When a material changes phase—such as ice melting into water, or molten metal solidifying in a 3D printer—the transition does not happen instantaneously everywhere. Instead, a moving boundary separates the solid and liquid phases.
Modeling this computationally is notoriously difficult. Traditional approaches try to track the exact position of this sharp interface over time (known as the Stefan problem). However, explicitly tracking a complex, twisting boundary in three dimensions requires intense computational overhead and often fails when the interface deforms rapidly, splits, or merges.
Furthermore, phase transitions involve a massive energy discontinuity: the latent heat of fusion. As noted by S.L. Birla et al. in their work on microwave thawing (Development of Novel Microwave Cooking Model for Not-Ready-to Eat Foods), failing to accurately capture these latent heat effects can cause simulations to miss critical phenomena like thermal runaway. Similarly, Kiumars Aryana et al. demonstrated in Toward Accurate Thermal Modeling of Phase Change Material‐Based Photonic Devices that omitting the enthalpy of fusion leads to massive overestimations of temperature rises and phase volumes.
The enthalpy-porosity method solves these challenges by transforming a discrete topological tracking problem into a continuous, physics-based material property problem.
The Intuition of the Mushy Zone
How do we avoid tracking a sharp, moving interface? We change our perspective: instead of viewing the transition as a discrete, zero-thickness boundary, we acknowledge that real-world materials often melt over a temperature range, creating an intermediate state where solid and liquid coexist. This region is called the mushy zone.
In Thermal Modeling of Laser Powder Bed Fusion during Printing on Temperature-Unstable Materials Considering Local Sintering, Felix Frölich highlights that simplifying simulations into purely discrete phases ignores these critical intermediate states (such as a sintering phase), which possess distinct physical and thermal characteristics.
By treating the boundary as a continuous, volumetric zone rather than a sharp line, we can solve a single set of governing equations across the entire domain without ever needing to know the exact location of the moving interface.
The Enthalpy Formulation: Capturing Energy
Instead of tracking temperature alone and trying to patch a boundary condition at the interface, the enthalpy method reformulates the energy equation using total enthalpy ($H$), which is the sum of sensible heat ($h$) and latent heat ($\Delta H$):
$$H = h + \Delta H = \int C_p dT + f_L L$$
where $f_L$ is the liquid fraction (ranging from 0 for a pure solid to 1 for a pure liquid), and $L$ is the latent heat of fusion.
This formulation allows the simulation to naturally "absorb" or "release" energy at the phase change temperature. When the material reaches its melting point, the temperature stops rising, and the incoming energy is instead funneled into increasing $f_L$.
As Mohammad Golzarijalal et al. showed in The Effect of Freezing and Thawing on Mozzarella Cheese: Insights from Industrial-Scale Experiments and Mathematical and Digital Analysis, this enthalpy-based tracking seamlessly handles large-scale industrial phase transitions where internal thermal gradients are highly complex. When modeling semi-crystalline polymers, Felix Frölich et al. in A submodeling approach for efficient prediction of local temperature profiles in component-scale additive manufacturing similarly emphasize that integrating this released energy is vital for maintaining simulation accuracy during cooling.
The Porosity Trick: Coupling Fluid Flow
Melting isn't just about heat conduction; it also involves fluid dynamics. When a material melts, the liquid phase can flow due to natural convection or external forces, which drastically alters heat transfer rates. How do we simulate fluid flow in the liquid zone while forcing the velocity to zero in the solid zone, all within a single fixed mesh?
The enthalpy-porosity method solves this elegantly by treating the mushy zone as a pseudo-porous medium. The porosity of each computational cell is defined exactly by its liquid fraction ($f_L$).
- In a fully liquid cell ($f_L = 1$), the porosity is maximum, and the fluid flows freely according to standard Navier-Stokes equations.
- In a fully solid cell ($f_L = 0$), the porosity is zero, meaning no fluid can pass through.
To enforce this physically, the method introduces a Darcy-like momentum sink term ($S$) into the fluid flow equations:
$$S = - \frac{C (1 - f_L)^2}{f_L^3 + q} \mathbf{v}$$
where $\mathbf{v}$ is the velocity vector, $C$ is a large constant representing the mushy zone morphology, and $q$ is a small number to prevent division by zero.
As a cell solidifies and $f_L \to 0$, this sink term approaches infinity, effectively overriding the momentum equation and forcing the fluid velocity ($\mathbf{v}$) to zero. As explained by Ying Xu et al. in Heat transfer characteristics of cascade phase change energy storage composite pipeline, this allows for the simultaneous calculation of heat transfer and fluid flow across the entire material domain.
Broad Applicability Across Scales
Because it avoids the geometric bottleneck of interface tracking, the enthalpy-porosity method is incredibly robust and scalable. It has become the workhorse technique across vastly different domains:
- Thermal Energy Storage: Kailiang Huang et al. used it in Macro-Encapsulated PCM Cylinder Module Based on Paraffin and Float Stones to simulate transient melting behaviors in paraffin composites, while Chunwei Zhang et al. applied it in Numerical Analysis of Heat Transfer Behaviours of Melting Process for Ice Thermal Storage Based on Various Heat Source Configurations to optimize heat source geometries.
- Metallurgy and Manufacturing: In heavy industrial processes like steel casting, Guifang Zhang et al. leveraged the technique in Numerical Simulation of Fluid Flow, Solidification, and Solute Distribution in Billets under Combined Mold and Final Electromagnetic Stirring to model how electromagnetic stirring alters solute distribution within the coexisting solid-liquid zone. Similarly, Hong Nam Nguyen et al. used it in Simulation and Experimental Study on the Steady Conduction Solution for Continuous Rheo-Casting for A356 Alloy to capture velocity damping during continuous alloy casting.
While enthalpy-porosity handles macroscopic transport, related implicit methods like the phase field method—adapted by Theodore Kim et al. in Visual simulation of ice crystal growth—rely on similar continuous phase variables to capture microstructural evolution like dendritic ice growth without explicit tracking.
By unifying the thermal effects of latent heat and the physical constraints of fluid flow into continuous variables, the enthalpy-porosity method makes the simulation of complex phase-change phenomena computationally tractable, stable, and highly accurate.
Key Foundational Work
Development of Novel Microwave Cooking Model for Not-Ready-to Eat Foods (2009) — contributed 1 concept elements to this direction.
Visual simulation of ice crystal growth (2003) — contributed 1 concept elements to this direction.