Schedule for: 16w5074 - EPIC - Enabling Process Innovation through Computation

Talk: Plenary Abstract: 1. It is imperative for any academic institution to achieve a fine balance between the generation and dissemination of new knowledge. There are various ways of dissemination which include undergraduate education, publications in various forms and active partnership with industry and society. 2. A careful global survey has shown that the national wealth is largely generated through the knowledge generated in academic institutions. Several case studies will be described which not only results into huge wealth but creates win-win-win-win …. situations for society, industry, environment, students, faculty and institutions, the most important being of having very satisfying profession. A systematic effort is possible to generate creative and enjoyable ambience in the institutions. Some personal thoughts and experiences will be shared. 3. Multiphase reactors are very common in chemical industry, such as bubble columns, trickle beds, fluidized beds, stirred tanks, packed columns of various types and sizes of packings, etc. Currently, there is high degree of empiricism in the design process of such reactors owing to multiphase turbulent three dimensional flow and its coupling with transport phenomena and chemical reactions. Hence, we focus on synthesizing recent advances in computational and experimental techniques that will enable future designs of such reactors in a more rational manner by exploring a large design space with high-fidelity models (computational fluid dynamics and computational chemistry models) that are validated with high-fidelity measurements (tomography and other detailed spatial measurements) to provide a high degree of rigor. Understanding the spatial distributions of dispersed phases and their interaction during scale up are key challenges that were traditionally addressed through pilot scale experiments, but now can be addressed through advanced modeling. 4. Through several examples of using computational and experimental fluid dynamics, this talk will explore the fundamental paradigm shift in the design process to enable process innovation. These include NOx absorption, design of annular centrifugal extractors, self inducing impellers, air-lift reactors, chromatographic separations, bio-mass based cook stoves, etc.

Talk: Keynote Abstract: TBA

Talk: Regular Abstract: Process modeling and simulation is a key enabling technology for chemical engineers to develop and design innovative and safe chemical processes. The underlying molecular thermodynamic models provide the scientific foundation essential for all process calculations involving thermophysical properties, heat and mass transfer, chemical reactions, phase and chemical equilibria, etc. Indeed, tremendous past accomplishments and recent advances in molecular thermodynamics have served chemical engineers well and contributed to the wide acceptance of process modeling and simulation especially in oil and gas industry and petrochemical industry. A prominent recent example is the development of a novel molecular thermodynamic model for chemical absorption of CO2 with a proprietary organic amine solution and the application of the model in the development, design, and optimization of the 1st commercial scale CO2 capture unit operated by Saskpower. However, many unsolved problems remain while new challenges emerge as chemical engineers explore new opportunities and frontiers such as carbon capture and sequestration, energy storage, saline water remediation, nuclear waste disposal, etc. This talk highlights several high impact research opportunities in the pursuit of sustainable energy. Specifically, we present new advances in molecular thermodynamics in support of modeling and simulation for high salinity produced water treatment processes in oil and gas production, defense nuclear waste treatment processes, noble metal extraction from recycled electronics, asphaltene aggregation in petroleum processing, and hybrid sulfur (HyS) cycle for hydrogen production.

Talk: Regular Abstract: Delivery of light to microorganisms plays a crucial role in the performance of algal photobioreactors, especially for reactors operated at high biomass concentrations. The important characteristics of light delivery include not only the intensity and wavelength spectrum of irradiation, but also temporal variations in these quantities. In particular, microalgal biomass productivity and light utilization efficiency can be significantly increased by causing the organisms to experience light/dark cycles with frequencies greater than 1 Hz (flashing light effect). As a result of these observations, we demonstrate that commonly used reactor designs (e.g. airlift reactors) are highly inefficient and that biomass growth rate in both batch and continuous flow photobioreactors can be significantly improved by employing reactor designs with flow patterns that coherently shuttle microorganisms between light and dark regions of the reactor. We also describe the development and validation of two computational frameworks for coupling models for the three fundamental phenomena that determine reactor performance, namely multiphase turbulent fluid flow, radiation transport, and algal growth kinetics. While these computational methods provide a rational basis for photobioreactor design, optimization, and scaleup, significant challenges limit their applicability, especially for reactors operated at desirable high biomass concentrations.

Talk: Regular Abstract: Structured reactors offer many opportunities for advanced computational modelling. The physical and chemical phenomena that are important in the reactor happen at several length scales and thus represents a multi-scale problem. As an illustration, consider the washcoated monolith reactor, a classical structure comprised of (usually) thousands of parallel channels. The smallest scale that might be considered is the molecular scale, where the various molecules interact with active sites to effect the reaction. It is certainly common to eliminate this scale through the use of global kinetic expressions, however, it is often necessary to use more detailed mechanisms, either because the global approach lacks accuracy or a more detailed description of the product distribution is desired. The second scale would be the washcoat, which has a thickness that varies between 10 and 150 microns. Structurally, it is a porous medium composed of pores that range from 10 nanometres in radius and larger. One may wish to model the actual microstucture, or to treat the washcoat as a continuum. Thus we have either one or two additional scales to consider that are orders of magnitude different. A typical monolith reactor channel has a dimension of the order of 1 mm, which is again two orders of magnitude different from the washcoat. Each channel has a spatially dependent distribution of species, temperature and velocity. Heat transfer occurs between channels, and the velocity distribution is not usually uniform. Finally, at the reactor scale, the dimensions are of the orders of many centimetres, which represents again an orders of magnitude scale shift. To incorporate information at all scales into a complete model requires a clever computing strategy and an appropriate selection of scale bridges, otherwise the solution will be extremely expensive from a computational point of view. Linking the different scales at the simplest level requires the passage of information among the various scales using a system of sub-models for each scale. We have recently had a good measure of success in enhancing computational efficiency using pre-computed data. The look-up table may contain the results of different sub-models, and thus acts as a scale bridge. Essentially, the look-up table can be considered to be a table of values computed at discreet points over a specified range of operating conditions. Intermediate values are normally computed using spline interpolation.

