You are on page 1of 5

Drexel University Design Report March 1, 2007 Introduction: For the 2007 competition season, Drexel Racing Formula

SAE is continuing to build upon its successful race-pro en design platform b! shifting focus on "e! areas to impro e performance# A thorough re ie$ of the 2007 competition rules re ealed that much of the "no$ledge and experience gained o er the past four !ears $ould still be rele ant# %ith this in mind, focus shifted to$ard identif!ing areas of the racecar that $ould be most easil! impro ed and that ha e the greatest influence on ehicle performance and o erall score# &he first area chosen $as mass placement, and in particular, the dri er's seating position# &he second area is suspension "inematics and steering geometr!, $here slight modifications help ma"e the static set-up less of a compromise bet$een cornering and acceleration# &he second area of focus $as the front frame, $here $eight could be reduced b! eliminating unnecessar! tubes and b! allo$ing the use of smaller hard$are b! designing the front bellcran"s in double shear# Again, this !ear's schedule includes a large amount of testing, $hich $ill be critical for the success of Drexel Racing# &o maximi(e the results of the time spent at the trac", data ac)uisition and d!namic simulation $ill be used extensi el!# As has become Drexel tradition, the manufacturing schedule $as compressed as $ell, to allo$ for more testing time# &he remainder of this document details the process that $as used to ma"e important design decisions and helped DR07 become the complex high-performance machine that it is#

Design fro

the ground up ! "ires and #heels:

&ires are arguabl! the ehicle component $ith the most influence on o erall performance# &he! are definitel! the component $ith the most influence o er the rest of the design# As such, the design process begins here# DR07 emplo!s *oosier 20#+x7-,- R2+A tires on all four corners# &his decision $as based on t$o main factors. /erformance as described b! tire data 0courtes! of FSAE &&12 and Drexel Racing's past experience and success $ith these tires# &he Formula SAE &ire &est 1onsortium data $as compared for four sets of tires. 3ood!ear's 20x4#+-,-, 20x7-,-, *oosier's 20#+x4-,- and 20#+x7-,-# 5oth manufacturers' 76 tires pro ided about 78 more cornering force than the narro$er tires $hen operating at an a erage temperature of ,-0 9F# From experience, it is "no$n that ,-0 9F is a reasonable temperature to expect the tires to reach on course# Also, re ie$ing past data has sho$n that it is reasonable to expect a Formula SAE car to spend about 7+8 of the time on course in a corner# Since the t$o main arguments for using narro$er tires are that the! are lighter 0and ha e less rotational inertia2 and the! $ill build up more heat, the narro$ tires $ere remo ed from the list of choices# &o compare both manufacturers' 7 inch $ide tires, plots of each tire's slip angle s# cornering force plots and slip ratio s# longitudinal force $ere examined# &he 3ood!ear tires pro ide +0 lbs more lateral force than the *oosier tires, but pea" at slip angles more than t$o degrees higher than the *oosier tires# Also, the pea" slip angle for the 3ood!ear tires mo es a fe$ degrees higher as the load on the tire is increased# Since the *oosier tires reach their pea" cornering forces at lo$er slip angles, more of the force generated $ill go into helping the car pull through the corner and a smaller component of the force $ill become drag# 1reating the pea" cornering force and aligning moment at smaller slip angles also means that less steering input is re)uired to get the tires to reach that pea" slip angle, $hich is effecti el! an increase in steering ratio# &his anal!sis sho$ed that from a performance point of ie$, the *oosier 20#+x7-,- R2+A tires ha e a slight edge# &he decision $as made to use *oosier's 20#+x7-,- R2+A tires# After choosing a tire, the next decision is ho$ to mount them to the suspension# &he choice of $heels $as determined b! $eight, cost, stiffness, and time to manufacture# Since Drexel Racing is traditionall! understaffed, the cost of purchasing an off-the-shelf $heel easil! out$eighs the time re)uired to manufacture a custom set# &o "eep the control arms as long as possible for a gi en trac", and to get a reasonable "ingpin inclination angle and scrub radius, a $heel $ith a moderate to large negati e offset is desired# For stiffness and $eight, the material $as chosen to be magnesium# Also, since Formula SAE cars are lighter than a t!pical full-scale racecar, $heels that are designed for full-scale cars $ill probabl! ha e ade)uate stiffness for Formula SAE application# &he lightest 4 inch $ide, ,- inch diameter $heel that $as readil! a ailable $ith a reasonable offset $as a standard Formula 1ontinental front $heel, $hich has a -2#+ inch offset#

