Methods for estimating X-ray machine output through measurement and simulation.

Ball grid arrays are increasingly being applied in the electronics industry and may require X-ray inspection to ensure the integrity and correct placement of solder pins. However, as the architecture of integrated circuits continues to narrow while simultaneously growing more complex, the risk of electronic failure due to radiation damage increases. While medical X-ray devices have been held to high standards and are repeatedly shown to be well characterized, devices used for electronic inspection are often lacking detailed characterization. This study presents unique methods to solve for important properties in X-ray inspection devices such as source to object distance and energy spectrum. This information can then be applied to Monte Carlo models to achieve better overall dose estimates to electronics, which will lead to superior manufactured products. Since X-ray devices can vary greatly in source characteristics, this work investigates spectral measurement and Monte Carlo representation of three X-ray devices. For a Philips SRO 33 100 medical diagnostic device, the spectral output followed expected trends given by the prediction software SpekCalc and Spektr. For the Dage XD7500NT, direct measurement showed a spectral artifact that through the use of Gafchromic films, was shown to be a contributing effect in the dose output. For the Rad Source RS1800, a high powered irradiation device, direct spectral measurement was not achieved. However, a Monte Carlo model using an assumed spectra was found to match ion chamber measurements to a high degree.

Contamination Measurements from Simultaneous Activated Potassium Bromide Radiological Dispersal Devices with a Collimated Vehicular Sensor.

ABSTRACT: Surface contamination was quantified over a distributed source of activated potassium bromide from three detonations of Radiological Dispersal Devices (RDDs) at the Idaho National Laboratory Radiological Response Training Range, with a maximum sampled area of 19,900 m2, to provide a baseline comparison with other rapid, remote mapping methods. Measurements were obtained with a cerium bromide sensor collimated to a field of view of 3.14 m2, using lead shielding, and towed behind a ground vehicle. Sensor response correction factors for activated potassium bromide were calculated through simulation with SWORD to obtain activity per meter-squared. Continuous maps were produced by interpolating coverage from lawnmower raster scans. Radiological data was overlaid with aerial imagery from an automated unmanned aerial vehicle flight to provide contextual geological information relative to contamination levels. The contamination distribution measurements will be compared to unmanned aerial vehicle methods in future work.