ABSTRACT

REV I EW OF HEAD INJURY TOLERANCE TO DI RECT IMPACT

By

Verne L . Roberts and James H . McEl haney H ighway Safety Research Insti tute

The Un ivers i ty of M i c h i gan Ann Arbo r , Mi chi gan U . S .A . 481 05

The human tol erance l i terature perta i n i ng to di rect head impact i s re­v i ewed and d i s cussed . The bas i s for the V i enna Index , the Effecti ve Di s­pl acement Index , the Severi ty I ndex , the Head Injury Cri teri a of USMVSS #208 , the Wayne State Head Inj ury Tol erance Curve and the Maxi mum Stra i n Cri ter ia wi l l be exami ned and the correl ation of these various i ndi ces wi th the avai l ­abl e data demonstrated . Add i t i onal l y , the predi ction of i njury for various i mpact pu l ses when used wi th the vari ety of avai l ab l e cri teri a wi l l be determi ned .

Where appropri ate , new data wi l l be i ntroduced to compl ete the exi sti ng data set . I t is fel t that on the bas i s of the currently avai l ab l e i n form­ation regardi ng the l i mi ts of head i mpact tol erance that tol erance can be establ i s hed for i mpacts of arbi trary di rect i o n .

INTRODUCTION

Wi th the advent of h igh s peed a i r and l and transportation , engi neers have become i ncreas i ng ly aware of the mechan i cal frangi bi l i ty of the human body . Thus , we have seen the evol ution of vari ous i sol ati ng and l oad di stri­buting devi ces rang i ng from seat bel ts and padded sun vi sors , to ejecti on seats , crash helmets , and accel eration couches . Whi l e there i s a l arge amount of i n fo rmation avai l abl e regard i ng the response of i nanimate sys tems to vi bration and i mpact , there i s a comparabl e dearth of knowl edge pertai n i ng to the mechani cal responses of b io l ogi cal sys tems . Therefore , the des ign of much support i ng and protective equi pment i s often based on i ntui ti on because of the l ac k of i n formation avai l abl e a bout the mechan i ca l behavior of the human body . I n addi t i o n , such knowl edge woul d be hel pful i n the treatment of i njury by serv i ng to i denti fy the mechan i sm of trauma . Thus , both a rational des i gn procedure for impact protection and a rational therapy for treatment of trauma cannot be devel oped unt i l a quanti tati ve descri pt i on of the mechan i ca l responses of the human body i s obta i ned .

I n order to properly des i gn devi ces ai med at min imi z i ng head i njury i n the automoti ve crash env i ronment , engi neers requ i re a means of pred icti ng potent ia l i njury o r a so-ca l l ed . Head Injury Cri teri a . Thi s cri ter ia mi ght be used i n rea l acci dent reconstructions , car crash and s l ed test experi ­ments o r mathemati cal s i mul ati ons .

The automoti ve crash envi ronment encompasses a wide range of impu l s e durations and d i recti ons . Thus , a v iab l e head i njury cri teri a mus t provi de appropriate mecbani sms that real i s t i cal l y a ccount for the frequently observed , but poorly documented , rel at i ons of head i mpact tol erance and impul se duration

2 4 1

and d i recti o n . In addi t i on , two di sti nct types o f l oad i ng are observed .

1 . An i mpact o r b l ow i nvol v i ng a col l i s i on of the head w i th another so l i d object at an apprec i a b l e vel oci ty . Thi s s i tuation i s genera l l y char­acteri zed by l a rge l i near accel erati ons and smal l angu lar accel erati ons dur-i ng the i mpact phase .

·

2 . An i mpu l s i ve l oading i nc l udi ng a sudden head mot i on wi thout d i rect contact . The l oad i s genera l l y transmi tted through the head-neck junction upon s udden changes i n the motion of "the torso and i s associ ated wi th l a rge angu l a r accel erati ons of the head .

I t has been shown ( 1 ) that i n the moderate but survi vab l e automotive crash envi ronment (30 mph barri er equ i val ent) no s i gni fi cant head i njuri es occur when the ful l y-bel ted occupant ri des the crash down wi thout head-to -ve­h i c l e contac t . Thi s does not mean that the Type Two l oadi ng descri bed above can not produce i njury , but only that the l evel s of l i near and angul ar accl er­ation requi red to produce head i njury wi thout contact do not occur i n moderate crashes . If , however, vehi c l es are sti ffened to provi de l ess compartment i n­trus ion at h i gher vel oci t ies , i njuries of the type descri bed by Ommaya and H i rsch ( 2 ) mi ght become commonpl ace .

Thus a rati onal head i njury cri teri a for current automotive des i gn may be concentrated on the first type of l oadi ng . There i s , of course , a pos­s i b l e defect in th i s rationa l e that concerns the pos s i b l e i ncreased potentia l for head i nj ury i nvol ved wi th a combi nation of the fi rst and second type of head l oadi n g . An attempt to reconci l e th i s probl em has l ead to the devel op­ment of the Mean Stra i n Cri ter i on . Th is head i njury cri teri a consi ders the total l i near accel eration h i s to ry of the head but ass umes a s i ng l e i nj ury mechan i s m .

MEAN STRAIN CRITERION (MSC )

The dynami c s tructural characteri s ti cs of monkey s kul l and brai n were determined over a wide frequency range by Sta l naker and McEl haney i n 1 971 (3 ) . These resul ts , reported as the change of mechani cal impedance ( force/vel oci ty ) wi th frequency , a l l owed the conceptual characteri zati on o f the head a s two masses coupl ed by a spring and das hpot . . The mathemati cal l y predi cted dynami c response of the model agreed wel l wi th the experimental data .

Experimental impacts del i vered to the heads of vari ous s i ze primates s howed that the dynami c model postul ated an the bas i s of vi bration studies accurately pred icted head-impact i njuri e s . I t was further fo und that for head i mpacts of a known magn i tude , the resul t ing i njuries cou l d be grouped by compar ing the mean strai n as predi cted by the theoreti cal model wi th i nj u ry l evel s , (where mean stra i n i s defi ned as the di spl acement of one s i de of the head rel ative to the other , d i v i ded by tbe d i s tance across the cran i um ) . T h i s experimenta l l y deri ved head i nj ury data for l i vi ng primates formed the bas i s for a Mean Stra i n Cri terion (MSC ) for head i nj ury to humans (4 ) . U s i ng the val ue of predi cted stra i n i n the Rhesus monkey head as a cri terion of i njury , a tol erance curve was deri ved wi th rel ated average accel erati on and time for a constant l evel of mean s trai n .

