Trap Design for Vibratory Bowl FeedersR. P. Berretty, L. Cheung, K. Goldberg, M. H. Overmars,
G. Smith and A. F. van der Stappen
Abstract:The vibratory bowl feeder is the oldest and still most common approach to the automated feeding (orienting) of industrial parts. In this project we consider a class of vibratory bowl filters that can be described by removing polygonal sections from the track; we refer to this class of filters as traps. For an n-sided polygonal part and an m-sided polygonal trap, we give an O (n2 m log n) algorithm to decide whether the part in a specific orientation will safely move across the trap or will fall through the trap and thus be filtered out. For an n-sided convex polygonal part and m-sided convex polygonal trap, this bound is improved to O ((n + m) log n). Furthermore, we show how to design various trap shapes, ranging from simple traps to general polygons which will filter out all but one of the different stable orientations of a given part. Although the run times of our design algorithms are exponential in the number of trap parameters, many industrial part feeders use few-parameter traps (balconies, canyons, slots); in these cases the running times of our algorithms range from linear to low degree polynomial. Implementation:We developed a Java applet to implement our trap design algorithms. Try the Trap Design Applet!
Experiments:How well do the resulting traps perform in practice? We discuss some of the differences between theory and practice and then describe laboratory experiments with a physical feeder. To study the behavior of our traps in practice, we tested two traps in the laboratory. Our experimental track uses a commercial inline vibratory platform from Automation Devices Inc. The platform generates an asymmetric vibration at variable amplitudes that moves parts along the track. The parts are standard uorescent light bulb sockets, approximately 2 inches in length. We assume that parts are singulated and that the same part face lies on the feeder track. Projecting the part onto the track yields a polygonal shape that we provide as input to our trap design algorithms. The two traps were designed along the line of thought; two traps were combined to enable us to choose the orientation to be fed. The balcony rejects orientations with radius larger than the fed orientation. The slot was designed to retain the fed orientation and reject the remaining orientations. The balcony and slot output from the algorithms was cut into sheet metal with a milling machine and attached to the vibratory platform.
In a controlled series of 100 trials with each of the part's stable orientations, we experienced no failures: all undesired part orientations were properly rejected by the pair of traps. The traps never jammed; they successfully rejected many out-of-plane part orientations not modeled by the algorithms. Failures were observed in cases when parts were not singulated: pairs of overlapping parts could be arranged to slip past the traps. Also, residual glue from a price label caused one part to violate our motion model.
Publications:
Acknowledge:This research was supported in part by the National Science Foundation under Award CDA-9726389, NATO Collaborative Research Grant CRG 951224, and the Dutch Organization for Scientific Research (N.W.O.). |