Three-Dimensional Validation of the McClain Roughness Model in GlennICE

The Aeronautics Research Mission Directorate at NASA is developing and applying tools to enable future technologies towards fuel efficient flight. As a part of this, investments were made through multiple experimental icing tests in the NASA Glenn Icing Research Tunnel to characterize icing roughness formation. The experimental campaigns leveraged multiple NACA 0012 geometries with various chord lengths and sweep angles along with the 60-in HAARP-II model to parametrically investigate the influence of cloud scaling properties, freestream air properties, and geometrically influenced surface properties on icing roughness. Through this decade long effort, an empirical correlation was developed that acts as a predictive model for roughness evolution in arbitrary icing scenarios. Separately, NASA began development of a new computational icing code called the Glenn Icing Computational Environment (GlennICE) to predict ice accretion on arbitrary geometries to inform early-stage concept development and improve safety for new and existing aircraft. As such, the McClain Roughness Model was incorporated into GlennICE to provide a generalized method of predicting roughness formation independent of geometric constraints. Previous evaluations of GlennICE on highly swept three-dimensional experimental models identified that the current workflow is not properly augmenting the convective heat transfer coefficient that is computed from the predicted roughness. To address this issue, the roughness for the GlennICE implementation of the McClain Roughness Model is evaluated on two separate geometries, the 21-inch NACA 0012 and the CRM65 Midspan Hybrid. The NACA 0012 is seen to largely align well with the experimental data over increasing freezing fractions with bulk trends being captured accurately. GlennICE tends to overpredict the roughness height on the NACA 0012, which is consistent with results seen previously, while the roughness extents and location of maximum roughness are modeled accurately. For the CRM65 Midspan Hybrid model, the GlennICE predictions for the roughness are improved in comparison to the NACA 0012. The GlennICE implementation of the model shows a strong ability to capture the roughness profile on the upper surface through the position of maximum roughness. GlennICE tends to underpredict the roughness height on the lower surface of the CRM65 Midspan Hybrid, especially for the max scallop conditions. The GlennICE underprediction as accretion time increases highlights nonlinear growth behaviors which provided insights into ice shape underpredictions at this condition.