1. Pipeline and Hazardous Material Administration (PHMSA) Department of Transportation EXPERIMENTAL SHOCK TEST DATA ON LARGE LITHIUM BATTERIES Presented at second UN informal working group on large lithium batteries September in Washington, DC Steve Hwang, Ph.D. steve.hwang@dot.gov 1

2. The design type tests specified by the UN manual of Tests and Criteria require large format batteries to be subjected to a half-sine shock of peak acceleration of 50 g and pulse duration of 11 ms. The force required to generate the test conditions may not be indicative of transport or reasonable abuse conditions for large format batteries some of which may exceed 400 Kg. Problem Statement: Study the dynamic loads experienced by large format batteries during transportation and evaluate whether the current UN/DOT 38.3 T4 shock test accurately represents transportation environments If the study found the current test not to be valid, propose criteria and methods for conducting shock test on large format batteries. Study the dynamic loads experienced by large format batteries during transportation and evaluate whether the current UN/DOT 38.3 T4 shock test accurately represents transportation environments If the study found the current test not to be valid, propose criteria and methods for conducting shock test on large format batteries. Objectives: 2

3. Batteries during the four modes of transportation (road, rail, air, and sea) experience a variety of dynamic forces. In general, these forces can be divided into two categories: Introduction The second category covers relatively infrequent, non-repetitive shocks encountered in handling. The most severe mechanical aspects of handling are usually associated with the shocks and transients arising from rough handling, and particularly from the materiel being dropped The first group encompasses the forces that are experienced due to vibration and repeated shocks due to road surface imperfections. These forces result in dynamic deflections of battery components. Dynamic deflections and associated velocities and accelerations may cause or contribute to structural fatigue and mechanical wear of battery components. 3

4. Transportation Mechanical Environments Responses were measured on the vehicle’s load bed over the rear axle The amplitude of transients experienced by restrained cargos is significantly lower than that likely to occur as a result of any mishandling i.e. being dropped Transient responses experienced on a Bedford 4x4 truck on a good quality road Propeller transport aircraft landing shock Transient excitations (shock) are only experienced during landing and peak in the case of fixed wing propeller aircraft. The amplitude of the transients can attain a two g experienced during air transportation will be less severe than that likely to occur as a result of any mishandling i.e. being dropped Transient excitations (shock) are only experienced during landing and peak in the case of fixed wing propeller aircraft. The amplitude of the transients can attain a two g experienced during air transportation will be less severe than that likely to occur as a result of any mishandling i.e. being dropped The maximum reported acceleration for switching operations is 15 g for traditional loose coupled wagons. The amplitude of the transients experienced during rail transportation will be less severe than that likely to occur as a result of any mishandling i.e. being dropped In sea transportation the payload experiences mainly quasi-static loading rather than dynamic motions. The quasi-static inertia loadings are usually of such low magnitude as not to cause any concern 4

5. REFERENCE AccelerationPulse Duration Number of ShocksPulse Form g n milliseconds UN38.3 T.4501118Half sine SAE J2464251518Half sine RTCA DO-160F Airborne Equipment20116Saw tooth USAF ASD-TR-76-30 December 1977 FAA 14CFR 25.5619 Crash data ISO/DIS 12405-1501066 Annex 8 to Regulation No. 100, 02 Series of amendments17-2880 Single step UL2580251518 UL 2271 Half sine ≤ 12 kg5011 >12≤100 kg2515 >100 kg1020 IEC61373 18 Class A & B Body Mounted3.1-5.130 Bogie Mounted30.618 Axle Mounted1026 LITERATURE INFORMATION 5

6. Shock Scenario Acceleration g n Pulse Duration milliseconds Number of ShocksPulse Form Typical wooden packages impacting a wood load platform during carriage over rough roads40 Not providedNot ProvidedNot provided Landing-fixed wing propeller aircraft2 Not provided Rail-Switching operations15 Not provided MIL-STD-810G, Method 516.6 Drop heights range from 18"- 48" depending on weight. The number of drops ranges from 5 to 26 depending on weight. Not provided ADDITIONAL DATA COLLECTED BY THE NAVAL RESEARCH CENTER 6

