# Publications

Either browse through the list of published material or use the search below to limit by, year, or author or document type.

## 2009 |

Stephens, Darrin W; Gorissen, Dirk; Dhaene, Tom Surrogate Based Sensitivity Analysis of Process Equipment Conference Seventh International Conference on CFD in the Minerals and Process Industries, 2009. Abstract | Links | BibTeX | Tags: RBF, Surrogate @conference{162009surrogate, title = {Surrogate Based Sensitivity Analysis of Process Equipment}, author = {Darrin W Stephens and Dirk Gorissen and Tom Dhaene}, url = {http://www.researchgate.net/publication/266138591_SURROGATE_BASED_SENSITIVITY_ANALYSIS_OF_PROCESS_EQUIPMENT}, year = {2009}, date = {2009-01-01}, booktitle = {Seventh International Conference on CFD in the Minerals and Process Industries}, abstract = {The computational cost associated with the use of high-fidelity Computational Fluid Dynamics (CFD) models poses a serious impediment to the successful application of formal sensitivity analysis in engineering design. Even though advances in computing hardware and parallel processing have reduced costs by orders of magnitude over the last few decades, the fidelity with which engineers desire to model engineering systems has also increased considerably. Evaluation of such high-fidelity models may take significant computational time for complex geometries. In many engineering design problems, thousands of function evaluations may be required to undertake a sensitivity analysis. As a result, CFD models are often impractical to use for design sensitivity analyses. In contrast, surrogate models are compact and cheap to evaluate (order of seconds or less) and can therefore be easily used for such tasks. This paper discusses and demonstrates the application of several common surrogate modelling techniques to a CFD model of flocculant adsorption in an industrial thickener. Results from conducting sensitivity analyses on the surrogates are also presented. NOMENCLATURE A radial angle between flocculant sparge and feed pipe j A Fourier coefficients b LS-SVM model parameter l j b neuron bias B vertical location of flocculant sparge j B Fourier coefficients C distance from feedwell wall to flocculant sparge D variance k e acceptable error () E y expectation of y F ANN activation function i G transformation functions K RBF kernel function N number of sample points or number of neurons in ANN p integer number s real number S main effect ST total effect u input to neuron () V y variance of y w LS-SVM model parameter ij w ANN weights x input sample points y true model output y mean true model output y % predicted model output k α LS-SVM model parameters β RBF model parameters ϕ input to output mapping function γ LS-SVM regularization constant ω Frequency Subscripts , , i j k indices Superscript l neuron layer INTRODUCTION }, keywords = {RBF, Surrogate}, pubstate = {published}, tppubtype = {conference} } The computational cost associated with the use of high-fidelity Computational Fluid Dynamics (CFD) models poses a serious impediment to the successful application of formal sensitivity analysis in engineering design. Even though advances in computing hardware and parallel processing have reduced costs by orders of magnitude over the last few decades, the fidelity with which engineers desire to model engineering systems has also increased considerably. Evaluation of such high-fidelity models may take significant computational time for complex geometries. In many engineering design problems, thousands of function evaluations may be required to undertake a sensitivity analysis. As a result, CFD models are often impractical to use for design sensitivity analyses. In contrast, surrogate models are compact and cheap to evaluate (order of seconds or less) and can therefore be easily used for such tasks. This paper discusses and demonstrates the application of several common surrogate modelling techniques to a CFD model of flocculant adsorption in an industrial thickener. Results from conducting sensitivity analyses on the surrogates are also presented. NOMENCLATURE A radial angle between flocculant sparge and feed pipe j A Fourier coefficients b LS-SVM model parameter l j b neuron bias B vertical location of flocculant sparge j B Fourier coefficients C distance from feedwell wall to flocculant sparge D variance k e acceptable error () E y expectation of y F ANN activation function i G transformation functions K RBF kernel function N number of sample points or number of neurons in ANN p integer number s real number S main effect ST total effect u input to neuron () V y variance of y w LS-SVM model parameter ij w ANN weights x input sample points y true model output y mean true model output y % predicted model output k α LS-SVM model parameters β RBF model parameters ϕ input to output mapping function γ LS-SVM regularization constant ω Frequency Subscripts , , i j k indices Superscript l neuron layer INTRODUCTION |

Jemcov, Aleksandar; Maruszewski, Joseph P Arithmetic Parameterization of General Purpose CFD Code Inproceedings CFD2009 CFD Society of Canada, CFD Society of Canada 2009. Abstract | Links | BibTeX | Tags: Arithmetic, C++, Differentiation @inproceedings{jemcov2009arithmetic, title = {Arithmetic Parameterization of General Purpose CFD Code}, author = { Aleksandar Jemcov and Joseph P Maruszewski}, url = {https://www.researchgate.net/profile/Aleksandar_Jemcov/publication/265160705_Arithmetic_Parameterization_of_General_Purpose_CFD_Code/links/54009abb0cf23d9765a3f563.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=%2Fqll%2FW4qkUHLdZjM%2BAZOylBf77fhfGpGUct%2Fb4KsWmba1UvawviumCqfgdLyjNMbgAho07zn%2BOc8nCcaENOr2A%3D%3D_B7A3OzkBPyCGlbXYafHvX2iP6ZlQXN40Ip4KfXZWJg%2FGO05QNleU9RyUKy%2F4WOhUX%2F77fmLhDx8RgH3yQbLtRQ%3D%3D}, year = {2009}, date = {2009-01-01}, booktitle = {CFD2009 CFD Society of Canada}, organization = {CFD Society of Canada}, abstract = {In this paper we describe the use of C++ templates to parameterize CFD code to produce specialized versions of the code at the compile time that is capable of defining different arithmetic types. This parameterization is primarily used to compute automatic derivatives in design optimization. However, it is possible to use the same code parameterization to define various arithmetics that can be used to produce additional useful side effects such as automatic stability characterization of the algorithms, uncertainty propagation, higher order derivative computations and interval arithmetics. Here we describe the implementation of the parameterization and the design of these classes and we show results of the use of different arithmetics in CFD and related fields.}, keywords = {Arithmetic, C++, Differentiation}, pubstate = {published}, tppubtype = {inproceedings} } In this paper we describe the use of C++ templates to parameterize CFD code to produce specialized versions of the code at the compile time that is capable of defining different arithmetic types. This parameterization is primarily used to compute automatic derivatives in design optimization. However, it is possible to use the same code parameterization to define various arithmetics that can be used to produce additional useful side effects such as automatic stability characterization of the algorithms, uncertainty propagation, higher order derivative computations and interval arithmetics. Here we describe the implementation of the parameterization and the design of these classes and we show results of the use of different arithmetics in CFD and related fields. |

Jemcov, Aleksandar; Maruszewski, Joseph P Multi-Step Saddle Point preconditioner for Incompressible Flow Problems Inproceedings CFD2009 CFD Society of Canada, CFD Society of Canada 2009. Abstract | Links | BibTeX | Tags: FEM, Incompressible, Linear Solver, Preconditioner, Saddle Point @inproceedings{jemcov2009multi, title = {Multi-Step Saddle Point preconditioner for Incompressible Flow Problems}, author = { Aleksandar Jemcov and Joseph P Maruszewski}, url = {https://www.researchgate.net/profile/Aleksandar_Jemcov/publication/265161023_Multi-Step_Saddle_Point_preconditioner_for_Incompressible_Flow_Problems/links/54009cfa0cf2bba34c1a4de4.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=8LbO91MAZ0gkiAx25dfpwCY7AhXJyhwIwFJNWx51hw9rbFYGsY0xW5tKANo0MGSZ3LKWdTL55v7TNqhJIHan1Q%3D%3D_%2BeVdOPh3Lmd4wyIRf7iDGiWVJSDwqpYtmpS3cJvFeWBljBhCOU7OpckbECda26qjZCvWGn7LgS6bb7tV%2Fm%2B%2BcA%3D%3D}, year = {2009}, date = {2009-01-01}, booktitle = {CFD2009 CFD Society of Canada}, organization = {CFD Society of Canada}, abstract = {A new multi-step block preconditioner for linear system of equations arising in finite element discretization of incompressible flow problems is described. The preconditioning method is based on block LU factorization of the matrix used in the iterative algorithm. The performance of the method is dependent on the choice of the approximate inverse used in construction of the Schur complement block and optimal choice of approximate inverse is discussed in this paper. The proposed method has good convergence properties and experimental results demonstrate that computational time scales almost linearly with the number of finite elements.}, keywords = {FEM, Incompressible, Linear Solver, Preconditioner, Saddle Point}, pubstate = {published}, tppubtype = {inproceedings} } A new multi-step block preconditioner for linear system of equations arising in finite element discretization of incompressible flow problems is described. The preconditioning method is based on block LU factorization of the matrix used in the iterative algorithm. The performance of the method is dependent on the choice of the approximate inverse used in construction of the Schur complement block and optimal choice of approximate inverse is discussed in this paper. The proposed method has good convergence properties and experimental results demonstrate that computational time scales almost linearly with the number of finite elements. |

