FAAM CORE Rosemount/Goodrich type 102 Total Air Temperature sensors, deiced and non-deiced models
FAAM uses two Goodrich/Rosemount type 102 probes, one with a deicing facility (deiced) and one without (non-deiced). Both sensors employ similar inlets to draw flow across an open-wire platinum loom, using inlets that have been designed to minimise water and particle ingress, as well as minimising interaction of the air with the inlet walls. The two sensors differ slightly in the following respects:
The deiced sensor contains a heater that is operated by the flight manager only during periods when icing is suspected. A correction is automatically applied to the data during these periods to account for the effect of deicing heat on the ambient measurement.
The type b deiced inlet is slightly larger and has a much higher airflow through the sensor to minimise the impact of deicing heat on the ambient air sample. As a result the scoop area on the front of the inlet is also greater on the deiced inlet.
The report from the first Temperature working group is available here:
The following document details the probes fitted to the FAAM aircraft, and is available from Goodrich
Data from both probes are reported at 32Hz. FAAM processing simply applies standard calibration corrections (including dynamic heating, recovery factors etc) to the measured resistance of the probe, and reports this as temperature. Owing to their design, the probes exhibit a relatively complex time response to changes in ambient temperature, comprising two distinct components. The platinum wire itself responds relatively quickly, but is affected by conduction from the loom supports throughout its length. Overall, 1s is FAAM's estimate for response time of the probes as per the standard processing.
Corrections have been developed to separate these time response components and allow extraction of fast-response data, allowing time responses of 150ms to be accessed. This does not form part of the core processing, for further details please contact FAAM.
Overall calibration uncertainties are of the order of ±0.3K for this probe.
|Deiced (right and) non-deiced housings||probe element||probe assembly|