Hubble Presentations
Induced Radioactivities of Returned Hubble Space Telescope Parts as Indicators of Radiation Exposure to the Spacecraft
Alan R. Smith / Lawrence Berkeley Laboratory
Donna L. Hurley / Lawrence Berkeley Laboratory
Richard J. McDonald / Lawrence Berkeley Laboratory

Abstract

We present a summary of results obtained at the Lawrence Berkeley Laboratory's Low Background Facility on the (mainly) proton-induced radioactivities in several parts from the Hubble Space Telescope (HST) modules that were returned to Earth in the recent repair mission. These induced-activity measurements permit us to estimate the integrated radiation exposure to the telescope during its 3 1/2 years in orbit. The estimates are based on comparisons of results from our very small data set with the much more extensive analysis done for the LDEF satellite mission. The analysis of additional samples in combination with relevant mission parameters will greatly improve confidence in these preliminary estimates of the HST radiation exposure.

Much information of value to future spaceflights was gained from radiometric analysis of materials obtained from the Long Duration Exposure Facility (LDEF) satellite after its return to Earth's surface in early 1990. In that mission a limited number of special material samples were on board for the express purpose of post-flight radiometric analysis, which results would permit independent estimates for the integrated exposure of the satellite to energetic nuclear particles from the solar wind and galactic cosmic rays. Of equal, or perhaps greater value than results obtained from these "intentional samples", were results obtained from radiometric analysis of spacecraft parts, including items made of aluminum alloy, titanium alloy, stainless steel, and lead. (Note that a number of small but very sensitive integrating passive nuclear radiation dosimeters were also onboard the LDEF.)

The radiometric analysis of LDEF materials has been accomplished over the past several years through a multi-laboratory collaboration (Ref. 1,2,3) headed by Dr. Thomas Parnell of the Marshall Space Flight Center (MSFC), Huntsville, Alabama. Since the radioactivities to be measured were typically at very weak intensities, much of this work could only be done with heavily shielded detector systems in which the "no-sample" or background (BKG) response is two to three orders of magnitude lower than the "unshielded in the laboratory" response -- the so-called "low-background" systems. The detector system of choice for such measurements (used by all the participants in this program) is the high-resolution high-purity Ge-crystal gamma-ray spectometer. Our Low Background Facilities (LBF) at the Lawrence Berkeley Laboratory (LBL) have played a major role in this program (see contributions or A.R. Smith and D.L. Hurley in Ref. 1,2,3), mainly as a consequence of the exceptional low-level counting capability existing at the underground Oroville Facility: a gamma-ray spectrometer system called "MERLIN II", which has both the highest sensitivity and the lowest BKG among the systems used by the participants in this program. And additional background reduction factor of 10 to 100 has been achieved with the MERLIN detector.

Hardware returned from the Hubble Space Telescope (HST) Repair Mission constitutes another potential treasure trove of materials to be analyzed for mission-induced radioactivities. Given the opportunity to analyse samples of aluminum and/or stainless steel, these HST activities can be compared to those measured in LDEF samples, and because of the comprehensive analysis done for the LDEF mission, may then be translated into approximate radiation exposures to the HST parts. The missions were of comparable duration: the returned HST items were in orbit for 3 1/2 years, whereas the LDEF voyage lasted 5 1/2 years, most of which time was spent at a similar altitude.

In Autumn 1993 we established contact with Dr. Lee Feinberg of the HST Office at Goddard Space Flight Center (GSFC) to explore the availability of suitable samples for this purpose. Following successful completion of the repair mission, we have so far received and analyzed three items from the returned modules:

  1. Four stainless steel screws from the exterior surface of the Wide Field Planetary Camera (WF/PC) module, weight 11.2 grams.
  2. An INVAR mirror mount from the WF/PC, weight 75 grams.
  3. A handle from the DF224 Module, at an interior location on the HST, weight 245 grams.


The screws and the mirror mount are alloys in which the major element is Fe; the handle is an alloy in which the major element is aluminum. Fe and Al are the major target elements for reactions that produce the two most important radionuclides discussed here (Mn-54 and Na-22), both of which are produced predominantly by energetic proton interactions. Results of the radiometric analyses are summarized here, and are interpreted in terms of radiation exposure by comparison to the LDEF experience.

