How to Prevent Rust and Corrosion in Injection Molds

A P20 core insert with 0.35 percent chromium ran 14 months on untreated cooling water. The return line came off the press with a pitted band 0.9 mm deep. Nothing leaked, so the tool stayed in service while four cavities ran a 12 °C hot spot. Cycle time grew from 22 to 31 seconds and the plant scrapped 340 parts before a technician pulled a hose.

The sections below cover the water specification, the steel and coating choices, the polymer chemistry and the shutdown routine that stop it.

The Snapshot

• Wall loss of 0.5 to 1.5 mm per year. Untreated water pits a P20 circuit at that rate, 1.2344 at 5 percent Cr lasts 2 to 4 times longer, and 13.6 percent Cr stainless still pits above 200 ppm chloride.
• Water at 200 to 500 microsiemens. A closed loop belongs at pH 7.5 to 9.5 with chloride under 25 ppm for stainless channels, and a 0.5 mm scale layer costs 5 to 10 percent of the cooling rate.
• Nickel at 25 to 50 micron. Electroless nickel protects a channel from the inside, hard chrome at 10 to 25 micron handles PVC and flame-retardant cores, and both beat welding a pitted water line.
• Dry air at 4 to 6 bar. Blow every circuit until the exhaust reads under 10 percent RH, film polished faces at 5 to 20 micron, and store at 40 to 60 percent RH in VCI film with 10 to 20 g of desiccant per 0.1 m3.

Where Mold Corrosion Starts

Corrosion in an injection mold concentrates on two surfaces. The first is the wetted steel inside every cooling circuit, and the second is any face that never dries. Both are hidden from the operator, and both fail before anything visible appears on a part.

A 12 mm channel drilled through a 400 x 300 mm plate exposes 0.5 to 1.2 m2 of steel to whatever the plant supplies. That surface sits behind a hose nipple, a baffle or a plug, so it is inspected only when a mold is stripped for other reasons.

Chromium content sets the rate. P20 at 0.3 to 0.4 percent Cr pits in untreated water inside a single warm season. Hot-work 1.2344 at 5 percent Cr lasts 2 to 4 times longer on the same water. Stainless S136 or 420 at 13.6 percent Cr holds at neutral pH but is attacked once chloride passes 200 ppm.

The failure signature is a temperature drift rather than a leak. Wall loss of 0.5 to 1.5 mm per year thins the channel, cuts flow and pushes the cavity surface 5 to 12 °C off setpoint. A leak is the last event in a sequence that started 10,000 shots earlier.

Two further sites corrode on the same tool. Steam pockets form where flow drops to a crawl. Condensate wets the parting line of a mold that sits cold between shifts, and grease in a pin hole holds water against steel for days.

The Water Specification That Stops Pitting

Corrosion control is a water problem before it is a steel problem. A loop that stays inside six limits keeps a P20 tool alive for a decade. A loop that drifts outside them pits 13.6 percent Cr stainless in a single season.

Conductivity belongs at 200 to 500 microsiemens per centimetre, chloride under 25 ppm for stainless channels, and pH between 7.5 and 9.5. Hardness above 150 ppm as calcium carbonate lays down scale that no biocide removes. Suspended iron settles in low-velocity zones and undercuts the steel it touches.

← swipe to scroll →

Table note: targets are typical values for a closed-loop mold temperature control system. Confirm the actual readings at the manifold with a calibrated meter before writing them into a tool specification.

Glycol earns its place only where a plant faces a hard freeze. A 30 percent glycol mix cuts heat transfer 5 to 15 percent. A tool that ran 8 °C above setpoint on water runs 10 to 14 °C above on the mix. Never top up a glycol loop with untreated tap water.

Sample the loop quarterly and log conductivity, chloride, pH and hardness in the tool file. A chloride reading that climbs from 12 to 40 ppm over two quarters means the makeup water changed upstream of the mold.

