Tip-based up-milling for smooth microchannel structures using Lissajous trajectories

Microfluidic channels offer significant advantages and have broad application potential in fields such as biomedicine and materials science. However, machining microchannels with dimensions on the order of tens of micrometers presents challenges related to machining quality and cost. Therefore, developing cost-effective and low-burr micromachining processes is crucial for advancing the application of microchannel structures. This paper proposes a method for fabricating microchannel structures using tip-based micro-milling with Lissajous trajectories, featuring a frequency ratio of 2. Leveraging the characteristics of Lissajous trajectories, the novelty of this method lies in its ability to achieve up-milling on both sidewalls of the channel in a single pass, without the occurrence of down-milling, thereby minimizing burr formation. By varying the phase angle to adjust the shape of the Lissajous trajectory, microchannels with different machining qualities are produced. The results reveal that material residue is the primary factor degrading machining quality, particularly in terms of sidewall smoothness and bottom surface roughness. The material residue caused by tip cutting is initially formed during the forward revolving movement of the tip (cutting path) and is subsequently reprocessed during the backward movement (non-cutting path). Numerical simulations, combined with experimental results, are performed to investigate the distribution of material residue on the microchannel bottom under different trajectories. Finite element (FE) analysis is used to simulate cutting processes with time-varying uncut chip thickness and cutting angles in various Lissajous trajectories, focusing on characterizing the primary shear zone. The optimal phase range of the Lissajous trajectory, between 45° and 60°, is identified, within which microchannel structures with widths of 10 μm and 20 μm are fabricated. The channel sidewalls exhibited improved smoothness, and the bottom surface roughness was minimized to S_a=20.4 nm.