Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12390
M. Afaq, S. Turek, A. Ouazzi, Arooj Fatima
We have developed a monolithic Newton-multigrid solver for multiphase flow problems which solves velocity, pressure and interface position simultaneously. The main idea of our work is based on the formulations discussed in [1], where it points out the feasibility of a fully implicit monolithic solver for multiphase flow problems via two formulations, a curvature-free level set approach and a curvature-free cutoff material function approach. Both formulations are fully implicit and have the advantages of requiring less regularity, since neither normals nor curvature are explicitly calculated, and no capillary time restriction. Furthermore, standard Navier-Stokes solvers might be used, which do not have to take into account inhomogeneous force terms. The reinitialization issue is integrated with a nonlinear terms within the formulations.The nonlinearity is treated with a Newton-type solver with divided difference evaluation of the Jacobian matrices. The resulting linearized system inside of the outer Newton solver is a typical saddle point problem which is solved using the geometrical multigrid with Vanka-like smoother using higher order stable FEM pair $Q_2/P^{text{disc}}_1$ for velocity and pressure and $Q_2$ for all other variables. The method is implemented into an existing software packages for the numerical simulation of multiphase flows (FeatFlow). The robustness and accuracy of this solver is tested for two different test cases, i.e. static bubble and oscillating bubble, respectively [2].REFERENCES[1] Ouazzi, A., Turek, S. and Damanik, H. A curvature-free multiphase flow solver via surface stress-based formulation. Int. J. Num. Meth. Fluids., Vol. 88, pp. 18–31, (2018).[2] Afaq, M. A., Turek, S., Ouazzi, A. and Fatima, A. Monolithic Newton-Multigrid Solver for Multiphase Flow Problems with Surface Tension. Ergebnisberichte des Instituts fuer Angewandte Mathematik Nummer 636, Fakultaet fuer Mathematik, TU Dortmund University, 636, 2021.
我们开发了一种求解多相流问题的单片牛顿-多网格求解器，可以同时求解速度、压力和界面位置。我们工作的主要思想是基于[1]中讨论的公式，其中指出了通过两种公式(无曲率水平集方法和无曲率截止材料函数方法)求解多相流问题的全隐式单片求解器的可行性。这两个公式都是完全隐式的，并且具有要求较少规则性的优点，因为法线和曲率都没有显式计算，也没有毛细时间限制。此外，可以使用标准的Navier-Stokes解，它不必考虑非齐次力项。重新初始化问题与公式中的非线性项相结合。用雅可比矩阵的分差求值的牛顿型求解器处理非线性问题。所得到的外牛顿求解器内部线性化系统是一个典型的鞍点问题，该问题采用具有Vanka-like光滑的几何多重网格，采用高阶稳定有限元对$Q_2/P^{text{disc}}_1$表示速度和压力，$Q_2$表示其他变量。该方法已在现有的多相流数值模拟软件包(FeatFlow)中实现。通过静态气泡和振荡气泡两种不同的测试用例对求解器的鲁棒性和准确性进行了测试[2]。[1]刘建军，刘建军，刘建军，等。一种基于表面应力的无曲率多相流求解方法。Int。J. Num.冰毒。液体。， Vol. 88, pp. 18-31， (2018).[2]Afaq, M. A, Turek, S.， Ouazzi, A.和Fatima, A.考虑表面张力的多相流问题的单片牛顿-多网格求解器。德国多特蒙德大学数学研究所(第636期)，德国多特蒙德大学数学研究所(第636,2021)。
{"title":"Monolithic Newton-Multigrid Solver for Multiphase Flow Problems with Surface Tension","authors":"M. Afaq, S. Turek, A. Ouazzi, Arooj Fatima","doi":"10.4995/yic2021.2021.12390","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12390","url":null,"abstract":"We have developed a monolithic Newton-multigrid solver for multiphase flow problems which solves velocity, pressure and interface position simultaneously. The main idea of our work is based on the formulations discussed in [1], where it points out the feasibility of a fully implicit monolithic solver for multiphase flow problems via two formulations, a curvature-free level set approach and a curvature-free cutoff material function approach. Both formulations are fully implicit and have the advantages of requiring less regularity, since neither normals nor curvature are explicitly calculated, and no capillary time restriction. Furthermore, standard Navier-Stokes solvers might be used, which do not have to take into account inhomogeneous force terms. The reinitialization issue is integrated with a nonlinear terms within the formulations.The nonlinearity is treated with a Newton-type solver with divided difference evaluation of the Jacobian matrices. The resulting linearized system inside of the outer Newton solver is a typical saddle point problem which is solved using the geometrical multigrid with Vanka-like smoother using higher order stable FEM pair $Q_2/P^{text{disc}}_1$ for velocity and pressure and $Q_2$ for all other variables. The method is implemented into an existing software packages for the numerical simulation of multiphase flows (FeatFlow). The robustness and accuracy of this solver is tested for two different test cases, i.e. static bubble and oscillating bubble, respectively [2].REFERENCES[1] Ouazzi, A., Turek, S. and Damanik, H. A curvature-free multiphase flow solver via surface stress-based formulation. Int. J. Num. Meth. Fluids., Vol. 88, pp. 18–31, (2018).[2] Afaq, M. A., Turek, S., Ouazzi, A. and Fatima, A. Monolithic Newton-Multigrid Solver for Multiphase Flow Problems with Surface Tension. Ergebnisberichte des Instituts fuer Angewandte Mathematik Nummer 636, Fakultaet fuer Mathematik, TU Dortmund University, 636, 2021.","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114590771","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12566