The deri ved tol erance curve for the subhuman primate was val i dated by pl otti ng the experimental data poi nts necessary to produce mi nor , but i denti fiabl e , bra i n i nj ury i n the l i vi ng test subject for a wide range of pul se durati ons ( F i gure 1 ) .

l t was determined that the heads of several s pecies of subhuman primates , squi rrel monkeys , Rhesus monkeys , the chimpanzee and the fresh human cadaver had mechan i c a l impedance characteri sti cs over a broad frequency band ( 5 to 5000 Hertz ) whi ch were s i mi l ar i n s hape but varied i n the mechan i cal char­acter i s ti cs of mas s , s t i ffnes s and damp i n g . U s i ng the maxi mum pred i cted s tra i n as the bas i s for i njury , experimental head i mpacts i n the l aboratory documented the val i di ty of the theory and formed the bas i s for establ i s h i ng mode l i ng rel at i onsh ips through wh i c h extrapol ation of the MSC to other s i ze heads may be made . Thi s approach was tested by compari ng the MSC to human vol unteer and fresh human cadaver head i mpacts ( F i gure 2 ) . The mean tol er­abl e head stra i n for humans (0 . 006 i n . / i n . ) was cal cul ated from the mathemati­cal model and the scal i ng techni que referenced above .

A s tudy was undertaken to document the val i d i ty o f the MSC for arb i tr­ari l y di rected head impacts . Thi rty careful l y sel ected Rhesus monkeys were i mpacted at i ncreasi n g l evel s i n vari ous d i rections ( front , s i d e , back ,top and mid-front ) . The i mpact l evel was i ncreased unti l autopsy studies i ndi cated that an Estimated Severity of Injury ( E S I ) of a moderate but revers i b l e type ( C l a s s 3 ) was obtai ned . Only cl osed bra i n i njuri es were consi dered .

The mechani cal i mpedance was determi ned for fi ve monkeys of approximatel y the same wei ght as the ones used i n the impact study . These impedance curves were obtai ned for the top , s i d e , rear , and front of the head ( F i gure 3 ) .

The val ues of tol erabl e accel erat i on and impedance data were then i nput to the MSC model . The pred icted mean stra i n val ues for each d i rection varied l es s than 7 . 5% . Thi s i ndi cates that , wh i l e w i dely vary i ng accel erati ons are requ i red to produce an i nj ury l evel of 3 i n the Rhesus monkey , the correspon­d i ng s tra i ns are approxi mately equal . The res ul ts of th i s s tudy are g i ven i n the form of an accel erat i on surface for a constant s tra i n l evel of 0 . 032 i n . / i n . for Rhes us monkeys s ubjected to r i g i d s tri ker i mpacts ( Fi gure 4) .

Prel imi nary studies on the fresh i ntact cadaver i ndi cate that a tol er­abl e mean s trai n l evel of 0 . 0061 as pred icted by the MSC model may be used for a rbi trary impact d i recti ons wi th model constants appropri ate for that d i rection . However , a suffi ci ent number of i mpact tests and dri v i ng poi nt i mpedance meas urements have not yet been made to veri fy a d i rect extrapol ation o f the Rhes u s monkey data . Fi gure 5 s hows pred i ct i ons and measured val ues of the MSC for human head impact i n the sagi ttal pl ane .

COMPARISON O F HEAD INJURY CRITERIA

The preced i ng di scuss i on i nvol ved the mos t recent work on the Mean Stra i n Cri terion for head impact. The fol l owing section conta i n s the res ul ts o f a series o f analys i s a i med at comparing various head i njury criter i a .

Val ues for various head i nj ury i ndi ces were computed for two crash s i mul a t i ons .

1 . Dummy resu l tant head accel erations i n frontal automotive crash en­vi ronments were used . The dummy was unbel ted , and s truck the wi nds h i e l d a cl ean b l ow . The windsh i e l d was not penetrated except a smal l tear i n the l ami nate was a l l owed ( C l as s A) , or several smal l tears ( C l ass B ) .

2 . Resul tant head accel erati ons from a recent series of h i gh s peed human vol unteer and dummy tests at Hol l oman Ai r Force Base us i ng a i rbag restra i nts .

Wayne State Tol erance Curve (WSTC) The Wayne State Tol erance Curve was i ntroduced by L i s s ner i n 1 960 ( 5 ) .

Ori g i nal l y thi s curve was devel oped from data obtai ned by droppi ng embalmed cadaver heads on to unyi el d i ng fl at su rfaces . L i nea r s ku l l fracture was used as the cri teri on of i njury . I n 1 962 Gurdj i a n publ i s hed the Wayne State Tol erance Curve as i t appears today ( 6 ) . Th is curve was devel oped by com­b i n i ng a wi de vari ety of pul se shapes , animal types and i njury mechani sms . The fa i l ure cri ter i a used was genera l l y s ku l l fracture and/or concus s i on , except for l ong pu l se duration from human vol unteers wi th no di scern i b l e i n­j ury. In 1 965 Patri ck et al ( 7 ) proposed that tbe ori g i na l hori zontal asymp­tote of 45 G ' s be ra i sed to 80 G ' s to adjust for addi tional data from tests aga inst y ie ld ing s u rfaces . Si nce that t ime numerous papers have been wri tten prov id i ng s upporting data for the tol erance curve (8 ,9 , 1 0 , 1 1 , 1 2 , 1 3 ) .

