All Figures

$\begin{figure} \par\includegraphics[width=7.8cm,clip]{7653fig1.eps} \end{figure}$ Figure 1: Limits of the instability domain in the (A^*, S) plane, for various values of $A_t/A_\mu$ . The stratification parameters are defined as $A_t=(l/n)^2 (N_t/\omega _A)^2$ and $A_\mu =(l/n)^2 (N_\mu /\omega _A)^2$ , whereas $A^*=\varepsilon A_t+ A_\mu$ . $\varepsilon =\eta /\kappa$ designates the Roberts number, and $S=2 \eta \Omega n^2/\omega _A^2$ defines the limits of the instability domain in vertical wavenumber n, for given A^*.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8cm,clip]{7653fig2.eps} \end{figure}$ Figure 2: How to close the dynamo loop involving the Pitts &Tayler instability. In dashed lines the loops proposed by Spruit (A) and Braithwaite (B). The only possible way to regenerate the mean toroidal and/or poloidal field is through the mean electromotive force $\langle\vec v \times \vec b\rangle$ produced by the non-axisymmetric instability-generated field. But the dynamo must regenerate the poloidal field, in order to be fed from the differential rotation, and this leaves only the loop drawn in solid line.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{7653fig3.ps} \end{figure}$ Figure 3: Time evolution of the energies of the mean poloidal (PME), mean toroidal (TME) and non-axisymmetric (FME) components of the magnetic field. Cases A and B refer respectively to higher and lower magnetic diffusivity (cf. Table 3). Note the steady decline of the poloidal field, which is not affected by the irruption of the m=1 Pitts & Tayler instability (at $t\approx 8000$ days in case A and $\approx$ 20 000 days in case B).
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8cm,clip]{7653fig4.eps} \end{figure}$ Figure 4: Azimuthal component of the instability-generated magnetic field, at the peak of the instability ( $\approx$ 10 000 days, case A), measured at $r/R_\odot = 0.7, ~ \theta = 75\hbox{$^\circ$ }$ . Note that this non-axisymmetric field is dominated by the wavenumber m=1, which is the signature of the Pitts & Tayler instability.
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{7653fig5CMJN.eps} \end{figure}$ Figure 5: Mean toroidal field $B_\varphi (r, \theta )$ (left) and meridional section of the fluctuating toroidal field $b_\varphi (r,\theta , \varphi )$ (right). Only the odd m have been kept in the latter, and the azimuthal angle $\varphi$ has been chosen such as to emphasize the m=1 component. Note that mean and fluctuating fields are of comparable strength. (Case B, t=50 000 days.)
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=7cm,clip]{7653fig6.ps} \end{figure}$ Figure 6: Evolution of the magnetic energies after suppressing the mean poloidal field at t=60 000 days (case B). The mean electromotive force due to the instability-generated field produces some amount of poloidal energy (PME, dashed line), but that field is too weak to prevent through $\Omega$ -efect the Ohmic dissipation of the toroidal field $B_\varphi$ (TME, dash-dotted line), which causes the decline of the instability-generated field (FME, solid line). There is no dynamo action, in spite of the high magnetic Reynolds number ${\it Rm}=10^5$ .
Open with DEXTER

In the text

$\begin{figure} \par\includegraphics[width=8cm,clip]{7653fig4.eps} \end{figure}$	Figure 4: Azimuthal component of the instability-generated magnetic field, at the peak of the instability ( $\approx$ 10 000 days, case A), measured at $r/R_\odot = 0.7, ~ \theta = 75\hbox{$^\circ$ }$ . Note that this non-axisymmetric field is dominated by the wavenumber m=1, which is the signature of the Pitts & Tayler instability.
Open with DEXTER

$\begin{figure} \par\includegraphics[width=7.4cm,clip]{7653fig5CMJN.eps} \end{figure}$	Figure 5: Mean toroidal field $B_\varphi (r, \theta )$ (left) and meridional section of the fluctuating toroidal field $b_\varphi (r,\theta , \varphi )$ (right). Only the odd m have been kept in the latter, and the azimuthal angle $\varphi$ has been chosen such as to emphasize the m=1 component. Note that mean and fluctuating fields are of comparable strength. (Case B, t=50 000 days.)
Open with DEXTER