Accuracy and Performance of Fluid-Structure Interaction Algorithms with Explicit versus Implicit Formulations of the Fluid Solver

Author(s):  
YiQin Xu ◽  
Yulia T. Peet
2019 ◽  
Vol 29 (05) ◽  
pp. 967-994 ◽  
Author(s):  
Alessio Castorrini ◽  
Alessandro Corsini ◽  
Franco Rispoli ◽  
Kenji Takizawa ◽  
Tayfun E. Tezduyar

Computational fluid–structure interaction (FSI) and flow analysis now have a significant role in design and performance evaluation of turbomachinery systems, such as wind turbines, fans, and turbochargers. With increasing scope and fidelity, computational analysis can help improve the design and performance. For example, it can help add a passive morphing attachment (MA) to the blades of an axial fan for the purpose of controlling the blade load and section stall. We present a stabilized Arbitrary Lagrangian–Eulerian (ALE) method for computational FSI analysis of passive morphing in turbomachinery. The main components of the method are the Streamline-Upwind/Petrov–Galerkin (SUPG) and Pressure-Stabilizing/Petrov–Galerkin (PSPG) stabilizations in the ALE framework, mesh moving with Jacobian-based stiffening, and block-iterative FSI coupling. The turbulent-flow nature of the analysis is handled with a Reynolds-Averaged Navier–Stokes (RANS) model and SUPG/PSPG stabilization, supplemented with the “DRDJ” stabilization. As the structure moves, the fluid mechanics mesh moves with the Jacobian-based stiffening method, which reduces the deformation of the smaller elements placed near the solid surfaces. The FSI coupling between the blocks of the fully-discretized equation system representing the fluid mechanics, structural mechanics, and mesh moving equations is handled with the block-iterative coupling method. We present two-dimensional (2D) and three-dimensional (3D) computational FSI studies for an MA added to an axial-fan blade. The results from the 2D study are used in determining the spanwise length of the MA in the 3D study.


2013 ◽  
Vol 767 ◽  
pp. 92-97
Author(s):  
Tei Saburi ◽  
Shiro Kubota ◽  
Yuji Wada ◽  
Masatake Yoshida

A multidimensional analysis code for reactive shocks (MARS), which is developed to solve various problems in the physical hazard analysis of high energetic materials, has been applied to such complex problems as multi-material problem and sympathetic problem because it can employ various types of equations of state and a materials database. However, it was difficult to meet a growing demand for large-scale analysis and fluid-structure interaction (FSI) analysis. To address these issues, this study reports a parallelization of the code and an implementation of the functional capability of FSI analysis, and performance results for sample problems were also shown.


2018 ◽  
Author(s):  
Marcel König

Eine sehr interessante und umfangreiche Arbeit aus der „Arbeitsgruppe Numerische Strukturanalyse mit Anwendungen in der Schiffstechnik“ aus dem Institut „M-10“ der Technischen Universität Hamburg – das lohnt sich!. Many engineering applications are governed by coupled multifield phenomena. In this thesis, a partitioned solution approach is followed to solve these kind of problems, which does not only enable the use of different discretization schemes for each of the subproblems but also allows to reuse specialized and efficient solvers, which enhances modularity, software reusability, and performance. A framework for the partitioned analysis of general multifield problems is proposed and implemented in the generic software library comana, which is verified against various benchmark problems and successfully applied to sophisticated fluid-structure interaction problems from the maritime industry. ...


2010 ◽  
Vol 65 (1-3) ◽  
pp. 271-285 ◽  
Author(s):  
Kenji Takizawa ◽  
Creighton Moorman ◽  
Samuel Wright ◽  
Timothy Spielman ◽  
Tayfun E. Tezduyar

Sign in / Sign up

Export Citation Format

Share Document