Nonholonomic Newmark method

Alexandre Anahory Simões,Sebastián J. Ferraro,Juan Carlos Marrero,David Martín de Diego

DOI: https://doi.org/10.48550/arXiv.2201.13162

2022-01-31

Abstract:Using the nonholonomic exponential map, we generalize the well-known family of Newmark methods for nonholonomic systems. We give numerical examples including a test problem where the structure of reversible integrability responsible for good energy behaviour as described in [16] is lost. We observe that the composition of two Newmark methods is able to produce good energy behaviour on this test problem.

Numerical Analysis,Mathematical Physics,Differential Geometry

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to generalize the classical Newmark method to nonholonomic systems. Specifically, the authors aim to develop a new numerical integration method - the nonholonomic Newmark method - for simulating and solving the equations of motion of mechanical systems subject to nonholonomic constraints. Such systems have wide applications in engineering and robotics, for example, in wheeled vehicles and satellite dynamics.

### Core problems of the paper

1. **Extension of the Newmark method**:
- The traditional Newmark method is widely used for numerical integration in nonlinear structural dynamics, but has not been extended to nonholonomic systems. Therefore, the goal of this paper is to construct a Newmark method applicable to nonholonomic systems.

2. **Maintenance of energy behavior**:
- Some numerical methods can produce good energy behavior because they preserve reversible integrability. However, in some nonholonomic systems, this property may be lost. By combining two Newmark methods, the authors observed that good energy behavior can be restored.

3. **Handling of linear and nonlinear constraints**:
- This paper mainly discusses the case of linear velocity constraints, but also mentions that its method can be naturally extended to nonlinear constraints, explicitly time - dependent systems, and nonholonomic systems with external forces.

### Main contributions

- **Introduction of nonholonomic exponential mapping**: Use the nonholonomic exponential mapping to define the nonholonomic Newmark method.
- **Construction of high - order methods**: Construct higher - order numerical integrators by combining low - order Newmark methods.
- **Numerical verification**: Verify the effectiveness of the proposed method through several specific nonholonomic system examples (such as nonholonomic particles, chaotic nonholonomic particles, CVT systems driven by pendulums), and compare it with classical methods (such as the Runge - Kutta method).

### Conclusion

Through these efforts, the authors have not only successfully generalized the Newmark method to nonholonomic systems, but also demonstrated the potential of this method in maintaining energy behavior, especially in those systems where reversible integrability is usually lost. This provides new tools and ideas for efficient numerical simulation of nonholonomic systems.