Cristian Lacey and Jackie Chen, together with researchers from Ulsan National Institute of Science and Technology (Korea) and Pukyong National University (Korea), have published a paper entitled “Time-dependent-bases with local CUR decomposition method for accelerating turbulent combustion simulations” in Combustion and Flame. This work presents a reduced-order modeling framework that extends previous work by introducing spatially local, rather than global, time-dependent-bases for manifold representations and an adaptive, eigenvalue-based method of identifying and penalizing fast-reacting species to reduce stiffness and function evaluations in the numerical system. The efficacy of the method is demonstrated on a variety of test cases, including premixed flames (1D), non-premixed ignition with vortex interaction (2D), and turbulent premixed flames (3D). The computational cost of high-fidelity turbulent reacting flow simulations is reduced by up to two orders of magnitude while retaining nearly full accuracy.
For more information, see: https://doi.org/10.1016/j.combustflame.2026.114848