🧪 PEEK
Turning PEEK: High-Performance Polymer on the Lathe Done Right
PEEK turns like a dream compared to the superalloys it often replaces, but treating it like just another plastic is how parts crack, warp, and miss tolerance. This is a high-performance thermoplastic with a glass transition near 290°F, real internal stress, and, in its filled grades, abrasive reinforcement that wears tools, so turning PEEK well is about heat management, stress control, and matching tooling to the grade.
ISO 13485ISO 9001AS9100
PEEK (polyetheretherketone) is a semi-crystalline engineering thermoplastic with excellent strength, chemical resistance, and a continuous service temperature around 480°F, which is why it substitutes for metal in demanding applications. On the lathe it machines readily with sharp tooling at high spindle speeds, much faster and with far lower cutting forces than any metal, but its low thermal conductivity means heat stays local at the cut. If the tool dulls or you dwell, frictional heat can take the surface above the glass transition (around 290°F) and cause gumming, melting, or a poor smeared finish.
The technique is sharp, polished tooling with positive rake (often the same geometry used for aluminum), high surface speeds, light-to-moderate feeds, and effective chip evacuation so hot chips do not re-weld to the part. Coolant or at least air blast helps carry heat away; many shops machine PEEK with air or a light mist rather than flood, and unfilled PEEK can be cut dry for light work if heat is managed.
Sharp edges matter more than with metal because a dull tool generates heat through rubbing rather than cutting, and PEEK is unforgiving of that heat. The same toughness and temperature resistance that make PEEK valuable mean it does not simply shave off cleanly when the edge degrades; it heats, gums, and tears.
Internal stress and the annealing question
The most common cause of out-of-tolerance and cracked PEEK parts is residual stress, and it is the issue least familiar to people coming from metal machining. PEEK stock, especially larger-diameter extruded or molded rod, carries internal stresses from how it was produced, and machining relieves those stresses asymmetrically, causing the part to warp, bow, or change dimension after it comes off the lathe. For tight-tolerance parts this is the dominant problem.
The fix is annealing. Stock is often annealed before machining, and tight-tolerance parts are commonly annealed again between roughing and finishing: rough the part leaving stock, anneal it through a controlled heating-and-slow-cooling cycle (specific to PEEK, reaching elevated temperature and cooling slowly) to relieve the stresses introduced by roughing, then finish to final dimension on the now-stable part. Skipping this on a precision PEEK part is a frequent and expensive mistake.
This annealing requirement is a real cost and lead-time factor unique to high-performance polymers like PEEK. It adds oven cycles and handling. For loose-tolerance parts you may machine in one pass without intermediate annealing, but for the tight-tolerance medical, aerospace, and semiconductor parts PEEK is usually chosen for, plan on a rough, anneal, finish sequence and discuss it with the supplier, because the dimensional stability of the finished part depends on it.
Unfilled, glass-filled, and carbon-filled grades
Unfilled (virgin) PEEK is the toughest and most ductile grade, with the best elongation and impact resistance, and it is the grade required for most medical and implantable applications (PEEK-OPTIMA and similar medical grades) and for parts needing the cleanest, most chemically pure surface. It machines the most easily of the three because there is no abrasive filler, sharp tools last well and finishes are good.
Glass-filled PEEK (typically 30% glass fiber) adds stiffness, dimensional stability, and reduced thermal expansion, used for structural and high-temperature mechanical parts. The glass fibers are abrasive and accelerate tool wear significantly, so you move to wear-resistant tooling and accept shorter tool life; finishes can also show fiber at the surface. Carbon-filled PEEK (typically 30% carbon fiber) gives even higher stiffness and strength, lower thermal expansion, plus electrical conductivity and better wear properties for bearing and structural parts. Carbon fiber is likewise abrasive on tooling, and the fine carbon dust requires good extraction.
The practical machining consequence is that filler choice drives tooling and cost as much as it drives part properties. Unfilled PEEK is the friendliest to machine and the medical default; the filled grades deliver stiffness and stability that unfilled cannot, at the price of abrasive tool wear. PCD or diamond-coated tooling is often used on the filled grades for production work to combat the abrasion, just as you would for glass- or carbon-filled composites.
Tolerances, finish, and why PEEK is worth the cost
Turned PEEK holds tight tolerances, ±0.001 in is achievable, but with caveats metal does not impose. PEEK's thermal expansion is high (several times that of steel, around 26 µin/in/°F for unfilled), so tolerance must be specified and inspected at a known temperature, and filled grades expand less, one reason to choose them for dimensional stability. Combined with the residual-stress and annealing issues above, holding tight tolerances on PEEK requires both thermal awareness and proper stress relief, not just a precise lathe.
Surface finish is good with sharp tooling, unfilled PEEK can reach fine finishes suitable for sealing and medical surfaces, while filled grades show fiber and are slightly rougher. Burrs are soft and easily removed but do form, and the heat-sensitive nature means a dull tool degrades finish quickly.
