所以減少磨損、提高使用效率以延長零件壽命是軸承研究重要議題。 流體動壓軸承在成本方面相對於其他形式之軸承較低，並且在系統設 計方面較其他軸承系統簡單，加上使用流體動壓軸承使用黏度較大的 潤滑介質可以使軸承有更好的抗震能力、更大的承載力及較低的噪音， 是現行常用於解決磨損的潤滑解決方案。 目前已知有實驗研究將動壓軸承中的潤滑介質換成磁流體，並做 與無外加磁場以及外加磁場排列不同的方式對於磁流體在動壓軸承 的潤滑效果的差異比較，而在所有外加磁場排列中交互排列的磁鐵排 列方式中對於磁流體潤滑液的潤滑效果為最顯著的提升，有鑑於將流 體動壓軸承中的潤滑液體替換成磁流體並加上磁流體是屬於較新的 領域，相關數值模擬較少文章。 本研究主要是利用數值模擬探討磁流體動壓軸承運動，此研究建 立的數值模擬模型採用以磁流體動壓軸承實驗作為數值模型建立的 依據，並且使用商用軟體 COMSOL Multiphysics 作為數值建模的工 具，在上述條件下直接對流體物理性質進行設定的方式來模擬薄膜磁 流體在磁場下的性質，以此方式建立薄膜流體的數值模型可以減少目 前常見的多重模型耦合所需要的計算資源，在減少運算需求的前提下 維持數值模型的可行性。 為驗證此數值模型之可行性，本研究取參照實驗的結果與數值模 型輸出之結果比較，並且對實驗與數值模擬的頻譜圖與軌跡圖進行比 對，比對重點會集中於發生不穩定之頻率與軌跡圖的形狀做判斷，此 數值模型經實驗驗證後可以說明直接對流體物理性質進行設定的方 式以模擬薄膜磁流體在磁場下的性質之可行性。

Abstract The main reason for bearing wear in mechanical structures is the shortened lifespan of mechanical components. Therefore, reducing wear, improving efficiency, and extending component life are important topics in bearing research. Fluid dynamic bearings have a lower cost compared to other forms of bearings and are simpler in system design. Additionally, using lubricants with higher viscosity in fluid dynamic bearings can provide better shock resistance, higher load-bearing capacity, and lower noise, making them a commonly used lubrication solution for addressing wear. Currently, there are experimental studies replacing the lubricating medium in hydrodynamic bearings with magnetic fluid. These studies explore the differences in lubrication effects of magnetic fluids in hydrodynamic bearings with and without external magnetic fields, as well as different arrangements of external magnetic fields. Among all the arrangements of external magnetic fields, the interleaved arrangement of magnetic poles shows the most significant improvement in the lubrication effectiveness of magnetic fluid. Since replacing the lubricant in fluid dynamic bearings with magnetic fluid and adding a magnetic field is a relatively new area, there are few related numerical simulation articles. This study primarily employs numerical simulations to investigate the motion of magnetic fluid in hydrodynamic bearings. The numerical simulation model established in this research is based on experimental data from magnetic fluid hydrodynamic bearing experiments. Commercial software COMSOL Multiphysics is used as the numerical modeling tool, and the fluid properties are directly set under the specified conditions to simulate the behavior of thin-film magnetic fluid under a magnetic field. This approach of establishing a numerical model for thin-film fluid reduces the computational resources required compared to common multi-model coupling, while maintaining the feasibility of the numerical model. To validate the feasibility of this numerical model, the study compares the results obtained from the model with the experimental outcomes, focusing on comparing the spectra and trajectory plots of both. The emphasis of the comparison is on the frequencies at which instability occurs and the shapes of the trajectory plots. After experimental validation, this numerical model demonstrates the feasibility of directly setting fluid physical properties to simulate the behavior of thin-film magnetic fluid under a magnetic field.