The sustainable and efficient use of high-strength micro-alloyed steels is of particular interest in today’s industry. Using modern steels with high strength and excellent toughness enables thinner constructions, for example in infrastructure applications, for pressure vessels or pipelines. Sufficient toughness is of special importance to avoid catastrophic brittle failure. Yet, the simplified, conventional design rules prevent a full exploitation of the properties of modern high strength steels. Modern simulation models are able to represent the actual failure characteristics of high strength steels and thereby help to fully exploit the potential of these materials. To include toughness properties, it is essential that the models are able to describe the transition behavior of these steels. The materials’ behavior in the transition region is characterized by a competition of ductile and cleavage failure mechanisms. To simulate the corresponding behavior, it is necessary to adapt damage mechanics models for considering both effects. While numerous studies on the simulation of either ductile or cleavage failure exist, only a limited selection is available on investigations in the transition range. Therefore, the presented study investigates two representatives of the most popular groups of ductile failure models. The Gurson-Tvergaard-Needleman model (GTN) is investigated as a micromechanical model while the Modified-Bai-Wierzbicki model (MBW) is selected as a phenomenological model. Both are combined with the Orowan cleavage fracture model. Investigations are performed on a thermomechanical rolled S355, for which the cleavage fracture stress and the parameters for the ductile failure models were experimentally determined. The study compares simulations of Charpy impact toughness tests to experimental results to determine which model class delivers results that are more accurate. The focus of the investigation is hereby set on the simulation of the lower shelf and the lower transition area. In addition, the efficiency of the simulations performed is also evaluated.