With simulation models integrated into production systems, it will be possible to control and adjust manufacturing processes in real time, avoiding product defects before they are finalized. The Stream-0D project coordinated by Itainnova will demonstrate this technology in three production processes: servo-brakes, bearings and sealing gaskets.
Production sets the pace. The part advances on the assembly line with its theoretical dimensions, defined by the design engineer, as a guide.
These ideal dimensions have some variability, but the simulation models are quite accurate. The dimension of each component is measured and entered into the model, which updates the prediction of the final characteristics of the product in real time. When the last component is reached, an optimization algorithm calculates the value it must have in order to comply with the design with maximum accuracy.
An automated cutting or precision machining process adjusts the length of the latter component. Mission accomplished: the defect has been avoided.
The European Stream-0D project coordinated by Itainnova integrates simulation models capable of interacting with the production process in real time into the production systems. Objective? Moving towards zero-defect manufacturing.
Having models capable of running at the pace set by the production line will allow them to be integrated with measurement or data acquisition systems and control systems to control and adjust manufacturing processes in real time.
Traditional simulation models have been used for years in the design and development phases of products and components. “Depending on the product, the models can be complex and need minutes, hours or even days to run,” Valdés points out. The novelty of this project is “the application of order reduction technologies to transform complex and computationally expensive simulation models into ‘simple’ parametric models-both to be run on a laptop, tablet or smartphone-but capable of detecting and correcting deviations in real time.”
The problem is that “the lines are moving very fast and the models are slow“. Therefore, “if models were available whose response was continuously updated with the real values of the parameters of each unit, measured on the production line itself, it would be possible to adjust some other parameter downstream of the process to ensure that each unit meets its specifications as accurately as possible, avoiding the appearance and propagation of defects”.
The main technological challenge is precisely “the development of simulation models that are as accurate as possible”. sufficiently precise to work with the small variations of the components in the line. and the generation of the corresponding reduced-order model that can be run in real time with the same accuracy as the original model, as well as the development of communication and control protocols between model, line and measurement and data acquisition systems,” Valdés explains.
The project will demonstrate this technology in three production processes in the automotive sector: brake servos, bearings and seals.
