A lining rarely fails for one reason. Three mechanisms act on it, usually in sequence, and the wear pattern left behind tells you which one led.
| Mechanism | What happens | Driven by |
|---|---|---|
| Penetration | Melt creeps into open pores | High open porosity, wetting, low slag viscosity |
| Dissolution | The refractory chemically dissolves | Basicity mismatch, temperature, unsaturated slag |
| Erosion | Surface is mechanically removed | Flow velocity, turbulence, suspended particles |
Penetration comes first
Slag enters the pore network and reacts at depth, forming phases with a different thermal expansion than the parent body. That creates a dense, brittle altered zone which separates during thermal cycling — structural spalling. The brick can look sound and still shed layers.
Dissolution follows chemistry
The rule is like dissolves like: basic slag attacks acidic refractory and vice versa. Rate depends on how far the slag is from saturation in the refractory oxide — an unsaturated slag is hungry. This is why saturating ladle slag with MgO extends a magnesia lining so effectively.
Erosion finishes the job
Moving metal, gas and particles scour the face. On its own it is modest; combined with dissolution it is not, because each removes the weakened layer the other created and exposes fresh material. The worst wear is always where chemistry, heat and movement meet — converter tuyeres, burner impingement, casting nozzles.
What to do about it
- Specify low, closed porosity — it blocks penetration before it starts.
- Match the refractory to the slag chemistry rather than buying on refractoriness alone.
- Saturate the slag where the process allows it.
- Use non-wetting additives such as carbon where penetration dominates.
- Control local hot spots and turbulence — they are process problems, not material ones.
- Record the wear profile each campaign; it is the best predictor you have.