/_ ?
`0
`S5-0-09
`Air Bag System for Side Impact Occupant Protection _;__:/_ce=--0~_() ____ _
`
`Toro Kiuchi, Kenji Ogata
`Toyota Motor Corporation
`Charles Y. Warner
`Collision Safety Engineering
`John Jay Gordon
`GMH Engineering
`
`Abstract
`Pilot and prototype designs of a door-mounted air bag
`system for occupant protection in side impact have been
`assembled and tested. The primary goal of the designs
`was to take advantage of the improved space utilization
`offered by the air bag when combined with the padding
`and structural benefits that are contemplated for torso
`injury. Another important goal of the project was the
`demonstration of the head-protection potential of such a
`system, attempting to interpose a pad between the head
`and side structures and intruding objects likely to cause
`impact injury.
`The pilot design was subjected to a test program,
`providing a preliminary evaluation of a system which
`incorporates both head and torso protection in a single
`air bag system. The pilot design showed sufficient
`promise that a preliminary prototype design program was
`undertaken.
`Full-scale crash tests of recent production 4-door
`sedans were conducted to establish baseline perfonnance
`over a range of side-impact conditions, Design objectives
`were analyzed and subsystem performance goals were
`established and proven by component testing. The proto(cid:173)
`type system incorporated two kinds of sensor switches,
`a production steering wheel air bag inflator module, a
`large. flat, tethered air bag, and a fabric air bag cover,
`all mouoted in a modified production door. The complete
`prototype system was evaluated in laboratory tests and
`full-scale crash tests, including FMVSS 214 crabbed
`moving deformable barrier (CMDB) tests employing the
`DOT/SID side-impact dummy, A very satisfactory per(cid:173)
`formance was achieved, as demonstrated by comparison
`of dummy indices measured in baseline and air bag(cid:173)
`equipped vehicles in full-scale crash tests. This paper
`outlines the designs and system configurations and
`discusses the resuHs of the pilot and preliminary design
`test series.
`
`Introduction
`Fatalities and fnjuries in Side lmpacr Accidents
`Many head injuries occur in side impacts, due to
`contacts with interior structures or exterior intruding
`objects. According to accident data collected by the
`National Crash Severity Survey (NCSS) and the National
`Accident Sampling System {NASS), side impact acci(cid:173)
`dents cause about 30 percent of all traffic accident
`occupant fatalities. Head injuries account for 40 percent
`of all fatalities to near-side front-seat occupants in side
`impacts, with chest injuries at 32 percent Objects exteri(cid:173)
`or to the vehicle arc involved in about 40 percent of
`head injuries, presumably due to partinl ejection or
`intrusion, while impacts with A-pillar (19%) and roof
`side rail (17%) structures make an almost equal contri(cid:173)
`bution
`[DOT/NHTSA, 1990-I; Viano,
`1987-1,23~
`Strother, 1990}.
`
`NHTSA Safely Rufemaking Activities
`The most recent .. Final Rule" of revised FMVSS 214
`was issued in October, 1990, aimed al reducing thorax
`and pelvis injury indices, as measured in CMDB crashes ..
`The test procedure includes the use of the DOT/Side
`Impact Dummy (DOT/SID) and the Thoracic Trauma
`Index (TII(d))- The rule currently applies to passenger
`cars produced after September, 1993 for sale in the U,S.
`[DOT/NHTSA, 1990-2). NHTSA is rightly considering
`means to reduce not only thorax and pelvis injuries but
`also head injuries in side impacts. The Advanced Notice
`of Proposed Rulemaking {ANPRM) issued in August of
`1988 suggested head protection by the use of pads on
`pillars and rails, glass-plastic glazing in compartment
`sides, strengthened door hardware, etc. [DOT/NHTSA,
`1988}.
`
`Upgraded Occupant Protection in Side Impact
`It is extremely difficult to provide adequate stroke for
`the absorption of occupant second collision energy in
`side impacts [Warner, 1989; 1990-2}. Attempts to
`upgrade occupant protection in this crash mode have
`generally involved modifications of vehicle body side
`structure and imposition of paddings between occupant
`and interior for thorax, abdomen. and pelvis protection.
