next up previous
Next: Nonholonomic Mapping to Spaces Up: Classical Motion of a Previous: Classical Motion of a

Equations of Motion

Consider the action of the particle along the orbit tex2html_wrap_inline4024 in a flat space parametrized with D rectilinear, Cartesian coordinates:

  equation250

It may be transformed to curvilinear coordinates tex2html_wrap_inline4028 , via some functions

  equation259

leading to

  equation262

where

  equation273

is the induced metric for the curvilinear coordinates. Repeated indices are understood to be summed over, as usual. For Cartesian coordinates, upper and lower indices i are the same. The indices tex2html_wrap_inline4032 of the curvilinear coordinates, on the other hand, are lowered by contraction with the metric tex2html_wrap_inline4034 or raised with the inverse metric tex2html_wrap_inline4036 .

The length of the orbit in the flat space is given by

  equation282

Both the action (3) and the length (5) are invariant under arbitrary reparametrizations of space tex2html_wrap_inline4038 .

Einstein's equivalence principle   amounts to the postulate that the transformed action (3) describes directly the motion of the particle in the presence of a gravitational field caused by other masses. The forces caused by the field are all a result of the geometric properties of the metric tensor.

The equations of motion are obtained by extremizing the action in Eq. (3) with the result

  equation298

Here

  equation313

is the Riemann connection or Christoffel symbol of the first kind. Defining also the Christoffel symbol of the second kind

  equation327

we can write

  equation337

The solutions of these equations are the classical orbits. They coincide with the extrema of the length l of a curve in (5). Thus, in a curved space, classical orbits are the shortest curves, called geodesics. The reason for the name shortest lines is that they minimize the invariant length (5) of all lines connecting two given points tex2html_wrap_inline4042 and tex2html_wrap_inline4044 .

The same equations can also be obtained directly by transforming the equation of motion from

  equation352

to curvilinear coordinates tex2html_wrap_inline4046 , which gives

  equation356

At this place it is useful to employ the so-called basis triads

  equation369

and the reciprocal basis triads

  equation376

which satisfy the orthogonality and completeness relations

  eqnarray382

The induced metric can then be written as

  equation393

Using the basis triads, Eq. (11) can be rewritten as

  equation400

or as

  equation413

The subscript tex2html_wrap_inline4048 separated by a comma denotes the partial derivative tex2html_wrap_inline4050 , i.e., tex2html_wrap_inline4052 . The quantity in front of tex2html_wrap_inline4054 is called the affine connection:

  equation426

Due to (14), it can also be written as

  equation435

Thus we arrive at the transformed flat-space equation of motion

  equation442

The solutions of this equation are called the straightest lines or autoparallels.

If the coordinate transformation tex2html_wrap_inline4056 is smooth and single-valued, it is integrable. This property is expressed by Schwarz's integrability condition, according to which derivatives in front of such a function tex2html_wrap_inline4056 commute:

  equation452

Then the triads satisfy the identity

  equation455

implying that the connection tex2html_wrap_inline4060 is symmetric in the lower indices. In this case it coincides with the Riemann connection, the Christoffel symbol tex2html_wrap_inline4062 . This follows immediately after inserting tex2html_wrap_inline4064 into (7) and working out all derivatives using (23). Thus, for a space with curvilinear coordinates tex2html_wrap_inline4046 which can be reached by an integrable coordinate transformation from a flat space, the autoparallels coincide with the geodesics.


next up previous
Next: Nonholonomic Mapping to Spaces Up: Classical Motion of a Previous: Classical Motion of a

Hagen Kleinert
Mon Feb 17 09:33:29 MET 1997