Experimental optimization and thermo-mechanical modeling of 4D-printed PETG structures

• Novel integrated framework combining experiments, regression modeling, and FEM to predict PETG morphing behavior. • Taguchi analysis identified layer height and printing speed as dominant parameters influencing deformation. • Regression models achieved R 2 up to 0.97, accurately capturing nonlinearities and interactions. • FEM with layer-specific CTE s accurately replicated curvature across 18 experimental runs. • Six-layer discretization provided the best balance between model fidelity and mesh quality. This study introduces a novel experimental–computational framework to analyze the thermally induced shape-morphing behavior of polyethylene terephthalate glycol-modified (PETG) beams fabricated by fused deposition modeling (FDM). By combining Taguchi L18 design, signal-to-noise ( S/N ) ratio analysis, regression modeling, and finite element simulations, this work presents a unique approach to predict and validate programmable deformation in PETG. S/N ratio analysis revealed that layer height and printing speed are the dominant factors, explaining up to 46% and 61% of the variance, respectively. Regression models demonstrated high predictive accuracy ( R ² up to 0.97), capturing nonlinearities and interaction effects. Finite Element Modeling (FEM) simulations, incorporating layer-specific coefficients of thermal expansion, replicated the morphing response with sub-millimeter agreement to experimental data. This integrated framework, which combines experimental design, statistical modeling, and FEM, provides a validated tool for optimizing thermally responsive PETG structures, offering promising applications in soft robotics, deployable systems, and adaptive materials.