Bibliographic Metadata
Bibliographic Metadata
- TitleMolecular modeling and simulation of real fluids for applications in process engineering
- Author
- Examiner
- Published
- Institutional NotePaderborn, Univ., Diss., 2010
- AnnotationTag der Verteidigung: 16.12.2010
- LanguageEnglish
- Document TypesDissertation (PhD)
- URN
Links
- Social MediaShare
- Reference
- IIIF
Files
Classification
Zusammenfassung
Für die Auslegung und Optimierung verfahrenstechnischer Prozesse ist die Kenntnis thermophysikalischer Stoffdaten, insbesondere Dampf-Flüssigkeitsgleichgewichte von Mischungen, unerlässlich. Da häufig nur sehr wenige experimentelle Daten vorliegen,werden Methoden benötigt, die eine quantitative Stoffdatenvorhersage ermöglichen. Hierfür werden üblicherweise Zustandsgleichungen oder GE-Modelle verwendet, die zwar sehr gute Korrelationswerkzeuge sind, aber erhebliche Schwächen bei der Stoffdatenvorhersage haben. Das Entwicklungspotenzial dieser Ansätze ist zudem weitgehend ausgereizt. Die vorliegende Arbeit ist in zwei Teile gegliedert.
Abstract
The first part of this work is on molecular modeling and simulation of 12 industriallyrelevant pure fluids as well as 12 binary mixtures of these components. Their economicalimportance and hazardous nature is a strong incentive for computer simulations.The investigation of these substances is separated into two groups. The Phosgene groupincludes Hydrogen chloride, Phosgene, Benzene, Chlorobenzene, Ortho-Dichlorobenzeneand Toluene. The Ethylene oxide group contains Ethylene oxide, Ethylene glycol andWater. The underlying force fields for these 12 pure substances are developed in this workon the basis of quantum chemical information on molecular geometry and electrostatics.The molecular models are individually optimized to experimental pure fluid data for vaporpressure and saturated liquid density. A comparison to other molecular models fromthe literature is given. The unlike dispersive interaction is optimized for ten of the 12studied binary mixtures. Previously unpublished experimental VLE data, measured byBASF in the vicinity of ambient temperature, are predominantly used for these fits. VLEdata, including dew point composition, saturated densities and enthalpy of vaporization,are predicted for a wide range of temperatures and compositions. The predictions arecompared to additional experimental binary VLE data that was not considered in themodel development. The good agreement shows the reliability of the molecular approachfor predicting thermophysical properties of hazardous fluid mixtures.In the second part of this work, by assessing a large number of binary systems, it isshown on a large scale that molecular modeling is a reliable and robust route to VLE of mixtures.
Stats
- The PDF-Document has been downloaded 43 times.
License/Rightsstatement
