Bubble-wall oscillation, in response to a step change in bias field, is reported in four bubble garnet samples with different compositions and characteristics. The wall velocity during oscillatory motion is found to be independent of the instantaneous drive on the domain wall, but does depend on the initial drive on the wall. The experimental relationship between the wall velocity vs and the initial drive Ha is given by vs=0.86 m/sec+(2.4 m/sec Oe) Ha for expanding motion and by vs =1.6+2.6Ha for collapsing motion. The half-period of oscillation, typically between 100 and 300 nsec, increases linearly with initial drive for low drive fields (∼1 Oe) and approaches a constant value at higher fields (≳2 Oe). Wall motion is analyzed in terms of Slonczewski’s formulation of wall mechanics and a velocity-momentum relationship of the form ?=a+b? is found to agree with the observed motion for a=3.2 m/sec and b=1.5 m/sec. Underdamped motion is observed in two ranges of initial drive field separated by a region in which wall motion is overdamped. The lower boundary between underdamped and overdamped motion, near a drive field of 2.3 Oe, is sharply defined, while the upper boundary, near 3.5 Oe, represents a more gradual transition. The transition to overdamped motion is accompanied by a sharp drop in effective domain-wall mobility, and the mobility returns to its original value in the second underdamped region. A comparison between theoretical and experimental parameters indicates that wall oscillation is most likely to be observed in materials with saturation velocities [v0=7.1γA/ h (K)1/2] greater than 10 m/sec and with resonance-damping parameters less than 0.04.

No abstract

No abstract

No abstract

Large, thin-film single domain areas have been observed, in the absence of a bias field, in garnets with magnetization perpendicular to the film plane.1,2 The domain stability in the work by Krumme1 was attributed to a combination of low saturation magnetization and a low Curie temperature. Uchishiba2 relates the stability in his double layer system to appropriate anisotropy fields in one layer compared to the magnetization in the other layer. A more complete model for large domain stability in a bias field free environment is given in this work. Three distinct stability regimes are predicted by the model and all have been observed experimentally. Areas 3.5-cm in diameter have been made into stable single domains. This was achieved in a material showing a zero bias strip width of 4.5 μm. The single domain diameter was, therefore, 7500 times the equilibrium energy domain width. The technique developed and the model have led to a new means for observing magnetic defects. More importantly, it also offers a means for measuring the strength of the defects. Possible applications of the model are also discussed.