This paper examines the novel combination of Taguchi methods and texture profile analysis (TPA) which produces a powerful problem-solving technique and predictive method to aid the manufacture of food industry products. Taguchi methods are well known in the mechanical and electronic engineering environments but have not been applied to the difficult domain of food-product baking. This research examined how this well-known method in mechanical engineering circles could be applied to a different domain when used in a novel combination with TPA, an important quality measurement tool used in the food sector. These industrially focused, laboratory-based trials were carried out in an actual food manufacturing company and were designed to mimic the experimental process which would be used in later trials on the actual production plant. As a consequence of the latter, it was important that the number of experiments was kept to a minimum in order to minimize production downtime costs when applied within the actual plant. The research shows that combining these techniques for the analysis of biscuit dough mix enabled the creation of an experimental methodology and empirical models for four quality characteristics: dough-up time, viscous force, applied-force-dependent-adherence ratio (AFDAR), and cohesiveness ratio; which give an indication of what will happen when a specific factor is changed. This provided useful information about food manufacturing processes and mixture rheology; especially with regard to how the ingredients interact. The predictive models developed use discrete variables for the ingredients and were tested to find the differences in their effects and were validated successfully against experimental trials. The success of this approach showed that the number of experimental factors and trials could be kept to a cost-effective minimum and that the methodology could subsequently be applied to the full-scale production plant process. © IMechE 2006.
|Number of pages||21|
|Journal||Proceedings of the Institution of Mechanical Engineers, Part B: Journal of Engineering Manufacture|
|Publication status||Published - 2006|
- Food industry
- Predictive model
- Texture profile analysis