$ngine %election and &reli inary Mass 'nalysis: Follo$ing the choice of tires and $heels, a preliminar! mass properties anal!sis $as conducted# &arget $eight distribution $as chosen to be :+;++ front;rear# &his decision $as based on pre ious testing and simple ehicle d!namics simulations, $hich both sho$ that a car $ith $eight biased to$ard the rear performs better in bra"ing and acceleration# 3oing be!ond ++8 rear $eight, ho$e er, ma"es achie ing a neutral balance in corners an unnecessar! challenge# %ith this goal, a target center of gra it! height of 7#7+ inches, and a total $eight of :70 lb, the anal!sis begins $ith the hea iest components. the engine and the dri er#

Drexel <ni ersit!

-,-

&eam =0:4

1ontinuing $ith the e olutionar! design philosoph!, the decision $as made to use a :-c!linder 400 cc motorc!cle engine as Drexel Racing has done in the past# >a"ing the assumption that proper design of engine peripherals $ill ma"e it possible to produce about the same tor)ue cur es on a number of different engines, engine selection becomes a mass and center of gra it! problem# Since the *onda 15R 400 F:i still maintains one of the lo$est cran"shaft positions 0$hich is approximatel! proportional to the height of the center of gra it!2 and are inexpensi e and readil! a ailable, the decision to use a 15R 400 F:i $as made# &o further reduce the height of the center of gra it! of the engine, a custom dr!-sump s!stem $as used, er! similar to past Drexel Racing designs# >ore detail on this s!stem $ill be pro ided later# Since the onl! $a! to modif! the contributions of the dri er to the height of the center of gra it! is to ad?ust the dri er's seating position, this became the next area of consideration# @n pre ious Drexel racecars, the dri er sat $ith his or her bac" ?ust slightl! off of ertical# @n an effort to reduce the height of the center of gra it! to meet the design goal, DR07 features a much more reclined seating position, $ith the dri er's bac" at :+ degrees, cantile ered o er the engine head# &his also helps "eep the dri er's center of gra it! close to the engine, reducing the moment of inertia in !a$ and helping to achie e the target $eight distribution $hile "eeping the $heelbase short# Additional components that $ere placed and gi en preliminar! shapes and olumes at this time include the radiator and $ater pump, electronics ba!, and fuel tan"# &he conclusion of this anal!sis $as that the design targets are reasonable and ma! be used for design of other s!stems, $hich $as expected based on past experience#

(asic )ehicle Di ensions and %uspension Design: 3i en that the ehicle is to be designed for an autocross st!le course, the si(e of the ehicle becomes a significant factor in maneu erabilit!, and ultimatel! in lap time# @n a four-cone slalom $ith cones spaced at 2+ feet, for e er! inch remo ed from the half-trac", the minimum path length is reduced b! as much as one foot# Although this ma! not seem significant, it is important to note that reducing the trac" also minimi(es the path length through other maneu ers as $ell# &his fact alone ma! not ma"e a significant difference in lap time, but it does ma"e the car easier to steer around the course $ithout hitting cones, and for the amateur dri ers that pilot Formula SAE cars, this could ma"e a er! significant difference# &he a erage trac" $idth $as chosen based on the maximum expected lateral acceleration 0based on pre iousl! recorded data A ,#7 g's2 and the target height of the center of gra it!# &o "eep the lateral $eight transfer due to these accelerations $ithin an acceptable range, an a erage trac" $idth of :,#+ inches $as chosen# &o aid $ith maneu erabilit! in slaloms and lane changes, the front trac" should be larger than the rear trac"# &he front and rear trac" $ere chosen to be, respecti el!, :-#2+ and :0 inches# >inimi(ing the $heelbase is also "e! to "eeping the o erall ehicle dimensions small# 5ased on a stud! of past Drexel racecars and the preliminar! mass placement anal!sis, the smallest $heelbase that $ill still allo$ for :+;++ $eight distribution is about 4+ inches# From the beginning, the decision $as made to e ol e, not re-design Drexel's race-pro en une)ual-length non-parallel double A-arm suspension# &esting has sho$n that lap times impro e $hen anti-roll bars are stiffened# &o compensate for the reduced roll angle, camber gain due to steering inputs $as slightl! reduced to pro ide the same d!namic camber desired in the corners# &his $as accomplished b! lessening caster and "ing-pin inclination# 5! re ie$ing trac" data and using custom >A&BA5 programs to anal!(e the motion of instant centers and roll centers as the racecar maneu ered around the trac", it $as determined that the instant center and roll center positions that $ere used in last !ear's design $ere er! $ell controlled laterall! and erticall!# As $as intended, the roll centers 0front and rear2 mo e up and do$n $ith the motion of the sprung mass to maintain a constant roll moment, and lateral roll center migration remained less than C0#020 inches# &his information pro ided the basis for decisions on front- ie$ irtual s$ing arm lengths, control arm lengths, and roll center positions# Front- ie$ irtual s$ing arm lengths $ere chosen to be 7+ inches in the front and 4+ inches in the rear, and the roll center heights $ere chosen to be ,#2D inches in the front and ,#:- inches in the rear# Final static "inematic ariables are gi en belo$, $ith the process b! $hich the! $ere obtained described belo$ them# *ront: E1amber gain, bump. 0#47 degree;inch E1amber gain, roll. 0#27 degree;degree E1aster angle. -#709 EFingpin angle. :#209 E>echanical trail. 0#06 EScrub radius. ,#+,6 E&rac". :-#2+6 ERoll 1enter. ,#2D6 abo e ground Rear: E1amber gain, bump. 0#77 degree;inch E1amber gain, roll. 0#-, degree;degree E1aster angle. +#D9 EFingpin angle. ,-#49 E>echanical trail. 0#+-26 EScrub radius. -,#+776 E&rac". :0#006 ERoll 1enter. ,#:-6 abo e ground

