« back to listing
Getting to grips with vibration requirements for EV batteries
Jim Hooper, senior engineer at Millbrook, discusses the lack of data to validate the impact of vibration on EV batteries, recent test and development work to identify these gaps and recommendations for future testing
Over the past decade the focus on developing EVs, all of which have some form of on-board high voltage electrical energy storage system, has increased significantly, due to the demand for passenger vehicles with increased fuel economy and reduced emissions. This demand is being driven by an increased customer awareness of green issues, the rising cost of traditional liquid fuels and government legislation.
A review of academic research suggests that some test standards and legislative requirements defined for the vibration and shock testing of EV battery packs have been derived of data from consumer electronics or transposed from data from more traditional IC engine vehicles. As a result there is every possibility that EV battery packs are being either over engineered (to compensate for vibration inputs that they would otherwise never be subjected to, thus increasing weight and cost) or under engineered (the battery component is not subjected to the correct in-market vibration and shock loads, therefore potentially increasing the risk of safety related failure).
In order to improve our understanding of vibration requirements in EV batteries, Millbrook has carried out a number of tests on a range of EV batteries, by subjecting them to ‘real world’ road surfaces and rough road abuse environments. This was conducted through the instrumentation and measurement of the vibration behavior of high voltage battery packs installed to a selection of current production HEVs and BEVs when driven over the repeatable surfaces at Millbrook.
The data was collected from tri-axial accelerometers at six locations of the vehicle, including both A Pillars, from the front and rear of the battery packs and from the front and rear of the chassis. These components were assessed over surfaces from Millbrook’s structural durability procedure, which simulates 10 years of durability for a typical European passenger car. To ensure consistency across the full range of tests, all surfaces were recorded at the same speed, using the same driver and carrying the same passenger.
Initial tests carried out in the study revealed some interesting results that differ considerably from what we currently know about vibration in electric vehicle batteries. For example, we identified that the bulk of vibration energy in EV battery packs occurs from 0 to 100Hz; it was previously thought to occur between 0 to 50Hz. It was also noted that current engineering specifications for battery vibration durability typically test from 7Hz upwards, however Millbrook’s study has highlighted the need to consider frequencies below 7Hz, as significant vibration energy occurred from 0 to 5Hz in all BEVs tested.
RESS vibration testing
An academic review and evaluation of current rechargeable energy storage system (RESS) vibration test procedures highlighted that typically the X and Y axis of EV battery packs are evaluated using the same test profile, however the X and Y vibration behavior of the EV battery packs evaluated in our research differed significantly. Therefore Millbrook would recommend that the X and Y axis of an EV pack be evaluated using specific vibration data, based on vibration measurements received in in-vehicle conditions.
Powertrain induced vibration testing
A mode typical of powertrain-induced vibration was noted in all EVs between 7 to 20Hz. High-energy spikes were also noted in some EVs above 300Hz, which is likely to be caused by either Powertrain-induced vibration from the electric motor or transmission assemblies or from the effect of vibration induced by different cooling strategies or power electronics.
Further research is required to identify the exact cause of these vibration energy spikes above 300Hz. This study also highlighted that if it was determined in future that vibration was being introduced to the pack through cooling strategies, the pack’s vibration life has the potential to be affected by different climates and driving styles.
There is evidence to suggest that due to the balanced weight distribution of EVs, lower vibration energy levels are typically witnessed within these vehicles than their ICV counterparts.
Body torsion testing
Evidence was also found to suggest that vibration behavior at 20 to 40Hz in EV battery packs is the result of a body torsion mode, due to the pack forming a large structural member of the chassis. This could be validated by further investigation of the pack and chassis vibration response via a modal analysis in a laboratory. However, some resonances between 20 to 40Hz were attributed to the vibration isolation strategies employed by manufacturers to isolate the pack from chassis vibration, therefore it is recommended that engineers consider the battery pack to chassis integration as the packs installation can affect the frequencies and vibration energy it will see during its service life.
Test standards J2380 and BS62660 have been compared to data recorded from current EV passenger vehicles currently on sale in the UK market. The recorded vibration data was sequenced to a known durability test schedule that correlated to represent 10 years durability and 100,000 miles of representative rough road surfaces. Both methods of data sequencing resulted in the conclusion that both test standards are too severe when used to determine the durability of a RESS of a European passenger vehicle.
An investigation into the effect of different weightings of road classifications was undertaken as part of Millbrook’s study and it has concluded that BEVs, designed for an urban and rural environment will be subjected to greater vibration energy from 0 to 399Hz than a BEV designed for a typical market operation weighting similar to that of an ICV.
Results from Millbrook’s initial investigation and ongoing research into the vibration requirements of EV batteries has the potential to benefit the accuracy of component level safety and durability testing and could aid the future weight reduction of EV battery packs. We are currently working with Dr James Marco at Cranfield University to roll out the results of our initial investigations and will be publishing further information about our research and test results later this year.
3 April 2013