Zheng Liu

Project 1: Design and Optimization for Battery Thermal Management Systems

Funding Source: NSF Center for Power of Optimization of Electro–Thermal Systems (POETS)

1.1 Control Co-Design Optimization for Battery Thermal Management Systems

• Established experimental setup with direct and indirect liquid cooling for cylindrical and pouch cells

• Developed COMSOL simulation models based on experiment testing to accelerate data generation

• Formulated surrogate models to investigate the influence of control and plant design on efficiency

• Achieved over 10% cooling system energy saving for the indirect liquid cooling system for pouch cell

1.2 Reliability-based Design Optimization for Battery Thermal Management Systems

• Investigated the uncertainties in the thermal management system to ensure the reliability of batteries

• Constructed high-fidelity Gaussian process models for the system with the estimation of uncertainty

• Adapted Monte Carlo simulations and Latin hypercube sampling methods for the complex system

• Devised a reliable thermal management system with 90% saving in system energy consumption 

1.3 Generative AI for Optimization of Battery Thermal Management Systems

• Implemented a generative design method based on variational autoencoder for battery cell layouts

• Analyzed useful properties from existing battery cell layouts and performance metrics autonomously

• Adapted cooling-guided diffusion model to generate 700,000+ battery cells layouts of a 40-cell pack

• Secured over 95% energy saving with the identified optimal layout design and validated via simulation 

1.4 Physics-informed Neural Networks for Battery Pack Temperature Estimation

• Simplified the complex 3D battery indirect liquid cooling system using a 2D representation 

• Proposed a multi-fidelity physics-informed convolutional neural network for temperature estimation

• Integrated heat transfer partial differential equation backbone with a high-fidelity projection head

• Accomplished high-accuracy temperature estimation with a small amount of training data 

Project 2: Life Cycle Assessment for Recyclable High Energy Density Batteries

Funding Source: NSF Future Manufacturing (Award CMMI-2037898)

2.1 Life Cycle Assessment and Decision Analysis for Cathode Active Materials

• Streamlined the hydrometallurgical and direct recycling methods with standardized processes

• Assessed the environmental impact of each process based on the ReCiPe method with unified reactors

• Evaluated the environmental impact of different cathodes based on specific capacity and lifespan

• Initiated online life cycle inventory of Li-ion battery comparing different recycling methods 

2.2 Life Cycle Assessment and Optimization for Solid-State Battery

• Standardized the manufacturing process for different solid-state electrolytes via tape casting

• Scrutinized the environmental impact of different solid electrolytes with different layer thickness

• Cultivating the failure model for solid-state batteries, considering dendrite nucleation and growth

• Optimizing the structure to balance the environmental impact, performance, and reliability 

2.3 Techno-Economic Analysis for Advanced Battery Manufacturing and Recycling Methods

• Conducted techno-economic analysis of lithium extraction from brine with CapEx and OpEx

• Contributed to the team, winning $1M DOE American-Made Geothermal Lithium Extraction Prize

• Inspected the cost and environmental impact of single-step electrochemical recovery for LCO battery

• Investigating the pretreatment process for cathode scraps for recycling of PVDF and carbon black 

Project 3: Design and Optimization for Additive Manufacturing Applications

Funding Source: NSF POETS (Cooperative Agreement No. EEC-1449548)

3.1 Uncertainty Quantification of Additively Manufactured Cellular Materials for Energy Absorption

• Accelerated uncertainty quantification by integrating physics-based and data-driven surrogate models

• Conducted compression testing of additively manufactured triply periodic minimal surface structures

• Established high-fidelity simulation model of deformation and coupled with data-driven method

• Delivered insights for the design and decision-making of cellular materials for energy absorption 

3.2 Reliability-based Design Optimization of Additive Manufacturing for Battery Silicon Anode

• Reviewed uncertainties in the manufacturing process and working conditions for silicon anode

• Originated multiphysics finite element models to investigate the volume fluctuation of silicon

• Employed reliability-based design optimization based on Gaussian process surrogate models

• Introduced new silicon anode design with over 10% improvements in total capacity and rate capability 

Project 4: Design for Reliable Semiconductor Devices 

Funding Source: NSF POETS (Cooperative Agreement No. EEC-1449548)

• Engineered high-fidelity simulation models to predict void formation through solder electromigration

• Devised surrogate models to predict the mean time to failure of the solder under stress and current

• Utilized conformal inference techniques to enhance the reliability of uncertainty management

• Configuring reliable solder design based on a data-driven approach and providing health diagnosis  

Project 5: Additive Manufacturing via Digital Light Processing

Funding Source: Meta Reality Labs

5.1 3D Printing of Viscoelastic Nanoparticle Suspensions via Digital Light Processing

