Experimental investigation and optimization of 3D-printed circumferentially corrugated multi-cell polygonal structures
Ahmet Alper Yanık, İbrahim Etem Karatay, Emre İsa AlbakPurpose
This study aims to investigate the crashworthiness of 3D-printed multi-cell thin-walled polylactic acid plus (PLA+) tubes under quasi-static axial compression. The objective is to explore the influence of inner geometry, radius application and diameter variations on crashworthiness indicators and to identify optimum geometric configurations that enhance energy absorption.
Design/methodology/approach
PLA + multi-cell tubes with hexagonal, octagonal and decagonal geometries are manufactured using Fused Filament Fabrication (FFF). Quasi-static axial compression tests are performed on a universal testing machine to evaluate crashworthiness parameters, including peak crushing force (PCF), energy absorption (EA), specific energy absorption (SEA) and crash force efficiency (CFE). The Taguchi method is applied to optimize geometrical parameters for improved crashworthiness.
Findings
The results demonstrated that internal geometry, radius ratio and inner diameter significantly influence crashworthiness. Decagonal structures exhibited superior EA and SEA compared to hexagonal and octagonal designs, owing to their increased number of edges and local buckling zones. Larger radius improved deformation stability and reduced stress concentrations, while increased inner diameters enhanced energy absorption. Taguchi optimization effectively identified configurations for optimum crashworthiness indicators. Notably, the optimized D_R3_22 model exceeded predicted EA values, achieving 638 J.
Originality/value
This study provides a novel contribution by experimentally and systematically analyzing the combined effects of internal geometry, radius ratio and inner diameter on the crashworthiness of 3D-printed tubes. Furthermore, the integration of the Taguchi method enabled the identification of optimal configurations that outperform the initial designs, including models that exceed predicted values.