Soft sensors estimate process variables that are difficult or impossible to measure directly by using mathematical models and available sensor data, e.g., product concentrations. Machine learning-based approaches have become popular for soft sensing tasks. These approaches offer automatic modeling using historical process data but lack basic process information, such as the process topology. This can lead to (1) modeling of correlations instead of causation between process measurements, (2) model deterioration in deployment due to unseen process scenarios, and (3) large data requirements. To overcome these shortcomings, we propose a novel ML modeling approach incorporating the process topology into soft sensor models for improved spatio-temporal modeling. For this, we propose process topology-aware graph neural networks. We combine process topology and sensor data by representing process data in a directed graph and leverage these process graphs to train graph neural networks. Our method demonstrates enhanced model robustness, reduced data requirements, and more intuitive data representations compared to standard black-box machine learning modeling approaches. Overall, this work introduces a new paradigm for soft sensing by directly embedding process information into the data, paving the way for more efficient and reliable digital twin applications.

The fractal structure of aggregates consisting of primary nanoparticles naturally arises during their synthesis. While typically considered to be a fully stochastic process, we suspect long-range interactions, in particular van der Waals forces, to induce an active pull on particles, altering the clustering process. Using an off-grid 3D model, we show that an active pull decreases the density and fractal dimension of formed clusters. These findings could not be reproduced by 2D models, which underestimate screening effects. Additionally, we determined the range within which van der Waals forces dominate the aggregation process.

Conventional fluidization of cohesive powders is challenging due to their strong interparticle forces, often requiring assistance methods. In this study, the hydrodynamics of pulsed and vibrated beds of cohesive Geldart C silica powder (Sauter mean diameter d₃₂=7.9μm) in a 19.2cm diameter column were investigated using X-ray imaging. The results show that low-frequency, moderate-amplitude gas pulsation improves fluidization by disrupting long, persistent gas channels. Higher-frequency pulsation is dampened throughout the bed, resulting in negligible improvement over unassisted systems. When coupled with mechanical vibration, gas pulsation slightly mitigates solid compaction at the bottom section, but the overall flow pattern remains largely unchanged compared to vibration alone. The findings highlight the potential of integrating gas pulsation with other assistance methods to enhance fluidization in practical applications.

Evaluation of the hydrodynamics of opaque multi-phase flows remains a challenging task, with implications for various industrial processes such as chemical processing, pharmaceutical, and mineral processing. Understanding how design and operational variables affect the complex behavior of multi-phase flow systems is essential for optimizing processing conditions and improving efficiency. Radioactive particle tracking (RPT) has been a proven measurement technique to evaluate hydrodynamics in multi-phase flow systems. However, a limitation of the classical RPT technique exists in the assumptions made in the simulation of the count rate received by the detectors in correcting for varying flow-induced fluctuations in the volume fraction of the dispersed phase, often encountered in industrial multi-phase flow systems. In this paper, we introduce a fundamentally novel experimental RPT method that directly uses detected incident photon hit locations for the reconstruction of the three-dimensional radioactive tracer particle position. We argue that this approach is inherently more robust as varying attenuation does not affect the reconstruction. The RPT setup consists of three identical γ-radiation slit collimator detectors that are placed equidistantly at 120° intervals. A subsequent calibration-experimentation procedure is established that allows reconstruction of the tracer particle position with spatial accuracy and precision in the order of 1 mm. We demonstrate the applications of this technique in evaluating hydrodynamics in multi-phase systems by characterizing the flow field of industrial-grade polypropylene reactor powder in a laboratory-scale horizontal stirred bed reactor.

Horizontal stirred bed reactors (HSBRs) are widely used in the commercial production of polypropylene (PP). Despite their commercial significance, a comprehensive understanding of the flow behavior in HSBRs remains elusive, primarily due to the lack of detailed experimental data. This study investigates the influence of operating parameters on the particle flow behavior of two types of PP reactor powder in a laboratory-scale HSBR using X-ray imaging. Our results indicate that the overall flow behavior and phase holdup in the HSBR are dominated by agitation. Moreover, gas injection through the inlet points at the bottom of the HSBR results in spouting behavior, which can lead to reduced gas–solid contacting and, in extreme cases, complete bypass. Finally, the presence of liquid (in this study, isopropyl alcohol) adversely affects the flow behavior of the PP reactor powder due to liquid bridging at the contact points of particles. Powders that comprise particles with relatively small sizes and dense surface morphology are particularly prone to reduced flow behavior when exposed to liquid.

