Jump to content




Modeling The 1G Crx Suspension As A Guide To Performance Upgrades


  • You cannot reply to this topic
11 replies to this topic

#1
Andy69

  • In the left lane
  • PipPipPip
    • Group: Members

I have an ongoing project on my 1st gen CRX regarding suspension setup.   I compete in local autocross events in SCCA F Street Prepared.  I’ve been trying to find the optimum setup for the car using knowledge from RPR making changes and testing, but I realized that I didn’t know enough about what the suspension is actually doing when it’s in motion to make changes that actually helped.  So I decided to model the suspension to gain a better understanding of this.  This series of posts represents the sometimes rambling path I’ve taken to get to the point where I think I reasonably understand the suspension,  how all the parts work together, and most importantly – if potential changes to the suspension will be helpful in terms of handling.

 

The car was originally an SCCA ITB car and came to me with the following setup:

 

August 2012

24mm front torsion bars

275# rear springs with Ground Control coilovers

16mm front ARB

22mm rear ARB

Tokico Illumina dampers

Strut bar

Roll cage

Total weight with driver 2190#

 

 

Early 2014 upgrade

Upped rear springs to 350#

 

Upgrades done soon after - 2014:

27mm front bars

450# springs

camber plates in place of the strut bar

 

Later in 2014

Upgrade rear to 500# springs

Added a  7” tall removable rear spoiler

 

2015

upgrade rear springs to 550# (Tokico dampers seriously over springed at this point)

 

2018

Removed a ton of weight from the car, about 250 lbs including roll cage, brake swaps, lightweight wheels, etc.

Upgrade dampers to Medieval Pro Teins  from Heeltoe, with proper valving to match the 27mm bars and the 400# rear springs that came with them.

Also took corner weights and balanced the car for the first time.

Set the front ride height such that the lower control arm was almost parallel to the ground with the inside slightly higher than the outside, and set the rear ride height about 1/8” lower.



#2
Andy69

  • In the left lane
  • PipPipPip
    • Group: Members

As you can see, I had been making adjustments to the suspension as I progressed as a driver, and I eventually arrived at the point where I thought I had a really good setup with a lot less weight and properly valved dampers. I had been relying on conventional wisdom from this site regarding car setup, as well as what limited information I could get on cars that were successful at the national level in SCCA Solo.  Such information was often incomplete and sometimes contradictory, and I kind of had to fill in the gaps. When I tested this last setup in early 2019, I quickly realized I had I inadvertently turned the car into an understeering and nearly undrivable pig.  I attempted to make various adjustments, namely lowering the front of the car, which helped tame some of the understeer but it still didn’t handle at all like I was expecting.

 

After doing some additional reading on the subject which included various websites dedicated to suspension setup, and Carroll Smith’s book Tune to Win, I realized I had no idea what the suspension was actually doing and didn't have enough information about the car to be able to make changes that helped. I decided I needed to go through the process of modeling the suspension, with the idea that I could use that to understand the car, what it was ultimately capable of, and to get a good baseline setup. With a well-developed model I would also be able to plug in potential changes to see what effect they would have on the suspension geometry.  I could then use real world experience with the car to validate the model – did it handle like I was expecting, and if not why not?

 

I used DG's Autocross to Win as a starting point.  They state in the suspension setup section that for a baseline autocross setup, you should get as close to natural suspension frequencies of 2.2 Hz front and 2.5 Hz rear.  Frequency in Hertz is how many times the loaded spring should bounce per second, undamped.  The first thing I did was to plug the numbers for the current setup of the car into their dynamics calculator, and the frequencies were around 2.4  and 2.6– that’s with the 27mm front bars and 400# rear springs.   The numbers for the original setup of the car were 1.8Hz and 1.9 Hz.  Two things immediately jumped out at me – the first was that the original frequency balance was off and I had corrected that balance with my first rear spring upgrade from 275 to 350.  That was based on the conventional wisdom of a 125# difference give or take in spring rate between the front and rear with the rear higher.  The other was that the 400# springs provided with the standard version on the Tein dampers did not seem to be high enough, which again matched with the conventional wisdom.  

 

One quick note on this process, as I went through it, I realized two complications – the aero, and the stock Si rear axle with the internal ARB – would complicate the accuracy of the process, with the aero shifting grip to the rear and the internal bar shifting it to the front.  I assumed they canceled each other out but kept it in the back of my mind.

