Garrick Forkenbrock & W.Riley Garrott, VRTC
2003 +SAE Government/Industry meeting


SLIDE 1:
PHASE IV AND NHTSA'S LIGHT VEHICLE DYNAMIC ROLLOVER PROPENSITY RESEARCH

 SAE Government/Industry Meeting
 May 13, 2003

- Garrick J. Forkenbrock
- W. Riley Garrott
	NHTSA/VRTC


SLIDE 2:
OVERVIEW OF NHTSA'S RECENT ROLLOVER RESEARCH PHASES:

PHASE IV:
[spring 2001]
 - response to TREAD act
 - consideration of many maneuvers

PHASE V:
[spring 2002]
 - research factors that may affect dynamic rollover propensity tests
 - rollover and handling ratiung development

PHASE VI:
[summer 2002]
 - evaluation of 26 vehicles using results from phase IV testing
 - rollover and handling tests performed

PHASE VII
[winter 2002]
 - refinement of rollover maneuvers using results from phase VI.

PHASE VIII (NCAP ROLLOVER DEMO)
[spring/summer 2003]
 - evaluation of 18 vehicles
 - rollover maneuvers only


SLIDE 3:
PHASE IV BACKGROUND

Tread Act Requirement: 
Develop dynamic rollover propensity tests to facilitate a consumer information program.

National Academy of Sciences:
"NHTSA should vigorously pursue the development of dynamic testing to supplement the information provided by SSF."


SLIDE 4:
PHASE IV OBJECTIVES
- Test many maneuvers with a limited number of vehicles

- Select maneuvers appropriate for use in a government rollover resistance rating system


SLIDE 5:
MANEUVER RECOMMENDATIONS

- recommendations received from government & industry

- NHTSA
  * VRTC
  * Safety Performance standards

- Alliance of Automobile manufacturers

- Consumers Union

- Ford Motor Company

- Heitz Automotive, Inc.

- ISO 2888 Part 2 Consortium
  * VW
  * BMW
  * DiamlerChrysler
  * Porsche
  * Mitsubishi

- MTS Systems Corporation

- Nissan Motors

- Toyota Motor Company

- UMTRI


SLIDE 6:
TEST CONDITIONS
- Test vehicles
  * 2001 Chrevrolet Blazer
  * 2001 Ford Escape
  * 2001 Toyota 4Runner
  * 1999 Mercedes ML320

- Front and rear mounted aluminum outriggers

- all tests performed on a dry, high-mu asphalt surface

- multiple configurations
  * nominal load
  * reduced rollover resistance


SLIDE 7:
TEST MANEUVERS

Characterization
 - constant speed, slowly increasing steer (SAE J266)*

Fishhooks
 - road edge recovery (Roll Rate Feedback fishhook)*
 - Fishhook (fixed timing)*
 - Nissan Fishhook

J-Turns
 - NHTSA J-Turn *

Double Lane Changes
 - ISO 3888 Part 2*
 - COnsumers Union Short Course*
 - Ford path-corrected limit lane change
 - Open-loop pseudo double lane change

	* discussed in this presentation


SLIDE 9:
J-TURN [diagram]

Vehicle / Handwheel input (degrees)
	Blazer 	/ 401
	4Runner / 354
	ML320	/ 310
	Escape	/ 287

note: steering rate was based on successful Phase II testing


SLIDE 10:
FIXED TIMING FHISSHOOK (symmetric)

vehicle	/ handwheel input (degrees)
	Blazer 	/ 326
	4Runner / 287
	ML320	/ 252
	Escape	/ 233


SLIDE 11:
ROLL RATE FEEDBACK FISHHOOK (symmetric)
vehicle	/ handwheel input (degrees)
	Blazer 	/ 326
	4Runner / 287
	ML320	/ 252
	Escape	/ 233


SLIDE 12:
QUESTION: Why use the handwheel angle at 0.3 g?


