Year: 2011
Type: dissertation
Abstract: Space radiation can cause temporary and permanent damage to electronic components residing in outer space.The awareness of radiation environments in space and the knowledge of radiation effects is important for reliable design of space electronics used in, for example, satellites and spacecrafts.Radiation-induced errors in the stored information of memory elements are called single-event upsets (SEU) and the prediction of SEUs is an important aspect of reliable space system design.The early, but still widely used, SEU rate prediction methods originating from the 1970s rely on the direct ionization, also called linear energy transfer (LET), of charged particles neglecting any other processes, such as nuclear reactions.Relying only on the LET means that similar particles always cause similar effect in the device.In reality, this is not true and the energy loss may vary significantly, especially in the case of a nuclear reaction, and this may lead to underestimation of the SEU rate and consequently cause the system to malfunction.SEU rate prediction can also be made by utilizing physics-based particle transport simulators, such as the Geant4, where all physical processes can be included.For example, the significance of nuclear reactions in SEUs can be analyzed with the help of Geant4.However, the accuracy of such tools must be known in order to make reliable SEU rate predictions.In order to test the accuracy of the nuclear reaction description in Geant4, a time-of-flight (TOF) experiment was conducted in the RADEF facility.In these experiments, heavy ion beam was accelerated and guided onto heavy ion targets and the TOF and kinetic energy of the outcoming particles were measured.The same scenario was simulated by using Geant4 and the results for heavy fragment production were compared to those obtained from the experiments.The simulations show clear underestimation, by 1-2 orders of magnitude, in the heavy fragment production cross-sections.In order to improve the underlying code responsible for the nuclear physics, research on the low energy heavy ion reactions is required.Despite the discrepancy between the Geant4 simulations and the experiments, the utilization of a particle transport simulator can improve the reliability of SEU rate prediction.In addition to LET, including other physical processes in SEU rate prediction makes it possible to assess situations which could not be analyzed with the traditional, LET-based, methods.One example of such a situation is a case where a signicifant amount of SEUs which are due to heavy fragments after nuclear reactions in the component.i
Author: Alex Hedlund
Institution: University of Jyväskylä
Language: English
Subfield: Electrical and Electronic Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: green
Funders: Vanderbilt University, Strong
OpenAlex ID: W2340676522