`The improvement of body side structure nnd interior
`padding configurations appears to be somewhat effective
`
`29
`
`Page 1 of 10
`
`KSS 1019
`
`

`
`13th lntemational Technical Conference on Experimental Safety Vehicles
`
`for reduction of chest injuries [Warner, 1990-1; Lau,
`1989; Viano, 1989-2; Ridelta, 1990]. It may also be
`partially effective for reduction of head injuries, but it is
`unresponsive to injuries caused by partial ejection or
`foreign body intrusion through the side glass. Our studies
`were directed at the feasibility of using air bags as
`supplemental side impact protection, with particular
`emphasis on head protection.
`
`Toyota/CSE Pilot Study of Side Air Bags
`A joint pilot study was initiated by Toyota and CSE in
`1989, to investigate lhe potential of improved side
`impact occupant protection by use of supplemental air
`bag systems [Warner, 1988], with demonstration tests in
`Toyota production vehicles, and injury index compari(cid:173)
`sons with baseline FMVSS 214 tests and DOT/SID
`dummy.
`
`Configuration of Pilot Side Air Bag System
`The pilot system shown in Figures I and 2 was con(cid:173)
`figured as an extension of findings from the early
`in 1980 Citation automobiles
`preliminary
`testing
`[Warner, 1988, 1990-3]. The bag and inflator were
`attached to a backup plate in the door inner panel, and
`covered with an energy-absorbing inner foam pad meant
`to reduce thoracic injury, The inflator was selected from
`among those already available for steering wheel air
`bags. The bag. was designed to deploy inward and
`upward rapidly, in order to accommodate the limited
`distance available. It included distributed venting and
`tethers to control lateral expansion and encourage
`vertical deployment over the window space, with a width
`of about 100 mm and a volume of about 60 liters
`(Figures 3 and 4). Mechanical intrusion sensors were
`mounted at two levels near the door outer skin" Switch
`closures resulting from small deflections of the door
`provided the inflation signal for the bag system (Figure
`5). Loadpath foams were included to support reaction
`forces and help with energy absorption. For lhe pilot
`study, no conventional interior trim was included on the
`door.
`
`Performance Evaluation of Pilot System
`The pilot system was evaluated in full-scale side
`impact test, using the FMVSS 214 procedure with the
`addition of a Hybrid-III dummy neck and head to the
`DOT/SID; results from the pilot system were compared
`with baseline tests from the same vehicle and proced(cid:173)
`ures. Figure 6 shows ratios of the comparative dummy
`index values. While the ITI(d) reduction of 5% is not
`very significant, other body areas show impressive
`reductions ranging as high as 65%.
`
`Prototype Air Bag System Performance
`Requirements
`The potential of the air bag system for the reduction
`of occupant injuries in side impacts and the technology
`for the bag upward deployment were confirmed in the
`
`loadpalh foam
`
`Sensors
`
`1nterior foam
`
`Figure 1. Configuration of the Pilot Air Bag System
`
`Figure 2. Frontal and Lateral Views of the Deployed Pilot
`Air Bag System
`
`Vent hole
`
`Bag base cloth
`
`Tether
`
`Figure 3. Shape of the Pilot Air Bag and Tether
`Ca nfig uration
`
`pilot system study. Since the top priority of the pilot
`study was given to questions of overall technical feasi(cid:173)
`bility. some of the basic door functions were ignored,
`making necessary further prototype design studies
`regarding the application of this type of system in mass
`production vehicles.
`
`Baseline Full-Scale Impact Tests
`Various types of side impacts, in terms of speed, angle
`and impact position may be observed in actual traffic
`accidents. Development of improved occupant protection
`
`30
`
`Page 2 of 10
`
`KSS 1019
`
`

`
`Folding line
`
`A-A SEC.
`
`B-B SEC
`
`B
`
`(tJ
`
`Figure 4. Pilot System Bag Folding Method
`
`Conventional
`side door beam
`
`Supplemental
`side door beam
`
`Section 3: Technical Sessions
`
`vehicle speeds of 30 mph and 15 mph, respectively.