The Merlin II Detector System

The Lawrence Berkeley Laboratory's Low-Background Facility operates low-level counting installations as two sites: at the LBL Berkeley site and in the underground power plant of the Oroville Dam (a multi-purpose facility of the California Department of Water Resources). At Oroville the 600-ft overhead thickness of bedrock reduces the surface Cosmic Ray intensity by 1000-fold, and is significantly lower than can be achieved at surface sites. The MERLIN II detector is operated at our underground Oroville site, to take advantage of this additional factor in BKG reduction.

The detector is a very large n-type high purity Ge-crystal, of dimensions 80 mm diameter by 85 mm length; its rated "efficiency" is 115%, which actually means it is 115% as effecient as the "standard" NaI(TI) scintillation crystal of dimensions 3" diameter by 3" length, under specific source-detector conditions. The cryogenic vacuum system is constructed of carefully selected low-radioactivity materials. The local shielding that surrounds the detector cavity is built of low-activity lead and copper. The detector cavity is a space 7" x 7" cross section by 18" height,that is shared by the centrally positioned 4" diameter detector cryostat and the various size samples.

The radiometric data are collected in the format of energy spectra which cover the gamma-ray energy range of 10 to 3600 kev, and contain 4000, 9000 or 16000 channels (bins). Our "peakprint" analytic technique utilizes sharp peaks in spectral data that represent total absorption of discrete-energy gamma-rays. The energies of "signature" peaks are used to identify the radionuclides, while their intensities provide data from which radionuclide quantities can be calculated.

Data and Experimental Results

The four screws from the WF/PC module were located on the exterior surface of the module. Each screw head was in space, while the shank and threaded portions were somewhat shielded by the screw head and the material in which it was seated. The screws are stainless steel, 8-32 x 5/8" with special socket heads. Total weight of the 4 screws is 11.2 grams -- a sample of very small mass for this sort of work.

The screws contained measurable amounts of the radionuclides Sc-46, Mn-54, Co-56, Co-57, Co-58, Co-60, determined in a one-week count time with the MERLIN system at Oroville. A plot of (almost) the entire spectrum is shown on Figure 1. The peaks of interest, as well as the BKG peaks, appear as narrow vertical lines rising from a slowly descending smooth continuum.



Figure 2 shows a narrow interval from this spectrum, on which four peaks of interest and one BKG peak are indicated. Detailed characteristics of the data can be seen here, including the shapes of peaks and the nature of the continuum. Table I lists the rates in diagnostic peaks for the above nuclides as observed at counting time. (No decay corrections have been applied to these values.) Uncertainties are expressed as single standard deviations on the actual data.



Table I

Nuclide: Sc-46, Diagnostic Peaks (KeV): 889+1120, WFPC Screws Observed Peak c/min: 0.029 +-0.002, Mirror Mount Observed Peak c/min: 0.029+-.004

Nuclide: Mn-54, Diagnostic Peaks (KeV): 834, WFPC Screws Observed Peak c/min: 0.515 +-0.008, Mirror Mount Observed Peak c/min: 1.24+-.02

Nuclide: Co-56, Diagnostic Peaks (KeV): 847, WFPC Screws Observed Peak c/min: 0.067 +-0.003, Mirror Mount Observed Peak c/min: 0.088+-.005

Nuclide: Co-57, Diagnostic Peaks (KeV): 122, WFPC Screws Observed Peak c/min: 0.962 +-0.001, Mirror Mount Observed Peak c/min: 2.51+-.03

Nuclide: Co-58, Diagnostic Peaks (KeV): 811, WFPC Screws Observed Peak c/min: 0.039 +-0.002, Mirror Mount Observed Peak c/min: 0.148+-.007

Nuclide: Co-60, Diagnostic Peaks (KeV): 1173+1332, WFPC Screws Observed Peak c/min: 0.010 +-0.002, Mirror Mount Observed Peak c/min: 0.041+-.004

All these induced-activity nuclides can be produced by reactions of energetic particles (mostly protons) on the major constituents of stainless steel -- iron and nickel. The dominant component of the activating flux is expected to be solar protons of energies in the range of a few 10's to many 100's of MeV. Activation by the more energetic galactic cosmic ray component, although contributing to the observed radionuclide inventory, is of minor importance for these surficial samples. Galactic cosmic ray activation plays a more dominant role in objects that are well-shielded from space-facing surfaces.

