48
E. Wilson and B. J. Holzer
find its solution numerically but to see an analytic solution for small amplitudes we
set φ s = 0 and φ ≈ sinφ :
¨
φ +
2πeV 0 hηf 2
E 0 β 2 γ
φ = 0.
(2.74)
The frequency of these synchrotron oscillations in longitudinal phase space is
f s =
|η| heV 0
2πE 0 β 2 γ
f,
(2.75)
or writing f rf = hf we could equally express
f s =
|η| eV 0
2πE 0 β 2 γ h
f rf .
(2.76)
In analogy to the transverse plane, we define a synchrotron tune, Q s , as the
number of such oscillations per revolution of the machine. This is analogous to
Q in transverse phase space.
Q s =
f s
f
=
|η| ehV 0 cos φ s
2πE 0 β 2 γ
.
(2.77)
Usually Q s is less than 10% of the revolution frequency. It drops down to zero at
γ transition where η is zero and then rises again. In large proton machines it can be
in the region 0 to 100 Hz and, but for the vacuum, one might hear it!
Close to γ tr we cannot strictly assume that β, γ , η, and f vary slowly in
comparison with the synchrotron oscillation which this equation describes. Hence
we should use a more exact form of the equation of motion and approximate only
when it seems that this is justified:
d
dt
E 0 β 2 γ ˙
φ
2πηhf 2
+ eV 0 (sin φ − sin φ s ) = 0.
(2.78)
In a stationary bucket, when φ s = 0, this exact differential equation for large
amplitude motion is identical to that for a rigid pendulum:
d 2 θ
dt 2 + g sin θ = 0.
(2.79)
There is an extra term, sinφ s , on the right hand side of the synchrotron equation
which is not there in the pendulum case but it could be introduced too for the
pendulum by using a magnetic ‘bob’ and biasing its equilibrium position to one
side by attaching a weight on a cantilever at right angles to the rod of the pendulum.
Précédent

- 58/867

Suivant