`Three relative positions along the struck vehicle side
`were evaluated, including the front and rear corner "L"
`configurations and the compartment "T" position speci(cid:173)
`fied in FMVSS 214. As in the pilot program, compari(cid:173)
`sons were directed at that baseline condition. and tests
`were set up to conform with its dummy. CMDB con(cid:173)
`cepts, and injury indices for the six tests representing
`car-to-car exposures .. For the last two tests, the CMDB
`was replaced with a light truck and a fixed pole, respect(cid:173)
`ively, with the pole impact carried out with the car
`moving laterally inlo the pole at 20 mph.
`Table 1 depicts the lest configurations and presents
`important results, expressed as ratios of the result
`obtained in the baseline FMVSS 214 test shown in the
`first column of data. Dummy kinematics and secondary
`impacts between dummy and interior were studied in
`detail by high-speed cinematography. Partial head
`ejections out the window area were observed in some
`cases, as noted below. Observe that head injury indices
`are greater than the NHTSA baseline in six of the seven
`alternate tests, and that all indices are greater in the 60
`degree compartment impact and the perpendicular light
`truck impact Head injury indices are particularly high in
`the concentrated pole impact, resulting from direct
`contact of the head against the pole.
`
`Table 1. Conditions and Results from Baseline Series of
`Full-Scale Crash Tests
`
`Figure 5. Structure of the Pilot System Sensor
`
`(%)
`
`lWCQl'lditXm
`
`SIOO ~.id t1.!)Ail":t1 MOD
`
`0 . .Cl
`0,71
`0.42
`O.Hi
`
`\'ES
`110
`Ito
`
`U&
`
`''"
`o.-'a
`0.17
`
`110
`YES
`110
`
`'"'"
`._,.
`
`1.20
`
`0.17
`
`JIO
`YES
`flO
`
`EjodJon
`ln11trkirC011bct
`Eiito~c:aibd.
`
`\'ES
`110
`HO
`
`'/}'
`.. ,.
`·~·
`
`2'.G9
`
`1.tl7
`
`t/O
`YES
`ND
`
`2.117
`
`,...,
`0,4.2
`om
`UD
`YES
`110
`
`~ ' ' f
`
`GU..~ =a.~
`
`......... l{; .... lf'<'l'o
`
`,.,.
`
`I.GO
`O.!lG
`0.(16
`
`2.1)<
`
`.,.
`'"
`
`1.14
`
`VES
`110
`VES
`
`YES
`110
`YES
`
`HIC Head Neck Neck TII CdJ Pelvis G
`3msG Mx
`Fy
`
`Figure 6. Dummy Injury Indices from Full-Scale Side
`Impact Test
`
`suggests the need for consideration and evaluation of
`dummy occupant behavior over a wide range of side
`impact accident conditions. A matrix of eight typical side
`impacts representing some prominent types of injurious
`side impacts was selected for comparative evaluation of
`the prototype side impact air bag system, and full scale
`impact tests were carried out to evaluate them. Impact
`angles of 60 degrees and 90 degrees were selected for
`seven vehicle-to~vehicle tests, with striking and struck
`
`31
`
`Target Air Bag Pe1jorma11ce
`During the pilot study, bag occupant protection per(cid:173)
`formance was given top priority at the expense of the
`normal door functions. Since the presence of an arm rest
`is considered essential, and since it has proven to be
`very difficult to deploy the bag both above and below
`the arm rest, it was decided to deploy the bag only above
`the arm rest The pelvis portion of the occupant was thus
`excluded from the intended direct coverage area of the
`bag. Relationships among door intrusion, dummy motion,
`and elapsed time were determined by analysis of high
`speed films. The most stringent intrusion rates among the
`six baseline vehicle tests was found lo occur under the
`FMVSS 214 test conditions, so that test was chosen for
`comparative evaluation of design goals. The relationship
`among bag thickness. deployment lime and impact sens(cid:173)
`ing can be ~pproximated as shown in Figure 7, where
`door intrusion D (t) and bag thickness W (t) are plotted
`
`Page 3 of 10
`
`KSS 1019
`
`

`
`13th lntematfonal Technical Conference on Experimental Safety Vehicles
`
`against time. The intrusion of the deploying bag surface
`B (t) can be expressed by Equation L
`B (t) = D (t) + W (t)
`t 1 : Sensing lime
`t2 : Air Bag Deployment Completion Time
`8(t) = D(I) + W(t)
`
`(l)
`
`c
`0
`'iii
`
`2 .:s
`
`!