7. DROP TEST SET-UP EXPERIMENTAL SET-UP 12/2013 7

8. Drop Test Set Up An accelerometer was placed on the base plate to measure the input acceleration. Two accelerometers were placed on the battery to measure the battery top plate response. Battery voltage and temperature were measured. The first set of drops were used to iteratively find the drop height and surface type that result in a 50 g, 11ms, half-sine input acceleration Next the battery underwent drops from 24”, 36”, and 48”. Other impact surfaces were tested besides the surface required for 11ms pulse: (2” wood plate, concrete and 1/2” steel plate) An accelerometer was placed on the base plate to measure the input acceleration. Two accelerometers were placed on the battery to measure the battery top plate response. Battery voltage and temperature were measured. The first set of drops were used to iteratively find the drop height and surface type that result in a 50 g, 11ms, half-sine input acceleration Next the battery underwent drops from 24”, 36”, and 48”. Other impact surfaces were tested besides the surface required for 11ms pulse: (2” wood plate, concrete and 1/2” steel plate) 8

9. Computer Monitors (g n, T, V, Pulse, visual) 9

10. VARIABLES TO BE TESTED Types of Surfaces Mass of Battery Height of Battery being Dropped 10

11. 50G 11ms Drop Test Impact Surface: 2” plywood + 2 rubber mats + 1” foam Battery temperature and voltage remained constant during the test. No venting or leakage was observed. No mechanical deformation was observed. Battery temperature and voltage remained constant during the test. No venting or leakage was observed. No mechanical deformation was observed. 11

12. Various impact surfaces Height: 11” Surface: 2” plywood + 2 rubber mattes + 1” foam Height: 11” Surface: 2” plywood Height: 11” Surface: ½” Steel 12

13. Various Heights Height: 24” Surface: 2” plywood Height: 36” Surface: 2” plywood Height: 48” Surface: 2” plywood X 10 -3 13

14. 14

15. Discussion of T4 Shock Test Pulse width of transients depends on material characteristics of the impact surface and the dropped object. The spectral content of the excitation energy has periodic peaks and notches in the frequency domain. All modes that coincide with the peaks of the frequency response function (FRF) will be preferentially excited, while the modes that coincide with the notches in the excitation FRF will not be excited. Pulse width of transients depends on material characteristics of the impact surface and the dropped object. The spectral content of the excitation energy has periodic peaks and notches in the frequency domain. All modes that coincide with the peaks of the frequency response function (FRF) will be preferentially excited, while the modes that coincide with the notches in the excitation FRF will not be excited. 15

16. Summary/Discussion Testing Indicated that that the fixed acceleration and pulse duration parameters defined in the current T4 shock test could induce responses in test items that are not representative of abuse conditions during transportation. Our data suggest that drop testing is more representative of worst case transportation conditions. Testing Indicated that that the fixed acceleration and pulse duration parameters defined in the current T4 shock test could induce responses in test items that are not representative of abuse conditions during transportation. Our data suggest that drop testing is more representative of worst case transportation conditions. Drop Test50G 11ms Shock Encompasses all the dynamic forces experienced during transportation Not universally representative of transportation environment for every battery design Simpler test apparatusEconomically impractical for large format batteries Repeatability is an issueRepeatability makes it more attractive from a regulatory stand point Further research is needed to define proper test parameters such packaging and drop height 16

17. OBSERVATIONS FROM EXPERIMENTS ● Type of Surface: Pulse duration remains about the same for the same type of drop surface even at different heights. ● Mass: As mass increases, g n decreases for a given surface and height. g n is inversely related to square root of mass. ● Type of Surface: The harder the surface, the higher the g n and the shorter the pulse duration. ● Height: As the height increases, g n increases for a given surface. g n is directly proportional to the drop height. 17

18. Handling Scenario at 24 inch drop on 2” Plywood backed by concrete for a 16 kg battery with a 0.95 ms pulse duration and 358 g n Weight (Kg) 1215202530354045>45 Height (cm) 8165493933282422 18

19. 19

20. Comparison to the US Military Standard (MIL-STD-810G, Method 516.6) 20

21. QUESTIONS?