Jemcov, Aleksandar; Maruszewski, Joseph P Shape Optimization Based on Downhill Simplex Optimizer and Free-Form Deformation in General Purpose CFD Code Inproceedings CFD2009 CFD Society of Canada, CFD Society of Canada 2009. Abstract | Links | BibTeX | Tags: FFD, Optimisation, Simplex @inproceedings{jemcov2009shape, title = {Shape Optimization Based on Downhill Simplex Optimizer and Free-Form Deformation in General Purpose CFD Code}, author = { Aleksandar Jemcov and Joseph P Maruszewski}, url = {https://www.researchgate.net/profile/Aleksandar_Jemcov/publication/265160789_Shape_Optimization_Based_on_Downhill_Simplex_Optimizer_and_Free-Form_Deformation_in_General_Purpose_CFD_Code/links/54009a100cf2bba34c1a4b54.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=FqjhrjBLkaNu2r4DeLufdDmsbnNkhAm%2BeTczveowvwba5s6mKSvhnLFB7iJdaS9ufA1B2Kwyl7f4Lq2ha6zPEQ%3D%3D_dpjOArJDei0szbKcEo7CrbsxXPt8E543IHgKcro8QvVtIxLhQ48PbqvpuHp2SjVfjxBbd7L%2BmYKHCHt%2F30LY7A%3D%3D}, year = {2009}, date = {2009-01-01}, booktitle = {CFD2009 CFD Society of Canada}, organization = {CFD Society of Canada}, abstract = {In this work the use of Free-Form Deformation tool together with Nelder-Mead Simplex algorithm applied to shape optimization problems in a general purpose CFD code is described. The main challenge is to find the effective representation of the shape that results in a small number parameters that are used by the optimizer. Free-Form Deformation (FFD) approach to solving the shape parameterization problem is described and it is shown that this representation results in a small number of active parameters that are required for the optimization procedure. Another aspect of this work involves the usage of the Nelder & Mead Simplex optimizer for finding optimal values of shape parameters. It was demonstrated that the combination of Free-Form Deformation and Nelder & Mead Simplex optimization algorithm is a very effective and versatile tool for industrial shape optimization problems.}, keywords = {FFD, Optimisation, Simplex}, pubstate = {published}, tppubtype = {inproceedings} } In this work the use of Free-Form Deformation tool together with Nelder-Mead Simplex algorithm applied to shape optimization problems in a general purpose CFD code is described. The main challenge is to find the effective representation of the shape that results in a small number parameters that are used by the optimizer. Free-Form Deformation (FFD) approach to solving the shape parameterization problem is described and it is shown that this representation results in a small number of active parameters that are required for the optimization procedure. Another aspect of this work involves the usage of the Nelder & Mead Simplex optimizer for finding optimal values of shape parameters. It was demonstrated that the combination of Free-Form Deformation and Nelder & Mead Simplex optimization algorithm is a very effective and versatile tool for industrial shape optimization problems. |

## 2008 |

Jemcov, Aleksandar; Maruszewski, Joseph P Nonlinear Flow Solver Acceleration by Reduced Rank Extrapolation Incollection 46th AIAA Aerospace Sciences Meeting and Exhibit, pp. 609, 2008. Abstract | Links | BibTeX | Tags: Acceleration, Algorithm, Extrapolation, GMRES, Nonlinear, RRE, SIMPLE, Solver @incollection{jemcov2008nonlinear, title = {Nonlinear Flow Solver Acceleration by Reduced Rank Extrapolation}, author = { Aleksandar Jemcov and Joseph P Maruszewski}, url = {https://www.researchgate.net/profile/Aleksandar_Jemcov/publication/275275199_Nonlinear_Flow_Solver_Acceleration_by_Reduced_Rank_Extrapolation/links/5536ca190cf2058efdea9401.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=mHcz4Kt3tcM0fKKvzoZAfJu5CZ7bJAZTebm7svkdI8HyTV2DqiloQBxWK3ZC9HHtc1VaKywIo53Gq4ACRkkbmQ%3D%3D_mZOVHfxww2SuKoJv0YEv1yJ19XXhbwqiKJ%2FuZ%2FsAeTe%2BZ2RdfWruA0KDkoFLyPO0ebx%2BKC9jrGwRj%2B%2F%2BKjsf7A%3D%3D}, year = {2008}, date = {2008-01-01}, booktitle = {46th AIAA Aerospace Sciences Meeting and Exhibit}, pages = {609}, abstract = {Convergence acceleration of nonlinear flow solvers through use of vector sequence ex-trapolation techniques is presented. In particular, suitability of the Reduced Rank Extrap-olation (RRE) algorithm for the use of convergence acceleration of nonlinear flow solvers is examined. In the RRE algorithm, the solution is obtained through a linear combination of Krylov vectors with weighting coefficients obtained by minimizing L2 norm of error in this space with properly chosen constraint conditions. This process effectively defines vector sequence extrapolation process in Krylov subspace that corresponds to the GMRES method applied to nonlinear problems. Moreover, when the RRE algorithm is used to solve nonlinear problems, the flow solver plays the role of the preconditioner for the non-linear GMRES method. Benefits of the application of the RRE algorithm include better convergence rates, removal of residual stalling and improved coupling between equations in numerical models. Proposed algorithm is independent of the type of flow solver and it is equally applicable to explicit, implicit, pressure and density based algorithms. Nomenclature Q Vector of conserved variables R Residual vector F c Vector of convective fluxes F v Vector of viscous fluxes S Source vector H Total enthalpy, J/m 3 p pressure, Pa τ ij viscous tensor ρ Density, Kg/m 3 u Velocity vector, m/s u X-component of velocity vector, m/s v Y-component of velocity vector, m/s w Z-component of velocity vector, m/s V Contravariant velocity, m/s f e,i Vector of external forces Φ Generic transport variable M Nonlinear preconditioning operator F Fixed-point function ∂ Q (·) Jacobian with respect to Q α ν Extrapolation coefficients c ν o t Time, s ∂Ω Boundary of computational domain Ω Computational domain}, keywords = {Acceleration, Algorithm, Extrapolation, GMRES, Nonlinear, RRE, SIMPLE, Solver}, pubstate = {published}, tppubtype = {incollection} } Convergence acceleration of nonlinear flow solvers through use of vector sequence ex-trapolation techniques is presented. In particular, suitability of the Reduced Rank Extrap-olation (RRE) algorithm for the use of convergence acceleration of nonlinear flow solvers is examined. In the RRE algorithm, the solution is obtained through a linear combination of Krylov vectors with weighting coefficients obtained by minimizing L2 norm of error in this space with properly chosen constraint conditions. This process effectively defines vector sequence extrapolation process in Krylov subspace that corresponds to the GMRES method applied to nonlinear problems. Moreover, when the RRE algorithm is used to solve nonlinear problems, the flow solver plays the role of the preconditioner for the non-linear GMRES method. Benefits of the application of the RRE algorithm include better convergence rates, removal of residual stalling and improved coupling between equations in numerical models. Proposed algorithm is independent of the type of flow solver and it is equally applicable to explicit, implicit, pressure and density based algorithms. Nomenclature Q Vector of conserved variables R Residual vector F c Vector of convective fluxes F v Vector of viscous fluxes S Source vector H Total enthalpy, J/m 3 p pressure, Pa τ ij viscous tensor ρ Density, Kg/m 3 u Velocity vector, m/s u X-component of velocity vector, m/s v Y-component of velocity vector, m/s w Z-component of velocity vector, m/s V Contravariant velocity, m/s f e,i Vector of external forces Φ Generic transport variable M Nonlinear preconditioning operator F Fixed-point function ∂ Q (·) Jacobian with respect to Q α ν Extrapolation coefficients c ν o t Time, s ∂Ω Boundary of computational domain Ω Computational domain |

## 2007 |

Morgans, Rick C; Doolan, Con J; Stephens, Darrin W Derivative free global optimisation of CFD simulations Conference 16th Australasian Fluid Mechanics Conference, 2007. Abstract | BibTeX | Tags: Algorithm, EGO, OpenFOAM, Simulation @conference{morgans2007derivative, title = {Derivative free global optimisation of CFD simulations}, author = {Rick C Morgans and Con J Doolan and Darrin W Stephens}, year = {2007}, date = {2007-01-01}, booktitle = {16th Australasian Fluid Mechanics Conference}, abstract = {This work reports on the use of numerical optimisation techniques to optimise objective functions calculated by Computational Fluid Dynamics (CFD) simulations. Two example applications are described, the first being the shape optimisation of a low speed wind tunnel contraction. A potential flow and viscous flow solver have been coupled to produce a robust computational tool, with the contraction shape defined by a two parameter B´ezier curve. The second application is a simplified test case with a known minimum calculated using a commercial CFD code. For the optimisation of complex CFD simulations, it is sometimes advantageous to use an efficient derivative free global optimisation algorithm because of potentially long simulation times, the objective function may contain multiple local minima and it is often difficult to evaluate analytical or numerical gradients. The Efficient Global Optimisation (EGO) algorithm sequentially samples results from an expensive calculation, does not require derivative information, uses an inexpensive surrogate to search for a global optimum, and is used in this current work. For both applications, the EGO algorithm is able to efficiently and robustly find a global optimum that satisfies any constraints.}, keywords = {Algorithm, EGO, OpenFOAM, Simulation}, pubstate = {published}, tppubtype = {conference} } This work reports on the use of numerical optimisation techniques to optimise objective functions calculated by Computational Fluid Dynamics (CFD) simulations. Two example applications are described, the first being the shape optimisation of a low speed wind tunnel contraction. A potential flow and viscous flow solver have been coupled to produce a robust computational tool, with the contraction shape defined by a two parameter B´ezier curve. The second application is a simplified test case with a known minimum calculated using a commercial CFD code. For the optimisation of complex CFD simulations, it is sometimes advantageous to use an efficient derivative free global optimisation algorithm because of potentially long simulation times, the objective function may contain multiple local minima and it is often difficult to evaluate analytical or numerical gradients. The Efficient Global Optimisation (EGO) algorithm sequentially samples results from an expensive calculation, does not require derivative information, uses an inexpensive surrogate to search for a global optimum, and is used in this current work. For both applications, the EGO algorithm is able to efficiently and robustly find a global optimum that satisfies any constraints. |