`I
`
`>-
`
`I /~ --::::::::::-~::W(tl
`t
`t t. ... --(cid:173)
`i
`-
`
`,
`
`_,,,..
`
`D(t)
`
`Time
`
`(msec)
`
`Figura 7. Intrusion of Air Bag Surface
`
`It is oi ~.::mrse desirable that contact between the air
`bag and the occupant occur at or after completion of air
`bag deployment, so that bag thickness may be used for
`energy management and its area may be used for force
`distribution over the occupant. As the initial distance
`belween the door inner panel of the vehicle used in this
`study and the normally-seated DOT/SID occupant is 130
`mm, Equation 1 becomes:
`130 s D (t2) + W (t2)
`(2)
`Figure 8 plots Equation 2, using the actual door intrn(cid:173)
`sion determined by the baseline FMVSS 214 test, and
`demonstrates that bag thickness and deployment comple(cid:173)
`tion time must be traded off against one another. The
`bag sliould be as thick as possible in order to use it as
`effectively for occupant energy absorption stroke, so the
`bag thickness at chest height was given top priority. A
`target deployed thickness of 100 mm was selected lo
`accommodate door intrusion, and tether length was set
`accordingly. Figure 9 shows the configuration of the bag
`at complete deployment, giving coverage from the arm
`rest to the roof side rail for head protection, as suggested
`by observed dummy kinematics and expected seating
`positions for a range of occupants.
`
`(mm)
`
`200
`
`Initial Distance "'130mrn
`
`0
`
`10
`
`l2
`
`TI me
`
`20
`
`(msec)
`
`Figure 8, Relationship Between Door Intrusion and
`Air Bag Thickness
`
`32
`
`Figure 9. frontal and Lateral Views of the Deployed Bag
`
`The amount of energy to be absorbed in the secondary
`impact speed between the dummy chesl and the door was
`calculated as 1000 1 from the baseline FMVSS 214 test ..
`Figure IO shows the conceptional load-displacement
`characteristics of the door with and without the bag.. A
`target energy absorption value of 500 J (50% of the total
`energy) was selected, assuming that the reduced door
`deformation by the bag will suppress Lhe maximum load
`somewhat With the bag thickness set al I 00 mm, the
`bag must complete its deployment within 13 msec.
`
`Displacement
`
`Figure 10. Load-Displacement Characteristics of Door
`Without and With Bag
`
`The total time to 100 mm deployment was arbitrarily
`divided into 2 to 3 msec. for sensing and 10 to 11 msec.
`for bag deployment. Bag actuation duration to give ade(cid:173)
`quate protection was targeted at 100 rnscc., based on
`dummy behavior in the lests. so that inflator charac-·
`teristics and venting behavior were balanced around that
`goa1.
`The most important requirement for the sensors is lo
`discriminate the need for bag deployment for occupant
`protection in side impact conditions, while avoiding
`inadvertent deployment in more normal conditions. The
`targeted 2-3 msec. sensing time is part of the 13 msec.
`deployment time, suggesting that G-sensors are impracli(cid:173)
`cal for this system .. Another type of sensor capable of
`more rapid sensing is required.
`
`Design of Prototype Side Air Bag System
`Individual components that constitute the side air bag
`system were designed to be assembled into a front door,
`as shown in figure 1 L
`
`Page 4 of 10
`
`KSS 1019
`
`

`
`Section 3: Technical Sessions
`
`Door Structure and Trim Modificatio11s
`A large opening was provided within the door inner
`panel for the installation of the backup plate, with a
`compensating window sill reinforcement. This allowed
`lhe system to function without major changes in overall
`door deformation characteristics. The door trim was
`similar to that of the baseline production vehicle, but a
`newly-designed trim cover for the air bag module was
`added.
`The upper portion of the trim cover was intended lo
`break away inward, encouraging upward bag deployment.
