[Paper Review] Multi-Frequency Implicit Semi-analog Monte-Carlo (ISMC) Radiative Transfer Solver in Two-Dimensions (without Teleportation)

Elad Steinberg, Shay I. Heizler|arXiv (Cornell University)|Aug 5, 2021

Radiative Heat Transfer Studies31 references16 citations

TL;DR

This paper presents a two-dimensional Multi-Frequency Implicit Semi-analog Monte-Carlo (ISMC) radiative transfer solver that eliminates teleportation errors inherent in traditional Implicit Monte-Carlo (IMC) schemes by using material-particles to carry energy and enabling implicit time-stepping. The ISMC method achieves accurate, teleportation-free solutions in both gray and multi-frequency problems across 1D and 2D geometries, converging faster than IMC in optically thick and high-gradient opacity regions.

ABSTRACT

We study the multi-dimensional radiative transfer phenomena using the ISMC scheme, in both gray and multi-frequency problems. Implicit Monte-Carlo (IMC) schemes have been in use for five decades. The basic algorithm yields teleportation errors, where photons propagate faster than the correct heat front velocity. Recently [Po\"ette and Valentin, J. Comp. Phys., 412, 109405 (2020)], a new implicit scheme based on the semi-analog scheme was presented and tested in several one-dimensional gray problems. In this scheme, the material energy of the cell is carried by material-particles, and the photons are produced only from existing material particles. As a result, the teleportation errors vanish, due to the infinite discrete spatial accuracy of the scheme. We examine the validity of the new scheme in two-dimensional problems, both in Cartesian and Cylindrical geometries. Additionally, we introduce an expansion of the new scheme for multi-frequency problems. We show that the ISMC scheme presents excellent results without teleportation errors in a large number of benchmarks, especially against the slow classic IMC convergence.

Motivation & Objective

• Address the long-standing teleportation error problem in Implicit Monte-Carlo (IMC) schemes, which causes unphysical heat wave propagation in optically thick materials.
• Extend the semi-analog Monte-Carlo (SMC) framework—previously limited to explicit schemes—into an implicit formulation to allow large time steps without sacrificing accuracy.
• Validate the ISMC method in two-dimensional Cartesian and cylindrical geometries across a range of benchmark problems, including both gray and multi-frequency (multi-group) cases.
• Demonstrate that ISMC achieves faster convergence than classic IMC, especially in regions with high opacity gradients or optically thick materials.
• Ensure stability and accuracy under large time steps and low spatial resolution, where IMC fails due to teleportation and convergence issues.

Proposed method

• Adopt the semi-analog Monte-Carlo (SMC) framework, where material-particles carry energy and photons are emitted only from existing material-particles, eliminating artificial photon propagation.
• Implement implicit time-stepping by solving for the energy exchange between material-particles and photons over a time step, using the material temperature at the start of the step.
• Use a frequency-dependent opacity model for multi-frequency problems, with energy-dependent cross-sections and temperature-dependent emission and absorption rates.
• Apply a fixed number of particles per time step, eliminating the need for population control in closed systems, and maintain particle conservation through emission and absorption tracking.
• Solve the radiative transfer equation in 2D using a discrete spatial mesh, with reflection boundary conditions and a black-body source term applied at t=0.
• Use a time step of Δt = 10⁻¹³ s in multi-frequency benchmarks and limit total particle count to 2.5×10⁷ to ensure numerical stability and convergence.

Experimental results

Research questions

• RQ1 Does the ISMC method eliminate teleportation errors in two-dimensional radiative transfer problems with optically thick materials and high opacity gradients?
• RQ2 How does ISMC perform in comparison to classic IMC in terms of convergence speed and accuracy across 1D and 2D gray and multi-frequency benchmarks?
• RQ3 Can ISMC maintain stability and accuracy under large time steps and coarse spatial resolution, where IMC fails due to teleportation and maximum principle violations?
• RQ4 How does the statistical noise in ISMC compare to IMC in optically thin regions, and can it be mitigated with increased particle counts?
• RQ5 Is the ISMC method scalable and robust in complex 2D geometries such as hohlraums and lattice structures with discontinuous material properties?

Key findings

• The ISMC method successfully eliminates teleportation errors in all 2D benchmarks, including optically thick and high-gradient opacity regions, even at coarse spatial resolution and small time steps.
• In the 2D Marshak wave and hohlraum problems, ISMC converges to the correct temperature and radiation energy density profiles much faster than IMC, especially in optically thick materials.
• For the 2D multi-frequency problem with aluminum blocks and foam, ISMC results show excellent agreement with the PN reference solution and with the original IMC implementation, with no observable teleportation effects.
• Despite higher statistical noise in optically thin regions due to discrete emission and absorption, ISMC remains stable and convergent, with noise reducible by increasing particle count.
• The method maintains comparable runtime to IMC when using the same number of particles, and shows similar tolerance to maximum principle violations under large time steps.
• In all benchmarks, ISMC achieves converged solutions with significantly lower spatial resolution requirements than IMC in opaque or high-gradient regions, demonstrating superior efficiency in such regimes.