Stirring has been recognized in the literature as a promising technique for facilitating fluidization of cohesive powders, via inputting additional energy to counteract interparticle forces. However, the influence of operating conditions and stirrer configurations on flow behavior remains largely unknown, which impedes the practical implementation of stirred fluidization. Utilizing X-ray imaging, this research demonstrates that stirring enhances fluidization in cohesive micron-silica powder (Sauter mean diameter [Formula presented]) by collapsing the powder packing structure, and transitioning channeling beds into bubbling states. Comb-like configurations featuring fewer stirrers and blades, placed in the bottom region, have shown to be highly effective. Excessive stirring can lead to air pockets and a compacted phase of particles on the column walls, undermining the interaction between particles and stirrers. Additionally, the experiments show that maximizing the sweeping coverage, employing complex asymmetrical configurations, and avoiding tortuous gas pathways are preferable.

In this study, we investigated the wettability and agglomeration characteristics of polymer microspheres coated with low-temperature deposited SiO₂ in a fluidized bed atomic layer deposition (ALD) setup. Surface characterization revealed the presence of a significant amount of deposited Si-OH groups within the first cycles. A drastic decrease in agglomerate size, water contact angle (WCA), and droplet absorption time of the powder was observed when coating was applied. Furthermore, we observed an increase in the amount of Si-OH present on the particle surface with increasing coating cycles, while no significant improvement in water affinity was found after the first coating cycles. Our findings suggest that surface coverage is the primary factor in improving the colloid stability of particles, coated at low temperatures. The low temperature operation of our system introduced a chemical vapor deposition (CVD) component to our coating process, which allowed full surface coverage to be achieved within the first two coating cycles.

In this study, we investigate the correlation between fluidization behavior and flow properties of 10 commercially available cohesive powders through fluidization and rotating drum experiments. The rotating drum was operated at high speeds to aerate the powder, creating flow fields and stress conditions comparable to those in gas–solid fluidized beds. We introduce a fluidization quality index (FQI) computed from X-ray imaging, which addresses the limitations of conventional analyses for assessing fluidization quality of cohesive powders. By analyzing flow patterns, the FQI integrates gas holdup and its temporal variation, providing a reliable measure of fluidization quality. The findings establish a positive correlation between flowability and fluidization quality, demonstrating how flowability measurements can predict the effectiveness of mechanical vibration in enhancing fluidization characteristics. These results suggest that fast, user-friendly flowability assessments in a rotating drum can effectively predict fluidization potential, contributing to process optimization and advancing fluidization studies for cohesive powders.

Horizontal stirred bed reactors are widely used in the commercial manufacturing of polypropylene. However, a comprehensive understanding of the particle dynamics in horizontal stirred bed reactors remains elusive, primarily due to the lack of detailed experimental data. In this work, we studied the influence of operating parameters on the particle flow dynamics in a laboratory-scale horizontal stirred bed reactor using single-photon emission radioactive particle tracking. The results show that the general solids flow behavior is strongly affected by both the agitator rotation speed and reactor fill level. Operation at low rotation speed and low fill level results in solids flow with poor radial and circumferential distribution due to internal bed circulation. On the contrary, at increased rotation speeds and fill levels, solids motion throughout the bed is continuous resulting in excellent solids distribution. The solids circulation was found to increase for both an increase in rotation speed and reactor fill level. The axial dispersion coefficient, on the other hand, shows a linear relation with the rotation speed, but no conclusive relation between the axial dispersion coefficient and the reactor fill level was found.

Stirred gas–solid fluidized bed reactors are commercially employed in polyolefin manufacturing, but the complex gas–solid contacting dynamics pose challenges in design, scale-up, and operation. In this study, the influence of agitation on the fluidization performance of Geldart B particles was investigated experimentally by X-ray imaging and pressure drop measurements and numerically by Computational Fluid Dynamics (CFD) - Discrete Element Method (DEM) - Immersed Boundary Method (IBM). The experimentally obtained minimum fluidization curve and time-averaged pressure drop show good qualitative agreement with the simulation results. Visual observations underscore that an increase in the angular velocity of the agitator results in reduced bubble size and improved bed homogeneity, as further evidenced by reduced pressure fluctuations. Furthermore, the simulations reveal that while the impeller enhances solids agitation, a proper design study is imperative, considering that static immersed bodies, such as the stirrer shaft, can adversely impact solids motion.

