Cr8 Cold-Roll Steel + 0.6% Niobium: a 40% Drop in Roll Wear, Validated by JMatPro
A Sendzimir 20-high mill second intermediate roll in Cr8 steel, quenched at 1090–1120 °C, tempered at 550 °C × 3 h, and micro-alloyed with 0.6% Nb, sees its wear volume fall 40% versus the baseline. The wear mechanism shifts from severe abrasive grooving to a protective, oxidation-controlled regime.
Cross-reference for procurement and standards teams. Cr8 (China GB/T 1299) is the 8% Cr cold-work family that sits between the 5% Cr A2 and the 12% Cr D2. The closest overseas relatives are the Cr8Mo2SiV variant (AISI D2-modified, DIN 1.2990, JIS DC53 / SKD11-mod) and the Russian GOST 8KhF. The 0.6% Nb addition in this study is a FUSHUN METAL design, not part of any standard grade.
Grade Cross-Reference: Where Cr8 Sits Between A2 and D2
Cold-work tool steels are usually classified by chromium content. The 5% Cr family (AISI A2, DIN 1.2363) gives the best toughness but only moderate wear life. The 12% Cr family (AISI D2, DIN 1.2379) gives the highest wear life but is brittle and prone to chipping. Cr8 is the deliberate 8% Cr compromise: enough chromium to form stable M7C3 and M23C6 carbides, low enough to keep the eutectic network fine and the matrix tough.
| Standard / Region | Designation | C | Cr | Mo | V | W | Other |
|---|---|---|---|---|---|---|---|
| China GB/T 1299 (standard Cr8) | Cr8 | 1.6–1.9 | 7.5–8.5 | — | — | — | — |
| China (this study, modified Cr8) | Cr8 + Nb | 1.1 | 8.0 | 1.3 | 1.2 | 0.8 | + 0.6% Nb |
| China (variant) | Cr8Mo2SiV | 0.9–1.05 | 7.5–8.5 | 1.8–2.1 | 0.15–0.35 | — | Si 0.8–1.1 |
| US ASTM A681 (closest) | AISI D2 modified | 1.5–1.6 | 11–12 | 0.7–1.0 | 0.7–1.0 | — | — |
| Europe W.Nr / DIN | 1.2990 / X38CrMoV5-3 | 0.38 | 5.0 | 3.0 | 0.6 | — | sibling: DIN 1.2379 (D2) |
| Japan JIS / Daido | DC53 (≈ SKD11-mod) | 0.9–1.1 | 7.5–8.5 | 1.8–2.0 | 0.2–0.5 | — | Si 0.8–1.1 |
| Sweden ASSAB / Uddeholm | ASSAB 88 / Sleipner | 0.9 | 7.8 | 2.5 | 0.5 | — | — |
| Russia GOST 5950 | 8KhF (8ХФ) | 0.7–0.8 | 7.5–8.5 | — | 0.3–0.5 | — | W ≤0.2 |
No single overseas standard is a one-to-one equivalent of the modified Cr8 in this study (C 1.1%, Cr 8%, Mo 1.3%, V 1.2%, W 0.8%). The closest neighbours are DC53 (lower V, higher Mo+Si) and AISI D2 (higher Cr). The +0.6% Nb in this study is a FUSHUN METAL design addition, not part of any published grade.
Why Sendzimir 20-High Mill Intermediate Rolls Demand a Better Cr8
In a Sendzimir 20-high mill, the rolling force applied to the work roll is transmitted radially through the first and second intermediate rolls, then into a cluster of backup rolls and finally into the mill housing. The second intermediate roll is the most fatigue-loaded of the small-diameter rolls: it takes the full rolling load, runs at high cyclic frequency, and any chipping or surface spalling shows up instantly on the strip as a thickness or shape defect.