Jasak, Hrvoje; Jemcov, Aleksandar; Tukovic, Zeljko OpenFOAM: A C++ library for complex physics simulations Conference International workshop on coupled methods in numerical dynamics, 2007. Abstract | Links | BibTeX | Tags: C++, CG-AMG, FSI, Iterative Solver, Object Oriented, POD, Scientific Computing @conference{jasak2007openfoam, title = {OpenFOAM: A C++ library for complex physics simulations}, author = { Hrvoje Jasak and Aleksandar Jemcov and Zeljko Tukovic}, url = {https://www.researchgate.net/profile/Zeljko_Tukovic/publication/228879492_OpenFOAM_A_c_library_for_complex_physics_simulations/links/00463528c618e93598000000.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=haZnOpj%2BiY3X9cXMdDDc1LcZEyKlShWo7%2BjsgCLomozNYgASvQXUlVhFDJm%2Fw23IpZVsMVmv%2BIu5Gf7ZOVVYNg%3D%3D_Y2eTjr5zkiEsBT%2FCqcce02aac722jLAiaEI1q2u1%2FC5WzTNFR%2FMfrg%2FyhkxiVVfy4LQqIvhREvlCMzh4LbT6Cw%3D%3D&inViewer=1}, year = {2007}, date = {2007-01-01}, booktitle = {International workshop on coupled methods in numerical dynamics}, pages = {1--20}, abstract = {This paper describes the design of OpenFOAM, an object-oriented library for Computational Fluid Dynamics (CFD) and struc-tural analysis. Efficient and flexible implementation of complex physi-cal models in Continuum Mechanics is achieved by mimicking the form of partial differential equation in software. The library provides Fi-nite Volume and Finite Element discretisation in operator form and with polyhedral mesh support, with relevant auxiliary tools and sup-port for massively parallel computing. Functionality of OpenFOAM is illustrated on three levels: improvements in linear solver technology with CG-AMG solvers, LES data analysis using Proper Orthogonal Decom-position (POD) and a self-contained fluid-structure interaction solver.}, keywords = {C++, CG-AMG, FSI, Iterative Solver, Object Oriented, POD, Scientific Computing}, pubstate = {published}, tppubtype = {conference} } This paper describes the design of OpenFOAM, an object-oriented library for Computational Fluid Dynamics (CFD) and struc-tural analysis. Efficient and flexible implementation of complex physi-cal models in Continuum Mechanics is achieved by mimicking the form of partial differential equation in software. The library provides Fi-nite Volume and Finite Element discretisation in operator form and with polyhedral mesh support, with relevant auxiliary tools and sup-port for massively parallel computing. Functionality of OpenFOAM is illustrated on three levels: improvements in linear solver technology with CG-AMG solvers, LES data analysis using Proper Orthogonal Decom-position (POD) and a self-contained fluid-structure interaction solver. |

Jasak, Hrvoje; Jemcov, Aleksandar; Maruszewski, Joseph P Preconditioned linear solvers for large eddy simulation Conference CFD 2007 Conference, CFD Society of Canada, 2007. Abstract | BibTeX | Tags: Algebraic Multi-grid Solver, Linear Solver, Preconditioner, Simulation, Solver @conference{jasak2007preconditioned, title = {Preconditioned linear solvers for large eddy simulation}, author = { Hrvoje Jasak and Aleksandar Jemcov and Joseph P Maruszewski}, year = {2007}, date = {2007-01-01}, booktitle = {CFD 2007 Conference, CFD Society of Canada}, abstract = {Efficient solution of linear systems of equations stemming from cell centred Finite Volume Discretisation in Large Eddy Simulation is critical in large-scale simulations. This paper presents a class of sparse matrix iterative solvers combining Algebraic Multigrid (AMG) and Krylov Space techniques with the idea of combining residual reduction techniques to improve efficiency over the current solver technology. Emphasis is placed on choosing combinations of a solver, a preconditioner and a smoother and setting control parameters that yield the most efficient solution. Results show consistent superiority of AMG-preconditioned Conjugate Gradient solvers for matrices under consideration}, keywords = {Algebraic Multi-grid Solver, Linear Solver, Preconditioner, Simulation, Solver}, pubstate = {published}, tppubtype = {conference} } Efficient solution of linear systems of equations stemming from cell centred Finite Volume Discretisation in Large Eddy Simulation is critical in large-scale simulations. This paper presents a class of sparse matrix iterative solvers combining Algebraic Multigrid (AMG) and Krylov Space techniques with the idea of combining residual reduction techniques to improve efficiency over the current solver technology. Emphasis is placed on choosing combinations of a solver, a preconditioner and a smoother and setting control parameters that yield the most efficient solution. Results show consistent superiority of AMG-preconditioned Conjugate Gradient solvers for matrices under consideration |

Jemcov, Aleksandar; Maruszewski, Joseph P; Jasak, Hrvoje Acceleration and stabilization of algebraic multigrid solver applied to incompressible flow problems Conference AIAA CFD conference, 2007. Abstract | Links | BibTeX | Tags: AMG, Linear Solver, Matrix, RPM, Solver, Velocity @conference{jemcov2007acceleration, title = {Acceleration and stabilization of algebraic multigrid solver applied to incompressible flow problems}, author = { Aleksandar Jemcov and Joseph P Maruszewski and Hrvoje Jasak}, doi = {10.2514/6.2007-4330}, year = {2007}, date = {2007-01-01}, booktitle = {AIAA CFD conference}, abstract = {Acceleration and stabilization of the Algebraic Multigrid solver (AMG) through n-th order Recursive Projection Method (RPM(n)) is described. It is shown that significant acceleration can be obtained if RPM(n) is applied to AMG during the inner iteration loop in a typical implicit incompressible CFD codes. In addition to accelerating the so- lution, RPM(n) provides increased stability to the AMG Solver extending it beyond its normal range of applicability in terms of matrix conditioning and M-matrix properties. RPM(n) algorithm allows the use of agglomerative AMG solver with simple smoothers to be effectively applied to matrices that are not an M-matrix. Theoretical foundations of RPM(n)-AMG algorithm are presented with some practical aspects of the algorithm im- plementation. Numerical experiments that involve pressure correction matrices of various sizes that appear in segregated pressure based algorithms together with coupled momen- tum and pressure matrices stemming from coupled pressure based algorithms are used to illustrate the effectiveness of the method.}, keywords = {AMG, Linear Solver, Matrix, RPM, Solver, Velocity}, pubstate = {published}, tppubtype = {conference} } Acceleration and stabilization of the Algebraic Multigrid solver (AMG) through n-th order Recursive Projection Method (RPM(n)) is described. It is shown that significant acceleration can be obtained if RPM(n) is applied to AMG during the inner iteration loop in a typical implicit incompressible CFD codes. In addition to accelerating the so- lution, RPM(n) provides increased stability to the AMG Solver extending it beyond its normal range of applicability in terms of matrix conditioning and M-matrix properties. RPM(n) algorithm allows the use of agglomerative AMG solver with simple smoothers to be effectively applied to matrices that are not an M-matrix. Theoretical foundations of RPM(n)-AMG algorithm are presented with some practical aspects of the algorithm im- plementation. Numerical experiments that involve pressure correction matrices of various sizes that appear in segregated pressure based algorithms together with coupled momen- tum and pressure matrices stemming from coupled pressure based algorithms are used to illustrate the effectiveness of the method. |

