LaCàN - Laboratori de Càlcul Numèric
http://hdl.handle.net/2117/2072
2016-09-30T13:36:53ZA Reduced Order Modeling approach for optimal allocation of Distributed Generation in power distribution systems
http://hdl.handle.net/2117/89506
A Reduced Order Modeling approach for optimal allocation of Distributed Generation in power distribution systems
García Blanco, Raquel; Díez, Pedro; Borzacchiello, Domenico; Chinesta, Francisco
This paper presents an “offline-online” strategy for optimal allocation and sizing of Distributed Generation. In traditional optimization approaches, each function evaluation requires the solution of a power flow problem, which makes global optimality a computationally challenging goal. In the proposed strategy the power flow solver is invoked only once and a parametric solution is constructed with a monolithic solver. Despite the fact that the parametrized power flow equations result in a high-dimensional problem, the proposed algorithm is specifically designed to circumvent the curse of dimensionality. This is achieved through the application of Model Reduction, in particular the Proper Generalized Decomposition combined with a nonlinear solver. Numerical examples are carried out for showing the validity of the proposed method.
2016-09-02T14:17:08ZGarcía Blanco, RaquelDíez, PedroBorzacchiello, DomenicoChinesta, FranciscoThis paper presents an “offline-online” strategy for optimal allocation and sizing of Distributed Generation. In traditional optimization approaches, each function evaluation requires the solution of a power flow problem, which makes global optimality a computationally challenging goal. In the proposed strategy the power flow solver is invoked only once and a parametric solution is constructed with a monolithic solver. Despite the fact that the parametrized power flow equations result in a high-dimensional problem, the proposed algorithm is specifically designed to circumvent the curse of dimensionality. This is achieved through the application of Model Reduction, in particular the Proper Generalized Decomposition combined with a nonlinear solver. Numerical examples are carried out for showing the validity of the proposed method.Building performance at low frequency range including flanking transmissions
http://hdl.handle.net/2117/89418
Building performance at low frequency range including flanking transmissions
Guigou-Carter, Catherine; Poblet-Puig, Jordi
Recent researches have proposed alternative formulas for the estimation of the vibration reduction index (Kij) at junctions. Among other aspects, the new approaches propose different expressions for the low, mid and high frequency ranges; the current recommendation in the Annex E of EN 12354 standard provides a single value for the whole frequency range. Vibration reduction index can also be evaluated for specific junctions (dimensions, characteristics) representative of those in a particular building. This work analyses the potential effect that the use of the new vibration reduction index expressions can have in the prediction of global sound insulation outputs at low frequencies for different French traditional concrete based buildings. Measured acoustic performance is compared to predicted performance obtained following the EN 12354 approach.
2016-08-30T11:52:11ZGuigou-Carter, CatherinePoblet-Puig, JordiRecent researches have proposed alternative formulas for the estimation of the vibration reduction index (Kij) at junctions. Among other aspects, the new approaches propose different expressions for the low, mid and high frequency ranges; the current recommendation in the Annex E of EN 12354 standard provides a single value for the whole frequency range. Vibration reduction index can also be evaluated for specific junctions (dimensions, characteristics) representative of those in a particular building. This work analyses the potential effect that the use of the new vibration reduction index expressions can have in the prediction of global sound insulation outputs at low frequencies for different French traditional concrete based buildings. Measured acoustic performance is compared to predicted performance obtained following the EN 12354 approach.Amplified catalogue of vibration reduction index formulas for junctions based on numerical simulations
http://hdl.handle.net/2117/89415
Amplified catalogue of vibration reduction index formulas for junctions based on numerical simulations
Poblet-Puig, Jordi; Guigou-Carter, Catherine
The vibration reduction index (Kij) is a key parameter in the prediction of flanking transmissions according to the EN-12354 standard. Formulas for the evaluation of Kij in L, T and X junctions that depend on the mass ratio are available in the Annex E. Junctions of straight elements with different thickness or thin elastic layers are also included. However, other junction types that are important for building industry are not considered: H-shaped junctions, L or T junctions not forming a right angle, asymmetrical T-junctions , X-junctions where only one of the parts is different (thickness or material) from the other three. In the current research, expressions for these non-covered junctions are provided. They are obtained by means of numerical simulations based on the spectral and/or the standard finite element method. Kij is predicted for a large population of junctions, considering usual thicknesses and material combinations. Afterwards, an statistical treatment of the data is done in order to obtain simple but representative formulas that can easily be used in the daily acoustic projects or included in design software without the need of large computations.
