How to Design Molds for High-Temperature Resins: PPS, PPA, PEEK & PEI

A PEEK connector mold came back from T2 with a hairline crack across the cavity face and a core that had lost 3 HRC. The tool had run 4,000 cycles at 200°C mold temperature against a 390°C melt. The steel was standard H13, tempered at 540°C, which looked safe on paper. It was not. The cavity surface cycled between 200°C and 380°C on every shot. After a few thousand cycles, the thermal stress exceeded the tempered yield strength of the steel. The mold cost $48,000. The rework, re-heat-treatment and lost production time cost more.

High-temperature resins punish molds that were designed for ABS or nylon 6. PPS processes at 300-330°C melt and 80-130°C mold temperature. PPA runs 320-345°C melt. PEI (Ultem) needs 380-420°C melt with mold surfaces at 120-160°C. PEEK demands 380-400°C melt and mold temperatures up to 200°C for maximum crystallinity. These numbers change every mold-side decision: steel grade and temper, heating system, ejection geometry, hot runner sealing, venting and expected tool life.

This guide covers the mold design rules for all four resin families. The focus is tooling, not processing. Melt and mold temperatures appear only as inputs to mold decisions. For material behavior itself, see our PEEK material guide and PPS material guide.

The Snapshot

• Tempering temperature must sit at least 50°C above the maximum working temperature of the steel, or the cavity surface softens in production. H13 at 46-52 HRC needs a temper above 590°C to survive a 200°C mold plus flash-heated surface layers.
• Mold surface uniformity target is ±5°C across the cavity. Oil temperature controllers hold ±2-3°C. Cartridge heaters drift ±10-15°C unless zoned and PID-controlled.
• PPS and PEEK parts drift 0.2-0.5% dimensionally during post-crystallization in the first hours to days after molding. Mold dimensions must target the annealed state, not the as-molded state.
• Draft angles of 1-2° per side are typical for high-temperature crystalline resins, versus 0.5-1° for amorphous materials, because shrinkage onto cores raises ejection force.
• Hot runners for PEEK and PEI need melt-tight seals rated above 400°C and thermal growth compensation of 0.3-0.6 mm per 100°C of manifold temperature rise.
• Expect a 30-60% mold cost premium over a comparable room-temperature-resin tool, driven by hot-work steels, heating circuits, hardened components and longer tryout cycles.

Melt Temperature vs Steel: The Thermal Fatigue Problem

The cavity surface of a high-temperature mold lives a violent thermal life. During injection, a 380°C PEEK melt touches a 200°C steel face. The surface layer of the steel flash-heats by tens of degrees in under a second. During cooling and ejection, it drops back. Each cycle imposes a compressive stress at peak temperature and a tensile stress at the trough. Steel fails by thermal fatigue long before it fails by static overload.

The 50°C temper rule.Tempered martensitic steels lose hardness when the working temperature approaches their original tempering temperature. The rule of thumb: the steel's tempering temperature must be at least 50°C above the maximum temperature the steel will see in service. A mold running at 200°C bulk temperature with surface excursions to 250°C+ needs steel tempered at 590-620°C. Standard P20, tempered around 590°C, sits at the edge. H13 tempered in the 540-590°C range at 46-52 HRC is the common choice for PPS and PPA tools. For PEEK production above 180°C mold temperature, specify H13 tempered at 600-620°C and accept 42-46 HRC. Better still, move to a hot-work grade like 1.2367 (H11-type) with higher tempering resistance.

Hardness and temper trade-off.Higher hardness resists wear from glass fiber but accelerates thermal fatigue cracking, because harder steel is more brittle under cyclic thermal stress. The typical working window:

• 46-48 HRC: better thermal fatigue resistance, preferred for PEEK and PEI cavities with mold temperatures above 160°C.
• 48-52 HRC: the compromise range for PPS and PPA tools at 80-130°C mold temperature.
• 52+ HRC: acceptable only for small inserts below 120°C mold temperature, never for large cavity faces in high-temperature service.

DieStrike holds heat treatment to HRC 62 for wear-critical components such as small cores and inserts. Cavity blocks for high-temperature service are tempered into the ranges above. Heat treatment specifications deserve their own review, covered in How to Specify Heat Treatment for Mold Steel.

