UPCommons. Portal del coneixement obert de la UPC
http://upcommons.upc.edu
The DSpace digital repository system captures, stores, indexes, preserves, and distributes digital research material.2020-06-06T06:04:16ZPràctiques de Química
http://hdl.handle.net/2117/190152
Pràctiques de Química
Departament d'Enginyeria Química
2020-06-05T22:37:13ZDepartament d'Enginyeria QuímicaEmbedding: a unifying concept for recursive program design
http://hdl.handle.net/2117/190151
Embedding: a unifying concept for recursive program design
Roselló Balanyà, Celestí; Balcázar Navarro, José Luis; Peña Marí, Ricardo
The concept of embedding one function into another is used to get a unified view of some well-known design techniques, such as loop derivation using inveriants, recursion removal, or folding-unfolding. A design method for recursive functions based on embedding is presented and its relation to those techniques stated. Efficiency transformations are also formalized using embedding. A few examples will clarify and complement the proposed technique.
2020-06-05T18:30:21ZRoselló Balanyà, CelestíBalcázar Navarro, José LuisPeña Marí, RicardoThe concept of embedding one function into another is used to get a unified view of some well-known design techniques, such as loop derivation using inveriants, recursion removal, or folding-unfolding. A design method for recursive functions based on embedding is presented and its relation to those techniques stated. Efficiency transformations are also formalized using embedding. A few examples will clarify and complement the proposed technique.Specification and verification of TCSP systems by means of partial abstract types
http://hdl.handle.net/2117/190150
Specification and verification of TCSP systems by means of partial abstract types
Peña Marí, Ricardo; Alonso de Armiño Pérez, Luís
A formal framework and a methodology for the specification, refinement and correctness proving of parallel systems are presented. Processes are objects in the TCSP model and are specified by means of an auxiliary partial abstract data type. Part of the proofs are made in the abstract data type framework and so more powerful deductive methods can be used. Examples of specifications and of proving a refinement correct are included.
2020-06-05T16:56:57ZPeña Marí, RicardoAlonso de Armiño Pérez, LuísA formal framework and a methodology for the specification, refinement and correctness proving of parallel systems are presented. Processes are objects in the TCSP model and are specified by means of an auxiliary partial abstract data type. Part of the proofs are made in the abstract data type framework and so more powerful deductive methods can be used. Examples of specifications and of proving a refinement correct are included.Highly parallel multi-physics simulation of muscular activation and EMG
http://hdl.handle.net/2117/190149
Highly parallel multi-physics simulation of muscular activation and EMG
Maier, B.; Emamy, N.; Krämer, A.; Mehl, M.
Simulation of skeletal muscle activation can help to interpret electromyographic measurements and infer the behavior of the muscle ﬁbers. Existing models consider simpliﬁed geometries or a low number of muscle ﬁbers to reduce the computation time. We demonstrate how to simulate a ﬁnely-resolved model of biceps brachii with a typical number of 270.000 ﬁbers. We have used domain decomposition to run simulations on 27.000 cores of the supercomputer HazelHen at HLRS in Stuttgart, Germany. We present details on opendihu, our software framework. Its conﬁgurability, eﬃcient data structures and modular software architecture target usability, performance and extensibility for future models. We present good parallel weak scaling of the simulations.
2020-06-05T15:42:07ZMaier, B.Emamy, N.Krämer, A.Mehl, M.Simulation of skeletal muscle activation can help to interpret electromyographic measurements and infer the behavior of the muscle ﬁbers. Existing models consider simpliﬁed geometries or a low number of muscle ﬁbers to reduce the computation time. We demonstrate how to simulate a ﬁnely-resolved model of biceps brachii with a typical number of 270.000 ﬁbers. We have used domain decomposition to run simulations on 27.000 cores of the supercomputer HazelHen at HLRS in Stuttgart, Germany. We present details on opendihu, our software framework. Its conﬁgurability, eﬃcient data structures and modular software architecture target usability, performance and extensibility for future models. We present good parallel weak scaling of the simulations.A novel method for magnetohydrodynamic simulations and its first applications in astrophysics and cosmology on high performance computational systems
http://hdl.handle.net/2117/190148
A novel method for magnetohydrodynamic simulations and its first applications in astrophysics and cosmology on high performance computational systems
Chetverushkin, Boris N.; Nikolaeva, Alexandra K.; Saveliev, Andrey V.
