FMVSS206

t-
c/ Memorandum
us Deportment
ot Tmnsportatim
DEPARTHLNT Or 1llAH;i’Oii IA?i!)t,i
National Highway
Tmfflc satety
Administmtbn
Subject:
INFORMATION: Submission to Docket Dale:
NHTSA-98-3705 -/
Reply to
gg$:::ds&q~ Attn, 0,’ APR - 7 lgg8
TO:
THRU: ~G~~~
Office of Crashworthiness Standards
John Womac
Acting Chief Counsel
Attached for inclusion in Docket NHTSA-98-3705 is a research report titled, “Regression

Analysis between NASS 1988-1996 Front Side Door Opening Rates and Agency Side Door
Latch Testing to Identify Possible Upgrades for FMVSS 206, Door Locks and Door Retention
Components.” This analysis was conducted to explore the existence of statistical correlations
between the series of agency side door latch tests and front side door opening rates seen in the
National Automotive Sampling System (NASS) in order to identify a potential test to upgrade
FMVSS 206.
Attachment
cc:
Steve Summers, Research and Development
Don Willke, Vehicle Research and Testing Center
Ellen Hertz, National Center for Statistical Analysis
#
Regression Analysis between NASS 1988-l 996 Front Side Door
Opening Rates and Agency Side Door Latch Testing to Identify
Possible Upgrades for FMVSS 206, Door Locks and Door Retention
Components
NATIONAL HIGHWAY TRAFFIC SAFE-W ADMINISTRATION

APRIL I, ,998
TABLE OF CONTENTS
Executive Summary ........ ........ ....... i
1,Background ._.._.__.__.__...__._.,, ........ ........ ...... 1
2. Objective ........ ........ ...... 2
3. Data Sources ........ ........ 2
4. Approach ........ ........ 8
5. Procedure ........ ........ 8
6. Discussion and Analysis of Results ........ ........ 9
7. Conclusions ........ ........ 14
8. References.. ........ ........ 16
9. Appendix I ........ ........ 17
10. Appendix II ........ ........ 23
1 I. Appendix III ........ ........ 50