The cost reality: PEEK is an expensive material, far more than common engineering plastics, and the medical and aerospace grades with certification and traceability cost more still, plus the annealing and careful machining add process cost. But PEEK is chosen precisely because it does things no cheaper plastic can: continuous use near 480°F, excellent chemical resistance, biocompatibility, low outgassing for vacuum and semiconductor use, and a strength-to-weight that lets it replace metal. When those properties are required, PEEK's machining cost is justified; when they are not, a cheaper engineering plastic like PTFE, Delrin, or Ultem may do the job, and over-specifying PEEK is a common and expensive mistake.
Frequently Asked Questions
Almost always residual stress, the issue most likely to surprise someone coming from metal machining. PEEK stock, particularly larger extruded or molded rod, carries internal stresses from how it was manufactured. When you machine it, you remove material asymmetrically and relieve those stresses unevenly, so the part bows, warps, or changes dimension after it leaves the lathe, sometimes hours later. For tight-tolerance parts this is the dominant failure mode. The fix is annealing: machine-grade PEEK stock is often pre-annealed, and for precision parts you anneal again between roughing and finishing. The sequence is rough the part leaving finishing stock, run a controlled PEEK annealing cycle (elevated temperature with slow cooling) to relieve the stresses that roughing introduced, then finish to final dimension on the now-stable part. Skipping this on a precision part is a frequent and costly mistake. PEEK's high thermal expansion (around 26 µin/in/°F for unfilled, several times steel) compounds the problem, so also specify and inspect tolerances at a known temperature. Discuss the rough-anneal-finish sequence with your supplier up front, because the dimensional stability of the finished part depends on it.
It depends on whether you need toughness, stiffness, or specific properties. Unfilled (virgin) PEEK is the toughest and most ductile, with the best impact resistance and elongation, and it is required for most medical and implantable applications (medical grades like PEEK-OPTIMA) and for parts needing the cleanest, chemically purest surface. It also machines the most easily because there is no abrasive filler, so tools last and finishes are good. Glass-filled PEEK (typically 30% glass fiber) adds stiffness, dimensional stability, and lower thermal expansion for structural and high-temperature mechanical parts, but the glass fibers are abrasive and significantly increase tool wear. Carbon-filled PEEK (typically 30% carbon fiber) gives even higher stiffness and strength, the lowest thermal expansion, electrical conductivity, and good wear properties for bearings and structural parts, with the same abrasive tool-wear penalty and fine conductive dust to manage. So choose unfilled for medical, impact, and purity; glass-filled for stiffness and stability; carbon-filled for maximum stiffness, wear, and conductivity. Remember the filled grades wear tooling much faster, often requiring PCD or diamond-coated tools for production, which affects cost and lead time.
No, in most respects it is far easier than the metals it replaces, with much lower cutting forces and higher achievable speeds, but it has its own pitfalls that catch people who treat it like an ordinary plastic. The main challenge is heat: PEEK has low thermal conductivity, so heat stays local at the cut, and if the tool dulls or dwells, frictional heat can push the surface above the glass transition near 290°F, causing gumming, melting, and a smeared finish. The defenses are sharp polished tooling with positive rake (similar geometry to aluminum), high spindle speeds, light-to-moderate feeds, good chip evacuation so hot chips do not re-weld, and air or mist cooling to carry heat away. The second challenge is residual stress, which causes warping unless the part is annealed, rough, anneal, finish for tight tolerances. The third is filler abrasion: glass- and carbon-filled grades wear tools quickly and often need PCD tooling. So PEEK machines readily and quickly with sharp tools and good heat management, but a dull tool, ignored internal stress, or the wrong tooling on a filled grade will produce cracked, warped, or smeared parts. Respect the heat and the stress and it turns beautifully.
PEEK is the wrong choice whenever you do not actually need its high-end properties, because it is expensive, far more than common engineering plastics, and the medical and aerospace grades plus the required annealing and careful machining add even more cost. PEEK earns its price only when you genuinely need continuous service near 480°F, excellent chemical resistance, biocompatibility for implants, low outgassing for vacuum and semiconductor use, or a strength-to-weight that lets it replace metal. If your part runs at moderate temperatures and does not need those, cheaper plastics do the job: Delrin (acetal) for general precision mechanical parts with good machinability and low cost, PTFE for chemical resistance and low friction at lower strength, nylon for tough general parts, or Ultem (PEI) for high-temperature needs that are demanding but below PEEK's level at lower cost. Over-specifying PEEK is a common and expensive mistake, sometimes a part is drawn in PEEK simply because it sounds high-performance when an acetal part would perform identically for a fraction of the cost. A good supplier will flag this and suggest the cheaper material when the application does not require PEEK's specific advantages.
Last updated: July 2026
Find PEEK Turning Suppliers
Search verified shops that handle PEEK turning.
No logins. No email gates. Just results.