Geometrically and physically coupled numerical model of freeze drying in lyophilizer

V doktorski disertaciji so predstavljeni rezultati raziskav na temo eksperimentalnega dela določanja kinetike sušenja med procesom liofilizacije in ovrednotenja procesne zmogljivosti naprave, ter razviti numerični modeli za celovito modeliranje procesa in rezultati le teh. V sklopu eksperimentalnega dela se je ovrednotila kinetika sušenja destilirane vode (sublimacija ledu), ter eksperimentalna določitev kinetike depozicije ledu, kjer se je za tok vodne pare, ki je v režimu dušenega toka, izmerilo masne tokove sublimata in tlake znotraj sušilne komore in kondenzatorja.\ Rezultati eksperimentov so nato služili za validacijo razvitih numeričnih modelov. Simulacije, ki vključujejo RDT preračune so bile pripravljene v računalniškem programu ANSYS Fluent ali v programskem jeziku Fortran. Razvilo se je več modelov različnih kompleksnosti, prvi model je model za ovrednotenje vpliva geometrije gumastega zamaška in vrata viale na tlačni padec in vpliv na proces, kjer se je tok sublimata modeliral v RDT programu, medtem ko se je kinetika sušenja računala s pomočjo 0D modela (sklopljen 0D-RDT model). Na podlagi rezultatov, se je določilo empirično korelacijo za opis le tega pojava. Naslednji model je večnivojsko sklopljen model z uporabo TCP strežnika, ki povezuje 1D program za simuliranje kinetike sušenja znotraj viale (napisan v programskem jeziku Fortran) s komercialno kodo ANSYS Fluent, ki je uporabljen za izračun tokovnih polj skozi sistem. Kinetika sušenja produkta v vsaki viali je simulirana z 1D modelom, v katerem se tlačni padec vodne pare, med tokom skozi zgornji prostor viale, obravnava z empirično korelacijo. Masni tokovi sublimata in temperatura plina se posredujejo RDT modelu, ki jih uporabi kot robne pogoje na lokalnih vtokih. Po izračunu novega tokovnega polja, RDT model vrne nove tlake nad vsako vialo. Na ta način dobimo vpogled v hidrodinamske razmere znotraj liofilizatorja in kinetiko sušenja vseh vial. Naslednji del je modeliranje depozicije ledu, kjer se je modeliral tok sublimata skozi celoten sistem, depozicija na hladnih stenah kondenzatorja pa se modelira z dodatnimi členi za ponor mase, gibalne količine in energije. V tem primeru se torej meja modeliranja prestavi v kondenzator, kar nam omogoči vpogled v dogajanje znotraj kondenzatorja in vzorec depozicije ledu, ki se odvija v kondenzatorju.The doctoral dissertation presents the results of research on the subject of the experimental determination the kinetics of drying during the lyophilization process and evaluating the processing capacity of the device, as well as developed numerical models for comprehensive modeling of the process and the results of these. As part of the experimental work, the drying kinetics of distilled water (sublimation of ice) was evaluated, as well as the determination of the kinetics of ice deposition, where the sublimate mass flow and the pressure inside the drying chamber and condenser were measured for the water vapor flow in the choked flow regime. \ The results of the experiments were then used to validate the developed numerical models. Simulations that include CFD calculations are made in the ANSYS Fluent program or in the Fortran programming language. Several models of different complexity have been developed, the first model is a model for evaluating the influence of the geometry of the rubber stopper and the vial neck on the pressure drop and the influence on the process, where the flow of sublimate is modeled in the CFD program, while the drying kinetics of the ice is simulated using a 0D model (coupled model). Based on the results, an empirical correlation was determined to describe this phenomenon. The latter model is multi-level coupled using a TCP server that interfaces a 1D in-vial drying kinetics simulation program (written in the Fortran programming language) with the commercial ANSYS Fluent code used to calculate the flow fields through the system. The drying kinetics of the product in each vial is simulated with a 1D model, in which the pressure drop of water vapor through the upper space of the vial is treated with an empirical correlation. Sublimate mass fluxes and gas temperature are input to the CFD model, which are used as boundary conditions at the local inlets. After calculating the new flow field, the CFD model returns the new pressures above each vial. In this way, we get an insight into the hydrodynamic conditions inside the lyophilizer and the kinetic drying of all vials. The next part is ice deposition modeling, where sublimate flow through the entire system is modeled, and deposition on the cold walls of the condenser is modeled with additional terms for mass, momentum, and energy sinks. In this case, the modeling boundary is in the condenser, which gives us insight into what is happening inside the condenser and the pattern of ice deposition that takes place in the condenser