Jemcov, Aleksandar; Maruszewski, Joseph P; Jasak, Hrvoje Performance improvement of algebraic multigrid solver by vector sequence extrapolation Conference CFD 2007 Conference, CFD Society of Canada, 2007. Abstract | Links | BibTeX | Tags: AMG, Extrapolation, Grid, MPE, Multigrid, PFE, RRE, Solver @conference{jemcov2007performance, title = {Performance improvement of algebraic multigrid solver by vector sequence extrapolation}, author = { Aleksandar Jemcov and Joseph P Maruszewski and Hrvoje Jasak}, url = {https://www.researchgate.net/profile/Aleksandar_Jemcov/publication/255577657_Performance_Improvement_of_Algebraic_Multigrid_Solver_by_Vector_Sequence_Extrapolation/links/0deec530e01e9a1e91000000.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=l%2FRqA1L7b0MsHj%2FU443TuP1z2sy1wGnZ36isKwJidqyfbvxfh1znfzdHGghuxvDZ1UsaTbrdodbLRM48tuw%2FSQ%3D%3D_0d7vGCwXITSmkPjwIoqTdFkhnJyIFX3cZPQVfLL9zoYt2B6bVpSn9PRtuwh7yMdk54t1LUp8kE3YNFzdQE15GA%3D%3D}, year = {2007}, date = {2007-01-01}, booktitle = {CFD 2007 Conference, CFD Society of Canada}, abstract = {Algebraic Multigrid Method (AMG) performance im- provement by vector sequence extrapolation is exam- ined. Projective Forward Extrapolation (PFE), Min- imal Polynomial Extrapolation (MPE) and Reduced Rank Extrapolation (RRE) are applied to the AMG resulting in a hybrid approach, vector extrapolated AMG. The impact of vector sequence extrapolation is shown to improve performance of the AMG in number of cycles and execution time, resulting in three new methods: PFE-AMG, MPE-AMG and RRE-AMG. Computational results of the application of vector ex- trapolated AMG to sparse matrices arising from dis- cretization of fluid flow equations are presented show- ing performance improvements compared to the tradi- tional AMG.}, keywords = {AMG, Extrapolation, Grid, MPE, Multigrid, PFE, RRE, Solver}, pubstate = {published}, tppubtype = {conference} } Algebraic Multigrid Method (AMG) performance im- provement by vector sequence extrapolation is exam- ined. Projective Forward Extrapolation (PFE), Min- imal Polynomial Extrapolation (MPE) and Reduced Rank Extrapolation (RRE) are applied to the AMG resulting in a hybrid approach, vector extrapolated AMG. The impact of vector sequence extrapolation is shown to improve performance of the AMG in number of cycles and execution time, resulting in three new methods: PFE-AMG, MPE-AMG and RRE-AMG. Computational results of the application of vector ex- trapolated AMG to sparse matrices arising from dis- cretization of fluid flow equations are presented show- ing performance improvements compared to the tradi- tional AMG. |

## 2006 |

Rackemann, Darryn W; Broadfoot, Ross; Stephens, Darrin W Improved CFD modelling of natural circulation vacuum pans Inproceedings Australian Society of Sugar Cane Technologists, pp. 462, 2006. Abstract | Links | BibTeX | Tags: Circulation, Heat Transfer, Modelling, Performance, Simulation, Vacuum Pan, Validation @inproceedings{rackemann2006improved, title = {Improved CFD modelling of natural circulation vacuum pans}, author = {Darryn W Rackemann and Ross Broadfoot and Darrin W Stephens}, doi = {10.13140/RG.2.1.1867.5046}, year = {2006}, date = {2006-01-01}, booktitle = {Australian Society of Sugar Cane Technologists}, journal = {PROCEEDINGS-AUSTRALIAN SOCIETY OF SUGAR CANE TECHNOLOGISTS}, volume = {28}, pages = {462}, abstract = {Numerical and especially CFD modelling are becoming cost-effective and reliable ways to develop improvements in vessel designs. Simulating boiling in crystallisation vacuum pans is a very complex process that needs to consider non-isothermal conditions, multi-component, multiphase boiling and condensation. This level of complexity which has been captured in the SRI CFD model was previously too complicated for most CFD software. With improvements to the software, these complex physical processes can now be modelled, albeit a little cumbersomely. This paper details the development of CFD models to predict the circulation patterns and heat transfer occurring in natural circulation crystallisation vacuum pans. Model validation involved checking the circulation velocities predicted by the CFD model with circulation velocity data measured on factory pans. The predictions were in reasonable agreement with factory measurements. The validated CFD model was used to investigate the effect of altering key dimensions on batch pans and on two different continuous pan designs. The batch pan investigations included: • Increasing the volumetric region above the calandria by flaring the pan body; • Reducing the clearance underneath the calandria; and • Changing the dimensions of the tubes (tube diameter and length) while maintaining the same heat transfer area and keeping the evaporation rate constant. The effect of variations in pan geometry, massecuite viscosity and operating level above the calandria were considered for the batch pan simulations. These results provide interesting insight into the complicated processes involved in the operation of natural circulation vacuum pans.}, keywords = {Circulation, Heat Transfer, Modelling, Performance, Simulation, Vacuum Pan, Validation}, pubstate = {published}, tppubtype = {inproceedings} } Numerical and especially CFD modelling are becoming cost-effective and reliable ways to develop improvements in vessel designs. Simulating boiling in crystallisation vacuum pans is a very complex process that needs to consider non-isothermal conditions, multi-component, multiphase boiling and condensation. This level of complexity which has been captured in the SRI CFD model was previously too complicated for most CFD software. With improvements to the software, these complex physical processes can now be modelled, albeit a little cumbersomely. This paper details the development of CFD models to predict the circulation patterns and heat transfer occurring in natural circulation crystallisation vacuum pans. Model validation involved checking the circulation velocities predicted by the CFD model with circulation velocity data measured on factory pans. The predictions were in reasonable agreement with factory measurements. The validated CFD model was used to investigate the effect of altering key dimensions on batch pans and on two different continuous pan designs. The batch pan investigations included: • Increasing the volumetric region above the calandria by flaring the pan body; • Reducing the clearance underneath the calandria; and • Changing the dimensions of the tubes (tube diameter and length) while maintaining the same heat transfer area and keeping the evaporation rate constant. The effect of variations in pan geometry, massecuite viscosity and operating level above the calandria were considered for the batch pan simulations. These results provide interesting insight into the complicated processes involved in the operation of natural circulation vacuum pans. |

Rackemann, Darryn W; Plaza, Floren; Stephens, Darrin W Steam side calandria modelling of vacuum pans and evaporators Conference Australian Society of Sugar Cane Technologists, 28 , 2006. Abstract | Links | BibTeX | Tags: Calandria, Heat Transfer, Modelling, Steam Flow, Vacuum Pan @conference{rackemann2006steam, title = {Steam side calandria modelling of vacuum pans and evaporators}, author = {Darryn W Rackemann and Floren Plaza and Darrin W Stephens}, doi = {10.13140/RG.2.1.2916.0801}, year = {2006}, date = {2006-01-01}, booktitle = {Australian Society of Sugar Cane Technologists}, journal = {PROCEEDINGS-AUSTRALIAN SOCIETY OF SUGAR CANE TECHNOLOGISTS}, volume = {28}, abstract = {The heat transfer and condensation of steam within the steam belt and the steam chest surrounding the calandria tubes of vacuum pans and evaporators were investigated using computational fluid dynamics (CFD) modelling techniques. The flow of steam in evaporators and vacuum pans is an aspect that is not usually given much attention but it can influence the productivity of these vessels. The latent heat of the steam provides the heat to the juice or massecuite which induces the formation of vapour bubbles and drives the circulation of the fluid within the vessel. Strong and uniform circulation of the massecuite in vacuum pans increases the production capacity and improves the quality of the sugar produced. Non-uniform heating by the steam on the outside of the calandria tubes can contribute to uneven and inconsistent heat transfer to the juice or massecuite within the vessel. As a consequence under these circumstances the installed heating surface is not effectively utilised. Inconsistent heating can affect the performance of vacuum pans since it influences the circulation of massecuite and the crystallisation rate of sugar. The CFD modelling investigation into the steam side operation of the calandria of vacuum pans and evaporators was preliminary in nature but has shown promising results. The results of the CFD simulations were compared against measured data to determine the applicability of the CFD model. The condensation physics of the CFD model currently has limitations, yet despite these, the CFD model has identified some deficiencies in the flow of steam within the calandria. The investigations into different geometries for steam flow into evaporators and vacuum pans and the results of CFD simulations are detailed and discussed. Some of the proposed modifications resulted in predicted improvements to the distribution of steam within the calandria. }, keywords = {Calandria, Heat Transfer, Modelling, Steam Flow, Vacuum Pan}, pubstate = {published}, tppubtype = {conference} } The heat transfer and condensation of steam within the steam belt and the steam chest surrounding the calandria tubes of vacuum pans and evaporators were investigated using computational fluid dynamics (CFD) modelling techniques. The flow of steam in evaporators and vacuum pans is an aspect that is not usually given much attention but it can influence the productivity of these vessels. The latent heat of the steam provides the heat to the juice or massecuite which induces the formation of vapour bubbles and drives the circulation of the fluid within the vessel. Strong and uniform circulation of the massecuite in vacuum pans increases the production capacity and improves the quality of the sugar produced. Non-uniform heating by the steam on the outside of the calandria tubes can contribute to uneven and inconsistent heat transfer to the juice or massecuite within the vessel. As a consequence under these circumstances the installed heating surface is not effectively utilised. Inconsistent heating can affect the performance of vacuum pans since it influences the circulation of massecuite and the crystallisation rate of sugar. The CFD modelling investigation into the steam side operation of the calandria of vacuum pans and evaporators was preliminary in nature but has shown promising results. The results of the CFD simulations were compared against measured data to determine the applicability of the CFD model. The condensation physics of the CFD model currently has limitations, yet despite these, the CFD model has identified some deficiencies in the flow of steam within the calandria. The investigations into different geometries for steam flow into evaporators and vacuum pans and the results of CFD simulations are detailed and discussed. Some of the proposed modifications resulted in predicted improvements to the distribution of steam within the calandria. |