EXECUTIVE SUMMARY
On August 15, 1995, the agency held a public meeting with industry, safety groups and other interested
parties, to discuss possible upgrades to FMVSS 206, “Door Locks and Door Retention Components.” In
this meeting the agency presented several possible test methods that were developed to simulate door
failures seen in real world crashes. By the conclusion ofthe meeting, however there was no resolutions
for selecting the appropriate test method(s) due to the lack of clear and convincing evidence of a
correlation between the latch test results from any of the procedures and their failure/injury rates seen in
the crash data files.
As a result, in 1995, the agency conducted an analysis of a series of vehicles tested in each of the
agency’s side door test procedures and their corresponding door opening rates in the NASS. The
analysis showed that hvo of the test procedures appeared to coincide with the ranking of latch failure
rates for similar vehicles in the NASS and thus indicated that a statistical correlation may possibly exist.
To further study this possible correlation, an additional analysis with more recent NASS data has been
completed and is presented in this report.
In this analysis, logistic regression was used to estimate the best fit wrves or models for the data. In
addition, a non-linear/logarithmic equation was used as a representation of each of the curves. Using
these techniques, three models having less than a 12 percent probability of occurring by chance were
chosen to identify the test procedures which had strong relationships to side door openings. The
procedures identified were the lateral and longitudinal full door and bypass tests. These models are also
used in this analysis to measure the effects of increasing the test loads on the reduction in the probability
of vehicles experiencing a side door opening in a crash. It was estimated that an increase in the test loads
up to approximately 4,000 newtons (in the full door and bypass tests) from the current FMVSS 206
requirement would decrease the probability of side door opening by 74 to 76 percent.
i
1. Backaround
Based on estimates from the 1988-1996 National Automotive Sampling System (NASS), there was an
annual average of I 1,804 fatally ejected occupants and 10,700 seriously injured ejected occupants in
light duty vehicles. Rollover crashes accounted for 58 percent of these ejection fatalities and 55 percent
of the serious injuries. In response to the high numberof rollover-related ejection fatalities and serious
injuries, the National Highway Traffic Safety Administration (NHTSA) was mandated by the 1991
Authorization Act to conduct rulemaking to reduce these types of injuries.
In an effort to reduce ejection fatalities and serious injuries, the agency has targeted ejection mitigation
through side doors and fixed glazing. As for ejections through doors, it was estimated that there was an
annual average of approximately 2,513 fatalities and 1,855 AIS 3-5 serious injuries (from the 1988-1996
NASS). Thirty-nine percent of these fatalities and 58 percent of the serious injuries occurred in rollover
crashes. As a solution to reducing ejections through side doors in rollover crashes as well as in other
types of crash modes, NHTSA has conducted research focused at identifying the causes for side door
opening, developing representative test procedures which simulate real world door opening and
performing analysis to determine the relationship between developed test procedures and agency crash
data.
In 1986, the agency contracted Chi Associates to determine if any correlation existed between the current
side door latch test procedure, Federal Motor Vehicle Safety Standard (FMVSS) 206, “Door Locks and
Door Retention Components,” and ejection rates for passenger cars from crash data files. The analysis
showed a “moderate” correlation (an inverse correlation coefficient as high as R = -0.67) between
ejection rates and ultimate latch loads in the longitudinal direction.’ In 1988, the agency conducted a
follow-up study to validate the finding of Chi Associates. The report concluded that the correlations
between latch strength and ejection rates were actually much lower than those found in the Chi study.*
Following these studies, the agency conducted numerous analyses comparing crash data rates and
existing and newly developed side door latch test procedures. The result of these analyses, however,
showed very little relationship between real world side door latch failure rates and any one single side
door latch test procedure.
To improve the accuracy for determining the performance of latches observed in the crash data tiles, in
1994, the agency conducted additional research which grouped vehicles with similar latch systems. The
latch systems were identified by common latch/striker/anchorage plate part numbers. From this analysis,
the agency recommended candidate make/model vehicles with high and low door opening and latch
failure rates for further laboratory latch strength testing. A report titled, ” Evaluation of Door Openings,
and Latch Failures in Recent Model Years Using Car Groupings,” contains the finding.’ Also in 1994,
the agency initiated research to test vehicles ranked from Reference 3 in a series of five side door latch
tests. The series of five side door latch tests included: (I) the full door lateral and longitudinal test
procedures; (2) the bypass test procedure; (3) lateral and longitudinal FMVSS 206 type tests; (4) the
inertial mini-sled test procedure; and (5) the General Motors (GM) horizontal rotation test procedure.
The goals of this research were to: I) test as many potential failure modes as possible by using test
methods which simulated all the identified failure modes; 2) test under loading conditions that are
common in NASS door opening cases; and 3) test under conditions that are not already covered by
FMVSS 206. The finding of this research were summarized in a report titled, “Door Latch Integrity
Study: Evaluation of Door Latch Failure Modes.” ’
In 1995, the agency re-conducted the research from Reference 4 expanding the test groups with
additional latching systems selected from Reference 3. The research was intended to evaluate whether
each test vehicle’s performance in real world crashes was comparable to its test results in any of the five
test procedures. The study showed that only the testing results from the bypass and the lateral full door
test procedures coincided with the high/low latch failure rate ranking established from the NASS for
each latching system.’ This indicated that it might be possible to find a statistical correlation between the
test and real world crash data, when the test procedures were coupled for comparison.
2. Obiective
This study was conducted as a follow-up study to Reference 5. It is intended to update the results of
Reference 5 and to validate its conclusions by identifying statistical correlations between a series of
agency side door latch tests and front side door opening rates seen in the NASS. From this information,
the agency may be able to identify a test procedure or group of procedures as suitable candidates for
upgrading FMVSS 206 or improving door retention performance in crashes.
3. Data Sources
2
The data for the analysis consisted of the results of agency side door latch testing and information from
the NASS crash data files. The NASS data were restricted to those cases reported from 1988 through
1996 for which the vehicle interior forms reported at least one front side door opening. This is indicated
in the NASS with a value of 2 for the codes OPENLF and OPENRF on the interior vehicle form. (Note:
1988 was used as the first calender year of NASS data due to the unavailability of side door opening
codes for prior years). As for the test data, they were also restricted to light duty vehicle latching systems
that were tested only in the longitudinal and lateral full door tests, the fork-bolt bypass test, GM’s
horizontal rotation test, and the FMVSS 206 lateral and longitudinal tests conducted by NHTSA’s office
of Safety Assurance and by VRTC (note, these procedures are referenced in subsequent sections as
LAT4, LONG4, 6650/LAT, GMROT, FTRANSl, FLONGl, LATS and LONG5 respectively). A brief
description and diagram for each of the test procedures is provided in Appendix 1. When a vehicle
latching system had more than one score for a particular test, the average of the scores was used for all
the vehicles of that type (similar make, model, and model year).
The data base of the test results is presented in Table I, pages 4 through 7, by vehicle make, model,
model year and test scores. mote, table I also includes scores for the mini and HYGE sled tests which
are shown for comparative purposes.] In total there were I86 vehicle latching systems selected to the
analysis. Vehicle were not grouped according to similar latching systems (similar in latch, striker and
anchorage plate numbers - as previously used in references 3 and 5). Each latching system was
considered separately to evaluate whether differences in other retention components as well as other
vehicle parameter influenced retention performance.
The final matrix of the data for the correlation consisted of a merge of the test data and the NASS
information. Table 2 (see Appendix II), shows the final matrix by vehicle group (from Table I) and by
the investigated (sample) and weighted (estimated) number of door openings, including the associated
sample errors for each vehicle/latching system. The sampling errors are provided to indicated the
reliability of the estimates. Note, several test vehicles were found to have no mated door opening cases
data, however they were still used in the analysis for the purpose of increasing the exposure of the
samples in the crash tiles as well as to measure the effects of zero door opening rates on the analysis.
For example the Ford Probe, a popular sedan model, was reported with no door openings. It should also
be noted that certain vehicles groups in Table 2 have very large sampling error, however this does not
3
Table 1. Test Results from agency side door latch testing
,,, , ,, , , ,
,,,, ,,, , ,, , , ,
log of the odds of a door opening as a function of the test score. Only variables that were statistically
significant are shown in Table 3 with an intercept and beta value.
Next, a multi-variate correlation analysis was conducted using all the significant variables in all
combinations to find any relationships which exist between crash rates and different combinations of
force loading. Together, however, it resulted in none of them being significant. Similarly, no
significance was noted for models with two variables. At best, when both 6650/LAT and LONG4 are
included in the model, only 6650/LAT remains significant. The model produced beta values of -
0.0000303 and -0.0001103 and p-values of 0.592 and 0.014 for the LONG4 and 6650/LAT, respectively.
This suggests that, in combination, LONG4 does not contribute very much and that the test procedures
can only independently influence door opening rates.
6. Discussion and Analvsis of Results
Table 3 shows that, when considered separately, the full door test procedures and the detent fork bolt
bypass tests were the only newly developed test procedures that are capable of measuring the
performance of the test latches in real world crashes (latch performance in real world crash was
measured by the number of system failures or inadvertent door openings that occurred in crashes). This
was determined by the column denoted as the p-value in Table 3 which represents the strength of the
relationship between the data and the test procedures (the statistical significance) or the probability that
9
their relationship is purely coincidental. As shown in this column, the p-values associated with the full
door and bypass test procedures range from a probability of 0.001 to 0.112. Other test procedures have
probabilities ranging more than twice this amount, starting at a probability of 0.205 and ending with a
probability of 0.772, especially, for the compliance FMVSS 206 procedures (FTransl and FLongl)
which had p-values which range from 0.21 to 0.64. Similar findings were also noted in other agency
analyses, such as reference 4, which also noted the bypass and full door test procedures as producing the
best distinctions between side door opening rates and smaller samples of test data. Other analyses
include references 2, and 5, which cite the FMVSS 206 type test as being unrepresentative of real world
door openings, as estimated by NASS.
However, what makes this analysis different from previous correlation studies (i.e., References I, 2 and
4) is that logistic regression was used here, which assumes a logarithmic relationship as the best tit curve
to the data. Previous agency correlation analyses have always assumed the relationship between side
door test results, particularly in the FMVSS 206 test, and latch failure, door opening and ejection rates,
as best fit by a linear equation. For this analysis, the equation of the best tit curve is as follows:
log(----
1 ” PI = a+Px Equation (1)
wherep = the absolute probability of door openings for NASS vehicles,

a is a constant or the vertical intercept of the curve (predicted by SUDAAN), and
x = test load amounts from agency testing
p = the slope of the curve or the change in the log of the odds of a door opening as a function of the test score (also shown in
Table 3)
Equation 1 establishes a direct relationship between the probability of door openings and the test load
amounts in the significant test procedures. Equation 1 not only establishes a direct relationship but also
by solving the equation in terms of p, we can measure the probability at any test load. The steps used for
solving Equation 1 in terms of p are as follows:
IO
log(---
1 p p) = a+Px Equation (1)
P ,,p’“‘Px) Multiplying both side by e”