S. Gregori, J. Gil, M. Tur, A. Pedrosa, Fco. Javier Fuenmayor
Overhead contact line or catenary is the structure composed of support elements and wires responsible for the power supply of the locomotive through sliding contact with the pantograph. This contact causes wear not only on the pantograph contact strips but also in the contact wire, which produces a reduction on its effective section and eventually its replacement, resulting in the stop of the railway traffic with its associate economical and operational drawbacks. For this reason, it is important for catenary designers to count with appropriate tools able to predict the contact wire wear behaviour for extending the service life of the system. This work proposes a strategy to simulate the long-term contact wire wear evolution considering the mutual influence between the dynamic behaviour and wear of the system. The method is based on two pillars: the efficient simulation of the catenary-pantograph dynamic interaction [1] and a heuristic wear model [2] which considers mechanical wear due to friction and electrical wear produced by Joule effect and electric arcs. With the proposed simulation tool, we analyse the effect on the long-term contact wire height irregularity caused by wear of different parameters such as the train speed, the pantograph uplift force or the presence of installation errors in the catenary.
{"title":"Simulation of the contact wire wear evolution in high speed overhead contact lines","authors":"S. Gregori, J. Gil, M. Tur, A. Pedrosa, Fco. Javier Fuenmayor","doi":"10.4995/yic2021.2021.12566","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12566","url":null,"abstract":"Overhead contact line or catenary is the structure composed of support elements and wires responsible for the power supply of the locomotive through sliding contact with the pantograph. This contact causes wear not only on the pantograph contact strips but also in the contact wire, which produces a reduction on its effective section and eventually its replacement, resulting in the stop of the railway traffic with its associate economical and operational drawbacks. For this reason, it is important for catenary designers to count with appropriate tools able to predict the contact wire wear behaviour for extending the service life of the system. This work proposes a strategy to simulate the long-term contact wire wear evolution considering the mutual influence between the dynamic behaviour and wear of the system. The method is based on two pillars: the efficient simulation of the catenary-pantograph dynamic interaction [1] and a heuristic wear model [2] which considers mechanical wear due to friction and electrical wear produced by Joule effect and electric arcs. With the proposed simulation tool, we analyse the effect on the long-term contact wire height irregularity caused by wear of different parameters such as the train speed, the pantograph uplift force or the presence of installation errors in the catenary.","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"26 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120929868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12367
I. Kalaimani, J. Dietzsch, M. Groß
Rotor-dynamical systems made of 3D-fiber-reinforced composites which are subjected to dynamical loads exhibit an increased fiber bending stiffness in numerical simulations. We propose a numerical modeling approach of fiber-reinforced composites that treats this behaviour accurately. Our model uses a multi-field mixed finite element formulation based on a dynamic variational approach, as demonstrated in [1], to perform long-term dynamic simulations that yield numerical solutions with increased accuracy in efficient CPU-time.We extend a Cauchy continuum with higher-order gradients of the deformation mapping as an independent field in the functional formulation, as suggested in [2], to model the bending stiffness of fibers accurately. This extended continuum also takes into account the higher-order energy contributions including the fiber curvature along with popular proven approaches that avoid the numerical locking effect of the fibers efficiently.We apply the proposed approach on Cook’s cantilever beam with a hyperelastic, transversely isotropic, polyconvex material behavior in a transient dynamic analysis. The beam is subjected to bending loads with a strong dependence of the overall stiffness on the fiber orientation. The spatial and temporal convergence as well as the conservation properties are analyzed. It is observed that the model needs an improved numerical treatment to conserve total momenta as well as total energy.REFERENCES M. Groß and J. Dietzsch, "Variational-based locking-free energy–momentum schemes of higher-order for thermo-viscoelastic fiber-reinforced continua", Computer Methods in Applied Mechanics and Engineering, (2019), 631-671, 343. T. Asmanoglo and A. Menzel, “A multi-field finite element approach for the modelling of fibre-reinforced composites with fibre-bending stiffness”, Computer Methods in Applied Mechanics and Engineering, (2017), 1037-1067, 317.