In this study, the impact of different vibrational modes on the fluidization characteristics of cohesive micro- and nano-silica powder was examined. Fractional pressure drop, bed expansion measurements, and X-ray imaging were utilized to characterize the fluidization quality. The densities of the emulsion phase at the top and bottom of the column were quantified and compared, providing insights into the solid distribution within the fluidized bed. In the absence of vibration, neither powder could be fluidized within the considered range of superficial gas velocities. Vertical vibration was found to initiate fluidization for both powders. In contrast, elliptical vibration failed to overcome the channelling behavior when fluidizing the micro-powder. For nano-powder, combined channelling and powder compaction occurred when the bed was subjected to elliptical vibration. For the micro-powder, it was observed that bed homogeneity was independent of vertical vibration intensity but improved with increasing superficial gas velocity. For nano-powder, intensifying vertical vibration led to segregation, likely due to agglomerate densification. Furthermore, fractional pressure drop measurements proved to be a strong tool in assessing fluidization quality, providing insights that could not be attained by conventional indicators.

Vibro-assisted fluidization of cohesive micro-silica has been studied by means of X-ray imaging, pressure drop measurements, and off-line determination of the agglomerate size. Pressure drop and bed height development could be explained by observable phenomena taking place in the bed; slugging, channeling, fluidization or densification. It was observed that channeling is the main cause of poor fluidization of the micro-silica, resulting in poor gas-solid contact and little internal mixing. Improvement in fluidization upon starting the mechanical vibration was achieved by disrupting the channels. Agglomerate sizes were found to not significantly change during experiments.

Mechanical vibration has been broadly used to assist fluidization of cohesive powders, because of its capability to disrupt gas channels and agglomerates. However, the improvement reported in literature is mostly deduced from bulk response and ex-situ measurements, whereas the induced fluidization behavior and underlying physics remain largely unexplored. In this work, the fluidization behavior of micron-sized cohesive silica (Sauter mean diameter D₃₂ = 7.9 μm) has been investigated experimentally under vibration of varying conditions. X-ray imaging was carried out to directly capture the temporal evolution of system hydrodynamics, and identify in-situ powder stratification, bubbling and channel formation. The study demonstrates that vibration effectively collapses gas channels, yet facilitates powder stratification and compaction, therefore developing three distinctive flow regions inside the bed with different fluidization states. Consequently, common measurements, such as pressure drop and bed expansion, tend to overestimate the improvement. In addition, increasing frequency, from 10 Hz to 30 Hz, is observed to increase the number of bubbles by 60 %, whereas a large amplitude (e.g., 2 mm) leads to a 10 % compaction in the top flow region.

The fluidization of cohesive powders has been extensively researched over the years. When looking at literature on the fluidization of cohesive particles, one will often find papers concerned with only micro- or only nano-sized powders. It is, however, unclear whether they should be treated differently at all. In this paper, we look at differences and similarities between cohesive powders across the size range of several nanometres to 10s of micrometres. Classification of fluidization behaviour based on particle size was found to be troublesome since cohesive powders form agglomerates and using the properties of these agglomerates introduces new problems. When looking at inter-particle forces, it is found that van der Waals forces dominate across the entire size range that is considered. Furthermore, when looking into agglomeration and modelling thereof, it was found that there is a fundamental difference between the size ranges in the way they agglomerate. Where the transition between the types of agglomeration is located is, however, unknown. Finally, how models are made and agglomerate sizes are measured is currently insufficient to accurately predict or measure their sizes consistently.

The water-resistance of cow-dung has made it a widely used stabiliser in traditional earthen structures in several Asian and African countries. Multiple studies have shown an improvement in water-resistance with the addition of cow-dung, but none provides insight into this behaviour. The present study investigates the water-resistance behaviour of cow-dung stabilised earthen blocks through an extensive experimental programme to identify and characterise the components of cow-dung responsible for its water-resistance. Fresh cow-dung was collected and separated into fibres (>63 μm), medium-sized microbial aggregates (1-63 μm) and small-sized microbial aggregates (0.5-7 μm). Each component was mixed with soil and samples were prepared at different water contents (optimum water content corresponding to the highest dry density and water content higher than optimum) and compacted with 2.5 MPa force to prepare compressed blocks. The water-resistance of these blocks was evaluated through the immersion and modified drip/rain test. It was found that the small-sized microbial aggregates are almost entirely responsible for water-resistance behaviour of cow-dung stabilised earthen blocks. Small-sized microbial aggregates were further characterised by gas chromatography, mercury intrusion porosimetry, N2- BET surface area, zeta potential measurement and electron microscopy. The results indicate that the small-sized microbial aggregates are composed of clay-sized negatively charged particles that are rich in fatty acids. The hydrophobicity of these particles is hypothesised to be responsible for water-resistance behaviour. These insights are further used to produce stabilised blocks that performed at least 30 times better than the unstabilised blocks in both water-resistance tests. The study concludes with practical recommendations for the use of wet cow-dung over dry cow-dung and a reduction of fibre content to increase the water-resistance of earthen blocks.

