Certified Rankings of Laminate Magnetoelectric Composites under Declared Parameter Uncertainty: Materials, Geometries, Modes and Electrodes

Abstract Reported magnetoelectric voltage coefficients of laminate composites span more than three orders of magnitude, and recomputing them on one engine makes them comparable without saying which comparisons are settled. Each figure of merit is therefore enclosed by two-sided bounds over a declared parameter box, and each comparative claim is decided by a bound on the RATIO of two configurations with the nuisance parameters they share held tied. The piezomagnetic coefficient and the interfacial coupling enter through the numerator only, so they cancel identically at fixed magnetic phase and structural claims survive an order-of-magnitude ignorance about the material. Five magnetostrictive phases (Terfenol-D, Galfenol, CoFe2O4, Ni, Metglas) and three piezoelectrics (PZT, BaTiO3 and LiNbO3 in four cuts) are compared across disk, bar and ring geometries and two electrode topologies. Fifteen of thirty-three headline claims are certified, two pairs certified indistinguishable, fourteen undecidable and two model-dependent: structural choices are settled by the evidence, rankings between materials are not. Blind validation on four published configurations, nothing calibrated, misses exactly the two devices with segmented electrodes, which the proper longitudinal-longitudinal circuit then contains. What is enclosed is parameter uncertainty at fixed model; model-form error cancels in the ratio for thirty-one of the thirty-three comparisons.