Thermal shock at start-up.Most thermal fatigue cracks nucleate during the first heating cycles, not during steady production. A cold mold at 25°C that jumps to 200°C in 20 minutes creates a surface-to-core gradient of over 80°C, and the surface yields in compression. Three design rules apply. Ramp rates below 5°C per minute for large blocks. Heating circuits no closer than 25 mm from the cavity face in hot-work steels. Preheating before the first shot. This is a mold-side decision, not a press-side one, because the circuit layout is frozen at tool design.

Mold Heating Systems: Oil vs Cartridge Heaters

Room-temperature molds remove heat. High-temperature molds must add heat continuously, because a 150-200°C tool loses energy to the platens, the ambient air and the cooler ejector side. There are two proven systems, and the choice drives mold structure.

Oil temperature controllers.Pressurized hot oil units run to 200°C (standard) or 350°C (pressurized high-temperature units). They deliver ±2-3°C stability at the mold inlet and circulate through drilled channels, so heat enters over a large surface. For PPS at 80-130°C and PPA at 90-120°C mold temperature, oil circuits in standard drilled channels are the default. Design rules: channels 9-11 mm in diameter, placed 25-40 mm from the cavity surface. Keep flow velocity above 1 m/s to stay turbulent. Use separate circuits per cavity in multi-cavity tools. Balanced oil flow keeps cavity-to-cavity temperature variation inside ±5°C.

Cartridge heaters.Electric heating rods reach 400°C+ and are required for PEI and PEEK tools that hold 150-200°C mold temperature with tight local control. But raw cartridge layouts drift ±10-15°C across the block. Holding ±5°C uniformity takes discipline. Use 6-8 mm cartridges at 30-40 mm pitch. Keep the heater-to-cavity distance at 20-30 mm. Zone the mold into blocks of no more than 100 x 150 mm per PID controller. Place thermocouples halfway between adjacent cartridges, 10-15 mm from the cavity face. Beryllium-copper inserts around cartridges improve heat spreading by 3-5 times versus direct drilled holes in tool steel.

Insulation and structural growth.A 200°C mold grows. Steel expands roughly 12 µm per meter per °C. A 500 mm mold plate at 200°C is therefore about 2 mm larger in each direction than at 20°C. Leader pins, guide bushings and shut-off faces must be sized for the hot state. Fit clearances designed at room temperature will bind or leak flash at production temperature. Insulation plates between mold and platen cut conductive losses by 60-80% and stabilize oil temperature. Specify ceramic or mineral-fiber plates rated 3-5 mm thick with compression strength above 30 MPa so clamping does not crush them.

Shrinkage, Post-Crystallization and Ejection

PPS and PEEK are semi-crystalline, and crystallization does not stop at ejection. A PEEK part molded at 200°C mold temperature keeps increasing crystallinity for hours to days. Service heat or an anneal at 200°C for 2-4 hours accelerates the drift. The dimension moves again: 0.2-0.5% additional shrinkage is typical for PPS and PEEK parts between as-molded and fully crystallized states. A 50 mm feature can grow out of tolerance by 0.1-0.25 mm after the CMM report says pass.

Design rules for dimensional drift.First, get the molding window fixed before cutting steel: mold temperature swings of ±20°C change crystallinity by 5-10 percentage points and shrinkage by 0.1%+. Second, cut the cavity to the annealed dimension, not the as-molded dimension, and record the expected as-molded size separately. Third, for tolerance-critical features at ±0.05 mm or tighter, plan a post-mold anneal with fixtures that hold the part flat. DieStrike works to ±0.005 mm on mold dimensions, which controls the tool side of this budget. The resin side must be managed in the design.

Shrink-onto-core ejection force.Crystalline resins shrink harder onto cores than amorphous ones. PPS shrinks 1.0-1.7% and PEEK 1.2-2.0%. Glass fill lowers shrinkage along the flow and raises it across the flow. The result is a part that grips cores with significant force at ejection temperature. At 150-200°C mold temperature, ejection happens while the steel and part are both hot and the polymer is weaker, but the grip is higher. Two rules follow:

• Draft angles: 1-2° per side minimum on cores and cavity walls for PPS and PEEK, versus 0.5-1° typical for amorphous resins. Add 0.5° per 10% glass fiber loading for textured or matte surfaces.
• Ejection area: size ejector pins for 1.5-2 times the ejection force of a room-temperature mold. Use 20-30% more pins than instinct suggests. Place them on ribs and bosses, never on thin unsupported walls that are soft at 150°C+ ejection temperature.

