Extrusion-based 3D printing processes of industrial plastic granulate expand the application possibilities of fused layer modeling (FLM) through higher deposition rates at lower material costs. This enables the economical production of large component structures, such as large tool molds or hull segments for ship constructions, which, however, places higher demands on process reliability and dimensional stability. Many parameters, such as the extruder speed, material quality or temperature, influence the shape of the extruded rope and therefore also the dimensional accuracy of the overall printed geometry. In addition, elementary influencing parameters are subject to process related fluctuations, which leads to imperfections and local deviations from the originally planned geometry. This results in the requirement to calculate and represent an exact knowledge of the actual geometry during the printing process in the one hand to localize defects and deviations and in the other hand to improve the actual printed geometry for the planning of finishing machining process. This article presents a tri-dexel-based simulation model that uses material data, extruder data and machine control data, which are acquired in a high-frequency interpolation cycle during the process, to calculate the material deposition and the resulting actual geometry and make it available as a CAD model. With this new method, complex geometric three-dimensional component measurements, e.g. with a laser line scanner, can be reduced and even substituted. The use of this actual geometry enables an adapted parameter and design optimization in the planning and design of extrusion-based 3D printing processes. In addition, this calculation method enables the generation of a complete digital twin in which the geometry with all its relevant properties is mapped in parallel to the physical product.