Propellant Mass Gauging in a Spherical Tank under Micro-Gravity Conditions Using Capacitance Plate Arrays and Machine...

Area of Science:

• Aerospace Engineering
• Applied Physics
• Sensor Technology

Background:

• Propellant mass gauging in micro-gravity is critical for spacecraft missions.
• Unpredictable fuel slosh and shape variations complicate traditional gauging methods.
• Capacitance sensors show promise but require detailed analysis of fuel body dynamics.

Purpose of the Study:

• To investigate the impact of varied propellant fill types and positions on capacitance sensor accuracy.
• To compare the performance of curve-fitting and machine learning approaches for propellant gauging.
• To identify the most effective method for precise fuel content estimation in micro-gravity.

Main Methods:

• Simulating and analyzing capacitance sensor responses for annular, core-annular, and stratified propellant fills.
• Evaluating multiple curve-fitting algorithms for propellant mass estimation.
• Implementing and testing a machine learning-based model for fuel gauging.

Main Results:

• The position and shape of the propellant significantly affect capacitance sensor readings.
• Curve-fitting methods show limitations in accuracy across diverse fuel configurations.
• The machine learning approach demonstrated superior performance in estimating propellant mass.

Conclusions:

• Machine learning offers a robust solution for propellant mass gauging under micro-gravity.
• Accurate fuel estimation is achievable despite complex fuel body dynamics.
• This research advances critical technologies for long-duration spaceflight and orbital operations.