2016-08-30T11:34:55ZPoblet-Puig, JordiGuigou-Carter, CatherineThe vibration reduction index (Kij) is a key parameter in the prediction of flanking transmissions according to the EN-12354 standard. Formulas for the evaluation of Kij in L, T and X junctions that depend on the mass ratio are available in the Annex E. Junctions of straight elements with different thickness or thin elastic layers are also included. However, other junction types that are important for building industry are not considered: H-shaped junctions, L or T junctions not forming a right angle, asymmetrical T-junctions , X-junctions where only one of the parts is different (thickness or material) from the other three. In the current research, expressions for these non-covered junctions are provided. They are obtained by means of numerical simulations based on the spectral and/or the standard finite element method. Kij is predicted for a large population of junctions, considering usual thicknesses and material combinations. Afterwards, an statistical treatment of the data is done in order to obtain simple but representative formulas that can easily be used in the daily acoustic projects or included in design software without the need of large computations.Sub-structuring of mechanical systems based on the path concept
http://hdl.handle.net/2117/89414
Sub-structuring of mechanical systems based on the path concept
Magrans Fontrodona, Francesc Xavier; Poblet-Puig, Jordi; Rodríguez Ferran, Antonio
Real mechanical systems are usually complex. Their modelling with discretisation techniques involve a large number of degrees of freedom/unknowns. This leads to high computational costs especially at high frequencies. An alternative, the statistical methods, are sometimes limited by strong behavioural requirements. So, sub-structuring can be considered very often as a good alternative. The present research studies the coupling between subsystems from the point of view of transmission path analysis. The interest is focused on the signal transmission rather than the energy distribution. A subsystem identification method is proposed. It is based on the expression of the solution in terms of the powers of the transfer matrix. This is related with the description of high-order paths which are more affected by damping and system properties. Consequently, the identification of subsystems can be more easily done. The method provides a quantification of the degree of coupling between subsystems and of the error caused by the detachment of a system part from the global system in the modelling phase. The output of the presented technique can be a sub-system definition valid for the SEA method or relevant information to be used in the design of measurement procedures.
2016-08-30T11:32:15ZMagrans Fontrodona, Francesc XavierPoblet-Puig, JordiRodríguez Ferran, AntonioReal mechanical systems are usually complex. Their modelling with discretisation techniques involve a large number of degrees of freedom/unknowns. This leads to high computational costs especially at high frequencies. An alternative, the statistical methods, are sometimes limited by strong behavioural requirements. So, sub-structuring can be considered very often as a good alternative. The present research studies the coupling between subsystems from the point of view of transmission path analysis. The interest is focused on the signal transmission rather than the energy distribution. A subsystem identification method is proposed. It is based on the expression of the solution in terms of the powers of the transfer matrix. This is related with the description of high-order paths which are more affected by damping and system properties. Consequently, the identification of subsystems can be more easily done. The method provides a quantification of the degree of coupling between subsystems and of the error caused by the detachment of a system part from the global system in the modelling phase. The output of the presented technique can be a sub-system definition valid for the SEA method or relevant information to be used in the design of measurement procedures.The solution of linear mechanical systems in terms of path superposition
http://hdl.handle.net/2117/89392
The solution of linear mechanical systems in terms of path superposition
Magrans Fontrodona, Francesc Xavier; Poblet-Puig, Jordi; Rodríguez Ferran, Antonio
We prove that the solution of any linear mechanical system can be expressed as a linear combination of signal transmission paths. This is done in the framework of the Global Transfer Direct Transfer (GTDT) formulation for vibroacoustic problems. Transmission paths are expressed as powers of the transfer matrix. The key idea of the proof is to generalise the Neumann series of the transfer matrix --which is convergent only if its spectral radius is smaller than one-- into a modified Neumann series that is convergent regardless of the eigenvalues of the transfer matrix. The modification consists in choosing the appropriate combination coefficients for the powers of the transfer matrix in the series. A recursive formula for the computation of these factors is derived. The theoretical results are illustrated by means of numerical examples. Finally, we show that the generalised Neumann series can be understood as an acceleration (i.e. convergence speedup) of the Jacobi iterative method.
