Experimental investigation of the performances of a wet...

18th International Symposium on the Application of Laser and Imaging Techniques to Fluid Mechanics・LISBON | PORTUGAL ・JULY 4 – 7, 2016

Experimental investigation of the performances of a wet separation process for traditional and bio- plastics

Floriana La Marca1, Emanuela Lupo2, Monica Moroni2,*, Alessandra Pomponi1 1: DICMA - Sapienza University of Rome, via Eudossiana, 18 – 00184, Rome, Italy 2: DICEA - Sapienza University of Rome, via Eudossiana, 18 – 00184, Rome, Italy

* Correspondent author: [email protected]

Keywords: Plastic separation, Particle Tracking, Two-phase flows, Coupling regimes

ABSTRACT

The hydraulic separator is a device employing a wet technology for particle separation. Due to the combination of a characteristic flow pattern developing within the apparatus and density, shape and size differences among two or more polymers, it allows their separation into two products, one collected within the instrument and the other one expelled through its outlet ducts. The geometry of the channel allows the formation of recirculation areas that play a major role in the material separation. The characteristic dimensions of those areas and their interaction with the main advective flow depend on the apparatus internal shape. As the geometry of the apparatus is one of the key elements to characterize the device separation capability, two different arrangements were examined. The kinematic investigation of the fluid flowing within the apparatus seeded with a passive tracer was conducted via image analysis. This technique was also employed to study the behavior of mixtures of plastic particles and passive tracer to understand the coupling regime between the flow field and plastic particles. For the operating conditions tested, two-way coupling takes place, i.e., the fluid exerts an influence on the plastic particle and the opposite occurs too.

1. Introduction In 2013, with a continuous growth over more than 50 years, the global production of plastic materials has amounted to 299 million tons with an increase of 4 % compared to 2012 (PlasticsEurope, 2014). The European production represents 20 % of the world’s total production. In 2012, 25.2 million tons of post-consumer plastic wastes ended up in the waste upstream. Roughly 62 % was recovered through recycling and energy recovery processes while 38 % went to landfill (PlasticsEurope, 2014). Recycling may be accomplished using chemical or mechanical processes. Chemical recycling is driven by thermal (e.g., pyrolysis) or chemical (selective solvents) processes that essentially break down the original polymer chains and molecules (Sinha et al., 2010). Mechanical recycling is generally considered as the best option for plastic waste management. The production of high-quality products, which favors virgin polymer substitution thus reducing environmental impact


and resource depletion (Hopewell et al., 2009), may be challenging due to the heterogeneous composition of plastic waste. For instance, bio-plastics which are substituting traditional plastics for the manufacture of several products, mainly in the food sector, may be wrongly conferred with plastic waste instead of wet waste, thus ‘polluting’ the plastic waste stream. Mechanical recycling requires several treatment steps, usually: cutting/shredding, to reduce particle size and to get a suitable shape for further processing; separation in dry conditions, to eliminate impurities, such as paper, dust and other non-plastic materials; polymer classification, to separate polymer per type; milling, to homogenize particle size of single-polymer plastics. Further steps, i.e., washing/drying, agglutination, extrusion, are designed to prepare the end-product according to the market standards (Al-Salem et al. 2009). The market value of recycled plastics is considerably affected by their purity, degree of decontamination and homogeneity (La Marca et al., 2012), which are strictly related to the effectiveness of the mechanical recycling process, in particular the polymer classification step. Nowadays the most common techniques employed in plastic recovery plants to classify polymers are based on flotation processes (Marques and Tenório, 2000), density differences (Gent et al., 2009), electrostatic forces (Wei and Realff, 2003; Wu et al., 2013) and optical properties (Ahmad, 2004). All processes present drawbacks related to cost, performance and environmental hazards. This work presents an innovative wet technology for separating plastic particles, which could be employed in the separation step of mechanical recycling plants, representing a valuable alternative to existing technologies. In fact it allows overcoming the typical problems of the most used separation methods, such as: the need of additives (that create secondary pollution) for flotation, dense sorting and cyclones; the influence of moisture, surface status and feeding speed of particles for electrostatic separation; the complexity and cost of equipment for optical methods; the need of finely grinded materials for cyclones. The separation process relies on the difference in density of the polymers and the characteristic flow pattern developing within the instrument. The device is suitable for the separation of polymers with a density higher than 1000 kg/m3 and employs plain water without chemical additives. Different tools employed for the characterization of the apparatus features, i.e., image analysis and separation tests, will be described in this contribution. The originality of the paper lies in two aspects. First, the paper investigates the separation efficiency as a function of the mass rate fed to the device. Second, the instrument performances with both traditional and bio-plastics are evaluated.


