Roy S. Rice Cal span Corporation Buffalo, New York 14221

Transcription

1 ACCDENT AVODANCE CAPABLTES OF MOTORCYCLES Ry S. Rice Cal span Crpratin Buffal New Yrk December

2 ACCDENT AVODANCE CAPABLTES OF MOTORCYCLES Ry S. Rice Cal span Crpratin Buffal New Yrk ABSTRACT This paper describes a study f mtrcycle accident avidance prperties invlving applicatin f experimental and analytical (simulatin) techniques t the determinatin f perfrmance characteristics.* Emphasis was placed n evaluating mtrcycle respnse and driver-vehicle interactin in tw basic maneuvers - steady-state crnering and lane-changing. Results frm full-scale tests f ne mtrcycle and simulatin evaluatins f six machines are discussed. Several supprting activities - measurement f mtrcycle physical characteristics and tire perfrmance capability definitin f equipment and instrumentatin requirements develpment f prcedures and perfrmance parameters and applicatin f a.nalytical methds t stability and cntrl evaluatins - are briefly utlined t give a brad verview f the apprach. The paper treats three aspects f mtrcycle dynamics f special interest t the safety prblem. These are: (1) the sensitivity f a mtrcycle's cntrl parameters t the perfrmance characteristics f its tires; (2) rider-vehicle interactins during maneuvering particularly with respect t rider cntrl inputs; and (3) the applicatin f analytical and simulatin methds t mtrcycle handling evaluatins. Other safety-related prblems are identified and recmmendatins fr additinal studies are given. * This wrk was perfrmed by Calspan Crpratin fr the Natinal Highway Traffic Safety Administratin under Cntract N. DOT-HS Rice

3 NTRODUCTON With the increasing interest in and use f mtrcycles in the United States cncern has grwn regarding the safe peratin f these vehicles. The Natinal Highway Traffic Safety Administratin (NHTSA) has respnded t this cncern with the prmulgatin f Federal Mtr Vehicle Safety Standards (FMVSS) Numbers and 123 which treat requirements fr lighting braking and cntrls respectively. t is als spnsring research in ther aspects f mtrcycle perfrmance -- the study described in this paper is cncerned with an investigatin f the accident-avidance capabilities f mtrcycles. Particular emphasis was placed n the lateral-directinal cntrl prperties f these vehicles and the bjective evaluatin f respnse characteristics and handling quality. The verall bjectives f the study were: 1. T develp a set f mtrcycle accident avidance test prcedures and t define the meaningful bjective respnse parameters that can be used t quantify accident avidance capabi 1i ty. 2. T evaluate the accident avidance capabilities f a representative sample f mtrcycles using the accident avidance test prcedures in a cmputer simulatin. The apprach used in meeting the abve bjectives cnsisted f perfrming full-scale experiments with a single representative mtrcycle in tw basic test prcedures - cnstant speed crnering and a lane change maneuver - which were develped fr this purpse. The results f these tests were used t evaluate the utility f selected perfrmance parameters and fr validatin f the Calspan digital cmputer simulatin prgram f mtrcycle dynamics which was used t investigate the characteristics f five additinal machines. Befre discussing the varius phases f the study it will be useful t give a brief backgrund descriptin f the general prblem f mtrcycle handling characteristics. Even befre 1900 varius researchers attempted t frmulate mathematical mdels f the lateral-directinal characteristics f tw-wheel vehicles. Many f these early effrts resulted in rather elegant mathematical descriptins but they suffered frm inadequate numerical infrmatin n varius cmpnents (mst ntably tires) r were simplified by neglecting terms r treating nly the steady-state fr mathematical tractability. Later studies during the perid vercame sme f these difficulties in the representatin f the vehicle but these treated the rider as nly a passive element in the system. Since 1970 increasing interest in tw wheel vehicles has led t a brader attack n the prblem and has resulted in a capability t investigate the dynamics f the rider-machine system. Amng the mst useful descriptins f the current state-f-the-art are papers by Sharp n the mtrcycle nly (Reference 1) Weir n the rider-vehicle system based in part n Sharp's vehicle mdel (Reference 2) and Rland n the clsed lp rider-bicycle system in a path-fllwing maneuver (Reference 3). 2 Rice

