Viscosity optimisation of photosensitive Al2O3 slurry for stereolithography based additive manufacturing

Abstract

Viscosity of alumina slurry is a key factor affecting the quality of ceramics formed by stereo photolithography, but it is closely related to the resin distribution ratio, dispersant content, plasticizer content and solid content. Most researchers utilize the single factor method to study the composition and ratio of the slurry. In this study, orthogonal experimental design and back propagation artificial neural networks methods were combined to solve the optimisation problem of multi-objective and multi-factor influence on alumina slurry performances. The results of optimal composition and content allocation were achieved by back propagation artificial neural networks and experimental testing. It was shown that the optimal conditions are: resin composition HDDA : PPTTA = 4 : 1, DS-165A dispersant content of 3.86wt.%, PEG plasticiser amount of 3.5wt.% and the solid content of 75.74wt.%. The predicted optimal viscosity value was 8787mPa•s and the shrinkage rate could reach 14.57%. The optimal values of viscosity and shrinkage were consistent with the experimental results, the viscosity and shrinkage errors were only 4.06% and 3.856%, respectively. The average density and bending strength of the sintered samples were 3.979 ± 0.005 g/cm3 and 365 ± 61MPa, respectively. According to the obtained data, stereolithography 3D printing alumina slurry with excellent flowability and low shrinkage was successfully prepared.

Keywords

• stereolithography
• al2o3 slurry
• orthogonal experiment design
• artificial neural networks