`
`Sensor Design
`In place of an ordinary G-sensor, contact switch
`sensor system to be installed to the outermost location of
`the door was considered to sense impacts as quickly as
`possible. According to the results of the preliminary
`the characteristics of the contact
`study regarding
`switches, it was found that the activation pressure of
`such switch should be set rather high level to avoid an
`inadvertent air bag deployment when, for example, the
`door was opened and hit against a pole, tree, etc. where
`the relatively high load was concenlrnted around the
`narrow contact area. And it was also predicted that such
`a switch with higher activation pressure might not acti(cid:173)
`vate even when an ordinary vehicle collided into the side
`door. Therefore, it was decided to develop a contact
`switch sensor system consisting of two switch systems(cid:173)
`the primary switch system to sense the impact with some
`other vehicle, and the secondary switch system that
`would not be activated in the impact on the pole when
`opening the door, but to be activated by a more severe
`side impact against a pole.
`The primary switch system consists of plural contact
`switches installed at given intervals, and each one of
`them turns on easily by relatively low activation press(cid:173)
`ure. The primary switch system itself turns on only
`where two or more contact switches are turned on simul(cid:173)
`taneously, but it remains off where only one of the con(cid:173)
`tact switches is activated by a mischievous action, etc.
`The secondary switch system was installed in the
`longitudinal direction of the door, which would not be
`turned on when the door hit a pole or a tree, even if the
`door was accidentally and strongly opened, but turns on
`in a side impact against a pole, tree, etc., where the air
`bag deployment is required to reduce the occupant
`injury.
`total contact switch sensor system was so
`The
`constructed that it would turn on if either the primary or
`secondary switch system described above turned on.
`the
`Figure 13 shows the sensor configuration and
`installation. The sensors were installed immediately
`inside the door outer panel along the pipe-wise side door
`beam so that they sense impacts as quickly as possible ..
`
`Door inner panel
`
`Door trim
`
`lollator::-0 ~ ~
`
`Backup plate ~
`Bag
`
`Figure 11. Configuration of the Air Bag System
`
`Air Bag, !njlator and Backup Plate
`The system consists of bag and inflator combined with
`backup plate, sensor, door panel and door trim, including
`air bag cover. Individual components were carefully de(cid:173)
`signed to be compatible with all normal door functions.
`The bag was designed to meet the thickness and cover(cid:173)
`age ranges decided, with tether length set at 100 mm, as
`shown in Figure 12. Bag volume is about 40 liters.
`
`Vent hole
`
`Tether
`
`/0,---.. \
`I
`I
`\
`'
`..... _,..,,.
`/
`
`I
`I
`I
`
`Figure 12. Shape of the Bag and Locations of the Tether
`
`Development of a custom-designed inflator would re(cid:173)
`quire a much longer time than available under the proto(cid:173)
`type project, so it was decided to use an existing or
`modified inflator designed for frontal air bag systems.
`Estimates of bag contact area with the occupant (500
`
`cm2), bag thickness (lOOmm) and the energy (500 J),
`suggest a dynamic internal pressure requirement of about
`10 N/cm2
`• An optimal inflator was selected from among
`several available inflators by means of a series of infla(cid:173)
`tion tests. The backup plate to which the bag and the
`inflator were assembled was designed so that the bag and
`the inflator were positioned within the limited inner
`space of the door without sacrificing normal door and
`window functions.
`
`33
`
`Page 5 of 10
`
`KSS 1019
`
`

`
`13th International Technical Conference cm Experimental Safety Vehicles
`
`(mm)
`
`200
`
`100
`
`(/J "' Ql c .x
`0
`:c
`I-
`
`0
`
`20
`
`40
`
`Time
`Figure 15. Bag Thickness at Occupant Chest Height
`
`60
`(msec)
`
`Figure 13. Configuration of the Sensor
`
`Evaluation of Bag Pe1formance in Static Bench Testing
`Deployment tests were carried out to evaluate folding
`techniques, internal pressure, venting, and overall bag
`performance, leading to design modifications in an effort
`to optimize performance within the criteria outlined
`above. Figure 14 shows the final configuration of the
`prototype bag. Figure 15 presents the dynamic bag thick(cid:173)
`ness at occupant chest height, taken from test films.
`L0"al bulging is obvious, ex.tending locally beyond 100
`. '\ gives the bag surface a mattress-like surface.
`,,;; folly acceptable, so long as the 13 msec. deploy(cid:173)
`ment time is achieved. Figure 16 shows the bag internal
`pressure vs. time, as recorded in a static deployment.
