A Framework for Combustion Chemistry Acceleration with DeepONets

A combustion chemistry acceleration scheme is developed based on deep operator nets (DeepONets). The scheme identifies combustion kinetics through a modified DeepONet architecture such that the solutions of thermochemical scalars are projected to new solutions in small and flexible time increments. The scheme utilizes a novel way to generate the training dataset and a novel model training mechanism that efficiently trains the modified architecture. The approach is designed to efficiently implement chemistry acceleration without the need for computationally expensive integration of stiff chemistry. An additional framework of latent-space dynamics identification with modified DeepONet is also proposed, which enhances the computational efficiency and widens the applicability of the proposed scheme. The scheme is demonstrated on the ``simple'' chemical kinetics of hydrogen oxidation and on the more complex chemical kinetics of n-dodecane low-temperature oxidation. The denoted React-DeepONet model is implemented on representative scalars or variables in the latent space based on these representative scalars. Other scalars can be recovered from the solutions of these representative scalars through artificial neural networks. The proposed framework accurately learns the chemical kinetics and efficiently reproduces species and temperature temporal profiles. Moreover, a very large speed-up with a good extrapolation capability is also observed with the proposed scheme.