Measurement of differential proton spectra onboard the space shuttle using a thermoluminescent dosimetry system

Nuclear Instruments and Methods in Physics Research A256 (1987) 393-397 North-Holland, Amsterdam

MEASUREMENT OF DIFFERENTIAL PROTON SPECTRA ONBOARD USING A THERMOLUMINESCENT DOSIMETRY SYSTEM R o b e r t G. R I C H M O N D

1),

G a u t a m D. B A D H W A R

393

THE SPACE

SHUTFLE

1) B e r n a r d C A S H 2), a n d W i l l i a m A T W E L L

3)

1) NASA/Johnson Space Center, Houston, TX 77058, USA 2) Lockheed Engineering and Management Services Company, Houston, TX 77058, USA 3) Rockwell International, Houston, TX 77058, USA

Received 5 August 1986

An experimental technique that permits the extraction of orbit-averaged, differential energy spectra of trapped radiation belt protons using simple passive detectors is described. An inversion technique is used for the data analysis. The basic principle of the described system is the measurement of the energy deposited in six thermoluminescent (TLD) detector assemblies behind various spherical absorbers. The technique has been applied to a detector assembly flown on four Shuttle flights. Although severe restraints were placed on the flight package, the differential energy spectra derived from these measurements are in good agreement with analytical results using a modified trapped proton environment model. The technique shows good promise for measuring the spectra in low inclination orbits where the flux of high energy galactic cosmic ray protons is small. Modifications to the detector assembly to improve the accuracy and to extend the range of the system to higher energies are suggested.

1. Introduction For over two decades the N A S A / J o h n s o n Space Center has developed analytical techniques and measurement systems to provide accurate and timely information on radiation exposures to crewmembers, experiments, equipment, and payloads for the manned space program. The results of the efforts have been documented in the literature [1-3]. Three experiments, viz., a p r o t o n - e l e c t r o n spectrometer on Gemini 11, an a l p h a - p r o t o n spectrometer used on several Apollo flights, and a proton-electron spectrometer flown on the Skylab, provided information on the differential proton energy spectra outside the spacecraft. For low-inclination, low-altitude missions, the bulk of the crew exposures inside the spacecraft come from trapped protons, primarily those located geographically in a region called the South Atlantic Anomaly. Energetic protons from galactic cosmic rays become more important as the inclination of the orbit increases. Trapped electrons generally contribute only to crew dose during extravehicular activities. For the purpose of estimating crew exposures, the differential energy spectra of protons inside the spacecraft have been calculated using analytical models of the radiation environment and the spacecraft shielding geometry. Recently, serious questions have been raised [4] concerning the applicability of the models of the proton environment to both the current time period and the expected Space Station era (1995-2025). It was clear that a measurement of the orbit-averaged differential 0168-9002/87/$03.50 © Elsevier Science Publishers B.V. (North-Holland Physics Publishing Division)

proton spectrum inside the Shuttle was needed. This paper describes a technique to measure proton spectra inside a spacecraft using a passive detector system.

2. Detector design and experimental approach The detector system was developed for use on selected Space Shuttle missions on a "space available" basis. Very severe restraints in terms of dimensions, materials, and weight were placed on the flight package. The outer envelope was limited to 3.5 x 3.5 x 2.0 inch 3. The system could use neither telemetry nor spacecraft power, and was limited to a total weight of 4.0 lb. All materials used in the construction of the system were required to pass flight qualification tests relating to outgassing, flammability, etc. After careful consideration, an approach was selected that used thermoluminescent dosimeters (TLD) centered inside spherical shields of various thicknesses. The detector system components are shown in fig. 1. Stacks of five annealed [3] lithium fluoride (TLD-100) detectors, each 0.125 × 0.125 × 0.035 inch 3, having a total thickness of 1.16 g / c m 2, were mounted into five separate spheres. The entire assembly was enclosed in a Lexan box of 0.375 g / c m 2 thickness (fig. 2). In addition, one stack was mounted outside the box, giving six measurement points. Physical parameters and effective energy cutoff of each component of the system are given in table 1. The detector package was assembled about a week

394

R.G. Richmond eta/. / Measurement of spectra onhoard the space shuttle Table 1 Physical parameters and effective energy cutoff of each compo-

nent of the flight package Sphere no.

