Application of uncertainty quantification to determine diffusion coefficient in droplet evaporation - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2017

Application of uncertainty quantification to determine diffusion coefficient in droplet evaporation

Résumé

Evaporation of a liquid droplet resting on a heated substrate has been the subject of recent intensive research. It is a complex free-surface advection-diffusion problem where heat and mass transfer processes occur simultaneously. Though the flow field within evaporating drops is largely controlled by surface-tension-induced (Marangoni) flow and the relative thermal conductivities of the substrate and liquid, the main driving force of evaporation is the vapor concentration gradient at the interface, thence the molecular diffusion coefficient of the vapor constitutes one of the most important physical input parameters in any computer simulations of an evaporation process. However, owing to limited resources, it's not feasible, in space experiments, to insert a large instrument allowing a precise measurement of diffusion coefficient. Given the uncertainty in the value of diffusion coefficient in a specific experiment in space, the question then arises as to what value should be used in numerical simulations. The purpose of this work is an attempt to address this concern. Specifically, we construct a continuous function of output using uncertainty qualification (QU) method based on a Kriging-based response surface, which allows to use the numerical results as a black-box with a limited number of inputs and outputs. The sensitivity analysis via Sobol' index is able to quantify the relative importance of each input parameter in determining the response variability. By the Kriging-based response surface, the numerical results for any input model data can be rapidly forecast without further numerical computation. Relevant values of diffusion coefficient can then be estimated by solving an inverse problem which is based on the experimental results and the proposed response surface. The combination of computer simulation and response surface method provides a really useful guide to interpreting experimental results, especially in an uncertain environment. This approach has been applied in a recent drop evaporation in microgravity environment onboard the Chinese scientific satellite SJ-10 launched in 2016.
Fichier non déposé

Dates et versions

hal-01612960 , version 1 (12-10-2017)

Identifiants

  • HAL Id : hal-01612960 , version 1

Citer

Xue Chen, Paul G. Chen, Qiusheng Liu. Application of uncertainty quantification to determine diffusion coefficient in droplet evaporation. 7th International Symposium on Physical Sciences in Space and 25the ELGRA, Oct 2017, Juan-les-Pins, France. ⟨hal-01612960⟩
48 Consultations
2 Téléchargements

Partager

Gmail Facebook X LinkedIn More