Modeling and Simulation of Complex Multiphase Flows in the Pharmaceutical Industry

Transcription

1 SIMNET Days 2010 Februar 10, 2010 Modeling and Simulation of Complex Multiphase Flows in the Pharmaceutical Industry D. Suzzi a, G. Toschkoff a, S. Radl a,b, Th. Hörmann a, M. Schaffer a, D. Machold a, C. Radeke a, R. Sungkorn b, J. G. Khinast a,b a Research Center Pharmaceutical Engineering, Graz/Austria b Institute for Process and Particle Engineering, Graz University of Technology, Graz/Austria K1 Competence Center Initiated by the Federal Ministry of Transport, Innovation & Technology (MMVIT) and the Federal Ministry of Economics & Labour (BMWA). Funded by FFG, Land Steiermark and Steirische Wirtschaftsförderung (SFG).

2 Overview 1. Introduction 2. Part A: Tablet Mixing Simulation 3. Part B: Detailed Spray/Tablet Simulation 4. Part C: Coater Flow Simulation 5. Conclusions and Outlook Page 2

3 Introduction! Multi-Scale Applications - 1 Numerical Simulation of particle transport and dissolution in stirred tank reactors Tank and impeller optimization Page 3

4 Introduction! Multi-Scale Applications - 2 Numerical Simulation of emulsions formation and drying Microscale 1D-Modeling Macroscale 3D Simulation of the whole system Mesoscale 3D local modeling Page 4

5 Introduction! Multi-Scale Applications - 3 Numerical Simulation of Bubble Columns and Bioreactors Page 5

6 Introduction! Multi-Scale Applications - 4 Numerical Simulation of Granular Wet Mixing Detailed Simulation of up to 7 Millions Particles Modeling of Spray and Fluid Spreading inside the Powder Bed Page 6

7 Target Process! Tablet coating Page 7

8 Target Process! Tablet coating Basic principles of the spraying and deposition processes on a single tablet or granule Page

9 Vision! Combined numerical approach B. Detailed Spray/Tablet Simulation (CFD) C. Coater Flow Simulation (CFD) C B A A. Tablet Mixing Simulation (DEM) Page 9

10 Vision! Objectives Tablet Coating Analyze Understand Optimize Page 10

11 Part A: Tablet Mixing Simulation! Particle Technology Numerical Simulation with DEM (Discrete Element Method) Software: EDEM and RPD (in-house development) Analysis of particles with arbitrary shape Optimization of the tablet mixing behavior in terms of: Surface velocity Residence time distribution Porosity distribution Angular velocity distribution Tablet stress (breaking and attrition of tablets) Page 11

12 Part A: Tablet Mixing Simulation! Coater Simulation: Example Tablets Translational Velocity Tablet form Tablets Rotational velocity Section of the drum coater Page 12

13 Part A: Tablet Mixing Simulation! Coater Simulation: Example Baffles design optimization Page 13

14 Part A: Tablet Mixing Simulation! Coater Simulation Numerical Analysis of a Real Industrial Coater Page 14

15 Part B: Detailed Spray/Tablet Simulation! Introduction Spray gun to inject the coating solution onto the moving particle bed. Drying air to enhance the film drying and to avoid sticking of the tablet to neighboring tablets Exhaust air exiting the pan through side opening, from inside the tablet bed (through an immersion tube system) or through a perforated rotary pan Rotating pan to ensure the tablet mixing Page 15

16 Introduction! Objectives 1 To understand the basic principles of the spray coating process by using sophisticated CFD multiphase simulations To model the spray deposition on the tablet, as well as the evaporation of the spray and the wall film in order to estimate the effects of the drying gas flow To numerically analyze the impact and deposition of droplets on particles with different shape To study the production and evolution of the liquid film on the surface of the tablets To investigate how different process parameters affect the coating process on a single tablet Page 16