Because parts eject hot and soft, ejector pin marks show more easily. Sleeves and blades distribute force over more area than round pins. Keep pin land lengths short, 3-5 mm, so the pins do not hang up on a shrinking part. Polish ejection faces to Ra 0.2 µm or better to cut sticking friction.

Hot Runner Rules Above 350°C

High-temperature resins need hot runners that hold 340-400°C melt temperature continuously, not intermittently. Standard hot runner systems are rated to 320°C. Beyond that, heater bands fail early, seal rings extrude, and manifold growth breaks alignment with the nozzles. Specify systems rated for engineering resins with documented maximum melt temperature of 400°C or higher.

Thermal growth compensation.A hot half at 380°C grows against a mold at 150°C. Across a 400 mm manifold, differential expansion reaches 1.2-1.5 mm versus room temperature. Two proven architectures handle it. Floating manifolds use radial expansion gaps of 0.3-0.5 mm with center-locating pads. Fixed manifolds use flexible nozzle tips that absorb axial growth. Rigid stack-ups with cold-fit dimensions are a guaranteed leak path. The nozzle-to-cavity seal face is the failure point. Molten PEEK at 390°C finding a 0.05 mm gap freezes into a seal-breaking shim. The leak grows every cycle.

Valve gate versus open gate.Valve-gated nozzles give a clean gate vestige and allow sequenced filling for weld-line control. The catch is the valve pin. It runs through a hot seal area and must tolerate mold temperature without galling. Pin materials need 50+ HRC with high-temperature coatings. Standard H13 pins at 46-50 HRC gall against PEEK-filled manifolds within weeks. Valve pin clearance is critical. Too tight and the pin binds as everything grows. Too loose and melt creeps past the shutoff. Typical hot-state pin clearance runs 0.03-0.05 mm. Open-gate (thermal or sprue-bushing) nozzles avoid moving parts entirely and are the robust default for PPS and PPA below 345°C. For PEI and PEEK above 380°C, choose open gates unless cosmetic gate quality forces valve gating, and then budget for pin maintenance every 200,000-500,000 shots.

Heater and thermocouple density.Maintaining 380-400°C through the nozzle requires 2-3 times the heater watt density of a 230°C ABS system, roughly 6-10 W/cm² on the nozzle body. Thermocouples must sit within 5 mm of the melt channel, not on the manifold surface, or control lag causes 15-30°C melt swings. Every swing shows up as viscosity variation and dimensional drift in the parts.

Steel Selection for High-Temperature Molds

Steel choice splits along two axes: thermal load and corrosion. High mold temperatures favor hot-work steels with high tempering resistance. Many high-temperature resins, especially flame-retardant PPS and PPA grades, release corrosive decomposition products that attack carbon steels, favoring stainless grades. The table shows the common selections.

← swipe to scroll →

Table note: Hardness ranges and temperature limits are typical industry figures for injection mold service. Final specification depends on part geometry, resin grade and cycle profile.

Corrosion versus abrasion.Flame-retardant PPS and PPA release halogen and phosphorus compounds that pit carbon steels within 100,000 shots. Glass fiber at 30-60% loading polishes steel away at gates and flow leaders. When both attack together, stainless matrix with a hard coating, or H13 with corrosion-resistant nitriding, outperforms either defense alone. Note the trade-off. S136 resists corrosion but conducts heat slightly worse than H13, roughly 24 vs 26 W/m·K at similar hardness. Slower heat extraction can add 2-5 seconds to cycle time in thick sections.

Polishability.PEEK and PEI parts often need glossy or optical surfaces. S136 and 420 stainless polish to SPI A1 (Ra 0.025-0.05 µm). H13 reaches A2-A3 at best because its carbides pull out during polishing. If the print calls for SPI A1-B1, the cavity material decision is made for you. Compensate for H13's thermal advantages with better cooling design.

