Wakefield Park matsuri was the turning point for the car, for a few events i had been running tiny amounts of caster to eliminate as much bind as possible, it made the car a pig to drive, but it was binding a little bit less worse than before, so it was actually driveable on full lock. I had barely any time to prep the car for the weekend as i’d just returned from Japan, so the car went pretty rough, without a proper check over. By the first day, half way through qualifying the front end was literally shaking apart, quite a few of the nuts and bolts came loose making the car near impossible to drive during the final stages of qualifying for both me and Simon who was borrowing the car. After getting the car to the pits and checking it out, almost everything was loose under there, and although it was tightened, by the 2nd day the front suspension was once again doing some weird things.
Without enough time to check under the car, i kept driving, until it just wouldn’t hold an entry right at the end of the day, it sent me off the track a few times, but it was one which sent me off, and back onto an access road which did a lot of damage, and finished my weekend.
After assessing the damage, it became clear that the caster rod bracket had come loose from the chassis, and was floating around, eventually after taking a big hit, it took the rad support with it which is the only other mounting point for the bracket. The rad support is held on with spot welds and is not designed to take anywhere near that kind of force, so it failed, taking my radiator with it, destroying a lot of my suspension and steering as well.
At this point, i decided the car needed a full rebuild at the front, the dampers had been completely stuffed for a long time, the rad support and rails at the front were a mess, the crossmemeber was ripped apart, the lca was shaped like a bannana, along with the tie rods and the crazy strong Cazman caster rods. There wasn’t much to salvage, so i decided to go from scratch.
Macpherson strut, based on the s-chassis became a better and better idea, it is simple, and easy to play around with, and the knuckles you can borrow from the s-chassis are easy to weld and use.
One of the biggest problems with macpherson strut front ends in R32’s, as it is certainly not a new concept, is the position of the strut tower, due to the fact the tower is not designed for the coilover to be an integral part of the geometry, it is positioned much further forward than the towers in the mac strut chassis’. This is generally acceptable in most applications and you position the new top as far back in the tower as possible, then dial the caster in with the tension rod, which pulls the lower balljoint forward, tilting the axis over and creating caster.
The issue with this, is the same problem i had with my old setup, a steering setup which is suceptible to bind, will get worse with your typical caster increase. This has nothing to do with caster itself, it has to do with pulling the lower outer pivot point, (along with the upright) forward in relation to the steering rack.
In this pic you can see the blue and purple lines represent tension/caster rods, the blue rod is shorter, pulling the lower control arm forward, a long with the upright/knuckle/wheel, creating higher caster. The picture shows how the outer pivot point of the tie rod which attaches to the knuckle, is able to move a lot sooner past the line drawn between the inner tie rod pivot point at the rack, and the outer ball joint on the lower control arm, even with the same amount of lock, which creates bind.
This is why you have the common modification of the offset rack spacer, which moves the inner tie rod pivot point forward, or the much better solution of cutting and shutting the crossmember to move the whole rack forward which essentially does the same thing without the issues associated with offset rack spacers, however it will require a custom steering column spacer. This gets you half way there, another solution is to move the whole crossmember forward, you will need custom engine mounts and make sure there is clearance from your sump, as well as an even bigger column spacer, this is the most extensive solution, but the best (along with the cutting and shutting of the rack mount together).
All these solutions deal with the inner tie rod pivot point, but there is also something you can do which will allow you to pull the upright further back, without losing all your caster. To do this, you have to move the upper mounting point of the steering axis, which for the macpherson strut, is the strut top’s bearing, ie. move the strut tower back even further. This is where the fundementals of my new design came from, the plan was to start from scratch and create a new strut tower that allowed the strut to sit further back, which increased the caster angle hugely, without the trade off of pulling the lower point further forward which would increase bind. In fact if i moved the strut tower far enough backwards, i could run a very healthy caster figure of around 9-11 degrees, with the upright even further back than standard. To give you an idea, standard setup runs around 5 degrees. More caster, less bind, it’s a win win.