Design goals for the suspension also include minimi(ing scrub 0achie ed b! "eeping the roll center close the ground2 and maximi(ing the installation ratio for the dampers# Scrub minimi(ation is necessar! to maintain tire compliance and a oid upsetting the tire-ground interface# >aximi(ing the installation ratio for the dampers helps reduce h!steresis caused b! the internal dis"s to deflect $hen changing the direction of motion and ma"es small ariations in

Drexel <ni ersit!

-2-

&eam =0:4

damper force less noticeable at the tire# &o achie e all of the abo e suspension "inematics, an iterati e approach $as ta"en using Racing b! the Gumbers soft$are 0%in3eo-2# Another uni)ue feature of DR07's suspension is the decoupled spring;damper s!stem# 5! placing the spring next to the damper instead of using a traditional coil-o er design, the bending forces generated $hen compressing the spring no longer affect the performance of the damper# 5! allo$ing these bending loads to act on the damper, additional friction is added to the s!stem that increases the h!steresis of the damper# &esting has sho$n that remo ing the spring from the damper can reduce the h!steresis of the assembled suspension b! as much as ,+8# &he dampers used on DR07 are /ens"e 7700 series internall! ad?ustable dampers# &he decision to use these $as based on cost and performance# Drexel alread! o$ns se eral sets of these dampers and all of the components 0and some experience2 re)uired to ad?ust their damping characteristics# *a ing been er! satisfied $ith the performance and ad?ustabilit! of these dampers in the past, Drexel Racing $as happ! to be able to re-use extra components in lieu of purchasing ne$ dampers# &his also helps increase efficienc! at the trac", $here ha ing more than one set of dampers means one set can be rebuilt in the trailer $hile the test engineers are e aluating another set on course# &he damping rates are initiall! chosen to be slightl! less that criticall! damped 0for the sprung mass2 in ?ounce and about t$ice that in rebound# &hese choices are based on $hat has t!picall! $or"ed $ell in the past, but are li"el! to change once trac" testing has begun# &he springs used on DR07 are custom $ound, and range in stiffness from -00 to 700 lb;in# 5! putting the springs next to the dampers, the installation ratio for the springs is reduced to 0#+7# &his ma"es the ad?ustable range for the $heel rate ,00 to 240 lb;in 0sprung mass natural fre)uencies from approximatel! ,#7 to 2#7 *(2# &he decision to ha e springs custom $ound rather than purchasing commerciall! a ailable components $as based on the need to ha e relati el! stiff springs in a relati el! small si(e# &he $or"ing range for the springs is less than ,#-+ inches, so er! short springs 0H-#+ inches2 $ill $or", but are not readil! a ailable# Also, the cost of ha ing springs custom $ound is er! similar to the cost of bu!ing high )ualit! springs off-the-shelf# Roll control is bolstered b! the use of front and rear tubular anti-roll bars, connected to the bellcran"s ia drop lin"s# 1ontrol arms are fabricated from 0#42+ inch diameter 0#0:D inch $all :,-0 steel tubing# Spherical bearings are double an il s$aged at all ?oints except the lo$er ball ?oints, $hich feature spherical bearings retained $ith a snap ring for easier replacement# &hese ?oints are smaller, lighter, and stronger than similar si(ed rod ends# &he front and rear uprights are designed to be as light as possible $hile being stiff enough to ma"e deflections reasonable# @n the front, uprights are machined boxed 404,-&4 aluminum $ith bonded shear plates# @n the rear, the uprights are also machined from billet 404,-&4, and use a triangular pattern of cuts to reduce $eight $hile maintaining part stiffness# @n both the front and the rear, camber ad?ustment is designed in b! using shims bet$een the upright and a brac"et that bolts to the upper control arm and the tie-rod# &he steering s!stem incorporates a front-upper steering la!out $ith the rac" mounted directl! in front of the upper tub mounts# &his allo$s the steering lin"s to reside in the same front- ie$ plane as the upper control arms to minimi(e bump steer# 5ump steer is easil! ad?ustable through interchangeable bushings at the outer steering lin" ?oint and;or shimming of the steering rac" in relation to the frame# A splined )uic" release steering $heel is utili(ed to reduce steering compliance o er a hex-st!le )uic" release# @n order to decrease the steering effort re)uired b! the dri er $ithout sacrificing steering feedbac", an upright design featuring (ero mechanical trail $as emplo!ed# Ac"ermann $as designed using an iterati e process $ith a custom >A&BA5 program# &he optimal cur e is generated based on pea" front and rear slip angles 0determined from tire data to be 7#0 degrees2 and rear toe# &his cur e is compared to the actual cur e, $hich is a function of the steering geometr! and the front toe# Steering geometr! is t$ea"ed until the cur es match throughout the range of steering rac" tra el# Final designed Ac"ermann is ,408 and it is almost constant through the range of tra el, $ith onl! a slight progressi e trend# Ac"ermann is ad?ustable b! replacing the brac"et at the top of the front upright, or b! simpl! changing the hole through $hich the steering tie-rod is bolted in the same brac"et# &his is something that $ill be aried and tested during the test phase of the racing season#

