Richard J. L'Abbe, James A. Newman and Andre M. St-Laurent Biokinetics and Associates Limited Ottawa, Canada

Transcription

1 DEVELOPMENT OF A NEW ATD THORAX Richard LAbbe ames A Neman and Andre M StLaurent Biokinetics and Associates Limited Ottaa Canada David A Dainty Department of Kinanthropoloy University of Ottaa Ottaa Canada Kevin Smyth and Anthony Bosik Davis Enineerin Limited Ottaa Canada A ne prototype ATD thorax has been developed to be compa tible i th the GM Hybrid test dummy This ne ehest provides a better opportuni ty to incorporate certain characteristics of live human beins hen subj ected to upper torso belt loadin This paper describes the deflection response characteristics of the ne surroate at subinjurious and injurious impact leve ls These results are compared to the responses obtained for other ATDs NTRODUCTON This paper is divided into three sections The first deals ith the continuation of the research proram first reported by the authors in 982 ( ) The second section describes the mechanical desin of a ne thorax structure The desin of the ne thorax is based partly on h uman volunteer response to be lt loadin () and partly on cadaver response to blunt sternal loadin (25) The third section describes the results of a sled test proram desined to quantify the response of the ne thorax and to study the overa l l ATD kinematic response in a documented injurious loadin environment (6 7) THORACC MPACT RESPONSE TO STRP LOADNG The testin described in this section is a continuation of the tes t proram initiated in 98 () The oriinal test apparatus as developed to assist investiators in mappin thoracic deflections under static point loadin and static and uasi static strip loadin (driver confiuration) These tests ere conducted ith the subjects lyin in a supine position ith a fully supported thoracic spine Durin the second phase of this proram a detailed analysis of thoracic deflections as performed only alon the path of the belt namely the riht 7th rib loer and upper sternum and clavicle locations t as assumed that deflections in other reions of the thorax ere of secondary importance in terms of overall body res p onse The surroates te sted ere The Part 572 Hybrid and MRA ( OPAT) ATDs and one human volunteer The human subj ect as chosen from the pool of voluntee rs used in the 982 study Durin the current phase he as retested for the tensed state Surroate Response to Belt Loadin Deflection vs time traces ere taken for three locations alon the path of the belt Belt force vs time as also recorded Upon detailed examination of the test results the human volunteers response to loadin as found to have poor repeatabili ty lt became apparent that the deree of muscular tensin from one test to the next could not be controlled to the deree oriinally anticipated by the investiators nterpretation of such data should thus be liberal Response to strip loadin as also examined on the MRA Hybrid (Part 572) and Hybrid dummies Fiures throuh 6 illustrate forcedeflection responses for a 8 oule impact yieldin an averae peak belt load of 32 kn (see Reference for measurement techniue) Each surroate as then retested ithout its standard soft tissue cover in an attempt to quantify the net effect of the soft tissue cover on the overall deflection response of the thorax The Hybrid ehest exhibited the loest amplitude deflection for all three locations in some cases as much as 5% loer than the human volunteer The Hybrid ehest response in the upper sternal reion as quite similar to the 3 3

2 Force Deflection Responses to Strip Loadin FG R f GH T R f B LOVER N O SK f N SOO llllll 7 +! ; 25 i ll c ei Oi y ;L_ DULECTOll UPPER STERNUM V f TH S K f N llll 2 FG 4 il Cl Q SD l lllll SllD pmrtl FG LOVER STERNUM NO SK f N 5 (; ( _A;! r SOO FG 6 j i _ DtfLECllON C M RA ( OPAT) HYBRD l _l +l L_L_ y; ornrc ; ;; ll = t t LOVER ST ERNUM V [ TH S K f N _ D ll UPPER ST ERNUM NO S K f N flg 3 7 ; llll ) FG l d S DffL(tllOll is Sllll SOO llll R f GHT R f B LOVER V f TH SK N Sllll ornro HYBRD HUMAN VOLUNTEER 34

