The physical mechanisms of frost formation have been widely studied, yet much empirism is still needed in numerical approaches. Indeed, accurate simulations of frost growth can be reached by setting up a specific combination of the model empirical inputs while using a method to accurately track the frost-air interface. This paper presents a finite volume ALE method which captures the air-frost interface using dynamic meshes. It is divided into two main sections. First, the search of a valid set of empirical correlations to correctly emulate frost growth under certain experimental conditions. An assessment of seven reference cases is carried out by comparing solutions using different empirical correlations against experimental data. As a result, a discussion on the performance of such parameters is made, emphasizing the fact of using diffusion resistance factors above 1.0 in order to capture the frost growth. Second, a 2D numerical test consisting of a duct flow with a non-homogeneously cooled lower boundary is performed. Aspects related to the frost thickness and growth rate are analysed, proving the method to be a valid candidate to simulate frost growth.