At this point, it’s probably worth mentioning the difference between caster angle, caster trail, and SAI (steering axis inclination) as they are extremely important in a large lock setup. Caster angle is the angle looking at the car from the side, between the 2 points of pivot concerning the steering; the upper bearing in the top of the strut; and the ball joint in the lower control arm, as shown in the image above. The caster trail is distance between where the centre of the tyre touches the ground, and an imaginary point on the road that is made by following this caster angle through the 2 pivot points all the way to the road (this is illustrated in images below). The SAI is whilst looking head on to the car, the difference between the angle of the steering axis between the 2 pivot points, and true vertical. the angle from true vertical to the camber angle of the wheel, is your included angle.
I feel like i’m watching an vhs back in high school with this vid, but gives you a good idea of how SAI and caster angle works together.
this one deals with caster angle/trail, the effect they describe is the effect caster trail has.
All 3 of these contribute to the self centreing effects of the steering, but by different methods, the caster angle and SAI work together to create camber gains with steering angle, they tilt the stub axle angle down towards the ground, effectively pushing the tyre down relative to the car, lifting the car. Naturally there is going to be resistance to lifting over a tonne of vehicle, so it will want to sit at centre, where the tyres are their flattest.
Caster trail works much different, by putting the steering axis in front of where the wheel actually intersects the road, the steering almost drags the wheels behind it, you can see this effect on your shopping trolleys which rely on caster trail to keep the wheels straight when pushing it along. Caster trail is created mostly by the caster angle, but can be introduced separate from the position of the stub axle.
You can see the caster trail is the distance between the 2 purple lines, one being the wheel centre, where the wheel contacts the ground, the other is where the imaginary line created by the caster angle (yellow) intersects with the ground in front of the wheel, the further in front of the wheel centre this line intersects, the more caster trail there is, and the more centering effect you get this way. In this example, number 1 has the least amount of caster angle and caster trail, the 2nd one uses increased caster angle by leaning the strut over further, which in turn creates more caster trail, the final example (3) has the same caster angle, but the stub axle is moved backwards on the upright, this has a dramatic effect on the caster trail without touching the caster angle at all.
The self centreing effect is important to any drift car, as you want the wheel to self correct as much as possible a lot of the time, you should never rely on a car to do all the work, but having a car where the wheel spins back on transitions nicely is always good.
The question is, which way do you go, if you look at the wisefab kits, they run low amounts of caster angle, and modified their knuckles with the stub axles backwards of their normal position, this introduces caster trail to offset the lack of centring effect with a low caster angle setup. I went the other way, and both ways have their merit, so i will talk about wisefab’s way first.
Caster angle, and the camber gains that come with it through the corner, work great with normal grip driving, as the way it is set normally, positive caster increases negative camber on the trailing wheel (the outside wheel on the corner) and decreases negative camber on the leading wheel (the inner wheel). This works as the outside wheel is usually loaded up more, and the car is leaning that way, so the increased camber has a positive effect. The problem with drift, is to go around the same corner, the geometry is flipped as you are at opposite lock, the loaded (outside wheel) is pointing in the opposite direction of what it was when it was just gripping, so now you are getting negative gain on the unloaded wheel, and less negative (to the point of positive camber) on the loaded wheel. Not only is the geometry flipped, it’s exaggerated due to the fact that with drift you typically have much more steering angle than when gripping, and since caster angle increases the camber gain more and more with steering angle, the effect is much higher.
For front grip, this is the opposite of what you want, but at the same time this camber gain is what is self centring the wheel, and creating good turn in. Since Wisefab’s kit increases caster trail giving you a similar self centering effect to SAI, without the camber gains, it means mid corner the wheels stay much more flat.
Here’s where more distinctions need to be made, lock is great, but it comes down to what you’re planning on using it for, lock is nothing without a car which has the chassis and mechanical grip balance to allow for big angle, but even with amazing balance to allow for angle, anything past the amount of angle that your average good cut and shut knuckle setup will allow you – is too much angle to be carrying mid-corner in my opinion, it looks cool for photos, but is slow as shit, and takes no skill to pull off. The issue with big angle mid corner, is your rear wheels are fighting your fronts too much, your rear tyres are pushing you to the inside of the corner, your front tyres are keeping you going to the exit, since you are relying on your rear tyres for speed mid corner, you cannot maintain good speed, at least not what i consider good.
