Title:Effective viscosity closures for dense suspensions in CSP systems via lubrication-enhanced DNS and numerical viscometry

Abstract:Dense particle suspensions are promising candidates for next-generation Concentrated Solar Power (CSP) receivers, enabling operating temperatures above 800 degrees C. However, accurate modeling of the rheological behavior of granular flows is essential for reliable computational fluid dynamics (CFD) simulations. In this study, we develop and assess numerical methodologies for simulating dense suspensions pertinent to CSP applications. Our computational framework is based on Direct Numerical Simulation (DNS), augmented by lubrication force models to resolve detailed particle-particle and particle-wall interactions at volume fractions exceeding 50\%. We conducted a systematic series of simulations across a range of volume fractions to establish a robust reference dataset. Validation was performed via a numerical viscometer configuration, permitting direct comparison with theoretical predictions and established benchmark results. Subsequently, the viscometer arrangement was generalized to a periodic cubic domain, serving as a representative volume element for CSP systems. Within this framework, effective viscosities were quantified independently through wall force measurements and energy dissipation analysis. The close agreement between these two approaches substantiates the reliability of the results. Based on these findings, effective viscosity tables were constructed and fitted using polynomial and piecewise-smooth approximations. These high-accuracy closure relations are suitable for incorporation into large-scale, non-Newtonian CFD models for CSP plant design and analysis.