Characterization of advected-surfaces. The sinews of a “MultiSECtioning” approach for non-premixed combustion

Abstract

Turbulent combustion has a particular relevance since in common devices involved flows are inevitably turbulent, because of their large flow rates and the request of intense heat and mass exchanges. The persistence of research on turbulent combustion over many decades reflects the formidable challenges of the subject, which yield slowly to our increasing understanding and technological capabilities in terms of computer power and instrumentation. The modeling paradigms of turbulent combustion are also relevant to an even broader range of physical phenomena, which exhibit multiscale and multiphysics natures. The advances needed in such computational modeling for practical combustion systems will require considerable investment in modeling research in this area, including the new paradigms of multi-scale modeling. It will also need an increased emphasis on research directed at improving experimental databases for model validation that are of a hierarchical nature and start with relatively simple problems that encapsulate the most important physics. For such reasons, strategies which select particular sub-models and which exploit them in such a way that split effects can be separately introduced in more and more complex processes, should be preferred to other ones. In fact, this rational approach allows more sensible evaluation of their effects as well as of their interaction with more complex processes. In this context, the PhD thesis presents a new framework to analyze non-premixed turbulent flames. A step-by-step systematic procedure to analyze different phenomena and their interaction in non-premixed combustion processes is discussed. The various steps of this approach (called MultiSECtioning) are worked out in detail. The first step (and the “sinews” of the approach) considers the stirring of material surfaces in transitional/turbulent flows. An experimental setup has been developed to study this. The optical diagnostics and the data processing procedure are clearly described. The limitations of the measurement technique are discussed. The experimental results represent a considerable database for the stirring/mixing processes. Moreover, the ignition part of the strategy has been discussed in relation to the possible initiation processes, i.e. assisted-ignition and autoignition. Specifically, the autoignition in 1D reaction diffusion layers has been reported. The different types of combustion regimes, which may occur in diffusion-controlled conditions, have been obtained to show that the chemical/diffusion time scales involved in the process are able to influence the quality of the ignition process that can lead or not to stable combustion conditions. Specifically, in MILD Combustion processes some behaviours different from those observed in standard combustion conditions have been obtained. In this respect, MultiSECtioning is a strategy able to adapt itself to the different reactive conditions that can occur in a combustion process. The most important conclusion from this PhD work is that the presented procedure is very robust, feasible, suitable and flexible. In particular, the thesis has permitted to identify and measure the main quantities that characterize the stirring process. Moreover chemical/diffusive time scales have been evaluated in relation to the different combustion regimes.