Food Science has always been largely dominated by an almost fully experimental approach, often based on trial and error, probably because that is how food is regularly prepared. The field is rapidly becoming a much more quantitative research area, with use of models like in other fields of applied science. Computer modelling is a relevant part of that, as food systems are in general very complex, and moreover the processing of food items is an essential and equally complicated aspect of their production. Analytical models, or even standard numerical tools, often fall short of providing good descriptions. So far scientist have mainly used tools from chemical engineering, such as heat and mass balances and unit operation approaches to control the large scale production of manufactured foods, thereby optimizing throughput and ensuring food safety, and done so successfully. In order to develop healthy and tasty foods, other aspects, such as structure, are in fact more important. An increasing number of food scientists now recognize the potential of soft condensed matter science and multiscale methods developed in that field, to understand and control food structure. A survey of the field to establish merits of existing modelling tools and specific needs for new tools that could be developed is highly relevant. Also the promotion of a modelling group within the food science community would greatly enhance the possibilities of further developing the application of computer models in the field.
Food is often a very complex material, with different physical properties at various levels of description. For instance wheat bread, one of the most common food items, is made from dough consisting of mixture of gluten, a collection of various macromolecules, some of which are among the largest molecules known, water, starch particles, sugars, and many other small molecules. Kneading of the dough irreversibly changes its physical properties and texture. Yeast cells next change the texture again to a foam like open cellular network, that consequently is fixated during the baking process. In hindsight it is remarkable that simple trial and error could have resulted in establishing this product, it has taken several millennia to get there. Bread is just one example of a complex food product, with a variety of physical processes relevant during its production, each of which can be subjected to a more quantitative (numerical) modelling approach.
Dairy science has been using soft matter descriptions and models for a much longer time already, and many ideas from colloidal models have been applied to various dairy products. Milk has been part of the diet of both young and older humans for many centuries, and of that of mammals in general for billions of years. Milk is a suspension of protein micelles and fat droplets in water, with salt and minerals and other small proteins. By processing the milk other diary products such as yoghurt and cheese are prepared from the natural substance. In cheese preparation the suspension is destabilised by adding an enzyme that severs a part of the hairy protein coating of the micelles, inducing them to cluster. Subsequent processing of the weak solid material removes much of the water, and many of the small proteins that would lead to rapid degradation of the milk. Yoghurt is made by acidifying the milk, which also destabilises the suspension and leads to coagulation of the protein micelles. Models describe various aspects of the structure and mechanical properties of the products, but still many details, like the actual structure and composition of the micelles, remain to be investigated. Soft matter computer simulation will help in these studies.
Joost H.J. van Opheusden
Ruud van der Sman
(University of Leeds)