To better understand the magnetic structure and eruptive process of solar filaments, a solar active-region filament (labeled F2) eruption associated with a B-class flare was investigated by using high-resolution Hα data from the 1 m New Vacuum Solar Telescope (NVST), combined with EUV observations of the Solar Dynamics Observatory (SDO). The filament F2 was disturbed by another filament (labeled F1) eruption that experienced a whip-like motion. Before the filament F2 eruption, the Dopplergrams show that the southern and the northern parts of the filament F2 body exhibit blueshift and redshift along the filament spine, simultaneously. This implies that the filament F2 was rolling from one side to the other. During the filament F2 eruption, the Doppler shifts of the filament body are opposite of those before its eruption. This demonstrates that the filament body exhibits an untwisting motion, which can also be identified by tracing the movement of the eruptive filament threads. Moreover, it is found that the material of the filament F2 was transferred to the surrounding magnetic field loops, which is caused by magnetic reconnection between the filament F2 and the surrounding magnetic loops. According to the right-bearing threads of the filament F2 before its eruption, it can be deduced that the filament F2 is initially supported by a sheared arcade. The following observations reveal that the twisted magnetic structure of filament F2 formed in the eruption phase.
Using the Hα data from the New Vacuum Solar Telescope at the Fuxian Solar Observatory together with multiwavelength images and magnetograms obtained by the Solar Dynamics Observatory, we study the detailed process of three homologous confined flares in active region NOAA 11861 on 2013 October 12. All of the three flares occurred at the same location, with similar morphologies and comparable classes. Through analyzing the evolution of magnetic field and flow field, we found an emergence of magnetic flux and a strong shearing motion between two opposite polarities near the following sunspot. The magnetic flux and the average transverse field strength exhibited a decrease before each eruption and reached the lowest point at the onset of each eruption. By calculating the shearing and the emergence energy in the photosphere, we found that the integral of energy injected from the photosphere, for a few hours, could provide enough energy for the flares. The reconnection between different loops was observed in Hα images during the occurrence of each flare. These results suggest that the emerging magnetic flux and the shearing motion in the photosphere can inject the energy to the sheared magnetic loops and the energy was finally released via magnetic reconnection to power the solar flares.