在数值模拟中，三维纤维增强复合材料转子动力系统在动力载荷作用下，纤维弯曲刚度增加。我们提出了一种精确处理这种行为的纤维增强复合材料的数值模拟方法。我们的模型使用基于动态变分方法的多场混合有限元公式，如[1]所示，执行长期动态模拟，在有效的cpu时间内获得精度更高的数值解。我们将变形映射的高阶梯度的柯西连续统扩展为泛函公式中的独立场，如[2]所示，以准确地模拟纤维的弯曲刚度。该扩展连续体还考虑了包括光纤曲率在内的高阶能量贡献，以及有效避免光纤数值锁定效应的常用方法。我们将提出的方法应用于具有超弹性、横向各向同性、多凸材料行为的Cook悬臂梁的瞬态动力分析。梁受到弯曲载荷与整体刚度对纤维取向的强烈依赖。分析了该方法的时空收敛性和守恒性。观察到该模型需要改进数值处理，以保持总动量和总能量。M. Groß和J. Dietzsch，“基于变分的热粘弹性纤维增强连续体的无锁定能量-动量格式”，应用力学与工程计算机方法，(2019)，631-671,343。T. Asmanoglo, A. Menzel，“基于纤维弯曲刚度的纤维增强复合材料多场有限元建模方法”，应用力学与工程学报，(2017)，1037-1067,317。
{"title":"Momentum conserving dynamic variational approach for the modeling of fiber-bending stiffness in fiber-reinforced composites","authors":"I. Kalaimani, J. Dietzsch, M. Groß","doi":"10.4995/yic2021.2021.12367","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12367","url":null,"abstract":"Rotor-dynamical systems made of 3D-fiber-reinforced composites which are subjected to dynamical loads exhibit an increased fiber bending stiffness in numerical simulations. We propose a numerical modeling approach of fiber-reinforced composites that treats this behaviour accurately. Our model uses a multi-field mixed finite element formulation based on a dynamic variational approach, as demonstrated in [1], to perform long-term dynamic simulations that yield numerical solutions with increased accuracy in efficient CPU-time.We extend a Cauchy continuum with higher-order gradients of the deformation mapping as an independent field in the functional formulation, as suggested in [2], to model the bending stiffness of fibers accurately. This extended continuum also takes into account the higher-order energy contributions including the fiber curvature along with popular proven approaches that avoid the numerical locking effect of the fibers efficiently.We apply the proposed approach on Cook’s cantilever beam with a hyperelastic, transversely isotropic, polyconvex material behavior in a transient dynamic analysis. The beam is subjected to bending loads with a strong dependence of the overall stiffness on the fiber orientation. The spatial and temporal convergence as well as the conservation properties are analyzed. It is observed that the model needs an improved numerical treatment to conserve total momenta as well as total energy.REFERENCES M. Groß and J. Dietzsch, \"Variational-based locking-free energy–momentum schemes of higher-order for thermo-viscoelastic fiber-reinforced continua\", Computer Methods in Applied Mechanics and Engineering, (2019), 631-671, 343. T. Asmanoglo and A. Menzel, “A multi-field finite element approach for the modelling of fibre-reinforced composites with fibre-bending stiffness”, Computer Methods in Applied Mechanics and Engineering, (2017), 1037-1067, 317.","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"87 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132575341","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12606
Jaime Gil-Romero, S. Gregori, M. Tur, F. Fuenmayor
Dynamics of repetitive structures subjected to moving loads is a common problem in railway engineering. Bridges, rails or catenaries are the most representative periodic structures, on which the train acts as a moving excitation. Usually, these structures are long enough to consider that their dynamic response is in permanent regime. In this work we present a method to obtain the steady-state solution of an infinite periodic structure subjected to a periodic moving load running at constant speed 𝑉𝑉.This problem has been dealt with in the literature by different approaches. Analytical models [1], two-and-a-half dimensional (2.5D) Finite Element models [2] and the Wave Finite Element Method (WFEM) [3] are found to be used. The method proposed in this work is valid for any generic periodic structure because it is modelled by the classical Finite Element Method. It is mathematically simpler and more efficient compared to WFEM, and it avoids the numerical problems that arise when WFEM is applied to catenaries.The proposed method consists of solving the dynamic interaction problem on a single repetitive block of the structure in which the periodicity condition is applied. Each block of length 𝐿𝐿 is excited by the same load. Thus, the periodicity condition states that the solution at the left boundary of the block is the same as at the right boundary but advanced a period 𝑇𝑇=𝐿𝐿/𝑉𝑉. This condition is imposed in the frequency domain and a procedure to shift into the time domain is presented.
