HPC systems and parallel applications are increasing their complexity. Therefore the possibility of easily study and project at large scale the performance of scientific applications is of paramount importance. In this paper we describe a performance analysis method and we apply it to four complex HPC applications. We perform our study on a pre-production HPC system powered by the latest Arm-based CPUs for HPC, the Marvell ThunderX2. For each application we spot inefficiencies and factors that limit their scalability. The results show that in several cases the bottlenecks do not come from the hardware but from the way applications are programmed or the way the system software is configured.
h i g h l i g h t s• TRACE modelization for PKL and ROSA/LSTF installations.• Secondary-side depressurization as accident management action.• CET vs PCT relation.• Analysis of differences in the vessel models. a b s t r a c t Experimental facilities are scaled models of commercial nuclear power plants, and are of great importance to improve nuclear power plants safety. Thus, the results obtained in the experiments undertaken in such facilities are essential to develop and improve the models implemented in the thermal-hydraulic codes, which are used in safety analysis. The experiments and inter-comparisons of the simulated results are usually performed in the frame of international programmes in which different groups of several countries simulate the behaviour of the plant under the accidental conditions established, using different codes and models. The results obtained are compared and studied to improve the knowledge on codes performance and nuclear safety.Thus, the Nuclear Energy Agency (NEA), in the nuclear safety work area, auspices several programmes which involve experiments in different experimental facilities. Among the experiments proposed in NEA programmes, one on them consisted of performing a counterpart test between ROSA/LSTF and PKL facilities, with the main objective of determining the effectiveness of late accident management actions in a small break loss of coolant accident (SBLOCA). This study was proposed as a result of the conclusion obtained by the NEA Working Group on the Analysis and Management of Accidents, which analyzed different installations and observed differences in the measurements of core exit temperature (CET) and maximum peak cladding temperature (PCT). In particular, the transient consists of a small break loss of coolant accident (SBLOCA) in a hot leg with additional failure of safety systems but with accident management measures (AM), consisting of a fast secondary-side depressurization, activated by the CET. The paper presents the results obtained in the simulations for both installations using TRACE, observing, in general, a good agreement with the experiments. However, ROSA/LSTF calculations underestimated the maximum PCT value, what might be explained by the higher core level predicted in the simulation compared with the experiment. In PKL calculations, PCT maximum value is slightly higher than in the experiment, and the core level predicted is lower. In the comparison of the evolution of both installations a different timing in the transient events is observed, due to the difference in the pressure vessel design. Thus, when PKL vessel is modified with some of the ROSA/LSTF features, the evolution of the new PKL model behaviour is closer the one observed in ROSA/LSTF calculations.
ElsevierQuerol Vives, A.; Gallardo Bermell, S.; Ródenas Diago, J.; Verdú Martín, GJ. (2015). Using lattice tools and unfolding methods for hpge detector efficiency simulation with the Monte Carlo code MCNP5. Radiation Physics and Chemistry. 116:219-225. doi:10.1016/j.radphyschem.2015.01.027. USING LATTICE TOOLS AND UNFOLDING METHODS FOR HPGE DETECTOR EFFICIENCY SIMULATION WITH THE MONTE CARLO CODE MCNP5A. Querol 1 , S. Gallardo, J. Ródenas and G. Verdú Instituto de Seguridad Industrial, Radiofísica y Medioambiental (ISIRYM)Universitat Politècnica de València, Valencia, Spain. AbstractIn environmental radioactivity measurements, High Purity Germanium (HPGe) detectors are commonly used due to their excellent resolution. Efficiency calibration of detectors is essential to determine activity of radionuclides. The Monte Carlo method has been proved to be a powerful tool to complement efficiency calculations. In aged detectors, efficiency is partially deteriorated due to the dead layer increasing and consequently, the active volume decreasing. The characterization of the radiation transport in the dead layer is essential for a realistic HPGe simulation. In this work, the MCNP5 code is used to calculate the detector efficiency. The F4MESH tally is used to determine the photon and electron fluence in the dead layer and the active volume. The energy deposited in the Ge has been analyzed using the *F8 tally. The F8 tally is used to obtain spectra and to calculate the detector efficiency. When the photon fluence and the energy deposition in the crystal are known, some unfolding methods can be used to estimate the activity of a given source. In this way, the efficiency is obtained and serves to verify the value obtained by other methods.
A realistic knowledge of the energy spectrum is very important in Quality Control (QC) of X-ray tubes in order to reduce dose to patients. However, due to the implicit difficulties to measure the X-ray spectrum accurately, it is not normally obtained in routine QC. Instead, some parameters are measured and/or calculated. PENELOPE and MCNP5 codes, based on the Monte Carlo method, can be used as complementary tools to verify parameters measured in QC. These codes allow estimating Bremsstrahlung and characteristic lines from the anode taking into account specific characteristics of equipment. They have been applied to simulate an X-ray spectrum. Results are compared with theoretical IPEM 78 spectrum. A sensitivity analysis has been developed to estimate the influence on simulated spectra of important parameters used in simulation codes. With this analysis it has been obtained that the FORCE factor is the most important parameter in PENELOPE simulations. FORCE factor, which is a variance reduction method, improves the simulation but produces hard increases of computer time. The value of FORCE should be optimized so that a good agreement of simulated and theoretical spectra is reached, but with a reduction of computer time. Quality parameters such as Half Value Layer (HVL) can be obtained with the PENELOPE model developed, but FORCE takes such a high value that computer time is hardly increased. On the other hand, depth dose assessment can be achieved with acceptable results for small values of FORCE.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.