An infrared and vision based temperature monitoring system for glass forming process.

One of the most important steps in optimising the glass container forming process is the determination of the glass temperature. Knowledge of the temperature is necessary because of the strong temperature dependence of the glass properties important to the forming process, especially glass viscosity. During the forming process, a few degrees change in glass temperature can change the glass viscosity, thus significantly affecting the moulding process. This is especially true for narrow neck glass bottle containers where viscosity is critical in maintaining the shape of the parison (unshaped mass of glass before moulded into final form) during the transfer from the blank to the mould, and in determining the forming characteristics during the final blow. By monitoring the glass temperature, it is possible to optimise the glass forming process, thus increasing production rates and/or improving product quality.The current methods of temperature measurement which utilise handheld thermocouples are used to infer the parison's surface temperature by taking measurements of the blank mould. However, large errors are not uncommon due to inherent variability of operator's measurements, as well as the interference they introduce into the heat flow path. Also, the mould temperature measurement leads to assumptions about the parison temperature, which can be inaccurate. The use of thermal imaging technology to measure the temperature of parison offers an attractive alternative. This thesis describes a proposed method to monitor the temperature of the glass during the glass forming process by measuring the temperature of the parison. Temperature measurement is performed using machine vision technology and a high speed infrared camera providing real time temperature data to an operator to monitor and control the glass forming process.The system was developed and tested in a glass manufacturing plant. It is able to aid operators in optimising the I.S. machine by monitoring trend of the parison temperature. While the system is capable of providing temperature trend of the parisons, it is not able to provide reliable temperature reading of the parisons. This is due to the blurring artefact from the image captured, these undesirable artefact exist because of fast parison motion in relative to the speedof the IR camera response time. This limitation causes the system to be less robust when it is used purely for temperature reading. However, when it is used to monitor trend of the parison temperature, the system could be used to improve and detect faults on the I.S machine. Faults in forming process are detected when the parison temperature starts to decrease or increase beyond a control limit. This information helps operator troubleshoot the I.S. machine and increase productivity.