1 -P
Solving for p
p = (l-p)exp@+PX)
,xp@‘P4
Normalizing the numerator yields
P=
,,p(c+ + 1
p= 1 Equation (2)
1 +exp(-+)
Equation 2, represents Equation I in terms of the probability of door openings. By substituting each
significant procedures’ beta and sigma values, and test loads from 0 to 54,000 newtons a range of
probability values can be determined. Table 5 is provided in Appendix III to give the probability of door
openings for the range of test limits from 0 to 54,000 newtons. In addition, the range of probability
values are graphically shown in Figure I. Figure I shows that the predicted probability of the door
openings decreases with more stringent load levels, which also coincides with an increase in door
retention performance. It also shows that for the range of test loads, the probability of door openings can
vary from approximately 0 to 4 percent. This predicted range of probabilities signifies that side doors
are a very rare event in relationship to the total number of accidents on the road ways; it is also a well
known fact that the associated ejections are even less common. However, this does not change the fact
that these ejections are more likely to result in sever to fatal injuries, especially in rollovers.
Other information that can be gained from Figure I is that for large values of the test loads, the
probability values change very little as the graphs approach infinity. This indicates that lower values of
the test loads can produce relatively the same probabilities as the larger load limits. In addition, Figure 1
contains three vertical lines which intersect the curves at the average test load amounts (from Table I).
These lines represent a reasonable approximation of the performance of the average tested retention
system and should also represent the average performance of most vehicles on the roadways. As
mentioned earlier, based upon the selection process of the particular sample of vehicles selected for
testing and their similarities in common driver usage patterns and vehicle class, it is assumed that
Equation I was applicable to the population of all passenger vehicles (excluding sports type vehicles),
and thus p represents a relatively close approximation of the probability of door openings for most of the
entire fleet of light duty vehicles. Therefore, from these average predictions, determinations can be
made as to what changes in current door retention performance will result from increased load criteria in
the bypass and full door test procedures. More importantly, however a reasonable approximation of
suggested test load limits, that will maximize the effectiveness of each procedure in reducing real world
side door openings, can be determined. [This same technique can be used to predict the number of lives
saved for the suggested load limits in each test procedure, since door openings have a direct relationship
to side door ejection injuries and fatalities; however, this will be explored in later analysis.]
For the purpose of demonstrating this method (determining the effects of changes in the test loads), a
suggested test load was chosen to represent the probability of door openings at 16,000 N in each
procedure. A test load of 16,000 N was considered as reasonable limit because they coincided with the
observed test results of better performing latches in the data; a value greater than the average test results
in each of the significant procedures. In addition, after this point the graph of the probability becomes
infinitesimally smaller as the test load curves approach infinity, which results in different values in the
test loads, causing little or no effect upon the probability of door openings. For example, the probability
I2
at a test load of 30,000 N is relatively the same at 50,000 N
Using a suggested load of 16,000 N in each procedure, the corresponding probabilities of door openings
at this limit are shown in Table 4. Also, for comparative purposes, Table 5 shows the values of the
probability of door openings at the suggested load amount, the average test loads (from Table I) and the
current FMVSS No. 206 test load requirements (Appendix III also contains the probabilities at these test
loads in each of the procedures).
Table 4. Prediction for suggested test load, 15,000 newtons I

Test Procedure Actual Probabilily of Suggested Test Load (N)
Door Opening (App. 111)
Comparatively, at the suggested test load, from Table 5, the probabilities of door openings were
considerably lower than those of the current 206 requirements and for the average loads, except for the
longitudinal full door test. This indicates that overall the suggested test loads will provide a more
stringent requirement. However, in order measure the percentage changes in the probability of door
openings or relative risk of increasing from the 206 requirements or the average loads to the suggested
load limit, the following approach must be used as follows:
13
Using Equation I, Equations 3 and 4 are generated by substituting p, as any initial probability, x, any
initial test load, a is the increase in the load amount from x,, and p, the probability at a.
I%-
,:;,,) = a + Px,, Equation (3)
log& a + lxx<, + 4 Equation (4)

l-p,) =
Subtracting Equation 4 from 3 yields Equation 5
l-p ) - I%( *) = -pa

log&
0 I
(p”) Equation (5)
’ -p<, =
log- -PO
PI
(--
I-p,)
Since the probability of a door opening is small, the relative risk is close to the relative odds
Substituting this relationship into Equation 5, yields Equation 6.
1 -PI _
!f
’ - P,
:. log(3 = -0p
PO = e,p-(m
-
PI
:. fhe % change for p by increasing lest load (a) =
P”-Pl
lOO[l -exp’Op)] = loo*-
PO Equation (6)
14
Equation 6, represents the percent change in the probability of a door opening or relative risk as a
function of the test loads, assuming any test load change, o. Using Equation 6 and substituting each
significant procedure’s beta and sigma values (from Table 3), and the change in test load (from either the
average values or the 206 requirements) predictions in the relative risk have been determined and are
shown in Table 6. Table 6 shows that by using a test load of 16,000 newtons in each of the significant
procedures the probability of door openings in average vehicle will approximately be reduced from 23 to
28 percent, except for the longitudinal full door test which has the risk of increase the probability by 6
percent. Table 6 also shows that if door performance design was based upon the current FMVSS 206
requirements, there was be an appreciable increase in the probability of door openings such that current
longitudinal latch performance is increased up to 65 percent and lateral performance up to 73 percent.
Table 7 is also provided in Appendix IV to give the percentage change in the probability of door
openings from current retention performance levels (evaluated at the average values) to a range of test
loads.
Table 6. Percent change in the probability of door openings from a change in the average
test load to the 206 requirements and suggested load, 16,000 newtons*
Test Procedure Avg-Suggested A”g-II,IZON Av&3,896 N Avg.4,448 N**
BYP= 28.27 -31.28 -72.73 NA
Full DOW Lateral 23.44 -31.06 -67.26 NA
Full Door Longitudinal -6.45 -64.66 NA NA

