Experimental Measurement Of Gas-liquid Interfacial Rayleigh-bénard-marangoni Convection And Mass Transfer

In the gas-liquid and liquid-liquid interfacial mass transfer process, surfacetension gradient and density gradient caused by temperature gradient andconcentration gradient can generate interfacial instability and ultimately induceinterfacial turbulence, which may significantly affect the mass transfer. The interfacialturbulence, also known as interfacial convection induced by surface tension gradientis referred to as Marangoni effect and that induced by density gradient is referred to asRayleigh-Bénard effect. In this desertation, the interfacial turbulence phenomena: theMarangoni effect and the Rayleigh-Bénard effect are investigated.Firstly, the Marangoni and Rayleigh-Bénard interfacial convections on ahorizontal gas-liquid interface are experimentally measured by Particle ImageVelocimetry (PIV). In the experiments, a simulator of horizontal gas-liquid interfacialmass transfer process is established. By making use of PIV, quantitativemeasurements of the velocity distributions are made for the Marangoni andRayleigh-Bénard interfacial convections generated in the liquid phase of a gas-liquidmass transfer of binary systems. The results indicate that Marangoni convection haseffect mainly on the vicinity region of the interface, while Rayleigh-Bénardconvection has effect mainly on the bulk liquid. With both effect of Marangoni andRayleigh-Bénard convections, interfacial liquid is forced by both gravity in verticaldirection and surface tension in horizontal direction, where an obvious roller eddiesare generated.Secondly, Rayleigh convection and its effect on mass transfer are studied byvortex average velocity, characteristic scale and other statistical analyses. The resultsindicate that Rayleigh convection induces obvious turbulent vortexes in the liquidbulk. Mass transfer coefficient and the enhancement factor are estimated and they areindicated that Rayleigh convection promotes surface renewal and intensifies masstransfer significantly. Two new methods are proposed to estimate mass transfercoefficient. One is by calculation with the surface residence time using characteristicscales and surface velocity, and the other is by the concentration distribution predictedby velocity vector with fluid dynamic models. And both results agree with theexperimental data.Thirdly, the thermal Marangoni flow fields in sessile droplets are measured byPIV, which encompass a wide range of droplet volatilities. There are converse thermal Marangoni convectional directions for different droplet evaporation rates. Anapproximate semianalytical solution is using to solve the flow field in the dryingdroplet in the present of Marangoni stress neglecting the effect of convection due totemperature distribution. Different simulation methods are used to predict the thermalMarangoni flow in the drying droplet with and without droplet distortion because ofevaporation.And lastly, solutal Marangoni convection induced by ethanol desorption from asessile ethanol-water bicomponent droplet is investigated experimentally by differentmethods and different measurement plane. A digital microscope qualitativemeasurement shows Marangoni convection sensitivity to surface concentration andviscosity. Micro-PIV quantitatively measures the velocity distribution on the top sheetof the droplet parallel to the glass slide. Velocity vector graphs display Marangoniconvection fluctuant and inordinance. Experimental results indicate that convectionalintensities reduce with the decrease in concentration and with the increase in viscosity.Mass transfer coefficients produce corresponding variations in flow velocities thatindicate Marangoni convection influence on mass transfer with droplet interfacebreakdown and renewal and the interfacial oscillation will inevitably enhance masstransfer efficiency. Velocity measurement in the plane perpendicular to the substrateindicates solutal Marangoni flow in dominant with higher ethanol concentration. Andstatistical analysis of fluctuant velocity shows no obvious period in Marangoni flow.