NASA Logo, National Aeronautics and Space Administration


The prognostic algorithms have been applied to historical pneumatic valve data from the Space Shuttle cryogenic propellant fueling system. The following movie demonstrates the prognostics application:

Demonstration Movie (23.3 MB)

In the demonstration, the first seven valve cycles are taken from historical data for the transfer line chilldown valve. A trend was observed in the valve timing attributed to the progression of valve damage. After these seven cycles, a maintenance action was performed to return the valve to specifications. An external leak at the bottom pneumatic port is the only fault that can explain the trends in valve timing. We determined the growth rate of this leak, and appended the actual data with simulated valve data that extended this failure mode to EOL, which is determined by thresholds in open and close times of the valve.

The demonstration covers a sequence of 38 fuelings, normally taking nine or more hours each. As the valve is cycled during fueling operations, damage accumulates, reflected in the change in valve behavior over time. The prognosis algorithm takes as input the valve position commands and the discrete open and close sensor signals. From these inputs, it estimates the degree of damage over multiple damage modes, and the progression rates of each. The demonstration shows how the prognosis algorithm determines the size of the external leak and its growth rate over time, and the resulting RUL prediction.

The prognostics interface is divided into the following sections.

  1. LH2 System Schematic: This is representative of the top-level interface viewed by a system operator. It shows a system schematic that updates dynamically as a refueling operation is underway. Beside component names are color-coded health indicators that change from green to yellow to red as damage is accumulated and the prognosis algorithm updates its estimate.
  2. LH2 Sensors: This shows plots of the main process variables.
  3. Prognostics: This aspect of the interface accesses the output of the prognostic algorithm. The user selects a component to see the current results of the prognosis algorithm. Here, we highlight the case of the transfer line chilldown valve. On the right side of the component selection is a summary of the current valve health. Within this part of the interface is shown the number of fueling operations the valve has undergone, its current overall health estimate, and its estimated RUL with confidence intervals. Below this information on the left side is shown the RUL plot at the 99% confidence level. The RUL may also be viewed at other confidence levels, with confidence intervals shown, or EOL may be viewed.
  4. Supportive Information: This aspect of the interface shows the information supporting the current predictions. At the top is shown the measured component signals (in this case, valve opening and closing times). At the bottom is shown the estimate of the fault magnitude and wear rate for a selected fault mode.
First Gov logo
NASA Logo -