CoatingIQ
← Course index

Wrap-up

88%

Lesson 08·Running Forge Dies

Reading wear: crazing, washout, abrasion, adhesive pickup, plastic deformation

Five wear modes, one visual signature each, and the rules that keep heat check from getting called mechanical fatigue and washout from getting called plastic deformation.

9 min readLesson 8 of 13

Tying it together

Distinguishing similar patterns

Five wear modes, four pattern-confusion traps that cost the most when called wrong.

Heat check versus mechanical fatigue. Heat check is a network, omni-directional, shallow, and follows the temperature map. Mechanical fatigue is a single crack or small set of parallel cracks, deeper than the network around them, aligned with the principal stress direction from forging load, and concentrated at geometric stress concentrators (inside corners, parting lines, edges of impressions). A crack that runs perpendicular to the material flow direction at an unradiused inside corner is mechanical fatigue regardless of whether a heat-check network surrounds it. The fix for heat check is thermal management. The fix for mechanical fatigue is geometry or substrate, and it usually means a die redesign with a generous corner radius the next time the part is built.

Hot tear versus service crack. A hot tear opens during the forging cycle itself, from material flow stress at high temperature, and shows discolored or scaled internal surfaces because the crack saw billet contact and oxidation while it was hot. A service crack from fatigue or thermal cycling shows clean internal surfaces, because it opened and propagated below the temperatures that scale the steel. On a microsection the hot tear shows oxide on the fracture face and the service crack does not. The two are confused on the bench because both can present as a single deep crack at a stress concentrator. Look at the fracture face under a 10x loupe before deciding which one is in front of you. If the crack walls are dark and scaled, the die was tearing under load. If they are bright and clean, the die was cracking under fatigue.

Washout versus plastic deformation. Both show dimensional drift in flow regions and both can present without cracks. Washout has an erosive, eroded valley shape with a matte or polished surface finish, and the dimension has grown by removal of material. Plastic deformation has rounded edges, dished surfaces, and the dimension has drifted by flow of material, not removal. Material lost (washout) versus material moved (plastic deformation). The decisive test is a microhardness reading on the affected feature. Washout leaves the surrounding substrate hardness intact. Plastic deformation usually shows substrate softening because the cause is over-temperature operation or wrong-grade substrate.

The clean band in a heat-checked face. A die with a uniform crazing network across the face except a clean band 5 mm wide running across the middle is not a die with one band in better shape than the rest. The band is not seeing the same thermal cycle as the rest of the face. Most often the band is a lube buildup zone, where graphite has accumulated thick enough to insulate the surface from the thermal shock that crazes everywhere else. The clean band reads as a good sign at the bench and is actually a sign of over-lube in that region. The buildup also changes the local heat balance, which can drive thicker buildup over time and eventually a thermal imbalance that affects the part finish. Less common causes for a clean band: the spray pattern is hitting that region harder than the rest, the die geometry has a relieved region that does not contact the billet, or a previous repair (weld, polish) altered the surface metallurgy in that band. Case (lube buildup) means reduce lube in that region. Case (geometry) is benign. Case (repair) needs metallurgical follow-up. Investigate before deciding what to do, because the obvious read (the clean band is healthy) is the wrong one in most shops.

What this means on the shop floor

For closed-die forge, the working bench protocol is a hand magnifier, a 10x loupe, and a portable HV tester. Walk the working face, name the modes you see, then read hardness on any feature that looks like plastic deformation before you call it washout. Photograph at standardized angles for the trend log.

For upsetters and hot-headers, adhesive pickup on the punch face is the diagnostic that most often drives the pull-now decision, because punch-face pickup degrades part quality immediately. Washout on the cavity floor is the diagnostic that most often drives the dimensional pull on schedule.

For ring-roll dies, plastic deformation of the mandrel is the failure to watch for, because the mandrel runs hot for long durations and substrate softening is the dominant failure path. Crazing on the work-roll face is expected and benign until cracks break out of the network.

Across all of them, the rule is the same. Name the mode before deciding what to do. The fixes for the five modes are different, and applying the wrong fix wastes the die.

Internal pushback questions

  1. Does every operator on the floor share the same five names for the five wear modes, or does each one use private vocabulary that does not translate at shift change?
  2. Is there a standardized photo protocol at die pull, with defined angles and lighting, so the trend log over hundreds of dies actually compares like to like?
  3. When a die comes off with a clean band in a heat-checked face, is the default read "healthy band" (wrong) or "investigate this band" (right)?
  4. Is the portable HV tester on the bench, with a known calibration, and does anyone read hardness on a suspected plastic-deformation feature before calling it washout?

Common confusions

A crazing network is not a failure mode on an H13 forge die past 5K hits. It is the surface accommodating the thermal cycle. Pulling on the network rather than on a propagating crack within the network wastes most of the die's life.

Washout is not a flow problem. It is a lube-and-flow problem at a high-velocity region. Welding the valley back without fixing the upstream cause produces a die that washes out the same way again on welded stock.

Adhesive pickup on a nitrided die surface is rare until the case wears through. Sudden pickup on a die that ran clean for tens of thousands of hits usually means the case is gone in the affected region, not that the lube needs changing.

Plastic deformation without cracks looks less dramatic than crazing but kills dies faster and costs more to recover from. Microhardness on the deformed feature decides whether to call it plastic deformation or washout, and the call drives a different fix.

The clean band in a heat-checked face is rarely the band in good shape. It is most often the band over-lubed enough to insulate from thermal shock.

Up next: the diagnostic workflow when a die starts losing tolerance.

Sources