`The average pressure is lower in this static test than the
`10 N/cm2 dynamic pressure target, but pressure will be
`higher due to occupant contact forces in a dynamic
`deployment.
`
`Tether
`
`Figure 14. Final Configuration of the Bag and Tether
`Locations
`Subsystem Deployment Tests on Vehicle
`The prototype bag system was installed in a vehicle
`equipped with DOT/SID for a series of static deploy(cid:173)
`ments. Figure 17 represents the developing configuration
`during static deployment at various times after ignition;
`Figure 18 is a photograph taken at 25 msec. No unfavor(cid:173)
`able interactions with dummy or seat were noted, except
`that the motion of the trim cover induced a 70 g rib
`acceleration in the dummy. A subsequent test without the
`trim cover showed a reduction to 30 g. The trim cover
`was removed for the remainder of testing in the proto-
`
`(Nlcm')
`
`10
`
`Time
`Figure 16. Bag Internal Pressure vs. Time
`
`50
`
`100
`
`The range of the deployment
`
`Figure 17. Process of Bag Deployment
`
`Figure 18. Condition of the Bag at 25 msec
`
`type project, with the idea that minor trim cover redesign
`efforts could resolve this dummy rib overload ..
`Impact tests were carried out to study the load(cid:173)
`displacement characteristics of the air bag system as
`installed in the door. The door was supported against a
`rigid barrier face while the inflating bag was struck al
`full thickness from the inside by an impactor which
`simulated the occupant. The result was the load-displace-
`
`34
`
`Page 6 of 10
`
`KSS 1019
`
`

`
`ment curve depicted in Figure 19. Target values of air
`bag energy and door energy were achieved.
`
`(kN)
`
`20
`
`"O "' .3
`
`10
`
`·36
`
`~
`
`The displacement
`of the bag
`
`The displacement
`of the door
`,
`I
`
`100
`Displacement
`
`200
`
`(mm)
`
`Figure 19. Load-displacement Characteristics of the Door
`With the Side Air Bag
`
`Evaluation of Sensor Performance
`The primary system used to sense striking vehicle side
`impact into the compartment must activated within 2 to
`.3 msec. Therefore it was decided to use switches which
`would be activated by a specified small activating stroke
`and load as the primary switch system.
`Primary sensor performance was evaluated to demon(cid:173)
`strate abilities to achieve sensing within 2-3 msec, by
`impacting a switch system subassembly attached to the
`rigid barrier by a FMVSS 214 moving barrier face at a
`speed of 3 m/s, with a resulting sensing time of 9 msec.
`Assuming an approximate inverse relationship of sensing
`time with speed, the required sensor performance was
`indicated. It was also verified that the primary switch
`system would not be activated when only one switch was
`turned on. Tests of the secondary sensor subsystem veri(cid:173)
`fied that it would not activate when impacted by a 160
`mm diameter pole at maximum foreseeable occupant
`door opening force, while other tests verified that the
`secondary sensor system will activate when struck by a
`barrier-mounted 305 mm pole system having the same
`mass as the vehicle and a speed of 2.8 m/s, yielding a
`sensing time less than the selected target of 7 msec, This
`speed was selected as an appropriate threshold for
`deployment in pole side impact.
