Particle size distribution real-time monitoring for air classification with dense-phase particle flow

Air classifiers separate particles in a gas stream using aerodynamic forces, with performance governed by cut size and separation sharpness. Real-time knowledge of the particle size distribution (PSD) of the product stream is important for stabilising operation, achieving target cuts, reducing energy use, and maintaining product quality. Conventional monitoring relies on intermittent off-line sampling or optical methods, which are difficult to apply in dense, opaque gas-solid flows because of time delay, sampling bias, fouling, and limited optical access. This study proposes an in-situ probe-based vibration sensing method for real-time PSD approximation in dense-phase dry particulate flow relevant to air classification. A probe inserted into the particle stream is excited by particle impacts, and the measured vibration spectrum is mapped primarily to the median size D 50 and percentile spread ratio D 75 / D 25 , while the Rosin–Rammler parameters ( P 50 , n ) are used secondarily for illustrative representation of the cumulative PSD. A single-impact excitation test and dense-phase spectral analysis showed that robust PSD-sensitive features are best extracted from intermediate response bands of the probe–sensor system. Ratio-based spectral features were used and shown capable of suppressing the influence of mass flow rate. Bench-top experiments on five size classes of poly(methyl methacrylate) beads, assessed across five repeat groups, showed that a calibration based on only the two endpoint samples could predict the intermediate size classes with promising accuracy. The mean absolute percentage errors were 7.58% for D 50 and 0.85% for D 75 / D 25 . Additional tests on six size classes of aluminosilicate ceramic particles showed that the same workflow remained effective for a second material class, although the calibration remained material-specific. For this second material, the mean absolute percentage errors were 5.70% for D 50 and 0.81% for D 75 / D 25 .

Authors

Institutions

Publication Details

Journal
Minerals Engineering
Published
2026-10-07
DOI
https://doi.org/10.1016/j.mineng.2026.110920
Primary Topic
Granular flow and fluidized beds
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Particle size distribution real-time monitoring for air classification with dense-phase particle flow

Difan Tang, Richmond Komla Asamoah, Lin Yang, Eric Jing Hu et al.
Minerals Engineering
Granular flow and fluidized beds
article

Particle size distribution real-time monitoring for air classification with dense-phase particle flow

Difan Tang, Richmond Komla Asamoah, Lin Yang, Eric Jing Hu, Lei Chen, Zhiwei Sun
article en

Abstract

Air classifiers separate particles in a gas stream using aerodynamic forces, with performance governed by cut size and separation sharpness. Real-time knowledge of the particle size distribution (PSD) of the product stream is important for stabilising operation, achieving target cuts, reducing energy use, and maintaining product quality. Conventional monitoring relies on intermittent off-line sampling or optical methods, which are difficult to apply in dense, opaque gas-solid flows because of time delay, sampling bias, fouling, and limited optical access. This study proposes an in-situ probe-based vibration sensing method for real-time PSD approximation in dense-phase dry particulate flow relevant to air classification. A probe inserted into the particle stream is excited by particle impacts, and the measured vibration spectrum is mapped primarily to the median size D 50 and percentile spread ratio D 75 / D 25 , while the Rosin–Rammler parameters ( P 50 , n ) are used secondarily for illustrative representation of the cumulative PSD. A single-impact excitation test and dense-phase spectral analysis showed that robust PSD-sensitive features are best extracted from intermediate response bands of the probe–sensor system. Ratio-based spectral features were used and shown capable of suppressing the influence of mass flow rate. Bench-top experiments on five size classes of poly(methyl methacrylate) beads, assessed across five repeat groups, showed that a calibration based on only the two endpoint samples could predict the intermediate size classes with promising accuracy. The mean absolute percentage errors were 7.58% for D 50 and 0.85% for D 75 / D 25 . Additional tests on six size classes of aluminosilicate ceramic particles showed that the same workflow remained effective for a second material class, although the calibration remained material-specific. For this second material, the mean absolute percentage errors were 5.70% for D 50 and 0.81% for D 75 / D 25 .

Minerals EngineeringVol. 250
Australian Research Council (AU), University of South Australia (AU), ARC Centre of Excellence for Enabling Eco-Efficient Beneficiation of Minerals (AU), Adelaide University (AU), The University of Adelaide (AU)
Openalex Percentile: Top 18%
Granular flow and fluidized beds
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.