Article révisé par les pairs
Résumé : The SODI-DCMIX thermodiffusion series experiments are part of the fluid research program carried out by the European Space Agency on board of the International Space Station (ISS). In particular, DCIMIX2/3 were conducted in the past inside the Microgravity Science Glovebox in the US Laboratory. Due to the physical nature of the processes implied, these kind of runs were very long and particularly delicate because the low vibratory limit requirements must be maintained for hours. This restrictive condition not always is achieved, therefore, an accurate surveillance of the acceleration levels along the different experiments is necessary, to ensure a correct interpretation of the experimental results. This work analyzes onsite vibrational environment of DCMIX2/3 covering the periods in which the experiments were going on. To do so, acceleration signals only coming from the es03 sensor, nearest to the experimental equipment and located in the Glovebox, were downloaded from the PIMS NASA website. To be as precise as possible the signals have always been treated minute by minute. To detect the transient disturbances along the experiments, several warnings were considered. First, 1 min RMS values, for the three acceleration components were evaluated, in time and in frequency domain. Additional information was obtained by plotting the power spectral densities of the signals, PSD, and their spectrogram with the aim of characterizing long periods of acceleration data. Due to great influence of low frequencies in this type of experiments, the Frequency Factor Index, FFI, was evaluated each minute. Complementary, the spectral entropy evolution was proposed as a fast new indicator of external perturbations. It has been found a good correlation between the spectrogram, temporal RMS and spectral entropy. Finally, a graphic representation of the points associated to the 1-min RMS values in one-third-octave frequency intervals which exceed the ISS limit curve requirements, was considered as a new and easy strategy for depicting the warnings that recognize the main disturbances along the experiment.