Suitability of low‐field nuclear magnetic resonance (LF‐NMR) combining with back propagation artificial neural network (BP‐ANN) to predict printability of polysaccharide hydrogels 3D printing
Chaofan Guo, Min Zhang, Huizhi Chen
Abstract
Summary The water state of polysaccharide hydrogel is closely related to its rheological properties, which is the critical parameter determining the printability of 3D printing. Low‐field nuclear magnetic resonance (LF‐NMR) results showed that T 23 (free water) values of polysaccharide hydrogels gradually decreased with increasing concentration, while the viscosity values exhibited a opposite trend. As concentration increased, the printed objects gradually changed from the fluid, which cannot form a shape, to a stable shape and then to a hydrogel of low fluidity that cannot print normally. Hydrogels became hard to extrude smoothly at the piston pressure higher than 428716 Pa. Back propagation artificial neural network (BP‐ANN) nonlinear models established by taking the fingerprint LF‐NMR signal as input variables yielded good predicting ability of the piston pressure values ( R 2 adj = 0.982) and printing scores ( R 2 adj = 0.988). The BP‐ANN model based on LF‐NMR might be a promising approach to quickly predict the 3D printability of polysaccharide hydrogels.