A REAL-TIME VISION SYSTEM TO MONITOR/ANALYZE THE CHANGES IN COMPOSITE SPECIMENS DURING MECHANICAL TESTING

RECENT ADVANCEMENTS IN MACHINE VISION TECHNOLOGY AND KNOWLEDGE-BASED EXPERT SYSTEMS HAVE RESULTED IN POWERFUL, COMMERCIALLY AVAILABLE SYSTEMS FOR IMAGE ANALYSIS AND INTERPRETATION. COMPOSITE MATERIALS RESEARCH HAS DEMONSTRATED THE EFFECTIVENESS OF EMPLOYING IMAGE PROCESSING TECHNIQUES TO PROVIDE REPEATABLE AND QUANTITATIVE MEASUREMENTS OF MATERIAL PROPERTIES USING MANUAL, INTEACTIVE TECHNIQUES ON DISCRETE IMAGES. PROPOSED IS THE DEVELOPMENT OF AREAL-TIME VISION SYSTEM TO ACQUIRE, ARCHIVE, AND SUBSQUENTLY ANALYZE IMAGES OF POLYMER COMPOSITE STRATEGIES WILL BE DEVELOPED THAT ALLOW THE OPTIMIZATION OF ACQUISTION SPEED, SPATIAL RESOLUTION, OR STORAGE CAPACITY. ACQUISITION AND STORAGE OF IMAGES DURING MECHANICAL TESTS LASTING SEVERAL HOURS IS POSSIBLE. IMAGES WILL BE ANALYZAED BY AN ESPERT SYSTEM TO INTERPRET THE VISUAL CHANGES, WHICH OCCUR DURING TESTING, IN TERMS OF AN APPROPIATE FAILURE MODEL AND TO DESCRIBE, IN A QUANTITATIVE MANNER, HOW DAMAGE DEVELOPS AND PROPAGATES WITHIN THE MATERIAL. AN EXPERIMENT WILL BE CONDUCTED TO VERIFY THAT A RELATIONSHIP EXISTS BETWEEN THE MECHANICAL CHARACTERISITICS OF THE TEST SPECIMEN AND THE VISUAL INFORMATION CAPTURED BY THE CAMERA. THE RESULTS OF THIS EXPERIMENT WILL BE COMBINED WITH A DISCUSSION OF KNOWLEDGE-BASED EXPERT SYSTEMS TO DESCRIBE HOW THE INTERPRETATIONN OF TEST SPECIMEN IMAGES WILL TAKE PLACE.