Mouthfeel of Beer: Development of Tribology Method and Correlation with Sensory Data from an Online Database

Abstract

Abstract Soft tribology has recently gained momentum as a technique to measure friction properties of food that can be correlated to mouthfeel. The mouthfeel of beer is an important driver of acceptance and liking of consumers, and especially non-alcoholic beers are often rated poorly in terms of palate fullness. In the present work, frictional parameters of ten beers (five alcoholic and their non-alcoholic counterpart) were measured, and a range of variables was extracted and subjected to dimension reduction analysis (Principal Component Analysis, clustering, and correlation analysis). Sensory data consisting of a numeric mouthfeel rating and written reviews from an online beer-rating website (www.ratebeer.com) were compiled, transformed, and correlated with the tribology data. Based on frictional parameters of the beers, clear differences were observed between alcoholic and non-alcoholic beers, as well as those beers with high or low mouthfeel rating. Text-mining and clustering of the written reviews led to the development of seven overall sensory descriptors ("watery", "smooth", "thick", "bitter", "foam", "astringent", and "sour") related to mouthfeel. Frictional parameters related to the static, boundary and beginning of the mixed regime were correlated with "watery", "smooth", and "thick", while "bitter", "foam", "astringent", and "sour" were represented later in the mixed regime. These results are significant in two ways; firstly, they indicate that online beer reviews represent a valuable resource when gathering large amounts of sensory data, and secondly, they demonstrate tribology as a tool to instrumentally determine important mouthfeel parameters of beer.