Computational study on microstructural optimization of multiferroic magnetoelectric composites

[Display omitted] •A multiscale optimization is proposed for magnetoelectric composite materials.•The proposed multiscale optimization successfully discovered an innovative microstructure.•The findings open new avenues for enhancing physical properties of functional materials. A multiscale optimization is presented for multiferroic composite materials to enhance magnetoelectric (ME) effect. The challenge of this study is to discover innovative microstructures beyond the conventional laminated structure which is the best of existing materials. The asymptotic homogenization theory was employed for scale bridging between the macrostructure and the microstructure. The homogenized ME coefficient of macrostructure was set to an objective function. The phase configuration and polarization directions in the microstructure were utilized as design variables. The computation yielded to an optimized microstructure, and found its macro homogenized ME coefficient is 21% larger than the conventional structure. The optimized microstructure consists of four rectangle regions and they are periodically repeated. Ferromagnetic and ferroelectric phases are alternately lined in each region and their polarization directions are determined by the special stated Euler angles. The computation successfully established that the optimal microstructure exists beyond the conventional layered structure. The findings open new avenues for enhancing physical properties of functional materials, and are expected to expand to digital fabrication by 3D printing.