For me, extreme lock is for one thing, and for one thing only, big fuck off entries, where you are using the momentum from the speed before the corner, to allow you to carry a lot of speed into the corner, but at large amounts of angle. Since you are no longer relying purely on your rear tyres for speed, you can go big without the same sacrifice of speed.
The thing about big entries, is that even with bulk amounts of lock, you never have enough for a proper 90 degree or backwards entry, so your fronts have to slide. The reason lock helps is because it allows you to get on the throttle earlier at that point where your front tyres are pointing in the direction you want to go, with more lock that point will be sooner, so you can get on the throttle sooner which not only looks good, but means you don’t lose as much speed. If the front tyres are still pointing back where you came from, you’re just going to loop it when you get on the throttle.
But back to the front tyres sliding, this needs to be done in a certain way, a car will always hold more angle with less grip at the front, as when a car is sliding nuetrally with no throttle or outside influence, if you have more grip on the front, the rear will want to keep moving the direction of momentum and the front will want to resist the way of momentum more which creates more angle and eventually a spin. if you have more grip on the rear, the rear wants to stop moving with momentum, and the fronts will want to keep sliding forward, this creates a handling balance which accommodates angle. It’s fairly basic, but not only that, the point where your front tyres grip back up after sliding past 60 degrees or whatever amount of lock you have, needs to be a smooth transition, the less grip you have at the front, the easier the transition will be. This transition is one of the hardest parts about a big entry, and you can always see it when one isn’t done right, it looks like a huge correction.
All this means is that at big lock on big entries, less is more, less grip that is. And It just so happens we have a certain element of the steering’s geometry we can play with which creates a dynamic change of grip the higher the steering angle you go, it’s back to caster angle/sai and it’s camber gains. What wisefab must have considered a bad thing, low grip at high lock, i consider to be good, for big entries at least, for mid corner control it’s rubbish.
This is why for my setup, i am running healthy caster angle 9-11 degrees, with standard positioned stub axle. I may introduce more caster trail later via an extensively modified knuckle to move the stub axle back, or a knuckle from scratch, just for additional wheel return effect, but for now i will be running it in the standard position.
Anyway, back to the construction, the crossmember was finished first, which allowed me to mock the new front end up, as mentioned the rack was moved forwards, the distance is filled in by this custom column spacer.
The next major part was the tower, the idea was to have something that was stronger than factory, but was also adjustable, or at least replaceable, for fine tuning. We went with a box layout for the bulk of the tower, with a plate setup that bolted to the box, where the strut could bolt to the plate. This meant the plate could be removed, modified or replaced, if i ever wanted to change the position.
The next big step was lower control arms, i was extremely happy with how my last custom LCA’s worked that i talked about in part 1, so i continued to use them, however there was a lot of length added. I also moved the caster rod mounting back, all of this was to increase clearance from the wheel hitting the caster rod which as many people know, is the limiting factor on a lot of modified knuckle setups. With all these modifications, the caster rod was now sitting in the region of 140mm further away from the hub than standard, this was a huge amount of clearance, but also a huge amount of strain on that part of the lca due to the fact that most of their strength is drawn from the triangulation with the caster rod, the leverage on this bit of LCA sticking out from where the caster rod used to connect was increased by roughly 5 times, 5 times the amount of force meant i needed 5 times the amount of strength. I think i have that covered with the reinforcement we went with…
Besides these modifications, as extensive as they were, the rest of the setup is quite normal, the rack ends and tie rod ends would obviously have to be longer, and due to the knuckle design i went with from Garage7, the tie rod ends required a 110degree balljoint, as opposed to the normal perpindicular/90 degree balljoint most use. Thankfully, R31 tie rod ends fit the bill in both areas, their taper is slightly off, but still holds, and their length is an extra 110mm over standard, so coupled with some slightly shorter than standard rack ends from a suzuki grand vitara, i had the perfect overall tie rod length i needed for less than $45 a side.
With all this completed, it was time to load the car and head off to Winton for their matsuri event with an untested, completely custom front end, i honestly had no idea what to expect, but to say i wasn’t nervous would be a pile of horseshit.
There is still a tonne more to go, but i will leave that to part 3…