Two Different Materials Are Hidden Behind One Search Term
Engineers searching for temperature forming of wear-resistant plate usually have one of two very different products in mind, and mixing them up is the fastest way to specify the wrong process. The first is quenched and tempered abrasion-resistant plate supplied to GB/T 24186 in the NM400, NM450, NM500 and NM550 grades, where the grade number states the nominal Brinell hardness level. That plate is already martensitic when it leaves the mill; any forming operation that heats it above its tempering range destroys the hardness that was paid for. The second is press-hardening sheet steel of the 22MnB5 type, a boron-microalloyed grade with roughly 0.22% carbon and about 1.2% manganese, which is deliberately heated into the austenite region and quenched inside the tool so that the finished part becomes martensitic. In the first case heat is an enemy of the specified property; in the second case heat is the mechanism that creates it.
Both families are discussed below, because the tooling, the furnaces and the quality documents overlap, while the acceptance logic does not. A buyer who orders NM450 and then asks for 950 °C forming has ordered a part that will measure far below the nominal hardness unless it is re-austenitised and re-tempered afterwards. A buyer who orders 22MnB5 and asks for room-temperature forming has ordered a part that will crack or spring back beyond tolerance. Process temperature must always be written on the purchase order together with the property that must survive it.
Why Cold Forming Runs Out of Capacity Above Roughly 1000 MPa
Formability is controlled by the ratio of yield strength to elastic modulus and by the strain-hardening exponent. As tensile strength climbs toward 1000 MPa and beyond, the elastic portion of the stress-strain curve occupies a larger share of the total strain available, so the material stores more elastic energy before it yields. When the punch is withdrawn, that stored energy is released as springback, and the part no longer matches the die. At the same time the flow stress at room temperature approaches the capacity of the press and of the tool, and edge cracking appears because the sheared edge has almost no ductility left to redistribute strain.
Warming the blank attacks all three problems at once. Between about 500 °C and 700 °C - that is, above the temperature where long-range dislocation motion becomes thermally activated but below the Ac1 temperature of roughly 727 °C for these low-carbon chemistries - yield strength falls substantially while the microstructure remains ferritic-pearlitic or tempered martensitic. The result is a lower forming load, a smaller springback angle and a better distribution of strain across the part. Nothing about the phase constitution changes, which is exactly the point: warm forming is a rheology tool, not a heat treatment.
Published forming studies on automotive anti-intrusion parts report the same trade-off from the opposite direction. When a warm-formed grade is compared with 22MnB5 at an equivalent 1500 MPa tensile level, the warm route is described as reaching its strength at a lower austenitising temperature, in the region of 800 °C instead of 950 °C, with a finer structure, no surface decarburisation and a total elongation improved from roughly 7.5% to about 10%. Those figures are the reported outcome of one research programme on specific chemistries and should be read as an indication of the direction of the effect, not as a guaranteed property of every plate carrying a similar grade name.
Comparing the Four Forming Routes on the Same Page
The table below puts the practical routes side by side. Load factors are expressed relative to conventional cold forming of the same thickness, and are indicative ranges rather than design values; the die design must always be calculated from measured flow curves for the actual heat.
| Route | Blank temperature | Typical final tensile strength | Forming load vs cold | Springback | Dominant risk |
|---|---|---|---|---|---|
| Cold forming | Ambient | Up to about 1000-1200 MPa | 1.0 (baseline) | High and variable | Edge cracking, tool overload |
| Warm forming | 500-700 °C | Set by the prior heat treatment | 0.4-0.6 | Moderate, more repeatable | Decarburisation, galling, die wear |
| Direct hot stamping | 900-950 °C, quenched in die | About 1500-1600 MPa | 0.3-0.4 | Very low | Scale, coating damage, die thermal fatigue |
| Indirect hot stamping | Cold preform, then austenitised | About 1500 MPa and above | Mixed | Low | Distortion during transfer, extra operation |
| Forming of QT abrasion plate | Ambient to about 250 °C | Unchanged if temperature stays below tempering range | 1.0 or higher, large radii | Low, high stiffness | Hardness loss and cracking if overheated |
Two consequences follow directly from the table. First, warm forming and hot stamping are not interchangeable: one preserves a microstructure, the other creates it. Second, for abrasion-resistant plate the useful working range is narrow and low, and every degree above the tempering temperature of the delivered condition is a permanent debit against wear life. A fabricator who needs a tight bend radius in NM500 should plan for a generous radius, a wide die opening and a stress-relief check, not for a torch.
