Optimization of polymer extrusion with genetic algorithms

In the manufacture of plastics parts by extrusion or injection moulding, polymers are usually plasticized in single-screw units. The mechanisms involved are complex and dependent on the material, the geometry, and the type of operation. Usually, process optimization is based on trial and error – a very inefficient method. A more efficient approach is to solve the inverse problem, i.e. to determine the operating conditions that produce the desired output and/or product characteristics. An alternative strategy consists in maximizing the value of an objective function, by solving the direct problem iteratively. The nature of the objective function – with conflicting criteria – and the characteristics of the search space make an approach based on genetic algorithms worth investigating. Therefore, a modelling package, an objective function, and a genetic algorithm are interrelated to solve the industrial extrusion problem. The advantages and disadvantages of this implementation are discussed.