|摘要: ||本篇研究是以直徑10 mm的水旋風分離器進行高分子溶液中懸浮粒子之分離，分別進行實驗量測與數值模擬分析，以實驗探討改變濃度之懸浮液與進料速度、壓力、分流比等操作條件對分離效率的影響，且改變不同物料之實驗結果進行比較，並使用Ansys套裝軟體，利用SIMPLE(Semi-Implicit Method for Pressure-Linked Equation)法則進行質量平衡與動量平衡等統御方程式之計算，紊流使用雷諾應力模型，並使用PRESTO!(PREssure STerring Option)進行壓力計算之疊代，分析流體之速度分布、壓降分布、以及粒子軌跡等，再藉以估算粒子的分離效率，經此研究探討最佳的操作方式。|
在分離懸浮液時，實驗結果顯示，總分離效率會隨著操作壓力、分流比和n值的增加而增加。當懸浮液在n值從0.51增加至1，總效率最多可以增加51 %。而當分流比自0.2增加至5，總效率會增加2倍。在PVC之懸浮液n值增加，其d50值也會跟著變大，範圍從25 m降至8 m，而底流口處的粒子濃度也會跟著下降5 %。比較三種物料的總效率值之順序為：Al2O3 > PVC > Kaolin。
在模擬方面之結果顯示，壓降會隨著進料速度的增加而增加，也會隨著流體之黏度的增加而增加，n值從0.51至1最多增加了149 %，。在粒子軌跡部分，n值越大，使粒子迴圈數與軌跡長度增長，導致不易有粒子從由溢流管流出。Kaolin的分級效率實驗與模擬值誤差較小約13 %，反之Al2O3及PVC之模擬與實驗結果誤差約40 %，由此可知密度和粒子形狀也會影響模擬之結果，而最後由分級效率與總效率可知實驗數據與數值模擬結果可知，推斷實驗與模擬相符合。
A 10-mm hydrocyclone was used for separating particles suspended in polymer solutions. The effects of operating conditions, such as concentration of PAA solutions, inlet velocity, pressure drop and split ratio, on the separation efficiency were studied using experimental and simulated methods. In simulation, the governing equations were coupled using the SIMPLE algorithm, and the Reynolds stress model was employed for the turbulent model in hydrocyclone. A numerical software, Ansys 14.0, was used for analyzing the distributions of fluid velocity and static pressure. The particle trajectories and separation efficiency were then simulated accordingly. Three kinds of particles, Al2O3, PVC and Kaolin, were used in experiments. The experimental results showed that the separation efficiency increased with increasing pressure drop, split ratio and fluid behavior index, n. The total separation efficiency increased 51% when n-value increased from 0.51 to 1.0, and it enhanced two-fold when the split ratio increased from 0.2 to 1.0. The index d50 increased with increasing n-value, it decreased from 25 m to 8 m as n-value increased from 0.51 to 1.0 for PVC particles. Comparing the total separation efficiency of three used particulate samples, the sequence was Al2O3 > PVC > Kaolin. The simulating results showed that the pressure drop increased 149% when n-value increased from 0.51 to 1.0. The rotation number and migration length of particles increased with increasing n-value. This causes that particles are more easily to be collected through underflow. The simulated particle separation efficiency approximately agree with available experimental data.