Fig. 2: Thermo-mechanical behavior and microstructural analysis of PLA and TPU disks.

PLA disks of varying thicknesses behave differently when exposed to an external thermal stimulus (a). Disks with h \(=\) 2 mm thickness are considered as the reference specimens in this study because in-plane stresses dominate the shape-shifting behavior in thinner disks (i.e., \(h\) < 2 mm), causing the disk to fold or roll. When the sheet thicknesses are large enough (i.e., h ≥ 4 mm), the increased bending stiffness prevents the disks from fully bending (a). The experimentally measured expansion factors (\({\beta }_{1},{\beta }_{2}\)) were used in our computational models (b). The results of dynamic mechanical tests (i.e., storage moduli in a logarithmic scale (c–top), and \(\tan \delta\) (c–bottom)) performed on the PLA specimens printed with the maximum (i.e., 80 mm s⁻¹) and minimum (i.e., 20 mm s⁻¹) printing speeds and for the TPU specimens printed with a speed of 25 mm s⁻¹. Constructed μCT images of a PLA disk were used to analyze the formation of micro-voids during the 4D printing process (d). A μCT image of a disk whose inner and outer parts were printed at 80 mm s⁻¹ and 20 mm s⁻¹, respectively. A morphological analysis of these specimens showed virtually no porosity for the PLA disks printed with the lowest speed (i.e., 20 mm s⁻¹) and TPU while the level of porosity was significantly higher for the disks printed at 80 mm s⁻¹ (e). The level of porosity also increased in the radial direction for the PLA disks printed at 80 mm s⁻¹ (f). See Supplementary Movie 1 for the visualization of the defect distribution in the specimens printed at different speeds. The lines in subfigures e and f are to guide the eyes of readers.