Engineering software helps to advance battery development
Batteries are complex devices. Everything from a battery’s electrical to chemical, thermal and structural properties must be considered during development. It’s a tough, multi-physics balancing act that simulation and modelling helps to address, providing engineers and scientists with key insights and information throughout the development process and across many, different length scales.
At the atomic level, for example, scientists often use simulation to explore combinations of new materials and novel technologies. Vaida Arcisauskaite, Senior Marketing Specialist at Synopsys, explained: “As new materials are needed to meet evolving standards for both sustainability and performance, atomistic materials modelling with Synopsys QuantumATK helps battery designers explore the countless combinations of materials and select the most promising ones without performing experiments, building, or testing batteries for each candidate, thus shortening development time and cost.”
These novel battery materials could be used in the cathodes and anodes, liquid and solid electrolytes, additives and solid electrolyte interphases (SEI) found in many batteries, for example.
Lithium-ion batteries dominate the market, powering portable devices such as smartphones and laptops to the larger-scale batteries found in electric vehicles (EV) and those used in renewable and grid energy applications to store generated energy.
While lithium-ion batteries are relatively cheap and rechargeable, longer-lasting, more reliable, safer and sustainable batteries need to be developed with faster charging times. This is where scientists focus on advancing current lithium-ion technologies or developing alternative technologies, such as solid-state batteries.
Anders Blom, Solutions Engineer, Senior Staff at Synopsys, explained: “Solid-state batteries in particular hold a lot of promise. Even though they are of similar cost to lithium-ion batteries, they are much safer (non-flammable), more reliable, have a higher capacity and faster charging time, require fewer raw materials, and allow for more charging cycles before degradation starts to happen.”
“Still, there are challenges that designers need to account for, including the suppression of dendrites which can lead to safety issues, lower mechanical stability during cycling, and electrical resistance.”
Lithium-air and lithium-sulphur battery technologies provide “increased capacity, lower cost, and environmental friendliness,” according to Soren Smidstrup, Director of R&D at Synopsys. “However, there is still a lot of R&D needed to address challenges such as improving the stability and cycle life of lithium-sulphur batteries and prolonging the life span and improving the stability of lithium-air batteries.”
Sodium and manganese ion batteries are less flammable, safer, more abundant, and easier to extract alternatives to lithium-ion batteries. Smidstrup added: “However, due to the larger size and thus lower mobility of sodium and manganese ions, there is a need to find suitable electrolytes for efficient ion transport and improve charging rates and capacity.”
This is one of the reasons why atomic-scale modelling is key to helping explore new battery materials and novel technologies. These tools allow for the faster and cheaper development of new batteries at the atomic scale and their associated chemistries. This work is vital to increase the energy density of batteries, helping scientists create faster charging batteries that are also safer and cheaper to develop.
Puneet Sinha, Senior Director of Battery Industry at Siemens Digital Industries Software, explained: “Maximising the energy density of a battery pack by design innovations can maximise the number of cells that can be packed in a given volume without sacrificing battery thermal and structural integrity. Cell-to-pack and cell-to-chassis technologies are such examples.”
At the micro-scale, Computer Aided Engineering (CAE) multi-physics simulations can optimise battery performance. “With simulations, companies can explore new battery designs, evaluate, and validate the impact of material chemistries and cell designs on performance, charging, safety and ageing much faster,” according to Sinha. “They can develop reliable algorithms to measure state of charge, state of health for batteries and ensure safe and secure implementation of these algorithms in their battery management system.”
Simulation software enables the evaluation of battery design and Battery Management Systems (BMS) in a system environment, which is faster, safer, and more cost effective than building and testing physical hardware prototypes.
Danielle Chu, Product Marketing Manager at MathWorks, explained: “The right simulation software provides great design insights at early stages, i.e. trade-off analysis, and exploring design space. Simulation software can also generate a digital twin of battery systems, enabling designers to easily modify battery design and test both normal and abnormal conditions based on simulation results.”
Xiao Hu, Senior Principal Application Engineer at Ansys, explained how this simulation process often works: “Electrochemistry processes occurring inside a pair of electrodes could be simulated using the electrochemistry model. Such a model could help cell manufacturers understand the electrochemical phenomena inside a cell including ageing, and the impact of different chemistry on cell performance.”
When we get to the battery module/pack level, different cooling system designs, including air cooling, liquid cooling, and even immersion cooling, could be evaluated by simulation, according to Hu. “The impact of different thermal management designs in the case of thermal runaway could also be evaluated using simulation software.”
Such work provides valuable insights that are tough to get from testing alone. Sinha added: “For instance, battery safety testing, though absolutely critical, can only tell if a cell or pack passes the safety test but doesn’t necessarily offer why it failed, if it failed, and what to tweak to meet the safety requirements. Battery thermal runaway simulations provide these insights to accelerate material/design optimisation and validation.”
