Zonal modelling of a counter-current spray drying tower

Abstract

A generic multi-zonal modelling methodology for spray drying towers is proposed to provide relatively quick (seconds) simulations on desktop computers for decision making on changing process parameters for automatic process control. A multi-zonal model comprises combinations of all or some of six differing zones: plug-flow, semi-plug-flow and well-mixed zones in either co-or counter-current flow. This approach is demonstrated by predicting the dried powder characteristics of a detergent powder from a pilot-scale counter-current spray drying tower. The types, sizes and locations of the different zones are obtained by a detailed analysis of predictions from a previous 3-dimensional Computational Fluid Dynamics (CFD) simulation of gas and particle flow dynamics and drying kinetics within the tower. The multi-zonal model consists of seven zonesfive in counter-current and two in co-current flow. The trends of the predicted gas temperature profiles are close to those from the CFD results. The particle exit temperature, moisture content and residence times over the full particle size distribution (PSD) range are very close to the CFD values unlike the previous simulated results from our plug-flow model. The outcome clearly demonstrates that the prerequisite to have a sound conceptual model of spray drying towers, such as the multi-zonal model developed here, is a detailed knowledge of the gas and particle flow fields within the tower.

The Use of Optimized Restitution Coefficients to Improve Residence Time Prediction in Computational Fluid Dynamics-Discrete Parcel Method Models for Counter-Current Spray Dryers

“…This improvement is ascribed to a better prediction of the vortex momentum flow and the added functionality to estimate wall restitution coefficients. Contrary to the air velocity profiles used by Ali, our previous work used as basis for the airflow the swirl intensity Ω to estimate the effect of wall roughness. A calibration based on Ω leads to a better overall prediction of the momentum flow and describes better the influence of wall conditions in the single-phase flow .…”

“…An important difference between both single-phase models is the different turbulent model used for the wall function. Contrary to the standard wall function used by Ali, our work suggested the use of a scalable wall function that provides more stable results, in terms of meshing, for the turbulent components . This wall function can play a key role on the dispersion of the small particles (low Stokes number) and influence on the variance observed.…”

“…The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.iecr.1c02415. Sensitivity analysis of time-step and forces involved in the discrete phase, sensitivity analysis of restitution coefficients, detailed description of framework for reducing the error of the mean RT and RTD curve, experimental results for RTD, analysis of mass losses, analysis of dimensionless groups, optimal restitution coefficients for each case, plots comparing the two methodologies and different pairs of restitution coefficients, air velocity profiles comparing refs with , Péclet numbers for a PFR model, Pearson correlation coefficients for the CSTR+PFR model, and errors in the equivalent reactor networks (PDF) …”

Computational Fluid Dynamics (CFD) Modeling of Swirling Flows in Industrial Counter-Current Spray-Drying Towers under Fouling Conditions

“…Previous studies for swirl towers used standard wall functions . However, standard wall functions are not robust under strong pressure gradients and the turbulent components are highly dependent on the y + .…”

“…The validation of CFD models for spray drying towers has traditionally been performed based on the study of velocity profiles. − However, the momentum distribution is not regular along the radius, concentrating the mass and momentum at large radial positions. Fluid momentum governs the particle momentum, concentrating them near the wall.…”