Prediction of heats of mixing of multicomponent systems using local composition models
Mining and Metallurgical Engineering
StatisticsView Usage Statistics
The parameters of the Wilson, NRTL and UNIQUAC local composition models were obtained from experimental binary heat of mixing data. The calculations involved minimizing the residual sum of squares (equation) using the Hooke-Jeeve algorithm. Heats of mixing of multicomponent systems were calculated with the estimated parameters and compared to available experimental data. For systems containing no alcohol or water, the three local composition models can be used to predict heats of mixing of multicomponent systems with errors normally not exceeding 15 %. For systems containing water, the empirical equation, proposed by Tsao and Smith proved to be more adequate for the prediction of ternary heats of mixing than the three local composition models used in this work.
Online access for this thesis was created in part with support from the Institute of Museum and Library Services (IMLS) administered by the Nevada State Library, Archives and Public Records through the Library Services and Technology Act (LSTA). To obtain a high quality image or document please contact the DeLaMare Library at https://unr.libanswers.com/ or call: 775-784-6945.