Gating High-Temperature, High-Viscosity Melts

PEEK at 390°C still runs a melt viscosity of 150-400 Pa·s (higher for glass-filled grades), and PEI at 380°C behaves similarly. These melts freeze fast against cooler steel and lose pressure through small gates. Gate design rules for high-temperature tooling:

• Gate cross-section: size 25-50% larger than you would for an equivalent nylon part. Edge gates 0.8-1.2 mm thick by 3-6 mm wide are typical for PPS and PPA. PEEK needs 1.0-1.5 mm thickness in unfilled grades and more in 30% glass grades.
• Gate freeze: high mold temperatures keep gates open longer. At 150-200°C mold temperature, a 1.0 mm gate may not freeze for 4-8 seconds, which extends hold time and cycle. Where cycle matters, use a hot runner so the gate never freezes, or accept the packing time in the schedule.
• Gate land: keep it short, 0.5-1.0 mm, because the high-viscosity melt generates shear heat and stress across the land. Long lands in PEEK tools cause freeze-offs and short shots.
• Gate location: avoid direct impingement on cores and pins. A 390°C jet erodes H13 core surfaces over hundreds of thousands of cycles, and the erosion shows as dimensional growth at the impingement point.

For multi-cavity high-temperature tools, gate balance matters more than usual because viscosity swings steeply with temperature: a 10°C melt difference changes PEEK viscosity by 5-8%. Cavities fed by nozzles running 10°C apart fill at measurably different rates. Verify nozzle-to-nozzle temperature spread during tryout with a pyrometer or melt-pressure trace, not just controller readouts.

Venting Decomposition Gases

PEEK begins to decompose above roughly 400-420°C if it stagnates in hot zones, and processing at 380-400°C leaves little margin. Decomposition releases small amounts of phenol, benzene and other volatiles. PPS and PPA release moisture and, in flame-retardant grades, corrosive gases. Mold-side venting rules:

• Vent depth: 0.015-0.025 mm for PPS and PPA, 0.01-0.02 mm for PEEK and PEI. High-temperature melts flash less readily than low-viscosity nylons. Still, keep vent lands short at 1-3 mm, followed by 0.5-1.0 mm relief pockets.
• Vent area: target 20-30% of the parting-line perimeter vented on PEEK tools. Trapped gas compresses and burns. The diesel effect chars the polymer instantly near 390°C and leaves black burn marks that no downstream process removes.
• Gas exhaust routing: decomposition gases are corrosive and toxic. Route vents away from sliding components and ejector pin bores, where gas condensate corrodes fits and causes sticking. Add vent cleaning to the maintenance schedule every 20,000-50,000 shots, because carbonized residue clogs 0.015 mm lands quickly.

Vent depth standards and placement logic for all resin families are covered in Mold Venting: Depth, Location and Design Rules.

Mold Life and Thermal Fatigue Inspection

A well-designed PPS mold in H13 at 48-50 HRC can run 500,000-1,000,000 shots before cavity surfaces need attention. A PEEK mold at 200°C typically reaches its first thermal fatigue checkpoints earlier, around 200,000-400,000 shots, because the temperature swing per cycle is larger. Life is not one number. It is a maintenance schedule with defined inspection points.

Thermal fatigue crack inspection.Heat checking appears as a network of fine surface cracks. Look at cavity faces, gate areas and core tips, where melt impingement and temperature swings concentrate. Inspection rules:

• At 100,000 shots: visual and dye-penetrant inspection of gate areas and core tips. Cracks below 0.05 mm wide can still be polished out, removing 0.02-0.05 mm of material.
• At 250,000 shots: repeat dye penetrant on all cavity faces. Measure any dimensional change at polished areas against the original CMM report, because polishing changes cavity dimensions by the amount removed.
• Every 50,000 shots after that: check sliding components. At 150-200°C mold temperature, standard greases carbonize and lifters and slides gall. Use high-temperature solid-film lubricants rated 300°C+ or self-lubricating bushings.

Hardness verification.Include a Rockwell check of cavity blocks at each major service interval. A drop of more than 2 HRC from the as-tempered value means the steel is over-tempering in service. The tool has reached its material limit. Re-tempering is not an option. Plan a new insert or cavity block. This check takes minutes on the bench and prevents the surprise failure that scrapes a production run.