How Steel Grade Changes Corrosion Resistance

Grade selection decides how much water chemistry a tool can tolerate. P20 at 0.3 to 0.4 percent Cr is the cheapest cavity steel on the market and the most sensitive to chloride. Hot-work 1.2344 at 5 percent Cr buys 2 to 4 times the time to first pit. Stainless at 13.6 percent Cr buys years, until chloride crosses 25 ppm.

Two costs come with a stainless cavity. The steel carries a premium over P20, and its lower thermal conductivity adds 3 to 8 seconds to warm-up on a cold start. Those seconds land once per shift, against a pitting rate that can retire an insert in a single season.

← swipe to scroll →

Table note: resistance bands are typical industry values for the listed chromium contents in a closed loop. Confirm with a coupon test in the actual plant water before releasing a steel choice for a high-volume tool.

Match the grade to the water, not to the price list. A pitted 1.2344 core is cheaper to run than a P20 core that needs a new insert every 18 months. A plated channel is cheaper than either when the loop cannot be cleaned up. The selection rules for resin-driven corrosion are in mold steel for corrosive resins.

Hardness matters in the same decision. A cavity at HRC 30 peens at a shut-off land and opens a gap that water exploits. Production inserts in the 1.2083 family run at HRC 50 to 54 and hold a seal land against a scratch that would start a pit.

Coatings and Anodes That Protect Mold Steel

Protective coatings buy time on the features that water and resin attack first. Four options cover most tools, and each one carries a geometry limit that decides where it can be applied.

Electroless nickel at 25 to 50 micron is the only coating that reaches inside a cooling channel. The process deposits without an electric field, so a 6 to 12 mm channel coats evenly from the inlet to the deepest return leg. At 25 micron the internal diameter shrinks 0.05 mm in total, which sits inside any flow allowance. Blind pockets under 5 mm wide and open threads remain uncoated and belong outside a protected circuit.

Hard chrome at 10 to 25 micron belongs on cores and cavity faces that see PVC, POM or flame-retardant grades. It holds at 190 to 210 °C and can be stripped and re-plated at 300,000 to 500,000 shots. It flakes outward from a pinhole, so the surface must be polished clean of every scratch first. Plating over a pit 0.2 mm deep leaves a defect that fails inside 100,000 shots.

Nitriding builds a compound layer of 0.05 to 0.12 mm over a diffusion zone that must exceed 0.2 mm. Gate areas and wear faces benefit most, because the surface dries between cycles and avoids standing water. A shallow diffusion zone cracks under the 150 to 250 °C cycling of a production tool, so specify zone depth as well as surface hardness.

Sacrificial anodes cover the manifold. A zinc plug on a cast water manifold dissolves in place of the steel around it and costs a few minutes to replace each year. Stainless quick couplings and brass-free fittings remove the couple between the plate and the hose circuit. The comparison of layer hardness and failure modes sits in nitriding and PVD coatings, and the plating thickness and adhesion tests sit in electroless nickel plating.

Polymer Chemistry That Attacks Mold Steel

Not every mold corrodes from its water side. Three resin families attack steel from the cavity, and each one leaves a residue that keeps working after the press stops.

PVC processed at 190 to 210 °C releases hydrogen chloride. That gas plus a trace of moisture forms hydrochloric acid, which etches an unplated 1.2344 cavity overnight. Vent the mold every 40 to 60 mm and run the melt at the low end of the window. Put a nitrided or chrome-plated surface on the cavity and the cores, with a 0.5 mm vent at 40 mm pitch to move gas out of the parting line.

Flame-retardant grades with brominated additives release hydrogen bromide, and antimony trioxide carries the attack into the hot runner. A P20 cavity running a brominated grade on two shifts can pit in 3 to 6 months. The same cavity in 1.2083 at 13.6 percent Cr usually survives the campaign with a wash and a film.

Moisture-carrying resins hydrolyse. PA, PC and PET processed above their moisture limit drop acidic residue into the parting line and the vents. Dry PA66 below 0.05 percent moisture and PC below 0.02 percent. Clean the cavity at the end of the run, not the next morning. A weekend of acidic residue plus 60 percent RH is when a parting line etches.

