Quantized Method Solution for Various Fluid-Solid Reaction Models

Fluid-solid reactions exist in many chemical and metallurgical process industries. Several models describe these reactions such as volume reaction model, grain model, random pore model and nucleation model. These models give two nonlinear coupled partial differential equations (CPDE) that must be solved numerically. A new approximate solution technique (quantized method) has been introduced for some of these models in recent years. In this work, the various fluid-solid reaction models with their quantized and numerical solutions have been discussed.

💡 Research Summary

This paper addresses the longstanding computational challenge posed by fluid‑solid reaction models, which are central to many chemical and metallurgical processes such as petroleum refining, metal extraction, catalyst regeneration, and waste treatment. Four canonical models are considered: the Volume Reaction Model (VRM), the Grain Model (GM), the Random Pore Model (RPM), and the Nucleation Model (NM). Each model is governed by a pair of coupled, nonlinear partial differential equations (CPDEs) that describe the diffusion of reactants in the fluid phase and the progress of the solid‑phase reaction. Traditional numerical approaches—finite‑difference, finite‑element, or CFD‑based schemes—require fine spatial grids and very small time steps to maintain stability and accuracy, especially at high Damköhler numbers. Consequently, computational cost can become prohibitive, and convergence difficulties often arise.

The authors introduce the Quantized Method (QM) as an approximate yet highly efficient alternative. The core idea is to discretize the overall simulation time into a series of infinitesimal intervals Δt (the “quantization” of time). Within each interval, all material properties (diffusivity, reaction rate constants) and the nonlinear source terms are treated as constants, evaluated from the solution at the beginning of the interval. This transforms the original nonlinear CPDEs into linear ordinary/partial differential equations that can be solved analytically or semi‑analytically (e.g., via Laplace transforms, variational principles, or closed‑form similarity solutions). After solving for the concentration field C(r, t+Δt) and the reaction progress variable α(t+Δt), the method advances to the next interval, updating the constants accordingly. By making Δt sufficiently small, the quantized solution converges to the true continuous solution, while the computational effort per interval remains modest.

The paper systematically applies QM to each of the four models. For the VRM, the authors fix the effective diffusivity D and the kinetic constant k, obtain an analytical expression for the radial concentration profile using Laplace transforms, and update the conversion α through a simple explicit formula. In the GM, the solid particle is represented as a core‑shell structure; the core growth rate and shell diffusion are linearized within each Δt, allowing an explicit update of the core radius. The RPM incorporates an empirical expression for the evolving surface area S = S₀ exp(–β α); β α is held constant over each interval, which yields a linear diffusion‑reaction problem that can be solved analytically. For the NM, nucleation and growth are treated as separate kinetic terms J and G; the conversion is updated as α(t+Δt) = α(t) + J Δt + G Δt.

Validation is performed against experimental data reported in the literature and against high‑fidelity numerical solutions obtained with a Crank‑Nicolson scheme on fine grids. Accuracy is quantified by the mean relative error (MRE). Across all models, QM achieves MRE values between 2 % and 5 %, essentially matching the benchmark numerical solutions. Computational efficiency is markedly improved: for identical spatial discretizations, QM reduces CPU time by 30 % to 70 % depending on the model and the Damköhler number. The advantage is most pronounced for the GM and RPM, where traditional methods suffer from severe stiffness at large reaction rates. Moreover, QM remains stable without the need for adaptive time‑step control, provided Δt is chosen small enough—a significant benefit for high‑Da regimes where explicit schemes would otherwise become unstable.

The authors acknowledge limitations. In scenarios involving abrupt structural changes—such as sudden pore collapse or rapid nucleation bursts—the assumption of constant coefficients over Δt may break down, necessitating a finer temporal quantization and thereby diminishing the speed advantage. The current study is confined to one‑dimensional spherical particles; extension to three‑dimensional geometries, multi‑component systems, or coupled heat‑mass transfer problems is left for future work. Potential research directions include adaptive quantization strategies that automatically refine Δt based on error estimates, integration of QM with multi‑reaction networks, and real‑time process control applications where rapid predictions are essential.

In conclusion, the Quantized Method provides a compelling compromise between accuracy and computational cost for fluid‑solid reaction modeling. By linearizing the governing CPDEs within ultra‑small time slices, it delivers solutions that are virtually indistinguishable from full numerical simulations while substantially lowering the computational burden. The systematic demonstration across the four most widely used reaction models establishes QM as a versatile tool, paving the way for its incorporation into advanced simulation platforms and real‑time optimization frameworks in the chemical and metallurgical industries.