File pdf 821KB

Vision System for Forged Crankshaft Defect Inspection

Vision System for Forged Crankshaft Defect Inspection

The objective of this project is to develop an inspection process using vision technology to identify and quantify imperfections in forged crankshafts and determine they meet acceptance criteria. During the forging processing of a crankshaft, a variety of surface imperfections can occur. If these imperfections are greater than the allowable threshold the crankshaft is rejected. Currently, these imperfections are manually evaluated in a black light booth. The crankshaft must be rotated 360⁰ to ensure complete inspection of all surfaces. The customer desires to automate the inspection process to gain operational efficiency, reduce human variation, and improve product quality.

The concept involved placing a box over the assembly line, which allows the crankshaft to stop inside the inspection area. The crankshaft sits on rollers on a pallet that allow it to be easily rotated. The rotation mechanism that was selected was based on an existing mechanism in the customer’s plant. It was determined that the quickest and most cost-effective way to rotate the crankshaft was a rubber wheel attached to a servo motor. A UV light would be casted from above to allow two cameras to capture images, which would then be processed using imaging software. Afterwards, a barcode would be sprayed on each defective part, linking the defect to the specific area of the crank. The crankshaft would then exit the box. The design has a monitor placed to the side to display the process and allow for service on the software.

One highlight is that the cost is affordable. A couple negatives were covering of the end of the crankshaft where the wheel would rotate and making sure that each crankshaft stopped in the same position. Also noted was the difficulty between communicating all 3 systems and the quick exchange.

The design involved outsourcing to camera companies who specialize in vision. Several cameras from different companies were tested for feasibility of crack detection. Out of the cameras tested, only one had software able to detect defects. Company A, an automation vision system company, will supply the cameras necessary for vision and provide the team with a controller to link into the PLC and image processing software to identify cracks.

A recent meeting with the customer revealed that it was of best interest to narrow the scope of the project. The customer wants the focus to be solely on the vision rather than an integrative system. The new goal is to have proof of detections via cameras. Although the goal has changed, the ambition to present a production ready solution is still present. As shown, a fully developed concept was developed that can work within the assembly line and is cost-effective. The hope is that once the team has proof of vision, other engineers will continue the concept and finish what was started this year.