Physics-Informed Neural Networks for Predicting Combustion Dynamics in Rocket Engine Chambers
Abstract
Accurately predicting combustion dynamics within rocket engine chambers is essential for ensuring propulsion system stability, efficiency, and safety. Traditional high-fidelity Computational Fluid Dynamics (CFD) methods offer detailed insights but are computationally expensive and often impractical for real-time applications. This study introduces Physics-Informed Neural Networks (PINNs) as a novel, efficient alternative for modeling combustion dynamics in rocket propulsion systems. PINNs integrate the governing partial differential equations (PDEs)—including mass, momentum, energy, and species transport—directly into the neural network's loss function, enabling data-efficient learning with limited reliance on extensive simulation datasets. We present a PINN framework trained on synthetic and benchmark CFD data from a hydrogen-oxygen combustor. The model captures complex physical phenomena such as temperature gradients, pressure oscillations, and flame front instabilities while significantly reducing computational cost. Our implementation employs a deep neural network architecture with eight hidden layers and 64 neurons per layer, using a combination of Adam and L-BFGS optimizers for training. The network achieves a validation root-mean-square error (RMSE) of 0.0029 MPa in pressure prediction, closely matching CFD results.Comparative analysis shows that the PINN model generalizes well to unseen temporal and spatial domains and offers predictions in under 0.5 seconds, in…
Read the paper
Related studies
- Improving the translation of search strategies using the Polyglot Search Translator: a randomized controlled trial
- Nutritional, Physical, Cognitive, and Combination Interventions and Frailty Reversal Among Older Adults: A Randomized Controlled Trial
- Clinical characteristics and risk factors associated with COVID-19 severity in patients with haematological malignancies in Italy: a retrospective, multicentre, cohort study
- A randomized, controlled trial of oral propranolol in infantile hemangioma
- Clinical phenotypes of delirium during critical illness and severity of subsequent long-term cognitive impairment: a prospective cohort study
- Increased physical activity improves sleep and mood outcomes in inactive people with insomnia: a randomized controlled trial
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.