I have been teaching myself magnetostatics for some time with some help from texts by Griffiths, Jackson, and other online sources. My approach is a little unorthodox in that I try to ignore electrostatics and free currents to the extent possible. Maybe not the best approach but one way or another, I think I have gained a pretty good understanding of the topic. However, some pockets of confusion remain.
One concept I still struggle with is knowing when $\nabla \times \boldsymbol{B} = \bf 0$ is applicable. I know it is true for a constant field and for a dipole field. But is there a general class of problems for which it is always true?
Another related question: I know that if there are no free currents, then $\nabla \times \boldsymbol{H} = \bf 0$ and we can write a Laplace equation for scalar potential $\Phi$. But is that also true for the analogous case of the vector potential $\bf A$ where $\nabla \times \boldsymbol{B} = \bf 0$? Or do we have to account for bound currents in the case of nonzero magnetization $\bf M$ ? The term "current" is not always defined clearly.
My two main applications of interest are 1) a permeable sphere or ellipse in a constant field, and 2) a permanently magnetized sphere or ellipse.
If these topics are summarized succinctly anywhere in a reference, I would appreciate knowing about it.
Thanks in advance.