Stochastic analysis of dynamical systems by phase-space-controlled Monte Carlo simulation

1999 ◽  
Vol 168 (1-4) ◽  
pp. 273-283 ◽  
Author(s):  
N. Harnpornchai ◽  
H.J. Pradlwarter ◽  
G.I. Schnëller
2020 ◽  
Vol 65 (1) ◽  
pp. 54-58
Author(s):  
T. Medjadj ◽  
A. Ksenofontov ◽  
A. Dalechina

Purpose: To develop an effective method of Monte Carlo simulation of the GammaKnife Perfexion system by rotating particles in the phase space file (PSF). This method does not require simulating of all 192 sources that are distributed in the conical form of the Perfexion collimator. The simulation was performed only for 5 out of 192 sources for each collimator size. Material and methods: Monte Carlo simulation of dose distribution for previous models of GammaKnife system requires phase space file for only one source, since this phase space is identical for all the 201 sources. The Perfexion model is more complex due to the non-coaxial positions of the sources and the complexity of the collimator system itself. In this work, we present an effective method to simulate the Perfexion model using a phase space file. Penelope Monte Carlo code was used to perform this simulation. In this method, the PSF was obtained for one source in each ring, resulting in five files for each collimator size. PSF for other sources were created by azimuthal redistribution of particles, in the obtained PSF, by rotation around the Z-axis. The phase space files of the same ring were then stored together in a single file. Results: The paper presented MC simulation using the azimuthal redistribution of particles in the phase space file by rotation around the Z-axis. The simulation has been validated comparing the dose profiles and output factors with the data of the algorithm TMR10 planning system Leksell Gamma Plan (LGP) in a homogeneous environment. The acceptance criterion between TMR10 and Monte Carlo calculations for the profiles was based on the gamma index (GI). Index values more than one were not detected in all cases, which indicates a good agreement of results. The differences between the output factors obtained in this work and the TMR10 data for collimators 8 mm and 4 mm are 0.74 and 0.73 %, respectively. Conclusion: In this work successfully implemented an effective method of simulating the Leksell Gamma knife Perfexion system. The presented method does not require modeling for all 192 sources distributed in the conical form of the Perfexion collimator. The simulation was performed for only five sources for each collimator and their files PSF were obtained. These files were used to create the PSF files for other sources by azimuthal redistribution of particles, in these files, by rotation around the Z-axis providing correct calculations of dose distributions in a homogeneous medium for 16, 8 and 4 mm collimators.


2013 ◽  
Vol 4 (1) ◽  
pp. 37-46 ◽  
Author(s):  
Kaihang Shi ◽  
Qianru Guo ◽  
Ann Jeffers

Author(s):  
Viviane Luise Silva de Lima ◽  
Israel Panazollo ◽  
Gustavo Cordeiro dos Santos ◽  
Daniel Pinheiro Bernardon ◽  
Mauricio Sperandio ◽  
...  

Bauingenieur ◽  
2015 ◽  
Vol 90 (09) ◽  
pp. 456-462
Author(s):  
Guido Morgenthal ◽  
Marcus Achenbach

Es werden Monte-Carlo-Simulationen brandbeanspruchter Stahlbetonstützen mit dem allgemeinen Verfahren nach DIN EN 1992–1–2 durchgeführt. Dabei werden zwölf Pendelstützen mit konstanter Lastausmitte und einer allseitigen Beflammung durch die Einheits-Temperaturzeitkurve untersucht. Die bei der Simulation berücksichtigte Feuerwiderstandsdauer wird mit den in DIN EN 1992–1–2 enthaltenen tabellarischen Daten bestimmt. Die Beispiele werden so gewählt, dass sie im experimentell abgesicherten Bereich liegen.   Bei der Monte-Carlo-Simulation werden zwei Rechenläufe betrachtet. Beim ersten Lauf werden die Unsicherheiten der thermischen Analyse und des mechanischen Modells berücksichtigt, beim zweiten werden diese vernachlässigt. Für beide Läufe werden die Versagenswahrscheinlichkeiten bestimmt und ausgewertet. Dabei zeigt sich, dass die Vernachlässigung der Modellunsicherheiten zu unsicheren Ergebnissen führen kann.   Die in der Monte-Carlo-Simulation verwendete Grenzzustandsfunktion wird mit einer globalen Sensitivitätsanalyse unter Berücksichtigung der Unsicherheiten der thermischen Analyse und des Widerstandsmodells ausgewertet. Die berechneten Sobolindizes belegen, dass das allgemeine Verfahren eine hohe Sensitivität gegenüber den Unsicherheiten der Temperaturberechnung und dem Widerstandsmodell aufweist.


2010 ◽  
Vol 452-453 ◽  
pp. 277-280
Author(s):  
Seiichiro Sakata ◽  
Fumihiro Ashida

This paper discusses a stochastic microscopic stress analysis of a composite material for a microscopic random variation. The stochastic stress analysis is performed via a stochastic homogenization and multiscale analysis. The homogenization method is employed for the multiscale analysis and the Monte-Carlo simulation or perturbation-based method can be employed for the stochastic analysis. In this paper, outline of the analysis and some numerical results are provided.


Sign in / Sign up

Export Citation Format

Share Document