Acetal and POM release formaldehyde at melt temperature and leave a film that holds water against the steel. It attacks less aggressively than hydrogen chloride but stays active over a shutdown. Wipe the cavity, blow the vents and spray a 5 to 20 micron film whenever a tool will sit idle for more than 72 hours.

Galvanic Corrosion Between Fittings and Plates

A mold cooling circuit behaves like a battery when it mixes metals. Brass fittings threaded into a stainless manifold sit 0.2 to 0.4 V apart on the galvanic series, and the water closes the circuit.

The damage shows as a ring of pits around the threaded port rather than across the whole plate. Wall loss at that joint runs 0.1 to 0.3 mm per year, several times the rate on the open channel. The cathode area is large and the anode area is small. A steel nipple in an aluminium manifold gives the same pattern over a wider band.

Three habits break the couple. Keep the manifold and every fitting inside one alloy family, preferably 304 or 316 stainless. Use stainless or polymer quick couplings in place of brass. Add an insulating bushing or a polymer nipple wherever a dissimilar fitting cannot be avoided.

Stray current does the same damage from outside the loop. A welding ground on a mold base drives current through the water. A heater circuit with a leaking relay does the same, and both pit the channel where it exits the plate. Check the loop after any repair that involved welding on or near the press.

Rust on Parting Lines and Polished Faces

Flash rust does not need a wet mold. A tool at 60 to 80 °C entering a 25 °C room at 60 percent RH condenses a film within 30 to 90 minutes.

That film etches on its own. A 0.5 to 2 micron rust bloom ruins an SPI A1 finish polished to Ra 0.012 to 0.025 micron. The pit reads as a speck on an optical part, and no polishing pass restores a pit 20 micron deep without changing the geometry around it.

Handling carries the same risk. Skin deposits 0.1 to 1 mg of chloride per square centimetre of contact. That fingerprint etches steel at 50 percent RH inside 24 to 72 hours. Wear nitrile gloves every time a polished face is touched, and never set a cavity insert down on a bare bench or a concrete floor.

Blind features stay wet longest. Screw holes, lifter pockets, ejector pin bores and slideways hold water for hours after the visible faces read dry. Blow each one with dry air before a tool is put away, and check the bores with a light after every washdown.

Shutdown and Storage Protocol

A shutdown routine returns more per minute than any other corrosion control. A tool dried and filmed in 20 minutes at the end of a run avoids the pitting that a wipe-down with a rag collects over a two-week break.

• Drain and blow every circuit with dry air at 4 to 6 bar until the exhaust reads under 10 percent RH.
• Spray a 5 to 20 micron anticorrosion film on every polished face, shut-off land and gate.
• Wipe slideways and ejector bores with the same film, then cycle the ejector plate twice.
• Seal the mold in VCI film with 10 to 20 g of desiccant per 0.1 m3 of enclosed volume.
• Store on a pallet 100 mm clear of a concrete floor, in a room held at 40 to 60 percent RH.
• Re-check and re-film at 6 months, and again after any move between buildings.

Seal the water ports. An open circuit breathes humid air and condenses on the channel wall. A stored mold with open ports can pit in an otherwise dry room. Fit blind plugs or cap the quick couplings, and label the tool with the date it was preserved. The handling and rigging side is covered in storing a mold.

Ambient conditions finish the job. Concrete holds a boundary layer at 60 to 90 percent RH, so a mold left on the floor rusts from the bottom plate upward. A room at 40 to 60 percent RH and a heater holding the tool 2 to 5 °C above ambient stop the condense and dry cycle.

Inspection and Flow Monitoring Loop

Corrosion is invisible from outside the mold base, so a scheduled inspection loop is the only early warning a tool gets. Three checks catch it inside the window where a re-plate still works.

Flow-check every circuit at 20,000 to 50,000 shots and compare the reading against the baseline logged at the trial. A drop above 10 percent on any circuit points to scale, sediment or a closing channel. Measure at the manifold with a meter covering 5 to 40 litres per minute per circuit.