 

My goal in this process was to get an optimal setup for the type of racing I do, with a list of changes I can make on the fly to balance the car if I’m having issues that may be site or course specific.  I’d like to have a pretty good idea of what effect those changes would have, so if the car is pushing a little or more than a little, I can make a specific change or changes to better balance the car.



#3
Andy69

  • In the left lane
  • PipPipPip
    • Group: Members

My initial setup when I started the process in January 2019:

 

1987 CRX Si

stock front brakes

HF rear drum brakes

stock Si rear axle

27mm front t-bars

Medieval Pro dampers

400# rear springs

16 mm front ARB

22 mm rear ARB

camber plates set to full negative (about 1.8 degrees)

1 degree negative camber in the rear

1/8" toe out front

1/16" toe in rear

Front ride height set for a LCA angle of about +0.5 degrees

Rear height set 1/8 lower than front

Corner balanced

 

Initial weight with 250# driver 1920#

 

Thoughts on the 2019 season

The conventional wisdom seems to be when lowering the car to drop it down only to a point where the LCAs are at a slightly positive angle, that is, angling upward slightly toward the centerline of the car. Any lower than that would create issues with bump steer, roll center height, etc.  This is where I started the 2019 season as far as ride height, with the rear set 1/8” lower and the car corner balanced.

I did not like how the car handled.  It pushed hard, but I still managed to finish in the top 25% of the field most events in my home region, and also a good solid 25th out of 102 and first in Street Prepared at St. Louis Region Event #1 in April.  I was however, down the field considerably in other regions.  Not satisfied since I think the car has more.  I attribute the halfway decent finishes locally to driver improvements as photographs and video have illustrated that I am getting much closer to cones than I used to, as well as having good hand position and looking ahead, but I think car setup was still holding it back.

 

I made some adjustments to try to tame the understeer, namely lowering the front.  I quickly discovered that the adjustment made the car handle better (no surprise there).  I was on to something here but again I still had no idea what the suspension was doing or why.  I know now looking back that it has a lot to do with the fact that you cannot adjust the roll center in the rear of the car much at all due to the panhard bar even though you can adjust the ride height and through that the center of gravity – height adjustment is very much a relative to the rear of the car activity since you have a much greater ability to affect the geometry of the front than the rear via height adjustment.   But all I knew at the time was that lowering the front ride height AND dropping the rear to match reduced understeer, which seemed counter intuitive as lowering the front RELATIVE to the rear will reduce understeer.  I was actually doing that at the time, I just didnt realize the front had more of an effect than the rear.  Another puzzling development that modeling would help to clarify.



#4
cbstdscott

  • Swap in HF drums, check your cam timing
  • PipPipPipPipPipPip
    • Group: 2015 Contributor
    • Location:CRX Heaven, Los Angeles
    • Drives: '87 CRX Si, '07 Civic Si Sedan, '15 Scion FRS
    • Image Gallery

I have found a lot of new rotation in the chassis by removing the internal rear sway bar that is inside the rear axle.

 

Removing the internal sway bar created a lot of over steer which I controlled with more front roll stiffness. I have set up my car to be "neutral," with no propensity to under or over steer unless I ask for it with steering wheel and throttle input.


Posted Image

Form Follows Function

#5
CSPCRX

I agree about the rear internal bar.

 

Confused by having the rear lower than the front. Where are you measuring this from? I have never run my car like this I usually run 1/4-1/2 in higher in the rear measure off the body line from the molding.


Victor
86 K24 Powered CRX SI (SMF Solo2, HPDE-4 NASA & TA-A Time Attack

85 CRX DX totally original
07 Harley FXST Softail

2021 Tacoma Tow/Daily


#6
Andy69

  • In the left lane
  • PipPipPip
    • Group: Members

I agree about the rear internal bar.

 

Confused by having the rear lower than the front. Where are you measuring this from? I have never run my car like this I usually run 1/4-1/2 in higher in the rear measure off the body line from the molding.

 

I usually measure from the top of the tire to the bottom of the wheel lip.

 

I'm still writing this up but I can tell you from going through this that it seems that the rear height doesn't really matter much as long as it's not something crazy.  All the action happens at the front and the rear is just along for the ride (and I think that's common knowledge among people who race these cars). You can change both the CoG and the roll center slightly, and the more you lower the car in the back the more the two will converge, because the CoG drops by an amount roughly equal to the drop in height and the RC by about half that.  What that means is the "lever arm" between the CoG and RC gets shorter the more you lower the car, with the result being less body roll.  But I think in an absolute sense you aren't going to change the weight transfer balance enough to really make a difference in handling unless you swap out the panhard bar for an adjustable one or something different like a Watt's link. But that's getting into something beyond what I've been trying to achieve.


