scholarly journals A noninvasive system‐level model order reduction scheme for flexible multibody simulation

2020 ◽  
Vol 121 (14) ◽  
pp. 3083-3107
Author(s):  
Frank Naets ◽  
Thijs Devos ◽  
Alexander Humer ◽  
Johannes Gerstmayr
Keyword(s):  
Model Order Reduction ◽  
Order Reduction ◽  
System Level ◽  
Reduction Scheme ◽  
Multibody Simulation ◽  
Model Order ◽  
Flexible Multibody ◽  
Level Model ◽  
System Level Model

Modeling of Capacitive Microaccelerometers with Squeeze Film Damping and Electrostatic Effects Using Model Order Reduction

2009 ◽  
Vol 60-61 ◽  
pp. 213-218
Author(s):  
Ya Fei Zhang ◽  
Wei Zheng Yuan ◽  
Hong Long Chang ◽  
Jing Hui Xu
Keyword(s):  
Model Order Reduction ◽  
Order Reduction ◽  
System Level ◽  
Squeeze Film ◽  
Electrostatic Effects ◽  
Model Order ◽  
Circuit Simulator ◽  
System Level Simulation ◽  
Squeeze Film Damping ◽  
Hardware Description

Model order reduction is an effective method to generate macromodels for system-level simulation. But it is difficult to deal with the electro-mechanical-damping coupling problems. So we presents a new approach to model the capacitive microaccelerometers with squeeze film damping and electrostatic effects using model order reduction (MOR) method. In this approach, the mechanical, squeeze film damping and electrostatic domains of the devices are modeled separately and then coupled at system-level. The macromodel for squeezed film damping effects could account for slip condition of the flow at low pressure and edge effects. In addition, some important parameters are preserved as symbol. The extracted macromodels are translated into the hardware description language and imported into a circuit simulator. An accelerometer is used to demonstrate the feasibility and efficiency of the proposed approach. Numerical simulation results show that the extracted macromodel can dramatically reduce the computation cost while capturing the device behavior accurately.

Model-Order Reduction of Flexible Multibody Dynamics Via Free-Interface Component Mode Synthesis Method

2020 ◽  
Vol 15 (10) ◽  
Author(s):  
Qinglong Tian ◽  
Peng Lan ◽  
Zuqing Yu
Keyword(s):  
Model Order Reduction ◽  
Multibody System ◽  
Order Reduction ◽  
Motion Equation ◽  
Component Mode Synthesis ◽  
Flexible Multibody System ◽  
Model Order ◽  
Free Interface ◽  
Equilibrium Configurations ◽  
Flexible Multibody

Abstract A new method of model-order reduction for the flexible multibody system which undergoes large deformation and rotation is proposed. At first, the flexible multibody system is modeled by absolute nodal coordinate formulation (ANCF), and then, the whole motion process of the system is divided into a series of quasi-static equilibrium configurations according to a given criterion. Afterward, motion equation is locally linearized based on the Taylor expansion. Therefore, the constant tangent stiffness matrix is obtained and does not need to be updated until the next configuration. Based on the locally linearized motion equation, the free-interface component mode synthesis (CMS) method is adopted to reduce the degrees-of-freedom (DOF) of the flexible multibody system molded by ANCF. The generalized-α integrator is used to solve the reduced motion equation. To verify the accuracy and efficiency of the proposed method, three examples including a free-falling pendulum, a flexible spinning beam and a deployable sail arrays are presented. Results show that the proposed method is able to reduce the computing time and maintain high accuracy.

Fast Multiphysics MEMS Simulation With Arnoldi Model Order Reduction

2008 ◽  
Author(s):  
David Binion ◽  
Xiaolin Chen
Keyword(s):  
Model Order Reduction ◽  
Krylov Subspace ◽  
Order Reduction ◽  
Full Scale ◽  
Solution Time ◽  
System Level ◽  
Computational Time ◽  
Mems Devices ◽  
Model Order ◽  
Reduced Order

Modeling and simulation of Micro Electro Mechanical Systems has become increasingly important as the complexity of MEMS devices increases. In particular, thermal effects on MEMS devices has become a growing topic of interest. Through the FEA, detailed solutions can be obtained to investigate the multiphysics coupling and the transient behavior of a MEMS device at the component level. For system-level integration and simulation, the FEA discretization often results in large full-scale models, which can be computationally demanding or even prohibitive to solve. Model order reduction (MOR) was investigated in this study to reduce problem size for complex dynamic system modeling. The Arnoldi method was implemented for MOR to improve the computational efficiency while preserving the input-output behavior of coupled MEMS simulation. Using this method, a low dimensional Krylov subspace was extracted from the full-scale system model. Reduced order solution of the transient temperature distributions was then determined by projecting the system onto the extracted Krylov subspace and solving the reduced system. An electro thermal MEMS actuator was studied for various inputs. To compare results, the full-scale analyses were performed using the commercial FEA program ANSYS. It was found that the computational time of MOR was only a fraction of the full-scale solution time, with the relative errors ranging from 1.1% to 4.5% at different positions on the actuator. Our results show that the reduced order modeling via Alnoldi can significantly decrease the transient analysis solution time without much loss in accuracy for coupled-field MEMS simulation.

Simulation in System-Level Based on Model Order Reduction for a Square-Wave Micromixer

2015 ◽  
Vol 16 (7-8) ◽  
pp. 307-314 ◽  
Author(s):  
Xueye Chen ◽  
Jienan Shen
Keyword(s):  
Model Order Reduction ◽  
Order Reduction ◽  
System Level ◽  
Order Model ◽  
Model Order ◽  
Reduced Order ◽  
Square Wave ◽  
Bonding Method ◽  
Proper Orthogonal ◽  
Sodium Solution

AbstractWith the aim to optimize design, a simulation in system level has been presented for the square-wave micromixer in this article. The square-wave micromixer is divided into straight channels and square-wave units. The reduced-order model based on proper orthogonal decomposition is applied in calculating concentration of the sample in the straight channels, and numerical simulation is applied in calculating concentration of the sample in the square-wave units. The data can mutually be transferred between straight channels and square-wave units by data fitting and interpolation. The maximal relative deviation is 1.52% between simulation in system-level and only simulation. The computational efficiency will be improved significantly with the numbers of straight channels increasing. The Polymethyl methacrylate (PMMA) micromixer is fabricated with mill and hot bonding method. The mixing experiment of fluorescein sodium solution with different concentrations is carried out to verify simulation. The relative deviations between simulation in and experimental results are below 8.26%.

Sign in / Sign up

Export Citation Format

Share Document