AISI D2 / DIN 1.2379 (12% Cr) delivers the highest red hardness, but its coarse eutectic carbide network makes it brittle in the cyclic-impact conditions of a 20-high mill. Cr8, with 8% Cr and roughly 1% Mo+V+W, lands in the middle: a fine, even carbide population, deeper hardenability than the 5% Cr A2 family, and enough matrix toughness to absorb the cyclic shock. The challenge is that unmodified Cr8, while tougher than D2, wears faster because it carries less total carbide volume.
| Property | AISI D2 / 1.2379 | Cr8 (baseline) | Cr8 + 0.6% Nb |
|---|---|---|---|
| Chromium content (%) | 11.0–12.0 | 8.0 | 8.0 |
| Eutectic carbide network | Coarse, brittle | Fine, uniform | Fine + NbC pinning |
| Quench window (°C) | 1000–1050 | 1050–1100 | 1090–1120 |
| Temper window (°C × h) | 500–530 × 2 | 520–560 × 2 | 550 × 3 |
| Wear volume (mm³, GCr15 ball, 20 N, 2 Hz, 120 min) | ~5.0 × 10⁻² (literature) | 4.719 × 10⁻² (measured) | 2.815 × 10⁻² (↓40%) |
| Dominant wear mechanism | Abrasive grooving | Severe abrasion + plastic deformation | Oxidative wear (protective) |
The 0.6% Niobium Sweet Spot
Niobium is a strong carbide former. In a Cr8 base it forms NbC, which is even harder than VC. But the value of Nb is not the carbide volume — it is where the NbC ends up.
Below 0.4% Nb, the NbC population is too small to pin the prior-austenite grain boundaries; the gain in hardness and yield strength is marginal. Above 0.8% Nb, the NbC precipitates in the liquid during solidification, grows coarse, and ends up clustered inside the austenite grains rather than at the grain boundaries. Once NbC sits inside the grain it loses its pinning role and instead becomes a stress concentrator — crack-initiation sites appear and the toughness collapses.
Between 0.4% and 0.8% lies a narrow window where NbC precipitates as fine particles with the austenite at the grain boundaries, and the grain-pinning effect is at its peak. JMatPro simulations in the source paper show that hardness and yield strength both rise steeply from 0.1% to 0.6% Nb, then fall. The peak sits at 0.6% Nb.
FUSHUN METAL view: 0.6% Nb is the engineering sweet spot. Below it, you are paying for niobium without getting the boundary pinning. Above it, you are paying for a metallurgical liability. The cost of 0.6% Nb is roughly one third of what 0.6% Mo would cost at today’s alloy surcharge — so the case for Nb is also a procurement case, not just a metallurgical one.
The 1090–1120 °C Quench Window: Higher than D2, for a Reason
AISI D2 is conventionally quenched from 1000–1050 °C. The 8% Cr Cr8 in this study needs 1090–1120 °C — roughly 80–100 °C higher. The reason is hidden in the CCT diagram and the austenite carbon-dissolution curve.
Martensite hardness peaks at about 0.6% dissolved carbon in the austenite at the moment of quenching. JMatPro was used to track the carbon content of austenite as a function of temperature for three Nb levels:
| Nb content (%) | 0.0 | 0.6 | 1.0 |
|---|---|---|---|
| Austenitization temperature for 0.6% C in solution (°C) | 1098 | 1108 | 1112 |
Three observations follow from this table:
- Every 0.6% Nb push lifts the required austenitization temperature by roughly 10 °C because Nb ties up some of the carbon as NbC that has to be re-dissolved.
- The martensite-start (Ms) temperature also climbs with Nb — from 158.6 °C at 0% Nb to 167.6 °C at 1.0% Nb. A higher Ms means more austenite transforms to martensite during the quench, leaving less retained austenite, which translates directly to higher post-quench hardness.
- Above 1120 °C, grain coarsening kicks in and offsets the carbide-dissolution gain. The practical window is therefore 1090–1120 °C, with 1100 °C as the engineering target.
Cooling medium is water quench. Oil or polymer leaves more retained austenite and misses the hardness target. Air cooling is out of the question for a roll of this section size.
550 °C × 3 h Temper: Slightly Past the 8.2% MC Peak, on Purpose
During tempering, the supersaturated martensite rejects carbon as fine MC carbides (mainly VC and NbC). JMatPro tracking across 500–600 °C and 3–4 h hold times gives a very clear picture:
| Temper parameter | MC content (%) | MC particle size | Reading |
|---|---|---|---|
| 500 °C × 3 h | ~1.0 | Fine | Below the precipitation threshold; most C still in solid solution |
| 540 °C × 3 h | 8.2 (peak) | Largest | Critical precipitation temperature; massive MC nucleation |
| 550 °C × 3 h | ~7.5 (slightly off peak) | Slightly smaller | Chosen operating point — see reasoning below |
| 580 °C × 3 h | ~5.0 | Coarsening starts | Coarsening and partial re-dissolution; hardness drops sharply |
| 600 °C × 3 h | < 3.0 | Coarse | Severe over-tempering; wear resistance collapses |
So why 550 °C if the peak is at 540 °C? Two reasons:
- The MC peak at 540 °C is also where the particles are largest. Just past the peak, the particles refine slightly even though the total MC volume drops a little — a better distribution more than compensates for the small volume loss in wear life.
