Within a recent research project (PON Safemeat) devoted to "innovative" fresh and fermented meat-based products, the project partners' chambers were used for salami ripening and a number of experimental tests were carried out under strictly controlled conditions. In addition to both process and product data obtained during ripening, the PON Safemeat has also allowed focusing on production cost data. Therefore, a mathematical model was developed to express costs associated to the industrial batch ripening of salami and an objective function was devised with the goal of optimization.The model considers the cost of raw materials, the operating costs (linearly increasing with time during maturing) and, finally, a cost for the "loss of quality" of an off-specification product, expressed as a "non-revenue" and referred to as "QL factor". The objective function is a nonlinear function of the nondimensional time t*.Two test cases have been constructed by considering Italian traditional sausage products, i.e., the spicy "Soppressata A" produced by Dodaro SpA and the "Salame of Felino", and the corresponding results have been analyzed and compared. The outcome of this work is interesting at present and promising in the future for further development and validation activities.

The paper provides an analytical and visual view of what actually happenedon the process sidein a fully instrumented, pilot-scale, air ascending-flow chamber of industrial type for salami ripening. Since ripening is always characterized by a-slow‖ dynamics and limited variations in process variables, the time course of the curing air temperature and humidity, as well as the sausage heart temperature, did not show rapid or incomprehensible transients. The monitored variables clearly showed limited amplitude oscillations due to the "go" and "stop" air circulation pattern, that is the sequence of phases with either forced or natural circulation in the cell as imposed by the supervision system for the automatic control of the chamber set points. The effectiveness of set point tracking was favorably assessed for the experimental tests. Then, comparisons were made between different measurements of the same variable, e.g., air temperature and humidity, monitored by probes at different heights in the chamber; similarly, the temperature measured by a TC at the heart of the sausage was matched to the curing air temperature in the chamber. From these comparisons and other crossed checks among data, it was possible to obtain static (e.g. the effect of the position on air temperature at equal height) and dynamic (e.g., the sausage temperature response to the temperature variations in the chamber) assessments of the process variables. All in all, the work done and its further exploitation offer a tool set for a real-time aid to a factory operator.