Experimental measurements and prediction of thermophysical properties of hydrocarbon mixtures using nonlinear models

摘要

Nowadays, thermophysical properties of hydrocarbon mixtures play a vital role in process industries. A thermophysical asset of the solvent mixture properties provides information for distillation, extraction operations, material balance and energy balance in process industries. They also play an important role in solving problems in heat transfer, mass transfer and fluid flow. Studies on different thermophysical properties of liquid mixtures within a wide ranges of composition and temperatures are valuable sources of information that can be used to examine the relationship between the internal structure of the system and its physical properties. Though a lot of work has been done in the case of pure liquids, data on thermophysical properties of liquid mixtures are not available for many systems. Hence the thermophysical properties of liquid mixtures have been studied for four binary liquid mixtures in the current investigation. Thermophysical properties of the following hydrocarbon binary liquid mixtures have been determined by using Oswald Viscometer and Pycnometer at 303.15, 308.15 and 313.15 K.ud1, 4 Dioxane + Bromobenzeneud1, 4 Dioxane + EthylbenzeneudAcetophenone + P-XyleneudAcetophenone + O-XyleneudThe thermophysical properties such as density, viscosity of binary liquid mixtures were determined experimentally over the entire composition range at 303.15K, 308.15K and 313.15K. The experimentally determined thermophysical properties of the binary liquid mixtures were used to calculate the excess molar volume, VE and viscosity deviations Δη. The excess thermophysical properties of liquid mixtures provide additional information regarding molecular interactions. Hence the intermolecular interactions of the mixtures are discussed with the help of excess properties. A perusal of the literature revealed that the predictions of thermophysical properties of liquid mixtures was scarce. With an aim, the thermophysical properties of density, viscosity, excess molar volume VE, and viscosity deviations Δη of liquid mixtures were predicted using various nonlinear models. The excess values of thermophysical properties of binary hydrocarbon liquid mixtures were correlated using Redlich-Kister polynomial equation to obtain their coefficients and standard deviations. Grundberg-Nissan, Krishnan-Laddha, McAllister and Jouyban-Acree viscosity models were used for predicting the viscosity of hydrocarbon liquid mixtures. The parameters of McAllister model were determined using a programming software and other models parameters were determined using polynomial equations. The experimental values and model predictions were compared to get the standard deviation for each model. The observed values of excess molar volume VE, viscosity deviations Δη data for the hydrocarbon mixtures support the main factor of gradual disruption of the self-associated aromatic hydrocarbon molecules and confirm that the hydrogen bonds and dipolar interaction in aromatic hydrocarbons makes VE and Δη positive and negative deviations. The weak physical intermolecular interactions between the aromatic hydrocarbon molecules dominate over the structure-breaking effect of hydrocarbon mixtures on the addition of aromatic hydrocarbons. Interactions between the liquids in hydrocarbon mixtures are strong and weak dispersive type interactions. Redlich-Kister polynomial equation represented the excess molar volume and viscosity deviations which is indicated accurately well by percent standard deviation of less than 1.25. The viscosity data tailored to Grundberg-Nissan, Krishnan-Laddha, Jouyban–Acree and McAllister models to derive the binary coefficients. Standard deviations have been considered between the fitted outcomes and the calculated data is helpful in deliberate mixing behavior of the binary mixtures. It can be concluded that in all cases, the data values found correlated with the corresponding models very well for this mixtures. The molecular interactions existing between the components and comparison of liquid mixtures were also discussed. The obtained experimental and correlation data is greatly reducing the difficulties in a tedious experimental work and provides a faster way of predicting the property values. The use of data correlation enables for prediction of data in future. Once a correlation is known with strong relationship between the variables of the study, then the prediction will be more accurate. Other than that, the use of these correlation models provides knowledge on the real behaviour of the liquid mixtures using only the experimental results and properties of the pure components.