Particle dynamics investigation by means of shadow imaging inside an air separator
Introduction
A zigzag air separator is an apparatus used for the separation of a solid dispersed particle phase into its fine and coarse fractions. It consists out of a cascade of connected vertical channels, alternatingly slanting left and right. Air is blown from the bottom, leading naturally to a transient flow behaviour in the channel. From a side opening the material to be separated is fed into the channel and, depending on the operating conditions of the channel (the mass flows of fluid and solid), the coarse fractions fall downwards while the fine fractions are lifted with the airstream to the adjacent cyclone. The characteristic settling velocity determined by parameters such as size, density, and shape of each particle thereby acts as separation property. The term “zigzag” refers to the geometrical pattern used to cascade the single duct elements in this apparatus, where a change of direction of the main flow under a certain angle and a repeated cross-flow separation in these elbow elements is achieved. The multistage approach affects the quality of the separation in a beneficial manner: Each particle is exposed to a separation stage several times during its flight through the apparatus. Thus, the overall separation sharpness can be considered high, whereas the amount of particles assigned to the incorrect fraction is small.
Air separators are typically used to separate particles in the range of 200 µm to a few millimetres and can realize high volume and mass throughputs. Often, they are part of complex, connected solids production systems where they are embedded into the interacting structure of the entire dynamic system. Applications can be found e.g. in the tea and tobacco industry as well as for the separation of cable scrap (Furchner and Zampini, 2012), municipal solid waste (Colon, 1976, Senden, 1980, Howell, 1992), shredded car bodies (Kox and Senden, 1982), building rubble (Tomas and Gröger, 1999, Tomas, 2004) and shredded lithium ion batteries (Georgi-Maschler et al., 2012).
Several investigations regarding the separation experiments have been published since the patent application in the 1930s by Stebbins (1930). Kaiser (1963) showed the general working principle and derived a first separation model based on a series connection of the separations in each stage. The geometry of the separator was first investigated by Fastov et al. (1975) where they preferred an inclination angle of 90° and a segment height-to-width ratio ranging from 1:1 to 3:2 based on separation experiments with granules. Senden (1979) researched the separation of municipal solid waste in paper and plastic fractions and identified an important influence of the impacts between particles and between particles and walls on the separator performance in addition to the force balance and the occurring flow turbulence.
Modelling approaches to describe the performance of the separator have been attempted for instance by Senden, 1979, Rosenbrand, 1986, Schubert et al., 1986, Tomas and Gröger, 1999, and Mann (2016). However, these models lack predictive character and can only be used to describe process modifications in a very limited and specific parameter range. In order to converge to the final aim of deriving a widely applicable modelling approach, better understanding of the process dynamics and involved process parameters is highly desirable.
Besides investigations of the single-phase fluid flow in such apparatus (see e.g. Gillandt et al., 1996, Roloff et al., 2015), attempts to characterise the particle dynamics can only be very rarely found in the literature, as the required methodologies are expansive and elaborate. Vesilind and Henrikson (1981) took few photographs of plastic and aluminium pieces which were given to the zigzag channel and derived local mass loadings. Kaiser, 1963, Senden, 1980 showed two long-exposure photographs of particles inside the zigzag channel giving a qualitative impression of the particle trajectories. Gillandt et al. (1996) interpreted video recordings of the particle motions at low and high mass loading and gave a qualitative description.
