OERLIKON TEXTIL PROCESSING
BV inspection in the textile industry, task description
Non-contact testing and measuring is unavoidable in many areas of industry, but in many cases is not possible for humans. A camera system replaces the human error source and works continuously at constant quality.
The company ISW GmbH has developed an image processing system which compares reference values of the filaments in a staple fiber production at the beginning and during the production of tows to exclude quality defects in the running process. In addition, a distance check is carried out between the fibre webs. The crack of a tow is also detected at an early stage.
The staple fiber production line used in this project produces polyester filaments. These filaments are spun into finger-thick tows by a fibre spinning mill.
The individual stations receive their own user interface which enables the plant operators to guarantee the required quality at any time without detailed knowledge of the test routines themselves.
The image processing stations operate as pure sensors without intervention in the process and without evaluation of the measured information. The evaluation of the measured information and the target/actual comparison is carried out by the higher-level controller, the PLC.
image processing solution
The image processing application determines the number of tows during start-up. The image acquisition is carried out by IP-67 housed Gig E cameras, which are used up to 100 m away from the BV computer at 2 stations. The number of tows is transmitted directly to a PLC at this point. The number of tows recorded there is the reference value for the further process. The total of three fibre webs each have a speed of approx. 280 m/min. At these high speeds, for example when the tows are wound up into a clew, great damage occurs to the system because the torn tows can hook or knot on the rollers.
In normal operation, the total of 5 cameras at 2 image processing stations, of which 2 cameras at the first (see picture 1) and 3 cameras at the second station, control the number of individual tows.
The respective recorded values are immediately transmitted to the PLC and compared with the reference value. In the event of a deviation, the system is switched off immediately to prevent major damage to the system.
So far, employees at this system component (see Fig. 2) have been checking whether tows are missing within the sliver and, in the event of a fault, have sent a signal through a tear rope so that the production process can be stopped.
An additional gap check is integrated at the second image processing station.
As soon as gaps in a fiber web are detected, it is assumed that a tow is torn. Due to special customer requirements, the image processing stations work as pure sensors without interfering with the process and without evaluating the measured information. The evaluation of the measured information and the target/actual comparison is carried out by the higher-level controller, the PLC.
Until now, employees at this system component have checked whether tows were missing within the sliver and, in the event of an error, sent a signal via a tear rope, thereby stopping the production process. In case of an error in production, the system is switched off immediately to prevent major damage to the system.
The individual stations receive their own user interface, which enables the plant operators to call up the tested quality without detailed knowledge of test routines.