Chemistry-Based Solidification Model
Traditional casting simulation relies on tabular material properties (solidus and liquidus temperatures, latent heat, and density) provided by the user. These inputs work well for standard alloy grades but tell engineers nothing about how microstructure varies across the casting, how mechanical properties relate to local solidification conditions, or how alloy composition changes could improve performance.
The Chemistry-Based Solidification Model in FLOW-3D CAST takes a fundamentally different approach. Rather than requiring users to supply material properties, it computes the full solidification path from the alloy’s chemical composition, accounting for nucleation, segregation, and local cooling conditions at every point in the casting. The result is a simulation that predicts not just where metal solidifies, but what microstructure forms, what mechanical properties result, and where porosity is most likely, all as a direct output of the casting process.
What the Model Predicts
Starting from alloy chemical composition, the Chemistry-Based Solidification Model in FLOW-3D CAST computes:
- Solidification path (solid fraction vs. temperature) based on nucleation, segregation, and cooling conditions, without requiring user-supplied tabular data
- All temperature-dependent material properties including latent heat, thermal conductivity, heat capacity, density, and viscosity, derived directly from the chemistry
- Microstructure features including secondary dendrite arm spacing (SDAS) and grain size, based on local composition and cooling rate
- Macro-segregation due to diffusion and advection of solute elements in the liquid phase
- As-cast and heat-treated mechanical properties including yield strength, tensile strength, and elongation, based on empirical relationships between microstructure and mechanical performance (AlSi- and AlCu-based alloys)
- Microporosity using the dimensionless Niyama criterion, which is enabled by the chemistry-based solidification path
- Resolved shrinkage porosity with improved visualization of leak paths, in the EXODUS output format
The following examples show predicted elongation for an AlSi9Cu3(Fe) alloy casting, illustrating how the model maps mechanical properties and microstructure spatially across the part.
Full and Simplified Models
The Chemistry-Based Solidification Model in FLOW-3D CAST is available in two modes to balance accuracy with simulation efficiency.
Full Model
The Full Model tracks chemical composition and phase evolution in every computational cell throughout the simulation. This provides the most accurate prediction of macro-segregation, where solute redistribution due to convection and diffusion varies significantly across the casting. The Full Model requires more computation time and memory than the tabular approach, typically two to three times longer, and is recommended when macro-segregation is a primary concern.
Simplified Model
The Simplified Model pre-computes the solidification path and material properties in the preprocessor based on the alloy composition and chosen solute redistribution model, then applies them uniformly across the simulation. This gives users the full benefit of chemistry-based microstructure and mechanical property prediction with no additional runtime or memory cost compared to tabular input. For most thin-walled castings and fast-cooled parts where macro-segregation is not significant, the Simplified Model is the recommended choice.
The two modes can be combined within a workflow: a restart simulation can transition from the Full Model to the Simplified Model or vice versa, giving engineers the flexibility to use the most appropriate approach at each stage of the simulation.
Supported Alloys
The Chemistry-Based Solidification Model currently supports:
- Aluminum-silicon alloys (AlSi-based), covering the majority of commercial AlSi grades including A356 and similar alloys
- Aluminum-copper alloys (AlCu-based), including A206 and similar grades
- Carbon and low alloy steels (CLAS), with alloying elements C, Si, Mn, P, S, Cr, Mo, Ni, and Cu, using either equilibrium or Gulliver-Scheil solute redistribution
For aluminum alloys, the effect of grain refiners (Ti and Ti-B additions) is included in the model, assuming optimum addition for maximum grain refining effect.
Validation
The aluminum-silicon and aluminum-copper alloy solidification model has been experimentally validated against A356 and A206 commercial ingots in collaboration with The University of Alabama at Birmingham. The results are detailed in a publication that received the Best Paper Award from the 2020 American Foundry Society Aluminum and Light Metals Division, confirming the accuracy of the microstructure and mechanical property predictions against measured experimental data.
Applications
The Chemistry-Based Solidification Model is used in FLOW-3D CAST to:
- Predict spatial variation in microstructure (SDAS, grain size) across a casting without physical sectioning
- Map as-cast and heat-treated mechanical properties to identify weak regions before physical testing (AlSi- and AlCu-based alloys)
- Evaluate the effect of alloy composition changes on microstructure and properties
- Predict microporosity using the dimensionless Niyama criterion, enabled by accurate solidification path computation
- Visualize resolved shrinkage porosity and potential leak paths in structural castings
- Optimize process parameters (pouring temperature, cooling rate, die temperature) for target mechanical property distributions (AlSi- and AlCu-based alloys)
- Support casting qualification by demonstrating predicted mechanical properties meet design requirements (AlSi- and AlCu-based alloys)