PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
Synapse
⌘+K
Synapse
May 6, 2026Physics of Fluids

Multi-task diffusion graph model for aero-electromagnetic analysis of blended-wing-body configurations

View Full Paper
Ask AI
Bookmark
Share

Authors

CLC G LiBSBowen ShuKZKe Zhao

Discussion

Loading...

Member takes

Overview

Graph neural network predicts aerodynamic coefficients and radar cross section in blended-wing-body configurations, suggesting enhanced efficiency in simulations.

Key Points

  • The aim is to develop a model that jointly evaluates aerodynamic and electromagnetic features in blended-wing-body designs.
  • Introduced a multi-task diffusion graph neural network defined on unstructured surfaces.
  • Predicted surface pressure coefficients and aerodynamic coefficients (CL, CD) along with radar cross sections (RCS) levels.
  • Evaluated performance on 3800 geometries at various Mach numbers and angles of attack.
  • Achieved R2=0.969 for surface pressure coefficients (Cp).
  • Reported mean absolute errors of 0.0065 for CL and 0.0013 for CD.
  • Demonstrated a wall-clock speedup of approximately 35× for outputs compared to traditional solvers.

Cite This Study

Li et al. (2026) studied this question.

synapsesocial.com/papers/69faa1eb04f884e66b532b57https://doi.org/10.1063/5.0326519
View Full Paper
Ask AI
Bookmark
Share