تعیین تئوری بازده سیستم های گاما طیف سنجی بر اساس توابع احتمالی. کاربرد برای اصلاح ضرایب خودکفایی

A generic theoretical methodology for the calculation of the efficiency of gamma spectrometry systems is introduced in this work. The procedure is valid for any type of source and detector and can be applied to determine the full energy peak and the total efficiency of any source-detector system. The methodology is based on the idea of underlying probability of detection, which describes the physical model for the detection of the gamma radiation at the particular studied situation. This probability depends explicitly on the direction of the gamma radiation, allowing the use of this dependence the development of more realistic and complex models than the traditional models based on the point source integration. The probability function that has to be employed in practice must reproduce the relevant characteristics of the detection process occurring at the particular studied situation. Once the probability is defined, the efficiency calculations can be performed in general by using numerical methods. Monte Carlo integration procedure is especially useful to perform the calculations when complex probability functions are used. The methodology can be used for the direct determination of the efficiency and also for the calculation of corrections that require this determination of the efficiency, as it is the case of coincidence summing, geometric or self-attenuation corrections. In particular, we have applied the procedure to obtain some of the classical self-attenuation correction factors usually employed to correct for the sample attenuation of cylindrical geometry sources. The methodology clarifies the theoretical basis and approximations associated to each factor, by making explicit the probability which is generally hidden and implicit to each model. It has been shown that most of these self-attenuation correction factors can be derived by using a common underlying probability, having this probability a growing level of complexity as it reproduces more precisely the geometric and attenuation configuration of the source-detector system. Experimental verification of the improvement produced by increasing the model complexity has been performed by measuring samples spiked with certified nuclide activities with a HPGe detector. The methodology can be extended to obtain the theoretical efficiencies and corrections corresponding to any geometry by defining adequately the physical model and the subsequent probability corresponding to the particular studied situation.