Trang, Simon CT; Stephens, Darrin W; Schwarz, Phil Modelling heat transfer in the dripper zone of a heap leaching operation Conference Fifth International Conference on CFD in the Process Industries, 2006. Abstract | Links | BibTeX | Tags: Heat Transfer, Leaching, Modelling, Solver @conference{trang2006modelling, title = {Modelling heat transfer in the dripper zone of a heap leaching operation}, author = {Simon CT Trang and Darrin W Stephens and Phil Schwarz}, doi = {10.13140/RG.2.1.1343.2167}, year = {2006}, date = {2006-01-01}, booktitle = {Fifth International Conference on CFD in the Process Industries}, journal = {a a}, abstract = {A computational fluid dynamics (CFD) solver ANSYS- CFX is used to model the heat transfer in the region near the surface of a leach heap when drippers are buried. The potential for natural convection to occur above the dripper level, thus substantially increasing heat loss from the heap, is investigated. A parameter analysis is performed which shows that the factors that may be important to the initiation of natural convection are permeability, the depth at which the drippers are buried and the space between each dripper. The current study shows that permeability is the only parameter which has a profound effect on heat loss by natural convection. }, keywords = {Heat Transfer, Leaching, Modelling, Solver}, pubstate = {published}, tppubtype = {conference} } A computational fluid dynamics (CFD) solver ANSYS- CFX is used to model the heat transfer in the region near the surface of a leach heap when drippers are buried. The potential for natural convection to occur above the dripper level, thus substantially increasing heat loss from the heap, is investigated. A parameter analysis is performed which shows that the factors that may be important to the initiation of natural convection are permeability, the depth at which the drippers are buried and the space between each dripper. The current study shows that permeability is the only parameter which has a profound effect on heat loss by natural convection. |

Jemcov, Aleksandar; Jojic, Branimir Numerical Modeling of Combustion Instability in Rijke Tube Inproceedings CFD2006 CFD Society of Canada, CFD Society of Canada 2006. Abstract | Links | BibTeX | Tags: Combustion, Modelling, Rijke Tube @inproceedings{jemcov2006numerical, title = {Numerical Modeling of Combustion Instability in Rijke Tube}, author = { Aleksandar Jemcov and Branimir Jojic}, url = {https://www.researchgate.net/profile/Aleksandar_Jemcov/publication/265160842_Numerical_Modeling_of_Combustion_Instability_in_Rijke_Tubes/links/54009ec70cf2bba34c1a4f59.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=qXvYaIM1NoaBwk22r3y7phomPdDKE1ybkMDPT6xQ6n7Gc6pSIC4e1VsVMR1ejNnMKtqgjwyg0q1a1AjIKzLbLg%3D%3D_5yTaQFK%2B%2FOwiX4OawttpSFVDsxGtrkqmBqvsoX6MESNK8PqdP70ULOlxLqfRHb4GZ0okByXELkL%2FpSt%2FcikNWw%3D%3D}, year = {2006}, date = {2006-01-01}, booktitle = {CFD2006 CFD Society of Canada}, organization = {CFD Society of Canada}, abstract = {Numerical modeling of combustion instability in Rijke tubes and the development of a new physical model for the fluctuating chemical reaction source term in the energy equation is the main focus of this paper. Although physical and numerical combustion instability model was developed and tested on Rijke tube, the final goal of this work is to develop the model that can be used for combustion instability simulation in rocket and air breathing jet engines. The approach taken here consists of the decomposition of the governing equations into a mean flow field equation that contains turbulence and chemical reaction terms, and a perturbation equation that contains nonlinear acoustic and interaction terms responsible for coupling between mean and perturbed flow. The resulting set of equations consists of the Navier-Stokes equations that describe mean flow, and nonlinear acoustic equations that describe perturbed flow. The nonlinear acoustic equations contain spatially varying coefficients whose values are determined by the mean flow. In addition, the nonlinear acoustic equations contain a forcing term that is a perturbation of the chemical reactions source term from the energy equation. This term together with spatially varying coefficients represents coupling between the mean and nonlinear acoustic fields. A new form of the forcing term that models coupling of fluctuations in the pressure field with fluctuations in energy release due to chemical reactions is proposed here. The basis of this model is in the evaporation rate controlled combustion mechanism that was singled out by time scale analysis and validated by numerical experiments of combustion instability in a Rijke tube with n-decane fuel. Results of the numerical simulations show the crucial role of the nonlinear nature of the fluctuating forcing term in the appearance of instabilities.}, keywords = {Combustion, Modelling, Rijke Tube}, pubstate = {published}, tppubtype = {inproceedings} } Numerical modeling of combustion instability in Rijke tubes and the development of a new physical model for the fluctuating chemical reaction source term in the energy equation is the main focus of this paper. Although physical and numerical combustion instability model was developed and tested on Rijke tube, the final goal of this work is to develop the model that can be used for combustion instability simulation in rocket and air breathing jet engines. The approach taken here consists of the decomposition of the governing equations into a mean flow field equation that contains turbulence and chemical reaction terms, and a perturbation equation that contains nonlinear acoustic and interaction terms responsible for coupling between mean and perturbed flow. The resulting set of equations consists of the Navier-Stokes equations that describe mean flow, and nonlinear acoustic equations that describe perturbed flow. The nonlinear acoustic equations contain spatially varying coefficients whose values are determined by the mean flow. In addition, the nonlinear acoustic equations contain a forcing term that is a perturbation of the chemical reactions source term from the energy equation. This term together with spatially varying coefficients represents coupling between the mean and nonlinear acoustic fields. A new form of the forcing term that models coupling of fluctuations in the pressure field with fluctuations in energy release due to chemical reactions is proposed here. The basis of this model is in the evaporation rate controlled combustion mechanism that was singled out by time scale analysis and validated by numerical experiments of combustion instability in a Rijke tube with n-decane fuel. Results of the numerical simulations show the crucial role of the nonlinear nature of the fluctuating forcing term in the appearance of instabilities. |

## 2005 |

Jemcov, ALeksandar; Mathur, Sanjay Nonlinear Parameter Estimation in Inviscid Compressible Flows in Presence of Uncertainties Conference CFD2005 Conference CFD Society of Canada, CFD Society of Canada 2005. Abstract | Links | BibTeX | Tags: AEM, Algorithm, C++, Compressible, Flow, SEM, Velocity @conference{jemcov2005nonlinear, title = {Nonlinear Parameter Estimation in Inviscid Compressible Flows in Presence of Uncertainties}, author = { ALeksandar Jemcov and Sanjay Mathur}, url = {https://www.researchgate.net/profile/Aleksandar_Jemcov/publication/265161046_NONLINEAR_PARAMETER_ESTIMATION_IN_INVISCID_COMPRESSIBLE_FLOWS_IN_PRESENCE_OF_UNCERTAINTIES/links/54009fa20cf2c48563ae5881.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=Eb6hxD11VAw8qlL87p7OgqEvootFE1qcP1%2B29eaQTXDZQJsA8p7cxhtWYujmiizIS6g9Ghe%2Fs%2FiKlYpqUkiuGg%3D%3D_xyhmRLK3POGJnefHrNX7GIlJ13cJFypuToDg6Wh3WROQGJDU9efiaEfVb4Ncmlvxom8DUrMXvQfMj5PjopEZKw%3D%3D&inViewer=1}, year = {2005}, date = {2005-01-01}, booktitle = {CFD2005 Conference CFD Society of Canada}, organization = {CFD Society of Canada}, abstract = {The focus of this paper is on the formulation and solution of inverse problems of parameter estimation using algorithmic differentiation. The inverse problem formulated here seeks to determine the input parameters that minimize a least squares functional with respect to certain target data. The formulation allows for uncertainty in the target data by considering the least squares functional in a stochastic basis described by the covariance of the target data. Furthermore, to allow for robust design, the formulation also accounts for uncertainties in the input parameters. This is achieved using the method of propagation of uncertainties using the directional derivatives of the output parameters with respect to unknown parameters. The required derivatives are calculated simultaneously with the solution using generic programming exploiting the template and operator overloading features of the C++ language. The methodology described here is general and applicable to any numerical solution procedure for any set of governing equations but for the purpose of this paper we consider a finite volume solution of the compressible Euler equations. In particular, we illustrate the method for the case of supersonic flow in a duct with a wedge. The parameter to be determined is the inlet Mach number and the target data is the axial component of velocity at the exit of the duct.}, keywords = {AEM, Algorithm, C++, Compressible, Flow, SEM, Velocity}, pubstate = {published}, tppubtype = {conference} } The focus of this paper is on the formulation and solution of inverse problems of parameter estimation using algorithmic differentiation. The inverse problem formulated here seeks to determine the input parameters that minimize a least squares functional with respect to certain target data. The formulation allows for uncertainty in the target data by considering the least squares functional in a stochastic basis described by the covariance of the target data. Furthermore, to allow for robust design, the formulation also accounts for uncertainties in the input parameters. This is achieved using the method of propagation of uncertainties using the directional derivatives of the output parameters with respect to unknown parameters. The required derivatives are calculated simultaneously with the solution using generic programming exploiting the template and operator overloading features of the C++ language. The methodology described here is general and applicable to any numerical solution procedure for any set of governing equations but for the purpose of this paper we consider a finite volume solution of the compressible Euler equations. In particular, we illustrate the method for the case of supersonic flow in a duct with a wedge. The parameter to be determined is the inlet Mach number and the target data is the axial component of velocity at the exit of the duct. |