Magnetic ﬁelds are one of the most important phenomena in science and engineering, as they are present on almost every scale in nature, ranging from atomic magnetic moments to the intergalactic space, and are used in applications ranging from Magnetic Resonance Imaging to nuclear fusion. In this work we ﬁrst present a novel powerful method for high performance magnetohydrodynamic (MHD) calculations which is based on kinetic schemes. In particular, using it, it is possible to derive the MHD equations directly from the Boltzmann Equation without the necessity of an ad hoc introduction of terms related to electromagnetic interactions. With that at hand, we were then able to apply the method to one of the most important problems in present day astrophysics and cosmology, namely to the question of the origin and time evolution of Intergalactic Magnetic Fields. As for their origin, there are mainly two scenarios discussed in the literature – on the one hand the cosmological one, where the magnetic ﬁeld is produced by some process in the very early Universe, and on the other hand the cosmological one, where a seed of the magnetic ﬁeld is created during structure formation and then ampliﬁed by some dynamo eﬀect. Here, we show ﬁrst results of the aforementioned application of our method – on the one hand, concerning the astrophysical scenario, the simulation of galactic winds, i.e. the ejection of matter from galaxies which might also carry magnetic energy, and on the other hand, for the cosmological scenario, the time evolution of primordial magnetic ﬁelds and their possible imprints on the Cosmic Microwave Background (CMB).
2020-06-05T15:38:25ZChetverushkin, Boris N.Nikolaeva, Alexandra K.Saveliev, Andrey V.Magnetic ﬁelds are one of the most important phenomena in science and engineering, as they are present on almost every scale in nature, ranging from atomic magnetic moments to the intergalactic space, and are used in applications ranging from Magnetic Resonance Imaging to nuclear fusion. In this work we ﬁrst present a novel powerful method for high performance magnetohydrodynamic (MHD) calculations which is based on kinetic schemes. In particular, using it, it is possible to derive the MHD equations directly from the Boltzmann Equation without the necessity of an ad hoc introduction of terms related to electromagnetic interactions. With that at hand, we were then able to apply the method to one of the most important problems in present day astrophysics and cosmology, namely to the question of the origin and time evolution of Intergalactic Magnetic Fields. As for their origin, there are mainly two scenarios discussed in the literature – on the one hand the cosmological one, where the magnetic ﬁeld is produced by some process in the very early Universe, and on the other hand the cosmological one, where a seed of the magnetic ﬁeld is created during structure formation and then ampliﬁed by some dynamo eﬀect. Here, we show ﬁrst results of the aforementioned application of our method – on the one hand, concerning the astrophysical scenario, the simulation of galactic winds, i.e. the ejection of matter from galaxies which might also carry magnetic energy, and on the other hand, for the cosmological scenario, the time evolution of primordial magnetic ﬁelds and their possible imprints on the Cosmic Microwave Background (CMB).Soil-Structure Interaction Simulations Taking into Account the Transient Propagation of Seismic Waves
http://hdl.handle.net/2117/190147
Soil-Structure Interaction Simulations Taking into Account the Transient Propagation of Seismic Waves
Schauer, Marco; Taddei, Francesca; Ríos Rodríguez, Gustavo A.
In this contribution we present a strategy to investigate the vibrations of build ings subjected to a transient seismic excitation, including the soil-structure inte ract ion . The proposed simulation method can be helpful during the design of earthquake-resistant structures in seismic active areas as well as for the design of vibration reduction measures for buildings subjected to surrounding emissions like vibrations induced by traffic or ma chine foundations. The structures and their foundation as well as parts of the soil arc modeled by Finite Element Method (FEM). The far field is idealized as an infinite half space and modeled with the Scaled Boundary Finite Element Method (SBFEM). Both methods arc coupled at the common inter face. This approach fulfills exactly the Sommer feld radiation condition. The seismic excitation is idealized as a plane wave propagating toward the structure with an arbitrary angle with respect to the soil surface. The 3D seismic wave field, caused by the wave passage at the near field boundary, is transformed into boundary tractions , which arc then applied at the interface between the near and far fields. We present an application of the proposed method for a group of three buildings, which interact with each other through the soil during the propagation of the transient seismic waves. Although not shown here, the proposed method can handle nonlinear material properties assigned to any clement of FEM part.