{"title":"Dynamic response of periodic infinite structure to arbitrary moving load based on the Finite Element Method","authors":"Jaime Gil-Romero, S. Gregori, M. Tur, F. Fuenmayor","doi":"10.4995/yic2021.2021.12606","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12606","url":null,"abstract":"Dynamics of repetitive structures subjected to moving loads is a common problem in railway engineering. Bridges, rails or catenaries are the most representative periodic structures, on which the train acts as a moving excitation. Usually, these structures are long enough to consider that their dynamic response is in permanent regime. In this work we present a method to obtain the steady-state solution of an infinite periodic structure subjected to a periodic moving load running at constant speed 𝑉𝑉.This problem has been dealt with in the literature by different approaches. Analytical models [1], two-and-a-half dimensional (2.5D) Finite Element models [2] and the Wave Finite Element Method (WFEM) [3] are found to be used. The method proposed in this work is valid for any generic periodic structure because it is modelled by the classical Finite Element Method. It is mathematically simpler and more efficient compared to WFEM, and it avoids the numerical problems that arise when WFEM is applied to catenaries.The proposed method consists of solving the dynamic interaction problem on a single repetitive block of the structure in which the periodicity condition is applied. Each block of length 𝐿𝐿 is excited by the same load. Thus, the periodicity condition states that the solution at the left boundary of the block is the same as at the right boundary but advanced a period 𝑇𝑇=𝐿𝐿/𝑉𝑉. This condition is imposed in the frequency domain and a procedure to shift into the time domain is presented.","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114676604","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12344
Z. Malinowski, Kamil Jasiewicz, A. Cebo-Rudnicka
The inverse solution tests have been performed to the experimental data obtained during the oxidised Armco steel plate cooling by the air nozzle. The three-dimensional numerical model of heat transfer during the plate cooling has been considered. Steel products cooled in the air from high temperatures are covered with the oxide layer having significantly lower conductivity and a different surface structure comparing to the non-oxidised metal surface. The Armco steel has been selected as the experimental material because it oxidised in a similar way to carbon steels but there is no microstructure evolution process in Armco steel below 900oC. It eliminates in the inverse solutions serious problems caused by a latent heat of microstructure evolutions encountered during carbon steel cooling. In the present study the steel plate has been heated to about 900℃ and cooled by the circular air jet. The plate temperature has been measured by 36 thermocouples. The test of the selected inverse solution models involving a different number of degrees of freedom have been performed. The influence of the scale layer on the results of the inverse solution to the heat flux and heat transfer coefficient has been investigated.
{"title":"Inverse solution to the heat transfer coeffcient for the oxidized ARMCO steel plate cooling by the air nozzle from high temperature","authors":"Z. Malinowski, Kamil Jasiewicz, A. Cebo-Rudnicka","doi":"10.4995/yic2021.2021.12344","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12344","url":null,"abstract":"The inverse solution tests have been performed to the experimental data obtained during the oxidised Armco steel plate cooling by the air nozzle. The three-dimensional numerical model of heat transfer during the plate cooling has been considered. Steel products cooled in the air from high temperatures are covered with the oxide layer having significantly lower conductivity and a different surface structure comparing to the non-oxidised metal surface. The Armco steel has been selected as the experimental material because it oxidised in a similar way to carbon steels but there is no microstructure evolution process in Armco steel below 900oC. It eliminates in the inverse solutions serious problems caused by a latent heat of microstructure evolutions encountered during carbon steel cooling. In the present study the steel plate has been heated to about 900℃ and cooled by the circular air jet. The plate temperature has been measured by 36 thermocouples. The test of the selected inverse solution models involving a different number of degrees of freedom have been performed. The influence of the scale layer on the results of the inverse solution to the heat flux and heat transfer coefficient has been investigated.","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129519706","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12219
R. Tielen, M. Möller, K. Vuik
