Abstract
The filling characteristics in Lost Foam Casting need to be monitored to minimize the fill-related defects. Capacitive sensors, based on the change in the coupling capacitance with the presence of grounded metal in its proximity, provide a simple non-destructive method of acquiring filling information. In the current work, the previously proposed capacitive sensor design is optimized for response sensitivity for metal fill in simple and complex foam patterns. The response from the multi-electrode sensor is determined for various locations of the metal piece. An ANFIS (Artificial Neural Fuzzy Inference System) numerical model is developed for the nonlinear sensor response and used to find the sensor responses for various metal rill conditions. Further, a least square error approach is suggested to estimate the sensor response for multiple pieces in front of the sensor.