Solidification model-based dual-objective optimization for dimensional accuracy and green strength in DLP 3D printing

Digital Light Processing (DLP) 3D printing enables high-resolution microfabrication by controlling light exposure with grayscale pixels. The energy exposed to the current layer also penetrates to prior layers, leading to overexposure commonly referred to as accumulated dosage. While accumulated dosage increases interlayer bonding, it also results in overcuring (curing beyond the intended layer boundary), reducing dimensional accuracy. Existing optimization efforts primarily focus on mitigating overcuring along the Z-axis, which inadvertently reduces green strength and causes microfeature failures. In this work, we formulate and experimentally validate a solidification model-based dual-objective optimization approach that simultaneously enhances dimensional fidelity and object green strength by spatially modulating voxel-level energy. The model computes 3D energy distribution from grayscale images using Gaussian distribution equations along XY and Beer Lambert’s absorption law along Z, followed by superposition. The optimization objective minimizes energy in the feature boundaries or negative regions to reduce overcuring, while maximizing energy in the core to improve interlayer crosslinking and structural integrity. We demonstrate the effectiveness of this approach in two applications. First, by increasing energy dosage through the core of re-entrant pillars, the optimized parts achieved both improved dimensional accuracy and ability to withstand higher adhesion forces during delamination, indicating enhanced mechanical strength. Second, microchannels with micro Z-direction holes were fabricated with dimensions closer to intended design while the rest of the part remained strongly bonded. These results highlight that dual-objective grayscale optimization effectively achieving both the intended dimensional accuracy and enhanced mechanical strength, enables fabrication of both positive and negative microfeatures with improved performance and reliability.