4 With the abve as a fundatin n which t base an apprach t accidentavidance capability evaluatins in the face f an almst cmplete lack f any frmalized test prcedures fr such evaluatins the study was aimed at defining reasnable testing techniques and meaningful perfrmance parameters and then applying these prcedures in full-scale tests and simulatin t demnstrate applicability. This paper gives a brief summary f the wrk which was perfrmed. A cmplete discussin f the study is given in Reference 4. METHODOLOGY Since the apprach used in this prgram invlved a number f separate activities each phase will be discussed individually befre presenting the cllective results f the experimental tests and simulatin studies. Mtrcycle Selectin Six mtrcycles were selected fr the simulatin studies t prvide a brad range f size weight and type f machine s that any trends in the perfrmance characteristics relatable t these factrs might be identified. Each f the mtrcycles selected fr this task is representative f a segment f the mtrcycle ppulatin with similar design parameters which determine their handling perfrmance. mprtant parameters fr cnsideratin in the selectin f a representative sample were weight frame design frnt and rear suspensin characteristics tires weight distributin and specific pwer utput. The mtrcycles range in weight frm 150 t 700 lbs and in pwer frm less than 10 t ver 70 hrsepwer. Fr the mst part frnt suspensin is by il-damped telescpic shck absrbers; at the rear there is usually a spring/shck cmbinatin cnstrained by a trailing arm. Three basic t)~es f tires are used: trials universal and ribbed. These characteristics can be used t divide the mtrcycle ppulatin int grups frm which a representative member f the grup was selected. The final selectins were: Heavyweight Street: Superbike: Harley Davidsn FLH-1200 El ect rg lide ntermediate Street: Hnda CB 360 Dual Purpse: Lightweight Street: Nrtn 850 Cmmand Radster Yamaha X F Kawasaki Hnda CB 125 Kl 3 Rice

5 Mtrcycle Physical Measurements Simulatin data input requirements call fr the numerical values f a set f physical characteristics f the mtrcycle which describe its gemetrical layut and its mass and inertia prperties. Althugh sme f this infrmatin is available frm manufacturer's publicatins mst f these data (especially the imprtant mment f inertia parameters) had t be measured. Many f the characteristics culd be simply measured as linear r angular dimensins r weights. The determinatin f mments f inertia f principal elements f the machine (as well as the values fr the cmplete mtrcycles) were made using a trsinal pendulum methd as indicated in the phtgraph Figure 1. The primary measurements which were made include ttal weight and weight distributin weight f frnt assembly rll and yaw mments f inertia f the cmplete machine rll and yaw mments f inertia f the frnt assembly wheelbase frk tube angle frk tube mass ffset and height f center f gravity. Tire Tests Perfrmance tests n representative riginal equipment tires fr each f the six mtrcycles were made n Calspan's Tire Research Facility (TRF). The principal factrs were: nflatin pressure - accrding t manufacturer's recmmendatin Nrmal lad - tw cnditins were tested: (1) nminal value with a 200 lb. rider and (2) 120% f the nminal value Slip angle range - sufficient t cver all reasnable perating cnditins; assuming symmetry f perfrmance fr ± values f slip angle. nclinatin angle range - full range f the tire test facility (withut mdificatin) assuming symmetry f perfrmance fr ± values f inclinatin angle. The fllwing measurements were made: three and three frces } mments vs. slip angle (a) and inclinatin angle (y) (1) at nminal frnt tire lad and inflatin pressure (2) nminal rear tire lad and inflatin pressure 4 Rice

6 Figure la Measurement f Yaw Mment f nertia ( zz ) Figure lb. Measurement f Rll Mment f nertia (XX) 5 Rice

7 (3) nminal frnt tire pressure and 120% lad and (4 ) nminal rear tire pressure and 120% lad fr a = deg. y deg. A sample data plt as generated by the TRF system is shwn in Figure 2. Nte that this plt differs frm the usual carpet plts f tire perfrmance (in which nrmal frce is included as an independent variable) by shwing the slip angle and inclinatin angle effects n side frce at a nminally cnstant value f nrmal frce. This frm f presentatin is very cnvenient (and useful) fr representing tire data fr tw-wheel vehicles fr which camber thrust is imprtant and lad transfer effects are small. n additin t side frce infrmatin which was f primary interest rlling resistance aligning trque and verturning mment data were btained as well. Fr use in the simulatin the test data were reduced t simple representatins f nrmalized crnering stiffness nrmalized camber thrust stiffness and effective pneumatic trail fr the nrmal lads at the tires used in the studies. Briefly the test results shwed:.. Variatin in crnering stiffness cefficient (nrmalized crnering stiffness) frm abut.33 lbs/deg/lb (at lightly laded cnditins) t abut.15 lbs/deg/lb. 2" Values f camber stiffness cefficient (nrmalized camber thrust stiffness) ranged frm.009 lbs/deg/lb t.021 lbs/deg/lb. 3. Values fr these cefficients were lwer fr the trials-type tires than fr the ther types. 4. Fr a given tire the value f the crnering stiffness cefficient decreased with increasing nrmal lad; the value f camber stiffness cefficient was relatively unaffected by nrmal lad. 5. Values f pneumatic trail based n aligning trque measurements ranged between 0.5 and 1.0 inches. 6 Rice