3 numan vounteer The Hybrid appeared too stiff in the riht 7th rib and loer sternum locations The MRA ave a more humanlike response in the loer rib area than both the Hybrid and Hybrid hoever it appeared too stiff in the sternal area The use of soft tissue cover has an effect on both the belt path and on the structures response to loadin n the loer rib and upper sternum location the Hybrid and MRA responses ere markedly inf luenced by the soft tissue n the loer sternum reion the soft tissue cover appears to play a prominent role in the Hybrid response increasin its externally measured deflection by 2% The externally measured ehest deflections are enerally hiher i th soft tissue cover althouh this is not alays the rule The MRA exhibits the opposite result in the upper sternum location MODFED CHFST DESGN Upon detailed examination of the Hybrid ehest responses to be lt loadin it as determined that an improvement could be made in the responses of the loer rib and clavicle reions t as further determined that the anthropometric measurements ere not 5th percentile () in terms of ehest depth ehest lenth and ehest breadth The thoracic assembly had also demonstrated to be sensitive to temperature (8) Hence a ne ehest structure as developed to more accurately simulate a 5th percentile human male Structural Desin Overall thoracic structure measurements ere based on anthropometric data obtained from a sample of 5th percentile male subj ects ( ) The Modified Hybrid (or Hybrid E) ehest has redesined rib elements to improve both overall anthropometry and thoracic response to strip loadin This ne desin is composed of ten rib elements hich are hined at the spine and constrained at the front by a spli t sternum hich resembles an inverted Y (Fiure 7) To telescopic clavicles ere also interated into the structure These are ball jointed at the outer extremities of the shoulder assemblies and on the top of the sternum assembly Soft tissue coverin as not used since Typical polymer foam materials serve larely to make contemporary dummies lock more realistic The present desin endeavours to provide suitable force deformation characteristics ithout the use of such external coverins nside the thoracic cavi ty to adj ustable linear decelerators ere added to rovide investiators ith the option of controllin ehest dampin characteristics Both dashpots are bilaterally supported by the spine box and are linked to the midsternal reion by means of a bellcrank lever arm mechanism FGURE 7 MODFED THORAX CONSTRUCTON Determination of Mechanical Characteristics usin Finite Element Mode llin The STARDYNE finite element modellin proram as used to determine the properties of the ehest The ehest structure as simulated usin a 52 node 47 beam data set Ten ribs ere modelled They ere hined at the spine hich as assumed to be restricted from motion n addition the top rib as 35

4 restricted trom lateral movement tnereby simulatin the presence of clavicles The first desin phase involved determinin the required sternum and rib properties The model as used to determine the size of the rib elements in order to achieve correct stiffness under applied static loads at various sites Member stresses ere restricted to alloable levels under a 75 cm deflection at each rib The thickness of certain ribs had to be doubled in order to meet th is cri terion n the second desin phase the dynamic response of the structure to lo and hih speed blunt sternal impacts as specified by the Hybrid ehest calibration procedures as determined The to linear variable hydraulic dashpots ere then incorporated into the model and the ir performance characteristics ere determined The DYNRE option of the STARDYNE proram as then utilized to model to dynamic events 427 ms and 667 ms blunt sternal impacts ith a 2336 k impactor The forcetime functions measured durin previous calibration tests on a Hybrid ehest ere used as input load approximations Forcedeflection corridors (3) for the hih and lo speed events ere used as taret curves The dynamic response as then adj usted by varyin the theoretical damper resistance to fall ithin the desired response corridors To validate the response the differential equations of motion of an approximate system ere solved Validation of Structure The ne thoracic structure as evaluated for def lection response to subinjurious strip loadin usin the tes t method previously described () No soft tissue cover as employed The rib structure as first tested i th dashpots adjusted to produce no dampin The response in the loer rib and clavicle reions ere demonstrated to be a marked improvement over the Hybrid as shon in Fiure 8 z lo 3 RGHT 7th RB MO STERNUM LEFT CLAV CLE z lo z lo 3 f 3 _ ; ; _ { 2 u 2 u 2 lll >< <( o 2 MODFED Dm cm Dm c m HYBRD HUMAN VOL UNTEER HYBRD = RELAXED (982) D m cm STATE FGURE 8 THORACC RESPONSE lo BELT LOAD Validation of Dashpot Settins Upon completion of a preliminary evaluation proram desined to determine the effect of dampin on ehest response a specific dashpot settin as chosen for hih speed impact impacts The thorax as then subj ected to the standard Hybrid ehest certification procedure Since the structure as not covered i th a soft tissue eo ve r in the force time response of the ches t exhibi ted some deree of hih frequency noise durin the hih speed impact Deflection response in the lo speed test as ithin specifications Hoever the hih speed impact deflection response as 3% belo specification The 36