2020-06-05T15:33:41ZSchauer, MarcoTaddei, FrancescaRíos Rodríguez, Gustavo A.In this contribution we present a strategy to investigate the vibrations of build ings subjected to a transient seismic excitation, including the soil-structure inte ract ion . The proposed simulation method can be helpful during the design of earthquake-resistant structures in seismic active areas as well as for the design of vibration reduction measures for buildings subjected to surrounding emissions like vibrations induced by traffic or ma chine foundations. The structures and their foundation as well as parts of the soil arc modeled by Finite Element Method (FEM). The far field is idealized as an infinite half space and modeled with the Scaled Boundary Finite Element Method (SBFEM). Both methods arc coupled at the common inter face. This approach fulfills exactly the Sommer feld radiation condition. The seismic excitation is idealized as a plane wave propagating toward the structure with an arbitrary angle with respect to the soil surface. The 3D seismic wave field, caused by the wave passage at the near field boundary, is transformed into boundary tractions , which arc then applied at the interface between the near and far fields. We present an application of the proposed method for a group of three buildings, which interact with each other through the soil during the propagation of the transient seismic waves. Although not shown here, the proposed method can handle nonlinear material properties assigned to any clement of FEM part.Pressure-displacement coupling in poroelasticity. further details of a stable finite volume formulation
http://hdl.handle.net/2117/190146
Pressure-displacement coupling in poroelasticity. further details of a stable finite volume formulation
Maliska, Clovis R.; Honório, Herminio T.
This paper further explores fundamental issues on the behaviour of a finite volume technique using staggered grids for solving poroelasticity problems. Attention is given to the well-known drawback of pressure instabilities, which arises in certain conditions, as in low permeability media, fast transients and undrained conditions. Finite volume techniques are not the first choice for solving poroelasticity problems, and the reason is cultural, since finite elements have a successful history in solving solid mechanics problems. It has been demonstrated, however, that the finite volume strategies can be successfully applied to poroelasticity problems, with the advantage of offering a single method, stable, and fully conservative for both, fluid mass and forces balance.
2020-06-05T15:17:16ZMaliska, Clovis R.Honório, Herminio T.This paper further explores fundamental issues on the behaviour of a finite volume technique using staggered grids for solving poroelasticity problems. Attention is given to the well-known drawback of pressure instabilities, which arises in certain conditions, as in low permeability media, fast transients and undrained conditions. Finite volume techniques are not the first choice for solving poroelasticity problems, and the reason is cultural, since finite elements have a successful history in solving solid mechanics problems. It has been demonstrated, however, that the finite volume strategies can be successfully applied to poroelasticity problems, with the advantage of offering a single method, stable, and fully conservative for both, fluid mass and forces balance.Numerical and analytical investigation of subcycling in the flow problem of a strongly-coupled partitioned fluid-structure interaction simulation
http://hdl.handle.net/2117/190145
Numerical and analytical investigation of subcycling in the flow problem of a strongly-coupled partitioned fluid-structure interaction simulation
de Moerloose, Laurent; Degroote, Joris
Fluid-structure interaction (FSI) simulations can be used to quantify the frequency, damping constant and amplitude of the vibration of equipment such as piping and heat exchangers. Typically, the time step is the same in the ﬂow and structural equations, but this causes long computational times when the time step is restricted due to stability requirements of only one solver. In that case, a more eﬃcient approach is to use so-called subcycling with a diﬀerent time step in the ﬂow and structural solver. In this paper, only subcycling with a smaller time step in the ﬂow solver compared to the structural solver is analyzed. The research presented here is split into two parts: an analytical study and a numerical computation of the one-dimensional ﬂow in an elastic cylindrical tube. Firstly, a monolithic analytical FSI calculation is analyzed with a Fourier stability analysis. This allows to verify the stability of the solution by considering the eigenvalues of the problem as a function of the perturbation wavenumber. The conclusions drawn from the analytical study are subsequently veriﬁed in a partitioned numerical FSI simulation, coupling the ﬂow solver Fluent with the structural solver Abaqus. The implicit coupling is achieved using an interface quasi-Newton method with an approximation of the inverse of the Jacobian (IQN-ILS), implemented in the in-house code Tango. The research shows that a stable solution is attained for signiﬁcant subcycling in the ﬂow problem: the results indicate that the solution remains temporally stable even if the time step in the ﬂow solver is only one tenth of the structural time step. However, some (temporally stable) oscillations in the resulting pressure proﬁle on the pipe wall arise when the time discretization schemes applied in the ﬂow and structural solvers are diﬀerent. These oscillations do not persist when the same time discretization scheme is applied.
