The research programme studied the cure reaction of a phenolic novolak resin and the effects of various additives and fillers on the reaction. The programme utilised the recently developed thermal analysis technique of temperature-modulated differential scanning calorimetry (TMDSC) performed in conjunction with other available thermal analysis techniques. TMDSC enables the signal for the heat of reaction to be separated from the underlying specific heat change in the resin. This meant that the reaction could be studied without interference from any physical changes in the resin. The manufacture of composite brake materials required the use of numerous additives and fillers to produce the desired properties. The influence of such additives on the cure rate and final properties of the resin was known to occur but had not previously been measured due to the difficulties presented by the presence of opaque additives. Some additives also underwent thermally induced physical changes in the temperature range of the cure. The final properties and the processing of new brake materials undergoing development often required trial and error adjustments to compensate for changes in cure rate. An understanding of the influence of additives would enable more rapid commercial development of brake materials through an improvement in the ability to predict both the properties of the product and the optimal processing parameters. Processing efficiency could also be improved through detailed knowledge of the kinetics. Moulding cycle times and post-baking times and temperatures were longer than necessary in order to ensure adequate cure at the end of each stage because of the lack of kinetic data. The cure of phenolic resin has been shown to be highly complicated with numerous alternate and competing reactions. For the manufacture of composite materials, knowledge of the kinetic parameters of individual reactions is not considered to be important; rather the overall kinetic parameters are required for prediction. Therefore the kinetic model parameters that best described the observed behaviour were chosen even though the model had no basis in the molecular interaction theory of reaction. Rather it served as a convenient tool for predictions. Characterisation of the resin proved to be difficult due to the presence of overlapping peaks, and volatile reaction products. TMDSC was successfully used to determine the reaction kinetics of the pure resin and the influence of certain additives on the reaction kinetics. The determination of the kinetic parameters using TMDSC agreed well with the traditional Differential Scanning Calorimetry isothermal and non-isothermal techniques. Both the Perkin-Elmer and TA Instruments were utilised for the research and were found to provide reasonably good agreement with each other. The capabilities and limitations of the individual instruments were critically examined, frequently beyond the manufacturers' specifications. TMDSC suffers from a limitation in the heating rate of the sample compared to DSC. However, it was observed that valuable information could still be obtained from TMDSC despite using heating rates that were higher than specified by manufacturers. Hot Stage Microscopy and thermogravimetry were additional experimental techniques used to aid in the characterisation of the resin. Some inhomogeneity of the resin was identified as well as differences in the behaviour of the cure between open (constant pressure) and closed (constant volume) environments were observed. A novel method of determining the orders of the cure reactions and their kinetic parameters was utilised. Reaction models for the overall cure reactions were postulated and tested by fitment to sections of experimental data in temperature regions which appeared to be free of interference from overlapping peaks. Once an individual peak was reasonably well modelled, adjacent overlapping peaks were able to be modelled both individually and in combinations by fitment to experimental data. The Solver function in Microsoft Excel was utilised to find the best fitting model parameters for the experimental data. The model parameters were able to be refined as overlapping peaks were progressively incorporated into the calculations. This method produced results that agreed well with the traditional method of analysing reaction peak temperatures at multiple scanning rates. Model fitment was shown to be of benefit where overlapping reactions occur. Various model scenarios could be tested and optimised to particular sections of experimental data. This enabled the researcher to easily identify areas of possible anomalies and postulate alternative scenarios. The accuracy of the postulated model was able to be determined by its successful fitment to experimental data from experiments run under different conditions.
| Identifer | oai:union.ndltd.org:ADTP/210111 |
| Date | January 2006 |
| Creators | Lele, Stephen, slele@bigpond.net.au |
| Publisher | RMIT University. Applied Sciences |
| Source Sets | Australiasian Digital Theses Program |
| Language | English |
| Detected Language | English |
| Rights | http://www.rmit.edu.au/help/disclaimer, Copyright Stephen Lele |
Page generated in 0.0017 seconds