Forecasting lift and drag on a circular cylinder at Re=10^6 using point pressure data and a fuzzy ARTMAP neural network

Abstract eng:
The prediction or simulation of long term time series is important in many engineering applications. In the current work, the fuzzy ARTMAP neural network is used to simulate surface pressure time series for a long circular cylinder in cross flow at Re = 106. It is found that accurate lift and drag fluctuations can be obtained (through the integration of the pressure coefficients) only if the trained networks have an input structure containing information from all areas of the surface. This is due to the vortex shedding phenomenon. This may be a significant limitation to using this technique more generally for the task of obtaining aerodynamic load information for other bluff bodies.

• Export as: BibTeX | MARC | MARCXML | DC | EndNote | NLM | RefWorks
• Add to your basket