Stefan Radl

Unexpected trapping of particles at a T junction.

Significance Most flow systems we encounter in everyday life, such as cardiovascular blood flow in our body or the flow in typical domestic or industrial piping systems, present T junctions that split the flow into two streams. Another characteristic that these systems often have in common is the presence of a dispersed phase made of low-density particles or air bubbles. In this paper we present evidence that in these situations particles can get trapped at the junction, accumulate, and ultimately change the flow distribution. The unanticipated trapping mechanism can affect health and diagnostics in human beings or create malfunctions, and consequently unexpected dangers, in industrial environments. Nonetheless, the same effect can also be exploited to design new particle separation devices. A common element in physiological flow networks, as well as most domestic and industrial piping systems, is a T junction that splits the flow into two nearly symmetric streams. It is reasonable to assume that any particles suspended in a fluid that enters the bifurcation will leave it with the fluid. Here we report experimental evidence and a theoretical description of a trapping mechanism for low-density particles in steady and pulsatile flows through T-shaped junctions. This mechanism induces accumulation of particles, which can form stable chains, or give rise to significant growth of bubbles due to coalescence. In particular, low-density material dispersed in the continuous phase fluid interacts with a vortical flow that develops at the T junction. As a result suspended particles can enter the vortices and, for a wide range of common flow conditions, the particles do not leave the bifurcation. Via 3D numerical simulations and a model of the two-phase flow we predict the location of particle accumulation, which is in excellent agreement with experimental data. We identify experimentally, as well as confirm by numerical simulations and a simple force balance, that there is a wide parameter space in which this phenomenon occurs. The trapping effect is expected to be important for the design of particle separation and fractionation devices, as well as used for better understanding of system failures in piping networks relevant to industry and physiology.

Modeling and simulation of polyacrylic acid/protamine nanoparticle precipitation

In this study a model for the prediction of polyacrylic acid/protamine nanoparticle precipitation was developed, which constitutes the first numerical approach for modeling the precipitation of organic nanoparticles. Due to the complexity of the process, a spatial resolution (i.e., a simulation via computational fluid dynamics) was not the target. For the description of the precipitation process, population balance equations, accounting for nucleation, growth and aggregation were used and coupled with the engulfment model for mixing. Furthermore, experiments have been carried out by turbulent mixing of precursors in a vortex mixer. Measurements of the resulting particle size distributions and of the final concentrations in the liquid were performed. Finally, by selecting unknown parameters, it was possible to achieve a good agreement of experimental and numerical results. It was found that the formation of the electrostatic surface charge, caused by a layer of protamine on the particle surface, cannot be described instantaneously. Small deviations in the final liquid concentrations and particle size distributions are probably caused by the significant simplifications introduced by the engulfment model and by the description of the solid/liquid equilibrium. However, the presented model may serve as a basis for further development, i.e., by replacing the engulfment model by more sophisticated approaches, e.g., via computational fluid dynamics.

The Engineering of Hydrogen Peroxide Decontamination Systems

In this article, the latest developments for designing hydrogen peroxide decontamination systems are analyzed. Specifically, focus is given to the accurate calculation of hydrogen peroxide condensation phenomena and discussion of a new correlation for its accurate prediction. A procedure for calculating the condensate composition or the dew point out of this correlation is detailed, and an h–x diagram for moist, hydrogen peroxide-laden air, which is of fundamental importance for the rational design of hydrogen peroxide decontamination systems, is proposed. Also presented are theoretical results that illustrate the effect of condensation and evaporation in these systems. Finally, some perspectives for improving hydrogen peroxide systems, and the role computational fluid dynamics (CFD) may have in this field, are provided.

Use of the Direct Compression Aid Ludiflash® for the preparation of pellets via wet extrusion/spheronization

Objective: Conventional solid oral dosage forms are unsuitable for children due to problems associated with swallowing and unpleasant taste. Additionally, the limit of tablets lays in the patient adapted dosing. Therefore, the suitability of Ludiflash®, a direct compression aid for orally disintegrating tablets, was investigated for the preparation of individually dosable pellets. Materials and methods: Micropellets consisting of Ludiflash® and small amounts of microcrystalline cellulose were prepared via the wet extrusion/spheronization technique. Paracetamol and ibuprofen were applied as model drugs. The obtained pellets were characterized with respect to drug release and disintegration characteristics, mechanical properties, as well as size and shape. Results and discussion: Drug loading was possible up to 30% for ibuprofen and even up to 50% for paracetamol. Higher ibuprofen loadings resulted in considerably slowed drug release and higher paracetamol contents yielded in non-spherical pellets. In vitro release studies revealed that more than 80% of the drug was released within 30 and 60 min for paracetamol and ibuprofen, respectively. Drug release rates were highly influenced by the pellet disintegration behavior. Investigations of the release mechanism using the Korsemeyer-Peppas approach suggested Super Case II drug transport for all paracetamol formulations and anomalous drug transport for most ibuprofen formulations. All pellets exhibited a low porosity and friability, as well as a sufficiently high tensile strength, which was significantly influenced by the type of model drug. Conclusion: Ludiflash® can be applied as main excipient for the preparation of individually dosable pellets combining fast drug release and a high mechanical stability.