Isogeometric Analysis [1] has become increasingly popular as an alternative to the Finite Element Method. Solving the resulting linear systems when adopting higher order B-spline basis functions remains a challenging task, as most (standard) iterative methods have a deteriorating preformance for higher values of the approximation order p.Recently, we succesfully applied p-multigrid methods to discretizations arising in IsogeometricAnalysis [2]. In contrast to h-multigrid methods, where each level of the multigrid hierarchycorresponds to a different mesh width h, the p-multigrid hierarchy is constructed based on different approximation orders. The residual equation is then solved at level p = 1, enabling the use of efficient solution techniques developed for low-order standard FEM. Numerical results show that the number of iterations needed for convergence is independent of both h and p when the p-multigrid method is enhanced with a smoother based on an Incomplete LU factorization with dual treshold (ILUT). However, a slight dependence on the number of patches has been observed for multipatch geometries.Since the resulting system matrix has a block structure in case of a multipatch geometry, weconsider the use of block ILUT as a smoother. Results indicate that the use of block ILUT can be an efficient alternative to ILUT on multipatch geometries within a heterogeneous HPC framework. Prelimenary results for p-multigrid methods adopting a block ILUT smoother will be presented in this talk. Furthermore, we investigate the use of alternative multigrid hierarchies, in particular when considering time-dependent problems.REFERENCES[1] T.J.R. Hughes, J.A. Cottrell and Y. Bazilevs, Isogeometric Analysis: CAD, Finite Elements,NURBS, Exact Geometry and Mesh Refinement, Computer Methods in Applied Mechanicsand Engineering, 194, 4135 - 4195, 2005[2] R.Tielen, M. Möller, D. Göddeke and C. Vuik, p-multigrid methods and their comparison toh-multigrid methods within Isogeometric Analysis, Computer Methods in Applied Mechanicsand Engineering, 372, 2020
{"title":"Multigrid Reduced in Time for Isogeometric Analysis","authors":"R. Tielen, M. Möller, K. Vuik","doi":"10.4995/yic2021.2021.12219","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12219","url":null,"abstract":"Isogeometric Analysis [1] has become increasingly popular as an alternative to the Finite Element Method. Solving the resulting linear systems when adopting higher order B-spline basis functions remains a challenging task, as most (standard) iterative methods have a deteriorating preformance for higher values of the approximation order p.Recently, we succesfully applied p-multigrid methods to discretizations arising in IsogeometricAnalysis [2]. In contrast to h-multigrid methods, where each level of the multigrid hierarchycorresponds to a different mesh width h, the p-multigrid hierarchy is constructed based on different approximation orders. The residual equation is then solved at level p = 1, enabling the use of efficient solution techniques developed for low-order standard FEM. Numerical results show that the number of iterations needed for convergence is independent of both h and p when the p-multigrid method is enhanced with a smoother based on an Incomplete LU factorization with dual treshold (ILUT). However, a slight dependence on the number of patches has been observed for multipatch geometries.Since the resulting system matrix has a block structure in case of a multipatch geometry, weconsider the use of block ILUT as a smoother. Results indicate that the use of block ILUT can be an efficient alternative to ILUT on multipatch geometries within a heterogeneous HPC framework. Prelimenary results for p-multigrid methods adopting a block ILUT smoother will be presented in this talk. Furthermore, we investigate the use of alternative multigrid hierarchies, in particular when considering time-dependent problems.REFERENCES[1] T.J.R. Hughes, J.A. Cottrell and Y. Bazilevs, Isogeometric Analysis: CAD, Finite Elements,NURBS, Exact Geometry and Mesh Refinement, Computer Methods in Applied Mechanicsand Engineering, 194, 4135 - 4195, 2005[2] R.Tielen, M. Möller, D. Göddeke and C. Vuik, p-multigrid methods and their comparison toh-multigrid methods within Isogeometric Analysis, Computer Methods in Applied Mechanicsand Engineering, 372, 2020","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131270250","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12473
M. Zlatić, M. Čanađija
With the recent surge in neural network usage, machine learning libraries have become more convenient to use and implement. In this paper we investigate the possibility of using neural networks in order to faster process displacements obtained from finite element calculation and replace existing post-processing procedures. The method is implemented on 2D finite elements for their relative ease of usage and manipulation. A speed up is observed in comparison to traditional methods of post-processing. Possible further applications of this method are also presented in this paper.
