Two metallurgical treatments — grain refinement and silicon modification — have the potential to measurably improve the mechanical properties of aluminium castings at very low addition rates and modest cost. Together, they can increase elongation, improve fatigue strength, and produce a more uniform microstructure that responds more consistently to heat treatment. Yet in many small and medium aluminium foundries, they are either not practised at all, applied inconsistently, or misunderstood in terms of what they actually do and when they are beneficial.
Grain Refinement — What It Is and Why It Matters
Aluminium castings solidify by nucleating solid crystals — grains — from the liquid melt. In the absence of deliberate grain refinement, these grains grow large and coarse. A coarse grain structure means fewer grain boundaries per unit volume, and grain boundaries are where deformation mechanisms are arrested — they are the barriers that give polycrystalline metals their ductility and fatigue resistance. A coarse, unrefined microstructure produces castings with lower elongation, more variable mechanical properties from casting to casting, and greater sensitivity to the location and orientation of any residual porosity.
Grain refinement works by introducing a large number of artificial nucleation sites into the melt just before pouring. Each nucleation site spawns a grain during solidification, multiplying the number of grains per unit volume and reducing average grain size dramatically. The result is a finer, more uniform microstructure with more grain boundaries, better ductility, more consistent properties, and improved response to heat treatment. The standard grain refiners for aluminium alloys are Al-Ti-B master alloys — typically added as TIBORAL rod or NUCLEANT waffle to the melt at low addition rates of 0.1 to 0.3 percent.
Why Grain Refinement Underperforms in Practice
Three errors consistently undermine grain refiner effectiveness in production foundries. First, the addition rate is insufficient — operators add grain refiner by estimate rather than by weight, and the estimate is typically low. Second, the temperature at addition is incorrect — grain refiner added at the wrong temperature dissolves differently and the nucleant particles may be less effective. Third, and most critically, fading — the nucleant effect diminishes with time after addition, and if the metal is held for an extended period before pouring, grain size returns toward the unrefined condition.
The practical requirement is consistent: weigh the grain refiner addition by calculation from the melt weight, add at the specified temperature, stir adequately to distribute the addition, and pour within the specified time window after addition. These are not complex requirements — but they demand process discipline that is often absent in high-pressure production environments.
Silicon Modification — Transforming the Eutectic Structure
Modification is a separate treatment applicable specifically to aluminium-silicon alloys. In unmodified Al-Si alloys, the eutectic silicon solidifies as coarse, acicular plates — elongated needle-like particles with sharp tips that act as stress concentrators in the microstructure. Under tensile or fatigue loading, cracks initiate at the tips of these silicon plates and propagate through the aluminium matrix. The result is low elongation and poor fatigue resistance — the alloy fails in a brittle manner even though aluminium itself is a ductile metal.
Modification transforms the eutectic silicon from coarse plates into fine, fibrous, rounded particles distributed uniformly through the aluminium matrix. The sharp tips disappear. The stress concentration sites are eliminated. The result is a dramatic improvement in elongation — often doubling it in LM25 — and a corresponding improvement in fatigue strength. The two practical modifiers are sodium and strontium. Sodium modification is effective but transient — the effect fades within 30 to 45 minutes. Strontium modification using Al-Sr master alloys is more durable, persisting for several hours, and is compatible with a wider range of alloys and downstream processes.
The Interaction Between Modification and Heat Treatment
The interaction between silicon modification and T6 heat treatment is significant and often overlooked. In the T6 condition, the solution treatment step causes the eutectic silicon particles to spheroidise and coarsen regardless of their initial morphology — the microstructural difference between modified and unmodified castings becomes less pronounced after full T6 treatment. This means that for components specified in the T6 condition, the primary benefit of modification is in the as-cast condition before heat treatment. For components in the as-cast or T5 condition, modification delivers its largest property improvement and is most worth the process discipline it requires.
Combining Both Treatments
Grain refinement and modification are complementary and fully compatible — they address different phases of the microstructure and can be applied sequentially in the same melt treatment sequence. The standard practice for quality-critical LM25 gravity die castings is to apply grain refiner first, allow adequate time for distribution, then apply strontium modifier, degas, and pour within the specified time window.
The combination of fine grain structure and modified eutectic silicon consistently produces the highest elongation and most uniform mechanical properties achievable in gravity die cast LM25 — measurably superior to either treatment applied alone. For foundries currently producing LM25 T6 structural castings without systematic grain refinement and modification, this is one of the highest-return metallurgical improvements available without any capital investment.