8 :;LP ~t~gle ([ ~Er..a.i H~C ~NATOi ~ ANGLE (DEG) 400 e.290 ~ ' ~- 0. e 0'0.1 /' ~ 1 ~. ~ J 2 ~ Q' RUN: n.n.' il. Q' Aflil ~- ~ -3 /. -~ Oftn -"...~ l L ~ "-~ ~ ~ ~~ r [-i... ~ -/ ~ /' " ~ ll ~ J -s. ~ / JfC ~. 'XC' -1 AGO ORQ:2 (Jl.[):2 Figure 2. Typical Tire Test Plt 7 Rice

9 Test Prcedure Develpment n attempting t define suitable test prcedures with which t determine first-rder perfrmance differences amng mtrcycles emphasis was placed n investigating lateral-directinal cntrl characteristics at nminally cnstant speed. t is recgnized that braking characteristics are nt cvered in this apprach but it was believed that extensin f the simulatin studies t include detailed investigatin f this aspect f perfrmance wuld have cmprmised the degree t which the directinal cntrl characteristics culd be studied. The fllwing factrs were used in defining the test prcedures: Cmpatability with simulatin - fr validatin purpses mdificatin requirements cst f peratin dependence n rider mdel. Full-scale test peratin - cverage f perfrmance range instrumentatin and equipment requirements test area cntrl input cverage cst f testing test safety. Tw prcedures were identified fr use in this prgram. One was t be cncerned with the determinatin f basic steady state cntrl respnse characteristics; the ther was t invlve rider-vehicle interactin effects under transient maneuvering cnditins. Emphasis was placed n utlining a basic stability and cntrl test which wuld yield first rder directinal cntrl characteristics f mtrcycles. The prcedures used fr autmbile evaluatins which are described in SAE XJ 266* were studied fr pssible adaptatin t mtrcycle testing. Each f the fur general methds - cnstant radius cnstant velcity cnstant thrttle and cnstant steer angle - was cnsidered and a cnstant speed type f test was finally selected. * Prpsed Recmmended Practice SAE XJ Passenger Car and Light Truck Directinal Cntrl Respnse Test Prcedures. 8 Rice

10 The primary purpse f this test is t measure the steady-state cntrl gains r sensitivities f the vehicle. The parameter f special interest is lateral acceleratin gain as given in three frms: (1) psitin cntrl sensitivity - the fixed cntrl respnse f lateral acceleratin t steering angle (2) trque cntrl sensitivity - the lateral acceleratin respnse t a steering trque input (3) rider lean cntrl sensitivity - lateral acceleratin as a functin f the rider's lean angle with respect t the machine. t is desirable that these parameters be determined ver a fairly brad speed range s that any perating cnditins f reduced stability (r instability) is identified. n effect the test shuld be aimed at determining cnditins at which the peratinal safety f the machine might be cmprmised r wuld impse severe demands n the rider fr cmpensatin. Varius frms f a lane change maneuver were cnsidered fr use as the primary transient handling task in the prgram. t was intended that this maneuver prvide baseline infrmatin fr rider skill differentiatins as well as mtrcycle perfrmance discriminatin. After reviewing several versins f this maneuver (single and duble lanes variable gemetry and dimensins etc.) a single lane change prcedure that is believed t address each f the requirements was adpted. The ratinale fr its selectin was: (1) The single lane change represents a maneuver frequently perfrmed by cyclists n the rad. By varying the lngitudinal distance ver which the fixed lateral displacement can be develped it prvides fr a range f speeds t be investigated. t calls fr the rider t apply bth steer and lean cntrl inputs and ffers a means fr cmparing bth stability and cntrllability characteristics. (2) t is cmpatible with present capabilities f the simulatin (i. e. it is initiated frm a trimmed straightahead cnditin) and affrds a gd basis fr validatin f transient behavir. These tw prcedures are in sme sense a cmplementary set. The respnse characteristics determined in the directinal cntrl tests are presumed t have sme bearing n the perating cnditins fr which the lane change can be successfully accmplished. These relatinships are cncerned with magnitudes f input cntrl levels limitatins assciated with the perfrmance envelpe f the test machine r dynamic cmpensatin required f the rider. Fr the purpses f this study mst f the tests (bth full-scale and simulated) were perfrmed at a nminal test speed f 40 mph and lateral acceleratins up t.5g. 9 Rice

