Most rubber compounds for a seal profile consist of numerous individual raw materials, Benjamin Klie explains. “Twenty components often come together here.” They are mechanically and chemically mixed in complex mixing procedures in internal mixers and roller mills depending on the formula. “This is where weaknesses are found with regard to the quality of the extrusion products”, says the Head of Department Processing Methodology at the Deutsche Institut für Kautschuktechnologie (DIK), which specializes in rubber (caoutchouc) research. He knows that, “On the way to the extruder, there is quite a lot you can get wrong.”
As project leader for the BMBF-funded “Digit Rubber” research project, the scientist, together with research partners wants to remedy this. The partners hope to achieve this in the future through digitalization of the process chain and the use of artificial intelligence. The research scenario: An extruder is fed the rubber compound data in real time. Artificial intelligence (AI) intervenes in the event of deviations from the reference compound. “The AI should decide how to adapt the process parameters so the compound being processed stays within the specification”, as Klie puts it in a nutshell. During the three-year project, the aim is to systematically generate typical compounds to feed data to the AI. The major promise of the intelligent technology: “Reject production could ultimately be reduced to zero.”
At the DIK technical center, the scientists have installed a research extrusion line with a 60-millimeter extruder. “This is a near-industrial scale”, adds Klie. For quality control, the researchers decided in favor of the iProfilControl (iPC) inspection system from PIXARGUS. Klie: “The PIXARGUS system ensures that we can perform 100% controls of the extrudate surfaces online and in real time.” Testing will primarily be carried out on traditional sealing applications. “Rubber profiles are geometrically very demanding,” as the rubber expert knows. But it is also about stability and surface effects for the researchers. In addition to the Pixargus system, other measurement devices will record the temperature distribution, for example. “All these data measured from the extrudate will then be fed back and integrated into the AI”, explains the DIK expert.
PIXARGUS measurement technology was not chosen by chance. The Würselen-based measurement technology producer has been a valued practice partner for training courses at DIK. And the iProfilControl system from PIXARGUS has already proven itself here, too. Klie: “Our industry partners that use the PIXARGUS systems for quality control in their production also have good things to report about the system.” And last but not least, the reasonable price of the system fits the research budget of the Institute.
For the cost-optimized, downsized iProfilControl series, PIXARGUS has adapted the system performance, range of functions and hardware to the process requirements so that maximum efficiency can be achieved using the minimum equipment. An introductory model equipped with four cameras can be scaled up to six or eight cameras for any process requirement. Depending on what the customer wishes, the models are designed for 360° dimension measurement or 100 % surface inspection – or, as a combination of both functions, available as an All-in-One solution. The compact desk-top device with an integrated computing unit is operated at the touch of a button by tablet, notebook or PC. The system software is based on proven ProfilControl 7 inspection technology. The iPC series’ repeat accuracy of 0.01% of the field of vision (FOV) is quite impressive in comparison with competitor systems. Any profile surfaces and geometries, no matter how complex, are scanned and inspected in seconds. The system flags the smallest irregularities – even before the reject tolerance limit is reached. This allows the production process to be optimally controlled.