Talk: Plenary Abstract: The OSU chemical looping platform technology can convert various carbonaceous feedstock such as coal, biomass, natural gas, and syngas into electricity, syngas, hydrogen, chemicals and/or liquid fuels. This platform technology embodies all elements of particle-science and technology including particle synthesis, reactivity and mechanical properties, multiphase flow characteristics and contact mechanics and fluidization. This presentation will focus on two aspects of mutliphase flow phenomena that are crucial to the successful operation of the chemical looping process system. They are solids bridging in the moving bed flow and L-valve behavior in the solid circulation control. On solid bridging (or arching), it is of an operational hazard when it occurs in the moving bed flow. The analysis of this arching mechanism was originally developed to characterize the discharge of particles from a hopper based on the particulate media mechanics with particle layer stress analysis. However, bridging can occur in a moving bed system that is correctly designed to allow desired continuous mass flow of solids particles. Experimental studies have revealed that the appearance of solids bridging normally accompanies with the appearance and accumulation of fine powders among the moving coarse particles along with an observed change in the interstitial gas flow. The counter-current interstitial gas flow provides an opposing resistance in terms of interphase drag preventing a continuous downward solids flow driven by gravity. Its change may account for stress redistribution inside the layered particles, which yields an interlock among particles allowing for bridging to form. The accumulation of fine powders may alter the hydrodynamic characteristics of the solids particles and decrease the local voidage inside the moving bed. The decrease in voidage may significantly increase the inter-phase drag while the change of solids characteristics may alter the tendency for solids particles to bridge. In the presentation, a theoretical stress analysis on layered particles distributed in a vertical moving bed with opposing interstitial gas flow will be described to account for the appearance of bridging phenomenon. The analysis will consider the characteristic effects of fine powder such as size distribution, volume fraction, and gas flow rate. Experimental validation of the theoretical predictions will be made for different combinations of fine powder distribution, and gas flow rates. The L-valve is used extensively in the operation of fluidized beds and circulating fluidized beds for solids circulation rate control. Knowledge on L-valves, however, is mostly limited to ambient flow conditions, particularly with respect to characterization of the Geldart A and B types of particles. Studies of the L-Valve under high temperatures which are common in industrial applications are rare. This presentation will describe the experimental studies of the characteristics of the L-valve operation under high temperature conditions with Geldart D particles. The relationships among the solids flow rate, aeration rate and pressure drop will be discussed along with the comparison under ambient conditions for different types of particles. The extent of the solids flow rate controlled with the L-valve aeration depends on the actual internal flow rate of the gas that is flowing through the L-valve rather than the external aeration flow rate of the gas that is introduced into the system. The extent of the actual aeration gas flow rate needed for generating the same extent of the solids flow rate is the same for different aeration tap locations as long as the tap locations are above the height-to-diameter ratio of 1.5. Operating L-valve with Geldart D particles needs more gas aeration compared to that with Geldart A and B particles for a given solids flow rate. However, under high temperature conditions, Geldart D particles flow behaves like Geldart A particles. The presentation will illustrate the aeration gas flow rate that is required under high temperature conditions is much smaller than that under ambient conditions for Geldart D particles.

Talk: Keynote Abstract: Fluidized bed reactors are widely used in the chemical processing, petroleum refining, power generation, specialty materials and waste conversion industries for carrying out gas-solids catalyzed and gas-solid non-catalytic reactions. However, fluidized beds also present outstanding challenges in understanding the gas-solids mixing, contacting patterns, gas-particle and particle-particle interactions as well as the combined effects of these phenomena on gas-phase homogeneous and gas-solid heterogeneous reactions. Computational fluid dynamics (CFD) has emerged as an approach for more rational design and scale-up of fluidized bed reactors. However, CFD models require parameters based upon phenomenological models. Two types of simulation methods, i.e., Eulerian-Eulerian and Eulerian-Lagrangian, are described in this presentation. For the Eulerian method (also called two-fluid model, or TFM), in which the discrete solid particles are assumed as a continuous granular phase interpenetrated with the gas, constitutive models are needed for closure of the solids phase stress tensor. The most commonly used models are derived from the kinetic theory for granular flow (KTGF), (e.g., Jenkins and Savage, 1983; Lun et al., 1984; Gidaspow, 1994). These models assume that during an inelastic collision, only the normal component of the relative velocity is decreased, while the tangential component remains unaltered due to the assumption of smooth particles. Here, it is assumed that the velocity magnitude that is dampened is due to an inelastic collision between two smooth or nearly smooth particles, rather than just the normal velocity component. Based on this assumption, new phenomenological models arise for solids pressure, shearing and bulk viscosity, thermal conductivity, and most importantly for the dissipation rate, which is found to be almost twice as large as the original one. The proposed models reduce to those for dense gases (Chapman and Cowling, 1970) when the coefficient of restitution e is equal to one, as does Gidaspow (1994). The stress tensor now becomes asymmetric without the need to assume a different restitution coefficient in the tangential direction, and to add transport equations for angular velocity and rotational fluctuation energy, as indicated by Lun and Savage (1987) and Lun (1991). Therefore, the proposed models predict less underestimated dissipation rate without extra complexity. The models are implemented in OpenFOAM to simulate a fluidized bed. These show that the new collision model provides an improved the prediction of the solids volume fraction without extra computational cost. In the Lagrangian or DEM (discrete element method) approach, the effort is motivated by Bhusarapu’s (2005) pioneering finding using CARPT (computer-aided radioactive particle tracking) that the traditional tracer method for measuring solids residence time distribution (RTD) cannot capture the actual residence time due to inability to distinguish the time that a particle temporarily spends out of the riser after first entry to or before last exit from the riser. By implementing the algorithm developed in this work to record separately the time that each particle actually spends in the riser, Lagrangian simulation is performed for a small circulating fluidized bed in OpenFOAM. The results clearly demonstrate the difference between RTD obtained from simulated traditional tracer method and that from the Lagrangian approach. Other information can be extracted as well such as the first passage time distribution, macromixing index, and interchange coefficient between a core region and an annular region, if a core-annulus model is used. Opportunities for other advancements by including other multiphysics will also be summarized.

