Machine learning-aided optimization of 3D-printed MWCNTs/NiFe/PETG osteospermum stepped-structure via full-wave simulation for broadband microwave absorption

The microwave absorption materials with broadband and wide-angle absorption characteristics represent a crucial prospect for addressing electromagnetic pollution. This study employs CST Studio and machine learning model to predicate the broadband absorption performance and corresponding optimal structure parameter. The specific optimum structure, called bio-inspired Osteospermum stepped structure, fabricated by 3D printing technology. The stepped structure demonstrates exceptional isotropic behavior and broadband absorption performance: across the 2–40 GHz frequency range, it achieves an impressive EAB of 31.4 GHz (spanning 8.6–40 GHz), while maintaining outstanding absorption capability even at large oblique incidence angles up to 65°. Macroscopically, the step-like interfaces generate multiple reflected waves with varying phase differences, leading to phase cancellation interference and effective electromagnetic wave attenuation. Microscopically, the incorporation of multi-walled carbon nanotubes (MWCNTs) and NiFe particles into the PETG matrix introduces additional conduction loss and magnetic loss. Ultimately, combination of 3D printing and theoretical simulation present a novel approach to design and fabricate unique structures with simultaneous broadband and wide-angle performance.