The i nj ury assessment i s based on the average accel erati on and pu l se dur­ation . A g i ven average accel eration at a parti cul ar pul se durati on wh i ch l i es bel ow the WSTC i s cons i dered to . cause at mos t cerebral concus s i on wi thout permanent after-effects , wh i l e any poi nt whi ch l i es above the curve i s con­s i dered to be dangerous to l i fe . For s i ngl e head i mpacts i nto a ri g i d fl at su rface the average accel erati on and pul se du rati on is qui te easy to determi ne , but for s l ed test i n g where mul t i p l e i mpacts are q u i te common the 11effecti ve1 1 pul se , that i s , the part of the pul se upon wh ich the average i s based , i s not wel l defi ned . I n spi te of the many i nterpretive d i ffi cul ties associ ated wi th th i s curve , i t has been the pri nc ipal source for head i njury tol erance i nform­ation used by the automoti ve safety commun i ty .

Gadd Severi ty I ndex (GSI) Because of various i nterpretive d i ffi cul ties associated w i th the use of

the WST curve , C . W . Gadd i ntroduced the Gadd Severity I ndex as a general ­i zati on of the Wayne curve ( 1 4 , 1 5 , 1 6 ) . More recen tly the GSI has been ex­tended for l ong pul se duration by means of the Ei band tol erance data and other primate s l ed runs . The severity i ndex equati on has the fo l l owi ng form :

where a = head accel erati on response function n = we i ght factor , general 2 . 5 ' = pul se duration t = i ntegral parameter of t ime

( 1 )

The head i njury threshol d sever i ty i ndex number was determi ned from compari son wi th the WST curve and the n umber 1 000 recornmended .

The assessment of i njury hazard i s obtai ned perfonn ing the above cal cul ­ation . I f th i s number (GS I ) i s greater than 1 000 , the accel eration pul se i s con s i dered to be dangerous to l i fe . I f this i ndex i s l es s than 1 000 , the accel erat i on pul se i s then cons i dered not to be l i fe threaten i ng .

Head I njury Cri teri on ( H I C )

The Head Inj ury Cri teri on was first proposed by J . Versace ( 1 7 ) and then modi fied by NHTSA ( 1 8 ) . Thi s cri teri o n i s based on a new i nterpretation of the Gadd Sever i ty I ndex .

Versace poi nted out that because the WST curve was pl otted for average a ccel erati on , any compari son to the WST curve shoul d be made u s i ng the average accel eration of the pul se of i nterest .

The question of l ong pu l se head accel erations has posed some probl ems when us i ng the S . I . to predict head i njuries . I n order to provi de a better compari so n wi th human vol unteer tes ts , a head cri teria has been proposed as

where

( 2 )

t 1 = a n arbi trary time i n the pul se t2 = for a g i ve t 1 , a time in the pul se wh ich maximi zes the H IC a = resu l tant accel eration at the head center of g rav i ty

I f thi s i ndex i s l ess than 1 000 , the s i tuation i s consi dered not to be l i fe-threaten i ng .

V i enna I nsti tute Index (JTI)

The V i enna I nsti tute I ndex was i ntroduced by A . S l attenschek ( 1 9 ) and i s based on a s i ng l e degree-of-freedom vi bration model .

Wi th the damp i ng assumed to be cri ti cal ß = 1 , and two triang l e accel er­ation pul ses determi ned from the WST curve , the natural angu l ar frequency va l ue of w = 635 rad/sec and a maxi mum tol erance d i spl acement xT = 0 . 092 ( 2 . 35mm) i nches was obta i ned from the fol l owi ng equation : ow

where X

.

x , x w ß

y ( t )

=

= = = =

x + 2ß • + w2x = y ( t ) wx

rel ative di spl acement of bra i n mass to s kul l rel ati ve vel o c i ty and rel at ive accel eration natural angul ar frequency o f Vi brat i on 635 rad/sec v i s cous damp i ng coeffi c i en t of 1 .0 accel eration pul se meas ured at the head

( 3 )

2 4 5

The maxi mum devi ation between the model and the WST curve i s -4 percent . To access an impact , the ampl i tude x correspondi ng to the accel eration pul se to be anal yzed i s determi ned f�B� the model and compared to the tol er­abl e ampl i tude xTo i r = 0 . 092 i nches . A J tol erance i ndex i s then defi ned by :

xmax J = --XTo l r

( 4 )

where : xmax xTol r

= maxi mum x generated by the model for a g i ven accel eration pul se = tol erabl e ampl i tude from the Wayne State Tol erance Curve 0 . 092

i nc hes ( 2 . 35mm)

According to S l attenschek , i mpacts wi th a J tol erance of J 2 = 1 j us t reach the threshol d of human tol erance ; val ues J < 1 at warst cause cerebral con­cus s ion wi thout permanent after-effects , wh i l e val ues J > 1 are consi dered to be h azardous to l i fe ( 20 ) .

Effect i ve D i spl acement I ndex ( ED !) The effecti ve di spl acement i ndex was i ntroduced by J . Bri nn ( 21 ) and

is s imi l ar to the Vi enna Insti tute model with c hanged damping and the angu l ar frequency . New angu l ar frequency and dampi ng val ues for the S l attenschek model were determi ned by match i ng the model to the WST curve . The empha s i s was p laced on matchi ng for short duration events ( 3-5msec pul se durat i on ) . The best fi t of the model to thi s porti on of the WST curve was found by us i ng the model parameters w 482 rad/sec and ß = 0 . 707 . These model parameters and the Sl attenschek model were exerci sed for poi nts on the WST curve and a di s ­pl acement va l ue o f 0 . 1 5 i nches was determined a s a 1 1Des i g n Bogi e 11 for human AP he�d i mpacts . Because of the unhuman l i ke res ponse of dummy heads , the Des i gn Bog i e i s rai sed to a dummy Test L imit of 0 . 1 7 i nches i n the AP d i rec­tion . When the res ul tant accel eration i s used , a Des i gn Bog i e of 0 . 1 8 i nches and a d ummy Test Limi t of 0 . 20 i nches i s recommended .

Revi s ed Bra i n Model ( RBM )

The rev i sed b ra i n model was i ntroduced by W . R . S . Fa n ( 22 ) and i s a mod i ­fi cat i on o f the V i enna Insti tute model . Li ke the JTI , the RBM i s a s i ngl e degree-of-freedom mas s-spr ing-das hpot model of the brai n . The v i s cous damp i ng coeffi c i ent for thi s model was esti mated from publ i s hed val ues of b ra i n mater­i a l properti es . Wi th an estimated damping coeffi c i ent of 0 . 4 data from the WSTC for l ong duration i nputs , a natural angul ar frequency of 1 75 rad/sec and the theoreti cal tol erabl e brai n deformation ( Sd ) of 1 . 2 5 i nches was estimated . A tol erabl e brai n vel oci ty Sv was then cal cu l ated from the WST curve for s hort pul se duration and was found to be 1 35 . 3 i n/ sec .