2. Materials and methods The hydraulic separator is a channel with a characteristic internal geometry. The channel is constructed from a sequence of 8 parallel semi-cylindrical transparent Plexiglas tubes welded together in a plane and the lower complex is shifted relative to the upper complex (Figure 1). The apparatus is transparent allowing the optical access and the use of image analysis techniques for the detection of the velocity field. Both inlet and outlet are equipped with 8 ducts located along the transverse direction (Ii (i=1,…,8) and Oj (j=1,…,8), respectively). Further, the lower part of the device presents 8 collecting ducts.

HYDRAULIC SEPARATOR

PRIMARY TANK

SECONDARY TANK

Ii (i = 1, ..., 8)

Oj (j= 1, ..., 8)

C1

C8

C2 C3 C4

C6 C7

C5

Figure 1. Synthetic scheme of the experimental facility and three-dimensional schematic representation of the

hydraulic separation channel with inner, outlet and collecting ducts; reference system.

To ensure a constant hydraulic head, the apparatus is filled via a variable height tank connected with the 8 inlet ducts. The water level within the tank is controlled through an overflow exit. Digital images were acquired using two high-speed (400 Hz), high-resolution (1280×1024 pixels) cameras (Mikrotron EoSens), equipped with Nikon lenses with focal length of 105 mm. Two


high-speed Camera Link Digital Video Recorders operating in Full configuration (IO Industries DVR Express® Core) were used to synchronise recording from the two cameras, manage data acquisition and storage. Proper illumination was ensured through the use of a LED-based Linescan Illuminator (COBRA Slim) positioned above the upper surface. A light sheet 40 cm long and 1 cm thick oriented in the longitudinal direction, i.e., parallel to the mean flow field, was generated. The water was seeded, via the secondary tank, with a well-reflecting neutrally-buoyant passive tracer to allow qualitative visualization of the flow and quantitative estimation of the velocity field (Figure 1). HLPT (Hybrid Lagrangian Particle-Tracking) was employed to reconstruct the velocity field. It is a algorithm for particle identification and tracking, based on the solution of the optical flow equation via a sum-of-squared-difference method. Particles are detected through the identification of corner features, where image intensity gradients are not null in two orthogonal directions. It is thus possible to identify low intensity and overlapped particles. Furthermore, the feature selection criterion is optimal by construction because it is based on the optical flow solution and therefore a good feature is the one that can be tracked well. Therefore it is possible to obtain the local velocity, given by the approximate solution of the optical flow equation, which can be used as a predictor for the subsequent particle pairing step (Shindler et al., 2012). The plastic materials tested within the apparatus are among the most common polymers available on the market (PET, PVC, PC and PLA). These materials have density slightly larger than water (Table 1). Hence their entrance within the apparatus is ensured. Each material was characterized i) geometrically, to determine a characteristic particle size; ii) physically, to determine the density and iii) spectrally, to ensure the membership to a particular type of polymer. For each polymer, samples of material at different stages of a product life cycle were selected: primary raw or virgin (V) materials, wastes (W) and secondary raw or regenerated (R) material. Four samples of virgin plastic particles, consisting of granules (G) of nearly spherical or cylindrical shape with rather regular and homogeneous dimensions were collected. Three samples of urban and industrial plastic wastes, washed and purified from any impurities, were triturated with a knife mill to obtain irregularly shaped flakes (F) or pieces (P). Finally, one sample of regenerated plastics provided by an Italian plant for plastic recovery and recycling (“Rigenera S.r.l.” – Terni (TR)) was triturated with a knife mill to obtain irregularly shaped flakes. Each material was classified using standard sieves into two size classes (d is the mesh size):


o size class I: 2.00 10-3 m<d<3.36 10-3 m; o size class II: 3.36 10-3 m<d<4.76 10-3 m.

