Fig. 8: Illustration of the machine learning framework and training dataset in the case of a/c domains. | npj Computational Materials

Fig. 8: Illustration of the machine learning framework and training dataset in the case of a/c domains.

From: Machine learning surrogate for 3D phase-field modeling of ferroelectric tip-induced electrical switching

Fig. 8

a Example depicting the evolution of the a/c domain structure and electrostatic potential throughout a training trajectory. Domain states are depicted at time t0, t20, t41, and the final time t62, highlighting various voltage applications during tip scanning trajectory. b Illustration demonstrating the surrogate model operations over the \({{\mathcal{P}}}_{x}\), \({{\mathcal{P}}}_{y}\), and \({{\mathcal{P}}}_{z}\) polarization components, along with the \({\mathcal{V}}\) electrostatic potential, specifically in the scenario of a/c ferroelectric domain switching. Distribution in the training dataset of c and d tip locations (ytip−ztip), e prescribed voltages of the AFM tip (uT), and f tip application times (tapp).

Back to article page