| About this Abstract |
| Meeting |
Materials Science & Technology 2020
|
| Symposium
|
Computation Assisted Materials Development for Improved Corrosion Resistance
|
| Presentation Title |
Applying Machine Learning to Determine the Corrosion Resistance of Alloys |
| Author(s) |
Szu-Chia Chien, Wolfgang Windl, Gerald Frankel |
| On-Site Speaker (Planned) |
Szu-Chia Chien |
| Abstract Scope |
Designing good corrosion resistant alloys is of utmost importance in many applications. The current guide for determining the corrosion resistance, pitting resistance equivalence number (PREN), provides an empirical value based on the chemical compositions without considering the environmental parameters such as the acidity or temperatures that are believed to be critically important in corrosion. Therefore in this work we have compiled and analyzed a large number of experimental data from the literature on corrosion of alloys. This database includes the alloy composition, electrochemical parameters as well as polarization testing parameters. Machine learning approaches (lasso and ridge regression) were used to find the unexpected correlations between aforementioned experimental parameters and the materials’ properties as well as to identify key features that govern the corrosion resistance. We anticipate that this work can provide a new guidance for rational design of corrosion resistant alloys. |