Granular materials are abundant in nature and industries, and their flow has been subjected to various experimental, analytical and numerical studies over the last decades. The granular flow applications is extended from sorting, packing, and transporting grains and pharmaceutical particles to the study of avalanches and soil erosion. Most of the studies focused on spherical grains. However, most grains in actual applications are geometrically ellipsoidal. Despite being common, the behaviour of the ellipsoidal grains and their response to the flow remain challenging to predict since their ellipsoidal shape brings the effect of the grain orientation to their mechanical response, which has been ignored in most studies. To study the tendency of ellipsoidal grains to align, this work first investigates the performance of a kinematic model that relates the flow to the evolution of the grain alignment. This is done by performing a numerical analysis for discharging ellipsoidal grains in a flat bottom hopper with various opening sizes. After that, two model parameters that can determine the tendency to align and the misalignment of the grains due to collision are generalized according to the shape of the grains and the level of the mutual alignments that is called the ordering factor. These two model parameters should be used along with the model to capture the effect of the grain orientation. The validity of the generalized model is supported by applying it in discharging prolate (rice) and oblate (lentil) grains from a flat bottom hopper and then comparing the outcome with the available experimental results. To reveal the improvement of the generalized model compared to the original model, the differences of the calculated orientational fields are quantified and visualized for the studied grains and conditions. The outcome of this work indicates that the generalized model is in a better agreement with experimental results in capturing the orientational flow of the ellipsoidal grains. This work also advances the capability of the model in capturing the flow type.