Slide the bulk copper content of a binary Al-Cu alloy and watch the Scheil solidification path, Kou's terminal-slope cracking index, and where that composition falls on the classic Pumphrey & Lyons "lambda" susceptibility curve, all updating live.
Model constants: Tm = 660.3°C, Te = 548.0°C, Ce = 33.0 wt% Cu, k = 0.17.
Low/Moderate/High bands are relative to this model's own computed peak, not an absolute imported threshold.
S. Kou, "A criterion for cracking during solidification," Acta Materialia 88 (2015) 366–374; P.H. Pumphrey & J.V. Lyons, "Cracking during the Casting and Welding of the More Common Binary Aluminium Alloys," J. Inst. Metals 74 (1948) 439. Liquidus/eutectic data and k are simplified, literature-typical values; see disclaimer above.
Solidification cracking happens when nearly solid grains are pulled apart before the last liquid can fill the gaps. Alloys with around 1% copper are the worst, which gives the classic lambda-shaped curve.
This is why higher-solute fillers are used: they move the weld away from the peak.
Pure aluminum freezes over a narrow range, so there's little time for cracks to form. High-copper alloys have enough eutectic liquid to backfill any cracks that open. In between, around 1% Cu, there's a long, steep end to solidification with little liquid left to heal cracks.
Kou's index captures this: a steep T vs. √fs curve means grains need a large temperature drop to bond, while liquid must feed shrinkage through long, narrow channels.
The model assumes a binary Al-Cu alloy, a straight-line liquidus from 660.3 °C to the eutectic (33 wt% Cu, 548 °C), a constant k, and Scheil conditions (full liquid mixing, no solid diffusion). Commercial alloys contain Mg, Si, and Mn, and real cracking also depends on restraint and grain size, so use it for trends, not to rank real alloys.
This calculator is free to use. If you need a multicomponent alloy model, a proprietary composition, or procedure-specific cracking analysis for your program, that's exactly the kind of work Jeffrium takes on.