* Nate, the probability at the FMVSS 206 requirement is based upon use in the new procedures
l * Secondary latch performance wa not evaluated in the testing.
To further evaluate the feasibility of the suggested load value, the test scores in Table I were re-
examined, identifying the number of test scores which fall below our new given criteria. In the
6650/LAT test, it was noted that only I5 bypass tests out of a total of 23 tests would not have met the
criteria. Similarly, 22 out of 29 lateral full door test would not meet this criteria. This would indicate
that 16,000 newton may be to high of a criteria for the bypass and lateral full door test. However, this
criteria may be suitable for the longitudinal full door test. Only 16 out of 34 test would not have passed
this criteria which is less than half the vehicles.
7. Conclusions
The analysis shows that a good correlation exists between the bypass and full door tests and NASS side
door opening rates, substantiating the conclusion of Reference 5. However, relationships only existed
when the variables were considered independent of one another. This was found to be acceptable due to
the distinct differences in the test setup and test load directions among the three procedures. The
analysis also provided a means of evaluating the effects of each test procedure in reducing the probability
of door openings. In the analysis, an arbitrary test load of 16,000 newtons was chosen, to demonstrate
the possible reductions in side door opening. At this level, there is less than a I percent probability of
door openings as shown in Table 4. In addition, it was also predicted that this value would decrease the
current probability of door openings by 23 to 28 percent in the bypass and lateral full door test.
However, for the longitudinal full door test it was predicted that 16,000 newtons may not be stringent
enough. This would suggest that different load criteria in each significant procedure may be appropriate.
However, it is important to note that 16,000 netwons was used to demonstrate a possible criteria, yet
other values can be used and may be evaluated at a later time.
Other important information that is not discussed in this report, is that these procedures simulate an
environment which more closely resembles the actual vehicle’s door support structure and the force
loading characteristics experienced in crashes in which door openings due to latch/striker failure. The
bypass test procedure, in particular, has test conditions which are representative of the combination of
compressive longitudinal and tensile lateral force loads seen in real world latch failure cases and no other
agency tests simulates this type of latch failure loading directions. The full door test procedures (lateral
and longitudinal) offer the option of evaluating the entire door as a system as opposed to other
procedures which are only component level tests.
Finally, it was also noted that these procedures are capable of producing failure loads below the current
FMVSS 206 requirements. For example, during longitudinal full door testing, three latches failed
structurally at levels below FMVSS 206 which suggests that even though the door latch may pass the
standard, it might not withstand a similar load when subjected to a full door test. We believe this is due
to the differences between the actual door mounting and the stiffer test fixture mounting used in the
FMVSS 206 test procedure.
16
References
I. Chi Associates, “Side Door Latch/Hinge Assembly Evaluation,” DOT HS 807-234, October
1986
2. Cohen, D., “Door Latch Integrity Program; Laboratory/Accident Data Correlation,” NHTSA
Research and Development, September 1989.
3. Stucki, L., “Evaluation of Door Latch Failures in Recent Model Years Using Car Grouping,”
NHTSA Research and Development, Door Latch Integrity Project, Task I, Subtask 1, May 1992.
4. J. Gavin Howe, Michael Leigh, Donald T. Willke; “Door Latch Integrity Study: Evaluation of
Door Latch Failure Modes;” final report for project VRTC-87-0149; DOT-HS-808-188; January
1994.
5. Willke, D., Howe, G., “Door Latch Integrity VRTC-80-0149: Evaluation of High and Low Rate
Latches.“, NHTSA Vehicle Research Test Center, VRTC-80-0149, Event Report, Docket 94-70.
N02, June 1995.
References
I. Chi Associates, “Side Door Latch/Hinge Assembly Evaluation,” DOT HS 807-234, October
1986
2. Cohen, D., “Door Latch Integrity Program; Laboratory/Accident Data Correlation,” NHTSA
Research and Development, September 1989.
3. Stucki, L., “Evaluation of Door Latch Failures in Recent Model Years Using Car Grouping,”
NHTSA Research and Development, Door Latch Integrity Project, Task I, Subtask I, May 1992.
4. J. Gavin Howe, Michael Leigh, Donald T. Willke; “Door Latch Integrity Study: Evaluation of
Door Latch Failure Modes;” final report for project VRTC-87-0149; DOT-HS-808-188; January
1994.
5. Willke, D., Howe, G., “Door Latch Integrity VRTC-80-0149: Evaluation of High and Low Rate
Latches.“, NHTSA Vehicle Research Test Center, VRTC-80-0149, Event Report, Docket 94-70-
N02, June 1995.
17
ADDendix I
Brief descriptions and diagrams of the full door longitudinal (Long4) and lateral (LAT4) tests, the
fork-bolt bypass (66501LAT) test, the FMVSS 206 lateral and longitudinal tests (FTRANSI,
FLONGl, LATS and LONGS) and the GM rotation test (GMROT).
18
Appendix I
Figure 1. GM Rotational Test Fixture
GM developed a static test procedure which allows the striker and latch to rotate relative to one
another, to simulate rotation of the surrounding structure which causes striker twist failures.
The GM rotational test fixture (Figure 8, Appendix II) was instrumented with: (1) a load cell to record
the force applied in the vertical directions; (2) a displacement “string” potentiometer to measure the
vertical displacement ofthe tensile machine cross head; and (3) a rotational potentiometer to record
the rotation of the striker relative to the latch. The test is conducted at a load application rate of 0.8
inches/minute, a striker rotation limit (relative to the latch) of 45 degrees and a latch mounting plate
thickness of ‘A inch.
19
Auoendix I
Figure 2a. Full Door Lateral Tension Test Fixture
Figure 2b. Full Door Longitudinal Tension Test Fixture
The lateral and longitudinal full door tension test fixtures are shown above as Figures 2a and 2b. The
lateral test consists of applying a lateral force to the latch/striker system by moving the striker
laterally at a rate of 2 cm per minute. The longitudinal test consists of applying a constant lateral
force of 1,000 N and a longitudinal force to the latch/striker system. The longitudinal force is applied
by moving the striker longitudinally at a rate of 2 cm per minute.
20
Aupendix I
LONGITUDINAL
APPLIED FORCE
Figure 3. Detent Lever - Fork Bolt Bypass Test Fixture
The by-pass test procedure is a static bench test which consists of applying and maintaining a
constant lateral force and then applying a longitudinal force by moving the striker at a constant rate of
displacement. The longitudinal displacement rate was 1 cm per minute. The tests were conducted at
4,450 and 6,650 N (1,000 and 1,500 Ibf) lateral load.
21
Amendix I
DOOR LATCH STAT/C LOAD TEST FCfTURE (LONGITUDINAL LOAD)
Figure 4. FMVSS No. 206 Longitudinal Test Procedure
!iI puLL
Figure 5. FMVSS No. 206 Transverse Test Procedure
Appendix I
FMVSS 206 Longitudinal Load Test: The door latch and striker assembly, when in the fully latched
position, shall not separate when a longitudinal load of I 1,000 newtons (2,500 pounds) is applied.
When in the secondary latched position, the door latch and striker assembly shall not separate when a
longitudinal load of 4,450 newtons (1,000 pounds) is applied (see Figure 4.)
FMVSS 206 Transverse Load Test: The door latch and striker assembly, when in the fully latched
position, shall not separate when a transverse load of 8,900 newtons (2,000 pounds) is applied. When
in the secondary latched position, the door latch and striker assembly shall not separate when a
transverse load of 4,450 newtons (1,000 pounds) is applied (see Figure 5).
These requirements are described in terms of the minimum latch failure loads in the longitudinal and
transverse directions. With respect to side doors, the longitudinal load test is applied parallel to the
front-to-back axis of the vehicle, perpendicular to the latch velocity vector at the instant of
latch/striker engagement perpendicular to the face of the latch. For the transverse load test, the test
machine applies a load in the direction of the door opening (see Figures 4 and 5). These practices are
defined in FMVSS No. 206, according to SAE Recommended Practice J839.
23
Appendix II
Table 2: Data for the analysis shown by vehicle group (shown in Table I) and by the reported actual
and weighted number of door openings including the sample errors associated for each
vehicle/latching system
24
A!mendix II
Table 2. Vehicle Latch Groups by Actual and Ueighted Door Openings.
SUOAAN
Software for the Statistical Analysis of Correlated Oata
Copyright Research Triangle Institute, April 1996
Release 7.00
Nunber of observations read 5448 Ueighted count : 2239516