`
`Full-Scale Side Impact Tests of the
`Prototype System
`Full-scale FMVSS 214 side impact tests were carried
`out to evaluate the overall performance of the prototype
`system. The sensor system activated within 2 msec., and
`the bag deployed properly, with the dummy acceleration
`results shown in Figure 20. The most significant differ(cid:173)
`ence in the upper rib acceleration was that the first peak
`occurred earlier than in the baseline test, coinciding with
`the contact between dummy and air bag, The second
`peak occurred when the dummy contacted the door inner
`panel through the bag, the third by boltoming of avail(cid:173)
`able door structure deformation. The highest peak was
`reduced by 20 percent, compared with the baseline test
`Accelerations of the lower spine do not show much
`
`35
`
`Section 3: Technical Sessions
`
`differences from the baseline, except for timing .. This is
`attributable to the reduction of the rib acceleration and
`the reduction of the pelvis acceleration which will be
`described later. Overall thoracic trauma should have been
`reduced by the air bag, as suggested by the ten percent
`reduction in TTI(d) provided. Notably, pelvis accelera(cid:173)
`tion was decreased by 24 percent compared to baseline
`test results, despite the fact that the bag did not contact
`the pelvis directly, probably due lo pelvis motion in
`response to transmission of thorax accelerations through
`the spine, moving the pelvis away from its later door
`contact. Figure 21 shows reductions achieved by the pro(cid:173)
`totype air bag system in several dummy injury indices,
`
`(g) Upper rib acceleration
`
`c
`0
`~ w
`@
`
`<(
`
`with Bag
`without Bag
`
`(msec)
`
`(g) Pelvis acceleration
`
`(msec)
`Time
`Figure 20. Dummy Acceleration from Full-Scale Side
`Impact Test With the Side Air Bag
`
`(%)
`
`100
`
`x
`.l!.J
`E
`i':;'·
`:J
`£
`>
`E
`E
`:J
`0
`0
`
`0 = "' a:
`
`Figure 21. Dummy Injury Indices from Full-Scafe Test:
`Prototype Side Impact Side Air Bag Compared to
`Baseline
`
`Page 7 of 10
`
`KSS 1019
`
`

`
`13th lntemational Technical Conference on Experimental Safety Vehicles
`
`Although head injury indices in the prototype system
`test are slightly higher than baseline, they are well below
`injury levels of concern. No head contact occurred with
`anything in the baseline test, nor with anything other
`than the air bag in the prototype system. so this compari(cid:173)
`son is somewhat mute regarding head injury effective(cid:173)
`ness. Figure 22 compares head position at 80 msec., with
`and without the bag. Note the risk suggested by the
`partial ejection of the head in the baseline test and the
`protection provided by the bag. As a comparison, Figure
`23 shows similar views from the baseline tests at the
`30/15 mph condition with a light truck and with the pole
`at 20 mph. In both of these test cases, the dummy head
`contacted objects from outside the vehicle with serious
`injuries indicated.
`
`Figure 23. Impact of Dummy Head Against Harmful
`Objects: Against Light Truck (Top), Against Pole (Bottom)
`
`protection from torso and head injuries have been
`assembled and evaluated in side impact testing
`conducted with the DOT/SID dummy and FMVSS
`214 CMDB procedures and criteria.
`• Folding methods and tether locations for deployment
`of the head-protection aspects of the bag have been
`clarified.
`• Two kinds of sensor switches have been incorpo(cid:173)
`rated into a successful side impact intrusion sensing
`system.
`2. A preliminary prototype system was developed by
`careful refinement of the design principles embodied
`in
`the pilot design. As compared
`to baseline
`vehicles, this improved design demonstrated sub(cid:173)
`stantial improvements in dummy injury measures in
`various vehicle-to-car and car-to-barrier crash
`testing.
`• Thoracic injury risk is reduced by ten percent as
`measured by the DOT/SID and the TTI(d) criteria.
`• While not specifically addressed as a design goal of
`the prototype air bag system, reductions in pelvic
`injury seem to be indicated by the test results.
`• The test results have provided a clear demonstration
`of the potential effectiveness of the prototype system
`in the prevention of head injury in side impacts due
`to contact with vehicle interior surfaces and objects
`near or protruding inward through the side glass.
`
`Figure 22. Comparison of Dummy Head Behavior With
`Bag (Top) and Without Bag (Bottom}
`
`The dynamic tests confirmed that the air bag system
`can be effective in overaH occupant protection in side
`impact. Although the head protection potential has not
`yet been fully evaluated in a numerical sense. it is clear
`that the prototype air bag system can provide substantial
`and meaningful protection from this important injury
`source as well.
`
`Conclusions
`1. A pilot design and a more sophisticated preliminary
`prototype design for air bag systems for side impact
`
`540
`
`36
`
`Page 8 of 10
`
`KSS 1019
`
`

`
`Future Research Objectives
`The prototype side impact air bag system tested and
`studied in this project has demonstrated promise for
`further research in the following areas:
`• The effect of interposition of an inflating bag on
`head injuries deserves further study
`in various
`impact modes .. This may well prove to be the great(cid:173)
`est potential benefit of the side impact air bag
`system ..