1 2 3 4 5

Fig. 1. Experiment package components. (Bar equals 1 inch.) before the flight to minimize the exposure to background radiation. The device was delivered to the spacecraft approximately four days prior to scheduled launch and was actually installed in the spacecraft in the last 24 h. Although the detailed spacecraft geometries varied slightly between the selected missions, the detector package was placed in the same mid-deck locker location. The localized shielding in this locker was essentially the same for the four missions. The configuration described above was flown on four selected Shuttle flights, 51-D and 51-J (both 28.5 ° inclination orbits), 51-F (49.5 ° inclination), and 51-B (57 ° inclination). These lnissions were chosen since they gave the range of expected source terms and ex-

Material

Gelatin Lexan Lexan Aluminum Brass

Thickness (g/cm 2)

0.0001 1.2 2.4 5.4 17.0

Cutoff energy nominal (MeV) Bare

inside Lexan box

5 35 53 71 121

19 41 57 75 123

posures representative of Shuttle flights to date. Premission calculation of doses from the South Atlantic Anomaly and galactic cosmic ray protons showed that the lower inclination orbits were high enough in altitude to produce acceptable doses in all the detectors. A large contribution to total dose at the higher inclinations would be attributable to galactic cosmic ray protons. U p o n completion of the mission, the detector package was returned within forty-eight hours following touchdown. Readout of the TLDs was accomplished within two days of their return on a commercial T L D reader coupled to a desktop microcomputer for glow curve acquisition and analysis, [5]. The main glow peaks were extracted, corrected for reader background and integrated. Adjustments were also made for accumulated, nonflight radiation background and estimated galactic cosmic ray contribution during flight.

3. Data analysis The basic concept of this experiment is that materials of various thicknesses absorb protons below the corresponding cutoff energies (table 1). The protons incident on the T L D stacks are greatly reduced in number with increasing absorber thickness, resulting in lower absorbed energy (dose). For the flight instrument, the cutoff energy varies from essentially zero for the detectors located outside the Lexan box, to approximately 120 MeV for detectors located inside the brass sphere. The spectrum, denoted J ( E ) , inside the spacecraft and incident on the instrument, is further modified by the material surrounding the T L D stacks. The resulting spectrum f(E) is responsible for the energy T deposited in the detector stacks and is given by

Fig. 2. Top view of the flight package. (Bar equals 1 inch.)

where (dE/dx) is the ionization loss in the T L D material (lithium fluoride), and i is the index denoting

R.G. Richmond et al. / Measurement of spectra onboard the space shuttle the particular sphere-stack combination. Particles entering a stack with energy insufficient to penetrate the stack are assumed to deposit all their energy in the stack. This assumption would cause the calculated dose in the finite detectors to be 5-7% lower than that expected from typical point dose calculations. Eq. (1) assumes that losses due to nuclear interactions are negligible. Evaluation of the nuclear interaction probabilities indicated that the effect on measured dose in this experiment was less than four percent. Since eq. (1) is a first order Fredholm integral equation [6], the incident proton spectrum J ( E ) can be obtained by inversion. There are several formal methods of solution. These are generally not applicable for cases such as this, where the deposited energy T is not known analytically. In this experiment, T is known at only six discrete points; therefore, additional constraints are required to make the solution unique. Our approach was to parameterize the form of the spectrum J ( E ) incident on the instrument with less than five parameters, while attempting to accommodate large variations in shape. Although the proton spectrum outside the spacecraft has a very large flux of low energy particles, the spectrum is dramatically modified as it passes through the shielding of the Shuttle. Fig. 3 shows a typical calculated proton spectrum inside the spacecraft for a 28.5 o inclination orbit using the AP-8 proton model [7]. Calculations for other environments (solar maximum, solar minimum, different magnetic field models and epochs) show the same general shape. Fig. 3 shows.that these spectra can be well represented by the Beta function

