Consider a mechanics problem with Lagrangian  (kinetic energy minus potential energy). When the solution is required to honor a holonomic constraint F(x) =0, it is often taught that the following modified Lagrangian:

 

 (where  is the celebrated Lagrange Multiplier) is the Lagrangian to use in the action integral. It is often suggested that the same procedure is valid even when the constraint is non-holonomic. I shall show using simple examples that the above is not correct.

 

报告人简介: Professor Sau-Hai Lam (林寿海)，普林斯顿大学工程与应用科学学院（SEAS）荣休教授，美国工程院院士, 现为我校热能工程系海外讲席教授组成员。研究兴趣包括：fluid mechanics, plasmas, chemical kinetics, Lagrangian dynamics, nonlinear control theory, and singular perturbation methodologies. 林教授是化学动力学中计算奇异摄动（CSP）方法的创始人。

 

个人主页：http://www.princeton.edu/~lam/