For example, electrochemical impedance spectroscopy (EIS) could be used with modelling and simulation to determine the parameters responsible for cell deterioration and ageing in a specific battery cell. Engineers at Comsol recently used simulations to determine how different battery designs and operating conditions are susceptible to ageing.
“The alternative would be to open the cell and examine the possible changes in the structure, which would be a very difficult and expensive procedure,” according to Ed Fontes, VP of development at Comsol. “It is more efficient to combine experimental measurements with modelling and simulation for characterisation to determine the internal status of a battery than trying to investigate the battery post-mortem. If a battery is failing, the same methods may be used for diagnosis.”
Another way modelling and simulation can be used in battery development is to determine the utilisation of the battery electrodes during charge and discharge. This provides information about how the cell should be designed and operated for a uniform utilisation and, eventually, for a longer life.
“It also provides a detailed distribution of the state-of-charge of a cell at any given point of operation, which is very difficult to achieve without modelling and simulation,” Fontes added. “Using a digital twin to track a battery or battery pack is an efficient method for continuously gathering and quantifying information about its status.”
At the macro-scale, simulations can help companies optimise their battery packing and manufacturing processes to reduce scrap rates and accelerate time-to-market.
Hu explained: “Predictive maintenance can be achieved through the Ansys framework of high-end 3D simulations and physics-based Digital Twin (ROM or Reduced Order Model) technologies. Hybrid analytics by using physics simulation and data-driven models in Ansys can help to create lifetime modelling.”
These simulations can be scaled up to cover entire production plants. Sinha added: “With Siemens digital manufacturing solutions, companies can design plants, optimise and validate cell and pack production processes virtually before implementing on the factory floor, thereby de-risking investment and shortening scale-up time.”
Advancing electrification is driving most of today’s battery developments, with the automotive industry leading the charge. Chu said: “Automotive (electric vehicles, hybrid electric vehicles, and automobiles), is a big market for lithium-ion battery innovation due to the increasing number of electric vehicles and favourable government policies.”
Chu added: “Energy storage and consumer electronics also have a high demand for battery usage, which tends to drive innovation. The main innovation we see is in both electrical management and thermal management. Both aspects are critically important to improve overall efficiency and safety and to extend the operational life of the battery system.”
In automotive, many manufacturers are targeting the problem of recharging speeds. This is because drivers are demanding electric vehicles that recharge in times comparable to the time taken to fuel a gasoline car. Then, drivers also want that rechange to last for hours. This is a big ask.
“The difference in magnitude of the recharge and discharge currents is a problem,” Fontes explained. “The battery, heating, cooling, current collectors and feeders, and many other components have to be able to cope and operate at very different conditions during recharge and discharge. This problem remains unsolved; therefore, a large part of battery research is dedicated to addressing it.”
The aerospace industry is also exploring the use of batteries for hybrid-electric and all-electric aircraft, satellites, and spacecraft. Here, the market demands high-energy-density batteries with a compact size and low weight.
“The growth of EV and consumer electronic markets are driving the research and development of battery technologies that will benefit other industrial applications as well,” Smidstrup added. “They demand batteries that offer higher capacity (thus, longer range), faster charging, and improved safety. In return, novel battery technologies, such as solid-state, could significantly improve the performance and safety factors, and become a game changer in exponentially growing the EV market globally.”
Innovation in wearable/implanted medical devices, such as battery-powered pacemakers, hearing aids, insulin pumps, and neurostimulators, also require the development of even smaller and more efficient batteries.
However, wind and solar power are “probably the largest future areas of application,” according to Fontes, who added: “These are so-called ‘intermittent’ sources of electricity. A possible future application is to use large battery systems to stabilise the grid. This would mean improving the quality of grid electricity, not necessarily storing electricity for the purpose of selling it, since doing so would require very large systems.”
To develop tomorrow’s batteries, simulation and modelling tools will come to increasingly rely on artificial intelligence (AI) and machine learning.
But there’s “a lot of buzz” around AI and machine learning, according to Fontes. “Almost every research institution working with batteries has programs for applying AI to batteries, mainly to predict the status of a battery based on data from operation history. We envision AI being used for different purposes.”
“One obvious use is to create very compact and fast digital twins to make predictions during the operation of battery systems,” Fontes explained. Then, modelling and simulation can be used together with training data to train an AI. This, in principle, is “not very much different” from the traditional validation processes that you will find in parameter estimation and optimisation, where a subset of data is used to train the software. Then another set of data is used to validate the predictions.
Fontes added: “The difference is that we would invest large amounts of computing power in developing and training the AI. Once this is done, an AI-based model could be very fast and compact. It could be used for real-time decision-making on an operating battery. It could also help make decisions regarding the selection of materials and for characterisation, diagnosis, and other methods involved in battery design and operation.”