The recommended measure of brai n i njury potenti a l i s x<S for i mpact pul se durati ons l ess than 20 msec and x<S for pul se duration� greater than 20 msec , as cal cul ated from the d i fferent9a1 equation of the S l a ttenschek model wi th revi sed coeffi ci ents .

RESULTS OF COMPARISON

The res u l ts of the computati ons of the various head i nj ury cri teri a are presented i n Tabl e 2 . F i gure 6 shows a l l of the model s and the i r constants for purposes of compari son wh i l e Fi gure 7 g i ves normal i zed val ues of the head i njury i ndi ces obtai ned by d i v i d i n g the part i c u l ar computed i ndex by the appropri ate cut off val ue ( i . e . 1 500 for the SI and 1 000 for the H I C ) .

Based an acci dent stati sti cs , i t i s fel t that head i mpacts of thi s type i nto the HPR windsh i e l d sel dom i nvol ve seri ous head i nj ury and an appropri ate head i njury . cri teri a s hou l d so i nd i cate . Study of Tabl e 2 shows the GSI to be qui te cons ervati ve i n th is s i tuati on . The H IC i s l es s conservati ve , but sti l l i nd i cates four l i fe-threateni ng s i tuations in the wi ndshi el d tests . The RBM , EDI and MSC a l l pred ict essenti a l l y the same i njury l evel s for both seri es .

Mathemati ca l Mode l i ng

Whi l e the previous authors have d i rected the i r efforts primary toward the acqu i s i ti on of experimental data and i ts expl anat i on i n terms of ei ther l umped parameter model s or mathemati cal correl ates to publ i s hed head i nj ury data , others have concerned themsel ves wi th mathemati cal model s of the head wh ich defi ne i ts response to impact . Both approaches can yi e l d i n s i ght i nto head i nj ury tol erance wi th the former prov i d i ng near term tool s whi l e the l ater . because of the i r greater attenti on to model i ng the enti re· system hol d great hope for future eval uation of protecti ve devices wi thout the extens i ve l aboratory werk requi red at th i s time .

Anzel i us ( 2 3 ) model ed the head as a fl u i d fi l l ed r i g i d shel l brought s uddenly to res t whi l e Gutti nger (24) so l ved essenti a l l y the s ame probl em wi th the shel l ach iev i ng a vel oc i ty from res t . Further attempts to model the head duri ng impact l ay dormant unti l proposed by Gol dsmi th ( 2 5 ) i n h i s revi ew of the phys i ca l processes of head i njury . Eng i n ( 26 ) sol ved the case concern­i n g the asymmetri c response of a s pheri cal shel l fi l l ed wi th an i nvi sc id fl u i d subjected to a l ocal rad i a l impul s i ve l oad . Recently L i u et a l ( 27 ) extended Engi n ' s work for the axi symetri c so l ution for a Di rac-del ta time funct i on to the case for a fi n i te time funct i on . Chan (28) has devel oped the theory further i n hi s study of the asymmetri c res ponse of a fl u i d fi l l ed s hel l . Benedi c t et a l ( 29 ) sol ved a pro b l em s imi l ar to that of Engi n ' s al l ow­i ng for on ly membrane effects whi l e H ick l i ng and Wenner ( 30 ) model ed the head as a two l ayered v i s coel asti c sphere s ubjected to axi symmetri c i mpact . Kenner and Gol dsmi th ( 31 ) experimenta l l y i nvesti gated the probl em so l ved by the various analyti cal stud ies .

Another series of i nvesti gators h ave model ed the response of the head to impact for the case where the pu l se durati on i s l ang rel ative to the trans i t time of a press ure wave through the contai ner . Grass (32 ) i n 1 958 deve loped a gl ass , fl u i d fi l l ed model and compared the resu l ts of cavi tation i nduced i n the f l u i d wi th h i s analys i s . Unterharnschei dt and Sel l i er ( 33 ) a l so per­formed exten s i ve studies of th is nature uti l i zi n g analyti cal and experimental model s extend i n g the i r work to ani mal s subjects to confi rm thei r analys i s . Concurrently L i n dgren (34) devel oped an exten s i ve s tudy o f mechani cal i nputs to a model of the head when s truck from di fferent d i rections wi th vary i ng

2 4 7

boundary cond i ti ons . Kopecky and Ri pperger ( 35 ) extended the model i ng treat­ment to i nc l ude a deformabl e fl u i d fi l l ed cyl i nder and confi rmed the previ ous res u l ts ( 1 2 ) i n d i cati ng that the l ocation of the nodal po int i s a functi on of the conta i ner deformation .

D ISCUSS ION

The dri v i ng poi nt i mpedance studies i n d i cate that the mechani cal response of the pri mate head may be approximated as a two-mas s sys tem . Injury l evel s for b l unt i mpacts have been rel a ted to the compress i on of a spring i n the s impl e two-mass model . The MSC model i ndi cates decreas i ng tol erance w i th i m­pul se duration for pul ses of approxi mately 1 0 t imes the resonance period or l es s . For pul ses l anger than 1 0 t imes the resonance period , a quas i -s tati c response i s i ndi cated that i s unaffected by further i ncreases i n pul se dur­ati on . The predi ct i on i s qui te s i mi l ar to that produced by ED! , JTI and the RBM , but cons i derably d i fferent from the SI and H IC , wh i c h i ndi cate that a quas i - s tat i c response i s never obtai ned . Obvious l y , as wi th a l l s i mple model s of compl ex phenomena , extrapol ation of model pred i ctions beyond the range of val i dation or to new s i tuations i s dangerous and shoul d be done wi th cauti on . The MSC model has been devel oped for bl unt i mpacts where the amount of bone and s ca l p i n contact wi th the i mpactor approximates that l oaded by the coupl i ng cl amp during the i mpedance tests . When the l oads are appl i ed to l a rge sections of the head or through the neck , many of the arguments used i n the model devel opment do not apply . I n add i t i o n , i t i s probabl e that the i nj ury mechani sms change cons i derably wi th these di fferent types of l oadi ng , and a s i ng l e mecha n i sm mode l , as are a l l the ones d i s cussed i n thi s paper , woul d be i nadequate .

