North American winters are notorious for aircraft icing, which is a direct threat to aviation safety. Existing high-fidelity codes successfully model icing/de-icing simulations and calculate convective heat transfer, with a penalty of a high computational cost. For a conceptual design of an ice protection system, however, correlations offer a quick approach to determine the average heat transfer rate on an airfoil. The objective of this paper is to introduce a novel correlation for the Frossling number on a NACA 0012, under fully turbulent flow conditions. The correlation is the result of curve fitting a heat transfer database, created using computational fluid dynamics simulations and valid for a wide range of Reynolds numbers and angles of attack. Moreover, the effects of using two different turbulence models as well as two different thermal boundary conditions are investigated. A flat plate test case is used to verify the heat transfer prediction, and airfoil simulations are validated with experimental measurements at the stagnation point. The final form of the correlation shows that a cubic variation of $\alpha$ fits the computational fluid dynamics data best, with an average error of 2.14%.

北美冬季因飞机结冰而臭名昭著，这是对航空安全的直接威胁。现有的高保真代码成功地对结冰/除冰模拟进行建模并计算对流热传递，但计算成本很高。然而，对于防冰系统的概念设计，相关性提供了一种确定翼型平均传热率的快速方法。本文的目的是在完全湍流条件下引入 NACA 0012 上 Frossling 数的新相关性。该相关性是对传热数据库进行曲线拟合的结果，该数据库使用计算流体动力学模拟创建，适用于各种雷诺数和攻角。而且，研究了使用两种不同湍流模型以及两种不同热边界条件的影响。平板测试案例用于验证传热预测，并通过在停滞点的实验测量验证翼型模拟。相关性的最终形式表明$\alpha$ 最适合计算流体动力学数据，平均误差为 2.14%。