• Synthesized resins from different functionalized silica nanoparticles suspended in polydimethylsiloxane

• Increased the weight fraction of functionalized silica nanoparticles in shear-thinning resin

• Tested the viscosities, storage moduli, and loss moduli of synthesized resins using a rheometer

• Adjusted printer’s light intensity and increased resin temperature for tough silicon manufacturing 

5.2 Acoustophoretic Liquefaction for 3D Printing Ultrahigh-Viscosity Nanoparticle Suspensions

• Applied acoustic agitation to ultrahigh-viscosity nanoparticle suspensions to improve printing quality

• Measured the movement of resin tray under acoustic agitation by a confocal chromatic system

• Simulated the resin’s velocities under 40 different acoustic frequencies during printing in COMSOL

• Improved the printed feature resolution by 25% and increases printable object sizes by over 50 times 

5.3 Digital Light Processing of Liquid Crystal Elastomers for Self-Sensing Artificial Muscles

• Rebuilt durable resin tray with polytetrafluoroethylene and build window with polymethylpentene

• Reprogrammed printing algorithm to achieve sliding motion for shear flow–induced alignment

• Calibrated the approaching speed of sliding motion to enhance uniaxial shear-induced alignment

• Satisfied the high specific work capacity and energy density of self-sensing artificial muscles 

Project 6: Design and Manufacturing for Soft Robotics

6.1 Shear Thickening Fluid Valve for Independent Control of Haptic Devices

• Utilized acoustic agitation to cause shear thickening fluid jamming to create a solid-like stiffness

• Processed 200+ shear thickening fluids from silica and polyethylene glycol, tested their viscosities

• Inspected the flow rate under piezo’s different acoustic frequencies provided by a function generator  

6.2 RoboBee Actuated by Combustion and Fabricated by Soft Materials

• Designed and tested the rubber valves, enabling exhaust out and fuel in during intervals of combustion

• Experimented the inflation frequency of the membranes using a high-speed camera and force sensor

• Refined the structure to reduce the weight and increase the inflation frequency for RoboBee 

SE 101 Engineering Graphics & Design (Taught Lab Sessions)

Graduate Teaching Assistant       Department of Industrial & Enterprise Systems Engineering, UIUC

Instructor: Professor Molly H. Goldstein

• Delivered guest lecture on additive manufacturing, inspired students to explore the engineering field

• Hosted sketching lab sessions, helped students master freehand sketching techniques in engineering

• Conducted CAD lab sessions, guided students using Autodesk Fusion 360 to draw complex machines

• Facilitated office hours to help students with problems in course material and homework

• Graded homework, quizzes, and exams, provided detailed feedback for students to improve

Mentoring for Battery Related Research Projects

Mentor               Reliability Analysis and Safety Assurance Laboratory, UIUC

• Guided 3 master’s students, 3 undergraduate students, and 1 high school student on research projects

• Navigating the students in understanding battery systems on chemistry and management systems

• Tutoring the students in experimental testing, finite element simulations, and data analytics

• Helping the students explore the electric vehicle industry for job application, providing career advice 

Mentoring for Institute of Industrial and Systems Engineers

Mentor                    Institute of Industrial and Systems Engineers, UIUC

• Mentored over 10 undergraduate and graduate students, enhancing their professional readiness

• Providing tailored advice on industry insights, networking strategies, and skill development

• Assisting the students in crafting personalized study plans and navigating their career paths

• Tracking mentees' progress actively, offering feedback and adjustments for continuous improvement 

Physics-informed Reinforcement Learning for Supply Air Temperature Settings

AI Scientist Intern   Nantum AI. New York, NY

Manager: Dr. Ali Mehmani

• Established an adaptable Modelica model of a 22-floor building with variable air volume systems

• Applied reinforcement learning coupled with physical information to optimize supply air temperature

• Reduced the energy consumption of the building HVAC to comply with Local Law 97 of NYC

• Implemented fault detection algorithm to existing HVAC control systems for skyscrapers 

Low-cost On-device Detection for Boating Equipment Devices Based on ML

Electrification Engineering Intern       Brunswick Corporation. Champaign, IL

Manager: Brennan Kelly, Trevor George

• Formulated the pipeline for hardware assembles, tests procedure, data collecting, and data processing

• Completed comprehensive tests of devices under different environments

• Generated machine learning algorithm to achieve the detection of different events for boating devices

• Optimized data processing procedure to balance the performance and cost of STM32 microcontrollers

• Adapted optimization algorithm for real-time path planning of electric boat autonomous trailering