11 Simulatin Descriptin The vehicle--rider mdel is a system f three rigid masses with eight degrees f freedm f mtin: six rigid-bdy degrees f freedm f the rear frame a steer degree f freedm f the frnt wheel and a rider lean degree f freedm. The analysis f the mdel is based n the fllwing assumptins: (1) The mass distributin f the vehicle is assumed t be symmetrical with respect t the vertical-lngitudinal plane thrugh the gemetrical center f the vehicle. Thus the X-Y and Y-Z prducts f inertia are assumed t be zer. X-Z prducts f inertia and all mments f inertia f each rigid mass are included. (2) The vehicle is assumed t be mving thrugh still air n a flat level surface. The aerdynamic drag the frnt t rear weight transfer due t aerdynamic drag and the pitching mment aerdynamic lift and steer mment due t windshield aerdynamic drag can be included as apprximatins. (3) A driving thrust n the rear wheel is included t vercme the aerdynamic drag. Thus the vehicle is initially mving at cnstant speed. Frnt tire rlling resistance is assumed negligible. (4) Tire lateral frces as functins f slip angle inclinatin (camber) angle and vertical lad are mdeled independently fr frnt and rear tires. (5) External trques acting abut the steer axis include the mments due t the lateral and vertical tire frces tire aligning trque and a cuple due t the aerdynamic drag frce n the windshield. The gyrscpic mments f the wheels and engine are included. (6) Viscus steering damping can be included between the frnt assembly and the rear frame. T analyze the handling f a tw-wheel vehicle in the nnlinear regin f peratin the equatins f mtin are written in cmplete nnlinear frm. All inertial cupling terms between the rider the frnt assembly and the rear frame are included. The digital cmputer simulatin prgram fr this analysis slves the equatins f mtin fr prescribed rider cntrl inputs and/r disturbance inputs and prduces time histries f the resultant vehicle mtins. The general frm f the rider cntrl mdel is shwn in Figure 3. This mdel invlves a rll stabilizatin lp and a path-fllwing guidance lp which are cnnected thrugh a simple human peratr transfer functin t the vehicle dynamics mdel. The basic frm f cntrl fr bth stabilizatin and path-fllwing assumes matching actual rll angle f the system with a "cmmand rll angle" -- a rider-generated term which crrespnds t a desired lateral acceleratin. 10 Rice

12 4 p fie Nt! OTHER MOTON VARABLES ~ ~:;(t;) / X ~ 'i~ u ~ RDER RDER DYNAMCS MODEL MODEL OF GUDANCE STABLZATON BiCYCLE FUNCTON FUNCTON AND Npd r4 ~ RDER 'r+ i - ri4 ; / ;t y t- ~? 91" ~ f-' f-' j J x;.~; - BCYCLE COORDNATES 5 d - RDER LEAN ANGLE :;0 1-'. () (D u - FORWARD VELOCTY - RDER LEAN ANGULAR VELOCTY. " Y - YAW ANGLE e. - COMMAND ROLL ANGLE (; - ROLL ANGLE ~ - STEER TORQUE fj - ROLL VELOCTY.. Nla' - RDER LEAN TORQUE - ROLL ACCELERATON :i= (;:.p)- DESRED PATH P Figure 3. Blck Diagram f Bicycle Rider Cntrl Mdel