5 ehest foree responses ere also 25 and 33% lo for the lo and hih veloeity impaets respeetively One should note that ehest forees are measured at 9 ms from time of probe eontaet on the ehest strueture This method of evaluation may not be totally appropriate sinee the response is in faet eomparable to some of the eadavers upon hieh the Hybrid speeifieations are based SLED TEST PROGRAM n 982 Ka llieris et al (6) reported on a series of sled tests usin a pulse similar to the pulse from a Volvo P 4 impaetin a fixed barrier at 5 kmh in an a ttempt to link eadaver and aeeident da ta n 983 Saul ( 7) expanded upon the ork by obtainin response eomparisons for the Part 572 Hybrid and APR dummies in the same three point belt restraint system environment n an attempt to ve rify the response of the ne modified Hybrid and MRA dummies the same test proeedures as employed by Saul ere adopted in this proram Experimental Results The averae peak aeeeleration of the half sine pulse in the present test series as 3 % hiher than in Sauls test pro9ram and the delta v as 9% loer The res traint system data is summarized in Table and the ATD ins trumentation measurements are summarized in Table 2 Observations are summarized as follos The mean values of peak shoulder belt load for the Hybrid E dummy are hiher than either the Part 572 or MRA The Part 572 in the eurrent test proram produeed head aeeelerations 534% loer than those reeorded in Sauls test series n this test series the Hybrid E peak resultant head aceeleration as approximately 9 s hiher than the Part 572 The Hybrid E head aeeeleration as comparable to Sauls results for Part 572 and APR dummies The peak resultant ehest aeeeleration for the Part 572 in the eurrent series is similar to that obtained in Sauls proram (387 s vs 38) The Hybrid E ehest aeceleration as only 23 s hiher than the indo containin results of the APR Hybrid and Hybrid in Sauls proram Peak thoraeie defleetions for Hybrid E are 9% loer than for the Hybrid as measured in Sauls proram The Part 572 is much stiffer ith 5 in deflection This measured deflection is substantially hiher than results obtained by Saul (5 in vs 28 in) The deflectiontime histories for both the Part 572 and the Hybrid E exhibit substantial differences in terms of event duration The deflection event duration as 5 ms for the Hybrid E as opposed to ms for the Part 572 The shape of the curve response as also markedly different as seen in Fiure 9 +HY3 E VCHE_STDEFL3=z; P TS72 CH E_STDE FL UN ;+ Fiure 9 4 ooe o2e o? TM E < S EC Part 572 and Hybrid E Deflection Response Peak resultant pelvic accelerations are ithin the indo of values obtained in Saul s proram Thouh to clavicle failures ere recorded on the initial tests usin the Hybrid E all measurements ere ithin a eoefficient of variation of 49% 37