The current research aims to derive quantitative insights into the dynamics of the particulate phase as well as into the characteristic features of the global separation process. Therefore, this paper introduces the development of a suitable multi-camera shadow particle imaging system (also called shadowgraphy system) to the environment of a pilot-scale zigzag air separator. Shadowgraphy is an image-based particle detection system, where a camera and a light source are installed face to face and particles occurring in between produce a shadow image at the camera sensor. The shadow images are processed by specific algorithms and thus allow for counting, sizing and in special configuration also for conducting velocimetry of the particulate phase (see e.g. Kim and Kim, 1994, Kerst et al., 2017). The applied shadowgraphy technique with four simultaneously used cameras allows for spatially as well as temporally resolved analysis of quantitative particle dynamics inside the zigzag channel. The paper further explains the calibration procedure required for estimating quantitative particle dynamics and discusses possible error sources as well as their mitigation. Finally, resulting particle velocity distributions along the measurement sections as well as their dependencies on the location inside the channel, the mass loading as well as the air flow rate are documented and discussed. The presented data provide novel and quantitative insights into particle motion in such a zigzag separator which can be used as validation basis for complex multiphase CFD-DEM models or for the further development of dynamic process models, which could be integrated into dynamic flowsheet simulation tools to facilitate design and optimization of connected process systems (Skorych et al., 2017).
Section snippets
Pilot-scale zigzag air separator
The pilot-scale air separator used in this research consists of the zigzag channel featuring four exchangeable channel modules aligned vertically and housing two segments each, which are connected under a prescribed inclination angle as can be understood from the sketch in Fig. 1. A controllable blower (1) drives the air in a closed circuit through the apparatus, i.e. air flows through the inflow pipe (2) and then from bottom to top through the zigzag channel (3), through an aero-cyclone (4), a
Separation experiments
A total of 60 separation experiments with the pilot-scale zigzag air separator were carried out during the campaign with varying solid and fluid mass streams. For the solid particles, glass spheres were chosen which were composed out of six different monomodal distributed fractions in different size classes ranging from approx. 1 mm to 4 mm. The resulting size distribution was measured with a Camsizer (Retsch Technology, Haan, Germany) and is presented in Fig. 3. Further, the Camsizer delivered
System description and Installation
In order to monitor the particulate phase inside the zigzag channel, a multi-camera shadowgraphy system was installed at the facility. As for most optical systems, its main advantage is its non-intrusive character which can provide an undisturbed insight into the characteristics of the observed process. Its severest disadvantage is its limited applicability to processes involving very high particle concentrations, as the detection algorithm may fail to accurately handle overlapping particle
Results and discussion
Using the double-frame shadowgraphy approach, it was possible to detect the projected particle tracks and their velocities in the x-z-plane of the observed volumes. These velocity data depending on the applied operating parameters will be presented and discussed below.
Conclusions
Multi-camera shadowgraphy measurements have been successfully conducted at a pilot-scale zigzag air separator. Four simultaneously working camera systems at different heights of the zigzag channel were adapted to the conditions expected during classifier operation, including appropriate calibration and systematic error correction. This enabled simultaneous recordings of glass particle populations featuring six different size classes at four zigzag positions with variable particle mass loadings
Acknowledgements
We are very grateful to Réka Podráczki for her dedicated support of the measurement campaign.
Funding
This work is part of the German priority program SPP 1679 “Dynamische Simulation vernetzter Feststoffprozesse” (“Dynamic simulation of interconnected solids processes”) that is financially supported by the Deutsche Forschungsgemeinschaft (DFG), Germany.
References (29)
Recycling of paper
Conserv. Recycling
(1976)- et al.
Drop size distribution in sprays by image processing
Comput. Chem. Eng.
(1990) - et al.
Development of a recycling process for li-ion batteries
J. Power Sources
(2012) A ten year review of plastics recycling
Waste Minimization
(1992)- et al.
CFD-DEM simulations of a fluidized bed crystallizer
Chem. Eng. Sci.
(2017) - et al.
Theory and practical application of multicomponent separation in solid waste technology, illustrated by a process for recovering materials from car shredder waste
Resour. Conserv.
(1982) - et al.
Novel system for dynamic flowsheet simulation of solids processes
Powder Technol.
(2017) - et al.
Effect of feed rate on air classifier performance
Resour. Conserv.
(1981) - et al.
Image processing system to analyze droplet distributions in sprays
Opt. Eng.
(1985) - et al.
Experimental determination of droplet collision rates in turbulence
New J. Phys.
(2013)