The Process Window That Has to Be Written Into the Work Instruction
For press-hardening grades the window is defined by four numbers: austenitising temperature, soak time at temperature, transfer time from furnace to die, and cooling rate through the martensite start temperature. If any one of them drifts, the part either fails to transform fully or transforms into a softer constituent, and the tensile result falls below the 1500 MPa target even though the hardness spot check may still look acceptable.
| Parameter | Typical industrial setting | Why it matters | How to verify |
|---|---|---|---|
| Austenitising temperature | 900-950 °C | Complete austenite without excessive grain growth | Furnace record plus prior-austenite grain size on a coupon |
| Soak time at temperature | 3-10 minutes depending on thickness | Carbide and boron dissolution | Thermocouple on the blank, not on the furnace wall |
| Transfer time | Ideally below 5-10 seconds | Prevents ferrite or bainite forming before the die closes | Time-stamped transfer log |
| Die quenching rate | Above roughly 27 K/s through the martensite range | Full martensitic transformation | Hardness mapping across the part, dilatometry on coupons |
| Die surface temperature | Controlled, generally below about 250 °C | Balances cooling rate against tool life | Embedded thermocouples in the insert |
| Post-form hardness | Roughly 430-520 HV for a 1500 MPa class part | Indirect but fast check of transformation | HV10 to ISO 6507 or ASTM E384, conversion per ISO 18265 |
Warm forming has a different window. The blank is heated to a temperature that is high enough to soften the material but low enough that no austenite forms, typically 500-700 °C, and it is then transferred to a tool that is often heated in its own right to 150-300 °C to slow the cooling of the sheet during the stroke. Holding time at temperature should be the minimum that gives a uniform through-thickness reading, because every extra minute at 600-700 °C grows the decarburised layer. The depth of that layer is measured to ISO 3887 on a cross-section, and it is the single most useful acceptance test for a warm-formed part.
Tooling, Lubrication and Surface Condition
Warm and hot forming move the wear problem from the workpiece to the tool. At 500-700 °C the oxide that forms on a bare steel surface is abrasive and tends to weld to the die radius, producing galling that is then pressed into every subsequent part. Tooling for these routes is therefore built from hot-work die steels of the 4Cr5MoSiV1 type covered by GB/T 1299, hardened and usually surface treated by nitriding or a hard coating, with polished radii and, where production volume justifies it, replaceable inserts at the draw radius.
Lubrication must be chosen for the temperature, not for room-temperature convenience. Conventional drawing oils carbonise above roughly 250 °C and leave a sticky residue that increases friction rather than reducing it. Solid-film lubricants, graphite-free high-temperature pastes and dry-film coatings applied to the blank before heating are the practical options, and the coating weight has to be controlled because an uneven film produces uneven friction and therefore uneven strain.
Surface protection of the blank is a separate decision. Uncoated press-hardening steel forms a mill-scale layer that must be removed by shot blasting after forming, and the blasting operation is a real cost that is often forgotten in the piece-price comparison. Hot-dip aluminium-silicon coated sheet, supplied under EN 10346, forms a thin intermetallic layer that protects the base metal during austenitisation and remains on the part as a corrosion barrier, but the coating must survive the die without being scored, and its presence changes the welding window. Zinc coatings are avoided in direct hot stamping because the liquid metal can penetrate the austenite grain boundaries under stress, so coated blanks of that type are generally processed by the indirect route and the coating condition is checked at the weld.
Springback, Forming Limits and Die Compensation
Springback never disappears, it only becomes small enough to compensate. The practical workflow is to measure the springback angle on a first-off part, compare it with the prediction from a simulation that uses an anisotropic yield function and a measured hardening curve, and then apply overbend and die-surface correction in two or three iterations. Simulation without a measured flow curve for the actual heat is little better than a guess, because heat-to-heat variation in yield strength of 30-50 MPa will shift the springback angle measurably on a long, shallow part.
