For online participants: Meeting Link
Increased driving range and enhanced fast charging capabilities are acknowledged as the immediate goals of transport electrification. However, these two objectives are at loggerheads with each other, since they place demands on improving two contrasting aspects of vehicular pouch-cell design, viz. their energy and power densities. By varying the number of layers versus the volume of active electrode material, bespoke pouch-cell designs targeting either of these goals can be obtained. Attempting this design trade-off through iterative empirical testing of layer choices is expensive and often leads to sub-optimal designs. This talk presents the author’s research towards developing a computational framework that employs a model-based methodology for determining the optimal number of electrochemical layers. The modelling objective is to maximise the usable energy whilst satisfying specific acceleration and fast charging targets.
Currently, the model developed thus far is able to handle the critical need to avoid lithium plating during fast charging and accounts for a range of thermal conditions. The modelling framework also takes into account the electrochemical and kinetic phenomena at the micro-scale using a hybrid Finite Element (FE)-spectral scheme, whilst propagating the results upwards to higher length scales for cell-level system design. Drawing upon inferences from his recent research into Hierarchical Multi-Scale Modelling (HMM) of composite materials, the author shall conclude the talk by presenting preliminary results from his hypothesis that coupling the micro-scale FE quadrature points to nano-scale phenomena shall, for the first time, help to quantify the influence of electrode cracking on cell capacity degradation.
Dr Krishnakumar Gopalakrishnan is a Senior Research Software Engineer at the Dept of Advanced Research Computing (ARC), University College London (UCL) in the UK where he works on high performance scientific computing (HPC) applications across a range of computational modelling research projects. Prior to this, he held a 2-year post doctoral research fellowship in scientific computing at the Centre for Computational Science (CCS) at University College London (UCL), UK, and has served as a visiting researcher at the University of Konstanz, and Rutherford Appleton Laboratories (RAL), UK.He holds a BTech degree in Electrical and Electronics Engineering from College of Engineering, Thiruvananthapuram, an MS degree in Electrical Engineering (power electronics systems & control) from the Center for Power Electronics Systems (CPES) labs at Virginia Tech. Later, he won a US Dept of Energy GATE fellowship to complete a graduate certificate program in electric drivetrain automation. Dr Gopalakrishnan received his PhD degree in Mechanical Engineering (mathematical modelling of lithium ion batteries) from Imperial College London. He was formerly employed at ABB Innovation Labs (Bangalore, India) and ABB Corporate Research Center (Baden-Dättwil, Switzerland). He has also served as a power management algorithms and systems engineer at Qualcomm Inc. (San Diego, USA) where he successfully filed corporate patents on novel Battery Management Systems (BMS) designs. He was awarded the President’s PhD Scholarship at Imperial College London and is a Mathworks Certified Matlab Associate.
Dr Gopalakrishnan has over a decade of teaching experience at various levels. At UCL, he currently teaches the University’s scientific computing with C++ course and is leading the course development effort and teaching plan for UCL ARC’s first Massively Open Online Course (MOOC) on the FutureLearn platform. He has also had the privelege of teaching Imperial College London’s first MOOC on Mathematics Essentials (for business majors) hosted on the EdX platform. He has also served as a teaching fellow at Imperial College London’s Computational Methods hub, wherein he was the lead instructor for several scientific computing courses taught to a campus-wide audience.
Dr Gopalakrishnan has published several well-cited research articles in peer-reviewed journals (including in Nature Computational Science) and technical whitepapers, and has presented at UK and international conferences. His research software engineering interests include heterogeneous and GPU computing, parallel and threaded programming, linear algebra libraries, low-latency network communications, and Unix systems administration. His scientific research interests include computational modelling of dynamic systems, power electronics and control, energy storage, non-linear optimisation, feedback control, signal processing, numerical methods and state estimation.