- Tempering at exactly 540 °C in a production furnace is hard to hold. A ±10 °C drift at that temperature swings MC content by several percent. 550 °C sits on the flatter shoulder of the curve, so furnace variability hurts less and the heat-treat operator has margin.
Hold time is fixed at 3 h because MC content is essentially flat from 3 h to 4 h — extending the hold gives no metallurgical benefit and only burns furnace capacity.
The 40% Wear Drop — and Why Oxidation Beats Abrasion
Pin-on-disc tribo-tests against a GCr15 ball (φ6 mm, 2 Hz, 8 mm stroke, 20 N normal load, 120 min, dry room-temperature) give the headline result:
| Indicator | Cr8 (no Nb) | Cr8 + 0.6% Nb | Change |
|---|---|---|---|
| Wear volume (mm³) | 4.719 × 10⁻² | 2.815 × 10⁻² | −40.4% |
| Wear depth (μm) | ~25 | ~20 | −20% |
| Wear width (μm) | ~330 | ~330 | ~0% (depth, not width, drives the volume drop) |
| Dominant mechanism | Abrasive grooving, plastic deformation | Oxidative wear, near-zero grooving | Mechanism shift |
| Surface oxide layer | Thin, discontinuous | Dense, continuous | Protective |
A 40% volume drop on a laboratory tribo-test is a strong signal. It does not translate one-to-one into 40% longer service life in a real mill, because the mill adds bending fatigue, thermal cycling, and emergency stop events that the test does not capture. But the mechanism shift is the part that matters: the baseline Cr8 fails by deep abrasive grooving and ploughing, while the 0.6% Nb variant forms a dense oxide layer that resists further material loss.
FUSHUN METAL view: in service, oxidation-controlled wear is almost always better than abrasion-controlled wear. An oxide layer slows heat build-up at the contact zone, prevents adhesive transfer, and stops the surface from micro-spalling. When a roll is running at the limit of its wear life, the question is not “how much material is left” but “what is the wear mechanism” — and oxidation is the friendlier mode.
Industrial Takeaways
- 0.6% Nb is the sweet spot for Cr8 cold-roll steel. Below 0.4% there is no grain-boundary effect; above 0.8% NbC clusters inside the austenite grains and degrades toughness.
- Quench 1090–1120 °C × 2 h, water quench. The 8% Cr base needs about 80 °C more than D2 to get 0.6% C into austenite solution. Under-austenitizing leaves undissolved M23C6 and cuts hardness; over-austenitizing coarsens the grain.
- Temper 550 °C × 3 h, air cool. Slightly past the 540 °C MC peak gives a flatter response to furnace variability and a finer MC distribution.
- Expect a mechanism shift, not just a number change. The 0.6% Nb variant moves the wear mode from severe abrasion to oxidation-controlled wear — the longer-term benefit in a real mill is greater than the 40% number suggests.
- Process a trial bar before the production roll. JMatPro is a strong guide but it does not catch raw-material variability. A 50 kg trial bar through the full heat-treatment cycle, followed by a small pin-on-disc test, is a low-cost insurance policy before committing a 1.5-tonne roll.
About the author
Harris is part of the FUSHUN METAL technical team. FUSHUN METAL supplies alloy steel, stainless steel, and nickel-based alloy bar, plate, wire and strip to global B2B customers across oil & gas, power generation, semiconductors, medical, and turbine-rotor applications. Process data in this article is sourced from the open study “Heat-treatment process and alloying-composition optimization of Cr8 steel for cold rolling rolls” (Hou Xiaohui, Taiyuan University of Science and Technology, published in Metal Heat Treatment, 2026.2) and cross-checked against the public grade data of AISI A2, AISI D2, DIN 1.2363, DIN 1.2379, JIS DC53 / SKD11, ASSAB 88 / Uddeholm Sleipner, and GOST 8KhF.
