A Parallel Algorithm for Stiffness Matrix Decomposition Using Threadpool Method

2012 ◽  
Vol 594-597 ◽  
pp. 2880-2885
Author(s):  
Jun Tao Chen ◽  
Ming Xiao ◽  
Hui Bo Liu
Keyword(s):  
Finite Element ◽  
Parallel Algorithm ◽  
Stiffness Matrix ◽  
Decomposition Method ◽  
High Efficiency ◽  
Matrix Decomposition ◽  
One Dimensional ◽  
Numerical Tests ◽  
Element Simulation ◽  
Nested Loops

To shorten calculation time in finite element simulation by using multithreading computer, a parallel algorithm for stiffness matrix decomposition based on threadpool method is proposed. Firstly, a decomposition method of applicability to parallel computation is put forward by transferring the Cholesky's LLT method. Then, the threadplool is employed to generate multithreading for repeating use and the optimization is conducted considering load-balancing of each thread. Finally, numerical tests by using proposed algorithm in decomposition of one-dimensional array stored stiffness matrix are carried out on different calculation platforms with multi-processors. It is shown that the parallel algorithm can overcome the limitations of OpenMP when being applied in nested loops and is of high efficiency on stiffness matrix decomposition with low platform demands. The algorithm has explicit concept and minor programming difficulty and is applicable to solve problems caused by limitation of OpenMP in particular.

Analysis of Stochastic Space Frame with Elementary Stiffness Matrix Decomposition Method

2010 ◽  
Author(s):  
G. K. Er ◽  
S. W. Lan ◽  
V. P. Iu ◽  
Jane W. Z. Lu ◽  
Andrew Y. T. Leung ◽  
...  
Keyword(s):  
Stiffness Matrix ◽  
Decomposition Method ◽  
Matrix Decomposition ◽  
Space Frame ◽  
Stochastic Space

Finite Element Analysis of Grinding Temperature Field for NMQL in Nickel-Base Alloy Grinding

2020 ◽  
pp. 102-131
Keyword(s):  
Finite Element ◽  
Temperature Field ◽  
High Efficiency ◽  
Base Alloy ◽  
Grinding Temperature ◽  
Element Analysis ◽  
Element Simulation ◽  
Step Algorithm ◽  
Nickel Base ◽  
Grinding Temperature Field

Temperature is not only an important parameter in machining, but also an important basis for process optimization. Accurate prediction and reasonable analysis of grinding temperature is of great and far-reaching significance to the development and promotion of nanofluid micro-lubrication. In this chapter, the mathematical model of finite element simulation of temperature field of high efficiency deep grinding under four kinds of cooling lubrication conditions is established, and the three boundary conditions and the constraints of simulation model are established, and the mesh division and time step algorithm are determined respectively. Using ABAQUS simulation platform and theoretical model to simulate grinding temperature field, the distribution characteristics of grinding temperature field under different working conditions are analyzed from different directions, different grinding depths, and different workpiece materials.

Stochastic Plane Stress Analysis with Elementary Stiffness Matrix Decomposition Method

2010 ◽  
Author(s):  
G. K. Er ◽  
M. C. Wang ◽  
V. P. Iu ◽  
K. P. Kou ◽  
Jane W. Z. Lu ◽  
...  
Keyword(s):  
Stress Analysis ◽  
Plane Stress ◽  
Stiffness Matrix ◽  
Decomposition Method ◽  
Matrix Decomposition ◽  
Plane Stress Analysis

Improved Rotative Mid-Point Mensuration and Newton-Raphson Method in 2-D Flat Rolling Simulation

2011 ◽  
Vol 328-330 ◽  
pp. 1436-1439
Author(s):  
Shu Ni Song ◽  
Jing Yi Liu
Keyword(s):  
Finite Element ◽  
Simultaneous Equations ◽  
Rigid Plastic ◽  
Cpu Time ◽  
Numerical Tests ◽  
Element Simulation ◽  
Flat Rolling ◽  
Newton Raphson ◽  
Time Required ◽  
Raphson Method

Newton-Raphson (N-R) method has been employed to solve the system of simultaneous equations arising in Rigid-Plastic finite element simulation. The combination of the improved rotative mid-point mensuration and the N-R method, named the M-P method is designated to solve the equations of velocity increment in Rigid-Plastic FEM. The CPU times required for calculation by the M-P method and the N-R method are compared and it is found that the CPU time required for calculation of the N-R method is more than the M-P method. The calculated rolling forces by the M-P method and the N-R method are compared and it is found that the former correlates better with the measured value. Numerical tests and application show that the M-P method is feasible and steady.

A Finite-Element Simulation of Pulsatile Flow in Flexible Obstructed Tubes

1982 ◽  
Vol 104 (2) ◽  
pp. 119-124 ◽  
Author(s):  
E. Rooz ◽  
D. F. Young ◽  
T. R. Rogge
Keyword(s):  
Finite Element ◽  
Pulsatile Flow ◽  
Continuity Equation ◽  
Element Model ◽  
Momentum Equation ◽  
Cross Sectional ◽  
One Dimensional ◽  
Element Simulation ◽  
The One

A finite-element model for pulsatile flow in a straight flexible partially obstructed tube is developed. In the unobstructed sections of the tube the model considers the continuity equation, the one-dimensional momentum equation, and an equation of state relating tube cross-sectional area to pressure. For the obstructed region, a nonlinear relationship between the flow and the pressure drop across the stenosis is considered. The applicability of a model is checked by comparing predicted flow and pressure waveforms with corresponding in-vitro experimental measurements obtained on a mechanical system. These comparisons indicate that the model satisfactorily predicts pressures and flows under variety of frequencies of oscillation and stenosis severities.

FINITE ELEMENT SIMULATION OF ELECTRON OPTICAL SYSTEMS

1992 ◽  
Vol 02 (03) ◽  
pp. 339-356
Author(s):  
L. AVALDI ◽  
A. QUARTERONI ◽  
F. SALERI
Keyword(s):  
Finite Element ◽  
Finite Elements ◽  
Neumann Boundary ◽  
Optical Systems ◽  
Convergence Properties ◽  
Beam Shape ◽  
Electron Trajectories ◽  
Numerical Tests ◽  
Element Simulation ◽  
Domain Decomposition Algorithm

Electron optical systems are of interest in several fields of instrumentation and technology. Here we consider a device and approximate beam shape and electron trajectories throughout the whole system. From the mathematical viewpoint, the problem is modeled by Laplace equation with mixed Dirichlet-Neumann boundary conditions. This problem is approximated by finite elements, and solved by a domain decomposition algorithm. We investigate the theoretical convergence properties of this method, and we carry out several numerical tests on significant examples.

Sign in / Sign up

Export Citation Format

Share Document