Forming limits for these grades are read from a forming limit diagram determined to ISO 12004-2, not from the tensile elongation alone. The sheared edge is the usual initiation site, so edge quality matters as much as the parent properties: a cut edge produced to a poor quality class on a guillotine has a work-hardened, micro-cracked zone that will fail long before the base metal reaches its limit. Specifying the cut edge quality and, where the part allows, adding a shaved or laser-cut edge on the critical contour is often cheaper than changing the steel grade.
Hardness, Microstructure and Acceptance Testing
Three checks cover most of the risk on a formed high-strength part. The first is hardness, taken as HV10 or HV30 at defined locations - flat areas, the draw radius, and the punch corner - and converted to tensile strength only for information. The second is microstructure on a cross-section etched to reveal prior-austenite grain boundaries, looking for a fully martensitic core, an absence of ferrite bands at mid-thickness, and a decarburised layer within the agreed limit measured to ISO 3887. The third is a tensile test to ASTM E8/E8M or ISO 6892-1 on a coupon cut from the same blank and given the same thermal cycle, because a hardness conversion is not a substitute for a tensile result on a safety-critical part.
Dimensional acceptance follows the general tolerances for plate and sheet, with thickness tolerances to EN 10029 where plate is supplied, and flatness checked after forming rather than before, since the thermal cycle usually relieves residual stress and changes the as-supplied flatness. Where parts are thermally cut after forming, the cut edge quality class should be stated explicitly; thermal cutting that is acceptable on mild steel can locally over-temper a martensitic part and leave a soft band along the edge.
Failure Modes and What They Usually Point To
Cracks that open at the draw radius on the first stroke, before the part is fully formed, usually indicate insufficient ductility for the strain path - either the temperature was below the intended window or the material was supplied at a higher strength than assumed. Cracks that appear only after the part has cooled point at hydrogen or at a coating interaction rather than at formability. A part that measures correct hardness in the flat but soft at the radius indicates that the die did not close on that area, so the cooling rate there fell below the critical value; the fix is in the tool, not in the furnace setting.
Inconsistent springback between shifts is almost always a temperature-control issue: furnace zone drift, a blank that waits too long on the transfer table, or a die that has not reached thermal equilibrium at the start of the shift. Logging blank temperature at the moment of entry into the die, rather than furnace setpoint, converts this from an argument into a measurement.
FAQ
Q: Can NM400 or NM500 abrasion-resistant plate be hot formed without losing its hardness?
No. These grades are delivered quenched and tempered, and heating above the tempering range of the delivered condition softens the martensite. If a shape can only be produced hot, the part must be re-austenitised, quenched and tempered after forming, and the final hardness re-verified against GB/T 24186.
Q: What is the difference between warm forming and hot stamping?
Warm forming takes place at roughly 500-700 °C, below the Ac1 temperature, and simply lowers flow stress while leaving the microstructure unchanged. Hot stamping heats above 900 °C to form austenite and then quenches the part inside the die, so the finished part is martensitic. The first preserves properties, the second creates them.
Q: Why is a 27 K/s cooling rate often quoted for press-hardening steels?
It is the order of magnitude of the critical cooling rate needed to miss the ferrite, pearlite and bainite regions in a boron-microalloyed 0.22% C steel and reach a substantially martensitic structure. Die design, cooling channel layout and contact pressure all have to be able to sustain it in the slowest-cooling region of the part.
Q: How is decarburisation measured, and what limit is reasonable?
Decarburisation depth is measured metallographically on a polished cross-section to ISO 3887, distinguishing the complete ferrite layer from the partial decarburisation zone. A reasonable limit for a formed structural part is that no continuous complete-decarburisation layer remains on the functional surface after scale removal; the exact number belongs in the drawing or the technical delivery condition.
Q: Does a warm-formed part need stress relief after forming?
Usually not as a separate operation, because the forming temperature is itself a tempering treatment for many grades, but the decision depends on the amount of cold work imposed afterwards, on any subsequent welding, and on the service environment. Where dimensional stability matters, a low-temperature stress relief below the tempering range of the delivered condition is the safe choice.
Q: Which documents should accompany a formed high-strength plate order?
The purchase order should state the grade and standard, the delivery condition, the forming route and its temperature window, the hardness or tensile requirement after forming, the decarburisation limit, the dimensional standard, and the inspection documents. The mill certificate proves what was shipped; only a coupon tested after the same thermal cycle proves what the part actually is.




