Scientific models are epistemic tools used in scientific practice to study natural pheno-mena. The model design integrates different idealizations deliberately employed to fa-cilitate the use of the model. However, we could question the role of such idealizations and distortions since these usually represent the phenomena in a way that does not coincide with our background knowledge. In this paper, I affirm that scientific models do not fulfill their epistemic function despite being distortions; instead, they accomplish such function because they are idealized representations of the phenomena. Thus, this paper aims to explain how such idealizations represent the phenomenon, facilitating the accomplishment of their epistemic functions. I use the liquid drop model (LDM) in nuclear physics as my case study since the idealized representation of the nucleus is essential for the LDM to accomplish its epistemic function optimally. The LDM ac-counts for the stability of the nucleus relative to its mass number and predicts nuclear fission through the binding energy calculus. This nuclear model represents the nucleus as a classical liquid drop; such distorted representation is deliberate and indispensable to fulfill its epistemic purpose. I use the theory of representation Z (representation-as) to explain how the idealization represents the target phenomenon and allows the attri-bution of properties to it, which is a relevant aspect of explaining the epistemic function of scientific representation.