Thermal analysis of complete toolpaths in Powder-Bed-Fusion (PBF) processes is necessary for detection and correction of local overheating issues. Moreover, it is also very helpful in the assessment of data-rich and noisy monitoring signals, as they are provided by modern process monitoring apparatus. It is well-known, that direct, time-resolved solutions of the heat equation (e.g. using 3D- FEM) are not feasible for realistic build size and necessary resolutions of <0.1mm, due to the extremely long toolpaths and the very high toolpath aspect ratio. Previous publications [1] presented numerical approximation solutions based on FEM-trained Machine-Learning methods, which can perform fast full-part thermal analysis even with very limited computational resources. However, “Black-Box” Machine Learning (ML) models have several serious drawbacks, based on the fact, that they rely on numerical model training using the specific exposure strategy and process parameters. An alternative approach is to approximate the thermal dynamics of each point analytically, based on Green’s Function solutions for the heat equation [2] In this talk, we will a present novel semi-analytic method based on a Quasi-Green’s-Functions approach. The method delivers a computational speed comparable to ML-models yet showing larger versatility and no need for exposure-pattern specific training. We also present validation experiments showing the comparison between calculation results and high spatio-temporal resolution emission measurements.