Solar electric propulsion (EP) is a key technology for human and robotic space missions, and is part of NASA’s vision for expanding human presence beyond low earth orbit. The high specific impulse but at the price of EP enables reductions in propellant mass, of long burn times. Deep space missions require operating times of many 104 hours. Demonstrating that the thruster meets this requirement is a challenge. Multiple life tests of the full mission duration are not practical. The life capability must be demonstrated by combining physics‐based modeling and short duration testing.
JPL developed the CEX2D and CEX3D codes to model erosion of ion accelerator systems in ion engines, a dominant failure mechanism. The codes model a primary ion beamlet, and charge exchange (CEX) ions from the beamlet. Impingement of main, beamlet, and CEX ions on the grids then determine erosion rates. The models predict time‐to‐failure, but key questions include: What is the uncertainty in those estimates? How much margin is needed to account for the uncertainties? Estimating uncertainty in experiments is routine, but the modeling community is still developing techniques for estimating errors. In this talk we discuss the physical processes of ion engine grid erosion, how they are modeled, and methods for quantifying model uncertainty and required life margins.