`• A complete study of the trade-offs in injury pre(cid:173)
`sented by other effects is called for, including:
`potential hearing damage due to the near proximity
`of the inflating bag, potential arm damage for the
`reported minority of occupants who may lean an arrn
`on the window sill (Viano, 1989], potential arm or
`chest damage due to rapid opening of the air bag
`cover, etc. These concepts must be considered care(cid:173)
`fully for both inlended and inadvertenl deployment
`situations.
`• Sensing strategies and hardware must be refined
`somewhat before proceeding to production .. This will
`require careful study of the various impact situations
`and great effort to shorten the sensing time.
`• Inevitable weight and cost penallies of incorpornlion
`of the side impact air bag system will need careful
`study before introduction of such systems in produc(cid:173)
`tion vehicles.
`References
`DOT/NHTSA, 1990-1: Final Regulatory Impact Analy(cid:173)
`sis: New Requirements for Passenger Cars to Meet a
`Dynamic Side Impact Test, FMVSS 214, U.S. Depart(cid:173)
`ment of Transportation, Washington, D.C., August
`1990.
`DOT/NHTSA, 1990-2: CFR 49 Parl 571. Docket No. 88-
`06, Notice 8, Federal Motor Vehicle Safety Standards:
`Side Impact Protection. U_ S. Department of Trans(cid:173)
`portation, Washington, D.C., October 1990.
`DOT/NHTSA, 1988: CFR 49 Part 571. Docket No, 88-
`06, Notice 3, ANPRM FMVSS 214 Amendment to
`Side Impact Protection - Passenger Cars. U. So
`Department of Transportation, Washington, D.C.,
`August 1988,
`Lau, 1989: "Design of a Modified Chest for EUROSfD
`Providing Biofidelity and Injury Assessment" Ian V,
`Lau, David C ... Viano, Clyde C. Culver and Edward
`Jedrzejczak, GlvfRL, [SAE Paper # 890881) SAE
`International Congress and Exposition, Detroit March,
`1989.
`Ridella, 1990: "Determining Tolerance lo Compression
`and Viscous Injury in Frontal and Lateral Impacts"
`Stephen A. Ridella and David C. Viano, [SAE Paper
`
`Section 3: Technical Sessions
`
`# 902330] Thirty Fourth Stapp Car Crash Conference
`Proceedings Orlando, Fla .. November 4-7. 1990 ..
`Strother, 1990: "Reconstruction and Side Impact Societal
`Benefit" Charles E. Strother and Charles Y, Warner,
`SAE International Congress and Exposition, Detroit
`[SAE Paper# 900379]. March, 1990 ..
`Viano, 1987-1: "Evaluation of the SID Dummy and TTI
`Injury Criterion for Side Impact Testing" David C
`Viano, GMRL [SAE Paper # 872208] Thirty First
`Stapp Car Crash Conference Proceedings New Orleans
`La. November 9-1 l, 1987
`Viano, 1987-2: "Evaluation of the Benefit of Energy
`Absorbing Material in Side Impact Protection: Part I"
`David C. Viano, GMRL [SAE Paper# 872212] Thirty
`First Stapp Car Crash Conference Proceedings New
`Orleans La. November 9-11, 1987
`Viano, 1987~3: "Evaluation of the Benefit o( Energy
`Absorbing Material in Side Impact Protection: Part II"
`David C. Viano, GMRL [SAE Paper# 872213] Thirty
`First Stapp Car Crash Conference Proceedings New
`Orleans La, November 9-11, 198 7
`Viano, 1989-1: "Patterns of Arm Position During Normnl
`Driving" Research Note: Human Facwrs, 1989. JI (6),
`Viano, 1989-2: "Biomechanical Responses and Injuries
`in Blunt Lateral Impact" Duvid C. Viano, GMRL
`[SAE Paper# 892432] Thirty Third Stapp Car Crash
`Conference Proceedings Washington. D.C .. October 4-
`6, 1989.
`Warner, 1988: "Inflatable Structures for Side Impact
`I,
`Crash Protection," Final Report, SBIR Phase
`Contract# DTRS-57-86-C"00089, U ,S. Department of
`Transportation,