Jemcov, Aleksandar; Mathur, Sanjay Sensitivity Analysis and Uncertainty Propagation in Compressible Inviscid Flows Journal Article 2005. Abstract | Links | BibTeX | Tags: Algorithmic Differentiation, Sensitivity, Uncertainty @article{jemcov2005sensitivity, title = {Sensitivity Analysis and Uncertainty Propagation in Compressible Inviscid Flows}, author = { Aleksandar Jemcov and Sanjay Mathur}, url = {https://www.researchgate.net/profile/Aleksandar_Jemcov/publication/265160947_Sensitivity_Analysis_and_Uncertainty_Propagation_in_Compressible_Inviscid_Flows/links/5400a0380cf2bba34c1a4fdd.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=vZLKH47vEcTWJxRZQIzhD%2FFCAYWbPF4WVSUeVQsEVY1%2FFtLcB6GkuTAgaT1GYh1iZ9yQRCyIUDm5EPG%2FiQ7M4Q%3D%3D_5Bmx1BL6eVwwoja7TFmp0QxBWSuqYKDwsLcn%2BRrRrT96ibINXSI4zhpffZndjLkeCf1WWsQn2HFRPRNFW4cBqg%3D%3D}, year = {2005}, date = {2005-01-01}, booktitle = {6th World Congress of Structural and Multidisciplinary Optimization, Rio de Janeiro, Brazil}, abstract = {Techniques for sensitivity analysis and uncertainty propagation in complex inviscid fluid flows are developed in this work. Algorithmic differentiation is used to obtain the directional derivatives in both sensitivity analysis and uncertainty propagation. This approach allows for accurate calculation of the directional derivatives and when used in the context of object oriented and generic programming, it results in efficient and non-intrusive implementation. We outline the algorithmic differentiation implementation and present results of sensitivity analysis and uncertainty propagation calculations with respect to flow parameters and shape.}, keywords = {Algorithmic Differentiation, Sensitivity, Uncertainty}, pubstate = {published}, tppubtype = {article} } Techniques for sensitivity analysis and uncertainty propagation in complex inviscid fluid flows are developed in this work. Algorithmic differentiation is used to obtain the directional derivatives in both sensitivity analysis and uncertainty propagation. This approach allows for accurate calculation of the directional derivatives and when used in the context of object oriented and generic programming, it results in efficient and non-intrusive implementation. We outline the algorithmic differentiation implementation and present results of sensitivity analysis and uncertainty propagation calculations with respect to flow parameters and shape. |

## 2004 |

Jemcov, Aleksandar; Mathur, Sanjay R Algorithmic Differentiation of General Purpose CFD Code: Implementation and Verification Conference ECCOMAS 2004, 2004. Abstract | Links | BibTeX | Tags: Algorithmic Differentiation @conference{jemcov2004algorithmic, title = {Algorithmic Differentiation of General Purpose CFD Code: Implementation and Verification}, author = { Aleksandar Jemcov and Sanjay R Mathur}, url = {https://www.researchgate.net/profile/Aleksandar_Jemcov/publication/242444170_ALGORITHMIC_DIFFERENTIATION_OF_GENERAL_PURPOSE_CFD_CODE_IMPLEMENTATION_AND_VERIFICATION/links/53ff342d0cf283c3583c8549.pdf?origin=publication_detail_rebranded&ev=pub_int_prw_xdl&msrp=KkE86%2FPDiEEYsRVbZyQHRTSPK3pCNPh7rn0mfRMk1pcnWbFGMAv%2Fm0znKf1veMgtTy4rjYmutKFIgUTkuiweVw%3D%3D_HEE41aP2HaVpH4LLPiBkOYWJV74h4bGPeclJQO%2FumnBAKHk76IUoFa%2BojwKrD0GO2xwgtUE7UApGJ4K2vBu%2BlQ%3D%3D}, year = {2004}, date = {2004-01-01}, booktitle = {ECCOMAS 2004}, abstract = {This paper describes the implementation of algorithmic difierentiation in a general purpose CFD code for calculation of forward sensitivities with respect to input pa- rameters. In addition, new techniques for tangent code veriflcation that use special prop- erties of non-dimensional forms of inviscid compressible equations and thermodynamic relationships are presented. Results of calculation of the forward sensitivities with respect to boundary conditions for the compressible inviscid ∞ow over airfoil are presented as an illustration of the general applicability of the newly developed framework.}, keywords = {Algorithmic Differentiation}, pubstate = {published}, tppubtype = {conference} } This paper describes the implementation of algorithmic difierentiation in a general purpose CFD code for calculation of forward sensitivities with respect to input pa- rameters. In addition, new techniques for tangent code veriflcation that use special prop- erties of non-dimensional forms of inviscid compressible equations and thermodynamic relationships are presented. Results of calculation of the forward sensitivities with respect to boundary conditions for the compressible inviscid ∞ow over airfoil are presented as an illustration of the general applicability of the newly developed framework. |

## 2003 |

Moller, David; Wagner, Steve; Broadfoot, Ross; Stephens, Darrin W Modification to the evaporator station in preparation for a cogeneration factory at Broadwater Mill Inproceedings Australia Societ of Sugar Cane Technolgists, 2003. Abstract | Links | BibTeX | Tags: Cogeneration, Evaporator, Expansion, Steam Efficiency, Vapour Bleeding. @inproceedings{moller2003modification, title = {Modification to the evaporator station in preparation for a cogeneration factory at Broadwater Mill}, author = {David Moller and Steve Wagner and Ross Broadfoot and Darrin W Stephens}, doi = {10.13140/RG.2.1.2260.7208}, year = {2003}, date = {2003-01-01}, booktitle = {Australia Societ of Sugar Cane Technolgists}, journal = {PROCEEDINGS-AUSTRALIAN SOCIETY OF SUGAR CANE TECHNOLOGISTS}, abstract = {Broadwater Sugar Mill commissioned a 4000 m 2 SRI-designed Roberts evaporator for the 2002 crush. The evaporator is the crucial first stage of a cogeneration plant, which will comprise a high-pressure boiler and turbo-alternator producing renewable electrical energy for sale. To achieve this, steam consumption by all process plant will need to be minimised and condensate quality for boiler feed water will need to be guaranteed. This paper describes the installation, control system and operating performance of the new SRI Roberts evaporator and the associated arrangement of vapour bleeding to pans and heaters used to optimise the steam efficiency in the 2002 season.}, keywords = {Cogeneration, Evaporator, Expansion, Steam Efficiency, Vapour Bleeding.}, pubstate = {published}, tppubtype = {inproceedings} } Broadwater Sugar Mill commissioned a 4000 m 2 SRI-designed Roberts evaporator for the 2002 crush. The evaporator is the crucial first stage of a cogeneration plant, which will comprise a high-pressure boiler and turbo-alternator producing renewable electrical energy for sale. To achieve this, steam consumption by all process plant will need to be minimised and condensate quality for boiler feed water will need to be guaranteed. This paper describes the installation, control system and operating performance of the new SRI Roberts evaporator and the associated arrangement of vapour bleeding to pans and heaters used to optimise the steam efficiency in the 2002 season. |

## 2002 |

Rackemann, Darryn W; Stephens, Darrin W Determining circulation velocities in vacuum pans Inproceedings Australian Society Sugar Cane Technologists, 2002. Abstract | Links | BibTeX | Tags: Circulation, Hot Film Anemometer, Simulation, Vacuum Pan @inproceedings{rackemann2002determining, title = {Determining circulation velocities in vacuum pans}, author = {Darryn W Rackemann and Darrin W Stephens}, doi = {10.13140/RG.2.1.2785.0087}, year = {2002}, date = {2002-01-01}, booktitle = {Australian Society Sugar Cane Technologists}, journal = {PROCEEDINGS-AUSTRALIAN SOCIETY OF SUGAR CANE TECHNOLOGISTS}, volume = {24}, abstract = {A method for using 'hot film' anemometers to determine circulation velocities within vacuum pans has been refined and tested on a number of pans during trials performed over the 2001 season. The anemometers have been used to characterise circulation velocity profiles that can be used to validate computer models and illustrate the differences among pan geometries. The method has also been used to show the effects of adding pan boiling aids and to examine the effect of changing operating parameters. The method has been shown to be effective in producing reproducible velocity profiles within vacuum pans. The values for circulation velocities obtained are considered reasonable and there are promising indications of the ability to use this technique as a qualitative tool to investigate the causes of poor circulation and the effects of varying operating parameters.}, keywords = {Circulation, Hot Film Anemometer, Simulation, Vacuum Pan}, pubstate = {published}, tppubtype = {inproceedings} } A method for using 'hot film' anemometers to determine circulation velocities within vacuum pans has been refined and tested on a number of pans during trials performed over the 2001 season. The anemometers have been used to characterise circulation velocity profiles that can be used to validate computer models and illustrate the differences among pan geometries. The method has also been used to show the effects of adding pan boiling aids and to examine the effect of changing operating parameters. The method has been shown to be effective in producing reproducible velocity profiles within vacuum pans. The values for circulation velocities obtained are considered reasonable and there are promising indications of the ability to use this technique as a qualitative tool to investigate the causes of poor circulation and the effects of varying operating parameters. |

