In a vast number of industrial applications, particle-laden flows consist of irregularly-shaped particles with various sizes and shapes and thus, the drag coefficient is strongly affected and different for each particle. The key issue is to incorporate these characteristics within numerical calculations. Therefore, a sample-set of irregularly-shaped combustion residues or rather clinker flakes, with sizes between 50 mu m and 250 mu m was experimentally and numerically investigated. Particle geometry, terminal velocity measurements and drag analysis were carried out within the present work. A reflected-light microscope and an optical particle analyzer were used for geometry analysis. Terminal velocities were measured in a water filled glass column with a commercially available digital camera with a recording rate of 1000 frames per second and a frame-by-frame tracking method. At first a simple particle classification method based on this geometrical analysis and terminal velocity measurements in a static fluid was introduced. As a result, several particle groups were defined based on the obtained data. As a second step, this data were used as input parameters for two simple drag models from literature. The required particle shape parameters were determined by velocity recalculations and data fitting. Furthermore, the applicability of these drag models in combination with the experimentally determined particle shape descriptors was evaluated within a simple test rig, where freely falling particles are exposed to a lateral compressed airflow. The particles would be deflected and settle at different positions at the bottom of the test rig. Numerical calculations of the test rig were performed using a commercial Computational Fluid Dynamics (CFD) code and a numerically efficient Euler-Lagrangian approach in order to predict the accurate motion of particles. The particle mass dispersion at the bottom of the test rig was compared to numerical results. It was shown that the used drag models in combination with the experimentally determined particle shape descriptors improved the prediction of particle drag coefficients and trajectories significantly without an increase of computational cost and time. (C) 2019 Elsevier B.V. All rights reserved.