This project concerns the evaporation of liquid films composed of multiple hydrocarbon components (to simulate gasoline) and is motivated by the debate regarding the significance of fuel films to hydrocarbon emissions from direct-injection, spark-ignition (DISI) engines.  The goals of this research are 1) to improve the understanding of the roles of the thermal and mass transport processes of film evaporation, 2) to provide a thorough characterization of the changing film composition during evaporation, and 3) to obtain experimental data taken under well-specified conditions representative of internal combustion engines that may be used for model validation.  While the motivation of the program is the practical problem caused by fuel films in DISI engines, we are interested generally in the fundamental principles that govern film evaporation and therefore will study film evaporation under a wide variety of conditions, not just those pertinent to engines.

            During the 2006-2007 reporting period, we conducted new experiments which led to an insight regarding the role of buoyancy in film evaporation.  Also, we continued to refine the spectroscopic and gravimetric experiments that we reported last year.  The spectroscopic experiments use a Fourier transform infrared spectrometer (FT-IR) to measure the transient composition of an evaporating film and the gravimetric experiments use an analytical balance to record the changing film mass as the film evaporates.

            An underlying assumption of our spectroscopic experiments is a constant interfacial surface area between the film and vapor.  To validate this assumption shadowgraph imaging, and subsequently schlieren imaging, were used.  These imaging techniques not only enable us to measure the interfacial area, they also enable us to measure the film volume and thickness as a function of time.  From these measurements, the film evaporation rate and evaporation flux (evaporation rate per surface area) are computed.  Additionally, schlieren imaging enables us to investigate the vapor layer that forms above the film.  The vapor layer thickness and the vapor concentration distribution are computed.

            A shadowgraph or schlieren imaging experiment results in thousands of images that must be analyzed in order to obtain the desired data.  This task is not feasible to do by hand and so computer programs were written to automate the image analyses.

            In addition to our experimental activities, a computational model of evaporation in a quiescent environment was developed by Dr. Leon Phillips, who traveled from the University of Canterbury in New Zealand to work with us.  Using this model as a context for analyzing the schlieren images led to the insight that natural convection of the film vapors significantly affects the film evaporation rate under nominally quiescent, ambient test conditions (P = 1 atm, 23°C<T<29°C).

            Our investigation of film evaporation first focused on films composed of pure hydrocarbon solvents.  The solvents used were alkanes from pentane to decane (excluding nonane) as well as cyclohexane, 3-methylpentane (3MP), and 2,2,4 trimethylpentane (isooctane).  The following are our primary findings from our pure film experiments, which were conducted under nominally quiescent, ambient conditions.  Our experiments were limited to measuring the film from 100% to approximately 50% of its initial mass.

·        The evaporation flux of each solvent is nearly constant.

·        A vapor layer immediately forms just above the film surface.  The thickness of this layer is approximately constant.

·        The evaporation fluxes appear to be controlled by a combination of diffusion through the vapor layer and a buoyancy-induced horizontal flow.

            The same solvents were used to study the evaporation characteristics of bi-component films.  The experiments were conducted under the same ambient conditions as those for the pure films.  Our analyses of the experimental results continue but our preliminary findings, which pertain only to bi-component mixtures of 3MP and hexane (mixture A) and 3MP and isooctane (mixture B), are listed below.

·        Each component's initial evaporation rate equals its ideal rate, which is the product of its mole fraction and its evaporation rate in a pure film.

·        The evaporation rates of the components vary with time.  The changing evaporation rates are believed to be due to the changing film composition.

·        As time passes, the components' evaporation rates diverge from their ideal values.  This divergence may be due to the development of a non-uniform film composition, for which the surface becomes concentrated in the lower volatility component.

·        A vapor layer immediately forms just above the film surface. The thickness of this layer is approximately constant.

·        The evaporation fluxes appear to be controlled by a combination of diffusion through the vapor layer and the buoyancy-induced horizontal flow.

            We continue to work to elucidate the roles of the various transport processes involved in hydrocarbon film evaporation.  In the future, we intend to focus on the evaporation of mixtures and on the effects of increased temperature and pressure.