Stephens, Darrin W; Harris, Jonathan A Prediction of evaporation, pressure driving force, and heat transfer in calandria tubes Conference Australian Society of Sugar Cane Technologists, 24 , 2002. Abstract | Links | BibTeX | Tags: Boiling, Calandria Tube, Evaporation, Heat Transfer, Model, Vacuum Pan @conference{stephens2002prediction, title = {Prediction of evaporation, pressure driving force, and heat transfer in calandria tubes}, author = {Darrin W Stephens and Jonathan A Harris}, doi = {10.13140/RG.2.1.3571.4403}, year = {2002}, date = {2002-01-01}, booktitle = {Australian Society of Sugar Cane Technologists}, journal = {PROCEEDINGS-AUSTRALIAN SOCIETY OF SUGAR CANE TECHNOLOGISTS}, volume = {24}, abstract = {A heat transfer model of boiling flow in steam heated calandria tubes is presented. The model predictions for vapour formation, heat transfer and pressure difference are compared with experimental data and yield reasonable agreement. Characteristic curves are presented showing predicted evaporation rate, heat transfer and pressure difference for a set of parameters representative of a high grade vacuum pan near the start of a strike. }, keywords = {Boiling, Calandria Tube, Evaporation, Heat Transfer, Model, Vacuum Pan}, pubstate = {published}, tppubtype = {conference} } A heat transfer model of boiling flow in steam heated calandria tubes is presented. The model predictions for vapour formation, heat transfer and pressure difference are compared with experimental data and yield reasonable agreement. Characteristic curves are presented showing predicted evaporation rate, heat transfer and pressure difference for a set of parameters representative of a high grade vacuum pan near the start of a strike. |

## 2001 |

Stephens, Darrin W Studies on modelling circulation in sugar vacuum pans PhD Thesis James Cook University, 2001. Abstract | Links | BibTeX | Tags: Axi-Symmetric, Calandria, Circulation, Flow, Heat Transfer, Mass Flow Rate, Model, Modelling, Quasi-Static, Saturation, Vacuum Pan @phdthesis{stephens2001studies, title = {Studies on modelling circulation in sugar vacuum pans}, author = {Darrin W Stephens}, doi = {10.13140/RG.2.1.4325.1048}, year = {2001}, date = {2001-01-01}, school = {James Cook University}, abstract = {This thesis presents an investigation into mathematical modelling of natural circulation in high grade batch vacuum pans. Batch vacuum pans are an important part of a sugar factory, with the circulation in such vessels being a key factor in successful sucrose extraction. The flow within a batch vacuum pan is laminar with three phases (molasses, crystal and vapour) present, and is driven by buoyancy, which results from vapour formation due to boiling. Numerical modelling of natural circulation in batch vacuum pans has been limited in the past by computational power and available computer software, and has suffered from the necessity for very restrictive assumptions to make modelling possible. The thesis uses computational fluid dynamics (CFD) as a tool to develop an improved batch vacuum pan model to investigate the detailed distribution of velocity and temperature within a batch vacuum pan at various stages throughout the strike. A segmented modelling approach has been developed where the vacuum pan is divided into two segments: the space inside the calandria tubes (the calandria tube segment), and the remaining part consisting of a downtake and the space above and below the calandria (the external flow segment). The external flow segment is modelled using the standard CFD approach, whereas the calandria tube segment is represented by a one-dimensional finite volume model. The two segments are coupled together to obtain the overall model of the entire vacuum pan. The calandria tube segment is the key to the vacuum pan model as the majority of the driving force for natural circulation is developed from the vapour formed due to boiling within the calandria tubes. The one-dimensional constant wall temperature tube model developed within this thesis demonstrates, for most parameters, reasonable agreement with previous experimental data. The tube model results have been presented in the form of characteristic curves showing pressure difference, heat transfer and evaporation rate as functions of mass flow rate. These curves provide a new insight into the boiling process within calandria tubes. Improvement of the one-dimensional model predictions would require more experimental data pertaining to the volume fraction distribution in the axial and radial directions, as well as an improved correlation for the boiling heat transfer coefficient. Quasi-static, two-dimensional, axi-symmetric CFD simulations of the vacuum pan were performed for three discrete levels of filling, representing the start, middle and end of the batch process. It was found that the magnitude of the flow speed through the tubes decreases drastically with increasing level within the vacuum pan. This reduction has two causes: first, the effect of increased viscosity with increasing head; and second the effect of the increased boiling point with increasing head. Both of these effects combine to give a much lower heat transfer rate within the tubes, thus producing less vapour to drive the flow. As the head above the calandria increases, the size, strength and existence of recirculation zones also increases. These recirculation zones do not provide any assistance in circulating the flow through the tubes. The simulations provide an improved understanding of the mechanisms producing natural circulation and allow suggestions of possible improvements to vacuum pan designs. With the large change in fluid viscosity from start to finish of the boiling process, combined with the change in saturation profile due to the increasing head, it is difficult to conceive a batch vacuum pan design that will operate at the maximum heat transfer and evaporation point for all times during the strike. The model developed produces a preliminary tool for analysis of vacuum pan operation, and may be applied to both batch and continuous pans. The segmented modelling approach, which is a novel contribution of this work, also provides a framework for future model improvements as new experimental data becomes available.}, keywords = {Axi-Symmetric, Calandria, Circulation, Flow, Heat Transfer, Mass Flow Rate, Model, Modelling, Quasi-Static, Saturation, Vacuum Pan}, pubstate = {published}, tppubtype = {phdthesis} } This thesis presents an investigation into mathematical modelling of natural circulation in high grade batch vacuum pans. Batch vacuum pans are an important part of a sugar factory, with the circulation in such vessels being a key factor in successful sucrose extraction. The flow within a batch vacuum pan is laminar with three phases (molasses, crystal and vapour) present, and is driven by buoyancy, which results from vapour formation due to boiling. Numerical modelling of natural circulation in batch vacuum pans has been limited in the past by computational power and available computer software, and has suffered from the necessity for very restrictive assumptions to make modelling possible. The thesis uses computational fluid dynamics (CFD) as a tool to develop an improved batch vacuum pan model to investigate the detailed distribution of velocity and temperature within a batch vacuum pan at various stages throughout the strike. A segmented modelling approach has been developed where the vacuum pan is divided into two segments: the space inside the calandria tubes (the calandria tube segment), and the remaining part consisting of a downtake and the space above and below the calandria (the external flow segment). The external flow segment is modelled using the standard CFD approach, whereas the calandria tube segment is represented by a one-dimensional finite volume model. The two segments are coupled together to obtain the overall model of the entire vacuum pan. The calandria tube segment is the key to the vacuum pan model as the majority of the driving force for natural circulation is developed from the vapour formed due to boiling within the calandria tubes. The one-dimensional constant wall temperature tube model developed within this thesis demonstrates, for most parameters, reasonable agreement with previous experimental data. The tube model results have been presented in the form of characteristic curves showing pressure difference, heat transfer and evaporation rate as functions of mass flow rate. These curves provide a new insight into the boiling process within calandria tubes. Improvement of the one-dimensional model predictions would require more experimental data pertaining to the volume fraction distribution in the axial and radial directions, as well as an improved correlation for the boiling heat transfer coefficient. Quasi-static, two-dimensional, axi-symmetric CFD simulations of the vacuum pan were performed for three discrete levels of filling, representing the start, middle and end of the batch process. It was found that the magnitude of the flow speed through the tubes decreases drastically with increasing level within the vacuum pan. This reduction has two causes: first, the effect of increased viscosity with increasing head; and second the effect of the increased boiling point with increasing head. Both of these effects combine to give a much lower heat transfer rate within the tubes, thus producing less vapour to drive the flow. As the head above the calandria increases, the size, strength and existence of recirculation zones also increases. These recirculation zones do not provide any assistance in circulating the flow through the tubes. The simulations provide an improved understanding of the mechanisms producing natural circulation and allow suggestions of possible improvements to vacuum pan designs. With the large change in fluid viscosity from start to finish of the boiling process, combined with the change in saturation profile due to the increasing head, it is difficult to conceive a batch vacuum pan design that will operate at the maximum heat transfer and evaporation point for all times during the strike. The model developed produces a preliminary tool for analysis of vacuum pan operation, and may be applied to both batch and continuous pans. The segmented modelling approach, which is a novel contribution of this work, also provides a framework for future model improvements as new experimental data becomes available. |

