An artificial intelligence approach for biomass devolatilisation in an industrial CFD model with advanced turbulence-chemistry interaction

[摘要] ENGLISH SUMMARY: The ground work to include more detailed chemistry than global approaches ina combustion simulation was completed. A reduced-order model of the BiomassChemical Percolation Devolatilisation model, ANN-Bio-CPD, was developed andimplemented with artifcial neural networks in order to achieve ease of executionand computational cost reduction with regard to an industrial computational fluidsdynamics application. ANN-Bio-CPD was validated with wire-mesh reactor anddrop-tube furnace experiments from literature. Subsequently, the Eddy DissipationConcept (EDC) turbulence-chemistry interaction model was implementedand validated with ANN-Bio-CPD in a bagasse- fired boiler simulation. The EDCmodel constants were adjusted to achieve the correct temperature and intermediatespecies results in combination with a two-step global reaction mechanism.

[发布机构] Stellenbosch University