13 RESULTS A few typical results f mtrcycle perfrmance are shwn here t demnstrate the type f infrmatin btained in the study. Simulatin Apprximately 40 runs were made with the simulatin prgram in the tw prcedures rectmlended fr use in this study. The majrity f these were devted t the evaluatin f steady state lateral-directinal respnse characteristics --- partly t emphasize the fundamental nature f these parameters and partly because f prblems f executin and interpretatin f the transient perfrmance task and its results. Fr use in the simulatin studies the prcedures were adapted fr cmpatability with the simulatin prgram in rder t maximize the efficiency f its utilizatin. Fr the directinal cntrl tests these adaptatins cnsisted f specifying a run array f several nminal cmmand rll angles at cnstant speed and restrictin f rider lean cntrl t a passive rle. This apprach allwed fr full cverage f the range f lateral acceleratin values f interest (but avided a requirement fr path cntrl) and fr emphasizing the steer cntrl mdes in the analysis. Typical time histries fr the primary input and utput variables in the directinal cntrl test simulatin are shwn in Figure 4. The cnditins fr this run were a speed f 40 mph and a desired (cmmand) rll angle f 25 degrees. This rll angle crrespnds t a lateral acceleratin f abut.5g. The rider remained apprximately in-plane with the mtrcycle thrughut the run -- reaching a maximum f.18 d.egree (lean--ut) early in the maneuver and settling t a steady-state value f.07 degree (lean-in). The respnse is rapid and well damped at this cnditin fr the rider cntrl mdel cefficients selected fr this maneuver. The initially-applied reverse steering trque prduces a small reverse steering angle. The rll angle (lateral acceleratin) builds up t a slight versht f the final steady state cnditin within.7 secnd f the time at which the cmmand rll angle is established. The ffset in actual rll angle frm the cmmand value f abut 2 degrees is due t the trque requirement withut a feed-frward integratin term. Results fr a series f runs ver a range f lateral acceleratins fr all machines with this test prcedure are shwn in Figure 5. They give values fr steer angle sensitivity at a reference perating cnditin 40 mph speed 200 pund rider and recmmended tires. The range f data fr the Hnda 360 shwn in the figure was extended t demnstrate the reasnable linearity f the characteristics ver the lateral acceleratin perfrmance envelpe. n general the steady state input requirements f steer angle and steer trque tend t be related t vehicle size. 12 Rice

14 G_ ~ 51frDY $la;e LCr..~[8nlc. - r"jjda CB-3GOC HCH1RCYCLE - ~O MPi ghar' 7S BRDGES10lJE 3.DO-16 ~RQ~l TRES. E;RlCE51Qt 3.S!'l-l8 REAR ljres STEER RND ROL~ RNGLES f) T~] ROLL.C) STR Ri'JC '6 CO~l ROLL 9 TORQUE ~ V.J ;:0 1-" n (J) 0 f';. rr) LJ LL! c.:: C' :... J.~ ~:; 0 r-j / ~ -' /'--'./ /' / d \il. ;./ _._ -() rr.-- 2J E9 \. 0-- \ ~ (... ) "\. -- / e) \00. ft 1 ~. SO: 2 w 10 & 0[. " S; ; ~-r-- r? r ----T TT G TTHF [SF~J Figure 4. Simulated Directinal Cntrl Test Respnse

15 f ' ; r ~ r ~ ~ : : : : : : : : : : : : : : : r----- : : : : : : : : : : : : : : -i : : : : : : : : : : :: : i ~ --6--T ~-~~ ~-~r;~~---r r T r T i i : : A H d :: : : ~ n a: 125 : : : t ~ ; ~ ~ r ~ ~ ~ ~ ~ ~-- ~ ~ ~ 0 : : ; Kawasaki 250 : : : : ' A' :: ~ : V Yamaha 650 : : : : : : _ L _~----. '- -:.. '- ~ : V : Nrtn 850 :. : : : : : : : :0: : : : : : HarleiY Davi~sn 1l200: : : : ~ ' ~ r t l r ~ i t ~ i i :: : :: : " ' ~ ~ ~ ~ ~ : ~ ~ : : 2.5 : ~ ~ ~ ~ ~ ~----- L ' ' ' ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ : : : : : ~ ~ ~ ' ~ ~----- : ' ~ *-0---~ : : ) : ~ ~ ~---J!-~..-0: " ~ ~ ' ~ ' l.. : t - ' ~ ~ ~ ~ ~- : ' ; : : : : : : : 1 ' t : ~ ~ ~ ~ :-.6--i ~ ~ '. tt 1 t"'" : : :~ ----; : : : : : ~ : ' : : :.u.: : : : : : ' til ~ ~ ~ r ~ r ~----- : ' ~ : :~A!:~~~;.~1- _~_~~_~J~-r:aJ ~f9!l-= g -~ ~.L..!. _..!.--- ~ l.. Figure 5..:4.;5 Simulated Directinal Cntrl Respnse Characteristics Steer Angle : i ~ ; _ --i ~ " _.! ~ ~ _ ~ tit --- l! ~ J ~ ~ J j ~ ~_ 14 Rice

