Development of an Experimental Digital Twin for the Analysis and Automated Adaptation of Textile Manufacturing Processes using Tufting Technology as an example

The increasing variety of products in German textile production combined with small batch sizes and high costs for machine changeover increase the prices per m² of textile produced. Expert knowledge and experience are required for retooling and setting up production machines as well as for product development. However, due to a shortage of skilled workers and increasing personnel turnover, this knowledge is steadily draining away. Increasing demands on finished products require increased reproducibility of manufacturing processes as well as continuous process optimization.

The research objective was to develop an Experimental Digital Twin (EDT) as a demonstrator exemplary for a tufting machine, to use it for increasing the effectiveness and efficiency of the setting process and to prepare it for further applications on textile machines.

For implementation, the interaction of the tufting tools needle, looper, knife was analyzed. This led to the construction of an EDT that incorporates e.g. kinematics, dynamics, sensors, man-machine interaction of the physical asset (tufting machine). It was used to visualize the movements in three dimensions and to realize an interactive assistance system. Furthermore, an automated adaptation of the machine parameters by means of simulation was designed and implemented. In parallel, sensors for measuring the position of the tufting tools were developed and integrated.

The involvement of experts from other textile manufacturing technologies from science and industry in the project committee guarantees a rapid transfer of the knowledge generated in the project to the entire textile sector.

Results

The movement curves of the tools, which are decisive for the tufting process, were made measurable with range sensors. Furthermore, the main shaft was equipped with an incremental encoder so that it could act as a reference variable for the overall system. Based on the design data of the tufting machine, a visualization system was developed that allows the motion sequence of the tools to be displayed realistically. In addition, an interface was installed with which sensor data can be read out, stored and evaluated. With the aid of this data, it is possible to evaluate the machine timing settings based on data and adjust them if necessary. Previously, this required the subjective visual assessment of the machine operator. This was neither quantified, measured nor reproducible.

The experimentable digital twin (EDT) enables the measurement-based recording of tool movements, provides assistance for setting up the tufting machine and makes settings reproducible. It offers a simulation of the real machine, which is able to represent the settings one-to-one and offers angles of view of machine parts that are not possible in reality due to, for example, the machine body. Furthermore, it is possible to either adjust settings manually in the operating panel of the EDT and display the effects in a simulated way or to have new parameter sets calculated by the EDT. These can then be transferred to the real twin with the help of the setup wizard.

The final report on the project is available on request from the research institution.

Funding program and project number

IGF 21166 N

Duration

01.04.2020 – 31.03.2022