Number of observations skipped i 501
(WEIGHT variable nonpositive)
Oenominator degrees of freedom : 24
WOW OPENED cl)* (2)"

1.. _ _ _ _ _ _ _ _ _ _ _ Weight Tot Actual Tot 1 Sample Err 1
.., _.
Total Sample Size 5448 164 5284

Weighted Size 2239516.49 15520.39 2223996.10
SE Weighted 379458.82 2525.78 377475.12
. ...........
1 Sample Size 11 0 11
Veighted Size 3870.46 0.00 3870.46
SE Ueighted 1931.80 0.00 1931.80
. ..
2 Sample Sire 0 0 0
Weighted Sire 0.00 0.00 0.00
SE Ueighted 0.00 0.00 0.00
3 Sample Size 1 0 1
Ueighfed Size 192.05 0.00 192.05
SE Ueighted 192.05 0.00 192.05
.................
4 Sample Size 0 0 0
Veighted Sire 0.00 0.00 0.00
SE Weighted 0.00 0.00 0.00
._.._____...........___.............~.~.....................
5 Sample Size 0
Weighted Sire 0.00
SE Weighted 0.00
6 Sample Size 0 0 0
Ueighted Size 0.00 0.00 0.00
SE Veighred 0.00 0.00 0.00
NOTE: On subsequent pages, for the Last two colurms in the tables the fallowing applies
* 1 represents - The ectual total number of door openings
l * 2 represents - The sampling error associated with weighting each sample size
25
Atxwndix II
Dare: 08-08-97 Research Triangle Institute Page : 2

Time: 14~17~17 The CROSSTAB Procedure Table : 1
"ass 88.96,c:Latch.prc
by: SUDGP, OPENED.
SUOGP
2
. . . . .._____ I
1
493.87
493.87
8 Sample Size 4 2 2
Weighted Size 580.91 167.95 412.96
SE Weighted 283.85 133.92 292.17
9 Sample Size 2 1 1
Ueighted Sire 184.57 93.11 91.45
SE Veighted 130.51 93.11 91.45
. .
Weighted Size 2730.49 0.00 2730.49
SE Ueighted 2240.55 0.00 2240.55
__.___........................
Weighted Size 260.31 0.00 260.31
SE Weighted I 171.96 0.00 171.96
. .
12
13
I’ Sample Size
Ueighted Size
SE Weighted I
2
168.06
168.06
0
0.00
0.00
2
168.06
168.06
Sample Size 1 0 1
Veighted Size 488.32 0.00 488.32
SE Weighted 488.32 0.00 488.32
26
Atmendix II
oate: 08-08-97 Research Triangle Institute Page : 3

Time: 14:17:17 The CROSSTAB Procedure Table : 1
"ass 88-96,c:Latch.prc
by: SUOGP, OPENED.
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..~~~~~~~~~.........................~~~~~~.
.
SUOCP OPENED
Total I ' I 2
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...~~~.......
...
Yeighted Size 26718.03 6.61 26711.43
SE Weighted 19103.58 6.61 19103.45
SE Ueighted 0.00 0.00 0.00
16 Sample Sire 2 0 2
Weighted Size 31.28 0.00 31.28
SE Weighted 23.90 0.00 23.90
Weighted Size 15512.08 61.39 15450.69
SE Ueighted 6122.20 49.47 6121.71
................ . ............
SE Ueightec 33.35 0.00 33.35
.. . . . _..
Ueighted Sire 0.00 0.00 0.00
SE Weighted 0.00 0.00 0.00
Weighted Size 1020.45 72.27 948.18
SE Weighted 718.71 72.27 715.07
27
ADDendix II
Date: 08-08-97 Research Triangle Institute Page : 4

by: SUOCP, OPENED.
OPENED
Total I 1 I 2
0 0 0
0.00 0.00
0.00 I 0.00 I 0.00
Weighted Size 15186.95 26.28 15160.66
SE Weighted 5379.72 26.28 5379.53
Weighted Size 1974.04 0.00 1974.04
SE Ueighted 1153.93 0.00 1153.93
. . . . . . . . . . . . . . . . . . . . . . . . . . . .. .
24 Sample Sire 0 88
Ueighted Sire 20824.:: 0.00 20824.63
SE Ueighted 5097.25 0.00 5097.25
__._______....... . . . . . . . .

Ueighted Size 85528.59 21.oP 85507.50
SE Ueighted 20473.ll 21.09 20470.37
____............. _ _ _ _ . _ _ _ _ _ _ _
26 Sample Size
Ueighted Size 33230.66
73 73
33230.66
SE Weighted 12165.92 12165.92
Weighted Sire 2214.22 0.00 2214.22
SE Weighted 1171.23 0.00 1171.23
_____________._...............----..........................................
28
-
Awendix II

"ass 88.96,c:latch.prc
by: SUDGP. OPENED.
SUOGP OPENED
I Total I 1 I 2 I
. . . .._.......................~~..............................................
Weighted Size
SE Yeighted I
8984.90
2816.27 326.83 I 2850.64
8658.07 I
Weighted Size 22149.34 64.49 22084.85
SE Weighted 7247.49 64.49 7212.63

Veighted Size 121361.84 344.47 121017.37
SE Weighted 25999.19 201.71 25888.71
Weighted Size 28143.82 41.52 28102.31
SE Weighted 14372.09 41.52 14370.94
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . ..~~.
Veighted Sire 17870.92 20.83 17850.09
SE Ueighted 5762.05 20.83 5760.27
. . ...._____..................... .._ ____________............
Weighted Sire 722.55 0.00 722.55
SE Weighted 722.55 0.00 722.55
Veighted Sire 11247.31 194.11 11053.20
SE Ueighted 3912.32 138.15 3912.11
________________.. . . . . . . . . . . . . . . . ..._.___________ . .
29
ADDendix II

nass 88.96,c:latch.prc
by: SUDtP, OPENED.
SUOGP OPENED
I Total I 1 2 I
35 Sample Size 0 0
Weighted Size 0.00 0.00
SE Weighted 0.00 0.00
3 0 3
373.24 0.00 373.24
254.92 0.00 254.92
Weighted Size 6055.27 0.00 6055.27
SE Weighted 2343.01 0.00 2343.01
Weighted Sire 5549.26 0.00 5549.26
3272.47 0.00 3272.47
I
SE Weighted
.................
40 Sample Size
Weighted Sire
SE Weighted
I 6070.06 2300.95 18 43.72 37.97 2
. . . . . . ..__
16
6026.34
2300.62
._
.................
Weighted Size 32940.71 620.88 32319.83
SE Weighted 9967.07 490.31 9992.56
30
Awendix II

