Dr. Robert Martinuzzi

Fundamental focus is on developing low-order models of the dynamics of turbulent bluff body wakes. This work aims to understand the influence of body shape and on-coming flow on: (i) aerodynamic loading on structures; (ii) mixing properties of the wake and (iii) turbulence characteristics. Low-order models are developed to model the wake dynamics and predict loading and mixing characteristics. The physics of vortex structure interactions and inter-scale energy exchanges is investigated for the purpose of the role of large-scale instabilities in the generation of turbulence and dispersion of scalar quantities (temperature or contaminants). The experimental research includes development of optical diagnostic techniques, vortex and feature identification techniques and real-time visualizations for virtual environments.

The objective of the work is to develop model-free closed-loop flow control strategies. The basic principle is based on genetic search algorithms and evolutionary algorithms to guide actuation of flow control devices to achieve drag reduction, mixing control or manipulation of unsteady forces. In this context, genetic programming is used to design the controller, a mathematical model describing how to use a sensor signal to guide the actuation signal, to achieve a desired effect. Examples include suction and/or blowing at critical locations on a body surface to reduce drag or increase mixing. The techniques are well suited for multi-objective optimization.  An important aspect of this work is robust implementation in real systems and reliability.

Semi-empirical and analytical techniques are developed for real-time estimation of flow dynamics and large-scale forces from sensor data obtained from locations removed from the flow domain or body of interest. Briefly, estimation algorithms are trained using experimental or CFD data to predict flow behaviour in full-scale environments. Examples of applications include: prediction fluctuating wind patterns in the wake of large structures from a small number ground measurements, flow patterns in large scale vessels based on wall-pressure measurements, prediction of flow-induced vibrations in pipelines or bridges; or wind gusts based on wall measurements.