-- E m ]

Em]

.,~

1

(2)

,

395

where E m is the maximum energy, a, and fl determine the shape, and A is the normalization constant related to the integral flux. The peak energy Ep is determined by E

a

1

Additional functional forms, with the constraint that the number of free parameters be three or less, were evaluated. As an example, the MaxweUian energy distribution function J ( E ) = A E exp( - E / E o ),

(4)

also fits the model calculation well. This function has only two free parameters A and the peak energy E0, and thus has much less flexibility in the form of its shape. The Beta function always gave a better fit than the Maxwellian form. Having satisfied the smoothness criterion on J ( E ) , and using the analytical representation of the kernel d E / d x , eq. (1) can be inverted to calculate J(E). The inversion algorithm minimizes the merit function F ( A , ct, fl, E m ) ~ g i=1 ~

°i

where D, and o~ are the measured dose and the standard deviation, in the measurement for sphere i, respectively, C is a factor to convert energy deposited in MeV to dose in mrad. The parameters A, a, fl and E m are varied to find the global minimum of the merit function by a modification of a technique due to Tyapkin [8]. Due to the limited space available for the package, the spheres were in close proximity to each other, resulting in undesirable mutual shielding. The solid angles subtended by each of the spheres were calculated and the effective transmittance for each sphere combi-

÷ --

se81

A P - 8 MAX. C A L C U L A T E D BETA F U N C T I O N FIT

> ><

1888

I ~4

~*

588

O

I¢1 Z

-~k_..~ ~..~

Oe¢

'''''I'''''I'''''I'''''I'''''I

8

68

128

188 KINETIC

,,,,,i,,,

248

388

368

,,i,,,,,i

428

488

ENERGY ( H E V / N )

Fig. 3. Fit of the beta function compared with the orbit-averaged AP-8 MAX proton differential spectrum.

396

R.G. Richmond et al. / Aleasurenlent oj spectra onhoard the spa
358

/ 380

~j

250

~J

J

[

I l

/ ~ z-

4

/~f/""

200 250

200

300

350

CALCULATED TLD DOSE IN MRAD

Fig. 4. Observed TLD dose versus calculated TLD dose relative to a 1 : l line.

nation was calculated as a function of proton energy . The resulting transmittance was applied to the inversion procedure.

4. R e s u l t s a n d d i s c u s s i o n s

For the purposes of this paper, data from a single, typical 28.5 ° inclination mission are reported and discussed. The mission selected was 51-J. This mission was preferred because of its circular orbit and high altitude (resulting in higher doses and better statistics). Preliminary analysis of data from mission 51-D, also at this inclination, indicates similar results. The doses measured in the individual T L D stacks respond to all incident ionizing radiation. For the 28.5 o inclination 51-J mission described here, the galactic

~, i

cosmic ray protons contribute approximately 4 m r a d / day [9]. Although this component was less than five percent of the total dose for mission 51-J, it was subtracted from each observation, leaving the remainder of the dose attributable for trapped protons. Fig. 4 shows a comparison of the measured and calculated doses. The agreement between the two doses is quite good. The best fit values for which F is the minimum, are A = (3.129 + 0.024) × 105 p r o t o n s / c m 2 day MeV, c~ (2.042 + 0.95), ,8 = (8.799 + 2.83) and E m of 1004.5 + 101 MeV. However, for most practical cases E m can be fixed at 1000 MeV. Fig. 5 is a comparison of the orbit-averaged proton spectrum using the AP-8 M I N model and that obtained by applying the inversion technique to the T E D measurements. The AP-8 is currently the best trapped proton environment model. The calculated proton spec-

3500 4 3808 2500

I

£ m Z o o m

2008 i

+ ++/

+++.