Time: 14:17:17 The CROSSTAS Procedure Table : 1
by: SUDCP, OPENED.
wow OPENED
1.. _ _ _ _ _ _ _ Total I 1 I 2 I
. ..___________.........................
Weighted Sire 16377.72 132.36 16245.36
SE Weighted 4787.88 96.15 4778.72
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..~~.....................................~~~. .I
SE Weighted 0.00 0.00 0.00
Weighted Sire 13559.31 0.00 13559.31
SE Weighted 12593.49 0.00 12593.49
.
SE Weighted 78.58 0.00 78.58
. .
46
................
Sample Size
Weighted Size
SE Weighted
.................
I 0
0.00
0.00
0
0.00
0.00
..
47 Sample Size
Weighted Size
SE Weighted I 0
0.00
0.00
0
0.00
0.00
. .
0
0.00
0.00
SE Weighted 0.00 0.00 0.00
31
Amwndix II

nsss 88.96,c:latch.prc
by: SUDGP, OPENED.
SUOGP OPEWEO
I __________________! . . . . . . . . . . . . . .._.___-.................
Total I 1 I 2
.
Weighted Size 164772.92 444.04 164328.88
SE Weighted~ 26892.90 182.71 26808.19
52 Sample Sire 8 1
Weighted Size 2026.09 17.26 2008.8:
SE Weighted 983.09 17.26 994.16
. . . . . . . . .._.... .
53 Sample Size I 143 15 128
I ’lei! Shted Size I 62449.18 I 677.75 1 60771.43 1

SE Weighted 1 15016.60 961.51 1 15025.20
..............................
54 Sample Size 0 10
Weighted Sire 0.00 1705.71
55 Sample Sire 0 18
Weighted Size 7700./Y 0.00 7700.41
SE Weighted I 4028.40 0.00 4028.40
32
Appendix II
Oate: 08-08-97 Research Triangle Institute Page : 9

nass 8%96,c:Latch.prc
by: SUOtP, OPENED.
12
287.67
940.28 I 257
21257.36
96093.09 I
Weighted Size 5574.94 0.00 5574.94
SE Weighted 3322.44 0.00 3322.44
__._........................~~~~...........................................,
77 1 76
33463.28 81.11 33382.16
9638.04 81.11 9631.05
.._._______
85 85
36819.50 36819.50
5822.52 5822.52
__,.____......
20
10098.60
4362.91
Weighted Six 0.00 0.00 0.00
SE Weighted 0.00 0.00 0.00
. . ..____..
SE Weighted 0.00 0.00 I 0.00
..._._............................................ . . .._____...............
33
Amxndix II

by: SUDGP, OPENED.
SUOGP
Sample Size 0 0 0
SE Weighted 0.00 0.00 0.00
. ..
I
0 0 0
0.00 0.00 0.00
0.00 0.00 0.00
.
10 0 10
6139.19 0.00 6139.19
3048.04 0.00 3048.04
.:.. .
52 1 51
42540.81 125.99 42414.82
19068.94 125.99 19068.52
._. ._
Weighted Size 17889.83 18.33 17871.49
SE Weighted 6300.31 14.99 6301.41
Weighted Size 10680.09 0.00 10680.09
SE Weighted 5150.26 0.00 5150.26
...................................

Weighted Size 56816.96 171.05 56645.91
SE Weighted 8826.96 71.46 8815.02
.
34
Amendix II
oate: OS-OS-97 Research Triangle lmtirute Page : 11

"a?.* 8%96,c:Latch.prc
tly: SUOGP, OPENEO.
SUOGP OPENED
Total
!.i
Ueighted Size 27158.33 192.12 26966.20
SE Ueighted 8912.78 136.71 8834 .a9
..
Ueighred Sire 13443.94 46.64 13397.30
SE Veighted 8082.09 12.77 8082.08
Weighted Size 5034.51 42.80 4991.71
SE Veighted 2179.02 42.80 2176.95
73 Sarrple Size 0 0 0
SE Weighted 0.00 0.00 0.00
74 Sample Sire 5 5
,.., . . _____________.................
75 Sample Sire 2 2
Ueighted Sire 399.30 399.30
SE Veighted 386.99 386.99
.... . .
Ueighted Size 1428.21 0.00 1428.21
SE Veighred 993.55 0.00 993.55
__.
35
A!mendix II

"a?.~ 88.96,c:latch.prc
by: SUDCP, OPENED.
Sample Size 0 0 0
SE Veighted 0.00 0.00 0.00
...-........................~~~......
78 Sample Sire 244 236

Weighted Size 111623.67 682.0: 110941.64
I77 I SE Ueighted 25828.77 287.85 25776.79
. . . . ._ _ _ _
Ueighted Size 1119.72 0.00 1119.72
SE Weighted 967.95 0.00 967.95
.._________....... __. .----......
SE Weighted 0.00 0.00 0.00
al Sample Size 1 0 1
SE Weighted 12.02 0.00 12.02
a2 Sample Size
Ueighted Size 16.82
1 01
0.00 16.82
1
SE Weighted 16.82 0.00 16.82
.----............_.......~..~.~~~~. .----._..........._______
a3 Sample Size 3 0 3
Weighted Size 1159.06 0.00 1159.06
SE Weighted 917.90 0.00 917.90
.----...... ............
36
ADDendix II

"ass aa-96,c:latch.prc
by: WOW, OPENED.
SE Weighted 0.00 0.00 0.00
.
a7 Sample Size 3 0 3
Ueighted Size 465.33 0.00 465.33
SE Ueighted 297.08 0.00 297.08
. . . .._...._...... _.........
Weighted Size 1880.30 0.00 1880.30
SE Weighted 1017.11 0.00 1017.11
.................. .:. . . . . . . . . . . .
n 0
a.06 0.00
0.00 0.00
. .
SE Veighted 59.52 0.00 59.52
.....---......................................~~..
31
Amendix II

by: SUoCP, OPENED.
SE Weighted 0.00 0.00 0.00
Ueighted Size 2411.19 0.00 2411.19
SE Veighted 1378.06 0.00 1378.06
Weighted Size 336.35 0.00 336.35
SE Ueighted 271.11 0.00 271.11
SE Weighted 0.00 0.00 0.00
38
A!mendix II
oete: 08-08-97 Research Triangle institute Page : 15

Time: 14:17:17 The CRoSSTAB Procedure Table : 1
"es5 aa-96,c:Latch.prc
by: WOW, OPENED.
SUOGP OPENED
Total I I 2 I
____._............ . . . . . . . ..__.
I
SE Ueighted 0.00 0.00 0.00
. . . .
SE Weighted 0.00 0.00 0.00
. .

Weighted Size 1124.03 0.00 1124.03
SE Weighted 662.90 0.00 662.90

Veighted Size 0.00 0.00 0.00
SE Ueighted 0.00 0.00 0.00
_____.....______. .......... .__

Weighted Sire 2592.31 30.79 2561.52
SE Weighted 1707.26 30.79 1706.98
. . . . . . . .._______ ________._....... . . . . . .._____._..........

Weighted Size 11084.96 161.51 10923.45
SE Weighted 6466.72 116.87 6470.81
................