Prosperetti/Wachs/Derksen/Juric

Talk: Plenary Abstract: Direct numerical simulations (DNS), where every continuum length and time scale is fully resolved, allow us to follow the evolution of complex flows for sufficiently long time so that meaningful statistical quantities can be gathered. Results for relatively simple multifluid and multiphase systems with bubbles and drops in turbulent flows are now available, but new challenges are emerging. First of all, DNS of very large systems are yielding enormous amount of data that, in addition to providing physical insights, opens up new opportunities for the development of lower order models that describe the average or large-scale behavior. Recent results for bubbly flows and the application of statistical learning tools to extract closure models from the data suggest one possible strategy. Secondly, success with relatively simple systems calls for simulations of more complex problems. Multiphase flows often produce features such as thin films, filaments, and drops that are much smaller than the dominant flow scales and are often well-described by analytical or semi-analytical models. Recent efforts to combine semi-analytical models for thin films using classical thin film theory, and to compute mass transfer in high Schmidt number bubbly flows using boundary layer approximations, in combination with fully resolved numerical simulations of the rest of the flow, are described.

Talk: Keynote Abstract: Many engineered processes rely on mass transfer between a solids phase and a liquid phase. Agitation and fluidization of dense solid-liquid suspensions are common ways of enhancing transfer rates. Suspensions have large interfacial area and the slip velocities between the phases help in transporting chemical species towards and away from the interfaces where in many cases surface-reactions occur. We present detailed simulations of flow dynamics and mass transfer in dense suspensions with explicit resolution of the solid-liquid interfaces. There are a number of challenges that will be discussed in this presentation: (1) Particle shape: resolving the flow around non-spherical particles and dealing with the frequent collisions between non-spherical particles in dense suspensions. (2) High Schmidt numbers associated with liquids ask for dedicated methods for sufficient resolution of the (fine) mass-transfer boundary layers. (3) Industrial liquids often exhibit complex rheological behavior that needs to be incorporated in numerical simulations.

Talk: Regular Abstract: A large class of hydrodynamic problems involves interfacial instabilities and within that class, phase change is an important physical process. Examples of phase -change phenomena are evaporation, solidification, and electrodeposition. They play an important role in materials processing, food preservation and in energy management. The interfacial instability associated with phase-change processes lead to pattern formation. Determining the conditions and form of pattern selection depend on the judicious use of analytical and computational schemes. This talk will focus on the physical and mathematical analogies between various phase-change instabilities. We will relate the results of theory arising from analysis and computations and compare them to experiments, showing the formation and selection of patterns.

Talk: Regular Abstract: Since the advent and popularization of the lattice Boltzmann method in the mid-nineties, this novel computational fluid dynamics approach has emerged as the method of choice for the simulation of flows through porous media. The success of this method in this context stems from its ease in discretizing complex geometries by means of a simple structured lattice on which the fluid and solid phases are encoded in a Boolean manner, and the inherent locality of its scheme, which makes it straightforwardly suitable for parallelization on distributed computers. Flows through computational domains involving billions of lattice cells can now readily be simulated on reasonably sized supercomputers and this opens up never-seen-before opportunities for improving classical porous media theory. Examples on how the lattice Boltzmann method has been used in our research group for predicting the impact of particle size polydispersity on the permeability of sphere packings and fibrous filters will be presented, as well as some practical applications.

Talk: Regular Abstract: Low-order modelling of the flow of fluid films has typically relied on the assumption that the thickness of a fluid is `small’ relative to both the characteristic wavelengths involved in the flow, and relative to the radius of curvature of the substrate. This is a significant imposition as many flows do not obey these constraints. Under a long-wave approximation, we use a symbolic algebra package to derive low-order models for low-order models of thick flows on substrates that are highly curved. We demonstrate that these models have a high degree of accuracy via comparison against direct numerical simulations.

Talk: Plenary Abstract: Since many years CFD (computational fluid dynamics) is applied for numerical calculations of the very complex flows in bubble columns using the two-fluid approach as well as the Euler/Lagrange approach, both based on the point-particle approximation. Numerous different models have been proposed and used for describing the forces on bubbles, modelling bubble induced turbulence and transport of bubbles by turbulence structures. The dynamics of bubbles, i.e. oscillations and tumbling motion have not been considered so far. Such a model has been developed, implemented in the frame of Euler/Lagrange calculations and validated based on detailed experiments (Sommerfeld and Bröder 2009). In this framework, a CFD model is developed and implemented in an open source platform (OpenFOAM®), based on the Euler/Lagrange approach. Flow field and turbulence of the carrier phase was modelled by the Large Eddy Simulation (LES) approach, considering also bubble induced modification of sub-grid-scale (SGS) turbulence, described by the Smagorinsky model (Lain et al. 2002). In bubble tracking all relevant forces such as drag, gravity/buoyancy, transverse lift, pressure and added mass are considered. The time step for Lagrangian tracking is dynamically adapted according the local relevant time scales. The effect of sub-grid scale turbulence on bubble motion was described by a stochastic single-step Langevin equation, based on Lagrangian and Eulerian time scales (Lipowsky and Sommerfeld 2007). LES and Lagrangian tracking is done with different time steps governed by the respective time scales, so that within each LES time step multiple Lagrangian time steps are conducted. This procedure reduces computational time. Moreover, the bubble dynamics in the point-particle approximation was modelled by stochastic variations of bubble shape and orientation according to experimental observations. For that purpose a Langevin model is used for describing the development of the eccentricity and orientation having a correlated and a random part. The correlation function depends on the Lagrangian time step and the time scale of oscillation (Lunde and Perkins 1998). Drag and lift coefficients are calculated by considering bubble eccentricity. Additionally, the bubble direction of motion is altered by a Langevin random process in order to mimic bubble tumbling motion. Numerical simulations, including comprehensive parameter studies (e.g. different force correlations coupling between the phases and oscillation model parameters), were compared with experiments from Sommerfeld and Bröder (2009) using a laboratory bubble column with a diameter of 140 mm and a height of the water level of 650 mm. The conclusion from this study is that bubbly dynamics needs to be modelled in numerical calculations in order to correctly predict their behaviour which will be of immense importance when also considering mass transfer and chemical reaction.