2020-06-05T15:14:00Zde Moerloose, LaurentDegroote, JorisFluid-structure interaction (FSI) simulations can be used to quantify the frequency, damping constant and amplitude of the vibration of equipment such as piping and heat exchangers. Typically, the time step is the same in the ﬂow and structural equations, but this causes long computational times when the time step is restricted due to stability requirements of only one solver. In that case, a more eﬃcient approach is to use so-called subcycling with a diﬀerent time step in the ﬂow and structural solver. In this paper, only subcycling with a smaller time step in the ﬂow solver compared to the structural solver is analyzed. The research presented here is split into two parts: an analytical study and a numerical computation of the one-dimensional ﬂow in an elastic cylindrical tube. Firstly, a monolithic analytical FSI calculation is analyzed with a Fourier stability analysis. This allows to verify the stability of the solution by considering the eigenvalues of the problem as a function of the perturbation wavenumber. The conclusions drawn from the analytical study are subsequently veriﬁed in a partitioned numerical FSI simulation, coupling the ﬂow solver Fluent with the structural solver Abaqus. The implicit coupling is achieved using an interface quasi-Newton method with an approximation of the inverse of the Jacobian (IQN-ILS), implemented in the in-house code Tango. The research shows that a stable solution is attained for signiﬁcant subcycling in the ﬂow problem: the results indicate that the solution remains temporally stable even if the time step in the ﬂow solver is only one tenth of the structural time step. However, some (temporally stable) oscillations in the resulting pressure proﬁle on the pipe wall arise when the time discretization schemes applied in the ﬂow and structural solvers are diﬀerent. These oscillations do not persist when the same time discretization scheme is applied.Modelling the effects of oscillating stripe cooling during a loca event in a vver440/213 reactor pressure vessel
http://hdl.handle.net/2117/190144
Modelling the effects of oscillating stripe cooling during a loca event in a vver440/213 reactor pressure vessel
Gálik, Gabriel; Kutis, Vladimir; Paulech, Juraj; Murin, Justin
A Loss of Coolant Accident (LOCA) is initiated by damage in the primary circuit and subsequent coolant leak of a pressurised water reactor (PWR). The loss of coolant is then compensated by the Emergency Core Cooling System (ECCS) [1]. This system supplies the primary circuit with cold high pressure coolant during emergencies. The mixing of this cold coolant results in stripe cooling of the RPV wall. This article focuses on the modelling of the mixing processes of this cooling stripe and their effects on the RPV wall in a VVER440/213 Russian type PWR. The mixing processes are modelled in a transient thermohydraulic analysis which models the mixing of the coolant flows in the reactor pressure vessel and results in the overtime temperature and pressure fields within the RPV [2]. The analysis results show that the cooling stripe is not stationary. The turbulent mixing causes an unstable oscillatory motion of the cold stripe which has a notable effect on the RPV wall temperature distribution. Selected results were subsequently transferred into a thermo-mechanical analysis via one way coupling method. This analysis was performed to evaluate the PRV loading state during the LOCA transient. Where the loading stresses were shown to correspond to the oscillatory nature of the cooling strip. In conclusion, an oscillating cooling strip can result in multiple loading cycles [3,4] of the reactor pressure vessel wall during a single LOCA transient or ECCS high pressure coolant injection.