1see -~" / / 1800 "

~ .

+*+++*

*++++++++~~ ++

,

AP-8 MIN

~-

500 8 0

108

280 KINETIC

300 ENERGY

480

500

(MEV)

Fig. 5. Differential proton spectrum vs proton energy for AP-8 MIN, using IGRF 1965, epoch 1964, and the inversion result.

R.G. Richmond et al. / Measurement of spectra onboard the space shuttle

trum used the I G R F 1965 magnetic field model [10], an October 1964 epoch, the appropriate time in the solar cycle and the Shuttle shield mass distribution. Based on the recent work of Konradi et al., [4], this appears to be the most likely calculated proton spectrum inside the locker. The agreement between the two spectra is particularly good considering that both spectra represent absolute flux values and the the AP-8 model is based on data as old as two decades. There is, however, one very significant difference; the trapped radiation model calculated spectrum peaks at 85 MeV whereas the spectrum determined from T L D measurements peaks at 118 MeV. This indicates that our measured spectrum is harder than the trapped proton model calculated spectrum. The fact that a harder proton spectrum than the AP-8 model is required by the observed dose measurements can be seen from the fact the AP-8 model produces a progressively larger difference between calculated dose and observed dose with increasing absorber thicknesses (gelatin to brass). Thus, it is suggested that the differences between the two spectra can be attributed to (1) insufficient knowledge of the shield mass distribution, particularly near the Lexan box, and (2) uncertainties in the trapped environment model. The trapped proton model does not consider the independence of the atmospheric cutoff energy from the magnetic field and assumes that the angular distribution of protons for random orientation of the Orbiter, is isotropic, an assumption also made in eq. (1). The two facts and uncertainty in the shield distribution will certainly affect the calculated spectrum. It should be emphasized that the spectrum estimated using the T L D measurements does not depend on any knowledge of the shield distribution

5. Conclusions An experimental technique that permits the extraction of orbit-averaged, differential proton energy spec-

397

tra inside the Space Shuttle has been developed. This measured spectrum and the one calculated from the AP-8 model show differences that could be due to model deficiencies. Additional experiments, particularly at locations where local shield mass distributions are well known, would be needed to clearly establish the reasons for the differences. It is further suggested that additional detectors with spheres of different absorber thicknesses and cleaner geometries be flown to extend the energy range and improve the accuracy of the estimated spectrum. These detector systems are simple enough to be flown on a routine basis and could provide the necessary data to improve trapped proton model environments.

References [1] R.G. Richmond, Aerospace Medicine 40 (1969) 12. [2] R.G. Richmond, NASA Technical Note TN-D 6695 (1972). [3] R.G. Richmond, K.L. Jones, and B.L. Cash, NASA Technical Paper, in publication (1986). [4] Andrei Konradi, Alva C. Hardy, and William Atwell, submitted to, J. Spacecraft and Rockets (1986). [5] B.L. Cash, and R.G. Richmond, NASA Technical Memorandum, In Publication, (July 1986). [6] F.W. Byron Jr., and Robert W. Fuller, Mathematics of Classical and Quantam Physics, Vol. II (Addison-Wesley, Reading, MA, 1970) pp. 518-563. [7] D.M. Sawyer and J.I. Vette, NSSDC/WDC-A-R&s-76-06, National Space Science Data Center, NASA/GSFC, Greenbelt, MD 20771 (1976). [8] A.A. Tyapldn, Proc. 1960 Ann. Int. Conf. on High Energy Physics at Rochester, NY Interscience, New York, 1960) pp. 138-140. [9] James H. Adams Jr., Naval Research Laboratory, private communication (1984). [10] J.C. Cain, S.J. Hendricks, R.A. Langel and W.V. Hudson, J. Geomagn. Geoelec. 19 (1967) 335.