FrankenFerrari - V12 Ferrari 308

In reply to Ian F (Forum Supporter) :

We tried to go to harisburg a couple weeks ago and same issue. 

Its been a WHILE since I've been to one, I had no idea they even could sell out before we hit register and were added to a wait list.  So this time we signed up a couple weeks early.

In reply to mke :

IIRC, the Franklin Mills event is limited due to paddock size. Warminster can get limited as well due to the size of the lot. I haven't been to an event at Pocono, so I'm not sure what that's like. 

Years ago when they ran at Citizens Bank Park, there was a ton of parking so the numbers were large - but they also needed a lot of participants to help pay for the lot fee, which was substantial - and then CBP more than doubled the fee which priced us right out.  It's a shame as those are great lots and the designers set up some really fun and interesting courses there. 

I went out the the shop and picked up a little, I need to get the engine set together and set into the car to get weights but I just wasn't feeling it 

Mentally I've decided I'm making new stub axles so I need to rear A-arm mounts.  I need a rotation, the bottom is easy but up top its welded.  The needs to move up/back, the rear down/back....or looking at it maybe just move the rear back  and let the front come up a bit more so I don't need to raise the rear as much and maybe also use a bolt-on fork like the lower for the toe adjust...that choice would make at least 1 person happy I believe 

Also I'm thinking about the rear roll high and how high it is compared to the front in the factory setup...F25, R103 is a huge difference and and cause the front outer corner to drip quite a bit.  Also I remember really struggling with rear grip coming out of the corners, the height roll center doesn't help that either.  So my new plan is leave the rear roll center right where it is down around 53mm, at least for now.

Got an email for head gasket guy, they are in production!  I guess the issue was he wasn't sure how to do the sealant silk screening but he came up with something and gaskets are a go!

I've been playing around a little with and online program, Vsusp....the best I can tell the 308 suspension was always a hot mess.  Up front looks fine at rest

But it got super soft spring so hard brake and you have this

Then try to throw it into a corner and WFT

This basically says its going to want to plow like a mother because the rear roll center is up near 4" so basically all the corning load is on the outer front tire.....and I recall my first autoX with the car total stock there was no braking into the corner allowed.  

The car at its current ride height I have this

the same as stock with brakes, hot mess

Adjusting the bolt-on lower A-arm mounts frame mounts  up about 16mm (there is a kit available from super performance) gets it here

then brakes with about 60% anti-dive also baked into the lower A-arm mounts and roll controlled with stiffer springs and swat bar gets it here.

not great but okish.

In the rear it starts with a roll center up at 103mm, so when its lowered that actually makes the roll center much more rational

then when you brake (with anti-squat/lift) and corner nothing very bad happens

I want to play around a little more before I start cutting, particularly to see what can be done about the front roll center migration (moving right/left in corners) but I think its workable as is and I could very skip the rear anti squat and not be too bad.

I did get the engine kind of slapped together so I can set it in for a weight....hopefully this week.

I found a lot of generally helpful info on this web site.  They do a nice job explaining suspension stuff.

Suspension secrets

House project is stuck on color choice so thinking about the car's stub axles again.

So a 308 uses a 40mmID outer bearing and it looks like modial8 uses the same axles...they are well known to break with race use and I know for sure have also broken in autoX use.

A 348 axles uses a 45mm ID bearing and early cars the same CV joint as 308 so are a possible upgrade option. Its a nice looking part but I'm not sure how hard to get or make fit.  I had a guy offer me a set in exchange for making honda k20 heads fit a 308 block...we'll see if that goes anywhere 

328 and 3.2Mondial use a different axle from 308 and 348....but I don't know what size the outer bearing size is   I don't ever recall hearing about broken 328 stub axles so I'm guessing 45mm but would love to know for sure if these are an upgraded part.

I found an OLD fchat thread from 2001 that was talking about 308 stub axles and included a 328 that lost a wheel at the track and blamed a poor design with inadequate radiusing and no heat treat, apparently fixed on the 348.  I saw a couple 348 left side wheel carrier assemblies on ebay for around $1000-$1300, no rights to go with.  It will likely be $2.5-3k if I go down this path.  I'll likely end up making stub axles....