Talk: Keynote Abstract: Solid-liquid mixing plays a key role in numerous industrial processes ranging from chemical reactors with solid catalyst, crystallizers to the production of a large variety of consumer goods such as pastes, paints, cosmetics, pharmaceuticals and food products. Despite its industrial relevance, this field still faces considerable challenges. Indeed, it remains highly difficult to predict critical parameters such as the impeller torque, the just-suspended speed (Njs) and the homogeneity of the suspension. To shed light on these issues related to solid-liquid mixing, numerical and experimental work is essential. In the present work we present a novel numerical model to study solid-liquid mixing. This model is based on an unresolved CFD-DEM strategy that combines CFD techniques for the fluid flow and the discrete element method (DEM) for the particle dynamics. This model, based on the CFDEM framework, is implemented using OpenFOAM for the fluid phase and LIGGGHTS for the solid phase. It can simulate efficiently and in a robust manner solid-liquid flows in stirred tanks for any flow regime and a wide range of particle concentrations. First, the governing equations for the liquid and solid phases are introduced along with the two-way coupling strategy between the particles and the fluid. The model is applied to the study of non-dilute solid-liquid mixing in a stirred tank provided with a pitched-blade turbine in the laminar, transitional and early turbulent regimes. A qualitative comparison of numerical results and experimental flow patterns is used to assess the adequacy of the CFDEM framework in the case where the fluid is viscous and the flow is laminar or transitional. The model is also validated quantitatively by comparing for different impeller speeds the fraction of suspended solids to experimental data obtained via the pressure gauge technique (PGT). The model is then used to study the case of a less viscous fluid, for which the flow is turbulent. There again, the fractions of suspended solids predicted with our model are validated against experimental measurements obtained with the PGT.

Sommerfield/Zoric/Matar/Ranade

Talk: Plenary Abstract: Particulate flows are ubiquitous in environmental, geophysical and engineering processes. The intricate dynamics of these two-phase flows is governed by the momentum transfer between the continuous fluid phase and the dispersed particulate phase. When significant temperature differences exist between the fluid and particles and/or chemical reactions take place at the fluid/particle interfaces, the phases also exchange heat and/or mass, respectively. While some multi-phase processes may be successfully modelled at the continuum scale through closure approximations, an increasing number of applications require resolution across scales, e.g. dense suspensions, fluidized beds. Within a multi-scale micro/meso/macro-framework, we develop robust numerical models at the micro and meso scales, based on a Distributed Lagrange Multiplier/Fictitious Domain method and a two-way Euler/Lagrange method, respectively. Collisions between finite size particles are modeled with a Discrete Element Method. Many real-life processes and/or flows involve non-spherical particles. Although there is still a lot to learn about flows laden with spherical particles, there is also a strong incentive to develop new modeling tools to account for non-spherical, angular, convex or even non-convex particles. At the micro scale, the main challenge is the modeling of collisions. At the meso scale, the consideration of non-spherical particles implies to derive appropriate closure laws for hydrodynamic interactions. For the former, we describe the numerical ingredients of our granular solver that handles both convex and non-convex particles. The latter is an open field in the literature. The classical two-way Euler/Lagrange approach for spherical particles normally extends straightforwardly to non-spherical particles, provided the particle aspect ratio is not too large. However, the determination of the dominant part of the hydrodynamic interaction and of the corresponding closure laws has not received a significant attention from the community so far. Finally, we shortly discuss some of the high performance computing issues related to our massively parallel numerical tools and illustrate their modeling capabilities on the two following problems relevant of applications from the chemical engineering and process industry: (i) a rotating drum filled with non-convex particles and (ii) fixed and fluidized beds of multilobic (and hence non-convex) particles.

Talk: Keynote Abstract: The paper will start by briefly describing the Physalis method for the fully resolved simulation of Navier-Stoked fluid flows with suspended spherical particles. Several applications will then be described. A method to extract the speed of concentration waves from the simulations of a fluidized-bed-like system is outlined and the results are compared with information available in the literature. The statistical geometry of the evolution of particle tetrads in the same system is described. The statistics of the rotation of a particle immersed in an incident turbulent flow is compared with the statistics of the flow vorticity. It is shown that the particle is only affected by the flow vorticity at scales larger than itself. Several other results will be briefly shown to demonstrate the capabilities of this simulation method.

Talk: Regular Abstract: This study presents a development of the direction splitting algorithm for problems in complex geometries proposed [1] to the case of flows containing rigid particles. The main novelty of this method is that the grid can be very easily fit to the boundaries of the particle and therefore the spatial discretization is very accurate. This is made possible by the direction splitting algorithm of [1]. It factorizes the parabolic part of the operator direction wise and this allows to discretize in space each of the one-dimensional operators by adapting the grid to fit the boundary only in the given direction. Here we use a MAC discretization stencil but the same idea can be applied to other discretizations. Then the equations of motion of each particle are discretized explicitly and the so-computed particle velocity is imposed as a Dirichlet boundary condition for the momentum equations on the adapted grid. The pressure is extended within the particles in a fictitious domain fashion. Finally, the presentation will demonstrate the accuracy and stability of the method on various benchmark problems involving rigid particles (see [2]). In addition, some results of direct simulations of fluidized beds involving thousands and millions of particles will be presented. Further details of these simulations can be found in [2]. [1]. Ph. Angot, J. Keating, P. Minev, A Direction Splitting Algorithm for Incompressible Flow in Complex Geometries. Comp. Meth. Appl. Mech. Engng. 217 (2012), 111–120. [2]. J. Keating, P. Minev, A Fast Algorithm for Direct Simulation of Particulate Flows Using Conforming Grids. J. Comp. Phys. 255 (2013), 486–501.

