| Summary: | The use of friction energy to deposit metals in the solid phase on a substrate, called Friction Surfacing, is typically used to provide hardfacing and corrosion resistant coatings. The process involves rotating a rod under pressure onto a moving substrate and a coating is generated from the rod material at the rubbing interface. The complex events occurring during the process have so far been studied only on flat surfaces with constant geometry. The introduction of standard machine tools in the friction surfacing process has allowed applications using complex shapes. This work aimed at developing new thermal models in order to control the parameters of the friction surfacing process for new applications. It was postulated in this work that the thermal events in the process have to be modelled in order to control the temperature of the bond between the coating and the substrate. Consequently the development of a three dimensional transient thermal model was undertaken. It was identified that the thermal response of the process is directly linked to the shape of the substrate below the rod. As the considered applications for the process involved non-uniform shapes of substrates, the finite volume method was chosen to build this model, which was integrated into a control algorithm. In order to simplify the model, new studies of the rod were undertaken. Analytical and finite volume methods were used and important results linking the dimension of the rod to the process parameters were established. These models were evaluated against experimental data. New ways of representing the rod that simplified the modelling of the friction surfacing process were established. Using the results from these studies, a new transient thermal model of the friction surfacing process was created, based on the finite volume method. Experimental runs were simulated and, by comparison with experimental data, the accuracy of the model was established. The different assumptions that were made during the design of this model were also discussed. Using the model, an iterative calculator was designed to predict the machine parameters for new shapes allowing a constant bonding temperature to be maintained and producing a trajectory for the rod. The error in this prediction was calculated, and it was shown that great accuracy is required in the measurement of the real bond temperature in order to minimise prediction errors.
|
|---|
