Self-optimizing rubber injection molding.

Manufacturers of rubber products work in a complex field of ever changing customer requirements. On the one hand, there are the customers who expect even lower prices along with more service and product efficiency, as well as an extremely tough competition within the global markets. On the other hand, they have to live with the loss of experience caused by the departure of employees and by relocations, just to mention a few of the various possible scenarios within this tension-filled environment.

The problems of manufacturers

The pressure towards cost savings leads many companies to reduce their staff of experts and to look for ways to compensate the effects of such losses by novel tools and aids. One such aid can be, for limited periods of time, external consultants who offer solutions for existing problems, but never or rarely take responsibility for putting them into practice. This job, then, has to remain with the few, often overburdened internal specialists.

Another solution is offered by systems which, as part of new process controls, can duplicate the jobs of these specialists in adjusting and monitoring the units. These systems resemble (without wishing to overstrain the expression) those expert systems which, in various forms, have often been promoted in the past. They have in common to transform, in the form of software and hardware, expert approaches and strategies into practical recommendations.

Such a system is introduced here which, starting from an initial situation, will successively modify the existing control and operation settings of a rubber injection molding machine in a way that the curing operation for the manufacture of molded rubber articles in this machine reaches an economical and quality-assured optimum.

[Cure.sup.2]

[Cure.sup.2] aims towards an ideal situation whereby, with the compound incubation period reduced to a minimum and the highest possible compound injection temperature, the onset of cure takes place, unaffected by thermal conduction considerations, practically simultaneously in all parts of the molding, and then continues almost at the identical rate. If this objective is attained, the result will be the shortest possible curing.

[FIGURE 1 OMITTED]

[Cure.sup.2] considers in its calculation all stages of the injection process, beginning with the strip intake area over plasticizing to the point of the cavity gate.

This is feasible because the system makes the individual process steps transparent, so that the state of the material is known at every stage. Only under such a condition is it possible to significantly reduce the commonly used safety margin and adjust it to the new circumstances.

The clear visualization of the fundamental process in its entirety, however, represents the basis for its optimization.

The Maplan control system PC5000touch, in combination with the CAS process computer, satisfies all these requirements and makes it possible to arrive independently at the optimum machine settings shortly after start-up, and to adjust them fully automatically during production to respond to varying external influences.

The process model used offers a high degree of transparency of the process, which helps the user in these optimization efforts.

This optimization first aims at minimizing the historical safety margins, and subsequently at reaching the actual distribution of crosslinks (equalling mechanical quality of the molded article) in a novel way, i.e., with a different machine setting in a shorter time.

The basis for the process optimization system is provided through usage of the Maplan FIFO injection units with the short flow distance, and therefore reduced pressure loss (figure 2). This reserve of injection power allows input of heat (by dissipation) directly near the cavity. This goal is also supported through the usage of valve gated cold runners (controllable nozzle diameter). Due to the optimized technical design of the injection unit, further applications for sophisticated compounds are possible.

[FIGURE 2 OMITTED]

by Gert Kain, (gert.kain@.maplan.at) Maplan