NASA Dryden Flight Research Center has ruggedized and qualified an inexpensive commercial off-the-shelf (COTS) oxygen sensor that accurately and reliably aids assessment, in flight tests, of the hazards associated with propulsion systems supplemented by oxidizers. Future flight-test vehicles continue to rely on such energetic propellants as liquid and/or gaseous oxygen and hydrogen for purposes of demonstration because these propellants deliver high specific impulses.

Figure 1. An Oxygen Sensor and Its Temperature Controller were electrically connected and packaged separately.

In preparation for flight testing in the Linear Aerospike SR-71 Experiment (LASRE) program, commercial sensors intended originally for medical and automotive application were qualified for flight. After a rigorous process of qualification and calibration, the sensors proved extremely reliable and repeatable, and gave engineers the confidence they needed to assess the hazard of flammability during test operations. These sensors have now been added to the Hyper-X program and are scheduled for use in future flight-test projects. (The Hyper-X is a proposed experimental hydrogen-fueled hypersonic aircraft.)

The sensor in question is a commercial-grade miniature fuel cell — in other words, a small transducer that converts chemical energy into electrical energy. Like all electrochemical transducers, a sensor of this type contains an anode, a cathode, and an electrolyte. At the cathode, during operation, oxygen molecules are reduced to hydroxyl ions. The anode is made of lead; during operation, lead in the anode becomes oxidized to lead oxide, producing two electrons for each atom of lead thus consumed. This particular transducer has cross-sensitivity only to other gases (e.g., halogens) that oxidize leads.

The sensor life expectancy has been estimated by the vendor to be approximately 2.4 years. However, because of the criticality of these sensors, it was decided that in the LASRE project, each such sensor will be replaced when its response deteriorates or it reaches a calendar life of 1.5 years.

Figure 2. The Uncertainty of Measurement as a function of pressure altitude was computed for various concentrations of oxygen for a 99-percent confidence level after correction.

In the laboratory calibration tests, the sensor was actively maintained at a temperature ≈115 °F ≈46 °C) to prevent freezing of the electrolyte (which is aqueous) as well as to maintain constant performance of the transducers throughout the flight envelope. Each sensor was inserted into an aluminum tube shell for structural support. Attached to the shell was a thin-film resistive heating element, a power transistor, and two resistance temperature devices (RTDs). In addition, thermal-insulation material was wrapped around the shell. The power to the heating element was regulated by a heater-controller circuit in a box separate from the sensor (see Figure 1), in response to temperature feedback from the first RTD. The second RTD was used for monitoring and to provide a temperature reading as part of the sensor output.

At the initial application of power, the heating element was activated until the transducer assembly reached the set temperature. Once the set temperature was reached, the power to the heating element was cycled on the basis of the reading of the first RTD. During application of power to the heating element, the heater controller draws an electrical current <1/2 A.

In initial flight tests, it was found that two-point calibrations were not satisfactory for the use of these instruments in flight tests. A calibration test was designed to enable the attainment of the following three test objectives: (1) characterize sensor behavior with increasing altitude (decreasing partial pressure), (2) perform a statistical analysis of laboratory-test results (which analysis would give a quantitative measure of the level of confidence in the mean of the untested sensors), and (3) quantify the uncertainty of the sensor measurements with decreasing partial pressure. Several of the sensors were installed in a small test chamber and exposed to the following range of O2 concentrations (volume percentages) with nitrogen balances: 0, 1.02, 2.06, 5.05, 10, and 21. Each sensor was cycled through a range of pressure altitude from sea level up to 80 kft (24 km) and then back down to sea level. During this cycle, the sensor output was measured at intervals of 10 kft (3 km), using the concentrations listed above.

Because of a nonlinear response that occurred during operation above 30 kft (9 km), an altitude correction factor was devised. This correction factor was computed in the following procedure: The ratio between (1) the true partial pressure of oxygen (as determined by use of the calibration gases) and (2) the transducer output was determined for each altitude. This ratio was then plotted against altitude, and the data of this plot were fitted with a fifth-order curve. During the flights, the correction factor was computed from the fifth-order curve. This correction factor was then multiplied with the linearly calibrated transducer output to produce an altitude-corrected partial-pressure reading. Normalization of this value by the static-pressure measurement yielded an output volume fraction. On the basis of the t distribution of the corrected readings, the uncertainty was found to remain less than one percent for altitudes less than 60 kft (18 km), as shown in Figure 2.

Twelve of the sensors and their respective thermal controllers were integrated into the LASRE flight-test fixture and were flight-tested out to a speed of mach 1.6 and an altitude of 52 kft (16 km). Eight of the original twelve sensors remained on board after removal of the LASRE for some follow-on flight tests at speeds up to mach 3.03, altitudes to 73 kft (22 km), and exposure to internal temperatures >130 °F (>54 °C). Even though the temperatures rose above the temperature-controller limits and the manufacturer's specifications for operation, the oxygen sensors exhibited no significant drift.

Four sensor suites are to be incorporated into the Hyper-X airplane for safety during the flight test in which the Hyper-X will be carried by a B-52 airplane. These sensors will be used to verify the desired and expected chemical inertness (more specifically, the lack of oxidizing gas) in the interior of the vehicle. They are being calibrated in much the same manner as in the LASRE program, but dynamic-response and sensor-recovery tests will be performed in addition, for comparison with results of tests in the boost and free phases of the flight of the Hyper-X.

This work was done by Neal Hass, Michele Jarvis, and Kim Ennix of ryden Flight Research Center DRC-00-04 .