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

PEEK sheet from 1 mm to 100 mm: how thickness changes the part

Thickness is not a scale factor in PEEK. Cooling rate sets crystallinity, crystallinity sets shrinkage, and shrinkage sets what the part does after you machine it.

Material Selection 9 min read
PEEK sheet from 1 mm to 100 mm: how thickness changes the part

Buyers treat PEEK sheet thickness as a number on a stock list: 1 mm, 5 mm, 20 mm, 60 mm, all the same polymer, just more of it. In a machine shop that assumption fails within the first cut. A 2 mm sheet and a 60 mm plate differ in crystallinity, in how much stress is locked inside them, and in what happens to a part when one face is milled away.

This article explains where that difference comes from, what it changes in practice between 1 mm and 100 mm, and how to specify and machine each thickness band so the finished part matches the drawing.

Thickness is really a cooling rate

PEEK is semi-crystalline, melting around 343 C and settling at 30 to 35 percent crystallinity in normal semi-finished stock. How much of it crystallises depends on how long the material spends in the temperature window between the melt and about 200 C, and that time is set by section thickness.

Heat leaves a plastic slab from its surfaces, and PEEK conducts heat poorly, roughly 0.3 W/mK. A 1 mm sheet can lose its heat through both faces in seconds. A 100 mm plate needs hours, and its core stays near the melt temperature long after the surface is solid. The same polymer therefore leaves the press with two different structures:

  • Thin sections cool fast. Chains do not have time to organise, crystallinity lands at the low end, and the surface can be noticeably less crystalline than the core. Fast-cooled PEEK is tougher, slightly more ductile, lower in modulus and higher in thermal expansion.
  • Thick sections cool slowly. Crystallinity is high and uniform, spherulites are coarser, the material is stiffer and harder, and it has already shrunk most of what it is going to shrink.

The practical consequence is that a 1 mm sheet and a 60 mm plate are not interchangeable at equal stress, and a thin part has a dimensional behaviour over temperature that a thick part does not share.

Skin, core and the stress that lives between them

In a thick plate the skin solidifies first while the core is still molten and free to move. When the core finally crystallises it wants to occupy less volume than the melt it came from, but the skin is already rigid and holds it. The result is a locked-in pattern: the core is in tension along the plate, the skin in compression, and the two balance only as long as the plate keeps its original section.

Machine one face and the balance is gone. The remaining stressed layer pulls the plate into a bow, usually concave on the machined side. The thicker the plate and the more material removed, the larger the movement. Thin sheet barely shows the effect because it had little gradient to begin with.

This is why stress-relieved stock exists, and why it is not a marketing detail. Annealing before machining converts the locked pattern into dimensional change at the annealing temperature instead of at the customer’s inspection bench. For heavy cuts the sequence is rough, anneal, finish, and for parts with tight flatness it is anneal both before and between.

Thickness band How it usually cools Dominant risk What to do about it
0.5 to 3 mm Fast, through both faces Warpage and fixturing, not internal stress Full support, vacuum or tape, light depth of cut, no re-clamping mid-feature
3 to 12 mm Moderate Stress shows when more than a third of the section is removed Balance cuts between faces, anneal if removal is heavy
12 to 40 mm Slow at the core Bowing after single-face milling, thickness tolerance drift Rough, anneal, finish; measure after stabilisation
40 to 100 mm Very slow, hours Core crystallinity, shrinkage, coarse structure Buy near-net, allow 3 to 5 mm per face, anneal twice

Machining 1 to 3 mm sheet without turning it into a potato chip

Thin PEEK does not fail from internal stress, it fails from clamping. The forces that hold a 1 mm sheet flat are larger than the forces of the cut, and releasing them at the end lets the part spring.

  • Support the whole area. A vacuum table with a sacrificial medium-density board, or double-sided tape on a flat plate, beats clamps every time. Clamps concentrate load at the edge and lift the middle.
  • Cut with sharp, two-flute carbide and climb milling. PEEK is abrasive in filled grades and gummy in unfilled grades; a dull tool pushes the sheet instead of cutting it.
  • Keep depth of cut small and speed high. Thin sections deflect under axial force. A 0.5 mm depth per pass at high spindle speed removes material without lifting the sheet.
  • Never machine half a profile, unclamp and re-clamp. Every re-fixture is a new stress state. Plan the operation so the part is cut in one setup, or leave tabs thick enough to survive the second one.
  • Let it sit. A thin sheet that has just been machined on one face can move for hours as it equalises moisture and temperature. Measure after 24 hours, not after a coffee.

