Iterative model of collar height in hole rolling for application in a closed-loop model predictive product property control

Closed-loop controls enable improvements of quality and resilience in production processes. In hole-rolling, there are currently no suitable models for predicting collar height that could be used in a closed-loop model predictive property control. In this paper, we present a novel model that enables this use case. To implement this approach, a novel state variable is defined, capturing the shape of the formed collar in addition to its height. This enables the prediction for changing process parameters. Finally, the model is deployed in a closed-loop control within the simulation program Simufact Forming to test its projection quality.