{"title":"Reducing computational time for FEM postprocessing through the use of feedforward neural networks","authors":"M. Zlatić, M. Čanađija","doi":"10.4995/yic2021.2021.12473","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12473","url":null,"abstract":"With the recent surge in neural network usage, machine learning libraries have become more convenient to use and implement. In this paper we investigate the possibility of using neural networks in order to faster process displacements obtained from finite element calculation and replace existing post-processing procedures. The method is implemented on 2D finite elements for their relative ease of usage and manipulation. A speed up is observed in comparison to traditional methods of post-processing. Possible further applications of this method are also presented in this paper.","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"25 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127535692","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12220
Josep Chordà- Monsonís, Mª Dolores Martínez Rodrigo, Pedro Galvín Barrera, Antonio Romero Ordóñez, Emmanuela Moliner Cabedo
Abstract. In the present contribution the coupling effect of the railway track on the dynamic response of single-track bridges with several simply-supported spans is investigated. The response of such structures is of interest due to the possible appearance of high vertical accelerations at the platform, with adverse consequences such as ballast deconsolidation, loss of track stability etc., especially at resonance. Single track railway bridges due to their inherent low mass and damping, are especially critical in this regard.This paper provides a detailed sensitivity analysis over a single track railway bridge catalogue considering lengths of interest from 10 to 25 meters with different typologies: concrete filler beam and concrete slab decks, taking into account and neglecting the vertical flexibility of elastic bearings. To this end, a 2D Finite Element numerical model is implemented admitting a three-layer discrete representation of the track components mass, stiffness and damping, based on [1].The effect of the track on the bridge acceleration response is evaluated. A parametric analysis is presented with the aims of (i) detecting the track parameters that affect the most the bridge vertical acceleration response under railway traffic and (ii) determining what bridges are the most affected by the rigidity and the damping induced by the track components. Additionally, the effect of including several isostatic spans in the model is also evaluated with the goal of concluding where the maximum response occurs and what models are on the safe side.In addition, the experimental response of a real two-span railway bridge belonging to the Spanish railway conventional network is presented under operating conditions. The bridge modal properties identified in a previous work by the authors [2] are employed and the previously described numerical model is calibrated and used to simulate the structural response under railway traffic.Finally, preliminary conclusions are extracted regarding the coupling effect of the track components both at resonant and not resonant conditions.
{"title":"ANALYSIS OF THE INFLUENCE OF THE BALLAST TRACK IN THE DYNAMIC BEHAVIOUR OF SINGLE-TRACK RAILWAY BRIDGES OF DIFFERENT TYPOLOGIES","authors":"Josep Chordà- Monsonís, Mª Dolores Martínez Rodrigo, Pedro Galvín Barrera, Antonio Romero Ordóñez, Emmanuela Moliner Cabedo","doi":"10.4995/yic2021.2021.12220","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12220","url":null,"abstract":"Abstract. In the present contribution the coupling effect of the railway track on the dynamic response of single-track bridges with several simply-supported spans is investigated. The response of such structures is of interest due to the possible appearance of high vertical accelerations at the platform, with adverse consequences such as ballast deconsolidation, loss of track stability etc., especially at resonance. Single track railway bridges due to their inherent low mass and damping, are especially critical in this regard.This paper provides a detailed sensitivity analysis over a single track railway bridge catalogue considering lengths of interest from 10 to 25 meters with different typologies: concrete filler beam and concrete slab decks, taking into account and neglecting the vertical flexibility of elastic bearings. To this end, a 2D Finite Element numerical model is implemented admitting a three-layer discrete representation of the track components mass, stiffness and damping, based on [1].The effect of the track on the bridge acceleration response is evaluated. A parametric analysis is presented with the aims of (i) detecting the track parameters that affect the most the bridge vertical acceleration response under railway traffic and (ii) determining what bridges are the most affected by the rigidity and the damping induced by the track components. Additionally, the effect of including several isostatic spans in the model is also evaluated with the goal of concluding where the maximum response occurs and what models are on the safe side.In addition, the experimental response of a real two-span railway bridge belonging to the Spanish railway conventional network is presented under operating conditions. The bridge modal properties identified in a previous work by the authors [2] are employed and the previously described numerical model is calibrated and used to simulate the structural response under railway traffic.Finally, preliminary conclusions are extracted regarding the coupling effect of the track components both at resonant and not resonant conditions.","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"83 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125385358","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12250
A. Ouazzi, N. Begum, S. Turek
This talk is concerned with the application of Finite Element Method (FEM) and Newton-Multigrid solver to simulate thixotropic flows.The thixotropy phenomena are introduced to yield stress material by taking into consideration the internal material micro structure using a structure parameter. Firstly, the viscoplastic stress is modified to include the thixotropic stress dependent on the structure parameter. Secondly, an evolution equation for the structure parameter is introduced to induce the time-dependent processof competition between the destruction (breakdown) and the construction (buildup) inhabited in the material. Substantially, this is done simply by introducing a structure-parameter-dependentviscosity into the rheological model for yield stress material, as for instance the Houska model based on a viscosity approach for the Bingham model [2].The modified viscoplastic stress w.r.t. the structure parameter which is integrated, in quasi-Newtonian manner or lagrangian multiplier manner, into the generalized Navier-Stokes equations and the evolution equation for the structure parameter constitutes the main core of full set of mod-eling equations, which are creditable as the privilege answer to incorporate thixotropy phenomena.The nonlinearity of the problem, related to the dependency of the diffusive stress on the material parameters, is treated with generalized Newton’s method w.r.t. the Jacobian’s singularities having a global convergence property. The linearized systems inside the outer Newton loops are solvedusing the geometrical multigrid methods with a Vanka-like smoother taking into account a stable FEM approximation pair for velocity and pressure with discontinuous pressure and biquadratic velocity spaces.We analyze the accuracy, robustness and efficiency of the Newton-Multigrid FEM solver [1] throughout the solution of thixotropic flow problems of benchmarking character in channel and Couette device [3].
