Automated Screw Thread Inspection
Inspect interior and exterior screw threads without contact
Screw threads are a common manufacturing element. They are found on pipes, bolts, bottles, and other manufactured goods in automotive, oil and gas, medical, aerospace, consumer goods, and a wide range of other industries. Monitoring the presence and quality of threads has typically been done by a contact inspection of a sample with mechanical gauges. Increasingly, other contact techniques such as rotary contact gauging have supplemented or replaced manual gauging.
Conventional machine vision systems are used for presence/absence inspection of threads and have been improving at determining pitch diameters and other metrics. However, they are still not as reliable as the much-slower contact gauging technique, and so are of limited use in applications that require higher speeds.
If screw threads do not meet the specified tolerances and allowances, called thread fit, the result can be attachment failure, fluid leakage, misalignment, vibration, and a variety of other functional problems, some of them subtle and hard to diagnose.
Cognex Deep Learning accurately inspects screw threads to ensure adherence to exact specifications. The defect detection tool learns the appearance of an acceptable external or internal thread from a set of images of good parts. As the thread is rotated 360 degrees, the defect detection tool instantly detects any anomalies, including missing threads, burrs, chips, incorrect thread pitch, incorrect thread depth, incorrect thread diameter, and a range of other defects.
Any errors are caught before the part moves on to a further assembly step, preventing inadequate screw connections in final assembly.