The shape and stacking method of the packing elements will affect the overall performance of the packed bed structure. In this study, the convection heat transfer in packed beds was simulated using the customized OpenFOAM “porousFoam_pimple” solver, which could provide the direction for the actual design and optimization of packed bed. Blender was used to generate the packed beds of spheres and fibers. Moreover, the influence of packed bed geometry on the convective heat transfer performance was investigated. The average pressure gradient and comprehensive heat transfer efficiency vary greatly with the parameters, such as fiber diameter, fiber axial, inlet velocity angle, and fiber interleaving degree. The customized “porousFoam_pimple” solver can be used for the numerical simulation of the convective heat transfer in packed beds with particles of different shapes under different operating conditions, providing a reliable and practical tool for the design and optimization of packed beds. Display omitted •The study was conducted using an all-open-source approach.•The geometric modeling of packed bed was realized using Blender.•The solver for convective heat transfer in packed bed was developed based on OpenFOAM platform.•The fibrous diameter and degree of interlacing were studied for the optimization of convection in fibrous packed bed.