No treatment of th i s s ubject wou l d be compl ete w i thout acknowl edg i ng the numerous s tud ies which poi n t out the combi nation of l i near and angu l ar motion of the bra i n subs tance i n contri buti ng to the head i nj u ry proces s . However the exi s tence i n th i s sympo s i um of a separate paper on head i njury i n the absence of i mpact wi l l undoubted l y cover th i s materi a l i n i ts ful ly expanded form .

REFERENCES

1 ) McEl haney , J . ·H . , V . L . Roberts and J . W . Mel vi n , 1 1 B iomechani cs of Sea t Be 1 t Des i gn , 11 Pro c . l 6th Stapp Conf . , 1 972 .

2 ) Ommaya , A . K . and A . E . H i rsch , 1 1Protection o f the Bra i n From Injury Duri ng Impact : Experi mental Studies i n the B i omechani cs of Head Injury , 11 AGARD Conference Pre- Print No . 88 on L i near Accel eration ( Impact Type) , 1 969 .

3 ) S cal nake r , R . L . , J . L . Fog l e and J . H . McEl haney , 11Dr iv ing Poi nt Im­pedance Characteri s t i cs of the Head , 11 J . of B i omechan i cs , Vol . 4 , No . 2 , pp . 1 27- 1 39 , March 1 97 1 .

4 ) McEl haney , J . H . , R . L . Sta l naker , V . L . Roberts and R . G . Snyde r , 1 1 Door Crashworthi ness Cri teri a , 1 1 Proc . l 5th Stapp Conf . , Paper 71 0864 , p . 39 , New York : Soci ety of Automotive Engi neers , I nc . , 1 97 1 .

5 ) L i s sner , H . R . , M . Lebow and F . G . Evans , " Experimental Stud i es o n the Rel at i on Between Accel erati on and Intracran ia l Pressure Changes i n Man , 11 Surgery, Gynecol ogy and Obstetri cs , 1 1 1 : 329-338 , 1 960 .

6 ) Gurdj i an , E . S . , H . R . L i ss ner and L . M . Patri ck , 1 1 Protect i on of the Head and Neck i n Sports , 1 1 Journal of Ameri can Med i c a l As sociation 1 82 pp . 509-51 2 , November 1 96 2 .

7 ) Patr i ck , L . M . , H . R . L i ssner and E . S . Gurdj ian , 1 1Surv i va l by Des i gn -Head Protecti on , 11 Proc . 7th Stapp Conf . , Soci ety of Automot i ve Engi neers , I nc . , New York , pp . 483-499 , 1 96 5 .

8 ) Gurdj i a n , E . S . , V . R . Hodgson , L . M . Thomas and L . M . Patri c k , 1 1Si gni ­fi cance of Rel at ive Movements of Scal p , Skul l , and Intracran i al Contents Duri ng Impact Inj ury of the Head , 1 1 Journal of Neurosurgery, Val . 29 , P P . 70-72 , 1 968 .

9 ) Hayas h i , T . , 1 1Study o f Intracran i a l Press ure Caused by Head Impact , 11 ( 2nd Report ) , Journal of the Facul ty of Engi neeri ng, U n i vers i ty of Tokyo , Vol . XXX , No . 2 , 1 969 .

1 0 ) Hodgson , V . R . and L . M . Patrick , 1 1Dynami c ResponsP. o f the Human Ca da ver Head Compared to a S imp l e Mathemati cal Model , 11 Proc . 1 2th Stapp Conf . , 1 968 , Soci ety o f Automoti ve Engi neers , Inc . , New York, pp . 280-301 , 1968 .

1 1 ) Hodgson , V . R . and L . M . Thomas , 11Test ing the Val i d i ty and L i mi tati ons of the Severi ty I ndex , 1 1 Proc . 1 4th Stapp Conf . , Nov . 1 7-1 8 , 1 970 , Paper 70090 1 , Soci ety of Automotive Engineers , Inc . , New York , pp . 1 69-1 87 , 1 970 .

1 2 ) Roberts , V . L . , V . R . Hodgson and L . M . Thomas , 1 1Fl u i d Pres sure Gradi ents Caused by Impact to the Human Skul l , 1 1 Proceedi ngs of the Human Factors Conference , ASME , Paper No . 66-HUF- l , 1 966 .

1 3 ) Hodgson , V . R . and L . M . Thomas , 11Compari son of Head Accel eration Injury I ndi ces i n Cadaver Skul l Fracture , 11 Proc . 1 5th Stapp Conf . pp . 1 90-206 , 1 972 .

1 4 ) Gadd , C . W . , 1 1 Criteri a for Injury Potenti a l , 11 i n , Impact Accel erati on Stres s . A sympos i um hel d at Brooks A i r Force Base , Texas , 27-29 November 1 961 . National Academy of Sci ence/Nati onal Research Counc i l ( 1 962 ) . Pub 1 i cati on 977 .

1 5 ) Gadd , C . W . , 1 1Use of a Wei ghted- Impul se Cri terion for Estimating Injury Hazard , 11 Pr') c . l Oth Stapp. Conf„ Uni vers i ty of Minnesota , Mi nneapol i s , pp . 1 64 - 1 7 4 . Society of Automoti ve Engi neers , I nc . , New York , SAE Paper No . 660793 , 1 966 .

1 6 ) Gadd , C . W . and L . M . Patri ck , "System Versus Laboratory Impact Tests for Esti mati ng Injury Hazard , 1 1 Soci ety of Automoti ve Engi neers , I n c . , New York . SAE Paper , No . 680053 , 1 968 .

1 7 ) Vers ace , J . , 1 1 A Rev i ew of the Severi ty I ndex , 11 Ford Techn i c a l Report, No . S-71 -43 , November 1 2 , 1 97 1 .