6 TABLE RESTRANT SYSTEM MEASUREMENTS TEST SERES TEST# PEAK SODER PEAK AP coment BELT LOAD BELT LOAD N (lb) N (lb) B)!>RD Part ( 785) 673 (53) N> clothiß ( 73) 742 ( 583) 4 ap Eelt soore ( 837) 6695 ( 55) +4 ap Eelt Score Part 572 tean 76 ( 7 ) 6734 (54) Fran Reference 7 (Saul) Stil Dev 73 (39) 25 (28) MRA (648) 557 ( 238) N> clothin used ( 729) 6356 ( 429) Clothin used ( 796) 6276 ( 4) Clothin used Mean 7669 ( 724) 645 ( 359) Stil Dev 33 (74) ( 54) All tests 469 Mean 7842 ( 763) 636 ( 42) lests usin clothin Stil Dev 2 (474) 57 ( 2 7) on AD Hybrid E ( 97) 7744 ( 74 ) eft Clavicle Brcke (at shoulder) (868) 7855 ( 766) Riht Clavicle Brcke (at sooulder) (879) 86 (82) Mean 8398 ( 888) 7893 (77) Stil Dev 43 ( 258) 37 (37) Hybrid Mean 8 ( 84 ) 895 ( 82) Fran reference 7 (Saul) Stil Dev 37 (69) 32 (72) TABLE 2 ATD NSTRUMENTATON MEASUREMENTS TEST SERES TEST # PEAK UPPER SPNE PE7K OEST PE7K RESULT RESULTAN PEAK RE3ULT DEFLECCN PELVC NXN HEAD NXN <lest NrN an (in) ( s) (s) (s) PAR (4) ( ) (9) 586 Part 572 tean o 7 ( 28) (Saul) Stil Dev (3) MRA 456 NA 526 NA NA NA NA NA NA Mean NA 56 NA NA Stil Dev Hybrid E ( 5) ( 4) ( 4) 568 Mean (43) 55 Stil Dev (OCXi) 5 Hybrid tean ( 6) 576 (Saul) Stil Dev (2) 9 38

7 Overall the ATD thoracic responses ere quite similar As pointed out by Saul (7) these response similarities occur despite very different dummy cons tructions Photoraphie analysis in the current study revealed that the thoracic spines underent a displacement of approximately 3 cm Since ehest compression accounts for only 8 3% of the ATD excursion This supports Sauls conclusion that the restraint system is a primary factor in controllin the dummy response Comparison to Previous Work Avai lable data from Kallieris (6) and Saul (7) are included in Table 3 The findins of the current study for the Part 572 are in close areement ith Ka llieris results in terms of head and thorax displacements and lap belt loads The Part 572 resultant thorax acceleration variance as only 44% amon all three investiators TABLE 3 ATD DATA COMPARSON TO KALLERS AND SAUL RESULTS HFAD lhrax WlKlARD lllrax SODER AD DSPUCEMEN DSPLACEMEN UP BELT RESULTAN BELT WAD an (in) an (in) N (lb) s N (lb) 4 arliver Kallieris 56 (22) ( 6) 5293 (9) 4 NA Part 572 Kallieris 48 ( 9) 25 () 69 7 (555) NA Part 572 Saul 58 (23) 43 ( 7) 6734 (54) ( 7) Hybrid Saul (82) (84 ) Part 572 present 49 (93) 234 (92) 6824 ( 534) ( 775) MRA present 526 (27) (359) (763) Hybrid E present 528 (82) 28 (82) 7873 (77) (888) (235) (59) (9) 37 lt is interestin to note that the MRA and Hybrid E dummies exhibited hiher head displacements than the Part 572 but exhibited substantially different lap belt loads and peak thoracic resultant accelerations Also based on the findins in this study the effect of clothin on ATD kinematics are neliible in terms of head excursion but accounts for approximately a % decrease in thorax displacement The lap belt load is also approximately 5% hiher and the shoulder belt load is approximately 7% hiher The Part 572 and Hybrid E responses are qui te similar to the results obtained by Sauls (7) ATD samples excet for the pelvic responses The peak pelvic accelerations ere typically hiher in the present study The MRA sternal responses ere different from the other ATDs thouh no explanation is proposed The MRA upper spine response thouh repeatable as not as smooth as the Part 572 and Hybrid E Comparison of Overall Responses of Hybrid E to Hybrid Data obtained in this proram for the Hybrid E as compared to the responses of the GM Hybrid as measured by Saul (7) Mean shoulder belt and lap be lt responses varied by a maximum 24% from the Hybrid to the Hybrid E the latter havin the hiher values Mean peak resultant head acceleration as 25% hiher (745 s) on the Hybrid but peak resultant ehest acceleration as % loer (4 s) in comparison to the modified Hybrid Peak midsternal deflection as 9% loer for the Hybrid E Mean peak resultant pelvic acceleration ere 4% hiher for the Hybrid dummy Maximum head displacement as 2% hiher for the Hybrid compared to the Hybrid E Chest displacement as 5% loer for the Hybrid E lt is interestin to note that Sauls thoracic displacement for the Part 572 as also approximately 5% hiher than for the Part 572 dummy tested in the present proram The reasons for these latter differences are not apparent at the time of ritin Comparison to Cadaver Data A detai led comparison as made ith a data base usin 3 cadaver te sts as chosen by Saul (7) based on the Heidelber cadaver data described by Kallieris (6) From these tests ATD comparisons ere made for belt loads sternal (upper and loer) spinal (upper and loer) and pelvic accelerations Typical comparisons appear in Fiures to 3 39