Ueighred Sire 4243.69 0.00 4243.69
SE Ueighted 1639.52 0.00 1639.52
39
ADDendix II

"a** aa-96,c:latch.prc
by: SUDGP, OPENED.

SE Ueighted 0.00 0.00 0.00

SE Ueighted 32.84 0.00 32.84

SE Ueighted 0.00 0.00 0.00
.............. . . . . . . . . ...__.............................. .
109 Sample Sire 0
0.00
0.00
110 Sample Size

Weighted Size
SE Weighted I 1
487.01
487.01
0
0.00
0.00
487.01
487.01
1
._

SE Weighted 0.00 0.00 0.00
40
ADDendiX II
Ode: 08-08-97 Research Triangle Institute PBge : 17

"Bee aa-Pb,c:Latch.prc
by: SUDGP, OPENED.
SUOGP
112 Sample Size

Weighted Size
SE Ueighted
29
13166.32
4774.43 12.19
12.19
1
I
4767.11
13154.13
.__.......................
28
I
113 Sample Size 1
Ueighted Size 4428.5: 11.01
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ..~~~...........
114 Sample Size 22 21

Veighted Size 4230.55 4192.85
SE Weighted 1215.99 1215.40
_._..........
115 Sample Size 1 1

__..__........
116 Sample Size

Veighted Size
SE Weighted I 10
2438.49
1156.83
I
10
2438.49
1156.83
ii8 Sample Sire 0 0 0

SE Weighted 0.00 0.00 0.00
.................
41
ADDendix II
Date: 08-08-97 Research Triangle Institute Page : ia

"ass aa-96,c:lstch.prc
by: SUOGP. OPENED.
SUDGP OPENED
I . . . . . . . . . . . . . . . . ..j . Total
I1 I
.._________________~...~.............
2

SE Weighted 0.00 0.00 0.00

Ueighred Size 0.00 0.00 0.00
SE Veighted 0.00 0.00 0.00
. . . . . . . . . . . . . . . . . . ..____________________.................................... .

Ueighted Size 11025.56 0.00 11025.56
SE Weighted 4288.24 0.00 4288.24

SE Weighted 24.08 24.08 0.00
.........

Ueighted Size 3013.59 0.00 3013.59
SE Weighted 1681.87 0.00 1681.87
.......................
124 Sample Size

Weighted Sire
SE Yeighted I 0
0.00
0.00
0
0.00
0.00
0
0.00
0.00

Weighted Size 16636.91 0.00 16636.91
SE Weighted 11049.24 0.00 11049.24
42
ADDendix Ii

"ass aa-96,c:Latch.prc
by: SUDGP, OPENED.
OPENED
Total I 1 I2 I
126 Sample Sire a 0 a

Weighted Size 5264.53 0.00 5264.53
SE Weighted 2932.25 0.00 2932.25

SE Ueighted 0.00 0.00 0.00
. . . . . . . . . ..__.........

Weighted Size 9855.65 336.50 9519.15
SE Ueighted 4492.91 238.35 4282.57
.______................

Weighted Sire 13297.33 0.00 13297.33
SE Weighted 11436.65 0.00 11436.65
. . . .
130 Sample Size 482 24 458

Ueighted Sire 193074.79 2460.44 190614.35
SE Ueighted 38776.26 765.77 38742.05

Weighted Size 8148.25 0.00 8148.25
SE Weighted 4272.58 0.00 4272.58
............
I
. . .

Weighted Size 6115.70 0.00 6115.70
SE Veighted 2285.15 0.00 2285.15
43
ADDendix II

Time: 14:17:17 The CRoSSTAg Procedure Table : 1
"a** 88-96,c:Latch.prc
by: SUOGP, OPENED.
SUDGP
I
__......._._.........................
4 0 4
2132.85 0.00 2132.85
1566.87 0.00 1566.87

SE Weighted 0.00 0.00 0.00

SE Weighted 0.00 0.00 0.00
__.__.....
137 Sample Sire 2 0

weighted Size 1467.54 0.00 1467.5:
SE Weighted 1118.47 0.00 1118.47
_________........ . . . . . . . . . . . . .._____.......

Ueighted Size 2840.08 0.00 2840.08
SE Weighted 1450.11 0.00 1450.11
___________...... ___.___._...............

Ueighted Size 268.49 0.00 268.49
SE Ueighted 268.49 0.00 268.49
44
Aooendix II
oaf=: 08-08-97 Research Triangle Institute Page : 21

by: SUOGP, OPENED.
SUOGP OPENED
Total I 1 I2 I
0 1
0.00 6.39
0.00 6.39
..,

l&sighted Sire 15879.20 0.00 15879.20
SE Ueighted 5626.78 0.00 5626.78
..,

Ueighted Size 1901.91 0.00 1901.91
SE Weighted 808.21 0.00 808.21
._

Weighted Size 9973.02 168.04 9804.97
SE Weighted 3986.79 168.04 3968.95

Weighted Sire 3499.52 0.00 3499.52
SE Veighted 1985.77 0.00 1985.77

Weighted Size 53271.82 0.00 53271.82
SE Weighted 14237.53 0.00 14237.53

Weighted Size 36635.47 308.19 36327.28
SE !&sighted 23944.85 308.19 23942.82
4s
ADDendiX II

nass 88.96,c:latch.prc
by: SUooP, OPENED.
SUOGP OPENED
Total I1 I2
______................................
2 34
14356.86 303.85 14053.04
4687.01 246.85 4687.50

Weighted Size 9812.72 0.00 9812.72
SE Ueighted 4707.19 0.00 4707.19

SE Ueighted 0.00 0.00 0.00
.............. .................

Weighted Size 14733.61 1184.45 13549.15
SE Weighted 5915.90 1184.45 4911.23
.:. ______........... .._ __.........

Weighted Size 1272.31 0.00 1272.31
SE Weighted 913.23 0.00 913.23
153 Sample Size 0 0

46
ADDendix II

nass a&96,c:Latch.prc
by: SUOGP, OPENED.
__________...................................................................
SUOSP OPENED
I Total I1 I2 I
Sample Sire
Weighted Size
SE Weighted I 1
42.06
42.06
0
0.00
0.00
1
42.06
42.06
Sample Size
Ueighted Sire
SE Ueighted I 25
6089.71
2541.24
0
0.00
0.00
25
6089.71
2541.24
156 Sample Size

Ueighted Size
SE Ueighted I 15
2500.41
910.20
0
0.00
0.00
15
2500.41
910.20

SE Weighted 0.00 0.00 0.00
. . . .. . ..__.....

Weighted Sire 343.67 0.00 343.67
SE Weighted 343.67 0.00 343.67
................ . .

Ueighfed Size 16910.30 0.00 16910.30
SE Ueighted 5965.92 0.00 5965.92
I
. .

Ueighted Sire 5999.84 0.00 5999.84
SE Veighted 1951.00 0.00 1951.00
41
Atmendix II

Time: 14~17~17 The CROSSTAB Pracedure Table : 1
"a** aa-96,c:latch.prc
by: SUDCP, OPENED.
SUOGP OPENED
I Total I 1 I 2 I
161 0 4
0.00 331.48
0.00 207.44
._.._......... . . . . . . .._.

