About this Abstract |
Meeting |
TMS Specialty Congress 2024
|
Symposium
|
2nd World Congress on Artificial Intelligence in Materials & Manufacturing (AIM 2024)
|
Presentation Title |
Modeling the Damage Healing of Self-healing Polymer Using Zero Bias Deep Learning Approach |
Author(s) |
Palvi FNU, Deepak Kumar, Foram Madiyar, Yongxin Liu, Sirish Namilae |
On-Site Speaker (Planned) |
Sirish Namilae |
Abstract Scope |
Intrinsic self-healing polymers are inherently self-repairing materials that do not require external stimuli such as heat or pressure. These materials find diverse applications in domains like aerospace and healthcare. In this study, we created a self-healing polymer recovery dataset using in-situ microscopic characterization of the damage healing process. Subsequently, an explainable AI approach, the zero-bias deep neural network (ZBDNN) model, is employed to analyze this data. ZB DNN modifies the final dense layer of the standard DNN model into a dimensionality reduction layer and a similarity-matching layer. This model provides a metric for characterizing the level of damage in each image based on the Mahalanobis distance between the feature vectors of healed and non-healed inputs. The model can be utilized to analyze and predict the temporal evolution of the healing process. |
Proceedings Inclusion? |
Definite: Other |