I’ve always noticed how race car tires have a very different attitude than street car tires. They are often wider and sit under beautiful fender flares. But they also seem to have a bit more tilt in them where the bottom of the tire sticks out further than the top. I’ve lately discovered this to be called “negative camber”. Camber is the tilt of the wheel around an imaginary axis line that would run either between the two front wheel centers or the rear wheel centers. When the top of the wheel moves in toward the center line of the car and the bottoms move outward, that is “negative camber”. The opposite is positive camber.
Why does camber matter? When a car corners, the car puts its weight onto the two outer tires and rolls somewhat toward the outside of the turn. If a tire starts off being completely vertical and then goes into the corner, the roll of the car will cause the top of the wheel to tilt outward and the tire will be up on its outer shoulder instead of still being flat on the road. Less tread (the contact patch) is touching the road. And grip is reduced. By tilting the tire inward, the contact patch will be tilted a little when the car is going straight (riding on the inner shoulders of the tire) but will get more flat as it goes into a corner – providing more cornering grip. Up to a point.
The thing I noticed recently was that I am beating up the outer shoulder of my tires with all of the autocross and track time I’ve done this season. Clearly I’m getting up onto those outer shoulders in turns and probably losing some traction/grip. I’m also wearing those shoulders out faster than the inside of the tires. So I’d like to get more even wear on my tires as well as a side benefit.
It’s a tough balance as too much camber is fairly bad for a street setup. And too little camber is tough at the track. There is a push-pull situation that I’ll have to work out over time.
My 2007 911 Targa 4 (997) is a very good car already. The handling and cornering are already well above most street cars. But the stock suspension is a bit limited. The front setup limits negative camber to about 1 degree (-1°). The rears can actually go a bit more – maybe -1.5° or so. I’ve read that having more than 0.5° difference between front and rear can create some instability so that’s one thing I’m keeping in mind. I’d like to max out the rear camber to -1.5° and take the fronts to -2.0°. Maybe -2.5° at some point if it feels good at -2. [edit: post alignment, I kept the -2.5° I put into the driver front and evened it out with passenger front. And I took the rears up to -2° to try to balance it out. I made the driver front too long initially but wasn’t going to pull it back apart to shorten it for more adjustability, but I’ll have to at some point. It felt pretty good on track, though, with some residual understeer but no surprises and pretty controllable. Better to have some understeer than too much unpredictable oversteer, for sure.]
But getting there requires a bit of work. There are a couple of ways to attempt to get more camber. The normal, easy, and inexpensive way is camber plates. These sit up on top of the shock/strut tower and tilt the strut inward or outward just a bit to change camber. I would have loved to go this route, but because the space in which these plates fit is small, they are limited as well in how much camber change they can effect. Most estimates seem to say that they could only add 0.5 to 1 degree. I didn’t want to spend a chunk of money and time for installation only to find out that I didn’t get enough adjustability.
The harder route, but common for Porsche’s, is changing the lower control arms (“coffins” – nicknamed due to their coffin-like shape, no doubt, and abbreviated as LCA).
The GT3 and Cup cars came with adjustable lower control arms that allowed for the more radical camber setups seen on the race cars. You can swap out the stock items on any 997 for the GT3 or Cup lower control arms. You get adjustability from them. But I also found that they require a shimming process to change the settings. I’m sure it works fine and has good repeatability since the shims are fixed sizes. But I wanted something a bit more fluid. RSS makes some fine units. But they have shims as well. I found the Elephant Racing lower control arms to be fairly nice units with simple turn-a-bolt adjustability. The Elephant Racing LCAs also come with full rubber boots over the inner monoball and outer tie rod monoball ends. Driving here in New England, even avoiding the depths of winter, will stress any exposed metal – especially moving parts like these. Race cars that get cleaned after each race, don’t see snow ever, and get torn down frequently can avoid these protections. My car cannot. Having to pay a little extra for the little bracket for my self-leveling headlights was a little tough to swallow, but in retrospect, I’m ok with it. The RSS LCAs include the bracket gratis.
Swapping out major suspension components isn’t something I do too often, so this was going to be a learning experience. In trying to find some online help, I didn’t see much. My books talk about swapping out these parts but don’t say how. One online post included all the torque specs for all of the bolts in the whole 997 suspension assembly, some of which I’d need, but was too confusing since I didn’t yet know all the terminology. I found one written tutorial with few pictures about changing the control arms on a Boxster due to torn tie rod end bushings. Close enough for me to get rolling. They said it would be easy to do this job. I took that with a grain of salt since just about every car thing I do take twice or three times as long as I think it should. I always hit a stuck bolt, rusted piece, or some other unexpected obstacle that presents a major challenge. Doing a simple brake rotor swap on a Cayenne, I once had to blowtorch bolts and drill out screws to perform a job that should have been fairly simple. Corrosion is a mean enemy here in the rust belt.
I stocked up on a few tools I figured I’d need – some extra metric sockets, a “pickle fork” (tie rod end separator – but I got the wrong size…), more rubber gloves, and an air socket wrench that I didn’t need per se but who doesn’t love the sound of air tools?!
First order of the day was removing the plastic underbody tray. Its a nice piece that has vents for the front differential and ducts to send cooling air to the front brakes. Looked easy to remove. It wasn’t turns out it underlaps another big underbody tray for the middle of the car which wasn’t worth the effort to remove. I let it hang and moved it a little when needed and it turned out ok.
I thought about taking off the other small plastic panels attached to the front locating arms that turn the air onto the front brakes, but they also weren’t really in the way much so they stayed. Following the Boxster process post that I read, I undid the chassis nut but left the bolt. Fairly easy.
