The assembly quality of high-torque nuts in the deep cavity (95mm-diameter, 800mm-deep) of aero-engine low-pressure rotors directly determines engine operational reliability. Key technical challenges include meeting the 1900–2100 N m torque requirement, \(\pm 0.2^\circ\) angular accuracy demand, and stable preload control under thermal-centrifugal coupling conditions. To address these issues, this study proposes an optimization method integrating fractal contact mechanics and multi-physics coupling, and develops a visual tightening system that combines visual monitoring with servo control. Utilizing the W-M (Weierstrass-Mandelbrot) fractal function and Hertz theory, the research reveals that a surface roughness of Ra=1.6 \(\upmu\) m limits the real contact area to 12%–18% of the nominal area. A thread stiffness model is established, with the calculated overall stiffness of \(5.06\times 10^5\) N/mm showing a relative error of less than 1% compared to experimental results. ANSYS simulations quantify preload attenuation under thermal-centrifugal coupling conditions. The developed system achieves an angular positioning accuracy of \(\pm 0.18^\circ\) . Experimental verification on 20 aero-engines test dummy shows that the developed system achieves an angular positioning accuracy of \(\pm 0.18^\circ\) , controls preload error within \(\pm 8\%\) (relative to the 200 kN design preload), reduces single assembly time from 4 to 2.6 h (35% efficiency improvement), and effectively avoids part collisions during the assembly process. These results fully meet the high-reliability assembly requirements of aero-engines.