I know this suspension/stub axles is a little ping-pongy but I don’t want to fix the suspension and then change the wheel carriers and need to fix it again.  The 348 stuff is a pretty pricey so kind of refocused on making stub axles and I have confused myself a bit.  The 308 uses a 40mm (1.575”) ID bearing, the 328 and 348 a 45mm (1.772”) ID bearings but the 328s still have issues, maybe material, maybe design, probably both. 

My plan was grab grab some US diff axles to use as blanks and start cutting knowing they rarely break and ASSUMING that was because they are big and heavy…wrong.  The big bearing ford 9” and all the mopar use a 1.564” ID bearing, not quite 40mm.  The small bearing 9” and a lot of GM stuff use 1.53”.  There are some GM trucks with 1.618”…WTF?  These things don’t break without REALLY trying and they are smaller?

Design wise the US axles all have a pretty decent radius between the wheel flange and bearing surface and that must be where the magic happens.  For sure the US axles are heat treated and the 308 are not for whatever reason.  The ferrari stubs have a nut torqued on to hold the CV flange, that is an extra force I guess.....but it feels like a slight design change will sort the issue.  Just not certain the off the shelf US axles have enough metal where I need it, I think so?  I probably need to pull mine apart and get some measurements so I can be strategic about what I order.

In reply to mke :

What alloy are the US  Axle blanks? Are the failures torque or cornering load generated? We used the 308 wheel bearing on the Shannon CSR and had plenty of bearing failures, but through 9 years of wins and crashes never hurt a stub axle. Full crush torque on the bearing races. Less weight and torque, but it made 2g steady state...Axles made in house by Shannon, 4340 by their spec., and by memory had an fillet radius nearly as great as the generous one on the bearing.

In reply to TurnerX19 :

Help me understand what you were running....308 bearing, in a 308 wheel carrier with a custom 4340 stub axle?

From what I've found US (ford, gm, mopar) are mostly 1040, some 1050 heat treated with a Tensile around  120kpsi. Replacements are mostly 1541H with tesile of 180kpsi.  4340/300M are available with Tensile around 220kpsi.  The 1541h I guess also hardeness deeper than 1040 and 4340 is though hardening, which has to help some.

The OEM 308 part has sharp corners and is rc20...probably as machined 1040?  The failures are fatigue in cornering.  Street cars never have a problem even boosted high hp as far as i know but a couple guys i know lost wheels on stockish hp cars they track and or autox regularly.  The 308 race cars have the 45mm outer bearing and are about rc32, nice big radius to match the bearing.  328 45mm with square corners, 348 45mm big radius.

My though is see if the 45mm bearing will fit in a 308 carrier and at least a 1541H, but I may pay up a bit more for 4130 or 4340 if I can find something useable...many already have multiple bolt patterns  so hard to add mine vs the 1541h are single pattern oem replacement so easier to make work.

You read the design correctly. I did not design it. The bearing choice was made originally by Eddie Jones, possibly Eddie Sr. At the time ZF would not sell me the bearing by number, they said must buy from Ferrari. Used them front and rear with inside out front stub axles. We had frequent rear bearing failures that were cured by popping the seal, cleaning and replacing the grease with Mobil 1 .Went from 1 race to 10!!  Pic of me in the car at Pocono, 1999.

In reply to TurnerX19 :

That is very helpful, thank you!

In reply to TurnerX19 :

Goes to show, ALWAYS replace the grease in an off-the-shelf bearing. I went from replacing rear wheel bearings basically every year in the S2000 to 5+ years on this set when replaced with Redline CV2

I have a big bearing 9" in my other RX-7.  31 spline axles by Dutchman.  Nice big radii from the flange to the bearing 2.5" away from the hub face.

The bearings last about two events before they are noisy, no matter what brand I buy (had opportunity to try many different brands...)   Given the amount of pad knockback, they must flex a lot, too.

