Fraunhofer Institute for Industrial Mathematics ITWM
Fraunhofer Institute for Industrial Mathematics ITWM
Schwarze gummilaminierte Metalldichtungen.
© Fraunhofer ITWM
Schwarze gummilaminierte Metalldichtungen.

Quality Assurance – Surface Inspection of Gaskets

The image processing system DISQ is developed for surface inspection of metal gaskets. DISQ uses digital camera images to detect defects in (coated) gaskets and classify them according to geometry, origin and size.

© Fraunhofer ITWM

MASC-DISQ – Surface Inspection of Gaskets

Gaskets are sensitive components. If a damaged gasket for example is integrated in a compressor the risk of malfunction or breakdown is high. Those damaged components are returned to the subcontractor immediately, in worst case the complete charge will be send back.

There are different surface defects, on the one hand defects in the feedstock, e.g. bubbles or impurity, on the other hand defects which results from the blanking process, e.g. dents or scratches. Other defects are colour and glue spots as well as geometrical defects, i.e. the punched components differ from their design form. The control of gaskets is strongly essential, but in manual control it is fault-prone and time-consuming.

 

MASC-DISQ Detects Defects on the Basis of Digital Images

The image processing system MASC-DISQ offers a solution for this problem. The system was developed especially for the surface inspection of black rubber coated metal-gaskets. On these gaskets different types of defects like bubbles, impurities, dents, scratches and geometrical defects can occur. MASC-DISQ is able to detect these defects on (coated) gaskets on the basis of digital images, classifying them according to geometry, cause, and size.

Methods with Sample Parts

Due to the partly very complex shape of the gaskets, which, e.g., show holes and embossment, the method uses samples of the same shape serving as reference for a go-part. After the image acquisition of the part to be inspected, the holes and embossment are detected, in order to exclude these areas from the detection process. Different algorithms are being used to detect the variety of defects. If the respective reference parts are available, a fast and automatic teaching of new types of gaskets (having a similar surface structure) is also possible.

Defect example: Metal gasket with a dent (red).
© Fraunhofer ITWM
Defect example: Metal gasket with a dent (red).
Defect example: Metal gasket with a scratch (red).
© Fraunhofer ITWM
Defect example: Metal gasket with a scratch (red).
Defect example: Metal gasket with a bubble (red).
© Fraunhofer ITWM
Defect example: Metal gasket with a bubble (red).

Related Links: