This thesis deals with problematics of air/fuel mixture diagnosis od older carburetor motors. Especially for motors with more than one cylinder. In this thesis are informations about commercial product for diagnosis, brief informations about function of CV caburetor and best setup of fuel mixture. The next step in this thesis will be design, construction and testing of device for the diagnosis air/fuel mixture in practical use. Device will be able to measure mixture and other values witch can be used for best carburetor setup.

La vaporisation explosive, ou flashing, par dépressurisation rapide du C6F14 au travers d'une vanne à boisseau sphérique est analysée expérimentalement sur une grande plage de surchauffe. Les visualisations rapides montrent un jet s’ouvrant très largement en aval, preuve de l'existence d'un fort gradient de pression. Des mesures locales par sonde optique et par vélocimétrie phase Doppler ont permis de caractériser tailles, vitesses, concentration et flux numérique des gouttes en fonction de la surchauffe. La faible influence du degré de surchauffe sur la remontée en pression et sur les tailles et flux de gouttes suggère que le champ de pression s’adapte, et que par conséquent le liquide est soumis à une surchauffe locale bien plus faible que la surchauffe globale imposée. Différents scenarii sont discutés pour expliquer les observations, dont le fait que les tailles de gouttes sont peu sensibles à la surchauffe. Celui basé sur l’existence d’un front d’ébullition est le plus probable. / Depressurization (flashing) experiments through a ball valve were conducted with C6F14 for a large range of superheat. High-speed imaging shows a rapid and wide expansion of the jet, which evidences strong pressure gradients. Local measurements with phase detection optical probes and phase Doppler velocimetry were used to characterize size, speed, concentration and volumetric flux of drops as a function of superheat. The level of superheat has little influence on the vaporized fraction and on drop size and flux: this suggests that due to strong pressure gradients the liquid sees a much weaker level of superheat than the global superheat imposed on the system. Different scenarii are discussed to explain observations, in particular the fact that drop size remains approximately constant independent of the superheat. A scenario based on the existence of a boiling front seems to be the most consistent.