&o+ertrain and Drivetrain Design: &he DR07 racecar is po$ered b! a naturall! aspirated 400cc *onda F:i, $hich Drexel Racing has de eloped o er the past fi e !ears# Iur engine s!stem design goals are focused on dri eabilit! and reliabilit!# Anal!sis of recorded data has sho$n that a $ell sorted car performs better than the one $ith the highest horsepo$er# &he highlights of the engine s!stem for this !ear include a tuned length inta"e manifold, custom dr!-sump s!stem, and a tuned :-2-, exhaust s!stem# A log-st!le inta"e manifold;plenum $ith a -020 cc olume and ,26 long x ,#006 @D inta"e runners $as fabricated out of aluminum# A con erging;di erging no((le, $hich houses the re)uired 20 mm air restrictor $as constructed out of 404,-&4 Aluminum# A -+mm butterfl! throttle bod! is mounted directl! upstream of the restrictor, $hich then transitions into the inner portion of the fabricated aluminum plenum# 5! using this design, the length of the di erging portion of the restrictor can be ad?usted during testing, $hile sta!ing $ithin the roll plane of the car, as mandated b! the rules#

Drexel <ni ersit!

---

&eam =0:4

&he engine management and fuel s!stem uses a stoc" *onda F:i fuel in?ection s!stem in combination $ith a D!noJet /o$er 1ommander /1-@@@r pigg!bac" s!stem# &his s!stem $as chosen in fa or of custom standalone s!stem for se eral reasons# First, the inexpensi e cost of the stoc" *onda fuel in?ection setup and the D!no-Jet s!stem better fits the o erall budget goal of the team and the competition# %ith the basic goal of the Formula SAE competition to construct cars that a $ee"end autocrosser could purchase and race, it is difficult to ?ustif! an expensi e and complicated engine management s!stem# 5! using the stoc" *onda F:i fuel in?ection s!stem there are additional benefits of ha ing pre-calculated fuel and ignition maps, including proper idle control, $hich can easil! be manipulated to $or" $ith the mandated restricted inta"e, $hile "eeping fuel econom! in the range of 20 liters;,00 "m 0,- mpg2# A dr!-sump s!stem is used on the DR07 to pre ent loss of oil pressure at high lateral accelerations, $here oil could flo$ a$a! from the pic"up tube, star ing the oiling s!stem# &he dr! sump s!stem uses a custom designed billet aluminum oil pan# &$o sca enge pumps po$ered off of the stoc" *onda $ater pump output shaft are used in con?unction $ith the internal stoc" *onda pressure oil pump to feed oil to the engine &his s!stem has been tested and pro en to be reliable b! tilting a running engine under load at -0, :+, and 40 degree angles to simulate 0#4, ,#0, and ,#7 3's respecti el!# &he stoc" mechanical $ater pump is replaced in the cooling s!stem b! a temperature dependant, flo$ ad?ustable electric $ater pump, $hich allo$s for greater control of engine $ater temperature# &he design of the dri etrain for DR07 centers on the integration of a chain dri e Salisbur! st!le differential into the rear chassis# /re ious cars ha e had chronic understeer problems that, through testing, ha e been attributed to the &orsen differentials $e ha e used in the past# &heir operation pro ides more po$er to the slo$er mo ing $heel, $hich is the inside $heel in a turn, causing understeer# %ith a more easil! ad?ustable Salisbur! differential the tor)ue bias and loc"up can be tuned in con?unction $ith the suspension rather than tuning the suspension to act as a crutch for the differential# &he center mount differential assembl! resides in aluminum pillo$ bloc"s $hich bolt directl! to a $elded steel bul"head attached to the rear box# &he pillo$ bloc" design of the rear differential allo$s for eas! chain ad?ustment b! the addition or remo al of shims bet$een the differential and the mounts on the frame# Axle halfshafts are fabricated out of high-strength Kasco>ax 1-+0 that has been gun-drilled and designed to eliminate excess mass in the dri e#

