THE EGS5 CODE SYSTEM

2.3 Simulating the Physical Processes—An OverviewIn some approaches, the Boltzmann transport equation is written for a system and from it a MonteCarlo simulation of the system is derived. This method gives correct average quantities, such asfluences, but may not correctly represent fluctuations in the real situation when variance reductiontechniques are employed. The reader is referred to Chapter 3 in the book by Carter and Cashwell[41]for details of this particular method.In all versions of EGS we have taken a different and more simple-minded approach in that weattempt to simulate the actual physical processes as closely as possible. We have not introducedany inherent variance reduction techniques, so that fluctuations in the Monte Carlo results shouldtruly be representative of real-life fluctuations. For the design of high energy particle detectors,this is an important consideration. On the other hand, fluctuations are not usually of interestin radiation shielding-type problems, and so EGS5 includes several variance reduction techniqueswhich may be optionally invoked to make certain classes of calculations run more efficiently. Noneof the variance reduction techniques are invoked by default, however, and so the method of EGS5can generally be described as analog 3 Monte Carlo.The simulation of an electromagnetic cascade shower can be decomposed into a simulation of thetransport and interactions of single high energy particle, along with some necessary bookkeeping.A last-in-first-out (LIFO) stack is used to store the properties of particles which have yet to besimulated. Initially, only the incident particle is on the stack (more correctly, the properties of theincident particle are stored in the first position of corresponding arrays). The basic strategy is totransport the top particle in the stack either until an interaction takes place, until its energy dropsbelow a predetermined cutoff energy, or until it enters a particular region of space. In the lattertwo cases, the particle is taken off the stack and the simulation resumes with the new top particle.If an interaction takes place, and if there is more than one secondary product particle, the particlewith the lowest available energy is put on the top of the stack. By “available energy” we meanthe maximum energy which can be imparted by a given particle to new secondary particles in acollision: E for photons; E − m for electrons; and E + m for positrons, where E is the particle’stotal energy and m is the electron rest mass energy. By always tracking the lowest energy particlefirst, we ensure that the depth of the stack will never exceed log 2 (E max /E cut ), where E max is thelargest incident energy to be simulated and E cut is the lowest cutoff energy. When the final particleis removed from the stack and none remain, the simulation of the shower event is ended. Thecomplete simulation of each individual shower event is commonly called a Monte Carlo “history”.3 The electron transport model in EGS5 is not strictly “analog” in that all scattering collisions are not treated onan individual basis, but it is “analog” in the sense that the models of the aggregate effects of the large numbers ofcollisions which are grouped together are analytic and can in most circumstances preserve the random nature of thefluctuations in showers.26