Purpose This study aims to design a novel strategy to avoid thermal defects in three-dimensional (3D) printing processes. A combination of subroutines and utility routines allows for in situ variations of the key process parameters (KPP) in the thermal simulation and produces by post-processing the output an updated machine code for enhanced quality of the printed part in a single iteration. Design/methodology/approach The input data for the thermal simulation were obtained by characterising an acrylonitrile butadiene styrene filament and calibrating an Ultimaker S5 printer. Abaqus subroutines were used to accurately simulate the continuous deposition of molten material. A utility routine extracted the nodal temperatures during simulation and detected regions that were prone to overheating. We developed a method, enabling the introduction of variable process parameters in the thermal simulation and validated it by printing several test geometries. Findings The results of the thermal simulation indicate that decreasing the print speed enhances the amount of heat dissipated to the environment. Therefore, it is an efficient way to avoid overheating of the printed geometry. The validation geometries showed improvements when adapting the print speed. Originality/value This approach is the first to demonstrate the potential of variable process parameters using implemented routines in the thermal process simulation. Here, the focus is on predicting and avoiding thermal hotspots. However, more potential can be obtained by enabling the prediction and adaptation of KPP to ensure mechanical performance. This would be an alternative way to tackle the cost- and time-inefficient post-processing of 3D-printed parts.