Detection and evaluation of the fibers' deposition parameters during wet filament winding

Abstract

Wet filament winding technology has been extensively used for the manufacture of rotationally symmetric parts made of fiber‐reinforced plastics (FRP). As the design and modeling of FRP‐parts require numerous assumptions, deviations between calculated and achieved mechanical properties are expected. One aspect that contributes to this discrepancy is the assumption of a homogeneous rectangular cross‐section of the fiber‐band. In this work, the fiber‐band geometry in a wet‐winding process of carbon fiber cylinders is analyzed. An Infrared‐optical system for the detection of the fiber bandwidth and winding‐angle is implemented. The influence of the winding speed and the resin temperature is analyzed. An image processing algorithm for the automatic measurement of the fiber's bandwidth and winding‐angle is developed. Manual and adaptive gamma corrections are implemented to improve image quality. A parameter study for the suitable selection of image processing parameters is performed.

Development of an image processing algorithm (IPA‐Delfin) for the digital reconstruction of composite overwrapped pressure vessels

“…In comparison to previous research, [ 5,7 ] the current work extends the application of an in‐line infrared optical system to a new robot‐based winding machine. In this study, two infrared cameras are used to record and analyze the vessel's circumferential winding and one of both dome sections.…”

“…The second camera (Camera‐2) is located at a fixed position on the winding unit to record the front dome region (Figure 2). Hopmann et al [ 5 ] have proven the suitability of the in‐line detection system for analysis of the wet filament winding process with winding speeds below 1 m/s. To enable the advantages of the robot‐based winding machine, that is, winding speeds above 1 m/s and better process control, a new image processing algorithm for analysis in real‐time is developed.…”

“…In this work, the image processing algorithm previously presented by Hopmann et al [ 5 ] is further developed and the movements of the robot‐based winding machine are integrated. The infrared images are reinterpreted into an 8‐bit grayscale system.…”

Impact of winding parameters on the fiber bandwidth in the cylindrical area of a hydrogen pressure vessel for generating a digital twin

“…Particularly in the case of band‐to‐band deposition of the warm fiber band in the circumferential layers, this results in limitations with regard to high winding speeds. [ 5 ] With the aim of reliably detecting the influence of process changes on the fiber bandwidth, the edge detection of the fiber band deposition in the circumferential winding pattern has to be secured at any time during the winding process. Therefore, the fiber band is deposited in the circumferential layers with a gap between two adjacent fiber bands.…”

“…An infrared‐optical camera system is therefore integrated into the production process on the principles outlined in a previous publication of the researcher. [ 5 ] The principle of in‐line determination of the fiber band geometry is a key element for the generation of a digital twin of the manufactured pressure vessels. As presented in Figure 1, the location‐resolved information in the digital twin is created by a continuous target‐actual‐comparison between the position data of the winding machine and the in‐line measured fiber band geometry and its position.…”