The development of knowledge and cognition in physics and other fields of contemporary science, technology, engineering and mathematics (STEM) is based on modelling processes increasingly requiring advanced methods of scientific computation. Physics education for STEM education should then involve learning sequences featuring modelling activities with computational knowledge and technologies. Such sequences should manifest epistemological and cognitive balance between theory, experimentation and computation, be interactively collaborative, and ensure the development of meaningful knowledge in physics, mathematics and scientific computation, appropriately considering the diversity of STEM contexts. To address this challenge we have proposed an approach based on the creation of sequences of interactive engagement learning activities with computational modelling that explore different kinds of modelling, introduce scientific computation progressively, generate and resolve cognitive conflicts in the understanding of physics and mathematics, and comparatively analyze the various complementary representations of the mathematical models of physics. In this work we discuss a learning sequence about fluid mechanics for introductory physics students of STEM university courses, during which they built and explored in the computer mathematical physics models and animations helping them resolve difficulties persisting after theoretical lectures and problem-solving paper and pen activities.