Mazda Suspension Tuning – bushing durometer & dynamic rear toe

Alright so… I autocross an over-powered ’95 ford escort.  I’m looking for ways to vanquish some understeer & improve rotation.  First step is replacing my worn out struts and “coilover sleeve kit” with actual purpose built coilovers and an alignment (still racking by brain about where best to adjust my front camber/caster, and whether or not changes in KPI are going to affect handling significantly… but I’ll save all that for another post).  Second step will be rolling my rear fenders so I can run the same race rubber x 4 (I ran HoHo SM7’s up front and NT01’s in back last year, will be SM7's x 4 this year). But of course… my brain has been over-thinking other things this winter :)

I did a little reading on Mazda’s “twin trapezoidal link” rear suspension a while ago, which this car shares with the 323, Protégé, MX3, etc. It sounds like Mazda originally used varying bushing durometers to control the rear toe changes under cornering load.  Neat!  Well, except that 10+ years ago I removed ALL of the stock rear bushings and replaced them with, if I remember correctly, 80A polyurethane throughout.

What I’m wondering now is… could I improve my rotation by replacing some of the bushings with either a softer or harder material?  Obviously it’s easier/cheaper to go softer (back to stock) than harder (custom bushings or spherical bearings).

If I am correctly visualizing how this suspension works, I believe the trailing link pushes the base of the spindle rearward during a bump/roll situation.  It seems that, with all bushings being equal, this would actually result in TOE IN at the outside rear wheel while cornering.  While good for stability it sounds like this might actually be the opposite of what I want for autocross??  If I wanted to try a bushing change, which arm should I try “softening”?  The front link bushings, or rear link bushings?

Finally, I’m told the car was three-wheeling a lot last season.  I think part of that was the way I had to chuck it into corners in an attempt to break rear traction and get it to rotate.  I’m hoping that making the rear suspension geometry/behavior a bit more tail-happy on its own will allow me to drive the car more smoothly, and keep all 4 tires on the pavement.  Thoughts?

Here are some photos of the suspension layout (not my car... this car has custom lateral links with rod ends, i still have stock arms with poly bushings)

In reply to therieldeal :

Any pics of it on 3-wheels?

Did you check tire temps to see how your front camber is?

You generally want to limit body roll and maximize negative camber up front, then tune the rear to make it rotate the way you want. If you already have the front dialed in, then you can move to the rear. 3-wheeling isn’t necessarily bad, but you may need more toe-out to help it rotate?

Regarding your original question about bushing durometer, the best thing you could do is remove a rear spring from one side, then manually articulate the suspension through its full range of motion, and check for any bind/weirdness. 

In reply to therieldeal :

Also, try for 3 to 4 degrees negative camber up front, and don’t worry too much about caster. Try to get your front contact patch temps as even across the tire as possible. 

Make sure you have a bit of front toe-out to help with turn-in, 1/8” or a bit more is probably a good start, then adjust rear toe-in/out to help balance the car. 

Remember - on most FWD autocrossers the rear tires are just ballast. 

Thank you for the tips!  I think you’ve confirmed that I’m really over thinking this rear suspension right now.  I’ll try disconnecting the swaybar endlink, removing a spring, and checking out the smoothness of travel.  If there’s no significant binding I’ll leave it alone for now and focus on the front first; if it binds I’ll try to find the source.  I know the sway bar setup on these cars causes some binding… it uses a stacked bushing/washer/tube/washer/bushing setup that attaches to the center of the lateral link, like a lot of cheap older cars.  I do have a later Protégé rear sway bar that’s longer and would correct directly to the strut via a ball-joint type drop link… but that’s on my future project list as I still need to create a bracket for the strut, and source drop links of a suitable length.

I definitely do not have the front dialed in yet, as those coilovers are all new vs. last season too.  The front struts were actually in worse shape than the rears when I pulled them off the other day (one actually leaking).  I’m hoping to be able to keep my static front camber under ~2 degrees or so.  This is because I’m trying to put down quite a bit of power, so the more static negative camber I dial in the less acceleration traction I’ll have (and braking traction, for that matter… flat spotted a couple front tires last year).  Since it’s very easy to get unintended wheelspin when my turbo spools, I’m not sure how useful the tire temp data would be (spinning in a straight line over-heating the inside edge of the tires).  Last season I ran about 1/8” of toe-out up front, and 0 toe in the rear.

