An efficient thermal assessment on motorized spindle heating with the structural constraint impact

• A full heat transfer capacity assessment function with structural constraints. • A constructed thermal equivalent resistancebased on heat transfer equivalence. • A novel dimension reduction approach of thermal networks. • A comprehensive thermal network for a motorized spindle with heat radiation. Substructures not only affect the heat conduction and contact heat transfer of spindle units, but also impact the radiation and convection heat dissipations, which were extensively studied. More and more integrated structural details for higher thermal forecasting precision raise the complexity of prediction models significantly. This will inevitably result in a contradiction between the accuracy and efficiency of thermal assessment. Considering the high real-time and deep integration requirements of CNC (Computer Numerical Control) machining systems, to perform a dimension pre reduction of complex models and then reconstruct a lightweight and precise thermal grid is an ideal solution to the current challenge. This paper aims at developing a simplification approach of complex thermal grids to assess the spindle heating truly and efficiently, and meanwhile providing an effective way in the prediction accuracy gap improvement of thermal network modeling. For this, a typical spindle unit was selected as an analysis object. The sub structures of its bearings and their external constraints were first simplified and abstracted, and then the full heat network of a typical feature unit (hollow cylinder) was planned. The heat transfer capacity evaluation function with structural constraint was subsequently constructed. By it, the roles of axial heat dissipation, radiation and convection as well in temperature reduction were discussed in detail, which provides a strong theoretical base for next thermal model simplification. The thermal equivalent resistance on a hollow cylinder was accordingly proposed, and a novel dimension reduction approach of thermal networks was constructed. Next, a novel spindle thermal prediction network with radiation heat transfer was planned. It, as a typical application, also was employed to explain the dimension reduction in detail. In this process, we employed the oil-air separation and heat transfer equivalence to further improve the latest multi-node mode of bearings, and meanwhile the double counting on the convective heat transfer of oil-air also was avoided. Finally, the above-proposed work was validated by experiments and contrast. The temperature predictions based on the current and newly developed models with no heat radiation, meanwhile, were performed and compared with the simplified grid for further verification. Both comparative analysis and test verification indicate that the thermal forecasting is more lightweight with high accuracy. This dimension pre reduction way evidently also benefits improving the forecasting precision shortcoming effectively because there is no worry about the complexity of the developed thermal assessment networks by employing it. These above contrast researches, in the meantime, also prove our developed spindle heating forecasting grid model considering the thermal radiation effect.