| About this Abstract |
| Meeting |
Materials Science & Technology 2020
|
| Symposium
|
Advances in Surface Engineering
|
| Presentation Title |
Application of Artificial Neural Network and Statistical Modeling to Study Water Contact Angle of Ductile Iron: Iron-graphite Composite |
| Author(s) |
Amir Kordijazi, Hathibelagal Roshan, Pradeep Rohatgi |
| On-Site Speaker (Planned) |
Amir Kordijazi |
| Abstract Scope |
The effect of graphite percentage, surface roughness, time, and droplet size on the water contact angle (CA) of ductile iron was examined. For design of experiment a full factorial design was utilized including 120 combinations of all factors and their levels. Contact angle values averaged 72°±11° with maximum of 92° and minimum of 46°. Multilayer Perceptron Neural Network Model was used to investigate the correlation between the predictor factors and CA. The results indicate the linear correlation between the predicted and observed values to be 0.756. The result also shows that the surface roughness is the most important predictor in CA variation followed by elapsed time, droplet size, and graphite percentage. In addition to ANN, multi linear and polynomial regression analysis were carried out. The result shows that CA increases by increasing surface roughness, graphite percentage, and time. This suggest that the ductile iron surface follow a quasi Cassie-Baxter regime. |