Or maybe I am just pathologically hard on parts 

Dutchman will make an axle to any length and any bolt pattern you want, as long as you provide the dimensions, which is why I went with them. They made axles to take a 4 on 4 1/2 pattern with a Mazda pilot.  

In reply to Pete. (l33t FS) :

Yes they flex, but that radius keeps them alive. "Our" wheel bearing is a double row angular contact ball set with large OD by the standards of my experience. They are larger than any other "S2000" type sports racer. Most other chassis used the Audi 5000 bearing. Those folks had less trouble than we did, but replacement was a service item on a Lola or Swift.

I sent a note to strange engineering who make nice axles to see if they have interest in quoting my stub axles or if I can buy a  partially finished set to at least save the aneal step.  It looks like they like 1550 for street and induction harden to rc58-62.  They have non-chrome alloy they use for drag only that through hardens to rc45-48 and 240kpsi.  

+1 on Dutchman. I'm happy with what they made for my MG. 

In reply to mke :

I might be reading this wrong, but it looks like you might have a positive camber curve in the front. Ie, camber goes positive on compression. 

I mention this because it seems to me like an overly high rear roll center relative to COG should act like a stiff rear roll bar and, thus, make your car a hot mess of oversteer rather than a hot mess of understeer on trail brake. The front roll center also seems very low, which should in theory act like a very soft swaybar. 

I could of course be misreading, but if I'm not, I would think you've basically got a "race" between two handling disasters. The disaster that seems to be "winning" is the positive camber curve in the front suspension, which washes you out into understeer. The disaster that would seemingly manifest if you fixed the front camber curve, is a roll center situation in which the rear gets increasingly stiff in roll (rising RC) and the front gets increasingly soft (falling RC) while cornering, and thus the car gets increasingly oversteery and might even start to do the stinkbug, ass up thing the higher the overall roll angle gets. 

In reply to DaewooOfDeath :

Here is baseline, camber is -0.981 and I reset the graph to be left bump v left camber

Here is 1" left bump, camber is now -2.03 so it has camber gain on bump

Here it is with 2 deg roll, left camber is now -0.056, so its loosing camber at 1/2 the rate the car rolls....not loosing any or even gaining would be ideal I guess but the geometry is working in a helpful direction at least.  I changed the graph to roll deg v left camber.

The whole roll center height vs how roll bars work honestly has been baking my brain a bit.  In the one video I post he talks about want the rear roll centers higher than the front and says that creates a roll axis that points down toward the front so that when the car rolls, it rolls onto the outer front tire and that makes it naturally under steer.  That makes sense to me, roll onto the the outer front you overload it, and understeer. 

Comparing that to bars though, its kind of the opposite?  raising the REAR roll height is like a heavy FRONT bar as both over load the outer front tire if I'm understanding this right...and I may not be.   My first thought was exactly what you're saying, high roll center the same as a big bar on the same end....but a bar doesn't change the roll axis so I've now convinced myself that its the opposite.

I've also kind of convinced myself that the whole idea of the car having inherent understeer base on roll heights is a time suck at the autoX.  I'm thinking the goal is to get much closer to neutral with the roll centers then then use the bars to tune so 1 set of tuning isn't fighting the other.  This might be misguided but in my head its making sense.

In reply to mke :

Your last paragraph is on point. This is the direction most have been moving towards for decades. Start neutral and adjust with bars.

mke said:

In reply to DaewooOfDeath :

Here is baseline, camber is -0.981 and I reset the graph to be left bump v left camber

Here is 1" left bump, camber is now -2.03 so it has camber gain on bump

Here it is with 2 deg roll, left camber is now -0.056, so its loosing camber at 1/2 the rate the car rolls....not loosing any or even gaining would be ideal I guess but the geometry is working in a helpful direction at least.  I changed the graph to roll deg v left camber.

