1	Bulb Usage Analysis of LED Automotive Lighting John C. Steiner Roger Clark Associates, LLC Neil E. Clark Roger Clark Associates, LLC David R. Thom Head Protection Research Laboratory
2	Overview History of Incandescent Light Analysis
3	Overview Extensive Research into Bulb Filament Analysis has been published: “Advances in the Analysis of Aircraft Crash Impacted Light Bulbs”; T.W. Heaslip, M. Vermij, M.R. Poole; International Society of Air Safety Investigators (ISASI) Baker, J. Stannard; Aycock, T.; Lindquist, T.; “Lamp Examination for On or Off in Traffic Accidents”; Topic 823 of the Traffic Accident Investigations Manual; Traffic Institute, Northwestern University; 1985 “Accident Reconstruction – Response of Halogen Light Filaments during Vehicle Collisions”; T.R. Fries, R.O. Lapp; SAE #890856; 1989 Just to list a few…
4	Introduction to LED Lamps LED or Light Emitting Diode composed of: Anode Post Bond Wire Semiconductor (LED) Chip Cathode
5	Introduction to LED Lamps LEDs operate on voltages of 1 to 4 Volts LEDs draw 10 to 40 milliamperes LED body is plastic and solid throughout and acts as a lens Internal components are very well supported
6	Benefits of LED Lamps Advantages over incandescent bulbs Draw very low current Reach full intensity almost instantaneously Extremely long service lives Shock Resistant
7	Passenger Vehicles Equipped with LED Exterior Lamps Cadillac Deville
8	Passenger Vehicles Equipped with LED Exterior Lamps Infinity G35 Coupe
9	Motorcycles equipped with LED Lamps Honda ST1300 ABS Touring Bike
10	Commercial Vehicles Equipped with LED Lamps
11	Test Overview “Accelerations and Shock Load Characteristics of Tail Lamps From Full-Scale Automotive Rear Impact Collisions”, L. Johnson, 2002 Four 1989 Ford Taurus automobiles were used as target vehicles in two full engagement rear-end tests and two 50% offset rear-end tests The Target vehicle used was a Class 8 1960 Kenworth 6x4 truck for two tests and a 4,000 lbf flat-faced moving barrier were used for the other two tests Accelerometers were mounted in the taillamp assemblies to measure accelerations during the crash
12	Test Overview Johnson recorded acceleration levels in the taillights
13	Test Samples Ten (10) identical LED CHMSL’s were used in this study Each lamp was composed of 44 individual LED bulbs mounted on a common circuit board The bulbs were wired in series in discrete sets of four, for a total of eleven groups of four LEDs
14	Test Equipment Impacts were completed on the test subjects using an instrumented twin-guide wire drop test device at the Head Protection Research Laboratory
15	Test Fixture Spherical Impactor was modified to support a lamp test fixture which a pair of LED CHMSL lamps were mounted Two standard trailer tail lamp housings containing one 1157 bulb each were also mounted with the CHMSL LED lamps The fixture was constructed to allow the spherical impactor to contact a hard rubber anvil mounted on the impact tester, without direct contact to the lamps
16	Test Equipment The Anvil, or Modular Elastomer Programmer (MEP) is a cylindrical-shaped pad with a durometer value of 60 ± 2 Shore A
17	Instrumentation Uniaxial Endevco accelerometer National Instruments Data Acquisition Card Data Acquisition software by Biokinetics version 3.0 Sampling rate of 10 kHz for 50 msec, and was filtered per SAE J211 Class 600 (SAE) Filter
18	Lamp Power Supply Power supplied by a Deep Cycle 12-Volt Battery Battery voltage maintained at 12.6V ±0.1V throughout the testing by using Battery Charger
19	Test Modes 5 tests completed at increasingly greater drop heights Each test completed one pair of LED CHMSL lamps and one pair of 1157 lamps – one ON and the other OFF
20	Test Results Test #1 Volts=12.6 Drop Height = 1 m V=4.4 m/s g=231
21	Test Results Test #2 Volts=12.6 Drop Height = 2 m V=6.2 m/s g=396
22	Test Results Test #3 Volts=12.6 Drop Height = 3 m V=7.1 m/s g=502
23	Test Results Test #4 Volts=12.6 Drop Height = 4 m V=8.63 m/s g=684
24	Test Anomalies Test #5 Involved Direct Contact of two LED CHMSL Lamp assemblies One powered ON and the other OFF Impact of lamps with two extruded steel box sections mounted on 2”x4” pieces of wood Duration exceeded allowable 50 msec window, impact acceleration data is unknown
25	Test Results In all non-contact tests, LED lamps were not effected by the subjected impacts Lamps subjected to direct contact impact still operated unless completely crushed Crushed LED lamps were analyzed
26	Post-Test LED Lamp Analysis Stereo Microscope Analysis of severely damaged LED bulbs revealed no noticeable indicators Guide wire missing or too small to observe
27	Post-Test LED Lamp Analysis LED Samples cut into 1” lengths for analysis using a Scanning Electron Microscope (SEM) SEM used was a Cambridge 360 at approximately 5.1kV
28	Post-Test LED Lamp Analysis Internal LED components could not be viewed with SEM due to the LED bulb being plastic and non-conductive If the bulb was fractured exposing the internal components then the die wire, LED die, anode and cathode post may be viewed In the cases where internal components could be viewed, we were unable to compare to an undamaged LED
29	SEM Image of LED Die
30	Conclusions Due to the fully potted construction of LEDs, the internal components are well supported and protected from shock Due to the physical size of the LED internal components, analysis is difficult Even when LEDs are fractured so as to expose the internal components, witness marks to indicate the status of the LED at impact could not be found If further research is unable to outline an analysis method of LED-type lighting, monitoring of this data could be completed with Event Data Recorders
31	Suggestions for future research Controlled destructive tests should be performed on powered LEDs The anode posts could be extracted from bulbs in both states of operation, and the conditions of any remnants of the bond wires at the tips of the post could be examined for differences in fracture characteristics It is possible that the delicate bond wire that connects the anode post to the cathode die might exhibit detectable differences in ductile fracturing, depending on whether the wire fails when it is conducting current or not
32	Contact Information For further questions, please contact the authors at: John C. Steiner – jsteiner@kevaeng.com Neil E. Clark – neclark@rc-recon.com David R. Thom – dthom@ci-dynamics.com