Below about 0.5 mm the product is normally film rather than sheet, made by extrusion or skiving, and the rules change again: it is supplied on a carrier, it wrinkles rather than bows, and flatness claims should be agreed in writing before the order.

12 to 40 mm: the band where things go wrong quietly

Mid-thickness plate is where most PEEK parts are made and where most dimensional arguments start, because the material is thick enough to carry stress and thin enough that a buyer expects tight tolerances without paying for stress relief.

Two habits prevent most of it. First, remove material symmetrically: if a pocket takes 10 mm out of one face, take 10 mm off the other, or expect the plate to curve. Second, allow for the fact that a 30 mm plate is not 30.00 mm. Semi-finished PEEK plate is supplied to a thickness tolerance, not to a grindable flat, and a part that must be 29.95 mm finished should start from 32 mm stock, not 30 mm.

Above 40 mm: buy the shape, not the block

At the top of the range, the economics flip. Removing 80 percent of a 60 mm plate to reach a 12 mm part wastes material that was expensive and releases stress that then has to be annealed anyway. Cast or pressed near-net shapes, or a tube where the geometry allows it, usually cost less than the block plus the machining hours and the scrap.

Thick stock also needs a longer anneal. Residual stress in a 60 mm section sits deeper, and a short cycle only relaxes the outer layer. A workable cycle ramps at 2 to 3 C per minute to about 200 C, holds one hour per 10 mm of section, and cools at the same slow rate; rushing the cool-down re-introduces exactly the gradient the anneal was meant to remove.

Stiffness scales with the cube, stress does not

Designers who normally work in metal underestimate how fast PEEK sheet loses rigidity as it thins. Bending stiffness goes with the cube of thickness: 2 mm sheet is eight times as stiff as 1 mm, and 3 mm is twenty-seven times. A 1 mm PEEK panel is a wear liner, a liner for a chute, an electrical barrier or a shim, not a structural plate.

The reverse also holds. A part that works at 20 mm will not simply work at 12 mm with the same load, and the deflection that appears is often read as a material problem when it is a thickness decision.

Specifying thickness on a drawing

Three lines on the drawing prevent most confusion. State the nominal thickness and the tolerance band, not a single value. State whether the tolerance applies to the supplied stock or to the finished part after machining. And state whether the material must be stress relieved, and at what point in the process.

For thin sheet the useful tolerance is typically plus or minus 0.1 to 0.15 mm on nominal, and flatness should be quoted as a number over a defined length rather than as a word. For plate the thickness tolerance widens with section, which is why a 60 mm plate is not ordered as 60.00 mm.

Questions we get about PEEK sheet thickness

Is 1 mm PEEK sheet available from stock? Yes, in unfilled and some filled grades, usually 500 x 1000 or 610 x 1220 mm. Natural and black are the common colours; other grades are made to order.

Does a thicker sheet hold a tighter tolerance? No, the opposite. Thickness tolerance widens with section, and flatness is harder to hold because the material has more stress to release.

Will thin PEEK sheet stay flat at 200 C? Only if it has been annealed. Fast-cooled thin sheet can crystallise further in service at elevated temperature and change dimension. Ask for annealed stock when the part runs hot.

Can I machine a 1 mm pocket in a 2 mm sheet? You can, but the remaining 1 mm floor will deflect and the wall will taper. For floors under about 1.5 mm, support the part from below with a sacrificial backing and take the last passes with a small radial engagement.

Top Plastic supplies PEEK sheet from 0.5 mm to 100 mm, unfilled, glass filled, carbon filled and bearing grades, with stress-relieved options and cutting to size. Send the drawing with the thickness band and the flatness requirement and we will tell you which stock thickness and which annealing sequence the part actually needs.

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