Dataset on machine learning analysis of droplet spreading and splashing for various liquids and different surface wettability

This dataset contains processed data from droplet-impact experiments investigating maximum spreading and the transition from spreading to splashing across liquids and surfaces with widely varying physicochemical and wetting properties. The maximum-spreading dataset comprises 707 experiments, while a further 44 measurements describe the critical conditions for the onset of splashing. The tested liquids include deionized water, water–glycerol and water–ethanol mixtures, ethylene glycol, ethanol, and isopropyl alcohol. Experiments were performed on aluminum, titanium, stainless steel, glass, PETG, and PEEK substrates ranging from hydrophilic to superhydrophobic. Droplet impacts were recorded at 4000 fps, and the extracted parameters include initial droplet diameter, impact or critical impact velocity, maximum spreading diameter, maximum spreading factor, liquid density, dynamic viscosity, surface tension, Weber number, Reynolds number, and static and advancing contact angles. The dataset also contains three trained Gaussian Process Regression models in MATLAB format: two models for predicting the maximum spreading factor using either dimensional or dimensionless input variables, and one model for predicting the critical impact velocity at the onset of splashing. The experimental data were collected between April 2024 and December 2025, and the machine-learning models were developed between December 2025 and March 2026 at the Faculty of Mechanical Engineering, University of Ljubljana, Slovenia. Detailed information on the experimental procedures, data reduction, model inputs, training, cross-validation, and use of the exported models is provided in the dataset description file. The dataset is valuable for research on droplet impact dynamics, liquid spreading, splashing, surface wettability, and data-driven modelling of multiphase flows. It can be used to develop, train, benchmark, and independently validate empirical, theoretical, numerical, and machine-learning models for predicting maximum droplet spreading and splashing thresholds across a broad range of liquids and surfaces. The included trained models also enable direct prediction of these outcomes for input conditions within the represented parameter space. Potential applications include spray cooling, coating and spraying processes, inkjet and additive manufacturing, surface treatment, fuel injection, agricultural spraying, and anti-icing or self-cleaning surface design, where control of droplet spreading and splashing is important.