-chain polyamide when dry stability is requiredWater absorption several times lower than PA6 · for humid environments and precisionRequest supply terms →

Examid® PA610 CF30 — Bio-based polyamide: a compromise between dryness and priceDrinks less than PA6, cheaper than PA12 · a compromise for humid assembliesRequest supply terms →

Material Wizard supplies Examid® engineering polyamides — from the baseline PA6 GF30 to the "dry" PA12 and PA610 — provides a TDS and a CoA per batch and supports grade selection for the tolerance class of your part: with dry/conditioned rows, a dimensional change calculation and a specimen conditioning plan, rather than a single water absorption figure. The company is located in Derazhnia and Kharkiv. The material is available with delivery across Ukraine — to verify the water absorption and tolerance of your part, please consult our specialist.