Mon–Sat 08:30–18:00 (GMT+8) Founded 2009 Dongguan, China

PEEK at cryogenic temperature: what changes below -100 C

PEEK is one of the few engineering plastics that survives liquid nitrogen service. The failure modes below -100 C are not the ones designers expect.

Material Selection 8 min read
PEEK at cryogenic temperature: what changes below -100 C

PEEK appears in cryogenic equipment for a simple reason: it is one of very few engineering thermoplastics that does not become glass at liquid nitrogen temperature. That does not mean it behaves like the PEEK in the datasheet. Below about -100 C, three properties change in ways that matter to a design, and one of them changes in the direction most designers do not expect.

This article covers what happens to impact behaviour, thermal contraction and sealing between -100 C and -196 C, how PEEK compares with PCTFE in the applications where both are specified, and where the material should not be used at all.

Why -100 C is the number to watch

PEEK is semi-crystalline with a glass transition around 143 C, far above any cryogenic service, which is why the amorphous phases are already frozen at room temperature and do not embrittle further on cooling. What does change around -100 C is the mobility of small-scale segments within the amorphous regions: a secondary relaxation that is active at room temperature progressively shuts down, and the material loses the small amount of internal damping that absorbed energy in impact.

In practice this means impact strength falls noticeably as the part passes through that window, then levels off. The fall is not a cliff. Unnotched PEEK remains tough at -196 C compared with a glass filled grade, which is brittle long before it reaches cryogenic temperature. Notched impact is where the difference shows, and a sharp machined corner at -196 C behaves very differently from the same corner at 20 C.

Impact: the design rule changes, the material does not

The mistake in cryogenic design is to treat low temperature as a strength problem. Tensile and compressive strength of PEEK actually rise as it cools, because the polymer chains have less thermal energy to move. The problem is energy absorption, not load capacity.

  • Remove notches. A radius that is generous at room temperature is marginal at -196 C. Every internal corner in a cryogenic PEEK part should be checked against the wall thickness, not against a general fillet standard.
  • Avoid press fits that generate hoop stress. A shrink fit designed for room temperature puts a permanent tensile stress into the part, and that stress is what makes a cold impact fail. Where a fit is needed, design for a controlled clearance and retain mechanically.
  • Do not use glass filled grades for cold impact. GF30 is brittle at room temperature in notched impact, and cooling does not improve it. Unfilled or carbon filled grades are the only sensible options below -100 C.
  • Support the part. A thin unsupported span that survives handling at 20 C can fail under its own contraction at -196 C if one end is restrained.

Contraction: the arithmetic that decides clearances

PEEK contracts on cooling, and it contracts more than steel but less than PTFE. Over the range from 20 C to -196 C, a PEEK part loses roughly one to one and a half percent of its length. A 200 mm dimension becomes about 197 to 198 mm.

That figure alone is manageable. The problem is that the mating metal contracts by a different amount, so the clearance between them changes. Stainless steel over the same range contracts by about 0.3 percent, aluminium by about 0.4 percent. A PEEK bore around a stainless shaft therefore closes as the assembly cools, because the plastic shrinks several times as much as the metal.

Material Contraction, 20 C to -196 C Consequence for a PEEK-to-metal fit
PEEK, unfilled 1.0 to 1.5 percent Reference for the plastic side
Stainless steel 316 about 0.3 percent PEEK bore closes on the shaft by roughly 0.7 to 1.2 percent of diameter
Aluminium about 0.4 percent Slightly less closure, but aluminium also yields earlier
PTFE 2.5 to 3.5 percent Why PTFE seals need a spring or a gland that follows the contraction

The practical rule: calculate the fit at the cold temperature, not at 20 C, and add the clearance needed for the contraction differential. A running clearance that is correct at room temperature will be an interference fit in liquid nitrogen.

Sealing: where PEEK is the right answer and where it is not

Cryogenic sealing splits into two jobs, and PEEK only does one of them well.

