Transit-Length Distribution for Particle Transport in Binary Markovian Mixed Media
Brian C. Kiedrowski · 2024
Read Transit-Length Distribution for Particle Transport in Binary Markovian Mixed Media by Brian C. Kiedrowski (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-19359.txt.
Introduction
Brian C Kiedrowski and Emily H Vu work out how far a particle travels inside one zone of a mixed material before it leaves. The mix holds two materials in random Markov order, like layers dealt by chance. Such mixes show up in porous rock, shielding, and reactor design.
The authors link the mix to a classic telegraph random process, then derive the full spread of transit lengths. For finely mixed media they give a sturdy approximate form that runs fast and stays accurate. The math slides into the atomic-mix limit as mixing grows extreme, as it should. A Monte Carlo check confirms the spread. They then apply it to light crossing a porous slab and to charged particles stopping in one, with closed forms for beam passage odds.
Catalog record
| CREATOR | Brian C. Kiedrowski |
|---|---|
| YEAR | 2024 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 74 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2412-19359
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Transit-Length Distribution for Particle Transport in Binary Markovian Mixed Media Brian C. Kiedrowski Abstract Transit-Length Distribution for Particle Transport in Binary Markovian Mixed Media Brian C. Kiedrowski and Emily H. Vu Department of Nuclear Engineering and Radiological Sciences, University of Michigan, 2355 Bonisteel Boulevard, Ann Arbor, MI 48109 Abstract The correspondence between the telegraph random process and transport within a binary stochastic Markovian mixture is established. This equivalence is used to derive the distribution function for the transit length, defined as the distance a particle moving along a straight-line trajectory travels through a specific material zone within the random mixture. A numerically robust asymptotic form of this distribution is obtained for highly mixed materials and the convergence to the atomic-mix limit is shown. The validity of the distribution is verified using a Monte Carlo simulation of the transport process. The distribution is applied to particle transport in slab geometry containing porous media for two cases: the transmission of light and the stopping of charged particles. For both of these applications, analytical forms using the approximate asymptotic model for the transmission probability of beam sources are obtained and illustrative numerical results are provided. These results show that in cases of highly mixed materials, the asymptotic forms are more accurate than the atomic-mix limit. 1 Introduction Accurately describing the transport of particle radiation through porous or turbid media requires modeling its underlying stochasticity. The naive description is the atomic-mix model, which is a simple volumetric homogenization, but this crude model often proves too inaccurate for applications. As such, several researchers have devised more faithful mathematical descriptions incorporating the material randomness. Levermore et al [ 1 ] derived the Levermore-Pomraning closure for the ensemble-averaged distribution of radiation under the assumption of a binary mixture with exponentially distributed, i.e. Markovian, chord lengths that provides a more faithful description of random media than atomic mix. Since then, several researchers including Levermore and colleagues [ 2 ] , Sanchez [ 3 ] , Davis [ 4 ] , Pautz and Franke [ 5 ] , and d’Eon [ 6 ] have developed more general mathematical descriptions. In this article we model binary Markovian stochastic media using a system of first-order partial differential transport equations relating two state variables, the total distance traveled within a material and the difference of the distances between the two different materials, that is identical to the telegrapher’s equations. We use the solutions of this system of differential equations to obtain the transit-length distribution, the probability density of the distance that a particle travels along a straight trajectory through a particular material type given that the starting material type is known. The telegrapher’s equations represent a random process describing a particle traveling at two fixed velocities and switching directions at exponentially distributed times. The connection of the telegrapher’s equations to this random process was first formalized by Kac [ 7 ] . A result similar to the transit-length density obtained in this article was derived by Ratanov [ 8 ] in the context of quantitative finance for the occupation time of a particular state of a telegraph process. To our knowledge, however, this is the first time the telegraph process has been applied to transport in stochastic media, with the novel insight of this research being the observation of an equivalence between the temporal and spatial variables of a standard telegraph process to the previously mentioned state variables and the switching of directions to the transition between material types. In our related conference summaries [ 9 , 10 ] , we have given this resulting transit-length distribution but without the derivation that we provide in this article and applied it differently as a weighting function to homogenize material properties of binary Markovian mixtures to produce distribution functions that efficiently model the transport of electrons. The transit-length distribution involves modified Bessel functions of the first kind that evaluate to very large numbers for media with a high degree of mixing, leading to numerical overflow. To address this, we derive an analytical asymptotic form of the transit-length distribution that is numerically stable in this highly mixed case. Using this asymptotic form, we then show that the transit-length distribution satisfies the atomic-mix limit. We provide implementations in C++, Python, and Matlab in a public software repository 1 1 1 https://github.com/bckiedrowski/transit_dist that also includes all the scripts used to generate the results within this article. Then, we show results from a Monte Carlo direct simulation of the transport process that provide evidence that the mathematical model is consistent and has been correctly implemented. We also give numerical evidence that the asymptotic and exact forms converge in the limit of large mixing. We then apply the transit-length distribution to the special case of porous media, which consists of a binary mixture of some material and void, in slab geometry. First, we consider the transport of light, but the results are applicable to any neutral particle where the material is effectively a pure absorber. We obtain a form of the expected, i.e. ensemble-average of the angular flux in terms of the transit-length distribution. This angular flux must be integrated numerically using the exact form of the transit-length distribution. However, we obtain an approximate analytical form for the angular flux using the asymptotic form that is valid when the material is not too optically thick. We also integrate these expressions numerically to determine the transmission probability, or equivalently the outgoing partial current. We provide numerical results that illustrate the behavior of these quantities and quantify the accuracy of the asymptotic form and the atomic-mix limit. The results show that when the conditions for its validity are met (highly mixed and not too optically thick), the analytical result using the approximate asymptotic form is more accurate than the result obtained with the atomic-mix model. Next, we show results for the slowing down or stopping of charged particles. (In this paper we only assume the continuous slowing down approximation and neglect stochastic effects of energy straggling and angular deflection, which we treat in other conference publications [ 9 , 10 ] .) We derive expressions for the transmission probability and energy spectrum. For the former, we also apply the asymptotic form of the transit-length distribution and discuss the limitations of its applicability in this problem. We then provide illustrative numerical results for electrons incident onto a porous slab of iron. 2 Transit-Length Distribution This section gives the derivation of the probability distribution for the transit length, the amount of distance a particle moving through a binary Markovian stochastic mixture along a straight trajectory is within a particular material type, given that it started within a prescribed material type. Asymptotic forms are obtained for the case where the number of zones traversed is large and the atomic-mix limit of this distribution is demonstrated. Considerations for a practical implementation are discussed. Finally, the distribution and implementation is verified using Monte Carlo sampling. 2.1 Derivation The problem being considered is a random mixture of two different materials, A and B . The distance a particle would travel along a straight trajectory within a region containing a particular material or stochastic zone is randomly distributed and given by an exponential distribution characterized by a mean