Update of SAE J 2908 Activities Mike Duoba

Update of SAE J 2908 Activities Mike Duoba Argonne National Laboratory 19 October 2015 1

SAE J 2908 History J 2907 – Motor Ratings J 2908 – System Rating • J 2907/J 2908 originally given SAE “J-doc” numbers more than 10 years ago, due to lack of progress, committee was tabled • 2013: some meetings were held to restart J 2907 • Dec 2013: J 2908 committee started recruiting members, regular monthly meetings in 2014 and 2015 • July 2014: Argonne started doing exploratory chassis dyno testing • March – July 2015: rented hub dyno for testing 2

Basic Elements of J 2908 good progress 1. Run a Test good progress 2. Collect Data problems 3. Report Rating – Based upon test data & possibly other information 3

1. Run Test • Prep requirements varied slightly for each HEV design (SOC and Temps) • Found a few test procedure options to best achieve peak total powertrain power for different types of HEVs • Hub dyno and chassis dyno (w/axle torque) were found to be equivalent Hub Dyno or Chassis Dyno or Wheel or axle torque sensor needed for power direct power measurements 4

Different Test Types Were Developed for Different Powertrain Operation Fixed-Speed mode Road load mode Low-Inertia, High ABC loads Road load mode Notes: - When speeds are changing quickly, inertia effects lower measured power. - Fixed Speed and Terminal Velocity tests are preferable. 5

2. Collect Data • Measure DC electric power – from battery – From accessories • Monitor Powertrain Parameters – Engine: RPM, MAP, MAF, injection timing, reported torque – Other: gear #, powertrain states, reported RPM and torques of motors • Measure Axle or Wheel Power – May or may not be required Current and Voltage Measurements Vehicle Network Data Dynamometer Data 6

3. Report Rating • J 2908 is Currently stuck here! • Conflict between two objectives: – Value to Consumer (rating allows comparisons to other vehicles) – Ability to Verify (is information from OEM required? ) • Reporting options do not satisfy both objectives Ability to Verify Useful to Consumer (can compare to ICE Vehicles) 7

From What Information is the “Net System Power” Rating Reported? ? Three methods emerged Battery Acc Engine Powerestimated + Method 1 DC Powermeasured MG 2 MG 1 Eng S Shaft Powersestimated Method 2 Any Configuration Axle/Wheel Powermeasured Method 3 8

Deciding on a Solution • If verifiability is of most importance, Method 3 will provide best data without reliance on estimations • If 100% compatibility with conventional vehicle (engine-only) ratings are of most importance, Method 2 provides that rating • Method 1 is partially compatible with existing ratings and partially verifiable 9

Application of Methods on Parallel HEV • Engine appears to be at rated power (from data) • Method 1 and Method 2 are similar in value (small battery assist) • As expected, Method 3 provides much a lower power rating than the other two. Axlemeas = 125 k. W Engine assumed to be at rated power here Elecmeas = 35 k. W Motorest = 29 k. W Current OEM Catalog Ratings Engine: 123. 7 k. W Motor: 30 k. W (mechanical) Total: 153. 7 k. W Method (1) 123. 7 + 35 = 158. 7 k. W (engine rating + measured battery power) “Method (2) 123. 7 + 29 = 152. 7 k. W (engine rating + reported motor power) Method (3) Total: 125 k. W Time (measured axle k. W) 10

Application of Methods on Series HEV • • Engine appears to be at rated power (from data) Method 2: Total power = Motor Power (CAN bus RPM and torque) Method 1 and Method 2 are 23 k. W different Method 2 and Method 3 are different by only 8 k. W Motorest = 124 k. W Axlemeas = 116 k. W Battmeas = 42 k. W Current OEM Catalog Ratings Engine: 105. 1 k. W Motor: 123. 8 k. W Total: 123. 8 k. W Method (1) 105. 1 + 42 = 147. 1 k. W (rated engine k. W + measured battery k. W) Point at expected rated Eng k. W - Peak fuel, peak RPM Method (2) Series HEV output = only Motor k. W Motor output: 124 k. W (from CAN bus) Method (3) Total: 116 k. W (measured axle power) 11

Application of Methods on (another) Series HEV • Engine power taken from CAN bus, motor power at rated power • Method 1 has very high result, 30 k. W higher than Method 2 Current OEM Catalog Ratings Engine: 63 k. W Motor: 111 k. W Total: 111 k. W 105 k. W 80 k. W 61 k. W Method (1) 61 + 80 = 141 k. W (CAN bus engine k. W + measured battery k. W) Method (2) Series HEV output = only Motor k. W Rated motor = 111 k. W Method (3) Total: 105 k. W (measured axle power) 12

Final Issue: Processing Results • • • All methods will require some filter applied to results Data signals must be filtered for robust result (avoid transient spikes or signal noise) Filter duration defines the duration period of peak power We must decide duration of filter size recognizing this choice is arbitrary but uniform Other viable methods besides window filter? 13

J 2908 Summary / Conclusions • Test procedures should be flexible to achieve highest and most representative peak power • Hybrids are transient in nature, definition of power is linked to duration (filter size) • Must resolve Key Issue: – Should hybrid rating be molded to fit current rating? – Should all vehicle power ratings switch basis to wheel/axle power 14