Table 1. Polymer sample characteristics: life cycle stage (virgin (V), waste (W), regenerate (R)); shape (granular (G), flake (F), piece (P)); density and size class.

Name Life cycle

stage Source Shape

Density (g/cm3)

Size class

PC_VG virgin raw

material granular 1.18 I

PC_RF regenerated industrial

waste flakes 1.20 I-II

PVC_VG virgin raw

material granular 1.30 II

PVC_WP waste industrial

waste pieces 1.61 I-II

PET_VG virgin raw

material granular 1.31 I

PET_WF waste urban waste

flakes 1.35 I-II

PLA_VG virgin raw

material granular 1.24 II

PLA_WF waste urban waste

flakes 1.22 I

These size classes were selected to verify the influence of the particle size on the separation process. 2. Results and discussion The purpose of the device is to separate the useful fraction from a mixture of (useful and useless) plastics and water introduced inside. Understanding the properties of two-phase flows (liquid-solid) is then required to identify the main characteristics of the system under investigation. To understand the interaction among the two phases and define the way of coupling among them it is mandatory to compute the characteristic parameters of both the fluid and solid phases. Flow phase characterization The velocity field within the apparatus depends on the internal geometry and the operating flow rate which can be tuned by changing the tank height and the number of the opened outlet ducts. In this paper, we investigate the apparatus performances for a given apparatus internal geometry and 9 different flow rates (Table 2).


Lagrangian data provided by HLPT have been used to reconstruct time averaged Eulerian velocity fields through a resampling procedure. The mean velocity components along the longitudinal and transverse directions, u and v, were evaluated in the knots of a regular grid of 71 rows and 256 columns. The phenomenon is assumed stationary, and then time averages over a time interval equal to the entire recording time of about 40 s were computed. The Eulerian description of the velocity field allowed computing the streamlines, which schematically represent the typical flow structures within a single chamber. Streamlines are presented in Figure 2.

Table 2. Flow rate within the apparatus at different hydraulic head and number of opened outlet ducts.

Height at the apparatus inlet ducts

(m)

Number of opened outlet ducts

Case Flow rate

(l/s)

Q1= 1 2 (O3,O6) 1 0.72

3 (O2,O4,O6) 2 0.92 4 (O2,O4,O6, O8) 3 1.08

Q2= 1.5 2 (O3,O6) 4 0.84

3(O2,O4,O6) 5 1.09 4 (O2,O4,O6, O8) 6 1.24

Q3= 2 2 (O3,O6) 7 0.94

3(O2,O4,O6) 8 1.22 4 (O2,O4,O6, O8) 9 1.36

Figure 2. Flow field structure for case #5: lower recirculation zone (red streamlines), upper recirculation zone (green

streamlines) and principal flow (blue streamlines). Configurations (a) ASYM and (b) SYM.

For better understanding the flow field, a zoom in a single chamber of the apparatus (chamber C3) for 3 hydraulic configurations (namely cases #1 (Figure 3), #5 (Figure 4) and #9 (Figure 5)) is presented. Though the streamlines reconstructed for cases 1 to 9 are very similar in shape to


those presented in Figure 2 (for this reason they were not reported), both velocity components present significantly different magnitude values for the different hydraulic configurations. The characteristic behavior of the fluid flowing through the apparatus is viewed by:

o the transport flow (blue streamlines), where the fluid is characterized by a positive value of the velocity component along the x axis in the entire longitudinal section; the principal transport flow is responsible for the transport of particles from one chamber to the next; it eventually drives material to the outlet nozzles without separation.

