Predicting System Performance with Aged Li-ion Batteries Using GT-AutoLion and GT-SUITE

As demonstrated in a previous blog, GT-AutoLion and GT-SUITE can be used together to predict how a Li-ion battery will degrade over time while considering any use case, weather condition, and even charging patterns.

Up to this point, battery degradation has primarily been presented as the change in the capacity of a battery over time.  This only tells a small portion of the full story for two reasons.  First, batteries also see an increase in resistance over time (which GT-AutoLion is able to capture).  Second, consumers are not interested in how the capacity and resistance of their battery change over time, they are interested in how fast their vehicle can accelerate to 60 mph, how many logs their electric chain saw can cut between charges, how many pictures can be taken and posted between charges on their cell phone, etc.

Luckily, with GT-AutoLion and GT-SUITE, understanding not only how Li-ion batteries degrade over time, but how that affects system-level performance is very straightforward.

Inserting an Aged Battery into A System-Level Model

When running an aging simulation of any kind, GT-Autolion stores an external file (.state) that details the state of the Li-ion cell at every cycle of the aging scenario (here a “cycle” can be a charge-discharge cycle, a vehicle’s drive cycle, an aircraft’s flight cycle, or a metric of time like days or weeks).  This external file generated by the AutoLion aging model can then be inserted into a system-level model to predict how the aged battery will influence product performance.  Using this, system-level models have a straightforward workflow to predict system performance after 100, 200, or 300 cycles;  1, 2, or 3 years of realistic operation; or even, in the case of automotive applications, after 12,000, 24,000, or 36,000 miles. That is, engineers utilize physics-based models to gain insight into the performance of their products once the battery is aged.

Combining this concept and the ideas presented in a previous blog post about using simulation to predict battery aging in real-world applications, GT-AutoLion and GT-SUITE work together seamlessly to capture how system-level performance will degrade over time. The figure below shows how this process works.

GT-AutoLion-and-GT-SUITE-Workflow
Summary of workflow using GT-SUITE and GT-AutoLion .state file to predict performance degradation over time

Example: BEV Performance Degradation

To demonstrate this workflow workflow, we have selected to build a model of a battery electric vehicle (BEV) in GT-SUITE.  This BEV model contains an accurate model of the powertrain, heating ventilation air and conditioning (HVAC) system, and cabin to understand tradeoffs between vehicle range and passenger comfort.  This BEV model can accurately capture the power draw from the battery and the temperature of the battery during any load condition (for example a drive cycle or commute pattern from GT-RealDrive).

BEV-HVAC-and-battery
Summary of process presented for a Battery Electric Vehicle application

With this workflow, automotive OEMs have the ability to take standard cell-level laboratory tests, such as calendar and cycle aging tests, and predict more meaningful system-level performance metrics over the lifetime of a battery electric vehicle, such as the BEV’s range and acceleration performance (here measured in 0 to 60 mph time).

From-Battery-Model-to-Meaningful-System-Performance
Summary of process presented for a Battery Electric Vehicle application

With the seamless workflows available between GT-SUITE’s advanced system-level modeling and GT-AutoLion’s accurate Li-ion battery modeling, engineers understand how system-level decisions (such as vehicle topology decisions, thermal topology decisions, and controls decisions) can affect the long-term degradation of not only the battery but the performance of the entire system.

Written By: Joe Wimmer