"Health check" would be built into pressure transducers, according to a proposal, to enable occasional, rapid, in situ testing of the transducers between normal pressure-measurement operations. The health check would include relatively simple devices that, upon command, would provide known stimuli to the transducers. The responses of the pressure transducers to these stimuli would be analyzed to quantify (at least approximately) deviations from the responses expected from previous rigorous calibrations. On the basis of such an analysis, a given pressure transducer could be removed from service, rigorously recalibrated, or continued in use with corrections applied for calibration drift. The use of the health check could provide timely warnings of pressure-transducer malfunctions and make it possible to retain confidence in the calibrations of pressure transducers while reducing the frequency with which they are replaced or subjected to full laboratory recalibration.
A pressure-transducer health check would typically include (1) a circuit that would, on command, insert a shunt calibration resistor in a bridge circuit within which the pressure transducer normally functions, (2) a device that would compress or expand a known amount of trapped gas to effect an increase or decrease of input pressure, and (3) optionally, a gas-temperature transducer. The health check must be subjected to an initial rigorous calibration along with the pressure transducer.
During the operation of the health check, the change in the output voltage of the bridge circuit occasioned by the connection or disconnection of the shunt resistor would be measured and compared with the corresponding change in voltage measured during the rigorous calibration. Any difference between these voltage changes would be attributed to a change in sensitivity of the pressure transducer or, equivalently, in the amplification in its signal-processing circuitry.
Figure 1 depicts the compression/expansion portion of the health check device installed on the input tube of a typical pressure transducer. To perform a health check, first the pressure transducer would be exposed to ambient pressure, then a ball valve would be actuated to trap gas in the pressure transducer, then a piston would be actuated by an electromagnet to change the volume of the trapped gas. A first pressure-transducer reading would be taken before the piston stroke; a second pressure-transducer reading would be taken long enough after the stroke that any measurable transient heating from compression or transient cooling from rarefaction would have dissipated, but not so long after actuation that the ambient temperature would have changed. By use of Boyle's law, the actual pressure after the piston stroke could be calculated from the ambient pressure and the ratio between the volumes, which would have been determined in the rigorous calibration. Pressure-transducer readings taken over several cycles of exposure to ambient pressure, compression, and rarefaction could be analyzed to determine any deviations from calibration and to characterize the response of the pressure transducer with respect to repeatability, hysteresis, and linearity.
The inclusion of a gas-temperature transducer would make it possible to extract additional information from the decay of the temperature transient that follows the piston stroke (see Figure 2). One would measure the temperature at intervals much smaller than the characteristic time of the temperature transient and would then attempt to fit the measured temperatures to a decaying exponential.
The reason for attempting this fit is that to a first approximation, the temperature of the gas would decay exponentially toward the ambient temperature with a characteristic time
t = mcv/hA,
where m is the mass of the trapped gas, cv is the specific heat of the gas at constant volume, h is a heat-transfer coefficient, and A is the total area of all surfaces that enclose the trapped gas. The value of cv is known and the values of h and A would be determined during the rigorous calibration. Fitting the temperature measurements to an exponential would provide the value of t, from which one could calculate m.
Using m, the known volume of the container, and the ideal-gas law or any other equation of state that is appropriate, one could calculate the pressure of the trapped gas as a function of temperature. This calculation would provide an additional calibration of the pressure transducer against the temperature transducer, making it possible to use the temperature transducer to perform an additional health check.
If either the pressure or the temperature transducer were malfunctioning and it could be determined which was malfunctioning, then by use of the ideal-gas law, one could still estimate the pressure or temperature from the temperature or pressure reading, respectively. As another health check, if either the temperature or the pressure measurements during the transient were to fit a decaying exponential poorly, then the temperature or pressure transducer, respectively, could be assumed to be responding nonlinearly and malfunctioning. As yet another health check, temperature and pressure readings could be compared through the equation of state to determine whether there was an offset error in the output of one of the transducers (though it would not be possible to determine which had the offset).
This work was done by Richard T. Deyoe, Pedro Medelius, Christopher Immer, Stanley Starr, and Anthony Eckhoff of Dynacs Engineering Co., Inc., for Kennedy Space Center. For further information, access the Technical Support Package (TSP) free on-line at www.nasatech.com/tsp under the Physical Sciences category.
This invention is owned by NASA, and a patent application has been filed. Inquiries concerning nonexclusive or exclusive license for its commercial development should be addressed to the Technology Programs and Commercialization Office, Kennedy Space Center, (321) 867-4879. Refer to KSC-12139/12077.