16 Transient Handling Maneuver The transient handling maneuver used fr this study was the single lane change. This sirrlulated maneuver requires peratin f the cmplete ridermtrcycle mechanizatin. Fr the rider mdel used in the simulatin several errr cefficients are emplyed in the path-fllwing guidance mde f peratin. Successful perfrmance requires that values f these cefficients be prperly balanced fr cmpatability with the desired path (as previewed) and predicted curse. Optimizatin f these cefficients and evaluatin f the six reference mtrcycles was nt accmplished in this study but the applicability f the simulatin t this maneuver was demnstrated. Figures 6a and 6b shw pltted time histries f the primary cntrl and mtin variables in the maneuver. Figure 7 shws a cmparisn f an ideal path and actual path fr a set f rider mdel cefficients that prduce a welldamped stable executin f the lane change but which shw a need fr a slightly lnger gap distance than desired t achieve the lateral displacement required. This run was made at a nminal speed f 40 mph with the Hnda 360 mtrcycle nrmal cnfiguratin characteristics and passive rider lean cntrl. Full Scale Testing The full-scale test wrk in the prgram invlved the measurement f significant cntrl input and utput mtin variables fr ne mtrcycle in the tw basic test prcedures. A Hnda 360G was selected fr this wrk. A cmplete instrumentatin system utilizing a telemetering link t minimize n-bard equipment weight was used. Figure 8 cntains phtgraphs f the sensrs. All tests were perfrmed by a single experienced rider n high skid-number asphalt surfaces under dry cnditins. Tw test prcedures were used -- ne cncerned with the measurement f input-utput relatinships in steady state directinal cntrl and the ther with rider/mtrcycle perfrmance and interactin in a transient cntrl maneuver. Briefly the directinal cntrl maneuver cnsisted f entering and maintaining a cnstant radius turn frm an initial straight path. Data were acquired starting with the straight path segment (which established a zer reference fr all variables except velcity) and cntinuing thrugh the transitin int the curved path and thrugh several secnds f the steady state turn. This maneuver was perfrmed at speeds frm 20 t 50 mph and turn radii frm feet. Varius cmbinatins f speed and turn radii were run giving lateral acceleratins between.08g and.ssg. nvestigatins f the maximum perfrmance capabilities f the mtrcycle r rider were nt undertaken using this prcedure. 15 Rice

17 0- m 'LANE CHANGE MANEuVER - H[WDA CB-360G MDll)RcYCLE - 40 HPt 25JlJ.J'7S BF;lDGE51OtlE FRONlllRE. BRtDGES11)'E E REAR 11RE ~lteef\ R~D ROLL R\~GLES ~J 0 -< ~ CJ ROLL 0 S TF RNC 6 CO~1 ROLL U") 0 w ~ 0: Q\ C) W 0 ~6.... '.-. L ~tr1 r- \... ~ ~~~ c N ;;>::J ~. () (1) (Y) "'' "" ~ Figure 6a. Simulated Transient Maneuver Test Respnse G T tv'e l SEC J

18 a 7 LANE CH8NGE rli1l~eu\"er - HONDA CB-':;GOG t-101qrcycle - 40 t-lph 25JUN'7S 8R1DGESTOfE FFiDNl l1re. 8RDGES10;~E 3.5D-18 REAR ilre STEER NCj T OROUE a rn c:j '" STR TORQ 0.---< X a t"\j ~ ~ >-' ---J :z... CO ---.J W ::;J d Si V \ 0 a: 0 r- ('u;! tv 1<\;\1... ' (v1y""-~ J... ~ /~... "~. / / ~ / ;:0 f-'. (') <1l..; r"\j_.. 1.: 0 Oi_ 0 Figure 6b. Simulated Transient Maneuver Test Respnse (Cntinued) G ~-~'-T-' 5 r--- ~ ~. '1F l SEC 1 1.: '"

19 ..{:: f~t"-~-'-r~'-"~ 1 r.---~.. i \...~-- -" ~ t -...:~--... ;~._-- --; "1! ~ P.. P.l 5 ~ H "---' U)... ~ n... r' ':>! f \:..~.. ' L';! i:~ ~ - ~i"'-'-' --- i~ -.-.~ i : Q.l bll ~ r..c:: u Q.l ~ r...:l "1j Q.l -.J r M ~ M U) t-- Q.l -< ;::j bll M '-L '.L i L!..j i 18 Rice

