Numerical model of the supralinear response of heavy charged-particle TL dosemeters

Abstract

This paper reports the performance of newly developed pre-calibrated Thermoluminescent Dosimeters (TLD) with a description and performance results for two commercially available TLD readers. Introduced for use in basic research, medical diagnostics and therapy applications, the Harshaw Model 5500 Automatic TLD Reader evaluates up to 50 dosimeters (rods, chips, disks, μ-cubes) per loading; the Harshaw Model 3500 Manual TLD Reader evaluates single dosimeters and powder. Tests were conducted using the International Electrotechnical Commission (IEC) guidelines for TLD readers to determine system performance and compliance. Both instruments passed the compliance standards in all tests established by the IEC: detection thresholds for both instruments were less than 10μGy, reproducibility was better than 0.7%, sensitivity to ambient light was less than 0.2Hmax (where Hmax is the maximum detection threshold), TL residue was less than 1% of the total integral charge, stability less than 1% deviation after a 24 hour warm-up period, linearity was within specification, test light stability was under 1%, and power leakage requirements set forth by UL-544 (less than 500 μA AC RMS leakage) was less than 42 μA AC RMS in all cases. The characteristics of newly introduced, factory calibrated, extruded TLD-100 rods were investigated by studying the consistency of the Element Correction Coefficients (ECCs) under a variety of conditions. The ECC (〈x〉 / xi) relates an individual dosimeter's response, xi, to the mean response, 〈x〉, of the group. Supralinearity effects and the validity of ECCs generated at 1 cGy for use in the 50–900 cGy dose range were characterized and reported. The results for two specific applications, phantom studies with 60Co and at four Orthovoltage energies are also reported. Portability of Element Correction Coefficients from reader to reader was achieved with a relative standard deviation of less than 1.7%. The relative standard deviation of five TLDs exposed to various doses in the range of 50–900 cGy, with ECCs applied using a 1 cGy dose, yielded results under 2%. At Orthovoltage energies ranging from 100 kVp to 300 kVp, the corresponding relative standard deviations were less than 3%.