The design exploration phase at the beginning of a product lifecycle determines to a large extent its full timeline and overall costs. During such phase, ideally a large number of simulations is run with parametric combinations of geometric shapes, material behaviors and operating condtions, in order to drive the selection of the optimal basic design choices, where functional requirements and feasibility meet the cost targets. As industrial leaders face competition challenges, this preliminary phase is tentatively shrinked to a few months for a family of products and the need to reduce computational costs becomes cardinal, as it is traded with reasonable levels of accuracy. This is how computational tools such as surrogate modeling become industrially relevant. An appealing class of surrogate modeling techniques include the so-called projection-based reduced order modeling techniques, that based on training data from full order model simulations enable users to simulate faster new parametric. Multiple techniques exist and differ from each other in the treatment of non-linear behavior. Within the presented works, the Authors investigate the application of parametrized loads on multiple non-linear statics cases making use of the Discrete Empirical Interpolation Method and obtaining encouraging results.