Search

International audience; The morphological description of fractal agglomerates is generally reduced to only two parameters,namely the mass fractal dimension and its prefactor. In the most evolved approaches, a stretching exponent is also introduced, while a packing factor is preferred to the fractal prefactor. In any case, the current analytical description of agglomerates morphology is accurate only for sufficiently large agglomerates, which is due to the limited spatial extension of the clusters that are actually quasi-fractal. In the present study, a cutoff function of the pair correlation function is considered for both larger and smaller scales. This enables a more accurate morphological description valid for any cluster size is to be given taking into account the polydispersity of the primary spheres. This new analytical morphological description relying on 5 parameters, is presented here for the first time. The physical range covered by these morphological parameters is determined based on virtually generated Diffusion Limited Cluster Agglomeration. Finally, the model is used to express the fractal prefactor and structure factors and their dependence on agglomerate size and morphological parameters is investigated.

WOS:000358626000003; International audience; Effective density measurements are used extensively to convert submicron particle sizes based on a particle's mobility diameter into mass. Measurements of the effective density also provide information concerning the particle morphology. For example, the effective density curves of fractal aggregates reveal a scaling factor that seems to correlate with the fractal dimension of the particles. The present paper proposes a simple semi-empirical model that permits the quantitative interpretation of these measurements to determine parameters such as the fractal dimension, the primary particle size, and the bulk density of an aggregate particle. The proposed model is based on the assumption that the hydrodynamic drag force of an aggregate is proportional to the drag force applied to isolated primary spheres and to the number of primary spheres in the aggregate at power a. The model was applied to soot particles produced by either a spark discharge (PALAS GFG1000) or by combustion (miniCAST 5206)-both mechanisms enable the generation of aggregates or agglomerates with very different primary sphere diameters. The proposed model showed a good fit for all of the effective density measurements obtained in this study; the a parameter was driven by the aggregate fractal dimension and by the Knudsen number that was determined based on the primary particle diameter. Finally, for a known primary particle diameter, the fractal dimension and the bulk density were determined successfully with the proposed model. (C) 2015 Elsevier Ltd. All rights reserved.

Specific surface areas of particles produced at small-scale diffusion flame burners and also in pilot and large-scale fires involving complex fuels are reported and compared. Specific surface area SBET is determined by BET (Brunauer-Emmett-Teller) analysis and also according to previously developed approach STEM relying on the primary sphere based on TEM (Tansmission Electron Microscopy) images analysis. True density is also determined and the respective influences of primary particle diameter (Dpp) and organic carbon to total carbon ratio (OC/TC) are discussed. The present study is the first to propose a careful comparison between TEM and BET specific surface areas of 20 different samples of carbonaceous particles emitted under realistic fire conditions and to bring validity range for such TEM based approach. For samples containing low and moderate OC content (less than 20%), a good agreement between STEM and SBET is reported (within ± 20% for a confidence interval of 95%), confirming the significant influence of primary particle diameter and the relevance of a purely geometrical description of the surface specific area of soot particles. For larger OC/TC, such approach fails to predict the particles specific surface area within a reasonable confidence interval.