20 Figure 8. nput and Mtin Sensrs 19 Rice

21 The transient cntrl maneuver was a single lane change right t left with a lateral displacement f 12 feet. This prcedure was designed t shw the input-utput characteristics f the mtrcycle during a transient cntrl maneuver at cnstant speed and als t investigate rider task perfrmance and rider/mtrcycle interactin. Tests were run using this prcedure with a lngitudinal gap between the entrance and exit lanes set at 30' 4S' 60' and 80' and test speeds between 20 and S4 mph. Bth rider task perfrmance and mtrcycle capabilities were evaluated. Data recrds fr a typical directinal cntrl maneuver and lane change maneuver are presented in Figures 9 and 10. T facilitate interpretatin f the data the sense f each variable has been dented by "right" r "left" as the rider wuld view them. F:r each run the start is indicated by an event mark belw the yaw rate trace which was activated by a tape switch placed at the start f the curse layut. Fr the directinal cntrl curse this was the pint at which the straight path was tangent t the curve. n the lane change curse it was the last cne pair encuntered in the entrance (right) lane. Full Scale Test - Simulatin Cmparisn Representative results affrding an pprtunity t cmpare simulatin utput with experimental data wen~ btained fr the Hnda 360 in bth the directinal cntrl and transient handling tests. These cmparisns are shwn in Figure 11 (directinal cntrl) and Table 1 (transient handling). n general the steering angle values shw reasnable agreement between the tw appraches; differences are f the rder f a few tenths f a degree. A substantial difference in applied steering trque hwever is shwn in the directinal cntrl test results. CONCLUSONS AND RECOMMENDATONS The apprach and methdlgy used in the prgram have been described and typical results frm bth the experimental and analytical phases have been given t illustrate the capabilities and effectiveness f the chsen appraches t meet the prgram bjectives. With respect t these bjectives test prcedures that prvide discriminating measures f mtrcycle lateraldirectinal respnse fr bth steady-state and transient perating cnditins have been devised and successfully demnstrated in simulatin and full-scale experimental studies. The applicability f these measurements and the assciated perfrmance parameters t the definitin f accident avidance capabili ty has still t be established; in this regard hwever the mtrcycle state-f-the-art is n wrse ff than that f passenger vehicles. But n the psitive side the prgram has identified sme interesting characterizatins and special cnsideratins f mtrcycle stability and cntrl and handling quality which deserve further examinatin. These include: 20 Rice

22 Steer Angle (degrees) Right 3.5 "-~ki--l+j'iqi:'r:j4~j~~ t 1+ C' ':! '' " T!. t " '~:; ~ :'~r"-l-rl"--i---+'''':-:'.'h~-'-:'-~:-'f ~-. ";1:- ~fi-+"'"t-. rrjct". 1 -~-t ~-.4~~=~ -";~~~~+.':;:::' ''--::ih~'~~~'~:il~:2~=-': 1 ' ' ' " "j ~ -:t " ~:.:-t.' ; 'i'--c -~ jl ~j~i-; - -~-~t" i...;~ --;-1~1:' t~...;-; i 1"1.!'~lta±t: t -:!" j:; 't ~ji~ +---; : i;f~'; "~' j'" '..-'"'---- :t.....l..:._:...l. - j- ;- r-4 " 1'- - --t-. ~ 'T t- -r"; ": FT'T' ;-'-1- _~~:::.r..~ ;".L."...L._.'._.. i' "' t ' '. "i".... f---. : j.' r ::.. t. f _. " :. ~.'!.. :. _'..' _ 1.! Steer Trque (in-lbs) Rll Angle (degrees) Right 120 ~. :~ -'''.'-'-' i.~'i :i \ i..!... Right 45 Right 20 Yaw Rate (degrees/sec) Lateral Acceleratin Right Maneuver 40 mph 300 ft. Rider Lean Angle (degrees) Figure 9. Full Scale Test (Directinal Cntrl) 21 Rice

23 Steer Anile (degrees) Right 7 ~~ J -_cc ~~::-=~=---r~r-- T Steer Trque (in-lbs) Right 240 Lane Change Maneuver 40 mph Rll Angle (degrees) Right 4S Right Yaw Rate (degrees/sec) 0; - ; "'! L" "".~_.:-:. t "' LJ.:.:.:+-L.-J..:..;...:..L~---L:-..L-~-'T-'---. i -c-:l-; ~_---L Lateral Acceleratin Right.2Sg Rider Lean (degrees) FGURE 27: Angle FULL SCALE TEST (TRANSENT MANEUVER) Right 10 Hgure 10. Full Scale Test (Transient Maneuver) 22 Rice