Weighted Size 12650.47 146.99 12503.48
SE Weighted 4225.23 146.99 4234.09
. . . ..__...

Ueighted Sire 10944.66 0.00 10944.66
SE Ueighted 4852.23 0.00 4852.23
______..........., . .

Yeighted Size 10278.48 0.00 10278.48
SE Ueighted 3673.41 0.00 3673.41
._. ._.........
165 Sample Size I 332 11

321 I
166 Sample Sire

Weighted Sire
SE Weighted I 7
1176.72
674.86
0
0.00
0.00
7
1176.72
674.86
167 Sample Sire 6 0

Weighted Sire 1839.96 0.00 1839.92
SE Weighted 1525.70 0.00 1525.70
._.
48
Atmendix II
bate: 08-08-97 Research Triangle Institute Page : 25

by: SIJOGP, OPENED.
SUOGP OPENED
Total I1 I2 I
0 0
0.00 0.00
0.00 0.00

SE Ueighted 0.00 0.00 0.00
._.
Ueighted Sire 1307.81 0.00 1307.81
SE Ueighted 924.92 0.00 924.92
171 Sample Size

Yeighted Size
SE Ueighted
6
2841.76
1722.72 I 0
0.00
0.00
6
2841.76
1722.72
0 0
0.00 0.00
0.00 0.00
. .

Weighted Sire 8806.78 0.00 8806.78
SE Ueighted 4901.65 0.00 4901.65

Ueighted Size 1467.38 0.00 1467.38
SE Yeighted 1431.75 0.00 1431.75
. . . . . . . . . . ..________~~~~~~~~~.......................................~~~~~~~.
49
Armendix II

Time: 14~17~17 The CROSSTAB Procedure Table : 1
by: SUotP, OPENED.
Sample Size 3 0 3
Ueighted Size 1822.03 0.00 1822.03
SE Weighted 1538.30 0.00 1538.30

Weighted Size 569.36 0.00 569.36
SE Weighted 320.92 0.00 320.92
__. __._............. .........

Yeighted Size 15802.02 167.51 15634.50
SE Yeighted 3244.58 111.25 3260.77
__________....... ...._.______............

Yeighted Size 0.00 0.00 0.00
SE Weighted 0.00 0.00 0.00
________......... .1

Ueighted Size 11348.32 7.16 11341.16
SE Ueighted 4440.29 7.16 4440.96
50
-.
Amendix II
Oate: 08-08-97 Research Triangle Institute Page : 27

by: SUOGP, OPENED.
. . . . . . . ..___.....
WOW
182 Sample Size 0 0

Ueighred Size 0.00 0.00
SE Ueighted 0.00 0.00

Weighted Size 14611.96 52.44 14559.52
SE Weighted 5659.01 52.44 5663.96
_______.................--

SE Ueighted 0.00 0.00 0.00

Weighted Size 1892.66 15.15 1877.51
SE Ueighted 1818.78 15.15 1818.72
_______.............................. .__________

SE !&sighted 32.06 0.00 32.06
51
Aowndix III
Table of the probability of door opening for the range of tests limits from 500 to 54,000 newtons in
the bypass and lateral and longitudinal full door test..
52
Appendix IV
Table 7. Percentage Change in Probability of Door Opening from Current Performance to a range of
test loads
54
Amendix IV

FMVSS206

Uploaded by

Copyright:

Available Formats

FMVSS206

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

FMVSS206

Uploaded by

Copyright:

Available Formats

t-

Attached for inclusion in Docket NHTSA-98-3705 is a research report titled, “Regression

NATIONAL HIGHWAY TRAFFIC SAFE-W ADMINISTRATION

Executive Summary ........ ........ ....... i

1,Background ._.._.__.__.__...__._.,, ........ ........ ...... 1

2. Objective ........ ........ ...... 2

3. Data Sources ........ ........ 2

4. Approach ........ ........ 8

5. Procedure ........ ........ 8

6. Discussion and Analysis of Results ........ ........ 9

7. Conclusions ........ ........ 14

8. References.. ........ ........ 16

9. Appendix I ........ ........ 17

10. Appendix II ........ ........ 23

1 I. Appendix III ........ ........ 50

6. Discussion and Analvsis of Results

wherep = the absolute probability of door openings for NASS vehicles,

P ,,p’“‘Px) Multiplying both side by e”

Table 4. Prediction for suggested test load, 15,000 newtons I

log& a + lxx<, + 4 Equation (4)

Subtracting Equation 4 from 3 yields Equation 5

l-p ) - I%( *) = -pa

BYP= 28.27 -31.28 -72.73 NA

Full DOW Lateral 23.44 -31.06 -67.26 NA

Full Door Longitudinal -6.45 -64.66 NA NA

Figure 1. GM Rotational Test Fixture

Figure 2a. Full Door Lateral Tension Test Fixture

Figure 2b. Full Door Longitudinal Tension Test Fixture

Figure 3. Detent Lever - Fork Bolt Bypass Test Fixture

DOOR LATCH STAT/C LOAD TEST FCfTURE (LONGITUDINAL LOAD)

Figure 4. FMVSS No. 206 Longitudinal Test Procedure

Table 2. Vehicle Latch Groups by Actual and Ueighted Door Openings.

Nunber of observations read 5448 Ueighted count : 2239516

WOW OPENED cl)* (2)"

Total Sample Size 5448 164 5284

Dare: 08-08-97 Research Triangle Institute Page : 2

oate: 08-08-97 Research Triangle Institute Page : 3

Date: 08-08-97 Research Triangle Institute Page : 4

25 Sample Size 235 1 234

Date: 08-08-97 Research Triangle Institute Page : 5

30 Sample Size 326 6 320

Date: 08-08-97 Research Triangle Institute Page : 6

Date: 08-08-97 Research Triangle Institute Page : 7

Date: 08-08-97 Research Triangle Institute Page : 8

I ’lei! Shted Size I 62449.18 I 677.75 1 60771.43 1

Oate: 08-08-97 Research Triangle Institute Page : 9

Date: 08-08-97 Research Triangle Institute Page : 10

69 Sample Size 157 7 150

oate: OS-OS-97 Research Triangle lmtirute Page : 11

oate: 08-08-97 Research Triangle Institute Page : 12

78 Sample Sire 244 236

Date: 08-08-97 Research Triangle Institute Page : 13

Date: 08-08-97 Research Triangle Institute Page : 14

oete: 08-08-97 Research Triangle institute Page : 15

100 Sample Size 7 0 7

101 Sample Size 0 0 0

102 Sample Sire 7 1 6

103 Sample Size 26 2 24

104 Sample Sire 13 0 13

Date: 08-08-97 Research Triangle Institute Page : 16

1,Background ._.._......._.,, ........ ........ ...... 1