Predicting proton-induced single event upsets rates

Microelectronic devices are susceptible to single event effects in some space radiation environments. A traversal by a single energetic particle that leads to a change in the information stored on a single memory element is known as a single event upset. The cross section for producing an upset can vary with the angle of incidence of the proton, depending on the dimensions of the sensitive volume and the critical charge. Satellites encounter a high proton flux region when they traverse the inner radiation belts in space. The goal of this research was to develop an algorithm for predicting proton-induced memory upset rates for orbits that cross these belts that includes the angular variations in the upset cross section. This required extending an existing Monte-Carlo code that simulates proton-induced spallation reactions in silicon to be capable of handling protons incident at any angle of incidence. These codes were used to predict upset cross sections. The radiation environment was modeled as a set of monoenergetic omnidirectional proton beams where the relative flux at each energy was predicted by AP8. The shielding that surrounds a device was assumed to be a six sided shell. Protons that traverse the shell will enter the device with a new energy. A shielding code was developed that translated each omnidirectional exposure into a new spectrum of energies. Each spectrum was used as input into the Monte-Carlo code which determined the angle-average upset cross section for that incident proton energy (i.e., the proton's energy before traversing the shielding). The upset rate was predicted from the cross section predictions and the flux prediction of AP8. Applying the codes to specific devices requires knowledge of the dimensions of the sensitive volume and the critical charge, i.e., the threshold energy that must be deposited in the sensitive volume to upset the device. These parameters were determined for four devices flown as part of the Microelectronic Package (MEP) experiment on CRRES using techniques developed in this laboratory and the results were used to calculate SEU rate predictions. The experimental results from the CRRES MEP experiments were compared to the predictions. The predictions were in agreement with experimental results.