 |  |
Astron. Astrophys. 335, 134-144 (1998)
Appendix A: modelling of the particle acceleration process
We consider the particle diffusion in the jet-cocoon region in the
case of the mean magnetic field being parallel to the jet velocity
both inside and outside the jet. Modelling of the diffusive particle
trajectories is based on the small-amplitude pitch-angle scattering
approach (Ostrowski 1991). It assumes that the particle diffusion
coefficients parallel to the mean field, , and
the one perpendicular to the mean field, , are
proportional to the particle momentum. Due to this fact, the amount of
computations required for reproducing particle diffusive trajectories
for low energy particles is much larger than the respective amount for
particles with higher energies. Therefore, in order to deal with these
low energy particles, we introduced a hybrid approach to derive
particle trajectories. For particles near the surfaces of the flow
velocity change at the jet boundary and at the shock, and not further
than 2 particle gyroradii away in the former case and 1 diffusive
scale ( , 2) from the
shock surface, we use the exact form of the mentioned pitch-angle
scattering method. When the particle diffuses further away from the
respective discontinuity, we use the spatial diffusion model involving
large time steps. However, the step length is always limited in order
to prevent particles from crossing any discontinuity in an individual
`diffusive' step - the mean diffusive step allows only for moving a
particle 0.25 of the distance to its closest boundary. In these
computations we use the values of and
derived in independent simulations involving
the `exact' pitch-angle diffusion procedure. This way we are able to
consider particles with momenta which differ in a few orders of
magnitude. During the simulations we use a variant of the trajectory
splitting procedure described by Ostrowski (1991). A good test for
this simulation procedure is provided by the agreement of our shock
accelerated spectra fitted at low energies with the values derived
analytically by Heavens & Drury (1988).
Appendix B: on the spectrum of particles accelerated at the tangential discontinuity
Let us consider a simple model for particles accelerated at the
plane tangential discontinuity surrounded with infinite regions for
particle diffusion. We derive the stationary spectrum at the
discontinuity for the perpendicular (to discontinuity) diffusion
coefficient being proportional to particle energy,
. For relativistic particles with
and the same velocity ,
the mean time between the successive particle interactions with the
discontinuity is also proportional to the particle's energy. Thus the
mean rate of particle energy gain is constant, independent of energy,
= const. The transport equation for the
phase-space distribution function has
the following form,
![[EQUATION]](img172.gif)
In the above equation one assumes a continuous particle injection
at . For a general power-law form for
it yields the solution,
![[EQUATION]](img175.gif)
where is the Heaviside step function. As
long as there are no specific energy scales introduced into the
acceleration process the obtained form for is
independent of the velocity difference at the discontinuity (particle
anisotropy), and of possible correlation between the interaction time
and the energy gain. In the case considered in the present paper
and .
For the discontinuity formed at the jet boundary, the jet radius
and the escape boundary radius provide energy scales to the process.
As a result, a cut-off occurs at large energies in the spectrum.
However, at small energies, where the jet boundary curvature is
insignificant and the diffusive regions very extended, the solution
should be close to the one given in (B2). In fact, numerical fits for
low energy sections of cases (c) and (d) at Fig. 6 give the
respective fits and
(the difference in the obtained spectral indices is not significant,
depending on the momentum range used for fitting).
© European Southern Observatory (ESO) 1998
Online publication: June 12, 1998
helpdesk.link@springer.de  |