2016-08-29T10:40:45ZMagrans Fontrodona, Francesc XavierPoblet-Puig, JordiRodríguez Ferran, AntonioWe prove that the solution of any linear mechanical system can be expressed as a linear combination of signal transmission paths. This is done in the framework of the Global Transfer Direct Transfer (GTDT) formulation for vibroacoustic problems. Transmission paths are expressed as powers of the transfer matrix. The key idea of the proof is to generalise the Neumann series of the transfer matrix --which is convergent only if its spectral radius is smaller than one-- into a modified Neumann series that is convergent regardless of the eigenvalues of the transfer matrix. The modification consists in choosing the appropriate combination coefficients for the powers of the transfer matrix in the series. A recursive formula for the computation of these factors is derived. The theoretical results are illustrated by means of numerical examples. Finally, we show that the generalised Neumann series can be understood as an acceleration (i.e. convergence speedup) of the Jacobi iterative method.Regression curves for vibration transmission across junctions of heavyweight walls and floors based on finite element methods and wave theory
http://hdl.handle.net/2117/89391
Regression curves for vibration transmission across junctions of heavyweight walls and floors based on finite element methods and wave theory
Hopkins, Carl; Crispin, Charlotte; Poblet-Puig, Jordi; Guigou-Carter, Catherine
Sound insulation prediction models in European and International standards use the vibration reduction index to calculate flanking transmission across junctions of walls and floors. These standards contain empirical relationships between the ratio of mass per unit areas for the walls/floors that form the junction and a frequency-independent vibration reduction index. Calculations using wave theory show that there is a stronger relationship between the ratio of characteristic moment impedances and the transmission loss from which the vibration reduction index can subsequently be calculated. In addition, the assumption of frequency-independent vibration reduction indices has been shown to be incorrect due to in-plane wave generation at the junction. Therefore numerical experiments with FEM, SFEM and wave theory have been used to develop new regression curves between these variables for the low-, mid- and high-frequency ranges. The junctions considered were L-, T- and X-junctions formed from heavyweight walls and floors. These new relationships have been implemented in the prediction models and they tend to improve the agreement between the measured and predicted airborne and impact sound insulation.
2016-08-29T10:34:56ZHopkins, CarlCrispin, CharlottePoblet-Puig, JordiGuigou-Carter, CatherineSound insulation prediction models in European and International standards use the vibration reduction index to calculate flanking transmission across junctions of walls and floors. These standards contain empirical relationships between the ratio of mass per unit areas for the walls/floors that form the junction and a frequency-independent vibration reduction index. Calculations using wave theory show that there is a stronger relationship between the ratio of characteristic moment impedances and the transmission loss from which the vibration reduction index can subsequently be calculated. In addition, the assumption of frequency-independent vibration reduction indices has been shown to be incorrect due to in-plane wave generation at the junction. Therefore numerical experiments with FEM, SFEM and wave theory have been used to develop new regression curves between these variables for the low-, mid- and high-frequency ranges. The junctions considered were L-, T- and X-junctions formed from heavyweight walls and floors. These new relationships have been implemented in the prediction models and they tend to improve the agreement between the measured and predicted airborne and impact sound insulation.A modal-spectral model for flanking transmissions
http://hdl.handle.net/2117/89390
A modal-spectral model for flanking transmissions
Poblet-Puig, Jordi
A model for the prediction of direct and indirect (flanking) sound transmissions is presented. It can be applied to geometries with extrusion symmetry. The structures are modelled with spectral finite elements. The acoustic domains are described by means of a modal expansion of the pressure field and must be cuboid-shaped. These reasonable simplifications in the geometry allow the use of more efficient numerical methods. Consequently the coupled vibroacoustic problem in structures such as junctions is efficiently solved.
The vibration reduction index of T-junctions with acoustic excitation and with point force excitation is compared. The differences due to the excitation type obey quite general trends that could be taken into account by prediction formulas. However, they are smaller than other uncertainties not considered in practice. The model is also used to check if the sound transmissions of a fully vibroacoustic problem involving several flanking paths can be reproduced by superposition of independent paths. There exist some differences caused by the interaction between paths, which are more important at low frequencies.
2016-08-29T10:30:50ZPoblet-Puig, JordiA model for the prediction of direct and indirect (flanking) sound transmissions is presented. It can be applied to geometries with extrusion symmetry. The structures are modelled with spectral finite elements. The acoustic domains are described by means of a modal expansion of the pressure field and must be cuboid-shaped. These reasonable simplifications in the geometry allow the use of more efficient numerical methods. Consequently the coupled vibroacoustic problem in structures such as junctions is efficiently solved.