Driver ,ontrols, (ra-e %yste , and $rgono ics: &he rotating mass of the bra"e hat;rotor assembl! $as reduced b! 208 o er pre ious designs b! utili(ing a lighter rotor design along $ith a redesigned bra"e hat# &he bra"ing s!stem is comprised of t$o ,0#+-inch cast iron rotors in the front and a single D-inch cast iron rotor in the rear, $ith each rotor 5lanchard ground to 0#,76 thic"# &he rotors are machined from &!pe-:0 cast iron blan"s chosen for its high strength, thermal properties and frictional characteristics# &he single rear rotor, attached to the differential, reduces un-sprung $eight $ithout compromising rear clamping force# 5oth front and rear rotors are attached ia a dog-dri e bra"e hat, $hich allo$s the rotor to float axiall!# Dual piston %il$ood calipers are used front and rear due to their lo$ ,#: lb $eight, di erse pad selection, and ease of pac"aging $hich allo$s maximum rotor diameter# &he calipers are actuated b! independent master c!linders for the front and the rear, $ith a bias bar assembl! in the pedal for ad?ustment# &he entire pedal set is modular and has 4 inches of tra el fore and aft to accommodate a range of different dri ers# &he master c!linders are mounted belo$ the dri er's feet, $hich allo$ the o erall car length to be shortened fi e inches and further help lo$er the o erall 13 of the car#

*ra e Design: Drexel's 2004 ehicle chassis is a $elded tubular space frame that has been reinforced b! the inclusion of shear panels# &he space-frame has been constructed from A@S@ :,-0 steel $hile the shear panels are made from carbon fiber reinforced epoxies# &he frame is composed of t$o separable portionsL a front-frame that includes e er!thing from the front bul"head to the engine and a rear-box $hich lies behind the engine and houses dri etrain and rear suspension components# &he front-frame and rear-box are connected to each other and to the *onda F:i engine, $hich ser es as a semi-stressed member# &his configuration has been chosen for three reasons. 0,2 b! using the engine as a tension member steel tubes can be eliminated from the frame, 022 the remo able rear-box allo$s the motor to be easil! remo ed from the frame $hile "eeping frame tubes tight around the engine, and 0-2 the remo able rear-box is modular, allo$ing the next model's rear-box and dri etrain to be built and tested before the completion of the front-frame# &he first design goal for the frame $as to achie e a target torsional rigidit! of at least 2+00 ft-lbs;deg, and the second $as to reduce the $eight of the frame to ++ lbs# &o meet the $eight and stiffness targets for the frame, uncon entional design tools $ere emplo!ed# 1ustom soft$are tools $ere de eloped to pic" tube diameters, $all thic"nesses, and cross-section shapes using a 3enetic Search Algorithm# &he soft$are uses the AGSMS Finite Element Sol er to e aluate potential designs according to their stiffness-to-$eight ratio# &he algorithm $as run for approximatel! ,2 hours on -0 parallel computers, e aluating o er :+,000 potential frames, before con erging on a famil! of high-stiffness, lo$ $eight designs#

Drexel <ni ersit!

-:-

&eam =0:4

Drexel <ni ersit!

-+-

&eam =0:4

You might also like