A new method for the automated design of cooling systems in injection molds

This paper presents a new method for the automatic design of the cooling system in injection molds, based on the discrete geometry of the plastic part. In a first phase the new algorithm recognizes the discrete topology of the part, obtaining its depth map and detecting flat, concave regions and slender details which are difficult to cool. The algorithm performs an automatic analysis of the heat transfer, taking into account functional parameters, in order to guarantee a uniform cooling of the part. Based firstly on the limit range distance from which the horizontal straight channels lose cooling effectiveness and secondly on the depth map data, the algorithm provides an optimal layout for the cooling system of the part by adapting it to its geometry. By means of adapting the precision of the algorithm to the molded geometry, both horizontal straight channels for low concavity areas and baffle matrices for concave regions are used. In a second phase, the parameters of the cooling system such as channel diameter, channel separation, etc., are dimensioned by means of genetic optimization algorithms. A second genetic optimization algorithm ensures uniformity and balance in the layout of the cooling system for the plastic part. The result is the design of the cooling system for the plastic part with the same performance as the conformal system. A constant distance between the cooling channels and the part surface is maintained, and at the same time the manufacturing of the mold using CNC techniques and traditional metal materials could be achieved. Complementarily, the algorithm performs an interference analysis with other parts of the mold such as the ejection system. The method does not need a subsequent CAE analysis since it takes into account functional and technical parameters related to heat transfer in its design, thus ensuring its functionality. The algorithm is independent of the CAD modeler used to create the part since it performs a recognition analysis of the part surfaces, being able to be implemented in any CAD system. The data obtained in the design can be used additionally in later applications including the automated design of the injection mold.