Assessing parameters impact in dropwise condensation heat transfer for an individual droplet on inclined/grooved surfaces: a sobol sensitivity analysis

Dropwise condensation (DWC) heat transfer models, at the level of individual liquid droplets, supply insights on a scientific basis for improving condensation processes. Therefore, heat transfer rates and parameters controlling these should be measured correctly. This investigation examined four distinct scenarios to assess parameter sensitivities. The analysis quantified the relative contributions of key variables to average heat flux and temperature distribution using Sobol's sensitivity indices, specifically evaluating inclination angles, roughness index, contact angle, Nusselt ([Formula: see text]) number, droplet height, and saturation temperatures. The study employed Design-Expert software to execute Response Surface Methodology (RSM), utilizing its regression approach and analysis of variance (ANOVA) capabilities to statistically validate the developed models and their constituent terms. Also, the sensitivity analysis was carried out employing the UQLab computational framework, which provides advanced capabilities for uncertainty quantification and global sensitivity analysis. Moreover, Sobol indices were computed utilizing two estimators with 95% confidence intervals from 1000 bootstrap resamples. Sensitivity analysis demonstrates that in general, the DWC heat transfer model is fairly highly sensitive to saturation temperature,[Formula: see text] number, and inclination angle. Furthermore, for rough surfaces, contact angle and saturation temperature dominate the impact on model predictions. Besides, the average heat flux and temperature distribution demonstrate significantly higher sensitivity to the [Formula: see text] number and saturation temperature, respectively compared to the other input parameters as evidenced by Sobol' sensitivity indices. Based on the outcomes, employing Janon estimator the maximum total effect Sobol' indices for scenarios of 1 and 4 are 99.22% and 92.7%, respectively. Such information is crucial in guiding the effort toward parameter accuracy in experimental characterization and computational modeling of DWC mechanisms.