Failure-Mode-A "On" RADMON

Introduction

Data Selection and Processing

For each orbit, I applied the current RADMON logic to the EPHIN data from the P4GM, P41GM, and E1300 coincidence channels to determine times when the RADMON process would identify a high-radiation environment. This provides a basis for comparison with an updated set of channels and limits relevant to the FMA-On configuration.

Reinhold Müller-Mellin provided guidance on how to approximate the response of the FMA-On configuration the E150 and E1300 coincidence data based on the historical record. There are two major effects to take into account when combining the electron, proton and helium rates. First, the overall geometric factor for the EPHIN is changed since the central segment of detector A is no longer providing a constraint on the particle direction. The front aperture is defined instead by the forward opening of the detector G (anti-coincidence detector) scintillator. This increases the E150(small), P4GM and H4GM channel geometric factors by 6.88; the E150(large) geometric factor would no longer be relevant but the historic data can be scaled by a factor of 0.23. Second, the detector B noise counts will contribute to the new E150 coincidence counts. This noise contribution can be approximated by subtracting appropriately scaled electron, proton, helium, and integral channel coincidence rates from the detector B00 rate. Similarly, a modified E1300 coincidence channel was produced by scaling the E1300 data and adding the P25GM and H25GM data. The RADMON logic was applied using these two modified coincidence channels, with a E150 threshold of 400000 counts/sample and a E1300 threshold of 1000 counts/sample. The following table lists the first time that RADMON would declare a high-radiation environment for orbits where either the current or the new system triggered.

The "RADMON Trip" and "FMA-On RADMON Trip" columns give the first time in the orbit that triggered the RADMON logic (time in seconds sine 1998.0). The "Cause" columns list the monitored channels that would have caused the trigger. The difference column give the difference in time between the two high-radiation triggers, with a positive value indicating that the new case triggered later, a value of "Missing" indicating the new case did not generate a RADMON trigger and "Extra" indication the new case triggered when the current logic did not. The orbit number links go to plots like the one shown here for orbit 841.

Figure 1: (Top) P4GM (black) and the modified E150 (red) counts/sample as a function of time for orbit 841 during the times when RADMON was scheduled to be enabled. Horizontal dashed lines indicate RADMON threshold used for the channel. Vertical dashed lines indicate the first RADMON trigger time of the orbit (black: current, red: new). (Bottom) Similar to top panel but for E1300 and the modified E1300 counts/sample. Only one threshold appears since the same one was used for both channels.

Discussion

The orbits where the new RADMON trigger is delayed or simply missed should be analyzed by the ACIS team for added proton fluence to access the risk to the instrument. I'll note that I have not screened these data for times when SCS-107 may have already been activated from the ground due to low-energy proton fluence.

The extra triggers on orbits 235, 312, 375, 585, and 686 are all associated with times of solar events. The first four of these occur on the tail of an event at exit from the radiation zone and would tend to delay the time we might be able to resume science operations. The last of these occurs at the start of an event but just on the way into the radiation zone; this could just be considered an advanced trigger. The remaining extra triggers are all associated with radiation zone entry and their time relative to the planned RADMON disable are given in the table below. Operating at higher temperatures might make these more likely due to increasing the noise-level on detector B. It may be possible to avoid unwanted RADMON triggers from the radiation zone entries by modifying the entry pad times.

The unwanted RADMON triggers just prior to radiation-zone entries could also be avoided by setting the E150 threshold higher. This would come at the expense of longer delays before reaching the safing threshold or additional missed safing altogether.

Update: E150 threshold of 800000

Lowering the E150 threshold could potentially pick up some of the missing triggers and reduce or eliminate the delay on the late triggers but will come at the cost of additional unnecessary triggers. The table below lists the comparison results for three FMA-On E150 threshold levels: 200000, 400000, and 800000 counts/sample. The total number gives the number of orbits in which either the current monitoring strategy or the FMA-On strategy with the listed E150 threshold and an E1300 threshold of 1000 count/sample would trigger the RADMON process. The other columns break the total number down into the types of differences between the triggers. There is no advantage to using the lower E150 threshold; any gains are overwhelmed but the cost of the additional unnecessary triggers.

Last modified: Fri Sep 19 14:40:47 EDT 2008

Dr. Michael Juda
Harvard-Smithsonian Center for Astrophysics
60 Garden Street, Mail Stop 70
Cambridge, MA 02138, USA
Ph.: (617) 495-7062
Fax: (617) 495-7356
E-mail: mjuda@cfa.harvard.edu