8 N po!) P) t [ U l t t f RN ;_ j _i _L f E R U M l A ; V f! i PT572 PR ; < Q; b!o oci <!; U u o u U ie ) c T! Mf > o 6 Part o i? n o ;) 6 < d i + j u ; j ca q _ ) MRA r ; } VA r f i = ; ;= ra _ r r o E RNc UP < t f ; 5 ; ; ;r n j _ ;!! M; Hybrid E r A OA V R P A V U?P S l EkUM n C A O V R up S! E R N U M U (Ar c 4 > Ol UPR S l ER N U M vpr S T ERNUM U TT ;_ i ; ; Q; i )i V CAOAVR HTlf u Fi Cadaver and Dununy Upper Sternum Acceleration Responses ooa i c r r ; ro S T E R N U M A CC _ o 2 c Cadaver r o v R PR S T E R N U M U < A UA V R vp S T ERNUM + a oo oj s ;; e= il r [ ;!! j l l 4 V R

9 N CADHR _ l 2 O<? Mf tsfc > > u > o ; t y ; i! L 4 ACr AC l vpr p NE Al vpr 5 P N E v P R S P NE oj F = ( ( L l l l P R SP o N t AC( d M RA L L + o o o oo n o l o o Hybrid E o M RA o V UPR SP R A C C C A O A V R o v P R S P N E A C C U C A O H R c Q oc n l o j = i i 7 )l l A O V upf >c o; L LLL V UPR ; P N [ A C n O H R vpr P j N [ R f S U Y 3f ; Fi Cadaver and Dununy Upper Spine Acceleration Responses!E c OP r RACC A C L i c AllA V R c CAOAVR u C A O A V R r=rrr rrrr b Part o oo ; z c v P F S P j N t r]r==== r Cadaver P S l 2 vpr SP NE l c a ; c t == < _ o % ;; r S P N E ACC

10 N N «T5!< O A V R! o i <> o >l llr i P N E ;? NE i L r SP l f Nf b Part lj _ N Hybrid ) 4 r } d M RA ) > o u < o < _ o c ) ); RA l llr S P R r A DA V R o L llr SP NE i ooo ; o < i <> o T c ACC AC C E (! S i [ AC c LllP! i < O P [ r uh R f < f ooe H u O Of V L llh U E L n ( A O V R L W R CP N E R E S u C A DA V R Y f Fi 2 Cadaver and Dummy Loer Spine Acceleration Responses r > A C L llr! n N [ A C R A C C ACC u A OA V R ; u r A C ( L llr L llr TlrTlrrr ) (! } ; c < o ; ] z Cadaver )4 T $ P N[ r R l a l ic c j < C < r ; o z r ( [H< o A lla V R o O A V

