Why Dies Wear Out and Fail

A die is discarded for one of nine reasons

The published account of die and mold wear separates the ways a tool is lost into nine mechanisms. Three of them, adhesive wear, abrasive wear and erosive wear, appear in most forming processes in some proportion. The remaining six end the life of the die through cracking, softening or an overload that was never a wear phenomenon at all. A die shop that treats every worn die as an abrasive wear problem will change the grade and see no improvement, because the mechanism was never abrasion.

Die and mold wear and failure mechanisms and the processes each one is associated with, summarised from the article on friction and wear of dies and die materials in ASM Handbook, Volume 18. The source notes that the first three are common to most forming processes, and that erosion rather than abrasion is the leading loss mechanism in die casting.

Two of these deserve naming on their own account. Thermal fatigue, which the shop floor calls heat checking, is the most common single reason a die casting die is scrapped, and it is followed closely by soldering, the adhesion of casting metal to the die surface. In hot forging heat checking is the second reason a die is rejected, behind abrasive wear. Both arrive as a network of fine cracks on the working surface rather than as a measurable change in dimension, so neither shows up in the maintenance record as wear.

What decides heat checking

A heat check is a low cycle fatigue crack driven by the strain amplitude of each thermal cycle, which is set by the temperature swing on the die surface rather than by the load. The properties that resist it are a low coefficient of thermal expansion, high thermal conductivity, high hot yield strength, resistance to temper softening, high creep strength and enough ductility to absorb the strain without cracking. Every die steel with a martensitic, bainitic or ferritic structure has roughly the same expansion coefficient, so that term is a constant across the grades a die shop normally buys. Austenitic structures are about 50 percent higher and are not die casting material on that account.

The practical lever is the temperature swing itself. A published comparison of the two ends of that swing shows that lowering the peak surface temperature by 100 C extends the crack initiation life far more than raising the minimum bulk temperature by the same 100 C, and the reason is that the peak sets the strain amplitude. Peak temperature can be pulled down by spraying a lubricant that insulates as well as lubricates, by lowering the workpiece temperature, or in forging by keeping the die preheated so that the swing starts from a higher floor. Keeping the peak below the Ac1 temperature also avoids the volume change of a re-austenitising and re-hardening cycle at the surface, which adds transformation strain on top of thermal strain.

Cleanliness and grain size enter through crack initiation. A steel refined to remove inclusions resists the start of a heat check because inclusions are where the crack starts, and a finer grain performs better for the same reason. A die steel that is chemically and structurally uniform has uniform thermal properties, so it does not build internal strain out of its own segregation, which is the argument for vacuum arc remelted die blocks on long run work. The reheating advice for the shop is short. Nitride after the final temper, never before, and temper above the nitriding temperature.

Service temperature sets the material family

Before grade selection there is a simpler question. The published service temperature bands for die materials in forging separate the four families by roughly 200 to 300 C at each step, and they are the reason a low alloy die block cannot be pushed into a press operation that a hot work grade would survive.

Typical service temperature of die materials in forging, reproduced from ASM Handbook, Volume 18. The bands overlap because the useful ceiling of a grade depends on how long it spends at temperature and on the hardness it was tempered to.

Above the hot work band the alternatives stop being tool steel. Maraging grades reach two to two and a half times the toughness of a chromium hot work steel at the same hardness and resist heat checking at least as well, but their wear resistance is low and the coating or nitriding step is what makes them economic. Cast dies in H12, H13, H19, H21, H26 and H42 hold hot hardness and wear resistance above their wrought equivalents and are less likely to crack at sharp corners, and plant experience of more than 50 percent longer life has been reported, but the pattern cost is only worth carrying when at least three castings come off one mould. Nickel base superalloys hold strength to 760 to 925 C and belong in isothermal forging, where a die steel cannot operate at all.

Normal life of a die casting die, and what ends it

Die casting is the clearest case of a process where the dominant mechanism is fixed by the melt rather than by the die steel. Zinc, aluminium and copper base melts sit in different temperature and chemical regimes, and the tabulated normal life follows the casting temperature rather than any property of the die.

Life-limiting factors for die casting dies, reproduced from ASM Handbook, Volume 18. The life figures are thousands of shots. The core of an aluminium or brass die gives up before the cavity does because it runs hotter and is surrounded by melt on more faces.

Hot work tool steel is not recommended for casting zinc except on long runs, and the prehardened mould grades handle it at a typical working hardness of 29 to 34 HRC. Aluminium and magnesium take H11 and H13 at about 44 to 48 HRC. Copper alloys take the tungsten hot work grades H20, H21 and H22 at 38 to 45 HRC, and the reason is the casting temperature rather than the hardness, because a brass melt at around 1000 C is beyond the temper resistance of a 5 percent chromium steel. The stock and grade side of that choice is set out on the zinc and magnesium die casting page, and the H13 case on why H13 is the die casting grade.

Erosion in a die casting die is the washing away of die material by the incoming molten metal jet, and the published results reduce it to three statements. Melt temperature is the strongest single variable, so the melt should be run as cold as the casting allows. The difference between common hot work grades is not significant, so changing grade does not buy erosion resistance. The third finding is the useful one, that hardened and tempered die blocks lose less material than soft annealed ones and that if metallic contact between steel and melt can be prevented, for instance by keeping an oxide film intact, the risk of severe attack drops sharply. That is why a nitrided surface helps, and also why a nitrided case on a die casting die should not exceed 0.13 mm, since a deeper case spalls at parting lines and sharp edges and the spalled pit tears the casting on ejection.