17 Numerical Method! Computational approach (1/3) Computational Fluid Dynamics (CFD) Code AVL FIRE Turbulent gas flow described with a RANS (Reynolds Averaged Navier-Stokes) finite-volume approach Euler-Lagrange DDM (Discrete Droplets Method) approach for the description of the spray droplets motion (Dukowicz, 1980) Continuity equation for each parcel i General Newton s Equation of Motion for the each parcel i m d..droplet mass...evaporation mass source..break-up mass source..coalescence mass source F G gravity force F p.. force due to pressure gradient F D.. drag force F L.. lift force F A.. added mass force Sketch of forces acting on a parcel Page 17

18 Numerical Method! Computational approach (2/3) Splashing model for the interaction between droplets and tablet surface (Mundo et al., 1995) Droplet Reynolds Number Droplet Ohnesorge Number Empirical critical curve delimiting the splashing and deposition regimes Two-dimensional transport of the liquid wallfilm Film continuity equation Film momentum equation!..film thickness s m film mass source "..shear stress on the film # m...boussinesq turbulent eddy viscosity Page 18

19 Numerical Method! Computational approach (3/3) Multicomponent evaporation of the spray droplets and the tablet film Evaporation mass source for each species j Spray " modified Abramzon-Sirignano approach (Brenn et al., 2007) Example of evaporating droplets and film Film " Fick s law of unidirectional diffusion corrected Sherwood Number B m,j Spalding mass transfer number $ g,j. binary diffusion coefficient in the gas phase D d..droplet diameter w I.. Mass fraction at the film/gas interface Page 19

20 Case Definition! Simulation domain Simulation with water droplets Page 20

21 Case Definition! Parameter study Water/Glycerol coating solution!" Basis point D d = 25 µm v d = 15 m/s M = 3.5 e -4 kg c = 20% Glycerol T L = K T TAB = K x H2O = 0 (dry air) %T inj = 0.1 s %T sim = 0.5 s Sphere Variation star Tablet Droplet diameter Air temperature Injection velocity Glycerol content Page 21

22 Results! Wall film evolution Time evolution of the film at different time steps for a sphere (top) and the tablet (bottom) for the base conditions Page 22

23 Results! Integral analysis of the deposition behaviour (1/2) Variation star Droplet diameter Air temperature Injection velocity Glycerol content Page 23

24 Results! Integral analysis of the deposition behaviour (2/2) Variation star Droplet diameter Air temperature Injection velocity Glycerol content Page 24

25 Results! Basic trends of the variation analysis Tablet Shape It determines the gas flow, thus heat and mass transfer, as well as the local shear forces affecting the film motion. Droplets size and velocity High values of both " strong splashing behavior " reduction of total mass and quality of the film. Gas Temperature Increasing the gas temperature " minor effects on the mean film thickness, high impact on the film quality and on the surface coverage. Glycerol Concentration Limited effects on the mean mass and on the quality attributes of the film Page 25

26 Current work! Validation of the film deposition on a levitating sphere (Experiments from Link et al., 1997) Application: fluid bed coating A. Top spray B. Bottom Spray (Wurster Coating) C. Tangential Spray Coating (Rotor Pellet Coating) Source Page 26

27 Part C: Coater Flow Simulation! Tablets are placed in a rotating drum! A tablet emerges at the top, rolls on the bed through the spray zone! Filming liquid is sprayed on by a twocomponent air-atomization nozzle! After spraying, the tablet is dried by the inlet air stream! The tablet rolls down the bed, enters the bed, and reemerges after some time! The drying air leaves via an outlet! Typical process duration is some hours! Of course, after each batch the coater has to be cleaned Rotating drum Drying air outlet Tablet bed Drying air inlet Air flow and spray play a central role Page 27

28 Objectives Allthough air flow and spray behavior is important for the process, litte research has been done so far Page 28

29 Definition of the coater model Drum rotation: 6 rpm Model Back wall Tablet bed Drying air inlet > 1 Millon cells! Scale-up is extremely difficult for a coating process, therefore a real-size coater is simulated! Capture the essential properties, leave out the details (for now)! Generation of a mesh with about one million cells of variable size Drying air outlet Page 29