{"title":"Monolithic Finite Element Method for the simulation of thixo-viscoplastic flows","authors":"A. Ouazzi, N. Begum, S. Turek","doi":"10.4995/yic2021.2021.12250","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12250","url":null,"abstract":"This talk is concerned with the application of Finite Element Method (FEM) and Newton-Multigrid solver to simulate thixotropic flows.The thixotropy phenomena are introduced to yield stress material by taking into consideration the internal material micro structure using a structure parameter. Firstly, the viscoplastic stress is modified to include the thixotropic stress dependent on the structure parameter. Secondly, an evolution equation for the structure parameter is introduced to induce the time-dependent processof competition between the destruction (breakdown) and the construction (buildup) inhabited in the material. Substantially, this is done simply by introducing a structure-parameter-dependentviscosity into the rheological model for yield stress material, as for instance the Houska model based on a viscosity approach for the Bingham model [2].The modified viscoplastic stress w.r.t. the structure parameter which is integrated, in quasi-Newtonian manner or lagrangian multiplier manner, into the generalized Navier-Stokes equations and the evolution equation for the structure parameter constitutes the main core of full set of mod-eling equations, which are creditable as the privilege answer to incorporate thixotropy phenomena.The nonlinearity of the problem, related to the dependency of the diffusive stress on the material parameters, is treated with generalized Newton’s method w.r.t. the Jacobian’s singularities having a global convergence property. The linearized systems inside the outer Newton loops are solvedusing the geometrical multigrid methods with a Vanka-like smoother taking into account a stable FEM approximation pair for velocity and pressure with discontinuous pressure and biquadratic velocity spaces.We analyze the accuracy, robustness and efficiency of the Newton-Multigrid FEM solver [1] throughout the solution of thixotropic flow problems of benchmarking character in channel and Couette device [3].","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123646079","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-07-07DOI: 10.4995/yic2021.2021.12684
Nanda Kishore Bellam Muralidhar, D. Lorenz
Fiber metal laminates (FML) are lightweight hybrid structural materials that combine the ductile properties of metal with high specific stiffness of fiber reinforced plastics. These advantages led to a dramatic increase in such materials for aeronautical structures over the last few years. One of the most common and vulnerable defects in FML is impact-related delamination, often invisible to the human eye. Guided ultrasonic waves (GUW) show high potential for monitoring structural integrity and damage detection in thin-walled structures by using the physical phenomena of wave propagation interacting with the defects [1]. The focus of this research project is on describing an inverse solution for the detection and characterization of defect in FML. Model-based damage analysis utilizes an accurate finite element model (FEM) of GUW interaction with the damage. The FEM is developed by project partners from mechanics at Helmut-Schmidt-University in Hamburg, Germany, and will be treated as a black-box for further analysis. A Bayesian approach (Markov chain Monte Carlo) is employed to characterize the damage and quantify its uncertainties. This inference problem in a stochastic framework requires a very large number of forward solves. Therefore, a profound investigation is carried out on different reduced-order modeling (ROM) methods in order to apply a suitable technique that significantly improves the computational efficiency. The proposed method is well illustrated on a simpler case study for the damage detection, localization and characterization using 2D elastic wave equation. The damage in this case is modeled as a reduction in the wave propagation velocity. The inference problem utilizes a parameterized projection-based ROM coupled with a surrogate model [2] instead of the underlying highdimensional model. This research is funded by the Deutsche Forschungsgemeinschaft Research Unit 3022 under grant LO1436/12-1.REFERENCES [1] R. Lammering, U. Gabbert, M. Sinapius, T. Schuster, P. Wierach (Eds)(2018) Lamb-Wave Based Structural Health Monitoring in Polymer Composites, Springer International Publishing. [2] Paul-Dubois-Taine A, Amsallem D. An adaptive and efficient greedy procedure for the optimal training of parametric reduced-order models. International Journal for Numerical Methods in Engineering 2014.