Figure 3. Eulerian velocity vectors overlapped onto the colormap of the horizontal velocity component (left-hand

side image), of the vertical velocity component (central image) and of the two-dimensional turbulent kinetic energy (right-hand side image) in C3 for case #1.





o the lower recirculation zone (red streamlines), visible below the principal current. Its

clockwise rotating motion is suitable for capturing particles from the principal current. It


is expected that plastic particles captured in this way will behave in one of the following ways: settle within the chamber if sufficiently heavy; follow the upward portion of the rotating motion without reaching the principal current, due to the fact that they are too heavy to perform a complete rotation; execute a complete rotation and be recaptured by the principal transport flow to settle in the next chamber or be expelled. The particle behavior within the recirculation zone is influenced by its density and dimension as well as by the presence of a vortex.

o the upper recirculation zone (green streamlines), visible above the principal transport current. The purpose of the upper recirculation zone is to subtract particles from the principal current and to transfer them to the preceding chamber. To undergo such transfer, particles must move across the principal current and settle in the lower recirculation area of the chamber. The particle physical attributes and the characteristic velocity of the principal current will influence the efficacy of this process.

Reynolds number, ν

FFdU=Re , was computed for all cases investigated.

The flow characteristic dimension, dF, was detected by analyzing the flow features in the vertical section passing through the upper cusp at the entrance of chamber C4. Along this section, the vertical profile of the horizontal velocity component was employed to compute the mean velocity UF. ν is water kinematic viscosity. Both dF and UF were provided by image analysis. The Reynolds number ranges between 2000 and 3400 (Table 3) and increases with the flow rate. Re values suggest the flow regime within the separator is in the transition region. Solid phase characterization Separation tests have been conducted with plastic mixtures of total solid volumes equal to 6.2 10-3 m3 and 12.42 10-3 m3. A useful parameter to characterize the two-phase flow concentration is the solid phase volume fraction αs, defined as the ratio between dispersed phase and total volumes. αs ranges between 2 10-5 and 5.23 10-5.

Table 3. Characteristic fluid phase velocity, dimension and time, Reynolds number.

Case UF

(m/s) dF

(m) Re

#1 0.078 0.026 2000 #4 0.084 0.024 2000 #7 0.095 0.020 2000 #2 0.100 0.024 2400 #5 0.110 0.023 2500 #3 0.112 0.023 2600


#8 0.123 0.025 3100 #6 0.130 0.024 3100 #9 0.136 0.025 3400

Finally, the particle Reynolds number, Rep, was computed for all plastic samples investigated. Table 4 and Table 5 show Rep for plastic samples belonging to size class I and II respectively. Being Rep on the order of 103, for a volume fraction of around 10-5, the coupling regime among the fluid and solid phase is expected to be two-way. Then, a reciprocal influence between the two phases is likely to occur.

Table 4. Characteristic size, settling velocity, Rep and relaxation time for plastic samples belonging to size class I.

Name Characteristic

size (m) Settling velocity

(m/s) Particle Reynolds

number Relaxation time (s)

PC_RF 0.00298 0.084 224.4 0.067 PVC_WP 0.00339 0.146 391.9 0.065 PET_VG 0.00264 0.104 279.4 0.064 PET_WF 0.00291 0.111 296.9 0.064 PLA_WF 0.00268 0.092 245.8 0.066

Table 5. Characteristic size, settling velocity, Rep and relaxation time for plastic samples belonging to size class II.

Name Characteristic

size (m) Settling velocity

(m/s) Particle Reynolds

number Relaxation time (s)

PC_VG 0.00317 0.098 396.9 0.108 PC_RF 0.00359 0.103 418.4 0.106

PET_WF 0.00388 0.136 553.5 0.101 PVC_VG 0.00319 0.126 512.5 0.102 PVC_WP 0.00441 0.180 730.7 0.102 PLA_VG 0.00395 0.113 458.4 0.104

