I"m looking at "Principles of Dynamics: Second Edition" by Donald T Greenwood. I'm trying to figure out how he obtains Eq. (6-64)

$$\frac{\partial\dot x_j}{\partial\dot q_i} = \frac{\partial x_j}{\partial q_i}\tag{6-64}$$

from Eq (6-54)

$$\dot x_j = \sum_{i=1}^n \frac{\partial x_j}{\partial q_i}\dot q_i + \frac{\partial x_j}{\partial t}.\tag{6-54}$$

where the transformation equations from a set of $3N$ Cartesian coordinates to a set of $n$ generalized coordinates are of the form given by Eq (6-1)

$$x_1=f_1(q_1,q_2,...,q_n,t)$$ $$x_2=f_2(q_1,q_2,...,q_n,t)$$ $$\vdots$$ $$x_{3N}=f_{3N}(q_1,q_2,...,q_n,t).\tag{6-1}$$

When I differentiate Eq (6-54) with respect to $q_i$ I get second derivatives and I have no idea how the term $\frac{\partial^2 x_j}{\partial t\partial \dot q_i}$ is dealt with. Any insight appreciated.


Greenwood is asking how does $\dot x_j$ change if we vary $\dot q_i$ at some specified time $t$ and point $q_i$ (or equivalantly a specified point $x_i$ as there is a - perhaps time dependent - 1-1 relation between the $q$'s and $x$'s). At that specified point and time all the quantities on the RHS of your second equation, with the exception of the $\dot q$'s are to be treated as constants. The answer then, is exactly your first equation.

| cite | improve this answer | |
  • $\begingroup$ Ahhhhh, I think I see now. Just to clarify, $\frac{\partial\dot q_i}{\partial \dot q_j}=0 $, because $q_i$ is independent of $q_j$ for $i \ne j$? $\endgroup$ – eball Jun 14 '19 at 17:14
  • $\begingroup$ Yes. You can vary the $\dot \q_i$'s independently of one another. $\endgroup$ – mike stone Jun 14 '19 at 21:40
  • $\begingroup$ Makes sense. Thanks again!! $\endgroup$ – eball Jun 15 '19 at 19:51

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.