Please use this identifier to cite or link to this item:
http://hdl.handle.net/11375/21918
Full metadata record
DC Field | Value | Language |
---|---|---|
dc.contributor.advisor | Hamed, M. S. | - |
dc.contributor.author | Afzaal, Umar | - |
dc.date.accessioned | 2017-09-14T15:12:53Z | - |
dc.date.available | 2017-09-14T15:12:53Z | - |
dc.date.issued | 2006-12 | - |
dc.identifier.uri | http://hdl.handle.net/11375/21918 | - |
dc.description.abstract | <p> In North America, heat treating adds about $15 billion per year in value to metal goods by imparting specific properties that are required if parts are to function successfully. Heat treating is an energy-intensive industry, requiring about 500 trillion BTUs (~ 0.5 trillion ft3 of natural gas) per year, which accounts for about 20% of the total cost ofthe business. Considering this huge demand on energy resources and its significant impact on the environment, in the year 1996, members of the heat treating industry represented by the ASM Heat Treating Society and the Metal Treating Institute (MTI) met and discussed the future of the heat treating industry in North America. A vision was developed known as "Heat Treating Industry Vision-2020". In that vision, the industry identified key research areas among which was the development of integrated process models. The industry recognized that most current heat-treating procedures are based on the experience of the heat treater. Trial-and-error often results in operations or components that are functional but not optimized. </p> <p> The present study is concerned with the development of process models of gas nitriding operations using Artificial Neural Networks (ANNs). Data required for the development of ANN s have been acquired from experiments carried out at the industrial partner site, V AC AERO International, Oakville, Ontario. Two types of ANNs have been developed and tested using the experimental data. The two models were able to predict various case depths produced by the nitriding process with reasonable accuracy in the ± 20% range. Predictions of the white layer thickness were in the ± 40% range. The sensitivity of predictions due to measurement errors has been investigated. The range of measurement error of the current study did not have a significant effect on the ANNs predictions. The effect of rate of cooling after the nitriding operation on the developed case depths has also been investigated. Cooling rates in the range of3° F/min to about 20 °F/min were tested. Results indicated that this range of cooling rates do not have a significant effect on the developed case depths. </p> <p> The present study has confirmed that ANNs models have the ability to be trained and applied to multivariable systems which renders ANNs the most suitable tool to develop integrated models for heat treating processes. </p> | en_US |
dc.language.iso | en | en_US |
dc.subject | Gas nitriding | en_US |
dc.subject | Neural Networks | en_US |
dc.subject | heat treating | en_US |
dc.subject | energy resources | en_US |
dc.title | Modeling of Gas nitriding using Artificial Neural Networks | en_US |
dc.contributor.department | Mechanical Engineering | en_US |
dc.description.degreetype | Thesis | en_US |
dc.description.degree | Master of Applied Science (MASc) | en_US |
Appears in Collections: | Digitized Open Access Dissertations and Theses |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
Afzaal_Umar_2006Dec_Masters.pdf | 17.37 MB | Adobe PDF | View/Open |
Items in MacSphere are protected by copyright, with all rights reserved, unless otherwise indicated.