To qualitatively and quantitatively study the interaction between discrete phase (plastic particles) and carrier phase (water), image analysis of mixtures passive tracer-plastic particles was also conducted on a subset of plastic samples. The aim of the tests was to confirm the coupling regime for plastic particles of both size classes to be two-way. Those tests were rather time consuming. For this reason, only case #7 (flow rate equal to 0.92 10-3 m3/s) was considered and results are not easily extended to other flow rates. Furthermore, the tests were conducted feeding a solid volume of 6.2 10-3 m3 (recalling separation tests were conducted also for solid volume equal to 12.42 10-3 m3). The hydraulic configuration was chosen according to the results of separation tests which have shown that particles of PC_VG and PC_RF (of both size classes) are completely expelled from the apparatus; particles of PVC_VG and PVC_WP completely settle


within the apparatus; 80 % of particles of PET_VG settle as well as 20% of PET_WF of size class I and 60% of PET_WF of size class II; PLA_VG completely settled within the apparatus whereas PLA_WF is completely expelled. Vertical profiles of both velocity components have been drawn at 5 characteristic locations within C2, where the polymer sedimentation occurs more evidently than in the other chambers. The profiles obtained from the analysis of images with only passive tracer have been compared to those resulting from the processing of images with both tracer and plastic particles, showing some discrepancies. A more careful data analysis is required. On the other hand, velocity profiles for plastic particles are significantly different than the corresponding tracer profiles. 4. Conclusions Image analysis with the particle tracking technique HLPT allows the reconstruction of the main characteristics of the velocity field within the separator. Three different characteristic structures can be identified: a transport current along the entire longitudinal section, and, in each chamber, an upper (smaller) and lower (larger) recirculation zone. The influence of the 9 hydraulic configurations is identified through both the mean velocity component magnitudes and TKE 2D visualization. The increase of velocity values is consistent with the flow rate. The reconstruction of the velocity field of plastic particles and passive tracer within a mixture allows understanding their behaviour as single phases and as interacting phases. The analysis of 8 different mixtures of passive tracer and PC, PET and PVC particles, coupled to the study of dimensionless numbers, demonstrates that there is an influence of the fluid phase on the particles, as well as the opposite, although in a less pronounced way. Then, a two-way coupling regime occurs. Although tests were carried out only for a volume of plastic particles equal to 6.2 10-3 m3, it is reasonable to infer that the same coupling regime will occur doubling the solid volume. Separation tests conducted for samples of volumes equal to 6.2 10-3 m3 and 12.42 10-3 m3 show that the same amount of material settles within the apparatus, confirming the possibility of feeding the apparatus with solid rate doubled respect to past investigations. References Ahmad SR (2004) A new technology for automatic identification and sorting of plastics for recycling. Environmental Technology 25(10):1143-1149.


Al-Salem SM, Lettieri P, Baeyens J (2010) The valorization of plastic solid waste (PSW) by primary to quaternary routes: From re-use to energy and chemicals. Progress in Energy and Combustion Science 36(1):103–129. Gent MR, Menendez M, Toraño J, Diego I (2009) Recycling of plastic waste by density separation: Prospects for optimization. Waste Management and Research 27(2):175-187. Hopewell J, Dvorak R, Kosior E (2009) Plastics recycling: challenges and opportunities. Philosophical Transaction of Royal Society B 364 (1526): 2115–2126. La Marca F, Moroni M, Cherubini L, Lupo E, Cenedese A (2012) Separation of plastic waste via the hydraulic separator Multidune under different geometric configurations. Waste Management 32(7):1306-1315. Marques GA, Tenório JAS (2000) Use of froth flotation to separate PVC/PET mixtures. Waste Management 20(4):265-269. PlasticsEurope (2014) Plastics – the Facts 2013: an analysis of European latest plastics production, demand and waste data. Shindler L, Moroni M, Cenedese A (2012) Using optical flow equation for particle detection and velocity prediction in particle tracking. Applied Mathematics and Computation 218: 8684–8694. Sinha V, Patel MR, Patel JV (2010) PET Waste Management by Chemical Recycling: A Review. Journal of Polymers and the Environment 18: 8–25. Wei J, Realff MJ (2003) Design and Optimization of Free-Fall Electrostatic Separators for Plastics Recycling. AIChE Journal 49(12):3138-3149 Wu G, Li J, Xu Z (2013) Triboelectrostatic separation for granular plastic waste recycling: A review. Waste Management 33(3): 585-597.

Experimental investigation of the performances of a wet...

Documents

Transcript of Experimental investigation of the performances of a wet...