透過您的圖書館登入
IP:3.147.103.8
  • 學位論文

己二酸晶種式批次結晶槽之模擬、操作及最適化

Modeling, Operation and Optimization of Experimental Seeded Batch Crystallization of Adipic Acid

指導教授 : 吳哲夫

摘要


本篇研究利用晶種式批次結晶去測試各種操作條件,並以此建立晶體的生長結晶動力模型。其後以晶體負載率及溫度曲線對產物晶體分布關係做實驗探討,對晶體負載率而言,結果顯示利用更多以及更小的晶體將有利於減少成核。而對相同的晶體負載率,凸型溫度曲線結果明顯優於線性溫度曲線;然而,最適化理想溫度曲線結果只有稍微優於其他凸型曲線,所以我們建議未來研究員在發展裡線溫度曲線時,結果應該與凸狀曲線做比較,而非線性跟自然冷卻溫度曲線。因為凸狀曲線是在不需要動力模型下能達到最好的結果,動力模型也用來調查不同目標函數的最佳化曲線的影響。最後,實驗用來測試一個理論,後期生長策略有助於減少新成核晶體重量,而初期生長策略會有比較大的成核重量。

並列摘要


Crystallization is a commonly used process for the production of pure substances or separating phases. Therefore, it is important to understand the characteristics of the crystallization process through kinetic considerations. General target of this thesis is the experimental examination of the batch-wise crystallization process of adipic acid in water and the determination of relevant kinetic parameters. In this work, a seeded batch crystallization experiment was developed to go through several operation policies. The experiment data at different experimental conditions were collected and were used to determine the kinetic parameters of crystal growth rate and crystal nucleation rate. Then the experiment was made use of investigate the effect of cooling trajectory and seed properties on the product crystal quality and crystal size distribution. In results, compare to the effect on the crystal size distribution between different seed loading and different temperature trajectories, to change the seed loading ratio is a more effective way to suppress nucleation and obtain better product equality. For a given seed loading, linear trajectory perform significantly worse than convex trajectories such as cubic, quartic, and Mullin-Nvylt trajectories. The performance of optimal trajectory is the best. Finally the experiment was used to test the phenomena proposed, which are called ‘early growth’ trajectories and ‘late growth’ trajectories. The ‘early growth’ trajectories give the larger supersaturation in the beginning of the batch than the end of the batch, which will produce a larger number of nuclei but a smaller nucleation mass. In contrast, ‘late growth’ trajectories give the smaller supersaturation in the beginning of the batch than the end of the batch, and it will produce a larger number of nuclei but a smaller nucleation mass. The results confirmed the theoretical prediction to be consistent again.

參考文獻


1. Jones, A. G., J. Budz, et al. (1987). "Batch Crystallization and Solid Liquid Separation of Potassium-Sulfate." Chemical Engineering Science 42(4): 619-629.
4. Mullin, J. W. and J. Nyvlt (1971). "Programmed Cooling of Batch Crystallizers." Chemical Engineering Science 26(3): 369-&.
5. Jones, A. G. (1974). "Optimal Operation of a Batch Cooling Crystallizer." Chemical Engineering Science 29(5): 1075-1087.
6. Miller, S. M. and J. B. Rawlings (1994). "Model Identification and Control Strategies for Batch Cooling Crystallizers." AIChE Journal 40(8): 1312-1327.
8. Jagadesh, D., N. Kubota, et al. (1996). "Large and mono-sized product crystals from natural cooling mode batch crystallizer." Journal of Chemical Engineering of Japan 29(5): 865-873.

延伸閱讀