Next up was the bracket for the “Litronic” (high-intensity-discharge) headlight self-leveling system. One simple bolt there so easy too. Tricky to get in there with clearances but not too hard.
Then came the tie rod end. Having done the steering tie rod ends on the 914 recently, I knew these could be tricky. They have a conical machined part that fits up into the conical housing. And these two pieces really stick together for some reason. You can bang away and these won’t easily come loose. Porsche makes a special separator tool for this job. I hate specialty tools as I don’t run an auto repair shop. So I got a pickle fork instead. Its like a two-pronged fork but each tine starts very narrow and becomes very thick where it meets the handle – hence how it separates two things. I knew it might destroy the rubber boots, but in this case I didn’t care so much. I wasn’t going back. The pickle fork I bought turned out to be too small. The inside of the fork didn’t fit around the shaft. I rented another from AutoZone that was a better size. Too big actually as it contacted the LCA on removal and did some surprising damage to it. Again, didn’t care as I wasn’t turning back. Getting the pickle fork to do its job was still very frustrating. That tie rod did not want to let go at all. Access to the side of the joint is tight in the wheel wells so you can’t get a good shot at the fork with a hammer. The big pickle fork I had was so long that I was barely able to get a blow on it at all. The angle of the fork I had to use to get a hammer on it was part of why it eventually contacted the old LCA and chewed off some of the ribbing bits during removal. I wasn’t happy about that and vowed to be a bit cleaner on the other side, knowing what I learned from this side. I got it apart eventually with some perseverance and some hard blows with a hammer on the fork. When it goes, you know it immediately as the wheel carrier drops down. What a relief when it finally came apart. There was some advice online to use a jack to raise up the wheel carrier as you try to push the tie rod down. I didn’t find it made a difference. I thought the worst was over.
Last thing to do is loosen up the easily accessible vertical bolt holding the locator arm in place. My new Elephant Racing LCA’s have a nice rotating connector for this bolt inside that spins around inside the LCA adding the ability to change caster angle (the angle the strut leans forward or backward relative to the direction of travel – think of a shopping cart wheel). The 997 doesn’t really allow for caster changes in a stock setup. But the wrinkle here was that I had to figure out where to set them up initially. I just aimed to replicate the stock setting as much as I could by eyeball. I’d be taking it to a shop I love – Roger’s Tires in Woonsocket, RI – for an alignment after it was all done. But I wanted to be able to drive it to the shop without too much trouble so I wanted it to be in the right ballpark. I’ll be shooting for about 8° of caster angle during the alignment – about the same as stock. Too much caster probably slows your steering down a bit but more caster also provides high-speed stability (like on a shopping cart, sort of).
After getting the two arms out, you can see them side by side.
You can also see how destroyed the boot on the tie rod end of the old LCA became. And just above there you can see where the pickle fork damaged the cast support fins on the old LCA. The new one is machined from billet and should be plenty strong. Even the GT3/Cup arms don’t have caster so that is a nice benefit.
Getting the new arms in looked like a piece of cake. Of course, I didn’t do a good job getting the install length sorted well before the first arm went in. Added negative camber right from the start! I wan’t going to pull it back out because it was not a piece of cake to get them in. The size of the new chassis mount portion was perfect – but too perfect. The fit was so tight and with the monoball connection now in place there, it was a serious pain in the butt to get it into the right spot to get the bolt through it. I kept having to gently nudge it with a long metal bar and a hammer. Only to overshoot at times and realize that I couldn’t hammer it back the other way. I found I could put in the locator arm bolt and hit that a bit to nudge it back. Good save there. I finally got the threaded part of the bolt into place but the main shaft of the bolt is a bit thicker and wouldn’t allow it through. Fed up at this point, I used the too-small but shorter pickle fork and a hammer to tap the bolt into place, straightening out the LCA in the process. It was hella difficult. From there things went together smoothly.
But I knew I had the other side to do too. I learned a bit from side one. I filed down the chassis mount points a little tiny bit on side two to get the LCA into place a little before hammering it in. Then I found a 16mm crescent wrench could be used to move the LCA end into place a bit more easily than hammering on it. This side went way faster and less frustratingly than side one. Chalk up another big lesson.
Once I got everything installed, I could see that the toe settings (how much the front of the tires relative to the direction of travel point in toward the center of the car or out away from it – think pigeon toed) were seriously toed-inward. You generally want a tiny amount of toe-in on the back tires to maintain stability under hard braking. The rear gets light as the weight shifts forward under breaking and having the tires pointed at each other a tiny bit helps keep the car pointed straight. Extra toe-in is normal for lowered cars and totally expected when the LCA effectively gets longer. But for a while I was stymied on how to correct it at least in a rudimentary fashion. I was playing with the caster adjustment for a while until I realized that it wasn’t going to help out. Finally I realized that I had to mess with the steering tie rods for this one. I wasn’t planning on touching them at all, but I had no real choice. I ended up having to add about 5/8″ to each side to get things at least visually ok. Yet another thing for my alignment shop to dial in later.
I took it out for a road test and everything seemed pretty much ok. Off to alignment shortly and all will again be right with the world – and hopefully with some added front negative camber to boot!
[edit: Post alignment specs: -2.5° up front camber, 0 toe, 8′ caster left, 9′ caster right (hard to get even). -2° neg camber rears. 0.1 toe in (not sure of units there… will have to look up). Stock caster on rears. Track results were solid. Some understeer but controllable. Might have too-stiff springs in back or might need to go more neg up front?]