Speaker
Description
Internationally, the use of wood in constructions is increasing due to its aesthetic appeal and environmental benefits. However, this trend poses significant challenges in terms of fire safety. Accurate fire simulations therefore require improved modelling of wood drying, pyrolysis, and combustion processes, with particular attention to the porous nature of wood materials.
In this work, the toolbox PATO[1] (Porous material Analysis Toolbox based on OpenFOAM) is employed to simulate properly the porous properties, pyrolysis and the induced degradation of these materials. It accounts for mechanisms such as thermal expansion, shrinkage, pyrolysis reactions of hemicellulose, cellulose, and lignin, moisture transport, gas generation, and gas flow within the porous structure
One of the major challenges in pyrolysis simulation lies identifying the physical parameters governing these processes: Some of them, such as density or humidity, could be directly measured, others, like thermal conductivity, are more difficult to determine experimentally. To address this issue, an inverse problem approach is adopted, using experimental data (e.g., temperature evolution and mass loss) obtained under different heating conditions for the same wood species. This procedure requires to have a rich dataset of experimental results and a physical model capable of capture the dominant mechanisms. So, it leads to a revised modelling of water in PATO, treating it as a distinct liquid phase rather than part of the virgin solid components like cellulose or lignin.
Experimental data are drawn from several published studies involving different wood species and heating configurations: In [2], structural members made of glued spruce timber (five pieces of 45 mm × 95 mm) are heated in small gas-fired furnace. In [3], a cylindrical pine wood sample is heated by wire resistance in an inert dinitrogen atmosphere. In [4], the beech samples are heated by dinitrogen introduced at 700°C; this article also considers 3 moisture percentages (0%, 14%, 44%), as well as a fully charred sample.
The inverse problem is solved using the optimization software Dakota[5]. The cost function to be minimized is the relative differences between simulated and measured temperatures at multiple locations, as well as mass loss when available. The inverse analysis targets highly influential uncertain parameters including the thermal conductivity of virgin ang char wood, formation enthalpy of each wood constituents (hemicellulose, cellulose and lignin). The other parameters are fixed before the optimization based on literature values. All of this makes it possible to predict the temperature field inside wood exposed to heat and to make estimation of the charring front evolution.
| References | [1] J. Lachaud et N. N. Mansour, « Porous-Material Analysis Toolbox Based on OpenFOAM and Applications », J. Thermophys. Heat Transf., vol. 28, no 2, p. 191‑202, avr. 2014, doi: 10.2514/1.T4262. [2] J.-C. Mindeguia, G. Cueff, V. Dréan, et G. Auguin, « Simulation of charring depth of timber structures when exposed to non-standard fire curves », J. Struct. Fire Eng., vol. 9, no 1, p. 63‑76, juin 2017, doi: 10.1108/JSFE-01-2017-0011. [3] F. Lahouze, W. Jomaa, C. Métayer, M. Meyer, F. Panerai, et J. Lachaud, « Pyromechanics: A solid mechanics approach to deformation during pyrolysis », Fuel, vol. 390, p. 134557, 2025, doi: https://doi.org/10.1016/j.fuel.2025.134557. [4] U. Sand, J. Sandberg, J. Larfeldt, et R. Bel Fdhila, « Numerical prediction of the transport and pyrolysis in the interior and surrounding of dry and wet wood log », Appl. Energy, vol. 85, no 12, p. 1208‑1224, déc. 2008, doi: 10.1016/j.apenergy.2008.03.001. [5] « dakota — dakota documentation ». Consulté le: 31 décembre 2025. [En ligne]. Disponible sur: https://snl-dakota.github.io/docs/6.20.0/users/ |
|---|---|
| Country | France |
| Acceptance of the Terms & Conditions | Click here to agree |