30 Spray Validation # The spray quality is a central factor in the coating process! Therefore, Phase Doppler Anemometry (PDA) was used to measure the size- and velocity-distribution of the spray droplets. The results are used for verification of the simulation. PDA Measurement principle:! A laser beam is split into two beams! The beams are crossed again! At the crossing point, an interference pattern emerges! A drop travelling through the beam crossing region reflects the light pattern! The reflected light is registered by two detectors! The fluctuation frequency gives the velocity! The phase shift gives the diameter Page 30

31 Spray Validation! Comparison of PDA measurement (red curve) with numerical simulation of spray (blue curve)! In the simulation, the spray droplets are initialized with the size taken from measurement! At a distance of 20 cm, the size and velocity of every droplet is registered 20 cm Velocity v / m/s Diameter d / um The simulated spray shows very good agreement with the experimental results Page 31

32 Results: Air Flow Inside the Coater Flow velocity Drying air outlet Eddie formation Drying air inlet Influence of gas flow on spray motion The spray is redirected by the eddie Left: jumping over the eddie Right: View from above (red square: inlet region) Page 32

33 Results: Streamlines! To get an idea of the three-dimensional flow, streamlines are drawn! They show the path a droplet would follow in the air stream Downstream streamlines show where the streamlines go after entering via the inlet Upstream streamlines show where the air was before leaving through the outlet Page 33

34 Parameter study! In the simulation (almost) everything is possible. But what parameters can be changed in reality? " Easily accessible: spray gun position and spray gun angle to the bed B the standard operational parameters tilted counter-clockwise to top of tablet bed tilted clockwise to air outlet translate to high position, Translate to low position, Translate to very low position Page 34

35 Results: spray mass impinged spray mass deserted through outlet Relative film liquid deserted / % #$%&'"(&))"*($*+,-."/"0" B #$%&'"(&))".-)-%1-."/"0" B Upper diagram! Mass of impinged droplets as a percentage of total introduced mass! The angle variations do not show much change.! Different positions: the higher the nozzle, the more particles impinge Lower diagram! Least spray loss for higher angle (1) and higher position (3) Open question: where do the droplets hit? Page 35

36 Results: spray masss on coater walls Wallfilm on the coater surfaces B 3 1!"#$%&'$((&)*&+,-&.)$+-#&/$00&1#-2&(-.3)*4&5&6& B Conclusion: The best results are obtained by tilting the spray gun away from the outlet Page 36

37 Conclusions The process of tablet spraying, as well as wetting, has been analyzed by means of a multiphase CFD solver Sophisticated models for the gas flow, the droplet motion and the flow of the liquid film on two different objects, i.e., a sphere and a diskshaped tablet, have been developed. Using these models we have performed a detailed variation study and analyzed the impact of the system properties on quality attributes of the resulting liquid film. Our variation study has been designed such as to provide a fundamental understanding of the importance of spray parameters with respect to film formation and coating efficiency Page 37

38 Outlook Page 38

39 Outlook Page 39

40 " Selected References " Suzzi, D., Radl, S., Khinast, J.G., Local Analysis of the Tablet Coating Process: Impact of Operation Conditions on Film Quality. Submitted to Chem. Eng. Sci. " Toschkoff, G., Suzzi, D., Khinast, J.G., 2010 Numerical and Experimental Analysis of Spray Losses in a Pharmaceutical Coating Process, in preparation. " Radl, S., Kalvoda, E., Glasser, B.J., Khinast, J.G., Mixing Characteristics of Wet Granular Matter in a Bladed Mixer. Submitted to Powder Technology. " Toschkoff, G., Suzzi, D., Reiter, F., Tritthart, W., Khinast, J.G., Improving drum coating processes by computational modeling and experimental evaluation AIChE Annual Meeting. " Suzzi, D., Egger, D., Radl, S., Reiter, F., Tritthart, W., Khinast, J.G., Detailed Numerical Analysis of Tablet Coating Processes AIChE Annual Meeting Page 40