Appendix C
FILM-BADGE DATA, CALIBRATION. AND ESTIMATES OF ERRORS
The nominally 24-hour gamma doscs for the individual film badges aboard the three target ships for both
shots are presented in Tables C.1 through C.19. The locations of the film-badge stations in the various
compartments or areas are presented in Figures C.1 through C.19.
€.1
CALIBRATION
Calibration expusures of film badges were made by TU~6 on their calibration range at EPG, using Co™
sources of knownstrength at various distances and for various exposure times.
Calculated doses were
checked by means of a Victoreen r-meter.
At EPG, the density of the developed film was read by means
of an Eherline-Angus densitometer which gave digital average-density readings for a fixed BAe bv ue inch
area of the film originally under the leud strip. A film-density-versus-dose plot. used for prelimvnary
results showed that there was considerable scatter in the data about the interim calibration curve.
Because damage to the film emulsion— such as pinholes. scratches. waterspots. and the like — would
increase light transmission. all films were reread at NRDL, using a Macbeth-Ansco densitometer which
permitted scanmng |,-:nch-diameter areas. in order tc find the maximumdensityof the film originally
under the lead strip. Standard density wedges were used frequently to check the calibration of the densitonjeter.
According to Reference 16. characteristic curves of film density versus dose for gamma ravs can be
obtained with beta-ray plaques calibrated with film to indicate an equivalent gamma ray exposure.
A
group of sources with several levels of activity will allow a complete curve to be reproduced in a short
period of time. The required activity is low and sources equivalent to manycuries of gamma rays can be
used directh in the laboratory without need for elaborate shielding.
Sr*®-y*° beta-ray sources were used
to establish the shape of the characteristic curve for the film used bythis project. The characteristic
curve for Sr“-¥" sources was then normalized to give a good fit with various calibration points obtained
bv use of Co® sources both at NRDL and EPG. This normalized characteristic curve was used as the final
calibration curve from which the film-badge doses presented in this report were determined. Only the re-
sults from the high-range film ‘DuPont $34) are presented. because manyinconsistencies were observed
between the results from the low- and high-range films (in the same badge) that were supposedly exposed
to identical doses.
C.2
ESTIMATES OF ERROR
Pairs of film badges were mounted at all stations. except that four badges were used at the GITR stations. In order to investigate random errors (not bias). the percentage difference in dose for each filmbadge pair was calculated. For Shot Wahoo data. the average percent difference for 276 film-badu. pairs
was 7.7 = 0.5 percent and the median value was 5.4 percent. For Shot Umbrella data. the average percent
difference for 311 film-badge pairs was 2.3 + 0.1 percent and the median value was 1.6 percent. The lower
values for the Shot Umbrella data reflect improved handling and processing of the film badges.
The standard errors of the film-badge dose averages, expressed as percentages of the average dose in
a compartment, are shown in Table C.20. These percentage standard errors were obtained from the ex-
pression: 100 Cz x n&° y/n (n-1)X7] 12. where x is the individual film-badge dose, n is the number of
film badges, and X is the average dose in the compartment.
All calibration films which had been exposed to known-strength Co™ sources (both at NRDL and EPG)
were usec to investigate the differences between the ‘‘actual” doses, i.e., calculated or measured on the
calibration range, and the “assigned” doses (based upon use of film densities and the calibration curve
discussed in Section C.1). The absolute magnitudes of the difference between the two doses varied from
0 to 32 percent of the assigned dose and had an average value of 7 percent in the 10-to-1,000-r dose range
99