Abrasive wear is what controls hot forging die life

In forging, the amount of die material removed by abrasion is proportional to the interface pressure and to the amount of relative sliding, and inversely proportional to the hardness of the die surface. Because a real forging die has varying pressure and sliding from point to point, the published work is built on simulative tests, and the results are consistent enough to quote.

Alloy content is the strongest die material variable. A controlled series of eleven die steels tempered to nearly the same hardness found that good wear resistance appears once the total alloy content passes about 3 percent, that molybdenum has a particularly strong effect, and that above about 2 percent molybdenum there is no further gain. The relative effectiveness of the carbide forming elements at die temperatures between 250 and 550 C is given in the ratio 2 to 5 to 10 to 40 for chromium, tungsten, molybdenum and vanadium, which makes vanadium and its carbides eight times as effective as tungsten and its carbides. A second ranking of tungsten, molybdenum and vanadium alone gives 10 to 20 to 40. The practical reading is that a high vanadium hot work grade wears measurably less than a plain chromium grade of the same hardness.

Hardness on its own works, but not equally across the field. The dependence of wear rate on hardness is steepest for the low alloy die steels, which is why raising the working hardness of a low alloy die block pays, and flatter for the hot work grades whose wear resistance comes mainly from their alloy carbides. Structure matters as well. A die hardened isothermally to lower bainite outlasts the same steel quenched and tempered to the same hardness, and the published explanation is that the isothermal treatment leaves fewer internal stresses and microcracks to start an abrasive failure. A comparison of flange forming dies makes the grade effect concrete. M4 and D7 working at 65 HRC gave more than twice the life of D2 at 61 HRC on 17-7 PH stainless, and a 4140 die that was nitrided gave about eight times the life of the same die left un-nitrided.

Three process variables move die wear as much as the steel does. Workpiece temperature has a peak, with wear rising to about 1100 C and falling again above it, because scale formation rises with temperature and then the flow stress of the workpiece drops enough to cut the interface pressure. Scale is the second, and the source estimates that poor scale control can cost 200 percent of die life, which is the argument for a controlled furnace atmosphere, an anti-scale coating or induction heating with the time at temperature kept short. Lubrication is the third and behaves in a way that surprises on first reading. In an open upsetting test a lubricated die wears more than a dry one, because the lubricant lets the workpiece slide further. Calculated against equal metal flow past a point, the same lubrication reduced wear by a factor of three.

Where the die is limited by abrasion the remedies are finite and can be listed. Move to a more highly alloyed hot work steel and accept the material and machining cost, or nitride it, which the source reports cutting wear rates by as much as half. Reduce scale. Redesign the blocker so that more squeezing and less lateral flow happens in the finishing impression. A hard coating will help only while it stays bonded, because a coating that cracks in service sheds fragments that then abrade the die faster than the uncoated surface would have worn. The hardness side of the same argument is collected on the hot hardness chart, and the heat checking side on the thermal fatigue resistant grades page.

Galling and pickup in sheet metal forming dies

Sheet forming dies wear in proportion to the distance the sheet slides for a given contact pressure, so thin annealed stock is the mildest case and thick hardened stock is the worst. Galling is the failure that stops production. It comes from frictional heat between die and sheet softening the die surface, so resistance to softening under heat is the property that ranks the grades, and the published remedies follow from that.

Measures against galling and pickup in sheet metal forming dies, from ASM Handbook, Volume 18. The source records that the amount of wear on a die radius can vary by a factor of twenty between the sharpest and the most generous radius, which makes radius design the first thing to check when a die that worked starts to pick up.

One detail is worth repeating because it is a specification the buyer can write. Hard chromium plating resists galling best when the substrate under it is harder than 50 HRC, when the coating is between 0.005 and 0.01 mm thick, and when the plated part is held at 150 to 205 C for at least three hours immediately after plating to drive off hydrogen. Coating thicker than that on a die spalls rather than wears. On stainless and high nickel work the published order of preference for the draw ring is aluminium bronze first, then D2, D3 or D4 smoothly ground, nitrided and polished, and then alloy cast iron quenched and tempered to 400 to 420 HB.

Grades and stock

The mechanisms above decide which group of grades to look at, and the group ranking itself is on the tool steel properties chart. The hot work range that carries hot forging, extrusion and die casting dies is listed on the page for hot forging, stamping and trimming dies. Where the die has already failed and the mechanism is not obvious from the wear pattern, the reading order is on the tool and die failure analysis page.

Two neighbouring mechanisms are covered separately. Where the damage is a shallow bruise pattern from small oscillating movement rather than a sliding wear scar, the diagnosis is on the fretting wear page. Where the tool lost a piece of its edge instead of wearing, the pattern belongs to impact wear and spalling. Where the question is what to do about a die that picks up and seizes rather than wears, the grade side is on the adhesive wear resistant grades page.

Source: ASM Handbook, Volume 18, Friction, Lubrication, and Wear Technology, ASM International, 1992.