Talk: Regular Abstract: Dense particulate systems can exhibit complex flow structures and diverse flow patterns similar to those observed in their fluid-fluid counterparts. The pattern/structure formation is attributed to two interactions: the particle-fluid (P-F) and the particle–particle (P–P) interactions. CFD-DEM has proven to be an effective numerical tool to capture complex flow behavior in dense particulate systems. In this talk, I will discuss the challenges and developments of modeling dense particulate flow under CFD-DEM framework from both numerical and physical perspectives. Ongoing work on development of a new drag law by considering the effect of local heterogeneous structures will be highlighted.

Talk: Regular Abstract: Predictive simulations based on multiphase computational fluid dynamics (CFD) have become indispensable for scaling up multiphase devices used in energy technologies. The predictive accuracy is subject to various uncertainties in the simulation. Quantifying them is essential for making engineering decisions based on simulations, which reduces the time and cost of scaling up new multiphase devices. Uncertainty quantification (UQ) and sensitivity analysis also help engineers to plan validation experiments and reduce the risk in scale up. UQ in multiphase CFD is currently a major research focus at National Energy Technology Laboratory (NETL). This presentation will review the methods explored at NETL and the recent progress. A validation methodology is applied to the simulations of a circulating fluidized bed. The overall pressure drop is the quantity of interest; the solids circulation rate and the gas velocity are the uncertain input quantities. From the known uncertainties in the input quantities, surrogate model, spatial discretization and time averaging, the uncertainty in the pressure drop is calculated, and the model-form uncertainty is determined with the help of validation data. The UQ results are expressed as a p-box plot, which can provide answers to various design questions. A Bayesian calibration methodology is applied to conduct predictive simulations of a carbon capture device. This methodology handles situations where many parameters describing different physical phenomena appear in a complex multi-physics problem, and a hierarchy of experiments is required for model validation. It enables progressively improving the model by learning from new information provided by the model validation hierarchy. Sensitivity and grid convergence analyses have been used to understand a gasifier model. Calculating the uncertainty resulting from grid convergence has been identified as an issue. An error transport equation formulation is being developed as an alternative to standard methods for calculating that uncertainty.

Talk: Plenary Abstract: The ability to predict the behaviour of multiphase flows accurately, reliably, and efficiently addresses a major challenge of global economic, scientific, and societal importance. These flows are central to micro-fluidics, virtually every processing and manufacturing technology, oil-and-gas, nuclear, and biomedical applications. Although significant advances have been made in the numerical procedures to simulate multiphase flows, there remains a large gap between what is achievable, and ‘real-life’ systems. As a result, empirical correlations remain widely used in order to bridge this gap. We will present the latest on the modelling framework that we are currently developing as part of the Multi-scale Examination of MultiPHase physIcs in flowS (MEMPHIS) programme in order to minimise the use of correlations and shift towards the use of numerical simulations as a truly predictive tool that can be used as a sound basis for design. The framework features model-driven experimentation (which we will not discuss here), massively-parallelisable interface-capturing methods, 3D, adaptive, unstructured meshes, and sophisticated multi-scale, multi-physics models. Support from the Engineering & Physical Sciences Research Council, UK (grant no. EP/K003976/1) is gratefully acknowledged.

Talk: Keynote Abstract: We present an approach for direct numerical simulations of multiphase flows on several to over 100 000 parallel processes. The code we call BLUE is based on a dual Front Tracking/Level Set approach and accurately handles capillary forces and complex interface dynamics including changes of topology resulting from breakup. Modules for flow, mass and heat transport are designed with particular attention to parallel memory management. We demonstrate solutions in a variety of challenging multiphase applications in microfluidics, free surface and interfacial flows as well as interactions of multiphase flows with solid structures. We also describe the architecture and discuss the parallel performance of the code.

Joshi/Jaffer/Syamlal/Pannala

Meet in the TCPL foyer for the group photo. Weather permitting, it will be taken outside, so you might want a jacket.

Talk: Plenary Abstract: Fluidized suspensions of particles manifest inhomogeneities spanning a wide range of length and time scales. Our research examines how one could capture the effects of these inhomogeneities on fluid-particle drag force. To probe this, we have performed simulations of flows at different scales: (a) at the particle-scale through Lattice-Boltzmann simulations; (b) at a larger scale through Euler-Lagrange Simulations tracking collisions between particles through Discrete Element Method, and (c) at an even larger scale through two-fluid model simulations. Snapshots generated through these simulations are analyzed to identify the pattern of dependence of the drag force on inhomogeneities that persist at all scales. It is found that the drag force model should include dependence on particle Stokes number and the local mean-squared fluctuations in the particle volume fraction occurring at a scale smaller than the filter scale. A dynamic, scale-similar model is proposed for the local mean-squared fluctuations in the particle volume fraction, which would not be available in coarse simulations. This improved model should lead to higher-fidelity simulations, thus enabling innovations to be tested initially through simulations before engaging in costly experimental testing.

Talk: Keynote Abstract: Recent progress on large-scale heterogeneities in high-velocity, vertical gas-solid flows will be surveyed. From a theoretical perspective, the quantative accuracy of kinetic-theory-based continuum descriptions will be critically assessed based on comparison with DNS data. A framework for the incorporation of cohesive forces that links micro-scale forces with macro-scale behavior will also be outlined. From an experimental perspective, recent shadowgraphy data of small agglomerates in riser flow will be presented, as well as accompanying CFD-DEM simulations illustrating differences between the nature of large-scale heterogeneities in non-cohesive vs. mildly cohesive flows. This talk will thus take the form of a survey of recent results from the group that can be discussed in greater detail throughout the workshop.