纤维金属层压板(FML)是一种轻质混合结构材料，它结合了金属的延展性和纤维增强塑料的高比刚度。这些优点导致在过去几年中，航空结构中使用这种材料的数量急剧增加。FML中最常见和最脆弱的缺陷之一是与冲击相关的分层，通常是肉眼看不到的。导波(GUW)利用波传播与缺陷相互作用的物理现象，在薄壁结构的结构完整性监测和损伤检测方面显示出巨大的潜力[1]。本研究项目的重点是描述FML中缺陷检测和表征的逆解。基于模型的损伤分析利用了GUW与损伤相互作用的精确有限元模型。该FEM由德国汉堡赫尔穆特-施密特大学的项目合作伙伴开发，并将作为进一步分析的黑匣子。采用贝叶斯方法(马尔可夫链蒙特卡罗)表征损伤并量化其不确定性。这种随机框架下的推理问题需要大量的前向解。因此，对不同的降阶建模方法进行了深入的研究，以期采用合适的技术来显著提高计算效率。通过一个简单的二维弹性波动方程的损伤检测、定位和表征实例，很好地说明了所提出的方法。在这种情况下，损伤被模拟为波传播速度的降低。推理问题使用了一个参数化的基于投影的ROM和一个代理模型[2]，而不是底层的高维模型。本研究由德国研究小组3022资助，资助号为LO1436/12-1。[1] R. Lammering, U. Gabbert, M. Sinapius, T. Schuster, P. Wierach(主编)(2018)基于lamb波的聚合物复合材料结构健康监测，Springer International Publishing。[2]张建军，张建军，张建军，等。一种基于自适应贪心算法的参数化降阶模型优化训练方法。国际工程数值方法学报，2014。
{"title":"A Model-Based Damage Identification using Guided Ultrasonic Wave Propagation in Fiber Metal Laminates","authors":"Nanda Kishore Bellam Muralidhar, D. Lorenz","doi":"10.4995/yic2021.2021.12684","DOIUrl":"https://doi.org/10.4995/yic2021.2021.12684","url":null,"abstract":"Fiber metal laminates (FML) are lightweight hybrid structural materials that combine the ductile properties of metal with high specific stiffness of fiber reinforced plastics. These advantages led to a dramatic increase in such materials for aeronautical structures over the last few years. One of the most common and vulnerable defects in FML is impact-related delamination, often invisible to the human eye. Guided ultrasonic waves (GUW) show high potential for monitoring structural integrity and damage detection in thin-walled structures by using the physical phenomena of wave propagation interacting with the defects [1]. The focus of this research project is on describing an inverse solution for the detection and characterization of defect in FML. Model-based damage analysis utilizes an accurate finite element model (FEM) of GUW interaction with the damage. The FEM is developed by project partners from mechanics at Helmut-Schmidt-University in Hamburg, Germany, and will be treated as a black-box for further analysis. A Bayesian approach (Markov chain Monte Carlo) is employed to characterize the damage and quantify its uncertainties. This inference problem in a stochastic framework requires a very large number of forward solves. Therefore, a profound investigation is carried out on different reduced-order modeling (ROM) methods in order to apply a suitable technique that significantly improves the computational efficiency. The proposed method is well illustrated on a simpler case study for the damage detection, localization and characterization using 2D elastic wave equation. The damage in this case is modeled as a reduction in the wave propagation velocity. The inference problem utilizes a parameterized projection-based ROM coupled with a surrogate model [2] instead of the underlying highdimensional model. This research is funded by the Deutsche Forschungsgemeinschaft Research Unit 3022 under grant LO1436/12-1.REFERENCES [1] R. Lammering, U. Gabbert, M. Sinapius, T. Schuster, P. Wierach (Eds)(2018) Lamb-Wave Based Structural Health Monitoring in Polymer Composites, Springer International Publishing. [2] Paul-Dubois-Taine A, Amsallem D. An adaptive and efficient greedy procedure for the optimal training of parametric reduced-order models. International Journal for Numerical Methods in Engineering 2014.","PeriodicalId":406819,"journal":{"name":"Proceedings of the YIC 2021 - VI ECCOMAS Young Investigators Conference","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-07-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115708903","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：