2 4 9

1 8 ) Department of Transportat ion Nati onal Hi ghway Traffic Safety Admi n i s tr­ation (49 CFR Part 571 ) [Docket No . 69- 7 ; Noti ce 1 7 ] , 1 10ccupant Crash Protection Head Injury Criteri on . 11

1 9 ) Sl attenschek , A . and W . Tauffki rchen , 1 1Cri teri cal Eval uation of Asses s ­ment Methods for Head Impact Appl i ed i n Appra i sa l of Bra i n Injury Hazard , I n Part i cu l a r i n Head Impact on Wi ndsh ie l ds , 1 1 I nternat i onal Automobi l e Safety Conference Compend i um , 1 97 0 , Paper 700426 , Soci ety of Automotive Engi neers , I nc . , New York , pp . 280-301 , 1 968 .

20 ) S l attenschek , A . , W . Tauffki rchen and G . Bened i kter , 11The Quanti fi cation of Internal Head Inj ury by Means of the Phantom Head and the Impact Assessment Methods , 1 1 Pro c . 1 5th Stapp Conf . , pp . 742-766 , 1 972 .

2 1 ) Bri n n , J . and S . E . Staffel d , " Eval uation of Impact Test Accel erati ons : A Damage I ndex for the Head and Torso , 11 Proc . l 4th Stapp Conf . , November 1 7- 1 8 , 1 970 , Paper 700902 , Soci ety of Automotive Engi neers , I nc . , New York , pp . 1 88-202 , 1 970 .

22) Fan , W . R . S . , 1 1 I nternal Head Inj ury Assessment , 11 Proc . 1 5th Stapp Conf „ pp . 645-66 5 , 1 972 .

23 ) Anzel i us , A . , 11The Effect of an Impact on a Spheri cal L i q u i d Mas s , 11 Acta . Path . M i crobio l . Scand . Suppl . 48 , pp . 1 53-1 59 , 1 943 .

24) Gutti nger , W . , "Der Stosseffekt aufeine Fl uss i gkei ts kugel a l s Grundl age ei ner Phys i kal i schen Theori e der Entstehung von Gehi rnverl etzungen . Z . Naturf . AS , pp . 622-628 , 1 950 .

2 5 ) Gol dsmi th , W . , 1 1The Phys i ca l Processes Produc i ng Head Injury , 1 1 Proc . Head Injury Conf . , L 1 ppi ncott , pp. 350-382 , 1 966 .

26 ) Engi n , A . E . , 1 1The Axi symmetri c Response o f a Fl u i d-Fi l l ed Spheri cal Shel l to a Local Rad ia 1 Impul se-A Model for Head Injury , 11 J . of B io­mechan i cs , Vol . 2 ,No . 3 , pp . 325-341 , 1 969 .

27 ) L i u , Y . K . , H . S . Chan and J . A . Nel son , 1 1 I ntracrani al Pressur.e Wave Pro­pagati on i n Head Impact , 1 1 Proc . Summer Computer S i m . Conf . , pp . 984-994 , 1 97 1 .

28) Cha n , H . S . , 1 1The Asymmetri e Response of a Fl ui d-fi l l ed Spheri cal Shel l ­A Mathematica l S i mu l ation of Head Injury , 11 Ph . D . D i s sertati on , Tul ane U . , New Orl eans , La . , 1 97 1 .

29 ) Bened i c t , J . V . , Harri s , E . H . and 0 . H . Von Rosenberg , "An Analyt ica l Invest i gati on of the Cavi tation Hypothes i s of Bra i n Damage , 11 J . of Bas i c Engr . , Vo1 . 9 2 , No . 3 , pp . 597 -603 , 1 970 .

30) Hi ckl i n g , Robert and M . L . Wenner , 11Mathemati cal Model of a Head Subject­ed to an Axi symmetri c Impact , 11 J . of B i omechan i cs ,Vol . 6 ,No . 2 , pp . 1 1 5- 1 32 , 1 973 .

31 ) Kenner , V . H . and W . Gol dsmi th , " Impact on a S impl e Phys i ca l Model of the Head , 1 1 J . of Bi omechan ics , Vol . 6 , No . l , pp . 1 -1 2 , 1 973 .

32) Gras s , A . G . , " Impact Thres hol ds of Bra i n Concu s s i on , 11 J . Am . Med . , Vol . 29 , pp . 725-732 , 1 958 .

33 ) Unterharnschei dt , F . and K . Sel l i er, " C l osed . Bra i n Injuries : Mechan ics and Pathomorphol ogy , 11 Proc . Head Injury Conf . , L i ppi ncott , pp . 32 1 - 341 , 1 966 .

34) L i ndgren , Sten 0 . , " Experimental Studies of Mechan ica l Effects i n Head Injury , 11 Acta Chi r . Scand . Suppl . 360 , 1 966 .

35) Kopecky , J . A . and E . A . Ri pperger , 1 1Cl osed Bra i n Injuries : An Engi neer­i ng Analys i s , '' J . of Bi omechani cs , Vol . 2 , No . l , pp . 29-34 , 1 969 .

2 5 1

TABLE 1 RESULTS OF RHESUS MONKEY HEAD IMPACTS AND IMPEDANCE TESTS

D i recti on Pul s e Model Constants Mean of Head Accel eration Durati on Stra i n Imoact (G ' s) (msec) wl k c w2 E:

Front 1 800 3 . 6 . 051 39 , 000 1 . 6 1 . 1 0 . 032

Si de 1 500 2 . 8 . 040 33 , 000 2 . 1 1 . 0 0 . 032

Top 980 7 . 0 . 030 1 8 , 000 1 . 2 0 . 9 0 . 032

Back 1 000 3 . 4 . 035 20 , 000 2 . 9 1 • 1 0 . 032

TABLE 2 SUMMARY OF HEAD INJURY INDICE COMPARI SONS

Accel . Pul se H IC Dura- Dura- Aver .

