[en] The design of digital gamma-ray spectrometer for environmental radiation monitoring has been presented. It detects gamma ray by NaI(Tl) gamma-ray detector. FPGA plays very import role in logical control and digital low pass FIR filter. FPGA controls programmable gain amplifier (PGA) to adjust the amplitude of nuclear pulse signal. After high speed ADC acquiring nuclear pulse signal, FPGA filters the digital signal by Hamming window FIR low-pass digital filter. And then DSP (TMS320C6713) is responsible for digital pulse trapezoidal Shaping, which can reduce pulse pile up and ballistic deficit, peak detecting, and obtaining spectrum. After calibration, the spectrometer can not only get the total count, but also roughly discriminate nuclides, which is meaningful to spot radiation monitoring in industrial areas. The design is very smart and scalable to meet spot radiation monitoring. (authors)

[en] By using Monte Carlo kit Geant4 it has been studied that both the shape of source and detecting distance are able to affect the deposited energy in NaI(Tl) scintillation detector. The simulation shows that the measurement distance of radionuclides with reaching maximum peak-to-gross ratio varies with different geometry of sources. (authors)

[en] BP neural network and quantum particle swarm optimization are widely used in many fields. By using Matlab to analyze the composition of geological samples, the performance of BP neural network is compared with that of BP neural network based on quantum particle swarm optimization. In the process of X-ray fluorescence data analysis and processing of geological samples, the prediction accuracy of QPSO-BP neural network is obviously higher than that of traditional BP model, and its stability is stronger. (authors)

[en] This paper makes use of analysis method of energy dispersion of X ray fluorescence, and combines with clustering analysis and wavelet neural network to build up an improved algorithm model. This model is able to quantify Fe and Ta in EDXRF accurately. The results show that the developed model lowers the average error of the content of Fe and Ta by 11. 89% and 9.94%, and average mean square root by 15.21% and 11.24% in comparison with the model of wavelet neural network without clustering analysis. This model reflects the nonlinear relation of element characteristic X ray count to content of the element, and thus improves the analysis accuracy. (authors)

[en] This work illustrated a wavelet threshold method by combining soft and hard threshold function of wavelet for γ energy spectrum under the natural radioactive condition as a kind of smooth processing. By using Matlab Program, we compared five different smooth de-noising methods including smoothing by five points, the Fourier transform method, soft and hard threshold method and average wavelet threshold method. We showed that the average wavelet threshold method can reflect the ratio of the peak and the peak-to-total ratio of characteristic peak of γ energy spectrum as 7.1609 and 0.3945, respectively. Thus, this method is superior to the other four methods in de-nosing, and it is able to keep the original signal properties with the largest extent and has a high value of application. (authors)

[en] Using the method of least square fitting curve based on wavelet soft threshold, the FeO spectral line of the energy dispersive X-ray fluorescence analyzer was smoothed and analyzed by Matlab software simulation. Respectively with unprocessed raw energy spectrum and using five point two order method, wavelet soft threshold method, Fourier transform method, least square method after processing of spectral quant Native analysis for comparison. The results show that the energy spectrum curve fitted by the least square method based on wavelet soft threshold can effectively retain peak information, the peak total ratio reaches 0.1036, the background is 42, the peak area is larger than other smoothing methods, the peak information is lost least, and the smoothness and de-noising effect of spectral line is significantly better than other methods, which is conducive to improving the accuracy of energy spectrum data processing. (authors)

[en] [Background] Micro target X-ray source has attracted much attention in X-ray microscopy imaging. The focus spot size of X-ray directly affects the spatial resolution of sample. [Purpose] This study aims to investigate the influence of target materials and target microstructures on the emission spot of X-ray tube by simulation. The influence of different target materials on the focus size of X-ray tube is also discussed. [Methods] Monte Carlo simulation code MCNP5 was employed to calculate the focus size with five different target materials, i.e., Mn, Mo, Silver, Rt, Wu and three different target microstructures, i.e., hemisphere, cone and cylinder. The full width at half maximum (FWHM) of the light intensity distribution line was used as the standard to measure the spot size. [Results] The results show that the minimum focus size can be achieved by micro cone structure target at the incident electron energy of 100 keV. Moreover, different target materials have obvious influence on the focus size behind the target. The focus size decreases with the increase of the atomic number of the target material at the beginning, but expands to a certain extent with the further increase of the atomic number, showing a suppression trend. The influence of different target microstructures on the focus size is obviously related to the target type. The difference of focus size between cylinder and cone is up to 30 μm. Compared with the cylindrical tungsten target, the spot size decreases by 17.93% by using micro cone target structure and the silver target. [Conclusions] Compared with the traditional transmission X-ray tube, the micro target X-ray tube is easy to implement because it does not need strong electron focusing equipment. The methodology and the Monte Carlo simulation results are valuable for guiding the design and optimization of micro target X-ray tube source. (authors)

[en] For locating radioactive source, a detector system that consists of three detectors and different shapes of shielding layer between detectors has been designed. The detection sensitivity of the detector under different shielding layer shapes was studied in this paper. The theoretical calculation value and the results of MCNP simulation show that the sensitivity of type C detectors is better than that of type A and B detectors at the same distance. And as the distance increases, the sensitivity of type A, B and C detectors all appear a downward trend, but it always maintains a characteristic that the sensitivity of type C detectors is better than that of type A and B. And the sensitivity of type C detectors decreases with the increase of the distance. In term of the size of model detector in this paper, the structure is applicable to the positioning of radioactive source within a radius of 5 meters. The type C detector can be used as a portable near-range radiation source accurate positioning detector by comparison. (authors)

[en] By using Monte Carlo method the simulation of emitting X-ray spectroscopy corresponding to different electron-injection direction in micro X-ray tube was performed and the optimal emitting angle was obtained. In the simulation there were electron beam with 50 keV from a point source, a silver target with 2 μm of thickness and an X-ray tube with 250 μm of beryllium window. The results indicated that in comparison with the structure of conventional transmission-type X-ray micro X-ray tube, after adjusting electron direction a decrease in interference of low energy X-ray background to characteristic X-ray of measured elements was observed in original spectroscopy. In addition the relative strength of both successive bremsstrahlung and characteristic X-ray increased in the bombardment, which was beneficial to accuracy and precision in the EDXRF analysis. (authors)

[en] A simulation of dose of ⁶⁰Co adsorbed by lung tumor concerning the distance in between was performed by using Monte Carlo method and introducing MIRD body phantasms, meanwhile the curve of adsorbed dose against distance was conducted. The results showed that theoretical calculation in dose adsorption (D) drops with increasing distance (r), and the power of distance (r) should be located between 1.9 and 2.1. The error between simulation and theoretical calculation is less than 10%, which indicates the theoretical calculation in this study is an effective and easy way to estimate the adsorbed dose. (authors)