Hall Thrusters and Erosion (Lifetime) Measurements

The goal of the fellowship research was to develop an in-situ Hall thruster sputter erosion rate sensor via cavity ring-down spectroscopy (CRDS), use that sensor to measure erosion rates of operating Hall thruster and, if possible, implement the sensor at a NASA facility. My work towards developing and implementing a CRDS erosion sensor can be broken down into four parts: 1. Improve the optical sensitivity and robustness of the CRDS sensor 2. Use the CRDS sensor to measure sputtered boron number densities of the SPT-70 thruster at CSU 3. Use the sensor at NASA Glenn to measure erosion of the HiVHAc thruster 4. Use laser-induced fluorescence to measure sputtered boron speed from the SPT-70 1. Based on the CRDS tests performed by my predecessors, two things needed to be improved about the CRDS sensor. First, the detection sensitivity needed to be improved. Second, the degradation of the sensor’s performance in the presence of the thruster plasma needed to be mitigated. The detection sensitivity was improved by greatly increasing the data acquisition rate, implementing a custom AOM driver, and adding data filtering. The result was a factor of ~5 improvement in sensitivity, which was outlined in an IEPC 2011 publication1. Additionally, measures were implemented to decrease the rate at which the sensor performance degraded as the thruster ran. The degradation appeared to be from a mixture of UV solarization of the cavity mirrors and sputtered particles adsorbing to mirror surfaces. These effects were reduced by sealing the volume in front of the cavity mirrors with the exception of a small iris to allow the beam to pass. A small argon purge flow was introduced into this volume to deflect sputtered particles from adhering to the mirror surfaces. During the HiVHAc erosion tests, the sensor was operated for approximately 11 hours without the need to vent the chamber or clean the mirrors, which was sufficient time to obtain the needed data. The degradation counter measures were outlined as part of a IEPC 2013 publication2. 2. A Hall thruster test facility was developed at CSU to allow for testing of an SPT-70 Hall thruster. The facility proved invaluable for troubleshooting the CRDS sensor, and ultimately, obtaining boron erosion measurements for the operating SPT-70, which were outlined in a 2013 JANNAF publication3. To test the accuracy of the erosion measurements, profilometry traces were taken of the thruster channel before and after a ~26 hour thruster test. Unfortunately, the comparison revealed systematic errors which caused the CRDS sensor to consistently underestimates the channel erosion rate. The SPT-70 erosion tests and profilometry comparison is outlined in a journal publication, which is currently under review4. 3. During one of my on-site experiences, I setup the CRDS sensor in NASA Glenn’s VF8 Hall thruster test facility and performed erosion measurements on the High Voltage Hall Accelerator (HiVHAc) engineering model thruster. The test was quite successful and was outlined in the 2013 IEPC publication2. The primary conclusion of the test was the erosion rate trend with operating condition. With many Hall thrusters, the erosion rate was observed to be proportional to I V2 , where I is the thruster current, and V is the thruster voltage. However, the HiVHAc erosion rate was observed to erode proportional to I V, which was consistent with the results of the optical emission spectroscopy tests by Williams et al.5 4. The majority of the discussion thus far has been on the CRDS sensor, which provides measurements of sputtered boron density in the plume of a Hall thruster. However, an erosion rate estimate (i.e. a flux measurement) requires both the number density and speed of the species of interest. The speed of the sputtered boron had been a large source of uncertainty. I used laser-induced fluorescence (LIF) to measure Doppler shifts, and therefore the speed, of boron atoms in the plume of the SPT-70 thruster at CSU. The tests revealed average boron speeds of 3840 +/- 180 m/s, much slower than were expected from the LIF measurements by Tao et al.6 The LIF measurements are given in my thesis and will be presented at the upcoming ICOPS conference.