Pull one hose per tool every 100,000 shots and inspect the first 50 mm of the return line. That section collects pits and iron sediment, and a photo taken from the same angle each time becomes a wear record. Rust colour names the source as well: orange-brown is iron, blue-green is a copper alloy insert under attack, and white is hardness scale.

Log pit depth with a gauge or an ultrasonic thickness reading on the channel wall. Strip and re-plate before any pit passes 0.3 mm deep, because past that depth the repair becomes welding and re-machining. Weld repair inside a water channel is possible, but the heat distortion usually costs more than a replacement insert. That work sits with repair and maintenance.

When to Repair or Replace a Corroded Mold

A pit map decides the repair route. Localized pits under 0.3 mm deep clean up with a polish and a re-plate. Pits from 0.3 to 1 mm need a weld build-up and re-machining. Wall loss over 1 mm, or pitting across more than 30 percent of a channel length, points to a new insert.

Remaining steel wall sets the limit. A 12 mm channel drilled 15 mm below the cavity face leaves 9 mm of steel. A 1.5 mm loss removes 17 percent of that wall. Under 4 mm the channel can deform at 60 to 120 MPa of injection pressure, and the plate becomes a pressure vessel with a thin lid.

Sequence the work against production. Standard components ship in 3 to 7 days and a replacement insert adds 2 to 5 weeks to a tool build. A corrosion finding logged at 300,000 shots should be quoted while the press still runs. Bring the pit photos, the flow baseline and the water analysis to that quote.

Mold Corrosion FAQ

Q1. Why do my cooling channels rust when the water looks clean?

Clean to the eye is not clean to steel. Untreated tap water near 400 microsiemens per centimetre carries enough chloride and dissolved oxygen to pit P20 inside one warm season. Test the loop, then hold it at 200 to 500 microsiemens with pH between 7.5 and 9.5.

Q2. What water quality stops mold corrosion?

Hold conductivity at 200 to 500 microsiemens per centimetre and chloride under 25 ppm for stainless channels. Keep pH at 7.5 to 9.5, hardness under 150 ppm and flow at 1.5 to 3 m/s in every circuit.

Q3. Will glycol in the loop protect my mold?

Glycol protects against freezing, not corrosion, and it cuts heat transfer 5 to 15 percent. Use it only where the loop sees below 5 °C, and replace it on the supplier interval.

Q4. How do I stop PVC from corroding my mold?

Vent the mold every 40 to 60 mm and run the melt at the low end of the window. Put a nitrided or chrome-plated surface on the cavity, and wipe and film the tool at the end of every run. Hydrogen chloride keeps working overnight.

Q5. Should a pitted core be re-plated or welded?

Re-plate while pits stay under 0.3 mm deep and the base steel is sound. Above 0.3 mm a 25 to 50 micron layer cannot bridge the defect. The choice becomes a weld build-up with re-machining or a new insert.

The Bottom Line

Mold corrosion is a water, chemistry and storage problem, and it announces itself as a 5 to 12 °C temperature drift long before it becomes a leak. Every limit in this guide is a number a tool specification can carry.

Hold the loop at 200 to 500 microsiemens, chloride under 25 ppm and pH 7.5 to 9.5, and keep flow between 1.5 and 3 m/s. Coat channels with electroless nickel at 25 to 50 micron, chrome or nitride the cores that see PVC and flame-retardant grades, and break every galvanic couple in the manifold. Blow each circuit dry at 4 to 6 bar, film the polished faces at 5 to 20 micron, and store at 40 to 60 percent RH. A P20 tool that would pit in one warm season then runs for years.

DieStrike builds injection molds to IATF 16949 with ±0.005 mm mold tolerance, ±0.002 mm on EDM and jig grinding, cavity hardness to HRC 62 and 120+ machines. Standard components ship in 3 to 7 days and tool lead time runs 2 to 5 weeks. The same shop supports tier-1 programs with TE, Amphenol, Luxshare and Dongshan.