Tracer percentage prediction of dive reflex samplers S. Bensmann a , E. Lockow b , P. Walzel a, ⁎, C. Weihs c a Laboratory of Mechanical Process Design, Department of Biochemical & Chemical Engineering, TU Dortmund University, 44221 Dortmund, Germany b Business and Social Statistics, Department of Statistics, TU Dortmund University, 44221 Dortmund, Germany c Computational Statistics, Department of Statistics, TU Dortmund University, 44221 Dortmund, Germany abstract article info Article history: Received 24 November 2009 Received in revised form 21 November 2010 Accepted 4 December 2010 Available online 14 December 2010 Keywords: Bulk solids Otical probe measurement Solids mass fraction Statistics Instead of the frequently applied monochromatic light probes a whie light fibre optic system was employed at the Laboratory of Mechanical Process Design, TU Dortmund University, in order to exploit the color in formations for concentration measurements within bulk solid. The system is applied to obtain local particle concentrations of blue- and red-colored quartz sand within the bed of a rotary drum. 16 solid mixtures with one or two particle sizes from 100 μm to 2000 μm and different species concentration were analyzed and the relationship between probe measurement values and red sand content was determined by statistical regression methods. After transformation of the data, linear models were found to derive the red sand content from given measurement values. Based thereupon, an all-purpose scheme for mono- and bi-disperse solid mixtures was developed and verified in an example with a mean error of 5%. © 2010 Elsevier B.V. All rights reserved. 1. Introduction Rotary drums are frequently applied for thermal treatment of bulk materials. Different inserts may help to mix the particles and to provide a better contact between the gas and the particles. Frequently, drums are used without any inserts in the case of sticky materials and/ or high temperature treatment. Drums or kilns in shape of a long cylinder L/D N 5 are usually inclined by a few degrees to favor the particle transport into one direction and to adjust the residence time of the particles. They may also be aligned horizontally when long residence times are needed. To ensure a minimum hold up, the drum may additionally be equipped with annular weirs at the inlet and outlet. The solid material is fed into the drum at the front end and is transported towards the rear end by repeated revolutions. The gas providing the thermal treatment mostly flows in opposite direction (counter-currently) to the particles and has direct contact with the surface of the particle layer and the wall of the drum. During flow of bulk solids in rotary kilns segregation can appear due to differences in particle size or density, which could lead to significant differences in the product quality of the product. Published studies are mostly focused on segregation in batch drums whose L/D ratio is near to one and which are in rolling mode. This mode is also considered in this paper. In this mode two layers of different particle movement can be found, see e.g. Liu et al. (2006) [1]. In the passive layer the particles move upwards with the same angular velocity as the kiln. No inter-particle solid movement takes part in this layer. Passing a critical angle the solids enter the active layer and roll down. During the rolling process free particle movement is possible and segregation can occur, see Fig. 1. It is well known that larger particles are collected at the surface of the cascade and circulated on outer orbits, see e.g. McGlinchey (2005) [2]. The opposite happens to the small particles that are predominantly segregating into the core of the cascade layer penetrating through the voids of the large particles. After a low number of revolutions typical ring core patterns can be found. Additional to the transversal movement of the particles in rotary kilns, an axial movement takes place, enforced by a kiln slope and/or by different bed heights at inlet or outlet respectively. It is not quite clear so far whether the axial movement influences the formation of the ring-core-pattern. Also the feedback of the ring- core pattern to the transportation behavior is still unknown. Detailed analysis would require to locally read the concentration of species within the particle bed. Whereas in lab-sized plants the solid bed can be preserved with synthetic resin and later on analyzed “off-line” for the ring-core- formation, this technique cannot easily be adapted to technical sized plants. In this case an “on-line” measuring methods (without sampling and sample preparation) is required. In the literature a wide range of different systems is presented [3], which can be distinguished in three groups: “non-invasive, non-optical methods”, e.g. nuclear magnetic resonance spectroscopy (NMR), X-ray, positron emission tomography (PET), “non-invasive, optical methods”, e.g. particle image velocimetry (PIV), particle tracking velocimetry (PTV), and “invasive, optical methods”, usually optic probes using different Powder Technology 208 (2011) 63–71 ⁎ Corresponding author. TU Dortmund University, Department of Biochemical & Chemical Engineering, Laboratory of Mechanical Process Design, Emil-Figge-Str. 68, D-44227 Dortmund, Germany. Tel.: +49 231 755 6088; fax: +49 231 755 3961. E-mail address: P.Walzel@bci.tu-dortmund.de (P. Walzel). 0032-5910/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.powtec.2010.12.004 Contents lists available at ScienceDirect Powder Technology journal homepage: www.elsevier.com/locate/powtec