<p>In multi-axis ball-end milling, tool posture plays a key role in determining surface quality, especially when machining difficult-to-cut nickel-based superalloys such as GH4169 (equivalent to Inconel 718). Improper tool postures can induce surface adhesion defects and degrade the service performance of machined components. Most existing studies evaluate surface quality using discrete tool posture parameters, making it difficult to capture the influence of continuous posture variations. To address this issue, a semi-continuous tool posture optimization method is proposed. First, a series of 3-axis ball-end milling experiments on inclined planes with radial tool paths was conducted to analyze the effects of workpiece inclination angle and feed direction angle on surface topography and roughness. Then, based on the geometric mapping relationship between 3-axis and 5-axis machining, a surface roughness prediction model was established with the lead and tilt angles as continuous variables, and optimal tool posture regions for suppressing surface adhesion were identified. Finally, a 5-axis plane milling verification experiment was carried out. The results show that the optimal tool postures determined by the proposed model result in lower surface roughness and smoother surface topography compared to the non-optimal postures. This study provides a theoretical basis for tool posture optimization and surface quality control in multi-axis ball-end milling.</p>