Heating system service.Cartridge heaters in 200°C service have finite life, typically 5,000-15,000 hours. Resistance-test every heater annually or every 10,000 hours: a reading more than 10% off nominal means replacement before failure. Oil channels coke internally above 180-200°C oil temperature. Flow-test circuits annually and flush when flow drops below 80% of the commissioned value.

What High-Temperature Tooling Costs

Expect a 30-60% premium over an equivalent mold built for nylon or ABS. The premium is not arbitrary. It breaks down into specific line items, and knowing where the money goes helps you decide what to spend and what to save.

• Steel and heat treatment: 10-20% of the premium. Hot-work grades like 1.2367 cost 30-50% more per kilogram than H13, and double or triple tempering cycles add furnace time.
• Heating system: 10-20%. Cartridge heaters, thermocouples, PID zones and insulation plates are absent from cold molds. A four-zone heated tool with beryllium-copper spreading inserts is a real budget line.
• Hot runner: 15-25%. High-temperature-rated systems with 400°C heaters and premium seals cost 30-60% more than standard 320°C systems.
• Tryout and validation: 10-20%. Heating uniformity mapping takes hours per mold. Temperature profiling with embedded thermocouples or infrared checks is standard practice on PEEK tools. Each profile run adds machine time at 150-200°C mold temperature, which is slow and expensive per shot.
• Components: 5-10%. High-temperature ejector pins, coated slides, high-temp springs and seals all carry premiums over standard catalog parts.

What does not save money: downgrading steel on a PEEK mold. A P20 or under-tempered H13 cavity in 180°C service fails within 100,000 shots. The rework plus the stopped production line exceeds the premium several times over. Where budgets are tight, cut the right things. Drop cavity count from 4 to 2. Choose an open-gate hot runner instead of valve gating. Accept SPI B2 instead of A1. Never cut steel grade or temper specification.

FAQ

Q1. Can H13 handle PEEK molding temperatures?

Yes, if specified correctly. H13 tempered at 600-620°C (42-46 HRC) survives 200°C mold temperature plus surface excursions. The common failure is H13 tempered at 540°C for hardness, which over-tempers in service. Keep the 50°C temper rule: tempering temperature must exceed the maximum steel working temperature by at least 50°C.

Q2. Why do my PPS parts measure in tolerance at T1 but drift out later?

Post-crystallization. Semi-crystalline PPS and PEEK keep crystallizing after ejection, adding 0.2-0.5% shrinkage over hours to days. Measure annealed parts and cut the cavity to the annealed dimension. Lock the mold temperature window to ±5°C. Crystallinity, and therefore final size, then stays repeatable.

Q3. Oil heating or cartridge heaters for a 180°C mold?

Both work. Pressurized oil to 200°C is cheaper to install and gives ±2-3°C stability over large areas. Cartridge heaters are required above 200°C mold temperature and give better local control. To hold ±5°C uniformity they need 6-8 mm cartridges at 30-40 mm pitch, zoned PID control, and thermocouples 10-15 mm from the cavity. Hybrid layouts, oil for bulk and cartridges for hot spots, are common on PEI and PEEK tools.

Q4. What draft angle do I need for glass-filled PPS?

Start at 1.5-2° per side for 30-40% glass-filled PPS, versus 1° minimum unfilled. Glass fibers lock into textured surfaces and shrink onto cores harder than amorphous resins. On deep cores over 50 mm, add 0.5° and verify ejection force at tryout while the tool is hot.

Q5. How often should a PEEK mold be inspected for thermal fatigue?

Dye-penetrant check gate areas and core tips at 100,000 shots, all cavity faces at 250,000, then every 50,000 shots with hardness verification on cavity blocks. Replace high-temp cartridge heaters every 5,000-15,000 hours or when resistance drifts more than 10% from nominal.

The Bottom Line

High-temperature molds fail from the same three causes: under-tempered steel, uneven heating and ignored post-crystallization drift. Design the temper first, specify ±5°C heating uniformity, size gates and vents for 380°C+ melts, and inspect on a shot-count schedule. DieStrike builds IATF 16949 precision molds with ±0.005 mm dimensional accuracy and 24-hour DFM feedback for connector OEMs including TE Connectivity, Amphenol and Luxshare. Bring your PPS, PPA, PEI or PEEK part print for a tooling review at our contact page.