Computational optimization of diesel particulate filter geometry: coupled effects of cell density and filter diameter on filtration efficiency and back pressure

Diesel particulate filters (DPFs) are widely used in diesel exhaust after-treatment systems to reduce particulate emissions while maintaining acceptable exhaust back pressure. This study examines the coupled effects of filter diameter and channel density, expressed as cells per square inch (CPSI), on clean-filter DPF performance using computational fluid dynamics (CFD), response surface methodology (RSM), analysis of variance (ANOVA), and equal-weight multi-response desirability analysis. Three-dimensional CFD simulations were conducted for filter diameters of 50 to 250 mm and channel densities of 100 to 300 CPSI. Back pressure exhibited pronounced nonlinear dependence on diameter, CPSI, and their interaction, whereas filtration efficiency was influenced predominantly by the individual geometric variables, with CPSI exerting the stronger effect. The desirability analysis identified a preferred constrained configuration at 50 mm diameter and 250 CPSI, corresponding to a predicted back pressure of 7.29 kPa and filtration efficiency of 86.8%. Because 50 mm represents the lower boundary of the investigated diameter range, this result is interpreted as a boundary solution rather than an unconstrained global optimum. A confirmatory CFD calculation at the selected condition differed from the RSM predictions by less than 1% for both responses, providing a local numerical check of the surrogate models but not experimental validation of the CFD framework. Flow-field analysis showed inlet velocity maldistribution followed by downstream redistribution, while a quantitative uniformity index was not available. Under measured exhaust conditions representing public utility jeepney, public utility van, and public utility bus cases, predicted filtration efficiencies remained above 83% and back pressures below 11 kPa. These results describe the clean-filter numerical response of the selected geometry under the prescribed conditions and should not be interpreted as commercial qualification. The framework provides a basis for examining diameter-CPSI trade-offs, while future work should include experimental benchmarking, extended design domains, transient soot loading, regeneration, and global multi-objective search methods. CFD, RSM, ANOVA, and desirability analysis quantify DPF geometric trade-offs. Diameter and CPSI exert distinct effects on back pressure and particle capture. A preferred 50 mm, 250 CPSI boundary solution predicted 86.8% filtration efficiency. The selected configuration produced 7.29 kPa back pressure under baseline conditions. Filtration efficiency remained above 83% across three public utility vehicle cases.