We also note the (unexpected) presence of Y-88. Although the evidence consists of very low-intensity peaks, both major peaks from decay of this radionclide are observed, and with appropriate relative intensities. This nuclide cannot be produced by reactions on the major constituents of stainless steel. Its presence could be explained if the screw contain Y, Zr, Nb, or Mo. Otherwise, it must be presumed to be associated with some exotic contaminant. Additonal information from the HST personnel or contractors may provide a solution to this puzzle.

Useful comparisons can be made between the level of activities measured in the WF/PC screws and those observed in samples from one of the LDEF stainless steel trunnions. (Samples had been taken at various depths in the solid 3.25-in diameter trunnion to provide a depth profile of induced activities.) The most useful is the comparison between activities of 312-day halflife Mn-54, an isotope that is produced mainly by reactions of protons on Fe in steel. The Mn-54 level in the WFPC screws is about three-fold greater than was observed in comparably located LDEF samples (see Table III). The smaller LDEF activities may be due partly to the decaying orbit, in the sense that the activating flux decreased as the satellite moved to lower altitudes during its final year aloft.

The WF/PC mirror mount is roughly an "oblong" shape, about 2 1/2" in maximum length and 2" maximum width, weight 75 grams. It is made from the dimensionally-stable alloy INVAR, which has a composition 64% Fe and 36% Ni. Observed induced activity nuclides are Sc-46, Mn-54, Co-56, Co-57, Co-58, and Co-60, as were seen in the WF/PC screws. The specific activity of Mn-54 is about 1/2 that seen in the screws. There is no evidence of Y-88 in this sample.

One aluminum alloy sample, the DF224 Module handle from an interior location in the HST, has also been analyzed at Oroville. The handle is of welded construction, made from both tubular and plate stock, and is 16" in length. Because of its "unusual" size a special calibration was required to permit conversion of observed peak intensities into absolute activity rates. The only induced-activity radionuclide observed is 2.62 year halflife Na-22, which can be prduced in aluminum by protons with energies greater than about 30 MeV. Most of the other gamma-ray peaks seen in this spectrum (Figure 3) belong to the Th-232 series of natural terrestrial radionuclides. Shown on Table II are the only major peaks that accompany Na-22 decay, and two of the peaks (of many) that are commonly used to measure Th-series concentrations. Although the 511 kev peak is the more intense of the pair, it cannot always be used as diagnostic for Na-22, as it arises from positron annihilation -- a phenomenon common to the decay of many other radionuclides.

Table II

Nuclide: Na-22, Diagnostic Peaks (KeV) 511, DF224 Handle Observed Peak c/min: 2.58+-0.02

Nuclide: Na-22, Diagnostic Peaks (KeV) 1274, DF224 Handle Observed Peak c/min: 0.724+-0.011

Nuclide: Th-232, Diagnostic Peaks (KeV) 583, DF224 Handle Observed Peak c/min: 0.490+-0.010

Nuclide: Th-232, Diagnostic Peaks (KeV) 911, DF224 Handle Observed Peak c/min: 0.414+-0.008

The Th-232 mass concentration in the handle is about 4 ppm, a value that falls toward the high end of the expected range for the Th-content of aluminum alloys.

Again, comparisons can be made with the Na-22 activities in LDEF aluminum alloy samples. The HST sample is from an interior (shielded) location in contrast to the LDEF samples which were all from the satellite's exterior surface. This difference in shielding makes interpretation of the Na-22 activities not as direct as was the case for Mn-54 activities in the stainless steel samples. Comparisons of both Mn-54 and Na-22 activities are shown on Table III, where all activities have been corrected to the times of return to Earth.