24 ~ ~ ~ ~ f f ~ ~ r f ~ : ~ : ~ : ~ ~ ~ ! ~ ~ ': : ' ~ '- --!O ANGLl~ --i : jc : TORQ:JE : ----\ : _.\ ~ ~ ~ l ; _~TES1! ' : ~ ' ~ \ : : ~-- : ~ ~ : ~ ~ ~ ~ ~ ~ ~ : _.: ! ~ ~ ~ : :- ' i _\.._ ~ ! : ~ " ~ : : ~ ~ : ~ ~ : ~ -- - ()(; _ h : e2- f-- -zttr --' 0 0': U:::: Q 0 r' ~ L : i ~ ~ ~ : _ : ~ : : : 0 : : : ~...-; ' : ~ ' ~.L ~-;;.;.--:C-. :0--10 : : ~ : : : : ; ~ V---~ : y: 0: : t-/---r- : - ' ~ i ~ ---~---~---~-~--r---~ : : : ---""l T~-5-- -f j ~-----~-- -T-n-~-~------: 0 C) :~ : ~;: : :0-1 OlJ :z :<r: -:5-----: ~ : '.l.2. j.4: ' : : : : : : : : : : : : : : ~ \ ~ L L ~ ~ 1 iii iii L.dERAL ~CCELErtATON:-g i i! i : : : : : : : : : : : : : : ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ l ;--- ~----- Figure 11. Directinal Cntrl Test Cmparisn ~ ; : i : : : : ; ii ~ ~- ----""l ~. ~ ~ : - : : : : : : : : : : : : : : : : : : : : : : : : : ' : : : tit : : ' tt tit tit t : : _! L -- J!. L : 1 1. "_t L J. j 23 Rice

25 Table 1 Lane Change Cmparisn VARABLE EXPERMENT SMULATON Steer Angle (deg) Peak Psitive Peak Negative Steer Trque (in-lbs) Peak Psitive Peak Negative Rll Angle (deg) Peak Psitive Peak Negative Yaw Rate (degjsec) Peak Psitive Peak Negative Peak Recrded Lateral Accelermeter Output (g) Rider Lean Angle (deg) Peak Psitive Peak Negative Rice

26 1. Develpment f a steady state directinal cntrl test prcedure with which t evaluate the principal perfrmance respnse parameters. This prcedure has been demnstrated in full-scale tests t be suitable fr mtrcycles t prduce repeatable data t discriminate amng effects f different cntrl inputs and t be highly flexible fr studying perfrmance in any perating regime. 2. Demnstratin f the capability f the currently available simulatin f tw-wheel vehicle dynamics t prduce useful results n mtrcycle perfrmance characteristics. Althugh it is clear that certain imprvements in the mdel are essential fr brad applicatin t studies f mtrcycle accident avidance capability the simulatin has been shwn t yield reasnable representatins f mtrcycle-rider behavir in selected applicatins. 3. Cmpilatin f baseline infrmatin n mtrcycle physical characteristics and tire perfrmance that has nt previusly been available. These baseline d.ata (especially the dynamic inertial prperties and tire perfrmance) encmpass a wide ran:~e f machines and can be used. fr ther studies f additinal perfrmance characteristics. 4. dentificatin f the very significant rle f tire characteristics in mtrcycle respnse. n particular the sensitivity f the respnse parameters t camber thrust cefficient (with respect t abslute value and t any differences between frnt and rear tires) the imprtance f pneumatic trail t steer trque requirements and the initial categrizatins f steer requirements at trim may be cited. The majr effrt n the prgram has been devted tward utlining tw cnstant speed test prcedures and assciated perfrmance measurements which can be applied t discriminatin f mtrcycle respnse characteristics. This was an essential first step in evaluating accident-avidance capabilities. Much remains t be accmplished. n the lng term it will be necessary t investigate the fllwing: 1. Rider-machine interactin. 2. Operating cnditins. 3. Additinal maneuvers. 4. Crrelatin f perfrmance parameters with accident invlvement - identificatin f prblem machines accident statistics critical maneuvers and cnditins expansin f perfrmance data base. 5. Upgrading f simulatin. 25 Rice

27 REFERENCES 1. Sharp R. S. The Stability and Cntrl f Mtrcycles. J. Mech. Eng. Sci. Vl. 13 N.4 August Weir D. M. Mtrcycle Handling Dynamics and Rider Cntrl and the Effect f Design Cnfiguratin n Respnse and Perfrmance. University f Califrnia Ls Angeles Rland R. D. Cmputer Simulatin f Bicycle Dynamics ASME Winter Annual Meeting December Rice R. S. Davis J. D. and Kunkel D. T. Accident Avidance Capabilities f Mtrcycles. Calspan Reprt N. ZN-557l-V-l and -2 (tw vlumes). June Rice