Quantifying Absorption Effects during Hydrogen Peroxide Decontamination

Understanding mass transfer effects (e.g., condensation and absorption) of hydrogen peroxide is essential for clean room decontamination technology. For example, absorption of hydrogen peroxide in polymers often causes unwanted effects in the final aeration phase of a decontamination cycle. This currently leads to significant challenges in the design and operation of clean rooms or isolators. We address here the absorption of H2O2 in polymers, which is a key for an understanding of mass transfer effects. Specifically, we developed a novel experimental setup to measure the desorption rate from a polymer sheet. Together with a model for the diffusion in a polymer, we are able to theoretically predict the absorption and desorption kinetics of hydrogen peroxide. We then performed measurements of still unknown properties of hydrogen peroxide in polymers, i.e., the saturation concentration and the diffusion coefficient using our experimental setup. As expected, the diffusion in the polymer is the rate-limiting step for the absorption and the release of H2O2. We find that considerable amounts of H2O2 can be absorbed in certain polymers on a time scale of less than an hour and may lead to a temporary violation of the Occupational Safety and Health Administration safety level in a typical clean room.

Closure Development for Multi-Scale Fluidized Bed Reactor Models: A Case Study

Abstract Chemical looping reforming (CLR) processes offer textbook examples for challenges in chemical engineering with respect to transport limitations. Phenomena that potentially need to be considered in a rigorous reactor model include (i) diffusion in solids and nanometer-scale pores, (ii) heat and mass transfer between suspended particles and the ambient gas, (iii) meso-scale phenomena such as clustering, as well as (iv) large-scale phenomena such as particle and gas-phase dispersion in the reactor’s axial direction. Here we summarize key scientific advances made in the “NanoSim” project, which established a computational platform that enables modelling a large range of these phenomena. Specifically, we show that already at the particle scale significant uncertainties are introduced when modelling chemical reactors in very detail. This is due to the nature of gas-particle flow, i.e., the spontaneous formation of heterogeneities (i.e., so-called meso-scale structures) that impact flow, species transport and reactions. The key finding is that these heterogeneities must be accounted for in typical CLR applications to correctly predict reaction rates in an industrial-scale reactor.

Size-Based Particle Separation in Coiled Channel Flow of Non-Circular Cross-Section

Abstract Helically coiled separation channels are applied in size-based particle fractionation in (i) mineral processing, and (ii) microfluidics for the separation of differently-sized cells. In such coiled channels a pressure difference between the inner and outer bent promotes secondary fluid motion, and channels with circular cross sections have been extensively used in the past. Recently, it has been shown that channels with a rectangular or trapezoidal cross section offer an interesting playground for process optimization, since the particles’ equilibrium position is highly sensitive to (i) channel shape, (ii) flow Reynolds number, and (iii) particle size. The present study utilizes detailed simulations of fluid flow and particle motion in such toroidal channels to gain quantitative insight into what causes particles to accumulate at preferential positions. Therefore, we first present details on the speed of secondary fluid motion for different cross sectional shapes. Also, we present results of Euler-Lagrange suspension flow simulations that predict the effect of fluid flow topology on particle trajectories. This allows us to quantify size-based particle separation. This new understanding is key for a future rational design of particle separation processes.

MODELING OF NON-SPHERICAL, ELONGATED PARTICLES FOR INDUSTRIAL SUSPENSION FLOW SIMULATION

Euler-Lagrange (EL) simulations of particulate suspension flow are an important tool to understand and predict multiphase flow in nature and industrial applications. Unfortunately, solid-liquid suspensions are often of (mathematically) stiff nature, i.e., the relaxation time of suspended particles may be small compared to relevant flow time scales. Involved particles are typically in the size range from μm to mm, and of non-spherical shape, e.g., elongated particles such as needle-shaped crystals and/or natural and man-made fibres. Depending on their aspect ratio and bending stiffness, those particles can be treated as rigid, or flexible. In this paper we present a recent implementation into the open-source LIGGGHTS and CFDEM software package for the simulation of systems involving stiff non-spherical, elongated particles. A newly implemented splitting technique of the coupling forces and torques, following the ideas of Fan and Ahmadi (J. Aerosol Sci. 26, 1995), allows significantly larger coupling intervals, leading to a substantial reduction in the computational cost. Hence, large-scale industrial systems can be simulated in an acceptable amount of time. We first present our modeling approach, followed by the verification of our code based on benchmark problems. Second, we present results of one-way coupled CFD-DEM simulations. Our simulations reveal segregation of fibres in dependence on their length due to fibre-fluid interaction in torus flow.