The vibration reduction index of T-junctions with acoustic excitation and with point force excitation is compared. The differences due to the excitation type obey quite general trends that could be taken into account by prediction formulas. However, they are smaller than other uncertainties not considered in practice. The model is also used to check if the sound transmissions of a fully vibroacoustic problem involving several flanking paths can be reproduced by superposition of independent paths. There exist some differences caused by the interaction between paths, which are more important at low frequencies.Tutorial on Hybridizable Discontinuous Galerkin (HDG) for second-order elliptic problems
http://hdl.handle.net/2117/88137
Tutorial on Hybridizable Discontinuous Galerkin (HDG) for second-order elliptic problems
Sevilla Càrdenas, Ruben; Huerta, Antonio
The HDG is a new class of discontinuous Galerkin (DG) methods that shares favorable properties with classical mixed methods such as the well known Raviart-Thomas methods. In particular, HDG provides optimal convergence of both the primal and the dual variables of the mixed formulation. This property enables the construction of superconvergent solutions, contrary to other popular DG methods. In addition, its reduced computational cost, compared to other DG methods, has made HDG an attractive alternative for solving problems governed by partial differential equations. A tutorial on HDG for the numerical solution of second-order elliptic problems is presented. Particular emphasis is placed on providing all the necessary details for the implementation of HDG methods.
2016-06-17T14:10:02ZSevilla Càrdenas, RubenHuerta, AntonioThe HDG is a new class of discontinuous Galerkin (DG) methods that shares favorable properties with classical mixed methods such as the well known Raviart-Thomas methods. In particular, HDG provides optimal convergence of both the primal and the dual variables of the mixed formulation. This property enables the construction of superconvergent solutions, contrary to other popular DG methods. In addition, its reduced computational cost, compared to other DG methods, has made HDG an attractive alternative for solving problems governed by partial differential equations. A tutorial on HDG for the numerical solution of second-order elliptic problems is presented. Particular emphasis is placed on providing all the necessary details for the implementation of HDG methods.Master-slave approach for the modelling of joints with dependent degrees of freedom in flexible mechanisms
http://hdl.handle.net/2117/87479
Master-slave approach for the modelling of joints with dependent degrees of freedom in flexible mechanisms
Muñoz Romero, José; Jelenic, Gordan; Crisfield, M. A.
The analysis of multibody systems requires an exact description of the kinematics of the joints involved. In the present work the master–slave approach is employed and endowed with the possibility of including several more complex types of joints. We present the formulation for joints where some relation between the different released degrees of freedom exists such as the screw joint, the rack-and-pinion joint or the cam joint. These joints are implemented in conjunction with geometrically exact beams and an energy-momentum conserving time-stepping algorithm
2016-05-30T08:41:00ZMuñoz Romero, JoséJelenic, GordanCrisfield, M. A.The analysis of multibody systems requires an exact description of the kinematics of the joints involved. In the present work the master–slave approach is employed and endowed with the possibility of including several more complex types of joints. We present the formulation for joints where some relation between the different released degrees of freedom exists such as the screw joint, the rack-and-pinion joint or the cam joint. These joints are implemented in conjunction with geometrically exact beams and an energy-momentum conserving time-stepping algorithmSpectral analysis of the Beznar dam accelerogram. Comparison with results in the northeastern of Spain
http://hdl.handle.net/2117/87083
Spectral analysis of the Beznar dam accelerogram. Comparison with results in the northeastern of Spain
Canas Torres, José Antonio; Barbat Barbat, Horia Alejandro; Pujades Beneit, Lluís; Egozcue Rubí, Juan José; Sarrate Ramos, Josep
Numerical analysis applied to the Beznar dam accelerograms (longitudinal and transversal components) shows that the Fourier amplitude spectrum is a good approximation to the pseudo-velocity spectra. Thís fact makes possible to use displacement ar velocíty seismograrns to generate acceleration of the ground in places
where acceleration data are not available.
Maximae accelerations determined near and at the Beznar dam are compared wíth the predicted acceleration values far the regían compressed by the Pyrenees Mountains, the lberic System and the Catalonia Coastal Mountains usíng a established theoretical formula
far this region. Theoretical and observad values are consistent among them.
2016-05-17T07:55:41ZCanas Torres, José AntonioBarbat Barbat, Horia AlejandroPujades Beneit, LluísEgozcue Rubí, Juan JoséSarrate Ramos, JosepNumerical analysis applied to the Beznar dam accelerograms (longitudinal and transversal components) shows that the Fourier amplitude spectrum is a good approximation to the pseudo-velocity spectra. Thís fact makes possible to use displacement ar velocíty seismograrns to generate acceleration of the ground in places
where acceleration data are not available.
Maximae accelerations determined near and at the Beznar dam are compared wíth the predicted acceleration values far the regían compressed by the Pyrenees Mountains, the lberic System and the Catalonia Coastal Mountains usíng a established theoretical formula
far this region. Theoretical and observad values are consistent among them.