The reason I’m looking at increased caster is that, first of all, the stock caster is something miniscule like 1*-2*.  Second, positive caster should in theory give me some extra camber gain in tight corners, which is what I really need as most of our autocrosses are at a very small lot.  I’ve also increased my ride height about 0.5” all around, to try and get the front control arms to droop a little (aiming for at least a little neg camber gain in roll).  Last but not least, I’m still working on a way to extend the front lower ball joint stud, again to get more control arm droop, but I haven’t actually done this yet as I need to work out what sort of effect that’s going to have on bump/roll steer, and how to compensate for it.

I do not have any photos of it three-wheeling last season, but I do have a couple photos of it cornering hard from the front.  I was getting a fair bit of body roll, so I’m hoping that this year with stiffer shocks and a 12% increase in spring rate all around it will corner a little flatter.  I’m already running the largest-available stock front sway bar that will fit, and the 2nd largest aftermarket rear bar.

Anyway, here I’m exiting a slalom & essentially turning into T11a at TSMP.  From this point on, the course essentially followed the racing line of the road course… a second or two after this shot I was spinning 3rd gear & fighting understeer all the way to the crest of the hill where the straight begins:

Here are a few mid-corner shots from tight turns at our usual autocross location:

In reply to therieldeal :

Your body roll doesn’t look too excessive, but you definitely need WAY more negative camber in the front - do you see how far positive it is in the first pic?

I totally understand your forward traction issue, but autox is 90% about the corners, and even more so on smaller/tighter courses. Certainly an LSD would help forward motivation, but that’s also $$ to replace. Patience with the throttle may be your best option for now - I can guarantee you’ll turn faster autox laps by dialing in more negative camber in front & backing off the throttle on corner-exit vs. working to put down more power. 

BTW I had a Sunrise Red EGT in 94, but sold it before I discovered autox. Even at stock power the inside wheelspin was a problem around corners! Pre-internet I called a company in Indianapolis I saw in the back of Turbo magazine & asked them about adding a turbo to it - the guy on the other end actually laughed at me for wanting to turbo it! I’d still like to find another someday...

I echo Pete's comments. And I also agree with you, you're overthinking it.

Why not just put the new suspension on and see how it does before you start messing with anything else? Get as much positive caster and negative camber as you can, then drive it. I think you need a good baseline. What are spring rates are you going to run?

Cool car!

In reply to Pete Gossett :

What you're saying makes a lot of sense.  When my friend can get around the course faster than me with his civic, with similar tires, and a bone-stock ~80 horsepower d-series motor... power obviously isn't as important as I want it to be!

I do have an LSD in there, a Quaife ATB.  So that helps a lot, but it could still be better.  I'm hoping maybe, just maybe, having dampers that actually dampen will help me put power down little a bit better too.  The hardest part about trying to put power down in this car is that the turbo is, quite literally, an old school 90's rallycar turbo setup.  Low compression pistons make it totally gutless... gutless... and then BAM it shoots up to full boost/torque within a few hundred RPM band.  I'm considering softening the wastegate spring and trying to tune in a more gradual spool up with electronic boost control.... would take a while to get that tuned right, though.

Hilarious that you were laughed at for wanting to turbo your old one, seeing as how Mazda sold the BP with a factory turbo in the rest of the world :).  My engine/turbo are actually bone stock... out of a 323 GTR.  Bolts right in!

In reply to CyberEric :

Thank you both for bringing me back down to earth here.  It's been a long winter of not racing... I haven't even driven the car since October.

My old spring rates were an off the shelf Ground Control sleeve kit on stock style struts, 300# front, 200# rear.  Definitely too much spring for the struts, and with the coilover springs being so much shorter than stock things got weird anytime a wheel was unloaded... zero spring preload other than the weight of the car.

The new suspension is a true coilover setup from YCW that allows me to adjust height independently of spring preload.  The springs are 6K front/ 4K rear, which is something like 335# / 225#.  Digressive valving, adjustable rebound.  Hopefully they work out well!  I considered going even stiffer, but the pavement at our primary lot is pretty bumpy/broken/uneven, so I need to be able to soak up some amount of bumps.