ow,

x?

“uo

~

oe

+

7 ee Ee eT eS

NN hae

Appendix B

GITR CALIBRATION

B.l

BIASED-FIELD CALIBRATIONS

All instruments were initially calibrated at NRDL with Co® sources accurate to within 3 percent.

All

calibrations were made with a standard orientation: the longitudinal axes of the detector and the radiation
beam were parallel. and the electrometer housing faced away from the source. In this orientation, dose
increments of 0.243 mr and 0.243 r were established for the low- and high-range chambers. respectively.

The linearity of the detector had been checked over a wide range of gamma intensities and is shownin
Figures B.J and B.2,

To assure optimum reliability and accuracy in the data each detector was recalibrated in the field,

before and after each shot. with the 120~curie Cs! source installed in the project's instrumentation
waler. This source was standardized to the Co™ sources by means of the Victoreen 70-A r-meter and
various calibrated chambers. To assure maximum reproducibilty of calibration, a jig was fabricated to

ww ©

LAR AABR

control positioning of all detectors in the radiation beam. For personne) protection. the beam was directed
sc ruically through the roof of the trailer. A calibration radiation field of 56.4 r/hr was used for the adjustmet of the detector output~pulse periods to 0.616 and 15.5 seconds for the low-range and the high-range
channels, respectively. The low-range~channel pulse period of 0.016 second (instead of the expected value

of 0.0153 second to give 0.243 mr) compensated for the 0.0005-second recycling time of the circuit.

The

calibration radiation field was too low to require a similar compensation for the high-range chamber.
lt was estimated that all field calibrations were made with a precision of about +2 percent.

Upon re-

calibration following an event, the random shifts in calibration were noted to be about +3 percent. Evaluation of all phases of instrument operation indicated that the relative precision of almost all detectors
was about +7 percent throughout an event.

However. it was known that the detector orientation used for

calibration, and chosen because it assured reproducibility. biased the results because of the nonuniform
cirectional response of the detectors. Figures B.3 and B.4 showthe results of pretest studies of energy
response and directional response characteristics.

E..

CORRECTIONS FOR CALIBRATION BIAS

After Operation Hardtack, a more-extenSive investigation of GITR directional characteristics as a
function of energy was undertaken at NRDL for three conditions: (1) detector in the aluminum jacket.
representing interior GITR stations, (2) detector inside the aluminum drum, representing exterior GITR

Stations; and (3) detector mounted inside the recorder case. Figure B.5 and Tables B.1 through B.6 show
the results in relationship to the biased field-calibration condition. The actual responses of the shielded
detectars (simulating the station mountings) to the several monoenergetic gamma-radiation beams for
Various detector orientations were divided by the responses of the unshielded detectors to Csradiation
beamed at the top of the detector (the biased field-calibration re sponses).

The directional responses indicated above were used to calculate integrated responses to four idealized
Tahation-source geometries: (1) horizontal radiation incidence, simulating remote pretransit radiation,
‘2; hemispherical] radiation source above station, simulating the transit phase; (3) spherical radiation
Source around station, simulating interior stations affected by radiation from both the overhead decks and

ac-acent water; and (4) radiation source presenting solid angle of 1.7-7 steradians belowstation, simulating

exterior stations exposed only to contaminated decks and/or adjacent water.

Figures B.6 through B.9 show

hese integrated responses in relationship to the biased field-calibration condition. However, these values
4pply only for monoenergetic radiation sources.
In the absence of measured gamma-energyspectra for these shots, the sensitivity of calculated correchon factors to various assumed spectra was investigated. Six un-degraded energy spectra for various
mes after fission were considered: 9-second and 6.8-minute spectra from Reference 10; a 31-minute
Spectrum from Reference 12; 1.1- and 5.2-hour spectra from Reference 13; and a 9-hour spectrum from

87

Select target paragraph3