Implicit Numerical Integration for Euler-Lagrange Equations via Tangent Space Parametrization∗
Florian A. Potra, Jeng Yen
Abstract
Florian A. Potra, Jeng Yen
Abstract
A new class of methods for solving the equations of motion of constrained mechanical system dynamics is presented. The tangent space local parametrization is used to form an index one system of mixed differential-algebraic equations (DAEs) that describes the constrained motion. Implicit numerical integration formulas are applied to the reduced ordinary differential equations (ODEs) in the tangent space. The resulting system of nonlinear equations is solved by Broyden's method. In the framework presented, the computational complexity of solving the implicit integration equations for the reduced ODE is about same as that of some explicit integration methods for the reduced ODE. In fact, in some cases the implicit method for solving the Euler-Lagrange equations can be implemented more efficiently than the explicit scheme.
OpenAlex reports 42 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
A new class of methods for solving the equations of motion of constrained mechanical system dynamics is presented. The tangent space local parametrization is used to form an index one system of mixed differential-algebraic equations (DAEs) that describes the constrained motion. Implicit numerical integration formulas are applied to the reduced ordinary differential equations (ODEs) in the tangent space. The resulting system of nonlinear equations is solved by Broyden's method. In the framework presented, the computational complexity of solving the implicit integration equations for the reduced ODE is about same as that of some explicit integration methods for the reduced ODE. In fact, in some cases the implicit method for solving the Euler-Lagrange equations can be implemented more efficiently than the explicit scheme.
Key concepts: Mathematics, Ode, Differential algebraic equation, Ordinary differential equation, Nonlinear system, Euler equations, Equations of motion, Numerical integration