2020-06-05T15:10:47ZGálik, GabrielKutis, VladimirPaulech, JurajMurin, JustinA Loss of Coolant Accident (LOCA) is initiated by damage in the primary circuit and subsequent coolant leak of a pressurised water reactor (PWR). The loss of coolant is then compensated by the Emergency Core Cooling System (ECCS) [1]. This system supplies the primary circuit with cold high pressure coolant during emergencies. The mixing of this cold coolant results in stripe cooling of the RPV wall. This article focuses on the modelling of the mixing processes of this cooling stripe and their effects on the RPV wall in a VVER440/213 Russian type PWR. The mixing processes are modelled in a transient thermohydraulic analysis which models the mixing of the coolant flows in the reactor pressure vessel and results in the overtime temperature and pressure fields within the RPV [2]. The analysis results show that the cooling stripe is not stationary. The turbulent mixing causes an unstable oscillatory motion of the cold stripe which has a notable effect on the RPV wall temperature distribution. Selected results were subsequently transferred into a thermo-mechanical analysis via one way coupling method. This analysis was performed to evaluate the PRV loading state during the LOCA transient. Where the loading stresses were shown to correspond to the oscillatory nature of the cooling strip. In conclusion, an oscillating cooling strip can result in multiple loading cycles [3,4] of the reactor pressure vessel wall during a single LOCA transient or ECCS high pressure coolant injection.FSI simulation of an axially moving flexible cylinder entrained by a supersonic flow
http://hdl.handle.net/2117/190143
FSI simulation of an axially moving flexible cylinder entrained by a supersonic flow
Delcour, Lucas; van Langenhove, Lieva; Degroote, Joris
Air jet weaving looms are widely used to weave fabrics because of the high production speed that can be attained. This is directly linked to the high insertion speed of the yarns. The yarn is accelerated into the reed by a main nozzle and its motion is subsequently supported by underexpanded jets emanating from relay nozzles. The contact with the reed is the only mechanical guidance that the yarn experiences along its path and its motion depends heavily on its interaction with the air ﬂow. The yarn can thus deviate from its envisaged path and cause a failed insertion. Furthermore, the tension in the yarn, induced by the traction of the air and the inertia and resistance of mechanical components, can cause yarn breakage. Failed insertions and broken yarns are undesired as they require the machine to be restarted. Due to the high speed of the yarn and the mechanical components, optical accessibility to the yarn inside the main nozzle is very limited. Furthermore, the complex geometry experienced by the air ﬂow makes it hard to assess the inﬂuence of adaptations. Fluidstructure interaction (FSI) simulations might assist in understanding the behaviour of the air ﬂow and the yarn. However, the ﬂexibility of the yarn in combination with the high speed ﬂow presents its own challenges. In this research an attempt is made to simulate the launch of a yarn by a single main nozzle into the atmosphere. To better approximate reality, the yarn is considered to be stored on a drum in front of the nozzle. A two-way ﬂuid-structure interaction simulation is performed using Fluent for the ﬂow side, Abaqus for the structural side and the inhouse code Tango for the coupling. Continuum elements or beam elements are used for the structure. The axial motion and large transversal displacement of the yarn pose signiﬁcant challenges for a single deforming grid in the ﬂow solver. To avoid these complications a Chimera approach, which superimposes several meshes, is opted for. The FSI simulations show that the yarn can indeed be represented by beam elements. The gain in computational time by switching to beam elements is evaluated and the results from the FSI calculation are compared to experimental results in terms of yarn velocity. Stresses in the yarn are examined to identify high tension regions.
2020-06-05T15:07:33ZDelcour, Lucasvan Langenhove, LievaDegroote, JorisAir jet weaving looms are widely used to weave fabrics because of the high production speed that can be attained. This is directly linked to the high insertion speed of the yarns. The yarn is accelerated into the reed by a main nozzle and its motion is subsequently supported by underexpanded jets emanating from relay nozzles. The contact with the reed is the only mechanical guidance that the yarn experiences along its path and its motion depends heavily on its interaction with the air ﬂow. The yarn can thus deviate from its envisaged path and cause a failed insertion. Furthermore, the tension in the yarn, induced by the traction of the air and the inertia and resistance of mechanical components, can cause yarn breakage. Failed insertions and broken yarns are undesired as they require the machine to be restarted. Due to the high speed of the yarn and the mechanical components, optical accessibility to the yarn inside the main nozzle is very limited. Furthermore, the complex geometry experienced by the air ﬂow makes it hard to assess the inﬂuence of adaptations. Fluidstructure interaction (FSI) simulations might assist in understanding the behaviour of the air ﬂow and the yarn. However, the ﬂexibility of the yarn in combination with the high speed ﬂow presents its own challenges. In this research an attempt is made to simulate the launch of a yarn by a single main nozzle into the atmosphere. To better approximate reality, the yarn is considered to be stored on a drum in front of the nozzle. A two-way ﬂuid-structure interaction simulation is performed using Fluent for the ﬂow side, Abaqus for the structural side and the inhouse code Tango for the coupling. Continuum elements or beam elements are used for the structure. The axial motion and large transversal displacement of the yarn pose signiﬁcant challenges for a single deforming grid in the ﬂow solver. To avoid these complications a Chimera approach, which superimposes several meshes, is opted for. The FSI simulations show that the yarn can indeed be represented by beam elements. The gain in computational time by switching to beam elements is evaluated and the results from the FSI calculation are compared to experimental results in terms of yarn velocity. Stresses in the yarn are examined to identify high tension regions.