Talk: Regular Abstract: Developing new and efficient commercial scale chemical reactors is extremely difficult as one needs to consider the complex interactions over a wide range of both temporal and spatial scales encountered in these systems (from molecules to macroscale). Therefore, it is important to codify and automate the knowledge collectively acquired so far, reserving human resources for the creative solutions that build upon codifying past learnings. Computational science (algorithms, theory and modeling, computer science, etc.) combined with exponential growth in computing hardware has great potential to revolutionize the way science and engineering has been performed. For this reason, computational science is often called the third pillar of modern science, complementing observations and theory. The complexity of the problem is introduced using biomass gasifier/pyrolyser, nuclear fuel coater, and fluidized bed polymerization reactors as examples and I will provide an overview of the various models currently used at the different scales. The coupling across the scales will be introduced through few different approaches: a) Discrete-Continuum Coupling using Discrete Element Method for particles b) A wavelet based technique for coupling multiple modeling approaches at different scales c) Upscaling data from Discrete Element Modeling results to continuum

Talk: Regular Abstract: Boundaries for hydrocarbon exploitation are being increasingly stretched towards remoter and deeper spots around the Globe. This entails recourse to floating production systems as an alternative to conventional off-shore oil platforms such as Deep Draft Semi Sub, Extendable Draft Platform, and so forth. Floating platforms are commonly integrated with floating units such as Floating Storage and Offloading (FSO), Floating Production Unit (FPU), Floating Liquefied Natural Gas (FLNG) and Floating Production Storage and Offloading (FPSO) which are used to replace costly pipeline infrastructures and onshore refining-treating facilities. It is thus not a surprise that development and application of floating units, e.g., FPSO and FLNG in deep-water oilfields, are subject to vivid research by the petroleum industry. One of the challenges confronting well-designed units resides on how the efficiency and performance of offshore facilities correlate with the restless sways caused by marine swells and how these latter impact the hydrodynamic characteristics of the reactors that are embarked on-board. In this presentation, we will discuss our recent results on multiphase packed-bed reactors mounted on a six-degrees-of-freedom hexapod robot to emulate swell movements and to analyze the hydrodynamic alterations brought about by separate or combined degrees of freedom under yaw and pitch rotations, and jerky swell movements versus stationary (straight and inclined) bed configurations. A twin-plane capacitance Wire Mesh Sensor (WMS) installed on the moving packed beds is used to measure the dynamic features of local phase distribution patterns, local and averaged liquid saturations and velocities, and flow regime changes under various configurations, e.g., concurrent two-phase upflow, downflow and drainage mode. Deviations from well-known behavior of straight and stationary packed-bed two-phase flows will be highlighted, quantified and interpreted.

Talk: Regular Abstract: In most technical applications the particles are not spherical, as usually considered in experiments and numerical calculations, but are non-spherical having either a regular shape (e.g. fibres, cylinders, granulates or disc-like particles) or are even completely irregular such as quartz sand and pulverized coal. With respect of the flow transport of such particles they mostly exhibit larger Reynolds numbers so that a creeping flow or Stokes assumption is not valid for obtaining the flow resistance coefficients (Loth 2008). Numerous experimental studies are available where the drag coefficient of non-spherical particles with a certain orientation is evaluated for higher particle Reynolds numbers based on wind tunnel or sedimentation experiments (see e.g. Haider and Levenspiel 1983). However, when dealing with non-spherical particles in a numerical computation (e.g. by the Euler/Lagrange approach) particle tracking involves determination of their location as well as orientation in the flow. Consequently, drag coefficients for higher particle Re are needed which are depending on particle orientation (Hölzer and Sommerfeld 2008). So far the determination of orientation-dependent resistance coefficients (i.e. drag, lift and torque) was mainly based on DNS (direct numerical simulation) as for example done by Hölzer and Sommerfeld (2009) and Zastawny et al. (2012) for regularly shaped non-spherical particles. Since the resulting dependence of resistance coefficient on orientation are usually rather complex mostly approximations are used in an Euler/Lagrange calculation (see e.g. van Wachem et al. 2015). Therefore, detailed experimental studies are needed on the transport and turbulence response of non-spherical particles for validating such numerical calculations, which are so far very rare. In the present study a fibre laden small jet (di = 5 mm) issuing into a fully developed turbulent cross-flow through a channel of square cross-section (100 mm x 100 mm) was considered. The objective of this study was providing data on the dispersion of fibres in the vortical structures of the cross-jet for different Stokes numbers and also in comparison with spherical particles with the same Stokes numbers (Pasternak and Sommerfeld 2015). The imaging technique applied is based on shadow imaging for avoiding disturbing light scattering effects when using a light sheet. A pulsed LED array was used for illumination with a pulse duration as low as 20 µs for avoiding motion-induced burring. As tracer almost neutrally buoyant 40 µm PMMA particles were used. Images of both phases were collected using a high-speed camera at a framing rate of 3600 Hz and an image size of 768 x 640 pixels. The applied objective yielded an imaging field of 40 mm x 33 mm with a small depth of field providing an effective imaging layer thickness of only 1.5 mm in connection with the adopted image filters. These were a LOG-filter for removing all out-of-focus images and an object-based filter using object size and sphericity to obtain separate images of tracer and fibres. The velocity fields of the tracer were obtained by the MQD (minimum quadratic difference) algorithm (similar to PIV) and the fibre velocities were determined by particle tracking (PIV). In addition fibre orientation and angular velocities were evaluated throughout the flow field. Furthermore, various correlations were determined for characterizing the fibre response in the cross-jet. These detailed data are available for the validation of numerical computations.

