This work introduces a numerical method to iteratively extract parameters of a rectangular parallelepiped (RPP) sensitive volume (SV) from experimental proton direct ionization SEU data. The method combines two separate numerical models. The first model estimates the average LET values for energetic ions, including protons and also heavy ions, in elemental solid targets. The second model describes the statistical variance in the energy deposition events of projectile-induced primary ionization within a RPP shaped target volume. To benchmark the method, simulated cross-section values based on RPP parameters derived with this method are compared with literature data from four SRAM devices. The RPP geometries determined by this method reproduced the experimental crosssection values in the literature with good accuracy, therefore showing that this method can be used to reliably and quickly determine the RPP parameters for SVs in memories sensitive to proton direct ionization (PDI). The method is currently strictly limited to direct ionization effects, i.e. not taking into account any nuclear reaction mechanisms, and elemental materials due to the underlying models' definitions.Index Terms-proton direct ionization (PDI), single event upset (SEU), linear energy transfer (LET), straggling, Monte Carlo (MC) method, rectangular parallelepiped (RPP)