Optimize projected mask images for improving three-dimensional printing accuracy for digital light processing based vat photopolymerization

Abstract

Additive manufacturing technology, specifically digital light processing (DLP) based vat photopolymerization (VPP), is extensively employed for fabricating complex constructs. Nevertheless, the accuracy of DLP-VPP is hindered by challenges such as light diffraction during propagation and light penetration along the exposure axis within the photopolymeric resin. To mitigate these issues, regulating the grayscale of projected mask images offers a reliable approach to reduce undesired curing. In this study, we introduce a voxel-based multi-layer solidification model to predict the exposure dose distribution of printed parts. Based on this model, we propose a pixel-level grayscale optimization method using the gradient descent technique. This method allows us to control the accumulated exposure dose of each voxel by modifying the grayscale of individual pixels in the projected mask images, thereby improving the accuracy of the printed parts. To validate our optimization method, we conducted a series of experiments involving the printing of four representative structures using both the traditional method and our proposed optimization method. The results demonstrate that the structures produced using our optimization method exhibit superior consistency with the target geometry. This indicates a significant improvement in accuracy achieved through our optimization technique. Furthermore, our optimization method has successfully enabled the fabrication of microchannels that are previously unattainable using traditional method. This achievement highlights the immense potential of our optimization method in the realm of printing microchannels.