Three-dimensional printing is a transformative technology in the manufacturing industry which provides customization and cost-effectiveness for all walks of life due to its fast molding speed, high material utilization, and direct molding of arbitrary complex structural parts. This study aims to improve the molding accuracy of 3D printed polyether ether ketone (PEEK) samples by systematically studying key process parameters, including printing speed, layer thickness, nozzle temperature, and filling rate. The 3D printing nozzle has an important impact on the extrusion rate of the melt, and the fluid simulation of the nozzle was carried out to explore the variation characteristics of the melt flow rate in the nozzle and optimize the nozzle structure parameters. In order to effectively optimize the process, considering its inherent efficiency, robustness, and cost-effectiveness, the L9 orthogonal array experimental design scheme was used to analyze the effects of printing speed, layer thickness, nozzle temperature, and filling rate on the molding accuracy of the test sample, and the optimal combination of process parameters was optimized through the comprehensive weighted scoring method so as to improve the molding accuracy of the 3D printed PEEK sample; finally, the molding accuracy of the components printed using the Sermoon-M1 3D printer with the optimized nozzle structure was printed. The results show that the nozzle structure is optimal when the convergence angle is 120° and the aspect ratio is 2, and the outlet cross-section velocity is increased by 2.5% and 2.7%, respectively. The order of influence strength on the dimensional accuracy of the test sample is layer thickness > filling rate > nozzle temperature > printing speed. The optimal combination of parameters is: a printing speed of 15 mm/s, a layer thickness of 0.1 mm, a nozzle temperature of 420 °C, and a filling rate of 50%. The insights derived from this study pave the way for predicting and implementing the selection of optimal process parameters in the production of 3D printed products, with important implications for the optimal molding accuracy of printed components.

Particle-based 3D printing shows great potential in high-performance composite fabrication due to high raw material utilization and flexible material compatibility. However, constrained by conventional extrusion system structures, critical issues (non-uniform melt conveying, insufficient mixing efficacy, poor extrusion stability, etc.) remain. To address these, this study proposes a novel separate-type pin screw integrating solid–liquid separation (from split screws) and high-efficiency mixing (from pin screws) to improve CF/PLA composite extrusion efficiency and mixing homogeneity in particle-based 3D printing. Three-dimensional modeling, static strength/stiffness analysis, and POLYFLOW-based numerical simulation of particle melt conveying/mixing in the screw channel were conducted to analyze structural parameter effects on pressure field, shear rate, and mixing. Experiments assessed printer extrusion rate (different screws) and printed specimen mechanical properties. The simulation and experiment confirmed the optimized screw has better pressure distribution and mixing at 20 rpm, with optimal pin parameters: diameter 2 mm, height 1.6 mm, radial angle 60°, and axial spacing 10 mm. This work offers theoretical/structural support for particle-based 3D printing extrusion system optimization.