Plasma heating methods for space propulsion

The future of space explorations will not register a big evolution until new families of engines will be qualified. High specific impulse has been achieved with ion thrusters but significant improvements are still waited such as high thrust, variable impulse and long lifetime. Starting with the analysis of a new phenomenon, the helicon double layer, the thesis investigates new space propulsion techniques, principally testing and developing computer simulations for the analysis of thrusters performances and diagnostic devices for their laboratory tests. The helicon Double Layer (DL) is a narrow potential jump, recently measured at the open end of a helicon source, able to produce a supersonic ion beam. From an extent bibliographic review the main characteristics of the phenomenon are: constant axial magnetic field into the source tube (>4.5 10-3 T) which diverges in the external chamber, low neutral pressure (<0.4Pa), ion beam velocity which is two times the ions sound speed, potential jump lower then 20V and source walls somehow able to charge with respect the plasma potential. Utilizing the Object Oriented Particle In-cell (OOPIC) software the experimental behaviour is reproduced in cylindrical geometry and with electrostatic approximation. Suggestions to apply the fluid Boltzmann electrons (hybrid PIC) are found in bibliography, this could simplify and make the simulations faster, but tests of the OOPIC nonlinear Poisson solve have revealed critical problems. It appears that the floating source wall is a key point which permits to increase the plasma potential inside the helicon tube. The potential jump is confirmed around 10-20V. The ion beam is measured with velocities around two times the sound speed, as expected, however the ions trajectories seam to have a higher divergence to respect the experimental data. Utilizing a left source wall with constant biased potential it is possible to increase the plasma potential and, consequently, the DL jump. The exhaust flux performances have then been analyzed. Another very promising engine concept is the Variable Specific Impulse Magnetoplasma Rocket (VASIMR) which uses a Ion Cyclotron Resonance Heating (ICRH) section to increase the exhaust velocity and thrust. Again OOPIC could be utilized to reproduce some aspects of this new device. Here we have to deal with the full electromagnetic solve and a complex geometry that make the simulations more difficult. The software parallelization has been tested obtaining a gain in the performances but the calculations remain prohibitive in terms of time and memory. Consequently a new method has to be applied and the most interesting and useful is a wave equation solve integrated with a Monte-Carlo particles' trajectories calculator, to have a full self-consistent model. The model is constructed starting from fundamental simplified hypothesis: cold plasma theory, negligible electric field along the magnetic field lines, time and poloidal Fourier expansions. The particle model is not a PIC but a trajectory predictor which propagates the particles from a given starting position and distribution considering the external and antenna fields under steady state hypothesis. From the trajectories the plasma density distribution can be constructed, which is the main input of the partial differences wave solve. The first version of the software has been written and its test and utilization is started. Following the interesting experiments done by Charles our CISAS Space Propulsion Team is developing a new device to test a micro engine for satellite applying the Double Layer phenomenon. This new experiment utilizes a reduced scale helicon source. The contribution of the thesis is the design and characterization of probes for the plasma discharge diagnostic. After have analyzed the basic physics of Langmuir probes and retarding field energy analyzers, PIC simulations have been performed to verify the particles behaviour for our helicon source geometry. An innovative RFEA model is here proposed. It is composed of successive layers of conducting and dielectric rings inserted in a grounded stainless steel tube. The main parameter of a RFEA is the grid to grid distance which must be of the order of the Debye length and as little as possible and it has been dimensioned here around 0.5 mm.