## 2000 |

Atkinson, Bruce J; Stephens, Darrin W; Harris, Jonathan A; Schneider, Phil A The net pressure driving force due to boiling in calandria tubes. Inproceedings Hogarth, DM (Ed.): Australian Society of Sugar Cane Technologists, 2000. Abstract | Links | BibTeX | Tags: Boiling, Calandria, Evaporation, Flow, Mass Flow Rate, Model, Pressure, Temperature, Thermodynamic, Two-Phase, Vacuum Pan @inproceedings{atkinson2000net, title = {The net pressure driving force due to boiling in calandria tubes.}, author = {Bruce J Atkinson and Darrin W Stephens and Jonathan A Harris and Phil A Schneider}, editor = {DM Hogarth }, doi = {10.13140/RG.2.1.1998.5762}, year = {2000}, date = {2000-01-01}, booktitle = {Australian Society of Sugar Cane Technologists}, journal = {Proceedings of the 2000 Conference of the Australian Society of Sugar Cane Technologists held at Bundaberg, Queensland, Australia, 2 May to 5 May 2000.}, volume = {22}, abstract = {Vapour formation due to boiling in calandria tubes provides the driving force for natural circulation in vacuums pans. The net pressure difference generated across a calandria tube is determined by the average density deficit in the tube relative to the downcomer (i.e. the amount of vapour in the tube) and the pressure loss due to friction and acceleration. This paper presents a mathematical model of the two-phase flow of molasses and vapour in calandria tube assuming equilibrium thermodynamics and steady state conditions. The model can predict the net pressure driving force and the evaporation rate produced by a tube as a function of parameters such as heat input, mass flow rate, liquid height above the calandria and boiling point elevation. Additionally, the model yields detailed profiles of temperature, absolute pressure, volume fraction and other variables as a function of distance along the calandria tube. Results are presented in the form of characteristic curves representing the net pressure difference available to drive natural circulation as a function of applied heat and mass flow rate. The circulation rate and evaporation rate in natural circulation lop may be determined by matching the appropriate characteristic curve to the system response curve for the loop. An example is presented to illustrate the application of the model.}, keywords = {Boiling, Calandria, Evaporation, Flow, Mass Flow Rate, Model, Pressure, Temperature, Thermodynamic, Two-Phase, Vacuum Pan}, pubstate = {published}, tppubtype = {inproceedings} } Vapour formation due to boiling in calandria tubes provides the driving force for natural circulation in vacuums pans. The net pressure difference generated across a calandria tube is determined by the average density deficit in the tube relative to the downcomer (i.e. the amount of vapour in the tube) and the pressure loss due to friction and acceleration. This paper presents a mathematical model of the two-phase flow of molasses and vapour in calandria tube assuming equilibrium thermodynamics and steady state conditions. The model can predict the net pressure driving force and the evaporation rate produced by a tube as a function of parameters such as heat input, mass flow rate, liquid height above the calandria and boiling point elevation. Additionally, the model yields detailed profiles of temperature, absolute pressure, volume fraction and other variables as a function of distance along the calandria tube. Results are presented in the form of characteristic curves representing the net pressure difference available to drive natural circulation as a function of applied heat and mass flow rate. The circulation rate and evaporation rate in natural circulation lop may be determined by matching the appropriate characteristic curve to the system response curve for the loop. An example is presented to illustrate the application of the model. |

McBain, Geordie D; Stephens, Darrin W Low Grash of number convective heat transfer across a spherical cavity Inproceedings Proc. of the 7th Australasian Heat and Mass Transfer Conference, 2000. Abstract | Links | BibTeX | Tags: Cavity, Flow, Grashof, Heat Transfer, Model, Modelling, Temperature @inproceedings{mcbain2000low, title = {Low Grash of number convective heat transfer across a spherical cavity}, author = {Geordie D McBain and Darrin W Stephens}, doi = {10.13140/RG.2.1.4194.0324}, year = {2000}, date = {2000-01-01}, booktitle = {Proc. of the 7th Australasian Heat and Mass Transfer Conference}, abstract = {The increase in heat transfer rate across a spherical fluid-filled cavity embedded in a highly conducting solid with a horizontal temperature gradient is investigated analytically and numerically. Analytically, the low Grashof number asymptotic expansion is extended to second-order for the temperature field. This provides the lowest order correction to the overall Nusselt number. This prediction and the associated flow and temperature fields are compared with specially obtained numerical solutions of the full nonlinear equations. Agreement is excellent for Rayleigh numbers less than a few thousand.}, keywords = {Cavity, Flow, Grashof, Heat Transfer, Model, Modelling, Temperature}, pubstate = {published}, tppubtype = {inproceedings} } The increase in heat transfer rate across a spherical fluid-filled cavity embedded in a highly conducting solid with a horizontal temperature gradient is investigated analytically and numerically. Analytically, the low Grashof number asymptotic expansion is extended to second-order for the temperature field. This provides the lowest order correction to the overall Nusselt number. This prediction and the associated flow and temperature fields are compared with specially obtained numerical solutions of the full nonlinear equations. Agreement is excellent for Rayleigh numbers less than a few thousand. |

Hsu, Kwen; Jemcov, Aleksandar Fluids 2000 Conference and Exhibit, American Institute of Aeronautics and Astronautics, 2000. Links | BibTeX | Tags: Chemical, Detonation, Hyrdogen, Mechanism @conference{hsu2000numerical, title = {Numerical investigation of detonation in premixed hydrogen-air mixture-assessment of simplified chemical mechanisms}, author = { Kwen Hsu and Aleksandar Jemcov}, doi = {doi:10.2514/6.2000-2478}, year = {2000}, date = {2000-01-01}, booktitle = {Fluids 2000 Conference and Exhibit}, journal = {AIAA}, publisher = {American Institute of Aeronautics and Astronautics}, keywords = {Chemical, Detonation, Hyrdogen, Mechanism}, pubstate = {published}, tppubtype = {conference} } |

## 1999 |

Stephens, Darrin W; Harris, Jonathan A Modelling convective boiling of molasses Conference 2nd International Conference on CFD in the Minerals and Processing Industries, CSIRO, Melbourne, Australia, 1999. Abstract | Links | BibTeX | Tags: Boiling, Flow, Heat Transfer, Model, Modelling, ODE, Pressure, Two-Phase @conference{stephens1999modelling, title = {Modelling convective boiling of molasses}, author = {Darrin W Stephens and Jonathan A Harris}, doi = {10.13140/RG.2.1.2359.0242}, year = {1999}, date = {1999-01-01}, booktitle = {2nd International Conference on CFD in the Minerals and Processing Industries, CSIRO, Melbourne, Australia}, abstract = {One-and two-dimensional numerical models of forced convective boiling of molasses in a calandria tube are described. The flow in the tube is considered to be composed of two phases (molasses and steam). The one-dimensional model solves a simplified set of ODEs describing the non-equilibrium boiling process. The two-dimensional model is based on the Eulerian/Eulerian multi-phase approach as implemented in the CFX-4.2 CFD code, and solves for the distribution of volume fraction and the temperature and velocity of each phase, along with global parameters such as pressure drop and evaporation rate. Solutions are presented for a case with similar conditions to those expected in a batch vacuum pan. The results show that the flow in the tube is complex and multi-dimensional. Vapour forms both at the wall (due to direct heating) as well as in the centre (due to bulk boiling). The observed features of the flow from the numerical simulation are qualitatively similar to available experimental observations made by previous investigators, although quantitative agreement has yet to be achieved.}, keywords = {Boiling, Flow, Heat Transfer, Model, Modelling, ODE, Pressure, Two-Phase}, pubstate = {published}, tppubtype = {conference} } One-and two-dimensional numerical models of forced convective boiling of molasses in a calandria tube are described. The flow in the tube is considered to be composed of two phases (molasses and steam). The one-dimensional model solves a simplified set of ODEs describing the non-equilibrium boiling process. The two-dimensional model is based on the Eulerian/Eulerian multi-phase approach as implemented in the CFX-4.2 CFD code, and solves for the distribution of volume fraction and the temperature and velocity of each phase, along with global parameters such as pressure drop and evaporation rate. Solutions are presented for a case with similar conditions to those expected in a batch vacuum pan. The results show that the flow in the tube is complex and multi-dimensional. Vapour forms both at the wall (due to direct heating) as well as in the centre (due to bulk boiling). The observed features of the flow from the numerical simulation are qualitatively similar to available experimental observations made by previous investigators, although quantitative agreement has yet to be achieved. |

## 1998 |

Stephens, Darrin W; Harris, Jonathan A Numerical modelling of two-phase flow a benchmark solution Conference Australasian Fluid Mechanics Conference, 13 , 1998. Abstract | Links | BibTeX | Tags: Eulerian, Flow, Heat Transfer, MATLAB, Model, Modelling, Multiphase, ODE, Two-Phase @conference{stephens1998numerical, title = {Numerical modelling of two-phase flow a benchmark solution}, author = {Darrin W Stephens and Jonathan A Harris}, doi = {10.13140/RG.2.1.2621.1681}, year = {1998}, date = {1998-01-01}, booktitle = {Australasian Fluid Mechanics Conference}, journal = {continuity}, volume = {13}, abstract = {This paper explores a simple test problem that can be used for the validation of Eulerian two-phase flow CFD models. Validation results are presented for a commercial CFD code with multi-phase flow capability, namely CFX-4.2, produced by AEA Technology. The results show that CFX-4.2 exhibits virtually perfect agreement with all 12 test cases over a full range of input parameters using a very moderate mesh.}, keywords = {Eulerian, Flow, Heat Transfer, MATLAB, Model, Modelling, Multiphase, ODE, Two-Phase}, pubstate = {published}, tppubtype = {conference} } This paper explores a simple test problem that can be used for the validation of Eulerian two-phase flow CFD models. Validation results are presented for a commercial CFD code with multi-phase flow capability, namely CFX-4.2, produced by AEA Technology. The results show that CFX-4.2 exhibits virtually perfect agreement with all 12 test cases over a full range of input parameters using a very moderate mesh. |

## 1996 |

Jemcov, Aleksandar; Matovic, Darko Wavelet Preconditioner for Pressure Equation in Domain Decomposition Inproceedings CFD96 Conference, CFD Society of Canada, CFD Society of Canada 1996. Links | BibTeX | Tags: Decomposition, Preconditioner, Pressure, Wavelet @inproceedings{jemcov1996wavelet, title = {Wavelet Preconditioner for Pressure Equation in Domain Decomposition}, author = { Aleksandar Jemcov and Darko Matovic}, url = {https://www.researchgate.net/publication/265160953_Wavelet_Preconditioner_for_Pressure_Equaiton_in_Domain_Decomposition}, year = {1996}, date = {1996-01-01}, booktitle = {CFD96 Conference, CFD Society of Canada}, organization = {CFD Society of Canada}, keywords = {Decomposition, Preconditioner, Pressure, Wavelet}, pubstate = {published}, tppubtype = {inproceedings} } |