Pul se t ion Peak t i on Accel . r . D . {msec} (g' s} GSI {msec} {g' s} HIC JTI RBM EDI MSC Sine 1 0 1 00 4S8 7 . 8 83 41 S 0 . 8S3 0 . 1 46 . 0039 Tri ang 1 0 1 00 286 S . 7 72 247 0 .691 0 . 1 1 8 .004S Square 1 0 1 00 1 000 1 0 . 0 1 00 1 000 1 . 026 0 . 1 72 . 0061

22 1 28 1 62 1 1 70 40 49 680 0 . 749 0 . 909 0 . 1 20 . 0037 23 l OS 207 1 609* 4 l 2S 702 1 . 087* 0 . 90S 0 . 1 80 . 0053 24 1 87 1 44 922 33 S4 704 0 � 7 1 8 0 . 950 0 . 1 23 . 0027

-0 2S 1 88 1 09 7 1 7 28 S2 SS5 0 . 683 0 . 937 0 . 1 1 1 .0039 .--26 1 82 1 1 1 825 43 45 597 0 . 764 0 . 934 0 . 1 22 . 0028 Q)

.,... ...c 41 1 87 248 2080* 38 61 1 082* 1 . 1 1 7* 0 . 890 0 . 1 82 . 0056 \/l c:( -0 42 2 1 1 290 3066* 2 254 2057* 1 . 608* 1 . 283* 0 . 269* . 0065* s:: \/l .,... \/l 43 202 l SO 91 7 46 47 71 6 0 . 668 0 . 741 0 . 1 09 . 0041 3 '°

.-- 44 2SO 1 1 7 l l S4 32 S9 841 0 . 839 0 . 863 0 . 1 37 .0033 :;:.., u E 4S 2SO 1 1 1 82S 1 8 66 644 0 . 901 1 . 098 0 . 1 47 .0037 E ::J Cl

-0 1 1 300 4 1 8 2229* 36 7S 1 273* 1 . 01 8* 1 . 373* 0 . 1 64 .0046 .--Q) 1 2 l S5 1 51 1 020 50 46 701 0 . 721 0 . 909 0 . 1 1 5 . 0030 .,... ...c 1 3 1 53 1 74 1 1 94 24 68 903 0 . 81 2 1 . 087 0 . 1 30 . 0028 \/l CO -0 21 l OS 1 50 l 27S l S 61 438 0 . 922 0 . 940 0 . 1 53 . 0038 s:: \/l .,... \/l 31 88 85 39S 43 31 232 0 . 603 0 . 685 0 . 1 00 . 001 8 3 '°

.-- 32 62 S8 400 31 42 3SS O . S27 0 . 733 0 . 088 . 0024 >,U � 47 2SO 98 1 577* 34 69 1 360* 0 . 9 1 3 0 . 1 1 6 0 . 1 46 . 0060 ::J Cl O'> '° 51 1 62 80 1 246 26 60 7 1 8 0 . 79S 1 . OSl 0 . 1 30 . 0047 CO S- S3 1 67 66 91 S 1 2 49 544 0 . 666 0 . 858 0 . 1 07 . 0038

.,... S4 1 89 S6 697 36 41 392 O . S68 0 . 782 0 . 092 . 0029 c:( s:: \/l s s 1 69 7S 1 249 30 S5 682 0 . 759 0 . 980 0 . 1 22 . 0043 '° .µ S7 l S4 76 1 324 26 61 76S 0 . 764 0 . 997 0 . 1 22 . 0043 E \/l ::J Q) 58 1 73 7Q 1 21 2 24 63 763 0 . 798 1 . 059 0 . 1 28 . 0044 :c 1--

S9 1 87 78 1 224 28 S9 751 0 . 783 1 . 031 0 . 1 26 . 0041 s:: '° SB 1 59 78 1 446 30 61 875 0 . 754 1 .Ol l 0 . 1 20 . 0044 E 0 SC 1 54 67 1 077 1 26 29 563 0 . 666 0 . 866 0 . 1 09 . 0038 .--

.--0 :c

S- SA 1 47 78 1 305 1 24 34 848 0 . 683 0 . 892 0 . 1 1 0 . 0042 .,... c:( S2 1 58 65 987 1 22 29 546 0 . 628 0 . 81 9 0 . 1 02 . 0036 � E O'>

S6 1 44 7 1 1 394 1 24 34 832 0 . 7 1 6 0 . 975 0 . 1 1 6 . 0041 ::J '° Cl CO

*Exceeds tol erab l e val ue of appl i cab l e cri teri on .

2 5 3

. ·. �

Dummy Wi nd-s h i e l d Tests

C l as s ' A '

Dummy W i nd-s h i e l d Tests

C l as s ' B '

Ho 1 1 o-man A ir Bag Tests

Human

Ho l l o­man A i r Bag : Dummy

TABLE 3 COMPARISON OF NORMAL IZED HEAD I NJURY INDICES

TEST I . D .

22 23 24 25 26 41 42 43 44 45

1 1 1 2 1 3 2 1 3 1 32 47

5 1 5 3 54 55 57 58 59 58 5C

52 56 SA

S I HIC

. 780 0 . 680 1 . 073 0. 702

. 6 1 4 0 . 704

. 478 0. 555

. 550 0 . 597 1 . 386 1 . 082 2 . 043 2 . 0 57

. 6 1 1 0 . 7 1 6

. 769 0 . 841

. 550 0 . 644

l . 485 l . 273 . 680 0 . 701 . 746 0 . 903 . 850 0 . 438 . 263 0 . 232 • 267 0 . 355

1 . 051 l . 360

. 831 0 . 7 1 8

. 61 0 0 . 544

. 46 5 0 . 39 2

. 832 0 . 682

. 882 0 . 765

. 808 0 . 763

. 81 6 0 . 7 5 1 • 964 0 . 87 5 . 71 8 0 . 563

. 658 0 . 546

. 929 0 . 832

. 870 0 . 848

JTI

0 . 749 1 . 087 0 . 7 1 8 0 . 683 0 . 764 1 . 1 1 7 1 . 608 0 . 668 0 . 839 o. 901

l . 0 1 8 o . 721 0 . 8 1 2 0 . 922 0 . 603 0 . 527 0 . 9 1 3

0 . 79 5 0 . 666 0 . 568 0 . 759 0 . 764 0. 798 0 . 783 0 . 7 54 0 . 666

0 . 628 0 . 7 1 6 0 . 683

RBM

0 . 727 0 . 724 0 . 760 0 . 750 0 . 747 0 . 7 1 2 1 . 026 0. 593 0 . 690 0 . 878

l . 098 0 . 727 0 . 870 0 . 752 0 . 548 0 . 586 0 . 893

0 . 841 0 . 686 0 . 627 0 . 784 0 . 798 0 . 847 0 . 825 0 . 809 0 . 693

0 . 655 0. 780 0 . 7 1 4

ED I

0 . 600 0 . 900 0 . 6 1 5 0 . 555 0 . 6 1 0 0 . 9 1 0 1 . 345 0 . 545 o. 685 0 . 735

