Process control parameters for skipjack tuna (Katsuwonas pelamis) precooking

The purpose of this research was to define the critical process control parameters that influence texture and yield for the precook unit operation in the commercial canning of Skipjack tuna. To accomplish this goal, the impact of precook temperature and time combinations on the Instrumental Texture Profile Analysis (ITPA) texture parameters, protein state, weight loss, and moisture content of hydrothermally treated tuna loin meat was investigated. It was found that temperature was the primary influence for all ITPA parameters, however time influenced texture when samples were held at 55°C. Auto proteolysis was suspected at this temperature, as some ITPA parameters declined with increased time.; Data from small steamed and small hydrothermally treated samples were compared with data from whole steamed fish to ascertain whether small sample results could be extrapolated to data from whole precooked fish. Weight loss, moisture content, and ITPA values reacted similarly, regardless of experiment method. For all treatments, weight loss and moisture content decreased with increased temperature, and hardness, instantaneous springiness, and retarded springiness increased with increased temperature. Cohesiveness did not vary with temperature. Linear conversion equations were written to predict texture, weight loss, and moisture content results of whole precooked fish from small steamed and small hydrothermal samples.; Process inputs which are available from a commercial precooking unit operation were used to model effects of precooking Skipjack tuna. Fuzzy logic, neural network, and multiple linear regression models were written to predict precook time, weight loss, friability, and edible weight of precooked fish from final backbone temperature, frozen weight, and storage time inputs. Both fuzzy logic and neural net models perform better than traditional multiple linear regression models in predicting cook time and weight loss. Friability and edible weight were difficult to quantify with all models, and more data was needed to better quantify these variables.