Field Precision title

Target optimization and forward dose calculation for a pulsed radiographic facility (2)

The GamBet calculations discussed in another note generated a bremsstrahlung distribution of photons from a 6.0 MeV, 1.0 A point electron beam striking a tunsgsten target of thickness 0.8839 mm. The escape file from the calculation was filtered to include only photons at z = 20.0 mm within 10.0 mm of the axis. The photon parameters were recorded in the file TARGET6004.SRC. In this post, we shall estimate the dose resulting from the photons in a detector a distance 1000.0 mm from the target face. As is often the case with numerical calculations, we will see the benefits of using scaling relationships instead of a literal approach.

We direct the photon distribution on an 3.0 mm thick aluminum phantom with the front face 20.0 mm from the target face. The mesh defined by DOSE60.MIN has a single region corresponding to aluminum with a radial width of 10.0 mm. For convenience, the mesh extends from 0.0 mm to 3.0 mm in z. I picked an element size of 0.10 mm in z and 0.25 mm in r. The Source section of the GamBet input file DOSE60.GIN has the following entries:

Shift 0.000 0.000 -19.999
SFile Target6004
NPMult = 50
TPulse: 60.0E-9

The Shift command (which must appear before the SFile command) subtracts 19.999 mm from the z coordinates in the file TARGET6004.SRC so that the photons enter at the front face of the aluminum sheet. The value of NpMult specifies 50 showers for each of the 177,482 primary photons for good statistics. The value of TPulse specifies a 60 ns pulse length for the photon flux. In this case, values are recorded in the output file in Gy (grays) rather than Gy/s. At the completion of the run, we load the information into GBView2 and apply a few cycles of smoothing to create the dose distribution show in the figure below.

An inspection of the figure shows that we must be careful defining dose. Because of the buildup of knock-on electrons, the surface dose is much lower than the peak dose (hence the use of intensifier screens in radiography). The figure shows that if we had a thin detector, we should put a mass equivalent of 2.3 mm of aluminum in front for the highest sensitivity with the 6.0 MeV bremsstrahlung spectrum. The peak on-axis dose of 0.1227 Gy (12.27 rad) ocuurs at a distance of 22.3 mm from the target face. We can apply the 1/r2 law to determine the dose for a 1.0 A beam at z = 1000 mm:

D(1000) = 12.27*(22.3/1000)^2 = 6.10175 mrad.

A beam current of 1.64 kA would be required to achieve a dose of 10 rad.

We have assumed that there is vacuum between the exit of the target and the aluminum witness plate. Suppose the space is actually filled with air and we want to estimate the effect on the photon distribution at a distance of 1000 mm. The strategy is to fill the 20 mm void with air at 50 times normal density. To model the special material can use the alternate form of the GamBet Material command. Here is the modified Composition section of TARGET60.GIN:

COMPOSITION
Material
Name Air50
Component N 0.80
Component O 0.20
Density: 0.06145
Insulator
End
Material W
Region(1) = 1
Region(2) = 2
END

Checking the photon distribution in the filtered escape file, we find that the effect of the air is relatively small. The number of photons drops from 178,842 to 178,669 and the average energy changes from 1.108 MeV to 1.106 MeV.

X-ray dose

Please use this link if you'd like more information about GamBet: http://www.fieldp.com/gambet.html


LINKS