Effect of reboosting manoeuvres on the determination of the Soret coefficients of DCMIX ternary systems

Abstract

Abstract Experiments for measuring the thermodiffusion and diffusion coefficients in ternary liquid mixtures need to be performed in a gravity free environment to avoid convective flows. If convection can be neglected, the resulting concentration field can be used to fit the analytical solution and subsequently obtain the values of the two Soret coefficients. The so-callled Diffusion Coefficients Measurements in ternary mIXtures (DCMIX) experiments carried out onboard the International Space Station use a parallelepipedic cavity ( 10 × 10 × 5 m m ) called Soret cell, with two opposite heated walls at different temperatures to attain a linear concentration between them. When a reboosting (to keep, for instance, the station in the correct orbit) occurs, the acceleration level increases considerably and, thus, the induced convection can potentially distort the experimental results. In the present study numerical simulations have been carried out using five different reboosting scenarios (reboosting occurring at equally distributed times from beginning up to the end of the thermodiffusive step), and the experimental procedure for measuring the Soret coefficient has been reproduced. Three different mixtures have been considered, corresponding to the DCMIX1, DCMIX2 and DCMIX3 campaigns. It has been found that the thermal Grashof number determines the sensitivity of the system to the reboosting in these three cases. The reboosting maneuvers can strongly affect the concentration field, and therefore, the calculation of the Soret coefficients. The closer to the end of the thermodiffusive step the reboosting occurs, the larger the error is. The DCMIX2 is the most distorted mixture reaching an approximately 100% of error when reboosting acts at the end. But if the reboosting takes place at the beginning, the above-mentioned error is smaller than 1%. Also, the results depend on the method used to measure the concentration field. When the concentration difference averaged at the hot and cold walls method is used, the error is generally lower.