Forum Springer Astron. Astrophys.
Forum Whats New Search Orders

Astron. Astrophys. 329, 895-905 (1998)

Previous Section Next Section Title Page Table of Contents

7. Discussion and conclusions

An inherent problem with 2D dynamical simulations of the type described in Sect. 3 is that the dissipation is rather too large, and this may result in overestimation of the radial velocities. This appears to be one, at least, of the reasons why we found it necessary to reduce the value of [FORMULA] from its canonical value. Nevertheless, we are encouraged by the general robustness of our results. The main features of the magnetic field structure displayed in Fig. 2 persist when both the dynamo parameters and the velocity data reduction method are altered. Moreover, the same general (but not particular) features appear in the magnetic field structure when we use velocities from a dynamically distinct simulation (Sect. 5.3). Moreover our 2D dynamo model appears to capture the essential features of a more general, 3D, calculation, at least with the 2D velocity data. This has enabled an exploration of parameter space at reasonable computational cost. Nevertheless, the 3D code will certainly be needed when we use data from a 3D dynamical simulation.

We have chosen the simplest possible representation of the [FORMULA] -effect - that [FORMULA] is a scalar that does not vary with r and [FORMULA]. More plausibly, [FORMULA], or even is proportional to the local vorticity (e.g. Brandenburg and Donner 1997). Test calculations with [FORMULA] proportional to the vorticity show that the overall changes introduced in the field structure are small. These and other experiments support the idea that in our simulations the magnetic field structure is almost completely determined by the recent history of the gas velocity field - see Fig. 1. Field lines closely follow streamlines. In this, our results are consistent with those of Otmianowska-Mazur & Chiba (1995). An [FORMULA] -effect is essential to prevent overall field decay, but relatively small field variations caused by changes in the form of [FORMULA] are masked by advection by the strong noncircular velocities.

Whilst our model does not purport to represent any `real' spiral galaxy, we believe that it possesses some generic features, and that the general features of the magnetic field structure - appearance and disappearance of rings and short arms, a mixture of [FORMULA] and [FORMULA] nonaxisymmetric structure, etc - may be widely valid. This view is supported by the overall similarity between the dynamo calculations described in Sects. 5.1and 5.3. An interesting and important step would be to model a barred galaxy for which both good quality velocity and magnetic data are available.

Previous Section Next Section Title Page Table of Contents

© European Southern Observatory (ESO) 1998

Online publication: December 16, 1997