Edited by Andy69, 31 August 2020 - 09:07 AM.


#7
Andy69

  • In the left lane
  • PipPipPip
    • Group: Members

I agree about the rear internal bar.

 

Confused by having the rear lower than the front. Where are you measuring this from? I have never run my car like this I usually run 1/4-1/2 in higher in the rear measure off the body line from the molding.

 

That's the big thing about the rear - I don't know exactly how to quantify that internal bar in terms of adding it to a model.  It's so unique in that it's not attached to the body and acts only on one side of the car.  What I need to do is to swap out the HF axle and see what kind of a difference it makes in the handling balance.  My initial hunch is that it would have the opposite effect of what Scott claims to have experienced - less roll stiffness in the rear results in more understeer or less oversteer.  BUT - and this is a big but - the internal bar does not act on the body of the car the way a normal ARB does, so you can't just plug it into a model like you can a normal one.  I would almost think that you would have more roll stiffness in a left turn than a right turn, which would necessitate a slightly different spring rate on the left side to compensate.  It's something that's been occupying my thoughts in this whole thing, seeing as I've been experiencing something different with the car than I would expect based on my model.


Edited by Andy69, 31 August 2020 - 09:21 AM.


#8
anjin

  • redlinefever
  • PipPipPipPipPip
    • Group: 2014 Contributor
    • Location:Perth, Western Australia
    • Drives: 83 crx - kakaburra,83 gen1 Noir, 87 EF7 Glasstop crx, 83 gen 1 rebuild replica mugen racecar
    • Image Gallery
    Garage View Garage

I don't think you have the rear axle correct. The internal arp acts on the axle not the body. so it will react to movement of the wheel on the other side.  For info, I did have the internal arp removed and a 14mm interga extenal one in its place. I have now cut and pinned the normal internal arb to replicate the mugen one vis shorter length rather than increased diameter.  It is about half the normal one's length. I have removed the additional external 14  integrs rear arp.  The car seems to have about the same balance but I haven't taken it onto a track for the limit testing.


anjin aka Ian

"I can't believe it - Ive just been passed by a sh!tbox"

#9
Andy69

  • In the left lane
  • PipPipPip
    • Group: Members

I don't think you have the rear axle correct. The internal arp acts on the axle not the body. so it will react to movement of the wheel on the other side.  For info, I did have the internal arp removed and a 14mm interga extenal one in its place. I have now cut and pinned the normal internal arb to replicate the mugen one vis shorter length rather than increased diameter.  It is about half the normal one's length. I have removed the additional external 14  integrs rear arp.  The car seems to have about the same balance but I haven't taken it onto a track for the limit testing.

 

I think there is certain amount of inaccuracy in the model due to this, hence my note that I really need to replace it with the HF axle.  But interestingly, the real world handling results indicate an understeer bias where the model suggests I should be experiencing overrsteer.

 

BUT, as I mention in the write up I'm working on, aero and the internal bar are not accounted for.  I don't know the effect of the rear bar but the spoiler alone I think would account for it.  Further testing would include swapping the HF axle and removing the aero.

 

But at the moment, the car is out of commission.  It's got a transmission/drivetrain problem and I'm not sure when or if I can get to it.  I've actually been looking around for a Fiesta ST for autocross and putting this one aside until I have time for a full on dedicated FSP build, which would include disassembly of the entire car to fix all the tired crap from 30 years of racing.



#10
CSPCRX

The HF axle is what I have run in mine fore the last 20+ years.  Has a custom adjustable sway bar that attaches to it.  Less weight.


Victor
86 K24 Powered CRX SI (SMF Solo2, HPDE-4 NASA & TA-A Time Attack

85 CRX DX totally original
07 Harley FXST Softail

2021 Tacoma Tow/Daily


#11
Andy69

  • In the left lane
  • PipPipPip
    • Group: Members

Ok, it’s been a long time since I worked on this.  I parked the car after it developed a serious vibration and instead auto crossed my Caterham, before buying a Fiesta ST in 2022.  I’ve got the 7 down for repairs and upgrades, and I’ve been focusing on finishing my long running 1969 Chevelle project.  I’m about done with it and I’m starting to think about the CRX again.



#12
CSPCRX

Nice!  Share some pics of that American muscle!


Victor
86 K24 Powered CRX SI (SMF Solo2, HPDE-4 NASA & TA-A Time Attack

85 CRX DX totally original
07 Harley FXST Softail

2021 Tacoma Tow/Daily