Modeling of Multi-Nozzle Plume Physics for Supersonic Retropropulsion

We model multi-nozzle plume physics for supersonic retropropulsion (SRP), a crucial technology for landing humans on Mars. SRP is where rocket engines decelerate a spacecraft by thrusting in the opposite direction of its descent. We address the complex interactions between jet plumes and the freestream flow, which are essential for accurate deceleration modeling.

To improve modeling efficiency and accuracy, we develop a physics-based model using the method of characteristics (MOC) to predict single jet plume geometries. We introduce the blockage ratio (BR) concept to quantify plume interactions and their aerodynamic effects. The BR, positively correlated with the number of nozzles and exhaust mass flow rate, helps minimize undesirable forces on the rocket heatshield.

Additionally, we integrate data-driven methods with physics-based models, using wind-tunnel test data and edge-detection algorithms for Schlieren images to capture flow features and validate the models. We conduct parametric analyses that reveal insightful trends.

The combined approach of physics and data-driven models offers more accurate, and faster predictions of SRP plume behavior and aerodynamic forces. The integration of neural network regression further enhances model robustness and reduces computational costs.