11 N c Ci r b < o n Part = r r;r C oo r f! L ; t o l t c LD i L r > H L D R fl >L Ol ] }] P L l O t ME ö s r Q O d M RA l r j < f> c D c ( } ; + r ; > O Ot m N ; o D ; L tse ooe T ME L f r! ; oc(!ad R SHLDP 9 L T L D Hybrid E O< Lll!lc f L D u Rt l L l l T R A f A O A V P! l c u SHP fi i o _ < = V r HTf ; CAUA fl Fi 3 Cadaver and Dummy Shoulder Belt Responses fc o o F rao A V R H < ( i ; HlfR B L! L O t c B ) ;>!LDR B L T L D Ü ; r== u o D n ; D A VR A D A V R rir ; 9L L Ü a Cadave r Pl57? j ] HLDP o ODAVR _ i i < 5 i! D R r

12 SUMMARY AND CONCLUSONS Finite element modellin has proven to be a very useful tool in the qesin of the ne ATD thorax The nely desined ATD thorax comprises ten rib elements each of hich are pivoted at the spine box junction nternally adj ustable linear decelerators provide the necessary viscous properties Statically and quasistatically the desin has been shon to faithfully emulate the relaxed human thorax structure response to strip loadin Dynamically the response can be made more or less humanlike by adj ustment of the viscous elements n a realistic 3point belt system impact environment the primary factor in controllin the dummy response the restraint system is ACKNOWLEDGEMENlS The authors ish to thank Mr Roer Saul of VRTC of Colombus Ohio for his assistance and uidance in the preparation of the sled test proram We ould also like to thank Dr Tim Boden and Mr Don Day at the HyGe Sled Facility at DCEM in Toronto Canada for their constant support and suestions We ish to acknolede the Road Safety and Motor Vehicle Reulation Directorate of Transport Canada for supportin this proram Fina l ly the authors ish to thank Helen OHara and Maro Williams for their patience in preparin this manuscript REFERENCES LAbbe R Dainty DA and A Neman An Experimenta l Analysis of Thoracic De f lection Response to Bel t Loadin Seven th nternational RCOB Conference September Neathery RF Analysis of Chest mpact Response Data and Scaled Performance Recommenda tions Paper Eihteenth Stapp Car Crash Conference Ann Arbor Michian December Neathery RF Kroell CK and H Mertz Prediction of Thoracic nj ury from Dummy Response Paper 75 5 Nineteenth Stapp Car Crash Conference San Dieo California November Schmidt G Ka llieris D Barz Mattern R and Klaiber Neck and Thorax To lerance Leve ls of BeltProtected Occupants in HeadOn Collisions paper Nineteenth Stapp Car Crash Conference San Dieo California November Foster K Korte A and M Wolanin Hybrid A BiomechanicallyBased Crash Test Dummy Paper TentyFirst Stapp Car Crash Conference Ne Orleans Louisiana October Kallieris D Mellander H Schmidt G Barz and R Mattern Comparison Beteen Frontal mpact Tests i th Cadavers and Dummies in a Simulated True Car Restrained Environment Paper 82 7 TentySixth Stapp Car Crash Conference October Saul RA Sullivan LK Marcus H and R M Moran Comparison of Current Anthropomorphic Te st Devices in a ThreePoint Belt Restraint System Paper TentySeventh Stapp Car Crash Conference San Dieo California October Saul R Frontal mpact Component Test Eva luation of Current Anthropomorphic Test Device Technoloy Ninth ESV Conference Proceedins November