Talk: Regular Abstract: Multiphase flows where two or more fluids have interfacial surfaces are often found in industrial engineering applications, including bubbles and droplets. The objective of this study is to investigate the fluid dynamics of three-dimensional two- and three-phase flow problems, such as falling liquid films, liquid jets, droplet impacting upon a gas-liquid interface and bubble rising through a liquid-liquid interface. An adaptive unstructured mesh modelling framework is employed here to study two- and three-phase flow problems, which can modify and adapt unstructured meshes to better represent the underlying physics of multiphase problems and reduce computational effort without sacrificing accuracy. The numerical framework consists of a mixed control volume and finite element formulation, a ‘volume of fluid’ type method for the interface capturing based on a compressive control volume advection method and second-order finite element methods, and a force-balanced algorithm for the surface tension implementation. Numerical examples of some benchmark tests and the dynamics of two- and three-phase flows are presented to demonstrate the capability of this method.

Talk: Plenary Abstract: Multiphase flows are ubiquitous in environmental and engineering applications. One category of flow of great importance in energy conversion devices is the formation of a liquid spray, a process called atomization. Due to their nonlinear and multiscale nature, such liquid-gas flows present a significant modeling challenge, especially when novel control strategies such as electro-hydrodynamics are considered. In addition, flow variables exhibit discontinuities across the phase interface, leading to numerical difficulties. Another category of flow of importance for energy conversion is dense particle-laden flows, as found in fluidized bed reactors. These flows are strongly multiscale, and momentum coupling between the gas carrier phase and finite-sized particles can lead to the production of gas-phase kinetic energy fluctuations, leading to cluster-induced turbulence which needs to be modeled. With the advent of more powerful computing resources, simulating such flows from first principles is becoming viable. As with single-phase flows, numerical methods need to be carefully designed to guarantee convergence under grid refinement, primary conservation of key quantities such as mass and momentum, and excellent parallel performance. We will discuss how such properties can be obtained in the context of various multiphase turbulent flows, including atomizing liquid jets and fluidized beds, as well as three-phase flows.

Talk: Keynote Abstract: Multiphase systems are ubiquitous in industrial applications aimed at the generation of products either by chemical/biological reaction or physical separation based on density, electrical charge or surface properties such as hydrophobicity. The physical processing of these multiphase systems is carried out at all scales of operation and within an endless variety of vessel shapes and ancillary devices. Underpinning each process is a complex interaction between phases involving hydrodynamic, heat and mass transport. At Newcastle, we are focusing on visualizing, quantitative measurement and theoretical and computational modelling of the phenomena that are taking place at the phase boundaries in multiphase systems. In particular, we are developing both modelling and measurement techniques to evaluate both spatial and temporal distribution of energy dissipation rates. This information is then being fundamentally related to the rate of heat/mass transfer, dispersion, breakup-coalescence, and hydrodynamic stability. Finally, these relationships are being used as the foundation for the development of systems that provide ideal energy dissipation rate at desired locations and at desired length scales. Our latest research, including innovative measurement approaches as well as analytical, CFD, DEM and DNS modelling approaches, will be presented for fluidised beds, novel mineral flotation approaches, and high temperature reactors.

Talk: Regular Abstract: The effects of polydispersity of particles on thermodynamic and hydrodynamic behavior of solid particle solar receivers (SPR) are studied. SPRs are alternative designs for thermal solar receivers, in which solid particles, laden in a carrier gas, are used to directly absorb concentrated solar radiation. For energy conversion, the heated gas drives a turbine and/or the heated solid particles are used in chemical reaction processes. The main advantages compared to the conventional thermal solar receivers are volumetric heat absorption and transfer by particles, more uniform heat transfer in the gas, and significantly less heat loss to the ambient. However, the challenge is to maximize the efficiency of radiation absorption and heat transfer by particles given their short residence time and interaction with turbulence. The flow in SPRs involves a turbulent mixture of gas and particles in a radiation environment. Investigating the performance of SPRs requires understanding the complex three-way coupling between the gas flow, particle transport and interfacial heat transfer. Here, we use direct numerical simulations of turbulent flows interacting with point particles subject to a heating source to study physical processes in SPRs. The particles are polydisperse, i.e. they have a specific cumulative distribution function (CDF) for particle size. A thermodynamically equivalent monodisperse particle size is calculated for this CDF so that the mass loading ratio of particles to gas and the total frontal area of particles available for radiation absorption are matched between the polydisperse and monodisperse particles. Our results show that the energy balance of the gas and particle phase are significantly different between monodisprese and polydisprse particles. Polydisperse particles are more efficient in transferring heat to the gas phase and generating a uniform gas temperature field. The spatial inhomogeneity of particle concentration due to particle clustering adversely impacts the heat transfer from particles to gas. By quantifying the inhomogeneity in particle concentration by radial distribution functions (RDF), we show that polydisperse particles are more uniformly distributed compared to their counterpart monodisperse particles. We show this is mainly because of the different preferential concentration patterns in different particle size classes in polydisperse particles. However, the heat transfer in polydisperse particles cannot be represented as linear superposition of different classes of monodisperse particles.

Talk: Regular Abstract: Petrochemical industry is facing significant economic, environmental and societal challenges. Technology research, development and deployment will be crucial to meeting these challenges and capturing future market growth opportunities. With an ever-increasing focus on the development of products that generate revenues, the industry will need to find ways to enhance profitability and capital productivity. Long-term and strategic investments in R&D and new technologies can drive the industry towards higher levels of financial performance. An important element of this R&D is the development of enabling technologies that advance the application of fundamental chemical sciences throughout the industry’s process environment. Technologies to assist industrial computations may be considered “enabling” in this regard since they are used in nearly every aspect of chemical research, development, design and manufacture. Among these computational technologies, Computational Fluid Dynamics (CFD) and Process Simulation & Modeling (PSM) have grown over the years as highly sophisticated integration of applied computer science, physics, chemistry, and engineering science put together to enable innovation throughout the process industry. In this presentation, we give an example of both CFD and PSM have being used to optimize and troubleshoot two important petrochemical processes, namely, fluid catalytic cracking and acid gas stripping deployed to produce fuel and chemical feedstock in the region of Abu Dhabi

