Figure 2 – Castrol On Thermal Rig schematic layout Figure 3 – Heated representative cells Advertising feature module over time which further impacts test results. To combat this, a timeconsuming cell balancing procedure must be completed regularly which significantly reduces test throughput. Live cells have a finite energy capacity and limits on charge and discharge rate and temperature. This limits the steady state test time and test conditions that can be investigated. Although rare, the separator within a live cell can fail for several reasons which causes an internal short circuit. This is likely to initiate combustion of the cell materials which can spread to the surrounding cells. The provision of safety systems required to deal with an event of this type considerably increases the cost and complexity of the test facility and recovering from such an event can be expensive and result in significant rig downtime. Using heated representative cells produced from non-combustible materials eliminates this risk. Castrol ON Thermal Rig The Castrol ON Thermal Rig has been developed entirely in-house and is shown in Figure 1. It allows for interchangeable model modules, enabling production or 3D printed development parts to be easily integrated and tested. The fluid handing circuit, shown schematically in Figure 2, includes a chiller, heat exchanger, fluid pump, control valves and comprehensive instrumentation to accurately control the boundary conditions of the thermal management fluid delivered to the test module. The rig supports up to 48 heated cells that can be independently controlled, generating heat internally at up to 200W each. Instrumentation includes 12 pressure and 122 temperature channels as well as mass flow meters and pump power consumption measurement. The integrated automation system enables comprehensive testing and systemlevel analysis with full test cycle control and can precisely emulate transient and steady state vehicle operating conditions. In addition, it controls the temperature of the test fluid and test chamber. The data acquisition system can log at a rate of up to 100Hz. Heated Representative Cells While removing live cells from the testing environment offers many advantages the main challenge with the technique is the need to carefully engineer the representative cells to replicate the thermal characteristics of the model they are replacing. This involved first mapping the surface temperature of the real cells under controlled conditions when charged at a range of C rates. The next step was to design an electrically heated version to replicate this performance, supported by simulation models and testing of multiple physical prototype iterations. After manufacture the cells were individually calibrated to determine the heater power required to emulate operation of the real cell at range of C rates. These data were then input into the rig control system. As previously discussed, temperature measurement instrumentation can be incorporated into the representative cell allowing high fidelity insights into the performance of the battery thermal management design such as intra-and inter cell temperatures. This is typically very difficult to achieve with live cells. In this case 3 thermistors were added to opposite sides of each cell as shown in Figure 3. As with all components that are in contact with the thermal management fluid, material compatibility must be ensured to avoid premature failure. To this end all materials used in the cell construction including any seals, adhesives and wiring are assessed for compatibility typically using an accelerated aging test. Conclusions The thermal rig with heated representative cells developed by Castrol is powerful experimental facility to study the performance of battery thermal management systems and fluids, particularly suited to the pack and module development phase. The approach overcomes many inherent difficulties with live cell testing to produce highly repeatable measurements although significant upfront effort is required to develop and control heated cells that faithfully replicate real cell performance. In Part 2 of this series, we will discuss how the method was applied in conjunction with simulation models to assess the performance of a prototype direct cooled battery module and how fluid properties affect this.
RkJQdWJsaXNoZXIy MjI2Mzk4