Current hub location models consider a single type of capacity constraint and assume that both import and export flows are processed at the same rate. This is not realistic in general, especially for intermodal hub networks. We posit that hub network design models should consider in more detail the characteristics of hub facilities and their main resources. To achieve a more realistic representation of hubs, we propose a hub location model that includes different types of hub resources and generalized capacity constraints. Generalized capacity constraints are necessary when the export and import processing rates are not equal or when there is a lack of synchronization between import and export vehicles, as is common in intermodal hub networks. We present and evaluate two mixed-integer linear programming formulations for the hub location problem with generalized capacity constraints. Formulations are evaluated under the well known AP data set and we also discuss the case study of a projected intermodal transportation system in Colombia. We conduct an analysis of the impact of train-truck synchronization on the rail-road hub network design. It was found that synchronization has an important impact on total costs affecting the hub network structure.