Cleaning In Place (CIP) systems are prominent in the food, beverage, and pharmaceutical industry to ensure reproducible hygienic conditions in production lines. Similar cleaning processes are also used as important pre-processing steps in many production processes. Despite the ubiquity of simulations in today’s engineering workflows, nozzle design for spray cleaning systems is still dominated by experiments and experience. The main reason for this is that conventional CFD methods cannot provide meaningful results within a time frame that would allow for a viable iterative design process. In the project SpraySim, supported by the Federal Ministry for Economic Affairs and Climate Action, we developed an efficient simulation workflow for spray cleaning based on a meshfree Generalized Finite Difference Method (GFDM). The software used for this purpose, MESHFREE, has already shown promising performance for various industrial applications where classical mesh-based methods have difficulties [1-4]. The advantages of our approach in the context of free surface flows already provide a considerable increase of performance compared to mesh-based methods. In addition, we consider the transition from a fully resolved 3D simulation in the vicinity of the nozzle to a reduced model which can simulate the motion of the liquid layer at the cleaning surface. This enables us to achieve computation times on moderate hardware that allow parameter variations in a time frame that is acceptable for product development. We validate the numerical models with the help of analogous experiments for the considered nozzles – full cone and flat fan. Furthermore, correlations between local flow quantities and observable cleaning effects are established that allow reliable predictions of cleaning efficiency already during nozzle design.