On the compatibility of non-holonomic systems and certain related variational systems

TL;DR

This paper investigates the relationship between equations of motion derived from d'Alembert's principle and those from a specific variational principle in non-holonomic systems, showing they are generally incompatible in higher dimensions.

Contribution

It demonstrates the fundamental incompatibility between d'Alembertian and variational equations for non-holonomic constraints in spaces of dimension three or more.

Findings

Equations of motion from d'Alembert's principle and variational methods do not coincide for non-holonomic systems in N ≥ 3.

The incompatibility arises specifically when the non-holonomic constraint is genuinely non-holonomic.

This result clarifies the limitations of variational approaches in modeling certain constrained mechanical systems.

Abstract

I consider the equations of motion which follow from d'Alembert's principle for a general mechanical system in a space of N dimensions, constrained by a non-holonomic constraint which is linear and homogeneous in the generalised velocities. The variational equations of motion which follow for the same system by assuming the validity of a specific variational action principle, in which the non-holonomic constraint is implemented by means of the multiplication rule in the calculus of variations are also considered. It is shown that these two types of equations of motion are not compatible in a space of dimension N greater than or equal to 3, if the constraint is genuinely non-holonomic. This means that these two types of equations of motion do not have coinciding general solutions.