Properties of powders produced from drying solute-containing droplets arise from the dynamic redistribution of solute during drying. While insights on the dynamic redistribution are instrumental for the rational design of powders and for the optimized operation of equipment such as spray dryers, experimental techniques that allow measuring the spatio-temporal concentration of solute in drying droplets are scarce. In this work, we explore and demonstrate the use of optical coherence tomography (OCT) to measure the spatio-temporal concentration of solute in drying droplets and the development of a solidifying shell at the liquid-air interface, using aqueous droplets of maltodextrin as a model system. This work provides a solid foundation for the use of OCT to quantify the dynamic redistribution of solute and link it to the development of the morphology of the produced particles and agglomerates.

The biotechnological production of poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV) derived from organic waste streams by mixed microbial communities is well established at the pilot-level. However, there is limited research on the recovery of the biopolymer from the microbial biomass, while its impact on product quality and product costs is major. When applying solvent extraction, the choice of solvent has a profound influence on many aspects of the process design. This study provides a framework to perform a systematic solvent screening for PHBV extraction. First, a database was constructed of 35 solvents that were assessed according to six different selection criteria. Then, six solvents were chosen for further experimental analysis, including 1-butanol, 2-butanol, 2-ethyl hexanol (2-EH), dimethyl carbonate (DMC), methyl isobutyl ketone (MIBK), and acetone. The main findings are that the extractions with acetone and DMC obtained the highest yields (91-95%) with reasonably high purities (93-96%), where acetone had a key advantage of the possibility to use water as anti-solvent. Moreover, the results provided new insights in the mechanisms behind PHBV extraction by pointing out that at elevated temperatures the extraction efficiency is less determined by the solvent's solubility parameters and more determined by the solvent size. Although case-specific factors play a role in the final solvent choice, we believe that this study provides a general strategy for the solvent selection process.

Sticking of particles has a tremendous impact on powder-processing industries, especially for hygroscopic amorphous powders. A wide variety of experimental methods has been developed to measure at what combinations of temperature and moisture content material becomes sticky. This review describes, for each method, how so-called stickiness curves are determined. As particle velocity also plays a key role, we classify the methods into static and dynamic stickiness tests. Static stickiness tests have limited particle motion during the conditioning step prior to the measurement. Thus, the obtained information is particularly useful in predicting the long-term behavior of powder during storage or in packaging. Dynamic stickiness tests involve significant particle motion during conditioning and measurement. Stickiness curves strongly depend on particle velocity, and the obtained information is highly relevant to the design and operation of powder production and processing equipment. Virtually all methods determine the onset of stickiness using powder as a starting point. Given the many industrial processes like spray drying that start from a liquid that may become sticky upon drying, future effort should focus on developing test methods that determine the onset of stickiness using a liquid droplet as a starting point.

Atomic layer deposition in a fluidized bed is utilised for the coating of polyamide powders with thin films of silicon dioxide. By varying the number of exposure cycles the amount of deposited material can be precisely controlled, allowing for tailoring wetting, flowing, and melting behaviour of the powder. The favourable process operation characteristics enable a facile envisioning of scalable production in volumes complying with the demands of 3D printing applications.

Unwanted changes in powder flow behavior can unexpectedly occur when a product is exposed to certain conditions of temperature and humidity. This can happen during production, but also during transport or storage. The work reported here demonstrates the novel approach of using an amended powder rheology set-up for measuring and predicting such changes in powder flow behavior. The developed methodology makes it possible to vary in-situ the temperature and the relative humidity of the air to which the product is exposed, thereby mimicking realistic conditions of production or related unit operations. An air flow capable of fluidizing the powder particles is controlled at a specific constant temperature and its relative humidity can be altered while measuring the torque in the fluidized powder bed in real time. The fluidization is necessary for generating a homogeneous introduction of temperature and relative humidity. Results obtained using citric acid and commercial coffee whitener products have proven this methodology to provide both similar and in certain instances dissimilar results compared to the more established methodology such as measuring the vapour adsorption isotherms. These observations are explained. In this way, it can be predicted under which combinations of temperature and humidity a product does or does not become sticky. The main advantages of our approach are that the flow properties are directly assessed, the interpretation of the obtained data is more straightforward and that the measurement times are shortened substantially.