Table III

Mn-54 in Stainless, WFPC Screws: 380 pCi/Kg, 1500 Rads, LDEF Trunnion: 83-171 pCi/kg, 500 Rads

Na-22 in Aluminum, DF224 Handle: 250 pCi/Kg, 1000 Rads, LDEF Keel Plate: 86-140 pCi/kg, 500 Rads

Specific activities for the LDEF samples are double-valued, to indicate the magnitude of differences we observed in activation of north-facing and south-facing surface samples on this attitude-stablized satellite. The Hubble Space Telescope, on the other hand, points toward whichever astronomical object is currently being viewed. We use an average of each pair of LDEF activity values to correlate with the radiation dose integrals measured during the satellite's voyage (Ref. 4). The radiation dose to the HST can then be estimated. From these approximations we can make the following comparisons:

  1. The dose integral to the exterior of the HST was about 3 times larger than that delivered to the LDEF exterior, as determined from the induced activity of Mn-54 in stainless steel.
  2. The dose integral to the "interior" HST site was about 2 times larger than that delivered to the LDEF exterior, as determined from the induced activity of Na-22 in aluminum.


These two statements appear to be consistent, but need more evidence from the HST and the environment in which it travelled, in order to provide stronger confirmation (or refutation). High on a list of such items are: more complete details about the samples already analyzed; additional HST samples from locations with known duration that includes relevant experimental data, and comparison of the results of the HST modelling with similar studies done for the LDEF flight. The HST appears to have received a larger dose integral in a shorter time than did the LDEF. The 11-year solar cycle needs to be taken into account, as well as the sudden appearance of the anomalous long-lived low-altitude "Van Allen" radiation belt during the HST mission.

SUMMARY

We have presented results from measurements of induced-activity radionuclides in several parts from Hubble Space Telescope modules that were returned to Earth in the recent successful repair mission. Since the intensities of these radioactivities are quite weak, sophisticated gamma-spectrometry systems operated in very low background environments are required to achieve successful radioassay. The underground Oroville installation of the LBL Low Background Facility is particularly well suited for this type of analysis. It was used for these HST samples, as it had been used previously for many samples from the LDEF mission.

The HST sample results, when compared to the extensive analysis performed on the LDEF mission, are interpreted to provide a preliminary estimation of integrated radiation exposure to the telescope during its first 3 1/2 years in orbit. Our estimate indicates the HST received a larger integrated dose in a shorter time than did the LDEF satellite. An extension of the work reported here can produce a much more accurate estimate for the HST radiation exposure; in addition, this more accurate estimate will add significantly to the validation of using induced-activity measurements for radiation dosimetry purposes. This capability is especially important when no "actual" radiation dosimeters are on board a spacecraft.

Acknowledgements

We are very appreciative of the opportunity to participate in analysis of returned HST parts, and thank the HST team at Goddard MSFC headed by Dr. Frank Ceppolina for making it possible. Special thanks are due Dr. Lee Feinberg, who first of all took our initial request seriously, who then provided us with samples, and without whose support we could not have attended this conference. We are truly indebted to Dr. Al Schultz (CSC/STScI, Baltimore), who got us in touch with the right people at the outset.

Bibliography

  1. Various authors: Space Environments, Ionizing Radiation, in First LDEF Post-Retrieval Symposium, NASA CP-3134, conference proceedings, pp 199- 396, 1991.
  2. Various authors: Space Environments, Ionizing Radiation, in Second LDEF Post-Retrieval Symposium, NASA CP-3139, conference proceedings, pp 67-274, 1992.
  3. Various authors: Space Environments, Ionizing Radiation, in Third LDEF Post-Retrieval Symposium, NASA CP-, conference proceedings, in press, 1993.
  4. A. L. Frank, E. V. Benton, T. W. Armstrong, and B. L. Colborn, Second LDEF Post-Retrieval Symposium, NASA CP-3139, conference proceedings, pp 163-170, 1992.


Figure Captions



This work was supported by the U.S. Department of Energy under contract No. DE-AC03-76SF00098 and by the Hubble Space Telescope Project, Code 442, Goddard Space Flight Center, Greenbelt, MD.



Link back to Hubble Space Telescope Hardware Archive Return to Hubble Space Telescope Hardware Archive System