Using Machine Learning to Predict Damage Tolerance in Unstable Systems: a Case of Impacted Batteries and Exploding Materials

Interdisciplinary Areas: Data/Information/Computation, Power, Energy, and the Environment

Project Description

Unstable systems are defined as systems that respond to applied stimuli in a way that significantly endangers surroundings. This includes nuclear materials, lithium ion batteries, and energetic materials. Significant advances have been made in designing such systems in a way that initial design is safe and performance is guaranteed. However, influence of a damage event such as an impact or extreme vibrations at high temperatures etc. significantly changes the operational landscape of such systems. The proposed work builds upon recent advances made using mechanical Raman spectroscopy experiments by Tomar and co-workers, and mesoscale modeling and stochastics in energy storage materials by Mukherjee and co-workers, to (1) delve into classical non-equilibrium foundations for defining and separating critical vs. non-critical instabilities in such systems using experiments with specific emphasis on Li-ion batteries, (2) designing new breed to sensors that replicate gained knowledge using machine learning to predict influence of instabilities on unstable system operation, and (3) developing new machine learning informed framework that integrates predictive modeling with underlying physics based on knowledge in (1) and (2). Emphasis is on addressing and removing inherent brittleness in machine learning protocols to include problem physics in such important material systems.

Start Date

March 1, 2019

Postdoc Qualifications

Candidates should have earned a doctorate in Physics, Chemical Engineering, Chemistry, Aerospace Engineering, Materials or Mechanical engineering or a related field. The successful candidate will have experience in performing material characterization and analytical analyses. Experience with numerical analyses is desirable. Candidate should have a strong publication record. Good communication skills and ability to work well in a group are desired. 

Co-advisors

Vikas Tomar, tomar@purdue.edu, School of Aeronautics and Astronautics, www.interfacialmultiphysics.com

Partha P. Mukherjee, pmukherjee@purdue.edu, School of Mechanical Engineering, https://engineering.purdue.edu/ETSL/

References

1. M. Gan, V. Tomar, An in situ Platform for the Investigation of Raman Shift in Micro-scale Silicon Structures as a Function of Mechanical Stress and Temperature Increase, AIP Review of Scientific Instruments, 2014, 85, 013902.

2. D.P. Mohanty, H. Wang, M. Okuniewski, V.Tomar, A nanomechanical Raman spectroscopy based assessment of stress distribution in irradiated and corroded SiC, J. Nuclear Materials, 2017, 497, 128. 
 
3. B. Li, R.A. Adams, J. Kazmi, A. Dhiman, T.E. Adams, V.G. Pol, V. Tomar., Investigation of LiCoO2 Cathode Response to Dynamic Impact Using Raman Imaging Based Analyses, The Journal of The Minerals, Metals & Materials Society, 2018, 201805029.
 
4. R. A. Adams, B. Li, J. Kazmi, T. E. Adams, V. Tomar, V. G. Pol, Dynamic Impact of LiCoO2 Electrodes for Li-ion Battery Ageing Evaluation, Electrochimica Acta, 2018 (EA18-1601R1), in press
 
5. A. Mistry, K. Smith, and P. P. Mukherjee, “Secondary Phase Stochastics in Li-Ion Battery Electrodes,” ACS Applied Materials and Interfaces, 10, 6317 (2018).
 
6. N. Kotak, P. Barai, A. Verma, A. Mistry, and P. P. Mukherjee, “Electrochemistry-Mechanics Coupling in Intercalation Electrodes,” Journal of the Electrochemical Society, 165, A1064 (2018).
 
7. F. Hao, A. Verma, and P. P. Mukherjee, “Mechanistic Insight into Dendrite-SEI Interactions for Lithium Metal Electrodes,” Journal of Materials Chemistry A, DOI: 10.1039/C8TA07997H (2018).
 
8. P. Barai and P. P. Mukherjee, “Mechano-Electrochemical Stochastics in High-Capacity Electrodes for Energy Storage,” Journal of the Electrochemical Society, 163, A1120 (2016).