SLIDE 13:
USE OF 0.3 g HANDWHEEL DATA

- NHTSA needed an objective way of calculating J-Turn and Fishhook steering angles
  * vehicles respond didferently to the same steering inputs
  * maneuvers must adapt to the vehicle being evaluated

- Handwheel data at 0.3 g is repeatable and easy to measure
  * not necessarily true for data based on maximum lateral acceleration


SLIDE 14:
CLOSED-LOOP, PATH-FOLLOWING DOUBLE LANE CHANGES

[diagrams]

- consumers union short course
- ISO 3888 Part 2


SLIDE 15:
EVALUATION TECHNIQUE

EACH MANEUVER EVALUATED IN 4 CATEGORIES
 - objectivity and repeatability
 - perormability
 - discriminatory capability
 - appearance of reality

RATINGS ASSIGNED AS FOLLOWS
 - excellent
 - good
 - satisfactory
 - bad
 - very bad


SLIDE 16:
OBJECTIVITY AND REPEATABILITY
One of the largest disadvantages of the ISO and CU double lange changes
 - driver input variability unavoidable

use of a steering machine inusures accurate, repeatable, reproducible inputs


SLIDE 17:

OBJECTIVITY AND REPEATABILITY
(Example: steering inputs)

[graphs]
steering machine-based fixed timing fishhook
(6 tests are presented)

driver-based ISO 3888 part 2 double lane change
(9 tests are presented)


SLIDE 18:
PERFORMABILITY
- each procedure was well-developed
- ISO and CU double lane changes
  * simplest to perform
  * require little instrumentation
- CU short course does not adapt course layout to vehicle
- RRF fishhok offers better adaptibility than does the FT fishhook


SLIDE 19:
DISCRIMINATORY CAPABILITY

- Lack of discriminatory capability is another large disadvantage of ISO or CU double lane changes
  * entire range of max entrance speeds no more than 5.7mph
  * driver variability accounts for up to 70% of this range
  * ISO and CU double lane changes were not capable of producing two-wheel lift during "clean" runs

- J-Turn and fishhooks sensitive to changes that reduce rollover resistance


SLIDE 20:
DISCRIMINATORY CAPABILITY
(example: metric comparison)
[tables]

* Roll rate feedback fishhook (MINIMUM two-wheel lift entrance speeds)

* ISO 3888 Part 2 double lane change
(MAXIMUM "clean" run entrance speeds


SLIDE 21:
DISCRIMINATORY CAPABILITY
(two-wheel lift summary, nominal load)
[table]


SLIDE 22:
DISCRIMINATORY CAPABILITY
(two-wheel lift summary, RRR) 
[table]

SLIDE 23:
DISCRIMINATORY CAPABILITY
(video Comparison) - excised


SLIDE 24:
APPEARANCE OF REALITY
- Each rollover resistance maneuver related to a real driving scenario

- ISO and CU Double Lane changes emulate emergency crash avoidance maneuvers

- Fishhooks emulate road edge recovery maneuvers [also very similar to first two steering inputs of the double lane changes]

- J-Turn steering least likely to actually be used, but possible


SLIDE 25:
QUESTION:  Are atual drivers able to input the steering angles and steering rates used for the NHTSA J-Turn and Fishhook maneuvers?



SLIDE 26:
ANSWER: YES!
- The ranges of NHTSA J-Turn and Fishhook handwheel angles and rates were within those observed during CU short course testing

- Maximum steering inputs
  * J-Turn: 1000 deg/sec for up to 0.40 seconds
  * Fishhook: 720 deb/sec for up to 0.45 seconds
  * CU Short Course
	- 1187 deg/sec for up to 0.50 seconds
 	- 1026 deg/sec for up to 0.75 seconds
	- 831 deg/sec for up to 1.00 seconds 


SLIDE 27:
QUESTION: Can the NHTSA J-Turn and Fishhook maneuvers be performed on a two-lane public roadway?


SLIDE 28: 
ANSWERS 
	1. Yes (fishhook)
	2. Not Likely (J-Turn)
[graphs]


SLIDE 29:
OVERALL ASSESSMENT	
- Roll Rate Feedback Fishhook deemed the best overall maneuver

- J-Turn the most basic maneuver, can be a useful compliment to the Roll Rate Feedback fishhook.

- Both maneuvers selected for use in Phases V, VI and VII of NHTSA's rollover research.


SLIDE 30:
QUESTION: Can the Slowly Increasing Steer maeuver be abbreviated since only linear range lateral acceleration data is used?


SLIDE 31:
ANSWER: YES! (provided enough data is considered)

[graphs]


SLIDE 32:
CONCLUDING REMARKS
Phase VI and VII techinical reports
 - complete, awaiting approval
 - scheduled to be released with the next rollover notice

NCAP Rollover demo is presently underway
 - ratings to be released as 2004 model year ratings


SLIDE 33:
ADDITIONAL INFORMATION
 - Phase IV Technical report (DOT HS 809 513)
 - SAE papers
	* 2003-01-1008
	* 2003-01-1009
 - www-nrd.nhtsa.dot.gov/vrtc/ca/rollover.htm
 - Rollover Docket
	* http://dms.dot.gov/
	* "simple Search" for number 9663