0 . 820 0 . 575 0 . 650 0 . 76 5 0 . 500 0 . 440 0 . 81 1

o . 722 0 . 59 4 o . 5 1 1 0 . 677 0 . 677 0 . 7 1 1 0 . 700 0 . 666 0 . 605

0 . 566 0 . 644 o . 61 1

MSC

0 . 607 0 . 869 0 . 443 0 . 639 0 . 459 0 . 9 1 8 1 . 066 0 . 672 0 . 54 1 0 . 607

0 . 7 54 0 . 49 2 0 . 459 0 . 623 0 . 295 0 . 393 0 . 983

o . 770 0 . 623 0 . 47 5 0 . 705 0 . 705 0 . 721 0 . 672 o . 721 0 . 623

0 . 590 0 . 672 0 . 689

.06

c ' .S .05

z .04 <i a: 1-

.03 U> z <l w .02 :::E

.01

0

0

Rigid lmpactor

A

--1 J

ESI

o 1 No lnjury

• 2 A 3

o 5 Fatal

� 1 � D 1 &

_ _ _ _ _A _ _ _ _ _ _ _ _ _ _

• '! • .t>. 1 TOLERABLE MEAN STRAIN

0 LI & t = 0.0320 in/in • •• f' 1 o

•o

• o

o e 1 • • • o

1 2

1 3

1 4

1 0 0

1 1 5

00

1 . 6

1 7

PULSE DURATION (msec)

1 8 1 9

1 10

Fi gure 1 . MSC Stra i n Level s for Rhesus Head Impacts Vari ab l e Di rection and Pul se Duration .

. 016

0(5) .014 0(5) 0 HUMAN CAOAVER IMPACTS WITH ESTIMATEO

o(5) SEVERITY OF INJURY ( )

0(5) A COLONEL STAPP VOLUNTEER SLEO RUNS

.012

:s 0(5) • HUMAN VOLUNTEER SLEO RUNS ' ! .010 (/) "' 3 ;; .008

z 0(3)

C( 0:: TOLERABLE MEAN STRAIN c • 0.0061 in/in 1- .006 (/)

0 (/) � :i; l .004 0(2) •

0(2) 0 ( 1 ) • . 002 0 ( 1 ) " "

0 ( 1 ) " 0

50 100 500 1000 2000

PULSE DURATION (msec)

Fi gure 2 . Mean Stra i n Criterion for Humans vs Pul se Durati on .

2 5 5

t - 180 � 1 : :� f +180 1000

s,,„,"e 500 /6/ ... � L 100

50

10 ö � 5 r" } 1

0.5

0.1

.--.--�--.,,.---��---�--,,��----..-,��-.--�

10 50 100 500 1000 Frequency, hertz

'>

\0 o� ,..,,.

ll'o•'· Q) p'>

5000

Fi gure 3 . Mechani cal Impedance of Rhesus Monkey Head .

F i gure 4 . Cri t i cal Accel eration Surface for Rhesus Monkey , lnjury I ndex = 3 .

H.S.R.I. Cadaver Tests (no skull fractures): A Rigid Striker

3000 D. Padded Striker

� '2 1000 :>

• Human Volunteer Runs 0 Wayne State Cadaver Test (skull fracture)

1 cn 500 c ·� l:! ..!! GI V 100 V <

lnjury c Index (in/in) - 5 .00964 - 4 .00787 - 3 .00610 - 2 .00433

GI so � - 1 .00256 . . . GI cl

10 0.6 5 10 50 100 200

Pulse Duration (msec)

Fi gure 5 . Mean Stra i n Cri teri on for Humans , Sagi ttal Pl ane Loadi ng .

S I H I C J T I R B M E D I MSC HIAO INJUU J • TOUIANCE ltf\llSEO lltAIN UHCTIVE MAX IM.UM SllAIN

SIVUITY INOIX CllTUION INDEX MOO(l Ol"lACIMENJ INOIX CltlTU:ION

IGADDI I VlllSACI 1. NHUA l ISlATU NSCH l l( I 1 FAN ) l llUNN) ( SlAlNAKfl)

. ' . Weighted lmpvlu Weighted Impulse � � a� 93+8 ol oltl

SI = f fam]'" dt •

11"'' III tHDllth

Au. I" 1-1111iti

s1,,, = 1000

ol olfl

Let ö -{�[l)lt "-Tl."='iJ

H I C = r„1, a„ c1,1,) t �-· · � 0( t,( 11( T

HIC1,1= 1000

l�l w,• .JkTm (roill1H

ß • c/c<

W,•635. ß ' 1 .0

J = � � O.Ot2S • h•

J101 ::= LO

1 ' . X(I)

'

----1 W,' 175. ß .o.4

�·� ,,, ,,, -lU.3 .fc ! 1.:U in

: ( ' lWI �- · -:

W,•482. ß ' 0.707

�� � 1 0.2 .„ _J

< l(ll) ,----

m,,0.6 llbtl

m,,10.0 Ob1I

c ' 2.0 (lb •••i„J

k ,50 000. 1 „ ,.,

c = x,..o./l

HUMAN: l ; S,1S i11IA·":

C101" 0.0061 ;„ ;„

Fi gure 6 . Summary o f Head Injury Cri teri a .

2 5 7

VI ... u ö z >-°' ::::> ..... a; Q c ... ::c Q ... .... ::; c � °' 0 z

x:Windshield o•Airbag

1 . 8

1.4

1 .0 . . :1 •• . • • 1 • . . . · � ;l +i : . '

�1 y� • • . . 0.6 �. 1 8• • . . � . . . � · . . „ . .. .

0.2 S I H I C J T I RBM E D I MSC

F igure 7 . Compari son of Head I nj u ry Cri teri a .