Static seals that must be leak-tight at -196 C are usually PCTFE, not PEEK. PCTFE stays ductile at cryogenic temperature and conforms to the sealing face, which is exactly the property a static seal needs. PEEK is stiffer, does not conform as readily, and needs higher seating stress to seal. A PEEK gasket in a liquid nitrogen flange is usually a compromise made for strength or chemical resistance, not for leak rate.

Dynamic and structural parts — valve seats, bushings, wear rings, compressor piston rings, pump bearing pads — are where PEEK earns its place. Low temperature raises its strength and hardness, dimensional stability is good, and it does not cold-flow the way PTFE does under a sustained load. The limiting factor is impact and clearance, not sealability.

The common cryogenic valve design uses PCTFE for the seat sealing element and PEEK for the guide, the bearing or the structural insert. That combination exists because each material is doing the job it is actually good at.

Hydrogen and the newer cryogenic applications

Liquid hydrogen at -253 C is a step further, and it brings a second problem: small-molecule permeation. PEEK has relatively low permeability for a thermoplastic, and it is used as a liner and as an insulator support in hydrogen systems, but pressure cycling with hydrogen can produce blistering at high pressure and rapid decompression. Where a part sees fast pressure drops in a hydrogen environment, the grade and the wall thickness need to be qualified by test rather than selected from a table.

The same caution applies to LNG at -162 C: the temperature is within PEEK’s capability, and the design issue is almost always the thermal gradient across the part during cool-down, not the steady-state temperature.

Machining and inspection for cold service

A cryogenic part usually fails for a reason set before it ever gets cold.

  • Anneal before final machining, and again after roughing. Residual stress from machining is locked into the part; when it contracts at -196 C, that stress adds to the thermal strain and a crack opens from a machined corner. Stress-relieved stock is not optional for thin sections.
  • Measure at the temperature the drawing specifies. A 0.05 mm tolerance measured at 23 C means something different at -196 C, and a supplier who inspects warm and ships to a cold drawing is handing over a dimensional argument.
  • Do not ream a tight bore expecting it to stay round. Boring a PEEK part that will run cold should leave a clearance calculated for contraction, and the bore should be checked for roundness after a thermal cycle, not only after machining.
  • Watch moisture. PEEK absorbs a small amount of water, and a part that goes from a humid workshop into a cryogenic line can develop surface frost and, in a closed volume, ice expansion. Dry the part before assembly.

When PEEK is the wrong material

Choose something else when the requirement is a leak-tight static seal at the lowest temperature, when the part will take impact loads while cold, or when the assembly cannot tolerate a differential contraction of one percent against its metal housing. In those cases PCTFE, modified PTFE or a metal seal is the correct answer, and specifying PEEK because it is the high-performance option produces an expensive part that fails at the acceptance test.

Questions from cryogenic projects

Can PEEK be used in liquid nitrogen? Yes, for structural, guiding and wear parts. It is not the first choice for a static seal at that temperature.

Does PEEK get brittle at -196 C? It loses impact energy, particularly in notched sections, but it does not become glass-brittle like a glass filled grade. The design rules for corners and fits matter more than the headline impact number.

Should I use carbon filled PEEK for cold wear parts? Often yes, because carbon keeps some toughness and adds thermal conductivity, which shortens cool-down gradients. Glass filled is the grade to avoid.

What clearance should I leave for a cold bore? Start from the contraction differential against the mating metal, add the running clearance you need warm, and verify by a thermal cycle test on a first article. Send us the shaft material and the cold temperature and we will return the starting dimensions.

Top Plastic supplies unfilled, carbon filled and bearing grade PEEK in stress-relieved sheet and rod, machined and annealed for cryogenic service, with dimensional reports measured at the temperature your drawing specifies.

Need a material recommendation?

Send us the application and we will reply with two or three candidate grades and a price.

Ask an engineer
Share this article

Send us your drawing, get a price in 24 hours

Tell us the material, the size and how many pieces. If you are not sure about the material, describe the application and we will recommend one.

Request a quote Browse stock shapes