Étude de l'usinabilité par électroérosion des alliages d'aluminium à faible coefficient de dilatation thermique pour application spatiale

Abstract

6061-T6 aluminum alloy, is presently used in the manufacturing of on-board radio frequency equipment of telecom satellite. , The latter are exposed, on a daily basis, to temperature variations of around 300 °C, thereby affecting the performance of radio-frequency equipment due to the severe dilatation and contraction (high thermal expansion coefficient) of 6061-T6 Al alloy. In order to minimize this effect, the replacement of this aluminum alloy by new alloys with low coefficient of thermal expansion is studied. On the other hand, the lack of knowledge about the machinability by die-sinking electrical discharge machining of those alloys appears. The purpose of this project if, first, to validate the exact mechanical properties of three candidate materials which are the ALM-450 from Adamco (USA), the MS43 from AMT Technologies (Germany) and the MS61 from the same place. From those three materials, the one that presents the best performances will be selected as the potential successor to the 6061-T6. The optimal material selected, the machinability optimization process will ensure that the material can be machined in a spec-effective and a cost-effective way. From ten parameters that guide the electrical discharge machining (EDM) process, five appear, from the completion of a sixteen-run Plackett-Burman design of experiments (12 runs + 4 central points) to significantly affect the material removal rate, the electrode’s radius formation rate, the arithmetical mean deviation of the profile and maximum height of the profile. Those five parameters are the duration of an electrical discharge, the delay between two discharges, the discharge current, the delay between two series of discharges and the voltage. Therefore, a second design of experiments, consisting in a half factorial plan with sixteen runs plus two central points, was developed and carried out. This design of experiment resulted in the generation of four predictive models for the four answers studied in the Plackett-Burman design of experiments. Finally, an optimal methodology is proposed in order to optimize the machining program initially generated by the EDM-Controller.

Date	4 Jun 2018
Original language	French
Awarding Institution	École de technologie supérieure
Supervisor	Mohammad Jahazi (Supervisor) & Victor Songmene (Co-supervisor)