Food Engineering, Materials Science, & Nanotechnology Evaluation of Models for Temperature-Dependent Viscosity Changes in Dairy Protein Beverage Formulations During Thermal Processing Clodagh M. Kelleher, James A O’Mahony, Alan L. Kelly, Donal J. O’Callaghan, and Noel A. McCarthy Abstract: Rheological modeling as a function of temperature is a useful tool for describing products undergoing thermal processing. The rheological behavior of a range of dairy-based (4%, w/w) protein beverages was investigated for applicability to semi-empirical temperature-dependent viscosity equations. The viscosity at 16.8 rad/s of the beverages was measured during heating, holding, and cooling over a temperature range of 25 to 90 o C using a rheometer with starch pasting cell geometry. Five established fitting methods were applied based on the Arrhenius and Williams–Landel–Ferry (WLF) equations using nonlinear regression analysis. A two-parameter WLF (WLF 2 ) model, using viscosity at a reference temperature of 25 o C resulted in high R 2 values (0.974 to 0.988) and a statistically superior fit compared to the Arrhenius, Generalized Arrhenius, and exponential equations (P < 0.001). Deviation from the WLF 2 modeled equation was used to describe and investigate the effect formulation had on the changes in viscosity during thermal heating. This study successfully applied the WLF equation to a liquid protein system, proving that a consistent and close fit can be achieved across a range of formulations. A rapid, quantitative method for viscosity–temperature profile evaluation is presented, which can ease product development and optimization of product processing stability. Keywords: Arrhenius, dairy protein beverages, thermal processing viscosity, Williams–Landel–Ferry Practical Application: This study validated the use of the Williams–Landel–Ferry equation to describe the behavior of dairy beverages during thermal processing, providing a better fit to rheological data than the widely used Arrhenius-based equations. In conjunction with the WLF equation, a method was presented which reduced the complex rheological data to a single value, which can aid in the comparison of formulations for product development and optimization in both research and industry. Introduction The use of whey protein ingredients in beverages for specialized applications is growing rapidly. These specialized beverages include nutritional products for the elderly, meal replacement drinks, low- sugar drinks for diabetic patients, and highly functional sports foods for high-performance athletes and body-builders (Shiby, Radhakrishna, & Singh, 2013). In addition to protein, which pro- vides amino acids for muscle recovery and repair, these beverages often contain carbohydrates as a source of energy. Formulating such heat-stable protein-carbohydrate nutritional beverages can be challenging (Chen & O’Mahony, 2016). Heat treatment is carried out on dairy beverages with the aim of reducing the microbial population, inactivating enzymes, while minimizing chemical reactions and physical changes in the prod- uct during storage (Lewis & Deeth, 2009). During heat treatment, a number of thermally induced physical changes occur in dairy- based beverages. Whey proteins undergo conformational changes during heating, due to unfolding of their native compact globular structures (that is, protein denaturation and aggregation), which result in technical challenges that may negatively impact process ef- JFDS-2017-1403 Submitted 8/28/2017, Accepted 1/30/2018. Authors Kelleher, O’Callaghan, and McCarthy are with Teagasc Food Research Centre, Moorepark, Fermoy, Cork, Ireland. Authors Kelleher, O’Mahony, and Kelly are with School of Food and Nutritional Sciences, Univ. College Cork, Cork, Ireland. Direct inquiries to author McCarthy (E-mail: noel.mccarthy@teagasc.ie). ficiency and product quality (Joyce, Brodkorb, Kelly, & O’Mahony, 2017; Wijayanti, Bansal, & Deeth, 2014). These denaturation and aggregation mechanisms, at temperatures greater than 75 ˚C, can lead to fouling of heat-exchangers, increased turbidity, sedimen- tation, and viscosity of beverages with a protein concentration greater than 3.5% (w/w; Joyce et al., 2017). Fouling is a major processing issue in the dairy industry, where up to 80% of op- erational costs can be related to fouling, shutdown, and cleaning processes (De Jong, 2008). Attempts to reduce fouling within ther- mal processing include increasing product heat stability, reducing temperature and residence time, increasing flow velocities and tur- bulence, and monitoring pH (Feldman, 2016; Santos, Nylander, Paulsson, & Tr¨ ag˚ ardh, 2006). Rheological characterization is used in process engineering, quality control and product development (Messaˆ adi et al., 2015), and changes in rheological properties, such as viscosity, have been long-established as indicators of protein denaturation, aggrega- tion, and fouling (Wallh¨ außer, Hussein, & Becker, 2012). Heat stability can also be determined from viscosity measurements at high temperatures, as the onset of coagulation can be detected by rheological analysis (Huppertz, 2016). Thus, characterizing the rheological behavior of a dairy protein beverage under a defined heating cycle can provide useful insights into its behavior during thermal processing, aiding formulation design. Mathematical modeling of viscosity can be employed to reduce a large quan- tity of rheological data to mathematical equations that can be re- lated to physical changes, easing the description of this rheological behavior (Saguy, 2016; Steffe, 1996). C  2018 Institute of Food Technologists R  doi: 10.1111/1750-3841.14097 Vol. 00, Nr. 00, 2018  Journal of Food Science 1 Further reproduction without permission is prohibited