Detection of cast residue on high pressure die casting die surface

This thesis aims to develop a temperature feedback system that could provide the basis for automated detection of anomalies on the die surface (e.g., soldering) or for control systems that could improve product quality. A key factor in the provision of temperature feedback was the application of temperature probes positioned within the casting die. Ideally, one would have liked to have as many probes as possible in order to acquire a meaningful temperature profile. In production environments, this was a difficult situation to achieve due to the die geometry and because of the practicalities of embedding numerous temperature probes into a commercial die. The development of such a feedback system would have, in fact, represented several research projects because of the die complexity and the production factors involved in the process. This thesis therefore represented the first step in such a process. As a starting point, a simplistic die geometry (i.e., a block) was chosen to initiate the research. This was fitted with a number of temperature probes and the objective became to: investigate the maximum distance that a temperature probe could detect the development of an unwanted adhered layer on a periodically heated metal block that is water-cooled. From the literature review, there were two analysis methods that were compared in terms of their ability to detect the unwanted adhered layer based upon thermocouple reading(s). These were: specific temperatures and times within a temperature profile of a HPDC (High Pressure Die Casting) cycle; and the theory pertaining to measurement distance from a periodically heated surface. The method that was employed to conduct this research program was based upon the development of an experimental rig in which to simulate an HPDC die. The development of the unwanted adhered layer on the HPDC die could not accurately be controlled in the HPDC process. The experimental rig was designed to allow the location of the unwanted adhered layer (via an artificial layer on the die surface) to be varied relative to the thermocouple position(s) which were at fixed locations of 3 mm from the periodic heat source. Thus, some of the 11 variables associated with the artificial layer could then be varied (i.e., width, thickness and time of occurrence). The results obtained from the experiments showed that the theory pertaining to measurement distance from a periodically heated surface and not the specific temperatures and times within a temperature profile of a cycle was the most stable data representation for all the experimental variables. This data representation was compared statistically against the theoretical equations and noise function to determine the maximum practical distance that a thermocouple could be away from the artificial layer for the two different cycle times (40 and 80 seconds). The maximum practical distance from the artificial layer where a thermocouple could detect the presence of an artificial layer and for this experimental set-up were approximately 0.017 and 0.026 meters for a 40 and 80 second cycle time respectively.