Talk: Regular Abstract: Multiphysics CFD simulations on HPC platforms provide a great opportunity to learn about the complex processes during drilling and completions operations of oil & gas wells. Several computational fluid dynamics (CFD) models with different features are presented for cuttings transport, cement placement, and production through completions in this talk. All simulation cases are both verified and validated against available experimental data for their corresponding physics. In order to get accurate flow predictions while optimizing computational resources requirements, unsteady shear stress transport (SST) k-ω turbulence model is used to model turbulence closure while solving Reynolds-averaged Navier-Stokes (RANS) equations using unstructured finite volume method (FVM) for discretization. Discrete phase is modeled with discrete element method (DEM) by including particle-particle and particle-fluid interactions with two-way coupling in Eulerian-Lagrangian simulations. Volume of Fluid (VOF) model is used to model displacement of different fluid types with non-Newtonian fluid rheology for cement placement applications. Specifically, during the drilling of highly deviated wellbores, the cuttings transport becomes difficult due to the rolling/sliding transport of the cuttings due to settling around the lower side of the annular region between wellbore and drillpipe. Inefficient cuttings transport may lead to several critical problems such as stuck pipe, increased torque and drag, damaged material and poor quality of cementing jobs. Increasing mud flowrates and improving mud properties for a proper wellbore cleaning is usually limited due to the hydraulic and mechanical thresholds for wellbore formation integrity. Further, understanding of cement placement process remains a critical step in achieving zonal isolation between casings and hydrocarbon bearing formations in all types of well construction operation. Lastly, a gravel-packed completion is modeled to showcase the capabilities of CFD simulations by gaining new insights into modeling and representation of high-rate producer wells in reservoir simulators.

Talk: Plenary Abstract: Multiphase flows with phase change are used in industrial practice for a large variety of applications. Applications of flows with phase change may be broadly classified into three categories: • Applications which release energy: combustion, gasification • Applications which make released energy more useful: boiling, cavitation, condensation • Applications which create specialized solid products: crystallization, solidification Combustion of solid or liquid fuels is extensively used for industrial applications which involve phase change from solid to gas or liquid to gas. Electricity generation from coal, oil or nuclear essentially relies on boiling and steam generation. Cavitation can be used for variety of applications ranging from process intensification to water dis-infection. Large numbers of pharmaceutical products are in solid form and are produced by crystallization. Solidification processes are also important for many large systems from fast breeder reactors to polymerization reactors. The large process equipment used for managing multiphase flows with phase change contribute significantly to the overall capital cost, reliability and overall operating effectiveness & margins. In this talk, I will share our experiences of modelling multiphase flows with phase change. Attempt will be made to highlight industry needs and to provide some suggestions on path forward. Co-author: Dr Gopal Kasat, Tridiagonal Solutions (www.tridiagonal.com)

Talk: Keynote Abstract: Many natural or industrial processes are of extreme complexity, where the time- and length scales range from an atomistic level to years and kilometers [1]. Often the processes or phenomena consist of multiple sub processes in which each comprises its own length and time scales. To optimize such production processes with respect to economic and environmental parameters we will have to develop pragmatic models, which can give the overall picture and at the same time be accurate enough to support the optimization process [1]. In this presentation we discuss pragmatic industrial modeling paradigm by three perspectives: (1) process-centric perspective, (2) data-centric perspective, and (3) knowledge/information service perspective: • Ref. [1] suggested a process-centric perspective on a frameworks (FWs) for pragmatic industrial modeling and demonstrated it on industrial use cases. We identified and discussed the tools needed for such analyses, including the analyses process itself and the frameworks needed for such analyses. We applied our pragmatic modeling framework to the aluminum production process and discussed the implications of our proposed concept. • Ref. [2] elaborated on data-centric perspective of pragmatic industrial models, by focusing on modeling and experimental data and metadata, their organization, syntax and semantics. We exemplified it on a use case, related to drilling of oil & gas wells. The analytical problem was to produce a model that can predict the motion of a spherical particle embedded in laminar non-Newtonian flow. An overview of tasks, procedures, organization, structure and flow of data (to/from various modeling and experimental phases), required metadata, and technical and quality requirements was provided. • Ref. [3] reports a work in progress and takes a knowledge/information service perspective. We discuss relations between modeling/simulation FWs and industrial decision support systems (DSS): o Requirements that pragmatic models and simulations have to satisfy (to become a part of the Industrial DSS). o Structure and organization of the modeling FWs: architecture entities and interfaces. o Knowledge/information service view (Modeling-as-a-Service): service design and implementation alternatives. We summarize the findings of these three research efforts [1-3] and discuss how the "pragmatism in industrial modeling" concept can help building consistent industrial modeling frameworks, answering to customer needs for actual accuracy and delivery speed. References: [1] Zoric J., Johansen S.T., Einarsrud K.E., Solheim A., “On pragmatism in industrial modeling.”, Proceedings of 10th International Conference on Computational Fluid Dynamics in the Oil & Gas, Metallurgical and Process Industries (S. T. Johansen and J. E. Olsen, eds.) pp.1-16, SINTEF, Trondheim, ISSN: 2387-4295, ISBN: 978-82-536-1433-5. [2] Zoric J., Busch A., Meese E.A., Khatibi M., Time R.W., Johansen S.T., Rabenjafimanantsoa H. A., “ On Pragmatism in Industrial Modeling, Part II: Workflows and Associated Data and Metadata, Proceedings of 11th International Conference on CFD in the Minerals and Process Industries, Melbourne, ISBN 978-1-4863-0620-6. [3] Work in progress.

5-day workshop participants are welcome to use BIRS facilities (BIRS Coffee Lounge, TCPL and Reading Room) until 3 pm on Friday, although participants are still required to checkout of the guest rooms by 12 noon.