← All publications

Layerwise spectral element method with frequency-dependent material identification for band-gap analysis in periodic polymeric sandwich structures

Lotfan, S., Fallah, A., Yıldızdağ, M.E., Bediz, B., Koç, B.

Composite Structures, 402, pp.121117, 2027. Journal paper

Figure for Layerwise spectral element method with frequency-dependent material identification for band-gap analysis in periodic polymeric sandwich structures

Abstract

Accurate band-gap prediction in additively manufactured polymeric sandwich structures with core cut-outs requires both efficient treatment of locally disconnected layers and realistic representation of polymer viscoelasticity. To address the former, this study develops a new layerwise spectral element method (LW-SEM) combining spectral Chebyshev discretization, layerwise first-order shear deformation kinematics, Bloch–Floquet periodicity, and pointwise transformation-based coupling. The formulation enforces inter-layer continuity while accommodating cut-out-induced disconnections without artificial void properties. To address the latter, an identification strategy based on experimental frequency response functions (FRFs) updates dynamic mechanical analysis (DMA)-derived properties into frequency-dependent storage modulus and loss factor functions. The framework is applied to polylactic acid-thermoplastic polyurethane (PLA-TPU) sandwich beams. PLA properties are calibrated using a single-material beam, while TPU properties are identified from a baseline sandwich beam through modal strain–energy partitioning. Compared with three-dimensional finite element method (FEM) results, LW-SEM reproduces natural frequencies, mode shapes, and dispersion curves with substantially fewer degrees-of-freedom (DOFs). Without further adjustment, the identified effective properties reduce the mean error relative to the measured resonances of beams containing one, three, and five periodic cells from 12.7% to 1.9%. The efficiency of LW-SEM is particularly advantageous for iterative dispersion analysis with frequency-dependent properties, where repeated high-DOF FEM solutions can become computationally prohibitive.

Read paper

Related publications