The whole roll center height vs how roll bars work honestly has been baking my brain a bit.  In the one video I post he talks about want the rear roll centers higher than the front and says that creates a roll axis that points down toward the front so that when the car rolls, it rolls onto the outer front tire and that makes it naturally under steer.  That makes sense to me, roll onto the the outer front you overload it, and understeer. 

Comparing that to bars though, its kind of the opposite?  raising the REAR roll height is like a heavy FRONT bar as both over load the outer front tire if I'm understanding this right...and I may not be.   My first thought was exactly what you're saying, high roll center the same as a big bar on the same end....but a bar doesn't change the roll axis so I've now convinced myself that its the opposite.

I've also kind of convinced myself that the whole idea of the car having inherent understeer base on roll heights is a time suck at the autoX.  I'm thinking the goal is to get much closer to neutral with the roll centers then then use the bars to tune so 1 set of tuning isn't fighting the other.  This might be misguided but in my head its making sense.

I think that might be a good idea, and it's relatively close to what I ended up doing with the munchkin-made Daewoo abomination, but there are two ways to get "neutral" roll centers. 

The first is to lower the rear, which seems to be what you're starting on. This will require bigger bars. The second is to raise the front, which will not require bigger bars. 

In reply to DaewooOfDeath :

I'm thinking some of each.  When I lowered the car it dropped both roll centers significantly so I'm thinking maybe leave the rear right about where it ended up, and then get the front back up into positive territory at least.  That will still leave a little rear bias but way less than it had.

I was looking at bars I thought I could useat.. Coleman?  I remember LOTS of options.

mke said:

In reply to DaewooOfDeath :

I'm thinking some of each.  When I lowered the car it dropped both roll centers significantly so I'm thinking maybe leave the rear right about where it ended up, and then get the front back up into positive territory at least.  That will still leave a little rear bias but way less than it had.

I was looking at bars I thought I could useat.. Coleman?  I remember LOTS of options.

I've been happiest when I don't use bars at all, but that's a whole can of worms. ;)

DaewooOfDeath said:

I've been happiest when I don't use bars at all, but that's a whole can of worms. ;)

Years ago on an FSAE car I convinced myself the bar were unnecessary and designed the shock linkage with an adjustable and aggressive progressive rate.   I'm not sure if it was a good idea, we didn't win, but did well, 17/120? overall, top 10 in driving events? The car drove good, very minimal roll but no real idea if my linkage was helping.

The springs I have in the 308 are pretty stiff, 2.5hz iirc?  I removed the rear bar years ago and it's the stock front bar with is pretty light, 0.70"?.  The "upgrade" is usually 1", but I've seen 1.125 also.... but they are usually running spring rates about 1/2 what I settled on so I am on the more spring less bar end of the spectrum....again that to say I've tested and this is faster, it just guess work really.

In reply to mke :

When I was making the Daewoo abomination, I read a series of setup articles by Steve Hoelscher on his no-bars X19 and MR2 autocross cars. He makes the claim that swaybars convert roll into weight transfer, though every math/physics person I've posed this question to has given me a different answer. 

Regardless, he won a cajillion titles with those cars so I thought it would at least be worth a try. I basically took the spring rate and roll center calculations he used on the MR2 and turned them backwards for the Daewoo - FWD, all struts Daewoo vs MR all struts MR2 - since an FWD car is basically an mid engined car driving in reverse. Or, in your case, your car can be interpretted as a V12 Civic driving in reverse. 

I don't have the equipment to tell you if this resulted in less weight transfer but I can tell you with 100% certainty that even with high roll centers (roughly the middle of the oil pan in front, about 4 cm higher than that in the back) this setup required quite stiff springs. It did end up handling way better than it had any right to, and better than a lot of dedicated performance cars I've tried before. It also made the suspension truly independent at all four wheels, and it took bumpy corners like a demon. 

I did a similar setup on a 2002 Hyundai Elantra to similar effect, and a 2004 Tuscani which was too softly sprung to really work.  I've never tried it on anything RWD, though it's on my to-do list. I'm not sure you'd see a big difference, considering how much spring and how little bar you already have.