Why ZGMn13 Liners Crack: A Failure Analysis Case
A cement plant installed a new set of ZGMn13 liners. Within days — not months, days — plates were shattering into multiple pieces. The mill had barely settled in.
Failures this fast are rarely about the mill. They are about the liner’s metallurgical history. What the lab found in this case is worth memorizing, because it is the same fingerprint that shows up in early liner failures all over the world.
The evidence
| Check | Result | What it should be |
|---|---|---|
| Manganese content | 8.47% | 11.0–14.0% (ZGMn13 spec) |
| Carbon | 1.05% | 0.9–1.4% (in range, barely) |
| Interior hardness | 46 HRC | ~20 HRC (water-toughened austenite) |
| Fracture surface | cleavage + quasi-cleavage, secondary cracks | — |
Two findings do all the explaining. First, the manganese was three points below specification. Second — and this is the one that ends arguments — the plates had never been water-toughened at all. Water-toughened ZGMn13 measures about 20 HRC inside; these plates measured 46 HRC. That hardness does not come from austenite; it comes from hard, brittle carbonides.
The mechanism: carbides at the grain boundaries
As-cast high-manganese steel is full of blocky carbides precipitated along grain boundaries — hard, brittle, and connected into near-networks. Water toughening (≈1,050 °C, hold, quench) exists precisely to dissolve them back into the austenite. Skip it, or start with too little manganese (which pushes carbide formation), and the plate ships with a glass-brittle grain-boundary network under a deceptively soft-looking surface.
Under ball impact, cracks nucleate on those carbide films and run clean through the plate — the classic human-arrow fracture pattern, cleavage faces under the microscope.
The kicker: the failure analysis re-treated a broken plate at 1,050 °C for one hour, water quenched — and the carbide networks dissolved, the structure returning to proper austenite. The steel was fixable. It simply never received the treatment the buyer paid for.
The wider pattern of early manganese-steel failures
This case is the extreme. Four chronic versions appear across mill failure records:
1. Bolt-hole cracking (the statistical leader). One tracked mill saw 60% of liner fractures run through bolt holes — square holes concentrate stress, and the surrounding metal solidifies last and hottest in the casting. Fixing the hole design and casting sequence cut the rate to 5%.
2. Swelling and distortion. Under-hardened plates flow instead of wearing. Documented mechanism stack: thermal expansion 50% above plain carbon steel, phase-transformation expansion between 300 °C and Ar₃, coarse grains with low yield strength. One plant’s liners grew 272 mm in a 240 mm envelope in under six months and sheared their M30 bolts.
3. Thermal runaway at the hot end. ZGMn13 has a practical ~300 °C service ceiling. In one coal mill with overheated inlet air, the front rows entered a self-feeding loop — spalling, thinning, running hotter, wearing faster — scrapping entire sets at two-thirds of design life while cooler rows wore normally.
4. Foundry-side cracks. At one large foundry, 80–90% of high-manganese casting rejects were crack defects — from shrinkage restraint, thermal gradients, and grain-boundary carbides. The seven-control fix list (gating, chills, low-P chemistry, rare-earth inoculation, controlled shakeout, slow heat-up, sub-2-minute transfer to quench) is standard practice at any foundry that ships fewer rejects than it makes.
And a replacement-strategy trap worth knowing: one plant replaced only the front half of a worn liner set. The new, thicker plates re-distributed the load; within a month the old half started failing in cascade, and the whole set had to come out anyway.
The buyer’s inspection checklist
If you source Hadfield/manganese liners, these five checks catch most of the failure modes above:
- Mill test certificate: Mn within 11–14%, P ≤ 0.06%. Sub-spec manganese is the classic cheap-material tell.
- As-supplied hardness: 180–230 HB. Significantly higher means carbides never went into solution.
- Heat treatment records: water-toughening temperature, hold time, and transfer-to-water interval (should be well under 2 minutes).
- Microstructure on a sample plate: carbide rating per the accepted scales — network carbides or more than a few carbides per field at 500× is a reject.
- Aging step (if specified): the 300–350 °C low-temperature aging that adds real wear life — see our 350 °C aging guide. A foundry that skips aging usually skips other things too.
The economics
None of this is exotic. Every item above is standard metallurgy, documented for a century. The difference between a manganese liner that lasts six months and one that lasts two years is rarely the grade name — it is whether the chemistry and heat treatment behind that name were actually executed. That distinction is precisely where “cheap” liners stop being cheap.
QM Mill Liners ships mill-test certificates, heat treatment charts and carbide ratings with every Hadfield order — because we would rather prove it than argue about it. Send your current liner failure photos for a read on root cause — contact here.
FAQ
Why do manganese steel liners crack? Most often: carbide networks from missing or poor water toughening, sub-specification manganese, bolt-hole stress concentration, or service above ~300 °C. The case above combined the first two.
How can I check if liners were properly water-toughened? As-supplied hardness near 180–230 HB and interior microstructure free of network carbides. Interior hardness of 40+ HRC is a red flag — proper austenite is around 20 HRC.
How long should ZGMn13 liners last? In suitable high-impact duty with correct processing: 6–12 months is typical; modified grades with proper aging reach 2× that. If they fail in weeks, the metallurgy failed, not the grade.